DEVELOPING AGRICULTURAL BIOTECHNOLOGY IN THE NETHERLANDS
D.H. Vuijk J.J. Dekkers &H.C . van der Plas (eds.)
Pudoc Scientific Publishers 1993 Cip-Data Koninklijke Bibliotheek, Den Haag
Developing
Developing agricultural biotechnology in the Netherlands / D.H. Vuijk, J.J. Dekkers & H.C. van der Plas (eds.). - Wageningen : Pudoc Scientific Publishers. -111 . With index ISBN 90-220-1081-3 bound NUGI821 Subject headings: agriculture and biotechnology; The Netherlands.
ISBN 90-220-1081-3 NUGI821
© Centre for Agricultural Publishing and Documentation (Pudoc), Wageningen, the Netherlands, 1993.
No part of this publication, apart from bibliographic data and brief quotations embodied in critical reviews, may be reported, re-recorded or published in any form including print, photocopy, micro film, electronic or electromagnetic record without written permission from the publisher: Pudoc, P.O. Box 4, 6700 AA Wageningen, the Netherlands.
Printed in the Netherlands Developing agricultural biotechnology in the Netherlands This book has been prepared with the help of an editorial committee. The members of the board have been of great help in selecting authors and subjects, examining manuscripts and writing introductory articles for specific chapters. The editors wish to express their gratefulness to these people.
Editorial committee W.IJ. Aalbersberg retired from: Netherlands Institute for Dairy Research, Ede
M.C. Horzinek Department of Infectious Diseases and Immunology, Veterinary Faculty, University of Utrecht
K. van 't Riet Department of Food Science, Wageningen Agricultural Univer sity
A. Rip School of Philosophy & Social Sciences, Technical University of Twente
S. Tamminga Department of Animal Nutrition, Wageningen Agricultural Uni versity
L. van Vloten-Doting Agricultural Research Service (DLO)
A.G.J. Voragen Department of Food Science, Wageningen Agricultural Univer sity
PJ.G.M. de Wit Department of Plant Pathology, Wageningen Agricultural Uni versity
A.J. van der Zijpp DLO-Research Institute for Animal Production, Zeist Preface
The past decade has seen an enormous increase in biotechnology research. In particular, the novel possibilities ofgeneti c engineering technology, i.e. the identification, isolation and transfer of genes, have given rise to a completely new type of research in scientific institu tions and in industry. Moreover, biotech companies have sprung up which provide special ized goods and services on a commercial basis. Genetic engineering has proved to be a great scientific success and in many cases an excellent tool for research in various disci plines. The important question for the coming years iswhethe r the new biotechnology will also be a commercial success. In our opinion, the coming decade will be decisive for applications of genetic engineer ing. What will be the attitude of society towards the new biotech products and, closely associated with this, what will be the authorization policy of the various governments? There isa general feeling that sound knowledge isneede d more than ever. The responsibil ity of scientists in various disciplines, both in natural sciences and in humanities is being severely tried. The Programme Committee on Agricultural Biotechnology in the Netherlands, an ad hoc Committee set up by the Minister of Agriculture, Nature Management and Fisheries, has stimulated and coordinated research in the area of agricultural biotechnology in the past years. The agricultural biotechnology research programme, in combination with the industrial and the environmental biotechnology programmes, has formed the core of the innovative research programme on biotechnology in the Netherlands. The interim results ofth e agricultural biotechnology programme have already been described1. This first book was warmly welcomed and received many positive reactions. There was obviously a need for a second book containing a more general overview of agricultural biotechnology in the Netherlands dealing with interesting developments, both in science and in social topics. This rather ambitious plan could not have been realized without the kind cooperation of many people. We would like to thank all the authors for their contributions, and for their time and energy spent on this book. The help of Pudoc, in particular J.C. Rigg and R. Moeliker, as well as Ann Chadwick is also greatly acknowledged. This book would not have been possible without the financial support of SENTER and the directorate of Science and Technology of the Ministry of Agriculture, Nature Management and Fisher ies. We are especially indebted to Professor K. Verhoeff for his help and interest in the project. We hope that this book on the development of agricultural biotechnology in the Nether lands will attract an even more extensive readership than the first one and will stimulate international cooperation in the field of agricultural biotechnology. Wageningen, the Netherlands D.H. Vuijk J J. Dekkers H.C. van der Plas
1 J J. Dekkers, H.C. van der Plas & D.H. Vuijk, (Eds.), 1990. Agricultural Biotechnology in Focus in the Netherlands. Pudoc, Wageningen. Contents
Preface v
1 Agricultural Biotechnology in the Netherlands D.H. Vuijk,J J . Dekkers & H.C. van der Plas 1 1.1 Introduction 3 1.2 Research and industry 7 1.3 Government policy 11 1.4 International cooperation 14 1.5 References 16
2 Biotechnology in plant breeding and crop protection 17 2.1 General introduction to biotechnology in plant breeding and crop protection PJ.G.M de Wit & L. van Vloten-Doting 19 2.2 Molecular biometrical genetics: a vigourous hybrid P. Stam 23 2.3 Genetic modification of plants: new developments in transformation procedures J.J.M. Dons &WJ . Stiekema 31 2.4 New approaches for identifying and isolating plant disease resistance genes J. Hille &P . Zabel 39 2.5 New strategies for obtaining fungal-resistant plants B.J.C. Cornelissen &J.A.L. van Kan 44 2.6 Molecular breeding for insect-resistant plants L. Visser & M.A.Jongsm a 50 2.7 'Plantibodies': a versatile approach to engineer resistance against pathogens J. Bakker, A. Schots, WJ. Stiekema & FJ. Gommers 57 2.8 New developments in molecular flower breeding J.N.M. Mol 61 2.9 Industrial application of plant biotechnology: promises and pitfalls E. Veltkamp 66 2.10 Market introduction of products obtained through plant biotechnology, aspects of consumer acceptance T.A.W.M. Saat 71 2.11 References 74
vu Biotechnology in animal breeding, husbandry and animal health 81 3.1 General introduction to biotechnology in animal breeding, husbandry and animal health M.C. Horzinek & A.J. van der Zijpp 83 3.2 Identification and characterization of a major histo- compatbility complex in a teleost fish and its relevance to improving disease resistance E. Egberts, S.H.M. van Erp, G.F. Wiegertjes & RJ.M. Stet 90 3.3 Polymorphism of the bovine major histocompatibility complex genes and the relevance of this polymorphism for the study of infectious diseases J.J. van der Poel, Chr.J. Davies, Ph.R. Nilsson & M.A.M. Groenen 99 3.4 Gene mapping in farm animals M.A.M. Groenen, R.P.M.A. Crooymans, D. Ruyter, AJ.A. van Kampen &J J. van der Poel 104 3.5 Regulation of expression of milk protein genes M.A.M. Groenen, RJ.M. Dijkhof, E.J.M. Verstege, C.P. Spira & J.J. van dtr Poel 112 3.6 The molecular biology of genetic variation in a functional gene and its use in selection strategies of breeding programmes M.F.W. te Pas &J.H.F . Erkens 122 3.7 More calves from high merit cows using new reproduction technologies A. van der Schans 127 3.8 Anti-GnRH monoclonal antibodies to prevent boar taint in pork T. van der Lende, L. Kruijt & M. Tieman 131 3.9 Development of a 'biotech' vaccine against bovine herpes virus type 1 that can also serve as a vaccine vector J.T. van Oirschot, A.LJ. Gielkens, F.A.M. Rijsewijk, F.A.C. van Engelenburg, M.J. Kaashoek & B. van den Burg 134 3.10 Safety of recombinant DNA virus vaccines T.G. Kimman, A.LJ. Gielkens, K. Glazenburg, J.M.A. Pol & W.A.M. Mulder 142 3.11 Putting a'bar code'on bacteria? B.A.M. van der Zeijst 150 3.12 References 156
Biotechnology in animal nutrition 167 4.1 General introduction to biotechnology in animal nutrition S. Tamminga, S.F. Spoelstra & GJ.M. van Kempen 169 4.2 Nutritional value and physiological effects of D-xylose and L-arabinose in poultry and pigs J.B. Schutte,J . deJong , P. van Leeuwen & M.W.A. Verstegen 175 vm 4.3 Antinutritional factors: aspects of the mode of action in the animal and inactivation of ANF-activity J. Huisman & F.H.M.G. Savelkoul 177 4.4 Phenolic monomers and digestibility of plant cell walls RJ. Hogendorp 182 4.5 Prospects of feed additives produced by biotechnology J.L. van Os 186 4.6 References 190
Biotechnology in the food industry 195 5.1 General introduction to biotechnology in the food industry W.IJ. Aalbersberg, K. van 't Riet & A.G.J. Voragen 197 5.2 Improving of the baking quality of gluten by enzymatic modification RL. Weegels & RJ. Hamer 201 5.3 Enzymatic hydrolysis of milk proteins: some fundamental and practical aspects S. Visser, D.G. Schmidt & RJ. Siezen 207 5.4 Plant cell wall degrading enzymes in the processing of food G. Beldman & A.G.J. Voragen 213 5.5 Enzymatic processing of triglycerides and fatty acids. Enhancing reaction kinetics by continuous product removal J.T.P. Derksen, G. Boswinkel, W.MJ. van Gelder, K. van 't Riet & F.P. Cuperus 220 5.6 Thermodynamic principles of enzymatic acylglycerol synthesis A.E.M.Janssen , A. van der Padt & K. van 't Riet 226 5.7 Engineering lactic acid bacteria for improved food fermentations W.M. deVos 231 5.8 Engineering and application of nisin, a natural food preservative O.P. Kuipers, H.S. Rollema,J . Hugenholtz, W.M. de Vos & RJ. Siezen 236 5.9 Molecular genetic approaches in plant cell wall degradation M.A. Kusters-van Someren, L.H. de Graaff&J. Visser 242 5.10 Detection and identification of food-borne bacteria by means of molecular probes A.H. Weerkamp, N. Klijn & W.M. de Vos 245 5.11 The development of immunoassays for the rapid detection of moulds G.A. De Ruiter, HJ. Kamphuis, S.H.W. Notermans,J.H . van Boom & F.M. Rombouts 254 5.12 Safety aspects of foods produced by recombinant DNA technology H. Hofstra 262 5.13 References 266
6 Societal aspects of agricultural biotechnology 275 6.1 General introduction to societal aspects of agricultural biotechnology A. Rip 277 6.2 Diffusion of genetic modification in Dutch agribusiness T.H.J.M. Hutten 281 6.3 Protecting and exploiting biotechnological inventions in agriculture: patent rights versus plant breeders rights? T.H.J.M. Hutten 291 6.4 The development of risk-evaluation for the deliberate release of genetically modified organisms: An outline of the international discussion H.E.N. Bergmans 297 6.5 Safety and acceptance of biotech-food D.A. Toet 302 6.6 Technology assessment and agricultural biotechnology A. Rip 305 6.7 Ethics and agricultural biotechnology J.M.G. Vorstenbosch 307 6.8 References 319
The Authors 321
Index 336 Agricultural biotechnology in the Netherlands
Contents
1.1 Introduction 3 1.1.1 Modern biotechnology: innovations by using ofbiologica l processes 3 1.1.2 Importance of agricultural biotechnology to the Netherlands 4
1.2 Research and industry 7 1.2.1 Agricultural biotechnology research infrastructure 7 1.2.2 Agricultural biotechnology industry in the Netherlands 9
1.3 Government policy 11 1.3.1 Innovation incentives 11 1.3.2 Authorization policy 13
1.4 International cooperation 14 1.4.1 The European Community 14 1.4.2 Eastern Europe 15 1.4.3 Developing countries 15
1.5 References 16 1 Agricultural biotechnology in the Netherlands
D.H. Vuijk,J J. Dekkers & H.C. van der Plas
1.1 Introduction
It is generally recognized that the change from hunting and gathering to organized agriculture has been of overriding importance to the evolution of human culture for the last 5000 years. Compared with the evolution that took place in the preceding period of several million years, this development can rightly be called dramatic. However, even in those days, man must have had a keen interest in nature because his survival virtually depended on an understanding of the plants and animals that were used to feed on. Man then began to grow crops and keep animals, thus gradually moulding nature in accordance with a three-stage process that the human race is inclined to follow: (1) interest and curiosity in nature which leads to (2) understanding of the processes of nature, which in turn is translated into (3) application of the knowledge. This last stage results in the control of nature by mankind.
1.1.1 Modern biotechnology: innovations by using of biological processes
The above-mentioned three-stage process is essential to the development of technology, including biotechnology, which in a sense can be labelled as one of the most illustrative examples of progress in technology. In former times, problems in agriculture involving the propagating of stock, growing methods and the storage and processing of harvests were addressed by this process (as well as by ritual or religious approaches). It should be noted that analysis-based scientific knowledge was relatively limited for a long time and did not gready expand until the last few centuries. Nevertheless, man has been engaged throughout the ages with improving crops, breed ing livestock, developing product storage 'and processing techniques, in order to be able to survive hard times as well as to enjoy the pleasure of abundant and varied food and drink. The basis for these activities has always been the genuine interest of man in his surround ings. With the enormous expansion of knowledge the possibilities for man to use and control nature have increased accordingly. In this sense modern biotechnology isjus t another step, although probably a discontinuous one, in a long series. In principle genetic modification techniques have enormous potential. The genetic contents of living organisms (plants, animals and micro-organisms) can be changed deliberately using foreign genes which add new, desirable characteristics. Large-scale production of enzymes allows agricultural raw materials to be modified towards a particular direction in industrial processing. The possibilities of new biotechnology or modern biotechnology which are character ized by the application of genetic modification technology have been described in many articles (Houwink 1989). Essential in modern biotechnology is the control of bioprocesses on the basis of acquired knowledge. Applications will mainly be found in the agricultural, health, pharmaceutical and environmental industries.
1.1.2 Importance of agricultural biotechnology to the Netherlands
Agriculture (including horticulture) has remained one of the major economic forces in Dutch society, despite the high densities of population in the Netherlands (Anon, 1992). Fig. 1.1.1 shows the development of Dutch agricultural imports/exports from 1968 to 1991. Due to the small-scale, water-rich structure of the agricultural area and the high pressure of population density, the specialized systems of production, often with little land attached to them, such as glasshouse growing of vegetables and flowers, intensive livestock production as well as dairy farming, have been the most successful. Apart from the factors mentioned above which have resulted in the specific structure of Dutch agriculture, the infrastructure of trade with the port of Rotterdam and the Euro pean concentrations of population at a relatively short distance from production centres have also been important. In order to be able to compete in European and world markets Dutch farmers, faced with expensive production factors (high land prices, expensive labour), must produce large quantities and grow high-quality products, both of which make tremendous demands on knowledge and skill. Fig. 1.1.2 shows the net value added per agricultural worker in each EC member state. The Netherlands and Belgium have a higher value added per worker than any other EC country. In order to achieve this theagricultural and agribusiness sectors have alwayspaid agreat deal ofattention totechnological innovations. There is ample interest in acquir ing knowledge within the agricultural communities in the Netherlands. Thus, technology plays an important role in Dutch agriculture. In addition, a healthy processing industry has developed (potato starch, dairy, sugar, and other food and drink industries) as well as ancillary industries (seed improvement, breeding, animal feed). It is generally felt that the one-track production-directed post-war approach has its limitations and is not without problems. Environmental problems in particular call for solutions, mainly the ones caused by pesticides, the use of fertilizers and manure surpluses and related ammonia and acidification problems. Modern agricultural biotechnology in the Netherlands, which has developed since the 1980s, is a technology which could contribute in various ways to solving the problems mentioned above. This book contains many examples, such as the development of disease-resistant plants, research into the genetic basis of immunological resistance in animals, the development ofmodern , effective vaccines, increasing the digestibility of feed using enzymes (resulting in reduced mineral surpluses) and the application of plants for industrial purposes (environmentally-friendly non-food uses). Besides solving specific environmental problems, agricultural biotechnol ogy may also contribute to several more traditional objectives. This book gives some examples, e.g. diversification ofvarietie s with varied colours, efficient selection of propaga tion stock, both in plant growing and animal breeding, more efficient processing technolo gies for agricultural products and new semi-finished and end products. It is obvious that modern biotechnology should not be seen as the solution to all problems, whether related to production or to the environment. Formulating the ecological structure of society Fig. 1.1.1. Major agro-biotechnological research centres in the Netherlands and trading ports: 1 Wageningen: Wageningen Agricultural University, DLO institutes (CPRO-DLO, ATO-DLO, IPO-DLO, CABO-DLO), TNO institute ILOB and (nearby at Ede) NIZO; 2 Utrecht: State University Utrecht: Faculty of Veterinary Sciences; 3 Zeist: TNO centre for Food Institutes, DLO institute TVO-DLO; 4 Lelystad: DLO institutes CDI-DLO and IWO-DLO; 5 Amsterdam: Free University and University of Amsterdam; 6 Leiden: Leyden University; 7 Lisse: Research Station for Bulb and Bulbflower Culture; 8 Schiphol: International airport; 9 Rotterdam: Major seaport; 10 The Hague: Ministry of Agriculture, Nature Management and Fisheries and Ministry of Eco nomic Affairs. 300
200
100
Po Gr Sp GFR Fig. 1.1.2. Net added value per agricultural worker in the EC. The EC average is set at 100. Po, Portugal; Gr, Greece; Sp, Spain; Ge, Germany; It, Italy; UK, United Kingdom; Dk, Denmark; Lu, Luxenburg; Fr, France; Be, Belgium; Nl, Netherlands. requires public consent. Technological innovations, such as biotechnology in agriculture and agribusiness, re quire a well-developed infrastructure of education and research as well as an adequate transfer ofresearc h results to the various industries and farms. For many years the Ministry of Agriculture, Nature Management and Fisheries has encouraged this with its combined approach of education, extension services and research tailored to the agriculture sector. Education is provided at the agricultural schools, colleges and at Wageningen Agricultural University. The university also has research responsibilities, as have the DLO-NL research institutes and the agricultural research stations. In addition, there are the extension serv ices for information at farm level. In recent years a distinct policy change has taken place: the influence of the government as the central controller and force is diminishing, the organizations are gaining more autonomy, and decisions are being made in a more decentralized way and on the basis of shared responsibilities. With the abolition of the internal borders of the European Community education and research structures will gradually lose their national character and cooperation between various groups in the EC member states will be consolidated, mainly as a result of fostering from Brussels. It is especially in this context that this book provides an international forum with a further introduction to Dutch agricultural biotechnology research. 1.2 Research and industry
In technological innovation the interaction between publicly and privately financed re search is essential. It is the size of the companies that co-determines the nature of the research of publicly financed institutions. In a sector consisting of many small companies with limited scope for research demand for applied research will be great. This is particu larly true for Dutch primary agriculture, which is characterized by family farms. Tailored applied research is carried out at the research stations. Small and medium-sized enter prises also have a great need for applied research. The large multinationals with their R&D divisions are mainly interested in fundamental research, and in suitably trained graduates.
1.2.1 Agricultural biotechnology research infrastructure
The research infrastructure relevant to agricultural biotechnology can be broken down into three categories: academic departments specializing in agriculture and biology, re search institutes, and research stations. Fig. 1.1.3 shows the geographical location of the research centres, and other relevant information. Table 1.1.1 gives capacity data on research inputs concerning agricultural biotechnology.
68 70 72 74 76 78 80 82 84 86 88 90 92
Year Fig. 1.1.3. Dutch agricultural import-export figures, 1968 - 1991.• , export; Organization Full-time equivalent DLO-NL Research Institutes plant production 140 animal production and veterinary 170 Wageningen Agricultural University food processing and biomolecular sciences 120 plant production 100 animal production 20 environmental biotechnology 30 University of Utrecht - Veterinary Faculty 60 Other universities' 50 TNO Research Institutes 70 NIZO Dairy Institute 40 Universities The bulk of agricultural biotechnology research in the Netherlands is carried out at Wageningen Agricultural University. It focuses on areas of application such as food tech nology, animal breeding, and crop protection, supported by fundamentally oriented groups for organic chemistry and biochemistry, microbiology, molecular biology, genetics, and physiology. The Faculty of Veterinary Medicine of the University of Utrecht is an important centre of animal biotechnology. It specializes in the reproductive physiology of livestock, development of diagnostic testing methods, veterinary immunology and vac- cinology. Various other university departments are also engaged in high-quality funda mental biotechnology research and education. Leyden University, the Free University of Amsterdam, the University of Utrecht and the Catholic University of Nijmegen are all involved with molecular plant biotechnology. Research institutes Two research organizations connected with agricultural biotechnology should be men tioned, one being the DLO-NL institutes and the other the TNO Applied Scientific Research institutes. The DLO-NL institutes are attached to and largely financed by the Ministry of Agriculture, Nature Management and Fisheries. Several large DLO institutes carrying out research in the field of agricultural biotechnology are: CPRO-DLO (plant breeding and reproduction research, Wageningen), ATO-DLO (agrotechnological re search, Wageningen), IPO-DLO (plant protection research, Wageningen), CABO-DLO (agrobiological research, Wageningen), IVO-DLO (animal production research, Zeist), CDI-DLO (veterinary research, Lelystad), and IWO-DLO (livestock feeding research, Lelystad). The TNO research institutes are largely financed by trade and industry. They focus on a wide range of scientific subjects. Important to agricultural biotechnology are the food group institutes (TNO-Food), e.g. TNO biotechnology and biochemistry (Zeist) and ILOB-TNO (animal feedstuffs, Wageningen). There is also the NIZO institute (dairy research) which is related to the dairy industry (Ede). Agricultural research stations The agricultural research stations focus on specific branches of agriculture, as wide- ranging as floriculture and pig production. Their main objective ist o address the problems and questions raised by the farmers and growers who provide an important share of their funds. Consequently, the research is highly practice-oriented. The most biotechnology- oriented isth e Bulb Research Station in Lisse, which focuses on tissue culture research and physiological research into bulbs and bulb flowers. Specific biotechnological research into consumable fungi is carried out at the Mushroom Experimental Station in Horst. 1.2.2 Agricultural biotechnology industry in the Netherlands Dutch agribusiness uses resources (material as well as human) from both the Netherlands and from abroad. Dutch agriculture is characterized by highly specialized and intensive farm management compared with other countries. Dutch agribusiness features a few large companies, particularly in the food sector and fermentation/chemical sector (enzymes, fine chemistry), several larger cooperatively established ancillary and processing industries, and medium-sized and small enterprises specializing in pharmacy, breeding, food and feed. Small companies, specifically concerned with biotechnology, have also emerged. Many Dutch biotechnology and agricultural biotechnology companies have joined forces in the NIABA Netherlands Industrial and Agricultural Biotechnology group. Table 1.1.2 shows the turnover of Dutch companies broken down into agribusiness sectors. Table 1.1.2. Turnover of Dutch industry in agribusiness sectors in 1991. Turnover in 106 NLG, excludingVAT . Sector Turnover Meat industry 13 220 Dairy industry 14 680 Fish processing industry 1 080 Flour industry 1 410 Sugar industry 2 010 Margarine and other plant/animal oil industry 4 600 Vegetable and fruit-processing industry 2 710 Bakeries 3 670 Cocoa, chocolate and sweets industry 3 400 Other food-processing industry 8 190 Feed industry 8 890 Beer brewers and liquor industry 3 880 Drink industry 1 400 Tobacco-processing industry 4 100 Plant breeding companies The Netherlands has several breeding companies oflon g standing, e.g. Zaadunie (horticul tural seeds), Van der Have Zaden (arable seeds), Royal Sluis (horticultural seeds), Baren- brug Zaden (maize and grass seeds) and Cebeco Veredeling (arable seeds). Specific bio technology companies active in plant improvement in the Netherlands are Mogen Inter national, Keygene, and Florigene. Moreover, the Ropta cooperative has established bio technology companies in addition to its traditional plant breeding enterprise. Cropprotection In the Netherlands reduction in the use of agrochemicals isdeeme d necessary. Biotechnol- ogical and biological pest and disease control is seen as promising. The Denka firm is active in the field of pheromones. The Groene Vlieg firm and the Koppert company produce predators commercially for biological control. Animal breeding Major companies engaged in breed improvement and breeding are Hypeco (poultry), Coveco (pigs), Euribrid (poultry), Fries Rundvee Syndicaat (catde), and Holland Genetics (catde). A specific biotechnology company is Gene Pharming, which is attempting to produce biomedical proteins in milk and, as a result, to increase the resistance of cows to bacterial infections (mastitis). Feed industries One of the reasons why animal production in the Netherlands was able to expand rapidly is the adequate supply of raw materials through the port of Rotterdam to produce com pound feeds. A strong feed industry has therefore developed, the main companies being the cooperatively organized Cehave and Cebeco and the private company of Hendrix International. There are also several smaller feed companies, e.g. Koninklijke N. Timmer mans' Veevoeders, Agrarische Unie/Vulkaan, Koudijs Wouda and UT-Delfia. Relevant pharmaceutical business companies in the veterinary sector The main companies engaged in the field of veterinary drugs and vaccines in the Nether lands are Intervet and Duphar. In addition, several small companies have emerged, especially in the development and commercialization of diagnostic testing methods, e.g. Livestock Control, MCA Development, Euro-Diagnostics, and Bio-Intermediair. Food and drink industry Many companies isactiv e in the food sector. In the dairy sector the cooperative companies of Campina-Melkunie, Coberco, Friesland Frico DOMO and the private company, Nu- tricia, play a leading part. The two leading sugar companies are the cooperatively organ ized Suikerunie and the privately owned CSM. 'Heineken' is a brewer's name of world wide renown, although there are other internationally operating breweries, e.g. Bavaria, Grolsch, and many small ones, sometimes still using traditional methods. Unilever and Wessanen are active in the field of oils and fats. 10 Other agricultural output processing industries Apart from the food sector companies mentioned above, a number of companies is active in the starch sector, e.g. Avebe (potato starch) and Latenstein (wheat starch); in addition, there are the Dutch branches of Cargille and Cerestar. Fermentation industry Gist-brocades plays a leading part in this sector. Main products are yeast, alcohol, and enzymes for many different applications. The company is interested in biotechnology applications on behalf of the agro and agro-food sectors. 1.3 Government policy 1.3.1 Innovation incentives Science andresearch In the 1980s economic competitiveness was the main objective of the government's science and technology policy. The government developed an Innovation Oriented Programme on Biotechnology (IOP-B) and earmarked extra funds to encourage biotechnology re search by universities and research institutions and have it linked up more to the demands of trade and industry. The Minister of Economic Affairs funded a programme on behalf of industry (food and drink, fermentation, chemical and pharmaceutical) and the Minister of Agriculture, Nature Management and Fisheries funded a programme on behalf of the agricultural industry. A recent international peer review organized by the Royal Netherlands Academy of Arts and Sciences shows that this promotion policy has been highly successful. A commis sion of international experts concluded that 'projected against the international scene,research into biotechnologyin theNetherlands has beendeveloped quite strongly and, taken as a whole, at agood scientific level. For its small size the Netherlands has become pre-eminent in several key areas and should seek to continue this investment for thefuture. The IOP-B has been implemented with remarkable success. Biotechnological research at the respective academic centresshows little, f any, undesired overlap. Accord ingly, biotechnology in the Netherlands has evolved as a strong network in a number of universities and governmental institutes and should,for themajor part, besupported as such becausethis network hasproved tobe very effective' (quotation from the Commission's report). Also at the request of trade and industry biotechnology promotion will be continued for a further five years with the explicit objective of guaranteeing its embedment in the shape of an adequately operating institution. Following the advice by the NWO Netherlands Science Organization, this will be in the form of an association of university biotechnology research school. The intention ist o step up fundamental research in several selected fields. Priorities in the agricultural field are in particular: - Applications in the field of crop protection. The Dutch government aims to have the use of agrochemicals in agriculture at least halved by the year 2000. Biotechnology techniques may contribute significanüy to crop resistance improvement and can provide quick, sensi tive detection methods. 11 - Applications in view of environmental improvement. This concerns in particular soil health, which is of paramount importance to efficient agricultural production. - Research into the development of fundamental knowledge of molecular structures and functions of food and drink. There is an enormous lack of knowledge in this field, which should be remedied as particularly the food and drink branch has scope for biotechnology innovations. - Research into non-food production of crops and upgrading of agricultural waste. In the field of non-food production especially, there is a demand for the expansion of economi cally interesting compounds, which requires adjusted research and the development of adequate biotechnology processing techniques. Industry Apart from stimulating biotechnology at the scientific level, the Ministry of Economic Affairs has launched a special programme to initiate support to biotechnology research at company level. This type of research may be carried out at any research institution, even a non-Dutch one by order of the company. This scheme aims to support especially smaller companies in their R&D activities. More commercially relevant projects are being set up within these programmes. A stock-taking by the PcLB Agricultural Biotechnology Pro gramme Committee has shown that the programme isgreatl y valued by trade and industry because of the near absence of red tape and the many possibilities for the financing of projects. Subsidies, though, are limited, which encourages firm commitment with the applicants. As many projects are carried out at universities and institutes, cooperation between companies and research institutions is moreover gready stimulated. Information concerning the past years of this subsidy scheme has indicated that agribusiness in particular is deeply involved in the introduction of biotechnology tech niques. The PBTS scheme was started in 1987. Between 1987 and 1992 240 applications were granted, 161 of which (67%) concerning agribusiness (see Table 1.1.3). Contextual measures Within the framework of the innovation promotion policy attention in recent years has focused, besides on the promotion of research, on the so-called contextual measures. That is, attention to a healthy development and application of new biotechnology such as Table 1.1.3. Distribution of company-orientated biotechnologically innovative projects. Economic sector Number of projects Subsidy in approved 106NLG Plant and animal breeding 88 49 Food and feed processing 42 29.2 Human and veterinary health 34 23.6 Chemical industry 25 10.5 Machinery and equipment industry 27 9.4 Environmentally orientated industry 11 2.4 Grand total 227 124.1 12 improving and harmonizing government policy, public education, establishing contacts aimed at widening industrial support, reviewing educational curricula, joining up with international developments, inviting foreign biotechnology companies to set up shop in the Netherlands, legislation, and increase of public awareness and acceptance and integration of biotechnology. This policy will be continued in the years ahead in the sense that, besides economic and scientific interests, other interests, e.g. environmental, health, welfare and ethical aspects, will increasingly obtain simultaneous and integrated attention. In research this will mean increasing attention to technology assessment, research into biosafety and research into ethical aspects, mainly concerning man and animal. In addition, not only the supply side but also the demand side of biotechnology will need to be highlighted, both technology transfer of industrial producers from and to agricultural producers and infor mation transfer from and to consumers. Research is needed to inform and advise policy makers. 1.3.2 Authorization policy Apart from innovation incentives and social aspects, the development and shaping of an authorization policy will be an important priority for the government. Biotechnology innovations are only successful if they are introduced to the market and the products are bought. A number of products, such as plant varieties, veterinary drugs, agrochemicals and food additives requires explicit government authorization, one of the difficulties being that there is relatively little scientific information available about the safety, effects and quality of genetically modified organisms and sometimes legislation for one or more of these specific aspects is even lacking. In recent years the European Community has been active in discussing and adopting directives aiming to realize harmonized legislation in the context of the Common Market. Two directives on the protection of man and the environment have been adopted, namely Directive 90/219/EEC on the contained use of genetically modified micro-organisms and Directive 90/220/EEC on the deliberate release into the environment of genetically modified organisms. Both directives depart from the case-by-case and step-by-step princi ple recommended by the OECD Organization for Economic Cooperation and Develop ment and are internationally accepted, implying that, on the basis of a risk assessment of the latest scientific findings, a decision ismad e to whether or not to proceed to the next step with less strict safety measures. If the product has been proved safe, it is released and marketed. In the Netherlands the EC directives are being translated into national legisla tion. Licenses are issued by the government, which is advised by the Provisional Commit tee on Genetic Modification. In addition, EC legislation on veterinary drugs has developed and is being discussed with regard to novel foods and intellectual property rights. In view of rapid authorization, research will have to support the new laws adequately and provide them with the latest scientific results. For the years ahead this means that more attention should be paid to the integration of natural scientific and socio-scientific research and to the understanding of the need to increase the social control of biotechnology developments. 13 1.4 International cooperation The Netherlands is a small country. Its favourable location with regard to the North Sea, has made it into a prosperous trading power where various multinationals, such as Shell, Unilever and Akzo have been established. Skilful entrepreneurship supported by agricul tural research and education, combined with soil made fertile by the estuaries of Rhine and Maas, have resulted in a strong export position ofagricultura l products throughout the world. Because of these specific circumstances the Dutch have always had an interest in developments abroad and shown a strong tendency towards international orientation. Science and technology have of old been strongly inclined towards the international scientific community. Although in the past technological competitiveness compared with other nations used to be an obstacle, the advantages are now so evident that international cooperation is being encouraged. The exchange of new methods and techniques and ideas often clarifies views and opens up new prospects for research. In addition, it may be desirable to form strategic alliances to realize new, improved products and services at a more rapid pace. Companies have set up new cooperation structures in order to facilitate the transfer of knowledge from universities and research institutes, mainly with regard to pre-competitive research. This applies in particular to biotechnology in trade and industry which is highly science-based and has to compete on world markets. 1.4.1 The European Community The creation of the European Community has been of overriding importance to Western Europe. Biotechnology is seen as a key technology in the development of the Community's competitive power. To strengthen cooperation biotechnology research programmes have been launched in the past years. The first of these programmes was the Biomolecular Engineering Programme (BEP, 1982-1986, 15 million ECU), aimed at removing of bot tlenecks which inhibit the application of molecular and cellular biology to agriculture and the agro-food industries. The second programme, the Biotechnology Action programme (BAP, 1986-1990, 75 million ECU) was directed towards medium and long-term objec tives essential to the strategic strength of European biotechnological industry and Euro pean agriculture. Innovative in supranational cooperation were the European Laborato ries Without Walls (ELWWs), projects in which laboratories from various member states cooperate on enzyme engineering, genetic engineering and cell culture technology and in transnational joint projects in bio-informatics, culture collections and risk assessment. The third programme, Biotechnology Research through Innovation and Development in Europe (BRIDGE, 1990-1994, 100 million ECU), introduced, apart from transnational cooperation, target projects (T-projects) aimed at developing larger cooperation structures to remove important botdenecks in structure and scale in several selected areas. From the management viewpoint the T-projects are considerably more complex than the ELWWs and have a structure of contracted laboratories supervised by a coordinator, a monitoring unit and, in a number of cases, a platform of interested companies. The BRIDGE pro gramme focuses on fundamental research into protein structures and gene analyses (ena bling technologies) and molecular and cell biology of organisms important to agriculture 14 and industry, modernization of database infrastructures and culture collections and the development of an EC research basis in the field of pre-normative research (i.e. in vitro toxicological tests and risk assessment). The current programme, BIOTECH (1992-1994, 160 million ECU), is being implemented. The programme focuses on three approaches, i.e. molecular approaches (proteins, genes), cellular and organism approaches (cellular regeneration, reproduction and development, metabolism studies and communication systems within living matter) and ecology and population biology (risk assessment and conservation of genetic resources). A new focus in the biotechnology programme is the possibility of obtaining funds for socio-economic and ethical studies. Concerning techno logical cooperation the possibility has been created to implement projects with technolog ical priority in which a considerable amount of money has been earmarked for extensive coordinated cooperation structures based on decentralized management. The first of these projects is AMICA (Advanced Molecular Initiative for Community Agriculture), which focuses on molecular plant biology. Dutch research groups and trade and industry make a sizeable contribution to coopera tion in EC context, which benefits the health and effectiveness of biotechnology in the Community. About ten percent of all laboratories participating in the mentioned EC biotechnology programmes are Dutch. 1.4.2 Eastern Europe In recent years, especially international cooperation has focused on Central and Eastern Europe. Central and Eastern European countries have high expectations of the develop ment of their economies and agriculture through new technologies, including biotechnol ogy. Western countries, the Netherlands among them, have much to offer to these coun tries and various forms of scientific cooperation have developed with Hungary and Bul garia. The Dutch government and the European Community intend to expand coopera tion in the years ahead. Promotion budgets have been earmarked to this end. 1.4.3 Developing countries The forms of cooperation described above concern the competitive power in an increas ingly interdependent world. Another form of cooperation, in principle of a non-economic nature, is development cooperation on behalf ofpoore r countries in Asia, Africa and Latin America. In the past decades the Netherlands has set great store by the development of these countries and is among the countries spending the highest percentage of the Gross National Product on projects to eliminate poverty in the world. Biotechnology is seen as a technological tool which may contribute to realizing the objectives of food security and disease prevention. Recent World Bank prognoses show that world food production must increase by 2.6% annualy to feed the world population in the next decades. Biotechnology, provided it is used correcdy, is in many ways an appropriate technology for developing countries. Application of biotechnology is often uncomplicated. Usually, production pro cesses based on biotechnological principles do not require much energy and are not very capital-intensive. Furthermore, a large turnover of biomass and a wide variety of biore- 15 sources are conditions that are favourable to devel The Special Programme Bi oping biotechnology in 'Third World' countries. How otechnology and Develop ever, biotechnology may also lead to social, economic ment Cooperation ofth e Di and environmental problems. In particular, the substi rectorate General Interna tution of raw materials, industrialization of agriculture tional Cooperation (DGIS) and privatization ofknowledg e may pose threats to the of the Dutch ministry of For export markets of developing countries and may eign Affairs was started in widen the R&D gap between the developing and the January 1992. This pro industrialized countries. gramme is intended to facili tate the accessibility of bio Several years ago the Dutch government had al technological know-how to ready recognized the necessity of an active policy to developing countries in what allow developing countries to benefit from the advan is called the 'Third World' tages ofbiotechnolog y and to realize more harmoniza and also to contribute to so tion between biotechnological research and education lutions for developmental in Dutch laboratories and in developing countries. Re botdenecks. The pro cently, a special programme on biotechnology was gramme is to run for five launched, aimed at enabling developing countries to years and will have a budget of 50 million Dutch Guilders develop local research capacity and to encourage the (± 22 million ECUs). The development of relevant research in industrialized programme covers agricul countries (see marginal note). An inventory has shown ture, health care and envi that six percent of the biotechnology research at ronmental management and Dutch universities and institutes has direct relevance aims to concentrate cooper to developing countries and twenty-five percent indi ation with Zimbabwe, rect, that is,wit h slight adjustment. Apart from studies Kenya and Columbia on be of research financing, studies are also being initiated half of the under-privileged. into the social and ecological aspects of biotechnology In close consultation with in developing countries, such as technology assess the partners, the choices of research topics will be made ment, socio-economic studies and risk assessment. for short-term projects like Moreover, the set-up of a biotechnology policy in de setting up tissue culture facil veloping countries is encouraged in particular by the ities as well as long term in dissemination of information, support in the develop vestments necessary for pro ment of adequate legislation, and organization of spe ducing transgenic plants and cific workshops. The starting point of the policy is to genetic maps. A research realize an international, coordinated approach to project on cassava has al gether with other donors and within the international ready been started in Zim organizations, such as the United Nations, CGIAR, babwe. the European Community and the World Bank. 1.5 References Anonymous, 1992. Facts and Figures on agriculture, nature management andfisheries. Ministr y of Agriculture, Nature Management and Fisheries of the Netherlands, The Hague. Houwink, E.H., 1989. Biotechnology, controlled use of biological information. Kluwer Academic Publishers, Dortrecht, Boston, London, pp. 1-4. 16 2 Biotechnology in plant breeding and crop protection Contents 2.1 General introduction to biotechnology in plant breeding and crop pro tection 19 2.1.1 The history of plant breeding and crop protection 19 2.1.2 Biotechnology in plant breeding and crop protection in the Netherlands 20 2.2 Molecular biometrical genetics: a vigourous hybrid 23 2.2.1 Molecular genetics 23 2.2.2 Biometrical genetics 24 2.2.3 Synthesis 24 2.2.4 Assessing associations, QTL mapping 25 2.2.5 Recent developments in gene mapping and QTL mapping 27 2.2.6 Application to plant breeding 28 2.2.7 Needs for the future 29 2.3 Genetic modification of plants: new developments in transformation procedures 31 2.3.1 Agrobacterium-mediated transformation 32 2.3.2 Transformation of monocotyledons 36 2.3.3 Improvement of genetic modification procedures 37 2.4 New approaches for identifying and isolating plant disease resistance genes 39 2.4.1 Transposon tagging of disease resistance genes 39 2.4.2 Map-based cloning of disease resistance genes 42 2.4.3 Conclusions and prospects 43 2.5 New strategies for obtaining fungal-resistant plants 44 2.5.1 Production of non-proteinaceous fungal toxic compounds in plants 44 2.5.2 Production of antifungal proteins in plants 45 2.5.3 Transfer of monogenic resistance genes 47 2.5.4 The two component sensor system 48 2.5.5 Conclusions 50 17 2.6 Molecular breeding for insect-resistant plants 50 2.6.1 Bacillusthuringiensis genes encoding insecticidal crystal proteins 50 2.6.2 Proteinase inhibitors from potato tuber for insect control 54 2.6.3 Future developments 57 2.7 'Plantibodies': a versatile approach to engineer resistance against pathogens 57 2.7.1 Monoclonal antibodies as inhibitors 58 2.7.2 Engineering monoclonal antibodies 58 2.7.3 Engineering durable resistance against potato cyst nematodes 60 2.7.4 Conclusions 61 2.8 New developments in molecular flower breeding 61 2.8.1 Manipulation of flower colouration by (anti)sense genes 62 2.8.2 White pollen of Petunia is male sterile 63 2.8.3 Studying the genes involved in flower development 64 2.9 Industrial application of plant biotechnology: promises and pitfalls 66 2.9.1 Biotechnology and agricultural challenges 67 2.9.2 Aspects related to commercial applications 67 2.9.3 Conclusions 70 2.10 Market introduction of products obtained through plant biotechnol ogy, aspects of consumer acceptance 71 2.10.1 What is 'better consumer acceptance', and how can it be promoted? 71 2.10.2 Public attitude towards food and food production 71 2.10.3 Previous introductions of new agricultural products onto the market 73 2.10.4 Product information, labelling = libelling? 73 2.11 References 74 2.1 General introduction to biotechnology in plant breeding and crop protection PJ.G.M de Wit & L. van Vloten-Doting 2.1.1 The history of plant breeding and crop protection Man has selected crop plants for thousands of years for properties such as higher yield, quality and resistance to disease and pests. Many crops were selected for adaptation to local environments and developed into land races. Over the centuries, many genes ex pressing these properties have been accumulated in different plant species. At the end of the last century, a few decades after the initial genetic studies by Mendel, a more scientific and systematic approach was used in crop breeding and crop protection. Since the begin ning of this century crop yields have increased steadily, mainly as a result of breeding efforts and this has led to the 'green revolution'. Modern cultivated crop plants differ significantly from their wild ancestors. Without modern breeding activities they would not exist. During the selection process for their high yielding and high feeding properties they have often lost many qualities related to competitive ability. The first positive results on resistance breeding were obtained by Biffen (1905), who discovered that resistance to yellow rust in wheat obeyed Mendel's laws. Since then, many genes for resistance to pathogens have been introduced in crop plants. During cultivation of resistant cultivars, produced by breeders, variants of pathogens able to evade newly- introduced resistance genes have appeared. Hence resistance breeding is a continuous process. The interplay between host and pathogen is often based on a gene-for-gene relationship which is characterized by the observation that a newly introduced resistance gene is eventually overcome by the pathogen (de Wit, 1992). Nevertheless, there have been more cases of success than failure in plant resistance breeding. Some sources of resistance have been effective for decades. To feed the ever-growing world population there is a need for crop improvement and crop protection. Each year around 30% of the world's agricultural products are lost because of pests, weeds and disease. Crop breeding over the past 40 years has gready improved agricultural production, which however has also often required the intensive use of fertilizers and pesticides. Since the intensive use of pesticides after World War II, the effort put into resistance breeding decreased significantly. Because the agrochemicals were often very effective and rather cheap, interest in resistance breeding decreased. However, during the last decade breeding for resistance isonc e again being given high priority as the need for agricultural production to become less reliant or in some cases even independent of agrochemicals increases. There is a strong call for environmentally friendly agricultural production which isbecomin g more consumer-driven than farmer-driven. The proportion of the population active in farming has decreased drastically during the second half of this century, resulting in the political power of the farmers being diminished. Political change and general concern among the population, including the farmers, about the continuous application of agrochemicals has greatly stimulated the search for alternative ways of growing plants and protecting them against disease and pests in sustainable agricultural 19 Systems. Immediate elimination of the use ofal lpesticide s would decrease the average yield of agricultural and horticultural products by more than 50%. Finding alternative solutions is more easily said than done. There is a need to apply fundamental knowledge when searching for the solution to agricultural problems. We are facing two tremendous tasks: (i) to find new and safe strategies for crop improvement and crop protection and (ii) to feed a world population which could double within the next forty years, reaching 10-15 billion people by the year 2030. It isclea r that classical and molecular plant breeding will be the most productive and safe ways of crop improvement and crop protection in the coming decades. In the future we will need to rely more and more on disease resistance genes, biological control and integrated pest management, using wherever possible only non-polluting agrochemicals. 2.1.2 Biotechnology in plant breeding and crop protection in the Netherlands In this chapter we will focus on research strategies in modern crop breeding and crop protection in the Netherlands. Crop breeding and crop protection have always been important issues in agriculture in the Netherlands, which is the world's second largest exporter of agricultural products. Not all strategies will be discussed in detail as some have already been elaborated upon in the first book of the Dutch Programme Committee on Agricultural Biotechnology (Dekkers et al., 1990). Some promising new technologies which will hopefully prove very useful in solving some of the obstacles encountered in modern crop breeding and crop protection will be discussed. These include: 1. determining and locating quantitative trait loci (QTL) using restriction fragment length polymorphism (RFLP) technologies; 2. constructing genetic maps with RFLPs and random amplified polymorphic DNAs (RAPDs) and related technologies; 3. identifying genetic loci in plants using transposon tagging; 4. searching for genes, the gene products of which interfere with or inhibit pathogens and pests; 5. unravelling the molecular basis of disease resistance, the principles ofwhic h can be used in modern crop protection. The different approaches will be briefly mentioned in this section and will be discussed in more detail in the following sections of this chapter. Molecular biometricalgenetics Molecular genetics have disclosed vast amounts of hitherto hidden genetic variation at the DNA level with the help of RFLPs, RAPDs and PCR products. Especially when using PCR products, almost any number of polymorphisms can be uncovered. Quantitative traits which have always been difficult to transfer in breeding programmes in a systematic way, now become amenable to study by biometrical genetics. As a vast array of markers is available and will grow continuously it is becoming imperative to obtain reliable statisti cal and computational tools for processing molecular marker information. Linkage rela tion among the different markers is required to optimally transfer as many positive traits 20 to the next generation as possible. This subject is discussed in Section 2.2. New developments in transformationprocedures In the early eighties, the production of the first transgenic kanamycin-resistant plants expressing the bacterial gene for aminoglycoside phosphotransferase was oudined. Since then other transgenic plants with new important traits have been described, some of which are about to be released on the market. However, only dicotyledonous plants can easily be transformed with the bacterium Agrobacterium tumefaciens and among those plants only some of them can be transformed efficiently. Many important dicotyledonous crop plants and virtually all monocotyledons are still difficult to transform (van Wordragen &Dons , 1992). Recent developments in transformation procedures will be discussed in Section 2.3. From this section, it will be evident that applying technologies from model plants to crop plants is often more difficult than anticipated. New approachesfor identifying and isolatingplant genes There is a growing need to identify and isolate genes encoding important traits. The number of genes with potentially valuable traits, which are currently available, is limited. Transposon tagging and map-based cloning are helpful techniques for identifying and localizing additional genes. These techniques are discussed in detail in Section 2.4. New strategiesfor obtainingfungal-resistant plants Transgenic plants, resistant to bacteria and fungi, have recently been reported in the literature. These plants express some members of the pathogenesis-related (PR) proteins. Other approaches are based on the expression of antibacterial or antifungal proteins from other organisms. These approaches based on the introduction of one gene encoding anti microbial activity, may include the risk of developing resistance after introduction into the plants, due to increased selection pressure. A novel approach based on the expression of the pathogen's avirulence genes exploits an array of natural plant defense responses initi ated by the hypersensitive response (de Wit, 1992). A number of different approaches, some of which some are still at a premature stage, is discussed in Section 2.5. Recent developments in molecular breedingfor insect-resistantplants Genes encoding insecticidal toxins from Bacillus thuringiensis (Bt-toxins) were some of the first to be introduced in plants. There are many of these toxins that are specific to different genera of insects. Globally considerable research has been focused on Bt-toxins and their application as bacterial preparations (formulation) or in transgenic plants expressing differ ent types of Bt-toxins (Feitelson et al., 1992). Transgenic plants expressing proteinase inhibitor genes (Pis) are currendy being tested for resistance to insects. The latest results on Bt-toxin research in the Netherlands are discussed in Section 2.6. An alternative strategy towards insect pest control in agriculture which has received considerable attention from Dutch agro-biotechnological research is the use of baculovirus, both recombinant and selected wild strains. This is not presented here as it has been discussed by Kool et al. (1990) in 'Agricultural Biotechnology in Focus in the Netherlands'. 21 Recent developments in engineering vims-resistant plants Transgenic plants resistant to viruses are not discussed in a separate section. However, considerable research is being devoted to this subject in the Netherlands. Initially, the research was mainly concentrated on expressing genes encoding coat proteins of RNA viruses in different plants. This resulted in approximately 12 transgenic plant species expressing coat proteins of different RNA viruses,whic h as a result, are now resistant to the respective viruses. The first application process to obtain plant breeders' rights for a genetically engineered cultivar (a potato virus X resistant potato) has been started. The work is being done in the Department of Virology at Wageningen Agricultural University and in the Department of Biochemistry at the Leyden University. Other strategies for obtaining virus-resistant plants are being investigated. These strategies, recently reviewed by Carr (1992), include transgenic plants expressing: satellite RNAs, antisense viral RNA sequences, nucleocapsid proteins of negative strand viruses, defective viral polymerases or defective viral transport proteins. Strategiesfor obtaining nematode-resistant plants No reports are yet available in the literature on transgenic plants resistant to nematodes. Approaches similar to those used to obtain plants resistant to bacteria and fungi are being tested for nematodes. Thus, many proteins are being tested for nematicidal activity and the genes encoding active proteins are being transferred to plants. In several laboratories another strategy is currently being employed. Genes encoding antibodies directed towards pathogenicity factors of nematodes are expressed in plants. The transgenic plants express ing 'plantibodies' will be assayed for nematode resistance (Schots et al., 1992). Many aspects concerning the expression of plantibodies and even more importandy aspects concerning the significance of pathogenicity factors in the infection process still need to be investigated. The plantibody approach which is being studied in the Netherlands and in many other European countries within the framework of the EC-Biotech programme, will be discussed in Section 2.7. New developments in molecular flower breeding The Netherlands is particularly well known for its flower production and flower breeding. Recent advances in molecular biology have led to a better understanding of the molecular basis of flower colour and flower development. The enzymes and genes involved in flower pigmentation have been isolated and cloned, respectively, and in many cases their regula tion has been studied. Expression of genes involved in flower pigmentation in the antisense orientation can dramatically change the pigmentation (Mol et al., 1989). Some of these genes also seem to play a role in gametophyte formation. Permission has been requested to release the first genetically altered chrysanthemum onto the market. These and other aspects of molecular flower breeding are discussed in Section 2.8. Industrial application ofplant biotechnology This chapter on current research in the biotechnology of plant breeding and crop protec tion would not be complete without discussing the economic aspects of plant biotechnol ogy and, just as important, the acceptance of biotechnology by the public and the Dutch 22 public in particular. The transfer of valuable new genes to plants and the subsequent production of high quality transgenic plants or seeds is a complicated and costly process. The length of time from isolating, cloning and transferring a desirable gene, followed by regulatory reviews, and scaling-up to a saleable product may be as much as 5 to 10 years. At present, often only one trait is introduced, while the economic value of seeds usually depends on many important traits. Of course the importance of the introduced trait determines its economic value, but the profits should not be overestimated. Notwithstand ing those important economic aspects, plant biotechnology still holds many promises for the future, and these are discussed in Section 2.9. Acceptance ofplant biotechnology by the consumer In parallel with the development of transgenic plants the public should be informed about all possible aspects of plant biotechnology. The public must not only accept but rather support plant biotechnology in the long run as a means of producing food in an environ mentally safe and durable way. If the consumer does not accept transgenic plants, then a transgenic product will never enter the market. These days, more than in the past, agricul ture and horticulture is consumer-oriented rather than producer-oriented. The consumer determines whether transgenic plants or their products are given the chance of entering the market. The various aspects related to the acceptance oftransgeni c plants by the public are discussed in Section 2.10. 2.2 Molecular biometrical genetics: a vigourous hybrid P. Stam This section focuses on the use of molecular markers of nuclear DNA in plant breeding without paying any attention to the molecular aspects as such, or the laboratory techniques involved. 2.2.1 Molecular genetics Since a molecular marker reflects chemical variation at a particular site of the DNA, such a marker is passed from one generation to the next in exactly the same way as genes are. In other words, molecular markers behave as classic Mendelian genes, and allow classic genetic analysis. In particular, by analyzing their co-segregation in segregating offspring generations, their genetic linkage can be established and measured (see Fig. 2.2.1). The smaller the physical distance between two sites of a chromosome, the tighter they are linked, and, as a consequence, the more likely they are passed together, via the gametes, to the next generation. Classic linkage experiments have resulted in genetic linkage maps with over 200 molecular markers in several major crops and model crops (Romero- Severson et al., 1989). 23 Fig. 2.2.1. Gametic output ofa n individual carrying distinct alleles attw olinke dloci .Th e parental allelic combinations (M,T and m,t) tend to persist among the gametes. The recombinant types (M,t and m,T) arise only through recombination at meiosis (crossing over). As a result, the allele pairs M,T and m,t will also be associated in the offspring. Tighter linkage means less recombination and increased association ofalleles . 2.2.2 Biometrical genetics Classic genetic analysis of quantitative characters is hampered by two factors: (a) their assumed polygenic nature (many genes with a small effect), (b) their sensitivity to environ mental variation. As a result neither the individual genes nor their effects can be identified; a continuous variation rather than the discontinuous variation, typical of Mendelian characters, is observed. In quantitative, or biometrical genetics, populations are described in statistical terms, i.e. means, variances, covariances, correlations, variance components, etcetera. The effects of both polygenes and environmental factors are described in the same statistical terms. Despite the lack of knowledge about individual gene effects on most quantitative characters, biometrical genetics has made great contributions to plant and animal breeding. This is due to the fact that the correlation between relatives (especially parents and their offspring) can be described in statistical terms, based on the Mendelian nature of the underlying (though unidentified) genes. 2.2.3 Synthesis The availability of numerous molecular markers has stimulated geneticists to develop new ideas to use these markers for the genetic location of known and unknown genes, and as a tool for indirect selection of various traits. The underlying idea is simple. When the 24 genome of an organism is covered with hundreds of marker positions, some of these markers willb e closelylinke d to (atleas t some)gene s and thevariant s ofthes emarker swil l showlinkag e associationwit hth echaracter s determined bythes egenes .Artificia l selection in a population on the basis of marker genotypes will then result not only in a change in frequency of marker variants (alleles) but, through 'linkage drag', will also change the frequencies of alleles of the genes that are of interest. The theoretical possibilities for this 'marker assisted selection' are very promising. It takes little imagination to envisage the future plant breeder who samples DNA from seedlings, 'fingerprints' the DNA for those markers that were previously shown to be associated with e.g. yield, and selects on the basis of the marker genotype, instead of growing a whole crop and assessing yield potential in laborious field trials. There are, however, a few steps to be taken before a breeder can really rely on markers rather than on the phenotype. These steps involve statistical and biometrical procedures, some of which are well known from classic quantitative genetics, and some of which require the development ofne w statistical tools.Ho w associations between agronomic characters and molecular marker variants can be assessed and how this information can be used to optimize plant breeding programmes are briefly sketched in the following sections. The current state of affairs is discussed together with today's bottlenecks and, based on the author's experience in thisfield, a fe w ideasar e presented with respect toth e current needs of everyday plant breeding. 2.2.4 Assessing associations, QTL mapping Assessing linkage association between molecular markers and agronomic traits basically requires the same type of experimental set-up as the classic genetic experiments for the detection oflinkag e between 'normal' Mendelian genes. To this end parents which carry distinct variants at a (preferably large) number ofmolecula r markers are chosen; in addi tion,thes eparent s shouldb egeneticall y different for the agronomic character(s) ofinterest . One parent, for example, might be a late-flowering, drought-tolerant cultivar, resistant to a certain disease, whereas the other parent is early-flowering, suffers from drought stress and susceptible to the same disease. These parents are crossed; in the second (or later) generation (an F2, F3, or a backcross) genetic segregation will occur for allgeneti c factors that made up the difference between the parents. Careful inspection of co-segregation of markerspa sean d ofmarker s and agronomic traits then gives a clue to the genetic linkage of markers per se and of markers and genes involved in the traits. For the markers, which behave as Mendelian genes, such an analysis amounts to a classic linkage analysis; the markers can be assigned to linkage groups (corresponding to chromosome pairs), and within linkage groups can be ordered linearly into a genetic linkage map. The same procedure can be applied tothos e traits which are monogenic, i.e.trait s with two or three clear-cut phenotypic classeswhic h correspond to a single gene difference. Examples ofth e latter are monogenic disease resistances. However, to assess the association between a continuously distributed trait and a marker, one has to rely, much the same as in a classic biometrical analysis, on means, variances, and possibly other statistical concepts. Soller, Weiler and co-workers were among the first to develop a theory on the power of 25 experimental designs to detect associations and procedures for estimating linkage between markers and quantitative trait loci (QTL) (Soller et al., 1976; Weller et al., 1989, and references therein). These estimation procedures aim at estimating the recombination frequency between a marker and a putative QTL, assuming a continuous Gaussian distri bution of the trait. The assumption concerning the distribution is required because these methods employ the concept of maximum likelihood. Also based on maximum likelihood (and thus requiring a specified probability distribution for the trait) is the 'interval map ping' procedure developed by Lander & Botstein (1989). This method currently seems to be the one most widely used, probably because it has been implemented in a computer package, MAPMAKER/QTL. Interval mapping can be described as scanning the genetic marker map for the presence of QTLs. For every possible position between a pair of adjacent markers the relative likelihood (LOD) of a QTL to be present can be calculated. A graphic representation as a plot of LOD values against map position yields a quick overview of the position of possible QTLs. Fig. 2.2.2 presents such a likelihood profile obtained from interval mapping; it also illustrates a situation where the method fails, i.e. in the detection of two linked QTLs. real QTLs at 'ghost' QTL 21- positions 20 at position 40 and 60 (approx.) 19 17 15 ° 13-1 _i 1H 9 7 5: 3: 10 20 30 40 50 60 70 map distance (cM) Fig. 2.2.2. Locating quantitative-trait loci (QTL) on ageneti c map requires sophisticated statistical procedures. The figure illustrates one of the draw-backs of 'interval mapping'. The horizontal axis is the map distance along the chromosome; markers on this chromosome are indicated by arrows. Plotted isth e relative likelihood (LOD) score for a hypothetical single QTL; thepositio n ofth e peak indicates the most likely location of the QTL. In the case shown here, data were generated by computer simulation of an F2 with twoQTL s segregating, at positions 20 and 60. It is seen that 'interval mapping' reveals a 'ghost' QTL in the wrong position. Artefacts such as these can be avoided by combining maximum likelihood and multiple regression methods Jansen, 1992; Stam, 1991). 26 2.2.5 Recent developments in gene mapping and QTL mapping The power of the successful Lander &Botstei n method for QTL mapping seems, however, to have been overestimated. Notwithstanding its elegance and user-friendliness it suffers from a few disadvantages. The first of these is the required Gaussian, or normal distribu tion of trait values. Applied biometricians are familiar with the fact that many agronomic traits do not obey a normal distribution, even after transformation of scale. This would be no major problem if interval mapping were robust, i.e. insensitive to violations of the underlying assumptions. Preliminary results ofva n Ooijen (1992, unpublished) and Jansen (1992) indicate the contrary; a non-Normal distribution of trait values may easily lead to false positives (i.e. apparent peaks in LOD graphs resulting from random error rather than QTLs) and/or wrong positioning of QTLs. But even when the requirement of normality is met, interval mapping does not provide an exact statistical test. In a recent paper, elucidating the computational aspects of the method, van Ooijen (1992) discusses the power of interval QTL mapping. Comparing his theoretical results with the putative QTLs in tomato (Paterson et al., 1991), it appears that some of the latter are very likely to be false positives. A second shortcoming of interval mapping is its lack of ability to detect linked QTLs with opposite effects (repulsion phase). Unless such a configuration is known in advance (which is purely theoretical) interval mapping will either detect no QTL at all, or assign a single one to the wrong position (see Fig. 2.2.2). This point was discussed by Stam (1991) whose results suggest that using the markers as regressors in a multiple linear regression may overcome this problem, provided it is a dense marker map (see alsoJansen , 1992). Jansen (1992) recently described the problem of QTL mapping in terms of a very general statistical model. He demonstrates that multiple regression and maximum likeli hood are easily incorporated when formulated in terms of'generalized linear models'; this implies that the model is not restricted to normally distributed trait values: in fact any type ofprobabilit y distribution can be used; in addition, experimental design factors can also be incorporated. The Jansen approach allows multiple QTLs (linked or unlinked) to be located simultaneously. Though very promising, this method needs further exploration; in particular, the aspects of stepwise regression (adding or omitting putative QTLs) and statistical testing need to be investigated before its merits as a general tool can be judged. Another recent contribution, though not directly related to QTL mapping, isth e devel opment of an algorithm (implemented in a computer package,JoinMap ) by Stam (1992) to generate 'integrated' genetic maps. Contrary to most other software for linkage analysis, JoinMap can cope with linkage data obtained from distinct types of experiments, such as F2s, backcrosses, or recombinant inbred lines (RILs).Usin g this approach, several separate RFLP maps and the classic genetic map ofArabidopsis thaliana have been integrated (Hauge et al. 1992). Fig. 2.2.3 presents part of this composite linkage map, together with its constituent maps. 27 CLASSICAL RFLP 1 RFLP 2 INTEGRATED sti -, y- 0.0 g4133 -, y- 0.0 pCITd112a -, y- 0.0 pCITd112a -, r 0.0 g4553 - - 4.2 m246 , ,11.1 PCIT1291 A I 13.6 m246 - -11.2 04133 v\ ', 16.3 pCIT1291 - - 13.8 sti i I 16.6 g4553 -- - 20.4 fve v - 23.0 a14G4 , ,31.7 hy3 - - 26.2 cp2 - - 27.4 pCITdIOO A I 34.4 cp2 " -» 28.9 a14G4 -, .- 31.8 m25t - - 33.1 m10 4 ,\ /, 38.9 pCITdIOO - - 34.5 fve A /, 39.9 er \ ^ 39./ m104 - - 39.3 hy3 > ', 42.8 hy1 - - 39.9 er - - 41.0 hy3 ' ^ 40.4 cp2 N / - 44.9 vv - m251 - - 46.9 m25l - , 45.9 - 49.9 ASA-2 - - 48.1 copl -- Gpa1 > - 47.8 s 54.7 re N as - - 53.9 Gpal , 57.2 er ^ , 56.2 as - - 56.3 N aux / ^ 57.5 er ^ •* 58.0 hy1 ' >>(, 57. 56.51 m336 - - 63.7 py - - 62.1 py - - 61.9 sul - - 65.9 cer8 - J- 67.7 ASA-2 - - 65.2 cerfl - - 69.7 m429 - - 71.5 copl ^ ^ 66.9 m429 't ^ 69.6 fpa - - 79.2 al \ 73.4 m336 - - 84.3 oux '/, m336 1. te sul '/ \* 83.0 cerfl ' 1 86.8 fpo -T~ 96.4 pCITN7-26 - - 101.2 PCITN7-26 /\ 99.5 Fig. 2.2.3. Integrated genetic maps are useful tools for molecular geneticists (map-based cloning) and plant breeders (choosing markers for indirect selection). Shown are three distinct maps of Chromosome 2 of Arabidopsisthaliana: th e 'classical' map and two DNA marker maps. The inte grated, or joined map was produced with the software package JoinMap (Stam, 1992). (Data courtesy E. Meyerowitz, H. Goodman and M. Koornneef.) Markers occurring on more than one map are shown in serif/bold face. Only part of the markers actually used in mapping are shown. 2.2.6 Application to plant breeding Once significant linkage associations between markers and agronomic traits have been established, the gate to indirect selection via markers is open. However, it should be emphasized that, stricdy speaking, the observed associations will only hold with certainty in the population in which they were found. An association found in the offspring of a particular cross does not apriori extend to other crosses, even when segregating for the same markers. A different genetic background could mask the effects of certain QTLs, whereas others, so far not detected, might become manifest. Also, the same QTLs could be involved, but the direction of association might be reversed. In ajoin t effort of European maize breeders, the 'EUREKA RFLP maize project', this question (among others) of generalization is currently being investigated. The question directly relates to the perspectives of indirect selection via markers; if associations need to be assessed separately for each of the crosses in a breeding programme, the cost would probably be prohibitive. This is especially true of quantitative traits which require large replicated field trials at several locations for linkage detection. Similar to the experiments in tomato reported by Paterson et al. (1991), the EUREKA maize project is also investigating the effects of environmental differences on the expres sion of QTLs in distinct crosses. 28 One of the most obvious ways to exploit molecular markers in breeding is in 'guided introgression'. Given a dense molecular marker map, specific favourable chromosome segments (favourable because they are shown to be associated with e.g. resistance or a yield component) can be 'traced' in a breeding programme. Selecting for markers, bracketing such segments, in a backcross programme substantially increases the efficiency of substitut ing the 'donor' genome (Young & Tanksley, 1989; van Ooijen, unpublished results). Bearing in mind that associations between markers and agronomic characters are not a prioritransferabl e to other crosses, the author is currently investigating the perspective of establishing and exploiting associations simultaneously over a number of crosses; possibly a (partial) diallel. The idea then is to identify, on the basis of markers, the most 'promising' combination of parents. A subset of two, three or four parents from a larger set can, in theory, be identified as the one which, upon crossing, will generate with the largest probability the most favourable combination of markers (and thus linked 'agronomic' genes). This approach to the use of markers for 'stacking' favourable genes from distinct sources will be further investigated in the near future. A potential field of application for this approach is combining genes for partial resistance into a single genotype. In a field test distinct genes for partial resistance may go unnoticed; when labelled with markers they will not. 2.2.7 Needs for the future When, some 10 years ago, the idea of using molecular markers in plant and animal breeding became widespread, some quantitative geneticists were sceptical, probably be cause of the long established concept of the polygenic nature of quantitative traits, which plays such an important role in the theory of quantitative genetics. In this concept the genetic basis of a quantitative character is a large number of genes, each with a small, unmeasurable effect. If indeed most agronomic traits were polygenic in this sense, little is to be expected from QTL detection via markers and indirect selection. In addition, the following argument could be heard in these circles: 'If a marker analysis really detects postulated QTLs, these must represent genes with large effects (major genes) that can easily be dealt with by traditional breeding and selection techniques'. However, the efforts of QTL detection have revealed that for various agronomic traits which typically show continuous variation (e.g. earliness and plant height in maize, pH, fruit size and soluble solid content in tomato), relatively few major genes are responsible for the genetic varia tion in a particular cross (Romero-Severson et al., 1989; Paterson et al., 1991). This by no means indicates that these traits are not polygenic; it could simply mean that some genes are more important than others. If a breeder can get hold of these genes by means of indirect selection via markers, he may have covered some 80 percent of the route to his selection goal. The question ofwhethe r or not this route would be cheaper and faster than the conventional one depends on crop- and/or character-specific factors, rather than on the methodology itself. If genotyping at a few marker loci can replace an expensive and time-consuming disease resistance test, the choice needs no debate. For other properties, requiring genotyping at more markers, the choice might be less obvious. So far, QTL location and indirect selection has been carried out on a limited experi- 29 mental scale; the published results describe test cases rather than routine work in running breeding programmes. The Population Biology group at CPRO-DLO has been (and still is) involved in a number of such research programmes, aimed at improving technology and methodology rather than actual breeding. From experience gained in these program mes it has become clear that, apart from the laboratory techniques, some conditions in the area of data collection and data processing must be fulfilled before the application of marker technology can be realized on an industrial scale in plant breeding. The first of these conditions is the availability of a computerized data management system. In view of the growing speed with which molecular data of the major crops are being accumulated, the need for electronic data storage and management is paramount. In applied plant breeding, existing databases will have to be re-designed or adapted so that molecular information at any level (probe identification, probe/enzyme combinations, RFLP patterns, molecular weights of fragments, primer sequences, genetic maps, etc.) can be incorporated and linked to the usual agronomic information. On a limited scale such a database is currently being developed at CPRO-DLO as part of a project financially supported by a group of Dutch breeding companies. Secondly, reliable statistical and computational tools for processing molecular marker information are needed. These comprise genetic mapping and QTL mapping procedures. Preferably, they should be implemented in software packages, either as stand-alones or as modules in widely used general statistical packages. In addition to existing mapping soft ware (e.g. LINKAGE, MAPMAKER), the newJoinMa p package (Stam, 1992), covering various types of linkage experiments, seems to be a good candidate for genetic mapping as such. Concerning QTL mapping however, the methodology still needs substantial funda mental research before routine-wise application to the wide variety of data types encoun tered in applied plant breeding is within reach. On the road from QTL mapping theory to practice one comes across obstacles not accounted for when initially developing the theory. Examples of these are systematic segregation distortion in wide crosses, time trends in the epidemics of pests and disease, the use of different (sometimes arbitrary) scales in scoring for resistance/susceptibility, and last but not least, missing observations in both molecular and agronomic data. In particular, the ordinal type of data, so often used in measuring resistance, need further attention by statisticians; a possible way out could be the use of threshold models for ordinal data. The ad hoc solution (a non-parametric statistical test) applied by van Ooijen (Van Oooijen et al. 1992) in a case of distorted segregation and ordinal scoring of resistance to Clavibactermichiganensis in tomato, is a good example of how such problems are currently being handled. Thirdly, indirect selection via markers requires statistical prediction ofagronomi c values on the basis of marker genotypes. In a demonstrative set-up it has been verified that indirect selection of a single quantitative trait via RFLP markers may indeed result in a considerable response to selection (Lindhout, 1992, pers. comm.). When it comes to statistical prediction, several approaches can be taken, depending on the breeding programme and the breeding goal. Predicting for a single character in a pedigree selection programme could be based on multiple linear regression, using the markers as regressors. Simulation studies by the author indicate that this method can easily outrival phenotypic selection, provided the associations between markers and the trait of 30 interest have been accurately assessed. Multiple trait prediction could be based on the same procedure, possibly by using a 'composite' trait the components of which are given economic weights. The statistical theory for classic index selection is well developed, but needs to be extended to cope with selection via markers. The basic idea of prediction via an index (a composite trait) or other linear predictors istha t the markers involved are given weights so that the response to selection is maximized (the more a marker contributes to the total genetic variation, the higher the weight it is given in an index). Stacking favourable genes from a variety of sources via linked markers requires a different approach. In this case marker-character associations need to be assessed across a number of preliminary crosses. The goal then is not to proceed immediately by means of indirect selection, but to identify those parents which, after crossing, will generate the most favourable offspring. This in turn requires knowledge of the linkage relations among the markers (genes on different chromosomes are easier to combine via crossing than genes on the same chromosome). Thus, a genetic marker map will be very useful in this case. Sophisticated computational and statistical marker-based prediction techniques are not available at present. Ad hoc procedures, based on a good sense of genetics and biometry may carry the breeder a long way along the road to his goal. But if the optimistic view of the advocates of marker-assisted selection is to be validated, considerable research efforts are required in quantitative genetics and statistics. The gap between the biotechnologist's perception of marker technology ('knowing where the genes are makes life easy') and everyday plant breeding must and can be bridged by biometrical genetics. In view of the investments in biotechnology research over the past years, it seems unwise to neglect this need and not to seize this part of the return on investments. 2.3 Genetic modification of plants: new developments in transformation procedures JJ.M. Dons & WJ. Stiekema The number of crop species for which protocols for the introduction of genes are well- established is still limited. Hence research has been continued and intensified over the years to improve these procedures and establish them for more plant species and more cultivars. Fundamental aspects of efficient integration and expression of transgenes have also been investigated. Some of the major new developments in the field of genetic modification of plants will be reviewed in this section and, in line with the aim of this book, related to research in the Netherlands. Transformation research in theNetherlands The development of procedures for specific varieties of crops is mainly carried out by biotech companies and the major seed companies. They focus on improving the efficiency of existing methods and on the introduction of genes for economically important traits. 31 Research institutes such as CPRO-DLO, often in collaboration with industrial partners, are working on the development of methods for crops for which procedures are not yet available. CPRO-DLO and University research groups at the Leyden University and Wageningen Agricultural University, are also investigating basic problems related to gene transfer and gene expression. Table 2.3.1 lists the crops for which research on regeneration and transformation is currently being carried out in various laboratories in the Netherlands. 2.3.1 Agrobacterium-mediated transformation TheAgrobacterium system The most widely used technique for the introduction of alien genetic information into plant cells is based on the natural gene transfer capacity of the soil bacterium Agrobacterium tumefaciens (for reviews, see Hooykaas & Schilperoort, 1992; van Wordragen & Dons, 1992). Part of the tumour-inducing (Ti)plasmid , the T-DNA is transferred to the plant cell. Another region of this plasmid harbours the virulence {vir) genes, which are responsible for the process of gene transfer. By deleting the oncogenic genes, the 'disarmed' Agrobacterium strains have been established. The T-DNA and virgene s were physically separated on two plasmids leading to the binary vector system which is currently the most widely used system for Agrobacterium-mediated gene transfer (Fig. 2.3.1). The basic protocol for gene transfer of plant cells involves the co-cultivation of expiants with the appropriate disarmed Agrobacterium strain which confers a selectable marker gene, usually the kanamycin resistance gene. Subsequently, regeneration of shoots in the pres ence of kanamycin leads to the production of transgenic plants. A classic example of this procedure is the 'leaf disk transformation' developed for solanaceous species. Transformation ofcrop plants Van Wordragen & Dons (1992) have recendy reviewed and summarized the literature on Table 2.3.1. Research on the genetic modification of crop species in the Netherlands. Not included in this list are plant species used as model species, such as tobacco {Mcotiana tabacum), petunia {Petunia hybrida) and arabidopsis {Arabidopsis thaliana). Transgenic plants obtained Research in progress Potato {Solanum tuberosum) Barley {Hordeum vulgare) Rapeseed {Brassica napus) Rye grass {Loliumperenne) Sugar-beet {Beta vulgaris) Rose {Rosa sp.) Maize ÇÇea mays) Lily {Liliumsp.,) Tomato {Lycopersicon esculentum) Tulip {Tulipagesneriand) Cucumber {Cucumis sativus) Apple {Malussp.) Lettuce {Lactuca sativa) Willow {Salix alba) Carnation {Dianthus caryophyllus) Cabbage {Brassica oleracea) Chrysanthemum {Chrysanthemum morifolium) Pepper {Capsicum annuum) Rice {Oryza sativa) Cassava {Manihot esculentd) 32 macrocamer stoppingplat e microcamer +DN A sample (expiant,cells ) newprotei n Fig. 2.3.1. Genetic modification of crop plants. Currently two techniques are used preferentially for the transformation ofdicotyledonou s and monocotyledonous.A )Th e Agrobacteriumbinar y system for dicotyledons; B)Th e particle gun for monocotyledons, the scheme represents the particle deliv ery system of Dupont. the transformation of plants in the period 1987-1991. During that period successful introduction of genes using disarmed strains ofAgrobacterium tumefaciens was claimed for 39 plant species not previously mentioned in the literature. Integration of DNA in transgenic plants, confirmed by molecular analysis, was reported for 27 species By applying the stringent criterion that the introduced genes should be transmitted to the next generation, the number of new transformable species dropped to fifteen, five belonging to the Sola- naceae. This clearly shows that progress is slow, but that the number of economically important crop species amenable to genetic modification byAgrobacterium isgrowin g stead ily. It is also obvious that Agrobacterium-mediatedgen e transfer is predominantly restricted to dicotyledonous plants, notwithstanding the fact that some monocotyledons have been shown to be sensitive to Agrobacterium infection (Hooykaas & Schilperoort, 1992). Recently this was also shown for the monocotyledonous ornamental tulip (Wilmink et al., 1992). For arable crops good procedures are now available for potato and rapeseed. Trans genic plants have even been obtained for sugarbeet, which is known to be extremely recalcitrant. However the most important arable crops are monocotyledons (wheat, barley and maize), for which Agrobacterium-m.edia.ted gene transfer cannot yet be applied. The situation is comparable for horticultural crops. Transformation procedures have been developed for the most important vegetable crops, e.g. tomato, cucumber, lettuce, but genetic modification of monocotyledons, e.g. onion and leek is not yet feasible. Considerable progress has recently been made in ornamental crops due to activities of biotech companies such as Calgene Pacific (Australia) and DNAP (USA, in collaboration with the Dutch company Florigene) and CPRO-DLO in collaboration with Dutch breed ing companies. Agrobacterium-ha&ed transformation protocols have been developed for chrysanthemum and carnation. Positive results have been obtained for woody ornamen tals such as rose, and for trees such as apple, willow and poplar. Development of proce dures is in progress. As these species are sensitive to infection with Agrobacterium the prob lems encountered mainly deal with low regeneration potential. 33 In conclusion, Agrobacterium-meaiated transformation is now possible for most of the important dicotyledonous crops and it ist o be expected that in the next few years strategies will be developed for the more recalcitrant species (Table 2.3.1). However, it is also clear that procedures are often time-consuming, inefficient and genotype dependent. In addi tion, the methods described are not usually reproducible, resulting in the establishment of laboratory-specific protocols. These factors urge for improvement in transformation pro cedures, and therefore more basic information on the process ofgen e transfer in plants will be essential. This isimperativ e in order to meet the requirements for economically reliable applications (Table 2.3.2). Improvement ofAgrobacterium transformation In their review, van Wordragen & Dons (1992) summarize the most important factors influencing the efficiency of gene transfer: the Agrobacterium strains and their virulence; the plant genotype and its sensitivity to infection; the virulence genes and their induction; the developmental stage and hormonal balance of the expiant material; the use of selective agents and reporter genes. Some of these factors will be considered in the following paragraphs. Agrobacterium stains andvirulence. A large number ofAgrobacterium tumefaciens strains has been disarmed and are being used for transformation. They are derived from the wild-type strains Ach5, C58, A281, B6S3, T37 and A6 (for references see van Wordragen & Dons, 1992). These strains are representative of the Agrobacterium tumefaciens collection and have been used for the plant species for which gene transfer has been reported. For the effi ciency of gene transfer the virregio n of these strains appears to be more important than the chromosomal DNA. A good example is the vir region of pTiB0542, present in the supervirulent Agrobacterium A281 strain. During the last few years the disarmed derivative of this strain EHA101 has been utilized successfully to improve gene transfer in various recalcitrant plant species. Super virulence is correlated with the presence of specific virG and virB loci, resulting in an enhanced transcription of the other virgenes , leading to a more efficient transport of the T-strand through the bacterial cell wall. Recent research on chrysanthemum transformation at CPRO-DLO by Mertens et al. (unpublished results) has shown that both the type of disarmed bacterial strain as well as the T-DNA binary plasmid influences the gene transfer as determined by the expression of the reporter gene gus. Table 2.3.2. Requirements for commercial application oftransformatio n procedures of crop plants. - high efficiency of procedure - reproducibility of procedure - fast selection without 'escapes' (non-transformants surviving selective culture media) - no changes in ploidy - no changes due to somaclonal variation (true to type) - preferably one copy ofth e transgene - predictable and stable expression of the transgene - applicable to all main cultivars 34 Theplant material. Besides the origin and age of the expiant, the developmental stage of the starting material is also important. In general, meristematic tissue composed of active, dividing cells seems to be extremely susceptible to gene transfer. A good example was shown by researchers of Plant Genetic Systems in Ghent, Belgium, on the improvement of Agrobacterium transformation of sugarbeet (D'Halluin et al., 1992). A pretreatment of seed lings on a medium with extremely high auxin concentrations resulted in rapidly dividing callus material which was amenable to transformation and regeneration of transgenic plants. Leaf expiants are suitable for transformation of Solanaceous species, but seedling expiants and embryos have been used effectively for other species. Apart from the origin of tissue to be used, the efficiency of transformation strongly depends on the genotype chosen. It is therefore questionable whether the published procedures for crop species are applicable to all cultivars, which means that in most cases procedures will have to be adapted. Reporting andselecting. Transformation events can be visualized easily by the introduction of a reporter gene. Expression of the bacterial gene coding for /^-glucuronidase (the ^.r-gene) leads to blue-stained cells after histochemical staining with X-Gluc (5-bromo-4-chloro-3- indolyl-y8-glucuronic acid) as a substrate. This assay (the GUS assay) iswidel y used and has been improved by the introduction of a eukaryotic intron in the bacterial gus gene. This leads to inhibition of the expression in the Agrobacteria allowing detection of eukaryotic expression in the early stages of co-cultivation, even in the presence of bacterial cells. The GUS-assay is very simple, can be used for all transformation techniques and can be generally applied to all plant species. Therefore, other reporter genes, for example the luciferase gene requiring a rather complicated detection method or the anthocyanin syn thesis genes involved in a genetically rather complicated pathway have only been used occasionally. The only drawback to using the gîw-gene is the lethality of the assay. Vital staining procedures have recendy been developed (Imagene GUS expression kits), which might allow the opportunity of visually selecting GUS-positive cells or shoot primordia. The aminoglycoside-3'-phosphotransferase II gene (aphA2 or nptll) is the most com monly used selectable marker gene. It confers resistance to aminoglucoside-like antibiotics (e.g. kanamycin, geneticin or G418, neomycin and paramomycin). Kanamycin has proven to be a very efficient selectable marker for dicotyledonous plants. However, differences in natural resistance between plant species or between tissues of the same species might give rise to high numbers of escapes. Many monocotyledons, especially Gramineae, show a high level of resistance to this antibiotic. Sometimes these problems can be solved by using more effective antibiotics or by carefully timing their application. For instance G418 can be used instead of kanamycin or the much more potent antibiotic hygromycin used in combination with the hpt(hygromyci n phosphotransferase) gene. Finally, genes conferring resistance to herbicides such as Basta or Roundup, have frequently been applied, espe cially for monocotyledons. 35 2.3.2 Transformation of monocotyledons Developments in alternative gene transfer The need for transformation procedures in combination with the experienced recalcitrancy of many plant species has resulted in the development of a large number of elegant and highly sophisticated techniques for the introduction of DNA into plant cells. These procedures have been summarized and discussed by Potrykus (1990). He presented a long list of strategies that have been developed, but only two have been applied successfully to many plant species. These strategies are: 1. direct gene transfer, in which naked DNA is introduced into protoplasts via PEG or electroporation. Such a procedure has recendy been elegantly applied for zygotic embryos of monocotyledons such as maize (Leemans et al, 1992); 2. biolistics or particle bombardment, in which DNA is introduced into intact cells or plant tissues through bombardment with DNA-coated particles. The Particle Delivery System (PDS, Fig. 2.3.1) and its upgrade helium version has been used frequendy and with success. The use of a micro-targeting system which enables the delivery of DNA in selected cells of apical domes of seedlings has recently been described. Introduction into cells of the L2 layer leads to gene transfer in the germ line of the plant. Transformation ofrice The Molbas research group at the Leyden University has thoroughly evaluated the trans formation of rice. Rice protoplasts are amenable to transformation making genetic engi neering of rice possible. However, this approach requires the induction and isolation of cell suspensions for each variety to be transformed, while further regeneration of transformed cells is laborious, time-consuming and unpredictable. Therefore, the potential of the Agrobacterium transformation system has been evaluated, but stably transformed rice plants could not be recovered (Hensgens & Schilperoort, 1992). In contrast, high transformation frequencies were obtained employing a particle gun using embryos from immature seed, while the obtained transgenic callus also regenerated into fertile transgenic plants (Hensgens & Schilperoort, 1992). Because rice embryo tissue can be regenerated easily and plants can be simply obtained from callus tissue of most rice varieties, it is anticipated that this procedure will be applicable to most varieties. Rice is resistant to high concentrations of kanamycin and G418 while it isver y sensitive to hygromycin. Therefore, the hpt gene can best be used as selectable marker gene in transformation experiments. Also different rice promoters to drive selectable marker genes have been characterized (Hensgens & Schilperoort, 1992). One of these promoters, the constitutive GOS2 promoter, was tested for expression in rice suspension cultures by bombarding it with microprojectiles coated with promoter-g&f constructs. Only in the presence ofintro n sequences of thegos2 gene , could GUS activity be detected in suspension cells, calli and meristematic tissue of regenerated rice plantlets. The activity measured was higher compared with the activity obtained using the maize alcohol dehydrogenase pro moter and the first intron ofthi s gene. Also, theAgrobacterium mannopine promoter showed high activity throughout all rice tissues but especially in trichomata and stomata of trans genic rice plantlets. 36 Fig. 2.3.2. Adventitious shoot formation from young flower stem segments of tulip (Photo A. Wilmink CPRO-DLO). Transformation of other monocotyledonous plants In the Netherlands, procedures for transformation are being developed for several other monocotyledonous crops such as barley, lily and tulip. Research on these two ornamental crops is being supported by the Dutch Urgency Programme for Research on Diseases and Breeding of Flower Bulbs. In tulip adventitious shoot formation proceeds very well from young flower stems (Fig. 2.3.2). Using this regeneration system, DNA transfer is now being investigated using particle bombardment (Wilmink et al, 1992). In ajoin t venture between CPRO-DLO and the Molbas research group, a project is being executed aimed at trans forming Lolium perenne L. (rye grass). Regeneration from protoplast to plant has been demonstrated before, but no direct gene transfer by the co-cultivation of protoplasts with DNA could be detected. The stable transformation of Lolium calli could be demonstrated by using the hpt selectable marker gene linked to the CaMV 35S promoter after particle gun bombardment of seedlings or suspension material. Expression of thegus reporte r gene linked to the constitutive rice promoter and present on the same plasmid, was also ob served. Plant regeneration is currently in progress. 2.3.3 Improvement of genetic modification procedures Position effect and transgeneexpression A frequently encountered problem in plant genetic engineering is the extensive variation in expression levels of newly introduced genes in different individual transformants. This 37 phenomenon is often referred to as a 'position effect', and indicates that the site of integration of the transgene influences its expression. In animal systems, the addition of 'nuclear scaffold attachment regions' or SARs has been shown to decrease the variability of transgene expression. SARs are thought to be involved in the formation of chromatin loops which may play a role in gene expression. Breyne et al. (1992 ) recendy found a decrease in variability of expression of reporter gene SAR constructs in tobacco callus. At CPRO-DLO a set of SAR-GUS combinations has been constructed using a chicken lysozyme A-element that has been shown to function as a SAR in animal cells. Transformation of these constructs to tobacco and analysis of transgenic plants is in progress. Gene targeting in plants Inactivation of transgenes may also occur by methylation or multiple loci inactivation of gene expression also called co-suppression (Mol et al., 1991;D e Carvalho et al., 1992). A way of avoiding inactivation by methylation, co-suppression or position effects could be targeted homologous recombination of the introduced DNA at a chromosomal location where expression is secured. Offringa et al. (1992) have recenüy shown that the Agro- bacterium vector system can be used for gene targeting in plants. However, frequencies of gene targeting are still low (approximately 10"4). Gene targeting via homologous recombination can also be utilized to modify gene activity or inactivate the expression of a specific gene involved in the biosynthesis of undesired compounds. The efficiency of this approach needs to be increased considerably especially in those cases where plants are used for which less efficient transformation protocols are available compared with tobacco and Arabidopsis thaliana. Competencef or regeneration and transformation Stable integration of DNA introduced either by Agrobacterium, electroporation or particle bombardment, should take place in cells which are able to regenerate into transgenic shoots. Recalcitrancy to genetic engineering might be related to difficulties in 'targeting' the DNA to the regenerable cells. Good histological practice has in fact shown that efficient gene transfer by Agrobacterium is sometimes directed at cells which do not take part in the formation of adventitious meristems. More fundamental research is needed to determine the parameters for the competence ofcell s for regeneration and transformation. This might be accomplished in model systems, e.g. the thin layer regeneration systems. On the other hand crop-specific research should continue because for each crop or each cultivar procedures need to be optimized to realize an efficiency useful for industrial application. Acknowledgements We would like to thank R Hooykaas (Leyden University), Mei li Tan (Zaadunie), M. de Both (Keygene), M. Akerboom (Florigene) and colleagues from CPRO-DLO for kindly sharing their ideas and results. 38 2.4 New approaches for identifying and isolating plant disease resistance genes J. Hille& P. Zabel The past decade has seen a major advance in the development of gene isolation strategies. Nowadays, essentially every segment of any organism's genome is accessible for detailed identification and manipulation, irrespective of its coding potential. When the protein product of a gene of interest is known, cloning of the gene is relatively simple using techniques that are in every molecular biologist's tool box. Thus, once the protein has been purified antibodies directed against it can be used to screen a cDNA expression library. Alternatively, oligonucleotides synthesized on the basis of corresponding amino acid se quences from parts of the protein may be applied to screen a cDNA library or to prime a polymerase chain reaction (PCR). The latter approach, in particular, has proven to be extremely powerful in isolating genes encoding low abundant proteins (Aarts et al., 1991). There is, however, an important class of plant genes that is not accessible through their protein product. For example, genes involved in developmental processes or disease resis tance are usually only known from their mutant phenotype. For those genes, basically two types of cloning strategies have been developed (Hille et al., 1991). The first approach, transposon tagging, involves the insertional inactivation of the gene by means of a trans- posable element and the second approach, map-based cloning, is based on a 'chromoso mal walk' starting from a nearby marker flanking the gene. In this section we will discuss both approaches in relation to our own research on tomato. 2.4.1 Transposon tagging of disease resistance genes The insertion of a transposable element into a gene can give rise to mutations and this has been shown to be extremely useful for generating plant mutants. In 1984, using the transposable element Ac from maize, Fedoroff and co-workers were able to clone a locus which until then had only been characterized genetically, suggesting that, in principle, any locus in maize identifiable with a transposable element can be cloned at the molecular level. This approach which was coined transposon tagging has successfully been applied to the isolation of various maize genes. Demonstrating that the transposable element Ac also transposes in dicotyledons has opened up the possibility of using this well-characterized element for isolating genes in other plant species (Haring et al., 1991). The aim of our research efforts in Amsterdam ist o develop a transposon tagging procedure in tomato that is generally applicable (see Fig. 2.4.1) and allows one to clone disease resistance genes. Here, we focus on the interaction between Altemaria altemata f. sp.lycopersici an d the cultivated tomato Lycopersicon esculentum. The fungus causes the disease Alternaria stem canker which is characterized by dark brown cankers on stems and necrosis of leaf tissue between the veins.AAL-toxin , produced by the fungus, plays a major role in pathogenesis, exhibits the same host-specificity as the fungus and is responsible for the development of leaf necrosis. In tomato there is a single dominant gene, the ^4.j<;-locus, which confers resistance to the harmful effects of this fungal infection and insensitivity to the AAL-toxin. 39 -c -m— -c -c Fig. 2.4.1. Transposon tagging. Basic-set up of transposon tagging in tomato. A tomato genotype is selected in which the transposable element (hatched box in upper line) is genetically linked to the gene of interest (open box in upper line). Subsequently the transposable element is induced to transpose (indicated by arrows). The progeny plants in which transposition occurred are screened for insertional inactivation of the gene of interest (the five lower lines). With the transposon as a probe the inactivated gene can be isolated. Since the transposon sequences are known, the sequences of the gene of interest can be identified. This information is sufficient for the isolation of the intact gene by more classical procedures. The ^.sr-locus has been genetically mapped on chromosome 3. Reliable and convenient bioassays have been developed to screen for the presence or absence of this locus (Witsen- boer et al., 1992). In applying AAL-toxin the ^JC-IOCUS can be recognized in the heterozy gous situation, which is a unique feature for disease resistance genes in tomato and offers possibilities to select for mutants due to dominant as well as recessive mutations at the Asc-\oc\x%. In maize and also in tomato it has been shown that Ac has a tendency to transpose into the vicinity of its original position rather than at random throughout the genome. There fore, with targeted transposon tagging, one must start with a plant containing the transpos able element close to the position of the desired gene. To this end, two component transposon systems have been developed comprising a stable but transactive transposase function and a cis-responsive non-autonomous transposable element (Rommens et al., 1992a; 1992b). In such transposon tagging strategies genetic linkage between the non- autonomous transposable element and the gene of interest can be selected for. In a European collaboration with laboratories in the U.K. and Germany we have pooled our resources to develop such a transposon tagging system in tomato optimizing both compo nents of the transposon for their function. We aim to generate a series of plants carrying transposable element insertions at different positions that are equally spaced over the 40 twelve chromosomes of tomato (see Fig. 2.4.2). Non-autonomous Ds-elements were intro duced into tomato byAgrobacterium-mediated transformation . Subsequently, the position of the Ds elements was established on the tomato RFLP-map using the Ds flanking plant DNAs as probes in hybridization analysis. These flanking DNA fragments were obtained either by using novel PCR-based methods like the newly developed 'supported PCR' or by plasmid rescue, facilitated by the presence of an origin of replication for E. coli present in the Ds element. So far seven transposable element inserts have been mapped on chromo some 3. The first experiments to tag the Arc-locus have been initiated (Overduin and Hille, in preparation). Using both a seedling assay and a leaf bioassay, conditions have been estab lished that permit the screening of large numbers of plants for the insertion of a transposa ble element in the Arc-locus. A transgenic tomato line, homozygous for the Arc-locus and a Ds element on chromosome 3 was crossed with a line also homozygous for Asc and containing Ac on chromosome 3. The resulting progeny in which both Ac and Ds can transpose, was crossed on a large scale with a sensitive male sterile tomato line {asc, asc). The seed population obtained, heterozygous for the Asc-locus, is analyzed for Alternaria susceptible mutants in order to identify transposon induced mutations at Asc.A number of putative mutants has been found that are now being analyzed in detail. 10 12 •-11914C •-FT33 t851P :-FT30A • -2995A •-721A ••601YY ••1515Q ••2962 B - -3041H • -ET530 • -721A --11B • •1402B B +ET517 1 TT21B - •3062B • -ET90 -•FT2 13715AJ.ET26 - -ET40 +3041M4. • -ET85 FT64 • -1515Q • • 105121 591-53 +1615C+ 1515Z •"13411V4.J5J5J - • 1561E 13715U •-851.1 - -ET905& • -ET16 7 • -ET100 Ä --1515X4. 3062C ST233 •-11A - -ST227 -•591-91 • -FT21 • -ST239 • • ST223 --1515W 851Q •-1515MM •-1515LL '•13411VV Fig. 2.4.2. Mapped transposon positions on the tomato genome. In an European Community BRIDGE project with laboratories in Cologne (Starlinger, Theres), Norwich (Jones), Nottingham (Taylor) and Amsterdam (Hille) more then 60 transposon positions have been established on the tomatoRFL Pmap .Thes e tomato genotypes can be used asstartin gmateria l for targeted transposon tagging fully exploiting the Actransposo n characteristic ofpreferentia l short-distance transposition. 41 2.4.2 Map-based cloning of disease resistance genes The approach referred to as map-based cloning, positional cloning or reverse genetics, allows one to isolate a gene on the basis of its genetic map position by means of a 'chromosomal walk' starting from nearby flanking markers. So far, most examples of genes approached by a map-based cloning strategy are to be found among human hereditary disease genes, but more recendy plant genes have also started to appear on the scene. Although simple in concept, gene isolation by a map-based cloning approach is labori ous requiring many steps before the gene of interest is finally at hand. Irrespective of the nature of the target gene, map-based cloning boils down to the following steps: 1. determining chromosomal location (genetic map position) of gene of interest by linkage analysis; 2. identifying tightly linked molecular markers; 3. determining the genetic order of molecular markers around the target gene; 4. constructing a long range physical map showing the physical order of, and distances between, molecular markers; 5. cloning all sequences between the two closest flanking markers; 6. identifying the cloned chromosomal fragment carrying the gene of interest by comple menting a genotype with the corresponding mutant phenotype. To illustrate how such a scheme translates into practice, our research efforts in Wagenin gen, in close collaboration with the Department of Nematology in Davis USA, on cloning the root-knot nematode resistance gene in tomato provides a good example. Root-knot nematodes are economically important plant pathogens that are causing severe damage worldwide to a wide variety of crops, including monocotyledons, dicotyledons and herba ceous and woody plants. In tomato, resistance against root-knot nematodes was found to occur in the wild tomato species Lycopersiconperuvianum and subsequently introduced to the cultivated tomato L. escukntum by breeding. The resistance is encoded by a dominant allele at a single locus, referred to as Mi, an acronym taken from the first letters of the root-knot nematode species Meloidogyne incognita. The Mi locus was mapped on chromosome 6 (see step 1) and found to confer resistance to the three most damaging root-knot nematode species. Conceivably, cloning of the Mi gene would allow its transfer into susceptible crops across genetic incompatibility barriers using recombinant DNA technologies. Apart from the potential impact on agriculture, cloning of Mi would also be a major step towards understanding the molecular events involved in the plant's defence response. As the product of Mi is unknown, we have chosen to gain molecular access to this resistance gene by following a map-based cloning approach. To start a collection of molecular markers surrounding the Mi gene (step 2) we initially isolated the acid phosphatase gene Aps-\, which is tightly linked to Mi and amenable to cloning using a product-based cloning approach. Furthermore, by screening a set of well-defined genetic stocks of tomato which differ only for a small chromosomal region carrying the resistance gene, we have been able to identify a variety ofadditiona l linked markers (Klein-Lankhorst et al, 1991a; 1991b; Ho et al., 1992). These markers have been ordered into a high resolution molecular linkage map of the Mi region (step 3)whic h provides the landmarks for a chromosomal walk in the 42 Mi region (Ho et al., 1992). As shown in Fig. 2.4.3, the nematode resistance gene Mi is located on a small (2000 kilobases) chromosomal segment derived from L. peruvianum. This segment is delimited by two internal markers (LC 379 and REX-1) and several flanking markers. We have recently constructed long-range physical maps (step 4) of the bracketing markers GP 79 and Aps-1, each spanning more than 700 kilobases (van Daelen and Zabel, in preparation). A chro mosomal walk starting at LC 379 and REX 1i n either direction should provide the bridge linking the flanking markers (step 5). To this end a genomic library of tomato constructed in Yeast Artificial Chromosomes (YAC) which carry tomato chromosomal segments of 100 - 500 kilobases is currendy being screened with markers from the Mi region. Eventu ally, this should result in the cloning of all tomato sequences contained between the two closest flanking markers of Mi and in the molecular identification of the Mi gene. 2.4.3 Conclusions and prospects Over the past couple of years new gene isolation strategies have been developed which hold exciting promises for experimental plant biology. In travelling the route from concept to product, transposon tagging and map-based cloning have now reached the stage at LC216B-v v £ C i TG232, H2D1,GP202 ,AFLP 6 B \ TG352 L. esculentum L.peruvianu m Fig. 2.4.3. Map of the Mi region of Chromosome 6. A map of Chromosome 6 is shown on the left with a 'blow-up' of the root-knot nematode resistance (Mi) region as present in the tomato cultivar Motelle on the right (see Ho et al., 1992). Chromosomal regions of Lycopersicum esculentum origins are represented as vertical lines. The region of L. peruvianum origin carrying the Mi gene is represented by a vertical black bar. Note that this region is delimited by two internal markers (LC379, REX-1) and several flanking markers providing landmarks in a chromosome walk. 43 which the isolation of plant disease resistance genes is at hand. As for tomato, it is antici pated that cloned resistance genes against fungi and nematodes will be available in the near future. This will open the avenue to an understanding of the molecular communica tion between plants and pathogens and to potential applications in crop protection. 2.5 New strategies for obtaining fungal-resistant plants BJ.C. Cornelissen &J.A.L. van Kan Although traits that confer plants with resistance to a broad range of fungi have not been isolated as yet, much effort is currently being put into the search for genes that may confer fungal resistance. A number of strategies and approaches initiated within university labo ratories and biotech companies will be discussed. In one approach genes encoding en zymes involved in the synthesis of non-proteinaceous fungal toxic compounds are being tested as resistance traits by Bayer AG in Germany (Section 2.5.2). Similarly, there is a search for genes encoding proteins with an inhibitory effect on the growth of fungi (Section 2.5.3). In the Netherlands this approach isbein g pursued by MOGEN International nv, a Leiden based plant biotechnology company, in collaboration with university groups in Leiden and Wageningen. Furthermore, various research groups, among which those of Hille in Amsterdam,Jone s in Norwich, England, and Zabel and Lindhout in Wageningen, Netherlands, are involved in the development of new methodologies to isolate monogenic resistance traits known from conventional breeding programme. Such genes may be introduced into target plants to serve as a specific fungal resistance trait (Section 2.5.4) or together with their corresponding avirulence gene to act as part of a necrosis-inducing system which may confer broad range pathogen resistance to plants (Section 2.5.5). This latter strategy is being developed by de Wit and co-workers at the Wageningen Agricul tural University, in collaboration with MOGEN International in Leiden. 2.5.1 Production of non-proteinaceous fungal toxic compounds in plants Under specific conditions plants synthesize non-proteinaceous compounds of low molecu lar weight which are toxic to micro-organisms. Often such compounds, known as phytoalexins, are synthesized locally and accumulate after exposure ofplant s to pathogens and/or stress. In many cases a correlation has been found between the concentration of phytoalexins and resistance to pathogens. This correlation suggests that the production in plants of specific phytoalexins may result in resistance to pathogens. To test this suggestion, Hain et al. (1990) recendy introduced the gene encoding stilbene synthase in tobacco. In groundnut stilbene synthase isth e key enzyme in the synthesis of resveratrol, a phytoalexin associated with resistance against the fungus Botrytis cinereai n grapevine. The precursors for the synthesis of resveratrol are available in tobacco but a gene encoding stilbene synthase is lacking. The constitutive expression of a groundnut stilbene synthase gene in transgenic tobacco plants results in the synthesis of resveratrol (Hain et al, 1990) and the transgenic plants appear to be more resistant to infection by B. cinereatha n the control plants (R. Hain, personal communication). 44 2.5.2 Production of antifungal proteins in plants Proteins with the ability to inhibit the growth offung i do so in various ways. Chitinases and y3-l,3-glucanases are thought to stop fungal growth by hydrolysing major components of the cell wall. Osmotins, osmotin-like proteins and thionines may interfere with fungal growth by disturbing the function of membranes. Plant ribosome-inactivating proteins inhibit protein synthesis. At present there is no clue as to how hevein or lectins inhibit fungal growth. However their mode of action, one can envisage that for all antifungal proteins the constitutive or pathogen-inducible expression of the corresponding genes may render plants resistant to fungi. Chitinases andß-1,3-glucanases Chitin and/M ,3-glucan are major constituents of the cell wall of many fungi (Wessels & Sietsma, 1981). They are hydrolysed by chitinases and ß-1,3-glucanases, respectively. These hydrolases have been isolated from many plant species and subdivided into classes on the basis of their molecular structure. In tobacco at least three of the four classes of endochitinases have been identified (Sela-Buurlage et al., 1992 and references therein), and three major classes of/M,3-endoglucanases (Ward et al., 1991). Both class I chitinases and class I/M,3-glucanases are potent inhibitors of growth invitro ofman y fungi. Combin ing the two hydrolases, in quantities insufficient to show an effect by themselves, results in very high antifungal activity (Leah et al., 1991; Sela-Buurlage et al., 1992). Apparently, class I chitinases and class I/M,3-glucanases act synergistically. In contrast to the class I hydrolases, Sela-Buurlage and co-workers (1992) at MOGEN in Leiden showed that class II proteins are not antifungal at all, neither alone nor in combination with other proteins. It is not yet known whether class III hydrolases and class IV chitinases have any antifungal activity in vitro. Broglie and co-workers (1991) recently introduced a bean class I endochitinase gene under control of the cauliflower mosaic virus (CaMV) 35S promoter into tobacco and canola {Brassica napus). Transgenic plants constitutively expressing the gene showed an enhanced resistance to the fungal root pathogen Rhizoctonia solani.A similar result was obtained byjach and co-workers (1992), who constitutively expressed a bacterial exochiti- nase gene in transgenic tobacco. In contrast, Neuhaus et al. (1991) found that the constitu tive expression in Mcotiana sylvestris of a tobacco class I chitinase gene did not alter the susceptibility of the transgenic plants to the fungal leaf pathogen Cercospora nicotianae. Many fungal infections of plants start intercellularly and proceed intracellularly only at a later stage. Some plant pathogenic fungi grow exclusively in the intercellular spaces. Therefore, to protect plants from fungal infections at an early stage antifungal compounds should be present in the intercellular spaces rather than inside the cells. However, the class I hydrolases showing antifungal activity in vitro are localized in the vacuoles of the plant cell. Hence, transgenic plants constitutively expressing these hydrolase genes will produce the antifungal proteins inside the cell and not in the intercellular spaces, the preferred site of action. To target the vacuolar proteins from the cell, Melchers and co-workers (1992) at MOGEN have modified both a class I chitinase gene and a class Iy3-l,3-glucanase gene. In transgenic plants expressing either of the two modified genes the two hydrolases were 45 found extracellularly. Both targeted proteins retained their fungal growth-inhibiting activ ity and together they still demonstrated synergy in the inhibition of fungal growth in vitro (Sela-Buurlage et al., 1992). To exploit this synergy inplanta transgenic tomato plants are currendy being produced which constitutively express both modified genes. Such trans genic plants will be tested for resistance against a broad spectrum of fungi. Osmotin, osmotin-likeproteins and thionines Not all fungi contain chitin in their cell walls. Oomycetes in particular seem to lack a detectable amount of this homopolymer. As expected their growth invitro is not affected by chitinase. In their search for proteins able to interfere with the growth of Oomycetes, Woloshuk et al. (1991) at MOGEN identified a protein with an inhibitory effect on the Oomycete Phytophthora infestans in tobacco. Further characterization revealed that this 24 kD anti-Phytophthora protein (AP24) was identical to osmotin, a basic protein that accu mulates under salt stress in the vacuoles of plant cells. AP24 also appeared to inhibit the growth of the human pathogen Candidaalbicans, the saprophyte Trichoderma reesi and the non-pathogenic fungus Neurospora crassa(Viger s et al., 1992). The same pathogens are inhibited in their growth by permatins, proteins that are very homologous to osmotin which have been isolated from several cereals. As with the class I hydrolases, AP24 could be targeted extracellularly by modifying the corresponding gene (Melchers et al., 1992). It has been suggested that AP24 or osmotin inhibits the growth of/? infestansb y interfer ing with the fungal membranes, hence disturbing their function. Thionines could act in a similar fashion. Thionines form a well-defined group of small (40-47 amino acids), basic proteins (for a recent review, see Bohlmann &Apel , 1991). Initially discovered in the seeds of many monocotyledonous plants, they have now been isolated from leaves and stems of both monocotyledonous and dicotyledonous plant species. Their toxic effect on both bacteria and fungi was discovered about fifty years ago. Ribosome-inactivatingproteins Plant ribosome-inactivating proteins (RIPs) inhibit protein synthesis in target cells by a specific RNA N-glycosidase modification of 28S rRNA. As a consequence elongation factor 2 binds less efficiently and the elongation step in protein synthesis is inhibited (for a recent review, see: Stirpe et al., 1992). RIPs do not affect the ribosomes of plants in which they are produced and show various degrees of specificity towards ribosomes of other plants. Targets of RIPs also include fungal ribosomes. In assays invitro a RIP isolated from barley has a lower antifungal activity than chitinases or /M,3-glucanases from barley. However, a strong synergetic effect is observed when barley RIP is mixed with either of these two hydrolases (Leah et al, 1991). A barley RIP gene, under the control of a wound-inducible promoter, has recendy been introduced into tobacco. Rj progeny showed increased resistance to R. solani (Logemann et al., 1992). The level of resistance of these plants appeared to be higher than that reported for transgenic tobacco plants ex pressing constitutively a bacterial exochitinase gene. 46 Hevein and lectins Hevein is a small (43 amino acids) long protein from Hevea brasiliensis, exceptionally rich in cysteine and glycine residues and capable of binding to chitin. It exerts an antifungal activity on a number of chitin-containing fungi. Two small (29-30 amino acids) proteins with chitin-binding capacity and antifungal properties similar to those of hevein have recently been isolated from Amaranthuscaudatus (Broekaert et al., 1992). Lectins are proteins that recognize and cross-link specific carbohydrates (see Chrispeels & Raikhel, 1991, for a recent review). Some of them specifically bind the sugar N- acetylglucosamine, its oligomers and its polymer chitin. It has recently been shown that the lectin from stinging nettle is a potent inhibitor of many fungi, whereas a lectin from potato is not (Broekaert et al., 1989). Apparendy, some lectins inhibit the growth of fungi whereas others do not. 2.5.3 Transfer of monogenic resistance genes Resistance breeding to pathogens has been performed in many agricultural crops over the last few decades. Resistance genes (R genes) specific to certain pathogens have been detected in a wide variety of wild monocotyledonous and dicotyledonous plant species (for reviews, see Pryor, 1987; Knogge, 1991). Resistance is often effectuated by localized cell death, visualized as a necrotic area. This so-called hypersensitive response (HR) is because the pathogen is recognized by the R gene of the plant. Due to crossing-barriers, R genes could only be used from sources that are related to the crop plant of interest (Niks, 1988), making it difficult to investigate whether R genes to a specific strain of a pathogen could function in a completely different genetic background. However, recent research has shown that non-host resistance of a specific plant to a certain pathogen-species can be based on the same mechanistic principles as a strain-specific resistance. Work carried out by Keen and Staskawicz and co-workers in the USA has provided insight into the molecular-genetic basis of resistance mechanisms in plants against phyto- pathogenic bacteria (see review by Keen, 1990). Genes which function as avirulence genes in a specific manner on some cultivars of several non-hosts were cloned from two bacterial pathogens of tomato, Pseudomonas syringae pv. tomato and Xanthomonas campestris pv. vesicatoria (Bonas et al., 1991;Whale n et al., 1991). This implies that non-host plants possess mecha nisms (R genes) that are able to recognize and counterattack these bacteria. Ifthi s principle proves to be true for several pathogens, non-host resistance would be caused by the simultaneous action of several strain-specific resistance mechanisms, resulting in a non- host phenotype. These two types of resistance cannot be discriminated statistically if more than five different R genes are involved (Parlevliet & Zadoks, 1977). The microbe would need to overcome or avoid each of the R genes in order to become a successful pathogen. On the whole, the data on phytopathogenic bacteria suggest that many R genes are probably functionally exchangeable between plants. R genes have been mapped genetically in several crop plants such as tomato, potato, and monocotyledons. In order to understand the mode of action of R genes, it will be essential to clone them and analyze their structure (for a review, see Bennetzen &Jones , 1992). If successful, cloned R genes could be transferred to other plant species by transfor- 47 mation, thus overcoming species barriers which have hitherto been obstacles in breeding programmes. As discussed above R genes from one plant species are likely to function with comparable pathogen specificity and efficiency in a heterologous plant background. Several methods are technically available to clone R genes: 1. RFLP or RAPiD markers can be obtained which co-segregate with a specific R gene (Hinze et al., 1991). These DNA markers can be used as probes to start a chromosome walking procedure (see Section 2.4 of this book). 2. A second approach to obtain R genes is by transposon tagging (see also Section 2.4 of this book). 3. A third approach is by the use ofbiochemica l techniques. It is assumed that the R gene encodes a receptor which specifically recognizes a fungal elicitor, encoded by an avirulence gene (Knogge, 1991; de Wit, 1992). Purified elicitors can be labelled with radioactive isotopes and used in receptor binding studies. If specific elicitor-receptor binding is observed, the receptor can be purified and its structure determined. On the basis of amino acid sequences of the receptor it might be possible to isolate the gene encoding the receptor. Irrespective of the method used, confirmation that an R gene has been cloned should be obtained by transforming a susceptible plant genotype to resistance with the isolated clones. 2.5.4 The two component sensor system Using isolated R genes it will be possible to transform a wide range of relevant crop plants to resistance against particular pathogens. However, the introduction of R genes, whether by recombinant DNA technology or by conventional cross-breeding, may lead to the selection of pathogen populations that overcome or avoid the resistance. This 'breakdown' of resistance has been observed in many plant-pathogen interactions, although some R genes have been 'overcome' more rapidly than others. A new approach is therefore being undertaken in Wageningen to obtain resistance based on the principles of elicitor-receptor recognition, which is predicted to have a broad application to a wide range of pathogens and will be less easily overcome or avoided. The fungal tomato pathogen Cladosporium fulvum contains an avirulence gene, avr9, encoding a protein which determines an avirulent phenotype on tomato cultivars contain ing the corresponding C/9 resistance gene (de Wit, 1992 and references therein). Resis tance is thought to result from the interaction of the avr9gen e product, the elicitor, and the Cf9 gene product, the receptor, leading to the transduction of a signal which activates a cascade of defence responses. Recognition of the AVR9 elicitor by the Cf9 gene product results in a rapid, localized hypersensitive response (HR), visualized by a necrotic area (de Wit, 1992). This HR is associated with the accumulation of a wide range of compounds with potential antifungal activity, such as phytoalexins and hydrolytic proteins, which were discussed in Sections 2.5.2 and 2.5.3. Some strains ofth e fungus have lostth e avr9gen e and hence avoid recognition, resulting in the ability to colonize tomato Cf9 genotypes (van Kan et al., 1991). Although these strains have been identified, they may be less suitable, because such strains have not become predominant in the fungal population. This implies 48 that the avr9 gene represents an important function to C.fulvumwhic h may not be lost, and that the Cf9 gene is a strong and relatively stable R gene. Indeed the C/9gen e has been used successfully by tomato breeders for over ten years, without a significant 'breakdown' of resistance. A similar strong and stable R gene, Bs2, has been detected in pepper, based on the reduced suitability of Xanthomonascampestris strains which lack the corresponding avirulence gene avrBs2(Kearne y et al., 1990). By transforming C.fulvum strains which are virulent on tomato Cf9 genotypes with the cloned avr9gene , it has been demonstrated that the avr9gen e is the single fungal genetic determinant interacting with the corresponding tomato Cf9 gene (van den Ackerveken et al., 1992). By expressing the fungal avr9 gene in a tomato Cf9 genotype, a necrosis- inducing system can be obtained based on the interaction of two components. The coordi nate expression of the two components in one cell of such transgenic plants would result in cell death (Fig. 2.5.1). Evidently, a constitutive expression of both components would be impossible because no viable plants would be obtained. Therefore the two components need to be regulated, so that coordinate expression only takes place when desired, such as during the course of a pathogen attack (Fig. 2.5.1). Since only one component (the avir ulence gene avr9) has so far been cloned, regulation should be imposed on this gene. Ideally, a plant promoter should be used to regulate of the aw9 gene, which meets three criteria: 1. The promoter should be inducible by a wide range of pathogens. 2. The promoter should be rapidly and locally inducible, to assure local necrosis and prevent induction of systemic cell death. 3. The promoter should not be activated during normal development of the plant or during any stress, other than attack by a pathogen. AvrTRANSGENI C PLANTS compatible induction » pathogen Avrgen e compatible interaction » pathogen gene products R -^ compatible non-specific » Avr pathogen resistance Fig. 2.5.1. Principle of the two-component sensor system. A transgenic plant contains a resistance gene (R) and the corresponding avirulence gene (Avr),unde r control of a pathogen-inducible pro moter. The challenge of this plant with a compatible pathogen results in the induction of the Avr gene, leading to the production of elicitor (<). The elicitor interacts with the receptor ( ) encoded by the R gene. The interaction of the two proteins causes the activation of hypersensitive response and non-specific resistance mechanisms. 49 When a promoter is found which meets these criteria, transgenic plants can be obtained which demonstrate resistance against a wide range of bacterial and fungal pathogens. Using this two-component sensor system, the whole array of defence reactions associated with the HR can be utilized (de Wit, 1992). This approach is in contrast to the utilization of individual defence genes to obtain fungal resistance, as discussed in Sections 2.5.2 and 2.5.3. It also avoids the problems encountered with the use of only R genes, that can easily be broken by new strains of the pathogen, as discussed in Section 2.5.3. Once the Cß gene has been cloned from tomato, the avr9-CF3 system can be introduced as a gene cassette into other crop plants, provided that the signal transduction pathway used in tomato to induce necrosis also functions in other plants. Ifth e avr9-Cj9syste m does not function in other plants, the principle of this method can be adapted by using other combinations of avirulence and resistance genes. 2.5.5 Conclusions In the last five years, many research laboratories have been attempting to engineer patho gen resistance in crop plants. Genetic engineering overcomes the need to obtain desired genetic traits from wild plant species related to the crop of interest. As discussed in this section, strategies to obtain pathogen resistance are being developed either by constitutive expression of individual compounds with potential antifungal activity, or by a modified inducible defence system such as the two-component avr9-Cß system. It is to be expected that the next five years will yield considerable scientific information as to whether the different strategies are feasible and effective under laboratory and greenhouse conditions. The most promising systems will need to be tested extensively on a large scale, under field conditions, before being applied to agricultural practice. Although all these scientific efforts will not solve the short-term problems of crop protection, genetic engineering should be able to provide a long-term alternative to crop protection using chemicals. 2.6 Molecular breeding for insect-resistant plants B. Visser & M.A. Jongsma This contribution aims to give an impression of the research being done on molecular breeding for insect-resistant plants. We will focus on the role that crystal proteins of the prokaryotic Bacillus thuringiensis, and proteinase inhibitors occurring in the potato tuber, might play in obtaining insect resistance by molecular breeding. In line with the aim of this book emphasis is on research being done in the Netherlands. 2.6.1 Bacillus thuringiensis genes encoding insecticidal crystal proteins Natural gene diversity and insectspecificity The spore-forming B. thuringiensis produces highly insecticidal proteinaceous crystals dur ing sporulation. Many different B. thuringiensis crystal protein genes (cry genes) have been 50 identified and the encoded proteins appear to display different insecticidal spectra. Several crystal proteins may occur in a single crystal. Novel activities of newly discovered crystal proteins, directed at mites, nematodes, flat worms and protozoa, have recently been claimed (Feitelson et al, 1992). Research at CPRO-DLO has focused on the identification and analysis of cry genes and the encoded crystal proteins specifically active against the insect genus Spodoptera, to which many of the world's most important pest insects belong. Spodoptera species appeared to be only marginally sensitive to the commercially available crystal protein products that were based on a single bacterial strain and mostly contained three closely related crystal proteins (all CrylA). By analyzing the DNA of several strains of B. thuringiensis showing high activity against Spodoptera species two additional gene classes were identified (crylCan d crylE).A t the DNA level, the homology of these genes with the cryLA genes appeared to be approximately 65%. However, like most crystal proteins, the encoded proteins need to be proteolytically con verted by the insect into smaller toxic molecules, and genetic variation appeared to be greater in those parts of the genes encoding the actual toxins. The overall DNA homology for the toxic parts of CrylA, CrylC and CrylE was only about 50%. The encoded crystal proteins displayed specific and high toxicity for the two species Spodoptera exigua and S. littoralis, but each in a different ratio (Visser et al., 1988, Visser et al., 1990). Preliminary evidence shows the existence of a third gene class with high activity against Spodoptera species. Analysis of the amino acid sequences of the different crystal proteins active towards lepidopteran species (CryI), suggested not only an evolutionary relationship, but also recent recombination events between different genes as the source for new gene classes (Visser et al., 1992). Based on homology and insecticidal spectrum, a total of nine different gene classes appear to cover a vast array of important lepidopteran pest insects. In view of the development of resistance to crystal protein demonstrated in some heavily exposed insect populations, a continuing quest for additional crystal proteins, shown to act inde pendently, is essential to guarantee the durability of crystal protein based control pro gramme. Domain-fiinctkm studies towards manipulation ofinsect specificity As addressed above, the large natural crystal protein diversity is gradually being discov ered. However, economically important pests such as leaf miners, white flies, and thrips cannot be controlled with the currently identified crystal proteins, and other insect species are only insufficiently controlled. Understanding the mode of action, in combination with the results of domain-function studies might enable us to manipulate the toxicity level and host range, thereby complementing natural diversity. The mode of action of the crystal proteins toxic for Lepidoptera has been largely elucidated. After oral uptake, the crystals are dissolved in the alkaline environment of the larval mid-gut and the solubilized crystal proteins of 130-140 kDa are then proteolytically processed by mid-gut proteases releasing the proteinase-resistant toxin. Binding of this toxin to receptors of the mid-gut epithelial cells of susceptible insect larvae is followed by pore formation across the cell membranes which disturbs the ion permeability of these 51 membranes, resulting in swelling and bursting of the cells. As a consequence of lysis the larvae eventually die (Honée & Visser, in press). Domain-function studies aim to link specific functions of a protein to specific structures (domains, amino acid sequences) of this molecule. In an attempt to identify the region(s) in the crystal protein determining insect specificity we constructed hybrid proteins composed of sequences of two crystal proteins, CryIA(b) and CrylC. Results demonstrated that the C-terminal part of the toxic fragment determines specific receptor binding, which in turn determines to a large extent the insect specificity (Honée et al., 1991).Additiona l data from other research groups (Honée et al., 1991, references therein) on different crystal proteins have also indicated the prominent role of the C-terminal part of the toxin in insect specificity. The three-dimensional structure of a different crystal protein (CryllLA) deter mined by X-ray crystallography, confirmed these genetic and biochemical data (Li et al., 1991). Studies with a second set of constructed hybrid genes involving crylA(b), crylCan d crylE suggest that it may indeed be possible to successfully manipulate the insect host range, thus allowing for fine-tuning and adaptation ofth e toxicity spectrum. A hybrid protein in which about 120 CrylE residues located at the toxin's C-terminus, were replaced by the homolo gous CrylC sequence expanded the host range of this toxin to include S. exigua (pers. comm. D. Bosch). Crystal protein gene expression and resistance in transgenic plants Insect-resistant transgenic plants have been obtained by introducing the genes crylA(b) or crylA(a) (Fischhoffet al., 1987). However, the level of expression of B. thuringiensis crystal proteins islo w compared with many other heterologous genes. Although plants expressing those low levels were protected against some lepidopteran insects, higher expression levels and more specific crystal proteins are required to control agronomically important lepi dopteran insects. Modifying the coding sequence of the ctylA(b) gene resulted in dramati cally increased expression levels in tobacco, tomato, and cotton, even when this modifica tion was limited to a small percentage (3%) of the nucleotides constituting the gene. The partially modified genes appeared to be expressed five-fold to ten-fold higher than the wild-type gene. Full modification of the gene, a procedure too costly and laborious to apply to many different genes, resulted in a hundred-fold improvement of expression (Perlak et al., 1991). Although the involvement of potential polyadenylation signals and specific instability motifs in the poor expression ofcry genes in plants has been suggested, the question remains whether the criteria used to design a partially modified crylA(b) gene could also be success fully applied to other crystal protein gene classes. In an attempt to minimize the efforts to obtain transgenic plants well-protected against S. exigua, we introduced slightly modified versions of the cryIA(b) and crylCgene s in tobacco and tomato, featuring only 21 and 45 nucleotide substitutions over a total of approxi mately 2000. The frequency of both transgenic tomato and tobacco plants showing high toxicity against the insects Manduca sexta, Heliothisvirescens and S. exigua showed an impres sive improvement (Table 2.6.1; Fig . 2.6.1). This result strongly suggests the feasibility of an approach involving only a minimal number of substitutions to obtain higher expression 52 Table 2.6.1. Insect resistance of transgenic plants. Values tabulated are number of insect-resistant plants to the left and total number of kanamycin-selected plants to the right. Insect resistance was defined as death of all larvae within four days on excised leaves from transgenic plants. Larvae of Manduca sexta, Heliothis virescens and Spodoptera exigua. wt, wild-type nucleotide sequence; md, modified sequence. Plasmid M. sexta H. virescens S. exigua tomato tobacco tobacco tobacco wtcrylA(b) 1/12 — — — mdciylA(b) 18/20 7/10 1/10 - wtctylC - 0/47 0/47 0/47 mdcrylC - - - 3/17 levels ofa t least several crystal protein genes. The approach may be similarly feasible and essentialfo r the application ofsom eothe r prokaryotic genesi nplants ,suc ha sT 4 lysozyme for bacterial control (During, Hamburg, pers. comm.). Fig. 2.6.1. Bioassay results with the tobacco horn worm Manduca sexta of transgenic (left) and control (right) tobacco leaf tissue, after four days feeding of newly hatched larvae. 53 2.6.2 Proteinase inhibitors from potato tuber for insect control Resistance in transgenics: interplay between endogenous and heterohgous proteinase inhibitors? Recent reports have provided data showing resistance to insects in transgenic tobacco constitutively expressing heterologous trypsin inhibitors (Hilder et al., 1987;Johnson et al., 1989). However, potential changes in the endogenous inhibitor levels during the 5-7 days of insect infestations were not monitored, as it was believed that wounding (i.e. crushing) does not induce proteinase inhibitors (Pis) in wild-type tobacco. During the course of our studies we discovered that, in contrast to wounding by crush ing, wounding by insects (larvae of M. sexta) induced high levels of Pis in wild-type, non-transformed tobacco. Therefore, an experiment was designed to establish whether tobacco distinguished between these two types of wounding (Fig. 2.6.2, manuscript in preparation). Leaves were damaged either by cutting (avoiding or not avoiding leaf veins) or by crushing, which clarified that tobacco (in contrast to tomato) specifically responded to damage by cutting and not to crushing. In other words, this experiment confirmed that the type of damage certainly determined whether or not Pis were induced in tobacco. As a next step, we determined the time course of endogenous Pi-induction in tobacco, to be able to relate this pattern to the constitutive expression of heterologous transgenic inhibitors. For this purpose a quantitative radial inhibitor diffusion assay was developed, directly visualizing activity levels of trypsin, chymotrypsin (predominant proteases in the Induction ofchymotrypsi n inhibitors IU/mg \oo^> crush •(^P- serial '/ // \^^" herringbone Z^ZZZZ ZZZZ 7o — Petit Havana Moneymaker TOBACCO TOMATO Fig. 2.6.2. Diagram of the chymotrypsin inhibitor activities in leaves 3, 5 and 7 of tobacco 'Samsun NN' and 'Petit Havanna SRI' and tomato 'Moneymaker', induced 7 days after subjecting leaf 5 each day to various types of leaf damage. Leaves are numbered from bottom upward. Inhibitor levels are expressed as inhibitor units (IU) per milligram of total soluble leaf protein. One IU is defined as inhibiting one microgram of proteinase. 54 guts of lepkiopterans) and subtilisin (Fig. 2.6.3, manuscript in preparation), and enabling fast quantification of inhibitor concentrations of large sample numbers. After one week of daily removal of a leaf strip from a single leaf the measured inhibitor levels in the wounded leaf approached the reported transgenic levels. This experiment will be repeated using insects instead of scissors to estimate the contribution of endogenous induced inhibitors to the total Pi-level in transgenic plants during insect attack. Additional experiments are needed to elucidate whether the insect species used are at all sensitive to the endogenous Pis. In conclusion, endogenous levels of Pis induced during insect attack on transgenic tobacco will probably add both to the level and diversity (additional Pi-classes next to trypsin inhibitors) of proteinase inhibitor activities. We suggest that induced, endogenous inhibitors present in wild-type, non-transformed tobacco plants may have enhanced the reported growth-limiting effects of the transgenic inhibitors on larvae of several insect species. Alteringproteinase inhibitor specificity Four cDNAs encoding four different PI families (PI-1 to PI-4) were isolated from potato tuber. A member of the PI-2 gene family, specifically inhibiting both trypsin and chymot- rypsin, was subsequently shown to effectively reduce larval growth of the tobacco hornworm M. sexto, upon transfer to tobacco (Johnson et al, 1989). In contrast, the transgenically expressed chymotrypsin inhibitor PI-1 appeared ineffective in reducing time (days) 01 23456789 10 Wounded Control chymotrypsin inhibition Fig. 2.6.3. Radial inhibitor diffusion assay showing the increasing chymotrypsin inhibitor levels in tobacco cv. Petit Havanna SRI in response to daily wounding ofth e leaves with a pair ofscissors . 55 insect growth. In vitro,onl y trypsin inhibition has thus far been shown to correlate with inhibition of larval growth. These observations may be explained by assuming that success of the proteinase inhibitors in a transgenic background will at least partly depend on their ability to inhibit the major proteinases ofth e insect gut. Alteration ofvariou s inhibitors into trypsin-specific analogues may thus lead to an improved performance of these molecules in insect control. Since the most relevant criterium for PI specificity has appeared to be the nature of the Pl-residue of the active site,w e changed the nucleotide codes for the relevant residues in the PI-1 and PI-2 genes to convert chymotrypsin specificity into trypsin specific ity which replaced the PI leucine residue by lysine or arginine (Fig. 2.6.4). Transgenic tobacco plants containing copies of these modified genes are currently being regenerated. Bioassays with insects on whole plants will provide data to confirm the relative importance of trypsin over chymotrypsin inhibitors in protection from insect attack. Potato Inhibitor I and chymotrypsin serine proteinase Ser ''W His.57 v*-«*-^;*^ Fig. 2.6.4. Three-dimensional representation of the molecular structure of a member of the Potato Inhibitor I family. The site of interaction including the PI residue of the inhibitor with the serine proteinase has been indicated. 56 2.6.3 Future developments Although insect-resistant plants expressing B. thuringknsiscrysta l protein genes may be among the first transgenics on the market within a few years, this will only mark the start rather than the completion of a development. The number of pests whose control will appear amenable to crystal protein expression in plants, will gradually increase, either by discovery or by construction of new genes. However, the number of genes transferred to a single crop variety will not only increase for this reason. To further improve durable resistance towards those pests that have shown rapid adaptability under selective pressure, multiple crystal protein genes will be needed. With this in mind, the transferred genes will all be adapted, either at a minimal level or more extensively. The dynamics of populations in the field may require a few sensitive plants to be maintained along with the resistant crop for control of balanced populations of plague insects to reduce drastically the risk of appearance of resistant insects. Proteinase inhibitors may play an important role in com plementing the use of crystal proteins, since inhibitors act on partially different pest species along a different mechanism. Research at CPRO-DLO will remain involved in these fields. In addition to the introduction of transgenic crops new bio-insecticide formulations based on B. thuringiensis crystal proteins will appear on the market that will either improve host range, toxicity or product stability. Bio-insecticide formulations will thus offer the user a greater flexibility in the control of pests for which the transgenic crops do not provide protection. A large number of these formulations will be based on the use of recombinant DNA technology. Research on proteinase inhibitors has been less intense, and therefore knowledge about the mechanism of action and consequently about the possible potential of these proteins in insect control is less extended. The primary effect of the inhibition of gut proteases is followed by a range of secondary physiological responses in the insect which probably attempt to compensate for the reduction in digestive ability. Scattered evidence suggests that the insect is capable of monitoring protease activity levels in the gut and of compensat ing for sudden reductions by secreting more enzyme molecules into the gut. This mecha nism may depend on the binding of a monitoring peptide to a receptor present on the cells lining the gut. We have started a programme to investigate this putative regulatory system in order to design more effective anti-digestive insect control strategies. 2.7 'Plantibodies': aversatil e approach to engineer resistance against pathogens J. Bakker, A. Schots, W.J. Stiekema & FJ. Gommers Engineering resistance against various diseases and pests is hampered by a lack of suitable genes. To overcome this problem we started a research programme aimed at obtaining resistance by making plants which produce specific monoclonal antibodies. The ability of antibodies to inhibit the function of enzymes is well known. The idea is that plants 57 producing specific monoclonal antibodies are able to inactivate molecules that are essen tial for the survival of the pathogen. In view of the potential of animals to synthesize antibodies to almost any molecular structure, this strategy should be feasible for a wide range of diseases and pests. Our research programme isfocuse d on engineering resistance against potato cyst nema todes. It is thought that monoclonal antibodies are able to block the function of the saliva proteins of this parasite. These proteins are, among others, responsible for the induction of multinucleate transfer cells upon which the nematode feeds. To illustrate the feasibility of this approach a brief overview concerning antibodies as inhibitors and the possibilities for antibody expression in plants are given. For more details on antibody technology we refer to a review article by Schots et al. (1992). 2.7.1 Monoclonal antibodies as inhibitors Monoclonal antibodies have often been used to study the function of enzymes (Schots et al., 1992). In this approach various monoclonal antibodies are raised against the enzyme by using hybridoma technology. The structure of the enzyme isinvestigate d by performing inhibition assays. For example, 11 monoclonal antibodies have been raised against aspar- agine synthetase. Inhibition experiments demonstrated that 4 of the 11 monoclonal anti bodies blocked the activity of the enzyme. Similar results have been obtained for other enzymes (see Table 2.7.1). The high incidence of inhibiting monoclonal antibodies is remarkable (Schots et al, 1992). The percentage of inhibiting antibodies ranges from approximately 30% to 75% (see Table 2.7.1). Antibodies are also able to inhibit biological processes in situ.A profound effect on cell morphology and DNA synthesis was found by blocking oncogene products through intra cellular microinjection of antibodies. Similarly, chromosome condensation was prevented by intranuclear injection of anti-actin antibodies into oocytes of frogs. 2.7.2 Engineering monoclonal antibodies Antibodies are flexible molecules as far as protein engineering is concerned. The basic structure of an antibody is shown in Fig. 2.7.1. The molecule consists of four polypeptide Table 2.7.1. Examples of inhibition of protein function by monoclonal antibodies (MAbs). For each of the enzymes, the number of MAbs raised and the number of inhibiting MAbs are shown. A relatively high proportion of the MAbs raised inhibit the activity of the enzyme. Protein antigen Number of Number of MAbs raised inhibiting MAbs Asparagine synthetase 11 4 Human pancreatic elastase 2 3 1 Terminal deoxynucleotidyl transferase 4 3 RNA Polymerase II 5 3 /J-Lactamase 9 5 58 »H VU :.^^ .CH1 .'7^ F(ab)2 \ v» MAb scFv dAb Fig. 2.7.1. Basic structure of an antibody molecule (left) and different fragments carrying the antigen-binding parts (right). The complete antibody molecule is composed of four polypeptide chains, two heavy (H) and two light (L) chains. The heavy chain consists of four parts, one variable part (VH) and three constant parts (CH1, CH2 and CH3). The light chain consists of two parts, one variable (VL) and one constant (CL) part. The VL and VH parts bind the antigen and determine the specificity. The VH and VL parts can be used with or without the constant parts. All fragments have already been expressed in bacteria. One aim of our research programme is to express single-chain antibodies (scFv) in plants, in which the VH and VL parts are linked by a peptide (further explana tions in text). chains, two heavy (H) and two light (L) chains. The antibody is a Y-shaped molecule, in which the variable parts of the light (VL) and heavy (VH) chain form the tips of the arms. The VL and VH parts bind the antigen and determine the specificity of the antibody. VL and VH fragments can be used separately from the other fragments. An interesting and valuable molecule is a single chain antibody (scFv), in which the VH and VL parts are linked by a peptide. A major advantage of this molecule is that VH and VL fragments are not able to combine at random. In a normal mammalian cell of the immune system the heavy and light chains combine at random. In mammalian cells this causes no problems, because each cell produces only one type of monoclonal antibody. However, iftw o distinct types of monoclonal antibodies are expressed in plants, a random combination of VL and VH regions will drastically reduce the affinity and specificity of both antibodies. With single chain antibodies, it ispossibl e to express the antigen binding fragments ofmor e than one monoclonal antibody without loss of affinity and specificity. By using single chain antibodies it isi n theory possible to simultaneously engineer resistance in plants against two or more pathogens. 59 Successful expression of whole antibodies and antibody fragments has been described for yeast, myeloma cell lines, plants, algae and bacteria (Winter & Milstein, 1991). A good example is a bacteriophage vector expression system to express a combinatorial library of Fab fragments (see Fig. 2.7.1) of the mouse antibody repertoire. This Fab expression library was constructed from messenger RNA isolated from stimulated mouse spleen cells. Immunoglobulin heavy chain fragments and light chains were randomly combined in a bacteriophage vector expression system. Using this procedure several million Fab frag ments can be constructed and examined for antigen binding. However, the random combination ofV H and VL fragments isa major problem in obtaining Fab fragments with a high affinity and specificity. Plants have suitable properties which enable them to produce monoclonal antibodies. Hiatt et al. (1989) transformed tobacco plants by using complete immunoglobulin genes and obtained remarkably high expression levels.Agrobacterium plasmid s containing the gene for either the heavy or the light chain were used to transform tobacco plants. The transfor mants expressing individual immunoglobulin chains were then sexually crossed to produce plants expressing both chains. In contrast with plants expressing either the light chain or the heavy chain, the expression level in plants having both chains was high, up to 1.3% of the total leaf protein. Also, During et al. (1990) described the synthesis of antibody in tobacco. They con structed an Agrobacterium plasmid containing both the heavy and light chain gene. Localiz ation and correct folding of antibodies in plant cells was studied using anti-idiotype anti bodies i.e. antibodies against the VL and VH regions. Correctly folded antibodies were detected in the endoplasmic reticulum (ER) and the chloroplasts. The flexibility of plants to express monoclonal antibodies was also demonstrated by Benvenuto et al. (1991). They produced plants expressing 'single domain' antibodies (dAb). Single domain antibodies comprise only the VH antigen-binding domain (see Fig. 2.7.1). 2.7.3 Engineering durable resistance against potato cyst nematodes Sedentary plant parasitic nematodes such as potato cyst nematodes (Globodera sp.) and root-knot nematodes (Meloidogyne sp.)induc e specialized feeding cells in the roots of their host plant upon infection. The feeding cells are the result of a redifferentiation of existing root cells, in which the metabolism has been changed to produce transfer cells. Saliva proteins from the nematode, which are injected via the stylet into the root cells play an important role in both induction and exploitation of the feeding cells. Blocking the func tion of these saliva proteins may prevent the nematodes from creating a feeding site. With this in mind, we raised monoclonal antibodies against proteins from the salivary glands of G. rostochiensis. The genes encoding for these monoclonal antibodies will be transferred to potato plants. It is expected that expression of these monoclonal antibodies in the feeding cells will inhibit the saliva proteins and thus induces resistance. Resistance obtained in this way is most likely to be durable. This can be inferred from experience with the H[ resistance gene against G. rostochiensis, a natural gene from Solanum tuberosum ssp. andigena. Durability of a resistant cultivar depends on the number of virulent 60 individuals present in a population before growth of the resistant cultivar (Bakker & Gommers, 1989). In the U.K. Hl resistance isextremel y effective and durable. No virulent G. rostochiensis populations have been found after 20 years of repeated growth of potato cultivars with the H, resistance gene. Apparently no virulent individuals of G. rostochiensis have been introduced from South America. This shows that the reproduction capacity of potato cyst nematodes is too small to break resistance by mutation and therefore keeps to the premise that antibody-mediated resistance is durable. 2.7.4 Conclusions Advances in gene technology have presented various novel ways of improving the disease resistance of crops. This has resulted in engineered resistance against insect species, bacte ria and viruses. However, suitable resistance factors against potato cyst nematodes have not yet been found. From the information given in this article it can be inferred that antibodies directed to pathogenicity factors may function as resistance factors against potato cyst nematodes. Because of the almost unlimited capacity of the immune system of mammals to synthesize specific antibodies to almost any molecular structure, this strategy should be feasible for a wide range of pathogens. 2.8 New developments in molecular flower breeding J.N.M.Mo l Many plant species serve as providers of food, fibre, oxygen and a considerable variety of secondary products. In addition, many plant species have been appreciated over the years for the beauty of their flowers, foliage and fruits. Flower breeding has become increasingly popular in the twentieth century, especially in the Netherlands. Export statistics show that more than 50% of the global cut flower exports proceed via the Netherlands. Rose, chrysanthemum and carnation together contribute about 45% of the total Dutch flower exports and have therefore been important targets for classic breeding (Mol et al. 1989). Recent advances in molecular biology, especially in gene isolation, manipulation and transfer between species make it possible to alter to the purpose the aesthetic and commer cial properties of plants. Research at the Free University in Amsterdam has focused on the genetic modulation of flower pigmentation using a technique called (anti)sense gene mod ulation. For the first time, in 1988, van der Krol showed that inverse transcription of a gene involved in flower pigmentation leads to inhibition offlowe r coloration (van der Krol et al. 1988). More recently, van der Meer et al. (1992) reported that the depression of pigmentation in reproductive organs leads to male sterility. This points to a novel function for flavonoids in male gametophyte development. Several reports have already been published in the literature about the successful application of antisense genes in primary (biosynthesis starch, ethylene, oils and fats, fruit ripening) and secondary (biosynthesis flavonoid and carotenoid pigments) metabolism. More recently, a search for genes that shape a flower has been initiated by two distinct approaches: random transposon tagging (see Hille and Zabel Section 2.4 in this book) and random isolation of genes encoding transcription factors. 2.8.1 Manipulation of flower coloration by (anti)sense genes Once flowers are formed, their beauty is increased by the deposition of brighdy coloured pigments in the vacuoles of their petals. At the beginning of this century, the chemistry of flavonoid pigments was elucidated. Today we know most of the enzymology of pigment biosynthesis and are beginning to understand the tissue-specific regulation of the expres sion of the genes involved. Flavonoids derive from the amino acid phenylalanine. Via a cascade of enzymatic conversions the seemingly infinite spectrum of flower colours is created ranging from orange (pelargonidins), red (cyanidins) to purple (delphinidins). The final flower colour is not only determined by the type of compound produced but also by vacuolar pH and possible interactions between pigments (co-pigmentation). Flower colour can be manipulated by producing mutations in 'the genes encoding the biosynthesis machinery. Before the advent of recombinant DNA technology mutations were produced by chemical treatments or irradiation. A major disadvantage of these procedures is their randomness; one is unable to target a specific gene. In 1988 van der Krol demonstrated the ability of antisense genes to target a specific plant gene. In this case a key step of the pigment biosynthesis was blocked leading to accumulation of colourless precursors. Petunia plants were transformed with a flower colour gene whose coding region was inverted relative to the start of transcription. Conse quently an RNA copy of the 'wrong' DNA strand is synthesized. This copy is called antisense RNA. It interferes with the process in which the messenger RNA from the resident copy of the same gene is translated into a protein. The plants obtained showed diminished floral pigmentation (see Fig. 2.8.1). Extended analyses have shown that only the targeted gene had diminished activity and that other genes from the pathway were unaffected. Surprisingly, some plants showed flower colour patterns in rings or stars (Fig. 2.8.2). It is not yet understood how these patterns arise and why the size of the pigmented areas can be influenced by light and by the plank hormone gibberellin. Another unexpected surprise was that normal 'sense' genes whose coding sequence was in the right orientation produced similar phenotypic effects (van der Krol et al. 1990). The experiment was designed to discover whether more genes would lead to more enzymes Fig. 2.8.1. Independent petunia plants transformed with an antisense version of a key flower pigmentation gene show different levels of pigmentation in the flower corolla. 62 and to a deeper colour. Since the endogenous as well as the incoming gene are eliminated this phenomenon is called co-suppression. Many genes even from bacterial origin are subject to co-suppression in plants and therefore the mechanism of action is of interest. The Dutch company Florigene B.V. in collaboration with the Californian biotechnol ogy company DNAP has applied the 'sense' strategy to develop pure white chrysanthe mums of the variety Moneymaker. Florigene regards this engineered cut flower as a good model for testing the complete procedure necessary for obtaining a delisted commercially approved product. Florigene has now completed field trials in the Netherlands and the USA. Breeders' rights were applied for in December 1991. In March 1992, applications were filed for delisting as a regulated product in the Netherlands and the rest of Europe. It is apparent that there are no wild strains of chrysanthemum in Northern Europe and that it is not used as a food source. Moreover, no foreign genes are used to manipulate the plant. This may facilitate acceptance of the genetically modified organism through the regulatory framework. 2.8.2 White pollen of Petunia is male sterile In the initial experiments conducted by van der Krol the pigmentation of the reproductive organs was unaffected. It appeared that the antisense gene used was not expressed at a sufficiently high level in the epidermis and in the tapetum cells ofth e anthers. By modifying the antisense gene construct van der Meer et al. (1992) obtained Petunia plants that carried white anthers and contained white pollen (see Fig. 2.8.3). When the plants carrying white anthers were self-fertilized viable seed could not be obtained. Attempts to grow pollen tubes in vitro were unsuccessful. Microscopic analysis of the anthers revealed that unripe precursors ofpolle n grains had accumulated. This indicates a novel function for flavonoids in male gametophyte development. Fig. 2.8.2. Petunia plant transformed with an antisense flower pigmentation gene showing a ring-shaped pigmentation pattern in the corollas. 63 Fig. 2.8.3. White petunia pollen ismale-sterile . Note that white pollen does not form pollen tubes in vitro. Nuclear male sterility is an important trait in hybrid seed production. The biotechnol ogy company Plant Genetic Systems (PGS) in Ghent, Belgium has recently developed a novel strategy to kill pollen grains by expressing an RNA-degrading enzyme exclusively in the tapetum cells of the anther. By inbreeding an inhibitor of this enzyme, full fertility can be restored. We are strongly convinced that inhibition of anther pigmentation is a good alternative to the PGS procedure. We are currently testing the feasibility of this approach by introducing antisense gene constructs in a variety of commercially important food and ornamental crops and evolving ways of restoring fertility. This project will be carried out in collaboration with CPRO-DLO in Wageningen and the Dutch Biotechnology Com pany MOGEN N.V. 2.8,3 Studying the genes involved in flower development Flower formation is a crucial event in the plant reproductive cycle and provides an excellent system in which to study cell differentiation and pattern formation. A basic model for determining flower organ identity has been proposed based on genetic and molecular analyses of Antirrhinuman d Arabidopsis mutant flowers. This model postulates that organ identity is specified by the combined action of three types of homeotic genes each being 64 expressed in two consecutive whorls ofth e flower (see Pruitt, 1991 for a review). In Petunia two homeotic mutants are known, both affecting the development of the petal. In the mutant 'green petals' (gp)flower petal s are transformed into sepals; in the mutant 'blind' (bl) the limb ofth epeta l istransforme d into an antheroid structure (seeFig . 2.8.4;va n Tunen et al. 1990; Geräts, 1991).Th e action ofBl an d Gp genes cannot readily be explained by the current model for determining organ identity (van Tunen & An- genent, 1991).Therefore , thesegene sma y represent asye tunidentifie d genefunction s that are required during floral determination. Fig. 2.8.4. Flower phenotype of two petunia developmental flower mutants: top, 'blind' mutant; bottom, 'green petal' mutant. 65 Two approaches were used to identify and clone the genes involved in flower morpho genesis in Petunia. In collaboration with the Free University group, van Tunen and co-workers at CPRO-DLO isolated a number of flower-specific genes containing the so-called MADS box domain. Such genes are candidate transcription factor genes that may be active during development. Recent data indicate that the expression ofon e of these MADS box genes fbpl) is regulated by the Gpgen e (Angenent et al. 1992). A second approach used to isolate flower developmental genes, is via random trans- poson-mutagenesis. We used a line of Petunia in which the recently isolated transposable element dTphl is active. A plethora of mutants was obtained in either flower development, embryo/seedling development, plant growth or flower pigmentation. The genes Gp and Bl have not yet been targeted. However, two new mutants in development were obtained: Paf and Nam. Paf. Mutants in Paf (Petals and anthersyûsed ) bear flowers in which the amount of petal tissue has increased (mostly doubled) and parts of the anthers are missing. Carpel development is also affected. This indicates that the paf gene exerts a key function in the development of the 3rd and 4th whorls of the flower. Nam: When plants heterozygous for the gene Nam (No apical meristem) are self-polli nated, twenty-five percent of the progeny seedlings are arrested in development. The cotyledons appear to be fused and development subsequendy stops. Presum ably, the Nam gene product is required during the early stages of embryogenesis. We plan to identify more mutants following this approach and to isolate the genes targeted by the transposable element dTphl. The commercial outlet for these genes is to transform them to commercially important cut flowers such as chrysanthemum, carnation, rose and gerbera. 2.9 Industrial application of plant biotechnology: promises and pitfalls E. Veltkamp The impact on industrial processes and products has been anticipated ever since 'Biotech nology' first emerged and was developed. In some areas, such as the pharmaceutical industry, applications of biotechnology have already been realized and new, improved products have been successfully introduced onto the market. There is no doubt that this new science is of vital importance for the future of clinical medicine. In other areas, such as the agricultural industry, we are still in the developmental phase. Although in the past high and sometimes overwhelming expectations were launched regarding the speed at which commercial applications could be realized, applications of plant biotechnology are taking a considerable amount of time. Several factors, some specific to this industry, are critical for the successful application of plant biotechnology. Most of them are not related to the technology itself but to the conditions created for industrial application. 66 2.9.1 Biotechnology and agricultural challenges For more than 100 years the Seeds Industry, in collaboration with the various institutes working in this field, has been active in developing new varieties of plants that can fill the world demand for food and plant-derived products. Agricultural productivity has in creased more in this century than ever before in the entire history of farming. This achievement has been made possible by improved farming practices, the development of agents that protect crops from disease and last but not least new plant varieties arising from breeding programmes. Taking into account the continuous increase in the demand for food and the need, from an environmental point of view, to limit the chemical input in agricultural practices, a major improvement in agricultural productivity must now come from improved genetic traits carried by crop plants. Plant biotechnology can contribute significantly to achieving this goal by improving product quality, providing new approaches for introducing agronomically important traits into crop plants, such as resistances to various diseases and insects, and by overcoming existing barriers in breeding. Several of these opportunities are reviewed elsewhere in this book. When making comparisons with biotech applications in other areas, such as the chemi cal and pharmaceutical industries, one should bear in mind that in agriculture the final product is not a chemical substance but an entire plant with the complexity normally associated with a living organism. 2.9.2 Aspects related to commercial applications R&D Costs It is obvious that like other industrial investments, investments in plant biotechnology should be considered when added values, in terms of social, economic or environmental benefits, are expected from the technology. For a new technology such asplan t biotechnol ogy, judging costs versus benefits, as well as the risks of the technological programmes, needs particularly careful analysis. Analysis which is required to deal with some uncertain ties. For instance, when calculating the level of investment for a specific product develop ment programme, all the cost components required to introduce the final product to the market place must be considered (see Table 2.9.1). Table 2.9.1. Components in the cost of developing a commercial transgenic plant variety. Cost components - basic research (laboratory work) - breeding (product adaption and fixing) - field testing and screening - permits (regulatory frame work) - royalties and licences (rights of third parties) - costs for registration and industrial property rights (acquiring patents) - product introduction costs (advertisements, information packages) 67 Costs I II III IV V Developmental phase > Fig. 2.9.1. Time coarse of product-development costs. Phases: I exploration phase; II basic re search/feasibility; III development of a product adaption; IV large-scale screening and producibility; V preparation of introduction to market. Because plant biotechnology is a new dimension in agriculture, there are no existing, well-established industrial practices in areas such as the licensing of the technology. Conse quently, some of the cost components will be based on guesswork. In any case it will be clear that the total costs associated with the development of a new product will be signifi cantly higher than just the costs for doing the basic biotech research in the laboratory (see Fig. 2.9.1): 'A transformed cell line is not a product but a starting material'. In the agricultural industry R&D costs are already relatively high, and many companies, active in the development of new plant varieties, are spending 8 to 12% of their total turnover to research. Since plant biotechnology does not replace current R&D activities such as breeding, but is more of an additional 'tool kit', the costs of plant biotechnology research will, in many cases, have an increasing effect on R & D budgets. It is not surprising, therefore, that many companies have only limited R&D program mes with a more fundamental character and are focusing on plant biotechnology targets where the application possibilities are not scheduled on a long-term-basis. In this context it is crucial for the sustainable development of agriculture that at both national and international level, stimulation programmes are executed so that research programme in government institutes and universities can be continued in order to strengthen our funda mental knowledge of plant science and so that the existing, basic bottlenecks for future applications can be solved. Company strategies and legislation Agricultural companies are following different strategies in the use ofplan t biotechnology. For instance, some companies are applying a 'wait and see' attitude whereas others are allocating significant resources to plant biotechnology in addition to their current breeding and technology programmes. The emergence ofplan t biotechnology has also triggered the establishment of numerous 'biotech' companies aiming to build a proprietary position in 68 certain technologies which should ultimately create a positive cash flow through licensing arrangements with partners in the market. As in other areas of technology it is anticipated that after this 'first wave', concentration and rationalization will occur over the years among these biotech companies due to factors such as critical mass, cost structure, product level times and market penetration. As such, plant biotechnology is entering the agricul tural industry in a period when the industry is in the process of internationalization and concentration. Technological opportunities with global and multi-crop potential will strengthen these dynamic changes. Recent surveys made amongst the industry have shown that the potential impact of plant biotechnology isno t in question. The lack ofclarit y in legislation and factors involved in downstream application and product introduction are major stumbling blocks for some companies to go 'full speed'. Today industry isconfronte d with different legislation in the various countries regarding plant biotechnology, and an ongoing debate on intellectual property rights. Reasons enough for various industrial bodies, such as the Senior Advisory Group Biotechnology (SAGB), to underline the need to establish more harmonized legislation, e.g. in the Euro pean community, as well as for an adequate system to protect intellectual property. For the agricultural industry this latter aspect is of crucial importance. So far, newly developed plant varieties are protected under the Plant Breeders' Rights system in countries subscrib ing to the UPOV convention. This system however, does not adequately cover new technologies and related inventions generated in the area of plant biotechnology. It is therefore crucial for the future of agricultural development that the agricultural industry, like other industries, can rely on adequate, long-term protection for its investments in terms of plant variety, protection and use of the patent system. During recent years, hundreds of officially approved field trials of transgenic plant have been conducted. The number of these trials being done per year is increasing rapidly. In 1991 around 200 field trials, covering approximately 25 different crop species, were conducted; the majority were executed in the United States, Canada and European countries including France, Belgium and the United Kingdom. In Europe, the fact that field trials of the same transgenic crop must undergo separate application procedures for each European country is still a stumbling block, one that does not exist in the US where recently the government also established clear rules for the release of commercial, trans genic plant varieties onto the market. Based on registered field trials, a trend can be monitored from tests dealing with marker genes towards field trials covering agronomic genes such as virus resistance, insect resis tance and herbicide tolerance. This clearly indicates that products now in the testing pipeline will enter the market place during the second half of the nineties. Specific agricultural aspects Regarding the benefits, the added value, generated by the use of plant biotechnology, comparisons are made in some articles with applications in the pharmaceutical industry including anticipated royalties. These comparisons are usually inadequate. First of all the cost structure of agricultural and pharmaceutical products are different and many agricul tural products have lower margins than pharmaceuticals. Another important factor is 69 market segmentation. For instance for a specific, single new trait in a crop plant one cannot simply calculate the world acreage of that crop and determine the added value on that basis. To be more specific on this point the following observation should be kept in mind: In contrast to a new pharmaceutical product with a broad, world market potential, a new agronomic trait usually has to be introduced in different plant genetic backgrounds (germplasms) adapted to different environmental conditions or users' demands for a given crop plant. This segmentation of the agricultural market implies higher introduction costs than for just one 'world' product whereas the added value can only be captured from certain segments in the market benefiting from the new trait. With respect to the different kinds of technologies needed for making a final, new plant variety it is unlikely that one player in the industry will be in possession of all the related proprietary rights. Various parties in the market will require a reimbursement, such as royalties, for their contribution. In setting up these types of transactions an understanding of the specific economics of the industry is crucial; a cumulative summing-up of different royalties to be paid to technology suppliers will inevitably run the risk of margin erosion. Public acceptance When reviewing the application of plant biotechnology, the acceptance of products cre ated by the use of biotechnology is of vital importance. More than in the past, industry should play an active role in providing relevant information with an open mind so as to avoid misunderstanding and misinterpretation due to lack of knowledge. The establish ment of industrial guidelines for the use of biotechnology and the release of products mediated by biotechnology can contribute to this aim. In this context both GIFAP and OECD have released position documents which are already widely supported. 2.9.3 Conclusions Plant biotechnology isno t a product in itself. It isa powerful set oftool s that will contribute significantly to our ability to meet agricultural challenges. In spite of some sceptics, I am convinced of the vital impact this technology will have on improving agricultural produc tivity. As with any innovative technology, plant biotechnology will change economic and competitive conditions in the market. For industries active in this area the price ticket will be high. The development ofne w plant varieties, and the applications of new technological approaches need significant resources and time. In addition, based on the structure of the market, specific adaptations are required to fulfil the demands of the various segments of the market. Successful implementation of plant biotechnology in the development of improved agricultural products and productivity does not only depend on the willingness of the industry to invest in this area. Supportive policies by governments are also needed to create the appropriate conditions for industrial applications. International coordination of legisla tion is essential so as to avoid unreasonable restraints on certain industries or competitive disadvantages of one region versus others. Transparency will be the key word; transpar ency in procedures, regulations and real information on the technological possibilities and limitations. Transparency will strengthen public confidence, confidence in the progress of new technologies which, when used prudently, can benefit society. 70 2.10 Market introduction of products obtained through plant biotechnology, aspects of consumer acceptance T.A.W.M. Saat 2.10.1 What is 'better consumer acceptance', and how can it be promoted? During thefirst tw o decades ofit sexistence , modern plant biotechnology was confined to laboratories, greenhouses and small-scalefield plots .No wtha t thistechnolog y has come of age itmus t go out into the world and find itselfaccepted , or.... rejected. Even though data from field trials and greenhouses show no specific risks or unexpected negative effects of plant biotechnology (Coulston & Kolbye, 1990; Landsmann, 1992), consumers are ex pected to be reserved. Some consumers are already expressing dissatisfaction (Hamstra, 1992) over: - plant breeding (everything used to be tastier with old crop varieties); - plant growing (everything used to be healthier before agrochemicals came into use); - plant biotechnology (everything used tob e more natural, without any riskt oma n or the environment). These signals are sustained and deserve to be taken seriously. One of the main rules of the market is that consumers are always right, even if they are wrong. One way to help consumers to overcome some of their fears is to provide information (Verrips, 1991; Giddings, 1989). This should not aim at selling any particular product, but should cover a broader area, including the conventional breeding and growing of crops. Information about goodqualit ywil lenabl eth econsume r todecid ewhethe r toaccep t orrejec t products on the basis of'just reasons' (Hamstra, 1991). 'Better acceptance' does not and should not therefore necessarily mean more acceptance. It isa challenge to biotechnologists, breeders and regulators alike to provide consumers with good, reliable information. But notjus t anybody can deliver this information to the public. The message must be of high quality. The source of the information must also be credible,competen t and attractive, alla t the sametim e(d eLoor , 1990).Fro m Eurobarom eter, the EC-wide survey, the 'Danish conclusion' was distilled: 'Extensive, open and honest information/debate can raise awareness and acceptance of biotechnology' (Candey, 1991). 2.10.2 Public attitude towards food and food production Conventional breedingand common sense There can be no serious doubt that during this century our food supply has steadily improved in quality, variety, nutritional value and safety, thanks to plant breeding (Coul ston & Kolbye, 1990).Eve n so,i ti sa myththa t fruit andvegetable s are alwayshealthy .A s Janzen wrote: 'plants are notjus t food for animals....Th e world isno t green. It iscoloure d lectin, tannin, cyanide, caffeine, aflatoxin, and canavanine'. In the course ofevolution , to fight off insects and other herbivores, plants not only took up a wide range of stinging defence systems,bu t developed a huge retaliative potential over at least 500 million years. 71 During the development of agriculture, man gradually turned to crop protection. This enabled him to reduce levels ofth e now redundant anti-nutritional substances by selection, much to the benefit of mankind (Coulston & Kolbye, 1990;Ame s et al., 1990b). Neverthe less, no more than 0.01 percent of dietary pesticides consists of relatively well-character ized and evaluated synthetic substances, the rest being natural (Ames et al., 1990a). This ratio, however, has hardly been noticed, judging from public concern and the regulatory effort resulting from this fact. Public attitude towardsfood production, agrochemicals Since the seventies, consumers have shown concern about the chemical industry and its products: 'there is a tendency for non-scientists to think of chemicals as being only syn thetic and to characterize synthetic chemicals as toxic, as if every natural chemical were not also toxic at some dose' (Ames et al., 1990b). This discrimination of synthetic versus natural chemicals has been dubbed chemophobia. A similar public bias has developed towards favouring products considered 'natural' over products regarded as 'manipulated'. Surveys onfood and the attitude tobiotechnology The lack of knowledge about conventional food forms an obstacle in spreading compre hensive information about genetic modification (Hamstra, 1991). An American survey showed that only 20 percent of the respondents answered in the affirmative when asked whether they had ever eaten food from hybrid crops (McCammon, 1992). In a Dutch survey, more than half the respondents stated that they had never eaten food that con tained DNA. 'The public is seriously worried about sterilization by radiation, food addi tives to improve colour, flavour and shelf-life and contamination by synthetic chemicals' and, according to Huis in 't Veld & Hoogenboom-Verdegaal, 'they pass by the real problem of food safety: bacterial contamination'. In 1989 only 57 percent of a representative group of the Dutch population had ever heard of biotechnology. Of these, only 64 percent, that is 37 percent of the group as a whole, chose the right definition from four possibilities (Hamstra & Feenstra, 1989). A list of 10 products was presented to the group of people who had chosen the right definition of biotechnology and they were asked to mark which individual product was produced with the help of biotechnology. On average only 4.4 answers out of 10wer e right. General knowledge of biotechnology would seem to be extremely limited. Some applications of plant biotechnology, such as the introduction of resistance against herbicides, even when they are environmentally superior, can give rise to chemophobia (Hamstra, 1991, 1992). The very moment, however, that biotechnology can realize the desire of the consumer to reduce the use ofpesticides , then a positive attitude from a large portion of Dutch consumers can be anticipated (Hamstra, 1992). Improvement in taste will also be welcomed, according to the same source. These results from various surveys indicate that although consumers have a genuine feeling of unease about yet another modern, high-impact technology, they have but limited knowledge of plant biotechnology. Their current judgement of biotechnology is related to their view of technology in general, and might change to either side depending on what kind of transgenic product first hits the market. 72 2.10.3 Previous introductions of new agricultural products onto the market Not all plant breeding firms are experienced in communicating with the final purchasers of their products. Only breeders of flowers and ornamentals develop varieties about which the general public can distinguish and comment. Few people are able to discern and judge the individual varieties of fruit and vegetables when filling their shopping basket; and with processed foods such as sugar, starch and edible oil the consumer is unable to tell from which variety or sometimes even from which crop it was produced. Formerly, providing technical information to consumers was not essential for new developments to become accepted. Hybrid ornamentals and tomatoes can be found in many allotment gardens, Petri-dish sown orchids and mass-propagated anthuriums fill today's florist's shop. In Canada, more than 1500 new plant varieties, bearing new traits caused by random mutations, have been accepted without hesitation (Downey, 1992). Completely new food items, introduced onto the Dutch market, such as kiwis, carambolas and aspartame became almost as readily accepted as many oriental and South American products in the past. In fact, very few plants that man eats today would have been present in the diet of our ancestors, the African hunter-gatherers. There has not been enough time for evolution by natural selection to develop specific defence mechanisms in man against the unique constituents of these novel foods (Ames et al., 1990b). Over the millennia, until very recently, the development of man and his food could proceed without legislation. It is a good-natured challenge to imagine how human evolution would have worked out if our African ancestors had decided to ban every new development until proven absolutely safe. Assuming, of course, that survival and any evolution could possibly have taken place under such restrictions in the first place. 2.10.4 Product information, labelling = libelling? It is, of course, principally wrong to label foods derived from genetically modified organ isms as no specific risks can be attributed to recombinant DNA techniques perse (Coulsto n & Kolbye, 1990; Verrips, 1991). Nevertheless, one should seriously consider labelling products if the public at large demands it. Some companies expect that there will be cultural differences in accepting products that are labelled. The British firm ICI is to transfer market introduction of its transgenic, longer shelf-life tomato to the United States for fear of lack of acceptance in Europe. Calgene, who plans to introduce a similar product onto the US market under the name of 'flavr savr', even expects to benefit from labelling its transgenic tomato, since research has shown that American consumers are willing to pay a bonus for this specific product of biotechnology. In China, public acceptance of transgenic tomatoes has never been an issue at all (Chen, 1992; Landsmann, 1992). The Chinese and American people seem to regard recombinant DNA food products mainly as an 'opportunity' while in this part of the world they are mainly seen as a 'threat'. The second international symposium on biosafety, summarized it as follows (Landsmann, 1992): 'We are still in need of data about real as well as perceived risks. But on the other hand we must be careful not to put our scientific credibility at stake by designing risk- related research and precautionary measures with no probability for useful results but 73 solely for public appeasement. This will backfire when we least need it. We are highly aware: genetic engineering is not a game for scientists, we must and we do think profession ally. If we do not take it seriously how can we ask the public to be reasonable'. 2.11 References Aarts.J.M.MJ.G.,J.G.J . Hontelez, P. Fischer, R. Verkerk, A. van Kammen & P. Zabel, 1991. Acid phosphatase-1' , a tightly linked molecular marker for root-knot nematode resistance in tomato: from protein to gene using PCR and degenerate primers containing deoxyinosine. Plant Molecu lar Biology 16:647-661. Ames, B.N., M. Profet & L. Swirsky Gold, 1990a. Dietary pesticides (99.99% all natural). Proceed ings of the National Acadamy of Sciences of the USA 87:7777-7781. Ames, B.N., M. Profet & L. Swirsky Gold, 1990b. Nature's chemicals and synthetic chemicals: comparative toxicology. Proceedings of the National Acadamy of Sciences of the USA 87:7782- 7786. Angenent, G.C., M.Busscher, J.Franken, J.N.M.Mol & AJ. van Tunen, 1992. Differential expres sion of two MADS box genes in wild-type and mutant Petunia flowers. The Plant Cell 4:983-993. Bakker, J. & FJ. Gommers, 1989. Characterization of populations of pathotypes of potato cyst nematodes. In: H.D. Loxdale andJ . den Hollander (Eds.): Electrophoretic studies on agricultural pests. Clarenton Press, Oxford, pp. 415^30. Bennetzen, J.L., & D.G.Jones, 1992. Approaches and progress in the molecular cloning of plant disease resistance genes. In:J.K . Settlow (Ed.): Genetic Engineering, Principles and Methods, Vol. 14: in press. Benvenuto E., RJ. Ordas, R. Tavaza, G. Ancora, S. Biocca, A. Cattaneo & P. Galeffi, 1991. 'Phytoantibodies': a general vector for the expression of immunoglobulin domains in transgenic plants. Plant Molecular Biology 17:865-874. Biffen, R.H., 1905. Mendel's laws of inheritance and wheat breeding. Journal of Agricultural Science 1:4—48. Bohlmann, H. & K. Apel, 1991. Thionins. Annual Review of Plant Physiology and Plant Molecular Biology 42:227-240. Bonas, U., R. Schulte, S. Fenselau, G.V Minsavage, B.J. Staskawicz & R.E. Stall, 1991. Isolation of gene cluster from Xanthomonas campestris pv. vesicatoria that determines pathogenicity and the hypersensitive response on pepper and tomato. Molecular Plant-Microbe Interactions 4:81-88. Breyne, P., M. van Montagu, A. Depicker & G. Gheysen, 1992. Characterization of a plant scaffold attachment region in a DNA fragment that normalizes expression on tobacco. The Plant Cell 4:463-471 Broekaert, W.F,J . van Parijs, F. Leyns, H.Joo s & WJ. Peumans, 1989. A chitin-binding lectin from stinging nettle rhizomes with antifungal properties. Science 245:1100-1102. Broekaert, W.F, W. Marien, F.R.G. Terras, M.F.C. De Bolle, P. Proost,J . Van Damme, L. Dillen, M. Cleas, S.B. Rees, J. Vanderleyden & B.P.A. Cammue, 1992. Antimicrobial peptides from Amaranthuscaudatus seeds with sequence homology to the cysteine/glycine-rich domain of chitin- binding proteins. Biochemistry 31:4308^314. Broglie, K., I. Chet, M. Holliday, R. Cressman, Ph. Biddle, S. Knowlton, CJ . Mauvais & R. Broglie, 1991. Transgenic plants with enhanced resistance to the fungal pathogen Rhizoctonia solani. Sci ence 254:1194-1197. Cantley, M., 1991. Eurobarometer biotechnology opinion survey, spring 1991, summary report. Congress Proceedings, acceptatie van biotechnologische producten, november 7, N.S.C., Vlaar- dingen, The Netherlands. Carr, J.P., L.E. Marsh, G.P Lomonossoff, M.E. Sekiya & M. Zaitlin, 1992. Resistance to tobacco mosaic virus induced by the 54-kDa gene sequence requires expression of the 54-kDa protein. Molecular Plant Microbe Interactions 5:397-404. 74 Chen, Z-L, 1992. Field releases of recombinant bacteria and transgenic plants in China. Lecture at 2nd international symposium on the biosafety results of field tests of genetically modified plants and micro-organisms, Goslar, Germany, 11-14 may. Chrispeels, M.J. & N. V. Raikhel, 1991. Lectins, lectin genes, and their role in plant defense. The Plant Cell 3:1-9. Coulston, F. & A.C. Kolbye (Eds.), 1990. International food biotechnology council, Biotechnologies and food: assuring the safety of foods produced by genetic modification. Regulatory toxicology and pharmacology, volume 12 (3), part 2. D'Halluin, K., M. Bossut, E. Bonne, B. Mazur, J. Leemans &J . Botterman, 1992. Transformation of sugarbeet {Betavulgaris L.) and evaluation of herbicide resistance in transgenic plants. Bio/ technology 10:309-314. de Wit, PJ.G.M., 1992. Molecular characterization of gene-for-gene systems in plant-fungus inter actions and the application of avirulence genes in control of plant pathogens. Annual Review of Phytopathology 30:391-418. De Carvalho F., G. Gheysen, S. Kushnir, M. van Montagu, D. Inzé & C. Castresana, 1992. Suppression of /M,3-glucanase transgene expression in homozygous plants. EMBO Journal 11:2595-2602 de Loor, M., 1990. Voorlichting over biotechnologie, lecture at congres 'Genetische modifikatie: op weg naar maatschappelijke acceptatie?', October 29, N.S.C., Vlaardingen, The Netherlands. Dekkers, J J., H.C. van der Plas & D.H. Vuyk, 1990. Agricultural biotechnology in focus in the Netherlands, Pudoc, Wageningen pp. 285. Downey, R.K., 1992. Biosafety of transgenic oilseed Brassica species. 2nd international symposium on the biosafety results of field tests of genetically modified plants and micro-organisms, Goslar, Germany, 11-14 may. During, K., S. Hippe, F. Kreuzaler &J. Schell, 1990. Synthesis and self-assembly of a functional monoclonal antibody in transgenic Mcotiana tabacum.Plan t Molecular Biology 15:281-293. Feitelson,J.S. ,J . Payne, & L. Kim, 1992. Bacillus thuringiensis: Insects and beyond. Bio/Technology 10:271-275. Fischhoff, D.A., K.S. Bowdish, FJ. Perlak, P.G. Marrone, S.M. McCormick, J.G. Niedermeyer, D.A. Dean, K. Kusano-Kretzmer, EJ. Mayer, D.E. Rochester, S.G. Rogers & and R.T. Fraley, 1987. Insect tolerant transgenic tomato plants. Bio/Technology 5:807-813. Geräts, A.G.M., 1991. Mutants involved in floral and plant development in Petunia. Plant Science 80:19-25. Giddings, L.V., 1989. The last word/face the challenge. Bio/technology 7:1200. Hain, R., B. Bieseler, H. Kindl, G. Schröder & R. Stöcker, 1990. Expression of a stilbene synthase gene in Mcotiana tabacum results in synthesis of the phytoalexin resveratrol. Plant Molecular Biology 15:325-335. Hamstra, A.M. & M.H. Feenstra, 1989. Consument en biotechnologie: kennis en meningsvorming. SWOKA research report number 85, SWOKA, The Hague, the Netherlands. Hamstra, A.M., 1991.Biotechnolog y in foodstuffs, towards a model of consumer acceptance. Swoka research report number 105, SWOKA, The Hague, the Netherlands. Hamstra, A.M., 1992. Acceptatie van transgene produkten door de consument, een analyse uit consumentenonderzoek. Lecture within the framework of'Problems in plant breeding', Wagenin gen Agricultural University, February 28, SWOKA, The Hague, the Netherlands. Haring, M.A., C.M.T. Rommens, H.J.J. Nijkamp &J. Hille, 1991. The use of transgenic plants to understand transposition mechanisms and to develop transposon tagging strategies. Plant Molec ular Biology 16:449-461. Hauge, B., P.Stam, E.M. Meyerowitz, M. Koornneef, M. Cherry & H.M. Goodman, 1992. The integrated RFLP and classical genetic linkage map of Arabidopsis thaliana. Submitted. Hensgens, L.A.M. & R.A. Schilperoort, 1992. Expression of transferred genes in transgenic rice (tissues) and tobacco. In: R. Appels et al. (Eds.): Gene conservation and Exploitation. Plenum Press, in press. 75 Hiatt, A., R. CafFerkey & K. Bowdish, 1989. Production of antibodies in transgenic plants. Nature 342:76-78. Hilder, V.A., A.M.R. Gatehouse, S.E. Sheerman, R.F. Barker & D. Boulter, 1987. A novel mecha nism of insect resistance engineered into tobacco. Nature 330:160-163. Hille,J. , P. Zabel & M. Koornneef, 1991. Genetic transformation of tomato and prospects for gene transfer. In: Kalloo (Ed.): Monographs on Theoretical and Applied Genetics 14: Genetic im provement of tomato. Springer Verlag Heidelberg, pp. 283-291. Hinze, K., R.D. Thompson, E. Ritter, F. Salamini, & P. Schulze-Lefert, 1991.Restrictio n fragment length polymorphism-mediated targeting of the ml-o resistance locus in barley (Hordeum vulgare). Proceedings of the National Academy of Science of the USA 88:3691-3695. Ho,J-Y. , R. Weide, H.M. Ma, M.F. van Wordragen, K.N. Lambert, M. Koornneef, P. Zabel, & V. Williamson, 1992. The root-knot nematode resistance gene (Mi) in tomato: construction of a molecular linkage map and identification of dominant cDNA markers in resistant genotypes. Plant PathologyJournal , in press. Honée, G., D. Convents, J. van Rie, S.Jansens, M. Peferoen, & B. Visser, 1991. The C-terminal domain of the toxic fragment of a Bacillusthuringiensis crystal protein determines receptor binding. Molecular Microbiology 5:2799 - 2806. Hooykaas, PJJ. & R.A. Schilperoort, 1992. Agrobacterium and plant genetic engineering. Plant Molecular Biology 19:15-38. Huis in 't Veld,J.HJ . & A.M.M. Hoogenboom-Verdegaal, 1992. Besmette etenswaar. Natuur en Techniek 60:335-345. Jach, G., S. Logemann, G. Wolf, A. Oppenheim, I. Chet,J. Schell &J. Logemann, 1992. Expression of a bacterial chitinase leads to improved resistance of transgenic tobacco plants against fungal infection. Biopractice 1:33-40. Jansen, R.C., 1992. A general mixture model for mapping quantitative trait loci using molecular markers. Theoretical and Applied Genetics, 85:252-260. Janzen, D.H., 1977. Annales of the Missouri Botanical Gardens, 64:706-736. Johnson, R, J. Narvaez, G. An & C.A. Ryan, 1989. Expression of proteinase inhibitors I and II in transgenic plants: effects on natural defense against Manduca sexta larvae. Proceeding of the National Academy of Science of the USA 86:9871-9875. Kearney, B., & B.J. Staskawicz, 1990. Widespread distribution and fitness contribution of Xantho- monas campestris avirulence gene avrBs2. Nature 346:385-386. Keen, N.T., 1990. Gene-for-gene complementarity in plant-pathogen interactions. Annual Review of Genetics 24:447-463. Klein-Lankhorst, R., P. Rietveld, B. Machiels, R. Verkerk, R. Weide, C. Gebhardt, M. Koornneef, & P. Zabel, 1991a . RFLP markers linked to the root knot nematode resistance gene Mi in tomato. Theoretical and Applied Genetics 81:661-667. Klein-Lankhorst, R., A. Vermunt, R. Weide, T. Liharska & P. Zabel, 1991b. Isolation of molecular markers for tomato (L.esculentum ) using random amplified polymorphic DNA (RAPD). Theoret ical and Applied Genetics 83:108-114. Knogge, W., 1991. Plant resistance genes for fungal pathogens-physiological models and identifica tion in cereal crops. Zeitschrift für Naturforschung 460:961-981. Kool, M., F.L.J, van Lier,J.M . Vlak &J. Tramper, 1990. Production of (recombinant) baculovirus in a two-stage bioreactor system using insect-cell cultures. In:J J. Dekkers, H.C. van der Plas & D.H. Vuyk (Eds.): Agricultural biotechnology in focus in The Netherlands, pp. 64 - 69. Pudoc, Wageningen. Lander, E.S. & D. Botstein, 1989. Mapping Mendelian factors underlying quantitative traits using RFLP linkage maps. Genetics 121:185-199. Landsmann,J. , 1992. The rapporteur's report. 2nd international symposium on the biosafety results of field tests of genetically modified plants and micro-organisms, Goslar, Germany, 11-14 May. Leah, R., H. Tommerup, I. Svendsen &J. Mundy, 1991. Biochemical and molecular characteriza tion of three barley seed proteins with antifungal properties. TheJourna l of Biochemical Chemis- 76 try 266:1464-1573. Li,J. ,J . Carroll & D. Ellar, 1991. Crystal structure of insecticidal J-endotoxin from Bacillus thuringien- sisa t 2.5  resolution. Nature 353:815-821. Logemann,J. , G.Jach, H. Tommerup,J. Mundy &J. Schell, 1992. Expression of a barley ribosome- inactivating protein leads to increased fungal protection in transgenic tobacco plants. Bio/Tech nology 10:305-308. McCammon, S., 1992. Commercialization review process for plant varieties developed through biotechnology and subsequent panel discussion. Proceedings of the 2nd international symposium on the biosafety results of field tests of genetically modified plants and micro-organisms, Goslar, Germany, 11-14 May. Melchers, L.S., M.B. Sela-Buurlage, S.A., Vloemans, CR Woloshuk, J.S.C. van Roekei, J. Pen, PJ.M. van den Elzen & B.J.C. Cornelissen, 1992. Extracellular targeting of the vacuolar tobacco proteins AP24, chitinase and /M,3-glucanase in transgenic plants. Plant Molecular Biology, in press. Mitten, D., 1992. Safety assessment and commercialization of transgenic fresh tomato food prod ucts, transgenic cotton products and transgenic rapeseed oil products, and subsequent panel discussion. 2nd international symposium on the biosafety results of field tests of genetically modi fied plants and micro-organisms, Goslar, Germany, 11-14 may. Mol,J. , A. Stuitje, A. Geräts, A. van der Krol & R.Jorgensen , 1989. Saying it with genes: molecular Flower breeding. Trends in Biotechnology 7:148-153. Mol, J., R. van Blokland &J. Kooter, 1991. More about co-suppression. Trends Biotechnology 9:181-182. Neuhaus,J.-M. , P.Ahl-Goy , U. Hinz, S. Flores & F. Meins, 1991. High-level expression of a tobacco chitinase gene in Mcotiana sylvestris. Susceptibility of transgenic plants to Cercospora nicotianae infec tion. Plant Molecular Biology 16:141-151. Niks, R.E., 1988. Non-host plant species as donors for resistance to pathogens with narrow host range. II. Concepts and evidence on the genetic basis of nonhost resistance. Euphytica 37:89-99. Offringa, R., PJ.M. van den Elzen & PJJ. Hooykaas, 1992. Gene targeting in plants using thé Agrobacterium vector system. Transgenic Research 1:114—123 Parlevliet, J.E. &J.C. Zadoks, 1977. The integrated concept of disease resistance; a new view including horizontal and vertical resistance in plants. Euphytica 26:5-21. Perlak, FJ., R.L. Fuchs, D.A. Dean, S.L. McPherson & D.A. Fischhoff, 1991. Modification of the coding sequence enhances plant expression of insect control protein genes. Proceedings of the National Academy of Science of the USA 88:3324-3328. Potrykus, I., 1990. Gene transfer to cereals: An assessment. Bio/technology 9:535-542. Pruitt, R.E., 1991.Molecula r genetics of floral development in Arabidopsis thaliana. BioEssays 13:347- 349. Pryor, T, 1987. The origin and structure of fungal disease resistance genes in plants. Trends in Genetics 3:157-161. Romero-Severson, J., J.Lorzer, C. Brown & M. Murray, 1989. Use of RFLPs for analysis of quantitative trait loci in maize. In: T. Helentjaris & B. Burr (Eds.): Development and application of Molecular markers to problems in plant breeding, 97-102. Cold Spring Harbor, New York: Cold Spring Harbor Laboratory. Rommens, C.M.T., MJ J. van Haaren, A.S. Buchel,J.N.M . Mol, AJ . van Tunen, HJ J. Nijkamp &J . Hille, 1992a. Transactivation of Ds by Ac-transposase fusions in tobacco. Mol. Gen. Genet. 231:433-441. Rommens, C.M.T., G.N. Rudenko, P.P. Dijkwel, MJJ. van Haaren, P.B.F. Ouwerkerk, K.M. Blok, HJJ Nijkamp &J. Hille, 1992b. Characterization of the Ac/Ds behaviour in transgenic tomato plants using plasmid rescue. Plant Molecular Biology, in press. Schots, A., T. Hermsen, S. Schouten, FJ. Gommers & E. Egberts, 1989. Serological differentiation of the potato-cyst nematodes Globoderapallida and G. rostochiensis: II. preparation and characteriza tion of species specific monoclonal antibodies. Hybridoma 8:401^413. 77 Schots, A.,J . de Boer, A. Schouten,J . Roosien,J.F . Zilverentant, H. Pomp, L. Bouwman-Smits, H. Overmars, FJ. Gommers, B. Visser, WJ. Stiekema &J . Bakker, 1992. Plantibodies: a flexible approach to design resistance against pathogens. Netherlands Journal of Plant Pathology 98 sup 2 183:191. Sela-Buurlage, M.B., A.S Ponstein, S.A. Vloemans, L.S. Melchers, PJ.M. van den Elzen & B.J.C. Cornelissen, 1992. Only specific tobacco chitinases and/?-l,3-glucanases exhibit antifungal activ ity, Plant Physiology, submitted for publication. Her, M., T. Brody & M. A. Genizi, 1976. On the power of experimental designs for the detection of linkage between marker loci and quantitative trait loci in crosses between inbred lines. Theoreti cal and Applied Genetics 47:35-39. Stam, P., 1991. Some aspects of QTL mapping. In: Pesek et al. (Eds.): Proceedings 8th Meeting Eucarpia Section Biometrics in Plant Breeding,Jul y 1991, Brno, pp. 23-31. Stam, P., 1992. JoinMap: a computer package for the construction of integrated genetic linkage maps. Submitted. Stirpe, F., L. Barbieri, M.G. Battelli, M. Soria & D.A. Lappi, 1992. Ribosome-inactivating proteins from plants: present status and future prospects. Bio/Technology 10:405-412. van den Ackerveken, A.FJ.M, J.A.L. van Kan & PJ.G.M. de Wit, 1992. Molecular analysis of the avirulence gene avr9o f the fungal tomato pathogen Cladosporiumfalvum fully supports the gene-for- gene hypothesis. The PlantJourna l 2, 359-366. van der Krol, A.R., P.E. Lenting, J. Veenstra, I.M. van der Meer, R.E. Koes, A.G.M. Geräts, J.N.M.Mol &A.R . Stuitje, 1988. An antisense chalcone synthase gene in transgenic plants inhibits flower pigmentation. Nature 333:866-869. van der Krol, A.R., L.A. Mur, M. Beid, J.N.M. Mol & A.R. Stuitje, 1990. Flavonoid genes in Petunia: Addition of a limited number of gene copies may lead to a suppression of gene expres sion. The Plant Cell 2:291-299. van der Meer, I.M., M.E. Stam, A.J. van Tunen, J.N.M. Mol and A.R. Stuitje, 1992. Antisense inhibition of flavonoid biosynthesis in Petunia anthers results in male sterility. The Plant Cell 4:253-262. van der Meiden, A., 1991. De acceptatie van het geheime knutselen, over biotechnologie en de publieke opinie. Congres Proceedings, acceptatie van biotechnologische producten, november 7, N.S.C., Vlaardingen, the Netherlands. van Kan,J.A.L., A.F.J.M. van den Ackerveken & PJ.G.M. de Wit, 1991.Clonin g and characteriza tion of cDNA of avirulence gene aur9o f the fungal pathogen Cladosporiumßilvwn,causa l agent of tomato leaf mould. Molecular Plant Microbe Interactions 4:52-59. van Ooijen,J.W., 1992.Accurac y of mapping quantitative trait loci in autogamous species. Theoret ical and Applied Genetics, in press. van Ooijen, J.W.,J. M Sandbrink, C. Purimahua, R. Vrielink, R. Verkerk, P. Zabel & P. Lindhout, 1992. Mapping Clavibacter michiganensis resistance gene: A piece of cake? In: Yoder et al. (Eds.): Proceedings symposium: The Molecular Biology of Tomato, Davis, California, August 1992. van Tunen, A.J., A.G.M. Geräts &J.N.M.Mol, 1990. Flavonoid gene expression follows the changes in tissue development of two Petunia hybrids homeotic flower mutants. Plant Molecular Biology Reporter 8:50-60. van Tunen, AJ . & G. Angenent, 1991. How general are the models describing floral morphogene sis? Flowering Newsletter 12:34-37. van Wordragen, M.F. & J.J.M. Dons, 1992. Agrobacterium tumefaciens-medi&ted transformation of recalcitrant crops. Plant Molecular Biology Reporter 10:12-36. Verrips, C.T., 1991. Biotechnology to meet the consumer demands for high quality natural food products. Agro-industry hi-tech, November:37-41. Vigers, A.J., S. Wiedemann, W.K. Roberts, M. Legrand, CR SelitrennikofF & B. Fritig, 1992. Thaumatine-like pathogenesis-related proteins are antifungal. Plant Science 83:155-161. Visser, B., T. van der Salm, W. van den Brink & G. Folkers, 1988. Genes from Bacillus thuringiensis entomocidus 60.5 coding for insect-specific toxins. Mol. Gen. Genet. 212:219 - 224. 78 Visser, B., E. Munsterman, A. Stoker & W.G. Dirkse, 1990. A novel Bacillus thuringimsis gene encoding a Spodoptera mgHtf-specific crystal protein.Journa l of Bacteriology 172:6783-6788. Visser, B., D. Bosch & G. Honée, 1992. Domain-function studies of Bacillus tkuringiensis toxins: a genetic approach. In: P. Entwistle, M.J. Entwistle, MJ . Bailey,J . Cory & S. Higgs (Eds.): Bacillus thuringiensis: Its uses and future as a biological insecticide,Joh n Wiley & Sons, England, in press. Ward, E.R., G.B. Payne, M.B. Moyer, S.C. Williams, S.S. Dincher, K.C. Sharkey,J J. Beck, H.T. Taylor, P. Ahl-Goy, F. Meins & J.A. Ryals, 1991. Differential regulation of /M ,3-glucanase messenger RNAs in response to pathogen infection. Plant Physiology 96:390-397. Weller,J.L. , M. Soller & T. Brody, 1989. Linkage analysis of quantitative traits in an interspecific cross of tomato (Lycopersicon esculentum Lycopersicon pimpinellifolium) by means of genetic markers. Genetics 118:329-339. Wessels,J.G.H . &J.H . Sietsma, 1981. Fungal cell walls: a survey. In: W. Tanner & F.A. Loewus (Eds): Encyclopedia of plant physiology, New series, vol 13B: Plant carbohydrates. Springer, Berlin, pp. 352-394. Wilmink, A., B. van der Ven &JJ.M. Dons, 1992. Expression of the gus gene in the monocot tulip after introduction by particle bombardment and Agrobacterium. Plant Cell Reports 11:76-80. Winter, G. & C. Milstein, 1991. Man-made antibodies. Nature 349:293-299. Witsenboer, H.M.A., K.M. Kloosterziel, G. Hateboer, HJ.J . Nijkamp, &J. Hille, 1992. Tomato susceptibility to Altemaria stem canker: parameters involved in host-specific toxin induced leaf necrosis. Plant Science 81:127-134. Woloshuk, CR, EJ.S. Meulenhoff, M. Sela-Buurlage, PJ.M. van den Elzen & B.J.C. Cornelissen, 1991. Pathogen-induced proteins with inhibitory activity toward Phytophthora infestons. The Plant Cell 3:619-628. Young, N.D. & S.D. Tanksley, 1989. Restriction fragment length polymorphism maps and the concept of graphical genotypes. Theoretical and Applied Genetics 77:95-101. 79 Biotechnology in animal breeding, husbandry and animal health Contents 3.1 General introduction to biotechnology in animal breeding, husbandry and animal health 83 3.1.1 Biotechnology in animal production and health in the Netherlands 83 3.1.2 Prevention of virus diseases by vaccination 86 3.1.3 Animal biotechnology and public opinion 89 3.2 Identification and characterization of a major histocompatibility com plex in a teleost fish and its relevance to improving disease resistance 90 3.2.1 Disease control in aquaculture 91 3.2.2 General aspects of the major histocompatibility complex 92 3.2.3 Relevance of the MHC for disease susceptibility 93 3.2.4 Do fish have an MHC? 94 3.2.5 Towards a characterization of the teleostean MHC 95 3.2.6 The MHC of carp: state of the art 96 3.2.7 Carp MHC polymorphism and disease resistance 97 3.3 Polymorphism of the bovine major histocompatibility complex genes and the relevance of this polymorphism for the study of infectious dis eases 99 3.3.1 The current status of BoLA genetics 100 3.3.2 New developments and future research opportunities 102 3.3.3 Conclusions 104 3.4 Gene mapping in farm animals 104 3.4.1 Genetic markers 105 3.4.2 Mapping of markers and genes 108 3.4.3 Reference and resource populations 110 3.4.4 Gene mapping initiatives in farm animals 111 3.4.5 Concluding remarks 112 3.5 Regulation of expression of milk protein genes 112 3.5.1 Major milk proteins in ruminants and rodents 113 3.5.2 Milk protein genes 113 3.5.3 Mammary gland epithelial cells 115 3.5.4 Protein-binding sites in the promoter regions of milk protein genes 116 81 3.5.5 Sequences shown to be essential for the expression of the milk protein genes 117 3.5.6 Expression of transgenes in the mammary gland of transgenic animals: levels of expression of transgenes using milk protein gene regulatory elements 119 3.5.7 Conclusions 119 3.6 The molecular biology of genetic variation in a functional gene and its use in selection strategies of breeding programmes 122 3.6.1 The halothane gene 123 3.6.2 Conclusions 126 3.7 More calves from high merit cows using new reproduction technologies 127 3.7.1 In vitro embryo production 127 3.7.2 /n ww oocyte collection 128 3.7.3 Further research in reproduction technologies 130 3.7.4 Conclusions 131 3.8 Anti-GnRH monoclonal antibodies to prevent boar taint in pork 131 3.8.1 Production and use of anti-GnRH monoclonal antibodies 132 3.8.2 Conclusions 134 3.9 Development of a 'biotech' vaccine against bovine herpes virus type 1 that can also serve as a vaccine vector 134 3.9.1 Vaccinology in perspective 136 3.9.2 Characteristics of bovine herpes virus type 1 infections 137 3.9.3 Designing a BHV1 deletion mutant vaccine 139 3.9.4 Concluding remarks 141 3.10 Safety of recombinant DNA virus vaccines 142 3.10.1 Safety of live genetically engineered non-carrier vaccines 142 3.10.2 Safety of live genetically engineered carrier vaccines 143 3.10.3 Legislation 145 3.10.4 Aim of the research 145 3.10.5 Oudook 146 3.11 Putting a 'bar code' on bacteria? 150 3.11.1 Bacterial phylogeny and evolution 150 3.11.2 Identification and classification of unculturable bacteria 153 3.11.3 The detection of transmission routes of bacteria 154 3.11.4 Prospects for the future in veterinary bacteriology 155 3.12 References 156 82 3.1 General introduction to biotechnology in animal breeding, husbandry and animal health M.C.Horzinek & AJ. van der Zijpp Since the dawn ofagricultura l civilization,th e health and reproduction ofanimal sbre d for food has been dependent on man, who has always tried to reduce the incidence ofdiseas e by providing shelter, preventing genetic aberrations and infections, and giving treatment. In retrospect, heha sbee nwildl ysuccessful : inth e industrialized nationsanima l production has reached an industrial scale, and a dream of one of the Bourbon kings has long come true '... que le dimanche chaque paysant ait sa poule au pot'. However, with increasing population densities, both of man and domesticated animals, the problems have snow balled: genetic defects have accumulated, infection pressure has built up, animal suffering has increased, environmental issueshav e emerged - all due to an ever-increasing appetite for animal protein. Societyha s responded to this challenge. Retreat into idyllic life (ifther e ever was one)i s advocated by some, militant anti-scientific action is taken by others. Neither approach is fertile, both rather distract from the real issue,whic h is to produce animal protein: - in sufficient amounts; - ofpredictabl e quality; - at acceptable cost; - with minimum environmental damage; - while respecting the health and welfare of the animals. Throughout the ages animal production and breeding have been matters of experience and empirism.Whil e much ofthi s isstil ltrue , the genetic roulette ofbreedin g by selection and of vaccine strain attenuation will probably be frowned upon as primitive by future generations. Nopropheti c gifts are required for the statement that biotechnology willpav e the way towards a sustainable animal production, and that it isher e to stay.Althoug h the term presently carries an emotional load- hope for some,horro r for others- iti sexpecte d to become as dispassionate as 'electricity'. However, education ofth e public and ofpoliti cians is essential, a responsibility ofacademic , governmental and industrial research insti tutions. 3.1.1 Biotechnology in animal production and health in the Netherlands In this chapter we will focus on research strategies in modern animal production and health in the Netherlands. Animal breeding, nutrition and health have traditionally been important issuesi n agriculture inth e Netherlands. The majority ofou r agricultural export is derived from animal production. In the first book published by the Dutch Programme Committee on Agricultural Biotechnology (Dekkers et al., 1990), animal biotechnology was discussed in relation to breeding, animal health and food applications, animal feeds and manure treatment. Progress in animal biotechnology has been considerable and has changed the research agenda with new technologies and newgoals . Based on the need for improvement in animal health and welfare, food quality, efficiency and protection of the 83 environment the following topics are receiving research attention: 1. Unravelling the molecular basis ofth e major histocompatibility complex (MHC) and its association with disease resistance. 2. Constructing genetic maps in cattle, pigs and poultry and development of statistical techniques for association with health and production traits. 3. Studies of the organization and regulation of gene expression. 4. In vitro embryo production without hormonal interference. 5. Use of monoclonal antibodies for the control of physiological processes with undesired results. 6. Development of host species-specific live virus vectors for use as vaccines. 7. Methods to assess the safety of DNA recombinant virus vaccines. 8. Methods to monitor bacterial evolution and development of undesirable traits (e.g. antibiotic resistance). The different approaches will be briefly outlined and discussed in more detail in the following sections of this chapter. TheMHC, itsrelevance todisease resistance andvaccine production High stocking densities have led to an increased risk of infection and outbreaks of disease. As medical treatment, especially with antibodies, is only a last resort, preventive measures such as improving genetic constitution and vaccination are strategically based choices. In fish, in the carp species, there is now evidence for the MHC at gene level, and with advanced molecular techniques the MHC gene product will soon be isolated. Future research will focus on associating genetic variation in immune response and disease resis tance with MHC polymorphism. Using gynogenesis many homozygous lines of carp can be produced, thus facilitating these studies. In catde the MHC (BoLA) is now one of the best characterized MHCs after those of man and mouse. Progress has been made in the serological identification of class II molecules, and molecular characterization of class II genes using PCR is now also avail able. Combining the best typing techniques for classes I and II disease association studies could yield better results, compared with the crude data obtainable with class I serology. Another area of interest is the presentation of immunogens by different class I and class II alleles. For vaccine development the role of the MHC in presentation could be crucial. This subject is discussed in Sections 3.2. and 3.3. Gene mapping infarm animals Genetic linkage maps are important tools in the breeding of farm animals. Genetic maps can be used to localize, isolate and characterize the genes that control phenotypic variation (QTL or quantitative trait loci). DNA markers associated with QTL can be valuable in animal breeding, allowing testing at an early age, because of sex limited traits and of traits that are difficult or expensive to measure. Expertise in quantitative genetics and statistics for marker-assisted selection in the Netherlands warrant maximum utilisation of markers. Available technology now includes RFLPs, VNTRs, mini-satellites, RAPDs. International collaboration is a precondition for obtaining linkage and physical maps. The most ad vanced is PIGMAP, an EC programme in which the development of markers, mapping 84 and utilisation of resource and reference populations have been efficiently organized with input from many laboratories. Many QTLs will be produced in the next decade and -provided that suitable statistical tools are made available - application in breeding pro grammes will follow quickly. Detailed information is presented in Section 3.4. The organization and regulation ofgene expression In Sections 3.5. and 3.6. so-called functional genes are discussed. Alteration of milk composition and the detection of the halothane gene associated with stress susceptibility produce effects that reflect on food quality. Understanding the regulation of these genes is difficult, because of the complex physiological processes involved. It has also proved to be difficult to develop cell lines for continuous experimentation. Expertise in this field is the basis for transgenesis, utilising milk protein and flanking genes for exclusive expression in the mammary gland. With advances in the fields ofgen e maps and QTLs research into the regulation of expression will probably be intensified. Reproduction technologiesin cattle In vitro embryo production would be greatly improved if in vitro fertilization, oocyte collec tion, sexing of the embryo or sperm and cloning of the embryo could be carried out efficiently. Since the publication of Dekkers et al. 1990, the ultrasound-guided follicle aspiration technique has replaced hormone treatment for multiple ovulations. The poten tial of this technique is about twenty-fold, although its practical feasibility has still to be established from pregnancy rates. The separation of X- and Y- chromosome-bearing sperm is also being studied in close collaboration with USDA-ARS . Both developments are promising although our knowledge of embryonic cell differentiation is insufficient to make in vitro embryo production feasible in the short term. For further details, the reader is referred to Section 3.7. Monoclonal antibodies and theprevention of boar taint Monoclonal antibodies can be applied for diagnostic purposes, but also to interfere with physiological (e.g. endocrine) processes. The prevention of boar taint in pig carcasses would improve the Dutch market position and contribute to an increase in meat produc tion. Boars are more efficient converters than castrates and therefore also contribute to decreasing the environmental load. Active and passive immunization against GnRH both have potential, but the age at which application must take place influences the decision in favour of passive immunization. Results of this research are given in Section 3.8. Further research will lead to immunization products and protocols, the final usefulness of which will depend on its acceptance by both pig producers and the public. Bovine herpes virus as a vaccinevector The success formula in herpes virus vaccine development has been to delete specified genes, notably those coding for the thymidine kinase (responsible for virulence) and for non-essential glycoproteins; by doing so mutants are obtained that are avirulent, protect against the homologous disease, and may accommodate genes from other viruses -thereby serving as vectors. A spectacular success has been the construction of a Pseudorabies 85 virus/swine fever virus recombinant which protects pigs against both Aujeszky's disease and hog cholera with one preparation. Details have been presented in Dekkers et al. 1990. This formula has now been extended to include bovine herpes virus type I, the cause of infectious bovine rhinotracheitis . Again both goals are being targeted, and bovine respira tory syncytial virus, the cause of acute respiratory disease in calves would be the recombi nation partner. The advantage of having screenable markers provided by deletion in the herpes virus would make a combined vaccination/test-and-removal strategy an attractive possibility of disease control, perhaps even of eradication. The possibilities are discussed in Section 3.9. Safety of recombinant vaccines In an increasingly litigious society which demands risk insurance and quality assurance for every service and product, the safety of recombinant vaccines will be a permanent issue. The biological insight that Gaussian distributions apply to every conceivable parameter -the virulence of a replicating agent included - cannot be expected from farmers and consumers: the public wants a zero-risk guarantee. While this is intrinsically impossible, the demand for an assessment of risks isjustified . The best classical vaccine ever produced for man, the yellow fever 17D strain, has been given to millions of people over several decades and has led to only a few fatalities; only a few - but still fatalities! Had its risk been assessed it would probably have been given the qualification of absolute safety. Risk assessment has become a science of its own, and the fact that evaluations go considerably further than offering 'educated guesses' is shown in Section 3.10. Unculturable bacteria - their evolution and detection What we know about microbiology may be compared to the tip ofth e iceberg: only agents that can be grown in culture, that can be visualized by (electron) microscopy and/or those that cause disease are conspicuous enough to attract the attention of the scientists. Most microbial life exists below the surface of scientific and public awareness — but it represents a gigantic reservoir for evolution. The article in Section 3.11 emphasizes the importance of refining laboratory techniques for the detection and comparison of micro-organisms. While determining the phylogenetic position of bacteria may be conceived as purely academic 'art-pour-1'art' research by some, it has immediate consequences when patho genicity develops. Insight into the genetic basis of antibiotic resistance and diagnostic refinement are the avenues for future progress in bacterial pathogenesis and epidemiology. 3.1.2 Prevention of virus diseases by vaccination Of all the attempts to control viral infections in man and animals, only immunoprophy- laxis by vaccines, induction or the application of interferon and chemotherapy have been effective; in veterinary medicine, however, the last two methods have only been used in exceptional cases and on a small scale. Vaccination on the other hand has been spectacu larly efficacious - it resulted in the worldwide eradication of smallpox, the infamous disfiguring and highly fatal virus disease that haunted mankind for centuries. The success ful completion of the WHO vaccination campaign in 1979 is considered to be the most 86 effective medical measure ever taken: more lives were saved than by all other interventions of the medical profession put together. Also in veterinary medicine vaccination is the most common and popular preventive measure. Data from the small animal field indicate that vaccination of dogs and cats forms the economic basis of many a practice, furnishing approximately 30% of its income. Traditional vaccines Anti-viral vaccines may contain either biologically active, replication competent virus particles (live vaccines) or virions whose nucleic acid has been damaged by chemical or physical means (killed vaccines). In the past, unmodified, virulent field strains were also used for immunization. In order to prevent symptoms of disease occurring at the natural pork d'entrée topic application was used, e.g. in different herpes virus infections. Thus, chickens have been protected against infectious laryngotracheitis by swabbing the virus into the cloacal mucosa, and cats have been inoculated intramuscularly with feline rhi- notracheitis virus. Had these virus preparations been given via the oculo-nasal route then disease would have been the consequence. Another vaccination principle employing the unmodified virus isheterotypi c immuniza tion. Here an antigenically related, avirulent virus is used to protect against its disease- producing sibling. Thus chickens have been immunized against Marek's disease using turkey herpes virus and against fowl pox using pigeon pox virus; dogs can be partially protected (in the presence of maternal distemper antibodies) by inoculating them with the related morbilli virus which causes measles in man. Smallpox vaccination also made use of the heterotypic principle of protection: vaccinia is closely related to the variola virus but causes only mild symptoms in most vaccinées. The origin of vaccinia virus is shrouded in mystery - it may be a product of genetic recombination in the field, a new species derived from the cowpox virus or the variola virus by serial passage or the living representative of a now extinct virus. By far the most live vaccines have been obtained by adapting a field strain of a wild type virus to a heterologous host. Using large inocula and serial passages in cell culture sponta neous mutants are selected which have a multiplication advantage under the conditions chosen. At different passage levels the virus is tested for virulence and immunogenicity, and a passage level is selected for use as seed stock where minimal clinical reactions are accompanied by maximum protection against challenge. Mutants with selectable marker properties, e.g. a temperature sensitive (ts) phenotype have been repeatedly used in modified live vaccine preparations. The technique is intellec tually appealing for respiratory viruses: these replicate at the lower (permissive) tempera ture prevailing in the upper respiratory tract mucosa while being unable to do so at the higher internal body temperature. The problem is the well-known leakiness of ts mutants, i.e. the occurrence of a certain fraction of virus progeny which displays the parental wild phenotype. Modem vaccines The techniques described above do not allow any prediction about the size and site of the changes in the viral genome. Progress in DNA recombinant technology has allowed 'live 87 carrier virus vaccines' to be tailored at veterinary requirements. Again vaccinia virus paved the way: a recombinant pox virus expressing an immunogenic and protective rabies virus protein has been constructed and evaluated in vitro and in animal experiments. The cDNA corresponding to the gene coding for the 524 amino acid glycoprotein G of rabies virus strain ERA was cloned (Anilionis et al., 1981) and inserted into the 180kb double- stranded DNA genome of vaccinia virus strain Copenhagen under control of the 7.5 kDa vaccinia protein promoter in the thymidine kinase gene (Kieny et al., 1984). The safety of the recombinant was tested in more than thirty species including labora tory, domestic and wild mammals as well as birds (Brochier et al., 1989). No excretion of the recombinant was shown, and no horizontal transmission from the vaccinées to contact animals was observed. Protective activity was evaluated by oral administration of the recombinant to foxes, raccoons, skunks, dogs and cats under laboratory conditions. When these trials had proved successful, field experiments were started. Data from Belgium and France are now available showing that the vaccine issafe , potent and stable and that it can be used for rabies control in the field (Pastoret et al, 1988; Desmettre et al., 1990). To meet safety regulations, other live virus vectors with a host spectrum restricted to a few animal species and excluding humans are being investigated; these include not only DNA viruses (e.g. herpes and adeno viruses) but recently also RNA viruses (e.g. alpha viruses). It is predicted that live virus carriers will determine the future of vaccinology, as far as modified live viruses are concerned. Several requirements must be fulfilled before a virus is acceptable as a live vaccine strain. The attenuated virus - should not cause clinical reactions or damage the embryo; - should be able to replicate in the host to provide sufficient antigenic mass for immune stimulation; - should not revert to virulence after a large number of passages; - should be antigenically close to virulent field strains; - should be free from contaminating adventitious viruses, bacteria and fungi; - should not be excreted by the vaccinée, at least not in quantities sufficient to infect other individuals (also from other species) and to cause disease. fulled vaccines Vaccines of another category contain only the protein case of the virion, be it intact or in fragments, or even single polypeptides. These vaccine preparations may include capsid surface antigens or viral envelope components, but also epitopes from internal, normally non-exposed proteins able to induce T cell responses. Inactivated vaccines against rabies, influenza, poliomyelitis, foot-and-mouth disease, Newcastie disease and others have been obtained by the inactivation of infectivity using formaldehyde, beta-propiolactone, acetyl ethylene-imine. The Netherlands have played a central role in the development of an inactivated vaccine against foot-and-mouth disease; Dr. Frenkel working at the Central Veterinary Institute in Amsterdam was the first to grow a virus in cultures of surviving tongue epithelium on a scale large enough for mass production of vaccine. Later, BHK-21 cells in suspension or microcarrier cultures were exploited for vaccine preparation. Three categories of viral sub-unit vaccines can be distinguished. The first makes use of viral antigens spontaneously produced by infected cells (e.g. the gp70 feline leukaemia virus glycoprotein shed by cells ofth e FL74 line) or those occurring in the blood of patients (e.g. the HBsAg in hepatitis B virus carriers). The second class of sub-unit vaccines is obtained for enveloped viruses, by chemical dissociation of the viral membrane with the aid of detergents. Influenza splitvaccine s give lessloca l reactions than inactivated prepara tions of whole virions. Split vaccines are being used e.g. to vaccinate horses against influenza A. A novel possibility for vaccine production employs DNA recombinant techniques. Iso lated viral genes coding for virion surface components are cloned into Escherichia colian d expressed as fusion proteins. This approach was successful with the VP1 protein of foot- and-mouth disease virus, hepatitis B virus surface antigen, the G-protein of rabies virus and the haemagglutinin protein of the influenza virus. Only recently, a recombinant DNA vaccine against feline leukaemia has been licensed which contains the gp70 glycoprotein of FeLV - only in its unglycosylated form - and which has resulted in good protection under field conditions. Concluding remarks The historically older modified live virus vaccines are generally more efficient than those based on inactivated virions. They can also be manufactured at a lower price per dose. Modified live virus vaccines (and the same seems to be true for recombinant DNA live carrier vaccines) have several advantages over inactivated preparations; they - can be applied less frequently; - can be applied to large numbers of vaccinées via a natural porte d'entrée (orally, by aerosol); - are able to stimulate local immunity; - can lead to a displacement of wild virus strains in a population. Live virus vaccination can be regarded as simulating a natural infection which addition ally results in short-term interferon induction. 3.1.3 Animal biotechnology and public opinion The time has come for veterinary scientists to take a firm stand in the debate on biotech nology. Genetic modification (previously, and sometimes still, referred to by the more emotional term 'manipulation') is being practised worldwide in hundreds of laboratories using viruses, bacteria and yeast. As a result, recombinant vaccines have been obtained and used under field conditions with considerable success. When balancing the argument of a hypothetical biological risk against the benefit of controlling rabies in fox populations or leukaemia in domestic cats, public opinion obviously favours the benefits. However, public opinion can quickly change; the advertising campaign launched by animal protec tionists in this country vilifies 'biotechnology' as such and would thereby include the genetic modification of micro-organisms. Safety issues need to be taken extremely seriously. Constant vigilance is indicated - as it is with conventionally produced attenuated vaccines — to minimize the hazards to both 89 man and animals. There is no such thing as an innocuous vaccine, be it live or attenuated, be it obtained by conventional or by biotechnological means. In the short history of vaccinology, however, disasters have occurred only after the use of contaminated vaccine lots (e.g. with the adventitious bovine diarrhoea virus) or of vaccines in host species for which they had not been designed; thus a live distemper vaccine destroyed a colony of the endangered black-footed ferret in the USA. It can be predicted that biotechnology will play a guiding role in the development of future vaccines. The reasons may be epitomized as technical, economical and epidemiol ogical. Let us only consider protective microbial expression products. The main technical advantage in producing recombinant virus vaccine proteins is that the production facilities for growing kilogram amounts of Escherichia coli and Saccharomyces cerevisiae are all in place. The gene whose products carry the desired properties can be selectively inserted into the vector, undesirable domains can be deleted, and predictable products with reproducible efficacy and innocuity can be obtained. The economical aspect is linked to the above considerations. Recombinant DNA products can be produced inexpensively, and mar keted accordingly, after the costs of research and development have been covered. The most important argument comes from the point of view of the epidemiologists: after vaccination with a recombinant product, infection with a wild type field virus can be easily demonstrated; in the infected vaccinée, antibody specificities appear that are absent in the non-infected individual. For example, pigs vaccinated with the deletion mutant Pseudora bies virus will only produce antibodies against the virus gl protein after infection by the wild type field virus and cats vaccinated with the FeLV gp70 (recombinant E. coli) protein will not produce antibodies against feline leukaemia virus proteins, except ofcoars e against the gp70 protein, unless infection by the wild type field virus has occurred. Where feasible, vaccination can be accompanied by a test-and-removal programme, and infection can be potentially eradicated from a population. 3.2 Identification and characterization of a major histocompatibility complex in a teleost fish and its relevance to improving disease resistance E. Egberts, S.H.M. van Erp, G.F. Wiegertjes & R.J.M. Stet In industrialized countries, fish cultivation constitutes a minor part of the total animal production, and most of the attention in aquaculture focuses on those species that are considered a delicacy such as salmon, trout and eel. In less industrially developed coun tries, people depend for a much larger proportion of their protein need on the production and consumption of fish. Under these circumstances, cyprinids prove to be the main species grown in aquacultural systems, resulting in a contribution of 2.5 millions to the total of 3.5 million tons of fresh water fish produced worldwide. Irrespective of the species adapted to large scale cultivation, however, the high stocking densities will lead to an increased risk of infections and disease outbreaks, with a concomitant reduction in produc tion yields. In some Scandinavian countries, for example, the yearly losses in aquaculture 90 from infections with a single pathogen may rise to the equivalence of S100 million in marketable product. 3.2.1 Disease control in aquaculture To date, most of the methods used to counter disease outbreaks in aquaculture deal with the control of the pathogenic load by therapeutic treatment, i.e., eradication of diseased fish stock and the application of antibiotics. Although sometimes successful in terms of controlling a particular disease, this approach results in a substantial loss of animals. In addition, considerable quantities of antibiotics may end up in the environment, or on the consumer's dinner table. For these reasons, increasing attention is being paid to improving disease resistance. Diseases are countered by animals through mechanisms which depend on specific as well as aspecific factors (Fig. 3.2.1). The latter include physical and chemical barriers which prevent intrusion by a pathogen. For example, the skin and other epithelial tissues, and body secretions such as mucus and sweat. Also, animals possess a specialized cell type which aspecifically destroys invading pathogens by phagocytosis, i.e., the macrophage. In addition, they have a specific defence system, which leads to an immunological response. This system is characterized by its capacity to recognize a particular invader and to memorize it, resulting in a more rapid and stronger response to any subsequent infection by the same pathogen. Thus, if we wish to increase disease resistance, we need to know which genetic factors are involved, and how they exert control over the various defence mechanisms of the animal. In addition, information should be obtained on the immunogenic characteristics of the pathogen. The former requires either the identification of the factors proper, or of polymorphic marker systems closely linked to them genetically. The latter requires a thorough understanding of the handling of a pathogen by the immune system. These approaches both merge in the study of the major histocompatibility complex (MHC). environment animal < > <- aspecific defence pathogen body skin (phagocytosis) antigen secretions epithelia specific defence (cellular and humoral immunity) I Fig. 3.2.1. Schematic representation of the animal's consecutive barriers to the intrusion of a potentially health-threatening element from the environment. 91 3.2.2 General aspects of the major histocompatibility complex The initial identification of the MHC in mammals and birds was based on its involvement in tissue transplantation, but it has since become evident that this phenomenon is of little relevance to the biological significance of the MHC, and simply represents an artificial expression of its pivotal role in the immune response proper. If a 'non-self element, also called an antigen, is introduced into the body, a specialized class of cells, the lymphocytes, are activated and differentiate into effector cells that are able to cope with the attack. There are two types of effector cells, plasma cells which originate from B lymphocytes and produce antibodies that are able to inactivate the antigen by binding to it, and cytotoxic T lymphocytes that can destroy any cell of the body in which the antigenic substance may have sought refuge. The production of effector cells is under the control of a distinct subpopulation of lymphocytes, the regulatory T helper lymphocytes. It has recently been established that the latter are activated through the interaction of a specific receptor, the T cell receptor for antigen or TCR, with a fragment of the antigen bound to an MHC molecule on the surface of another cell (Fig. 3.2.2). T-cell TCR Antigen Fragment MHC class ÏÏ Antigen Presenting Cell Antigen uptake and processing Fig. 3.2.2. Diagram of the essential step in the generation of an immune response, the presentation of an antigen fragment complexée! with an MHC molecule on the surface of a cell, to the specific receptor of a T lymphocyte. 92 The MHC encompasses a number ofloc i coding for membrane glycoprotein molecules, which demonstrate an unsurpassed degree ofpolymorphism . Based on their structural and functional characteristics, the MHC molecules can be assigned to one of two categories designated class I and class II. Class I MHC molecules are present in differing densities on all somatic cells. They are involved in the activation of T helper cells which control the differentiation of cytotoxic T lymphocytes, in response to the binding of an intracellularly derived antigenic fragment. Structurally, they are composed of a variant transmembrane glycosylated polypeptide a-chain of 40-45 kDa, complexed with an invariant non-MHC encoded polypeptide of 12-14 kDa designated/^-microglobulin . Class II MHC molecules are expressed mainly on cells of the immune system such as B lymphocytes and certain macrophages. These cells are able to phagocytize and process extracellular material, and to present the resulting antigenic fragments bound in MHC class II molecules to the T helper lymphocytes which regulate the generation of antigen-specific antibody pro duction. Biochemical analyses of MHC class II molecules have shown two variant trans membrane glycosylated polypeptides, an a-chain of 30-33 kDa, and a /J-chain of 28- 30 kDa, respectively. Within a given individual, all the MHC genes present are expressed, whether they are from paternal or maternal origin; a situation which is also known as co-dominance. In addition, the specific combination of alleles on either ofth e homologous chromosomes that incorporate the MHC behaves genetically as a Mendelian factor, only rarely showing redistribution of its components by crossing-over events. Therefore, such a unique constel lation of alleles of the different MHC genes is designated a haplotype. 3.2.3 Relevance of the MHC for disease susceptibility The biological significance of the MHC can thus be defined as a binding element for antigenic fragments, which is a crucial step in the initiation of an immune reaction. Indeed, evidence from crystallographic data indicates the possibility for binding short peptide fragments in a groove formed by the a-helices and antiparallel jff-strands of the two membrane distal extracellular domains of MHC molecules (Fig. 3.2.3). In addition, short peptide fragments have been eluted from MHC molecules isolated from cells after 'feed ing' excess antigen, and were shown to correspond to a linear amino acid sequence ofthat antigen. Considering the limited number of MHC molecular variants an individual can dispose of, and the great variety of antigenic elements it may be exposed to, each MHC molecule is likely to bind a range of antigenic peptide fragments, albeit with different affinities. Following this line of reasoning, individual variants of the MHC molecules then may differ in their capacity to bind a particular fragment. This current view of the crucial role of the MHC in antigen presentation as a limiting step in the initiation of an immune response is sustained by data from a number of animal species that are indicative of an association between the prevalence or paucity of a certain disease, and the presence of specific variants of the MHC molecules. Thus, MHC polymorphism may serve as an efficacious genetic marker to assess the disposition of an individual to contract a particular disease. 93 Fig. 3.2.3. Schematic representation ofth e structure of the extracellular part of the human MHC classI molecule HLA-A2, as determined from X-ray diffraction analysis (from Bjorkman et al., 1987). The two domains closest to the plasma membrane (a3, and/?2 microglobulin) both resemble a typical immunoglobulin domain. The two membrane distal domains (a, and (X2)togethe r form a groove at the top of the molecule towhic h isboun d a linear oligopeptide fragment ofth e antigen is bound. The N-terminal and C-terminal endso fth ea-chai n are designated byNo tan d Ca, and those of the ^2"microgl°bunn by NyS and Cß, respectively. The pairs of black dots indicate the cysteine residues that form disulphide bridges. 3.2.4 Do fish have an MHC? Studies with several representatives of the Teleostei, or bony fish, have resulted in our current belief that their immune system is not principally different from that of higher vertebrates such as mammals. Thus, fish possess of lymphocyte subpopulations function ally equivalent of T and B cells (Secombes et al., 1983;Mille r et al., 1987), and are able to develop a primary as well as a secondary humoral immune response characterized by the production of antigen-specific immunoglobulin molecules (Smith et al., 1966; Rijkers, 1980). Reactivities such as allotransplant rejection (Hildemann, 1970; Rijkers &va n Muis winkel, 1977; Botham et al., 1980) and mixed leucocyte responses (MLR) (Caspi & Av- talion, 1984; Miller et al., 1986) indicate that the integrated immune system offish also incorporates an MHC. Despite a continuous search for its genes or gene products, how ever, evidence of a genuine MHC in this class of animals is still conjectural (reviewed in Stet & Egberts, 1991). Thus, it has so far been impossible to demonstrate (co)segregation patterns of MHC-associated functions, e.g., MLR or acute graft rejection, in the progeny of outbred teleosts (Gloudemans et al., 1986; Kaastrup et al., 1988). With respect to the 94 gene products, none of the presently available serological reagents with established speci ficity for vertebrate MHC molecules have been found to identify fish MHC-like molecules (Kaufman et al., 1990a; 1990b). Also, cross-hybridization of heterologous MHC cDNA (sub)probes to fish genomic DNA has not resulted in identifying MHC-like sequences. Recendy, however, Hashimoto et al. (1990) have successfully employed the polymerase chain reaction (PCR) technique to the identification of a putative carp MHC. They observed a high degree of conservation of amino acid sequence among higher vertebrates in those regions of the MHC molecules that contain the cysteine residues involved in the formation of an intra-chain disulphide loop. With degenerate oligonucleotide primers, the sequence of which was based upon these short stretches of similarity between the MHC molecules of man, mouse, and chicken, they were then able to isolate two genomic fragments from carp DNA which may include the fish homologues of the MHC class la and class 11/?gene s for the following reasons: i) the gene sequences identified demonstrate an exon-intron organization which corresponds to what is known of the MHC genes of eutherian species; ii) some degree of conservation of specific amino acid residues derived from the gene sequences, with those of mammalian and/or avian MHC molecules can be observed, notably for key amino acids such as cysteine and tryptophan; and iii) the predicted tertiary structure of the protein fragments deduced from both the MHC-like gene sequences can be superimposed onto the proposed structure of mammalian MHC molecules, with the conserved amino acid residues of the first and second domains partic ipating in the framework of the antigen-binding groove. 3.2.5 Towards a characterization of the teleostean MHC Although the results from Hashimoto et al. (1990) constitute a major technical break through in the search for a teleost MHC, the question remains whether the gene elements described really represent the teleostean homologues of the eutherian MHC class I and class II genes. In principle, several issues need to be dealt with before this question can be answered. First, since the carp genomic sequences analyzed so far incorporate only part of the information necessary to generate a complete MHC protein, data should be obtained with respect to the transmembrane, cytoplasmic, and untranslated elements for the genes assumed to represent a carp MHC class la and class Uß chain, as well as with respect to the /^-microglobulin and class IIa genes which in mammalian species provide the com plementary protein chains for the MHC class I and class II molecules. Second, it has to be demonstrated for putative carp MHC genes that they are transcribed into proteins which localize on the surface of all somatic cells (class I), or a restricted category thereof (class II). In this respect, our own detection with alloantisera of an allelic cell surface antigen polymorphism in carp that is associated with skin allograft survival (Kaastrup et al., 1989), provides an independent alternative approach to the identification of putative teleost MHC proteins. As a result of the identification of the carp MHC gene products, possibil ities are created to determine their biochemical characteristics by analogy to their euth erian counterparts. Third, and most crucial to establishing a true teleost MHC, evidence should be obtained that the products of the genes function as antigen-presenting molecules leading to T cell activation, as part of the generation of an immune response in carp. 95 3.2.6 The MHC of carp: state of the art Using PCR primers deduced from the genomic sequences described by Hashimoto et al. (1990), we have also been able to obtain fragments of the carp MHC genes (Stet et al., 1993). In addition, we have successfully applied the same technique to the cloning and sequencing of the carp /^-microglobulin gene transcript (Dixon et al., 1993).A t present, a /Igt1 1 cDNA library from carp pronephros is being screened for full-length messenger sequences from the genes representing the MHC class la and class \\ß chains. Such sequences should include the transmembrane, cytoplasmic, and untranslated elements, so as to provide information with respect to the complete MHC protein molecules. Most recently, this approach has resulted in the identification of two full-length MHC class II sequences of carp (Ono et al., 1993). These sequences probably represent two alleles of a single MHC class \\ß gene, Cyca-DAB,bot h of which are transcribed into messenger RNA within the hemopoietic tissues of a single individual. The Cyca-DABlocu s is orthologous to a previously identified DAB locus of the zebrafish (Ono et al., 1992), but differs considera bly from the carp MHC class II locus identified by Hashimoto et al. (1990). In this respect, it is also noteworthy that analyses of the /^-microglobulin cDNA sequence that has been isolated from the carp spleen/pronephros Agtl 1cDN A library demonstrate an extensive degree of homology at the amino acid level with corresponding sequences from mammal ian vertebrates (Fig. 3.2.4), much more than has so far emerged from an analogous examination of, e.g., the carp class la sequence that has been identified by Hashimoto et al. (1990). -20 -10 1 10 20 30 40 I I I I I I I TILAPIA QVYWRHPGEYGKEDVLICHVSNFHPPDITITLLKNGE CARP MRAIITFALFCVLWT-VQGKTSSPKVQVYSHFPGEYGKENTLICHVAGFHPPDITIELLKDGEILPNT RABBIT VQRAPNVQVYSRHPAENGKPNFLNCYVTSGHPPQIDIELMKNGVKIENV MOUSE MARSVTLVFLVLVSLTGLYAIQKTPQIQVYSRHPPENGKPNILNCYVTQFHPPHIEIQLMKNGKKIPKV HUMAN MSRSVALAVLALLSLSGLEAIQRTPKIQVYSRHPAENGKSNFLNCYVSGFHPSDIEVDLLKNGERIEKV Consensus * * ************** ********** 50 60 70 80 90 I I I I I CARP QQTDLAFEKGWQFHLTKSVTFKPERGQNYACSVRH—MNNKNIYSWEPNM RABBIT EQSDLSFNKDWSFYLLVHTEFTPNNKNEYSCRVKHVTLKEPMTVKWDRDY MOUSE EMSDMSFSKDWSFYILAHTEFTPTETDTYACRVKHASMAEPKTVYWDRDM HUMAN EHSDLSFSKDWSFYLLYYTEFTPTEKDEYACRVNHVTLSQPKIVKWDRDM Consensus ***** ** **** * Fig. 3.2.4. Comparison of amino acid sequence in the teleost fish carp /^-microglobulin molecule with the corresponding proteins from three mammalian species (man, mouse, rabbit), and a partial sequence from another teleost (Tikpid). The amino acids, which were deduced from the gene nucleotide sequences, are represented in the single-character code. The numbers above the se quences denote the relative position of an amino acid in the peptide chain starting from the amino-terminal end, taking the rabbit sequence as reference. Asterisks below the sequences indicate positions of complete identity. Note that a gap of two amino acids has been introduced in the carp sequence at positions 85-86 for optimum alignment. 96 From the nucleotide sequences of a specific carp MHC class I or class II cDNA, proper choices can be made for the construction of oligonucleotide primers to amplify any other mRNA sequence encoded for by a particular gene. The technique involves the application of a PCR cDNA made from cytoplasmic RNA preparations of carp of distinct genetic origin, with reverse transcriptase in combination with, e.g., random hexamer primers. Thus, it will be possible to obtain information on the nucleotide sequence variability at the MHC gene level that is likely to result into a polymorphism at the level of its protein product. In addition to our studies of the carp MHC at the gene level, we are also focusing on the isolation of the MHC gene products proper. Since attempts to apply alloantisera to the immunoprecipitation of MHC-like molecules from carp lymphocyte membrane prepara tions have been without success, our current strategy is to make use of partial or complete MHC cDNA sequences that result from screening a cDNA library. The insertion of such sequences in the correct orientation and reading frame into the cloning site of a suitable expression vector is expected to result in a fusion protein product. To this end, we make use of a vector where the protein to be characterized becomes linked through a short X-factor sensitive stretch of amino acids to protein A. Consequently, the resulting fusion protein product can be purified by immunoaffinity on an IgG-Sepharose column, and the original protein is recovered from it by the action of the protease factor Xa. Following this procedure, the protein proper may then be applied to the production of polyclonal xenoantisera with reactivity for public determinants of MHC class I or class II molecules in carp. With such antibodies, it should be possible to immunoprecipitate and isolate the corresponding native MHC cell membrane protein(s). Two-dimensional gelelectro- phoresis of these precipitates will provide information on the number of MHC class I and class II allelic protein molecules present in an individual animal, the number of their polypeptide chains, and their relative molecular weights and charge characteristics. The degree of intraspecies polymorphism displayed by these molecules may then be deter mined from the immunoprecipitate IEF-profiles among animals of distinct genetic con stitution. 3.2.7 Carp MHC polymorphism and disease resistance In analyzing the genetic mechanisms which control a species' susceptibility or resistance to disease, it is of considerable advantage to have genetically defined material available from, e.g., cloning or inbreeding. With carp, conventional inbreeding to obtain homozygous strains, which usually requires 20 generations to become established, will take at least 30 years due to the relatively long generation time. In lower vertebrates, however, it is possible to reduce the level of heterozygosity by gynogenesis, since only the maternal genome present in the ova will contribute to the offspring (Purdom, 1983). To initiate the second meiotic division, ova are fertilized with irradiated sperm. Diploidization of the activated ova is obtained by retention of the second polar body, or prevention of spindle formation in the first cleavage division, by applying a temperature shock. The first method will result in some residual heterozygosity because of cross-overs in the first meiotic division. The second method will give rise to essentially homozygous all-female offspring 97 in one generation. In our laboratory, such ploidy manipulation techniques have provided a rapid alternative approach for generating homozygous lines of carp (Komen et al., 1991). Analyses of restriction digests of genomic DNA from outbred and gynogenetic families of carp with the MHC gene fragments as probes suggest a considerable polymorphism for MHC class I genes, but a more restricted polymorphism for MHC class II genes in this species (Stet et al., 1993). As expected, RFLP patterns among a mitotic gynogenetic off spring proved to be identical, and were shown to include 9 to 12 fragments for the MHC class I probe and 3 to 5 fragments for the MHC class II-like probe, respectively (Fig. 3.2.5). This suggests that the carp MHC consists of multiple class I and class II genes. The levels of polymorphism of these genes, calculated from the number of polymorphic hybridizing fragments in a comparison of six carp strains from different geographical origin, has been estimated at 70% for class I, and 40 to 65% for class II genes. In addition, two related gynogenetic carp families that differed in their reactivity with a set of alloantisera specify ing allelic traits of a single locus with histocompatibility characteristics designated K, were also shown to differ in their MHC class I RFLP by one fragment. Future studies will concentrate on the identification of MHC-generated RFLP patterns in carp that have been characterized with respect to their disease resistance. To this end, the individual variation in antibody production to the hapten-carrier antigen complex DNP-KLH has recently been investigated within a hybrid population of common carp 20.2 kbp * - ^ t -5.2 - 4.3 i 2.0 Fig. 3.2.5. RFLP patterns of 7ä^I-digested DNA from the offspring of gynogenetic (W11.52) , partially inbred (R8), and outbred (WK) carp after hybridization with an MHC class Iprob e (A), and a class IIprob e (B). 98 (Wiegertjes et al., 1993b). The data from these experiments suggest that the DNP-specific humoral immune response in carp is characterized by a very high variability between individuals, allowing for the classification of 7% of the animals as high, and 13% as low responders (Fig. 3.2.6). Reproduction of these carp using gynogenetic techniques should result in homozygous clones with high and low immuno-responsive genotypes, respec tively.Result s from a recently developed bath challenge systemwher e the animals are also exposed to a subtle infection withAeromonas salmonicida spp. nova through a natural route, indicate that genetic factors are involved in the resistance ofcar p to the disease caused by this bacterium (Wiegertjes et al., 1993a). Likewise, a gynogenetic clone was shown to be much more susceptible to an infection with Trypanopksma borreli, a blood flagellate known to induce a parasitaemia-dependent antibody response in carp (Jones & Palmen, 1993), than a group of outbred animals (Wiegertjes et al., unpublished data). MHC haplotyping ofcar p classified in either ofth e experiments described above, should allow for investiga tions ofth e genetic and biological relationships between the MHC, the humoral immune response, and disease resistance in this species ofteleosts . 3.3 Polymorphism of the bovine major histocompatibility complex genes and the relevance of this polymorphism for the study of infectious diseases J.J. van der Poel, ChrJ. Davies, Ph.R. Nilsson &M.A.M . Groenen The numerous associations between the major histocompatibility complex (MHC) and disease susceptibility in man (Tiwari 1985),th e wish to improve disease resistance in cattle and the role of MHC molecules in immune recognition have stimulated researchers to study the bovine MHC. Since the first descriptions of the Bovine Lymphocyte Antigen IM 0.5 • a o 0.0 O 12 21 days post—immunization Fig. 3.2.6. ELISA antibody titre specific for DNP hapten (mean ± standard deviation) of a total group of carp (a), and the high (•) and low (») responders among them, after intramuscular immuni zation with the hapten-carrier complex DNP-KLH at 25°C. 99 (BoLA) complex, the MHC of cattle, knowledge of the genetic organization and polymor phism of the different BoLA complex loci has grown rapidly. In recent years the BoLA complex has been studied intensively using molecular techniques. As a direct result, the BoLA complex is now one of the best characterized MHCs after those of man and mouse. Knowledge of the fine structure of the MHC will allow more precise description of the genetics underlying immune recognition/regulation and hence provide a basis for study ing disease resistance or susceptibility. Functional studies on the consequences of the role of MHC class I and II polymorphism in antigen presentation, thus may open new avenues for vaccination and therapeutic intervention strategies or alternatively provide selection criteria for improvement of disease resistance in cattle. 3.3.1 The current status of BoLA genetics Initially serology and the Mixed Lymphocyte Culture assay were used to define two genetic regions called BoLA-A and BoLA-D, these regions encode the class I and class II molecules, respectively. Over the years serological typing techniques, and class I antigen definitions, have been standardized through international serum and cell exchanges (BoLA Workshops). In the last BoLA workshop 50 class I specificities were defined (Ber- noco et al. 1991). More recently BoLA class II serology has been developed (Davies & Antczak 1991). To date 18 serological class II specificities can be described (Nilsson unpublished data). Biochemistry Other methods of analyzing MHC polymorphism have also been used to study the BoLA complex. Biochemical typing of class I and class II gene products based on the determina tion of their iso-electric points (pi) (Joosten et al. 1988 and 1989) confirmed the extensive polymorphism of both types of molecules. Iso-electric focusing (IEF) patterns for class II are much easier to interpret than class I patterns. Currently 20 distinct BoLA-DRB patterns have been defined and several BoLA-DO_alleles can alsob e distinguished (Joosten personal communication). Although the IEF technique provides useful information on the polymorphism of expressed class I and class II gene products, because of difficulty with reproducibility and interpretation, this technique is best suited as a complement to other typing methods. For class I, serology isclearl y the typing method of choice. Class I serology is well established and is cheap and reliable. However, for class II the story is quite different. Class II typing sera are very difficult to produce and, consequently, very few laboratories are interested in investing in the development of class II specific antisera. Today most laboratories are turning to molecular biological techniques, notably the PCR technique, for the definition of class II polymorphism. Molecular biology Molecular characterization of the BoLA complex was initially based on restriction frag ment length polymorphism (RFLP) studies using human MHC cDNA probes, which cross-hybridize with bovine DNA. These studies showed that the BoLA class II region is 100 highly polymorphic and enabled BoLA class II genotyping based on RFLP patterns (Sigurdardottir et al. 1988). The RFLP studies gave indications of the number of genes in the BoLA complex and suggested that the organization of BoLA complex is similar to the organization of the HLA complex, the MHC of man (Andersson et al. 1988) (Fig. 3.3.1). Subsequent pulsed field electrophoresis studies showed that the overall genetic organiza tion of the BoLA complex is indeed quite similar to the organization of the human MHC (Bensaid et al. 1991). A number of BoLA class II genes was cloned from a genomic library by Groenen et al. (1989). As a result homologous, locus specific, bovine class II probes became available. As expected, these probes proved much better than heterologous MHC probes for RFLP analysis (Sigurdardottir et al. 1991a). Nevertheless, RFLP patterns are still difficult to interpret. The availability of nucleotide sequences for different class II genes, notably DRB, DQB and DQA (Groenen et al. 1990, van der Poel et al. 1990), has enabled the design of locus specific primers and the use of the polymerase chain reaction (PCR) for detailed analysis of class II polymorphism at the nucleotide level. This analysis has now progressed to the point were sequences for the second exon, the exon that codes for the most polymorphic part of the class II molecule, of many BoLA-DRB, -DQA and -DQB alleles are known (Andersson et al. 1991,Sigurdardotti r et al. 1991b, 1992, van der Poel unpublished observations; see Table 3.3.1). The latest advancement has been the development of PCR based typing techniques and the use of restriction enzyme digestion of PCR products (PCR-RFLP, van Eijk et al. 1992) or sequence specific oligos (SSO) (Giphart &Verduij n 1991). In the case of the BoLA-DRB3 locus an intronic hypervariable simple (TG)n repeat can also be used as a marker for specific DRB alleles (Ellegren et al. 1992), interestingly a similar repeat is present in the human DRB gene. The PCR based class II typing techniques are relatively quick and accurate and are much less laborious BoLA class II class III class 1 nn •• nn nnnn nnn n n B A TAP LMP B A B B B A DY DQ DR C4 C2 Bf A B HLA class II class III class I nn n H B nnnn nnnn nnn nn n n n BA BABABBBA AB DP DNA TAP LMP DQ DR C4 Bf C2 TNF BC A Fig. 3.3.1. Genetic map of the MHC region of man and cattle. Class I, II and III indicate regions where different MHC encoded genes are located. Individual genes are denoted by boxes, the black boxes denote the recendy described peptide transporter (TAP) and proteosome (LMP) genes. Al though many more genes have been mapped in man (HLA), the similarity in the genetic organiza tion of the MHC of cattle (BoLA) is apparent. 101 Table 3.3.1. Polymorphism ofbovin e lymphocyte antigen (BoLA) complex genes. 'Number of alleles recognized at the 4th International BoLA workshop (Bernoco et al. 1991). Many additional alleles have been defined by individual laboratories. Locus Method Alleles BoLA-A serology 43' BoLA-B IEF unknown BoLA-DRA IEF 1 RFLP 4 BoLA-DRB IEF 20 RFLP 31 exon 2 sequences 34 PCR-RFLP 30 TG repeat alleles 16 BoLA-DQA RFLP 25 exon 2 sequences 12 BoLA-DQB RFLP 23 exon 2 sequences 18 DYA RFLP 2 DYB RFLP 2 DNA RFLP 2 than RFLP analysis. In the future most laboratories will probably turn to PCR based techniques for class II typing. Compilation Due to the shortcomings of the techniques that have been available, the most accurate way to define the polymorphism of BoLA class II antigens and genes has been to combine several of the aforementioned techniques (Joosten et al. 1990, Bernoco et al. 1991, Davies et al. 1992, Nilsson unpublished data). In the Fifth International BoLA Workshop (Inter- laken, Switzerland; 1Augus t 1992) these different techniques were further evaluated and a successful effort was made to define BoLA class II haplotypes. Currently PCR-RFLP typing in combination with determination of the (TG)n repeat variation in the DRB3 gene provides an excellent means of defining class II haplotypes. Using these methods in con junction with class I serological typing currently provides the best option to investigate the immunogenetic basis of diseases in cattle. 3.3.2 New developments and future research opportunities With the advent ofgoo d class I and class II typing techniques the stage isno w set for proper evaluation of MHC polymorphism in relation to disease susceptibility. To date virtually all disease association studies in cattle have used the class I genes as a marker for the entire MHC (Lewin et al. 1988). However, in man the majority of MHC associated diseases correlate most strongly with class II antigens (Tiwari 1985). One would predict that in catde disease associations with class II antigens will eventually be documented. In addition, it is now known that other genes that are involved in antigen presentation and the regula- 102 tion of immune responses are located within the MHC. The effects ofthes e genes must also be taken into account. Cells process antigenic proteins into short peptides which are then displayed on the cell surface bound to MHC class I or class II proteins. These MHC-peptide complexes can then be seen by the T cell receptors of circulating T lymphocytes. Recognition of MHC- peptide complexes by T lymphocytes is a requirement for the initiation of a T cell medi ated immune response. If one wants to improve disease resistance, basic knowledge about T cell responses and the involvement of specific MHC class I or class II alleles in these responses, i.e. the underlying genetics involved in the immune response, are prerequisites. Studies involving T cell responses in cattle have demonstrated both class I (Goddeeris et al. 1986) and class II (Glass et al. 1990) mediated MHC antigen presentation and T cell restriction. For the development of component or peptide vaccines it is essential to have detailed information about the presentation of specific immunogenic peptides by different class I and class II alleles. There is currendy a lot of interest in the development of component vaccines for domestic animals and, consequendy, several laboratories are now studying MHC mediated presentation of immunogenic peptides in cattle (Glass andjoos- ten personal communication). Recently it has been established that the binding of peptides to freshly synthesized class I molecules is needed for the normal assembly and stability of the molecules. Peptides for MHC class I presentation are produced in the cytoplasm by proteasomes, encoded by the Large Multifunctional Protease (LMP) genes, and are translocated into the endoplasmatic reticulum by 'peptide transporters', encoded by the 'Transporter associated with Antigen Processing' (TAP) genes (for reviews see DeMars & Spies 1992, Monaco 1992, Neefjes & Ploegh 1992). Both the LMP and TAP genes are located in the class II region of the MHC. Furthermore, there are indications that there are also class II specific 'transporter' genes located within the MHC region (DeMars & Spies 1992). Consequendy, although the biochemical pathways and the underlying mechanisms of antigen processing are not yet well established, we need to take another look at the results of MHC disease association studies. The possibility of linkage between a transporter gene phenotype and susceptibility to diabetes, both in humans and mice, was recendy suggested by Faustman et al. (1991). Perhaps other MHC associated diseases will show stronger associations with TAP or LMP alleles than with class I or II alleles. Furthermore, other MHC genes involved in immuno- regulation and the amplification of immune responses, such as the genes encoding tumour necrosis factors (TNF) alpha and beta or complement factors C2, C4 and Bf, may affect disease resistance. It is important to realize that typing for class I and class II polymor phism is merely the basic framework for disease susceptibility studies. Typing for addi tional MHC linked genes like TAP, LMP, C4, C2,Bfmà TjVFmay be at least as valid. One should strive to define entire MHC haplotypes, i.e. combinations of alleles at the different MHC loci, so that the effects of individual genes or subregions can be evaluated for their contribution to specific diseases. In the human population it has been observed that certain MHC haplotypes are present at a much higher frequency than would normally be predicted. These 'protected' MHC haplotypes have been referred to as: supratypes (Dawkins et al. 1983), haplotypes with preferential allelic association or extended haplotypes (Awdeh et al. 1983). It has been 103 suggested that the conservation of such haplotypes is of importance and that one should look at the MHC in terms of chromosomal segments rather than as a complex of separate genes. The standard approach has been to assume that the 'immune response genes' of the class II region are the most important genes in determining disease susceptibility. How ever, this view may be too restricted. A supratype is composed of specific alleles at multiple loci. Accordingly it is important to consider the possible role of all polymorphic genes within the segment. Some disease associated HLA supratypes have been shown to have deletions, insertions or atypical gene duplications. Deletions and insertions in the region between the HLA B and HLA DQloci are particularly common (Tokunaga et al. 1988). Supratypes may also have characteristic levels of gene expression, i.e. quantitative as well as qualitative characteristics. The quantitative differences between haplotypes may pro vide an alternative explanation for some MHC associated disease susceptibilities (Faust- man et al. 1991). In view of the structural similarity between the HLA and BoLA complexes, it seems likely that supratypes exist in cattle. If bovine supratypes exist, some of them may carry insertions or deletions between the class I and class II genes similar to those that have been found in some disease associated HLA supratypes (Tokunaga et al. 1988). These 'defective' bovine supratypes might be associated with disease. Before bovine supratypes can be defined additional BoLA complex genes need to be cloned and mapped. Recendy the bovine homolog of the LMP7 gene (Trowsdale et al. 1990) was cloned from a bovine cDNA library (Davies unpublished observation). Efforts to clone additional genes are continuing. 3.3.3 Conclusions The disease association studies performed in cattle so far are hampered by the fact that the genetic analysis, which could be performed, was rather crude. The developments outiined in the previous sections imply that there isno w good potential for defining and unravelling the genetic basis for MHC linked disease susceptibility in cattle. However, a lot of basic scientific labour still needs to be done before practical applications could be expected. In particular, it will be necessary to do well controlled model studies where emphasis is put both on sophisticated genetics and relevant immunological parameters like pathogen specific T and/or B cell responses. 3.4 Gene mapping in farm animals M.A.M. Groenen, R.P.M.A. Crooymans, D. Ruyter, A.J.A. van Kampen & JJ. van der Poel The detection of a large number of highly polymorphic sites in the genomes of vertebrates has made it possible to develop a considerable amount of DNA markers that can be used to construct linkage maps for these species. Many of such markers have already been developed in man (Nakamura et al., 1987; Weissenbach et al, 1992) and mouse (Love, 104 1990; Cornall, 1991) an d have subsequently been used to construct genetic maps (Weissen bach et al., 1992; Guyer et al. 1992; Gopeland &Jenkins, 1991). Although a genetic linkage map can be an important tool for the development of a physical map, it also has other applications especially in farm animals and plants. A genetic map is a powerful tool in the localization and, in the longer term, the isolation and characterization of genes that control phenotypical variation (quantitative trait loci or QTL). Phenotypical variation is often controlled by a number of different genes (QTL), each having a small effect, resulting in continuous or quantitative genetic variation in a trait between animals. An animal with a genetically high growth rate, for example, might have 'high growth rate' alleles at a number of different QTL throughout the genome. A genetic map will allow variation in economically important traits to be dissected into variation caused by individual QTL. If a QTL for a particular trait is closely linked to a marker, different marker alleles will appear to be associated with different levels of per formance for the trait. This association, which is likely to occur within families, can be detected by statistical techniques such as regression or maximum likelihood. If a complete genetic map is available and sufficient animals are analyzed, any QTL with an appreciable effect on performance can be located between a pair of linked markers (Soller, 1990 and references there in). DNA markers associated with QTL can be of great value in animal breeding for a number of reasons: (1) Animals can be tested directly after birth (or even before birth in some cases) (2) Traits having a phenotypical effect only in female animals can be tested in male animals (3) Typing for traits that are difficult to measure or are very expensive to measure such as disease resistance and fertility. In this paper an overview is given of the main techniques used for gene mapping followed by a brief summary of the present state of the art for the three main livestock species. 3.4.1 Genetic markers A genetic marker is used to follow a part of a chromosome within a family tree. The characteristics of a good marker are, therefore, a known chromosomal location, a high degree of polymorphism (many different alleles all with a reasonable frequency), easily scored in readily available tissue (such as blood) and it should be codominant. Markers based on DNA sequence variation fulfil these criteria. Restrictionfragment length polymorphisms Restriction fragment length polymorphisms (RFLPs) are caused by DNA sequence varia tion at restriction sites and often only detect two alleles. RFLPs, nevertheless can be very informative in crosses between two divergent lines with very different allele frequencies, as is the case in the reference populations used within the PIGMAP project and the chicken projects mentioned below. Often, cloned cDNA sequences are used to detect RFLPs. An advantage of using cDNA probes to detect RFLPs is that they represent functional ex pressed genes that have often been mapped in other species (man, mouse). Therefore, these markers provide a skeleton oflandmar k loci that are useful in comparative mapping. 105 Variable number oftandem repeat Within the genome of most (or all) vertebrates many different DNA elements are found to be repeated and dispersed throughout the whole genome.These elements are referred to as satellite DNA. One class of these elements are the minisatellites, sequences of base pairs 20 to 60 (bp) in length Jeffreys et al, 1985). At a given locus these elements occur as direct repeats and the number of repeats varies between different alleles (Fig. 3.4.1). Hence they are called variable number of tandem repeat (VNTR) loci (Nakamura et al, 1987). Because of the variation in length and, therefore, the large number of different alleles, they provide another type of (very polymorphic) marker based on DNA sequence variation. If a minisatellite sequence itself is used as a probe (probe A in Fig. 3.4.1) in an RFLP analysis a 'fingerprint pattern' is obtained (Fig. 3.4.2). Because of the large number of different polymorphic bands fingerprinting is extremely useful in pedigree testing. However, al though fingerprinting can be used as a set of multiple markers in crosses between two inbred lines (Hubert et al., 1991), in general it is of limited use in mapping studies (Haley, 1991). If the sequences directly flanking the repeat (probes B and C in Fig. 3.4.1) are used as the probe in an RFLP analysis, only one or two bands are detected depending on whether the animal ishomozygou s or heterozygous. Therefore, due to the extensive variation in the number of repeats, probes B and/or C are useful locus-specific polymorphic markers for mapping studies. Microsatellites Microsatellites or simple sequence repeats are sequences that consist of a direct repeat of di, tri or tetra-nucleotides such as (TG)n, (TA)n or (CAC)n where n can vary from eight to more than thirty. Microsatellites are abundant in the genomes of vertebrates (Hamada et al, 1982; Tautz &Rentz , 1984). They are estimated to occur at least once in every 105 bps, allele 1 -^ —» » » » »— » » » allele 2— ö * > >>»»>»> > iwrfîWMWwm allele n—Q >>>>>» e^^zzza B AC Fig. 3.4.1. Schematic representation of VNTR or microsatellite-containing region. Left-right arrow, repeat; vertical arrow, recognition site for a restriction enzyme. The use of the repeat (A)a s a probe in Southern hybridizations will generate a fingerprinting pattern (Fig. 3.4.2). Fragments B and C can be used as locus-specific probes and, for microsatellites, oligonucleotides can be synthe sized homologously to these sequences (B and C) that can be used as specific primers in a PCR typing. 106 $ cr"1 2 3 4 5 6 7 8 9 10 11 12 f* :.f .1 "6 Fig. 3.4.2. Fingerprint of a full-sib family in chicken. Chromosomal DNA was isolated from the blood of chicken, digested with the restriction endonuclease Hae III and hybridized with a DNA fragment of a mouse clone homologous to part of the Drosophila Pergen e (Georges et al., 1987). which means that in the genome of most vertebrates more than 104 and probably as many as 105 microsatellites are present. As with the larger minisatellites, the number of repeats varies between different alleles. Generally, the greater the number ofrepeat s (n)th e greater the number of alleles. One advantage of using microsatellites as markers, is that their total length, including the flanking DNA, is short enough (100-300 bp) to make them amenable to polymerase chain reaction (PCR) amplification. Polymorphism isdetecte d by separating the PCR amplified fragments on high resolution Polyacrylamide gels. This typing method is faster than conventional RFLP analysis, requires only very small amounts of DNA and is suitable for automatization. Other advantages ofmicrosatellite s are: their random distri bution throughout the genome, their relative ease of isolation, and the high percentage that ispolymorphic . Fig. 3.4.3 shows a typical PCR analysis of a microsatellite marker on a small family in chickens. Random amplifiedpolymorphic DNA The random amplified polymorphic DNA (RAPD) method uses short arbitrary primers, usually 10 bp long, one at a time to amplify random genomic fragments by PCR. These PCR products are separated on an agarose gel and the fragments made visible using simple 107 £OP1 23 45 67 8 Allele A - m m Allele B Allele C Fig. 3.4.3. Segregation of a chicken microsatellite in a pedigree. Typing of the chickens for different alleles of the microsatellite marker was done by PCR (R.P.M.A. Crooymans, T. van Kampen,J.J . van der Poel and M.A.M. Groenen, unpublished results). Three alleleswer e detected within this family. The female is heterozygous for alleles A and Ban d the male is homozygous for allele C. As a result, the offspring have either type A/C (chickens 1, 2, 6 and 8)o r B/C (chickens 3, 4, 5, and 7). staining techniques. The result is a number of fragments with different lengths, and the polymorphisms are observed as either the presence or absence of one or more of these fragments. Advantages of this method are: a large number of reactions can be conducted at a time, little input is needed to develop the markers, and the polymorphism is easily detected. A disadvantage of RAPDs istha t they are dominant and the other alleles are not detected. Therefore, this method is probably most useful in crosses between two inbred lines. 3.4.2 Mapping of markers and genes Genetic mapping (linkage analysis) Two genetic loci or markers are said to be linked if they segregate together in pedigrees more often than would occur by chance. In other words, both loci are situated close to one another on the same chromosome. Linkage maps are based on recombination frequencies (range 0-0.5%) between pairs of loci. The recombination frequency is a measure of the distance between the two loci, the smaller the recombination frequency the smaller the distance between the two loci. The distance between two loci is expressed in centiMorgans (cM), where 1 cM represents a recombination frequency of 1%. The recombination frequency can be determined by studying the frequency of genotypes within one or more full sib families (van der Beek & van Arendonk, 1992). The physical length that corre sponds with 1c M is highly dependent on the amount of recombination within the species involved (e.g. in Arabidopsis 1 cM=140 kb and in human 1 cM=l 100 kb). The primary objective in farm animal mapping projects is to produce a genetic map with markers evenly spaced approximately every 20 cM. For each species around 150 marker loci will be needed; however, two or three times this number will probably have to 108 be mapped, as some will be rejected for technical reasons or for being closely linked to more informative markers. Physical mapping In situ hybridization is the most direct way to physically map genes or markers. A cloned DNA fragment is labelled and directly hybridized to metaphase chromosomes. The start ing point for this method is a standard chromosome preparation from cells treated with colchicine (colchicine arrests cell division in the metaphase). After fixation, the chromo somes are stained with Giemsa stain resulting in the characteristic banding patterns from which each individual chromosome can be recognized (G-banding). After hybridization and autoradiography the gene or marker is physically mapped to a specific chromosome by examining the result under the microscope. The most widely used physical mapping method in human genetics, isth e use of somatic cell hybrids. Polyethylene glycol treatment is used to fuse mouse cells with cells from the animal species to be studied. This results in heterokaryons that are unstable and tend to lose chromosomes of the animal under investigation. Eventually, this results in a set of stable cell lines that contain a full set of mouse chromosomes and a few chromosomes of the other animal. The chromosomes present in the cell lines can be identified using G-banding. Subsequently, such a panel of well-characterized cell lines can be used to map genes and markers to specific chromosomes. A third method of assigning genes and markers to specific chromosomes is by FACS (fluorescence activated cell sorter) sorting of chromosomes. Metaphase chromosomes, stained with an appropriate fluorescent dye such as ethidium bromide, are passed through a laser beam. The resulting fluorescence emission of the various chromosomes will be dependent upon the amount of dye bound which in turn is proportional to the DNA content (and hence size) of the chromosome. A series of detectors can monitor and sort the chromosomes which are associated with the correct amount of fluorescence. Better results are obtained if two dyes and two laser beams are utilized. Due to the small amounts of DNA needed, the PCR technique is especially useful for mapping of markers on FACS sorted chromosomes. Alternatively, the technique allows the development of chromo some-specific libraries, and the isolation of chromosome-specific markers. Comparative mapping Although comparative mapping (Nadeau, 1989) is not particularly useful for mapping markers, it can give valuable information on the possible location of certain genes and on candidate genes for mapped QTL. During the evolution and divergence of mammalian species, numerous recombinations and translocations have occurred. These events have lead to different numbers of chromosomes, and the distribution of previously linked genes over different chromosomes, in different species. However, certain genes have remained linked to one another on the same chromosome in different species. For example approx imately 37% of the 1500 cM of the autosomal genome ofmous e islocate d within segments where both syntony and gene order have been conserved compared with the human genome (Nadeau, 1989). Therefore, if it is known which segments of the chromosomes have been conserved between two species, the position of genes mapped in that region in 109 only one of these species has a high probability of also being present in that region in the other species. 3.4.3 Reference and resource populations Several factors need to be considered when choosing families to use in the mapping study, and these differ depending on whether the family will be used to generate a genetic map or to study associations between markers and QTL. For the generation of a genetic map it is important that the parents are heterozygous for the marker loci. Therefore, a family is most informative if the parents differ at as many loci as possible. This argues in favour of a family where parents come from two divergent lines. A disadvantage of using two widely divergent lines is that markers that are polymorphic in this cross might be mono- morphic in the normal breeding population and therefore will be of no further use. However, since microsatellite markers are very polymorphic even within breeds, the former considerations are probably more important for cDNA type RFLP markers than for the other types of markers. One possible disadvantage of using a cross between two widely divergent lines as the resource population for mapping QTL, might be that QTL will be identified that have a considerably effect on the production trait and that have already been fixed in the breed ing population. Of course the importance of this factor ishighl y dependent on the produc tion trait and the population being studied. Furthermore, it is not always economically possible to generate a specific cross between two divergent lines, i.e. it isusuall y impossible to generate a large resource population solely for mapping QTL in cattle. Finally, al though not obligatory, the reference and resource population can sometimes be combined to reduce the number of animals that have to be typed. For association studies the use of a grand-daughter approach (Weiler et al, 1990) can also considerably reduce the number of typings. In this approach, in a breeding population, production traits are measured routinely on a large number of animals and only the sires and grandsires are typed for the markers. Compared with cattle, both pigs and chickens are exceedingly amenable to mapping purposes. Both animals have short generation intervals and many offspring can be ob tained from a single pair of parents. In addition pigs can be karyotyped relatively easily, which is of great importance for physical mapping. For the PIGMAP project (see below) the reference family is part of a much larger resource population (Haley et al, 1990; Archibald, 1990). Here crosses have been made between genetically very different breeds; i.e. between European commercial breeds (Large White) and the Chinese Meishan or wild boars. These breeds are phenotypically very different. The Meishan reaches puberty at half the age at which the Large White does, and produces a litter of about three to four piglets more than the Large White. However, the Meishan has a low growth rate with a fat carcass. The QTL controlling these differences will be segregating and hence can potentially be mapped in an F2 cross. For chicken gene mapping, specific families have also been generated. One family was produced from the widely differing breeds RedJungl e fowl and White Leghorn (Critten den et al., 1992), and another from two more closely-related White Leghorn lines, which 110 differ in their susceptibility to a number of diseases, including salmonellosis. The latter population has already been used to generate a preliminary map (Bumstead & Palyga, 1992). 3.4.4 Gene mapping initiatives in farm animals As mentioned in the introduction, a genetic map is important for the localization of QTL and genes involved in genetic diseases. Therefore, a lot of research isbein g done to obtain genetic maps for the three major farm animals: pigs, cattle and chickens. Recently (Bras- camp et al., 1991), a working group from the Genetic Commission of the European Association for Animal Production, summarized the involvement of at least 127 different laboratories throughout the world in gene mapping in farm animals. Pigs For pig, more than 16 European laboratories are collaborating on the EC funded 'PIG GENE MAPPING PROJECT' to obtain a 20 cM map of this animal by the end of 1993 (Archibald et al, 1990; Haley et al., 1990). Hundreds of polymorphic markers have already been developed (RFLP, VNTR and microsatellites) and, at the moment, a refer ence population isbein g typed for these markers. As part of this project many markers and genes are being physically mapped onto specific chromosomes by insitu hybridization , by using somatic cell hybrids and by FACS sorting of chromosomes. Cattle In Europe, initiatives for a similar project for the generation of a genetic map in cattle, also targeted for financial support from the EC, are in progress. In the USA, Genmark Inc. started to isolate polymorphic DNA markers some years ago and has succeeded in isolating hundreds of polymorphic markers, again primarily of the VNTR and microsatellite types (Georges et al., 1990). Furthermore, comparative mapping and in situ hybridization are being used to generate a physical genetic map for cattle (Fries et al., 1989; Threadgill & Womack, 1991). Of all the genetic maps in farm animals, that for cattle is the most advanced. However, as a result of the PIGMAP project, this situation is changing rapidly. By the end of 1991, there were as many as 301 bovine loci with at least some mapping information. The vast majority of this information has come from the analysis of a somatic cell hybrid panel, developed in the laboratory ofJ . Womack of Texas A&M, USA. Of the 301 loci, 263 are related to coding genes, and 75 have been shown to be polymorphic. In addition, more than 123 other polymorphic loci have been described but have not yet been mapped. Chickens Although collaboration on generating a genetic map for chickens is currendy less well- organized than the efforts for cattle and pigs, several laboratories have started to develop polymorphic markers for chickens. Many markers of the RFLP type (Bumstead & Palyga, 1992), VNTR type (Bruford et al, 1990) microsatellite type (Crooymans, van Kampen, van der Poel & Groenen, unpublished results) and RAPDs (I. Levin & L. Crittenden, 111 personal communication) are already available and several groups have started to produce reference populations for mapping purposes (Bumstead et al., 1991; Crittenden et al., 1992; Brascamp et al, 1991). A crude genetic map consisting of 100 loci (RFLP type), located in 18 linkage groups has already been produced using a backcross between two inbred chicken lines (Bumstead & Palyga, 1992). Several laboratories that have developed markers in chicken, have recently started to map their markers on the reference popula tions produced by Crittenden and by Bumstead. Finally, since the chicken has been used as a model system for a long time, many chicken genes have already been cloned and studied in detail at the molecular level. 3.4.5 Concludin g remarks Given the amount of research being done in the field of gene mapping, it is reasonable to expect that within the next few years genetic maps will be available for the three farm animals described in this paper. Such maps will form the basis of the much more compli cated task of identifying and localizing the specific QTL for important production traits. Once the QTL have been mapped, the markers flanking them can be used to improve estimates of breeding and shorten generation intervals (marker assisted selection). Litter size in pigs for example has proven difficult to improve by traditional selection and is the sort of trait with a low heritability, expressed in only one sex, for which marker-assisted selection may be of value. Furthermore, comparison of the map of a farm animal species with those of man and mouse (comparative mapping) might identify candidate genes for QTL or might indicate genes that could be used as additional markers. The isolation or identification of additional markers more closely linked to the QTL might eventually lead to the isolation and characterization of the QTL itself (Wicking and Williamson, 1991). 3.5 Regulation of expression of milk protein genes M.A.M. Groenen, RJ.M. Dijkhof, EJ.M. Verstege, CR Spira &JJ. van der Poel The potential for milk production in cattle has increased enormously as a result of breed ing strategies and improved feeding. New molecular genetic techniques will enable an increase in genetic change resulting in increased milk production and also, even more important, will make it possible to alter the composition of the milk, and thereby improv ing its quality. To genetically improve milk composition and yield, the typing of different protein variants and knowledge of the regulation of expression of the different milk protein genes are both important. Some of the processing properties of milk are dependent on the composition of the milk. Composition, however, not only depends on the different variants but also on the degree of transcription of the different milk protein genes. Knowledge of the DNA sequence elements and genes involved in the expression ofth e milk protein genes is, therefore, of great importance for the genetic improvement of these traits in breeding programmes. To date the genes encoding all six major milk proteins of cattle have been cloned and (partially) sequenced, and are the object of intensive studies by an increasing 112 number of laboratories. Furthermore, the promoters of several different milk protein genes have been used to direct the expression of transgenes to the mammary gland in transgenic animals. Expression of the milk protein genes however, is under complex multihormonal control and is only found in the highly specialized epithelial cells within the mammary gland. Accordingly, studies on milk protein gene expression have been extremely difficult and although data on factors that interact with the promoter of these genes have accumu lated over the past few years many of the binding sites have not yet been proven to be functional in vivo. In this section an overview of what is known about the organization and regulation of expression of these genes will be given, with particular emphasis on the data that has accumulated on this subject (Bonsing & Mackinlay, 1987; Mercier et al., 1990). 3.5.1 Major milk proteins in ruminants and rodents Although both quantitative and qualitative differences occur in milk of different species, the milk proteins of all mammals can be divided into two classes; the caseins and the whey proteins. The caseins which make up nearly 80% of the proteins in cow's milk, consist of XT-casein and the calcium-sensitive caseins asl, 3.5.2 Milk protein genes The genes coding for the seven major milk proteins described above have all been cloned from a number of different species and several of them have also been sequenced (Mercier et al, 1990; Groenen & van der Poel, 1993 and references therein). The organization of the bovine casein genes is shown in Fig. 3.5.1 and that of the genes coding for the whey protein genes from different species is shown in Fig. 3.5.2. Some of the genes have been sequenced for two or more species. The ^-casein and a-lactalbumin genes for example have been sequenced in five species. Comparison of these sequences shows that the overall organization of these genes has been conserved during the evolution of these species whereas the sequence has diverged extensively. From the sequences of the calcium-sensi tive caseins it is clear that these genes probably evolved from one ancestral gene (Bonsing & Mackinlay, 1987; Yu-Lee et al., 1986). In these genes the sequence coding for the leader peptide and the promoter region up to position -150/-200 are both highly conserved. It has also been shown that the as2 and ß casein genes are more closely related to each other than either of them ist o the asl-casein gene, and that the 113 53 63 333 92 42 42 42 43 32 45 4 422 44 22 72 4 155 385 asl-casei- tfWWWWWVWiYiW^ 17508 bp 44 63 27 21 42 27 27 27 24 45 123 27 27 24 45 120 45 266 as2-casei . tfWWiWiWWWVWb^ 18483 bp 44 63 27 27 24 42 42 322 6-casein IIII^^H 8498 bp 65 62 33 517 173 K-casein 'i'li^^M 13 kb Fig. 3.5.1. Organization ofth e bovine casein genes. Numbers above theexon s indicate the number of base pairs (bp). Thelengt h of the complete gene isindicate d toth e right, thelengt h ofth e introns is not shown. Open boxes, 5' and 3' non-coding regions; hatched box, region coding for the signal peptide; black boxes, regions coding for the protein; kb, kilobase pairs. For references giving the sequences of these genes see Groenen & van der Poel, 1993.Reprinte d with permission from Groenen &va nde r Poel, Livestock Production Science, in press. It has been shown that the four bovine casein genes are clustered on a 200 Kbp fragment on chromosome 6 (Fig. 3.5.3; Threadgill & Womack, 1990; Ferretti et al., 1990) and in mice the a and ß casein genes have been shown to be clustered on chromosome 5 (Geissler et al, 1988). This clustering of the genes supports the hypothesis of common hormonal regulation of the entire complex and the possibility of the presence of a locus control region analogous to that found for the yS-globin genes. 136 140 74 111 105 42 180 ovine ß-lactoglobulin 4662 bp 160 159 76 330 bovine a-lactalbumin 2023 bp rat WAP 2.8 kb Fig. 3.5.2. Organization of whey protein genes from different species. Numbers above the exons indicate thenumbe r ofbas e pairs (bp). Thelengt h ofth e complete gene isindicate d toth eright ,th e length ofth e introns isno t shown. Open boxes, 5' and 3' non-coding regions; hatched box, region coding for thesigna l peptide; black boxes, regions coding forth eprotein ; WAP, whey acidic protein. Reprinted with permission from Groenen &va nde r Poel, Livestock Production Science, in press. 114 ocsl ß as2 K o 50 100 150 200 250 kb Fig. 3.5.3. Schematic representation of the casein locus in cattle. Reprinted with permission from Groenen & van der Poel, Livestock Production Science, in press. 3.5.3 Mammary gland epithelial cells The milk protein genes are specifically expressed during lactation into the epithelial cells of the mammary gland, and are under complex multihormonal control (Fig. 3.5.4; for a review see Vonderhaar & Ziska, 1989). The minimal hormonal requirements for lacto- genesis are increased secretion of prolactin, glucocorticoids, and estradiol-\,lß and de creased secretion of progesterone. Other enzymes such as insulin and growth hormones and several other growth factors also seem to influence lactogenesis. Many of these are probably involved in the development of the mammary gland during gestation. However, there is no single hormone that initiates lactation. Rather a cascade of events occurs in the endocrine system during the third period of gestation which prepares the mammary gland for the secretion of milk. Furthermore, in addition to hormones, interactions between the epithelial cells and the extracelluar matrix play a vital role in the expression of the milk protein genes (for a review see Aggeler et al., 1988). For this reason it has proved to be extremely difficult to develop mammary gland epithelial cell lines that are consistently capable of expressing transfected milk protein genes. Expression of transfected /?-casein gene constructs in primary epithelial cells (Yoshimura & Oka, 1990) and cell lines (Doppler et al.,1989 ; Schmidhauser et al., 1990) peptide hormones (e.g. prolactin . . , . s^/^^ and insulin) steroid hormones ' ' ' (e.g. glucocorticoids and progesterone) ^hormone receptors transcription fac G proteins second messengers phosphorylation Fig. 3.5.4. Multiple hormones and components ofth e extracellular matrix influence the expression of milk protein genes in the epithelial cells of the mammary gland. Reprinted with permission from Groenen & van der Poel, Livestock Production Science, in press. 115 have been described. Unfortunately, these cell lines cannot be induced to express the a and K casein genes. A bovine mammary gland epithelial cell line, MAC-T (Turner & Hung, 1989), capable of expressing the a, ß, and K casein genes has recently been isolated. However, induction of these cells is difficult (unpublished results). Induction of the milk protein genes in these cells requires them to be grown on floating collagen gels in the presence of prolactin, insulin and hydrocortisone. 3.5.4 Protein-binding sites in the promoter regions of milk protein genes Sequence comparisons of the promoter regions of several of the milk protein genes have revealed sequence elements that are conserved in some of these genes (Yu-Lee et al, 1986; Mercier et al., 1990). In the promoter region of all calcium-sensitive caseins sequenced so far, several elements have been highly conserved, some of them have been shown to bind nuclear factors (discussed below). However, these conserved sequences are not found in the other milk protein genes. A 30 bp sequence showing a 70 to 80% sequence similarity was found in both caseins and whey proteins and has been termed 'milk box' (Laird et al., 1988). Studies on DNA protein interactions have been performed with six different milk protein gene promoter regions; mouse WAP (Lubon & Hennighausen, 1987), rat a- lactalbumin (Lubon & Hennighausen, 1988), rat/?-casein (Schmit-Ney et al, 1991), ovine yS-lactoglobulin Watson et al., 1991), mouse /9-casein (Lee & Oka, 1992) and bovine as2-casein (Groenen et al., 1992b). The results of these studies are summarized in Fig. 3.5.5. DNase footprinting experiments showed binding of a nuclear factor to bps -124 to -107 and -104 to -79 of the rat a-lactalbumin gene. Although the protected region is part of the 'milk box', the results presented indicate that the nuclear factor that binds to this region is a general transcription factor with CTF/NFl-like activity. Many sequences within the mouse WAP promoter region have been shown to interact with nuclear factors (Lubon & Hennighausen, 1987); most of which appear to be ubiquitous nuclear factors. However, one mammary gland specific complex was found in gel retardation experiments using a fragment containing nucleotides -88 to -175 from the promoter region of the mouse WAP gene. Unfortunately, it is not clear from the results presented to what se quences this factor binds. Two different mammary gland specific factors, MPBF and MGF, have been shown to bind to the ovine /?-lactoglobulin promoter region and to the promoter region of the calcium-sensitive casein genes respectively (see Fig. 3.5.5). Although sequences with a weak homology to the MPBF site have been described in other milk protein genes, MPBF has only been shown to bind weakly to nucleotides -151 to -138 of the mouse WAP gene. The relevance of the 'related' sequences in the other genes remains doubtful. In addition to the mammary gland specific factors a number of ubiquitous nuclear factors have been shown to bind to several positions within the promoter regions of some of the milk protein genes (Fig. 3.5.5). Two of these factors have been identified as NF1 and Oct-1. In all the calcium-sensitive casein genes the strong Oct-1 binding site at position -50 and the MGF binding site at position -90 have been highly conserved. Finally, binding studies on the mouse /?-casein promoter region (Lee & Oka, 1992) 116 -480 -O—//—£M;K> -250 -200 -150 -100 -50 +1 ^^ TATA-box <^> transcriplion initiation site Mamillary gland-specific factors: H MGF [__] MI'l lF Ubiquitous factors: ^NFl(related) (jj) Oct-1 Q unidentified Negatively acting factors: |;| I'MF ® unidentified Fig. 3.5.5. Summary of DNA-binding sites of different milk protein gene promoter regions. Only those binding sites that have been identified in the original studies (see text for details) are indicated. Therefore additional binding sites may be present in these genes. For example, the MGF-binding site at Position -90 and the strong Oct-1 binding site at Position -50 are both also conserved in the mouse /?-casein gene. Numbers are in base pairs. Reprinted with permission from Groenen & van der Poel, Livestock Production Science, in press. showed, in addition to four not further identified binding sites between nucleotides -545 and +7, the presence of two binding sites for a pregnancy-specific nuclear factor (called PMF) at positions -350 and -8. At both positions the palindromic sequence TGATN5 8 ATCA is found. Mutations in these sequences abolished binding of PMF. This factor may also be involved in repression of the /?-casein gene by progesterone during pregnancy (discussed below). These sequences have been conserved and are located at the same positions in the rat jS-casein gene but are not found in the bovine jS-casein gene or any of the other calcium-sensitive caseins that have been sequenced. 3.5.5 Sequences shown to be essential for the expression of the milk protein genes Calcium-sensitive caseins Because of the absence of good mammary gland epithelial cell lines the information on the sequences, essential for the expression ofth e milk protein genes, israthe r limited. The most detailed information is available for the rat and mouse /?-casein gene, which have been expressed in cultured mammary gland epithelial cells. All other information concerning proven functional sequences comes from studies using transgenic animals. 117 Transfection experiments with fusion constructs of the promoter region of the rat /?-casein gene and the reporter gene chloramphenicol acetyltransferase (CAT), using the mouse cell line HCl 1 (Doppler et al., 1989), showed that the region -285 to +487 was sufficient for expression and induction by prolactin and dexamethasone. Constructs con taining less than 170 bps of the promoter region were not expressed in HCl 1 cells. In agreement with this, two lines of transgenic mice containing the region from -2300 to +490 and from -524 to +490, of the rat jS-casein gene were both found to express specifically the transgene in the mammary gland during lactation and to do so at the same level (Lee et al., 1990). Finally, in primary mouse mammary gland epithelial cells, trans- fected CAT constructs containing nucleotides -545 to +7 of the mouse jS-casein gene were expressed and inducible by hormones (Yoshimura and Oka, 1990). Using primary mouse mammary gland epithelial cells it was further shown that the two PMF binding sites at positions -350 and -8 are probably involved in repression by progesterone of this gene during pregnancy (Lee & Oka, 1992). In this model system, the addition of progesterone usually had an inhibitory effect on the induction of expression of the transfected mouse yö-casein gene by 60%. y5-casein constructs with mutated PMF sites, to which PMF no longer bound, were inhibited by progesterone by only 15%. Using the mouse HCl 1cel l line, it was shown (Schmit-Ney et al., 1991) that the MGF binding site at position -90 is involved in expression of the rat /?-casein gene and that the binding activity of MGF is probably regulated during lactation by means of phosphoryla tion (Schmit-Ney et al., 1992). Because this MGF binding site has been conserved in all calcium-sensitive casein genes, and MGF has been shown to bind at that position in the bovine Wheyproteins All three major whey protein genes have been studied in transgenic animals. Studies with transgenic mice containing bovine a-lactalbumin transgenes with varying parts of the 5' region (Soulier et al., 1992) showed that sequences between nucleotides -477 and -220 are involved in the tissue-specific expression of this gene. A similar result was found for the ovine /Mactoglobulin gene in transgenic mice (Whitelaw et al., 1992). Here the region 118 between nucleotides -406 and -149 of the 5' region of the gene was essential for high-level, tissue-specific expression. Even more interesting results were obtained with transgenic mice containing the gene for the rat whey acidic protein (Dale et al., 1992; Bayna & Rosen, 1992). At the 5' end, again, similar results were found as with the other whey protein genes. The smallest construct tested in transgenic mice contained the 5' region of the rat WAP gene up to position -949. This construct was expressed at a high level and its expression was develop- mentally regulated and tissue-specific. Interestingly, constructs that lacked the 3' end of the WAP gene were either expressed at a very low level or were not expressed at all. It was shown (Dale et al., 1992) that the last exon of the WAP gene, which consists primarily of the 3' untranslated region, contains an element that is involved in copy-number-depend ent, integration-site-independent high level expression of the WAP gene in transgenic mice. 3' Sequences of the WAP gene, between positions +2020 and + 3250 may also contain some sort of repressor element. In constructs containing this region the expression was lower and integration-site-dependent. Finally, another striking observation was that the WAP gene was expressed efficiendy in CID 9 mammary gland epithelial cells in the absence of ECM and hormones (Dale et al., 1992). 3.5.6 Expression of transgenes in the mammary gland of transgenic animals: levels of expression of transgenes using milk protein gene regulatory elements Numerous transgenic animals have been generated that carry transgenes containing the 5' and 3' flanking sequences of one of the milk protein genes. These animals have been produced, not only to study the elements involved in expression of these genes, but also to express and produce biologically interesting proteins in the milk of these transgenic ani mals. Even in the early days of this technique, the prospect of using transgenic livestock for the production of therapeutic proteins as an alternative to expensive cell culture systems was apparent. In fact, this is probably the main reason for the increase in the number of laboratories working on the regulation of expression of these genes. Today several pharmaceutically interesting human proteins have been produced in the milk of transgenic animals including: anti-haemophilic factor IX, ^-antitrypsin, tissue plasminogen activator, lactoferrin, the cystic fibrosis transmembrane conductance regula tor and urokinase (for references see Groenen & van der Poel, 1993). The level of expression in these animals varies by several orders of magnitude depend ing on the construct used, the gene to be expressed, and the site of integration in the transgenic animal. The results obtained with the genes described above and those obtained with heterologous milk protein genes in transgenic animals are summarized in Table 3.5.1. 3.5.7 Conclusions Although several mammary gland specific factors that play a role in expression of some of the milk protein genes have been identified, none of them have been shown to be involved 119 Table 3.5.1. Expression of transgenes in the mammary gland of transgenic animals. Abbreviations: bp, base pairs; kb thousand base pairs; FR, flanking region; WAP, whey acidic protein; CFTR, cystic fibrosis transmembrane conductance regulator; tPA, tissue plasminogen activator; LAtPA, longer acting tissue plasminogen activator; CAT, chloramphenicol acetyl transferase; n.c, not comparable. For references see Groenen & van der Poel, 1993. Gene Animal Regulatory sequences Level of expression Rat /?-casein mice Complete rat /?-casein gene 0.01-1% of the en including 3.5 kb of the 5'-FR dogenous mouse ß- and3.0kbofthe3'-FR casein gene Ha-nu oncogene 2.5kbofthe5'-FRofthe mouse WAP gene Sheep /Mactoglobulin mice Complete sheep /Mactoglob 3-23 g/1 (up to 5 ulin gene including 4 kb of the times as high as in 5'-FR and 7.3 kb of the 3'-FR sheep) Human tPA 2.6 kb of the 5'-FR of the 0.1-0.46 mg/1 mouse WAP gene Bacterial CAT reporter mice -2300 and -524 respectively, gene to +490 of rat /J-casein gene Human factor IX sheep Complete sheep y8-lactoglob- 0.08% of the en ulin gene including 4 kb of 5'- dogenous yS-lactog- FRand 1.6 kb of the 3'-FR lobulin gene Rat WAP Complete rat WAP gene in 1-95% of the en cluding 949 bp of the 5'-FR dogenous mouse and 1.4kbofthe3'-FR WAP gene Human a [-antitrypsin mice 4 kb of the sheep jS-lactoglob- 0.08-7.7 g/1 ulin gene Mouse WAP Pigs Complete mouse WAP gene 1-2 g/1 (50-100% of including 2.6 kb of the 5'-FR the endogenous and 1.6kbofthe3'-FR mouse gene) Human ^-antitrypsin sheep 4 kb ofth e sheep lactoglobulin 0.3-63 g/1 gene Variant of human tPA goats 2.6 kb of the 5'-FR of the 3-6 mg/1 (LAtPA) mouse WAP gene Human lactoferrin cattle 15 kb of the 5'-FR and 6 kb of not yet known the 3'-FR of the bovine (Continuedonnext page) 120 Table 3.5.1. (Continued) Gene Animal Regulatory sequences Level of expression Human urokinase mice half ofth e bovine asl-casein 1-2 g/1 (8-18% gene including 20 kb of the compared to cattle) 5'-FR and 1.5 kb ofth e 3'-FR CFTR mice half ofth e goatyö-casei n gene including 2.8 kb of the 5'-FR and2.5kbofthe3'-FR Bovine (X-lactalbumin mice complete bovine a-lactalbu- 0.25-0.45 g/1 (31- min gene including 750 bp of 56% ofth e en the 5'-FR and 336 bp of the dogenous mouse 3'-FR gene) Goat ce-Iactalbumin mice complete goat a-lactalbumin 1.2-3.7 g/1(150 - gene including 750 bp of the 460% of the en 5'-FR and 0.5 kb ofth e 3'-FR dogenous mouse gene) Guinea-pig Ct-lactalbu- mice complete guinea-pig ce-lactal comparable to that min bumin gene including 1195 in the guinea-pig bp ofth e 5'-FR and 398 bp of the 3'-FR goat/J-casei n complete goaty8-casei n gene 3-24g/ 1 (10-100% including 3 kb of the 5'-FR ofth e endogenous and6kbofthe3'-FR mouse gene) Rat WAP complete rat WAPgen e in 80-500% (ofth e en cluding 949 bp ofth e 5'-FR dogenous mouse and70bpofthe3'-FR gene) in the expression of all or even the majority of these genes. However, in spite of the considerable différences and the large number of different transcription factors involved, the general picture that emerges is that the promoter regions of the milk protein genes are all composed of at least one strong binding site for a mammary gland specific factor around position -100 to -150 flanked by binding sites for ubiquitous transcription factors. In addition, other weaker binding sites for mammary gland specific transcription factors may also be present. As well as these positively acting factors, binding sites for negatively acting factors are probably present to specifically regulate expression of these genes in the mammary gland only during lactation. For all of the milk protein genes studied in trans genic mice, the promoter region up to nucleotides around -400 to -500 is sufficient for developmentally regulated tissue-specific gene expression. When comparing the expression of the different gene-constructs that have been used to direct the expression of heterologous proteins to the milk of transgenic animals, it is 121 apparent that the levels of protein produced vary over several orders of magnitude. There are many different reasons for this fact e.g the site of integration in the genome of the animal, the transgene used (cDNA clones are known to be expressed very poorly) and the regulatory sequences present in or absent from the injected construct. However, in gen eral, it seems that constructs containing the regulatory sequences of the whey proteins are expressed at much higher levels than those containing casein regulatory sequences. In the natural situation the whey protein genes exist as single copy genes whereas the four caseins are clustered on a small 250 kb segment on the chromosome. It is possible that certain specific regulatory elements such as a locus control region or matrix attachment sites are present within or flanking the entire locus but that these elements are missing from the micro-injected constructs. Given the increase in the amount of research, and the availability of cell lines suitable for studies on milk protein gene expression, it is to be expected that many questions addressed in this paper will be answered in the near future and that many genes involved in the regulation of expression of the milk protein genes will be identified and cloned. This information will not only have an impact on the expression of heterologous proteins in the milk of transgenic animals, but will also create new possibilities for improving milk quality by breeding. 3.6 The molecular biology of genetic variation in a functional gene and its use in selection strategies of breeding programmes M.F.W. te Pas &J.H.F. Erkens Breeding strategies for meat-producing animals, such as pigs and cattle, are focused on optimizing growth rate during the growing period and on carcass composition of the animals at slaughter. Breeders select and combine the best males and females to produce the next generation. Quantitative geneticists measure growth rate and carcass composition traits on a large number of animals. It has been assumed that the traits are the result of the combined action of many genes, some having a major influence on the trait while others are of minor importance. Neither the number of genes for each trait nor their identity is known. Molecular biology can elucidate the underlying genetic variation used for selection, thereby giving a better understanding of the nature and extent of genetic variation. Furthermore, using molecular biology to select the animals from which to breed the next generation changes selection criteria from phenotypic towards genotypic. This could ac celerate genetic improvement and shorten the generation interval to a minimum (Meuwis- sen & van Arendonk, 1992). At IVO-DLO we are working with DNA probes of functional genes derived from man or animals. Since mammals show a high degree of homology, probes of different species can be used directly. By using probes for the gene itselfw e hope to concentrate on the most relevant fraction of the total genetic variation (detected polymorphisms). 122 Polymorphisms in genes can be detected in a number of ways: (1) by isolating and sequencing the gene causing phenotypical differences from phenotypically different ani mals, (2)b y restriction fragment length polymorphism (RFLP) analysis, (3)b y PCR analysis of hypervariable repeats near genes, or (4) PCR analysis of mutations. The first method is time-consuming but with the other techniques a considerable number of polymorphisms can be found more quickly. Not all detected polymorphisms may influence the trait directly. Association studies relating polymorphisms and quantitative variation are needed to select the meaningful genetic variation in the gene. The aim of this contribution is to show genetic variation in a functional gene, the halothane gene, and to associate functional traits of the gene with this genetic variation. 3.6.1 The halothane gene Breeding for increased meat percentage and reduced fat percentage in pigs has resulted in stress-susceptible animals, which can die during environmental stress-attacks. The disease iscalle d the porcine stress syndrome (PSS). PSS is associated with a number of quantitative production traits, both positive and negative. Food conversion rate, eye-muscle area, visual conformation scores, lean meat percentage and carcass lean proportion are posi tively influenced by PSS while the syndrome is accompanied by disadvantages in adlibitum daily food consumption, carcass length, litter productivity, post-weaning mortality and meat quality due to an increased incidence of pale, soft, exudative (PSE) meat (Webb & Simpson, 1986, Simpson et al., 1986). Furthermore, elucidation and elimination of the disease improves animal welfare since stress susceptible animals are less viable. Although PSS renders net economic advantages for meat production, the increased incidence of PSE meat and reduced litter productivity is unacceptable for commercial exploitation (Simpson & Webb, 1989). Breeding against the stress-susceptibility trait needs a reliable test to detect affected animals before they can be selected to raise the next generation. IVO-DLO was the first to be involved in the development of the first test to detect stress sensitive animals: the halothane vapour test (Eikelenboom & Minkema, 1974). Homozygous affected animals react strongly to inhaling the anaesthetic vapour, halothane, by displaying leg-stiffness. However, homozygous non-sensitive animals and heterozygous animals are not detected in this test since the disease is recessively hereditable. It was therefore impossible to eliminate the disease allele completely from a breeding population. After a long period of selection for stress-resistance some 1t o 5% of the animals in the breeding population were still carriers of the disease and, when mated, produced stress-susceptible offspring. Analogically, tests based upon allele differences in closely associated genes of the ha lothane gene linkage group on porcine chromosome 6 (van Zeveren et al., 1988a) associ ated more or less to the disease allele, but it was not possible to eliminate all affected animals from the breeding population (van Zeveren et al., 1988b). Only the use of molec ular genetic techniques could provide a test suitable to eliminate the disease-associated allele from breeding populations since this changes selection from phenotypic towards genotypic criteria. 123 Molecular genetic research into this trait became possible once the human ryanodine receptor gene, responsible for malignant hyperthermia (MacLennan et al. 1990) had been isolated (Zorzato et al., 1990). Using this information, genetic variation in the gene can be detected and associated with the disease and the production traits associated with the disease. Here we describe some of the polymorphisms found in the halothane gene using different methods and experiments to associate the polymorphisms with the stress-suscepti bility trait. Genetic variation inporcine DMA withprobes for the halothanegene Using the polymerase chain reaction (PCR) with primers based on information from the human ryanodine receptor gene (Zorzato et al. 1990) we developed a porcine halothane gene probe consisting of nucleotides 5603-5776. Although we detected a Bam HI poly morphism with this probe (te Pas & Erkens, 1992), this polymorphism was not associated with the stress-susceptibility trait (te Pas & Erkens, submitted). Using the information of the halothane gene defect (Fujii et al. 1991) we synthesized another DNA-probe: a 39 base pair oligonucleotide covering the reported mutation in the ryanodine receptor gene associated with the halothane gene defect (Fujii et al. 1991). Using this probe we detected a remarkably high polymorphic RFLP with the restriction enzyme Rsa I (Fig. 3.6.1). We found 12 different genotypes in 33 animals from two landrace breeds with differences in genotype distribution between both these breeds (Table 3.6.1). However, again none of the polymorphic bands appeared to correlate with the stress-susceptibility trait (te Pas & Erkens, in preparation). Thus, although the probe is Halothane genotype NN NN nn rin nn NN NN Nn Nn nn nn 21 kbp - 15 kbp - 14 kbp - 13,5 kbp - Fig. 3.6.1. Southern blotting experiment showing a restriction fragment length polymorphism (RFLP) of the porcine halothane gene. Each lane contains 10ßg of porcine genomic DNA from animals of known halothane genotype (NN, homozygous normal; nn, homozygous PSS-afFected animals;Nn , heterozygous animals:carrier s ofth e disease) digested with Rsa I. The polymorphism consists of 4 bands, indicated on the left (kbp, 1000 base pairs). The probe is a 39 base pair oligonucleotide containing nucleotide 1823-1962 of the porcine halothane (ryanodine receptor) gene. 124 Table 3.6.1. Distribution and frequency ofth e halothane gene Rsa I genotypes in Dutch Landrace (DL) and Belgian Landrace (BL)pi g breeds. Genotype Band length (kb) Breeds DL BL A 21 — 5/18 B 15 - 1/18 C 14 - 1/18 D 13.5 6/15 - E 21+ 1 5 1/15 1/18 F 21+ 1 4 1/15 5/18 G 21+ 13. 5 3/15 1/18 H 15+14 - 1/18 I 15+13.5 2/15 2/18 J 14+13.5 1/15 - K 15+14+13.5 1/15 - L 21+ 15+13. 5 - 1/18 located around the described mutation in the halothane gene (Fujii et al. 1991) and the system is highly polymorphic, analysis showed that we could not associate the polymor phism with the stress-susceptibility trait under these conditions using this method. Possible association with other halothane gene-associated production traits needs to be evaluated. However, we were able to detect the halothane trait by using a different restriction enzyme Hgi AI instead of Rsa I. As described by Fujii et al. 1991), the mutation in affected animals creates a sequence recognized by the restriction enzyme Hgi AI. Using Hgi AI with the same oligonucleotide probe, we were able to detect a simple polymorphism which correlates well with the halothane genotype in a small number of animals in our RFLP- based DNA test (te Pas & Erkens, in preparation). Although the mutation destroys a recognition site for HinP I present in unaffected animals (Fujii et al. (1991) we were unable to find a Southern-blotting based polymorphism with our oligonucleotide probe when using the HinP I restriction enzyme, probably due to methylation of the GCGC HinP I recognition sequence. We are currently extending these experiments to more animals. A simple PCR based test todetect the halothanegenotypes associated with the stress-susceptibility trait Based upon detailed knowledge of the mutation, Fujii et al. (1991) developed a PCR based test to detect the stress-susceptibility trait. The PCR reaction amplifies a 74 base pair genomic DNA piece covering the mutation. The restriction enzyme HinP I detects a sequence disrupted by the mutation. Thus, HinP I digestion of the amplified DNA (Fig. 3.6.2, Fujii et al., 1991) results in either two bands (41 and 33 base pairs, Fig. 3.6.2) indicating homozygous non-affected animals (genotype NN), three bands (74, 41 and 33 base pairs, Fig 3.6.2) indicating heterozygous animals called 'carriers' (genotype Nn) orjus t one band (74 base pairs, Fig. 3.6.2) indicating homozygous affected animals (genotype nn) in which the mutation has disrupted the HinP I recognition site. The opposite results can be obtained using the restriction enzyme Hgi AI (data not shown). 125 Halothane genotype NN nn Nn Nn Nn NnM bp -587 -267 -124 - 51 Fig. 3.6.2. Polymerase chain reaction (PCR) analysis ofth e halothane genotype ofpig s ofth e three genotypes. The figure shows a family of a homozygous halothane negative and a homozygous halothane positive parent with 4heterozygou s piglets.Th e PCR product with 74bas epair s (bp)wa s digested with HinP I to result in a cleavage product containing 41 bp and 33 bp fragments in homozygous normal animals, fragments of 74bp , 41 bp and 33 bp in heterozygous animals and no cleavage in halothane-positive animals (Fujii et al. 1991).Th e lane on the right isa molecular-weight marker (pBR322 digested with Hae III) indicating DNA length in base pairs. This test enables us to detect the genotype of the animals shortly after birth before the breeding period. Thus it is possible to select the homozygous non-affected animals before puberty and to eliminate all carriers from breeding populations. This test is a good example of marker assisted selection (MAS) since selection is changed from phenotypic to genotypic criteria. The PCR based MAS-test is simpler, quicker and less costly to perform than the Southern-blotting based RFLP test. 3.6.2 Conclusions Molecular biological knowledge of functional genes leads to a better understanding of the fundamental processes underlying genetic variation used in breeding. Furthermore, an understanding of genetic variation at the DNA level can be used to select animals for breeding purposes using genotypic selection criteria instead of phenotypic criteria. Be cause the genotype can be evaluated directly after, or even before birth, this enhances genetic improvement and shortens the generation interval. We have described polymorphisms in the porcine halothane gene and investigated the significance of the polymorphisms for breeding. The results suggest that much of the knowledge about genetic variation at the DNA level is not directly applicable to breeding. 126 Acknowledgements The authors would like to thank Dr.J.K . Oldenbroek and Dr.J.A . Lenstra for their critical reading of the manuscript of Section 3.6. 3.7 More calves from high merit cows using new reproduction technologies A. van der Schans Improving the genetic properties of cattle is a continuous process based on the selection of the best bulls and cows within the population as parents for the next generation. Genetic progress in breeding populations is limited by the number of offspring of these selected bulls and cows. In general bulls are not the limiting factor in breeding programmes. From the moment the properties of a bull are detected, based on the production records of his first daughters, more than 200 000 inseminations can be performed annually. Cows, however, will nor mally produce their first calf at the age of about two years followed, at the most, by one calf per year. Superovulation has been used for many years to increase the number of offspring of cows. With superovulation the number of ovulations is increased and, following artificial insemination, several embryos can be recovered and transferred to recipients. With re peated superovulation, an average of 30 embryos, resulting in 15 calves, can be produced per cow per year. More embryos can be obtained with in vitro embryo production. Using this procedure, maturation and fertilization of the oocyte and the early culture of the embryo are per formed in vitro (= IVF). Large numbers of embryos from specific cows can be obtained if IVF is combined with a procedure in which oocytes are collected repeatedly from individ ual, live cows. The development of techniques such as the sexing of embryos or sperm and the cloning of embryos would enable an even greater increase in the number of offspring with the desired genetic properties. 3.7.1 In vitro embryo production The ovaries of a cow contain around 100 000 dormant oocytes. On each day of the oestrous cycle about 30 to 40 of these oocytes are present in follicles with a diameter of two millimetre or more (Rajakoski, 1960). These follicles result from a continuous process of follicular development and selection which proceed on each day of the oestrous cycle. Due to the selection process most follicles degenerate (= atresia) and so most oocytes are lost. This results in only one follicle developing to full maturity and ovulation during one oestrous cycle. When oocytes are collected before they are lost through follicular atresia, the progeny of a single cow could be greatly increased. Therefore, many groups worldwide are currently performing research programme to improve IVF. 127 In vitrofertilization Immature oocytes can be collected by the aspiration of follicles of two mm or more, present on ovaries collected from cows after slaughter. Under suitable conditions the oocytes start to mature spontaneously in vitro, immediately after release from the follicular environment. In general, only the best oocytes are used, i.e. those surrounded by a compact, multilayered cumulus and with an even cytoplasm. About 70% of the mature oocytes are fertilized within a few hours after addition of sperm. Fertilization isfollowe d by a culture period of five to seven days. During this period about 30% ofth e oocytes develop to an embryo with 50 to 100 cells (morula to blastula stage). Such embryos can develop into normal calves after transfer to recipients, as proven by the birth of several hundreds of these calves all over the world. Ovaries obtained from the slaughterhouse are the most appropriate source of oocytes for research purposes, because many oocytes can be obtained. Ovaries collected after slaughter from beef cows can also be used for the in vitro production of good quality beef embryos at a low cost. By transferring these beef embryos to dairy cows with the lowest milk production records (about 30% of the herd), farmers can increase the value of the calves born from these low producing cows by several hundreds of guilders, without loosing any calves when replacing culled dairy cows. 3.7.2 In vivo oocyte collection For breeding purposes it is important to obtain more offspring from the best cows within the population. This would be possible if large numbers of oocytes could be repeatedly collected from living cows, without any detrimental effects. Accordingly, in vivo oocyte collection procedures have been developed. In vivo oocyte collection technologies can be divided into laparoscopic and ultrasound-guided techniques. Laparoscopicfollicle aspiration Lambert et al. (1983) developed an oocyte collection technique using a laparoscope to make the ovaries visible. To grasp the ovaries, the laparoscope and forceps are inserted through the flank of the cow. Follicles can be aspirated by a puncture needle attached to the laparoscope. This method is in fact a surgical operation, with all the risks entailed. It is also laborious and only follicles larger than about five millimetre can be aspirated. While the ovaries of a cow contain less than 10 of such follicles (Pierson & Ginther, 1987a,b), follicular development must be stimulated using gonadotrophins to increase the number of follicles that are larger than five millimetre. Such a treatment interferes with normal follicular development, so both treatment and oocyte collection can only be performed with intervals of several weeks. Because of its disadvantages, the laparoscopic collection technique has never become very popular. Ultrasound-guidedfollicle aspiration Stimulated by the results of the laparoscopic aspiration technique, ultrasound-guided follicle aspiration was developed. The ultrasonography used to make the ovaries visible is based on high frequency ultrasound, which ispartl y reflected by the tissue under investiga- 128 tion. The reflection of ultrasound is proportionally related to the echo-density of the tissues. By processing the reflected ultrasound, which is transmitted and received by the transducer, an echoscope can produce a two-dimensional image of organs and tissues. Originally, ovaries were made visible and follicles aspirated through the pelvic wall (Callesen, 1987). Because of the considerable distance between the transducer and the ovaries and the variety oftissue s between the two, again only follicles larger than about five millimetre could be made visible using this procedure. This technique has therefore only a slight advantage over the laparoscopic method. The transducer, however, can also be inserted into the vagina of the cow. A hand in the rectum of the cow can position the ovary against the vaginal wall, opposite the transducer. When a follicle is being inspected on the ovary in line with the puncture needle attached to the transducer, the needle can be pushed through the vaginal wall into the follicle. Subsequendy, the follicular fluid containing the oocyte can be collected. The greatest advantage ofth e transvaginal procedure istha t the ovaries can be approached very closely. Because of the short distance, ultrasound of 7.5 Mhz with a high resolution, but also a low penetration, can be used. This allows the aspiration of follicles with a diameter as small as two millimetre. The first transvaginal follicle aspiration technique was developed by a research group in the Netherlands, at the Department of Herd Health and Reproduction of the Veterinary Faculty at Utrecht University (Pieterse et al., 1988). The collection rate was 2.6 oocytes per cow per week. Assuming that this number can be collected over a long period of time and that about 30% of the collected oocytes develop in vitro to the morula/blastula stage, the same number of embryos can be produced using the transvaginal follicle aspiration method as with superovulation. At IVO-DLO another system for transvaginal follicle aspiration with a higher resolution has been developed. Using this system, about 25 follicles of two millimetre or larger can be inspected and aspirated per cow per aspiration session. Around 9.5 to 10 oocytes can be collected per cow during one session. There were no significant differences in the number of oocytes collected per session if the aspiration sessions were performed once or twice a week. Therefore, with two sessions a week 19 oocytes can be collected per cow per week (van der Schans et al., 1991). Later research showed that the in vivo collected oocytes can develop very well in vitro, resulting in the development of almost 14 embryos per cow per week (Table 3.7.1). These results can be obtained without any hormonal treatment of the cows involved (van der Schans et al., 1992). Based on previously mentioned assumptions, these numbers allow the production of up to 20 times as many embryos compared with superovulation. The first pregnancies after transfer of embryos obtained by transvaginal follicle aspira tion and IVF were achieved in 1989 (Kruip, pers. comm.). In November 1991, the first calf from a frozen/thawed embryo, obtained by transvaginal follicle aspiration and IVF, was born in Lelystad at the IVO-DLO experimental farm, 't Gen. Experiments have shown that many embryos can be produced by transvaginal follicle aspiration in combination with IVF. The potential of ultrasound-guided transvaginal follicle aspiration has resulted in tremendous interest in this technique from all over the world. For a reliable evaluation ofth e viability ofembryo s obtained by transvaginal follicle 129 Table 3.7.1. Results of repeated ultrasound-guided transvaginal aspiration ofa follicle and devel opment invitro ofcollecte d oocytes. Oocytes were collected 4 times from each cow. s, standard deviation;ab, no significant difference at F<0.05 for values with the same letter. Cow Average number of oocytes Cleavage (%) Developing oocytes on day 7; per collection (N±s) % of total number collected 1 16.0±5.2a'b 67.2 36.0 2 26.5±7.4a 74.5 43.4 3 15.0±5.2a'b 70.0 43.3 4 9.3±3.5b 67.6 40.5 5 12.0±3.3b 79.1 54.2 Average 15.8±7.8 72.1 43.2 aspiration in combination with IVF, pregnancy rates after transfer of these embryos, either fresh or frozen/thawed, need to be established. Once good pregnancy rates have been obtained, the method can then be made available for practical application. 3.7.3 Further research in reproduction technologies Various research groups are investigating other reproductive techniques, such as sexing and cloning. Sexing can be used to increase efficiency in the production of offspring with the sex of choice, i.e. females for milk production and males for beef production. With in vitro sexing ofpre-implantatio n embryos, which is already being used in combination with superovulation, only embryos with the desired sex will be transferred. At present embryo sexing is only possible by detecting male embryos using Y-chromosome specific DNA- probes (McEnvoy, 1992). This invasive method gives a reliable result, but is laborious, decreases embryo viability and is therefore not generally implemented. Sexing sperm ismor e advantageous, because it can double the number ofoffsprin g with the desired sex by directing embryo development with separated sperm to the sex of choice. So far, only a flow-cytometric method detecting the difference in the amount of DNA between X and Y chromosome-bearing sperm is reliable Johnson, 1991). The main problems with this method are the low number of sperm cells that can be separated and the reduced viability of the sorted spermatozoa. However, using this technique as a reference, further research can be directed towards developing other ways of sexing larger numbers of spermatozoa without loss of viability. As part of its strategic research pro gramme, IVO-DLO is currently investigating differences in membrane-bound antigens between X and Y chromosome-bearing spermatozoa (Hendriksen et al, 1992). This research, which is being performed in cooperation with BARC (Beltsville, MA, USA), ultimately may provide an improved method of sperm-sexing. Another development in reproduction technology for catde breeding which has re ceived considerable interest iscloning . Enucleated oocytes fused with cells from genetically superior embryos will produce large numbers of identical, genetically valuable embryos. By cloning, genetic properties that have so far only been available to the top of the 130 breeding papulation, can be spread over the entire population. For dairy cows this will increase the net income per cow per year by about 300,- Dutch guilders (Meuwissen, pers. comm.). If stable lines of pluripotent embryonic stem cells can be obtained for fusion, almost unlimited numbers of identical embryos can be produced. Using methods to char acterize or even modify embryonic stem cells, it should be possible to produce only high quality clones with specific characteristics. Notwithstanding the enormous opportunities offered by embryo cloning, no successful method is yet available. Most of the procedures tried so far have an empirical basis. To make a breakthrough, systematic investigations of the fundamental processes involved in cloning are essential. Therefore, IVO-DLO is preparing a fundamental research programme which will investigate the fundamental aspects of the differentiation of embryonic cells with the ultimate goal of generating ESC. 3.7.4 Conclusions The new developments in reproduction technology can help speed-up genetic progress. During the last decade, Dutch breeding organizations have exploited the possibilities of artificial insemination and embryo transfer to take leading positions in (dairy) cattle breed ing. Using the possibilities offered by the new reproduction technologies, genetic progress can be accelerated. In the present climate of high agricultural production on the one hand and growing consumer awareness and public demands on animal welfare on the other, characteristics other than production properties are of interest. The challenge of the new technologies is to use them to improve the health and welfare of the cattle population besides increasing production parameters. Therefore, IVO-DLO ispreparin g a multidisci- plinary (ethology, genetics and reproductive physiology) research programme to: - evaluate possible deleterious side-effects of new reproduction technologies; - to develop criteria for breeding programme that will help improve health, welfare and reproduction. Systematic studies to develop new reproduction technologies, in combination with a suitable evaluation of the technologies and adaptations of breeding programmes, are essential steps before the use of these new developments can be justified. 3.8 Anti-GnRH monoclonal antibodies to prevent boar taint in pork T. van der Lende, L. Kruijt & M. Tieman Intact boars are more efficient and produce leaner carcasses than castrated boars (bar rows). The fattening of boars is therefore economically advantageous. One drawback, however, is that pork from boars frequently emits an unpleasant odour during heating, to the disgust of most consumers. This odour, known as boar taint, is usually a consequence of the presence of androstenone (5 131 Androstenone is produced by the testes and is used as a signalling sex pheromone to induce a mating response in sows (Reed et al., 1974). The steroidogenic activity of the testes is stimulated by a pituitary hormone. This hormone, the luteinizing hormone (LH), is in turn produced and secreted under the influence of the gonadotrophin-releasing hormone (GnRH) from the hypothalamus. Skatole is a derivative of tryptophan and is resorbed from the intestines under the influence of testosterone, another important testicu lar steroid. The function of skatole is unknown. The testicular origin of androstenone and the indirect role of the testes in the presence of skatole, explains why pork from barrows and sows does not have this undesirable taint. To prevent boar taint all male piglets destined for fattening - approximately 10 million per year in the Netherlands - are castrated early in life. It has long been recognized that an alternative to castration to prevent the boar taint problem is essential, for reasons of both animal welfare and economics. Accordingly, several reasonably successful attempts have been made to prevent the development of boar taint by means of active immunization against one of the hormones involved, e.g. androstenone (Shenoy et al., 1982; Williamson & Patterson, 1982; Daniel et al., 1984; Williamson et al., 1985), LH (Falvo et al., 1986) or GnRH (Falvo et al., 1986; Bonneau et al., 1987; Awoniyi et al., 1988; Hagen et al, 1988). When immunizing against androstenone, skatole accumulation isb y definition unaffected. When immunizing against either LH or GnRH, the steroidogenic activity of the testes is inhibited, thus inhibiting the production of androstenone and the resorption of skatole. The best results so far have been achieved with active immunization against GnRH, but this method has some shortcomings. First, the immunization procedure must be started early in life. Thus the pig producer hardly benefits from producing intact boars: the level of the anabolic male steroids, responsible for increased efficiency of boars compared with barrows and sows, is reduced from an early stage of growth onwards. Second, the immune response elicited in the boars is rather variable, so the efficacy of the immunization is not 100% in all boars. Taking into account the shortcomings of active immunization against GnRH and considering the potentials ofhybridom a technology to produce highly specific monoclonal antibodies against GnRH, a research project has been started to investigate the possibilities of preventing boar taint by passive immunization with a single injection of anti-GnRH monoclonal antibodies a few weeks before slaughter. Since the majority of male pigs are still boar taint negative at this stage, they will remain negative. In boars which are boar taint positive, inhibition of the steroidogenic activity of the testes will allow the already accumulated androstenone and skatole to be released from body tissues (mainly fat) and cleared from the circulation. The results obtained so far in this research project will be described and briefly discussed in the remaining part of this paper. 3.8.1 Production and use of anti-GnRH monoclonal antibodies GnRH is a phylogenetically well-conserved decapeptide. For the production of anti- GnRH producing hybridomas, Balb/c mice were immunized with GnRH, conjugated to Keyhole Limpet Hemocyanin with either benzidine or carbodiimide. Conjugation is nec- 132 essary because an attempt is made to induce an auto-immune response in these mice. The hybridomas were produced as described by Booman et al. (1984). Anti-GnRH antibody- secreting clones were selected using amongst others GnRH, GnRH-Bovine Serum Albu min (GnRH-BSA), GnRH-Bovine Thyroglobulin (GnRH-BTG), BSA and BTG Enzyme Immuno Assays (EIAs) and a 125I-GnRH Radio Immuno Assay (RIA). Of a total of 43 initially selected clones, four (32 IV, 34 VI, 35 V and 35 XIX) were selected for further use after thorough screening. These clones all produced antibodies of the IgG, isotype. After producing a stock of monoclonal antibodies in a serum-free medium, the bio- effectivity was first studied in mice. Briefly, mice were injected intraperitoneally on the day of di-oestrus with one of the four monoclonal antibodies or with a cocktail of two or all four. Control mice were given the same amount of an irrelevant IgG which had been checked for cross-reactivity with GnRH. These mice were mated during subsequent oestrus and 18±1 days later pregnancy was checked. The monoclonal antibodies which effectively neutralized GnRH prevented the occurrence of a pre-ovulatory LH surge, and thus ovulation and subsequent pregnancy. From the results it was clear that only 35 V had a reasonable bio-effectivity (Fig. 3.8.1). Although ultimately the monoclonal antibodies are to be used to suppress LH secretion in boars, we decided to do the initial bio-effectivity studies in gilts because the LH secretion pattern is more effectively described for gilts than for boars. Due to the results obtained in mice, the studies with gilts were restricted to 35 V. All gilts used showed regular oestrous cycles and were injected with monoclonal antibodies on Day 10+1 after the last oestrus. In the first experiment, gilts were given either 10 or 40 mg 35 V by intravenous injection. Control gilts were given the same amount of the irrelevant IgG which had also been used in the experiments with mice. All gilts had been catheterized (van de Wiel & Eikelenboom, 1977) at least seven days before injection of monoclonal antibodies to enable frequent blood sampling. The LH content of the collected samples was measured with the porcine LH RIA which has been described by van de Wiel et al. (1981). The experiment showed that, for the doses tested, the monoclonal antibody had no effect 100 u (0 >k. . o c ra c O) V Control 75pg 150pg 300|jg Experimental group Fig. 3.8.1. Effect ofa single intraperitoneal injection ofanti-GnR H monoclonal antibody 35V on pregnancy rate in mice. The mice were injected on the day ofdioestru s and were mated during the subsequent oestrus. The control mice were injected with an irrelevant monoclonal antibody. 133 whatsoever on LH secretion. Analysis of the collected blood samples showed that the injected monoclonal antibodies were present in the blood circulation for at least three weeks, which is significant for the ultimate aim of the research project. Since the doses used in the experiment with gilts, expressed as the amount of IgG per kg of body weight, were much lower than in the experiments with mice, two gilts were given 600 and 1100 mg 35 V, respectively. The LH profile of the gilt which received 600 mg showed only minor differences in comparison to control gilts (Fig. 3.8.2). In contrast, in the gilt which received the highest dose (1100 mg), the amplitude of the LH pulses was considerably reduced for at least the first few days after administration and the LH level remained reduced until the end of the blood sampling period 20 days after administration (Fig. 3.8.2). Furthermore, the pre-ovulatory LH peak was not observed in this gilt and her ovaries appeared to be completely inactive when slaughtered 21 days after the anti-GnRH injection. These results clearly indicate that in principle it ispossibl e to inhibit the secretion of LH by passive immunization, but that with the monoclonal antibodies available the necessary dose is too high to be of commercial interest. From an affinity study, it became clear that the four selected monoclonal antibodies all have a low affinity for GnRH (Ka in a RIA between 9.38 x 106 and 1.95x 108 M"1). The highest affinity was found for 35 V, indicating the significance of this parameter for bio-effectivity. 3.8.2 Conclusions To achieve the original goal, hybridomas which produce high-affinity monoclonal anti bodies against GnRH should be made available. At present we are in the process of producing new hybridomas. At the immunization level as well as at the screening level of the produced anti-GnRH monoclonal antibody secreting clones, all possible measures are taken to enhance the chance of producing and selecting hybridomas which secrete high- affinity monoclonal antibodies. The present availability in our laboratory of primary porcine pituitary cell cultures, which respond with distinctive LH productions to stimula tion by GnRH, allows us to determine the binding capacity of the produced anti-GnRH monoclonal antibodies in competition with GnRH receptors in vitro before commencing costly in vivo experiments. 3.9 Development of a 'biotech' vaccine against bovine herpes virus type 1tha t can also serve as a vaccine vector J.T. van Oirschot, A.L.J. Gielkens, F.A.M. Rijsewijk, F.A.C. van Engelenburg, MJ. Kaashoek & B. van den Burg Vaccination is one of man's most significant inventions. It has led to the global eradication of smallpox and has greatly diminished the incidence of devastating human diseases such as poliomyelitis, measles, yellow fever, and rabies in the developed world. 134 Gill 1 (control) * * 7 0 11 13 15 17 19 21 23 9.00 11.30 14.00 16.30 9.00 11.30 Days after onset of the experiment Day 7 (6.30-17.30) &Da y 9 (9.00-15.15) Gilt 2 (600 mg) 7 9 11 13 IS 17 19 21 23 25 27 9.00 11.30 14.00 16.30 9.00 11.30 Days after onset of the experiment Day 9 (6.30-17.30) & Day 11 (9.00-15.15) Gilt 3 (1100 mg) 3 I " 3 oi I ' X 2 - L X 2 - I —I I _i .1 n i -ifll jI I .I | JI I I •HUUtHMHIHim7 B a 11 13 15 17 1Q 21 23 25 27 9.00 11.30 14.00 16.30 0.00 11.30 14.00 Days after onset of the experiment Day 8 (6 30-17.30) & Oay 10 (9.00-15.15) Fig. 3.8.2. Effect of a single intravenous injection of the anti-GnRH monoclonal antibody 35 V (indicated by the arrow) on the luteinizing hormone (LH) profiles of gilts. The left panels show the LH concentrations during the entire experimental period, based on 4 blood samples per day. The right panels show the pulsatile LH secretion on the day of injection and on the second day after that. Blood samples were collected every 15 minutes. The control gilt was injected with an irrelevant monoclonal antibody. 135 Vaccines are also indispensable for economic livestock production. In the Western world, highly contagious diseases such as foot-and-mouth disease and classical swine fever have been brought under control by vaccination and eradication programmes. The sys tematic annual vaccination against foot-and-mouth disease in several countries of the European Community has decreased the prevalence to such a low level that the EC was able to adopt a non-vaccination policy in 1991. Vaccines have mainly been developed against viral diseases, but the increasing resis tance of bacteria to antibiotics and the problem of antibiotic residues in animal products has reactivated interest in vaccines against bacterial diseases. 3.9.1 Vaccinology in perspective Edward Jenner and Louis Pasteur (Fig. 3.9.1) are the fathers of vaccinology. Jenner performed the first scientific attempt to protect individuals against smallpox by deliberately inoculating them with material derived from a pustular lesion from a human affected by cowpox. He demonstrated that they were indeed resistant to smallpox by challenge inocu lation. Jenner named the material 'vaccine', from the Latin word 'vacca' for 'cow', and published his spectacular findings in 1798. Thejennerian approach to prevent smallpox proved to be so effective, that as a re sult, the World Health Organisation could officially announce the global eradication of the disease on May 8, 1980. The next major advance in vac cinology was Pasteur's work on attenu ation of the bacterium Pasteurella mul- tocida which causes fowl cholera. Pas teur observed that a culture of this or ganism which was accidently left ex posed to air during his absence, did not induce disease in chickens, but pro vided immunity against challenge ex posure with a fresh culture ofPasteurella multocida. Thus, he established the principle of attenuation, as he recog nized that organisms could be ren dered avirulent by various treatments, but were still able to provoke immu nity. This concept of attenuation also led to the successful development of the rabies vaccine which was first ad ministered to a human being in 1885. Since Pasteur, the principles of de Fig. 3.9.1. Pasteur (left) watching the vaccination of veloping live vaccines have essentially a boy against rabies in 1885. 136 remained unchanged, until the rise of biotechnology in the 1970's. Biotechnology has created tools for studying the molecular basis of virulence and immunogenicity of mi crobes, and the molecular pathogenesis and immunology of infectious diseases. The knowledge gained in this field offers new strategies for engineering well-defined vaccines. Various approaches for designing 'biotech' vaccines are outlined in Table 3.9.1. Scientists at (CDI-DLO) in Lelystad are performing research directed at developing a deletion mutant vaccine against bovine herpes virus type 1infection s that at the same time can serve as a vector for foreign genes. 3.9.2 Characteristics of bovine herpes virus type 1 infections Disease Bovine herpes virus type 1 (BHV1) induces diseases of the respiratory and reproductive tracts in cattle. It causes: - rhinotracheitis (Figure 3.9.2a), characterized by necrotic lesions and exudate formation in nose, eye (Fig. 3.9.2b) and trachea, often accompanied by fever, general depression, a drop in milk production and abortion; - vulvovaginitis/balanoposthitis, a venereal disease characterized by pustular lesions in genital mucosae. Because the infection temporary suppresses various functions of the immune system, secondary bacterial infections often aggravate the disease (Yates, 1982) Pathogenesis During the primary infection the virus first replicates in nasal and genital mucosae and is excreted from these tissues in enormous quantities for 10 to 15 days (Fig. 3.9.3). The animal is most infective during this period. After this primary phase, BHV1 persists in the animal in a latent state, in which the viral DNA resides in nervous tissues, probably for the host's entire lifetime (Rossi et al., 1982). Under influence of stress, the viral DNA may be Table 3.9.1. Modern approaches for vaccine development. Killed vaccines contain (parts of) micro-organisms that are unable to replicate, whereas live vaccines contain organisms that replicate in the vaccinate. 'Killed' vaccines 1. Subunit vaccines 2. Recombinant DNA vaccines - polypeptide expression in bacteria, yeast, viruses or mammalian cells 3. Synthetic peptide vaccines 4. Anti-idiotype vaccines 'Live' vaccines 1. Deletion mutant vaccines 2. Reassortment vaccines 3. Vector vaccines (vaccinia, avipox viruses) 137 Fig. 3.9.2a I'uiuluii exudate and small necrotic foci in the nose of a calf infected with BHV1. reactivated to form an infectious virus, which may spread through the body to nasal and genital cavities. Thus, latent infections may lead to virus excretion and transmission to other susceptible cattle. This latent state, which is a hallmark of many herpes virus infec tions, implies that in the framework of an eradication programme all (latently) infected animals must be identified and killed. Fig. 3.9.2b. A mucopurulent discharge from the eye due to conjunctivitis in a calf infected with BHV1. 138 Fig. 3.9.3. Virus shedding after infection with BHV1 in a non-vaccinated and a vaccinated calf. Control/eradication BHV1 is prevalent worldwide. In the Netherlands, approximately 50% of cattle possess antibodies to the virus. Recently, Denmark and Switzerland, which have never permitted the use of BHV1 vaccines, have eradicated the virus by a programme of identifying and subsequent slaughter of infected animals. The international trade in cattle and cattle products is jeopardized by the presence of BHV1 in some countries and its absence in others. Therefore, several European countries are currently running programmes to elim inate BHV1 from artificial insemination centres as a first phase in nationwide eradication. Vaccines Conventionally prepared live and killed vaccines are currendy being used to lessen eco nomic losses due to BHV1 infections. Although they reduce the severity of the disease, they do not uniformly prevent infection or subsequent virus excretion (Fig. 3.9.3) and establish ment of latency. Furthermore, various vaccines are not safe, i.e. they induce latency themselves and may give rise to abortion (Pastoret et al., 1980, Whetstone et al., 1986). Another drawback of using vaccines is that they interfere with eradication programmes, because vaccinated cattle cannot be distinguished serologically from cattle infected with the wild-type virus. Therefore, novel, more efficacious and safe marker vaccines need to be developed. The application of biotechnology offers tools, prospects to design modern BHV1 vaccines. 3.9.3 Designing a BHV1 deletion mutant vaccine Viral characteristics The size of the virion (Fig. 3.9.4) is 150-200 nanometre in diameter. The genome of BHV1 consists of a linear double-strand DNA molecule, with a size of approximately 140 139 Fig. 3.9.4. An electron micrograph ofa BHV1 particle. kilobases. It is composed of a unique long (Ul) sequence and a unique short (Us) sequence, bracketed by an internal repeat (IR) and terminal repeat (TR) sequences (Fig. 3.9.5). Glycoproteins play a crucial role in the interaction of the virus with its host. They are major components of the virion envelope and are expressed on the surface of the infected cell. The genes coding for seven viral glycoproteins have been mapped, among which those of four putatively non-essential genes gC, gG,gl and gE (Fig. 3.9.5). Viruses that fail to express gC or gE are still able to replicate in cell cultures and in cattle. The functions of Fig. 3.9.5. A schematic drawing of the genome organization of BHV1, indicating the location of four non-essential glycoprotein genes. 140 these non-essential glycoproteins in the virus replication cycle are not yet well understood. As for other herpes viruses, the thymidine kinase (TK) enzyme is also non-essential. The presence of non-essential genes in the BHV1-genome creates the possibility to construct mutant viruses with deletions in these genes. Research objective The first objective of the research, which isperforme d in close collaboration with research groups in Belgium and Germany and an international pharmaceutical industry, is to construct mutant viruses with deletions in either the gE, gl, gG or gC whether or not combined with a deletion in the 77i"gene. These deletion mutant viruses are assessed for virulence, immunogenicity and induction of latency in cattle. The mutant with the best vaccine properties is selected as a candidate vaccine, which will then be extensively tested for safety. Monoclonal antibodies against BHV1 are prepared and selected for reactivity with the non-essential glycoproteins using prokaryotic and eukaryotic cells that are manip ulated to express these glycoproteins. The monoclonal antibodies are examined for their suitability for use in glycoprotein-specific antibody assays. Whether BHV1 can serve as a vector for foreign genes will first be investigated with genes of the bovine respiratory syncytial virus (BRSV). Research progress BHV1-mutants with deletions in thegi sand/o r 7Xgenes have been constructed and tested in specified pathogen-free calves in comparison with the wild-type parent strain (van Engelenburg et al., 1992). Whereas the parent strain gave rise to the typical signs of infectious bovine rhinotracheitis, the deletion mutants hardly induced any sign of disease. Hence, gE and TKaxe involved in the expression of virulence of BHV1. After challenge exposure, the non-inoculated control calves developed rhinotracheitis, whereas the calves given the deletion mutants or the parent strain five weeks before challenge did not develop severe signs of disease. This indicates that the immunogenicity of BHV1 is hardly, if at all, reduced, by deleting the gE and/or TTCgenes . Monoclonal antibodies have been prepared which recognize at least four antigenic domains on the gE glycoprotein (Rijsewijk et al, 1992), and antibodies against gE have been detected in naturally infected as well as in experimentally infected cattle. Genes of BRSV, a virus which causes acute respiratory disease in calves during autumn, are the first to be inserted in BHV1 deletion mutants. For this purpose, cDNA clones of the immunodominant F, N and G proteins of BRSV have been made. 3.9.4 Concludin g remarks Freedom of BHV1 will improve productivity in the cattle industry and will facilitate international trade in cattle and cattle products. An efficacious and safe marker vaccine and a companion diagnostic test can greatly contribute to the eradication of BHV1, particularly in countries where the virus isendemic . Although there are several approaches for designing marker BHV1 vaccines (Table 3.9.1), the route leading to live deletion mutant vaccines is considered the most promising. There are four candidate genes (gE,gl, 141 gG and gQ to be used as a serological marker. The preliminary results on gE-deleted BHV1 mutants are encouraging. 3.10 Safety of recombinant DNA virus vaccines T.G. Kimman, A.L.J. Gielkens, K. Glazenburg,J.M.A . Pol & W.A.M. Mulder Animal production and welfare is severely hampered by infectious diseases, both in devel oped and developing countries. Vaccines are among the most effective and cheap prophy lactic tools to help prevent losses due to infectious diseases. Vaccines are not only used to control clinical signs following infection, but also to reduce transmission of the pathogen in the host species with the purpose of eliminating the pathogen. Examples of successful disease eradication using vaccines are smallpox among the human population and, more recently, rabies among foxes in several parts of Europe (Pastoret & Brochier, 1991). Attempts to eradicate Aujeszky's disease among pigs in commercial pig farms using vacci nation have also produced promising results. Disease eradication in modern animal hus bandry using vaccines could lead to more or less 'disease free' animal husbandry which needs to make only limited use of antibiotics. Recombinant DNA techniques are usually applied to generate live vaccines, but they have also been useful in generating dead vaccines from micro-organisms which do not multiply well or safely in vitro,o r from which the protective antigens cannot be easily purified. For example, there are vaccines available against the hepatitis Bviru s that consist of hepatitis B surface antigen produced by yeast cells. This vaccine carries fewer risks on contamination with unwanted pathogens in comparison with the old vaccine which was produced from blood from infected individuals. Provided that the rules of Good Manufac turing Practice, including checking the presence of live micro-organisms, have been well- performed, such dead genetically engineered vaccines carry no more risks than conven tional dead vaccines. 3.10.1 Safety of live genetically engineered non-carrier vaccines To develop a live vaccine, attenuation is of the virulent micro-organism to which immu nity must be induced is required. Spontaneous attenuation may occur during cultivation of the micro-organism and is due to genetic changes varying from point mutations to the loss of genes. Attenuation can also be achieved by deletion mutagenesis using rec. DNA techniques. In principle, such genetically engineered deletion mutants carry no more risks than conventionally attenuated mutants. On the contrary, rec. DNA techniques may be able to construct mutants that are less likely to cause safety problems. Well chosen, multiple deletions in several virulence genes may lead to vaccine strains which are very unlikely to revert to virulence, whereas conventionally attenuated vaccine strains may have genetic changes that are insufficient to cause attenuation or so small that they may easily revert to virulence. For example, the conventionally attenuated live Sabin vaccine strain against the polio virus contained only a single nucleotide substitution and could 142 therefore easily revert to virulence (Evans et al., 1985). Reversion to virulence has also occurred in rabies vaccines (Report WHO meeting 1990). Another potential drawback of conventionally attenuated vaccine strains is insufficient attenuation. An early measles vaccine still caused fever and rash in vaccinated children and vaccinia virus caused en cephalitis in a small minority of those who had been vaccinated (Mims, 1986). A thorough knowledge of gene functions is thus important in order to choose specific deletions for attenuation. In our laboratory the function of several gene products has been identified for the herpes virus in pigs,Aujeszky' s disease virus (ADV), which belongs to the subfamily of alpha-herpesvirinae. Inactivation of the genes encoding the 28K protein or glycoprotein gX did not reduce virulence, while inactivation of the genes encoding ribonu cleotide reductase (RR), protein kinase (PK), glycoprotein gp63, or gl strongly reduced virulence (Kimman et al. 1991, de Wind et al, in press). In particular, inactivation of the gene encoding gl strongly reduces virulence of the virus, while its immunogenicity is fully conserved. Further analysis revealed that deletion of two amino acids, valine-125 and cysteine-126, in a conformation-dependent antigenic domain of gl completely abolishes neurovirulence of ADV for pigs (Jacobs et al, submitted). Another virulence gene of herpes viruses is the gene encoding thymidine kinase (TK), which istherefor e often deleted from genetically engineered ADV vaccine strains. Knowledge of such virulence genes which are non-essential for virus growth in vitro thus allows a rational approach to the design of safe and effective deletion vaccines. Live vaccines may be able to recombine with other vaccines or wild-type strains of the same or related species. In vivo recombination of a gl negative ADV vaccine strain with another ADV vaccine strain deleted in another glycoprotein may result in the generation of virulent recombinants (Henderson et al. 1990). However, this event is more of a theor etical than practical importance, because pigs will not normally be vaccinated simultane ously with two vaccines. Nevertheless, because ADV vaccine strains may establish latent infections, vaccine strains which can complement each other's gene deletions should not be used in the same population of animals. Live ADV vaccines may also be able to recombine with field strains of ADV, which may also be considered of minor practical importance, because such a recombination will not lead to the emergence of more virulent strains than those already circulating in the field. An important safety requirement is that vaccines should be free from mycoplasmas and from extraneous viruses. Pestiviruses are notorious contaminants ofvaccines . Live vaccines are obviously more prone to contamination by infectious agents than inactivated vaccines. 3.10.2 Safety of live genetically engineered carrier vaccines Vaccination with live carrier vaccines which express genes of foreign pathogens is a promising approach. Complex antigens are amplified in vivoan d in a way that mimics antigen presentation after natural infection. This approach may be valuable for the devel opment of vaccines against pathogens which cannot be grown easily or safely invitro. Most attention has been focused on vaccinia virus, a pox virus, as vaccine vector. Vaccinia virus is effective in eliciting both B and T cell-mediated responses. Other vectors currendy under investigation are adeno viruses, herpes viruses, and several bacteriae including 143 Salmonella, Shigella and Escherichia coli species (Report WHO meeting, 1990). Recombinant vaccinia virus carrying the glycoprotein G gene of rabies virus induces complete and long-term resistance against rabies in foxes and raccoons. The recombinant virus has been shown to be innocuous for the target and non-target species. Recently, vaccine baits containing capsules with recombinant virus have been used successfully in a large field trial in Belgium to immunize foxes. The incidence of rabies declined rapidly over the trial period, and no cases of the disease in either foxes or domestic animals have been reported in the area (Pastoret & Brochier 1991). Other pox viruses presently under investigation as vaccine vector are fowlpox, raccoonpox and swinepox. Carrier vaccines based on adeno viruses are an alternative for vaccinia and other pox viruses. Attenuated strains of human adeno viruses types 4 and 7 have been administered orally or intratracheally, thereby inducing immunity on mucosal surfaces. One advantage of some adeno viruses is their narrow host range, while others have a wider host range. More information regarding immunogenicity, safety, and stability of adeno virus recombi nants is however required before practical application can be considered. Live attenuated herpes virus vaccines are applied worldwide to protect swine against Aujeszky's disease, poultry against Marek's disease, and cattle against bovine rhinotrachei- tis. Because herpes viruses have large genomes (varying from 130 kbp to 230 kbp) large segments of foreign DNA can be inserted without affecting infectivity. Foreign genes have usually been incorporated in the TK gene. Other insertion sites have also been used successfully, including the Us2 gene of Marek's disease virus, the major beta gene of human cytomegalo virus, and the gX and gl genes of ADV (Cantello et al., 1991; Spaete et al, 1987; van Zijl et al., 1991; Hooft van Iddekinghe et al., unpublished results). Scientists at the National Cancer Institute and CDI-DLO constructed an ADV vector carrying the envelope glycoprotein El of hog cholera virus (HCV) (van Zijl et al., 1991). Pigs vaccinated with this recombinant proved to be fully protected against a challenge with a lethal dose of HCV and were also well protected against a challenge with virulent ADV. Live carrier vaccines carry similar potential risks of harmful side-effects as live non- carrier vaccines, but the incorporation of foreign genes leads to additional safety consider ations before such vaccines can be used. Through the incorporation of foreign genes the carrier organism may acquire altered biological properties. The following potential risks of carrier vaccines should be examined: 1. Changes in cell, tissue, or host tropism, and virulence of the carrier organism through the incorporation of foreign genes. 2. Recombination with other vaccines or wild-type strains of the carrier organism and subsequent changes in biological properties. 3. Spread and survival in the environment. Only limited experimental data are available which illustrate that the biological behavi our of a micro-organism can change as a result of the incorporation of a foreign gene. For example, there are some indications that vaccinia virus carrying the attachment protein G of respiratory syncytial virus, a pneumotropic virus, replicates better in lungs of mice than vaccinia virus carrying other genes of this virus (Taylor et al, 1991). Likewise, incorpor ation of the g/virulence gene of ADV in vaccinia virus appeared to enhance its virulence 144 (Kost et al., 1989). To estimate the risks of recombination it is necessary to know the reservoirs of potential acceptors of foreign genes. For example, vaccinia virus might easily recombine with cowpox virus which is endemic in Europe in small rodents and which may infect domestic cats and man. 3.10.3 Legislation Worldwide, strict regulations are being applied concerning the construction of rec DNA vaccines, their use in small-scale animal experiments, their environmental release in large- scale field trials, and their registration for the market. In the EC, the registration of biotechnologically engineered vaccines in member states is coordinated by the Committee for Veterinary Medicinal Products (CVMP). However, even within the EC, rules appear to be applied differently in the member states. Further European and international coordi nation regarding legislation of rec DNA vaccines is therefore highly desirable. Despite considerable effort, decision-making and legislation regarding the environmental release of rec DNA vaccines are largely undefined. Risk assessment studies should therefore provide a solid scientific basis from which all the possible environmental problems in volved in the use ofthes e vaccines can be realistically identified and evaluated. This should guarantee their safety and protect the environment without imposing unnecessary bur dens. 3.10.4 Aim of the research Research at CDI-DLO is focused on potential risks connected with the use of live carrier and non-carrier herpes virus vaccines, using ADV in pigs as a model (Fig.3.10.1). Major steps in the pathogenesis of Aujeszky's disease in pigs are virus multiplication in the oropharyngeal and nasal mucosa, uptake and transport of the virus by axons, multiplica tion and spread ofth e virus in the central nervous system (CNS). In the CNS, the virus may cause neuronal necrosis and inflammation. It may also disseminate by viraemia, but this route seems less important. The virus further gives rise to latent infections of neurons, which is probably lifelong. After reactivation of latent ADV, infectious virus may again be excreted and transmitted to other pigs. These properties of wild-type ADV require thor ough evaluation when using attenuated ADV as vector. For both carrier and non-carrier vaccines it iso futmos t importance to know which gene products of the carrier organism determine properties such as virulence, spread from animal to animal, survival in the environment, host range, latency, and reactivation. So far RR, TK, PK, gp63, and gl have been shown to be important for virulence, while gll, gH, and gp50 are essential genes necessary for the virus to replicate (Kimman et al., 1992; Peeters et al., 1992). Glycoprotein gl appears to affect tissue tropism of ADV in pigs, presumably by facilitating the spread of the virus through the central nervous system (Fig. 3.10.2. a and b). Although still possible, strains deficient in TK activity are less likely to cause latent infections that can be reactivated (Mengeling 1991). No differences in virulence, cell or host tropism, or in the pathogenesis were detected among gl and TK negative strains of ADV expressing, or not expressing, the El gene of 145 Fig. 3.10.1. Electron micrograph of the herpes virus of pigs, the Aujeszky's disease virus. Bar represents 100 nm. HCV under the control of the gX promoter. However, pigs inoculated with the recombi nant vaccine virus shed less virus and over a shorter period compared with pigs inoculated with the control vaccine virus, presumably because the foreign gene functioned as a 'ballast' or affected the stability of the virus. Both strains proved avirulent for pigs. In contrast, a constructed mutant that could theoretically occur upon recombination of the vaccine carrier (ADV TK', gl', gX', El +) with wild-type ADV proved highly virulent for pigs. The pathogenesis of this strain (ADVTK+, gl+, gX', El +) was indistinguishable from a wild-type ADV (TK+, gl+, gX', El ")an d the virus killed three out of five pigs (Mulder et al., submitted for publication). The results demonstrate that thegX ofAD V is not involved in the virulence of ADV for pigs, that inserting a foreign gene into the gX locus only minimally, if at all, reduces the virulence of ADV, and that the gZ locus is therefore not a suitable insertion place for foreign genes (Fig. 3.10.3). Methods will be developed that can estimate the frequency of recombination after the in vivo administration of different strains using different routes of application. In addition, the basic reproductive rate (R0) of vaccine strains will be determined in order to quantify pig-to-pig transmission of vaccine strains, and to identify genes of ADV that determine transmission and thereby the survival in the host species. IfR 0 isles s than 1tha n the strain cannot establish itself in a population by transmission (Anderson & May, 1992). 3.10.5 Outlook Rec DNA techniques have allowed a more rational approach of generating vaccines against micro-organisms which cannot be easily attenuated or which cannot be grown 146 H "T-vr :>;i*ÄjK.*;.-- ..,;• . .. • .ai*-..: r f* F'. ••?&* -" • •• '• • • , '*•«-• 1fc: 'ï *>•:. Fig. 3.10.2. Immunohistological staining of brain tissue ofpig s inoculated with a ^/-negative strain ofAujeszky' s disease virus (ADV) (top) and a^/-positiv e strain of the same virus (bottom). Both strains were inoculated by nasal inhalation. Infected cells stain dark. The ^/-positive strain induces degener ation and necrosis of neurones with inflammation: meningoencephalitis. These lesions were absent in the brain ofpig s inoculated with the^/-negativ e strain, which replicated almost equally well at the mucosal site of application and in peripheral neurones. So the absence of virus and lesions in the brains ofpig s inoculated with the negative strain indicates that the gl protein isinvolve d in transport of the virus through the brain. ADV strains that lack gl are avirulent for pigs. 147 El \ / \ / \ / NONVIRULEN T I I ( ) + VIRULENT gX gp50 gp63 gi El \ / \ / VIRULENT I I E1 \ / \ / NONVIRULEN T I I 1 ! + VIRULENT gX gp50 gp63 gl \ / \ / MILDLYVIRULEN T \ / TK 148 simply or safely. Safe deletion mutant strains can be generated by introducing preferably multiple deletions in several genes determining virulence and replication. Because such live vaccines may induce suitable protection, their future appears promising. Examples of such vaccines are deletion vaccines against Aujeszky's disease and bovine herpes virus type 1. The use of carrier vaccines expressing foreign genes carries more potential risks because the phenotype of the carrier organism is altered and the transfer of the heterologous genes to wild-type micro-organisms may involve further risks. The risks related to the use of such vaccines should be evaluated, case-by-case. Thorough knowledge of the function of the inserted gene product and of the gene at the insertion locus may minimize such risks. For example, genes that might alter host tropism of a pathogen should generally not be incorporated in a carrier organism, or only in such a way that it is harmless, for instance because it is not expressed on the cell or virion surface or because a minimum of protective sequences are expressed. Carrier organisms may be modified so that their environmental spread is minimal or absent. For example, ADV mutants with a deletion of the essential gene encoding gp50 may infect cells, but infectious viral particles are not produced (Peeters et al., 1992). Likewise, fowlpox virus does not give rise to a productive infection in non-avian species, but induces immunity to inserted gene products (Taylor et al, 1988). Deletions in genes that determine latency, or reactivation, or both, may reduce the period during which the carrier is present in the environment, thereby reducing the risks on recombination. The choice of a vector with a small host range may further limit environmental spread of the vector, but preclude its general use. For example swinepox virus only infects swine. This virus would therefore be a suitable carrier candidate in swine in contrast to the vaccinia virus which has a broad host range. Possible exchange of genes can be further minimized by choosing a parental route of immunization. Should there be the possibility ofrecombination , then any possible harmful consequences can be minimized. For example, the expression of a foreign gene in a gene that determines virulence or replication of a micro-organism will yield a less virulent micro-organism after recombination with a wild-type of micro-organism. Finally, carrier organisms could be designed so that lethal mutations occur after recombination with wild-type strains. For this purpose, foreign genes should be inserted into vital genes so that non-viable mutants occur upon recombination. Fig. 3.10.3. Potential effects of recombination between different ADV-carrier vaccine strains and circulating wild-type strains of the virus. When a heterologous gene (i.e. the gene encoding the El protein of hog cholera virus) is cloned in a gene of the carrier (ADV TK, gT) that has no role in virulence (e.g. the gene encoding glycoprotein gX), then recombination with wild-type ADV (TK*, gX*, gl+, El ')ma y lead to a virulent strain (ADV TK*, gX', El *, gl+). In contrast, ifth e heterologous genes is cloned in a gene that governs virulence (i.e. the gene encoding glycoprotein gl), then recombination will lead to a mildly virulent strain (ADV TK*, gX*, El *, gT). 149 3.11 Putting a 'bar code' on bacteria? B.A.M. van der Zeijst The biomass on earth contains over 1029 bacteria (calculated from Fenchel, 1988). They belong to more than 2500 different recognized bacterial species. But the actual number of bacterial species is much higher. Only 0.1% of the bacterioplankton is culturable (Fer guson et al., 1984). It is assumed that the other 99.9% belong to at least 10 000 unknown species of bacteria (Wayne et al., 1987). Each species is adapted to the specific environ ment it inhabits such as the sea, the soil or other living organisms, like protozoa, animals and plants. Medical and veterinary bacteriology deal with much smaller numbers. There are only a few hundred species of pathogenic bacteria and even a few ingested bacteria can cause disease. Medical and veterinary bacteriologists are thus confronted with the difficult task of specifically detecting a small number of pathogenic organisms within an overwhelming majority of other bacteria. This art has been refined immensely during the past century by selective culture techniques and subsequent typing methods. Nevertheless, in some in stances the currently available methods fail or are very laborious. This, for instance, is the case with non-culturable bacteria or bacteria with fastidious growth requirements. Molec ular biological techniques, particularly the polymerase chain reaction (PCR, Fig. 3.11.1), can fill this gap. In a previous communication, we reported how we had started to develop novel diag nostic methods using these latest techniques (Wagenaar et al., 1990). We decided on a generally applicable approach, based on amplification of genes coding for bacterial 16S ribosomal RNA (rRNA genes). All bacteria contain 16S rRNA genes. These genes contain fragments that are highly conserved among bacteria, while other parts are specific to a smaller group of bacteria or even to one bacterial species (Fig. 3.11.2). After determining the 16S rRNA nucleotide sequence of the pathogenic bacteria of interest, primers for the polymerase chain reaction can be chosen that amplify specific regions of the 16S rRNA gene of that bacterium. We and others have used this approach successfully for a number of bacteria and will continue to do so. This section, however, looks at the results of these research efforts from a wider perspec tive. The results of the systematic worldwide analysis of 16S rRNA sequences have been deposited in computer databases (Olsen et al, 1992). Analysis of these data has yielded insights that have much wider implications than simply the diagnosis of diseases caused by bacteria, especially in the fields of: - bacterial phylogeny and evolution; - identification and classification of unculturable and uncultured bacteria in nature and disease; - the direct detection of bacteria in soil and other complex environments. 3.11.1 Bacterial phylogeny and evolution Man has always felt the need to classify living creatures. Early classification was based on 150 /w / w/ w/ \\ DNA double helix denature DNA and add primers ,—. ^^ i add nucleotides and enzyme i , ^^ ,. ^^m * ^•™ f ^^ ^^H cycle 2 f • "^ , - ^~ cycle 3 •—' • • ^^m f • • ^^ • • , , ^^™" ^^ cycle 4 ^^ ^^ 1 ,^^— 1 ^^ i 1 Fig. 3.11.1. Schematic representation of the polymerase chain reaction. Double-stranded DNA template is melted out, allowing the binding of the primers onto the individual DNA strands. The primers are elongated by the enzyme with the added nucleotides. After the first cycle, two double- stranded DNA molecules were formed. During the next cycles, the number of fragments with a fixed length increases exponentially. comparative anatomy. Later on the classification became phylogenetic, i.e. followed the lineages of evolution. Bacteriologists have in the past tried to apply the principles used in botany and zoology to develop a phylogenetic system for bacteria. This was a failure (van Niel, 1955) since the morphological variation of bacteria is extremely limited: a sphere, a rod or a spiral. The only other traditional criterion was a cell wall feature, either gram- negative or gram-positive. A solution was indicated by Zuckerkandl and Pauling (1965) in their paper 'Molecules as documents of evolutionary history' in which they argued that nucleic acid would be the 'master molecules' providing the basis for a molecular phylogeny. At that time it was still 151 CUIAJCCUU IGU "CCCGGUQ U GUUG UGCUfG cu AGGGCCGG 1070 C G CT 1020 UGAGAAUG ,CC 1040 C G —C G-C , G-C-' »e* 5 980 990 g - % 1Ö0 , Fig. 3.11.2. Structure and domains of the Escherichia coli 16S rRNA. The molecule is drawn so as to indicate the secondary structural elements (Stern et al., 1988). Regions highly conserved among bacteria that can be used to derive oligonucleotide primers for DNA sequencing and the PCR are indicated by a line (Weisburg et al., 1991). 152 impossible to determine DNA and RNA sequences. When RNA sequences became possi ble Woese argued that the 16Sribosoma l RNA (rRNA) subunit iswel l suited for measuring phylogenic relationships among the bacteria (Woese et al., 1983, Woese et al., 1990). The 16S rRNA molecules are: - present in all bacteria; - have long, highly conserved regions useful for investigating at distant phylogenetic relationships, interspersed with variable regions, valuable for close relationships; - unlikely candidates for lateral transfer, i.e. there is no exchange of 16S rRNA genes between different bacterial species, as often occurs with other gene(s), e.g. those involved in antibiotic resistance. Nowadays the genes for the 16S rRNA are the standard object of research, rather than the rRNA itself. The generation of 16S rRNA gene sequences and their comparison and analysis by the suitable computer programs have resulted in a completely revised phylo genetic system for bacteria. The unexpected finding was that there are three main groups of cellular organisms (Fig. 3.11.3): Bacteria, Eucarya (the 'higher' forms of life) and Archaea (archaebacteria) (Woese et. al, 1990). The latter group had not been recognized before. It consists of organisms adapted to extreme environmental conditions (heat, salt concentration, atmosphere). Not only is the overall shape of this 'universal phylogenetic tree' new, there are also many detailed differences with the former morphological classification in the subtrees. The new system has provided a useful framework for understanding the biological properties of bacteria. Also it has led to the renaming of many bacteria, a confusing but necessary process that will go on for a number of years. 3.11.2 Identification and classification of unculturable bacteria Recent analysis of 16S rRNA sequences of marine microflora and bacteria belonging to Bacteria Archaea Eucarya TTiemtotogales Fig. 3.11.3. Universal genetic tree determined from evolutionary distance matrix analysis of small subunit rRNA sequences (Woese et al., 1990). 153 'microbial mats' (stratified bacterial communities living in aqueous environments) has identified many unknown species (Britschgi & Giovannoni, 1991; Ward et al., 1990; Giovannoni et al., 1990). Several pathogenic bacteria that are poorly, if at all, culturable have also been identified in this way. Examples from human medicine are the bacteria causing Whipple's disease (Relman et al., 1992b) and bacillary angiomatosis (Relman et al, 1990; Relman et al, 1992a). To date, examples relevant to veterinary medicine are the classification of the agents causing cat scratch disease (O'Connor et al., 1991; Relman et al., 1992a), contagious endometritis (a venereal disease of horses, Bleumink-Pluym et al., 1993) and of Covudria mminantium.Th e latter agent is endemic in sub-Saharan Africa where it causes heartwater, a fatal disease of ruminants. This bacterium appeared to be closely related to the Ehrlichia species causing disease in dogs and humans (van Vliet et al., 1992; Fig. 3.11.4). Without doubt many more fastidious or unculturable pathogenic bacteria will be de tected and characterized by the 16S rRNA approach in the near future. 3.11.3 The detection of transmission routes of bacteria A longstanding problem in veterinary microbiology is to establish the way pathogenic organisms spread from one infected individual to the next. This is not a simple task. Many potentially pathogenic bacteria are virtually ubiquitous. They occur freely in nature (soil Ehrlichia sennetsu Ehrlichia risticii r Ehrlichia chaffeensis Ehrlichia canis - Cowdria mminantium Ehrlichiaphagocytophila / Ehrlichia equi - Anaplasma marginale Rickettsia prowazekii - Escherichia coli -Chlamydia psittaci 0.00 0.05 0.10 0.15 0.20 0.25 Evolutionary distance Fig. 3.11.4. Evolutionary-distance tree showing the relationship of Cowdria mminantium with other bacterial species (van Vliet et al., 1992). 154 or water) or in animals in which they do not cause disease. Examples are Salmonella, Yersinia, Campylobacter, Listeria, Actinobacillus, Streptococcus, Brucella, Bacillus,Mycobacterium and Leptospira species. Disease will occur after stress and/or after transmission of the infection from either human or another animal. However, sometimes there will be no disease at all because the animal is harbouring an apathogenic bacterial strain. It will be clear that to gather useful information on the spread of the clone causing a particular bacterial disease, each bacterium should ideally have its own 'bar code'. Traditionally the 'bars' contained information on the surface of the bacterium (serological properties, protein profiles, phage receptors), complemented by physiological properties (fermentation of nutrients, antibiotic resistance). The techniques used to obtain this information are usually labour intensive and the results are often ambiguous. Therefore, there is a tendency to switch to a 'genotypic' approach. A first step in such an approach is often based on the characterization ofth e 16S rRNA. This is a powerful method for determining of the bacterial species. Moreover, using fluorescently labelled oligonucleotides hybridizing with rRNA bacteria can be detected in situ (Hahn et al., 1992). However, characterization of the 16S rRNA isonl y a preliminary step and is insufficient to discriminate between clonal isolates of the same bacterial species as exemplified by our own work on Taylorella equigenitalis, the causative agent of equine contagious endometritis (Bleumink-Pluym et al, 1993). Additional 'bars' must be provided by other methods. As pointed out above, there is a tendency to use a 'genotype' approach such as e.q. electro- phoretic separation of the characteristic large fragments of the bacterial genome. This latter method provides considerable information in the case of T. equigenitalis (Bleumink-Pluym et al., 1990), but proved to be rather labour intensive. Employing prim ers to sequences occurring at several positions in the genome in PCR-techniques though, are simple and quick and have already been shown to work in a number of cases (Woods et al., 1992; Fekete et al; 1992, Versalovic et al, 1991). 3.11.4 Prospects for the future in veterinary bacteriology The expected new developments already indicated are: - understanding the phylogenetic position of a scarcely studied pathogenic bacteria may provide clues about how they cause disease based on analogy with more broadly studied organisms; - the ability to detect and characterize an unculturable bacterium by analyzing its 16S rRNA (genes) provides methods which will prove that this bacterium causes a certain disease; - new methods to put a 'bar code' on individual bacteria species or strains which will provide methods to follow their dispersal, especially through international livestock and meat transport and trade. In addition, the detection of genes causing resistance to antibiotics will become feasible. Many of these genes have been characterized Qacoby & Archer, 1991) and can be easily detected using PCR techniques (Courvalin, 1991). Usually antibiotic resistance is suddenly acquired from other, already resistant bacteria. This either occurs by the uptake of a plasmid carrying the gene for resistance or by the uptake and incorporation of a chromo- 155 somal gene encoding resistance from another bacterium. After the introduction of an antibiotic it is only a matter of time before resistance develops, sometimes gradually, as the result of consecutive mutations of resistance genes. This happened recently after introducing quinolones, a class of potent new synthetic antibiotics inhibiting bacterial DNA replication (Hooper et al., 1986). Mutations confer ring resistance have been mapped (Oram & Fisher, 1991; Goswitz etal , 1992).Thi s opens up the possibility of following upcoming resistance, both in the pathogens of interest and in commensal bacteria that can act as sentinel organisms. On the basis of this kind of monitoring it would be possible to switch, ifnecessary , to other antibiotics. To summarize: We are close to the possibility of putting a label with a 'bar code' onto bacterial clones.Ther e islittl e doubt that these technical possibilitieswil lb e used tocollec t epidemiological data and tocontro l and monitor movements oflivestoc k within Europe in the coming years. After an initial investment to develop and improve techniques the costs of these assays will be in the same order of magnitude, or even cheaper, as conventional tests based on culturing bacteria. 3.12 References Aggeler,J., C.S Park, & MJ . Bissell, 1988. Regulation of milk protein and basement membrane gene expression: The influence of the extracelluar matrix.Journa l of Dairy Science 71: 2830-2842. Andersson, L., & L. Rask, 1988. Characterization of the MHC class II region in cattle. The number of DQ_genes varies between haplotypes. Immunogenetics 27:110-120. Anderson, R.M. & R.M. May, 1992. Understanding the AIDS pandemic. Scientific American 266:20-26. Anilionis, A., W. Wunner & R Curtis, 1981. Structure of the glycoprotein of rabies virus. Nature 294:275-276. Archibald, A., C.S. Haley, L. Andersson, A.A. Bosma, W. Davies, M. Fredholm, H. Geldermann, J. Gellin, M. Groenen, I. Gustavsson, L. Ollivier, E.M. Tucker & A. van der Weghe, 1990. PIGMAP: A European initiative to map the porcine genome. Animal Genetics 22, Suppl. 1:82- 83. Awdeh, Z.L., L.D. Raum, EJ. Yunis & CA. Alper, 1983. Extended HLA/complement allele haplotypes: Evidence for T/t-like complex in man. Proceedings of the National Academy of Sciences of the USA 80:259-263. Awoniyi, C.A., V. Chandrashekar, R.D. Arthur, B.D. Schanbacher, A.G. Amador & R.E. Falvo, 1988. Pituitary and Leydig cell function in boars actively immunized against gonadotrophin- releasing hormone.Journa l of Reproduction and Fertility 84:295-302. Bayna, E.M. &J.M. Rosen, 1992. Tissue-specific, high level expression ofth e rat whey acidic protein gene in transgenic mice. Nucleic Acids Research 18: 2977-2985. Bensaid, A., A. Kaushal, J.R. Young & A.J. Teale, 1991. Pulsed field gel electrophoresis and its application in the physical analysis of the bovine major histocompatibility complex. In: L.B. Schook, H.A. Lewin & D.G. McLaren (Eds.): Gene Mapping Techniques. Marcel Dekkers Inc., New York, pp. 127-158. Bernoco, D., H.A. Lewin, L. Andersson, MA. Arriens, G. Byrns, S. Cwik, CJ . Davies, H.C. Hines, W. Leibold, O. Lie, D. Meggiolaro, R.A. Oliver, H. Ostergard, R.L. Spooner, J.A. Stewart- Haynes, AJ . Teale,J.W . Templeton & M. Zanotti, 1991.Join t report of the fourth International Bovine Lymphocyte Antigen (BoLA) Workshop, East Lansing, Michigan, USA, 25 august 1990. Animal Genetics 22:477^-96. Bjorkman, PJ., MA. Saper, B. Samraoui, W.S. Bennett, J.L. Strominger & D.C. Wiley, 1987. 156 Structure of the human class I histocompatibility antigen, HLA-A2. Nature 329:506-512. Bleumink-Pluym, N., E.A. ter Laak & B.A.M, van der Zeijst, 1990. Epidemiologie study of Taylorella equigenitalis strains by field inversion gel electrophoresis of genomic restriction endonuclease frag ments.Journa l of Clinical Microbiology 28:2012-2016. Bleumink-Pluym, N.M.C., L. van Dijk, A.H.M, van Vliet,J.W.B , van der Giessen & B.A.M, van der Zeijst, 1993. Yhylogzneûc^oùûonoï Taylorella equigenitalis determined by analysis of amplified 16S ribosomal DNA sequences. International Journal of Systematic Bacteriology, submitted. Bonneau, M. & C. Carelli, 1987. Essai de castration immunologique du porc mâle par un vaccin aqueux synthétique. Annales Zootechnique 36:265-270. Bonsing,J. & A.G. Mackinlay, 1987. Recent studies on nucleotide sequences encoding the caseins. Journal of Dairy Research 54:447^1614. Booman, P., M. Tieman, D.EM, van de Wiel,J.M . Schakenraad & W. Koops, 1984. Application of monoclonal antibodies in animal production: pregnancy diagnosis in cattle. In: E.H. Houwink & R.R. van der Meer (Eds.): Innovations in Biotechnology. Elsevier Science Publishers, Amster dam, p. 259-266. Botham,J.W. , M.F. Grace & MJ . Manning, 1980. Ontogeny of first set and second set alloimmune reactivity in fishes. In: M.J. Manning (Ed.): Phylogeny of Immunological Memory, Elsevier/ North-Holland Biomedical Press, Amsterdam, pp. 83-92. Brascamp, E.W, C.S. Haley, R. Fries, F. Grosclaude, R. Hanset &J.E . Womack, 1991. Genome mapping in farm animals, a survey of current research. 42nd annual meeting of the EAAP. Britschgi, T.B. & SJ . Giovannoni, 1991.Phylogeneti c analysis of a natural marine bacterioplankton population by rRNA gene cloning and sequencing. Applied and Environmental Microbiology 57:1707-1713. Brochier, B., B. Languet, J. Blancou, M.P. Kieny, J.P. Lecocq, F. Costy, P. Desmettre, & P.P. Pastoret, 1988. Use of recombinant vaccinia-rabies virus for oral vaccination of fox cubs(Vulpes vulpes) against rabies. Veterinary Microbiology 18:103-108. Brooks, R.I., A.M. Pearson, M.G. Hogberg, J J. Pestka & J.I. Gray, 1986. An immunological approach for prevention of boar odor in pork.Journa l of Animal Science 62:1279-1289. Bruford, M.W. et al., 1990. Hypervariable markers in the chicken genome. In: Burke,T., G. Dolf, J.A. Jeffreys & R. Wolff (Eds.): Proceedings of the 4th World Congress Applied to Livestock Production 13:139-142. Birkhauser-Verlag, Basel. Bumstead, N. &J . Palyga, 1992. A preliminary linkage map of the chicken genome. Genomics 13, 690-697. Callesen, H., T. Grève & F. Christensen, 1987. Ultrasound guided aspiration of bovine follicular oocytes. Theriogenology 27:217. Cantello,J.L., A.S. Anderson, A. Francesconi &R.W. Morgan, 1991. Isolation of aMarek's disease virus (MDV) recombinant containing the lacï gene of Escherichia coli stably inserted within the MDV US2 gene.Journa l of Virology 65:1584-1588. Copeland, N.G. & N.A.Jenkins, 1991. Development and applications of a molecular genetic linkage map of the mouse genome. Trends in Genetics 7:113-118. Cornall, RJ., TJ. Airman, CM. Hearne & J.A. Todd, 1991. The generation of a library of PCR-analyzed microsatellite variants for genetic mapping of the mouse genome. Genomics 10:874-881. Courvalin, P., 1991. Genotypic approach to the study of bacterial resistance to antibiotics. Antimi crobial Agents and Chemotherapy 35:1019-1023. Crittenden, L.B., L. Provencher, L. Santangelo, I. Levin, A. Dahlgren,J.B . Dodgson, H. Abplanalp & W.E. Briles, 1992. Characterization of a red jungle fowl by white leghorn backcross reference population for molecular mapping of the chicken genome. Submitted. Dale, T.C., MJ. Krnacik, C. Schmidhauser, C.L.Q. Yang, MJ. Bissell & J.M. Rosen, 1992. High-level expression of the rat whey acidic protein gene is mediated by elements in the promoter and 3' untranslated region. Molecular and Cellular Biology 12, 905-914. Daniel, M.J., E.V.B. Shenoy & P.G. Box, 1984. Immunization of pigs against the boar taint steroid 157 androstenone.Journa l of Comparative Pathology 94:319-321. Davies, CJ., & D.F. Antczak, 1991. Production and characterization of alloantisera specific for bovine class II major histocompatibility complex antigens. Animal Genetics 22:417-434. Davies C.J., L.Andersson , I.Joosten , P.Mariani , L.C. Gasbarre & E.J. Hensen, 1992. Characteriza tion of bovine MHC class II polymorphism using three typing methods: serology, RFLP and 1ER European Journal of Immunogenetics 19:253-262. Dawkins, R.L., F.T. Christiansen, P.H. Kay, M.J. Garlepp,J . McKluskey, PL. Hollingworth & PJ. Zilko, 1983. Disease associations with complotypes, supratypes and haplotypes. Immunology Reviews 70:5-23. de Wind, N., A. Berns, A. Gielkens & T.G. Kimman. Ribonucleotide reductase-deficient mutants of Pseudorabies virus are avirulent for pigs and induce partial protective immunity. Journal of General Virology, in press. Dekkers,J.J. , H.C. van der Plas & D.H. Vuijk (Eds.), 1990. Agricultural biotechnology in focus in the Netherlands. Pudoc, Wageningen. DeMars, R. & T. Spies T, 1992. New genes in the MHC encode proteins for antigen processing. Trends in Cell Biology 2:81-86. Desmettre, P., B. Languet, G. Chappuis, B. Brochier, I. Thomas, J.P. Lecocq, M.P. Kieny, J. Blancou, M. Aubert, M. Artois & P.P. Pastoret, 1990. Veterinary Microbiology 23:227-236. Dixon, B., R.J.M. Stet, S.H.M, van Erp & B. Pohajdak, 1993. Characterization of/?2-microglobulin transcripts from two teleost species. Immunogenetics, in press. Doppler, W., B. Groner & R.K. Ball, 1989. Prolactin and glucocorticoid hormones synergistically induce expression of transfected rat beta-casein gene promoter constructs in a mammary epithe lial cell line. Proceedings of the National Academy of Sciences of the USA 86:104—108. Eikelenboom, G. & D. Minkema, 1974. Prediction of pale, soft, exudative muscle with a non-lethal test for halothane induced porcine malignant hyperthermia syndrome. Netherlands Journal of Veterinary Science 99:421-426. Ellegren, H., C.J. Davies & L. Andersson, 1992. A polymorphic microsatellite within the bovine major histocompatibility complex DRB3 gene: Implications for microsatellite evolution and DRB3 typing. Genomics, in press. Evans, D.M.A., G. Dunn, P.D. Minor, G.C. Schild, A.J. Cann, G. Stanway, J.W. Almond, K. Currey &J.V Maizel, 1985. Increased neurovirulence associated with a single nucleotide change in a non-coding region of the Sabin type 3 poliovaccine genome. Nature 314:548-550. Falvo, R.E., V. Chandrashekar, R.D. Arthur, A.R. Kuenstler, T. Hasson, C. Awoniyi & B.D. Schanbacher, 1986. EfTect of active immunization against LHRH or LH in boars: reproductive consequences and performance traits.Journa l of Animal Science 63:986-994. Faustman, D., X. Li, H.Y. Lin, Y. Fu, G. Eisenbarth, J. Avruch &J. Guo, 1991. Linkage of faulty major histocompatibility complex class I to auto-immune diabetes. Science 254:1756-1761. Fekete, A.,J.A . Bande, S.M. Hailing & R.W. Stich, 1992. Amplification fragment length polymor phism in Brucella strains by use of polymerase chain reaction with arbitrary primers. Journal of Bacteriology 174:7778-7783. Fenchel, T., 1988. Marine plankton food chains. Annual Review of Ecological Systems 19:19-38. Ferguson, R.L., E.N. Buckley & A.V. Palumbo, 1984. Response of marine bacterioplankton to differential filtration and confinement. Applied and Environmental Microbiology 47:49-55. Ferretti, C, P. Leone & V. Sgaramella, 1990. Long range restriction analysis of the bovine casein genes. Nucleic Acids Research 18:6829-6833. Fries, R.,J . Beckmann, M. Georges, M. Soller &J. Womack, 1989. The bovine gene map. Animal Genetics 20:3-29. Fujii,J. , K. Otsu, F. Zorzato, S. De Leon, V.K. Khanna,J.E . Weiler, PJ. O'Brien & D.H. MacLen- nan, 1991. Identification of a mutation in porcine ryanodine receptor associated with malignant hyperthermia. Science 253:448-451. Garcia-Regueiro, J.A. & I. Diaz, 1989. Evaluation of the contribution of skatole, indole, andros tenone and androstenols to boar-taint in back fat ofpig s by HPLC and capillary gas chromatogra- 158 phy (CGC). Meat Science 25:307-316. Geissler, E.N., S.V. Cheng,J.F . Gusella & D.E. Housman, 1988. Genetic analysis of the dominant white-spotting (W) region on mouse chromosome 5: Identification of cloned DNA markers near W. Proceedings of the National Academy of Sciences of the USA 85:9635-9639. Georges, M., P. Cochaux, A.S. Lequarré, M.W. Young & G. Vassart, 1987. DNA fingerprinting in man using a mouse probe related to part of the DrosophilaPer gene. Nucleic Acids Research 15:7193. Georges, M., A. Mishra, L. Sargeant, M. Steele & X. Zhao, 1990. Progress towards a primary DNA marker map in cattle. Proceedings of the 4th world congress on genetics applied to livestock production XIII: 107-112. Giovannoni, S.J., T.B. Britschgi, C.L. Moyer & K.G. Field, 1990. Genetic diversity in Saragasso Sea bacterioplankton. Nature 345:60-63. Giphart, M.J. & W. Verduijn, 1991. The eurotransplant HLA-DRB oligonucleotide typing set. European Journaal of Immunogenetics 18:57-68. Glass, EJ. , R.A. Oliver & R.L. Spooner, 1990. Variation in T cell responses to ovalbumin in cattle: evidence for Ir gene control. Animal Genetics 21:15-28. Gloudemans, A.G.M., N. Cohen, G.E. Boersbroek, J.L. Grondel, E. Egberts & W.B. van Muis winkel, 1986. The major histocompatibility complex (MHC) in fish. Carp family studies using the mixed leucocyte reaction. In: K. Tiews (Ed.): Selection, Hybridization and Genetic Engineering in Aquaculture, Vol. II, Heenemann Verlag, Berlin, Federal Republic of Germany, pp. 265-268. Goddeeris, B.M., W.I.Morrison, A.J. Teale, A. Bensaid & C.L. Baldwin, 1986. Bovine cytotoxic T-cell clones specific for cells infected with the protozoan parasite Theileria parva: Parasite strain specificity and class I major histocompatibility complex restriction. Proceedings of the National Academy of Sciences of the USA 83:5238-5242. Goswitz,JJ., K.E. Willard, C.E. Fasching & L.R. Peterson, 1992. Detection oïgyrA gene mutations associated with ciprofloxacin resistance in methicillin-resistant Staphylococcus aureus - Analysis by polymerase chain reaction and automated direct DNA sequencing. Antimicrobial Agents and Chemotherapy 36:1166-1169. Groenen, M.A.M., J.J. van der Poel, R.J.M. Dijkhof & MJ. Giphart, 1989. Cloning of the bovine major histocompatibility complex class II genes. Animal Genetics 20:43-54. Groenen, M.A.M.J.J. van der Poel, R.J.M. Dijkhof &M J . Giphart, 1990. The nucleotide sequence of bovine class II DQB and DRB genes. Immunogenetics 31:37-44. Groenen, M.A.M., R.J.M. Dijkhof, AJ.M . Verstege & J.J. van der Poel, 1992a. The complete nucleotide sequence of the bovine as2-casein gene. Gene in press. Groenen, M.A.M., R.J.M. Dijkhof, J.J. van der Poel, R. van Diggelen & E. Verstege, 1992b. Multiple octamer binding sites in the promoter region ofth e bovine as2-casein gene. Nucleic Acids Research 20, 4311-4318. Groenen, M.A.M. & JJ. van der Poel, 1993. Regulation of expression of milk protein genes. Submitted. Guyer, M. et al., 1992. A comprehensive genetic linkage map of the human genome. Science 258:67-86. Hagen, G., 0. Andresen, T. Blichfeldt & K.A. Berg, 1988. Effects of immunization of young boars against gonadotropin releasing hormone. Proceedings of the 11th International Congress on Animal Reproduction and Artificial Insemination, Dublin, 1988. Vol. 4, abstract no. 493. Hahn, D., R.I. Amann, W. Ludwig, A.D.L. Akkermans & K.H. Schleifer, 1992. Detection of micro-organisms in soil after in situ hybridization with rRNA-targeted, fluorescently labelled oligonucleotides.Journa l of General Microbiology 138:879-887. Haley, C.S., A. Archibald, L. Andersson, A.A. Bosma, W. Davies, M. Fredholm, H. Geldermann, M. Groenen, I. Gustavsson, L. Ollivier, E.M. Tucker & A. van de Weghe, 1991. The pig gene mapping project - PIGMAP. In: Proceedings of the 4th World Congress on Genetics and Applied Livestock Production (Edinburgh) XIII: 67-70. Haley, C.S., 1991. Use of DNA fingerprints for the detection of major genes for quantitative traits 159 in domestic species. Animal Genetics 22:259-277. Hamada, H., M.G. Petrino & T. Kakunaga, 1982 A novel repeated element with Z-DNA-forming potential iswidel y found in evolutionary diverse eukaryotic genomes. Proceedings of the National Academy of Sciences of the USA 79:6465-6469. Hashimoto, K., T. Nakanishi & Y. Kurosawa, 1990. Isolation of carp MHC genes. Proceedings of the National Acadamy of Sciences of the USA 87:6863-6867. Henderson, L.M., J.B. Katz, G.A. Erickson & J.E. Mayfield, 1990. In vivo and in vitro genetic recombination between conventional and gene-deleted vaccine strains of Pseudorabies virus. American Journal of Veterinary Research 51:1656-1662. Hendriksen, PJ.M, L.A.Johnson,J.A . Grootegoed & T. van der Lende, 1992. Anti H-Y monoclo nal antibodies do not show different binding to X and Y enriched boar sperm fractions. Proceed ings ofth e International Congress on Animal Reproduction, The Hague, volume 1,pp . 478^80. Hilbert, P., K. Lindpainter, J.S. Beckmann, T. Serikawa, F. Soubrier, C. Dulbay, P. Cartwright, B. De Gouyon, C.Julier, S. Takahasi, M. Vincent, D. Ganten, M. Georges & G.M. Lathrop, 1991. Chromosomal mapping of two genetic loci associated with blood-pressure regulation in heredi tary hypertensive rats. Nature 353:521-529. Hildemann, W.H., 1970. Transplantation immunity in fishes: Agnatha, Chondrichthyes and Os- teichthyes. Transplantation Proceedings 2:253-259. Hooper, D.CJ.S. Wolfson, K.S. Souza, C. Tung, G.L. McHugh & M.N. Swartz, 1986. Genetic and biochemical characterization of norfloxacin resistance in Escherichia coli. Antimicrobial Agents and Chemotherapy 29:639-644. Jacobs, L., TG. Kimman, A.L.J. Gielkens &J.T van Oirschot, 1993. Deleting two amino acids in glycoprotein gl of Pseudorabies virus reduces plaque size in various cells and virulence for mice. Submitted for publication. Jacoby, G.A. & G.L. Archer, 1991.Ne w mechanisms of bacterial resistance to antimicrobial agents. New EnglandJourna l of Medicine 324:601-612. Jeffreys, AJ., V. Wilson & S.L. Thein, 1985. Hypervariable 'minisatellite' regions in human DNA. Nature 314:67-73. Johnson, L.A., 1991. Sex preselection in swine: altered sex ratios in offspring following surgical insemination of flow sorted X- and Y-bearing sperm. Reproduction in Domestic Animals 26:309- 314. Jones, S.R.M. & M. Palmen, 1993. The effects of inoculum route and dose on the immune response of common carp, Cyprinus carpio L., to the blood parasite Trypanoplasma borreli. Veterinary Immu nology and Immunopafhology, in press. Joosten, I., R.A. Oliver, R. Spooner,J.L. Williams, B.G. Hepkema, EJ . Hensen, 1988. Characteri zation of class I bovine lymphocyte antigens (BoLA) by one dimensional iso-electric focusing. Animal Genetics 19:103-113. Joosten I., M.E Sanders, A. van der Poel, J.L. Williams, B.G. Hepkema & E.J. Hensen, 1989. Biochemically defined polymorphism of bovine MHC class II antigens. Immunogenetics 29:213- 216. Joosten, I., E.J. Hensen, M.F. Sanders & L. Andersson, 1990. Bovine MHC class II restriction fragment length polymorphism linked to expressed polymorphism. Immunogenetics 31:37-44. Kaastrup, P., B. Nielsen, V Horlyck & M. Simonsen, 1988. Mixed lymphocyte reactions (MLR) in rainbow trout (Salmo gairdneri) sibling. Develop. Comp. Immunol. 12:801-808. Kaastrup, P., R.J.M. Stet, A. Tigchelaar, E. Egberts & W.B. van Muiswinkel, 1989. A major histocompatibility locus in fish: serological identification and segregation of transplantation anti gens in the common carp (Cyprinus carpio L). Immunogenetics 30:284—290. Kaufman,J.E , S. Ferrone, M.F. Flajnik, M. Kilb, H. Volk & R. Parisot, 1990a. MHC-like molecules in some nonmammalian vertebrates can be detected by some cross-reactive monoclonal antibod ies.Journa l of Immunology 144:2273-2280. Kaufman, J.F., K. Skjoedt, J. Salomonsen, M. Simonsen, L. DuPasquier, R. Parisot & P. Riegert, 1990b. MHC-like molecules in some nonmammalian vertebrates can be detected by some cross- 160 reactive xenoantisera.Journa l of Immunology 144:2258-2272. Kieny, M.P., R. Lathe, R. Drillien, D. Spehner, S. Skory, D. Schmitt, T. Wiktor, H. Koprowski & J.R Lecocq, 1984. Expression of rabies virus glycoprotein from a recombinant vaccinia virus. Nature 312:163-166. Kimman, T.G., N. de Wind, N. Oei-Lie, J.M.A. Pol, AJ.M. Berns & A.LJ. Gielkens. 1992. Contribution of single genes within the unique short region of Aujeszky's disease virus (suid herpesvirus type 1)t o virulence, pathogenesis and immunogenicity. Journal of General Virology 73:243-251. Komen,J. , A.B.J. Bongers, C.J.J. Richter, W.B. van Muiswinkel & E.A. Huisman, 1991. Gynogene- sis in common carp (Cj/prinus carpio L.). II. The production of homozygous gynogenetic clones and F, hybrids. Aquaculture 92:127-142. Kost, T.A., E.V. Jones, K.M. Smith, A.P. Reed, A.L. Brown & T.J. Miller. 1989. Biological evaluation of glycoproteins mapping to two distinct mRNAs within the 5amHI fragment 7 of Pseudorabies virus: Expression of the coding regions by vaccinia virus. Virology 171:365-376. Laird,J.E., L.Jack, L. Hall, A. Boulton, D. Parker & R.K. Craig, 1988. Structure and expression of the guinea-pig (Z-lactalbumin gene. Biochemical Journal 254:85-94. Lambert, R.D., C. Bernard,J.E . Rioux, R. Béland, D. D'Amours &A . Montreuil, 1983. Endoscopy in cattle by the paralumbar route: Technique for ovarian examination and follicular aspiration. Theriogenology 20:149-161. Lee, C.S. & T. Oka, 1992. A pregnancy-specific mammary nuclear factor involved in the repression of the mouse /8-casein gene transcription by progesteron. Journal of Biological Chemistry 267:5797-5801. Lee, K.F., S.H. Atiee &J.M. Rosen, 1990. Differential regulation of rat/J-casein - chloramphenicol acetyltransferase fusion gene expression in transgenic mice. Molecular and Cellular Biology 9:560-565. Lewin, H.A„ M.C. Wu, T.J. Nolan &J.A. Stewart, 1988. Association between BoLA and subclinical bovine leukaemia virus infection in a herd of Holstein-Friesian cows. Immunogenetics 27:338- 344. Love, J.M., A.M. Knight, M.A. McAleer & J.A. Todd, 1990. Towards construction of a high resolution map of the mouse genome using PCR-analyzed microsatellites. Nucleic Acids Research 4:4123-4130. Lubon, H. & L. Hennighausen, 1987. Nuclear proteins from lactating mammary glands bind to the promoter of a milk protein gene. Nucleic Acids Research 15:2103-2121. Lubon, H. & L. Hennighausen, 1988. Conserved region ofth e rat a-lactalbumin promoter is a target site for protein binding in vitro. BiochemicalJourna l 256:391-396. MacLennan, D.H., C. Duff, F. Zorzato,J . Fuji, M. Phillips, R.G. Komeluk, W. Frodis, B.A. Britt & R.G. Worton, 1990. Ryanodine receptor gene is a candidate for predisposition to malignant hyperthermia. Nature 343:559-561. McEvoy,J.D. , 1992. Alteration of the sex ratio. Animal Breeding Abstracts 60:97-111. Mengeling, W.K., 1991.Viru s reactivation in pigs latently infected with a thymidine kinase negative strain of Pseudorabies virus. Archives of Virology 120:57-70. Mercier, J.C.,J.L. Vilotte & C. Provot, 1990. Structure and function of milk protein genes. In: H. Geldermann & F. Ellendorf (Eds): Genome analysis in domestic animals. VCH, Weinheim, Germany, p. 233-258. Meuwissen, T.H.E. &J.A.M . van Arendonk, 1992. Potential improvements in rate of genetic gain from marker-assisted selection in dairy cattle breeding schemes. Journal of Dairy Science 75:1651-1659. Miller, N.W., A. Deuter & L.W. Clem. L.W., 1986. Phylogeny of lymphocyte heterogeneity: the cellular requirements for the mixed leucocyte reaction with channel catfish. Immunology 59:123- 128. Miller, N.W.,J.E . Bly, F.W. van Ginkel, CF. Ellsaesser & L.W Clem, 1987. Phylogeny of lympho cyte heterogeneity: identification and separation offunctionall y distinct subpopulations of channel 161 catfish lymphocytes with monoclonal antibodies. Develop. Comp. Immunol. 11:739-747. Mims, C.A., 1986. The pathogenesis of infectious diseases. Third edition, Academic Press. Monaco,JJ. , 1992. A molecular model of MHC class-I-restricted antigen processing. Immunology Today 11:173-179. Mulder, W.A.M.,J . Priem, K.L. Glazenburg, J.M.A. Pol, A.L.J. Gielkens & TG. Kimman. Viru lence and pathogenesis of attenuated and wild-type strains of Aujeszky's disease virus expressing glycoprotein El of hog cholera virus. Submitted for publication. Nadeau, J.H., 1989. Maps of linkage and synteny homologies between mouse and man. Trends in Genetics 5:82-86. Nakamura, Y, M. Leppert, P. O'Connel, R. Wolff, T Holm, M. Culver, C. Martin, E. Fujimoto, M. Hoff, E. Kumlin & R. White, 1987. Science 235:1616-1622. Neefjes,JJ. & H.L. Ploegh, 1992. Intracellular transport of MHC class II molecules. Immunology Today 11:179-184. Neu, H.C., 1992. The crisis in antibiotic resistance. Science 257:1064-1073. O'Connor, S.P., M. Dorsch, A.G. Steigerwalt, DJ. Brenner & E. Stackebrandt, 1991. 16S rRNA sequences of'Bartonella bacilliformis and cat scratch disease bacillus reveal phylogenetic relationships with the alpha-2 subgroup of the class Proteobacteria.Journa l of Clinical Microbiology 29:2144— 2150. Olsen, G.J., R. Overbeek, N. Larsen, TL. Marsh, MJ. Mccaughey, M.A. Maciukenas, W.M. Kuan, TJ. Macke, Y.Q. Xing & C.R. Woese, 1992. The ribosomal database project. Nucleic Acids Research 20:2199-2200. Ono, H., D. Klein, V. Vincek, F. Figueroa, C. O'hUigin, H. Tichy &J . Klein, 1992. Mhc class II genes of zebrafish: origin and polymorphism. Proceedings of the National Acadamy of Science of the USA, in press. Ono, H., C. O'hUigin, V. Vincek, RJ.M. Stet, F. Figueroa &J. Klein, 1993. Newß chain encoding Mhc class II genes in the carp. Immunogenetics, in press. Oram, M. & L.M. Fisher, 1991. 4-Quinolone resistance mutations in the DNA gyrase ofEscherichia coli clinical isolates identified by using the polymerase chain reaction. Antimicrobial Agents and Chemotherapy 35:387-389. Pastoret, P.P., L.A. Babiuk, V Misra, & P. Griebel, 1980. Reactivation of temperature-sensitive and non-temperature-sensitive infectious bovine rhinotracheitis vaccine virus with dexamethasone. Inf. Immunology 29:483-388. Pastoret, P.P., B. Brochier, B. Languet, I. Thomas, A. Paquot, B. Bauduin, M.P. Kieny,J. P Lecocq, J. de Bruyn, F. Costy, H. Antoine & P. Desmettre, 1988. First field trial offo x vaccination against rabies using a vaccinia-rabies recombinant virus. Veterinary Records 123:481—483. Pastoret, P.-P. & B. Brochier, 1991. Biological control of wild animal infections. Current Opinion in Biotechnology 2:465-469. Peeters, B., N. de Wind, M. Hooisma, F. Wagenaar, A. Gielkens, R. Moormann, 1992. Pseudora bies glycoproteins gp50 and gll are essential for virus penetration, but only gll is involved in membrane fusion. Journal of Virology 66:894—905. Pierson, R.A. & O.J. Ginther, 1987a. Follicular populations during the estrous cycle in heifers: I. Influence of day. Animal Reproduction Science 14:165—176. Pierson, R.A. & O.J. Ginther, 1987b. Follicular populations during the estrous cycle in heifers: II. Influence of right and left sides and intraovarian effect of the corpus luteum. Animal Reproduc tion Science 14:177-186. Pieterse, M.C., K.A. Kappen, Th.A.M. Kruip & M.A.M. Taverne, 1988. Aspiration of bovine oocytes during transvaginal ultrasound scanning of the ovaries. Theriogenology 30:751-762. Pieterse, M.C., P.L.A.M. Vos, Th.A.M. Kruip, Y.A. Wurth, Th.H. van Beneden, A.H. Willemse & M.A.M. Taverne, 1991. Transvaginal ultrasound guided follicular aspiration of bovine oocytes. Theriogenology 35:857-862. Purdom, CE., 1983. Genetic engineering by the manipulation of chromosomes. Aquaculture 33:287-300. 162 Rajakoski, E., 1960. The ovarian follicular system in sexually mature heifers with special reference to seasonal, cyclical and left-right variations. Acta Endocrinologica, Supplement 52, Copenhagen. Reed, H.C.B., D.R. Melrose & R.L.S. Patterson, 1974. Androgen steroids as an aid to the detection of oestrus in pig artificial insemination. British VeterinaryJourna l 130:61-66. Relman, D.A.,J.S . Loutit, T.M. Schmidt, S. Falkow & L.S. Tompkins, 1990. The agent of bacillary angiomatosis. New EnglandJourna l of Medicine 323:1573-1580. Relman, D.A., RW. Lepp, K.N. Sadler & T.M. Schmidt, 1992a. Phylogenetic relationships among the agents of bacillary angiomatosis, Bartonella bacilliformis & other alpha-proteobacteria. Mo lecular Microbiology 6:1801-1807. Relman, D.A., T.M. Schmidt, R.P. Macdermott & S. Falkow, 1992b. Identification of the uncul tured bacillus of Whipple's Disease. New England Journal of Medicine 327:293-301. Report WHO meeting, 1990. Potential use of live viral and bacterial vectors for vaccines. Vaccine 8:425-437. Rijkers, G.T. & WB. van Muiswinkel, 1977. The immune system of cyprinid fish. The development of cellular and humoral responsiveness in the rosy barb [Barbus conchonius). In:J.B . Solomon &J.D. Horton (Eds.): Developmental Immunobiology, Elsevier/North-Holland Biomedical Press, Am sterdam, pp. 233-240. Rijkers, G.T, 1980. The Immune System of Cyprinid Fish. Ph.D.-Thesis, Agricultural University, Wageningen, The Netherlands, 176 pp. Rijsewijk, F.A.M., J. Magdalena, J. Moedt, MJ . Kaashoek, M.A. Maris-Veldhuis, A.LJ. Gielkens, &J.T van Oirschot, 1992. Identification and functional analysis of glycoprotein E (gE) of bovine- herpes virus type 1. Proceedings of the 17th International Herpesvirus Workshop, Edinburgh, August 1-7, 1992, p. 245. Rossi, C.R., G.K. Kiesel, & P.F. Rumph, 1982. Association between route of inoculation with infectious bovine rhinotracheitis virus and site of recrudescence after dexamethasone treatment. American Journal of Veterinary Research 43:1440-1442. Schmidhauser, C, M.J. Bisell, C.A Meyers & F. Casperson, 1990. Extracelluar matrix and hormones transcriptionally regulate bovine /?-casein 5' sequences in stably transfected mouse mammary cells. Proceedings of the National Academy of Sciences of the USA 87:9118-9122. Schmidhauser, C, G.F. Casperson, C.A. Myers, K.T Sanzo, S. Bolten & M.J. Bisell, 1992. A novel transcriptional enhancer is involved in the prolactin- and extracellular matrix-dependent regula tion of/J-casein gene expression. Molecular Biology of the Cell 3:699-709. Schmitt-Ney, M., W. Doppler, R.K. Ball & B. Groner, 1991. /f-casein gene promoter activity is regulated by the hormone mediated relief of transcriptional repression and a mammary gland- specific nuclear factor. Molecular and Cellular Biology 11:3745-3755. Schmitt-Ney, M., B. Happ, R.K. Ball & B. Groner, 1992. Developmental and environmental regulation of a mammary gland-specific nuclear factor essential for transcription of the gene encoding /f-casein. Proceedings of the National Academy of Sciences of the USA 89:3130-3134. Secombes, CJ.,JJ.M. van Groningen & E. Egberts, 1983. Separation oflymphocyt e subpopulations in carp Cyprinus carpio L. by monoclonal antibodies: immunohistochemical studies. Immunology 48:165-175. Shenoy, E.V.B., M.J. Daniel & P.G. Box, 1982. The 'boar taint' steroid 5 163 Sigurdardottir, S., Borsch, K. Gustafsson & L. Andersson, 1992. Gene duplications and sequence polymorphism of bovine class II DQB genes. Immunogenetics 35:205-213. Simpson, S.P., AJ. Webb & I. Wilmut, 1986. Performance of British Landrace pigs selected for high and low incidence of halothane sensitivity. 1. Reproduction. Animal Production 43:485-492. Simpson, S.P. & A.J. Webb, 1989. Growth and carcass performance of British Landrace pigs heterozygous at the halothane locus. Animal Production 49:503-509. Smith, R.T., P.A. Miescher & R.A. Good, 1966. Phylogeny of Immunity, University of Florida Press, Gainesville, 277 pp. Soller, M., 1990. Genetic mapping of the bovine genome using deoxyribonucleic acid-level markers to identify loci afTecting quantitative traits of economic importance. Journal of Dairy Science 73:2628-2646. Soulier, S., J.L. Vilotte, M.G. Stinnakre &J.C. Mercier, 1992. Expression analysis of ruminant tï-lactalbumin in transgenic mice: developmental regulation and general location of important cis-regulatory elements. FEBS Letters 297:13-18. Spaete, R.R. & E.S. Mocarski, 1987. Insertion and deletion mutagenesis of the human cytomega lovirus genome. Proceedings of the National Academy of Sciences of the USA 80:7155-7217. Stern, S., B. Weiser & H.F. Nolier, 1988. Model for the three-dimensional folding of 16 S ribosomal RNA.Journa l of Molecular Biology 204:447-481. Stet, R.J.M. & E. Egberts, 1991. The histocompatibility system in teleostean fishes: from multiple histocompatibility loci to a major histocompatibility complex. Fish Shellfish Immunology 1:1-16. Stet, R.J.M., S.H.M, van Erp, T. Hermsen, H.A. Siiltmann & E. Egberts, 1993. Polymorphism of MhcCyca class I and class II genes in laboratory strains of common carp (Cyprinus carpio L.). De velop. Comp. Immunol., in press. Taylor,J. , R. Weinberg, B. Languet, P. Desmettre & E. Paoletti, 1988. Recombinant fowlpox virus inducing protective immunity in non-avian species. Vaccine 6:497-503. Taylor, G., E.J. Stott, G. Wertz & A. Ball. 1991. Comparison of the virulence ofwild-typ e thymidine kinase (tk)-deficient and tk+ phenotypes of vaccinia recombinants after intranasal inoculation of mice.Journa l of General Virology 72:125-130. te Pas, M.F.W. & J.H.F. Erkens, 1992. Porcine stress syndrome testing and the halothane gene. Proceedings of the 4th Netherlands Biotechnology Congress, Amsterdam 20-21 February 1992, G04 (abstract). te Pas, M.F.W. &J.H.F . Erkens, submitted. A Bam HI restriction fragment length polymorphism in pigs: differences between two Landrace breeds. Animal Genetics. Threadgill, D. W. &J.E. Womack, 1990. Genomic analysis of the major bovine milk protein genes. Nucleic Acids Research 18: 6935-6942. Threadgill, D.W. &J.E. Womack, 1991. Synteny mapping of human chromosome 8 loci in cattle. Animal Genetics 22:117-122. Tiwari,J.L. , & P.I. Terasaki, 1985. HLA and disease associations. Springer Verlag, New York. Tokunaga, K., G. Saueracher, PH. Kay, FT. Christiansen, R. Anand, R.L. Dawkins, 1988. Extensive deletions and insertions in different MHC supratypes detected by pulsed field gel electrophoresis.Journa l of Experimental Medicine 1988; 168:933-940. Trowsdale,J. , I. Hanson, I. Mockridge, S. Beck, A. Townsend &A . Kelly, 1990. Sequences encoded in the class II region of the MHC related the 'ABC' superfamily of transporters. Nature 348:741- 744. Turner,J.D. & H.T. Hung, 1989. Establishment of hormone responsive bovine mammary epithelial cell lines.Journa l of Dairy Science 72, Suppl. 1, 317. Usinger, W.R., M. Curie-Cohen & WH. Stone, 1977. Lymphocyte defined loci in cattle. Science 196:1017-1018. van de Wiel, D.F.M. & Eikelenboom, G., 1977. Improved method for permanent catheterization in pigs. Deutsche Tierärztliche Wochenschrift 84:12-16. van de Wiel, D.F.M.,J . Erkens, W Koops, E. Vos & A.AJ. van Landeghem, 1981. Periestrous and midluteal time courses of circulating LH, FSH, prolactin, estradiol-1Iß and progesterone in the 164 domestic pig. Biology of Reproduction 24:223-233. van der Beek, S. &J.A.M , van Arendonk. Estimation of recombination rates between marker loci from segregating populations containing several families. Submitted. van der Poel,J.J. , M.A.M. Groenen, RJ.M.Dijkhof, D. Ruyter & MJ. Giphart, 1990. The nucleo tide sequence of the bovine MHC class II alpha genes: DRA, DQA and DYA. Immunogenetics 31:29-36. van der Schans, A., L.A.J, van der Westeriaken, A.A.C, de Wit, WH. Eyestone & H.A. de Boer, 1991. Ultrasound-guided transvaginal collection of oocytes in the cow. Theriogenology 35:288. van der Schans, A., B.T.T.M. van Rens, L.A.J, van der Westeriaken & A.A.C, de Wit, 1992. Bovine embryo production by repeated transvaginal oocyte collection and in vitro fertilization. Proceed ings of the International Congress on Animal Reproduction, The Hague, volume 3, pp. 1366 1368. van Eijk, M.J.T.,J.A . Stewart-Haynes & H.A. Lewin, 1992. Extensive polymorphism of the BoLA- DRB3 gene distinguished by PCR-RFLP. Genomics, 23:483-496. van Engelenburg, F.A.C, MJ. Kaashoek, F.A.M. Rijsewijk, L. van den Burg, A. Moerman, A.L.J. Gielkens, &J.T. van Oirschot, 1992. Deletion ofth e glycoprotein E gene or thymidine kinase gene ofbovin e herpesvirus type 1 produces attenuated vaccine strains. Proceedings ofth e 17th Interna tional Herpesvirus Workshop, Edinburgh, August 1-7, 1992, p. 379. van Niel, C.B., 1955. Classification and taxonomy of the bacteria and bluegreen algae. In: A century ofprogres s in the natural sciences 1853-1953. California Academy of Sciences, San Francisco, pp. 89-114. van Vliet, A.H.M., F.Jongeja n & B.A.M, van der Zeijst, 1992. Phylogenetic position of Cowdria ruminantium (Rickettsiales) determined by analysis of amplified 16S ribosomal DNA sequences. InternationalJourna l of Systematic Bacteriology 42:494-^498. van Zeveren, A., A. van de Weghe, Y. Bouquet & H. Varewyck, 1988a. The porcine stress linkage group. I: The sequence of the linked blood marker loci in Belgian Landrace pigs. Journal of Animal Breeding and Genetics 105:177—186. van Zeveren, A., A. van de Weghe, Y Bouquet & H. Varewyck, 1988b. The porcine stress linkage group. III. Linkage disequilibria and application of blood marker analysis in selection against stress susceptibility in Belgian Landrace pigs.Journal of Animal Breeding and Genetics 105:195- 203. van Zijl, M., G. Wensvoort, E. de Kluyver, M. Hulst, H. van der Gulden, A. Gielkens, A. Berns & R. Moormann, 1991. Live attenuated Pseudorabies virus expressing envelope glycoprotein El of hog cholera virus protects swine against both Pseudorabies and hog cholera. Journal of Virology 65:2761-2765. Versalovic, J., T. Koeuth &J.R. Lupski, 1991. Distribution of repetitive DNA sequences in eub- acteria and application to fingerprinting of bacterial genomes. Nucleic Acids Research 19:6823- 6831. Vonderhaar, B.K. & S.E. Ziska, 1989. Hormonal regulation ofmil k protein gene expression. Annual Reviews in Physiology 51:641-652. Wagenaar,J.A.,J.W.B. van der Giessen, W Gaastra & B.A.M, van der Zeijst, 1990. Rapid diagnosis ofbacteria l diseases using DNA probes and gene amplification. In:J.J . Dekkers, H. C. van der Plas & D.H. Vuijk (Eds.): Agricultural Biotechnology in focus in the Netherlands. PUDOC, Wagenin gen, The Netherlands, p. 134-140. Ward, D.M., R. Weiler & M.M. Bateson, 1990. 16S RRNA sequences reveal numerous uncultured micro-organisms in a natural community. Nature 345:63-65. WatsonJ.W, K.E. Gordon, M. Robertson &A J . Clark, 1991.Interactio n of DNA-binding proteins with a milk protein gene promoter in vitro:identificatio n of a mammary gland-specific factor. Nucleic Acids Research 19, 6603-6610. Wayne, L.G., DJ. Brenner, R.R. Colwell, P.A.D. Grimont, O. Kandier, M.I. Krichevsky, L.H. Moore, W.E.C. Moore, R.G.E. Murray, E. Stackebrandt, M.P. Starr & H.G. Trüper, 1987. Report of the adhoc committee on reconciliation of approaches to bacterial systematics. Interna- 165 tionalJourna l of Systematic Bacteriology 37:463-464. Webb, AJ . & Simpson, S.P., 1986. Performance of British Landrace pigs selected for high and low incidence of halothane sensitivity. 2. Growth and carcass traits. Animal Production 43:493-503. Weisburg, W.G., S.M. Barns, D.A. Pelletier & DJ. Lane, 1991. 16S ribosomal DNA amplification for phylogenetic study.Journa l of Bacteriology 173:697-703. Weissenbach, J., G. Gyapay, C. Dib, A. Vignal, J. Morissette, P. Millasseau, G. Vaysseix & M. Lathrop, 1992. A second-generation linkage map of the human genome. Nature 359:794—801. Weller,J.I. ,J . Kashi & M. Soller, 1990. Power of daughter and granddaughter designs for mapping of quantitative traits in dairy cattle using genetic markers.Journa l of Dairy Science 73:2525. Whetstone, C.A., J.G. Wheeler, & D.E. Reed, 1986. Investigation of possible vaccine-induced epizootics of infectious bovine rhinotracheitis, using restriction endonuclease analysis of viral DNA. American Journal of Veterinary Research 47:1789-1795. Whitelaw, C.B.A., S. Harris, M. McClenaghanJ.P Simons & AJ. Clark, 1992. Position-independ ent expression of the ovine beta-lactoglobulin gene in transgenic mice. Biochemical Journal 286:31-39. Wicking, C. & B. Williamson, 1991. From linked marker to gene. Trends in Genetics 7:288-293. Wiegertjes, G.F.,J.G. Daly & W.B. van Muiswinkel, 1993a. Disease resistance ofcar p (Cyprinus carpio L.). Identification of individual genetic differences by bath challenge with atypicalAeromonas salmonicida. Submitted for publication. Wiegertjes, G.F., R.J.M. Stet & W.B. van Muiswinkel, 1993b. Variation in carp (Cyprinuscarpio L.) immune parameters. Detection of high and low immune responsive phenotypes. Submitted for publication. Williamson, E.D. & R.L.S. Patterson, 1982. A selective immunization procedure against 5a-andros- tenone in boars. Animal Production 35:353-360. Williamson, E.D., R.L.S. Patterson, E.R. Buxton, K.G. Mitchell, I.G. Partridge & N. Walker, 1985. Immunization against 5(X-androstenone in boars. Livestock Production Science 12:251-264. Woese, C.R., R. Gutell, R. Gupta & H.F. Noller, 1983. Detailed analysis of the higher-order structure of 16S-like ribosomal ribonucleic acids. Microbiological Reviews 47:621-669. Woese, C.R., O. Kandier & M.L. Wheelis, 1990. Towards a natural system of organisms: proposal for the domains Archaea, Bacteria & Eucarya. Proceedings of the National Academy of Sciences of the USA 87:4576-4579. Woods Jr., C.R.,J. Versalovic, T. Koeuth &J.R. Lupski, 1992. Analysis of relationships among isolates of Citrobacter diversus by using DNA fingerprints generated by repetitive sequence-based primers in the polymerase chain reaction.Journa l of Clinical Microbiology 30:2921-2929. Yates, W.D.G., 1982. Review of infectious bovine rhinotracheitis, shipping fever pneumonia and viral-bacterial synergism in respiratory disease of cattle. CanadianJourna l of Comparative Méde cine 46, 225-263. Yoshimura, M. T. & Oka, 1990. Transfection of/?-casein chimeric gene and hormonal induction of its expression in primary murine mammary epithelial cells. Proceedings of the National Academy of Sciences of the USA 87:3670-3674. Yu-Lee, L.Y, L. Richter-Mann, C.H. Couch, A.F. Stewart, A.G. MacKinlay, J.M. Rosen, 1986. Evolution of the casein multigene family: conserved sequences in the 5' flanking and exon regions. Nucleic Acids Research 14:1883-1902. Zorzato, F.,J . Fujii, K. Otsu, M. Phillips, N.G. Green, F.A. Lai, G. Meissner & D.H. MacLennan, 1990. Molecular cloning of cDNA encoding human and rabbit forms of the Ca2+ release channel (Ryanodine receptor) of skeletal muscle sarcoplasmic reticulum.Journa l of Biological Chemistry 265:2244-2256. Zuckerkandl, E. & L. Pauling, 1965. Molecules as documents of evolutionary history. Journal of Theoretical Biology 8:357-366. 166 4 Biotechnology in animal nutrition Contents 4.1 General introduction to biotechnology in animal nutrition 169 4.1.1 Characterization of feeds 169 4.1.2 Micro-organisms 169 4.1.3 Enzymes 171 4.1.4 Substrates 173 4.1.5 Active agents 174 4.1.6 Conclusions 174 4.2 Nutritional value and physiological effects of D-xylose and L-arabi- nose in poultry and pigs 175 4.2.1 Digestion of non-starch polysaccharides 175 4.2.2 Utilization of non-starch polysaccharides 176 4.2.3 Conclusions 177 4.3 Antinutritional factors (ANFs): aspects of the mode of action in the animal and inactivation of ANF-activity 177 4.3.1 Occurrence of ANFs in seeds 178 4.3.2 Major effects of ANFs in monogastric animals 179 4.3.3 Inactivation of ANFs 179 4.4 Phenolic monomers and digestibility of plant cell walls 182 4.4.1 Location of phenolic acids 182 4.4.2 Digestion process 184 4.4.3 Perspectives for biotechnology 185 4.4.4 Conclusions 186 4.5 Prospects of feed additives produced by biotechnology 186 4.5.1 Technological success 186 4.5.2 Commercial success 187 4.5.3 Fermentation products for the feed industry 187 4.5.4 Micro-organisms in animal feeding 189 4.5.5 Conclusions 190 4.6 References 190 167 4.1 General introduction to biotechnology in animal nutrition S. Tamminga, S.F. Spoelstra & GJ.M. van Kempen In animal production, feeds of mainly vegetable origin are hydrolysed in the digestive tract. The resulting monomers are absorbed and subsequendy utilized as sources of energy (ATP) or precursors for the synthesis of protein, lipids and occasionally carbohydrates. Digestion may be facilitated by micro-organisms inhabiting specific segments of the diges tive tract. Most pronounced are the presence of high numbers of bacteria, protozoa and sometimes fungi in the forestomachs of ruminants, and the presence of bacteria in the hindgut of horses and to a lesser extent in rabbits and pigs. Biotechnology in animal nutrition can be applied at three levels; in the feed, in the digestive tract and during intermediary metabolism. Application of biotechnology is possi ble through the micro-organisms, enzymes, substrates and active agents produced by this biotechnology. This paper elaborates on ideas concerning where and when biotechnology can be successfully applied in animal nutrition. 4.1.1 Characterization of feeds Feed is characterized by its ability to provide energy, protein, minerals and vitamins to animals. Frequently observed limitations of feeds are a high moisture content, resistance to enzymatic breakdown in the digestive tract and the presence of antinutritional factors (ANFs) (Tamminga et al., 1990). In feeds harvested at an immature stage, like many forages, a high moisture content is quite common. Storage of such feeds requires a suitable method of conservation, which involves microbes converting sugars into lactic acid. This conservation is achieved due to the low pH mainly resulting from the accumulation of lactic acid. Resistance to hydrolytic enzymes may result from the absence (or limited capacity) of the appropriate enzymes in the digestive tract. This situation normally occurs in monogas- tric animals. A discrepancy between the size ofpore s in the substrate to be hydrolysed and the appropriate hydrolytic enzymes may also prevent degradation. This situation is com mon in cell walls, particularly when they are encrustrated with lignin. The substrate to be hydrolysed may also be surrounded by a resistant layer. For example, the presence of waxes on many plant surface tissues or starch granules surrounded by a protein layer, rendering them resistant to amylolytic enzymes. Antinutritional factors (ANFs) often impair digestion by interacting with digestive en zymes (trypsin inhibitors, amylase inhibitors), with the gut wall (lectins) or with the feed itself (tannins) (Table 4.1.1). The negative effect of ANFs may not be restricted to the digestive tract, they may also impede intermediary metabolism. 4.1.2 Micro-organisms Micro-organisms play a positive role in animal production by supporting the conservation of moist feeds and by digesting structural carbohydrates in the forestomachs and hindgut. 169 Table 4.1.1. Nature, target and action of some antinutritional factors (ANFs). ANF Nature Target Result Trypsin inhibitor protein digestive enzymes inhibition Lectins glycoprotein gut wall damaging Amylase inhibitor protein enzymes digestive inhibition Tannins polyphenol protein complex formation On the other hand, pathogenic micro-organisms in the digestive tract endanger animal health and retard animal production, either directly or through the production of biogenic amines. These biogenic amines are thought to be responsible for impaired feed intake in ruminants, digestive disorders and reduced productivity in general. With regard to bio technology, two types of micro-organisms seem to have this potential: bacteria and fungi. Bacteria inforage conservation To increase the rate of acidification, starter cultures containing strains of lactic acid bacteria (usually belonging to the species Lactobacillusplantarum) ar e applied to the forage during ensiling. Several starter cultures, which give an inoculation rate of 105-106 live bacteria per gram of fresh forage are commercially available. If the crop contains enough fermentable sugars and the inoculation rate is sufficiently high to dominate the lactic acid bacteria naturally present on the crop, the use of such cultures leads to improved quality. In a literature review Spoelstra (1991) concluded that starter cultures applied to low dry matter (DM) grass and to wilted lucerne were, with a few exceptions, consistent in lowering pH, ammonia and acetic acid when compared with the untreated control silage. The effect on lactic acid was varied. On average its content increased, although in 13 out of 36 experiments a reduction was reported. Intake increased on average by 3 to 4% resulting in a 7% increase in liveweight gain and a 3% increase in milk production. However, only in 10 of the 32 production trials was a statistically significant increase in weight or milk reported. Data on the effect of starter cultures on high DM grass silages were lacking. In some segments of the digestive tract such as the forestomachs or the hindgut, cellu- lolytic bacteria are preferred, because they are the only means through which cell walls can be fully used. Lack of survival capability in the rumen limits the prospects of introduc ing into the rumen ecosystem genetically modified bacteria, like those capable of degrad ing one cell wall at low pH or bacteria capable of nitrogen fixation (Armstrong & Gilbert, 1991). Probiotics Many attempts have been made to improve production efficiency or the health status of veal calves and monogastrics by adding bacteria to the feed. The bacteria selected are intended to have a beneficial effect on the intestinal microflora and belong to different genera, including Lactobacillus, Streptococcus and Bacillus.Positiv e effects have been reported 170 on the zootechnical results, but these results are often contradictory (Van Belle et al., 1990). Yeasts andfimgi In ruminants, it has been claimed that some yeasts have a positive effect on rumen fermentation. Effects usually obtained with Saccharomyces cerevisiae or Aspergillus oryzae,in clude a rise in the numbers of cellulolytic bacteria both invitro and invivo. This is presumed to be the result of a stabilization of rumen pH due to a reduced rumen lactate concentra tion, arising from an increased uptake of lactate by Selelomonas ruminantium, a predominant strain of rumen bacteria when using concentrate- rich diet feeds (Martin & Nisbet, 1992). Of the 15 lactation studies reviewed by Williams and Newbold (1990), six showed a significantly increased yield of fat-corrected milk (FCM), and in one study a significant increase in milk fat content was observed. Average feed intake was increased by 2.4% (six studies) and FCM yield by 4.7% (15 studies). Cell wall degrading anaerobic fungi are also claimed to contribute to cell wall degrada tion, particularly poor quality cell walls which are encrustrated with lignin. Anaerobic fungi were shown to emulsify part of the lignin, but there is still no evidence that they can utilize lignin as a carbon source (Fonty &Joblin, 1991). 4.1.3 Enzymes The importance of enzymes in animal nutrition is their biocatalytic role, essential for the hydrolysis of feed components in the digestive tract (Table 4.1.2). Without their hydrolytic action no subsequent absorption could take place. Proteolytic enzymes Proteolytic activity in the digestive tract is usually abundant and mainly exceeds require- Table 4.1.2. The main enzymes in digestion, their origin, target of action and the result of their activity (end-products). NSP, non-starch polysaccharides. Enzyme Origin Target of action End-products Proteolytic - pepsin stomach protein peptides - trypsin pancreas protein peptides - chymotrypsin pancreas protein peptides - peptidases gut wall peptides amino acids Amylolytic pancreas starch glucose Lipolytic pancreas lipids fatty acids Cell wall degrading rumen/gutmicrobes cell walls NSP monomers Phytase feed, microbes phytate phosphorus 171 ment. As a result the capacity for protein digestion is generally not a limiting factor. Two exceptions seem possible. In the very young or newly-weaned animal, digestive organs such as the stomach and pancreas may not have adapted to the high supply or low digestibility of protein. In newly-weaned animals such as piglets, enzyme production drops temporarily and protein digestion often decreases for a while (Chappke et.al., 1989). Impedance of protein digestion by ANFs such as trypsin inhibitors, tannins or lectins is the second exception (Tamminga, 1990). Only the first situation can be eliminated by adding proteolytic enzymes, the second cannot, because ANFs may have the same negative effects towards exogenous proteolytic enzymes. Enzymes capable of degrading ANFs appear to be present in micro-organisms (both aerobic and anaerobic strains) and in germinating plants. It has been demonstrated that germination reduces the effect ofANF s (Savelkoul et al., 1992a). Lipolytic enzymes Limitations in fat digestion do not usually result from a lack of lipolytic capacity, but from a lack of emulsifying capacity. The addition of lipolytic enzymes would therefore be of very limited value. Amylolytic enzymes As with proteolytic enzymes, amylolytic enzymes may become limiting in the very young or newly-weaned animal. This problem appears to be more easily overcome by making the substrate more susceptible to enzymatic degradation and digestion than to supplement the substrate with exogenous enzymes. A practical but non-nutritional advantage of adding amylolytic enzymes is that it facilitates slurry feeding in pigs, because it reduces the water-binding capacity of the feed. Cell wall degrading enzymes Enzymes capable of hydrolysing non-starch polysaccharides (NSPs) including structural polysaccharides in plant cell walls, comprise cellulases, xylanases, pectinases, /?-glucosi- dases and yS-glucanases. Cell wall degrading enzymes are characterized by a wide variety of enzyme activities, which act synergistically (Beldman, 1986). Applying such enzymes in animal nutrition may have several advantages. One draw back is their considerable size, which may make it difficult or sometimes even impossible for them to penetrate pores in large complex structural carbohydrates. Measuring the activity of cell wall degrading enzymes against standard substrates may therefore produce misleading results.A combination ofchemica l (pre)treatment and enzyme treatment might also be beneficial. Cell wall degrading enzymes combining hemicellulolytic, cellulolytic and often other activities, have been applied as an aid to conserving high moisture feeds. Their anticipated advantage is the liberation of additional sugars for the preservating lactic acid bacteria, often resulting in an improvement of animal performance as a result ofpredigestio n of the plant cell wall in the silage. Studies on enzyme-treated silages have shown that compared with untreated silage, lactic and acetic acid increase and ammonia and pH decrease. The most significant effects 172 were found in silages based on immature low DM grass. When poorly fermentable grass is ensiled, the application ofenzyme s does not prevent butyric acid fermentation. Enzymes applied to such crops at commercial dosages, appear to liberate insufficient additional sugar during the onset of silage fermentation. Only a limited number of animal trials with enzyme-treated, low DM grass silages has been published (Spoelstra, 1991). Of 12 comparisons a non-significant improvement in digestibility by enzyme treatment was observed. Compared with untreated silage, only six out ofth e 24 experiments reported, showed a statistically significant increase in feed intake. In fact, compared with a formic acid treatment a decrease in intake occurred. It is concluded that in low DM silages, the rather undefined enzyme preparations currently existing, have only limited applicability as such, as well as in combination with starter cultures. Other applications are in the feeding of monogastric animals. Enzymes like /?-gluca- nases have already proven their value, particularly in poultry ifbarle y is an important feed ingredient (Classen & Bedford, 1991). Their efficacy can be explained in several ways. Viscosity resulting from glucans and impairing digestion may be eliminated. Disruption of large molecules encapsulating essential nutrients may also contribute to their positive effect. The liberation of monomers from structural carbohydrates may also provide extra energy. The potential advantage of the latter seems, however, largely restricted to the liberation of glucose and galactose, because other monomers such as xylose and arabinose are poorly utilized by pigs and poultry (Schutte, 1991). Research is currently being undertaken to study the digestive behaviour of NSP in different segments of the digestive tract in pigs and poultry and to study the effect of enzyme treatment on this behaviour. Phytase A promising example ofth e practical application ofenzyme s in animal nutrition is phytase, which hydrolyses phytate present in plant material and inorganic phosphorus (P). Phytate bound P is not available to monogastric animals because of a lack of endogenous phytase. Phytase of microbial origin (often from Aspergillus niger) added to the diet, was shown to be active while the feed was passing through the digestive tract, making phytate-P available to the animal Jongbloed et al., 1992). It was also shown that adding commercially availa ble microbiological phytase to the feed can largely, or completely, substitute for the addition of inorganic feed P to animal diets, thus contributing to a reduction in the emission of minerals to the environment by animal production. 4.1.4 Substrates Biotechnological conversions generally result from an interaction between enzymes (either direcdy or provided by micro-organisms) and substrate. The substrate may influence enzyme activity, to some extent, particularly when the enzymes are supplied by microbes. Microbial growth and competition may be influenced by characteristics of the substrate. For example it has been claimed that certain types of oligosaccharides have a positive effect on Bifidobacter, a micro-organism believed to have beneficial effects when present in 173 large quantities in the large intestine of humans. A second way in which substrates may be important isfo r instance when amino acids, such as lysine, threonine and tryptophan, essential for animal production, can be selec tively produced. 4.1.5 Active agents Many agents that are active against micro-organisms have been developed and are now being used in animal production. In the past the emphasis was on activity against patho gens, but antibiotics are being increasingly used to select a desired microbial ecosystem in one of the segments of the digestive tract. Examples are rumensin and avoparcin. Both result in positive production responses. The site of activity of rumensin is mainly in the rumen, whereas avoparcin isals o believed to be active at the intestinal level. Widespread introduction of active agents in animal production is hampered by consumer resistance against their application. Suchresistanc e isparticularl y directed towards active agents that interfere with intermediary metabolism, such as BST, PST and beta-agonists. 4.1.6 Conclusions Table 4.1.3 gives a summary ofth e potential ofth e different applications ofbiotechnolog y in animal nutrition, takingint o account their biological potential and practical acceptabil ity.I n animal nutrition, biotechnology ismainl y restricted to enzyme technology, either by the direct application of enzymes or through enzymes derived from added micro-organ isms. In the latter case, manipulation of enzyme activity is also possible through specific substrates enabling specific microbes to outgrow their competitors. A profitable applica tion of exogenous enzymes seems restricted to situations where there can be no simple development of a technological alternative. The combination of technological treatment and enzymatic treatment is, as yet, an under-explored area of research. Consumer resis tance will most likely prevent a positive role of active agents being developed and will favour the application ofharmles s micro-organisms and enzymes. Table 4.1.3. Application of biotechnology in animal nutrition. O, prospects nil; X, prospects very little; XX, prospects little; XXX, prospects good. Target feed digestive tract intermediarinl y metabolism Micro-organisms XXX XX O Enzymes XXX X o Substrates o XX o Active agents o XX X 174 4.2 Nutritional value and physiological effects of D-xylose and L-arabinose in poultry and pigs J.B. Schutte,J . deJong , P. van Leeuwen & M.W.A. Verstegen Non-starch polysaccharides (NSPs) can form a significant fraction of the carbohydrate content ofpractica l diets for pigs and poultry. These NSPs include a mixture of substances such as cellulose, hemicellulose, pectins and oligosaccharides, which contain hexose and pentose sugars and uronic acids. It is well known that NSPs are resistant to the digestive enzymes of farm animals. As a result, they pass to the hindgut where microbial degrada tion takes place. The end -products of microbial degradation ofNSPs , short chain organic acids, are readily absorbed and can be utilized by monogastric animals as an energy source, but with a lower efficiency than glucose (Agricultural Research Council, 1981;Jus t etal. 1983; van Es, 1987). There is conclusive evidence that the digestibility of NSPs can be improved by treat ment with enzymes which hydrolyse the NSPs to monosaccharides (Rexen, 1981; Ches- son, 1987; Schutte et al. 1990). The benefits of a hydrolysis of NSPs, however, are not determined solely by an improvement in digestibility, but also by the potential of the animal to utilize the products ofhydrolysis . In addition to glucose, other sugars will also be released by complete hydrolysis of NSPs in normal practical feed compositions. In quanti tative terms the pentose sugars D-xylose and L-arabinose are the most important ones. A review of the relevant literature has revealed that knowledge about the nutritional value of these sugars in pigs and poultry is incomplete (Schutte, 1991). 4.2.1 Digestion of non-starch polysaccharides A series of experiments was designed to obtain quantitative data on the digestion and utilization of D-xylose and L-arabinose in poultry and pigs. Feeding D-xylose and L-arabinose to pigs and poultry caused a series of physiological changes. These changes related to an increased production in volatile fatty acids and lactic acid at the end of the ileum, increased water retention in the ileal contents and in the faeces, and also to distention and increased weight of the caeca in chickens. In general, these changes are similar to those observed in rats fed with lactitol, lactose, xylitol and sorbitol, all of which are known to be poorly absorbed in the small intestine. The changes induced by D-xylose were less pronounced than those produced by L-arabinose. Data on water intake, dry matter content of excreta and caecal weight of birds fed on either D-glucose, D-xylose or L-arabinose are presented in Table 4.2.1. Ileal digestibility measurements in pigs and poultry demonstrated that L-arabinose was not absorbed completely from the small intestine. At a dietary level of 100 g/kg, approxi mately 25% of the ingested L-arabinose was recovered in the ileal chyme of poultry. In pigs this percentage amounted to approximately 33%. Ileal digestibility of D-xylose was found to be close to 100%. Thus, these results suggest that D-xylose and L-arabinose, in spite of their identical molecular size, have a different mode of absorption from the small intestine of monogastric animals. There were indica- 175 Table 4.2.1. Effect of dietary content of added D-glucose, D-xylose and L-arabinose on water intake, dry matter content of excreta and caecal weight of broiler chickens. Sugar Dietary content Water intake per Dry matter content Caecal weight ( (g/kg) chicken (g/d) of excreta (g/kg) D-glucose 75 166 279 5.5 D-xylose 25 192 226 5.6 50 206 203 no data 75 216 159 5.8 L-arabinose 25 207 213 7.2 50 229 191 no data 75 257 142 7.6 tions that both pentose sugars were partly absorbed as organic acids from the small intestine. These indications are based on the observed increase in ileal flow of organic acids in pigs fed on diets containing either D-xylose or L-arabinose. 4.2.2 Utilization of non-starch polysaccharides Both pentose sugars were partly excreted in the urine. The extent of this urinary excretion in percentage of intake, increased as the dietary inclusion ofeithe r D-xylose or L-arabinose was increased (Table 4.2.2). At equal dietary levels, more D-xylose was excreted into the urine from pigs than from poultry. In fact, the excretion of D-xylose and L-arabinose into the urine is higher than the figures presented in Table 4.2.2. There is conclusive evidence that both pentose sugars are also excreted in the urine in a form other than D-xylose or L-arabinose. Studies have shown that a proportion of an orally administered D-xylose is converted into a four-carbon polyol (D-threitol). L-arabinose appears partly to be con verted to L-arabitol and L-arabonic acid. These metabolites of both pentose sugars were Table 4.2.2. Urinary excretion of D-xylose and L-arabinose by poultry and pigs. Excretion as fraction ofdietar y levels. Dietary Urinary excretion(%) of content (g/kg) D-xylose L-arabinose poultry pigs poultry pigs 25 7.2 20.3 8.7 10.9 50 13.1 31.6 11.2 12.2 75 16.9 39.0 13.9 13.7 100 20.2 42.5 16.6 14.7 200 52.6 176 found to be excreted in the urine. There is no direct proof for the conversion of D-xylose and L-arabinose to glucose in monogastric animals. It appears that some D-xylose may be catabolized to carbon dioxide, but this pathway seems to be of no significance for L-arabinose (Segal & Foley, 1959). It is therefore concluded that the energy value of both pentose sugars mainly depends on their degree of fermentation. Taking into account the losses in energy arising from this fermentation process, it was estimated that the energy value of the two pentose sugars is approximately 25 to 35% ofthat ofD-glucose. 4.2.3 Conclusions In conclusion it can be stated, that both pentose sugars, D-xylose and L-arabinose, may provide only some energy to poultry and pigs. These sugars also induce an increase in water intake and, as a result, wet droppings in chicks. Considering these aspects, the benefits of a complete hydrolysis of NSPs fractions which release mainly these sugars (e.g. hemicellulose) are doubtful. Accordingly, it is not advisable to aim at a complete hydrolysis of all NSP fractions by enzyme supplementation. Literature data have shown that an incomplete enzymatic hy drolysis of NSPs into smaller polymers can also improve chick performance. In this case the improvements relate to an increased digestibility of dietary nitrogen and fat and to an elimination of the antinutritional effects of some NSP fractions. 4.3 Antinutritional factors (ANFs): aspects of the mode of action in the animal and inactivation of ANF-activity J. Huisman & F.H.M.G. Savelkoul The nutritional value of plant protein sources not only depends on the chemical composi tion, but also on the extent to which nutrients are digested, absorbed and utilized. Various factors can inhibit digestibility, absorption and utilization of nutrients. Two major classes of factors can be distinguished: a lack of appropriate enzymes present in the gastrointesti nal tract for optimal digestion and the presence of substances hampering optimal digestion, absorption or utilization of nutrients. The latter class of factors are called antinutritional factors (ANFs). ANFs are classified as non-fibrous natural substances which have negative effects on growth or on the health of man and animals. In this definition fibre is excluded because in human food, fibre may be classified as a positive health factor. Further, fibre may have some energy value when digested in the large intestine. Also excluded are mycotoxins because they are classified as contaminants, and any factors originating from processing (Yannai, 1980). Most ANFs offer the plant a natural protection against attacks of moulds, bacteria, insects and birds (Bond & Smith, 1989; Broadway et al., 1986; Etzler, 1986;Janzen et al., 1976; Ryan, 1983; Pistole, 1981). These reports show that the defensive effect of ANFs in 177 plants and seedsseem s tob e related to disturbances in digestiveprocesse s in farm animals, micro-organisms and insects. Because there are similarities in digestive processes in farm animals,micro-organism s and insects,t o a certain extent ANFs can be expected to disturb the digestive processes in farm animals in a similar way asthe y do inmicro-organism s and insects. It has been demonstrated that insects starve due to the inhibition of digestive enzymes after eatingleave s containingproteas e inhibitors (reviewed byRyan , 1978, 1983; Broadway et al., 1986; Birk, 1987). The inhibiting capacity ofa-amylas e inhibitors against amylases in bacteria, fungi and insects has been pointed out by Buonocore &Silan o (1986). Interactions ofplan t lectinswit hvariou smicro-organism s have been found (reviewed by Pistole, 1981). Moreover, a defensive function of plant lectins against fungi, bacteria and viruses has been postulated in'many studies (Etzler, 1986; Pistole, 1981). ANFs can be classified in different ways.I n the following scheme they are classified on the basis ofthei r effects on the nutritionalvalu e offeedstuff s and on the biological response in the animal: - factors with a depressive effect on protein digestion and on the utilization of protein (trypsin and chymotrypsin inhibitors, lectins, phenolic compounds, saponins); - factors with a negative effect on the digestion of carbohydrates (amylase inhibitors, phenolic compounds, flatulence factors); - factors with a negative effect on the utilization of minerals (glucosinolates, phytic acid, oxalic acid, gossypol); - factors that inactivate vitamins or cause an increase in the animals' vitamin require ment; - factors that stimulate the immune system (antigenic proteins). In thissectio n some aspectso fth e occurrence in seeds,th e mode ofactio n in the animals and inactivation ofANF-activit y are briefly discussed. 4.3.1 Occurrence of ANFs in seeds Recent reviews, reports and books on the presence and distribution of ANFs in various seeds have been presented by Cheeke & Shull (1985); Friedman (1986); Huisman & Jansman (1991); Liener (1980); Pusztai (1989), Rackis et al. (1986); Savage &De o (1989); Sissons &Tolma n (1991). Many seedscontai n several different ANFs (Chubb, 1982;Liener , 1981; Savag e &Deo , 1989). When such seeds are fed to animals the negative effects may be attributed to a combination of the effects ofvariou s ANFs. In Table 4.3.1 the ANFs in various seeds are listed. Table 4.3.1 shows that trypsin and chymotrypsin inhibitors and lectins are most impor tant in the legume seeds soya,pea s and beans, but some varieties ofry e and triticale may also contain moderate levels oftrypsi n inhibitors. Tannins are mainly present in sorghum, in some barley varieties, in the coloured flowering varieties of beans and peas and in rapeseed. Glucosinolates and sinapins are important in rapeseed. Alkaloids are present in lupins, whereas soya may be contaminated with alkaloids originating from Datura. Gos sypol ispresen t in cotton seed. Antigenicity ofprotein s isfoun d in soya, peas and beans. 178 Table 4.3.1. Antinutritional factors in cereals and seeds.- , below detection level; +, low level; ++, medium level; +++, high level. A, antigenic proteins; B,vicin/convicin ; c, alkaloids; , glucosinolates and sinapins; E, phenolic compounds (3-3.5%);F , gossypol; G, 16-18% in the shell around the nut. Cereals/seeds Antinutritionalfactors trypsin inhibitors lectins polyphenolic others compounds Cereal grains wheat, rice, maize -/+ rye -/+/++ triticale -/+/++ barley -/+ - -/+/++ - sorghum -/+ - +/++/+++ - Legume seeds soya ++/+++ ++ - ++/+++A-C Viciafaba bean + + +/++/+++ +/++/+++B Phaseolus vulgaris bean -/+/++ +/++/+++ +/++ +/++/+++A Pisum sativum +/++ +/++ +/++ - lentils, cowpeas, chickpeas +/++ +/++ -/+/++ - lupins - - - +/++/+++c Other seeds rapeseed - - +/++ +/++/+++D sunflower seed -/+ - +/++E - F cotton seed -/+ - - +/++/+++ peanut - - +/++G - 4.3.2 Major effects of ANFs in monogastric animals In Table 4.3.2 a survey is given of the major effects of ANFs on physiological processes in monogastric animals. 4.3.3 Inactivation of ANFs The nutritional value ofANF-containin g seeds can be increased by reducing ANF activity (van der Poel, 1989). There are various ways of doing this, through plant breeding, heat treatment, (bio)technological processes. The advantage of plant breeding is that no special treatments are needed to reduce ANF-activity. Examples of successful attempts are double zero varieties of rapeseed and low-tannin varieties of sorghum and faba beans. One disadvantage is that the absence ofANF s can give rise to more damage due to foraging by rabbits, deer and birds. Moreover, seed may be more susceptible to predators such as insects, viruses and bacteria, which evokes an increased use of pesticides and hence envi ronmental pollution. A point of discussion is therefore to what extent plant breeding can 179 Rhine water were very active and others were not. These observations demonstrate that the use of enzymes (ANF-ases) may have potential as an alternative to process technology in reducing ANF-activity. It is to be expected that continued research will, in the future, lead to further discoveries about enzymatic degradation of ANFs other than lectins. All in all biotechnological processes seem to be promising alternatives to plant breeding and heat treatment. On the other hand, there may be limits to the use of biotechnological processes, because in the case of glucosinolates and vicin/convicin the metabolites formed by enzymes may be toxic. 4.4 Phenolic monomers and digestibility of plant cell walls RJ. Hogendorp Phenolic compounds in plant cell walls are probably the most important factors limiting the digestion of forages and feed ingredients rich in structural carbohydrates by farm animals (Kerley et al.,1988) .Althoug h they are not classified as a real antinutritional factor (ANF, see Huisman & Savelkoul, section 4.3, this book), they certainly have a negative effect on nutritional value, and once solubilized in the gastrointestinal tract, especially of monogastrics, they may become an ANF in the true sense. Phenolic compounds that are covalently bound to cell walls can be subdivided into 'core'-lignin, which is made up of highly polymerized phenolic alcohols, and 'non-core'- lignin, which consists of monomeric phenolic groups. Because phenolic compounds hold key positions in the matrix structure of plant cell walls, their digestibility is evidendy important for animal performance. In contrast to previously held views on lignin, it has recently been postulated that the absolute amount ofligni n in plant cell walls isno t the predominant factor in determining forage digestibility (Jung, 1989). The manner and extent of the cross-linking of lignin to other structures appear to be equally important determining factors. However, information on the exact nature of the complexes formed between lignin and carbohydrate structures, such as hemicellulose and cellulose, is incomplete. 4.4.1 Location of phenolic acids Alkali treatment disrupts ester bonds thus releasing a variety of low molecular weight phenolic compounds from the cell walls. Some values of the contents of phenolic acids in animal feed products are given in Table 4.4.1. p-Coumaric acid (PCA) and ferulic acid (FA) (Fig. 4.4.1) are by far the most abundant phenolic monomers released. As well as being components of lignin, phenolic acids are also found in carbohydrate polymers, especially hemicellulose. Fragments of grass cell walls that have been isolated by chemical or enzymatic methods, have been found to contain a single phenolic acid (PCA or FA) linked to arabinose, which in turn islinke d to xylose monomers or xylose oligomers (Smith & Hartley, 1983). Dimers of phenolic acids can also be isolated after alkali treatment. They appear in two 182 Table 4.4.1. /î-Coumaric acid and ferulic acid content of dry matter of some feedstuffs in g/kg. (data: R.Hogendorp IWO-DLO). Sample Coumaric acid Ferulic acid Maize-gluten feed 0.38 5.70 Maize meal 0.14 2.22 Maize germs 0.46 4.30 Whole maize silage 16.25 8.35 Whole grass 0.56 2.46 Ensiled grass 0.63 2.30 Ensiled grass+enzyme 0.88 1.66 Soy meal solv.-extracted <0.09 <0.09 Palm kernel <0.10 0.32 classes, depending on whether they are dimerized by oxidative coupling (for example dehydrodiferulic acid), or by cyclodimerization of the C=C bond, resulting in a dimer with a cyclobutane ring. These dimers of phenolic acids probably cross-link different arabino- xylan polymers covalendy. Cross-linking fragments have recently been isolated and identi fied from bamboo shoot cell walls (Ishii, 1991). Calculations have shown that the numbers of dimers found can theoretically account for total cellwal l rigidity (Ford &Hartley , 1989). These findings indicate that phenolic monomers also play a key role in the structure of the plant cell wall and digestibility characteristics. Current research is directed towards revealing more detailed information on the distribution of phenolic acids in plant cell walls, with the ultimate aim of eliminating them and, as a result, improving the digestibility of animal feed. R3 O // — C =: C — c D> \ OH R/ RI R2 R3 Compound H OH H p-coumaric acid OCH, OH H ferulic acid Fig. 4.4.1. Molecule structure of/>-coumaric and ferulic acid, the most abundant phenolic acids in plant cell walls. 183 4.4.2 Digestion process Because the use of forages for animal feed is largely restricted to ruminants, information on the interactions between rumen microbes and phenolic substances isnecessar y in order to clarify the detrimental effects ofphenoli c compounds in forages (Akin et al.,1988) . While plant cell walls are degraded by rumen micro-organisms, soluble phenolic-carbohydrate complexes are released into solution. Griggs et al.(1989 ) purified several fractions of phenolic-carbohydrate complexes from maize after partial digestion with cellulase, and found one of these fractions to inhibit cell wall digestion by rumen micro-organisms in vitro. Because only minor amounts of these complexes are found in the rumen, an active detoxification process apparently exists (Jung & Fahey, 1983). A fundamental question which still needs to be answered is: 'What makes the indigesti ble fraction indigestible?'. Chemical analysis of digested feed has revealed that it is hardly different from the original material, except for a significant increase in the content of phenolic compounds. Chesson (1988) postulated that during digestion of feed, phenolic compounds accumulate on the outside of partially digested feed particles. Once a critical concentration has been reached, the phenolic units will tend to polymerize, thus forming an indigestible layer. Therefore, compounds inside a partially digested feed particle, being readily digestible as such, cannot be reached by microbial enzymes. Work in our laboratory has revealed that disappearance kinetics during the breakdown of feed particles is different for PCA and FA. When the alkali-labile phenolic acid content was determined on residues of maize silages, which were incubated in nylon bags in the rumen of dairy cows, it was shown (Fig. 4.4.2) that PCA is more resistant to digestive UL Ul 100, K Ul 80 < 60 z " "——^__T a J- - ± ec "" - - _ _ o 40 -~ —— - __ ^ ~- i"* • Ul •-— <3 < 20 Ul Ü ec 1 I i ui a. 100 200 300 400 INCUBATION PERIOD (h) PCA —•— FA - I NDF Fig. 4.4.2. Disappearance kinetics of PCA, FA and cell walls (expressed as NDF fraction) from maize silages (N = 5) during incubation of the samples in nylon bags in the rumen of dairy cows. 184 solubilization and FA is readily liberated with respect to cell wall disappearance as a whole. High correlation coefficients are found ifth e phenolic acid content ofth e residues is related to the percentage of original cell wall material still present (Fig. 4.4.3). Our data indicate that PCA is more frequently found in lignified tissues than ferulic acid, the latter being predominandy situated in carbohydrate-rich regions of the plant cell wall. Monitoring the relative rates of disappearance of individual phenolic acids in time is an essential tool for evaluating the control of plant cell wall digestion. 4.4.3 Perspectives for biotechnology At least two distinct biotechnological approaches are possible when dealing with the negative effects of phenolic monomers. The first one is the molecular biological approach to the plant material itself. Manipulation of phenolic biosynthesis appears to be a straight forward way of improving digestibility. Molecular biological research on the so-called Brown-Midrib-Mutants cultivars ofmaize , pearl millet and sorghum (Cherney et al. 1991), which have been produced by traditional selection breeding, has revealed that relatively simple mutations in the lignin biosynthetic pathways can have a considerable effect on digestibility. One of the main characteristics of these mutants is the structurally lower content of ester-bound phenolic acids, together with a significant increase in cell wall digestibility. The second approach isth e (biotechnological) production of specific enzymes which can be added to the feed or to the animal. In ruminants, manipulation of rumen bacteria, leading to an increase in the production of specific enzymes, might have some perspective (Armstrong & Gilbert, 1991), although any success would be highly speculative. Specific 40 < < Ü ••"• a 30 % . . R2 - 0,91 z LU S Q 20 - a • £ • < 10 R2 - 0,54 ... " -% *- . W T • < -M my •% • • • • • U a R2 -~Tft&**~-*t- ' i i . i 20 40 60 80 100 CELL WALL RESIDU (% OF ORIGINAL AMOUNT) • PCA • FA • PCA/FA Fig. 4.4.3. Correlation coefficients of the PCA, FA, PCA/FA ratio, and percentage of cell walls still present in the residue of msacco incubations of maize silage (N = 5). 185 p-coumaric acid-de-esterase and ferulic acid-de-esterase have recently been isolated from fungi and subsequently characterized (Borneman et al., 1991). So far these enzymes have only been found to exhibit substantial activity towards model substrates and hardly any towards native plant cell walls. Whether these enzymes will be capable of increasing overall digestibility remains to be seen. It has recently been claimed that some plants produce the de-esterases themselves in rather substantial amounts. This might open the way to manipulate the number of enzymes produced by the plant by, for example, introducing the genes responsible for enzyme synthesis into new species. 4.4.4 Conclusions Plant fibre is an important constituent of animal feed. The value of forage for animal feed is mainly limited by the degradability of the plant fibre. Many studies indicate that the presence of monomeric phenolics is an important constraint to the digestion of plant cell walls, because these compounds are located at strategic cross-linking places within the cell wall matrix. Biotechnology offers possibilities for improving digestibility of the plant cell wall. Ma nipulating lignin or phenolics biosynthesis seems the most promising strategy. The biotech- nological production of specific enzymes to be added to feed is also the subject of current research, although accessibility of the cross-linking spots within plant cell walls will impose an intrinsic limit on the maximally achievable improvement. 4.5 Prospects of feed additives produced by biotechnology J.L. van Os Novel feed additives are an important target of research and development in animal feed biotechnology. Such additives include products produced by industrial fermentation, as well as industrially produced micro-organisms. The prospects of commercial feed additives produced by biotechnology depend on the usual requirements for industrial product development, such as a fair chance of technolog ical and commercial success in combination with an acceptable cost-benefit ratio. Such aspects should therefore be included when the outlook of such products is evaluated. 4.5.1 Technological success To be a technological success, a product should at least show a significant effect under a variety of conditions, naturally without undesirable side-effects. Although this seems obvi ous, new products frequendy fail to meet this criterium sufficiently to justify their wide spread use, for instance because their beneficial effect is too dependent on specific condi tions. Furthermore, there should be a fair margin between the added value of the effect and the final cost price of production. For products such as enzymes a cost price can be 186 calculated beforehand. The cost price, however, could limit their use to applications which have significant added values. This means that any development of a feed additive should be based on a nutritional concept with sufficient scope tojustif y the cost of development. Although such a concept could originate completely from industrial research, this is the exception rather than the rule. Most significant developments emanate from an interaction between basic research at universities or institutes with industrial research. Any hypothesis for a potential nutritional concept should be followed by testing its validity. If positive, a product concept can be developed and tested. All this can be done by independent scientists on their own account, or continued as precompetitive research, ifbroade r industrial funding is needed. However, bilateral cooperation at an early stage is usually more fruitful in view of the patent protection that must be established in order to justify future large-scale industrial investment. One area of independent research that is highly relevant to the prospects of major new products is the development of appropriate methods of analysing desired effects. Perhaps even more crucial is the development of predictive in vivo models. Most relevant are those in which only minimal quantities of a test substance is necessary for screening, in view of the limited availability of such test substances in the early stage of development. 4.5.2 Commercial success To be a commercial success, most major new products, in view of the extensive investment in their development, require a large, if not a global market. They also usually need a sufficiently broad area of application, i.e. the number of species or categories of animals for which they can be used. Not uncommonly, the economic scope of the type of feed for which the product is of benefit, is a further decisive factor. For many new products there are alternative approaches, which already limits the use of such products. Government approval is an ever-increasing barrier, or at least a delaying factor, for the application of many new feed additives. Numerous registration guidelines must be adapted to the new developments in molecular biology, at a time when the EC is harmonizing regulations of the member states. Approval might also interfere with public acceptance, which could be decisive for commercial success or failure. Bovine Somatotropin (BST) is the most prominent example. In fact, this case has had a negative influence on the public acceptance of biotechnological products in general. However, even when all these aspects are taken into account, feed additives produced by biotechnology will certainly play an increasing role in animal nutrition. 4.5.3 Fermentation products for the feed industry The three major categories of fermentation products, relevant to the animal feed industry, are: (1) amino acids, (2) enzymes and (3) somatotropins such as BST and PST Only the first two groups can be used as an additive. Interestingly, the major industrial research for these three groups started in different parts of the world: the amino acids inJapan , the enzymes in Europe and the somatotropins in the USA. 187 Amino acids Amino acids, such as lysine and threonine and, to a lesser extent, tryptophan, are examples of completely developed products, the price of which is, or will further be, reduced by biotechnological developments. The amino acid composition of feed proteins often has a negative influence on nutritional value. Adding limiting amino acids to pig and poultry diets can increase protein utilisation, and hence reduce nitrogen excretion (Schutte, 1989). In areas with a surplus of nitrogen caused by animal production, the need for such an addition is obvious. So far, their large-scale use has been hampered by feed economics. In view of the ratio between their price and the added value for the environment one would expect a greater use of such products. However, both the increasing environmental pres sure as well as a growing overcapacity of amino acid production will further stimulate their use. Enzymes Enzymes can be used in feed to improve the economic utilisation of feed ingredients, or for other considerations such as health or environmental concerns. For a long time, virtually all the enzyme preparations tested for feed applications were developed for purposes other than their addition iOanima l feeds. Usually, such mixtures of enzymatic activities showed less conclusive results or a poor price performance in feed applications. However, the number of enzyme products, specifically formulated and marketed for feed applications, is increasing. Intensive testing of such products has led to better recom mendations for the restricted conditions under which consistent results can be achieved. Of the enzyme mixtures produced in the classical way, so far/?-glucanas e has been the only example of an activity with an acceptable price/performance ratio, especially in barley- based poultry diets. A substantial further extension in the use of classical enzymes in feed seems unlikely because of the small margin between the cost-price and the added value. In recent years a completely new approach has started. The screening and identification of required enzyme activities isfollowe d by using advanced biotechnological knowledge to increase the yield of such activities. This can be done to such an extent, that their use as feed additives becomes economically feasible. The development of a product based upon the enzyme phytase is one of the first examples of such an approach. The nutritional concept of this product was investigated some twenty years ago in the USA. This study indicated that microbial phytase produced by Aspergillusspecies , was able to liberate phosphorus from phytate, when added to the diet of chickens (Nelson et al., 1971). Such an addition could replace, at least partially, the normal required addition of inorganic feed phosphate to pig and poultry diets, and hence reduce the excretion of phosphorus (P) in manure. A considerable contribution was there fore to be expected of such an approach towards solving the P pollution problems in areas with intensive livestock farming. This expectation led to the unique cooperation between the Dutch Community Board of Animal Feed, several specialist research institutes and a biotechnological industry (Simons et al., 1990), which resulted in the introduction of a commercial phytase product (Beudeker et al., 1990). The new approach of developing enzymes with a highly specific activity will certainly lead to an increasing use of feed enzymes. It should be realised, however, that further 188 projects willlac k the unique combination of aspects which stimulated the phytase projects: an existing nutritional concept that was expected to contribute towards solving a major environmental problem in the Netherlands, leading to a unique cooperation between most of the expertise available at national level. In order to develop a clear new nutritional concept, knowledge generally has to be gathered in three areas: that of the animals' digestive physiology, that of the major feed substrates and their metabolites, and that of the specific enzyme activities which could cause a significant effect. In spite of the lacking knowledge, an endo-xylanase activity from Aspergillus nigerha s been identified, with such nutritional effects when added to wheat-based poultry diets, that the further development of a commercial product is justified (van Paridon et al., 1992). With the steadily growing knowledge in the three areas mentioned previously, further relevant nutritional concepts, and thus new enzyme products, are likely to follow. Growing knowledge on antinutritional factors (ANFs) must lead to products that are able to prevent or reduce some of the effects of ANFs. The present state of the art, however, does not justify short-term expectations. The negative effects are more well known than the underlying mechanisms, and better analytical methods are needed to investigate these mechanisms. A clear product concept is, therefore, still lacking. To evaluate the potential of a product a better quantification of the real damage is needed, in combination with more evidence on the relevance of effects of Phaseolus and other legumi nous seeds. In addition, any new product reducing the effects of ANFs must compete with technological alternatives, such as heating. 4.5.4 Micro-organisms in animal feeding Micro-organisms currently in use, or under development, as feed additives include bacte ria, fungi and yeast cells. There are three major areas of interest: (1) positive effects on the intestinal flora of monogastric animals, (2) effects on the ruminai flora in cattle, and (3) the use of micro-organisms as starters, mainly in silage. Probiotics For many years, various species of bacteria or bacterial mixtures have been tested, or even used in practice, as probiotics, to produce a positive effect on the intestinal flora. In spite of numerous publications showing positive effects, most field results were insufficiently consistent or relevant to allow a real breakthrough for this type of products. In fact, for several reasons such a breakthrough is unlikely to occur, at least not in the near future. Good models to test probiotics, although in development, are hardly available. The basic knowledge of the intestinal flora is insufficient. It is uncertain, for instance, whether colonisation of the added micro-organisms is feasible. Moreover, it is unknown to what extent the properties of bacteria remain intact in the intestinal flora. Finally, herd specific, if not animal specific, characteristics of the intestinal flora will probably prevent any widespread use of one particular product. The future of probiotics therefore seems at best unclear, as is that of other approaches, such as specific substrates that are able to stimulate the positive part of the intestinal flora (Fishbein et al, 1988). 189 The effects of yeasts and fungi on the ruminai flora of cattle seem to be more consistent, although the area of application appears to be limited by the type of ration (Williams & Newbold, 1990). A reasonably predictive model seems to be available for ruminai fermen tation. This could help increase the knowledge on the mechanisms of action under varying conditions, allowing the broadening of applicability. An acceptable ratio between cost- price and added value forms a further argument for expecting the possibilities for this type of product to increase. Starters in silage offorage Bacteria, especially lactobacilli, are widely used as starters in silage of forage. Their aim is to rapidly reach the low pH needed for good preservation. Although new silage products are being developed, including combinations of enzymes and bacterial starters, many cheaper substitutes are available. Therefore, optimization rather than real innovation seems to be the prospect for this area. Other applications Although somewhat of a side-issue, it seems relevant to give some attention to the way in which fermentation processes as such may contribute to animal production. Increasing environmental concern is interfering with the disposal of agricultural by-products. The recycling of ingredients is one of the most acceptable solutions, and fermentation is a promising technology for reaching this goal. A novel fermentation process to produce amino acids from manure is a good example of this new approach (Sanders, 1991). 4.5.5 Conclusions It may be concluded that although the rate of most biotechnological developments has been overestimated, major contributions of biotechnology to the animal feed industry are still to be expected. Delaying factors are: (1) insufficient basic knowledge and (2) lack of analytical methodology (for example lack of predictive in vivo models). Consumer accep tance of new approaches iso fgrowin g importance. Finally, the potential impact of fermen tation for the processing and recycling of agricultural by-products seems to be somewhat underestimated. 4.6 References Agricultural Research Council, 1981. The nutrient requirements of pigs. Slough: Commonwealth Agricultural Bureaux. Akin, D.E., L.L. Rigsby, M.K. Theodorou &R.D . Hartley, 1988. Population changes of fibrolytic rumen bacteria in the presence of phenolic acids and plant extracts. Animal Feed Science and Technolology 19:261-275. Armstrong,D.G . &H.J. Gilbert, 1991.Th e application ofbiotechnolog yfor futur e livestockproduc tion. In: T. Tsuda, Y. Sasaki & R. Kawashima (Eds.): Physiological aspects of digestion and metabolism in ruminants. Academic Press, San Diego, pp. 737-761. Beldman, G., 1986. The cellulases of Trichoderma viride. Mode of action and application in biomass conversion. Ph.D. Thesis, Agricultural University, Wageningen, 109pp . 190 Beudeker, R.F., C. Geerse & GJ. Verschoor, 1990. Biotechnological products for compound feed industry. In K. North (Ed.): Biotechnology International 1990, pp. 340-343. Century Press Ltd., London 1991. Birk, Y., 1987. Proteinase inhibitors. In: A. Neuberger & Kr. Brocklehurst (Eds): Hydrolytic en zymes. Elsevier, Amsterdam, Netherlands, 257-350. Bond, D.A. & D.B. Smith, 1989. Possibilities for the prediction of antinutritional factors in grain legumes by breeding. In:J . Huisman, A.F.B, van der Poel & I.E. Liener (Eds.): Recent advances of research on antinutritional factors in legume seeds. Proceedings of the first international workshop on antinutritional factors (ANF) in legume seeds, Wageningen, Netherlands. Pudoc, Wageningen, pp. 285-296. Borneman, W.S., L.G. Ljungdahl, R.D. Hartley & D.E. Akin, 1991. Isolation and characterization of p-coumaroyl esterase from anaerobic fungus Neocallimastrix strain MC-2. Applied and Environ mental Microbiology 57:2337-2344. Broadway, R.M., S.S. Duffey, G. Pearce & C.A. Ryan, 1986. Regulation of proteinase inhibitors: A defence against herbivorous insects? Entomologica Experimentalis et applicata, 41:33-38. Buonocore, V. & V. Silano, 1986. Biochemical, nutritional and toxicological aspects of alpha- amylase inhibitors from plant food. In: M. Friedman (Ed.): Nutritional and Toxicological signifi cance of enzyme inhibitors in foods. Plenum Press, New York, USA, pp. 483-507. Chappke, R.P., J.A. Cuaron & R.A. Easter, 1989. Temporal changes in carbohydrate digestive capacity and growth rate of piglets in response to glucocorticoid administration and weaning age. Journal of Animal Science 67:2985-2995. Cherney, J.H., DJ.R. Cherney, D.E. Akin &J.D. Axtell, 1991. Potential of brown-midrib, low- lignin mutants for improving forage quality. Advances in Agronomy 46:157-198. Chesson, A., 1987. Supplementary enzymes to improve the utilization of pig and poultry diets. In: W. Haresign & DJ.A . Cole (Eds.): Recent advances in animal nutrition. London, Butterworths. pp. 71-90. Chesson, A., 1988. Lignin-polysaccharide complexes of the plant cell wall and their effect on microbial degradation in the rumen. Animal Feed Science Technology 21:219-228. Chubb, L.G., 1982. Antinutritive factors in animal feedstuffs. In: W. Haresign (Ed.): Recent ad vances in animal nutrition. Butterworths, London, 21-37. Classen, H.L. & M.R. Bedford, 1991. The use of enzymes to improve the nutritive value of poultry feeds. In: H. Haresign and D.J.A . Cole (Eds.): Recent advances in animal nutrition 1991. Butter- worth/Heinemann, Oxford, pp. 95-116. Etzler, M. & M.L. Branstrator, 1974. Differential localization of cell-surface and secretory compo nents in rat intestinal epithelium by use of lectins. The Journal of Cell Biology 62:329-343. Fishbein, L., M Kaplan & M. Gough, 1988. Fructo-oligosaccharides: a review. Veterinary and Human Toxicology 30(2):104-107. Fonty, G. & K.N. Joblin, 1991. Rumen anaerobic fungi: Their role and interaction with other rumen micro-organisms in relation to fiber digestion. In: Physiological aspects of digestion and metabolism in ruminants. Academic Press, San Diego, pp. 655-680. Ford, C.W. & R.D. Hartley, 1989. GC/MS characterisation of cyclodimers from p-coumaric and ferulic acids by photodimerisation - a possible factor influencing cell wall biodegradability. Jour nal of the Science of Food and Agriculture 46:301-310. Friedman, M., 1986. Nutritional and toxicological significance of enzyme inhibitors in foods. Ple num Press, New York, 572 pp. Griggs, T.C.,J.H. Cherney, DJ.R. Cherney, &J.A. Patterson, 1989. Phenolic-carbohydrate com plexes in maize cell walls and influences on rumen microbes. Agronomy Abstracts pp. 169. Huisman,J . & AJ.M. Jansman, 1991. Dietary effects and some analytical aspects of antinutritonal factors in Pisum sativum, common beans and soyabeans (Glycinemax L.) in monogastric farm animals. A literature review. Nutrition Abstracts and Reviews (series B), vol. 61, 12:901-921. Huisman, J. & G.H. Tolman, 1992. Antinutritional factors in the plant proteins of diets for non- ruminants. In: P.C. Garnsworthy, W. Haresign & DJ.A. Cole (Eds.): Recent advances in animal 191 nutrition. Proceedings of the 26th University of Nottingham feed manufacturers conference, Nottingham 1992. Butterworth/Heinemann, Oxford, U.K., pp. 3-31. Ishii, T., 1991. Isolation and characterisation of a diferuloyl arabinoxylan hexasaccharide from bamboo shoot cell walls. Carbohydrate Research 219:15-22. Janzen, D.H., H.B.Juste r & I.E. Liener, 1976. Insecticidal action of the phytohemaglutinin in black beans on bruchid beetle. Science 192:795-796. Jongbloed, A.W., Z. Mroz, & P.A. Kemme, 1992. The effect of supplementary Aspergillus niger phytase in diets for pigs on concentration and apparent digestibility of dry matter, total phospho rus, and phytic acid in different sections of the alimentary canal. Journal of Animal Science 70:1159-1168. Jung, H.G. & G.C. Faheyjr., 1983. Nutritional implications of phenolic monomers and lignin: a review.Journa l of Animal Science 57:206-219. Jung, H.G., 1989. Forage lignins and their effects on fibre digestibility. Agron.J. 81:33-38. Just, A.,J.A . Fernandez & H.Jôrgensen , 1983. The net energy value of diets for growth in pigs in relation to the fermentative processes in the digestive tract and the site of absorption of the nutrients. Livestock Production Science 10:171-186. Kerley, M.S., G.C. Faheyjr., J.M. Gould & E.L. Iannotti, 1988. Effects of lignification, cellulose crystallinity and enzyme accessible space on the digestibility ofplan t cell wall carbohydrates by the ruminant. Food Microstructure 7:59-65. Liener, I.E., 1980. Toxic constituents of plant foodstuffs. Academic Press, New York, 502 pp. Liener, I.E., 1981. Factors affecting the nutritional quality of soya products.Journa l of the American Ou Chemists' Society 58, 3:406-415. Martin, S.A. &D.J . Nisbet, 1992. Effect of direct fed microbials and rumen microbial fermentation. Journal of Dairy Science 75:1736-1744. Nelson, T.S., T.R. Shieh, R.J. Wodzinsky &J.H. Ware, 1971. Effect of supplemental phytase on the utilization of phytate phosphorus by chicks.Journa l of Nutrition 101:1289. Paridon, P.A. van, J.C.P. Boonman, G.C.M. Selten, C. Geerse, D. Barug, P.H.M. de Bot & G. Hemke, 1992. The application of fungal endoxylanase in poultry diets. In:J . Visser, G. Beldman, M.A. Kusters-van Someren & A.G.J. Voragen (Eds.): Xylans and Xylanases, Proceedings of an International Symposium, Wageningen 8-11 December 1991. Elsevier Amsterdam 1992, Pro gress in Biotechnology 7. Pistole, T.G., 1981. Interaction of bacteria and fungi with lectins and lectin-like substances. Annual Review Microbiology 35:85-112. Pusztai, A., 1989. Biological effects of dietary lectins: In: J. Huisman, A.F.B, van der Poel & I.E. Liener (Eds): Recent advances of research in antinutritional factors in legume seeds. Pudoc, Wageningen, Netherlands, 17-29. Rackis,J.J. , WJ. Wolf & E.C. Baker, 1986. Protease inhibitors in plant foods: content and inactiva- tion. In: M. Friedman (Ed): Nutritional and toxicological significance of enzyme inhibitors in foods. Plenum Press, New York, 299-347. Rexen, B., 1981. Use of enzymes for improvement of feed. Animal Feed Science and Technology 6:105-114 Ryan, C.A., 1978. Protease inhibitors in plant leaves; Biochemical model for natural plant protec tion. Trends in Biochemical Sciences 5:148-150. Ryan, C.A., 1983. Insect-induced chemical signals regulating natural plant protection responses. In: R.F. Denno & M.S. McClure (Eds.): Variable plant and herbivores in natural and managed systems. Academic Press, New York, 43-60. Sanders, J.P.M., 1992. Verwerking varkensdrijfmest middels de produktie van essentiële ami nozuren, in het bijzonder lysine. Proceedings Symposium Mestverwerking, Veenendaal, Nether lands, 15 April 1992. Informatie Centrum Mestverwerking, P.O. Box 43, 6700 AA Wageningen. Savage, G.P. & S. Deo, 1989. The nutritional value of peas (Pisum sativum). A literature review. Nutrition Abstracts and Reviews (series A), vol. 59, 2:66-68. Savelkoul, F.H.M.G., A.F.B. van der Poel & S. Tamminga, 1992a. The presence and inactivation 192 of trypsin inhibitors, tannins, lectins and amylase inhibitors in legume seeds during germination. A review. Plant Foods for Human Nutrition 42:71-85. Savelkoul, H.M.G., H. Boer, S. Tamminga & M.G. van Oort, 1992b. Biotechnological degradation oflectins, tannins and trypsin inhibitors in legumes. In: Proceedings of the 1st European confer ence on grain Legumes, Angers, France, 1992. Schutte,J.B. , 1989. Practical application of (bio) synthetic amino acids in poultry in pig diets. In: E J. van Weerden & J. Huisman (Eds.): Nutrition and Digestive Physiology in Monogastric Farm Animals, pp 75 - 88. Pudoc, Wageningen. Schutte,J.B. , GJ.M. van Kempen & RJ. Hamer, 1990. Possibilities to improve the utilization of feed ingredients rich in non-starch polysaccharides for poultry. In: Proc. VIII. European poultry conference, Feria de Barcelona Avda. Reina M. Cristina, s/n. pp. 128-135. Schutte, J.B., 1991. Nutritional value and physiological effects of D-xylose and L-arabinose in poultry and pigs. PhD Thesis Agricultural University, Wageningen, 173 pp. Segal, S. &J.B. Foley, 1959. The metabolic fate of C14 labelled pentoses in man.Journa l of clinical investigations 38:407-413. Simons, P.C.M., H.AJ. Versteegh, A.W. Jongbloed, P.A. Kemme, P. Slump, K.D. Bos, M.G.E. Wolters, R.F. Beudeker & G.J. Verschoor, 1990.Improvement of phosphorus availability by microbial phytase in broilers and pigs. Br.Journa l of Nutrition 64:525. Sissons,J.W & G.H. Tolman, 1991. Anti-nutritional properties of soybean antigens in calves. In: J.P.F.D. Mello & C.M. DufTus (Eds.): Toxic factors in crop plants. Proceedings of the second spring conference Edinburgh, pp. 62-85. Smith, M.M. & R.D. Hartley, 1983. Occurrence and nature of ferulic acid substitution of cell wall polysaccharides in graminaceous plants. Carbohydrate Research 118:65-80. Spoelstra, S.F., A. Steg &J.M.W. Beuvink, 1990. Animal feed biotechnology: Application of cell wall degrading enzymes. In:J.J . Dekkers, H.C. van der Plas & D.H. Vuijk (Eds.):Agricultura l biotech nology in focus in the Netherlands. Pudoc, Wageningen pp. 165-171. Spoelstra, S.F., 1991. Chemical and biological additives in forage conservation. In: Forage conserva tion towards 2000 (G. Pahlow and H. Honig, eds.). Landbauforschung Völkenrode, Sonderheft 123:48-70. Tamminga, S., 1990. Biotechnology and improvement of animal nutrition. Med. Fac. Landb. Rijksuniv. Gent, 55:1373-1382. Tamminga, S., A.F.B. van der Poel, F.H.M.G. Savelkoul, J.B. Schutte & S.F. Spoelstra, 1990. Animal feed biotechnology: Improved utilization of protein and carbohydrates. In:J J. Dekkers, H.C. van der Plas & D.H. Vuijk (Eds.): Agricultural biotechnology in focus in the Netherlands. Pudoc, Wageningen pp. 155-164. Van Belle, M., E. Teller & M. Focant, 1990. Probiotics in animal nutrition: A review. Archives Animal Nutrition 7:543-567. van der Poel, A.F.B., 1989. Effects of processing on antinutritional factors (ANF) and nutritional value oflegum e seeds for non-ruminant feeding. In:J . Huisman, A.F.B.va n der Poel & I.E. Liener (Eds.): Recent advances of research in antinutritional factors in legume seeds. Proceedings of the first international workshop on antinutritional factors (ANF) in legume seeds, Wageningen, Neth erlands, 1988. Pudoc, Wageningen, pp. 213-229. van der Poel, A.F.B., 1990. Effects of processing on bean (Phaseolus vulgaris L). Protein quality. PhD-thesis, Agricultural University, Wageningen, The Netherlands, 160 pp. van Es, A.J.H., 1987. Energy utilization of low digestibility carbohydrates. In: D.C. Leegwater, VJ. Feron & RJJ. Hermus (Eds.): Low digestibility carbohydrates, Pudoc, Wageningen, pp. 121-127. Williams, P.E.V. & C.J. Newbold, 1990. Rumen probiosis: The effect of novel micro-organisms on rumen fermentation and ruminant productivity. In: W. Haresign & D.J.A. Cole (Eds.): Recent advances in animal nutrition. Butterworths, London. Williams, P.E.V, A. Walker &J.C. MacRae, 1990. Rumen probiosis: the effects of addition of yeast culture (viable yeast, Saccharomyces cereviciae, plus growth medium) on duodenal protein flow in wether sheep. Proceedings of the Nutrition Society 49:128A. 193 5 Biotechnology in the food industry Contents 5.1 General introduction to biotechnology in the food industry 197 5.1.1 Food manufacturing: history and needs 197 5.1.2 Potential of biotechnology in food manufacturing 197 5.1.3 Biotechnological research for the food industry in the Netherlands 201 5.2 Improving of the baking quality of gluten by enzymatic modification 201 5.2.1 Modification of structural and functional properties 202 5.2.2 Applications of enzymatic modification: quality control 203 5.2.3 Conclusions 206 5.3 Enzymatic hydrolysis of milk proteins: some fundamental and practical aspects 207 5.3.1 Enzymatic hydrolysis of caseins 208 5.3.2 Enzymatic hydrolysis of whey proteins 210 5.3.3 Concluding remark 213 5.4 Plant cell wall degrading enzymes in the processing of food 213 5.4.1 Fruits and vegetables 214 5.4.2 Cereals 216 5.4.3 By-products 219 5.4.4 Conclusions 220 5.5 Enzymatic processing of triglycerides and fatty acids. Enhancing reaction kinetics by continuous product removal 220 5.5.1 Specific hydrolysis of tricaprylin 221 5.5.2 Modelling 222 5.5.3 Applications 225 5.5.4 Conclusions 226 5.6 Thermodynamic principles of enzymatic acylglycerol synthesis 226 5.6.1 Theory 227 5.6.2 Results 227 5.6.3 Conclusions 231 195 5.7 Engineering lactic acid bacteria for improved food fermentations 231 5.7.1 Genetic tools 231 5.7.2 Biosafety considerations and food-grade selection systems 232 5.7.3 Phage resistant strains 233 5.7.4 Strains with improved proteolytic capacity 233 5.7.5 Improved production of diacetyl 234 5.7.6 Conclusions 235 5.8 Engineering and application of nisin, a natural food preservative 236 5.8.1 Structure and functional properties of nisin 236 5.8.2 Improvement of nisin by protein engineering 238 5.8.3 Development of nisin-producing cheese starter cultures 240 5.9 Molecular genetic approaches in plant cell wall degradation 242 5.9.1 Gene regulation 242 5.9.2 Influencing the composition of enzyme preparations 243 5.9.3 Towards enzyme engineering 245 5.9.4 Outlook 245 5.10 Detection and identification of food-borne bacteria by means of molecular probes 245 5.10.1 Nucleic acid hybridization probes 246 5.10.2 Specific applications of nucleic acid probes for the dairy industry 248 5.10.3 Conclusions 252 5.11 The development of immunoassays for the rapid detection of moulds 254 5.11.1 Development of immunoassays based on detection ofantigeni c extracellu lar polysaccharides 254 5.11.2 Specificity and practical use of immunoassays for mould detection 256 5.11.3 Laboratory analysis of foods using immunoassays 256 5.11.4 Analysis of fungal contamination with commercially available immunoas says 261 5.11.5 Future developments 261 5.12 Safety aspects of foods produced by recombinant DNA technology 262 5.12.1 The use of recombinant DNA technology in food biotechnology 263 5.12.2 Food safety aspects with regard to acceptance 264 5.12.3 Concluding remarks 266 5.13 References 266 196 5.1 General introduction to biotechnology in the food industry W.IJ. Aalbersberg, K. van 't Riet & A.G.J. Voragen 5.1.1 Food manufacturing: history and needs Food manufacturing in the Netherlands As long ago as the sixteenth century the Netherlands was a major producer of foods that had a long shelf-life. In those days fishing and animal husbandry were important activities. Accordingly, cured herring and Edam and Leyden cheeses were among the staple food products. Initially, these foods were for home consumption and were only exported to neighbouring countries. During the seventeenth century, when the Dutch share in global sailing increased, the staple food products were consumed by sailors during their long voyages and sold to tradesmen practically all over the world. In later years, the central position of food processing in the Netherlands further increased as raw materials from abroad, such as sugarcane, pepper, cinnamon and rice were taken home for further processing. Today in the Netherlands food processing is still an important industrial activity with a share in the national income of approximately 4%. In 1991 food processing in the Nether lands accounted for approximately 160 000 labour years, which is equivalent to 17% of the industrial labour activity and 3.2% of the total labour activity. In 1991, the export of processed foods amounted to 42.1 billion Dutch guilders, whereas the total agricultural export amounted to 61.6 million Dutch guilders. The spread in the use of agricultural raw materials and semi-finished goods over the various food processing sectors in 1988 is given in Table 5.1.1. Meedfor improving world production of food The production and distribution of food will face tremendous tasks in the coming decades. The world population is likely to grow from approximately 5 billion at present to approx imately 10 billion by the year 2050. At the same time, the energy required for food production will become more expensive and the need for a considerable reduction in pollution by carbon dioxide and waste materials will grow (Meadows, 1992). Furthermore, physical distribution will become increasingly complicated, in particular in larger urban areas. The needs of the consumer must also be considered. The existing tendency of diversification of consumption patterns will continue. At the same time, the demand for 'natural' products containing few, if any, additives will increase and the safety and hygiene of food products will be further emphasized. 5.1.2 Potential of biotechnology in food manufacturing Potential ofbiotechnology It is expected that biotechnology will offer tools to overcome, at least partly, some of the problems indicated in the preceding section. Asa n example the impact ofth e use of bovine somatotropin (bST) on animal numbers, feed requirement and waste production of dairy 197 Table 5.1.1. Classdistributio n ofth e useo fagricultura l raw materials and semi-finished goods over the various food-processing sectors in 1988 (in 106NLG) . Raw materials and semi- Origin Total finishedproduct s indigenous imported Meat 891 6 295 9211 Dairy 849 5 722 9217 Fish, vegetables, fruits 421 360 781 Cereals, cattle feed 200 7 473 4 674 1 Sugar 765 4 769 Flour 131 2 330 164 2 Cocoa, chocolate, confectionery 257 120 5 146 2 Margarine, fat, oils, starch, potatoes, coffee, tea 1610 4 686 6 296 Drinks 249 359 609 Tobacco - 852 852 cows to achieve 1988 US milk production levels indicated in Table 5.1.2 is extremely illustrative (OECD, 1992). Further opportunities for improving the feed requirement and waste production of dairy cows using biotechnology may be expected. For example, amino acids to improve dietary balance, novel antibodies to overcome resistance, cellulase en zymes to improve feed utilization, silage fermentation and cellulose breakdown, micro organisms as probiotics and transgenic micro-organisms to enhance ruminai digestion (OECD, 1992). Monoclonal antibodies and r-DNA- and r-RNA-probes will offer improved possibilities to control food contaminants which could be harmful to health. The increasing use of enzymes for the manufacture of cheese, fruit juices and flavours and fragrances can also be expected. It is anticipated that the scientific and technological progress will be concerned with enzyme reactors, encapsulated enzymes and protein engineering for improved and new enzymes. The production of mycoproteins and plant cell cultures for food purposes is also foreseen. However, due to the conservative attitude of the consumer only a few new products are likely to have entered the market by the mid-nineties. Such products could be genetically modified bakers yeast, cheese made with chymosin obtained by recombinant DNA-technology and special dietetic foods and food ingredients. Table 5.1.3 gives a more detailed picture of the OECD-expectations regarding food biotechnology developments. Although not mentioned emphatically in the OECD-report, it should be stressed that biotechnology may also be helpful in improving the composition of raw materials for better or new utilizations. This is of particular interest in plant breeding. Potential offermentation biotechnology For many years, specific food ingredients and foodstuffs have been produced by fermenta tion. These include alcohols, amino acids, and organic and fatty acids. Microbiologically 198 Table 5.1.2. Impact of use of bovine somatotropin (bST) on number of animals, feed requirement, and waste production of dairy cows to achieve the milk production in the United States in 1988. In the USA, 1988, population of cows was 10.24 million, production of milk per cow 6 460 kg and total production of milk 30 million tonnes (OECD, 1992). Variable Impact by bSTa Animals number of cows decrease by 10.7% milk yield per cow increase by 12.0% Feedb energy equivalent as maize grain decrease by 2.5 million tonnes protein supplement equivalent as 44% soya oil meal decrease by 56 thousand tonnes Waste decrease by 6 million tonnes decrease by 8 million m3 urinary nitrogen decrease by 80 thousand tonnes 3 methanee decrease by 80 million m Ifal lfarmer s used bST and ifit s use increased average milkyiel d per cowb y 12%. Ifcommerciall y approved, expected impact would be lessbecaus e technology rarely achieves 100% adoption Basedo nnutrien t requirements for dairycow sbod yweigh t averaging 650k gan d producing milko ffa t content 3.5% Based on an average diet composition of net energy content of 10.4 MJ/kg, and digestibility of 65% and content of dry matter in faeces of 16% Based on a volume rate ofurin e for a dairy cowo f 201/ dwit h a massconcentratio n ofnitroge n in urine of 10 g/1 If methane production represents 5% ofgros s energy intake produced polysaccharides are also increasing in significance. Xanthan produced by Xanth- omonas campestrisi sbein g increasingly applied. Gellan, with its interesting functional proper ties, has already been approved, and other promising microbial polysaccharides from lactic acid bacteria and from other bacteria, such aswelan , rhamsan, curdlan, scleroglucan and pullulan will become available in the near future (Voragen & Nout, 1991). By manip ulating their biosynthesis or treating them with enzymes these polysaccharides can be modified, and consequently can acquire new or improved properties. Natural or nature-identical flavours and fragrances are obtained by cultivating yeasts on precursor media and then spray-drying them. For instance vanilline, 2-heptanone and 2-undecanone, the main constituents of the Roquefort-aroma, and furanones such as 4-hydroxy-5-methyl-3(2H)-furanone, are already being produced biotechnologically. Modern biotechnology contributes to optimizing reactors and downstream processing. Cultures used for fermentations have been modified considerably. Starter cultures for accelerated cheese-ripening have become available (Exterkate & de Veer, 1990). Phage resistant starter cultures for the dairy industry and yeasts modified by r-DNA techniques to break down the dextrines have been developed. 199 Table 5.1.3. Developments in food biotechnology (OECD, 1992). In commercial operation - bioconversion of starch to sweet products - high fructose products - bioconversion ofvegetabl e oils - novel sweeteners - food flavours and enhancers - processing of fruit juice - amino acids and other special nutrient - novel structured foods from fermentation - cheese enzymes - yeast hybrids - new biotechnological methods of testing - lactose-free dairy product Scientific development; more widely commercial in next 5 years - new 'functional' foods for specific nutrition needs - food colours and ingredients by biotechnology - bioreactors for traditional foods; products of egg and soya - genetically modified baker's yeast - culture ofplan t cells and micro-algae for ingredients of high value Scientifically feasible commercial in 5-10 years - modified food enzymes - novel biocatalysis for food processing - bioconversion technology for unusual food environments - genetically modified food bacteria for flavour and quality - new biological preservation systems - rapid "dipstick" tests for all common food contaminants Potential ofenzymatic conversions For many years enzymes have been used in the production of foods. Biotechnology has recently given rise to new enzymes such as the starch-modifying enzymes with better properties, pullulanase for the improved manufacture of dextrose and maltose syrup or combined with amyloglucosidase for the manufacture of light beer. Glucanases and xyla- nases have been used in the manufacture of beer. Arabinanases and exo-glucanases have also been introduced, for example to clarify fruit juices. Fructosyl-transferase is recommended for the manufacture of fructosyl-oligosaccharides from saccharose. This low-calorie, non-cariogenic sugar, should promote the development of Bifidobacterium in the intestinal flora, act as dietetic fibre and have moisture-controlling properties (Voragen & Nout, 1991). Proteases are applied in the manufacture of cheese aromas, protein hydrolysates for special applications such as hypoallergenic food (de Koning, 1992 and Asselin et al, 1989) or modified proteins with specific properties (Siezen et al., 1990). Microbial lipases are applied for the hydrolysis of fats into fatty acids and glycerol, for the transesterification of palmoil with stearic acid as a cocoa butter substitute and for the preparation of food products with a specific taste. 200 Some of the aforementioned enzymes have been cloned in food-grade micro-organisms, by which means they can be produced. One example is the food-grade antibiotic nisin A and its improved version nisin Z obtained by site-directed mutagenesis (Rauch & de Vos, 1992). Since the application of glucose-isomerase, the first enzyme used in an immobilized form to produce fructose-rich syrups from cornstarch, other immobilized enzymes such as immobilized lactase and gluco-amylase have been applied (Carasik & Carroll, 1983; Scott etal, 1987). Potential of biotechnologyfor food safety In recent years a growing number of immunological test-kits for the rapid detection of e.g. Salmonella, E.coli, Campylobacter, Listeria, moulds and parasites has become available, as well as those for the rapid and specific detection of poisonous microbial products from Staphylo coccus aureaus and Bacillus cereus. These tests are performed either as ELISA or as latex agglutination reaction. Flow cytometry offers the possibility of detecting mould and bacte ria in a sensitive way via the microbial reductase systems. Immunological test-kits have also been developed to detect contaminants such as pesticides, proteins, hormones and anti biotics. The development of r-DNA and r-RNA-probes is still in an early stage. However, it is expected that they will be applied on a large scale within the next five years. 5.1.3 Biotechnological research for the food industry in the Netherlands The position of the Netherlands as a major producer of food products on the one hand and the potential of biotechnology for improved foods and improved food production on the other have given rise to considerable effort in biotechnological research for the food industry. Well before the start of the innovation-oriented research programme on agricul tural biotechnology (IOP-PcLB), biotechnological research for the food industry was al ready underway. For eight years this innovation-oriented research programme has given an important additional impulse to this research. In the rest of this chapter, specific subjects regarding biotechnological research for the food industry in the Netherlands are presented by way of examples. Some of these subjects have been subsidized under the auspices of IOP-PcLB, whereas others have not. Finally, a major barrier against applying biotechnology to the manufacture of food products isth e rather conservative attitude of the consumer. Therefore, the last part of this chapter is devoted to safety and to public acceptance. 5.2 Improving of the baking quality of gluten by enzymatic modification P.L. Weegels & R.J. Hamer Most wheat gluten (70 to 80%) is used to improve flour. The increase in the use of wheat gluten is largely due to the higher consumption of wholemeal bread and to the increasing 201 use of European wheat, the quality of which is less than that of US and Canadian wheat. In addition, gluten is used to replace milk or soya in veal-calf milk formulations. Gluten can also be used in a variety of non-food applications, such as pharmaceuticals, paper- coating or even plastics. However, wheat gluten isuse d very little in non-food applications. The increasing supply of wheat gluten and its relatively limited range of uses means that other areas of application must be found. In this section it will be shown that other applications of gluten are feasible. Examples will be confined to flour products. The role ofmodifie d gluten in improving baking quality will also be discussed (for a review on gluten modification, see Weegels 1990). 5.2.1 Modification of structural and functional properties For applications of modified gluten in food, we decided to use enzymatic modification since it is relatively mild, thus enabling some essential functional properties to be retained. Enzymes act in a highly specific way, so when food-grade enzymes are used, application in food is possible. In general, enzymatic modification is also environmentally safer than chemical modification. Unfortunately, relatively little is known about the relations be tween the structure and specific function of gluten proteins, which impedes targeted and site-directed modification. At present gluten is modified by the industry by trial-and-error research, which isno t very cost-effective and does not broaden our knowledge of structure- function relationships. We therefore aimed our research towards broadening the under standing of structure-function relationships before determining product quality character istics. To this end gluten was separated into its two main components: glutenins and gliadins. These proteins were hydrolysed by commercial enzyme preparations and the structural and functional properties determined. The main activity in these preparations was proteolytic, but there was also some hemicellulytic or amylolytic activity (Weegels et al., 1990). Some of these enzymes prefer to hydrolyse gliadins, others glutenins and still another group has no preference for either (Weegels et al, 1990). This implies that different groups of proteins can be modified selectively by choosing the right enzyme. This is important because glutenins are responsible for the elastic properties of gluten while gliadins account for the viscous properties. Not only can enzymes have a preference for either gliadins or glutenins, but differences in the rate of hydrolysis can also occur within these groups. It has been shown that enzymatic hydrolysis of gliadins may depend on the surface hydrophobicity of gliadins. Some enzymes hydrolyse hydrophilic gliadins first, others hydrolyse hydrophobic gliadins first and some have no preference (Weegels et al., 1990). This can be important since the hydrophobicity of gliadins is related to bread-making quality (Weegels et al., 1991a). When high-molecular-weight glutenins are hydrolysed preferences can also be observed for certain subunits. Fig. 5.2.1 shows that ifglutenin s are hydrolysed, a specific pattern can be observed. Some bands disappear more rapidly than others. Also specific bands are generated. Glutenins form the backbone of the gluten polymer and play an important role in the structure of dough (Graveland et al., 1985). Therefore, selective hydrolysis can be important for dough properties and bread-making quality. The glutenin subunits form a large three-dimensional network by means of disulphide bonds. We have developed a 202 t Mtagyw _ j* *S#fr7 ™-«î£ a *••* *;- '* -> -> n 10 to to 120 J40 t>«0 1440 min min min min min min min mfn Fig. 5.2.1. Immunoblot ofth e electrophoretic pattern ofhigh-molecular-weigh t glutenin subunits (top left) enzymically modified by enzyme 16Z06 for various periods of time and stained with polyclonal antibodies against these subunits. Specific fragments generated are indicated with an arrow. Reactive fragments labelled with a thiol-specific fluorescent label are indicated with an asterisk. method of detecting these reactive groups in fragments (Weegels et al., 1991b, indicated in Fig. 5.2.1 as asterisks). As can be seen, not all fragments contain such reactive groups. In addition to the study of structural changes, changes in some fundamental functional properties were also determined. For this purpose special techniques had to be developed or applied to gluten proteins and their digests for the first time (Weegels et al., 1990). Hydrophobicity of the digests changed dramatically upon hydrolysis. Depending on the enzyme chosen and hydrolysis time, protein surface hydrophobicity either increased or decreased. Surface tension of the digests decreased upon hydrolysis. In general, a low surface tension is related to good foaming properties. Emulsifying properties were also determined. These properties play an important role in various food systems, also in bakery products in which emulsifiers are often used. Depending on the enzyme, the ability to emulsify oil in water either increased or decreased with more extensive hydrolysis. Unfortunately, stability is not increased (Weegels et al., 1990). It can be concluded that enzymes are able to modify structural and functional properties in a highly specific and targeted way. The effects of the changes on product quality, especially of bakery products, are discussed below. 5.2.2 Applications of enzymatic modification: quality control Enzymes were used in two different ways to control the quality of bread, biscuits and crackers. First, gluten was hydrolysed by enzymes, and dried and heated to inactivate 203 them. Next, it was added to flour and product quality was assessed. The effect of enzymes on the quality of gluten added to dough was then determined. The loafvolum e of flour with one hydrolysed gluten preparation increased slightly when small amounts of hydrolysed gluten were added to flour (Fig. 5.2.2). Other hydrolysates had a strong, negative effect on loaf volume, especially at high dosages. In all samples, sensory properties (e.g. crumb colour and structure, appearance) were unsatisfactory. However, what isnegativ e for bread can be positive for products such as cakes, biscuits and crackers. Therefore, hydrolysates were added to the flour used for biscuits and crackers. In biscuits too much cohesion of the dough by protein is not desirable, because it produces springiness which results in thicker, oval biscuits and hence to problems during packaging. Additives not naturally occurring in flour are used to improve the dough properties of biscuits. Biscuits must have an open structure with an airy appearance and a low density. In Table 5.2.1 the effect of enzymatically modified gluten on both biscuit density and dough properties, as measured by strain of the dough, are given. As can be seen, hydrolysates are able to control dough properties and product quality both positively and negatively. Fortunately, improving the properties of dough is consistent with improv ing the density of the biscuits. Crackers differ from biscuits in that they need to have a layered, light structure. Fig. 5.2.3 shows, that depending on the dosage of hydrolysate, the density of the crackers can be changed in both a negative and a positive sense. These examples show the possibilities of controlling product quality by using an additive derived from wheat instead of using artificial improvers. Finally, two examples are given of enzymes added to dough containing extra gluten. The effect of some enzymes on the bread-making quality of the gluten added to dough c01 (0 JZ Ü 0.005 0.01 0.1 concentration of hydrolysate (% flour) Fig. 5.2.2. Effect of concentration of enzymically modified gluten on loaf volume. Gluten was hydrolysed by enzymes 30M15 (A) and 16M01 (•). 204 Table 5.2.1. Effect of addition of enzymically modified gluten on volume mass and dough property of biscuits as determined by penetrometer. Enzyme used Volume mass Dough property for modifying ofbiscuits(k gl" 1) strain (%)' gluten 16M01 0.42 0.30 None 0.39 1.12 16M04 0.39 1.42 16Z05 0.38 1.96 1 Percentage of penetration ofpenetromete r into the dough: the larger the strain value, the softer the dough was determined. The experimental set-up was chosen so that only the effect on the gluten was determined. In a range of baking tests the change in loaf volume could be estimated by adding the effect of gluten alone and that of enzyme alone (Fig. 5.2.4, left column). The actual increase in loaf volume caused by the addition of gluten in combination with enzyme was also determined (Fig. 5.2.4, right column). In all cases gluten increased loaf volume. Only hemicellulase is able to improve the bread-making quality of flour without extra gluten. Only one enzyme was able to improve the baking quality of gluten (Weegels and Hamer, 1992). This enzyme hydrolyses gliadins more rapidly than it hydrolyses glutenins, and has a preference for hydrophilic gliadins. Statistical and reconstitution studies (Weegels et al., 1991a) in our laboratory have shown that gliadins of this type have a negative effect on bread-making quality. We can conclude, therefore, that enzymatic hydrolysis of these proteins has a positive effect on the bread-making quality of gluten. 'J.J J ^* 0.31 - 0.29 >, 'cô 0.27' 0) / ~ ~ -A ^\ ____ -• 0.25 ~~» A , , i 0 12 3 4 5 concentration of hydrolysate (% flour) Fig. 5.2.3. Effect of concentration of enzymically modified gluten on the density of crackers. Gluten was hydrolysed by enzymes 16M01 (A), 16M04 (•) and 16Z05 (•). 205 130 tV ik o 100 o 70 40 m 10 CD Dl -20 hemicellula se gliadinase 1 ghadmase 2 glutenin enzyme to Camp Remy flour + 4 gluten WEM enzyme E*2<2§ gluten V///Ä gluten ^M gluten + 1 2 enzyme Fig. 5.2.4. Predicted cumulative effect on loaf volume of enzymes and gluten if added separately to flour (left bar) and the actual increase in loaf volume when gluten and enzyme were added together (right bar). If the baking quality of added gluten is improved by an enzyme it is indicated with an asterisk. The gliadin 2 was tested on two gluten samples differing in quality. An approach completely different from hydrolysis is enzymatic cross-linking of gluten proteins, for example by transglutaminase. If this enzyme is added during mixing, the amount of SDS-insoluble glutenins increases (Fig. 5.2.5). From the maximum in the mixing curve it can be seen that the dough development time decreases. Here again, many possibilities to modify gluten proteins are available. 5.2.3 Conclusions Gluten is a cheap, industrially produced protein suitable for a variety of applications. The properties of gluten, however, need to be modified. Enzymatic modification is the method of choice, since it is mild, food-grade-directed and site-directed. However, to prevent trial-and-error modification more insight into the structure-function relationships needs to be gained. Enzymatic modification can be a useful tool in this respect. Enzymes are able to specifically hydrolyse gliadins or glutenins, or even subunits of these fractions. Thus, structural and functional properties can be modified in a site-directed way. Both enzymatically modified gluten and enzymatic hydrolysis in dough with added gluten can efficiently control product quality as well as processing properties. This offers opportu nities for improving the baking quality of wheat gluten and to extend its applications to other food applications as well as to non-food applications. 206 o c CD O a 140 resting time of dough (min) Fig. 5.2.5. Effect ofcross-linkag e ofglute n proteins ofCam p Remy dough with (•) or without (A) transglutaminase as determined by the increase in SDS-insoluble glutenin (gel protein) during resting (top) and by the effect on mixing properties as determined by the power needed to deform the dough (bottom). 5.3 Enzymatic hydrolysis of milk proteins: some fundamental and practical aspects S. Visser, D.G. Schmidt & RJ. Siezen Enzymatic modification forms one of the various routes along which industrial proteins can be altered in order to increase their value and versatility (Arai & Fujimaki, 1991). Generally, enzymatic reactions can be performed under mild conditions of pH and tem perature. Moreover, because of their high degree of efficiency and specificity, a reasonable prediction can be made about the nature and quality of the end product. Therefore, the application of enzymes is an attractive tool with which to improve functional properties of food proteins without a significant deterioration in their nutritive value. Enzymatic modification of proteins can be subdivided into two classes: hydrolytic and non-hydrolytic reactions. Among the latter are enzymatic cross-linking between proteins, and the enzyme-induced incorporation or removal ofphosphate , carbohydrate and amide groups, which may strongly influence the properties of a protein. Hydrolytic reactions drastically change the structure of the protein, since large or small peptide fragments are formed which completely destroys the integrity of the original molecules. This may have far-reaching consequences on the functionality of the protein, for example its solubility, viscosity, foaming and emulsifying capacity, water binding, and flavour characteristics. A special application is represented by the plastein reaction, in which (enzyme-generated) peptide fragments, after having been concentrated to a high degree are, essentially ran- 207 domly, polymerized or aggregated under the influence of some proteolytic enzyme. This opens up the possibility of upgrading the amino acid composition of a protein or protein fraction by incorporating peptides or amino acids of high functional or nutritive value. In this section we will concentrate further on some aspects of enzymatic hydrolysis of milk proteins as it is applied in long-standing procedures or in new developments potentially suitable for various practical purposes. 5.3.1 Enzymatic hydrolysis of caseins Milk-clotting Based on their solubility behaviour at pH 4.6 and room temperature, the milk proteins can be divided into two classes: the caseins (insoluble) and the whey proteins (soluble). The caseins, constituting 80% ofth e milk protein fraction, can be subdivided into asl, as2,ß and K casein components, which in milk occur as a micellar complex in the approximate proportions of4:1:4:1 . This complex iskep t colloidally dispersed by the protective function ofth e K:-casein. By the action ofa milk-clotting enzyme this protective function is destroyed through the specific cleavage of one particular peptide bond in xr-casein. At an appropriate temperature (ca. 30 °C) the thus de-stabilized casein starts to coagulate under the influence of calcium ions in the medium. This forms the basis of curd formation in the normal cheese-making process. The specific cleavage of JC-casein by the milk-clotting enzyme chymosin has been the subject of extensive structure-function relationship studies at our institute. It appeared that all structural requirements for an optimal splittingb y the enzyme are located in a region around the susceptible peptide bond consisting of only 14 amino acid residues; furthermore, a minimum peptide length of five or six residues around that peptide bond was found to be necessary to obtain a measurable substrate cleavage (Visser, 1981). As a practical application of the latter a synthetic peptide mimicking this sequence, i.e. Leu-Ser-Phe-Nle-Ala-Ile methyl ester, is now being used as a standard substrate for measuring chymosin and pepsin activity and for determining the amounts of these en zymes in commercial rennets (Visser et al., \i Cheese-ripening During cheese-ripening most of the clotting enzyme, together with the whey components, is separated from the curd, but a small amount (about 10% in the case of Gouda-type cheese) is incorporated into the curd. Here, together with milk proteinases and proteolytic enzymes from added starter bacteria, it contributes to the further degradation of casein, finally creating a pool of small peptides and amino acids. These components may directly contribute to some flavour characteristics or may form precursors of cheese-flavour compounds. In general, the actual cheese flavour is thought to be mainly generated by the further non-proteolytic conversion of amino acids via enzymatic and non-enzymatic reactions. Hydrolytic breakdown of caseins in cheese may generate a bitter flavour, due to the accumulation of bitter peptides, which generally have a hydrophobic character. A major bitter peptide, isolated from a bitter Gouda-cheese, was identified as the extremely hydrophobic carboxyl-terminal part of the /?-casein component. This part can be split off by the (residual) milk-clotting enzyme chymosin as well as by the Pr type proteinase present 208 10 20 30 R-E-L-E-E-L-N-V-P-G-E-I-V-E-S-L-S-S-S-E-E-S-I-T-R-I-N-K-K-I- in the cell envelope of a number ofLactococcus strains (Fig. 5.3.1). Such strains often belong to the group of'bitter strains'. In contrast, 'non-bitter' strains have been found to contain a Pnl type cell-envelope proteinase, which preferably attacks/?-casein in the amino- terminal part of the molecule (Fig. 5.3.1). Unlike the breakdown of asl-casein, the degradation ofy9-casein , and consequently the development of a bitter flavour, is strongly suppressed by the presence of salt in the cheese. Moreover, bitter peptides once formed, can be degraded by the sequential action of various types of peptidases present in the interior of lactococcal cells. The availability and activity of such debittering peptidases also depend on the salt concentration of the cheese medium. The literature on proteolytic enzymes in relation to cheese-ripening and flavour formation has recently been reviewed (Visser, 1993). Altogether, it is obvious that fundamental studies on the proteolytic system of rennet and lactic acid bacteria (e.g. characterization, kinetics, specificity) have contributed considerably to a better understanding of cheese-ripening and have also led to progress in the technique of accelerated ripening (Exterkate, 1987). Other applications At our institute the study of the proteolytic system of lactic acid bacteria has led to the development of new cheese types, in one case based on the principle of accelerated ripening (Exterkate & de Veer, 1990). It has also resulted in the development of a natural flavour product from milk by directly making use of the proteolytic action ofenzyme s from selected micro-organisms or combinations thereof (Siezen et al., 1990). The utilization of isolated casein peptides, generated by enzymatic hydrolysis, forms 209 Table 5.3.1. Activities reported for various peptides isolated from bovine or human caseins. For details on the individual topics, see Meisel et al. (1989),Maubois & Léonil (1989),Yamauch i (1992), and references cited there. Kind of Fragment(s) from (bio)active effect casein component Surface activity (emulsifying properties) ß Immunostimulation &si,ß Inhibition of enzyme converting angiotensin I asl,ß,K (anti-hypertensive activity) Antithrombotic activity K Opioid activity asl, ß Opioid-antagonist activity asl, K Antigastric activity K (inhibition of acid secretion) Stimulation ofDN A synthesis ß Growth stimulation of Bifidobacterium K Mineral carrier activity (Xsl,a s2, ß (binding capacity for minerals; inhibition of calcium-phosphate precipitation) another application of great interest. Table 5.3.1 shows a brief selection of properties reported for casein peptides. Such products may be of industrial or medical/pharmaceuti cal importance. The inhibition potency of casein phosphopeptides towards calcium phos phate nucleation and precipitation, also studied by our group (Schmidt et al., 1987), seems promising in connection with the promotion of intestinal calcium absorption (Sato et al., 1991) and the prevention of caries (Reynolds, 1987). Different procedures have been used to isolate and purify casein peptides from digests obtained from a variety of proteolytic systems. In our laboratory, plasmin-generatedyS-casei n peptides were isolated using either a batch process or by making use of a continuous membrane reactor; the technique of high performance liquid chromatography (HPLC) was used for monitoring the isolation process and for the final purification of isolated peptides (Visser et al., 1989). 5.3.2 Enzymatic hydrolysis of whey proteins General The whey protein fraction of milk mainly contains yff-lactoglobulin and a-lactalbumin, together with a small portion of minor proteins (e.g. serum albumin, lactoferrin, immuno globulins). This class of proteins has long been considered solely as an unimportant by product of cheese and caseinate production. Being a relatively inexpensive product, most of the whey protein is utilized as animal feed with a minor part used in the food industry. Due to their high nutritive value much more effort is now given to applying whey proteins for human consumption. In many cases their functional properties can be improved by modification. For reasons described in the opening of this section, enzymatic hydrolysis is 210 often the modification method of choice. As in the case of casein hydrolysates, whey protein hydrolysates may also contain very bitter peptides. The extent of the bitter flavour defect depends on the degree of hydrolysis, the hydrophobicity of the proteolytic products and, therefore, on the nature and quantities of the proteolytic enzyme(s) used. Structure andsubstrate behaviour ofwhey proteins In contrast to caseins, which have a rather unordered structure, whey proteins, in their native state, are globular proteins which can undergo denaturation as a result of various physical processes, such as heat treatment. The state of denaturation strongly influences the susceptibility of a whey protein towards enzymatic hydrolysis. In the case ofyS-lacto- globulin it was found that the genetic variability may also influence the rate of hydrolysis by certain enzymes, such as trypsin and papain (Schmidt et al., 1990) (Fig. 5.3.2). The latter is probably caused by the fact that the substitution of a charged aspartyl group (A variant) by an uncharged glycyl (B variant) residue in a protruding loop of the molecule directly influences the electrostatically determined enzyme-substrate interactions. The sigmoidal shape of the hydrolysis curve of non-heatedjff-lactoglobuli n treated with papain (see Fig. 5.3.2) suggests a slow initial cleavage ofon e or only a few peptide bonds. This then DH (%) 6 ßLg- A^^-^ 5 - y\^ //v^- A (heated) 3 - / I ^^$Lg- B (heated) / / ßL9~ it-—-—— 1 - // / /^^ aLa U/^^^—" aLa (heated) 3 1 incubation time (h) Fig.5.3.2 . Curves showingdegre e ofhydrolysi s (DH)a sa functio n ofincubatio n timefor th e action ofth e enzyme papain on the individualwhe yprotein s a-lactalbumin(aLa )and/Mactoglobuli n(/?Lg ) at pH 8.0 and 40 °C. Influence of the genetic variability of/?Lg, and of heat treatment (10 min, 85 °C) ofth e substrates before incubation with the enzyme. 211 makes other bonds more accessible causing acceleration of hydrolysis. The sigmoidal shape might also be explained by the time needed for initial activation of papain (a sulphydryl-type proteinase) via the sulphydryl group of the substrate. For commercial purposes total whey protein products, for instance in the form of whey protein isolates or concentrates, are used. From the above it follows that the state of protein denaturation and aggregation in such products largely determines their suitability for modification by enzymatic hydrolysis. Therefore, this potential variability among whey protein products sometimes makes special experimental conditions necessary in order to facilitate accessibility of the proteins to enzymatic cleavage (Schmidt et al., 1990). In general, proteolysis can be promoted by factors such as variations in the enzyme/substrate ratio, temperature, ionic strength, by increasing or diminishing the calcium-ion concentra tion of the medium and/or by using combinations of enzymes showing different cleavage capacities towards the individual whey proteins (Schmidt & Poll, 1991). Special applications Enzymatic hydrolysis has been employed to enhance the solubility ofwhe y proteins (Monti &Jost, 1978) and to diminish their allergenic properties (Asselin et al., 1989). The latter has also been the subject of studies in our laboratory (Schmidt et al., 1990); the characteristics of various whey protein hydrolysates were examined and the relationship between para meters such as antigenicity, allergenicity and molecular weight distribution in the hydroly- sate determined (Fig. 5.3.3). Whey protein hydrolysates are also used for other dietary and therapeutic formulations (Roger et al, 1981; Wenner, 1982). In one application a diet of 220 time (min) Fig. 5.3.3. Molecular mass distribution pattern obtained from high-performance size-exclusion chromatography of a hypo-allergenic whey protein hydrolysate. The zones of molecular mass, indicated by vertical lines, are based on elution patterns of a series of reference substances. 212 low phenylalanine content was prepared, which was claimed to be suitable for patients suffering from phenylketonuria (Arai et al., 1986). Other possible medical/pharmaceutical applications are some isolated peptides showing opioid (Antila et al., 1991; Yamauchi, 1992) and opioid antagonist (Tani et al., 1990) activity, compositions for skin care and hair conditioning (Hidalgo &Jost , 1980; Tomita et al., 1991) and a preparation claimed to have a beneficial effect on various kinds of diseases, such as neurodermatitis, rheumatism, arthritis and glaucoma (Gauri, 1986). A peptide fragment of the minor whey protein lactoferrin was recently reported to exert antibacterial activity much higher than that of the parent protein itself (Bellamy et al., 1992). Peptide fragments isolated from /Mactoglobulin, after its tryptic hydrolysis in an ultrafil- tration-type membrane reactor, showed enhanced interfacial activity (Turgeon et al., 1992), which may be used to improve food emulsions and foams. In nutritional applications the use of food-grade bacterial enzymes has proved helpful in the debittering of hydrolysates (Parker & Pawlett, 1987). 5.3.3 Concluding remark Fundamental studies on the enzymatic hydrolysis of milk proteins have in many respects led to the development of new ideas for the practical application of milk protein hydroly sates and purified peptides obtained thereof. Many more utilizations of milk proteins are to be expected as a result of future basic and applied research. 5.4 Plant cell wall degrading enzymes in the processing of food G. Beldman & A.G.J. Voragen During the last decades polysaccharide degrading enzymes have been introduced into several food manufacturing processes in order to facilitate the process and to obtain products with higher yields and improved quality. Initially, undefined fungal (generally Aspergillus niger) excretion products, containing a broad spectrum of polysaccharidases, were used as enzyme preparations, for instance to clarify apple juice. Gradually, more knowledge of the role of some of the main activities in these preparations became availa ble. This led to the development of better defined preparations, designated for the degra dation of one of the cell wall constituents, or some of them. Thus, it is necessary to distinguish between modification and degradation of plant cell walls and their polysaccha ride building blocks. In Fig. 5.4.1 different processes are shown in which polysaccharide degrading enzymes can be applied. The arrows indicate to what extent the enzymes are used for degradation or for specific modification of the cellwal l or one ofit spolysaccharides . New developments in the research into relevant enzymes in these areas will be described in the following paragraphs. 213 Modification Degradation Fruits and Clarification — vegetables Cloud stability — Pressibility Liquefaction — < Maceration — Firming Cereals and Filtration seeds Malting/Brewing Baking Starch/gluten separation Extraction By-products Fibre modification Digestibility Saccharification - Miscellaneous Oligosaccharides - Retting of flax _ - Pulp and paper - - Gelling agent Thickening agent - Analysis Fig. 5.4.1. Enzymatic degradation and modification of plant cell wall polysaccharides. Arrows indicate to what extent polysaccharides are modified or degraded. 5.4.1 Fruits and vegetables Exogenous enzymes are used during the processing of fruits to increase juice yields and improve quality. Fungal pectinases have been used forjuic e clarification for more than fifty years. For some applications cellulases also are needed. It is estimated that in this industry about 20 million US dollars per year are spent on enzymes (Ulig, 1991). The pectinases can be added to the process at different stages, depending on the technologies to be applied. Four main processes can be distinguished (Pilnik and Voragen, 1989). In chro nological order of their introduction in industrial processes these are: juice clarification, pulp enzyming, liquefaction and maceration. Cloud particles present in the extracted juice consist of a positively charged core of protein, coated by a layer of negatively charged pectin. These particles are kept in suspension by electrostatic repulsion (Fig. 5.4.2). Partial removal of the pectin layer leads to the formation of larger aggregates, which can be removed by sedimentation (Lea, 1991). The viscosity of the juice also decreases, which is beneficial for final filtration. The pulp enzyming process was developed to produce juice from soft fruits, which are difficult to press, but was later also used for extracting the juice from apples and grapes. Two relatively new developments in the enzymatic processing of fruits are liquefaction and maceration. For liquefaction a combination ofpectinase s and cellulases is required. In this process the cell wall polysaccharides are degraded to such an extent that they are 214 Protein Pectin in suspension due to electrostatic repulsion Pectin Pectolytic enzyme action Exposure of dissimilarly charged surfaces Agglomeration due to electrostatic attraction Fig. 5.4.2. Mechanism for the formation ofpectin-protei n floes in applejuic e (Lea et al., 1991). almost completely disintegrated and solubilized. Model experiments with apple cell walls have shown that in addition to pectolytic enzymes a combination of an endo-/?-1,4- glucanase (Endo I or IV) and an exo-yS-1,4-glucanase (cellobiohydrolase, CBH) is essential (Voragen et al., 1980). From recent results obtained in our laboratory it can be concluded that for total liquefaction not only the breakdown of cellulose is important, but xyloglucan degradation is also essential, because this polysaccharide interlinks cellulose fibrils and shields them from cellulases (Fig. 5.4.3) (Renard et al., 1991). The fine structure and enzymatic degradation of xyloglucans from apple, pear and potato are currendy being investigated. The maceration process is based on enzymatic degradation of middle lamella pectin, which weakens cell cohesion and results in a suspension of intact cells. Compared with a convential mechanical maceration process, the enzymatic approach saves energy and gives rise to less undesirable reactions (e.g. enzymatic browning) because reactants remain compartmentalized. For enzymatic maceration well-defined enzyme preparations are needed, preferably with a high level of endo-polygalacturonase (endo-PG) or endo-pectin lyase (endo-PL) and void of pectin esterase. This can be achieved by genetic modification of the microbial source (Aspergillus species). Cloning of the endo-PG and endo-PL genes and transformation into A. niger has been established (Bussink et al., 1991; Kusters-van Someren et al., 1991). The pectolytic enzymes mentioned so far are only active towards the homogalacturo- 215 RHAMNOGALACTURONAN mmm CELLULOSE NEUTRAL SIDE CHAINS ]II1 1II II I HYDROGEN BONDS XYLOGLUCAN O ANCHOR Fig. 5.4.3. Schematic structure of the primary cell wall. nan part of the pectin molecule, which is mainly built up of galacturonic acid. In addition to these 'smooth' regions, the pectin molecule also contains ramified 'hairy' regions, with a backbone of alternating galacturonic acid and L-rhamnose units (rhamnogalacturonan), branched with side chains composed of neutral sugars. Also, some galacturonic acid residues are esterified by acetic acid. The rhamnogalacturonan is not degraded by most of the commercial 'pectinase' preparations. During enzymatic juice production using these enzymes, the non-degraded part accumulates in thejuic e and can be recovered by ultrafil tration. The polysaccharide fraction thus obtained is called 'modified hairy regions' (MHR). A rhamnogalacturonase, able to degrade this structure has recendy been identi fied and subsequently purified and characterized (Schols et al, 1990). The role of this enzyme in the liquefaction process is still unknown. A prerequisite for the degradation of the hairy regions by this enzyme is deacetylation of the substrate. In fact, an acetylesterase activity, specific for the acetylated hairy regions has recently been found (Searle-van Leeuwen et al., 1992). 5.4.2 Cereals The cell walls in cereals, surrounding the starch granules, are built up of polysaccharides which are generally known as non-starch polysaccharides (NSP). The major part of the NSP is arabinoxylan. A general structure of arabinoxylans is given in Fig. 5.4.4. The /?-l,4-linked xylopyranose units in the backbone can be substituted with arabinofuranosyl residues on 03 or both on 02 and 03. There is evidence of minor substitution on 02 only. In some cereal arabinoxylans (i.e. from sorghum and rice) substitution with glucu- 216 X=D-Xylose A =L-Arabinos e GA =D-glucuroni c acid [ R = H ) or 4-0- Methyl -•• - (R = CH3) Fig. 5.4.4. Structure of arabinoxylans. ronic acid is also found. In addition, ester linkages with acetic and phenolic acids can be present. It is known that arabinoxylans play an important role in cereal processing and in the properties of cereal products. Their effects on milling behaviour, starch-gluten separa tion, gluten quality, filtration of wheat starch hydrolysates, dough handling, volume, crumb and crust characteristics of loaves, are mentioned in the literature (Voragen et al., 1992). Arabinoxylans also forms a barrier for starch-degrading enzymes during brewing, has a negative effect on wort and beer filtration and may lead to haze formation. Several of these characteristics can be affected by technical enzyme preparations, containing xylanases. However, considering the complex nature of the materials and the fact that the enzyme preparations used are mixtures of several activities, it isdifficul t to gain any insight into the mechanisms involved. Therefore, current research is being directed towards elucidating the fine structure of cereal arabinoxylans and the mode of action of xylan- degrading enzymes on these substrates. Relevant xylanolytic enzymes have been isolated from the culture filtrate ofAspergillus awamori (Kormelink, 1992). Three different endo-xylanases (Endo I, II and III), a ß- xylosidase and an arabinoxylan arabinofuranohydrolase (AXH) have been found. Endo- xylanases I and III show clear differences in their mode of action on arabinoxylans extracted with alkali from the water-unextractable cell wall material of wheat flour. In Fig. 5.4.5 the elution patterns of Endo I and Endo III digests of this arabinoxylan are com pared. With Endo I, monomeric xylose and a range of oligosaccharides are formed, while Endo III produces only small amounts of xylose and oligomers with two, three or four sugar units. By isolating all the different oligosaccharides and identifying their structure by NMR (Gruppen, 1992; Kormelink, 1992), the preferential site of cleavage could be as signed for both endo-xylanases. Degradation of a hypothetical, moderately branched arabinoxylan (arabinose/xylose = 0.6) by Endo I and III is shown in Fig. 5.4.6. Endo I can split the glycosidic linkage at the non-reducing (left) side of a xylopyranose residue, inde pendent of the presence of a substituent. Endo III, however, needs at least one more unsubstituted xylose residue at the non-reducing side of a branch point, to be active. This accounts for single as well as for double substituted residues. To be able to split at the reducing (right) side of a single substituted xylose residue, Endo I needs at least one unsubstituted xylose residue and Endo III at least two unsubstituted units. For a double branched xylose, both enzymes need at least two unsubstituted residues. Considering these 217 10 9 a 7 6 5 4 3 2 1 cone (t-ig/ml) 800 i 1 1 1 1 1 1 r 0 20 40 60 80 100 120 140 160 18020 0 cone lug/ml) 800 600- 400- 200 0 20 40 60 80 100 120 140 160 180 200 fraction number (-) Fig. 5.4.5. Bio Gel P2chromatograph y ofalkali-solubl e wheat arabinoxylan digested by endoxyla- nase I (a) (Gruppen, 1992) and endoxylanase III (b) (Kormelink, 1992). modes of action, the appearance of larger oligomers in the Endo III digest is evident. It is known that enzyme preparations containing xylanases are able to influence baking characteristics of wheat flour. In baking experiments Endo I and III behaved differently. Although both enzymes produced increased bread volume, this effect was more pro nounced for Endo III than for Endo I (Table 5.4.1). Endo III was also better with respect to break and shred as well as to crumb structure (Gruppen, 1992). It is proposed by these investigators that the better performance of Endo III is caused by its higher degree of Table 5.4.1. Baking characteristics of endo-xylanase-treated dough and bread (Gruppen, 1992). Loafvolum e was determined by the rape seed displacement method. Break and shread quality and crumb structure was scored one day after baking (scale ofrating , 0 - 10). Stickiness Loaf volume Break & shread Crumb structure (ml) Blank no 489 6.5 6.5 Endo I no 511 6.5 6.5 Endo III slightly 557 7.0 7.3 218 Endo I Endo III Endo I V v V \y v v V Endo Fig. 5.4.6. Possible sites of cleavage by endoxylanases I and III on a hypothetical arabinoxylan (Kormelink, 1992). Circles indicate 1,4-linked/?-xylopyranosy l units; squares indicate 1,2/3-linked a-arabinofuranosyl units. randomness in arabinoxylan degradation, leading to a more rapid decrease in the water- holding capacity ofth e water-unextractable cellwal l material. As a consequence this could also result in a more efficient redistribution of (bound) water from the hemicellulose to the gluten, which is thought to be responsible for improved loaf volume and dough character istics (Maat et al., 1992). This is a good example of how fundamental enzymology can provide more insight into applications. It is therefore not surprising that Endo Ill-like xylanases from different Aspergillus strains have been cloned and brought to expression in Aspergillus (de Graaf et al., 1992; Maat et al., 1992). More fields of application for these cloned xylanases are certainly possible and are currently being explored. We have already mentioned applications in the brewery and wheat-starch industries. Applications in the non-food sector are, however, also possible. For instance, in improving the digestibility of cereal-based animal feed, bleaching of paper pulp, the up-grading of recycled paper and the total saccharification ofagricultura l biomass. It isexpecte d that these areas will provide a rapidly growing market for cloned xylanases. 5.4.3 By-products The processing of vegetable agricultural resources often results in a by-product containing cell wall material. Generally, such products are used as animal feed, but they could be of greater value if they were applied in food for human consumption. For instance, it is well known that pectin can be extracted from apple waste and citrus peel. Other residues containing interesting polysacharides are beet pulp (pectin and arabinan), potato fibre (pectin) and cereal wastes such as wheat bran, corn bran and brewers spent grains (arabi- noxylans). In many cases the polysaccharides which can be extracted from these materials do not have the desired properties. Enzymes can be of use to specifically modify these 219 polysaccharides. A patent from British Sugar (McCleary et al, 1989) describes the extrac tion of a (branched) arabinan from sugar beet and its gelling after removal of the arabinose side chains with an arabinofuranosidase. The arabinan degrading enzymes from A.niger have been isolated and characterized and genetic work is being carried out to clone the structural genes coding for these enzymes and to improve the expression levels in A. niger (van der Veen et al., 1990). Another example may be the enzymatic modification of pectin from sugar beet to improve its gelling response (Matthew et al., 1990). 5.4.4 Conclusions The following conclusions can be drawn from the above-mentioned information: - Plant cell wall degrading enzymes have a broad field of application, especially for the processing of food, but also in the non-food sector. - There is growing knowledge about the role of the individual enzymes, present in cocktails of technical enzyme preparations, during the processing of plant raw materials. - For processes aimed at the enzymatic modification of one of the cell wall polysaccha rides, preparations with one major activity are necessary. This implies cloning of the desired gene and transformation in a suitable production strain. - Cloning of relevant enzymes makes production of 'tailor-made' technical enzymes feasible. Futher optimization of the cloned enzyme (for instance improved pH and temper ature optimum) can be obtained by protein engineering. 5.5 Enzymatic processing of triglycerides and fatty acids. Enhancing reaction kinetics by continuous product removal J.T.P. Derksen, G. Boswinkel, W.M.J, van Gelder, K. van 't Riet & F.P. Cuperus In naturally occurring plant seed oils many unusual but industrially interesting fatty acids are often positioned predominantly at positions 1 and 3 of the triglyceride molecules. Examples are erucic acid in Crambe abyssinica and Brassica napus (rapeseed) seed oils (Muuse et al., 1992), dimorphecolic acid (a conjugated hydroxy-diene fatty acid) in Dimorphotheca pluvialis (marigold) seed oil (Muuse et al., 1992) and 7-linolenic acid, a dietary and phar- maceutically interesting a-linolenic acid isomer (Briggs, 1986) in Borago officinalis or Oenoth era spp. (evening primrose) seed oil (Muuse et al., 1988; Christie, 1991). To specifically hydrolyse the above desirable fatty acids from the triglyceride molecules, 1,3-specific lipases can be employed, both in emulsion systems (Boswinkel et al., 1990) and in membrane bioreactors (Derksen et al., 1991, 1992). An extremely versatile bioreactor system employs hydrophilic hollow polymeric membrane fibres onto which lipase is ad- sorptively immobilized (Kloosterman et al., 1987, Pronk et al., 1988). In this type of bioreactor the membrane not only provides a support for the enzyme, but also creates an interface with a constant surface area between the water phase and the oil phase on the lumen side of the membrane. Specific hydrolysis reaction takes place on this interface 220 (Pronk et al., 1991, 1992a, 1992b). Specific hydrolysis, however, leads to an equilibrium reaction mixture, composed largely of non-water-soluble products, i.e. fatty acids, monoglycerides, diglycerides and triglycerides. This incomplete triglyceride hydrolysis is unfavourable to an industrial specific fatty acid production process. The most elegant way of solving the problem of an unfavourable equilibrium position is to continuously remove one or more of the reaction products as they are being formed. In the case of specific hydrolysis one can choose to recover either monoglycerides or fatty acids from the reaction mixture. Such selective recovery will, on the one hand, allow the hydrolysis reaction to proceed towards completion and, on the other, will greatly facilitate the downstream processing of the desired product. In this section we describe some of our recent results in addressing this equilibrium problem. As a model system we have chosen to hydrolise specifically the 1,3 esterbonds of tricaprylin. The resulting caprylic acid is continuously eliminated from the reaction mix ture by either selective precipitation or by membrane-assisted solvent-solvent extraction. Mathematical evaluation of the results from this model system also allows us to assess important issues such as enzyme kinetics and mass transfer problems, thus aiding the practical implementation and scale-up of specific triglyceride hydrolysis. In addition, we will address some industrially interesting applications, for which the developed technology will be of considerable benefit. 5.5.1 Specific hydrolysis of tricaprylin If tricaprylin is allowed to react with 1,3-specific Rhizopus delemar lipase the reaction does not proceed towards complete hydrolysis: apart from fatty acids and monoglycerides, diglycerides and triglycerides are still present (Fig. 5.5.1a). This unfavourable thermody namic equilibrium interferes with an efficient production of specific fatty acids from triglycerides. We therefore tried to extract the enzymatically liberated caprylic acid contin uously by precipitation with barium-hydroxide (0.1 M) in a pH-stat at a pH of 7.0. Fig. 5.5.1b shows the results of such an experiment. It can be seen that the concentration of fatty acids is reduced and that the relative concentration of the monoglyceride end- product is increased, implying an enhanced reaction selectivity towards monoglycerides and specific fatty acids. Although the above experimental system has been very valuable for assessing the kinet ics of the specific hydrolysis reaction and of the end-product removal, it is difficult to scale-up. For this reason a more elegant system was developed utilizing membrane tech nology (Fig. 5.5.2). In this set-up the hydrolysis reaction takes place in a hollow-fibre membrane module in which 1,3-specific Rhizopus javankus lipase is immobilized. The tricaprylin isle d through the lumen of this reactor module, where it is split into monoglyc erides and fatty acids. Due to the presence of a potassium phosphate buffer at the shell side of the membrane the caprylic acid is simultaneously converted into caprylate ions. The caprylate then permeates through the hydrophilic membrane and is removed with the phosphate buffer phase. The remainder of the reaction mixture is recirculated through the reactor module to allow further hydrolysis. Fig. 5.5.3 shows the concentration of monocaprylin, dicaprylin and tricaprylin and 221 CONCENTRATION [mol/1org . phase] (A) 10 15 20 25 TIME [hours] Fig. 5.5.1. For legend see next page. caprylic acid with respect to reaction progress. In comparison with the normal equilibrium position, as exemplified in Fig. 5.5.1a, it is clear that the equilibrium is shifted towards an increased selective hydrolysis (i.e., a higher monoglyceride concentration is reached). From mass balance calculations we found that after 57 hours of reaction per mole of initial tricaprylin substrate, 0.127 mole of caprylic acid was extracted, or 6.4% of the potentially available fatty acid. These results, combined with the reaction kinetics, give valuable information on system configuration and membrane surface area required for optimal hydrolysis and extraction. This information has a direct bearing on industrially relevant applications as described below. 5.5.2 Modelling Reaction kinetics and the influence of the fatty acid extraction on the equilibrium positions were modelled. The conversion of triglycerides to diglycerides and monoglycerides, as depicted in Fig. 5.5.4, was considered as a pseudo-second order reaction resulting in equations 1 and 4. drr]/dt = -k1[T][W]+k1[D][F] (eq. 1) d[D]/dt = k,rT][W] -k.rDJtF] -k2rD][W] + k2[M][F] (eq. 2) d[M]/d t = k2[D][W ] - k2 [M][F ] (eq. 3) d[F]/dt = k,[T][vV]-k,p][F]-k2[D][W] + k2[M][F] (eq. 4) 222 CONCENTRATION [mol/1 org. phase] (B) 10 15 20 25 TIME [hours] Fig.5.5.1 . Lipase-mediated hydrolysis oftricapryli n with (A) and without (B) simultaneous caprylic acid extraction. (A) 30 g of tricaprylin was emulsified with water containing 250 mg of Rhizopus delemar lipase. Hydrolysis proceeded at 30 °C and was monitored by GLC. D, caprylic acid; *, monocaprylin; +, dicaprylin; O, tricaprylin. (B) Asi n (A),excep t that the pH ofth e water phasewa smaintaine d atp H 7.0b ycontinuou s addition of a Ba(OH)2 solution of0. 1 mol/1i n a pH-stat, resulting in a precipitate of barium caprylate. with equilibrium constants Kn = k„/k.n or, at equilibrium: K, = [D][F ]/ \Y\[W ] and K2=[M][F]/[D][W]. The hydrolysis of tricaprylin without product extraction was simulated using equations 1 - 4, in which the equilibrium constants K, and K2 were estimated using the concentrations at equilibrium. The rate constant k, was calculated from the intial rate of the reaction, hence d[T]/d t = -k,[T ][W] .Usin g a Gear integration routine and a modified least square analysis k.1; k2 and k_2 were estimated. When the above model was used in simulation experiments a poor fit with the experimental data was obtained. For a good mathematical reaction model it was found necessary to include a third hydrolysis term, taking into account the hydrolysis of monoglycerides to fatty acids and water: d[M]/dt = k2[D][W] - k2[M][F] - k3[M][W ] + kS[G]\F] (eq. 5) d[FJ/d t = (eq. 4) + k3\M\ [W] - k.3[G][F J (eq. 6) With this modified model a perfect fit with the experimental data points was obtained, completely overlapping the curve shown in Fig. 5.5.1a. The calculated rate constants are given in Table 5.5.1. These modelling experiments indicate, rather unexpectedly, that a 223 tank reaction mixture caprylin phase pump pressure control membrane reactor- extractor extraction phase tank pump fatty ^8H acids buffer pH =7.0 Fig. 5.5.2. Scheme of a membrane bioreactor unit with an in-line fatty acid extraction unit. (small)amoun t ofmonoglyceride s isfurthe r hydrolysed. Webeliev e this reaction stept ob e due to a slowintramolecula r 2-monoglyceride rearrangement: the 2-positioned fatty acid CONCENTRATION [mol/l org. phase] 3 2.5 D S 2( 1.5 \ x""^ 1 rV/ r ;H* ""~"^>\ *~znr^==—— \ =¥ 0.5 r *-/^ o ~ —O Of 0 10 20 30 40 50 TIME [hours] Fig. 5.5.3. Hydrolysis of tricaprylin in a membrane bioreactor. 150 ml of tricaprylin was recircu lated through the lumen of a hollow-fibre membrane bioreactor (Cuprophane, ENKA, Wuppertal), onto which Rhizopusjavanicus lipas e was immobilized. Through the shell of the membrane module a potassium phosphate buffer 0.1 mol/l (pH 7.0) was led single pass at a flow rate of 0.5 ml/min. (a), caprylic acid; *, monocaprylin; +, dicaprylin; O, tricaprylin. 224 wator fattyaci d water fatty acid water fatty acid vw V2^ Vk3^/ triolyceride ^> _ »- digryceride J" ^**. monoglyceride ^_ l+° glycerol T^7\ TTT\ 7^7\ water fatty acid water fatty acid water fatty acid Fig. 5.5.4. Simplified reaction scheme of a non-specific triglyceride hydrolysis. migrates to the 1-position of the triglyceride, whereupon the resulting 1-monocaprylin is hydrolysed by the 1,3-specific lipase to glycerol and caprylic acid. The above findings clearly illustrate how mathematical modelling can supplement experimental data. 5.5.3 Applications At ATO-DLO, we are currently involved with the development of bioreactor processes for the specific hydrolysis of triglycerides from several newly-introduced oilseed crops. One example is Dimorphothecapluvialis see d oil. This oil contains more than 60% dimor- phecolic acid (9-hydroxy,1 Ot ,12t-octadecadienoi c acid; for a structural formula see Fig. 5.5.5), a highly reactive hydroxydiene fatty acid, which has received considerable indus trial interest (Muuse et al., 1992). Due to its reactivity this fatty acid cannot be recovered by conventional Colgate-Emery-type high temperature steam splitting techniques. Moreo ver, dimorphecolic acid was found to be located predominantly on the 1,3-position of the triglyceride (Muuse et al., 1992). For these reasons specific enzymatic hydrolysis was considered the best method for producing dimorphecolic acid. Expanding on experimen tal data on specific hydrolysis and product removal, combined with its mathematical modelling, we have developed a novel bioreactor system for processing this oil (Derksen et al., 1991, 1992). Incorporating a newly-developed membrane that can hold high enzyme activities as well as an in-line cold trap, this bioreactor produces high yields of dimorph ecolic acid, with its hydroxydiene functions still intact. Further optimization of the reactor, as well as characterizing and derivatizing this unusual fatty acid is currently in progress (Derksen et al. 1993). A second example in which the results described in this section have provided us with valuable information isth e specific hydrolysis of Crambe abyssinica seed oil. This oil contains highly symmetrical triglycerides with more than 57% erucic acid, which is located exclu sively on the triglyceride's 1,3-positions (Muuse et al., 1992)(Fig. 5.5.5). Lipase-mediated Table 5.5.1. Calculated rate coefficients and equilibrium coefficients. They were calculated from the mathematical model for specific hydrolysis oftricapryli n as described in the text. The columns represent the kinetic coefficients of each ofth e individual reaction steps in the hydrolysis of trica prylin, as illustrated in Fig. 5.5.4. T, tricaprylin; D, dicaprylin; M, monocaprylin; G, glycerol. Step T^D D^M M^G 3 1 k+x 10" [l-mol"'-If ] 5.6 4.5 2.1 k_x 10"3 [l-mol"'h-'] 58 188 38.2 K[-] 0.1 0.024 0.055 225 COOH V V Erucic acid (13c-Docosenoic acid) OH COOH Dimorphecolic acid (9-Hydroxy-10t,12t-octadecadienoic acid) Fig. 5.5.5. Structural formulas of dimorphecolic and erucic acid. specific hydrolysis of Crambe oil yields a mixture of fatty acids which, due to the symmet rical nature of Crambe triglycerides, comprise more than 85% pure erucic acid, a purity that can otherwise be reached only by costly downstream vacuum distillation (Derksen et al., 1991, 1992). 5.5.4 Conclusions Lipase-catalysed specific hydrolysis of triglycerides can be an extremely useful way of producing fatty acids that are enriched on the 1,3-position of the triglyceride or are too thermally unstable to be produced by conventional means. Furthermore, there are very good prospects for developing new bioreactor systems within the next few years for use in enzymatic triglyceride hydrolysis on a commercial scale. 5.6 Thermodynamic principles of enzymatic acylglycerol synthesis A.E.M.Janssen , A. van der Padt & K. van 't Riet The esterification of glycerol and fatty acid consists of a sequence of three equilibrium reactions: Glycerol + Fatty acid <-» Monoester + H20 Monoester + Fatty acid <-> Diester + H20 Diester + Fatty acid <-> Triester + H20 These reactions can be catalysed by the enzyme lipase. Enzymatic synthesized acylglyc- erols, such as monoacylglycerols (monoesters) and triacylglycerols (triesters) are of com mercial interest, since these products are usually regarded as natural substances. Monoesters serve as emulsifiers for the food industry. It is desirable to obtain the monoester in a high purity state, since it then has better emulsifying properties than a mixture of esters. Triesters with characteristic properties, such as a specified melting range, 226 are also useful for the food industry. Contamination with fatty acids and diesters shoud be avoided for it will influence the melting properties and have a negative effect on yield. The aim of the paper presented in this section is to describe the equilibrium position of enzymatic acylglycerol synthesis. The reaction conditions to obtain pure monoester and pure triester are discussed. 5.6.1 Theory For a reaction: A + B <->C + H20 the reaction equilibrium can be described by: XC ' XH,0 YC ' YffjO K YR Where K: equilibrium constant thermodynamic activity mole fraction activity coefficient As shown above, three reaction steps are involved for the production of acylglycerols. Therefore, three equilibrium constants can be defined, an equilibrium constant for monoester production (KM), for diester production (KD) and triester production (KT). To calculate the mole fractions at equilibrium, it is important to know the activity coefficients of all components in relation to their concentrations. Experimental determination of all activity coefficients is physically impossible. However, methods are available for calculat ing activity coefficients and in this study the UNIFAC group contribution method is used (Fredenslund et al., 1977). To predict the ester mole fractions at equilibrium in a two-phase system, the computer program 'Two-phase Reaction Equilibrium Prediction' (TREP), was developed (Janssen et al., van der Padt et al. [a]). 5.6.2 Results The experimental and calculated results for the esterification of glycerol and decanoic acid are shown in Fig. 5.6.1. The markers in this figure are the experimental values, while the lines represent the calculations with the computer program TREP. To produce pure monoesters, it is to be expected that the glycerol to fatty acid molar ratio must be one or more. The results of the experiments where the molar ratio of glycerol to fatty acid is varied, while the water to glycerol ratio is kept constant, are shown in Fig. 5.6.1a. Indeed, the monoester mole fraction increases by increasing the glycerol to fatty acid ratio. How ever, it can be calculated that, even at a sixfold excess of glycerol, diesters and triesters are still formed. Therefore, adding an excess of glycerol is not an efficient way of producing monoesters. To produce pure triesters, stoichiometric amounts of glycerol and fatty acid are necessary. In the experiments shown in Fig. 5.6.1b, the glycerol to fatty acid molar 227 0 2 4 6 0 2 4 6 Initial glycerol to fatty acid ratio Initial water to glycerol ratio Fig. 5.6.1. Measured and calculated substance fractions of ester (x)a t equilibrium as a function of the glycerol to fatty acid (a) and the water to glycerol (b) ratio. In Fig. 5.6.1a, the reaction was with 20 mmol of decanoic acid and equal molar amounts of glycerol and water. In Fig. 5.6.1b, the reaction was with 30 mmol of decanoic acid, 10mmo l ofglycero l and a varying amount ofwater . For all tests 25 mg of lipase (Chromobacterium viscosum) was added and the samples were incubated for 200 h at 35 °C. For the calculations, the values used were KM=1.1, KD=0.5 and KT=0.4. Measured substance fractions of monoacyl (D), diacyl (O) and triacyl glycerol (A) and calculated substance fractions monoacyl ( ), diacyl (---) and triacyl glycerol (-—). ratio iskep t constant (1 to 3) and the water content isvaried . For the calculations, the same set of equilibrium constants is used as in Fig. 5.6.1a and some deviations are apparent. However, these are less than ±0.02 mole fraction. At a low water to glycerol molar ratio, a relatively high triester mole fraction was expected. But Fig. 5.6.l b shows that the triester mole fraction is not very dependent on the initial water to glycerol ratio. From Fig. 5.6.1 it can be concluded that by varying the ratios of glycerol to fatty acid or water to glycerol, a mixture ofmonoesters , diesters and triesters isalway s obtained. Neither pure monoesters nor pure triesters can be produced in this way. Solvent effects on the equilibrium position The addition of an organic solvent is found to influence the ester concentrations at equilibrium. The dielectric constant here is used as a measure of polarity of the solvent (Table 5.6.1); a high dielectric constant corresponds with a polar solvent and a low dielectric constant corresponds with a non-polar solvent. In Fig. 5.6.2a, the calculated ester mole fractions at equilibrium are plotted against the dielectric constant of the added solvent. It can be seen that in polar solvents, high mole fractions of monoesters and low mole fractions of triesters are calculated, while in non-polar solvents, the calculated differ ences in the mole fractions of monoesters, diesters and triesters are lower. In Fig. 5.6.2b, the experimental results for the esterification of glycerol and decanoic acid in several organic solvents are shown. The same trends can be seen as in Fig. 5.6.2a; the addition of a polar solvent results in a high mole fraction of monoester, while the non-polar solvents show relatively higher mole fractions ofdiester s and triesters. The measured mole fractions are not exactly equal to the predicted values. However, deviations are less then a factor of 228 Table 5.6.1. Dielectric constants oforgani c solventsuse d in thisstud y From Riddick & Toops(1955) . Solvent Dielectric constant l.Hexane 1.88 2. Isooctane 1.94 3. Nonane 1.97 4. Decane 1.99 5. Toluene 2.38 6. Pentyle ether 2.77 7. Isoamyl ether 2.82 8. Butyl ether 3.06 9. Phenyl ether 3.69 10. Isopropyl ether 3.88 11. Ethyl ether 4.34 12. Chloroform 4.81 13. 2-Methyl-2-butanol 5.82 14. Monoglyme 7.20 15. Triglyme 7.50 16. Dichloromethane 9.08 17. 1,2-Dichloroethane 10.4 18. 2-Methyl-2-propanol 10.9 19. 4-Methyl-2-pentanone 13.1 20. 3-Pentanone 17.0 21. Diacetone alcohol 18.2 22. 2-Butanone 18.5 23. Acetone 20.8 24. l-Methyl-2-pyrrolidone 32.0 25. Dimethylformamide 37.0 26. Acetonitrile 37.5 2. These deviations may be due to inaccuracies in the UNIFAC group contribution method. For the production of monoesters it is important to choose a polar solvent. At a high concentration of a polar solvent, almost no diesters and triesters are formed (Janssen et al., in press). The addition of polar solvents is an efficient way of producing monoesters as the only product. Effect ofwater activity In the calculations shown in Fig. 5.6.1b, the initial water activity is in the range of 0.0001 to 0.85. However, during the reaction glycerol is consumed and water is produced, both leading to an increase in water activity. At equilibrium the water activity is in the range of 0.75 to 0.92. These high water activities explain the low ester mole fractions, especially the low triester mole fraction, at equilibrium. To obtain high ester mole fractions, it is desirable to keep water activity low during the reaction. Calculations are made for a reaction system where the water produced is removed continuously. The results are shown in Fig. 5.6.3. At high water activities, a mixture of monoesters, diesters and triesters is formed and the triester mole fraction is low. However, at water activities below 0.1, the triester mole fraction increases rapidly, while the monoester and diester mole fractions decrease. There- 229 13 1o5 21 D 22 a D 0.2 D 20 23 8 10 Da 16 0.1 5%6 s 4B7 **„ 10 20 30 40 Dielectric constant 10 20 30 40 Dielectric constant Fig. 5.6.2. Calculated (a) and experimental (b) substance fractions of ester (x) in the organic phase as a function of dielectric constant. The reaction was with 10 mmol of decanoic acid, 10 mmol of solvent, 20 mmol of glycerol, 20 mmol ofwate r and 25 mg of lipase (Chromobacterium viscosum). The sampleswer e incubated for 300h at 35°C .Fo rth ecalculations ,th evalue suse dwer e KM =1.1; KD = 0.5 and KT = 0.4. Solvents numbers are listed in Table 5.6.1. For some solvents, the calculations are missing, since the required UNIFAC parameters are unknown. Substance fractions of monoacyl (D), diacyl (O) and triacyl glycerol (A). fore, a water activity of less than 0.1 during the reaction seems to constrain triester production. It is possible to control water activity in an immobilized enzyme pervapora- tion system by using a condenser (van der Padt et al. [b]). The ester mole fractions are measured as a function of water activity and it is shown that an excess of triester can only be produced at water activities of less than 0.1. 230 0.2 0.4 water activity Fig. 5.6.3. Calculated substance fractions of ester (*) with continuous removal of water as a function ofwate r activity. Input variables TREP (withoutwate r production):KM =1.1 ,KD=0. 5 and KT=0.4; t=35 °C; Initial amounts were 30mmo l ofdecanoi c acid, 10mmo l ofglycero l and varying amounts ofwater . Calculated substance fractions ofmonoacy l( ),diacy l (—) and triacyl glycerol (--)• 5.6.3 Conclusions The product of the esterification of glycerol and decanoic acid in a closed two-phase system, is always a mixture of monoesters, diesters and triesters. Pure monoesters can be produced by the addition of polar organic solvents. To produce pure triesters, it is impor tant to keep water activity as low as possible during the reaction. The computer program TREP is extremely useful for predicting the ester concentrations at equilibrium. 5.7 Engineering lactic acid bacteria for improved food fermentations W.M. de Vos In this section the present state of the art and perspectives of engineering lactic acid bacteria for improved food fermentations is reviewed. Most attention will focus on lacto- cocci that were the first to be characterized genetically and for which food-grade genetic modification methods have been developed. Novel lactococci have now been obtained by genetic, metabolic and protein engineering. In most cases similar methodologies are appli cable to other lactic acid bacteria including lactobacilli, leuconostocs, streptococci, and pediococci. 5.7.1 Genetic tools Initial genetic interest in lactic acid bacteria has been mainly focused on the mesophilic strains ofLactococcus lactis, because of their easy laboratory manipulation, simple metabolic pathways, and considerable economic importance as starter cultures for the production of 231 cheese, quark and fermented butter. Early studies have shown that lactococci contain a wealth of mobile DNA elements. These include a panoply of plasmids encoding key enzymes in important metabolic path ways (McKay, 1983), transposons such as Tn5276 encoding sucrose metabolism and nisin production (Rauch & de Vos, 1992), and a great variety of bacteriophages, which frustrate industrial fermentation processes (Klaenhammer, 1987). With the use of these elements basic genetic transfer processes such as conjugation and transduction have been estab lished, which are effective for constructing improved starter strains (Gasson, 1990). Various lactococcal plasmids have been converted into useful vectors (de Vos, 1987). A family of stable and high copy number vectors has been developed based on the replicon of pSH71, which appeared to be functional in all the transformable lactic acid bacteria, several other gram-positive bacteria and also in Escherichia coli (de Vos, 1987). In combina tion with efficient transformation procedures based on electroporation, and several plas- mid-free genetic model strains, a variety of host-vector systems has been developed for lactic acid bacteria (de Vos, 1987). These and other vectors have been converted into expression, secretion, and integration vectors that are now used for the genetic modifica tion of many lactic acid bacteria (Chassy, 1987; de Vos et al., 1989; Kok, 1990). 5.7.2 Biosafety considerations and food-grade selection systems Appropriate biosafety considerations are a prerequisite for the application of engineered lactic acid bacteria for the production of fermented foods. In all cases the nature of the genetic improvement should be tuned to the ultimate application ofth e lactic acid bacteria that may be either in foods to be placed on the market (uncontained use) or in industrial fermenters in which they are eventually separated from the desired products (contained use). Three types of genetic improvement may be distinguished that have different legal status (as defined by the Council directives, 1989) and that are presendy effective in each member state of the European Community: (i) classic genetic techniques, that do not involve genetic modification, such as mutation-selection and natural gene transfer meth ods, that have been used for decades and are excluded from the directives; (ii) self-cloning techniques, which when applied to food lactic acid bacteria to be used in contained systems are not considered to result in genetic modification and are therefore also excluded from the directives; and (iii)technique s that result in genetically modified lactic acid bacteria, the use of which is regulated by the directives. In many cases, the incorporation of transmissible antibiotic resistance genes in lactic acid bacteria is either not allowed or not preferred. Therefore, food-grade markers have been developed for selecting and maintaining the desired genetic event. To improve L.lactisstarte r cultures by conjugation, a natural genetic transfer mechanism, selection has been made using metabolic markers, such as plasmid-encoded lactose metabolism (Sand ers, 1988) or transposon-encoded sucrose fermentation (Rauch &d e Vos, 1992). A homol ogous food-grade system for use in self-cloning and genetic modification has been devel oped on the basis of the lactose genes of L.lactistha t have been extensively characterized (de Vos et al., 1990). Using the small (0.3 kb) lacFgene for Enzyme IIIlac in combination with various LacF-deficient L.lactis strains, an effective system has been developed and 232 applied in lactococci to allow both the selection of improved strains and the stable mainte nance of recombinant plasmids during industrial fermentation processes on lactose-con taining media (de Vos, 1990; see also below and Fig. 5.7.1). 5.7.3 Phage resistant strains A variety of bacteriophage insensitivity mechanisms have been identified in L.lactis(Klaen - hammer, 1987). Commercial successes have been realized by stacking into one L.lactis strain plasmids that encoded different bacteriophage-insensitivity mechanisms (Sanders, 1988; Klaenhammer, 1987). In these experiments conjugation was used to transfer plas mids and counter-selection was based on the capacity of the recipient strain to ferment lactose. In this way a food-grade selection was realized and genetic probing used to verify the successful transfer of bacteriophage-insensitivity plasmids. Various major culture pro ducers in the world now provide these improved starter strains, illustrating the application potential of advanced knowledge of plasmid biology. 5.7.4 Strains with improved proteolytic capacity Lactic acid bacteria are nutritionally fastidious and contain complex proteolytic systems to provide essential amino acids and peptides that support growth during fermentation in protein-rich environments such as milk or meat. The key enzyme in the cascade of proteolytic degradation of the milk protein casein in lactococci is a cell envelope-located serine proteinase, that belongs to the well-known family of subtilases (Siezen et al., 1991). Proteinase production has been studied extensively in the industrial strain L.lactis SKI1 and found to depend on the expression of two divergently transcribed genes,prtP encoding the pre-pro-proteinase and prtM encoding a lipoprotein that is required for the autoprote- Fig. 5.7.1. Schematic representation ofplasmidpNZl 125 obtained by self-cloning and containing the food-grade marker (lacF).Th e plasmid contains the pepNgene that codes for overproduction of the debittering aminopeptidase N. 233 olytic processing of the proteinase. The SKI 1proteinas e has been the subject of a great variety of studies aimed at engineering its processing, specificity, and production (de Vos et al., 1990). Interestingly, proteinase-overproducing lactococci were able to grow more rapidly in milk. This indicated that the initial step in the proteolysis is a limiting factor during starter growth that may be partially overcome by increasing proteinase production (Bruinenberg et al, 1992). In addition, by protein engineering based on knowledge-based modelling, various novel proteinases have been obtained that showed improved stability or substrate specificity (Vos et al., 1991). These and other proteinases may have application in the production of novel foods or may be used to produce specific flavour-generating peptides from casein. Following the initial degradation of casein by the lactococcal proteinases, further con version into amino acids or peptides is realized by an extensive set of peptidases that have been found in lactococci. One of these peptidases is aminopeptidase N, which has a broad substrate specificity and the capacity to debitter casein hydrolysates (Tan et al., 1992). Following cloning and characterization of the L.lactispepJV gen e encoding aminopeptidase N, lactococci have been constructed that highly overproduce this aminopeptidase (van Alen-Boerrigter et al., 1991). In subsequent studies, the food-grade marker based on the lacFcomplementatio n (de Vos, 1988) was incorporated, resulting in the first example of self-cloned lactococci that overproduce an industrially important enzyme. L.lactis strains harbouring the final constructs (one is illustrated in Fig. 5.7.1) have now been excluded from the biosafety directives for contained use. 5.7.5 Improved production of diacetyl Diacetyl is an important flavour compound present in buttermilk, fermented butter, and also included in some margarines. Small amounts of this flavour compound are formed by oxidative decarboxylation from a-acetolactate, an intermediate in the citrate and sugar metabolism of some lactic acid bacteria (Fig. 5.7.2). The production of a-acetolactate has been analysed by metabolic NMR studies (Verhue et al., 1991) which confirmed its earlier proposed metabolic pathway (Fig. 5.7.2). To allow for a high and stable a-acetolactate production strains should be used that have a low level of a-acetolactate decarboxylase. One such strain is present in the L.lactis culture NIZO 4/25 used in various industrial diacetyl production processes (Hugenholtz & Starrenburg, 1992). There are various approaches to increase production of a-acetolactate and hence diace tyl, by using fermentation or metabolic engineering. One is to allow for efficient citrate transport by fermenting at optimal pH. The characterization and functional expression of the L.lactiscitPgene for citrate permease in E.colimembran e vesicles has made possible the detailed analysis and optimization of this transport process (David et al., 1989). Moreover, improved production of a-acetolactate could be realized by growing L.lactisunde r aerated conditions (Starrenburg & Hugenholtz, 1991). This was explained by the high oxygen sensitivity of the enzyme pyruvate formate lyase that competes with a-acetolactate syn thase for available pyruvate (Fig. 5.7.2). Finally, the gene for a-acetolactate synthase has been cloned and characterized, allowing its overproduction in appropriate L.lactis strains (Verrips et al. 1990). It is expected that these and other changes in the metabolic flux 234 lactose lactate 1 ^^ « acetate v PTi T~ formate pyruvate ^m ethanol ^^ pdh V ac-tpp^ acetate t carbondioxide citrate acetoin ioc-acetolactate c=> diacetyl Fig. 5.7.2. Pathways for the production and conversion of pyruvate by lactic acid bacteria. Key enzymes are indicated: ldh, lactate dehydrogenase; pfl, pyruvate formate lyase; pdh, pyruvate dehydrogenase; als, a-acetolactate synthase; aid, a-acetolactate decarboxylase. The intermediate acetaldehyde-TPP is indicated by ac-tpp. The white arrow indicates the oxidative decarboxylation of a-acetolactate leading to the formation of diacetyl. leading to a-acetolactate (Fig. 5.7.2)wil l result in even further improvement in the diacetyl production in lactic acid bacteria by metabolic engineering. 5.7.6 Conclusions It is evident that genetic, metabolic and protein engineering approaches are now feasible in lactococci and other lactic acid bacteria. Some results of these approaches have been illustrated here by examples that include the construction of bacteriophage-resistant strains, food-grade overproduction of a debittering aminopeptidase, production of novel proteinases, and overproduction of the flavour compound diacetyl. In view of the signifi cant research efforts currently being undertaken in various parts of the world, including the European Community where there is a large programme focusing on the biotechnol ogy of lactic acid bacteria (Aguilar, 1991), further exploitation of lactic acid bacteria is to be expected. Acknowledgements I am grateful to Roland Siezen andJeroe n Hugenholtz for stimulating discussions. Part of the research is funded by BRIDGE contract BIOT-CT91-0263 of the Commission of European Communities. 235 5.8 Engineering and application of nisin, a natural food preservative O.P. Kuipers, H.S. RollemaJ. Hugenholtz, W.M. de Vos & R.J. Siezen Nisin is an antimicrobial peptide produced by various strains of the lactic acid bacterium Lactococcus lactis (formerly known as Streptococcus lactis)(Delves-Broughton , 1990; Hurst, 1981). The first three letters of nisin are short for 'Group JV Streptococci inhibitory Substance'. Nisin inhibits the growth of Gram-positive bacteria. Since it is not toxic to either humans or animals and is broken down in the gastrointestinal tract, its use as a preservative in the food industry is accepted by the World Health Organization (WHO) and the US Food and Drug Administration (FDA). In Western Europe nisin is used only on a limited scale, for instance in the preparation of cheese spreads and other processed cheese varieties. In Eastern European countries nisin is also used in canned fruit and vegetables. The limited use of nisin must be attributed on the one hand to the fact that it is relatively unknown, especially outside the dairy industry, and on the other to its limita tions, such as its rather limited solubility and lowered stability in strongly acidic and basic media, the insensitivity of (sometimes pathogenic) Gram-negative bacteria to nisin and the high sensitivity of most lactic acid bacteria to nisin. In order to overcome these limitations NIZO scientists are working within a multidisci- plinary research programme aimed at acquiring detailed knowledge about the production, activity and properties of nisin and at extending its application. Such applications could include not only prolonged preservation of foods and beverages such as meat, salads, (alcohol-free) beer and wine, but also pharmaceutical applications such as the control of acne in humans (Propionibacterium acnes) and infection of the udder in cows (Streptococcus agalactiae). This article will deal successively with the structure/function relationship of nisin, the use of protein engineering techniques to construct modified, and possibly improved nisin molecules and the development of a nisin-producing (cheese) starter culture. 5.8.1 Structure and functional properties of nisin The polypeptide chain of nisin is made up of 34 amino acids. Besides the common amino acids, the nisin molecule contains a number of special amino acids not frequently found in nature. It contains the unsaturated amino acid residues dehydroalanine (Dha) and dehy- drobutyrine (Dhb), as well as lanthionine and yS-methyl lanthionine residues which consti tute five thio-ether rings (see Fig. 5.8.1). For this reason nisin is classified in the group of lantibiotics (lanthionine-containing antimicrobial peptides) (Schnell et al., 1988). The spe cial amino acids are formed during a post-translational modification process, in which several serine and threonine residues are first converted into dehydroalanine and dehydro- butyrine residues, respectively. This is followed by a stereo-specific addition of the five cysteine residues to five Dha and Dhb residues, resulting in thio-ether linkages which form the characteristic cyclic structural elements in nisin. 236 H,N- R NH R I i l co CH 1 1 CH- -S- -CH2- II l -CH- -CH 1 • _NH — C— CO- NH CO 1 1 Dehydroalanine (R=H) Lanthionine (R=H) Dehydrobutyrine (R=CH3) 3-Methyllanthionine (R=CH3) Fig. 5.8.1a. Primary structure of nisin A and structural formulae of unsaturated amino acids and lanthionines: a. primary structure of nisin A. -S- indicates the sulphur atom that is involved in the thio-ether linkage, to form (/?-methyl)lanthionine; b. structural formula of dehydroalanine (R-H) and dehydrobutyrine (R-CH3); c. structural formula of lanthionine (R-H) and yS-methyllanthionine (R- CH,). From the amino acid composition ofnisi n it appears that it isa basic peptide. In aqueous solution the nisin molecule displays maximum stability and solubility in a weakly acidic medium. The solubility decreases with increasing pH as a result of the decrease in positive charge due to titration of the two histidine residues. Nisin is not stable in a strongly acidic or strongly alkaline medium where loss of biological activity occurs. In a strongly acidic medium the addition of a water molecule to the dehydroalanines takes place, followed by cleavage of the polypeptide chain due to the formation of an amide and a pyruvyl amino acid (Rollema et al., 1991). Little is yet known about the precise mode of action of nisin. It is assumed that under certain conditions this lantibiotic may damage the cell membrane of the bacterium, as a result of which a leakage of essential cellular constituents may occur (Ruhr & Sahl, 1985). Model studies have demonstrated that nisin is capable of forming pores in artificial membranes (Sahl et al., 1987; Gao et al., 1991). It has also been proposed that nisin can specifically bind to cell wall components and thus prevent cell growth (Morris et al., 1984). In various laboratories throughout the world investigations are being conducted into the functional properties and the mode of action of lantibiotics. Within the scope of the European BRIDGE programme, NIZO iscollaboratin g with the universities of Nijmegen, Bonn and Tübingen on various aspects of lantibiotics. 237 Fig. 5.8.Ib . Model ofth e structure ofnisi nA ,base d on one ofth e conformations asdetermine d by 2dimensiona l NMR techniques (van de Ven et al., 1991). Nisin Z differs at position 27 in the last of the five lanthionine rings, with a histidine-to-asparagine substitution. Residues 17 and 18 were chosen for protein engineering. 5.8.2 Improvement of nisin by protein engineering The previously mentioned insensitivity of Gram-negative bacteria to nisin and the de creased stability and solubility of nisin at particular pH values have up to now limited its application to weakly acidic food products. It is therefore desirable to adapt several properties of nisin with the aim ofwidenin g the field of application. Methods that might be considered include chemical modification of nisin or organic synthesis of altered nisin molecules for specific modulation of their properties. However, both approaches have the disadvantage that they are time-consuming and costly. Moreover, there is no possible way of directly using an L. lactis strain which is able to produce these altered nisins in the preparation of food-products. Therefore, these synthetic and modification methods hardly seem practicable from an industrial point of view. Another approach is the directed search in naturally occurring L. lactisstrain s for nisin variants which might have other properties. Recently, investigators at NIZO discov ered that a naturally occurring Lactococcus strain produces a nisin variant in which the amino acid histidine in position 27 is replaced by asparagine (Mulders et al., 1991). Subsequently, the occurrence of the gene for this natural variant was identified in numer ous other L. lactis strains (Kuipers et al., 1991). This nisin variant, which is referred to as nisin Z, can readily be separated from the other nisin, which is now referred to as nisin A, by reversed-phase HPLC. The properties of nisin Z differ somewhat from those of nisin A. For instance, in a so-called agar diffusion test nisin Z displays larger inhibition zones at high concentrations than nisin A (Fig. 5.8.2). From a practical point ofvie w this means that 238 nisin Z has certain advantages compared with nisin Afo r use in food products with a low water content such as cheese. A novel approach for making specific changes in the nisin molecule isth e use ofprotei n engineering techniques. This method is based on the fact that a gene, coding for a particular protein, can be expressed in a suitable (micro)organism, which leads to the production ofthi sprotei n by the host.Makin g changes (mutations) in specific places in the DNA coding for the desired protein makes it possible to selectively change the codons for particular amino acids, as a result ofwhic h specifically altered proteins can be produced. This type of protein engineering is referred to as 'site-directed mutagenesis'. To perform protein engineering it is required that: 1. the gene coding for the desired protein is isolated; 2. an expression system for the gene is developed; 3. a purification method for the protein is developed; 4. the protein isproperl y characterized; and 5. preferably also structural information on the protein isavailable . In the case of nisin all these conditions have been satisfied. In 1988 various research groups, including our team at NIZO, isolated the gene codingfo r nisinA (Buchma n et al., 1988, Rauch et al, 1991). Subsequently the nisin was localized on a large so-called conjugative transposon, a piece ofmovabl e chromosomal DNA which can be transferred from one bacterium to the other by natural gene transfer techniques (Rauch et al, 1991, Rauch and de Vos, 1992). Moreover, a purification procedure, the structure determina- Fig. 5.8.2. Agar-diffusion assay of antimicrobial activity of purified nisin A and nisin Z. The indicator strain Micrococcusflavus wa s grown on an agar medium. Nisin samples were put into wells and allowed todiffus e into the medium. The diameter ofth e clearing (ordiffusion ) zone isa measure for the antimicrobial activity. Identical amounts ofnisi nA (top )an d nisin Z (bottom)wer e compared; samples at the left contained 1 fig per well and samples at the right 0.05fig per well. 239 tion using 2D-NMR and various functional properties of nisin A and Z have been de scribed by different groups. Several papers by various authors on these subjects can be found in the book 'Nisin and Novel Lantibiotics' (Jung, G. and Sahl, H.G. (Eds.), 1991). Recendy, expression systems for (mutant) nisin genes were developed at NIZO, which has allowed the production of various nisin mutants by L. lactis (Kuipers et al., 1991). In some cases the production level of the plasmid-encoded nisin was found to be at least twice as high as that of the chromosomally-encoded nisin A. With this system on hand a start has been made with the production of mutant nisins (Kuipers et al., 1991). Initially, we are aiming to improve the solubility and action of nisin at neutral and higher pH values, to enhance the stability and change the antimicrobial spectrum. In addition, this expression system can also be used to answer more fundamental questions on the mechanism of the post-translational modifications to which nisin is subjected, its secretion mechanism and its mode of action. 5.8.3 Development of nisin-producing cheese starter cultures The 'late-blowing' defect observed in cheese-making is the result of contamination with spores of Clostridium tyrobutyricum which gives rise to butyric acid fermentation (see Fig. 5.8.3). It is the most frequent bacterial contamination in Gouda-type cheeses. The growth of this bacterium, which enters the milk as a spore from cattle feed, is effectively inhibited by nisin. In practice, however, this defect is usually avoided by adding nitrate to milk or by bactofugation. Objections to the addition of nitrate have increased in recent years, as it could be metabolized to nitrite and carcinogenic nitrosamines. Nisin has been shown to be a successful alternative to nitrate in experimental cheese preparation. The addition of 200 IU nisin per gram of cheese was found to be amply Fig. 5.8.3. Effect of contamination with Clostridiumtyrobutyricum o n the structure of Gouda cheese, the 'late-blowing' defect (top), compared to a normal cheese (bottom). 240 sufficient to suppress butyric acid fermentation at levels of Clostridium contamination which correspond to the large numbers ofspore s (occasionally) found in winter milk (Table 5.8.1). Although the separate addition of nisin to cheese milk has not yet been accepted in the Netherlands, it is possible to use nisin-producing cheese starter cultures. However, most nisin-producing lactic acid bacteria lack a number ofpropertie s essential to cheese produc tion, such as resistance to bacteriophages and the capability of decomposing milk protein. The characteristic of nisin production is known to be transferrable to other lactic acid bacteria. At NIZO it was demonstrated that the genetic elements for nisin production, nisin immunity and sucrose fermentation are closely coupled and reside on a transposon (Rauch and de Vos, 1992). This transposon can be transferred to other lactic acid bacteria at a frequency which is dependent on various factors, such as the type of bacterium, the presence ofplasmid s and the growth phase ofth e bacteria. Transfer ofth e nisin transposon is complicated by the fact that the donor bacterium produces nisin, whereas the acceptor bacterium is generally extremely sensitive to nisin. The transfer of nisin production or nisin immunity to present industrial starter cultures is unfeasible since most industrial starters used in the Netherlands are complex mixed cultures ofvariou s types and strains of lactic acid bacteria. The most popular cheese starter in the Netherlands, the Bos starter culture, consists of dozens of different strains which might each specifically contribute to the final flavour and structure of the cheese. The transfer of nisin production and immunity to all these strains is impracticable. At NIZO it was therefore decided to opt for a different strategy, the design of a well-defined, limited- strain starter. First, from the various cheese starters, strains were isolated which are suitable for use in the preparation of cheese as far as organoleptic and metabolic properties and resistance to bacteriophages are concerned. It was found that a suitable cheese could be made with a mixture of only three strains of the Lactococcus type. Such a mixture now permits a simple incorporation of nisin production and immunity. The next step is to transfer the production of nisin (preferably nisin Z) to at least one of the three strains and nisin immunity to all three. When this goal has been reached NIZO will have developed lactic acid bacteria cultures which produce nisin and can be used as starters in the production of cheese without there being any need to add nitrate. Table 5.8.1. Prevention of butyric acid fermentation ('late-blowing') in Gouda cheese, produced without nitrate, by inclusion of nisin-producing lactic acid bacteria (Hugenholtz & de Veer, 1991). Cheese was produced with different number fractions of the nisin-producing L. lactisstrai n NIZO R5 in the starter culture. The number concentration of Clostridium spores in cheese milk was 6000 LT1. Cheese number fraction nisin content in cheese start of no of strain R5 (IU/g) late blowing in starter (%) (weeks) after 3 wk after 26 wk 0 0 4 1 9 10 10 210 110 none 100 1080 630 none 241 5.9 Molecular genetic approaches in plant cell wall degradation M.A. Kusters-van Someren, L.H. de Graaff&J. Visser Plant cell walls are predominantly composed ofpolysaccharide s such as cellulose, hemicel- lulose and pectin. Fungi are the most common source of industrially used enzymes: Phanerochaete for the production of ligninases (Alic & Gold, 1991) and Trichoderma as well as Phanerochaetefo r the production of cellulases (Coughlan, 1991).Th e food-grade filamentous fungus Aspergillus nigerha s been the organism of choice for the production ofpectinase s and hemicellulases for many years. Being a saprophytic fungus, A. niger must be able to adapt to different sources of polysaccharides. Structural studies, for example on pectins from various plants have shown that there are differences with respect to esterification, acetyla- tion and molecular weight. This is reflected by the variety of pectinolytic activities pro duced by A. niger. Since the enzymes are thus expected to differ with respect to e.g. substrate specificity or pH or temperature optimum, they may also have different applica tions. Technical pectinase and hemicellulase preparations are always a mixture of many enzymes. For some applications, however, this is not desirable. For example, when making pulpy nectars a limited breakdown of the pectin leads to stabilization of cloud. This is accomplished preferably by pectin lyase (PL) or polygalacturonase (PG) preparations free of pectinesterase (PE) (Voragen, 1989). Another example is in the manufacturing of clear juices, where partial hydrolysis ofth e arabinans by arabinofuranosidase Blead s to undesir able haze-formation (Whitaker, 1984). To influence the composition of a pectinolytic or hemicellulolytic preparation to some degree, three approaches have been followed: 1) changes in the fermentation conditions of the fungus (differences in carbon source), 2) the use ofA n^ mutants producing more, or less, of certain components, 3) treatment of the enzyme cocktail after fermentation to selectively decrease certain enzyme activities. The development of molecular genetics of filamentous fungi is leading to a better understand ing of the regulation of the genes involved, to ways of influencing the composition of enzyme cocktails, to producing enzyme preparations which contain virtually one enzyme only and to altering physico-chemical properties and specificity of enzymes. 5.9.1 Gene regulation Fungal genes are probably regulated in at least two ways: 1) by pathway-specific regula tion, whereby several genes in a pathway are co-regulated, 2) by wide-domain regulation, whereby a whole group of genes is switched on or off in response to the availability of carbon or nitrogen in the medium. It is generally poorly understood how genes encoding pectinolytic or hemicellulolytic enzymes are regulated. Progress has recently been made in studying the regulation of the Aspergillusendo-1,4-y5 - xylanase-encoding xlnA gene (de Graaff et al., 1992). Promoter-deletions were made and the effects ofthes e deletions were studied in transformants. From the results it was deduced that the region -100 to -250 with respect to the TATA box contains an essential functional 242 dement involved in the induction of gene expression. Nucleotide sequence analysis of this region has revealed the presence of a triplicate repeat of 14 nucleotides. Cloning of this putative regulatory element upstream of the reporter gene glucose oxidase (gox) rendered a xylan-inducible gox expression. Carbon catabolite repression is also mediated via this regulatory element, which is shown by repression oï gox expression when the fungus is grown on a combination of xylan and glucose. It remains to be seen whether both types of regulation work via separate binding sites on the regulating fragment or whether carbon catabolite repression functions by modifying the activating protein. 5.9.2 Influencing the composition of enzyme preparations Table 5.9.1 shows three ways of upgrading technical enzyme preparations which are further discussed below. Selectiveoverproduction Cloning of pectinase and hemicellulase genes has led to the construction of enzyme overproducing strains. This has been realized in our laboratory for endo-arabinase and arabinofuranosidase A and B, endo-xylanase, several pectin lyases, polygalacturonases and for pectinesterase. It is also possible to detect new enzymes by first cloning the encoding gene. In commer cial Aspergillus nigerpectinolyti c preparations, usually only two pectin lyases (PLI and PLII) are found (Harmsen et al., 1990). Surprisingly, the pectin lyases are encoded by a gene family of at least six members (Harmsen et al., 1990). To identify the gene products that are absent in the commercial preparation, these genes have been isolated and used for transformation of A. niger.Strain s overproducing pectin lyase PLA and PLD are easily made, and these enzymes are identical to PLI and PLII found in the commercial prepara tion (Harmsen et al, 1990; Kusters-van Someren, 1991). The product of the pectin lyase pelB gene was shown to be proteolytically unstable, and is therefore rather difficult to detect in crude enzyme mixtures. By constructing a gene fusion between the promoter of the highly expressed pyruvate kinase (pki) gene and the structural part of the pelB gene, Table 5.9.1. Upgrading of technological enzyme preparations obtained by recombinant DNA techniques. Method Result Expression under a heterologous promoter Overproduction of a specific enzyme in the absence of other related enzymes Gene disruption Elimination of a specific undesired activity Protein engineering Altering enzyme characteristics: - enhancing thermostability - altering pH profile - altering substrate specificity 243 high production levels of PLB can be achieved (Kusters-van Someren et al., 1992). An other major advantage of using such a gene fusion is that the enzyme can be produced in fermentations using glucose as the sole source of carbon. Under these conditions, virtually no other contaminating extracellular enzymes are produced. PLB was purified from a pki-pelB gene fusion transformant and shown to have kinetic properties that are rather different from those of PLA and PLD (Kester et al., in preparation). To identify the gene products oipelC, E and F a similar approach will be followed. Thus, new enzymes, with possibly new applications, can be identified. Selective inactivatwn It is also possible to selectively abolish an enzyme activity by gene disruption (Fig. 5.9.1). The unwanted enzyme should first be identified and purified. Amino acid sequences from the N-terminus of the mature protein or from peptide fragments must be determined in order to design oligonucleotides which can be used as probes for the screening of a gene library or in PCR. Once the encoding gene has been cloned, a plasmid can be made which contains the gene of interest, disrupted by a selectable gene such as the pyrA. gene which encodes orotidine 5'-phosphate decarboxylase. Transformation with this construct of a pyrA' strain and selection for uridine auxotrophs will select for transformants which have integrated the plasmid in the chromosomal DNA. In some transformants the integration site will be the homologous pectinolytic gene, which may now have been replaced by the disrupted gene. In such a transformant, enzymatic activity encoded by the disrupted gene will be abolished. geneX pyrA gene plasmidwit h disrupted gene X gene X on chromosomal DNA transformation of pyrA' strain selectlo r pyrA+ disruptedgen eX transformants onchromosoma l DNA gene replacement bydoubl ecrossove r Fig. 5.9.1. Method for gene disruption in Aspergillus. 244 5.9.3 Towards enzyme engineering The A. nigerpolygalacturonase s are also encoded by a gene family. The production levels of PGs are higher than those of PLs, and therefore more iso-enzymes can be detected in a commercial preparation (Kester & Visser, 1990). Purification and characterization of these enzymes has made it possible for us to detect differences in maceration capacity and substrate specificity between the individual enzymes. The nucleotide sequence has been determined for three genes. It would seem that the individual members of this gene family have homologous amino acid sequences. Amongst all PGs (bacterial, fungal and higher plants) there is one conserved histidine residue. This histidine is thought to be present in the catalytic centre of PGs (Cooke et al, 1976). Solving the three-dimensional structure for a polygalacturonase would lead to more precise information about substrate binding and residues involved in catalysis. As part of the strategy to establish the role of particular amino acid residues, other defined protein mutants can be made by site-directed mut agenesis. The kinetic properties of such mutants can be further investigated in combina tion with a structural crystallographic analysis. This approach may be considered as a first step in modifying the properties of such enzymes. From an application point of view changes in pH profile and in thermostability are important targets. 5.9.4 Outlook It is expected that detailed basic knowledge on the molecular genetics of pectinolytic and hemicellulolytic genes and the structure and catalytic properties of the encoded enzymes will finally lead to: - modifying existing application processes, for example by replacing chemical steps by enzymatic ones; - more efficient and reliable upgrading of agricultural products within a broad range of applications (extraction, stabilization, maceration, liquefaction, digestibility); - reducing in waste volumes due to improved processing and - further upgrading of agricultural wastes, made suitable, for example, for animal feeds. 5.10 Detection and identification of food-borne bacteria by means of molecular probes A.H. Weerkamp, N. Klijn & W.M. de Vos Food-borne diseases of microbial origin are a major health problem, and microbial spoil age causes economically significant losses in food production. Accordingly, both the food industry and the public health agencies have a vested interest in rapid and reliable methods of detecting and identifying the specific micro-organisms present in food and starter cultures and in the environment of the production facilities. The classic methods used to detect and identify micro-organisms are slow and laborious and are often indiscriminating. Many approaches are being made to enhance the sensitivity of detecting microbial 245 activity. The specificity of detection can, in principle, be significantly improved by apply ing molecular probes, such as immunological or nucleic acid probes. A general review of these approaches is given by Huis in 't Veld et al. (1988). The present contribution focuses on the use of specific nucleic acid hybridization probes for various applications in the food industry and on increasing the sensitivity of detection using the polymerase chain reaction (PCR). 5.10.1 Nucleic acid hybridization probes All organisms contain portions of genetic material that are unique in terms of the sequence of the bases in nucleic acids. These sequences can be detected using gene probes in a hybridization assay (Walker & Dougan, 1989). Nucleic acid hybridization has long been a tool for taxonomie purposes in microbiological research. In its most simple form bacteria are lysed, the double-stranded DNA separated by heat or alkali treatment, the DNA attached to a membrane filter, and the relevant hybridization probe, tagged with a suitable radioactive or non-radioactive label, added and allowed to hybridize with the unknown counterstrand DNA on the filter. After incubation and removal of the unbound probe, the label is detected and quantitatively reflects the presence of the desired bacteria in the sample. The practical use of modern nucleic acid techniques for detection and identification was first applied in the clinical laboratory to detect and identify micro-organisms that were difficult to grow or to be characterized with classic physiological properties. Also, because such techniques may circumvent the need of a growth step, identification can be per formed much more rapidly. Finally, due to the fact that nucleic acid hybridizations detect the primary genetic information rather than the expression products of this information, much higher specificity can be realized. In the food industry, time is also a significant factor when pathogens or spoilage organ isms need to be detected in raw material, ingredients and processing equipment, and in the product before it can be released. With increased scale and process control of the industrial food fermentations, the need for rapid monitoring of the growth and activity of starter bacteria also increases. DNA-probes One of the first uses of DNA-DNA hybridization was in studies to determine the taxo nomie relationship between different micro-organisms using total radioactively labelled DNA from one strain hybridized with total unlabelled DNA from another strain. This, for example, led to the differentiation of the genus Streptococcus into the three new genera Lactococcus, Enterococcus and Streptococcus by Schleifer et al. (1985). Further refinement became possible with the development of gene cloning techniques and the ability to synthesize oligonucleotides of any nucleic acid sequence in large quanti ties. Thus, oligonucleotide probes were targeted towards specific genes or other DNA fragments. Strategies for designing DNA probes and some applications for agroindustrial purposes are shown in Table 5.10.1. Probes can also be based on non-chromosomal DNAs, such as bacterial plasmids. 246 Specific plasmid DNA sequences were for example used by Cocconcelli et al. (1991) to follow the colonization of grass silage by certain strains of Lactobacillus and Pediococcus. Functionally unknown DNA sequences may also be used to target specific oligonucleo tide probes. In this case, chromosomal fragments from the organism to be studied are cloned in a host. The cloned sequences are then screened for their usefulness in detection applications by hybridization with chromosomal DNAs of a variety of related and unre lated organisms. Accordingly, selected probes were developed to identify Lactobacillus del- brueckii (Delley et al., 1990) and Lactobacillus helveticus (Pilloud & Mollet, 1990). Apart from plasmid DNA and some cases of repeating chromosomal DNA sequences, the targets for DNA probes are present in the cell in only one single copy. This obviously limits the sensitivity of the test since each cell generates only one signal. To increase the number of targets per cell, two strategies can be followed, i.e. the use of multicopy nucleic Table 5.10.1. Examples of DNA hybridization probes for identification of micro-organisms poten tially useful for the food industry. Abbreviations: AP, alkaline phosphatase; r.a., radioactivity; PCR, polymerase chain reaction. Target organism Target DNA Location Detection specificgenes: Listeria monocytogenes haemolysin A {hlyA) chromosome luminescence L. monocytogenes delayed hypersensiti chromosome r.a. vity factor (DTH) Escherichia coli (general) /^-lactamase chromosome biotin/streptavidin E. coli (hemorrhagic) verotoxin chromosome r.a. E. coli (enterotoxigenic) heat-labile toxin plasmid PCR, r.a. Salmonella sp. DNA replic.origin chromosome PCR, digoxigenin Salmonella sp. IS200 insertion elem. chromosome AP/fluorochrome Staphylococcus aureus (general) thermonuclease chromosome r.a./digoxigenin S. aureus (enterotoxigenic) enterotoxin genes chromosome r.a. Lactococcus lactis (nisin+) nisin (nisA) transposon PCR, r.a. random DNA elements: Campylobacter sp. chromosome r.a. Escherichia coli(enteroinvasive ) plasmid r.a. Lactobacillusplantarum plasmid digoxigenin Salmonella sp. chromosome r.a. Yersinia enterocolitica plasmid r.a. ribosomal RNA: Campylobacter sp. PCR, r.a. Carnobacterium sp. 16S PCR, r.a/AP Clostridium tyrobuyricum 16S r.a. Lactococcus sp. 16S/23S PCR.r.a. L. monocytogenes 16S r.a./AP Salmonella sp. 16S luminescence 247 acid targets such as ribosomal or messenger RNA, or in vitro amplification of the target sequence via PCR (Bloch, 1991). Ribosomal RNAprobes The bacterial ribosome contains three RNAs of different sizes, the 5S, 16S and 23S RNA. The genes coding for ribosomal RNA are chromosomally located and often present in multicopies. Such RNAs contain various regions with sequences that strongly differ and sequences which are very much conserved among various micro-organisms (see Fig. 5.10.1). The variable regions are often different enough to allow discrimination between species and even subspecies of bacteria. On the other hand, the conserved flanking regions offer the opportunity to design general primers for PCR-amplification of the variable regions (Fig. 5.10.1). This allows identification of organisms using very small quantities of cellular material (Klijn et al., 1991). 16S rRNA sequences of many different bacteria species and strains (>1000 ) are already known, whereas a smaller number ofth e 23S rRNA sequences ofbacteri a are known. The latter molecule may be more suitable for the development of DNA probes because of the larger variable regions (Betzl et al., 1990). 5.10.2 Specific applications of nucleic acid probes for the dairy industry Identification ofstarter bacteria andspecific genes Lactococcus and Leuconostoc species are the main components of starter cultures used for the manufacture of soft and semi-hard cheeses, butter and buttermilk. Identifying these organ isms is difficult because of the morphological and physiological similarity between these genera and their species and subspecies. The distinction is,however , important in terms of process and product control. Also, specific detection and identification of the organisms is significant in studies directed to risk-assessment of genetically modified starter organisms in food fermentation (Klijn et al., 1992a). Analysis of the sequences of 16S RNA of both genera (Collins et al., 1989; Martinez- Murcia & Collins, 1990; Salama et al., 1991) showed that the VI and V3 regions were most variable for Lactococcus and Leuconostoc species, respectively. Based on the differences in length and sequence, specific probes were developed which allowed differentiation at the species and subspecies level (Figs. 5.10.2 and 5.10.3). Testing the probes with other, closely related organisms from the genera Enterococcus, Streptococcus and Lactobacillus demon strated that the probes made reliable identification possible. Combined with PCR-amplifi cation of the VI and V3 regions, the material from one single bacterial colony on an agar plate was enough to obtain reliable identification of the bacterium (Klijn et al., 1991). A similar strategy can be used to detect the presence of specific genes in starter bacteria or in food samples. Fig. 5.10.4 demonstrates the specific detection of the P-/?-galactosidase (lacG) gene of lactococci in Gouda cheese of different ages. The viable counts in the two weeks and eight months old cheeses were approximately 108 and 6xl02 cfu/g, respectively. The nisin gene (nisA) is only detected in cheeses prepared with a nisin-producing starter culture. 248 Fig. 5.10.1. Representation of the secondary structure of 16S ribosomal RNA. = 5'-terminus. Conserved regions are shown in thick solid lines and variable regions in thin lines (broken lines represent regions which have only been shown in a few organisms; Neefs et al., 1990). The positions of the PCR primers described in this paper (P1-P4) are indicated by arrows. Identification offouling bacteria in dairy equipment The operating time of industrial evaporators for milk and whey is largely limited by the development of fouling microflora on the walls of the equipment. Studies by Langeveld et al. (1990) demonstrated the growth of unidentified, non-sporulating thermophilic bacteria (NSTB) which could be subdivided into two groups on the base of the optimal growth temperatures (60-75 °C and 45-58 °C, respectively). Based on morphological and physio logical characteristics, the high-temperature group was suspected to be related to Thermus, a genus so far not described in similar environments. Recent results obtained by Bateson et al. (1990) suggested that the 16S rRNA V3 region contained a genus-specific sequence. Fig. 5.10.5 shows that by using a probe based on this sequence, it was possible to identify the high-temperature NSTB group as belonging to the genus Thermus. 249 123456789 _90b &^të^ Fig. 5.10.2. Identification of the starter and non-starter Lactococcus species. Upper panel: separation of PCR-amplified VI regions of 16S rRNA using primer set P1,P2 on agarose gel stained with ethidium bromide. Lane 1: Lambda-DNA digested with Hindlll; lane 2: L. lactis subsp. lactis NCFB 2597; lane 3: L. lactis subsp. lactis var. diacetylactis NCFB 176; lane 4: L. lactis subsp. cremoris NCFB 1200; lane 5: L. lactis subsp. hordniae NCFB 2181; lane 6: L. plantarum NCFB 1869; lane 7: L. gawieae NCFB 2155; lane 8: L. rqffinolactis NCFB 617; lane 9: V.fluviale NCF B 2497. Following fixation of the DNA to membrane filter, hybridizations were with specific probes for (lower panels): L. lactis subsp. lactis (A),L. lactis subsp. cremoris (B),L. plantarum (C), L. gawieae (D) and L. rqffinolactis (E). Identification and detection of cheese spoilage bacteria Production of ripened semi-hard and hard cheeses is threatened by the presence of spores of the anaerobic spore-forming bacterium Clostridium, tyrobutyricum in the cheese-milk, which survive cheese-milk pasteurization. Even extremely low levels of spores (1 spore per ml) may lead to excessive production of butyric acid and hydrogen gas, which affect both taste and structure of the cheese (late blowing). To avoid significant economic losses, early detection and quantification of this organism in the cheese-milk is of prime importance. However, current methods are too slow and lack sufficient specificity to distinguish C. tyrobutyricum from non-spoiling Clostridia. 250 1234567 1234567 A B 90b _ C D _360b Fig. 5.10.3. Identification of starter and non-starter Leuconostoc species. PCR-amplified DNAs of regions VI (lane 2 to 4) and V3 (lane 5 to 7) of 16S rRNA genes using primer sets P1,P2 and P3,P4, respectively, were separated on 2% agarose gel and stained with ethidium bromide (Panel A). Lane 1: pU C 18 digested with Hpall; lanes 2 and 5: Lc. mesenteroides NCFB 523; lanes 3 and 6: Lc. lactis NCFB 533; lanes 4 and 7: Lc.paramesenteroides NCFB 503 (Lactobacillus). Gels run in parallel that contained identical samples were blotted and hybridized with DNA probes specific for Lc spp. (VI) (panel B), Lc. lactis (V3)(panel C) and Lc. mesenteroides (V3)(panel D). 1 2 3 4 5 6 1 2 3 4 5 6 LacG NisA Fig. 5.10.4. Detection of lactococcal genes using PCR amplification in Gouda cheeses of various age. DNAs obtained after PCR amplification of DNA extracts from the cheeses, using primers based on specific sequences of the nisin (nisA; left panel) and P-/?-galactosidase (lacG; right panel) genes, were separated on 2% agarose gels and stained with ethidium bromide. Lane 1, negative control; Lane 2, positive control; Lane 3, Gouda cheese prepared with commercial mixed-strain starter at an age of 2 weeks; Lane 4, the same cheese at 8 months; Lane 5, Gouda cheese produced with a nisin-producing starter culture, 2 weeks of age; lane 6, the same cheese at 8 months. 251 co _D !E Q. O o E to c CD > o o O o *4— Oo o to CO m > 3 D in < L_ 4-» ^—^—- ,—*—. Z tu CO Ei_ E O) Q .*: O) r- CD tU 00 CD r--. O c 1- 1- Z u O O I -> c E 1 23456789 101 1 L_450bp Thermus sp. Fig. 5.10.5. Identification of thermophilic non-spore-forming bacteria from evaporating equip ment for milk and cheese whey. PCR-amplified DNAs from the 16S rRNA V3 regions of the bacteria using primer set (P3,P4) were separated on agarose gel and stained with ethidium bromide (upper panel). After blotting, the DNAs were hybridized with a 77i«TOZ«-specific DNA probe. N8, Gl, C6, C7, H andJ represent isolates with the optimum growth temperature shown (Langeveld et al., 1990). Determination of the ribosomal RNA sequence of C.tyrobutyricum (va n der Meer et al., 1992) and comparison with 16S RNA sequences of closely related Clostridia, made it possible to design probes specific for the target organism as well as for other Clostridia (Klijn et al., 1992b). Fig. 5.10.6 shows that the C.tyrobutyricum prob e can be used success fully to identify this organism from related Clostridia commonly found in raw milk and cheese. The probes will be used in a collaborative study under the EC Eclair programme aimed at developing quantitative detection methods for use in the dairy industry. 5.10.3 C onclusions Modern DNA techniques allow the development of microbial detection and identification techniques which are faster, less laborious and often more specific than the classic method ology. In combination with PCR-amplification the methods also promise high sensitivity and accuracy, requiring only very small amounts of sample. Various applications for the food industry that are currently being developed are directed towards improving the microbiological safety and quality of the products and the control of the manufacturing process. 252 probe: 16SrRN A •fc '4^4^641 <£ £ A £ probe: tyrobutyricum V2 :# probe: tyrobutyricum V6 A: C. tyrobutyricum B: C sporogenes C: C. butvricum Fig. 5.10.6. Identification of Cbstridiumtyrobutyricum. Do t blot hybridization of total DNA extracts ofClostridi a obtained from cheese,ra w milk, silage and other environmental materials. The bacteria were identified by using classical identification techniques. The DNAs were hybridized with 16S rRNA probes. Upper panel: general probe. Other panels: two different specific probes for C. tyrobutyricum. The positive signal in the lower lane appeared to result from the previous misidenti- fication of a C.tyrobutyricum strai n that was thought to belong to C. butyricum. Some commercially developed and relatively simple test-kits for detecting food patho genic bacteria, such as Listeriamonocytogenes and Salmonella species, are now available for routine laboratories of the food industry. Since these kits require relatively high concentra tions of the bacteria (>106), their main suitability is to confirm the identity of the bacteria. Further, more critical applications will depend on the development of sensitive non radioactive labelling procedures and the automatization of the sample preparation proce dure for the PCR technique. Considerable research will also need to be done to improve enrichment procedures that can be applied to various food products. 253 5.11 The development of immunoassays for the rapid detection of moulds G.A. de Ruiter, H.J. Kamphuis, S.H.W. Notermans,J.H . van Boom & F.M. Rombouts Many mould species are important for foods, because they can cause food spoilage result ing in huge economic losses (Pitt & Hocking, 1985). However, moulds are also used for fermentation purposes such as in the manufacture of tempe (Rhizopusspecies ) and many cheeses (Pénicillium species). Some species, mostly those belonging to the generaPénicillium, Aspergillus and Fusarium, are able to produce mycotoxins which cause a wide variety of mycotoxicoses in man (Frisvad & Samson, 1991). As the number of mycotoxins which can be present in foods is more than 400, detecting each of them accurately is almost impossi ble. Also, the processing of foods often includes filtration or heat treatment through which the viable parts of moulds are removed or inactivated but leaving behind mycotoxins. Methods commonly used for detecting moulds in industrial food products have been reviewed byjarvis et al. (1983) and Kamphuis (1992a). These include the frequendy used plating techniques (mould colony count) and microscopic techniques such as the Howard mould count. Plating techniques are based on the enumeration of viable parts, including asexual and/or sexual propagules. Mycelium, even when present in large numbers, usually leads to low mould colony counts. Heating or filtration leads to the inactivation or removal of viable mould parts, resulting in uncertainty about the mycological quality and safety of the products. These methods therefore have serious drawbacks for use in the food industry and food inspection services. Methods, which are able to recognize moulds in food under all circumstances, will be more feasible for evaluating the possible risks of fungal contami nation and may result in reducing the number of mycotoxin analyses. Accordingly, there is a continuous need to improve existing methods and to develop new ones for the detection of fungal contamination. In this section, we describe the development ofimmunoassay s for the rapid and sensitive detection of moulds in food and the evaluation of their industrial use. The immunoassays developed are based on the detection of antigenic extracellular polysaccharides produced by moulds which are heat-stable and water-soluble and thus suited for detecting moulds in foods, even after heat-processing or filtration. 5.11.1 Development of immunoassays based on detection of antigenic extracellular polysaccharides Intracellular or extracellular biopolymers of moulds are antigenically active, heat stable, mostly water soluble and mainly composed of carbohydrates with some protein and therefore often referred to as extracellular polysaccharides (EPS). It has been demon strated that EPS of Pénicillium and Aspergillus species contain sequences of yS(l-5)-linked D-galactofuranose residues which are responsible for their antigenic properties. This re sulted in a new structural model for the antigenic galactofuranose residues as shown in Fig. 5.11.1. (Notermans et al., 1988a; van Bruggen-van der Lugt et al., 1992). Also, EPS preparations isolated from moulds belonging to Mucorales contain one common antigenic 254 ri ri ß-Gal/-(l-»5)-ß-Gal/-(l-5)-ß-Gal/-(l-»5)-ß-Gal/-(l-5)-ß-Gal/-(l-Mannan L JK 6 L JL t r i i ß-Gal/-(l-»6)-ß-Gal/-(l-*6)-ß-Gal/' L JM Fig. 5.11.1. New structural model for the antigenic galactofuranose side-chains of the extracellular polysaccharides of Pénicillium and Aspergillus species according to van Bruggen-van der Lugt et al. (1992). Only the outer /?(l,5)-linked galactofuranoside chains are antigenically active. fraction (de Ruiter et al., 1991a). However, until now no information has been available on the composition of the epitopes of polysaccharides from mould species other than Aspergillus and Pénicillium. Many polyclonal antibodies against intracellular or extracellular polysaccharides of moulds have been raised in rabbits (Hearn &Mackenzie , 1980; Notermans & Soentoro, 1986; Kamphuis et al., 1989b; Cousin et al., 1990; de Ruiter et al., 1991b). Recently, the development of monoclonal antibodies (IgM) against several fungal polysaccharides has also been described (Stynen et al, 1992b; Fuhrmann et al, 1992). Using IgG antibodies raised against EPS preparations of different moulds, various formats of immunoassays could be developed. Sandwich ELISAs (enzyme-linked im munosorbent assays) were developed for detection of the mould species Pénicillium and Aspergillus, Mucor and Rhizopus, Botrytis and Monascus, Fusarium, Cladosporium and Alternaria which are reviewed by Cousin (1990). The development of a mould latex agglutination assay (MLA) with antibodies specific for Pénicillium and Aspergillus species was described by Kamphuis et al. (1989a). This assay format as shown in Fig. 5.11.2 is much more convenient for practical use in the food -V J** Latex beads coated EPS antigens Agglutination with IgG antibodies Fig. 5.11.2. Schematic principle of the mould latex agglutination assay (MLA). The assay was based on the recognition of antigens to extracellular polysaccharides by specific IgG antibodies coated onto latex beads resulting in a visible agglutination. 255 industry, because agglutination can be observed visually without any expensive laboratory equipment. The latex agglutination assay isslightl y less sensitive than the sandwich ELISA but satisfactory results can be obtained within 10 to 20 minutes. 5.11.2 Specificity and practical use of immunoassays for mould detection A successful use of immunoassays for detecting moulds in food requires antibodies, either polyclonal or monoclonal, with specificity for a certain group of related genera. Many antibodies are reported for this purpose as summarized in Table 5.11.1. However, some antibodies give cross-reactions with mould species belonging to other non-related genera. Most users will consider these to be inappropriate for reliable use. In general, antibodies raised against EPS from Pénicillium species are specific for nearly all species belonging to the genus Pénicillium and the related genus Aspergillus, due to their common /?(l-5)-galac- tofuranose epitope. Mould species belonging to the order of Mucorales (including the related genera Mucor, Rhizopus, Rhizomucor, Syncephalastrum, Absidia and Thamnidium) also contain a common antigenic determinant, resulting in specific antibodies for this group of moulds. Similar results have been obtained for mould species of the related generaBotrytis and Monascus.A promising approach to obtain specific antibodies was proposed by Kam phuis et al. (1989b). They raised antibodies towards synthesized/?(l-5)-linked galactofura- nose oligomers, representing the epitopes of Pénicillium and Aspergillus species, and coupled them to a protein. After immunizing rabbits with this conjugated protein, specific antibod ies could be isolated from the serum which were specific for Pénicilliuman dAspergillus (Kamphuis et al., 1989b). Such an approach isonl y possible ifth e epitopes of the antigenic EPS from different mould genera are known in detail. The practical use of immunoassays in the food industry requires an easy test format. However, all immunoassays are susceptible to non-specific reactions in food products as a result of their often very complex matrix. False-negative results can be recognized easily by adding purified EPS derived from the moulds involved to the test sample. Consequendy, a negative result will be converted into a positive one and a false-negative sample remains negative. Recognizing false-positive reactions is more difficult. One way of recognizing such reactions is to add a synthetic epitope to a test sample which prevents a reaction between EPS and the IgG molecule. Due to this specific inhibition a positive result should be converted into a negative one by specifically blocking the immunological binding sites (paratopes) of the antibodies used, as schematically shown for an ELISA in Fig. 5.11.3. This principle for the recognition of false-positive results can be used in both ELISA and MLA. 5.11.3 Laboratory analysis of foods using immunoassays Results of the reports on the laboratory analysis of foods and on the contamination of Pénicillium and Aspergillus species by immunoassays as shown in Table 5.11.2 are mainly based on the mould latex agglutination assay (MLA) as developed by Kamphuis et al. (1989a). The pre-treatment necessary for the successful use of immunoassays is simple. 256 Table 5.11.1. Specificity ofpolyclona l and monoclonal antibodies raised against different mould antigens tested by enzyme-linked immunosorbent assay (ELISA). , moulds against which monoclo nal antibodies (IgM)ha d been raised. Antibodies raised Specific for Cross- Reference against mould genera reactivity to other genera Pénicillium digitatum Pénicillium, Aspergillus no Notermans & Soentoro, 1986 Pénicillium verrucosum - var. cychpium Pénicillium, Aspergillus no Notermans & Soentoro, 1986 Pénicillium verrucosum - var. verrucosum Pénicillium, Aspergillus yes Fuhrmann et al., 1989 ß{1 -5)-Gah roligomers Pénicillium, Aspergillus no Kamphuis et al., 1989b Pénicilliumfréquentons Pénicillium, Aspergillus no Fuhrmann et al., 1992 P. aurantiogriseum Pénicillium, Aspergillus yes Banks et al., 1992 Aspergillusfumigates Pénicillium, Aspergillus yes Stynen et al., 1992a Pénicillium islandicum Pénicillium no Deweyetal., 1990 Mucor racemosus Mucor, Rhizopus yes Notermans & Soentoro, 1986 Mucor racemosus Mucor, Rhizopus, Rhizomucor,Absidia, Syncephalastrum, Thamnidium no De Ruiter et al., 1992b Mucorpiriformis Mucor, Rhizopus yes Robertson &Patel , 1989 Rh.iz.opus stoknifer Rhizopus no Lin &Cousin , 1987 Rhizopus stolonifer Mucor, Rhizopus yes Robertson &Patel , 1989 CI. cladosporioides Cladosporium no Notermans & Soentoro, 1986 Fusarium oxysporum Fusarium yes Notermans & Soentoro, 1986 Fusarium solani Fusarium yes Robertson &Patel , 1989 Fusarium culmorum Fusarium yes Banks et al., 1992 Fusarium oxysporum Fusarium no Ianelli et al., 1983 Fusarium oxysporum Fusarium no Kitagawa et al., 1989 Botrytis tulipae Botrytis no Cousin et al., 1990 Botrytis cinerea Botrytis yes Robertson &Patel , 1989 Monascuspilosus Monascus no Cousin et al.,199 0 Geotrichum candidum Geotrichum no Notermans & Soentoro, 1986 Geotrichum candidum Geotrichum no Lin & Cousin, 1987 Routinely, food products are diluted 10-fold with phosphate-buffered saline (70 mM sodium phosphate, pH 7.2, 150 mM NaCl) and homogenized with the stomacher for one minute. The supernatants after centrifugation are used in the immunoassays. Use of the citrate buffer (pH 7.2) is reported to be better for dairy products, especially cheese, due to their high content of calcium caseinate and fat (Tsai & Cousin, 1990). Probably citrate complexes with Ca2+-ions, resulting in the complete solubilization of caseinate. Immunoas- 257 Ig G < Ab. positive ELISA negative ELISA Fig. 5.11.3. Recognition of false-positive ELISA reactions by addition of synthesized epitopes to the sample, converting a positive result into a negative. IgG, immunoglobulin G coated onto the wall of a polyvinyl microtitre plate; a, antigenic extracellular polysaccharides; b, carbohydrate epitopes; c, peroxidase enzyme conjugated to IgG antibodies. says are extremely sensitive. Generally, antigenic EPS as low as 1 ng/ml can be detected in these supernatants using ELISA or MLA. The results are often expressed as the titre, defined as the reciprocal dilution of the supernatant which just shows a positive im- munoreaction. For example, if grapejuic e tested on Mucoralean moulds by ELISA shows a titre of 1000, approximately 1jUg/m\ antigenic EPS of these moulds is present in the sample. This amount can originate from only one mouldy grape per 10 litres ofjuice . For many products such as spices, cereals, fruit juices, dehydrated foods, animal feed, cheese, coffee and cocoa, these immunoassays are used successfully to estimate fungal contamina tion. However, a good correlation with the mould colony count, based on detection of the viable parts, is not always observed. The basic principles of both methods are completely different. It is to be expected that the immunoassay provides a better indication of the (previous) presence of moulds in most samples than plating techniques, as many food products are heat-treated, filtered or gamma-irradiated. In addition, the use of immunoassays can reduce the analysis time considerably, espe cially in dehydrated products. As shown in Table 5.11.2, no colony-forming units (cfu) were observed in the pasteurized fruit juices. In these samples, previous contamination of Pénicillium and Aspergillus species could be specifically detected with immunoassays, detect ing the extracellular polysaccharides of these moulds which are heat resistant. 258 Table 5.11.2. Detection of contamination of Pénicillium and Aspergillus species in different types of food products by the use of immunoassays. ' MLA, mould latex agglutination assay; ELISA, en zyme-linked immunosorbent assay; 2 false-negative results; 3 cfu, colony-forming units; 4 NT, not tested. Food product Format' Number of False Correlation Reference samples positive with other results methods Spices ELISA 6 NT4 aflatoxins Notermans et al., 1986 Spices MLA 9 0 cfu3 Kamphuis et al., 1989a Spices MLA 17 0 cfu Notermans & Kamphuis, 1990 Spices ELISA 9 0 cfu Notermans et al., 1988b Spices MLA 26 0 cfu van der Horst et al., 1992 Spices MLA 15 0 cfu Braendlin & Cox, 1992 Spices MLA 49 0 cfu Karman & Samson, 1992 Nuts ELISA 4 NT aflatoxins Notermans et al., 1986 Nuts MLA 7 0 cfu Kamphuisetal., 1989a Nuts ELISA 9 0 cfu Notermans et al., 1988b Nuts MLA 9 0 NT4 van der Horst et al., 1992 Nuts MLA 12 0 aflatoxins Braendlin & Cox, 1992 Nuts & oil seeds MLA 30 0 cfu Braendlin & Cox, 1992 Walnuts MLA 3 1 cfu Kamphuis et al., 1989a Walnuts ELISA 5 2 cfu Notermans et al., 1988b Walnuts MLA 10 3 cfu Notermans & Kamphuis, 1990 Walnuts MLA 23 192 cfu Karman & Samson, 1992 Cereals ELISA 16 NT aflatoxins Notermans et al., 1986 Cereals MLA 27 0 cfu Notermans & Kamphuis, 1990 Cereals MLA 78 0 cfu Karman & Samson, 1992 Cereals MLA 23 0 aflatoxins Braendlin & Cox, 1992 Cereals & pulses MLA 60 0 cfu Braendlin & Cox, 1992 Maize products MLA 35 3 cfu Kamphuis et al., 1992b Juices MLA NT Notermans & Kamphuis, 1990 Fruit juices MLA 117 0 NT Karman & Samson, 1992 Fruit pulps MLA 84 0 cfu Karman & Samson, 1992 Tomato products MLA 17 0 cfu Van der Horst et al., 1992 Dried products MLA 53 0 cfu Karman & Samson, 1992 Dried products MLA 22 0 cfu Braendlin & Cox, 1992 Animal feed MLA 13 0 aflatoxins Braendlin & Cox, 1992 Animal feed MLA 13 0 aflatoxins Braendlin & Cox, 1992 Coffee & cocoa MLA 16 0 cfu Braendlin & Cox, 1992 Dairy products ELISA 12 NT biomass Tsai & Cousin, 1990 Cheese products MLA 39 0 cfu Karman & Samson, 1992 259 Specific detection of Mucoralean moulds can be performed using an immunoassay as shown in Table 5.11.3. In most samples, no correlation could be established with the number of colony-forming units of these moulds as most food products were either pas teurized or heat-processed in another way. Samples of flour showed a good correlation between colony count ofspecie s ofMucorale s and the ELISA titre (de Ruiter et al., 1992a). An important quality aspect of foods contaminated with moulds is the possible presence of mycotoxins. Many assays, either chemical or immunochemical, are available for the detection ofvariou s mycotoxins in food. Since there are more than 400 known mycotoxins this is a very laborious approach. It is more feasible to detect the (previous) presence of the moulds that are able to produce mycotoxins as proposed by Notermans et al. (1985). Immunoassays can be used to assess the quality of the food and to identify those samples that need to be tested further for mycotoxins. As shown by the studies of Notermans et al. (1986) and Braendlin & Cox (1992) this approach is promising, as in most samples high amounts of aflatoxins correlate with high titres of the immunoassay. However, more fundamental research on this correlation is needed to establish the level of antigenic EPS which signals the possible presence of different mycotoxins in food. A successful immunoassay requires having the possibility of recognizing false-positive and false-negative reactions which may occur as a result of the complexity of the food samples. As shown in Table 5.11.2, samples of walnuts, for example, are extremely sensi tive to both false-positive and false-negative reactions using ELISA (Notermans & Kam phuis, 1990). This phenomenon is observed in many reports. As shown by de Ruiter et al. (1992a) this is not limited to immunoassays for Pénicillium and Aspergillus but is also found in immunochemical detection of species belonging to the order of Mucorales. In general, the occurrence offalse-positiv e reactions in food samples has been observed for all sensitive immunoassays. These findings underline the need to test for false-positive results in food analysis using immunoassays. Table 5.11.3. Detection of Botrytis,Monascus, Rhizopus and Mucor species in different type of food products by ELISA. Abbreviations: cfu, colony-forming units; NT, not tested. Food product Mould genera Number of Correlation Reference samples with other methods Spices Mucor, Rhizopus 10 cfu Notermans et al., 1988b Spices Mucor, Rhizopus 26 cfu de Ruiter et al., 1992a Nuts Mucor, Rhizopus 17 cfu Notermans et al., 1988b Nuts Mucor, Rhizopus 21 cfu de Ruiteretal., 1992a Nuts Mucor, Rhizopus 33 NT de Ruiter et al., 1992a Juices Mucor, Rhizopus 39 NT de Ruiter et al., 1992a Flour Mucor, Rhizopus 17 cfu de Ruiter et al., 1992a Dried products Mucor, Rhizopus 25 NT de Ruiter et al., 1992a Spices Monascus, Botrytis 7 NT Cousin et al., 1990 Fruit juices Monascus, Botrytis 33 NT Cousin et al., 1990 Fruit products Monascus, Botrytis 43 NT Cousin et al., 1990 260 5.11.4 Analysis of fungal contamination with commercially available immunoassays In the last three years test-kits for the detection of Pénicillium and Aspergillus moulds have been introduced onto the market by Holland Biotechnology (Leiden, the Netherlands) and Sanofi Diagnostics Pasteur (formerly Eco-Bio; Genk, Belgium). Both are based on the recognition of fungal antigens (EPS) by specific antibodies using the latex-agglutination format. The first experiences with immunological tests for the detection of moulds in food revealed that commercialization is hampered by many factors. As the intrinsic principles of immunoassays are completely different from the frequently used plating techniques, many industrial laboratories have little experience of these types of assays. Also, validation of immunoassays for routine analysis of fungal contamination of food is not easy. As there are no clearly defined widely applicable standards, each user must define the antigen concentration which corresponds to an unacceptable level of moulds in his or her own setting. However, the most important problem for implementing immunoassays in the food industry is the more or less conservative attitude towards new methods of detection. The severe drawbacks of the plating methods are inherendy accepted, probably because these methods are so widely used. A poor correlation between the methods often leads to extensive discussion about the negative aspects ofth e new method, without considering the relevance of the drawbacks of the old methods. Furthermore, there is no clear legislation in most countries for fungal contamination of foods, and the use of these immunoassays often results only from the internal motivation or quality assurance policy of the industrial laboratories. Nevertheless, immunological detection ofmould s iso f extreme importance to the food industry for the control of raw materials, semi-manufactured products and end products. To this end they could develop their own standards. 5.11.5 Future developments Fungal contamination affects not only the quality of food products, but also their safety, because of the production of mycotoxins. At present more than 400 mycotoxins are described, mainly produced by species belonging to the genera Aspergillus, Pénicillium and Fusarium.I n the future, the number of mycotoxins that can be analysed, will increase and the feasibility of testing raw materials or foods for all these mycotoxins will decrease. Detection of the (previous) presence of moulds will itself become more efficient. Methods of detection currenüy being used that are based on enumerating the viable parts present in food, are not very satisfactory because moulds can be inactivated or removed by filtration during food processing, without removing the mycotoxins. More sophisticated methods such as immunoassays may be considered to be much more appropriate for this purpose. As described above, immunoassays are sensitive and can be converted into simple test-kits, allowing reliable results to be obtained within 30 minutes. Widespread implementation of these techniques requires a change in attitude of the industrial control laboratories and food inspection services. The introduction of law- enforced guidelines for the quality of food-products might change this attitude in the 261 future. The availability of test-kits covering all important fungal food contaminants would make antigen detection more attractive. Also, potential customers would be more easily convinced if more published data were available on the practical application of these immunoassays. Acknowledgements These investigations were supported by the Netherlands Foundation for Chemical Re search (SON), with financial aid from the Netherlands Technology Foundation (STW). 5.12 Safety aspects of foods produced by recombinant DNA technology H. Hofstra Safety evaluation and societal acceptance of foods produced by biotechnological tech niques will be a subject of public debate during the next few years, as the first products will have to find their way to the market. After a slow start ofmoder n biotechnology in the late seventies, the rapid technological developments within molecular biology have given rise to numerous possible applications in the food sector. The general problem of acceptance of novel agro-biotechnological developments is comparable with the problems faced by any new technology within society: their acceptance does not only depend upon exact figures regarding risks and risk-assessment or on procedures and regulations accompany ing and determining their implementation into society, but also on more vague concerns about risks or feelings of mistrust which can equally influence societal or political accep tance. These concerns, which are also reflected in risk assessment procedures and decision schemes, will be the main subjects of this contribution. Which subjects will be important in discussions on the acceptance of agricultural bio technology and in particular foods produced by recombinant DNA technology? Basically, all applications of food biotechnology have positive effects, both on health and safety of the foods concerned and also on environmental aspects of the production process. The poten tials of foodbiotechnology are given in several publications. A brief overview of these potentials can be given: - a more effective production of foods and food-ingredients, more cost-effective and less harmful to the environment, - a more rapid and efficient control of food safety, regarding both microbiological and toxicological aspects, - more efficient plant breeding, aimed at higher nutritional value, improved technological properties, lower contents of antinutritional factors and lower production costs, - a reduction in the use of chemical crop protection agents by developing more disease- resistant crops or developing biological alternatives to chemical pesticides, - valorisation of agricultural by-products and wastes by biotechnological means, provid ing e.g. animal feeds with a higher nutritional value at low production cost. 262 Although the development of food biotechnology will lead to better product quality, produced more efficiently, and a reduction in the use ofpesticides , public acceptance is not only determined by such positive effects. Technological developments, also those achieved within the field of recombinant DNA technology, are generally allowed to proceed undisturbed as long as they are performed on a pilot scale inside a research laboratory. Once technology moves to the production stage and new products are approaching the market, regulations regarding their application and discussions regarding their acceptability become relevant. Subjects to be discussed then, are e.g. health hazards and environmental risks, and regulations need to be based on genuine demands with regard to health and environmental safety. Public discussions on acceptance, caused by the anxiety that many people have about genetic modification technology, influences the regulation process. The impact of biotechnology is enormous because, as a technology, it touches funda mental values oflife . This is especially true of food biotechnology as it combines manipula tion of life with perhaps the most intimate relation we have with other living organisms: eating them. Therefore implications of biotechnology, in particular of foods prepared using recombinant DNA techniques, are subject to thorough public discussion. Accep tance of novel foods and food production processes, implying modern biotechnological techniques, must be achieved by providing the right information to consumers. It is essential that society as a whole acquires a basic knowledge of the technology and its implications. This stresses the need for education in schools and universities and also for information to be given by the media. This should initiate or stimulate ample public discussion, leading to well-informed consumers, able to decide what to buy and what to eat. Availability of technical information and background knowledge will, however, not automatically enhance public acceptance. The various interest groups, the 'professionals' (the biotechnologists) on the one hand and the 'activists' (environmentalists, animal protec tionists, etc.) on the other do not normally interact very efficiently, if at all. 5.12.1 The use of recombinant DNA technology in food biotechnology Micro-organisms The two major applications for genetically modified micro-organisms (GEMS) in the food sector are starter cultures in fermented foods and feeds and their use as highly efficient producers of food ingredients. Previously, these micro-organisms had to be isolated from other natural sources in a tedious, time-consuming and costly way. A third application, the production of microbial primary biomass for food has so far been of limited importance, especially in industrialized countries. Starter cultures have always played an important part in the production of fermented foods and drinks for human consumption. They are also important for silage processes both for human food and animal feed production. Most producers of fermented products still rely on the traditional starter micro-organisms. Genetic modification of micro-organ isms offers a number ofchallengin g opportunities to improve starter strains. Improvements will imply e.g. a better taste of the fermented product, lowering or removing the produc- 263 tion of biogenic amines during fermentation, a beneficial influence on the gut flora, an anti-cholesterol activity, possible anti-tumour activity and a positive influence on the immune system. In silage for animal feeds, an important aspect is the detoxification of raw materials, through the removal of natural toxins (mainly mycotoxins) and antinutritional factors. A wide variety of food ingredients will be produced by GEMS in the near future. These may vary from enzymes, produced by micro-organisms (e.g. chymosin), to antimicrobial agents (e.g. nisin), stabilizers, emulgators, flavours. These components are of a relatively simple nature and can, after having been isolated and purified, be added to the foods in a pure form. Plantsand animals In contrast to the microbial products, foods from plant or animal origin are basically all complex mixtures. This implies that a complete evaluation of their properties, such as nutritional value, wholesomeness and toxicological aspects, is difficult. Ifgeneti c modifica tion has only been directed towards a certain aspect, e.g. disease or insect resistance in crops, or if the modification has been limited to the introduction or elimination of one or very few genes, the situation isrelativel y simple. Safety examination could be limited to the genetics of the aspect that has been changed, provided that the food in question has so far generally been regarded as safe. If, on the other hand, the purpose of genetic modification was to drastically increase or modify the amount of a certain natural ingredient in the organism, research is needed to evaluate the consequences of this increase. The same applies when a compound is introduced that does not naturally occur. In both cases, the concentration of the ingredient in relation to the approved daily intake is one of the main points to be addressed. Acceptance in these cases requires complete and open information to the consumer, e.g. a public discussion on the test results of research into safety aspects, as provided by independent laboratories, or on the decisions that led to the final admission of the product to the consumer market. 5.12.2 Food safety aspects with regard to acceptance Apart from many different aspects such as price to the consumer, environmental implica tion and animal welfare, the central theme in the public discussion on biotechnology is the overall safety of foods produced by biotechnological techniques. Foods or food ingredients produced by genetically modified organisms need not necessarily present a greater health risk than those produced by organisms improved by classical breeding and selection. The contrary might even be true as more knowledge is obtained on the molecular level of organisms produced by recombinant DNA techniques in comparison to classically selected organisms. In the case of classical breeding and selection, modification of the genetic content has been carried out as well, be it without control and without the possibility to monitor afterwards what has been changed, not to mention the natural starter cultures used in fermented foods, of which we know practically nothing at all. Therefore, it is more or less accepted in the scientific debate that improving foods by genetic modification is in 264 principle not very different from the classical way of selecting or breeding new varieties or races. Regarding food safety and the authorisation of newly-developed foods, feeds or food ingredients, a number of different categories of materials can be distinguished. Chemical ingredients Chemically well-defined and pure food ingredients, traditionally extracted or isolated from natural sources, such asplants , animals or micro-organisms, are mainly food additives such as flavours, fragrants, colourants, sweeteners. When such ingredients are made by means of genetically engineered micro-organisms, their composition in the pure form will not differ from that of the classical product. Once the absence of any contaminating material from the producing organism, such as microbial toxins or transferable DNA, can be guaranteed, their authorisation and acceptance will probably not present a major prob lem. Foods from plant or animal origin Foods of plant or animal origin present a more complex situation. Safety assessment of foods, feeds or ingredients from biotechnologically modified plants or animals, requires a thorough investigation of toxicological aspects and pleiotropic effects that could result from genetic modification. This is the case when, for example, pleiotropic effects cause a considerable rise in the concentration of certain natural ingredients. The same applies in principle to the situation when the yield offoo d products from animal origin are influenced by external stimulators, e.g. growth-stimulating hormones or other veterinary drugs that have been prepared in micro-organisms using recombinant DNA techniques. In this situation, some risk lies in the possible presence of residues. Safety assessment of these cases, therefore, would principally not be any different from the safety evaluation of chemically made pharmaceuticals. However, the emphatic discussions about the accep tance of BST in milk production, show that things are not this simple. From these discus sions it is clear that safety, or the guarantee of an identical product (milk) or even a lower consumer price, are not enough. Economic factors, such as an increase in milk production, coinciding with the surpluses in dairy production in Europe, adds to the anxiety among consumers. As a result, many consumers will prefer traditional milk, even at a higher price Foods from microbial origin Foods from microbial origin and fermented foods prepared by modified organisms present their own problems regarding safety and authorisation. When new producer microbes are developed by modifying existing food grade bacteria, the newly-developed organism should also be a food grade microbe. This implies that some research into possible plei otropic effects should be part of the procedure. The acceptance of foods produced by genetically modified starter micro-organisms will depend primarily on the proof of their claimed beneficial effects such as the lowering ofblood-cholestero l and anti-tumour effects. In these cases, consumer confidence and acceptance will probably also be increased by other positive side-effects of the biotechnological modification: a better taste, environmen tally safe production, a lower price to the consumer and a longer shelf-life. 265 5.12.3 Goncludin g Remarks Public anxiety regarding the acceptance of foods produced by biotechnological methods stresses the need for open debate on all aspects that might influence consumer confidence in these products. Not only safety aspects but also environmental implication, animal welfare and price to the consumer mustb e considered. Public education through informa tion presented by the mass media will not be enough. A public debate organised in such a way that the scientists themselves participate in the discussion and answer the questions that arise from (future) consumer panels, could be a successful formula. This isi n linewit h the model used in Denmark, where public conferences on biotechnology are organised, which apparently cause a considerable amount ofpublicit y and a lively debate on accep tance. In the end, this should lead to a well-informed consumer who knows what he buys and trusts what he eats. 5.13 References Aguilar, A., 1991. Biotechnology of lactic acid bacteria. A European perspective. Food Biotechnol ogy 5:323-330. Alic, M. & M.H. Gold, 1991. Genetics and molecular biology of the lignin-degrading basidiomycete Phanerochaete chrysosporium. In: J.W. Bennett & L.L. Lasure (Eds.): More gene manipulations in fungi. Academic Press, New York, pp. 319-341. Antila, P., I. Paakkari, A.Järvinen , MJ. Mattila, M. Laukkanen, A. Pihlanto-Leppälä, P. Mäntsälä &J . Hellman, 1991. Opioid peptides derived from in vitro proteolysis of bovine whey proteins. International DairyJourna l 1:215-229. Arai, S., A. Maeda, M. Matsumura, N. Hirao & M. Watanabe, 1986. Enlarged-scale production of a low-phenylalanine peptide substance as a foodstuff for patients with phenylketonuria. Agricul tural and Biological Chemistry 50:2929-2931. Arai, S. & M. Fujimaki, 1991. Enzymatic modification of proteins with special reference to improv ing their functional properties. In: P.F. Fox (Ed.): Food Enzymology, Volume 2. Elsevier Applied Science, London/New York. Asselin,J.,J . Hébert &J. Amiot, 1989. Effects of in vitroproteolysi s on the allergenicity of major whey proteins.Journa l of Food Science 54:1037-1039. Banks,J.N. , SJ . Cox,J.H. Clarke, R.H. Shamsi & B.J. Northway, 1992. Towards the immunological detection of field and storage fungi. In: R.A. Samson, A.D. Hocking,J.I . Pitt & A.D. King (Eds.): Modern Methods in Food Mycology. Elsevier Publishers, Amsterdam, pp. 247-252. Bateson, M.M., K.J. Thibault & D.M. Ward, 1990. Comparative analysis of 16S ribosomal RNA sequences of Thermus species. Systematical and Applied Microbiology 13:8-13. Bellamy, W., M. Takase, K. Yamauchi, H. Wakabayashi, K. Kawase & M. Tomita, 1992. Identifi cation of the bactericidal domain of lactoferrin. Biochimica et Biophysica Acta 1121:130-136. Betzl, D., W. Ludwig & K.H. Schleifer, 1990. Identification of lactococci and enterococci colony hybridization with 23S rRNA-targeted oligonucleotide probes. Applied and Environmental Mi crobiology 56:2927-2929. Bloch, W., 1991. A biochemical perspective of the Polymerase Chain Reaction. Biochemistry 30:2735-2747 Boswinkel, G., B. van der Meer,J.T.F . Keurentjes & K. van 't Riet, 1990. Production of specific fatty acids using membranes. In: JJ. Dekkers, H.C. van der Plas & D.H. Vuijk (Eds.): Agricultural biotechnology in focus in the Netherlands. PUDOC, Wageningen, pp. 228-233. Braendlin, N. & L. Cox, 1992. Immunoagglutination assay for rapid detection of Aspergillusan d 266 Pénicillium contamination in food. In: R.A. Samson, A.D. Hocking,J.I . Pitt & A.D. King (Eds.): Modern Methods in Food Mycology. Elsevier Publishers, Amsterdam, pp. 233-240. Briggs, C.J., 1986. Evening primrose. Canadian Pharmaceutical Journal 5:249-254. Bruinenberg, PJ., P. Vos & W.M. de Vos, 1992. Proteinase overproduction in Lactococcus lacks strains: Regulation and effect on growth and acidification in milk. Applied Environmental Micro biology 58:78-84. Buchman, G.W., S. Banerjee &J.N. Hansen, 1988. Structure, expression, and evolution of a gene encoding the precursor of nisin, a small protein antibiotic. Journal of Biological Chemistry 263:16260-16266. Bussink, H.J.D., F.P. Buxton &J. Visser, 1991. Expression and sequence comparison of theAspergillus tubigensis genes encoding polygalacturonase II. Curr. Genet. 19:467-474. Carasik, W. &J.O. Carroll, 1983. Development of immobilized enzymes for production of high- fructose corn syrup. Food Technology 37:85-91. Chassy, B., 1987. Prospects for the genetic manipulation of lactobacilli. FEMS Microbiology Re views 46:297-312. Christie, W.W., 1991. Fractionation of the triacylglycerols of evening primrose oil by high-perform ance liquid chromatography in the silver ion mode. Fat Science Technology 93:65-66. Cocconcelli, P.S., E. Triban, M. Basso &V . Bottazzi, 1991.Us e of DNA probes in the study of silage colonisation by Lactobacillus and Pediococcus strains.Journa l of Applied Bacteriology 71:296-301 Collins, M.D., C. Ash,J.A.E. Farrow, S. Wallbanks &A.M . Williams, 1989. 16S Ribosomal ribonu cleic acid sequence analysis of lactococci and related taxa. Description of Vagococcusfluvialis gen . nov., sp. nov.Journa l of Applied Bacteriology 67:453-460. Cooke, R.D., C.E.M. Ferber & L. Kanagasabapathy, 1976. Purification and characterisation of polygalacturonases from a commercial Aspergillus nigerpreparation . Biochimica et Biophysica Acta 452:440-451. Coughlan, M.P., 1991. Cellulose degradation by fungi. In: W.M. Fogarty & C.T. Kelly (Eds.): Microbial enzymes and biotechnology. Elsevier Applied Science, London & New York. pp. 1-33. Council Directives 90/219/EEC on the contained use of genetically modified micro-organisms and 90/220/EEC on the deliberate release into the environment of genetically modified organisms, 1990. Commission of European Communities, Brussels. Cousin, M.A., 1990. Development of the enzyme-linked immunosorbent assay for detection of molds in foods: a review. Developments in Industrial Microbiology 31:157-163. Cousin, M.A., J. Dufrenne, FM. Rombouts & S. Notermans, 1990. Immunological detection of Botrytisan d Monascusspecie s in food. Food Microbiology 7:227-235. David, S., M.R. van der Rest, AJ.M. Driessen, G. Simons, & W.M. de Vos, 1990. Nucleotide sequence and expression in Escherichia coli of the Lactococcus lactis citrate permease gene.Journa l of Bacteriology 172:5789-5794. de Graaff, L.H., H.C. van den Broeck, A.J.J, van Ooijen &J. Visser, 1992. Structure and regulation of an Aspergillus xylanase gene. In:J . Visser, G. Beldman, M.A. Kusters-van Someren & A.GJ. Voragen (Eds.): Xylans and Xylanases. Elsevier Science Publishers, Amsterdam, pp. 235-246. de Koning, PJ., 1992. Koemelk eiwitallergie. Voedingsmiddelentechnologie 25(5):12-14. de Ruiter, G.A., A.W. van der Lugt, A.GJ. Voragen, F.M. Rombouts & S.H.W. Notermans, 1991a. High-performance size-exclusion chromatography and ELISA detection of extracellular polysac charides from Mucorales. Carbohydrate Research 215:47-57. de Ruiter, GA., P. Smid, A.W. van der Lugt, J.H. van Boom, S.H.W. Notermans & FM. Rom bouts, 1991b. Immunogenic extracellular polysaccharides of Mucorales. In: J.P. Latgé & D. Boucias (Eds.): Fungal Cell Wall and Immune Response, NATO ASI Series Vol. H 53, Springer Verlag, Berlin Heidelberg, pp. 169-180. de Ruiter, G.A., T. Hoopman, A.W. van der Lugt, S.H.W. Notermans & M.J.R. Nout, 1992a. Immunochemical detection of Mucorales species in food. In: R.A. Samson, A.D. Hocking,J.I . Pitt & A.D. King (Eds.): Modern Methods in Food Mycology. Elsevier Publishers, Amsterdam, pp. 221-228. 267 de Ruiter, G.A., A.W. van Bruggen-van der Lugt, MJ.R . Nout, WJ. Middelhoven, RS.S. Soentoro, S.H.W. Notermans & EM. Rombouts, 1992b. Formation of antigenic extracellular polysaccha rides by selected strains of Mucor spp., Rhizopus spp., Rhiz.om.ucor spp., Absidia corymbifera and Syncephalastrum racemosus. Antonie van Leeuwenhoek 62:189-199. de Vos, WM., 1987. Gene cloning and expression in lactic streptococci. FEMS Microbiology Reviews 46:281-295. de Vos,W.M. , 1990. Process for selecting and maintaining recombinant DNA in lactic acid bacteria. European Patent Application 0355036. de Vos, W.M., I. Boerrigter, R.J. van Rooijen, B. Reiche & W. Hengstenberg, 1990. Characteriza tion of the Lactose-specific Enzymes of the Phosphotransferase System in Lactococcus lactis.Journa l of Biological Chemistry 265:22554-22560. de Vos, W.M., IJ. Boerrigter, P. Vos, P. Bruinenberg & RJ. Siezen, 1991. Production, Processing and Engineering of the Lactococcus lactis SKI 1 Proteinase. In: G. Dunny, P.P. Cleary & L.L. McKay (Eds.): Genetics and Molecular Biology of Streptococci, Lactococci and Enterococci. American Society for Microbiology, Washington, pp. 115-119. Delley, M., B. Mollet & H. Hottinger, 1990. DNA probe for Lactobacillus delbrueckii. Applied and Environmental Microbiology 56:1967 1970. Delves-Broughton, J., 1990. Nisin and its uses as a food preservative. Food Technology 4:100-112. Derksen, J.T.P., G. Boswinkel & W.M.J, van Gelder, 1991. Lipase-catalyzed hydrolysis of triglyc erides from new oil crops for oleochemical industries. In: L. Alberghina, R.D. Schmid & R. Verger (Eds.): GBF monograph 16, Lipases: Structure, mechanism and genetic engineering. VCH Publishers, Inc. New York U.S.A. pp. 377-380. Derksen, J.T.P., B.G. Muuse, F.P Cuperus & WMJ. van Gelder, 1992. New seed oils for oleo chemical industry: Evaluation and enzyme-bioreactor mediated processing. Industrial Crops and Products (in press). Derksen, J.T.P., Krosse, A.M., Tassignon, P. and Cuperus, F.P, 1993. Lipase-catalyzed production of functionalized fatty acids from Dimorphotheca pluvialis seed oil. Proceedings 6th Forum for Applied Biotechnology, 24—25 September, Brugge, Belgium (in press). Dewey, F.M., M.M. MacDonald, S.I. Philips & R.A. Priesdey, 1990. Development of monoclonal- antibody-ELISA and -DIP-stick immunoassay for Pénicillium islandicumi n rice grains.Journa l of General Microbiology 136:753-760. Exterkate, F.A., 1987. On the possibility of accelerating the ripening of Gouda cheese; a comment. Netherlands Milk and DairyJourna l 41:189-194. Exterkate, FA. & GJ.C.M. de Veer, 1990. Method for the preparation of cheese. European Patent Application 0359295. Fredenslund, A., J. Gmehling & P. Rasmussen, 1977. Vapor-liquid equilibria using UNIFAC. Elsevier, New York. 380 pp. Frisvad, J.C. & R.A. Samson, 1991. Filamentous fungi in foods and feeds: Ecology, spoilage, and mycotoxin production. In: D.K. Arora, K.G. Mukerji & E.H. Marth (Eds.):Handboo k of Applied Mycology. Vol. 3: Foods and Feeds. Marcel Dekker Inc., New York, pp. 31-68. Fuhrmann, B., M.F. Roquebert, M. van Hoegaerden & A.D. Strosberg, 1989. Immunological differentiation ofPénicillium species. Canadian Journal of Microbiology 35:1043-1047. Fuhrmann, B., V Lebreton, M. van Hoegaerden, HJ. Kamphuis & A.D. Strosberg, 1992. A monoclonal antibody specific for conidia and mycelium wall layer of Pénicillium andAspergillus. Microbiology and Immunology 36:1-12. Gao, F.H., T. Abee & W.N. Konings, 1991. Mechanism of action of the peptide antibiotic nisin in liposomes and cytochrome c oxidase-containing proteoliposomes. Applied and Environmental Microbiology 57:2164-2170. Gasson, M.J., 1990. In vivo genetic systems. FEMS Microbiology Reviews 87:43-61. Gauri, K.K., 1986. Protein hydrolysates, production process and drugs containing them. Interna tional Patent Application WO 86/04217. Graveland, A., P. Bosveld, WL. LichtendonkJ.P. Marseille,J.H.E . Moonen &A . Scheepstra, 1985. 268 A model for the molecular structure of the glutenins from wheat flour. Journal of Cereal Science 3:1-6. Gruppen, H., 1992. Structural characteristics ofwhea t flour arabinoxylans. Ph.D. thesis Agricultural University Wageningen. Harmsen, J.A.M., M.A. Kusters-van Someren, H.C.M. Rester & J. Visser, 1990. Cloning and expression of a second Aspergillus niger pectin lyase gene (pelA): indications of a pectin lyase gene family in Aspergillus niger. Current Genetics 18:161-166. Hearn, V.M. &D.W.R . Mackenzie, 1980. The preparation and partial purification of fractions from mycelial fungi with antigenic activity. Molecular Immunology 17:1097-1103. Hidalgo,J . & R.Jost , 1980. Skin care composition containing lactalbumin hydrolysate. UK Patent Application GB 2 046 591. Hugenholtz,J. & G.J. de Veer, 1991.Applicatio n of nisin A and nisin Z in dairy technology. In: G. Jung & H.G. Sahl (Eds.): Nisin and Novel Lantibiotics, pp. 440-447 Hugenholtz, J. & M. Starrenburg, 1992. Diacetyl production by different strains of Lactococcus lactis subsp. lactis var. diacetylactis and Leuconostoc spp. Appl. Microbiol. Biotechnol. 38:17-22. Hurst, A., 1981. Nisin. Advances in Applied Microbiology 27:85-123. Huis in 't Veld, J., B. Hartog & H. Hofstra, 1988. Changing perspectives in food microbiology: implementation of rapid microbiological analyses in modern food processing. Food Reviews International 4:271-329. Ianelli, D., R. Capparelli, F. Marziano, F. Scala & C. Noviello, 1983. Production of hybridomas secreting monoclonal antibodies to the genus Fusarium.Mycotaxonom y 17:523-532. Janssen, A.E.M., A. van der Padt, H.M. van Sonsbeek &K . van 't Riet. The effect oforgani c solvents on the equilibrium position of enzymatic acylglycerol synthesis. Biotechnology & Bioengineering 41:95-103. Jarvis, B., D.A.L. Seiler, AJ.L. Ould & A.P. Williams, 1983. Observations on the enumeration of moulds in food and feedingstuffs. Journal of Applied Bacteriology 55:325-226. Jung, G. & H.G. Sahl (Eds.), 1991. Nisin and novel lantibiotics, ESCOM, Leiden. Kamphuis, HJ., S. Notermans, G.H. Veeneman,J.H. van Boom & F.M. Rombouts, 1989a. A rapid and reliable method for the detection of molds in foods: using the latex agglutination assay. Journal of Food Protection 52:244-247. Kamphuis, HJ., G.H. Veeneman, F.M. Rombouts,J.H . van Boom & S. Notermans, 1989b. Anti bodies against synthetic oligosaccharide antigens reactive with extracellular polysaccharides pro duced by moulds. Food and Agricultural Immunology 1:235-242. Kamphuis, HJ., 1992a. Extracellular polysaccharides as target compounds for the immunological detection of Aspergillus and Pénicillium in food. Ph.D. thesis, Wageningen Agricultural University. Kamphuis, HJ., M.I. van der Horst, R.A. Samson, F.M. Rombouts & S. Notermans, 1992b. Mycological condition of maize products, International Journal of Food Microbiology (in press). Karman, H. & R.A. Samson, 1992. Evaluation of an immunological mould latex detection test: A collaborative study. In: R.A. Samson, A.D. Hocking, J.I. Pitt & A.D. King (Eds.): Modern Methods in Food Mycology. Elsevier Publishers, Amsterdam pp. 229-232. Kitagawa, T, Y. Sakamoto, K. Furumi & H. Ogura, 1989. Novel enzyme immunoassays for specific detection of Fusarium oxysporumf.sp . cucumerinum and for general detection of various Fusarium species. Phytopathology 79:162-165. Kester, H.C.M. &J. Visser, 1990. Purification and characterization ofpolygalacturonase s produced by the hyphal fungus Aspergillus niger. Biotechnology and Applied Biochemistry 12:150-160. Klaenhammer, T.R., 1987. Plasmid-directed mechanisms for bacteriophage defense in lactic strep tococci. FEMS Microbiololy Review 46:313-325. Klijn, N., A.H. Weerkamp & WM. de Vos, 1991. Identification of mesophilic lactic acid bacteria by using Polymerase Chain Reaction-amplified variable regions of 16S rRNA and specific probes. Applied and Environmental Microbiology 57:3390-3393. Klijn, N., A.H. Weerkamp & W.M. de Vos, 1992a. Development of sensitive techniques, based on DNA probes and PCR-amplification, for the study of the survival and genetic stability ofLactococ- 269 eusspp . in the environment and the gastrointestinal tract. Microbial Releases 1 (special supple ment), in press. Klijn, N., J. Dommes, C. Bovie, J. Hoolwerf, C.B. van der Waals, A.H. Weerkamp & F.FJ. Nieuwenhof, 1992b. Identification of Clostridium tyrobutyricum using DNA probes based on specific 16S rRNA sequences. Submitted for publication. Kloosterman, J., P.D. van Wassenaar & WJ. Bel, 1987. Membrane bioreactors. Fat Science Tech nology 89:592-597. Kok, J., 1991. Special purpose cloning vectors. In: G. Dunny, P.P. Cleary & L.L. McKay (Eds.): Genetics and Molecular Biology of Streptococci, Lactococci and Enterococci. American Society for Microbiology, Washington, pp. 97-102. Kormelink, F.J.M., 1992. Characterization and mode of action of xylanases and some accessory enzymes. Ph.D. thesis Agricultural University Wageningen. Kuipers, O.P., W.M.G.J. Yap, H.S. Rollema, M.M. Beerthuyzen, RJ. Siezen & W.M. de Vos, 1991. Expression ofwild-typ e and mutant nisin genes in Lactococcus lactis. In: G.Jung & H.G. Sahl (Eds.): Nisin and Novel Lantibiotics, pp. 250-259, ESCOM, Leiden. Kusters-van Someren, M.A., 1991. Characterization of an Aspergillus niger pectin lyase gene family. Ph.D. thesis, University of Utrecht. Kusters-van Someren, M.A.,J.A.M. Harmsen, H.C.M. Kester andJ . Visser, 1991. Structure of the Aspergillusniger pelA gene and its expression in Aspergillusniger an d Aspergillusnidulans. Current Genetics 20:293-299. Kusters-van Someren, M.A., M.J.A. Flipphi, L.H. de Graaff, H.C. van den Broeck, H.C.M. Kester, A. Hinnen & J. Visser, 1992. Characterization of the Aspergillusniger pelß gene: structure and regulation of expression. Molecular and General Genetics 234:113-120. Langeveld L.P.M., W. van der Klijn & C.B. van der Waals, 1990. Bacteriegroei in indampers voor melk en wei. Voedingsmiddelentechnologie 23:13-17. Lea, A.G.H., 1991. Enzymes in the production of beverages and fruit juices. In: G.A. Tucker and L.FJ. Woods (Eds.): Enzymes in food processing. Blackie and Son Ltd, Glasgow, London, pp. 194-220. Lin, H.H. & M.A. Cousin, 1987. Evaluation of Enzyme-linked immunosorbent assay for detection of molds in foods.Journa l of Food Science 52:1089-1094,1096. Maat, J., M. Roza,J. Verbakel, H. Stam, MJ . Santos da Silva, M. Bosse, M.R. Egmond, M.L.D. Hagemans, R.F.M. van Gorcom,J.G.M. Hessing, C.A.MJJ. van de Hondel & C van Rotterdam, 1992. Xylanases and their application in bakery. In:J . Visser, G. Beldman, M.A.R. Kusters-van Someren and A.GJ. Voragen (Eds.): Xylans and Xylanases. Elsevier, London. In press. Martinez-Murcia AJ . & M.D. Collins, 1990. A phylogenetic analysis of the genus Leuconostoc based on reverse transcriptase sequencing of 16S rRNA. FEMS Microbiology Letters 70:73-84. Matthew, J.A., S.J. Howson, M.HJ. Keenan & P.S. Belton, 1990. Improvement of the gelation properties of sugarbeet pectin following treatment with an enzyme preparation derived from Aspergillus niger- Comparison with a chemical method. Carbohydrate Polymers 12:295—306. Maubois,J.L . &J. Léonil, 1989. Peptides du lait à activité biologique. Le Lait 69:245-269. McCleary, B.V,J.M . Cooper & E.L. Williams, 1989. Debranched araban and its use as fat substi tute. Pat. Application GB 8828380.9. McKay, L.L., 1983. Functional properties of plasmids in lactic streptococci. Antonie van Leeuwen hoek 49:259-274. Meadows, D.H., D.L. Meadows &J. Randers, 1992. Beyond the limits. Confronting global collapse, envisioning sustainable future. Earthscan Publishers, London. Meisel, H., H. Frister & E. Schlimme, 1989. Biologically active peptides in milk proteins. Zeitschrift für Ernährungswissenschaft 28:267-278. Monti, J.C. & R. Jost, 1978. Enzymatic solubilization of heat-denatured cheese whey protein. Journal of Dairy Science 61:1233-1237. Morris, S.L., R.C. Walsh &J.N. Hansen, 1984. Identification and characterization of some bacterial membrane sulfhydryl groups which are targets of bacteriostatic and antibiotic action.Journa l of 270 Biological Chemistry 259:13590-13594. Mulders, J.W.M., I.J. Boerrigter, H.S. Rollema, RJ. Siezen & W.M. de Vos, 1991. Identification and characterization of the lantibiotic nisin Z, a natural nisin variant. European Journal of Biochemistry 201:581-584. Muuse, B.G., M.L. Essers & L.J.M, van Soest, 1988. Oenothera species and Borago officinalis: sources of gamma-linolenic acid. NetherlandsJourna l of Agricultural Science 36, 357-363. Muuse, B.G., F.P. Cuperus & J.T.P. Derksen, 1992. Composition and physical properties of oils from new oilseed crops. Industrial Crops and Products (in press). Neefs, J., Y. van de Peer, L. Hendriks & R. de Wachter, 1990. Compilation of small ribosomal subunit RNA sequences. Nucleic Acids Research 18(Suppl.):2237-2317. Notermans, S. & CJ. Heuvelman, 1985. Immunological detection of moulds in food by using the enzyme-linked immunosorbent assay (ELISA); preparation of antigens. International Journal of Food Microbiology 2:247-258. Notermans, S. & P.S.S. Soentoro, 1986. Immunological relationship of extra-cellular polysaccharide antigens produced by different mould species. Antonie van Leeuwenhoek 52:393-401. Notermans, S., CJ. Heuvelman, R.R. Beumer & R. Maas, 1986. Immunological detection of moulds in food: relation between antigen production and growth. International Journal of Food Microbiology 3:253-261. Notermans, S., G.H. Veeneman, C.W.E.M. van Zuylen, P. Hoogerhout &J.H. van Boom, 1988a. (l-5)-linked yS-D-galactofuranosides are immunodominant in extracellular polysaccharides of Pénicillium and Aspergillus species. Molecular Immunology 25:975-979. Notermans, S.,J. Dufrenne & P.S. Soentoro, 1988b. Detection ofmold s in nuts and spices: The mold colony count versus the enzyme linked immunosorbent assay (ELISA).Journa l of Food Science 53:1831-1833,1843. Notermans, S. & H. Kamphuis, 1990. Detection of moulds in food by latex agglutination: a collabo rative study. Food and Agricultural Immunology 2:37-46. OECD, 1992. Biotechnology, Agriculture and Food, OECD, Paris. Parker, D.M. & D. Pawlett, 1987. Flavour control of protein hydrolysates. European Patent Applica tion 0 223 560. Pilloud, N. & B. Mollet, 1990. DNA probes for the detection of 24 Lactobacillus helveticus. System atic and Applied Bacteriology 13:345-349. Pilnik, W. & A.GJ. Voragen, 1989. Effect of enzyme treatment on the quality of processed fruit and vegetables. In:J.J .Je n (Ed.): Quality factors of fruits and vegetables. Chemistry and Technology. ACS American Chemical Society Symposium Series 405: pp. 250-269. Pitt,J.I . & A.D. Hocking, 1985. Fungi and Food Spoilage. Academic Press, Sydney. 410 pp. Pronk, W., PJ.A.M. Kerkhof, C. van Helden, C. & K. van 't Riet, 1988. The hydrolysis of triglycerides by immobilized lipase in a hydrophilic membrane reactor. Biotechnology and Bioen gineering 32:512-518. Pronk, W. & K. van 't Riet, 1991. The interfacial behavior of lipase in free form and immobilized in a hydrophilic membrane reactor. Biotechnology and Applied Biochemistry 14:146-154. Pronk, W., G. Boswinkel & K. van 't Riet, 1992a. The influence of fatty acid and glycerol on the kinetics of fat hydrolysis by Candida rugosa lipase in a membrane bioreactor. Biocatalysis 5:305- 323. Pronk, W., G. Boswinkel & K. van 't Riet, 1992b. Parameters influencing hydrolysis kinetics of lipase in a hydrophilic membrane bioreactor. Enzyme Microbial Technology 14:214—220. Rauch, PJ.G., M.M. Beerthuijzen & WM. de Vos, 1991. Molecular analysis and evolution of conjugative transposons encoding nisin production and sucrose fermentation in Lactococais laclis. In: G.Jung, G & H.G. Sahl (Eds.): Nisin and Novel Lantibiotics, pp. 243-249, ESCOM, Leiden. Rauch, PJ.G. & W.M. de Vos, 1992. Characterization of the novel nisin-sucrose conjugative transposon Tn5276 and its insertion in Lactococcus lactis.Journa l of Bacteriology 174:1280 1287. Renard, C.M.G.C., MJ.F . Searle-van Leeuwen, A.GJ. Voragen,J.F . Thibault & W. Pilnik, 1991. Studies on apple protopectin. IL Apple cell wall degradation by pure polysaccharidases and their 271 combinations. Carbohydr. Polym. 14:295-314. Reynolds, E.C., 1987. Phosphopeptides. International Patent Application WO 87/07615. Riddick,J.A . & E.E. Toops, 1955. Organic solvents, physical properties and methods of purification. Interscience publishers, New York. 552 pp. Robertson, A. & N. Patel, 1989. Development of an enzyme-linked immunoassay for the quantifica tion of mould in tomato paste - an alternative to the Howard mould count. Food and Agricultural Immunology 1:73-81. Roger, L., G. Brule & J.L. Maubois, 1981. Nouvelles voies de valorisation des protéines lac- tosériques. La Technique Laitière 952:65-67. Rollema, H. S., P. Both & RJ. Siezen, 1991. NMR and activity studies on nisin degradation products. In: G.Jung & H.G. Sahl (Eds.): Nisin and Novel Lantibiotics, pp. 123-130, ESCOM, Leiden Ruhr, E. & H.G. Sahl, 1985. Mode of action of the peptide antibiotic nisin and its influence on the membrane potential of whole cells and on cytoplasmic and artificial lipid bilayers by the peptide antibiotic nisin. Antimicrobial Agents and Chemotherapy 27:841-845. Sahl, H.G., M. Kordel & R. Benz, 1987. Voltage-dependent depolarization of bacterial membranes and artificial lipid bilayers by the peptide antibiotic nisin. Archives of Microbiology 140:120-124. Salama M., W. Sandine & S. Giovannoni, 1991. Development and application of oligonucleotide probes for identification of Lactococcus lactis subsp. cremoris. Applied and Environmental Microbiol ogy 57:1313-1318. Sanders, M.E., 1988. Bacteriophage resistance in lactic acid bacteria. Biochemie 70:411-421. Sato, R., M. Shindo, H. Gunshin, T. Noguchi & H. Naito, 1991. Characterization of phosphopep- tide derived from bovine y3-casein: an inhibitor to intra-intestinal precipitation of calcium phos phate. Biochimica et Biophysica Acta 1077:413-415. Schleifer, K.H.J. Kraus, C. Dvorak, R. Kilpper-Bälz, M.D. Collins & W. Fischer, 1985. Transfer of Streptococcus lactis and related streptococci to the genus Lactococcus gen. nov. Systematic and Applied Microbiology 6:183-195. Schmidt, D. G., P. Both, S. Visser, K.J. Slangen & PJ. van Rooijen, 1987. Studies on the precipita tion of calcium phosphate. II. Experiments in the pH range 7.3 to 5.8 at 25 °C and 50 °C in the presence of additives. Netherlands Milk and DairyJourna l 41:121-136. Schmidt, D.G., B.W. van Markwijk, K.J. Slangen & S. Visser, 1990. Enzymatische afbraak van wei-eiwitten (Enzymatic hydrolysis of whey proteins). Voedingsmiddelentechnologie 23 (16/ 17):11-13. Schmidt, D.G. &J.K. Poll, 1991. Enzymatic hydrolysis of whey proteins. Hydrolysis of a-lactalbu- min and /Mactoglobulin in buffer solutions by proteolytic enzymes. Netherlands Milk and Dairy Journal 45:225-240. Schnell, N., K.D. Entian, U. Schneider, F. Götz, H. Zähner, R. Kellner & G.Jung, 1988. Prepeptide sequence of epidermin, a ribosomal synthesized antibiotic with four sulphide rings. Nature 333:276-278. Schols, H.A., C.C.J.M. Geraeds, M.F. Searle-van Leeuwen, FJ.M. Kormelink & A.G.J. Voragen, 1990. Rhamnogalacturonase: a novel enzyme that degrades the hairy regions of pectins. Carbo hydrate Research 206:105-115. Scott, D., F.E. Hammer & TJ. Szalkucki, 1987. Bioconversions: Enzyme Technology. In: D. Knorr (Ed.): Food Biotechnology Marcel Dekker Inc. New York, pp. 413-440. Searle-van Leeuwen, MJ.F, L.A.M. van den Broek, H.A. Schols, G. Beldman & A.GJ. Voragen, 1992. Rhamnogalacturonan acetylesterase: A novel enzyme from Aspergillus aculeatus, specific for the degradation of hairy (ramified) regions of pectins. Applied Microbiol. Biotechnol. Submitted. Siezen, R.J., S. Visser & D.G. Schmidt, 1990. Enzymatic modification of milk proteins. In: J.J. Dekkers, H.C. van der Plas & D.H. Vuijk (Eds.): Agricultural biotechnology in focus in The Netherlands, Pudoc, Wageningen, pp. 204-208. Siezen, RJ., W.M. de Vos,J.A.M . Leunissen & B. Dijkstra, 1991.Homolog y modelling and protein engineering strategies of subtilases, the family of subtilisin-like serine proteinases. Protein Engi- 272 neering 4:719-737. Starrenburg, M. & J. Hugenholtz, 1991. Citrate fermentation by Lactococcus and Leuconostoc spp. Applied Environmental Microbiology 57:3535-3540. Stynen, D., L. Meulemans, A. Goris, N. Braendlin & N. Symons, 1992a. Characteristics of a latex agglutination test based on monoclonal antibodies for the detection of fungal antigens in food. In: R.A. Samson, A.D. Hocking,J.I . Pitt & A.D. King (Eds.): Modern Methods in Food Mycology. Elsevier Publishers, Amsterdam, pp. 213-220. Stynen, D.,J . Sarfati, A. Goris, M.C. Prévost, M. Lesourd, H. Kamphuis, V. Darras &J.P. Latgé, 1992b. Rat monoclonal antibodies against Aspergillusgalactomannan . Infection and Immunity 60:2237-2245. Tan, P.S.T., 1992. The biochemical, genetic and physiological properties of aminopeptidase N in Lactococcus lactis.Ph.D . thesis, University of Groningen. Tani, F., K. Iio, H. Chiba & M. Yoshikawa, 1990. Isolation and characterization of opioid antago nist peptides derived from human lactoferrin. Agricultural and Biological Chemistry 54:1803- 1810. Tomita, M., Y. Fukuwatari, T. Kitazawa, M. Nojiri & S. Kawaura, 1991.Milk-protei n hydrolysates and compositions for use as hair and skin treating agent. European Patent Application 0 457 565. Tsai, GJ . & M.A. Cousin, 1990. Enzyme-linked immunosorbent assay for detection of molds in cheese and yogurt.Journa l of Dairy Science 73:3366-3378. Turgeon, S.L., S.F. Gauthier, D. Molle &J. Léonil, 1992. Interfacial properties of tryptic peptides of/Mactoglobulin .Journa l of Agricultural and Food Chemistry 40:669-675. Ulig, H., 1991. Enzyme arbeiten für uns: technische Enzyme und ihre Anwendung. Carl Hanser Verlag, München, Wien. p. 409. van Alen-Boerrigter, IJ., R. Baankreis & W.M. de Vos, 1991. Characterization and overexpression of the Lactococcus lactis pepJV gene and localization of its product, aminopeptidase N. Applied Environmental Microbiology 57:2555-2561. van Bruggen-van der Lugt, A.W., HJ. Kamphuis, G.A. de Ruiter, P. Mischnick, J.H. van Boom & F.M. Rombouts, 1992. New structural features of the antigenic extracellular polysaccharides of Pénicillium and Aspergillus species revealed with exo-/S-D-galactofuranosidase. Journal of Bacteriol ogy 174 (in press). van de Ven, F.J. M., H.W. van den Hooven, R. N. H. Konings & C. W Hilbers, 1991.NM R studies of lantibiotics: The structure of nisin in aqueous solution. EuropeanJourna l of Biochemistry 202, 1181-1188. van der Horst, M., R.A. Samson & H. Karman, 1992. Comparison of two commercial kits to detect moulds by latex agglutination. In: R.A. Samson, A.D. Hocking, J.I. Pitt & A.D. King (Eds.): Modern Methods in Food Mycology. Elsevier Publishers, Amsterdam, pp. 241-246. van der Meer J.R., W. Ludwig & W.M. de Vos, 1992. Characterization of a ribosomal RNA gene cluster from Clostridiumtyrobutyricum: phylogeneti c positioning based on the 16S and 23S nucleo tide sequences. Submitted for publication. van der Padt, A., A.E.M.Janssen,J J.W. Sewalt & K. van 't Riet, [a]. Acylglycerol equilibrium in a two-phase system. Accepted for publication in Biocatalysis. van der Padt, A.,J J.W. Sewalt & K. van 't Riet, [b]. Specific by-product removal in a membrane bioreactor. Submitted for publication. van der Veen, P., A.G.J. Voragen &J. Visser, 1990. Production, structure and function of some arabinases produced by Aspergillus niger. In:J.J . Dekkers, H.C. van der Plas and D.H. Vuijk (Eds.): Agricultural biotechnology in focus in the Netherlands. Pudoc, Wageningen, pp. 187-192. Verhue, W.M., & F.S.B. Tjan, 1991. Study of the citrate metabolism of Lactococcus lactis subsp.lactis biovar. diacetylactis by means of C-13 nuclear magnetic resonance. Applied Environmental Micro biology 57:3371-3377. Verrips, CT., F.S.B Tjan, J.B. Marugg, D. Goelling & W.M. Verhue, 1990. Metabolic studies of citrate metabolism in Lactococcus lactis subsp. lactis using NMR and rDNA techniques. In : H. Breteler, R.F. Beudeker & K.Ch.A.M. Luyben (Eds.): Proceedings of the Third Netherlands 273 Biotechnological Society, pp. 56-63. Visser, S., 1981.Proteolyti c enzymes and their action on milk proteins. A review. Netherlands Milk and DairyJourna l 35:65-88. Visser, S., H.S. Rollema, M.K. Friedenthal & GJ. van Alebeek, 1988. Spectrophotometric method for the determination of chymosin and pepsin in calf and adult bovine rennets. Netherlands Milk and DairyJourna l 42:221-232. Visser, S., H.J. Noorman, K.J. Slangen & H.S. Rollema, 1989. Action of plasmin on bovineyS-casei n in a membrane reactor.Journa l of Dairy Research 56:323-333. Visser, S., 1993. Proteolytic enzymes and their relation to cheese ripening and flavor: an overview. Journal of Dairy Science 76:329-350. Voragen, A.GJ., R. Heutink & W. Pilnik, 1980. Solubilization of apple cell walls with polysaccha ride degrading enzymes.Journa l of Applied Biochemistry 2:452-468. Voragen, A.G.J., 1989. Food science: basic research for technological progress. In: J.P. Roozen, F.M. Rombouts & A.G.J. Voragen (Eds.): Proceedings of the symposium dedicated to Professor W. Pilnik, Wageningen, Netherlands, 25 November 1988. Pudoc, Wageningen, pp. 59-81. Voragen, A.G.J. & MJ.R. Nout, 1991. Biotechnologie in de levensmiddelenindustrie: Biotechnolo gie in Nederland 8:7-9. Voragen, A.G.J., H. Gruppen, M.A. Verbruggen & RJ. Viëtor, 1992. Characterization of cereal arabinoxylans. In: J. Visser, G. Beldman, M.A.R. Kusters-van Someren and A.G.J. Voragen (Eds.): Xylans and Xylanases. Elsevier, London. In press. Vos, P.R.J. Siezen, WM. de Vos, J. Kok, G. Venema & AJ. Haandrikman, 1991. Modified proteases, process for their preparation and their use in foodstuffs. European Patent Application 0411715. Walker, J. & G. Dougan, 1989. DNA probes: a new role in diagnostic microbiology. Journal of Applied Bacteriology 67:229-238. Weegels, PL., 1990. Modification of wheat gluten as a means to reveal structure-function relation ships. TNO report nr B 91-019. 73 pp. Weegels, PL., A.M. de Jager, M.A.B. Voorpostel, M. Harte & RJ. Hamer, 1990. Enzymatic modification ofwhea t gluten. In:JJ . Dekkers, H.C. van der Plas & D.H. Vuijk (Eds.): Agricultural Biotechnology in focus in the Netherlands. PUDOC, Wageningen, pp. 211-217. Weegels, PL., J.P. Marseille & R.J. Hamer, 1991a. Large scale separation of gliadins and their breadmaking quality. In: W. Bushuk & R. Tkachuk (Eds.): Gluten proteins, 1990. AACC, St Paul, pp. 214-218. Weegels, PL., 1991b. Sensitive measurement of the contents and reactivities of thiol and disulphide in soluble and insoluble proteins. In: W. Bushuk & R. Tkachuk (Eds.): Gluten proteins, 1990. AACC, St Paul. pp. 414-419. Weegels, PL. & RJ. Hamer, 1992. Improving the breadmaking quality of gluten. Cereal Foods World 37:379-385. Wenner, V, 1982. L'utilisation des protéines du lait à des fins diététiques et thérapeutiques en alimentation humaine. Le Lait 62:549-565. Whitaker,J.R. , 1984. Pectic substances, pectic enzymes and haze formation in fruit juices. Enzyme and Microbial Technology 6:341-349. Yamauchi, K., 1992. Biologically functional proteins of milk and peptides derived from milk pro teins. International Dairy Federation (Brussels), Bulletin No. 272, pp. 51-58. 274 Societal aspects of agricultural biotechnology Contents 6.1 General introduction to societal aspects of agricultural biotechnology 277 6.1.1 Changing social relations of agriculture 277 6.1.2 Reshaping the relations between agriculture, biotechnology, and society 278 6.1.3 Technology dynamics 279 6.1.4 Concluding remarks 281 6.2 Diffusion of genetic modification in Dutch agribusiness 281 6.2.1 The dynamic concept of diffusion 282 6.2.2 State of the art of genetic modification 282 6.2.3 Different types of innovations in the product life cycle 283 6.2.4 Characteristics of markets for inputs and outputs 284 6.2.5 Characteristics of branches in agro-industries 285 6.2.6 Public acceptance and social issues 288 6.2.7 Government policy 289 6.2.8 Concludin g remarks 29 1 6.3 Protecting and exploiting biotechnological inventions in agriculture: patent rights versus plant breeders rights? 291 6.3.1 Current position of plant breeders' rights 292 6.3.2 European patent rights on plants and animals 293 6.3.3 EC-directive on biotechnological inventions 294 6.3.4 Balancing both rights 296 6.4 The development of risk-evaluation for the deliberate release of genetically modified organisms: An outline of the international discussion 297 6.4.1 Safety in biotechnology 298 6.4.2 Stepwise development 298 6.4.3 Risk analysis in biotechnology 299 6.4.4 The concept of familiarity 300 6.4.5 The case-by-case approach 301 6.4.6 Conclusion 302 275 6.5 Safety and acceptance of biotech-food 302 6.5.1 What is acceptance? 302 6.5.2 Present situation 303 6.5.3 Future developments 305 6.6 Technology assessment and agricultural biotechnology 305 6.6.1 Activities of the Netherlands Organization for Technology Assessment 306 6.6.2 Activities of the Netherlands Council for Agricultural Research 307 6.7 Ethics and agricultural biotechnology 307 6.7.1 The need for ethical reflection 308 6.7.2 Changes in the support for agriculture 309 6.7.3 Biotechnology: a novel development or a new way ofrealizin g old goals? 310 6.7.4 'Moral' and 'ethical' concerns: some problems of terminology 311 6.7.5 Subject and method of ethics 311 6.7.6 Five problem-areas of the ethics of biotechnology 312 6.7.7 Conclusion 318 6.8 References 319 276 6.1 General introduction to societal aspects of agricultural biotechnology A. Rip To inquire into the societal aspects of biotechnology is not just an academic exercise for economists and sociologists. It is important for biotechnologists and policy makers inter ested in improving the reception of biotechnology, and for other groups and organisations gradually becoming involved in the new technology who also need to know about the dynamics of developments and their repercussions. The public, or society in general, may not have a direct interest in such knowledge, but still profits from the insights. Societal aspects of biotechnology occur at two levels: at the level of specific technologies and their adoption and diffusion, and at the level of relations between biotechnology and society. What is happening can be viewed as a case of a novel input into society: new technical possibilities emerge and their implementation and prudent use are not immediately obvi ous. Both promising future options and the concerns about possible risks and about acceptability still need to be articulated. It can be a long drawn out and conflictual process, and for biotechnology, this certainly seems to be the case. Expectations about break throughs are raised high, but biotechnology firms are going through periodic shake outs. Ecological risks and the symbolic or real threat to human (and animal?) genetic integrity make the novel input of biotechnology much more conflictual than say, new materials, or telecommunications. It is important to recognize, in addition, that such novel inputs are not isolated inputs into an otherwise static society. Our society evolves continually, and responds to a variety of inputs, technological and otherwise. For biotechnology, one obvious example is the vicissitudes of the Common Agricultural Policy of the European Communities, and Euro pean politics in general, which may throw up barriers (as in the case of the isomerase process) or create specific incentives. Increasing claims ofparticipatio n in decision making, and the emergence of voluntary associations and spokespersons for environmental and ethical causes, are general phenomena in our society, but have particular repercussions on biotechnology. These general considerations are additionally important in the case of agricultural biotechnology, because agriculture and related activities and institutions are themselves in a transitional stage, and the introduction of new biotechnology will quicken the pace, even while the overall direction that agriculture is moving towards is still unclear. 6.1.1 Changing social relations of agriculture Agriculture as we know it now, with its practices, its knowledge, its institutions and its culture and politics, is the result of a broad, long-term process of modernization. The beginnings of modern agriculture can be traced back to at least the end of the last century. Rationalization and 'scientification' have been important features, with productivity and efficiency as dominant values. The institutions of agriculture, the formal and informal networks among actors, and the culture of this 'green' world are all geared towards these 277 goals and values. One can thusvie w the relation between agriculture and society as a social contract: ensure the production of food (in quantity, availability, and quality) in return for support and (relative) autonomy. It is this 'charter' which allowed agriculture in western countries to expand, and it also determined our ways of looking at the problems of developing countries. By the nineteen eighties, however, a transition had become noticeable: the drive for efficiency and effectivity started to falter, certainly in the Netherlands where there was overproduction (in the economic sense, as well as according to political criteria). Environ mental aspects became salient, definitely so in the Netherlands. And new ways of looking at farming systems emerged, so that the old 'charter' came to be recognized as only one of a range of possible ones. The public and political prominence of agriculture ishigh , sometimes uncomfortably so. New actors, in addition to the original participants and those from the agro-food sector who joined at an earlier stage of industrialization of agriculture, are making themselves heard. This implies that dimensions for evaluating agriculture and its development will broaden and shift. For the time being at least, a shared view is absent. But it is precisely through the antagonistic attempts to come to terms with the transitions and new challenges that a new social contract may emerge. This is the situation in which agricultural biotechnology isbein g developed, introduced, and adopted (or not). The existing institutions and networks of actors in agriculture are not adapted to the novel inputs: their roles were to sustain modernisation. In responding to biotechnology, they will start by using approaches geared to modernisation, for example seeing biotechnology as a means to improve productivity. During a transitional process, it is not clear what the new 'charter' will turn out to be, so there may, in fact, be good reasons to follow the old 'charter'. 6.1.2 Reshaping the relations between agriculture, biotechnology, and society The contributions to this chapter of the book start from such a diagnosis of the situation: what is at issue in the process of reshaping the relation between agriculture, biotechnology and society? In Hutten's discussion of diffusion of genetic modification technologies (Section 6.2), he emphasizes that most products (and industries) in the agribusiness are in a stage of matur ity, and innovation is only welcome if it reduces processing costs. New products run high marketing risks, and agribusiness is risk averse. Suppliers may be more innovative, but run up against the combination of risk-averse agribusiness and public acceptance problems, as is clear from the case study of chymosin, presented by Toet. The one exception, plant breeding, shows an interest in shortening development work and introducing new varieties. As Hutten indicates in Section 6.3, the existing arrange ments of patents and plant breeding rights are not adapted to the novel inputs of biotech nology. His attempt to sketch a way forward may not be the final answer, but is valuable for identifying the relevant dimensions ofthes e issuestha t must be taken into account when forging a solution. 278 Issues of risk, and behind risk, public acceptability, feature heavily in the discussions, and in the practice, of biotechnology, not in the least because of the way that governments and international organisations have stepped into these debates, and have attempted to discharge their regulatory responsibilities. Bergmans gives a useful overview of the present situation of risk assessment of the deliberate release of genetically modified organisms (Section 6.4). The achievement of a rational and robust approach to risk assessment has to be complemented with a sense of public acceptability, without which there will be no political support for regulation on this basis. Toet's case study of chymosin (Section 6.5), again, clearly illustrates these issues. The difficult issues of adoption, acceptability and diffusion, and the increasing recogni tion of the need to anticipate such downstream problems require some form of technology assessment. In the Netherlands, interactive forms of technology assessment are being pioneered by the Netherlands Organization for Technology Assessment, and the brief chapter by Rip summarizes what has been done (Section 6.6). From technology assessment to ethical aspects is a small step: assessment refers to values and criteria and, especially in the case of novel technological inputs, it will not be evident how existing views and values should be applied, or perhaps modified. Thus, there is a need for ethical reflection. In Vorstenbosch's concluding Section 6.7, the role of ethical reflection is discussed, and an agenda of five issues important for agricultural biotechnol ogy is presented. How decisions to adopt a new technology are taken, and what approaches to handle safety and risks emerge, can be viewed as indications of the extent and direction of the process of reshaping of relations. Recognizing the importance of technology assessment, and the increasing ethical debate about biotechnology, are further indicators. There, one sees explicit reflection about desirable and undesirable directions in the process of reshap ing. But it isth e aim of all chapters in this part ofth e book to discuss issues of adoption, risk assessment and ethics, with explicit reference to the process of reshaping that is going on. This is not just an intellectual, social-scientific point. New views are already emerging, issues crop up and are addressed, new actors become involved. While the various actors may not agree on what is desirable or not, they should have a common interest in a productive transformation of the world of agriculture. And in order to have a productive transformation, actors should realize what is happening and not fall back unthinkingly on older approaches and attitudes. The more reflective Sections in this part of the book contribute to this aim. 6.1.3 Technology dynamics To illustrate my point, I will briefly discuss some insights from my own field of technology dynamics. What ishappenin g in agriculture and biotechnology isno tjus t a transition to an unknown new world; it is a transformation with strong non-linear characteristics. That is, one cannot extrapolate present trends and so find the contours of the biosociety. Technol ogies branch out in different directions, partly in relation to existing or perceived niches (economic and socio-political). So-called network externalities and 'domino-effects' are important and 'lock-ins' can occur (Dosi et al., 1988). To give a few examples: 279 1. 'Impression management' rather than functionality can drive developments. Around 1980, the strong interest in the production of insulin and human growth hormone with the help of genetically modified organisms was related to the promises put forward earlier to defend recombinant DNA technologies against critics. It is indicative of the general atmosphere that the first rDNA product to come onto the market, Akzo's anti-dysentery drug for pigs, went almost unnoticed. 2. Non-critical markets can be crucial to get a new product through the period of teething troubles. High-density, regular polythene was saved for further development thanks to the Hula-Hoop craze in the mid-fifties, which took up the batches of low-quality product available at the time when they were glutting the storehouses and threatening to justify stopping the development process altogether (for further details and refer ences, see Rip 1992). The nature of such a non-critical market isno t correlated with the social importance of the eventual product. In the case of biotechnology, enzymes to stonewash jeans may well be an example. 3. While pharmaceutical products have remained important in biotechnology, the interest and activity in industrial biotechnology - originally an important option - has declined, while agricultural biotechnology has become a very active field with many innovations. This kind of branching out is hard to predict. 4. Shifts in R&D and product development occur in response to strategic assessments of developments in economic and political contexts. 'Hedging your bets' is one such phenomenon, where firms will argue against intended regulation and at the same time prepare themselves for the eventuality. Over time, such strategies add up to technolog ical development which follows the path of least regulatory resistance, Monsanto's work in the area ofplan t pesticides isa n example. Asjelsma(1991) has shown, themain trajectory was to genetically modify Pseudomonas so that it would produce a toxin. When authorization of such organisms became more and more difficult, Monsanto shifted to modification of the plant (to resist pesticides), which was expected to meet less regula tory resistance. 5. In arenas of public debate, foci emerge which draw attention to a specific set of issues, which will orient strategies of producers, regulators and consumers. Human food and the issue of acceptability of biotechnological inputs in food production issuc h a forceful focus, and all parties develop their own strategies. Producers, for example, are very careful in introducing biotechnological products, and consumers appear to be fearful (orjus t prudent). The technologies, and products and processes being developed thus tend to anticipate, and work around, broader issues in a selection environment, rather than just being offered on a market. While these examples and general considerations do not offer easy guidelines to actors concerned about biotechnology in society, they do create the basis for viewing and han dling the transformation process in a realistic way. It would actually be counterproductive to use traditional economics to evaluate biotech nology in society. In the case of a novel input, here biotechnology, into a system, there are diffuse expectations (hopes and concerns) rather than the articulated preferences of con sumers and cost-benefit optimalisations of producers assumed by traditional economics. Articulation of innovation and production possibilities, in interaction with articulation of 280 demand, must get priority. And such articulation processes have to work through trial and error. When new products are being tried out (and possibly fail) the action-reaction patterns and the forceful foci emerging in this way have to be taken into account. Industry and other near-market actors now seem to recognize this point and seek relations with relevant groups, even ifi t is not always clear how to handle the problems. (Scientists are in a somewhat different position: their task is to create new options, and this requires a protected environment in which there is room to explore.) Besides demand articulation, there is also acceptability articulation. Again, this is not a linear process. Acceptability cannot be enhanced simply through better PR and putting more resources into public understanding of science and technology. One important point that is often overlooked, is that biotechnology (and technology in general) will be seen as an intruder. Unknown because it is new, hopefully beneficial, and always dangerous, exactly because it is an intruder. More information about the 'intruder' will not help. Fear of burglars does not decrease if one knows the colour of their eyes and hair. Information about safety measures, trust that you can call the police and that they will come, is what reassures. (Studies of risk perception and attitudes have indicated the importance of this dimension of'institutional safety'.) Thus, it isbette r to inform the public about institutional safety procedures, care in decision making and in handling issues, and to make this part of the culture rather than to explain endlessly that DNA is DNA is DNA. Seen in this light, the role of the media takes on a different complexion. Criticisms by biotechnologists and industrialists focus on the lack of expertise and the tendency to highlight issues, often negative aspects. But this is an important route to articulate accep tance: bland information, or repeating the message that everything is well will not appear uneasy feelings. 6.1.4 Goncludin g remarks To conclude this brief introduction: insights offered by technology dynamics, and the analysis of adoption, risk assessment and ethics in the contributions that follow are based on scientific and scholarly studies, but they are more than just a presentation of research results. They are an input into our thinking about ongoing transformations in agriculture, biotechnology and society, and will help shape the overall agenda (as political scientists would call it). It is important to have societal aspects on the agenda: because they are essential for the long-term success of biotechnology, and because of the reflectivity, that is, the recognition of one's own situation and dynamics involved, that they introduce. 6.2 Diffusion of genetic modification in Dutch agribusiness T.H.J.M. Hutten It is impossible to evaluate the diffusion of biotechnology in general. The main reason being that biotechnology is an accumulation of many different technologies, sharing only the 'biological' aspect. The subject of this Section is therefore restricted to genetic modifi- 281 cation. Genetic modification is,however , one ofth e most striking features of biotechnology which is being met by great public concern. But even under this restriction, the issue of diffusion is a difficult one. This is mainly due to the concept of diffusion itself and to the very complex process that diffusion describes. It should be made clear that a complete and exhaustive picture cannot be presented. The diffusion of genetic modification in many branches of agribusiness is briefly reviewed with respect to the elements of the dynamic concept of diffusion. The plant breeding sector will be discussed more thoroughly. 6.2.1 The dynamic concept of diffusion In the last decade the concept of diffusion developed from a static to a dynamic one (Jacobs, 1990). The concept has been linked to: 1. the state of the art in the development oftechnolog y and its innovative results in various sectors of industry; 2. the different types of innovation in successive stages of product life cycles; 3. the characteristics of markets for innovation for inputs and outputs; 4. the characteristics of branches of industries and even individual firms; 5. public acceptance and social issues; 6. government policy; 7. interactions between the previously mentioned factors. It is,however , beyond the scope of this evaluation to discuss allthes e elements in detail. For our purpose it will suffice to summarize some conclusions with respect to the diffusion of genetic modification within branches of agribusiness. 6.2.2 State of the art of genetic modification Since the end of the seventies genetic modification of micro-organisms has developed into a sophisticated tool for pharmaceutical and chemical applications. It is even possible to incorporate and express genes isolated from plants, animals or humans into micro-organ isms. This development has been a powerful impetus for the development of new products especially designed for markets in agriculture (Stasse-Wolthuis & Rombouts, 1992). The development of products from genetically modified, undifferentiated cell strains of plants isver y promising but is still encountering scientific and technological problems. These are caused more by inadequate fermentation techniques and lack of physiological knowledge than by problems related to the modification process (Verpoorte 1991). There is also the question of whether large-scale production of plant metabolites should take place in a fermenter or by growing crops. From the technological point of view the first method requires more basic knowledge of plant physiology and differentiation processes while the second route will face processing and purification issues. Despite the lack of knowledge about the processing and purification of vegetable materials, large-scale production through field crops is seen as being the most promising. However, at present there are also some restrictions with respect to the modification and generation of plants. Commercial processes are available for the modification of dicotyledons. However, the regeneration of whole plants from cells or tissues is still a problem for some crops. Until recendy, modifica- 282 tion of monocotyledons was not commercially possible, but the introduction of the particle gun has largely solved this problem. The technique is spreading rapidly in the field of cereals and grasses. The use of (genetically modified) animal cells for specific applications is still a distant prospect. The culturing of animal cells in a fermenter is an insuperable barrier. Again, genetic modification of animals, like plant cell culture versus crops, seems more promising. It has recently been shown that the protein lactoferrin, which inhibits growth of bacteria, can be produced in milk from genetically modified cows. Also, the genetic modification of animal cells is commercially less developed, for the following reasons. First, only a few commercial transformation processes are available. Second, so far it has been impossible to regenerate animals from animal cell cultures. Genetic modification of animals is there fore only possible through genetic modification of reproductive cells. But even in the field of human and veterinary health care, in which products are obtained from genetically modified micro-organisms, there is only a handful of commer cially successful products and an even smaller number of enterprises. It can be concluded therefore that a firm producing inputs, intermediates or consumer products must make a choice about a technology of which the future is uncertain. Addi tionally, many agro-industries do not have long research traditions or do not have a research and development tradition at all. As well as the choice between technological alternatives, other main issues are expectations on the relevance of genetic engineering, development of expertise in-house or by partnership, protection of knowledge, acceptance by consumers and the potential market for a genetically engineered product and finally, the pay-back period for investment in technology and production. An inventor or early adopter risks investing in techniques which are proving to be quickly outdated by later developments, thus leading to bad investments or to commercially irrelevant products. This explains why the agro-industries are taking a more conservative attitude towards adopting the new technology than was predicted by the same industries ten years ago. 6.2.3 Different types of innovations in the product life cycle In the first stage of the product life cycle, the emergence stage, biotechnology evolves rapidly. There are no dominant technical designs, and scientific results need to be devel oped for or adapted to specific fields of applications. The risks in this first stage are high: investments might be made in wrong directions and successful technological developments do not necessarily lead to fast payoffs in a (global) market. Genetic modification of plants and animals was at this stage during the eighties. However, the genetic modification of plants in particular is now reaching the next stage, called the growth phase. About five years ago, this stage had already been reached by the genetic modification of micro organisms. But, even the genetic modification of micro-organisms has not been the com mercial success that was anticipated in the eighties. It is not surprising that the develop ment of genetic modification occurs in two types ofindustries : 1. Large industries that have strong market positions in chemical and fine-chemical markets, based on competitive R&D efforts; and 2. New biotechnology firms (NBFs) which have a strong human resource position and a unique know-how, but generally lack the appropriate entry to the market 283 or are often unable to manufacture their own products on a commercial scale. They sell, or in many cases intend to sell, embodied knowledge. Hence, these firms have to rely heavily on different forms of cooperation. The same applies to the breeding ofplant s and animals. However, having been involved in the agricultural sector forjus t one decade, some multinationals (e.g. Unilever, Shell) are already disposing of their recendy acquired plant breeding firms. So far, new biotechnol ogy firms in the agricultural sector have not produced a commercial product. Some of them have property rights on commercially relevant processes. In animal breeding, the NBF Gene Pharming has succeeded in introducing the lactoferrin gene in cows. Lactofer- rin is thought to be active in the human defence system and it may help to prevent mastitis infections in dairy cows. The amount of lactoferrin needed to prove its feasibility as a medical drug can be produced by milk from the genetically modified cows. Mogen and Florigene produce transgenic plant varieties that do not have the potential to generate substantial cash flows. They are, however, committed to these products for other reasons (see later on). Nevertheless, these NBFs survive! Some of them do so by revenues from a patented process (e.g. Mogen), others by the capital assignments of shareholders (e.g. Keygene and Florigene) or by selling physiologically interesting genes to plant breeders who have important market shares in major commercial crops. However, the interest of the capital markets for NBFs is decreasing. The Rabobank, the major agriculturally orientated bank in the Netherlands, has recently abolished its biotech venture fund. There fore, NBFs in the Netherlands demanding capital venture must meet certain requirements for participation loans. The initial optimism on the capital markets has been replaced by a more realistic approach. Small and medium-large agribusiness industries cannot afford the approach of either the NBFs or the multinationals. In almost every area of agribusiness the small and me dium-large industries have reached maturity and many of their products are heading towards the end of the product life cycle. Small profits, strategies focused on issues other than technologically superior innovative products and R&D directed to immediate prod uct and process improvements delay the introduction of radical technological changes. The development of new products and the prolongation of life cycles of existing ones is not a strong feature of the small and medium-large agribusiness companies, although the inert attitude to radical changes may lead to short-term solutions of their problems. Plant breeding, especially horticultural plant breeding, is to some extent an exception. In this sector the product life cycles are short and product innovation is crucial. 6.2.4 Characteristics of markets for inputs and outputs Although many branches of the agribusiness are mature, this does not mean that markets do not change. First, the successful common agricultural policy has led to too high budget ary burdens on the EC. The fast-growing production cannot be sold within the internal market, which is hardly growing anyhow, and thus the surplus can only be traded on the world market by high subsidies. High quality and added value become increasingly impor tant and output markets have to be served and segmented according to the developments on the side of consumers and retailers. For example, the United Kingdom requires a zero 284 tolerance for the residues ofpesticide s in flowers. Other countries do not. Similar questions arise on vegetables and fruits. Hormones, medicines, preservation chemicals and microbial contaminants are increasingly resented, thus leading to changes in the markets. In some cases biotechnology can respond accordingly to these changes by contributing towards solving new problems. Similar developments occur on the input side. Especially in the last decade the limits of the production systems in primary agriculture and in agribusiness became visible and the competition between agriculture and other users oflan d restricted an unhampered agricul tural growth development. Pollution prevention and control and conservation of natural ecosystems and landscapes induce more differentiation in agricultural production and necessitate diminished use of inputs, such as fertilizers and pesticides. Genetic modification could be crucial in solving these issues. However, the technology must be applied to specific problems. The development of expensive technologies, such as genetic modifica tion, is in many cases not the only solution possible and it has to compete on relatively small markets in which sometimes only low added values can be generated. One of the reasons why agro-industries and plant breeders pay so much attention to herbicide resis tance is that the potential market for a specific herbicide can become relatively large due to the fact that the resistance gene can be introduced in many crops, leading to wide applicability of the specific herbicide. This strongly contrasts with genetic modification for solving crop-specific or even variety-specific problems. The incentive must be economically interesting. It is therefore almost impossible for industries which are not familiar with these markets to enter them.Join t ventures, acquisi tion and co-makership are possible strategies. But many competitors within the agricul tural sector are not convinced of the advantages of genetic modification. Dutch dairy industries generally believe that microbial diversity still offers so many possibilities that genetic modification of micro-organisms will not contribute to product development in the coming decade. Some animal breeders believe the same with respect to the breeding of cow and pig varieties. Scepticism is also encountered in molecular plant breeding with respect to pest resistance by defence mechanisms determined by single genes. 6.2.5 Characteristics of branches in agro-industries Diversity among agro-industries in different branches is large. The same applies to indus tries within one branch. It is therefore only possible to present some of the headlines of these branches. One of the major developments is the very narrow interaction between industries on the input side and farmers, industries and/or trade organizations on the output side. Especially the chains from feed to meat and, to a lesser extent, from seed to vegetable food or industrial products are controlled by particular industries, which are represented by products and services at every link in the chain. Many of these industries arose from mergers of cooperatives, active in only one link of the chain. As a result, consumer needs and attitudes are more often translated at earlier links than in the past on the one hand, while on the other hand the controlling firm is less interested in optimizing single links of the chain, but more interested in optimizing the total result. This means for example that R&D is concentrated more and more in one laboratory studying the whole 285 chain or that close cooperation arises between different laboratories. Although R&D spending in many areas is less than one percent of the sales, the critical mass for new technologies may already have been reached. Nevertheless, R&D is still primarily focused on process development and processing steps and not on product development for con sumers or other users. Genetic modification does not have a high priority in agribusiness. New biotechnological inputs are mainly induced or developed by specific producers of enzymes, additives etc., from outside the agricultural sector. Agro-industries for example use phytase from Gist-brocades, amino acids fromJapa n or medicinal mixes from pharma ceutical industries from all over the world and so they are more interested in mix-formula tions and application research. Beside non-agricultural industrial laboratories, much research with more fundamental or high risk application character and product development in the field of meat and feed is done by private or public research institutes. The same is true for pesticides, minerals and vitamins for plants. The use of biotechnology and particularly genetic modification in the processing of meat or other consumer end-products from different agricultural chains is met with con siderable scepticism. Genetic modification is not used in the commercial breeding of cows. This is done in industries which are far less integrated in the production chain as are the industries involved in pig breeding and even more so in chicken breeding. Here the breeding stations are generally a part oflarge r companies offering a wide range of products such as varieties, feed, offtake of fully grown animals by slaughter-houses etc., to the farmers. Genetic modification to breed better varieties ofchicke n and pigs therefore seems more likely as long as consumer aspects do not play a dominant role. Catde breeders and dairy industries do not have tight bands. Genetic modification is done in both private and public laboratories, as in the case oflactoferri n secretion in milk. Pharmaceutical industries are extremely interested, but dairy industries show considerable reservations. In addition to consumer acceptance, this attitude is caused by the structure of these industries. They are not prepared to process small amounts of milk for special purposes (e.g. pharmaceuti cals), because they focus on large-scale processing. It is not only a question of production capacity, but also of traditional relations between the cooperative industries and their members (dairy farmers). But even ifth e processing of milk does not have to be extensively adapted, dairy industries do not promote the use of genetic modification or its products. Their commercially successful traditional calf-chymosin for cheese processing, the produc tion of which is limited, may have to compete with chymosin that can be produced by a genetically modified yeast in any required amount. None of the dairy industries believe that genetically modified micro-organisms will contribute to product development in the near future. Thus, they are not active in research on genetically modified organisms. This isdon e by a private research institute which islargel y funded by the Dutch dairy industries. The future of this institute is currently being discussed. One of the arguments to cut budgets is that although the research on biotechnology has been of high quality, commer cial application of results is still not clear and therefore does not fit in with the R&D strategies of the financing industries. The technology of genetic modification has been most successful in plant breeding. A market for commercially interesting genes is currently emerging. In addition to NBFs and 286 subsidiaries of multinationals (such as e.g. Sandoz Zaadunie), traditional breeding firms are also becoming involved in 'molecular' breeding. Some ofthes e firms or their subsidiar ies are cooperatives. Others are privately owned family enterprises. Although they may have a strong market position for one or more crops, many of them are still only classed as small businesses. Traditionally, plant breeding is very research intensive and R&D spending is normally about ten to fifteen percent of the sales. Expenditure ranges from a few million to some ten million ECUs. Only part of it (10-20%) can be spent on biotech nology because the major development costs are made in the selection procedures and field trials in different climates and on various soils.Thi s also has some advantage for these firms because it poses a very high entry barrier to new competitors. On the other hand, the suppliers of genes will preferentially develop genes that can be applied in more than just one crop or under various environmental conditions. Generally, traditional breeders have to develop their own products for more specific applications. This is usually done in joint ventures, such as Keygene and Florigene or in cooperation with universities and research institutes. But on the other hand, it is difficult and risky to invest in specific research, because the markets in which profits have to be made are small and the market shares are relatively modest for many varieties and fast-changing for many crops. Some examples: the sales in the Netherlands in 1989 by Van de Have for sugar-beets amounted to NFL 576.000, by ZPC for seed-potatoes NFL 4.68 million and by Cebeco for wheat NFL 1.32 million (Table 6.2.1). Even though domestic sales are about ten percent of the total, the markets are relatively small compared with the costs of R&D and production. It is therefore not surprising that many traditional breeders Table 6.2.1. Total area (ha) of winter wheat, consumer potatoes and sugar-beet grown in the Netherlands in 1989, average costs of seed per ha. in Dutch Guilders (NLG), number of varieties grown, names of the three major varieties and their share (%) and three major suppliers of seed and their share of the Dutch market (%). H.Z.Z., Hollands Zweedse Zaadmaatschappij. Winter wheat Table potatoes Sugar-beet Total area (ha.) 130 700 71 300 123 800 Average costs of seed 177 1100 240 (NFL/ha) Total seed-market (NFL) 22 • 106 78 • 106 2.4 • 106 Number of varieties grown 9 11 10 3 major varieties Obelisk 61 Bintje 78 Univers 32 (name; share, %) Pagode 12 Saturna 4 Accord 27 Arminda 6 Eigenheimer 3 Lucy 18 Irene 3 3 major seed suppliers Wiersum 73 farmers 83 H.Z.Z. 37 (name; share, %) V.d.Have 10 ZPC 6 V.d. Have 34 Cebeco 6 Agrico 5 Cebeco 18 287 become customers of NBFs or research institutes, universities or other gene suppliers. The incentive for taking the risks and high costs for genetic modification in breeding program mes, as for example AVEBEs subsidiary Kama did in starch potatoes, must be very strong. A substantial part of AVEBEs markets are specialties, and competitive strength can be enlarged by the production ofpur e amylopectine and its derivatives. Genetic modification has proved to be a faster method of breeding potatoes with the desired starch components compared with traditional breeding. Many traditional breeders try to restrict capital costs in biotechnology or try to spread the costs by using the same techniques or genes in as many crops as possible. Even so, some of them enter markets for crops which have not been bred by them before. The choice of whether to invest in biotechnology becomes even more difficult because the product life cycle of new varieties is shortening. The Dutch auctioneers' representative body VBN believes that the short product life cycles has delayed the use of genetic modification in the breeding of flowers and potted ornamentals. This technology seems more promising for increasing shelf-life and for pest and disease control. Many genetically modified plants of food crops are developed in laboratories. Some of them can improve the quality of the product (e.g. colour of flowers) and others the storage life (e.g. tomatoes and potatoes). Also, resistance to pests, diseases and herbicides isa major objective of genetic modification research in plant breeding. A new field is the use of the plant as a natural 'fermenter' for the production of non-plant compounds, such as human serum albumin (Mogen). This development is, however, given more attention by the pharmaceutical industries than by agribusiness. This is partly due to the orientation of agro-industries on the quantitative aspects of breeding and production. In the summer of 1992 14 field trials with genetically modified organisms were carried out by private firms and three by public research institutes in the Netherlands. The first introduction of genetically modified plants on the market will take at least another three years. One difficulty for commercialization in the plant breeding sector is also the attitude of industrial customers on the other side of the production chain. Many of them are not vertically integrated. For example, the introduction of a new traditionally bred potato variety (Agria) took several years. Although the growers could readily be convinced of the advantages, it took years to convince consumers who had become familiar with the old yellow potato varieties to purchase varieties for deep-frying. Similar problems can be expected if the processing has to be adapted by a customer, or if the effect of a new raw material is not clear. Products of genetically modified plants will not be an exception. It can therefore be concluded that the use of genetic modification is determined first mainly outside the agro-industries and second that in the near future it will be of great importance mainly for industries on the input side. 6.2.6 Public acceptance and social issues One of the major obstacles in introducing genetically modified organisms in food, feed and other non-medical applications, is acceptance by the public. However, the critical or sometimes even negative attitude of farmers, consumers and environmental organizations will also pose problems to their introduction. This is certainly true for products such as 288 BST and PST. Especially in the case of BST, dairy industries will oppose the introduction of this growth and milk production stimulating protein, because they will meet opposition from consumers. For PST the proposition is slightly different because the feed producers have interests in the whole chain. But here again, the attitude of the public prevents their introduction. But not only consumers and farmers contribute to the discussion, as shown in the case of chymosin produced by a genetically modified yeast. Dairy companies in Germany have attempted to libel the image of cheese from Dutch dairy companies following an amendment of a law by which the use of yeast-chymosin was accepted. Another example is the lactoferrin-producing cows. Many discussions took place before the ethical advisory board advised positively to the Minister of Agriculture, Nature Man- agment and Fisheries to give permission for further breeding. The reproduction of these animals is however not undisputed and is on the agenda of the Dutch parliament. So far the use of phytase in feed seems to have been accepted because of its benefit to the environment and the great distance between feed and consumer products. The use of vaccines and other therapeutics in the agribusiness is stimulated by progress in human health care where these products have been accepted. It is noteworthy that industries with considerable experience in the acceptance of prod ucts and with good marketing divisions, made some very expensive mistakes by introduc ing products onto the agricultural market. This may be of great importance in the develop ment of new products and will delay the diffusion of genetic modification or its products to the field of agriculture. For genetically modified plants and animals the future is unclear. Is it possible to convince the public of the advantages of genetically modified organisms without affecting the image of existing products that are, for example, more polluting? Some opponents maintain that an economic advantage for producers is not a good argument, only advan tages for consumers and society as a whole matter. Preceding these marketing aspects, the discussion on the environmental effects of commercial introductions of GMOs must be brought to a satisfactory conclusion. For the time being it has to be done case by case. Thus, it will take a long time and will cost a lot of money. To a large extent, circumstances restrict the use of genetically modified organisms by agro-industries: Nobody likes to be the first, and cooperation with other suppliers is extremely difficult. In the competition between multinationals and medium and small agribusiness industries the cost and risk factors are a disadvantage for the latter. It is possible that the uncertainty of the outcome of the social discussion is affecting the struc ture of agribusiness more than the technology itself. Key developments are made by those who can afford and/or will take the risks. Many firms however, will only introduce biotechnology if its applications are no longer argued. 6.2.7 Government policy The role of the government in the diffusion of genetic modification in agribusiness is very diverse and complex. First, in the past much research for agriculture, especially in the field of plant breeding and farm practices was done in public research institutes at the expense of the government 289 and special purpose programmes were introduced with extra government funds, such as biotechnology innovation programmes. The existing biotech programmes have recently expired and government policy towards special programmes is changing. The policy with respect to funding publicly financed institutes is also changing. On the one hand research at these institutes is becoming more directed towards the goals of government policy, such as aspects of the public perspective on environmental issues, food safety guarantees, con sumer interest and other political aspects, for example the ethical acceptability of biotech nology. On the other hand, the funding is shifting from standard overall institute budgets to programme-dedicated financing. As a consequence, the research institutes have to adjust to the new policies, and long-term research objectives are under pressure from short-term financial needs. This is especially true where the size of the government funds for genetic modification is being reduced. The public research institutes will have to be financed to a much larger extent by agriculture and agribusiness. Without going into detail, it can be concluded that the diffusion of genetic modification will be influenced by this development. The research will be lesspubli c and industries will have to pay more for research and know-how of these institutes and, in future, will be confronted with patents and restrictive contracts. In addition, the institutes will specialize more than in the past, meaning that strategic partnership will become more important. It may be possible that some industries will find biotechnology inaccessible or that industries will choose their own R&D instead of supporting the public one. Second, the legislative framework in which genetic modification must take place is of great importance for the diffusion of this technology. The early innovators continuously faced new regulations and other legislative rules. Although industries have no doubts on the importance of good, publicly well-understood rules, they fear the uncertainty of the outcome of the political discussions and the feasibility of procedures (e.g. EC Novel Food Regulation) and other restrictions (e.g. labelling) for market introduction. This trial and error period also puts a high financial burden on the early innovators and adopters. Diffusion is certainly delayed by this development and, even more important, it prevents the rapid spreading of genetic modification in agribusiness. It also means that industries with a large capital strength from outside agriculture have a competitive advantage with respect to small and medium sized agribusiness industries, whose own capital position is normally not too strong. Third, future developments in agriculture are very uncertain. Not only is the Common Agricultural Policy changing, but also pollution control and the concept of multi-func tional use of land (agriculture, nature conservation, recreation) make it difficult to focus research at the proper targets. Will herbicide resistance be accepted or not, or will zero tolerance of pathogenic organisms on flowers or in vegetables become a prerequisite in international trade? Or is the breeding of cereals of interest when their production levels are restricted and the pressure on prices also favours the use offarm-save d seed? Especially those firms which are not diversified or vertically integrated bear high risks. They will postpone investment in expensive R&D. Fourth, the question of intellectual property rights is not adjusted to the agribusiness' satisfaction. However, for diffusion of genetic modification this problem must be solved as soon as possible. Without a clear cut agreement between patent rights and plant breeders' 290 rights and without the establishment of patents in the field of agricultural breeding, diffu sion of the new technology will be delayed because the advantages cannot be commercial ized to the satisfaction and benefit of all concerned. 6.2.8 Concluding remarks About two decades ago the first genetic modification of micro-organisms was introduced, and after a decade of introduction in higher organisms it can be concluded that the initial highly perceived potential and expected fast introduction has diminished. Although 'agricultural markets' are large, though sometimes very segmented, products from genetically modified micro-organisms do not contribute substantially to technologi cal developments in agriculture today. To some extent this is caused by technical queries but to an even greater extent because diffusion is limited due to the risks and uncertainties concerning the choices to be made in the field of R&D, production and market introduc tion. Many of these choices deal with consumer acceptance and the reluctant attitude of the public. The highest diffusion has taken place in sectors in which the concept of genetic modifi cation has the potential ability to solve problems that cannot eventually be solved using traditional technology and which are not directly concerned with consumer food. With respect to the appropriate investments in this technology as well as to the costs of participating in political discussions, the present status quo is not advantageous to agro- industries and to agriculture. It is possible therefore that agro-industries will not be able to develop biotechnology as a key technology but will have to limit their activities in biotech nology as users ofbasi c technology. This hampers competitive strength, which is so crucial in the fast-changing agricultural world of today. 6.3 Protecting and exploiting biotechnological inventions in agriculture: patent rights versus plant breeders rights? T.H.J.M. Hutten The introduction of biotechnological products and processes has caused considerable discussion about intellectual property rights. Biotechnologically related patents are now entering the field of plant and animal breeding which used to be dominated by plant breeders' rights and animal pedigrees. Emotions sometimes run high in the debate and there has been much misunderstanding on the strategic position of business and social organizations on the one hand and the effects of the new developments on the other. The major pros and cons are presented in this Section. Both systems can be useful tools for the development of agricultural breeding and for the protection of the rights of either (contracting) party. However, patent and plant breeders' rights must be in tune with each other and mutual interference must be regulated to the benefit of all parties. The present EC proposals on breeders' rights and the protection of biotechnological inventions do not meet these conditions. 291 6.3.1 Current position of plant breeders' rights In the past, intellectual property rights on plant varietieswer e settled by the plant breeders' rights regulations. Other plant taxons were not included in this regulatory system as there was no need to do so. Furthermore, there was no need at all to introduce a property rights regulation in animal breeding as owners of excellent parental animals with successful pedigrees can determine how the offspring are to be used. The traditional technology used in breeding is extremely difficult to patent because it is almost impossible to meet all the requirements for a patent (novelty, invention, reproducibility and applicability). So far, breeders' rights have proved to be effective in those countries where they are practised. However, the little support given by other countries, the application limited to a restricted number of crops and the free use of farm-saved seeds by farmers (called farmers' privilege) can be seen as major drawbacks of the breeders' rights regulation. The advantages of the plant breeders' rights system for breeders are: - a proper return can be obtained, - a minimum of litigation occurs, - the regulations are clear, - the assistance of legal advisors or patent agents is not required, - the rights can be managed at a minimal cost. Farmers also benefit from this system as it leads to increasing yields with decreasing costs and to better quality and maximal diversity. The system is embedded in the production system of agriculture and fits the reproductive character and the nature of vegetative materials well. Besides the previously mentioned drawbacks, other developments have made it neces sary to change the latest version of the UPOV (International Union for the Protection of New Varieties of Plants) Act signed in 1978 by the present 17 member states. These developments are mainly due to developments in technology, especially but not exclusively in biotechnology. Major issues are the definition of variety, legal aspects of derived varie ties, applications in other crops, scope of the rights, the interphase between patent rights and breeders' rights and the use of farm-saved seeds. Solutions to problems in these issues must make the system more attractive to national regulatory bodies and to potential owners of rights. UPOV therefore invited many national representatives, industrial inter est groups and patent experts to discuss these issues. The results, laid down in the Geneva convention act of March 1991, are somewhat disappointing. Nevertheless, breeders' rights have been reinforced: varieties of all crops can be claimed, the definition of a variety is more up-to-date, essentially derived varieties become dependent upon the authorization of the initial owner of the breeders' right, the scope is enlarged to other uses and exhaustion occurs through a cascading principle, and so on. However, there is still no clear division between plant breeders' rights and patent rights, and the mutual dependency which can arise is not well regulated. The translation of the UPOV act into an EC-directive shows the same omission. This is surprising, for only DG VI (Directorate General for Agriculture) is responsible for its launching in the EC. Besides, the opposition from agriculture, especially the farmers' and cooperative unions do not treat patent and plant breeders' proposals as being mutually 292 dependent. Their main intention ist o secure a plant breeders' rights regulation which is as favourable as possible for farmers and 'classical' breeders. 6.3.2 European patent rights on plants and animals Patent rights have proved to be of great importance for all branches of industry, including agriculture. For tractors, greenhouses or pesticides many (licensed) patents play a role. Nobody in agriculture has ever worried about these patents nor have there been any complaints about the costs of patents or their exclusion. For patent rights on plants and animals however, such criticisms are being heard. A patent right is only valuable when it is used. Otherwise itjus t costs money. The only advantage that could justify investing in non-commercialised patents is the possibility to halt progress by competitors. However, this is a risky strategy. Competitors often use patents from each other or build new patents on an existing one held by a competitor. This particular criticism of agricultural interest groups and others is thus largely unfounded. Some aspects of patent rights, however, may threaten agriculture or can disturb existing structures. The latter does not necessarily have to be bad if it stimulates a more competi tive, safe and sustainable agriculture and agribusiness. Agriculture cannot allow itself to hold off patent rights, because it must adapt to very fast-moving frontiers in social, politi cal, environmental and economic issues. Biotechnological potential for example, is clearly visible in the field of human and animal pharmaceuticals and in microbial biochemistry. On the other hand, there are real limitations to the benefits of (unadapted) patent rights for living matter used in agriculture. Patent rights must allow for the exceptional position of living matter in agriculture and must adapt to the complex nature of agricultural production (see also Section 6.2. on diffusion of biotechnology). To a certain extent this can be shown by the developments in patent law and patent litigation. International collaboration takes place in the World International Property Organiza tion (WIPO). This organization is responsible for the fundamental agreement on patent policies, settled by conventions. However, there is not just one patent law. Member states who have ratified a convention, must introduce the results of the convention in their national laws. Apart from the national laws there is also (since 1978) a European Patent Convention (EPC), currendy with fourteen members. The European Patent Office (EPO) is located in Munich. EPC and EPO play an essential role in the development of patent legislation. These regulatory structures, developed relatively independent of EC-policies, are closely related to the needs of, for example, chemical industries to protect their R&D investments and products. For living materials, the EPC first concentrated on non-genetically modified micro organism strains and later on genetically modified micro-organism strains. The complex matter of reproducibility and variability have been major points of discussion. These questions have now been answered, but there is still considerable inconsistency among national regulations of different countries, also within the EC. The discussion in the field of micro-organisms now concentrates on the question of whether living organisms or their pure natural products that already exist, can be considered as inventions. The next point in the debate iswhethe r plants and animals can be patented. Experience 293 and the results of the foregoing discussion are translated into animal and plant cells and differentiated tissues by broadening the interpretation of what is meant by a micro organism. This move was necessary because of article 53b in EPC which excluded patents for essentially biological processes and for plant and animal varieties. But if plant and animal cells, or their products, can be claimed in a patent, then, in the opinion of many patent officers or experts, plants and animals themselves can also be claimed, as long as they are not claimed as varieties. In their opinion higher taxonomie groups than varieties are not excluded and article 53b can be bypassed. However, this logically consistent argumentation is based only on the EPC point of view. EPO has recently granted a patent on the Harvard myc-mouse, thus creating a prece dent for animal patenting. Although this mouse is not a farm animal and would not, at first sight, appear to influence agriculture, it is still worthwhile to look into this example, for patenting of agricultural organisms could be the next step. The broad interpretation of article 83 EPC not only makes a mouse patentable, but also other mammals possessing the same gene. Interpreting EPC article 53b as narrowly as possible, it states that only varieties are not patentable, but no statement is made on higher taxa. However, article 53b does not contain a good definition for variety and therefore the statement on exclusion is redun dant. In the case ofth e patent application for the phaseolin gene by the Lubrizol company, EPO granted this request based on a similar interpretation. This interpretation of EPC articles will probably become standard practice, given the way that EPC interpretation is not required to be tuned to breeders rights. But the consequences for plant breeding may be serious. The point is that mutual tuning of patent rights and plant breeders' rights can now only occur in practice. Patent claims that are too broad or insufficiently limited must be met by opposition and invalidation procedures. Since this implies high legal costs for the parties interested in narrowing down possibilities of patenting plants and animals and will take a long time, thus longer periods about uncertainty of the outcome, firms will be reluctant to engage in such a procedure. This isespeciall y a problem for small and medium sized firms who produce products with relatively small profits, which is the case in agricul ture. The real threat for agriculture may well be that the two systems of intellectual property rights on plant material will thus exist in parallel without clear regulation on mutual dependency or demarcation. In animal husbandry the situation is more of a risk because there are no breeders' rights and animal breeders seem rather unaware of possible effects of patenting on their present position. So far, the market has been served well by a system in which breeders exclusively possess excellent parent animals for successful pedi grees. 6.3.3 EC-directive on biotechnological inventions From the EPC and EPO discussions it would appear that the development of patent law takes place independently of politics and democratic control in the EC, its members or other governments. However, this is only partly true because the European Commission is also preparing protective measures for biotechnological inventions. But the EC is not clear in drawing up regulations in which the rights ofbreeder s and patent rights are settled 294 to the satisfaction of all parties mainly because the interest ofth e biotechnological industry, which is the EC's main concern. This is illustrated by the fact that it was the DG III, the EC Directorate General for industrial affairs and internal markets, that prepared the EC proposals on protection of biotechnological inventions, without consulting DG XII (sci ence and R&D), DG VI (agriculture) or DG I (external affairs). First, the EC made an effort to address the problems arising from the nature of living matter, the definition of what is patentable and the scope of protection. Second, the directive takes the mutual dependency between patent rights owner and plant breeder into account. Some key elements of the proposal (which is still under discussion) are as follows: - The principle of patents may be applied to all biological materials with probably one exception, namely human beings. Therefore, it may also apply to all progeny of patented organisms, thus also to all identical or, for example mutated progeny produced from a patented organism and also to uses of plant and animal varieties and processes with which they are produced. The proposal restricts protection of material that isintroduce d into the market, but which is used for purposes other than those described in the claims of the patent. This may be heavily disputed. - Only plant material for which a plant breeders' rights can be obtained, is excluded. But higher taxonomie groups are not excluded, neither are animals. The nature of living matter and the unknown diversity of nature are thus not taken correctly into account. Besides, it seems that the definition of a variety and of biological processes are still confus ing and different in the EC proposals for plant breeders' rights and patents. - Chapter 3 of the proposal is also important, in which the mutual dependence between a patent rights owner and a breeders' rights owner is described. The position of a breeder holding breeders' rights and facing infringement by a patent, is weaker than that of a patent rights owner using a variety on which a plant breeders' rights have been established. Some examples: 1) The owner of a patent, who uses his invention in a variety only becomes dependent on the original variety under the new UPOV act. But he may use any variety and can establish newly derived plant breeders' rights after the introduction of his patented material into the original variety. 2) A plant breeder who wants to use a patent depends on the interpretation and definition of an essential technological development for a licence. 3) A patent rights owner licensing his patent to a breeder who then uses it to produce a new variety and establishes breeders' rights on this new variety, always gets a licence for this variety in order to exploit his patent rights embodied in this newly devel oped variety. Thus, in this case a plant breeder is never the only one who exploits his breeders' rights. 4). Although perhaps of little practical importance, the patent rights owner is able to refuse a licence for a breeder for the first three to four years after the patent has been established. Summarizing, it seems that the scope of the patent proposals is as broad as possible and/or covers a whole range of applications, with which the owners of patents are able to defend their rights well. The extent ofth e rights only depends on the type ofpatent , but not on other property rights which are common practice in agriculture. The risks of a specific application of successfully patented inventions in a variety and the introduction of this variety into small markets for specific agronomical purposes are left to plant breeders and these applications can be excluded if it suits the patent owner. 295 The implementation of the UPOV convention may, on the other hand reinforce the position of breeders. This, however does not mean that the interface between patent and plant breeders' rights is regulated in an effective and efficient way for both parties. One consequence is the threat to survival of small and medium sized plant breeders operating independendy from the multinationals or other strong participants on the same markets. 6.3.4 Balancing both rights The advantages of plant breeders' rights for the agricultural sector are not disputed. The same is true for patents outside the field of living matter. Nobody worries about these patents because they have little negative influence on the availability of tools and equip ment and nobody complains about the costs of patents. Recent (bio-)technological developments and their economic and social consequences make it necessary to adapt both rights, though special complications are met at the interface and by the nature of living materials. A second conclusion is that neither the developments within the EPC and EPO nor the developments in the EC regulation on plant breeders' rights and the EC directive on biotechnological inventions answer the questions mentioned earlier. At least for agricul ture, it can be concluded that the existing and developing patent law and case law do not fulfil its needs. More sophisticated and stronger plant breeders' rights are not in themselves an adequate solution. Agricultural interest groups continue to focus their attention primar ily on plant breeders' rights. The maintenance or even the extension of the farmers' privilege to use farm-saved seed fundamentally undermines the breeders' rights and the demarcation and tuning against patents rights. In order to support proper demarcation and tuning ofbot h rights, preference should be given to plant breeders' rights as powerfully as possible. The result could then be laid down in a special regulation which would replace the EC directive on biotechnological inventions. The following issues need to be discussed so that they can be part of a regulation in which property rights on biotechnological inventions are established: 1. A uniform definition of a variety. 2. Resolution of the question of whether products of a higher taxonomie group than a variety can or cannot be patented. 3. Since process patents differ from product patents (the latter interfere with plant breed ers' rights) one should discuss whether process patents may be restricted to, for example the directly produced living material and products thereof. 4. The question of a restrictive or broad interpretation of applications in the field of plants and animals. Will a successful invention producing genetically recombinant mice with specific properties lead to patents on all mammals with the same properties? 5. The question of a restrictive or broad interpretation of the requirements for a patent. Can the identification of natural genes be regarded as an invention? 6. Will a patent right be lost if its effect is overruled by nature? This, for example, is the case when a biotechnologically introduced resistance against a disease is disregarded by the disease pathogen. 7. If a plant breeder infringes a patent in order to acquire new plant breeders' rights on 296 a variety, or a patent is acquired with an infringement on plant breeders' rights, there must be rules by which mutual dependency can be solved. Is an obligatory licence possible when the parties do not agree? Even though the discussion has been going on for about four years, the answers to these questions are still not clear. But they must be solved, otherwise agriculture and agribusiness are facing financial and time-consuming problems, which can also affect market positions and competitiveness. This should be avoided because biotechnology is able to supply new and sometimes much needed solutions to the many problems faced by agriculture. It is also of interest for other industries, because in the future they will need agricultural products more than they expect as sustainable inputs. 6.4 The development of risk-evaluation for the deliberate release of genetically modified organisms: An outline of the international discussion H.E.N. Bergmans The introduction in the early seventies of new molecular technologies started a new era in the discussion of safety in biotechnology: the recombinant DNA debate. The sudden emergence at that time of techniques that enabled scientists to identify, isolate and manip ulate genes essentially at will, gave rise to a feeling in the international arena that this was a powerful technique with possibilities to 'create' novel organisms, with unpredictable qualities. The general feeling was that genetically manipulated organisms might pose risks, albeit hypothetical. The discussion resulted in the adoption in many countries of some form ofguideline s for the safe use of this technique, most of them derived from the seminal guidelines of the US National Institute of Health. The discussion went on between the industrialized countries, member states of the OECD, in an ad hoc group of government experts on safety and regulations in biotechnology. In 1986 this group published another seminal report under the title Recombinant DNA Safety Considerations, also known as the 'OECD Blue Book'. In this report a first consensus view was presented on safety considerations for industrial, agricultural and environmental applications of organisms derived from recombinant DNA techniques, as the subtide of the report states. By that time it had been realized that genetic modification (the word manipulation was considered less suitable) and its products, the genetically modified organisms (GMOs) are neither inherently risky nor safe. Safe use of the technique and its products implies the availability of knowledge of the organism, the genes used to genetically modify the organ ism and the environment in which the GMO will be released. GMOs are governed by the same biological laws as organisms modified by more traditional breeding techniques, but genetic modification differs from traditional breeding methods in the respect that much broader genetic resources can be used, although it is recognized that the technique allows greater precision in the use of these resources. From this discussion it became clear that a distinction should be made between application of GMOs in laboratory settings, in which 297 adequate physical containment prevents the escape of GMOs into the environment, and applications of GMOs with low or essentially no physical containment, e.g. large-scale industrial use, or deliberate release of GMOs into the environment. It was argued that for these cases a great deal of knowledge should be available on the GMO and its interaction with the environment. The international discussion has been carried on in the OECD Group of National Experts on Safety in Biotechnology. This group has published several reports on safety considerations concerning GMOs, in large-scale industrial applications, introducing the concept of GILSP (Good Industrial Large-Scale Practice), GDP (Good Developmental Principles: guidelines for the design of small-scale field experiments for the deliberate release of GMOs); a report on safety considerations concerning the large-scale release of genetically modified crop plants is imminent. As work progressed it became necessary to once more give an oudine of the general principles for safety in biotechnology, as a reflection of the current knowledge, in order to provide a framework for further development. Here we will present an overview of the general principles, and of the way these principles are applied to genetically modified crop plants. It should be mentioned that the following is the personal interpretation of the author of the state of the art of the international discussion. It does not represent an official view of any of the participants in the international discussion. 6.4.1 Safety in biotechnology Biotechnology is that part of technology involving the use and exploitation of living organisms to meet the requirements of mankind. It differs from other parts of technology in the respect that its subject lives and proliferates, posing particular practical and ethical considerations. Safety in biotechnology is obtained by application of risk analysis, and subsequent appropriate risk management, at the different stages of development of a biotechnological product. In this paper we will focus on the safety of GMOs as one part of safety in biotechnology. We will consider some of the concepts that have been developed for the safe development and use of GMOs. Although these concepts have been developed in the GMO debate, they will be generally applicable to all biotechnology. 6.4.2 Stepwise development Like any scientific progress, the development of a GMO proceeds in a stepwise fashion through a number of stages. The concept of stepwise development has been a key concept in the discussion of safety in biotechnology. As each step is taken in this process, the relevant information of the previous stage is analyzed before the step into the next stage is taken. Safety considerations are always inherently part of this process, and at each step some form of risk analysis isdone , in order to decide whether the next stage can be entered and, if so, to design the risk management appropriate at the next stage. The production of a new genetically modified plant cultivar, e.g. a potato with a gene, derived from a bacterium, which codes for the production ofa n insect toxin, an insecticidal 298 protein, that will cause some degree of insect resistance in the plant may serve as an example of stepwise development. The process starts with a planning stage, putting together all available knowledge on the plant, the insects that are particular pests to the plant, the available insecticidal proteins and their specificities, the strategies to isolate the gene expressing the toxic protein and the methods available to obtain expression of the toxin in the plant at practicable levels. Subsequently, the gene is isolated from its genetic source, a variant ofBacillus thuringiensis in this case, it is characterized (usually its entire structure will be determined), and pieced together to signals that will govern its expression in its future host: the potato plant. The genetic information is then passed from the test tube to the bacterium Agrobacterium tumefa- ciens,whic h has the natural ability to pass on genetic information to plant cells. The resulting Agrobacterium line is used to infect potato cells, some of which will take up the foreign DNA and integrate it into their chromosomal DNA. In this way a genetically modified cell is made. From these cells new plants can be regenerated, which will possess the identical genetic makeup of the original modified cell. In the next stages, the performance of the genetically modified plants is studied. Only a small part of the plants modified in this way show the desired characteristic: production of the insecticidal toxin, without appreciable side-effects on general growth and development of the plant. The best performers are selected in growth chambers and glasshouse experi ments, and promising plant lines will eventually be tested in field experiments. 6.4.3 Risk analysis in biotechnology One should realize that any risk analysis starts with hazard identification. Only if a potential hazard can be identified will the associated risk be assessed, in order to design the appropriate risk management. Risk analysis in the development of a GMO is based on the characteristics of the organism used, the introduced traits, the environment into which it is introduced, and the interaction between them. In general, the hazards associated with the release of a new GMO into the environment are considered to be: the behaviour of the GMO, its potential to behave as a weed or pest through competition and damage to other organisms in the environment, and the possibility of the GMO to spread its new characteristics to other organisms by sexual crosses or parasexual processes in micro-organisms. Hazard identifi cation of a new biotechnological product is usually not straightforward because of a lack of experience with the interaction of these elements in the setting of the new product. Familiarity with the elements in some form may however compensate for the lack of experience. This concept of familiarity has developed into a key concept in the discussion on risk analysis in biotechnology. In the example given above, pest resistance in general is not a novel trait in potatoes, although the actual molecular mechanism of the resistance caused by expression of a bacterial toxin is novel. In the next Section this example will be used to show how the concept of familiarity may help in hazard identification and risk assessment of GMOs. 299 6.4.4 The concept of familiarity The concept of familiarity typically operates at different levels of comparison when differ ent issues of the same transgenic trait are discussed. The hazard identification and risk assessment of potatoes with a cloned Bacillusthuringiensis endotoxin gene are a clear exam ple of this. Influence of the toxin on weediness For this issue a comparison can be made with similar traits, basically irrespective of the molecular mechanism. Potatoes are very common in the Netherlands as a crop plant. They show no tendency at all towards weediness, and potatoes have never been found to settle outside managed agricultural settings. Insect resistance is a well-known trait in potatoes; in fact potatoes are relatively pest resistant because of toxic substances (glyco-alkaloids) that occur naturally through the genus Solanum. Wild Solanum species related to the potato {Solanum tuberosum) have been used as a genetic source to cross pest resistance into the potato. This has never noticeably increased the weediness of potatoes. Irrespective of the molecular mechanism insect resis tance is not assumed to cause weediness, if similar insect resistance has already been introduced into the same plant by other genetic means. Damage to non-target organisms For this issue it is necessary to be familiar with the actual molecular mechanism ofth e trait, although not necessarily in the same genetic background. The possibility that insect resistance could also damage innocuous insects, possibly even endangered insect species, is an identified hazard. Familiarity with the molecular mecha nism of the trait as it is expressed in the original Bacillus thuringiensis can be used here to assess the associated risk. Bacillus thuringiensis endotoxin is a class of toxins that has been studied extensively. It is known that the toxin is produced by the bacterium in an inactive form, as intracellular crystals. The toxin is activated by proteolytic cleavage in the gut of insects that swallow the bacterium. The active toxin then binds to epithelial cells of the insect gut, on specific receptor molecules on the cell surface. There are hundreds of different strains of Bacillus thuringiensis, which form toxins that use different specific receptor molecules on the cell surface. Familiarity with the molecular mechanism therefore leads to the reasonable certainty that damaging effects of the toxin will be restricted to the group of insects for which susceptibility to the toxin has been shown. Development of resistance to the toxin in the target organism For this issue the expression of the trait in the organism has to be considered. In the genetically modified plants that are equipped with a Bacillus thuringiensis en dotoxin, the trait is usually expressed continuously and at a relatively high level in the entire plant. This results in constant exposure of the target insect population to the toxin. Experiments in which the toxin produced by the bacterium itself is sprayed on plants have shown that such a constant high exposure may lead to resistance development in the 300 population. This is a scientific fact. Whether it is also seen as an identified hazard is in principle a political question. The bacterial endotoxins are seen as relatively environmen tally-friendly pesticides, which one would not like to loose through the unnecessarily fast development of resistance. If this is taken to be a hazard, then the associated risk can be assessed by familiarity with resistance development observed when the endotoxin from the bacterial source is used, and by knowledge of the level of expression of the toxin in the GMO. If it turns out that expression of the toxin in the GMO ishig h enough to cause resistance development, it may be deemed necessary to implement risk-management practices when the GMOs are re leased into the environment. It may even be preferable not to release the GMO at all, but to construct better GMOs, e.g. GMOs those in which the trait is not expressed continu ously, or with two different toxins that both have specificity for different receptors in the same target insect. 6.4.5 The case-by-case approach To a certain extent the procedure of hazard identification and risk assessment described above is a formalized approach. The number and intricacy of the factors to be considered in actual release situations is however quite high, as the properties of the GMO, the environment, and the interaction of these factors are involved. This precludes a general ized approach of the risk evaluation of the deliberate release of GMOs. Currently each release situation isjudge d on its own merits, in a case-by-case approach. Even so, in the international debate a certain degree of categorization becomes appar ent. Crop plants, being typically man-dependent for growth and survival, are deemed relatively safe host organisms for genetic modification. Traits that influence typical agro nomic properties of crop plants, and for which there is a certain degree of familiarity, will in general change the properties of the crop in a predictable way; it is a simple matter to check whether the prediction holds for the actual GMO. In some cases, these considerations have already led to categoric statements. For in stance, the use in plants of a certain bacterial antibiotic resistance gene, causing resistance to kanamycin, as selectable property during the development of a GMO, is generally thought to be safe. As the concentration of this antibiotic encountered by the GMO in the environment isno t high enough to cause any selective advantage for the resistant plant, the trait will not contribute to the weediness of any GMO. The trait is not likely to have any toxic effects; in fact bacteria expressing this trait are frequent, familiar, guests in our gut micro-flora. Similar risk evaluation is being done for other genes that are used as marker genes in GMOs, and which are generally thought to be safe for similar reasons. Examples are genes encoding resistance to other antibiotics, such as hygromycin, and genes coding for en zymes, e.g. /^-glucuronidase, that are easily assayed and therefore useful for quantitative estimation of the expression of such a gene in the GMO. 301 6.4.6 Conclusion The development of risk-evaluation for the deliberate release of genetically modified organisms is rapidly evolving towards a rational system of hazard identification and risk assessment. The driving force behind this process is the fact that large scale release of GMOs into the environment is imminent. As this can only be permitted if and when risk evaluation has shown that there is only a negligible associated risk, there is an acute need for clear evaluation of these risks. The international discussion, that has been treated briefly above, ismovin g towards such a clear evaluation. Familiarity with the novel trait at the appropriate level is a key concept in this risk evaluation. 6.5 Safety and acceptance of biotech-food D.A.Toe t Biotechnology has been part of food production ever since mankind started to modify primary crops and products for a variety of reasons such as improvement of yield, digesti bility, nutritional value, taste, appearance, shelf-life or just safety. Through the ages it has proven to be a safe method when applied according to good manufacturing practices. From a historical perspective it was quite unexpected that this technology would become the source of so much debate. This was even more unlikely because biotechnology is such a broad concept that the word itself is almost meaningless! It will be evident that the prime cause for debate and concern is the application of genetic modification or recombinant DNA technology in traditional food biotechnology. Public attention for this new development has been fuelled by unbalanced and exagger ated publications and predictions from champions as well as from adversaries. In addition, the complexity of the subject and the lack of knowledge and understanding among those who are not directly involved, makes the hesitation about the degree of acceptance of biotech products understandable. Before drawing conclusions from this apparent hesita tion, it should be emphasized that history has taught us that no single new technology has found its place without having gone through a period of sometimes strong public resistance and uncertainty. This of course does not mean that leaning back and waiting will end the present uncertainty about public acceptance. On the contrary, action isneeded . However, it requires an understanding of the mechanisms behind public acceptance and the willing ness of those involved in biotechnology to think and act in unorthodox ways when it comes to supplying and sharing information. In this Section, I will use the experience gained at Gist-Brocades as a case study to highlight a number of features concerning the public acceptance of biotech-food. 6.5.1 What is acceptance? Defining 'acceptance' in a scientifically sound way goes beyond the scope of this article. The best way to describe the situation where acceptance of a product has been reached is 302 that mutual trust and respect exist between consumer and producer. Emotions play a rather important role in that process. In this respect it is remarkable that the public acceptance issue largely seems to be limited to food. Pharmaceuticals have hardly met with any difficulties, not even in countries like Germany where emotions about genetic modifi cation run high. Increased public awareness is not limited to biotechnology. Public accep tance has recently acquired a much broader meaning and hence has become a critical requirement for a successful market introduction of a new product. The main reason for this isa n apparently increased interest ofconsumer s in ever more specific product informa tion, covering a wide range of aspects such as packaging, ingredients, origin. In relation to biotechnology several surveys on public acceptance have been carried out on a European as well as on a national basis (Eurobarometer, SWOKA, etc.). Although it is almost impossible to extract a general message from these surveys, a few things are more or less clear. On average there is a low level of understanding of biotechnology, with levels somewhat higher in north-western Europe. At the same time the public is sceptical of the technology as such, although it is very important to point out that there is no total rejection. There seems to be a rather changing attitude towards genetic modification depending on the objective or the resulting product. If certain modification results in a product offering distinct advantages for the consumer, which are e.g. environmental, nutritional or price, there seems to be an apparent willingness to accept these products. In relation to the subject there seems to be hardly any hesitation on the genetic modification of plants and micro-organisms. Applied to animals, the public tends to be more reluctant, especially from an ethical point of view. The two elements mentioned - increasing public awareness, low level of understanding ofbiotechnolog y - must be considered simultaneously; each one on its own is meaningless. The overall conclusion from the surveys can be summarised as being positive with a warning. The public isawar e ofbiotechnology , has shown interest, has its doubts but is also ready to accept beneficial products provided adequate safety and transparent information is guaranteed. 6.5.2 Present situation Never before has the food supply in Europe been as varied, diverse and healthy. When inspecting the total package, it ishardl y possible to identify a single product where biotech nology in its broadest sense has not been applied at some stage of its production. However, if the narrow meaning of biotechnology is considered, then only a few such products are actually on the market. The next few years will provide increasing numbers. According to a statement made by Henry Miller of the US-FDA, approximately 400 biotech products are in FDA's pipeline. At present, the number of biotech products is limited to a few enzymes produced with genetically modified micro-organisms. Among them are chy- mosin, a vital ingredient in cheese-making, and phytase, an enzyme which allows the amount of phosphor to be reduced in animal feed, thus reducing environmental pollution from manure. It is interesting to take a closer look, from an acceptance point of view, at the develop ment and introduction of both products, because they serve as a model for other biotech 303 products. Several parties have played a role here, the most influential being the regulatory authorities, user(s), consumers and producers(s). But, I should emphasize that (contrary to repeatedly heard statements) regulation for these types of biotech products already exists. General food legislation prohibits the sale and marketing of products unfit for human consumption. Before specific requirements for the application ofgeneti c modification were formulated, regulatory authorities followed a case-by-case approach. Authorities in the USA and UK have been especially critical and constructive in its application. At the European level the case-by-case experience has now been brought into the regulatory framework of the Novel Food Regulation, which is currently under discussion. Three elements play a role in the regulatory evaluation, namely safety, efficacy and product identity. In close cooperation with the authorities a safety evaluation programme for chymosin and phytase was developed and carried out, which led to the generally accepted1 conclusion that both products may safely be applied in food and feed produc tion. Efficacy has been proven in an extensive number of application trials under production conditions in many countries. Product identity is a more complex criteriom. Apart from establishing the identity of the product itself which, in the case of chymosin, is absolutely identical to that of its traditional counterpart extracted from the stomachs of calves, end product composition is also often analysed. This is done from the perspective that applica tion of the product may not change the end product in such a way that it could mislead the consumer, e.g. by using inferior raw materials. Product as well as end product analyses have clarified the fact that the end product characteristics have not been changed, either in nutritional or in organoleptic or compositional aspects, so as to be misleading for the final consumer. In the Netherlands, the complete set of data has also been discussed with consumer and environmental organisations. During these discussions it became evident that from a safety and efficacy point of view both products were quite acceptable to these organisations. Consultations with these groups revealed two other aspects, the first being the benefits of the specific product for the consumer. Given the nature of the food industry where progress comes in successive small steps this should be put in a broader context, e.g. improvement of production efficiency or qualitative improvement of quantitative minor ingredients should also be seen as advantageous. A second element is the need for con sumer-orientated information expressed by these organisations. Discussions are now going on between the biotech industry and consumer and environmental organisations to find mutually acceptable solutions for these problems with a preference for exploratory systems on a voluntary basis where need and effects can be evaluated. 1 Both products have been individually evaluated by both national and international health authorities. Specific authorizations for one or both of the products have been given a.o. by the USA, UK, Germany, the Nether lands, France, Belgium, Italy, Switzerland, Sweden, Finland, Norway, Australia and FAO/WHO. 304 6.5.3 Future developments A number of lessons is to be learned from the case study, and I will go a step further and formulate them as precepts. Taking full advantage of the opportunities for the food industry offered by biotechnology will only be possible after a set of conditions have been met. A very basic need for all parties involved is a harmonised European regulatory framework which on the one hand should set clear requirements for industry while on the other should satisfy consumer demands for safe and efficacious products. This framework should also be balanced against regulatory developments in the rest of the world in order to avoid competitive disadvantages for the European industries. The latter would inevita bly lead to a situation in the foreseeable future where Europe would find itself faced with an outdated and inefficient food industry. However, an acceptable regulatory framework, however important it may be, will not be sufficient. Consumers have expressed justified requests to be an equal partner in the process of exchanging information and opinion forming. It is difficult to envisage such a participation incorporated in a regulatory ap proach. The only practical way out seems to be an open dialogue between the industry and interested consumer organisations at a European level. It is obvious that for such a dialogue to be effective and fruitful an open atmosphere is needed. The first objective should be to build up mutual confidence and respect rather than aiming at complete agreement. Industry must be prepared to involve consumers at the earliest possible stage taking into account patent positions. As a consequence, the industry must also be prepared to accommodate justified consumer concerns. Consumer organisations as well as the industry should be equally prepared to drop traditional ways and to develop innovative thinking especially in relation to product information systems. Such a dialogue could imply a certain commitment from consumer organisations. Although at present this is somewhat against their nature its emergence is to be seen as a clear signal of their changing position in our society. Surveys have revealed that these organisations are the most trusted ones among the public. Such a position lends authority to an organisation but at the same time loads a heavy responsibility on its shoulders, which goes beyond the old idea of consumerism. Equally well it puts pressure on the industry to enter into the dialogue with an open mind and as much transparency as possible about its intention in and with biotechnology. Only if a transparent and effective dialogue can be established and maintained can biotechnology be developed to its full potential for the benefit of all. 6.6 Technology assessment and agricultural biotechnology A. Rip Technology assessment is becoming increasingly important, and increasingly varied in form and methods, and in its objectives. The Netherlands Organization for Technology Assessment (NOTA), in its programme for the 1990s, distinguishes three main approaches. 305 The oldest one being the approach emphasizing technological forecasting and impact analysis. In industrial firms, in health care and in other societal sectors, this approach is the dominant one. The methods involved here are also used in the other approaches when relevant. In the second approach, technology assessment issee n as a special kind ofpolic y analysis and/or ex ante evaluation. Governmental and high-level strategic decision-makers use varieties of this type of TA. The USA Office of Technology Assessment (OTA) has a record of such studies, which are generally available and often widely used. A third approach, especially in Europe, has evolved where societal actors and groups (ranging from industrial firms and government agencies to community and environmental groups) are to be reached with studies, workshops and other activities, so as to stimulate social learning and interactions about new technologies and their effects. This approach is particularly suited to take the variety of attitudes and responses to new technology into account, and can be seen as an input into the process of articulation of acceptance (see Section 6.1). 6.6.1 Activities of the Netherlands Organization for Technology Assessment NOTA has been active in the technology assessment of biotechnology since its inception in 1986. The Netherlands Council for Agricultural Research occasionally performed forecasting and impact studies, and now prepares itself for a more all-round effort in TA. Other organizations occasionally perform TA-type studies or activities; one example is SWOKA, the Institute for Consumer Research. NOTA's activities covered three areas. First, biotechnology and world food production which was discussed at a major conference in 1987, with an important follow-up study (by Bunders and collaborators) focusing on biotechnology for small-scale farming in develop ing countries. Impact studies and policy analysis were combined with articulation of new options for biotechnology, in relation to articulation and assessment of social needs. Also pro-active, but in a different way, were the activities in the area ofdeliberat e release of genetically modified organisms. Instead of biotechnology and its applications, the views and attitudes of various actors and social groups were the focal point. A carefully struc tured pair of workshops was held in 1988, with the OTA report on Field-Testing Engi neered Organisms as technical input, and with opportunities for participants from differ ent organizations and groups to argue for and against the various views. No consensus view emerged, nor was this expected. But some mutual understanding was created, to gether with possibilities for further interaction. NOTA continued its efforts in this direction by producing a discussion agenda for assessment and debate of deliberate release, and by trying generally to bridge the gaps that exist between promoters and critics of biotechnol ogy- In the third area, genetic modification of animals, ethical aspects were prominent, in addition to traditional issues of regulation. NOTA produced documents that summarised and articulated the positions, and produced an analysis of the issue of patenting, and an overview of progress of regulation in six countries and the European Community. 306 The focus of NOTA on social learning and interaction will be clear from this brief overview. The studies often start with a summary of views and attitudes, and are thus an alternative to opinion polls. The latter promise 'generalizability', but often suffer from methodological and interpretative limitations, as has been pointed out in an in-depth study of'PublicJudgemen t of New Technology' commissioned by NOTA. A recurrent problem is that the general public is not really aware of a new technology, and when asked to voice an opinion or air its views, will fall back on the general repertoire available in our culture ('don't mess about with nature'; 'if we can alleviate human suffering, we should allow ourselves more margins'; etc.). Social learning about the new technology still has to take place, and actors who are promoting the technology and therefore want to know what to expect as to acceptability will have to wait until this learning process has occurred. Or, preferably, contribute to it. Industrial firms and other relevant actors are increasingly collaborating with NOTA to this end. In addition to circulating its reports and other communication of its results, NOTA itself experiments with the possibilities of Public Debate. 6.6.2 Activities of the Netherlands Council for Agricultural Research While NOTA has a variety of audiences, and addresses society in general, NRLO, the Netherlands Council for Agricultural Research, directs itself in its TA activities to the world of agriculture. Two recent TA-studies of agricultural biotechnology, funded by Ministry of Agriculture, Nature Management and Fisheries illustrate this point. Safety of genetically modified foods was studied through a literature survey of (advisory) reports on assessment strategies and criteria, and the summary report was discussed in a workshop with a broad range of relevant actors. Societal aspects of genetically modified potatoes were analysed through a diagnosis of the sector and the issues, checking for the various social aspects. Again, the report has been discussed in a workshop with the relevant actors. Thus, on a smaller scale than with NOTA, social learning still is an important aim, operationalized as articulation of relevant knowledge, its diffusion and take-up, at least in workshops. 6.7 Ethics and agricultural biotechnology J.M.G. Vorstenbosch Human activities such as science, economics, technology and politics are built on a partic ular value-foundation that is often taken for granted. For instance, economics as a human activity is built on the value of human welfare or want-satisfaction. For practical reasons the interpretation of this value must be taken for granted in order to make a systematic effort for realizing this value possible. The market and modern technology are important ways in which this value is realized in our modern society. Ethics as a systematic and normative discipline starts with the concept of human prac tice as an activity that gains meaning and purpose on the basis of values. Ethics tries to 307 make explicit, articulate and interpret the values that enlighten this practice and studies the way they give rise to moral standards and how they are constrained by moral principles. Some of these values, for instance health or well-being, point to a concentration on consequences of actions as the main criterion of morality. Other values refer to some morally relevant property of human actions or to standards that should not be violated, for instance standards of integrity or justice. While philosophical ethics concentrates on the foundation of these standards and principles, applied ethics looks into actual problems, related to particular practices, and to the moral implications of decisions in these practices. To be sure, in applied ethics more attention needs to be paid to the practical and empirical aspects of the situations in which ethical problems arise and to the manner in which these problems are experienced and perceived in these particular practices. But the aim of applied ethics is still to arrive at an ethical assessment of human action. It is not a descriptive or empirical affair. Given the different orientation of practices such as economics and technology on the one hand and ethics on the other, it is no wonder that there is a certain tension between them. It is the tension between the sustained effort to realize one particular value in a consistent way and the reflective, distanced attitude which is required in ethics - not just in its philosophical but also in its applied form. Nevertheless, this reflective attitude is needed to set the practice and its value in its proper perspective, for instance by relating it to other important values and comparing it to other interpretations of the same value. 6.7.1 The need for ethical reflection Ethical reflection is a theoretical activity that has a value in its own right. But there is a special and urgent practical need for ethical reflection in times of technological, ideological or institutional change in society. These changes almost always give rise to uncertainty and to conflicts of interests and responsibilities, to disagreement about moral standards and about the course society ought to take. Ethical reflection and discussion in those cases can no longer be relegated to the philosophical classroom. In the case of modern biotechnol ogy, which has launched the possibility of using recombinant DNA-techniques to obtain 'new' organisms - the technological, ideological and institutional dimensions of change all play a role. In the first place the developments have provoked ethical debate on the moral acceptability of concrete applications of the new technology. In the second place biotech nology has generated or given new impetus to ideological discussions about the course agriculture has taken in the 20th century, for instance by bringing to public attention the ways in which animals are treated in intensive husbandry and providing a new case for the proponents of animal rights. On the institutional level there are also changes. For more than a century agriculture in the Netherlands developed successfully by a combined effort of government, agricultural corporations and farmers. In this way it gained a prominent position in the world. In the case of biotechnology other agents are entering the field, for instance international companies that invest heavily in the development of gene constructs and genetic techniques and may well have a 'moral culture' that isver y different from the agricultural corporations and farmers. The national government takes a strong interest in developments in the field of biotechnology, not only by subsidizing them but also by legal 308 regulation and ethical review by government committees. Regulation on the national level also has to reckon with developments within the European Community. In the following I will largely refrain from the broader institutional question concerning which agents have what moral responsibilities. I will try to describe more systematically the ethical issues surrounding biotechnology by taking up the idea, suggested in the introduction (see Sec tion 6.1), that biotechnology isa novel input and by applying this idea of novel input to the ethical discussion on agriculture, in particular animal husbandry. Has biotechnology pro vided novel ethical problems or has it necessitated new ways of dealing with ethical problems? 6.7.2 Changes in the support for agriculture Throughout history, certainly agriculture has been one of the most important human practices and in many ways still is. It is directly linked to the fundamental needs and interests of human beings. These interests concern not only survival, but also the relation ship of human beings to living nature, the role of agriculture in shaping social life and the meaning of improvement and consummation of'the fruits of the earth'. Historically a rich harvest - and as a result a growing population - was considered to be a sure sign of a healthy and well-governed society. In our age of agro-business and biotechnology the societal support for agriculture is no longer unambiguous. It is mingled with concerns about overproduction, questions ofjus t distribution, worries about the environment, calls for more respect for nature and for animal protection. In many respects modern agriculture is a victim of its own success. It has gained relative freedom from external influences of a political, moral and religious nature and has allied itself with science and technology in a way that has proved to be extremely successful in terms of productivity and efficiency. The ethical problems that are called forth by agricul tural biotechnology, reflect the tension that exists between certain values, such as knowl edge, welfare and freedom, and the principles of ethics that are intended to protect or promote these values. Science, technology and economics have realized these values in a way that is no longer in line with what society can subscribe to without any problem. One of these anxieties is that science is producing more and more knowledge about reality, while people are more and more at a loss about how and by whom this knowledge can or will be used, or even what this knowledge is good for. Another anxiety is that technology and economics are steadily producing a more varied product range and possibilities for consumer choice, while there is also strong scepticism about the autonomy of consumers. Many people even speak about 'the technological imperative' which forces people into a strait-jacket and makes 'real' choices impossible, not only at the micro-level of individual choice but also at the level of society. Awareness ofthes e contradictory tendencies can help in obtaining a better view of public reaction towards biotechnology. The balance between the way in which particular practices have developed, especially under the influence of science and technology, and the value-system of society as a whole is endangered by this process. The interest in ethics has much to do with the need for finding a new balance. But finding a new balance is not only thwarted by the fact that practices have their own dynamics, it ismad e much more difficult by the fact that speaking 309 of 'society as a whole' in our pluriform society is to a large extent unwarranted, at least in several ethical debates. The new balance cannot be found in an undisputed conception of the ethically good and desirable. On many points it is precisely the normative force and even the logical sense of concepts (such as 'natural', 'integrity', 'consciousness or rights of animals') that are contested. 6.7.3 Biotechnology: a novel development or a new way of realizing old goals? A case in point isprecisel y the concept of 'novel input'. Presumably, biotechnology consti tutes a novel input for society as a whole and for agriculture in particular. At the outset this statement seems sufficiently clear. For instance in the field of animal biotechnology a Dutch advisory committee to the government, in which all parties concerned were repre sented, stated in 1990 that three developments in animal biotechnology constitute a clear breach with existing practices. The committee indicated as such transgenesis (recombi- nant-DNA-technique), embryo-technology (cloning) and the use of substances and micro organisms, obtained by genetic modification (see note 1). However, on further interpreta tion even this statement may not go undisputed. Some people rightly ask what is the novelty of biotechnology from the point of view of the goals that have been pursued in livestock production for a long time already by all kinds of means such as stock-breeding, concentrates and computerized control of stable-conditions. It may seem that this dispute can only lead to petty discussions such as 'Is it new or not new?'. But there is more at stake for the ethical and political discussion in at least two respects. The first is that ajudgmen t on the novelty of a technique or a development sometimes carries important and far-reaching consequences. It isno tjus t a 'description'. In important ways it is a practical prescription ofho w to deal with it. For instance, in the practice of law, a substance or a technique that is considered to be 'new' is treated differently from other substances or techniques which are considered to be 'normal'. And the judgment of the above-mentioned committee that gene transfer constitutes a 'break in trend' with existing practices, has led to a stringent and extensive review procedure for proposals that make use of this technique. The second way in which a closer inspection of the question of novelty can help us to find our way in the ethical discussion is by sorting out the novel ethical questions that agricultural biotechnology involves. In the rest of this article we will follow this last road by describing some of the novel questions and reactions that agricultural biotechnology has generated in different contexts of ethical theory and argument. However, before doing that, I will first make some remarks on the use of the terms 'moral' and 'ethical'. This will not be done merely to avoid misunderstandings - for instance concerning the expectations ofth e contribution that ethics and ethicists can make towards solving moral problems. More important is that the discussion will enable us to construct a framework for the description ofth e ethical issues in biotechnology and to offer some starting-points for ethical thinking. 310 6.7.4 'Moral' and 'ethical' concerns: some problems of terminology There are several problems with the definition of terms like 'moral' and 'ethical', as they are often used in the debate about biotechnology. One of these problems is that the descriptive use of the term 'ethical', that is the way we use the term to identify or indicate ethical issues, is often confused with the normative or prescriptive use by which wejudge actions or practices. An example of prescriptive use is seen in expressions like 'genetic manipulation of animals is unethical'. This confusion stands in the way of a clear analysis of the problems. It also stands in the way of a rational discussion in a pluriform society where disagreement often results from different moral assumptions and principles. The confusion can lead to a general distrust of 'ethics', because the term 'ethics' will, wrongly, be identified with the position of particular parties in the debate, for instance animal protection agencies. This in turn creates a situation in which none of the parties involved is pressed to do what is the point of purpose of ethics as a reflective activity, namely to make explicit and defend on rational grounds one's own moral assumptions and beliefs. We all have moral beliefs and standards. The business of ethics is to test and assess their validity against facts, logic and principled argument. A second problem is that the term 'ethical' is exclusively associated with dilemmas and conflicts. As these dilemmas and conflicts result from standards and principles on which there is no substantial agreement no solution can be expected. We have just noted that moral disagreement is a reality. But there are also principles and standards, about which there is fairly general agreement. One example is the principle that scientists may not experiment on human beings without their consent, neither for the sake of science nor even if great benefits to mankind are expected. Another example is the principle that governments have a duty to promote the welfare of their citizens. The last example shows that it may well be that precisely those principles which are accepted by society, may become ethically problematic when considered from broader viewpoints like the future of our planet or justice on a global scale. In agricultural biotechnology these broader view points are certainly relevant, because of the global economic and environmental conse quences these developments may well have. 6.7.5 Subject and method of ethics These two problems - how to identify the ethical issues without taking some particular moral belief as the ethically right belief and how to account for the broader view that ethics demands - will guide us in presenting a definition of'ethical issues'. I think these problems can be solved by defining 'ethical issues' as those issues that 1) can be related in a meaningful way to the standards, values and principles that, implicitly or explicidy, are held by persons or groups in society as fundamental and 2) are claimed to be valid and obligatory for everybody. The first element of this definition will enable us to present an overview of the most important ethical issues that are brought forward in relation to agricultural biotechnology. It is important to notice that the second element - the claim of validity - points to at least one aspect of the solution of the second problem noted above - the problem how to realize 311 in ethics an encompassing view on the problems. Because, when a claim islai d that beliefs or standards are valid for everybody, there is a strong presumption that the claimant must be prepared to argue for this claim against other people — in principle all other persons -and that he is prepared to consider the relevance and rationality of his beliefs in the light of what others bring to the fore. It would lead us too far into the methodology of ethics to argue for this implication. But in reading the description of the ethical issues and the mentioning of some principles, the reader should bear in mind that this process of arguing and (re)considering is the most defensible and - in the literal sense - responsible way to deal with these issues. 6.7.6 Five problem-areas of the ethics of biotechnology On the basis of our definition I will identify a series of ethical issues that are prominent in the debate on biotechnology. The issues are related to principled and fundamental convic tions and considerations of different groups. I will group these issues under five headings: 1. issues concerning procedures of deliberation and decision-making; 2. issues concerning welfare and interests of human beings and animals; 3. issues concerning rights; 4. issues concerning nature and values of nature; 5. issues concerning ideals, resulting from religious and philosophical views. Of course these issues, let alone this sequence, do not follow logically from the definition. It is a limited and provisional list, based on the analysis of the present debate. In line with what has been stated earlier, I will point out the novel elements that biotechnology has introduced in the ethical discussions on agriculture. These novel elements will be illus trated by the case of transgenic animals. A similar exploration can be done for other areas in agriculture where biotechnological techniques are used, for instance molecular breeding of crops. In some cases, for example issues concerning the welfare, rights and interests of human beings, these elaborations are rather obvious. But the case of transgenic animals is the most actual and touches on all ethical dimensions. Problem-area 1. Procedureson deliberation and decision-making There are several ethical issues that relate to establishing a procedure for the control, assessment and review of genetic experiments and applications. In a democratic and pluralistic society one must consider the rights and interests of all kinds of groups. Whose rights and interests are affected? Who must be heard or has a right to take part in the decision-making process? How to handle the conflicts that will probably rise? These questions are central to democratic procedures. The answers to them can be seen as interpretations (or criticisms) ofth e liberal-democratic 'ethos'. This ethos iss o fundamental to our society that it is often not explicidy regarded as ethical, but it certainly is, if we consider it in the light of the definition of ethical given above. Two principles can be pointed to that can give a certain guidance in establishing procedures. The first is that of publicity. It connects to what has been said earlier about the method of ethics. The principle dictates that the facts, considerations and decisions regarding major develop ments in biotechnology (for instance gene-transfer or cloning of animals) ought to be 312 publicly available in order to make possible control, argument and discussion in society. The second principle could be called the principle of containability. This principle is of special relevance in any 'frontier' technology. The principle says that contested develop ments, if they are accepted, ought to be tried out first in a way and on a scale that is safe and controllable. One of the elements of this principle is that the consequences of the developments must be reversible, so that we may adjust our decisions in view of these consequences. This element has been emphasized by the Dutch Government Committee on Biotechnology and Animals (see note 2). The novel input agricultural biotechnology and especially genetic modification of farm- animals has brought about discussions on procedure (at least in the Netherlands where a structure for authorization of animal experiments already exists) on at least three points. The first is that individual proposals for genetic modification are not considered as something to be reviewed at the level of scientific institutions or corporate business (as in traditional animal experiments). They must be submitted for review to a national ethics committee with a rather broad mandate. This shows that the principle that animals are objects of moral concern (at least the higher vertebrate animals), istake n seriously in Dutch society and has gained political relevance. The second point is that genetic modification is not simply seen as a new stage in the development of livestock production, which has to be evaluated by its contribution to the internal values of livestock production, even if these values should include care for the health and welfare of animals. Biotechnology has led to an intensification of the discussion on the aims and presumptions of animal husbandry in general. Some critics demand explicitly that alternatives for biotechnology should be considered within a broader view on the future of livestock production and that ethical limits should be set to the ways the values of productivity and efficiency are realized. The third point is that public concern over genetic modification has led to further interest in the management of ethical issues in institutions and business. The existing procedures often cannot cope with these social and ethical issues: there is no separate address for them. The need to give attention to them, may lead in time to new structures and forms of interaction between institutions and companies on the one side and the general public on the other (see note 3). So one sees a movement away from existing organizations and groups following existing routines of internal regulation (in the end justifiable through the liberal-democratic 'ethos') to a broader discussion in which this ethos, is re-assessed for the new situation. Some actors are already anticipating the out come of this discussion and are trying to put new routines in its place, for instance by developing 'early warning' procedures for ethically sensitive developments or setting up advisory committees in which parties (or the views ofparties ) from outside are represented. Problema-area 2. Welfare and interests ofhuman beings and animals The ethical principles that can give a certain guidance in this area are those of beneficence ('Promote the well-being and interests of human beings and animals') and non-maleficence ('Avoid doing harm to (individual) human beings and animals'). But the most pressing ethical questions concern the balancing of interests (especially the interests of human beings versus the interests of animals) and finding criteria that can structure this weighing- 313 process. Here, we will leave aside the consequences of agricultural biotechnology for the welfare and interests of human beings (as consumers, farmers, workers or otherwise af fected) and concentrate on animal welfare. There is no question that there are many ethical issues in the case of human beings, but they are strongly interwoven with legal, economic and political questions. As far as animals are concerned, there is a set of rather new and difficult empirical questions about the actual effects on animal welfare of applying genetic techniques or biotechnological substances like Bovine Somatotropine. The difficulty that experts have in reaching agreement on the parameters by which to assess the health and well-being of animals, throws doubt on the trust society can have on the reliability of expert-judgment. There have also been early experiments with animals, as with pigs in Beltsville, Maryland, where health and welfare were severely affected through genetic modification. It seems to be an accepted standard, at least in the Netherlands, that experiments of which it can be predicted that they will lead to considerable welfare problems for the animals to be born, are morally unacceptable, no matter what the scientific value of these experiments may be. On a different level we are confronted with problems of a more conceptual and practi cal nature. In the case of genetic modification in the germ-line, it is not easy to be certain whether negative effects will not appear in the second generation where they did not appear in the first generation. Every transgenic animal must be considered as a unique 'founder' of a new genetic line. So, human responsibility stretches out over several genera tions of animals. Considering the fact that the moral responsibility towards future genera tions of human beings is one of the most hotly debated questions in applied ethics, the responsibility towards future animals is an even more complicated question to deal with. The most important 'novel input' of biotechnology however in the ethical discussion on the welfare and interests of animals has to do with the concept of'welfare of animals' and its implications. In modern times this concept has often been interpreted in a functional and 'experiential' way: what does matter ethically is how the actions of human beings affect the physiological functioning of the animal and its experiences, mainly negative experiences like pain and stress. But genetic modification may offer the potential to make animals feel well not as they 'are', but as they by human making, could be. So we could promote their welfare, not by creating the conditions that they need, given their needs and interests, but by changing their genetic make-up in such a way that they no longer have these needs and interests. The prospect is that it will become possible to create animals for livestock production that no longer have certain species-specific needs that prevents them from adjusting to the economic production-conditions. As a consequence, they will no longer suffer from the conditions in industrial livestock production which have developed in our society in order to meet the demand for (cheap) meat. However, for many people this prospect seems to clash with their moral intuition, especially in relation to our moral responsibility for the existing unfavourable position of animals. The crucial ethical ques tion iswhethe r animals may be adjusted to their actual situation - a situation that is heavily criticized on grounds of animal welfare — or that the situation should be adjusted to meet the needs of animals as they are - and as we increasingly come to know them in all their complexity from ethology and other sciences. This question directly connects to the objec tives for which the techniques of genetic modification may be used on animals. These 314 objectives can be distinguished in three categories. 1. Production-related objectives (for instance a fat-meat ratio in pigs that is more in accordance with the preferences of consumers). 2. Objectives related to the health and welfare of animals (for instance making cows more resistant to mastitis). 3. Objectives related to human health (for instance the production of pharmaceuticals in the milk of cows or sheep). On the principle that claims must bejustifiabl e to other persons (ideally all persons) to be ethically acceptable, it seems safe to say that thejustifyin g 'force' of the first category of objectives is lowest and that of the third highest, at least in the Netherlands and in other European countries. Problem area 3. Rights For the last twenty years the concept of 'rights' has been a bone of contention for propo nents and opponents of animal 'liberation', mainly in the United States and Great Britain. In the Netherlands it seems that the case for animal rights has been won by the defenders of more strict legal protection of animals, actually without being fought as heavily as in the Anglo-Saxon countries. In 1993 a law covering the protection of the health and well-being of animals will come into force. Although an important step, this law will not solve all moral problems. As a matter of fact, as we look closer at the content of this law, we will see that, for broad areas of human dealing with animals, it prescribes only a more stringent authorization procedure and provides sufficient consideration for human interest in mat ters of animal welfare. By the 'rights' that are meant under this heading, we mean moral rights, that is to say 'a right' is considered as a fundamental, 'deep', concept in moral argument, or as a strong claim that is justified on moral grounds (and not justified by referring to positive law). Rights in this moral sense have at least two characteristic features. First, they are used to 'trump' or 'overrule' the weighing-process of cost and benefits in so far as it is aimed at 'maximizing' results. In this respect they give a special sort of protection. Second, rights are described to individuals (or clearly identifiable groups of individuals) (see note 4). With regard to animal biotechnology what rights could be claimed for (or 'in the name of) animals and for (or by) human beings? As for animals we can think ofthre e things. The first has already been mentioned under the heading of welfare. It could be claimed that animals should have the right to be protected against harm to their health and welfare that could result from intervention in their genome. The second claim concerns the right to genetic integrity, that can logically be separated from the right not to be 'harmed'. These two rights involve individual (future) animals. A different and even more difficult ethical question has to do with the right of individual animal species to genetic integrity. Ifw e give some endangered species extensive protection, why not protect the integrity of the gene- pool which lies at the basis of the species? These claims, however, are far from clear, either conceptually or normatively. Respect for the integrity ofanimal s and species seems to have a strong intuitive appeal for many people, but as yet we have no theory to account for this appeal and no criterion to help us in deciding which interventions are acceptable on the basis of this principle of respect. Considering human rights, we can think of three kinds of rights that play a role in the discussion on biotechnology (see note 5). The first kind of right concerns the freedom of scientists to do research and the freedom of business to develop and invest in profitable 315 enterprise. These rights however have obvious limits in a range of ethical considerations and, at least for animal biotechnology, they do not bring us very far. The least we can ask is that the use that is made of this freedom is not in conflict with the moral standards of society. The second kind of right - often not explicitly brought forward, but more implied by proponents of biotechnology — is the claimed right of specific groups (or brought forward on behalf of those groups, for instance patient-groups or Third-World countries), that something 'ought' to be done about their situation. This is a much more problematic right, especially because these claims may welljeopardiz e the protective function of'rights' as we have defined them above. It may well be that this isth e point where the most difficult moral conflicts lie that are generated by biotechnology. The third sort of right, even more complicated, is the right of future generations to the (genetic) resources of the earth. Here the principle has been defended that a generation ought to leave to the next generation at least as many resources as it inherited. Problem-area 4.Nature andthe normative f one ofthe concept of 'the natural' Ideas about nature, the value of nature and what is 'natural' or 'unnatural' unquestionably belong to the core of the debate about biotechnology. The novel input biotechnology constituting for ethical discussions comes mainly under this heading. Of course, there is a long philosophical and ethical tradition, going back to Ancient Greece and Mediaeval Scholasticism, to which ideas of nature and natural are central. In this tradition of'essen- tialism' the given character of the genetic and biological constitution of the natural world and especially of the types of species were undisputed starting-points. With the acceptance of Darwin's theory of natural evolution however the belief that nature was divided once and for all into fixed species was abandoned. The fundamental importance of variety and flexibility instead of fixedness was emphasized from then on by biologists. But it was not before the possibility of changing the genetic make-up of organisms by human intervention and crossing the boundaries between species, that the consequences of this new view of nature had to be considered inpractical ethics . The paradox is that much argument against biotechnology still seems to thrive on old 'essentialist' presumptions. I will try to sketch some aspects of this debate injus t a few lines. In the eyes of many, biotechnology is the culminating-point of a process of mechaniza tion, reduction, objectivation and manipulation by science and technology of what was 'once' a spontaneous, self-regulated whole: nature. Unfortunately, there are three impor tant problems with the concept of 'nature' and 'natural' in ethics. The first is that it is notoriously ambiguous. So it is hard to discover what people mean when they say that something is 'unnatural'. Second, even if we can articulate the notion of'the natural', we cannot without further notice say it is good or right or conclude that we are not justified to interfere with it. Counter-examples to the reasoning from natural to good, let alone morally good, abound. For instance the striving of a virus to maintain itself in the human body, can be considered in most interpretations of 'natural' as a fairly natural process, but nobody would deny that we are entirely justified in trying to kill it and that doctors have a moral duty to so. The confusion of 'natural' and 'morally' good or right has given its name to the most famous fallacy in ethics: the naturalistic fallacy. The third problem is at what level should the normative force of the concept of the natural - ifi t has any force -be 316 applied. Is it at the level of the individual organism, ofth e species, of the ecosystem or even 'Mother Earth'. The intuitive appeal of the 'argument from the unnatural' may well come from the combination of three different strains of argument. Each one of which has different ramifications in the other dimensions that have already been mentioned. The first argu ment can be summarized as 'the unnatural is dangerous and harmful'. It starts from the idea that organisms and natural systems are balanced wholes that have evolved with time. Interference in some part of the whole can easily disturb this balance and have all kinds of unforeseen, unwanted and dangerous consequences, for ourselves as well as animals. This leads us back to the second problem-area of welfare and interests. The second argument can be summarized as 'the unnatural is the disrespectful'. It starts from the premise that natural systems and organisms have a value of their own, for instance on account of their spontaneity, autonomy, richness or complexity. Although only human beings can experience this value, this is no argument for saying that the value is ascribed to nature in an arbitrary way. We can give intelligible reasons for our valuing nature. To interfere with nature on this view can be seen in many cases as a wrong way of relating to nature, because it does not show respect for the valuable characteristics and possibilities of the natural system or organism. In its radical form, this appeal to respect would result in claims of right that belong to the third problem-area. The third argument is built on the premise that 'the unnatural is the artificial and man-made'. Of course, by this cannot be meant that what is man-made (for instance technology or art) has no value at all. Human practices have value in many respects. The normative force of this statement therefore comes only to the fore when we connect it with the idea that the natural is a fundamentally different dimension, in which we meet phe nomena that have their own purpose and value,jus t because they have not been produced or created by man. Nature is the 'otherness' that enriches human consciousness, experi ence and purpose in its own way. This source of purpose should be cherished by mankind and not reduced for the sake of more welfare. This approach of the concept of the natural points brings us to our fifth problem-area: that of ideals. Problem-area 5. Idealsresultingfiom religious andphihsophical views Even if the discussion on values of nature and the normative force of the concept of the natural, can be organised in a meaningful way, this will probably not, in the short run, lead to results that can be used in practice for assessment and regulation of biotechnological applications. It may even be expected that we will see a deepening of the conflict over the right way to treat animals, plants and ecosystems. Beliefs and arguments on these questions are often incidental to deep-seated beliefs of a religious, philosophical and moral nature. It is improbable that these beliefs will be abandoned, not even after long and honest discussion between people who respect each others viewpoint. There are also situations in which different people can recognize and even appreciate the values that are at issue, but where each comes to a different conclusion. This is a consequence of the priority that is given to certain goals or ideals (for instance knowledge for the scientist, animal welfare or integrity for the vegetarian or human health and welfare for the 'anthropocentrist'). So society has to develop a way to accommodate conflict over ideals so that the idea of 317 a 'moral community in dialogue' does not lose all its appeal. The strong feelings of the public, positive as well as negative, that biotechnology has evoked, indicate that it consti tutes a novel input in the discussion on the question of how to deal with ethical pluralism in a liberal-democratic society. The novelty does not only consist of the fact that these feelings are more intense or that biotechnology speaks more to the imagination. A strong case can be made that these reactions are more intense because biotechnology and genet ics touch on fundamental and crucial interests, values and experiences. The public should, therefore, have a say in these developments, in a way that is at least not obvious for other technologies. This seems to bring us back to the first problem-area, that of procedures. But the compromise that can be reached in this context, has its limits. It cannot solve the differ ences that hinge on deeper conflicts of a religious and philosophical nature. I also think that traditional principles of tolerance, respect, freedom and neutrality of the state, will no longer be sufficient, although they should not be given up of course. Biotechnology is one of the developments that forces us to reconsider the values and ideals we want to live by as a community, now that it can safely be said that as a system of legal and economic cooperation we have been successful. Our attitude towards animals the environment and nature may be one of the touchstones of this new ideal. Time will tell whether biotechnol ogy is a threat to, or a means of developing and realizing this ideal. 6.7.7 Conclusion As things stand, we cannot say in what way biotechnology will change the world or, for that matter, agriculture or livestock production in particular. I have indicated five prob lem-areas that are relevant to ethical discussion. These problems are not only bound up with the fundamental nature of genetics and the consequences genetic changes may bring, but also with the conceptions and values that we entertain about nature, animals and ourselves. It is clear that thinking about the possibilities, consequences and options of biotechnology while considering the interests, values and ideals that are at stake, is no luxury: it is an important duty if we want to live up to the idea of a human society that shapes itself in a reflective and conscious way. Notes 1. Report of the Advisory Committee on Ethics and Biotechnology with animals, Wagen ingen, May 1990, p. 10 2. Compare the descriptive use of'character' (for instance in 'He has an impulsive charac ter') with the normative use of the concept in ajudgmen t like 'What a character') 3. Compare A. Rip, Maatschappelijke aspecten van de landbouwbiotechnologie, Lezing gehouden te Ede, januari 1992. 4. For reasons of space, I do not consider here the important ethical question of the consequences of agricultural biotechnology for justice (on a national and global level) and the meaning of property-rights for this question. 5. This conception of rights owes much to the work of Robert Nozick (Anarchy, State and Utopia, New York, 1974) and Ronald Dworkin (Taking Rights seriously, London, 1977) 318 6.8 References Biotechnology and Dutch Industry, 1989. Report by Arthur D. Litde, Ministry of Economic Affairs, The Hague. Dosi, G., Chr. Freeman, R. Nelson, G. Silverberg & L. Soete (Eds.), 1988. Technical Change and Economic Theory. London: Pinter. Jacobs, D., 1990. The policy relevance of diffusion. Beleidsstudies Technologie Economie 8, Minis try of Economic Affairs, The Hague Jelsma,J . Constructive Technology Assessment In Action: The Case of Biotechnology. Proceedings Twente Workshop on Constructive Technology Assessment, Sept, 20-22, 1991. Forthcoming in the proceedings, Volume edited by A. Rip,J . Schot & T. Misa. In press. Rip A., 1992. Science and Technology As Dancing Partners. In: P. Kroes & M. Bakker (Eds.): Technological development and science in the industrial age. Dordrecht: Kluwer Academic, pp. 231-270. Stasse-Wolthuis, M. & F.M. Rombouts (Eds), 1992. Sleutelen aan ons voedsel; wat kan biotechnolo gie? Bohn Stafleu Van Loghum, Houten/Antwerpen Verpoorte, R., 1991. Producten uit plantecellen. In: W.G.J. Brouwer (Ed.): Plantaardige grondstof fen voor de industrie. Samson, Alphen aan de Rijn, Netherlands 319 The authors W.IJ. Aalbersberg was director general of NIZO (Netherlands Institute of Dairy Research) at, the Netherlands from 1973 to 1992. He studied chemistry and physics at the Free University at Amsterdam, where he obtained a doctorate in chemistry in 1960. Form 1954 to 1973 he held several research and management positions in industry. He was a member of the Programme Committee on Agricultural Biotechnology. Address: p/a NIZO, P.O. Box 20, 6710 BA Ede. J. Bakker is associate professor at the Department of Nematology. He obtained his doctorate in plant pathology from Wageningen Agricultural University. His special inter ests are engineering (pl)antibodies and elucidating the molecular basis of virulence in cyst nematodes. Address: Department of Nematology, Wageningen Agricultural University, P.O. Box 8123, 6700 ES Wageningen. G. Beldman is a lecturer in food enzymology in the Department of Food Science at the Wageningen Agricultural University. He studied biochemistry at the University of Gronin gen. He obtained his doctorate at Wageningen Agricultural University in 1986. From 1985-1990 he was head of the Biochemistry Department of the Netherlands Institute for Carbohydrate Research (NIKO-TNO). His major interest is the enzymology of plant cell wall polysaccharides. Address: Department of Food Science, Agricultural University, Bomenweg 2, 6703 HD Wageningen. J.E.N. Bergmans studied biology at the University of Utrecht, where he also obtained his doctorate. His research interests included microbial genetics, most recently the genetics ofvirulenc e factors of Gram negative bacteria. In 1990 he left active science for a function as secretary of the Provisional Committee on Genetic Modification, an independent advi sory committee for the Dutch Ministry of the Environment. Address: VCOGEM, P.O. Box 80.022, 3508 TA Utrecht. J.H. van Boom isprofesso r ofbio-organi c chemistry at Leyden University. He graduated as a chemist at the University of Utrecht where he obtained his doctorate in 1968. His special interest is the synthesis of carbohydrates and nucleic acids. Address: Gorlaeus Laboratory,: Department of Bio-organic Chemistry, P.O. Box 9502, 2300 RA Leiden. G. Boswinkel has studied chemistry at the College ofAdvance d Technology in Hengelo. After completing his study he was employed by Fokker B.V. for two years. From 1987 to 1992 he has worked on specific hydrolysis of fatty acids in the Food and Bioprocess Engineering Group at Wageningen Agricultural University and in the Agrotechnological Research Institute ATO-DLO. He is currently working for ATO-DLO in the Oils and Fatty Acids Technology group. Address: ATO-DLO, P.O. Box 17, 6700 AA Wageningen. 321 B. van den Burg studiedbiologica l sciencesa tth eUniversit y ofGroningen . He obtained his doctorate at the same university in 1991 on research in thefield o fprotei n engineering of Bacillus proteases. In 1991 he started as a researcher in the Department of Molecular Biologya t the CentralVeterinar y Institute.Hi sspecia lare a ofinteres ti s structure-function analysis of proteins. Address: Central Veterinary Institute, P.O. Box 365, 8200 AJ Lely stad. B.J.C. Cornelissen isprofesso r ofplan t pathology at the University ofAmsterdam . He studied at the University of Nijmegen and obtained his doctorate at the University of Leyden. From 1986 until the summer of 1992 he headed the crop protection group of MOGEN Int. nv., a Leiden based plant biotechnology company. His special interest isth e molecular basis ofplant-pathoge n interactions.Address :Plan t Pathology Section, Univer sity ofAmsterdam , Kruislaan 318, 1098 SM Amsterdam. R.P.M.A. Crooymans isa research assistant in molecular biology in the Department of Animal Breeding at Wageningen Agricultural University. He completed his studies as a biochemical technical assistant in 1986. He is involved in genome mapping in chickens. Address:: Department Animal Breeding, Wageningen Agricultural University, P.O. Box 338, 6700 AH Wageningen. F.P. Cuperus is group leader of the Oils and Fatty Acids Technology group at the Agrotechnological Research InstituteATO-DL O and iscurrentl y involved in the develop ment of new separation and reactor technologies for the processing of renewable re sources. He has obtained his doctorate on membrane technology from the University of Twente. Address: ATO-DLO, P.O. Box 17, 6700 AA Wageningen. C.J. Davies is an immunologist who isworkin g as a visiting scientist at the Department of Animal Breeding of Wageningen Agricultural University. His special interest is the immunogenetics ofth e immune response ofcattle .Address :Departmen t ofAnima l Breed ing, Wageningen Agricultural University, P.O. Box 338, 6700 AH Wageningen. J.J. Dekkers is currently working at the Directorate of Science and Technology of the Dutch Ministry ofAgriculture , Nature Management and Fisheries where he isengage d in agricultural biotechnology policy. He studied physical chemistry at the Free University in Amsterdam, where he also obtained his doctorate. His special interest isth e development ofagricultura l biotechnology and itsimplicatio n for society.Address :Ministr y ofAgricul ture, Nature Management and Fisheries, Directorate of Science and Technology, P.O. Box 20401, 2500 EK Den Haag. J.T.P. Derksen is Head of the Division of Industrial Crops, Protein and Fatty Acids Technology of the Agrotechnological Research Institute ATO-DLO in Wageningen. He studied biology at the University of Groningen, where he also obtained a doctorate in biochemistry. From 1988 to 1990 he worked as a research group leader for a biomedical company in Pasadena, USA. In 1990 hejoine d the Agrotechnological Research Institute 322 ATO-DLO. His special interest is the development of new products and process technolo gies based on renewable resources. Address: ATO-DLO, P.O. Box 17, 6700 AA Wagenin gen. R.J.M. Dijkhof is a research assistant in molecular biology at the Department of Animal Breeding of the Wageningen Agricultural University. She completed her studies as a technical assistant in analytical chemistry in 1982.Address : Department Animal Breeding, Wageningen Agricultural University, P.O. Box 338, 6700 AH Wageningen. J.J.M. Dons is head of the Department of Developmental Biology at the Centre for Plant Breeding and Reproduction Research (CPRO-DLO) in Wageningen. He studied biology and obtained his doctorate in biochemistry at Leyden University. His main interest is the molecular regulation and modification of plant developmental processes: regeneration, hormonal regulation and flower development. Address: Centre for Plant Breeding and Reproduction Research CPRO-DLO, P.O. Box 16, 6700 AA Wageningen. E. Egberts graduated in biology in 1971, and obtained a doctorate in biochemistry in 1977, from the University of Groningen. From 1971 to 1977, he was a research associate in cell biology at the Max Planck Institute, Wilhelmshaven, Germany. Since 1977, he has been working at Wageningen Agricultural University, recently as a senior lecturer. His current interests is the molecular and cellular characterization of the immune system of fish, with special emphasis on the major histocompatibility system. Address: Department of Experimental Animal Morphology and Cell Biology, Agricultural University, Zodiac Building, P.O. Box 338, 6700 AH Wageningen. F.A.C. van Engelenburg studied molecular sciences at the Agricultural University of Wageningen where he graduated in 1988. He then did research on the induction of glucose-oxidase in Aspergillus niger at the same university. In August 1989 he started as a researcher at the Central Veterinary Institute. His special interest is molecular pathogene sis of viral infections. Address: Central Veterinary Institute, P.O. Box 65, 8200 AB, Lelystad. J.H.F. Erkens was a research assistant at the DLO-Research Institute for Animal Pro duction 'Schoonoord' (IVO-DLO) Department of Reproduction for 14 years, where he was involved in developing radioimmunoassays for protein hormones. In 1989 he started work as a molecular biology research assistant at the IVO-DLO Department of Animal Breeding. Address: DLO Research Institute for Animal Production 'Schoonoord' (IVO- DLO), P.O. Box 501, 3700 AM Zeist. S.H.M, van Erp graduated in molecular sciences from Wageningen Agricultural Uni versity in 1991, and has since been working as a research student on a project involving the identification and characterization of the genes that belong to the major histocompatibility complex of carp. Address: Department of Experimental Animal Morphology and Cell Biology, Agricultural University, Zodiac Building, P.O. Box 338, 6700 AH Wageningen. 323 W.J.M, van Gelder was until recently head of the Division of Industrial Crops, Products and Process Technologies of the Agrotechnological Research Institute ATO-DLO. He is currently director of Technology Policy for the Federation of Netherlands Industries (VNO) in The Hague. He obtained a doctorate in phytochemistry from Wageningen Agricultural University. His special interest is the effect of technological innovations on economic growth. Address: VNO, P.O. Box 93093, 2509 AB The Hague. A.L.J. Gielkens is head of the Department of Molecular Biology of the Central Veteri nary Institute, Lelystad. He studied biochemistry and completed his doctorate on retro viruses in 1976 at the University of Nijmegen. He has longstanding experience in molecu lar virology and vaccine development of herpesviruses. Address: Central Veterinary Insti tute, P.O. Box 365, 8200 AJ Lelystad. A.L.J. Gielkens is head of the Department of Molecular Biology at the DLO Central Veterinary Institute CDI-DLO in Lelystad. He studied molecular biology at the University of Nijmegen, where he obtained his doctorate in 1976. His special interests are molecular virology and bacteriology. Address: Central Veterinary Institute, Department of Virology, P.O. Box 365, 8200 AJ Lelystad. K. Glazenburg isa research scientist in the Department ofMolecula r Biology at the CDI Central Veterinary Institute CDI-DLO in Lelystad. He studied biology at the University of Utrecht, and graduated in 1986. His special interest is animal herpes virus infections. Address: Central Veterinary Institute, Department of Virology, P.O. Box 365, 8200 AJ Lelystad. F.J. Gommers is associate professor at the Department ofNematology . He studied plant pathology in Wageningen and obtained a doctorate in biochemistry at the University of Groningen. He is mainly involved in research on host/parasite interactions and the genet ics of plant parasitic nematodes. Address: Department of Nematology, Wageningen Agri cultural University, P.O. Box 8123, 6700 ES Wageningen. L.H. de Graaffis a molecular geneticist working in the Molecular Genetics Section of the Department of Genetics at Wageningen Agricultural University where he obtained his doctorate in 1989. His special interest is the molecular genetics of gene regulation in filamentous fungi. Address: Department of Genetics, Wageningen Agricultural University, Dreijenlaan 2, 6703 HA Wageningen. M.A.M. Groenen is a molecular biologist and iscurrentl y working in the Department of Animal Breeding at Wageningen Agricultural University. He studied chemistry at the University of Nijmegen and obtained his doctorate at the University of Leyden in 1987. His special interest isth e regulation ofgen e expression in eukaryotes and genome mapping in farm animals. Address: Department Animal Breeding, Wageningen Agricultural Uni versity, P.O. Box 338, 6700 AH Wageningen. 324 R.J. Hamer is head of the Biochemistry and Physical Chemistry Department of the TNO-Biotechnology and Chemistry Institute. He obtained his doctorate from the Univer sity of Utrecht in 1981. His special interests are cereal protein/enzyme chemistry and immunochemistry. Address: TNO Biotechnology and Chemistry Institute, Utrechtseweg 48, 3700 AJ Zeist. J. Hille is a molecular geneticist and head of the plant disease resistance group of the Department of Genetics at Free University Amsterdam. He studied chemistry at the University of Leyden where he obtained his doctorate in 1983. His special interests are plant/pathogen interactions and the biology of transposable elements. Address : Depart ment of Genetics, Free University Amsterdam, De Boelelaan 1087, 1081 HV Amsterdam. H. Hofstra is head of the Molecular Biology Section in the Department of Microbiology at the TNO Institute for Biotechnology and Chemistry. He studied biology at the Univer sity of Groningen and obtained his doctorate from there in 1981. He is interested in the development and implementation of detection based on DNA hybridization and amplifi cation or based on immunological principles. Address: TNO Nutrition and Food Re search, Institute for Biotechnology and Chemistry, Utrechtseweg 48, 3700 AJ Zeist. R.J. Hogendorp has been is working at the Centre for Agro-Biological Research (CABO-DLO) since 1992. He studied biology at the University of Groningen and joined the Research Institute for Livestock Feeding and Nutrition (IWO-DLO) in 1990 where he was involved in research on phenolic compounds in plant tissues and their relevance to digestion. His special interest is the fermentation of proteins by ruminants. Address: CABO-DLO, P.O. Box 14, 6700 AA Wageningen M.C. Horainek isprofesso r ofvirolog y and virus disease at the Veterinary Faculty of the University of Utrecht and director of the Institute of Veterinary Research at the same University. He graduated from the Veterinary School in Hanover, Germany, where he obtained his doctorate in virology in 1970. His special interest is enveloped viruses with positive-stranded RNA genomes, as well as feline virus infections. Address: University of Utrecht, Department of Infectious Diseases &Immunolog y - Virology Division, P.O. Box 80.165, 3508 TD Utrecht. J. Hugenholtz is a project leader in the Department of Microbiology at the Netherlands Institute for Dairy Research (NIZO) in Ede. He obtained his doctorate in microbiology in 1986 at the University of Groningen. His special interest is physiology and metabolic engineering of lactic acid bacteria and their practical application. Address: Netherlands Institute for Dairy Research (NIZO) P.O. Box 20, 6710 BA Ede. J. Huisman is head of the department of Digestion Physiology in Husbandry at the Institute of Animal Nutrition and Physiology ILOB-TNO. He studied animal feeding at Wageningen Agricultural University and obtained his doctorate in 1991 with a the thesis on 'Anti-Nutritional effects of legume seeds in piglets, rats and poultry'. His special area of 325 interest is the anti-nutritional effects of certain compounds in feed. Address: Institute of Animal Nutrition and Physiology ILOB-TNO, P.O. Box 15, 6700 AA Wageningen. T.H.J.M. Hutten studied biochemistry at the University of Nijmegen and agricultural economics at Wageningen Agricultural University. He obtained his doctorate in biochem istry at the University of Nijmegen. He was deputy director of the National Cooperative Council for Agriculture and Horticulture of the Netherlands. In 1993 he joined the Research Management staff of the DLO-NL institutes. His special interests are technolog ical policies, R&D strategies and development of organizational structures at both micro and meso levels. Address: DLO-NL, P.O. Box 59, 6700 AB Wageningen. A.E.M. Janssen is currently working in the Department of Food Science, Food and Bioprocess Engineering Group, at Wageningen Agricultural University where she pres- endy is preparing a doctorate thesis. Her special interest is biocatalysis and biotransforma tions. Address: Wageningen Agricultural University, Department of Food Science, Food and Bioprocess Engineering Group, P.O. Box 8129, 6700 EV Wageningen. J. de Jong is a research assistant at the Institute of Animal Nutrition and Physiology ILOB-TNO. He graduated from agricultural college. His main interest are research activities on animal nutrition. Address: Institute of Animal Nutrition and Physiology ILOB-TNO, P.O. Box 15, 6700 AA Wageningen. M.A. Jongsma is a molecular biologist at the DLO Centre for Plant Breeding and Reproduction Research (CPRO-DLO). He iscurrentl y preparing a doctorate thesis on the physiology of stress response by plants, molecular biological aspects of the proteins in volved and digestion physiology of (crop-pest) insects. Address: CPRO-DLO, PO Box 16, 6700 AA Wageningen. M.J. Kaashoek studied veterinary medicine at Utrecht University, where she graduated in 1989. She was then active as a practitioner, specializing in large animals, before starting as a researcher at the Department of Bovine Virology in August 1990. Her main interest is pathogenesis and vaccinology of virus diseases in ruminants. Address: Central Veteri nary Institute, P.O. Box 65, 8200 AB, Lelystad. A.J.A. van Kampen is a research assistant in molecular biology in the Department of Animal Breeding at Wageningen Agricultural University. He completed his studies as a biotechnology assistant in 1990. He is involved in genome mapping in chickens. Address: Department Animal Breeding, Wageningen Agricultural University, P.O. Box 338, 6700 AH Wageningen. H.J. Kamphuis is a food technologist and head of the application laboratory of Gerkens Cacao Industry in Wormer. He studied food technology at Wageningen Agricultural University and obtained his doctorate in food microbiology in 1992. Address: Gerkens Cacao Industry, P.O. Box 82, 1530 AB Wormer. 326 J.A.L. van Kan is associate professor in the Department of Phytopathology at Wagenin gen Agricultural University. He studied biochemistry at Leyden University and obtained his doctorate in 1988 on virus-induced plant defence genes. His special interest is the molecular genetics of plant pathogenic fungi. Address: Wageningen Agricultural Univer sity, Department of Pathology, P.O. Box 8025, 6700 EE Wageningen. G.J.M van Kempen is head of the Department of Animal Nutrition and Physiology of the TNO Institute of Toxicology and Nutrition (ILOB-TNO). He studied husbandry at Wageningen Agricultural University where he obtained his doctorate in 1971 on anaemia in pigs. His special interest is phytopharmacology and toxicology. Address: Institute of Animal Nutrition and Physiology ILOB-TNO, P.O. Box 15, 6700 AA Wageningen. T.G. Kimman is a veterinary virologist and head of the Aujeszky's disease laboratory at the DLO Central Veterinary Institute CDI-DLO in Lelystad. He studied veterinary med icine at the University of Utrecht, where he obtained his doctorate in 1989. His special interests are the immunology and pathology of virus infections. Address: Central Veteri nary Institute, Department of Virology, P.O. Box 365, 8200 AJ Lelystad. N. Klijn is a researcher in the Department of Microbiology at the Netherlands Institute for Dairy Research (NIZO). She graduated as a molecular microbiologist from Wagenin gen Agricultural University in 1990. Her special interest is ecophysiology and molecular biology of lactic acid bacteria. Address: NIZO, P.O. Box 20, 6710 BA Ede. L. Kruijt is a research assistant and is currently working in the Immunobiology Group of the Reproduction Department at the Research Institute for Animal Production (IVO- DLO) 'Schoonoord'. His special interests are immunoassays and monoclonal antibody mediated irnmunomodulation. Address: DLO Research Institute for Animal Production (IVO-DLO) 'Schoonoord', P.O. Box 501, 3700 AM Zeist. O.P. Kuipers is a project leader in the Department of Biophysical Chemistry at the Netherlands Institute for Dairy Research (NIZO) in Ede. He studied biology at the Uni versity of Utrecht and obtained his doctorate in biochemistry at the same university in 1990. His special interests include bacteriocins, protein engineering and biotechnology of lactic acid bacteria. Address: Netherlands Institute for Dairy Research (NIZO) P.O. Box 20, 6710 BA Ede. M.A. Kusters-van Someren is currently working in the Molecular Genetics Section of the Department of Genetics at the Wageningen Agricultural University. She studied Biology at the University of Utrecht and obtained her doctorate in 1991. Her special interest is the molecular genetics of plant cell wall degradation. Address: Department of Genetics, Wageningen Agricultural University, Dreijenlaan 2, 6703 HA Wageningen. 327 P. van Leeuwen is a research assistant at the Institute of Animal Nutrition and Physiol ogy ILOB-TNO. He graduated from agricultural college. He is interested in research activities on animal nutrition. Address: Institute of Animal Nutrition and Physiology ILOB-TNO, P.O. Box 15, 6700 AA Wageningen. T. van der Lende is head of the Immunobiology Group of the Reproduction Depart ment at the Research Institute for Animal Production (IVO-DLO) 'Schoonoord'. He studied animal science at Wageningen Agricultural University, where he obtained his doctorate in 1989. His special interests are monoclonal antibody mediated immunomodu- lation and reproduction technology. Address: DLO Research Institute for Animal Produc tion (IVO-DLO) 'Schoonoord', P.O. Box 501, 3700 AM Zeist. J.N.M. Mol is professor of molecular plant genetics at Free University Amsterdam. He studied chemistry at the University of Amsterdam and specialized in biochemistry and molecular biology. Since 1980 his main interest has been flower development and in particular the molecular biology of floral pigmentation. Address: Department of Genetics, Free University, De Boelelaan 1087, 1081 HV Amsterdam. W.A.M. Mulder is a pathologist at the DLO-Central Veterinary Institute CDI-DLO in Lelystad. He graduated from Wageningen Agricultural University in 1990. His special interest is the pathogenesis of animal herpes virus infections. Address: Central Veterinary Institute, Department of Virology, P.O. Box 365, 8200 AJ Lelystad. Ph.R. Nilsson is a post graduate student in the Department of Animal Husbandry at Wageningen Agricultural University. She studied medical biology at the University of Utrecht, and is currently doing research on the definition of polymorphism of the bovine Major Histocompatibility Complex. Address: Department of Animal Husbandry, Wage ningen Agricultural University, P.O. Box 338, 6700 AH Wageningen. S.H.W. Notermans is a food microbiologist working at the National Institute of Public Health and Environmental Protection (RIVM) in Bilthoven. He obtained his doctorate in food microbiology in 1975 from Wageningen Agricultural University. His special interests are food microbiology, pathogenic organisms and safe food products. Address: National Institute of Public Health and Environmental Protection, PO Box 1, 3720 BA Bilthoven. J.T. van Oirschot is head of the Department of Bovine Virology and professor of veterinary vaccinology at Utrecht University. He studied veterinary medicine and ob tained his doctorate at Utrecht University in 1980. His special interest is veterinary vac cinology and viral infections of ruminants and pigs. Address: Central Veterinary Institute, P.O. Box 65, 8200 AB, Lelystad. 328 J.L. van Os obtained his doctorate in veterinary sciences at the University ofUtrecht . He worked for the Gist-brocades company for 22 years. Before he retired he was responsible for developing and marketing novel feed ingredients, such as the phytase enzyme. He is currently a consultant for Corporate Development at Gist-brocades. Address: Gist-bro cades, P.O. Box 1, 2600 MA Delft. A. van der Padt is currently working in the Department of Food Science, Food and Bioprocess Engineering Group, at Wageningen Agricultural University. His special inter ests are bioconversion and separation processes. Address: Wageningen Agricultural Uni versity,: Department of Food Science, Food and Bioprocess Engineering Group, P.O. Box 8129, 6700 EV Wageningen. M.F.W, te Pas is a molecular biologist at the DLO-Research Institute for Animal Production 'Schoonoord' IVO-DLO, Department of Animal Breeding. He studied biol ogy at Leyden University where he obtained his doctorate in 1991. His special interest is the development of molecular genetic markers for breeding in meat-producing animals. Address: IVO-DLO, P.O. Box 501, 3700 AM Zeist. H.C. van der Plas isRecto r Magnificus of Wageningen Agricultural University and was formerly chairman of the Programme Committee on Agricultural Biotechnology. He is professor of Organic Chemistry at the same university. His special interest is organic chemistry and the development of agricultural biotechnology. Address: Wageningen Agri cultural University, P.O. Box 9101, 6700 HB Wageningen. J.J. van der Poel is a molecular biologist in the Department of Animal Breeding at Wageningem Agricultural University. He studied chemistry at the University of Leyden where he also obtained his doctorate in medicine in 1985. His special interest is the genetics of the immune response and genome mapping in farm animals. Address: Depart ment of Animal Breeding, Wageningen Agricultural University, P.O. Box 338, 6700 AH Wageningen. J.M.A. Pol is a pathologist at the DLO Central Veterinary Institute CDI-DLO in Lely stad. He studied veterinary science at the University of Utrecht, where he obtained his doctorate in 1990. His special interests are the pathogenesis of virus diseases and the role of cytokines in virus infections. Address: Central Veterinary Institute, Department of Virology, P.O. Box 365, 8200 AJ Lelystad. K. van't Riet is professor of Food and Bioprocess Engineering in the Department of Food Science at Wageningen Agricultural University. His special interests are bioconver sion, separation and drying processes and information sciences. Address: Wageningen Agricultural University: Department of Food Science, Food and Bioprocess Engineering Group, P.O. Box 8129, 6700 EV Wageningen. 329 F.A.M. Rijsewijk is head of the laboratory of molecular virology in the Department of Bovine Virology at the Central Veterinary Institute CDI-DLO. He studied biology and completed his doctorate at the University of Amsterdam on the thesis 'The mouse onco gene Wnt-\ and its Drosophila homolog wingless1 in 1992. He is especially interested in molecular aspects of cytological and virological processes. Address: Central Veterinary Institute, P.O. Box 65, 8200 AB, Lelystad. A. Rip studied chemistry and philosophy at Leyden University. He obtained his doctorate in 1981 on the thesis 'Social responsibility of Chemists'. He became associate professor in the Dynamics of Science at the University ofAmsterda m in 1984. Since 1987 he has been professor of Philosophy of Science and Technology at the University of Twente. His research interests are dynamics of science, technology assessment, dynamics of technology and social perception of science and technology. Address: University of Twente, School of Philosophy and Social Sciences, P.O. Box 217, 7500 AE Enschede H.S. Rollema is a senior research scientist in the Department of Biophysical Chemistry at the Netherlands Institute for Dairy Research (NIZO) in Ede. He studied physical chemistry at Free University of Amsterdam and obtained his doctorate at the University of Nijmegen in 1976. His main research interest is the relation between structure and functional properties of biopolymers. Address: Netherlands Institute for Dairy Research (NIZO) P.O. Box 20, 6710 BA Ede. F.M. Rombouts is professor of food microbiology and food hygiene at the Department of Food Science of Wageningen Agricultural University. He obtained his doctorate at 1972 at Wageningen Agricultural University. He is presently involved in research related to food preservation, food pathogens and fermented foods. Address: Wageningen Agricul tural University: Department Food Science, PO Box 8129, 6700 EV Wageningen. G.A. de Ruiter graduated as a food technologist in 1988 at Wageningen Agricultural University where he is currently working on a doctorate thesis on immunoassays for the detection of moulds in food at the Department of Food Microbiology. Address: Wagenin gen Agricultural University: Department Food Science, PO Box 8129, 6700 EV Wagenin gen. D. Ruyter is currently working as a molecular biologist in the Department of Animal Breeding at Wageningen Agricultural University, where he studied animal breeding. He is involved in the PIGMAP project, a European initiative to generate a genetic and physical map of the pig genome. Address: Department Animal Breeding, Wageningen Agricultural University, P.O. Box 338, 6700 AH Wageningen. F.H.M.G. Savelkoul is presendy chemistry teacher at the Ds. Pierson College in s'Her- togenbosch. He studied cell biology at the Wageningen Agricultural University and is preparing his doctorate thesis on enzymatic inactivation of antinutritional factors, based on his own research at the Department of Animal Nutrition of the same university. His 330 special interest is the transfer of scientific knowledge to students. Address: Ds. Pierson College, Geraert ter Borghstraat 1, 5212 CZ s'Hertogenbosch. A. van der Schans ishea d of the Embryo section ofth e Reproduction Department at the DLO Research Institute for Animal Production 'Schoonoord' IVO-DLO in Zeist. He studied veterinary science at the University of Utrecht. His special interest lies in the field of reproduction technologies for cows, including welfare and health aspects. Address: DLO Research Institute for Animal Production (IVO-DLO) 'Schoonoord', P.O. Box 501, 3700 AM Zeist D.G. Schmidt is a physical chemist in the Department of Biophysical Chemistry at the Netherlands Institute for Dairy Research (NIZO) in Ede. He studied physical chemistry and chemical physics from the University of Amsterdam and obtained his doctorate in 1969 at the University of Utrecht. His present interest is the proteolysis of whey proteins. Address: Netherlands Institute for Dairy Research (NIZO), P.O. Box 20, 6710 BA Ede. A. Schots is head of the Laboratory for Monoclonal Antibodies in Wageningen. He graduated in plant pathology and obtained his doctorate from Wageningen Agricultural University. His special interests are immunology, engineering monoclonal antibody genes and serological assays to diagnose pathogens. Address: Laboratory for Monoclonal Anti bodies, P.O. Box 9060, 6700 AA Wageningen. B. Schutte is a researcher at the Institute of Animal Nutrition and Physiology ILOB- TNO. He studied at the Agricultural College in Wageningen. He obtained his doctorate from Wageningen Agricultural University in 1991 with a thesis entitled 'Nutritional value and physiological effects of D-xylose and L-arabinose in poultry and pigs'. His special interests are amino acid utilization in monogastric animals and the utilization of mono mers released after enzymatic treatment ofnon-starc h polysaccharides in pigs and poultry. Address: Institute ofAnima l Nutrition and Physiology ILOB-TNO, P.O. Box 15, 6700 AA Wageningen. R.J. Siezen is head of the Department of Biophysical Chemistry at the Netherlands Institute for Dairy Research (NIZO) in Ede. He studied biochemistry at the University of Groningen and obtained his doctorate there in 1973. His main interests are protein structures and protein engineering. Address: Netherlands Institute for Dairy Research (NIZO) P.O. Box 20, 6710 BA Ede. C.P. Spira is a post-graduate student working in the Department at Animal Breeding of the Wageningen Agricultural University where she studied animal nutrition. She is doing research on the regulation of expression of the bovine casein genes. Address: Department Animal Breeding, Wageningen Agricultural University, P.O. Box 338, 6700 AH Wage ningen. 331 S.F. Spoelstra is head of the department of Physiology and Biochemistry of the DLO Institute for Livestock Feeding and Nutrition Research (IWO-DLO) in Lelystad. He obtained his doctorate from Wageningen Agricultural University in 1978. His special interest is the use of enzymes in forage conservation. Address: Institute for Livestock Feeding and Nutrition Research (IWO-DLO), Runderweg 2, P.O. Box 160, 8200 AD Lelystad. P. Stam is head of the Population Biology Department at the DLO Centre for Plant Breeding and Reproduction Research (CPRO-DLO) and is senior lecturer in the Depart ment of Genetics at Wageningen Agricultural University. He was trained in plant breeding and genetics and obtained his doctorate in 1972 with a thesis on linkage drag. His interests are population and quantitative genetics and its application in plant breeding. Address: Centre for Plant Breeding and Reproduction Research, P.O. Box 16, 6700 AA Wagenin gen. T.A.W.M.Saat started working for the Van der Have plant breeding company after gaining his biology degree from the University of Groningen. He is responsible for genetic modification of cruciferous crops and vegetables. He also coordinates the company's field trials, both in the Netherlands and elsewhere, and isinvolve d with informing the public on biotechnology in plant breeding. Address: Van der Have Research, P.O. Box 1, 4410 AA Rilland. R.J.M. Stet obtained a degree in biology from the University of Amsterdam in 1980. In 1987, he obtained a doctorate from the University of Groningen for his work on graft- versus-host reactions in the rat. After a short period as a visiting scientist at the Marine Laboratory, Aberdeen, Scotland, where he worked on the immunogenetic typing of wild salmon stocks, he joined Wageningen Agricultural University in 1988 as a lecturer. His major interest is the immunogenetics of fish, particularly the major histocompatibility system. Address: Department of Experimental Animal Morphology and Cell Biology, Agricultural University, Zodiac Building, P.O. Box 338, 6700 AH Wageningen. W.J. Stiekema is head of the Department of Molecular Biology at the Centre for Plant Breeding and Reproduction Research (CPRO-DLO) in Wageningen. He studied Biology and obtained his doctorate at Free University of Amsterdam. His main interest is molecu lar breeding for pest and disease resistance in crop plants. Address: Centre for Plant Breeding and Reproduction Research CPRO-DLO, P.O. Box 16, 6700 AA Wageningen. S. Tamminga studied animal nutrition and agricultural biochemistry at Wageningen Agricultural University and obtained a post-graduate degree from the University of Newcasde (UK). He is now professor of Ruminant Nutrition at Wageningen Agricultural University and head of the Department of Animal Nutrition. His special interests are protein metabolism and rumen fermentation in ruminants and the application of enzyme technology in animal nutrition. Address: Department of Animal Nutrition, Agricultural University, Haagsteeg 4, 6708 PM Wageningen. 332 M. Tieman is a research assistant currently working in the Immunobiology Group of the Reproduction Department at the Research Institute for Animal Production (IVO-DLO) 'Schoonoord'. Since 1982 she has been working on hybridoma technology. Address: DLO Research Institute for Animal Production (IVO-DLO) 'Schoonoord', P.O. Box 501, 3700 AM Zeist. E. Veitkamp is corporate head of Research & Development at Sandoz Seeds, Basel, Switzerland and he is also associate professor of plant genetics at Free University Amster dam, where he studied biology before obtaining a doctorate in 1976. Address: Sandoz Seeds Ltd., CH-4002 Basel, Switzerland; contact address in the Netherlands: P.O. Box 26, 1600AAEnkhuizen. E.J.M. Verstege is a research assistant in Molecular Biology in the Department of Animal Breeding at Wageningen Agricultural University. She completed her study as a biochemical technical assistant in 1988. Address: Department Animal Breeding, Wagenin gen Agricultural University, P.O. Box 338, 6700 AH Wageningen. M.W.A. Verstegen is professor of animal feeding at Wageningen Agricultural Univer sity. He studied animal husbandry at the same university and obtained his doctorate in 1971 on the thesis: 'The influence of ambient temperatures on the energy metabolism of growing pigs'. His special interest is in nutrition physiology of non-ruminants. Address: Wageningen Agricultural University, Department ofAnima l Nutrition, Haagsteeg 4, 6708 PM Wageningen L. Visser studied Molecular Sciences at Wageningen Agricultural University and ob tained his doctorate on human molecular virology at the University of Utrecht. He was, until recently, head of the Gene Expression Group of the Molecular Biology Department at the DLO Centre for Plant Breeding and Reproduction Research CPRO-DLO where he was involved in gene cloning and analysis and plant gene transfer. In 1992 hejoine d the Special Programme Biotechnology and Development Cooperation of the Ministry of Foreign Affairs, Directorate General International Cooperation, as senior officer for insti tutional development. Address: Ministry of Foreign Affairs of the Netherlands, DGIS, P.O. Box 20061, 2500 EB The Hague. J. Visser is a senior staff member of the Molecular Genetics Section of the Department of Genetics at Wageningen Agricultural University and coordinating the Aspergillus bio technology programme. He studied biochemistry at Wageningen Agricultural University and obtained his doctorate in 1970. His main interests are in the regulation of carbohy drate metabolism and fungal polysaccharide degradation systems.Address : Department of Genetics, Wageningen Agricultural University, Dreijenlaan 2, 6703 HA Wageningen. S. Visser is currently working in the Biophysical Chemistry Department at the Nether lands Institute for Dairy Research (NIZO) in Ede. He studied chemistry at Leyden Univer sity where he obtained his doctorate in 1970. His major interests are the study of milk 333 proteins, milk protein derived products and enzymatic studies in relation to milk-clotting and cheese ripening. Address: Netherlands Institute for Dairy Research (NIZO), P.O. Box 20, 6710 BA Ede. L. van Vloten-Doting isDirecto r of Research at the Agricultural Research Department (DLO) in Wageningen and affiliate professor in Applied Genetics at the University of Nijmegen. She studied chemistry and obtained her doctorate on the genome constitution of alfalfa mosaic virus at the Leyden University. Her special interest is plant breeding, plant molecular biology and plant biochemistry. Address: Agricultural Research Depart ment, P.O. Box 59, 6700 AB Wageningen. A.G.J . Voragen is professor of food chemistry in the Food Science Department at Wageningen Agricultural University, where he studied food science and before obtaining his doctorate. His special interests are structure-function relationships of food polysaccha rides. Address: Department of Food Science, Agricultural University, Bomenweg 2, 6703 HD Wageningen. J.M.G. Vorstenbosch ishea d of the research group on animal biotechnology and ethics of the Centre for Bio-ethics and Health Law at the Veterinary Faculty of the University of Utrecht. He obtained his doctorate on a thesis on the justification and application of the principle of informed consent in research ethics. His special interest is in the ethics of biotechnology, animal ethics and the methodology of moral reasoning and argumentation. Address: Centre for Bio-ethics and Health-Law, Heidelberglaan 2, 3584 CS Utrecht. W.M. de Vos is head of the Molecular Genetics Research group at the Netherlands Institute for Dairy Research (NIZO) and is professor of Bacterial Genetics and Chairman of the Department of Microbiology at Wageningen Agricultural University. He studied biochemistry and microbiology at the University of Groningen, where he obtained his doctorate in 1983. His interests include the molecular and applied genetics of anaerobic, extremophilic and archaebacteria. Address: NIZO, P.O. Box 20, 6710 BA Ede. D.H. Vuijk studied biology at Wageningen Agricultural University. He was coordinator of the Dutch Research Innovation Programme on Agricultural Biotechnology. His special interest is the development of biological sciences and its social implications. Address: Pudoc Scientific Publishers, P.O. Box 4, 6700 AA Wageningen P.L. Weegels is currently working in the Biopolymers and Physical Chemistry section of the Biochemistry and Physical Chemistry Department at the TNO Biotechnology and Chemistry Institute. In 1986 he graduated in human nutrition at Wageningen Agricultural University. He is currently preparing a PhD thesis at King's College, University of Lon don. His special interests are in gluten protein chemistry and food and non-food applica tions of industrial proteins. Address: TNO Biotechnology and Chemistry Institute, Utrechtseweg 48, 3700 AJ Zeist. 334 A.H. Weerkamp ishea d ofth e Department at Microbiology ofth eNetherland s Institute for Dairy Research (NIZO). He graduated as a biologist at the University of Nijmegen where he also obtained his doctorate in 1977. His special interest ismicrobia l physiology and food fermentation. Address: NIZO, P.O. Box 20, 6710 BA Ede. G.F. Wiegertjes took a degree in fish culture at Wageningen Agricultural University in 1988. Since then, he has been working as a research associate on the genetics of disease resistance in carp. His special interests are the development ofgynogeneti c strainso f carp, and their characterization by immunological parameters.Address :Departmen t ofExperi mental Animal Morphology and Cell Biology, Agricultural University, Zodiac Building, P.O. Box 338, 6700A H Wageningen. P.J.G.M. de Wit isprofesso r ofphytopatholog y at Wageningen Agricultural University. He obtained his doctorate in 1989 on a thesis on biochemical and physiological basis of resistance reactions in plants against pathogens. His special interest isi n molecular phyto pathology and molecular resistance breeding. Address: Wageningen Agricultural Univer sity, Department ofPhytopathology , Binnenhaven 9, 6709 PD Wageningen. P. Zabel is associate professor of Molecular Biology at the Wageningen Agricultural University. He graduated asa biochemist at Free UniversityAmsterda m and obtained his doctorate in 1978 from Wageningen Agricultural University. His special interests are molecular genetic organization of plant chromosomes, mechanisms of disease resistance, genome projects and molecular evolution. Address : Department of Molecular Biology, Wageningen Agricultural University, Dreijenlaan 3, 6703 HA Wageningen. B.A.M. van der Zeijst isprofesso r ofbacteriolog y inth e Faculty ofVeterinar y Medicine in Utrecht. He was trained as a molecular biologist; his doctorate thesis was on yeast ribosomes. After obtaining his doctorate he worked in molecular virology in the Nether lands and in the USA. His special interests are bacterial gene regulation and molecular analysis of microbial pathogenicity. Address: Department of Bacteriology, Institute of Infectious Diseases and Immunology, Faculty of Veterinary Medicine, University of Utrecht, P.O. Box 80.165, 3508 TD, Utrecht. A.J. van der Zijpp graduated inhusbandr y science and obtained her doctorate in 1982. She is currendy director of the DLO-Research Institute for Animal Production (IVO- DLO) in Zeist. The aim of IVO-DLO is to improve animal welfare, health, production efficiency and product quality. It also aims to achieve results by applying modern tech niques, such as biotechnology. Her special interests include immunogenetics and animal diseases.Address : DLO-Research Institute for Animal Production (IVO-DLO), P.O. Box 501, 3700A M Zeist. 335 Index Agtl 1 96 aminoglycoside phosphotransferase 21 5S rRNA 248 aminoglycoside-3'-phosphotransferase II gene 16S rRNA 150, 154-155, 248-249 35 23S rRNA 248 aminopeptidase N 233-234 28S rRNA 46 ammonia 172 AAL-toxin 39-40 amylase inhibitors 169, 178, 180 Absidia 256-257 amyloglucosidase 200 Ac element 39, 41 amylolytic activity 202 acetate 235 amylolytic enzyme(s) 169-172 acetic acid 172, 216-217 amylopectine 288 acetoin 235 Anaplasma marginale 154 acetolactate, a- 234-235 androstenone 131-132 acetolactate decarboxylase, a- 234 ANF(s), see: antinutritional factor(s) acetolactate synthase, a- 234 angiotensin 1210 acetyl ethylene-imine 88 anther 63-64 acetylation 242 anthurium 73 acetylesterase 216 anti-cholesterol 264 acetylglucosamine, N- 47 anti-dysentery drug 280 acid phosphatase gene 42 anti-haemophilic faxtor IX 119 acne 236 anti-idiotype 60 actin 58 anti-nutritional substances 72 Actinobacillus15 5 anti-Phytophtera protein (AP24) 46 acylglycerol(s) 226-227 anti-tumour 264 adeno virus 88, 143-144 Anti-viral vaccines 87 ADV, see: Aujeszky's disease antigen(s) 59, 94-95, 130, 261 Aeromonas salmonicida spp. nova9 9 antigenic 178, 180-181, 254, 256-257, 260 anatoxin 71, 260, 262 antigenicity 178, 212 Agrobacterium 32-34, 36, 40, 60 antinutritional factor(s) (ANF(s)) 169, 172, 177- Agrobacterium tumefaciens 32, 34, 299 182, 189, 264 agrochemicals 19 Antirrhinum 64 alcohol(s) 11, 198,236 antisense 22 alkaline phosphatase 247 antisense genes 61-64 alkaloids 178, 180-181 antisense viral RNA 22 allele 25, 96 antithrombotic activity 210 allergenic properties 212 antitrypsine, a,- 119 allergenicity 212 AP24, see: anti-Phytophtera protein 46 alloantisera 95, 97 Apergillus oryiae 171 allotransplant 94 aphA2 gene 35 alpha virus 88 apple(s)33, 214-215, 219 alpha-herpesvirinae 143 Aps-l 42-43 Altemaria 255 Arabidopsis 64, 108 Alternaria alternata f.splycopersici. 3 9 arabinan(s) 220, 242 Amaranthuscaudatus 47 arabinase(s) 200 amide 238 arabinase(s), endo- 243 amines 264 arabinofuranosidase 220, 242-243 336 arabinofuranosyl, a- 219 black-footed ferret 90 arabinose217, 220 blastula 128-129 arabinoxylan arabinofuranohydrolase (AXH) boars 110, 131, 133 217 BoLa, see: bovine lymphocyte antigen arabinoxylan(s) 183, 216-219 Borago officinalis 220 archae (archaebacteria) 153 Botrytis255 , 260 arginine 56 Botrytiscinerea 44, 257 arthritis 213 Botrytistulipae 25 7 Ac-locus 39-41 bovine 104, 116-118, 209-210 asparagine 238 bovine diarrhoea virus 90 aspartyl 211 bovine herpes virus 85 Aspergillus 188, 215, 219, 242, 244, 254-256, bovine herpes virus type 1 (BHV1) 134, 137, 258-261 139, 141-142, 149 Aspergillus awamori21 7 bovine lymphocyte antigen (BoLa) 99, 102 Aspergillusfumigatus 257 bovine respiratory syncytial virus (BRSV) 86, Aspergillusniger 173, 189, 213, 215, 220, 242-245 141 atresia 127 bovine rhinotracheitis 144 Aujeszky's disease (ADV) 86, 142-146, 149 bovine serum albumin 133 autoproteolytic processing 233 bovine somatotropin (BST) 173, 187, 197, 199, autoradiography 109 265, 289, 314 auxin 35 bovine thyroglobulin 133 avirulence gene 48 Brassica napus45 , 220 avoparcin Brow-Midrib-Mutants 185 aw9 48-50 BRSV, see: bovine respiratory syncytial virus Brucella spp. 155 B cell, see: lymphocytes Bs2 49 bacillary angiomatosis 154 BST, see: bovine somatotropin Bacillus 155, 170 Bt-toxin 21 Bacilluscereus 201 butyric acid 173,240 Bacillus thuringiensis 21, 50-52, 57, 299-300 bacteriophage 60, 232-233 caeca 175 bacteriophage-resistant strains 235 caffeine 71 baculovirus 21 callus, calli 36-38 balanoposthitis 137 Campylobacter 155, 201, 247 Balb/cmice 132 canavanine 71 bamboo 183 Candida albicans 46 barley 37, 46, 178 canola 45 barrows 132 caprylate 221 Basta 35 caprylic acid 221, 223, 225 beans 178 capsid surface antigens 88 benidine 133 carambola 73 beta-agonists 173 carbodiimide 132 beta-propiolactone 88 carbohydrate 182-184 BHV1, see: bovine herpes virus type 1 carp 95-96, 98-99 Bifidobacter 173 cartenoid 61 Bifidobacterium 200, 210 casein hydrolysates 211 biomedical proteins 10 casein(s) (a, ß, K) 113-118, 208-210, 233 bioreactor 220, 225 caseinate 210, 257 biotin 247 cat scratch disease 154 bitter peptide(s) 208-209, 211 CAT, see: chloramphenicol acetyltransferase bitter strains 209 cat(s) 87-89, 90, 145 Bl gene 64, 66 cattle 110-112, 131, 144, 189,286 337 cauliflower mosaic virus (CaMV) 35S promoter colchine 109 45 colourants 265 cDNA 39, 95-97, 100, 104, 122 Committee for Veterinary Medicinal Products cell membrane 97, 238 (CVMP) 145 cell wall degrading enzyme(s) 171-172 conjugation 232 cell wall(s) 45, 171, 183-185, 213-216, 220 convicin 180-181 cellobiohydrolase (CBH) 215 core-lignin 182 cellulase(s) 172, 198, 214-215, 242 corn 219 cellulolytic activities 172 cotton 178 cellulolytic bacteria 170-171 coumaric acid (PCA),p- 182-185 cellulose fibrils 215 coumaric acid-de-esterase, p- 186 cellulose(s) 175, 182,215,242 Council directives 232 centiMorgans (cM) 108, 111 cow(s) 129, 236 Cercospora nicotianae 45 Cowdria ruminantium 154 Cfl 49-50 cowpox 136, 145 chemotherapy 86 Crambe abyssinica 220, 225 chicken(s) 38, 95, 110-112, 175, 286 Crambe oil 226 chitin 45^-7 crygenes 50-5 2 chitinase 45^6 Cryl-proteins, see: crystal protein chitinase, exo- 46 crystal protein 50-52, 57 chitinases, endo- 45 crystallographic analysis 245 Chlamydiapsittaci 154 cucumber 33 chloramphenicol acetyltransferase (CAT) 118 cumulus 128 chloroplasts 60 curdlan 199 Chromobacterium viscosum 228, 230 CVMP, see: Committee for Veterinary Medici chromosomal walk 39 nal Products chrysanthemum 22, 63 cyanide 71 chymosin 198, 209, 264, 278-279, 286, 289, cyanidin 62 303-304 Cyca-DAB 96 chymotrypsin 54, 56, 171, 180 cyprinids 90 chymotrypsin inhibitor(s) 55, 56, 178 cysteine 47, 95, 143 CID9 118-119 cystic fibrosis 119 cinnamon 197 cytoplasm 128 citPgene 234 cytoplasmic RNA 97 citrate 234, 235, 247 citrate permease 234 dairy cows 184 citrus 219 Datura 178 Cladosporium 255 de-esterase(s) 186 Cladosporium cladosporioides 257 debittering aminopeptidase 235 Cladosporiumfulvum 48-49 decanoic acid 228, 230-231 class I MHC 93, 95-97, 100 defective viral poymerase 22 class II MHC 93, 95-97, 100, 102 defective viral transport protein 22 Clavibacter michiganensis 30 dehydroalanine (Dha) 236-237 cloacal mucosa 87 dehydrobutyrine (Dhb) 236-237 Clostridium 241 dehydroferrulic acid 183 Clostridium tyrobutyricum 240, 247, 250-253 delayed hypersensitivity factor 247 cM, see: centiMorgans delphinidin 62 co-segregation 23, 25 dexamethasone 118 co-suppression 62 dextrines 199 cocoa butter 200 dextrose 200 codominant 105 Dha, see: dehydroalanine codon(s) 239 Dhb, see: dehydrobutyrine 338 diacetyl 235 endoplasmatic reticulum (ER) 60 diacylglycerols 226, 228, 230 endoproteinase glu-C 181 diallel 29 endotoxin(s) 300-301 dicaprylin221, 223-224 Enterococcus 246, 248 dielectric constant 228-230 enzymatic hydrolysis 177 diester 226-229, 231 enzyme immuno assays (EIA's) 133 dietetic fibre 199 enzyme-linked immunosorbent assays (ELISA) digestive tract 170, 174 201, 255, 256, 258, 259, 260 diglycerides 221-222, 225 EPC, see: European Patent Convention digoxigenin 247 epidemic(s) 30 dimorphecolic acid 220, 225-226 epidermis 63 Dimorphotecapluvialis 220, 225 epithelial cell strain COMMA-ID 118 Directive 13 epithelial cells 113, 115-119 DNA-coated particles 36 epithelium 88 DNHP-KLH 98 epitope(s) 88, 255-256 dog(s) 87-88, 154 EPO, see: European Patent Office domino effects 279 EPS, see: extracellular polysaccharide(s) dough 204 equine contagious endometritis 155 DQA 101 ER, see: endoplasmatic reticulum DQßlOl erucic acid 220, 226 DRB 101-103 ESC, see: embryonic stem cells Ds-element 40-41 Escherichia coli 89, 90, 114, 154, 201, 232, 234, dTphl 66 247 dubble zero varieties 179 esterification 226-228, 231, 242 Dutch Government Committee on Biotechnol estradiol-1,7/? 115 ogy and Animals 313 ethanol 235 ethylene 61 El gene 145-146 Eucarya 153 ECM, see: extracellular matrix 118 EUREKA RFLP maize project 27 eel 90 Eurobarometer 303 Ehrlichiaspp . 154 European Patent Convention (EPC) 293-294, EIA's, see: Enzyme Immuno Assays 296 electroporation 36 European Patent Office (EPO) 293-294, 296 electrostatic repulsion 214 evening primrose 220 elicitor 48 exogenous enzymes 174, 214 ELISA, see: enzyme-linked immunosorbent as exon 119 says extracellular matrix (ECM) 118-119 embryo technology 310 extracellular polysaccharide(s) (EPS) 254—256, embryo tissue 36 258, 260-261 embryo transfer 131 embryo(s) 84-85, 127, 129-130 FA, see: ferulic acid embryonic stem cells (ESC) 131 faba beans 179, 181 emulgators 264 FACS, see: fluorescense activated cell sorter emulsifyers 203 flpim emulsifying properties/capacity 203, 207 feline leukaemia virus 89 encephalitis 143 FeLV gp70 protein 90 endo I, II, II, IV, see: xylanases, endo- ferulic acid (FA) 182-185 endo-PG, see: polygalacturonase, endo- ferulic acid-de-esterase 186 endo-PL, see: pectin lyase, endo- fingerprints 25 endocrine (system) 85, 115 flatulence factors 178, 180 endogenous 54—55 flavonoid 61-62, 64 endometritis 154 flavour(s) 199, 207-208, 211, 264-265 339 flavr savr 73 gl protein 90 flour 204 gibberellin 62 flow cytometry/cytometric 130, 201 GIFAP 70 fluorescense activated cell sorter (FACS) 109, gll gene 145 111 guts 133-134 fluorochrome 247 glaucoma 213 foam 213 gliadin(s) 202, 205-206 foaming properties/capacity 203, 207 globin, ß-, genes 114 follicle aspiration 128-129 Gtobodera 60 follicles 127-129 Globodera rostochiensis 60-61 follicular atresia 127 glue, X- 35 food emulsions 213 glucan, yS-1,3- 45 food-grade bacterial enzymes 213 glucanase(s) 200 foot-and-mouth disease 88-89, 136 glucanase(s), endo-/M,3- 45 formaldehyde 88 glucanase(s), endo-/?-1,4- (Endo I, Endo IV) 215 formate 235 glucanase(s), exo- 200 formic acid 173 glucanase(s), exo-/M,4- 215 fowl cholera 136 glucanase(s), ß- 172-173, 188 fowlpox 88, 144, 149 glucanase(s), jS-1,3- 45^46 fox(es) 88-89, 142, 144 glucocorticoids 115 fragrants 199, 265 glucoronidase, ß- 35, 301 fructose 201 glucose 176 fructosyl-oligosaccharides 200 glucose, D- 175 fructosyl-transferase 200 glucose oxidase 243 furanones 199 glucosidase, N- 181 Fusarium254-255 , 261 glucosidases, ß- 172 Fusariumculmorum 257 glucosinolates 180-181 Fusariumoxysporum 257 glucuronic acid 216-217 Fusariumsolani 25 7 gluronic acid, 5-bromo-4-chloro-3-indolyl-/?- 35 gluten 201-207, 217 G418 35-36 glutenin(s) 202-203, 205-206 galactofuranose, ß-1,5-D- 254—256 glycerol 200, 225-231 galactosidase, V-ß- 248, 251 glycine 46 galacturonic acid 216 glycoprotein El 144 gametophyte 22, 61 glycoprotein G 88, 144 gastrointestinal tract 182 glycoprotein gl 145 ^Cgene 140-141 glycoprotein gp63 143, 145 £Ëgene 140-141 glycoprotein gX 143 gelelectrophoresis 97 glycosidase, N- 46 gellan 199 glycosidic linkage 217 gene-pool glycyl211 generalized linear models 27 GnRH, see: gonadotrophin-releasing hormone geneticin 35 gonadotrophin-releasing hormone (GnRH) 85, genital mucosae 135 131-134 genomes 104 gonadotrophins 128 genotype 42 gos2 gene 36 genotypic 123, 125 gossypol 178, 180 Geotrichum candidum25 7 Gouda cheese 208 germplasms 69 gox gene 243 ^Ggene 140-141 Gpgen e 64, 66 gH gene 145 gp50 gene 145, 149 glgene 140-141, 144-146 Gram-negative bacteria 151, 236 340 Gram-positive bacteria 151, 232, 236 hybridoma technology 58, 132 grapes 214 hybridoma(s) 132-133 grapevine 44 hydrolase(s) 45-46 gus reporter gene 37 hydrolysate(s) 204, 212-213 GUS 35-36 hydrophilic 202, 220 gXgene 144, 146 hydrophobic(ity) 202-203, 208-209, 211 gynogenesis 97 hygromicin phosphotransferase gene 35 gynogenetic families 98 hygromycin 35 hyoscyamin 180 haemolysin A 247 hypersensitive response (HR) 47-48, 50 haemolysis 180 hyperthermia 124 haemolytic anaemia 180 hypoallergenic food 200 halothane gene 123-125 hypothalamus 132 haplotypes 93, 103-104 HC11 cells 118 IEF, see: iso-electric focussing HCV, see: hog cholera virus IEF-profiles 97 heartwater 154 IgG (antibodies) 133-134, 255-256 Heliotyisvirescens 52 IgG-Sepharose column 97 helium 36 ileum 175 hemicellulase(s) 205, 242-243 immunoamnity 97 hemicellulolytic 172, 202, 242, 245 immunoassay(s) 257-258, 260-261 hemicellulose(s) 175, 177, 182, 242 immunogenic 88 hepatitis B virus 89, 142 immunogenicity 87 heptanone, 2- 199 immunoglobulin 60, 94, 210 herbivores 71 immunological test-kits 201 herpes virus 87-88, 138, 143-144 immunoprecipitation 97 heterologous 54 immunoprophylaxis 86 heterozygous 110 in vitro fertilization 127-130 Heveabrasiliensis 47 influenza 88-89 hevein 45, 47 insecticidal protein(aceous) 21, 50, 298-299 high performance liquid chromatography interferon 86 (HPLC)210 interval mapping 26 High Temperature Short Time process 181 intestines 132 histidine 237-238 iso-electric focussing (IEF) 100 HLA 5 104 IVF, see: in vitro fertilization HLAZ)£M04 HLA, see: Human Lymphocyte Antigen JoinMap 27, 30 hlyA gene 247 juice clarification 200, 214 hog cholera 85 hog cholera virus (HCV) 144-145 kanamycin 21, 32, 35-36 homogalacturonan 215 kiwi 73 homologous 244 homozygous 97 L-arabinose 175-177 horse(s)89, 154, 169 L-arabitol 176 HPLC, see: high performance liquid chroma L-arabonic acid 176 tography L-rhamnose 216 hpt gene 35-36 lacFgen e 232, 234 HR, see: hypersensitive response lacG gene 248, 251 human adeno virus 144 lactalbumin, a- 113, 116, 210-211 human cytomegalo virus 144 lactamase, ß- 247 human ryodine receptor gene 124 lactase 201 Human Lymphocyte Antigen (HLA) 101 lactate 171,235 341 lactic acid 169-170, 175 Listeriamonocytogenes 247, 253 lactic acid bacter(-ium,-a) 170, 199, 219, 231— liver 180 236, 241 LMP, see: Large Multifuctional Protease gene lactitol 175 locus, loci 24, 38-39 lactobacilli 190,231 Loliumperenne L. 37 Lactobacillus 170, 247-248 luminescence 247 Lactobacillus delbrueckii 247 lupin(s) 178, 180 Lactobacillus helveticus 247 luteunizing hormone (LH) 132-133 Lactobacillusplantarum 170, 247, 450 Lycopersicon esculentum 39, 42 lactococcal cells 209 Lycopersiconperuvianum 42^- 3 lactococci 231-235 lymphocytes, B,T cell 88, 92-95, 97, 102, 104, Lactococcus 209, 231, 241, 246-248 143 Lactococcusgarviae 250 lysine 56, 188 Lactococcus lactis 231-234, 236, 238, 240-241, lysozyme 38, 53 247,250-251 Lactococcus mesenteroides 251 MAC-T cell line 116 Lactococcusparamesenkroides 251 maceration 214, 245 Lactococcus raßnokctis 250 macrophage 91 lactoferrin 119-210, 213, 284, 289 MADS box genes 66 lactoglobulin, ß- 113, 116, 119-211, 213 maize 185 lactose 175, 232-233, 235 major histocompatibility complex (MHC) 84, lactose metabolism 232 90-104 lanthionine(s) 236-238 male sterility 61, 64 lantibiotics 236-237, 240 maltose 200 laparascope 128-129 mammary gland 113, 115-119, 121 Large Multifuctional Protease (LMP) gene 101, Manduca sexta52 , 54—55 103-104 mannopine 36 laryngotracheitis 87 MAPMAKER/QTL 26, 30 late-blowing 240-241 Marek's disease 87, 144 latex agglutination reaction 201 marigold 220 leaf miners 51 marker assisted selection (MAS) 25, 126 lectin(s) 45, 47, 71, 169, 172, 178, 180-181 MAS, see: marker assisted selection leek 33 mastitis 10 Leptospira spp. 155 maximum likelihood 105 lettuce 33 measles 134, 143 leucine 56 meiotic division 97 leucine aminopeptidase 181 Meishan, chinese 110 Leuconostoc 248, 251 Meloidogyne 60 leuconostocs 231 Meloidogyne incognita 42 leukaemia 89 membrane bioreactors 220 LH, see: luteinizing hormone membrane vessicles 234 lignin 169, 182, 185 membrane(s) 45, 130 ligninase(s) 242 meristematic tissue 35 lily 37 messenger RNA (mRNA) 60, 247 LINKAGE 30 metaphase 109 linolenic acid, a- 220 methylation 38 linolenic acid, T- 220 methyllanthionine, ß- 236-237 lipase(s) 200, 220- 221, 225-226, 230 MGF 116, 118 lipolytic enzyme(s) 171-172 MHC, see: major histocompatibility complex lipoprotein 233 Mi locus/gene 42-43 liquefaction 214, 245 mice 133, 144 Listeria20\, 155 microbial mats 154 342 microcarrier 88 264 Micrococcusflavus 239 nisin A 201, 238-240 microglobulin, ß2- 93, 96 nisin Z 201, 238-241 microsatellites 106-107, 111 nitrate 240-241 millet, pearl 185 nitrite 240 minisatellites 84, 107 nitrosamines 240 mixed leucocyte responses (MLR) 94 non-core-lignin 182 mixed lymphocyte culture 100 non-polar solvent 228 MLA, see: mould latex agglutination assay non-starch polysaccharide(s) (NSP(s)) 171, 172, MLR, see: mixed leucocyte responses 175, 177,216 molecular markers 23, 42, 57- 58, 60, 84- 85, novel input 277, 280, 310, 314 131-134, 141 Novel Food Regulation, EC- 290, 304 Monascus 255, 260 nptll gene 35 Monaseus pilosus 257 NSP(s), see: non-starch polysaccharide(s) monoacylglycerols 226, 228, 230 nuclear scaffold attachment regions (SARs) 38 monocaprylin 221, 223-225 nucleocapsid proteins 22 monoclonal antibodies 198, 256 monoester 226-229, 231 O-glucosidase 181 monoglyceride(s) 221-225 Oct-1 116 monosaccharides 175 OECD 13,70, 198,297 morula 128-129 Oenanthera spp. 220 mould latex agglutination assay (MLA) 255, oestrous cycle 127, 133 256, 258, 259, 261 oligosaccharide(s) 173, 175, 217 mouse 95, 104, 116 onion 33 mouse cell line HC 11118 oocyte(s) 58, 85, 127-130 mRNA, see: messenger RNA oomycetes 46 Mucor 255-256, 260 opioid 210, 213 Mucorpiriformis 25 7 orchid 73 Mucor racemosus 257 orotodine 5'-phosphate decarboxylase 244 Mucorales 254, 256, 258, 260 osmotin 46 mycelium 254 ova 97 Mycobacterium spp. 155 ovaries 129 mycoplasmas 143 ovine 116, 119 mycoproteins 198 ovulation 127 mycotoxicoses 254 mycotoxins 254, 260-264 Paf66 myeloma 60 pale soft exudative meat (PSE) 123 palmoil 200 Nam 66 pancreas hypertrophy 180 necrosis 49 pancreatic enzymes 180 necrotic 47 pancreatin 181 neomycin 35 papain 181,211-212 nervous tissue 138 paramomycin 35 Netherlands Organization for Technology As paratopes 256 sessment (NOTA) 279, 305-307 particle bompardment 36-37 neurodermatitis 213 particle delivery system (PDS) 36 Neurospora crassa 46 particle gun 36, 283 Newcastle disease 88 Pasteurella multicida 136 NF1 116 PCA , see:/)-coumari c acid Mcotiana sylvestris 45 PCR primers 96 nisA gene 247-248, 251 PCR, see: polymerase chain reaction nisin 232, 236-238, 240-241, 247-248, 251, PE, see: pectinesterase 343 pear 215 phytoalexins 44 peas 178 Phytophtera infestons 46 pectin esterase PI, see: proteinase inhibitor genes pectin lyase, endo- (endo-PL) 215 pig(s), piglet(s) 90, 110-111 ,146, 169, 171-172, pectin lyase(s) (PL) 242-245 175-177, 181, 188,280,286 pectin(s) 214-216, 219, 242 pigeon pox virus 87 pectinase(s) 172, 214, 216, 242-243 PIGMAP84, 105, 110-111 pectinesterase(s) (PE) 242-243 PK, see: protein kinase pectinolytic 242, 244-245 phi gene 244 pectolytic enzymes 215 PL, PLI, PLII, PLA, PLD see: pectin lyase pediococci 231 Plant Breeders Rights 69 Pediococcus 247 plantibodies 22 PEG, see: polyethylene glycol plasmid(s) 232-233, 240-241 pelargonidin 62 plasmin 210 ^«ZB/C/E/F gene 244 plastein reaction 207 pelvic wall 129 plating techniques 254 Pénicillium 254-256, 258-261 pleiotropic 265 Pénicillium aurantiogriseum 257 PMF 118 Pénicillium digitatum25 7 pNZ1125 233 Pénicilliumfréquentons 25 7 polar solvent 228 Pénicillium islandicum 257 polio virus 142 Pénicillium verrucosum 257 poliomyelitis 88, 134 pentose sugars 175 polyclonal antibodies 203, 255-256 pepJVgene 233-234 polyethylene glycol (PEG) 109 pepper 49, 197 polygalacturonase, endo- (endo-PG) 215 pepsin 171, 181 polygalcturonase(s) (PG) 242-243, 245 peptidase(s) 171 polygenes 24, 29 pestivirussen 143 polymerase chain reaction (PCR) 20, 39, 95,97, petunia 63-64 107, 124-125, 150, 155, 246-248, 251-252 PG, see: polygalcturonase polymorphic markers 111 phage resistant strains 233 polymorphic sites 104 phage-resistant 199 polymorphism 84, 93, 95, 97, 102, 105, 107, phagocytosis 91 123-126 Phanerochaete 242 polysaccharide(s) 199, 213-215, 220, 242, 254- Phaseolus 189 255, 258 Phaseolus vulgaris 181 polythene 280 phenolic 184 poplar 33 phenolic acid(s) 182-183, 185, 217 porcine pituitary cell 134 phenolic alcohol(s) 182 porcine somatotropin (PST) 173, 187, 289 phenolic compound(s) 178, 182, 184 porcine stress syndrome (PSS) 123 phenolic monomer(s) 182-183, 185 potato 33, 47, 52, 215, 219, 287-288, 298 phenotype 42, 47, 87 potato cyst nematodes 60 phenotypic 123, 125 potato starch 10 phenotypically 110 potato virus X 22 phenylalanine 62, 213 poultry 144, 175-177, 181, 188 phenylketonuria 213 pox virus 88, 144 pheromones 10 PR, see: pathogen related proteins phosphopeptides 210 probiotics 189 phylogenetic system 151, 153 product life cycle(s) 282-283 phylogeny 150-153 progestreon 115 phytase 171, 173, 188-189, 289, 303-304 prolactin 115, 118 phytate 173, 180 prolidase 181 344 propagules 254 RedJungl e fowl 110 Propinibacterium acnes 236 reductase systems 201 proteases 51, 54, 181, 200 rennet 209 protein engineering 234-236, 238-239 repulsion phase 27 protein kinase (PK) 143, 145 restriction enzyme Hgi AI 125 proteinase inhibitor (PI) 21, 54-57 restriction enzyme Rsa I 124 proteinase K 181 restriction fragment length polymorphism proteinase(s) 56, 208-209, 233-235 (RFLP) 20,48, 84-85, 98, 100-101, 105-107, proteinase-resistant toxin 51 111-112, 123-126 proteolysis 212, 234 resveratrol 44 proteolytic degradation 233 reversed phase HPLC 239 proteolytic enzyme(s) 171, 181, 208-209 RFLP, see restriction fragment length polymor proteolytic systems 233 phism prtM gene 233 RFLP-map 41 prtP gene 233 rhamnogalacturonan 216 PSE, see: pale soft exudative meat rhamnogalacturonase 216 Pseudomonas 280 rhamsan 199 Pseudomonas syringae 47 rheumatism 213 Pseudorabies virus 85, 90 rhinotracheitis 137 pSH71 232 rhinotracheitis virus 87 PSS, see: porcine stress syndrome Rhizoctoniasolani 45-46 PST, see: porcine somatotropin Rhizomucor 256-257 PTiB0542 34 Rhizopus 254-256, 260 pullulan 199 Rhizopus delemar 221 pulp enzyming 214 Rhizopusjavankus 221-224 pulsed field electrophoresis 101 Rhizopus stobnifer 257 pyrA gene 244 RIA, see: radio immuno assays pyroglutamate aminopeptidase 181 ribonucleotide reductase (RR) 143, 145 pyruvate 234-235 ribosomal RNA (rRNA) 150, 247-248 pyruvate formate lyase 234 ribosome inactivating protein (RIP) 45-46 pyruvate kinase 244 rice 36, 197, 216 pyruvyl 238 Rickettsiaprowazekii 154 RIL, see: recombinant inbred lines QTL, see: quantitave trait loci RIP(s), see: ribosome inactivating protein quantitave trait loci (QTL) 20, 26-30, 84, 105, root-knot nematode 42 109-112 Roquefort cheese 199 quinolizidin 180-181 rose 33 quinolones 156 Roundup 35 RR, see: ribonucleotide reductase rabbits 169 rRNA, see: ribosomal RNA rabies 88-89, 134, 137, 142-144 rumensin 174 raccoon(s) 88, 144 rye 37, 178 raccoonpox 144 radio immuno assays (RIA's) 133-134 Sabin vaccine 142 random amplified polymorphic DNAs (RAPDs) Saccharomyces cerevisiae 90, 171 20,48,84, 107, 108, 111 saccharose 200 RAPDs, see: random amplified polymorphic SAGB, See: Senior Advisory Group Biotechnol DNAs ogy rapeseed 33, 178-179, 218, 220 salmon 90 rat(s) 116, 118-119, 175 Salmonella144 , 155, 201, 247, 253 receptor 48 saponins 178, 180 recombinant inbred lines (RILs) 27 saprophytic 242 345 SAR(s), see: nuclear scaffold attachment regions sugarbeet(s) 33, 287 satellite RNA 22 sugarcane 197 scleroglucan 199 sulphydryl212 scopolamin 180 superovulation 127 SDS 206-207 supratypes 103-104 Seiemonas ruminantium 171 sweeteners 265 Senior Advisory Group Biotechnology (SAGB) swine 144, 149 69 swine fever virus 86 serine 236 swinepox 144, 149 serine proteinase 56, 233 SWOKA 303 serum albumin 210 Syncephalastrum 256-257 sex pheromone 132 sexing 130 T cell receptor (TCR) 92, 102 Shigella spp. 144 T cell, see: lymphocytes silage(s) 170-173, 190, 198, 247 T helper lymphocytes 92-93 sinapin(s) 178, 180-181 T-DNA 32, 34 single chain antibody 59 tannin(s)71, 169, 172, 179-181 skatole 131-132 TAP, see: transporter associated with antigen skunk 88 processing gene smallpox 87, 134-135 tapetum cells 64 Solanum 300 Taql 98 Solanum tuberosum 300 Taylorella equigenitalis 155 Solanumtuberosum ssp. andigena 60 TCR, see: T cell receptor somatotropins 187 telecommunications 277 sorghum 179, 185,216 tempe 254 Southern blotting 125 terminal repeat sequence (TR) 140 sows 132 testicular steroid 132 soya 178,202 testosterone 132 soya bean 181 Thamnidium 256-257 sperm 85, 97, 130 thermodynamic equilibrium 221 sperm-sexing 130 thermolysine 181 spermatozoa 130 thermonuclease 247 spleen cells 60 Thermus249 , 252 Spodoptera 51 thiol 203 Spodoptera exigua51-5 2 thionine 45-46 Spodoptera littoralis 51 threitol, D- 176 Staphylococcus aureus 201, 247 threonine 174, 188,236 starch 61, 217, 287 thrips 51 starch-degrading-enzymes 217 thymidine kinase 88, 141, 143 starter culture(s) 199, 232, 236, 240-241 thyroid 180 stearic acid 200 Tilapia 96 stilbene synthase 44 tissue plasminogen activator 119 stomata 36 77Tgene 141, 144-146 streptavidin 247 TK, see: thymidine kinase streptococci 231 Tn5276232 Streptococcus 155, 170, 246, 248 tobacco 38, 45-46, 52, 55-56, 60 Streptococcus agalactiae 236 tomato 30, 33, 39, 42, 49, 73, 288 Streptococcus lactis23 6 toxins 264 structural carbohydrates 173, 182 TR, see: terminal repeat sequence subtilases 233 transduction 232 subtilisin 55 transglutaminase 206 sucrose 232 transporter associated with antigen processing 346 (TAP) gene 101, 103 verotoxin 247 transposable element 40 veterinary drugs 10 transposon tagging 20-21, 39-40 vicin 180-181 transposon(s) 232, 239-241 virgen e 32, 34 transvaginal follicle aspiration 129 viral envelope 88 TREP, see: Two-phase Reaction Equilibrium virulence 87 Prediction viscosity 207 triacylglycerols 226, 228, 230 VNTR, see: variable number of tandem repeat tricaprylin 221-224 loci Trichoderma 242 volatile fatty acids 175 Trichoderma reesi 46 vulvovaginitis 137 trichomata 36 triester 226-231 WAP, see: whey acidic protein triglycerides 220-222, 225-226 welan 199 triticale 178 wheat 201-202, 206, 217, 219, 288 tropism 144-145 wheat starch 10 trout 90 whey acidic protein (WAP) 113, 116, 119 Trypanoplasma borelli 99 whey protein hydrolysates 211 trypsin 54, 56, 171, 180,211 whey protein(s) 208, 210 212-213 trypsin inhibitor(s) 54-56, 169, 172, 178, 180- Whipple's disease 154 181 white flies 51 tryptic hydrolysis 213 White Leghorn 110 tryptophan 95, 132, 173, 188 willow 33 tulip 33, 37 WIPO 293 tumour inducing (Ti) plasmid 32 tumour necrosis factors (TNF) 103 X-chromosome 130 turkey herpes virus 87 xanthan 199 Two-phase Reaction Equilibrium Prediction Xanthomonascampestris 47, 49, 199 (TREP) 227, 231 xlnA gene 243 xylan 243 ultrasonography 128 xylanase, endo-l,4-/J- 243 ultrasound-guided follicle aspiration 85, 128 xylanase(s) 172, 200, 217-218 undecanone, 2- 199 xylanase(s), endo- 189, 217-219 UNIFAC group contribution method 227, 229- xylanolytic enzymes 217 230 xylitol 175 universal genetic tree 153 xyloglucan 215 UPOV 69, 292, 295-296 xylopyranose 217 urokinase 119 xylopyranosyl, /?- 219 uronic acid(s) 175 xylose 182, 217 Us2 gene 144 xylose, D- 175-177 USDA-ARS 85 xylosidase, yS- 217 vaccinia virus 87-88, 145, 149 Y-chromosome 130 vacuolar proteins 45 YAC(s), see: Yeast Artificial Chromosomes vacuoles 45-46, 62 Yeast Artificial Chromosomes (YAC) 43 vagina 129 yeast(s) 11, 60, 142, 171, 190, 198-199 valine 143 yellow fever 86, 134 vaniline 199 Yersinia enterocolitica 247 variable number of tandem repeat (VNTR) loci Yersinia spp. 155 84, 106 variola virus 87 zero tolerance 284 veal calf/calves 181,202 347