Research Collection
Doctoral Thesis
Characterization of propionibacteria in Swiss raw milk by biochemical and molecular-biological methods
Author(s): Fessler, Denise Sophie
Publication Date: 1997
Permanent Link: https://doi.org/10.3929/ethz-a-001855109
Rights / License: In Copyright - Non-Commercial Use Permitted
This page was generated automatically upon download from the ETH Zurich Research Collection. For more information please consult the Terms of use.
ETH Library Diss. ETH No. 12328
Characterisation of propionibacteria in Swiss raw milk by biochemical and molecular-biological methods
A thesis submitted to the Swiss Federal Institute of Technology (ETH), Zurich for the degree of Doctor of Technical Sciences
presented by
DENISE SOPHIE FESSLER
Dipl. Lm.-lng. ETH born September 25th, 1969 citizen of Walzenhausen (AR)
accepted on the recommendation of Prof. Dr. Z. Puhan, examiner Dr. M. G. Casey, co-examiner Dr. S. Lortal, co-examiner
Zurich 1997 To my parents Acknowledgements
I would like to express my sincere gratitude to Prof. Dr. Z. Puhan for giving me the chance to carry out this thesis and for his supervision.
Very special and warm thanks belong to Dr. M. G. Casey for supporting me. He gave me invaluable suggestions and always had an open ear for my problems. Without his patience and confidence this thesis would not have been possible.
I thank Dr. Sylvie Lortal for taking over the external examination of this work.
I would like to extend my appreciation to the whole Department of Bio¬ chemistry, FAM. I thank Dr. M. Furst, Dr. J. Jimeno, Dr. A. Baer and Dr. J. Meyer for their backing and Mrs. M.-T. Raemy, Mrs. 1. Ryba, Mrs. N. Loosli, Mr. J. Gruskovnjak and Mr. A. Spahni for creating the friendly working environment.
I wish also to thank Dr. C. Steffen, manager of the FAM for providing me with the working place and all the employees of the FAM, who gave me their support.
I thank Dr. E. Wehrli for the analysis with the Raster Electron Micro¬ scope.
Particular thanks belong to my family and all my friends for their unfailing encouragement and friendship. Abbreviations
ATCC American Type Culture Collection, Rockville, Md. bp base pair CNRZ Centre National de recherches Zootechniques, Jouy- en-Josas, France cont. continued dist. distilled
DSM Deutsche Sammlung von Mikroorganismen und Zellkulturen GmbH FAM Federal Dairy Research Institute, Switzerland FAMb Biochemistry, Federal Dairy Research Institute, Switzerland Fig. Figure freud. freudenreichii kb kilo base
Lb. Lactobacillus MIBD/SICL Milchwirtschaftliche Inspektions- und Beratungsdienste Dairy Inspection and Consulting Services min minute P. Propionibacterium °SH acidity sherm. shermanii sp. species subsp. subspecies T type strain not analysed Contents
Summary / Zusammenfassung / Resume 1
1. Introduction and presentation of the problem 8
2. Literature 10
2.1. History and taxonomy of the genus Propionibacterium 10
2.2. Factors affecting growth of propionibacteria 14 2.2.1. Temperature 14 2.2.2. pH 14 2.2.3. Effect of sodium chloride 14
2.3. Metabolism 15 2.3.1. Propionic acid fermentation 15 2.3.2. Metabolism of aspartate 15 2.3.3. Proteinase and peptidase activity 17 2.3.4. Lipase and esterase activity 17
2.4. Some other features of propionibacteria 18 2.4.1. Plasmids 18 2.4.2. Bacteriophages 18 2.4.3. Bacteriocins 19 2.4.4. Interactions with lactic acid bacteria 19
2.5. Propionibacteria in practice 20 2.5.1. Natural habitat 20 .^ 2.5.2. Use of propionibacteria 21 2.5.3. Propionibacteria as probiotics 22 2.5.4. Propionibacteria in cheese manufacture 23 2.5.4.1. Swiss-type cheese 23 2.5.4.2. "Brown spots" defect 23 2.5.4.3. Split defect 24
2.6. Methods for enumeration, identification, differentiation and classification of propionibacteria 25 2.6.1. Isolation and enumeration 25 2.6.2. Identification and classification based on biochemical methods 26
2.6.3. Identification and classification based on genomic criteria 27 2.6.4. Identification and classification by other methods 28
3. Material and methods 30
3.1. Growth media for propionibacteria 30 3.1.1. Yeast extract lactate broth (YEL) and yeast extract lactate agar (YELA) 30 3.1.2. Minimum growth medium 30 3.1.3. MF95C 30 3.1.4. Peptone whey 31 3.1.5. Dilution solution 31
3.2. Media for other bacteria 31
3.3. Microorganisms 31 3.3.1. Propionibacteria 31 3.3.2. Other bacteria 33
3.4. Sugar fermentation 33
3.5. Growth curves 33
3.6. Survival rate at cheese manufacture conditions 33
3.7. Analysis of the soluble cell free protein extract 34 3.7.1. Preparation of protein extracts 34 3.7.2. Protein gel electrophoresis and staining of gels 34
3.8. Purification of genomic DNA 35
3.9. Restriction enzyme analysis 36
3.10. Ribotyping 37 3.11. Restriction analysis of 23S rRNA 37 3.11.1. Preparation of DNA extracts 37 3.11.2. Polymerase chain reaction and restriction analysis 38
3.12. RAPD 38
3.13. Plasmids 40
3.14. Emmental cheese 41
3.14.1. Model Emmental cheese manufacture 41
3.14.2. Analysis and sensoric tests performed on model cheeses 43
3.14.3. Emmental cheese manufacture 44
4. Results and discussion 46
4.1. Propionibacterial flora in Swiss raw milk from lowlands and alps 46 4.1.1. Introduction 46 4.1.2. Reproducibility of protein profiles and restriction analysis profiles 46 4.1.3. Propionibacterium species in lowland milk 47 4.1.4. Propionibacteria from alps 52 4.1.5. Classification of P. rubrum 53 4.1.6. P. freudenreichii subspecies 53 4.1.7. Differentiation between strains 54 4.1.7.1. Plasmid profile 55 4.1.7.2. RAPD 56
4.2. Propionibacteria used for Emmental cheese manufacture in Switzerland 60
4.2.1. Introduction 60 4.2.2. Protein profiles, plasmids and RAPD of commercial strains 60 4.2.3. Detection of propionibacteria of commercial cultures 63 4.3. Propionibacteria and cheese faults 66 4.3.1. Introduction 66 4.3.2. Brown spots 66 4.3.2.1. Emmental 66
4.3.2.2. Sbrinz 67 4.3.2.3. Appenzell 69 4.3.2.4. Raclette 70
4.3.2.5. Conclusions 70 4.3.3. Split defect 71 4.3.3.1. Introduction 71
4.3.3.2. Sbrinz 71 4.3.3.3. Gruyere 72 4.3.3.4. Conclusions 72
4.4. Model Emmental cheese manufacture with selected wild Propionibacterium strains 73 4.4.1. Introduction 73
4.4.2. Selection of strains 73 4.4.3. Quality of model Emmental cheese 75 4.4.4. Influence of subspecies 78 4.4.5. Chemical and microbiological analysis 79 4.4.6. Sensorics of model Emmental cheeses 82 4.4.7. Raster electron microscopy 83
4.5. Emmental cheese manufacture with three wild Propionibacterium strains 87 4.5.1. Introduction 87 4.5.2. Quality of cheeses 87
5. Conclusion 90
6. Bibliography 94
7. Annex 102 Summary / Zusammenfassung / Ftesumfe £1;
Summary
Characterisation of propionibacteria in Swiss raw milk by biochemi¬ cal and molecular-biological methods
For the manufacture of Emmental cheese in Switzerland raw milk is used. The formation of eyes, however, is achieved by addition of culture with selected strains of propionibacteria. The wild propionibacteria in raw milk and cheese have until now not been systematically analysed. How¬ ever, it is known that they are responsible for defects such as brown spots and split defect in various hard or semi-hard raw milk cheeses and that technological factors during cheese manufacture may influence the occurrence of these defects. The commercially available P-culture, a mixture of P. freudenreichii strains, has been used for many years and a new culture, Prop 96, is also commercially available since 1996. The strains included in these cultures have until now not been analysed by molecular-biological methods.
In the present study the propionibacterial flora in Swiss raw milk was in¬ vestigated and propionibacteria were classified by protein profile analy¬ sis, restriction profile analysis of the 23S rRNA gene, plasmid content and RAPD profiles. Propionibacteria isolated from various cheese types with defects were differentiated and classified, and compared with the propionibacteria strains isolated from raw milk.
The propionibacterial flora in Swiss raw milk was found to be extraordi¬ narily rich. All four dairy Propionibacterium sp. were found in lowland raw milk, 71% were P. freudenreichii, 19% P. jensenii, 8% P. acidipropionici and 2% P. thoenii. In alpine raw milk P. acidipropionici was not found, P. freudenreichii made up 55%, P. jensenii 15% and P. thoenii 30% of the total. Among the 278 P. freudenreichii strains 219 (79%) different strains were identified by RAPD at the strain level. For the other species strain diversity was even higher. Only 30% of all analysed strains carried plasmids.
Commercial cultures were analysed by RAPD and the presence of the strains making up the P-culture was looked for in Emmental cheese, in order to see, if all the strains grew in the cheese. Of the six strains com¬ prised in the P-culture, four strains were found to be identical, meaning that this culture contains only three different strains. One of these strains -2- Summary / Zusammenfassung / Resume could not be detected in commercially available premium grade Emmen¬ tal and one was only occasionally present. Consequently, it is possible that only one of the P-culture strains is responsible for the good cheese quality. The two strains in Prop 96 were found to be identical, but differ¬ ent from the P-culture strains.
In brown spots of Emmental and Sbrinz cheeses various strains of P. freudenreichii were detected by RAPD. The brown spots of Appenzell cheese also contained P. jensenii and P. acidipropionici. In Gruyere and Sbrinz cheeses with split defects, different P. freudenreichii strains were detected. Only one of the P. freudenreichii strains of Appenzell, Gruyere and Sbrinz with defects was identical with strains isolated from raw milk.
Model Emmental cheeses were produced with selected wild P. freuden¬ reichii strains. Cheeses manufactured two or three times with the same strain were of identical quality. Chemical, microbiological and molecular- biological methods detected no differences between cheeses of favour¬ able quality and cheeses with tendency towards split defect. Some chemical and microbiological differences, however, were detected be¬ tween cheeses of favourable quality and cheeses with brown spot forma¬ tion. The most striking difference was between the growth curves of strains producing good quality cheeses and strains producing cheeses with brown spots. The latter showed slower growth in model Emmental. Raster electronic microscope photos of cheese samples taken at magni¬ fications between 6000 and 75000 showed, that in cheeses with brown spots propionibacteria were only found in the spots and not in the spot free zones. It could also be shown, that model Emmental cheese of good sensory and overall quality can be produced with various single wild P. freudenreichii strains. This was confirmed with three wild strains by manufacturing commercial Emmental cheeses. Summary / Zusammenfassung / Resume ^
Zusammenfassung
Charakterisierung von Propionsaurebakterien in Schweizer Roh¬ milch mit biochemischen und molekularbiologischen Methoden
Zur Herstellung von Emmentaler Kase in der Schweiz wird Rohmilch verwendet. Die Lochbildung wird jedoch erreicht durch den Zusatz von Kulturen ausgewahlter Propionsaurebakterien. Die wilden Propionsaure¬ bakterien in Rohmilch und Kase wurden bis heute nicht systematisch untersucht. Es ist jedoch bekannt, dass diese fur Fehler wie braune
Tupfen und Nachgarung in verschiedenen Hart- und Halbhartkasen ver- antwortlich sind, sowie dass technologische Faktoren bei der Kaseher- stellung das Auftreten dieser Fehler beeinflussen kfinnen. Die P-Kultur, eine Mischung von sechs P. freudenreichii-Stammen, wird seit vielen Jahren fur die Herstellung von Emmentaler Kase eingesetzt, und eine neue Kultur, Prop 96, ist seit 1996 erhaitlich. Die Stamme dieser beiden Kulturen wurden bis jetzt nicht mit molekularbiologischen Methoden un¬ tersucht.
In der vorliegenden Arbeit wurden die Propionsaurebakterien-Flora in Schweizer Rohmilch untersucht und die Propionsaurebakterien klassifi- ziert anhand der Analyse von Proteinprofilen, Restriktionsprofilen der 23S rRNA, des Plasmidgehalts und RAPD-Profilen. Aus verschiedenen Kasesorten mit Fehlern isolierte Propionsaurebakterien wurden ebenfalls differenziert, klassifiziert und mit den Stammen aus der Rohmilch vergli- chen.
Die Propionsaurebakterien-Flora in Schweizer Rohmilch wurde als au- sserst reich befunden. Alle vier milchwirtschaftlich relevanten Propioni¬ bacterium sp. wurden in Rohmilch aus dem Flachland nachgewiesen, 71% waren P. freudenreichii, 19% P. jensenii, 8% P. acidipropionici und 2% P. thoenii. In Alp-Rohmilch wurden keine P. acidipropionici gefunden,
55% waren P. freudenreichii, 15% P. jensenii und der P. fhoen/7-Anteil erreichte 30%. Auf Stamm-Niveau wurden unter den 278 P. freudenrei¬ chii mittels RAPD 219 (79%) verschiedene Stamme identifiziert. Die
Vielfalt der Stamme war bei den anderen Spezies sogar noch gr6sser. Von alien analysierten Stammen enthielten nur 30% Plasmide.
Die Propionsaurebakterien-Stamme aus P-Kultur und Prop 96 wurden mittels RAPD analysiert und das Vorhandensein der Stamme der P- 4o Summary / Zusammenfassung / Ftesumfe
Kultur wurde im Emmentaler nachgewiesen, um festzustellen, welche der Stamme im reifen Kase vorkommen. Von den sechs in der P-Kultur ein- gesetzten Stammen wurden vier fur identisch befunden, und folglich be- steht diese Kultur nur aus drei verschiedenen Stammen. Einer dieser
Stamme konnte nicht in kauflichem Emmentaler von erster Qualitat nachgewiesen werden, und einer war nur gelegentlich vorhanden. Folg¬ lich kfinnte moglicherweise nur einer der P-Ku/fur-Stamme fur die Kase- Qualitat verantwortlich sein. Die zwei Stamme der Prop 96-Kultur wurden als identisch identifiziert, aber sie waren verschieden von den Stammen der P-Kultur.
In den braunen Tupfen von Emmentaler und Sbrinz Kasen wurden mittels RAPD verschiedene P. freudenreichii-Stamme nachgewiesen. Die brau¬ nen Tupfen in Appenzeller enthielten auch P. jensenii und P. acidipro¬ pionici. In Greyerzer und Sbrinz Kasen mit Nachgarung wurden ver¬ schiedene P. freudenreichii-Stamme nachgewiesen. Nur einer der P. freudenreichii-Stamme aus Appenzeller, Greyerzer und Sbrinz mit Feh- lern war identisch mit einem der Stamme, die aus Rohmilch isoliert wur¬ den.
Mit ausgewahlten wilden P. freudenre/ch/7-Stammen wurden Modell- Emmentaler Kase produziert. Kase, die zwei- beziehungsweise dreimal mit demselben Stamm hergestellt wurden, waren von identischer Quali¬ tat. Chemische, mikrobiologische und molekularbiologische Methoden stellten keine Unterschiede fest zwischen Kasen von guter Qualitat und solchen mit einer Tendenz zur Nachgarung. Einige chemische und mi¬ krobiologische Unterschiede wurden jedoch zwischen Kasen von vorteil- hafter Qualitat und solchen mit braunen Tupfen gefunden. Der auffallig- ste Unterschied bestand zwischen den Wachstumskurven der Stamme, die Kase von guter Qualitat und denen, die Kase mit braunen Tupfen lieferten, indem die letzteren in Modell-Emmentalem ein langsameres Wachstum aufwiesen. Raster-elektronenmikroskopische Fotos von Ka- seproben zeigten bei einer Vergrosserung zwischen 6000 und 75000, dass Propionsaurebakterien in Kasen mit Tupfen nur in den braunen Tupfen selbst, nicht aber in den tupfenfreien Zonen anwesend waren. Es wurde zudem gezeigt, dass Modell-Emmentaler von guter sensorischer und Gesamt-Qualitat mit verschiedenen wilden P. freudenreichii- Einzelstammen hergestellt werden kann. Dies wurde auch mit der Her¬ stellung von Emmentaler Kase mit drei Wildstammen in einer gewerbli- chen Kaserei bestatigt. Summary / Zusammenfassung / Resume -5-
Resume
Caracterisation des bacteries propioniques dans le lait cru suisse avec des methodes de biochimie et de biologie moleculaire
Le lait cru est utilise pour la fabrication du fromage d'Emmental, mais la formation des yeux est atteinte par I'addition de cultures de souches de
bacteries propioniques selectionnees. Les bacteries propioniques sauva- ges n'ont pas ete analysees systematiquement jusqu'a present dans le lait cru et le fromage suisse. II est toutefois connu, que ces bacteries sont responsables de defauts, tels que les points bruns et la fermenta¬ tion secondaire, qui se manifestent dans differents fromages durs et mi- durs et que des facteurs technologiques dus a la fabrication du fromage, peuvent influencer I'apparition de ces defauts. La P-culture est un me¬ lange de six souches de P. freudenreichii. Elle est utilisee depuis long- temps pour la fabrication du fromage d'Emmental. Une nouvelle culture, nommee „Prop 96", est commercialisee depuis 1996. Les souches de ces
deux cultures n'ont pas ete analysees par des methodes de biochimie et de biologie moleculaire.
Le present travail a consiste a etudier la flore propionique dans le lait cru suisse et a classifier les bacteries propioniques grace a lanalyse des profils proteiniques, des profils de restriction du gene rARN 23S, du contenu plasmidique et des profils du RAPD. Des bacteries propioni¬ ques, isolees de differents types de fromages presentant des defauts ont
ete differenciees, classifies et comparees avec des souches du lait cru.
La flore propionique dans le lait cru de Suisse s'est averee etre extraor- dinairement riche. Les quatre especes de Propionibacterium ont ete trouvees dans le lait cru de plaine, 71% d'entre elles etaient des P. freu¬ denreichii, 19% des P. jensenii, 8% des P. acidipropionici et 2% des P. thoenii. Dans le lait cru d'alpage, aucune souche de P. acidipropionici n'a ete trouvee, et la proportion des P. freudenreichii a atteint 55%, celle des P. jensenii 15% et celle des P. thoenii 30% de toutes les bacteries propioniques. 219 souches differentes (79%) ont ete identifies par RAPD parmi les 278 P. freudenreichii. La diversite des souches a ete encore plus grande dans les autres especes. Parmi toutes les souches analysees, seules 30% contenaient des plasmides. -6- Summary / Zusammenfassung / Resume
La P-culture et la culture Prop 96 ont ete analysees par RAPD et la pre¬
sence des differentes souches de la P-culture a ete verifiee par cette
methode dans le fromage d'Emmental, afin d'etablir, si toutes les sou¬ ches s'etaient effectivement multiplies dans le fromage. Des analyses
ont permis de determiner que quatre des six souches de la P-culture
sont identiques, et que par consequent, cette culture ne consiste effecti¬ vement qu'en trois souches differentes. L'une de ces trois souches n'a
pas ete detectee dans I'Emmental commercial de premiere qualite, et
une autre n'a ete que parfois presente. Ainsi, il est possible que la quali¬ te du fromage ne soit influencee que par une seule souche de la P-
culture. Quant aux deux souches de la culture Prop 96, elles se sont
averees etre les memes, mais cependant differentes des souches de la P-culture.
Diverses souches de P. freudenreichii ont ete detectees par RAPD dans les points bruns de I'Emmental et du Sbrinz. Dans I'Appenzell, les points bruns contenaient en plus des P. jensenii et des P. acidipropionici. Dans les fromages Gruyere, Sbrinz et Appenzell avec fermentation secondaire, des souches differentes de P. freudenreichii ont ete detectees. Mais
seule une d'elles etait identique avec une souche isolee du lait cru.
Des fromages d'Emmental modeles ont ete fabriques avec quelques
souches sauvages de P. freudenreichii. Les fromages, fabriques plu- sieurs fois avec la meme souche, ont ete tous de meme qualite. Les methodes chimiques, microbiologiques et de biologie moleculaire n'ont decele aucune difference entre les fromages de bonne qualite et les fro¬
mages ayant une legere fermentation secondaire. Cependant quelques differences chimiques et microbiologiques ont ete trouvees entre les fro¬
mages de bonne qualite et les fromages contenant des points bruns. La
difference la plus nette a ete percue au niveau des courbes de crois- sance entre les souches qui fournissaient du fromage de bonne qualite et les autres. Dans les Emmental modeles, la croissance des souches
isolees des fromages contenant des points bruns a ete plus lente que celle des souches isolees des fromages de bonne qualite. Des photos
obtenues par microscopie electronique a balayage d'echantillons de fro¬
mage contenant des points bruns ont montre, a un agrandissement de
6000 a 75000 fois, que les bacteries propioniques ne se trouvaient que dans les point bruns. Ce travail a permis egalement de montrer qu'une
seule souche sauvage de P. freudenreichii etait necessaire pour fabri- Summary / Zusammenfassung / Resume -7-
de bonne tout quer des fromages modeles d'Emmental qualite generate particulierement sensorielle. Ceci a ete confirme dans une fromagerie souches commerciale par la production d'Emmental avec trois sauvages differentes. i 1, Introduction
1. Introduction and presentation of the problem
Propionibacteria are used in the Swiss cheese industry for the manufac¬ ture of Emmental (or Swiss-type) cheese to achieve the characteristic eyes and nutty flavour (Langsrud and Reinbold, 1973a and 1973b). At the same time propionibacteria occurring naturally in the milk seem to be responsible for defects such as brown spots and split defects in hard and semi-hard raw milk cheese. Some technological factors influencing the occurrence of those defects have been described, but still many ques¬ tions regarding propionibacteria involved still remain unanswered. No systematical analysis and classification of the propionibacterial flora in
Swiss raw milk has been performed until now. Also, the identification and classification of propionibacteria is still under discussion among experts, and various methods have been applied to resolve this problem.
In Switzerland, two propionibacterial cultures P-culture and Prop 96 are presently used in Emmental manufacture and the search for alternatives continues. A better insight into the strain composition of the commercially available cultures is important for their improvement. The impact of the commercially used strains on the natural propionibacterial flora of raw milk and their involvement in cheese quality as well as defects is not yet sufficiently understood.
In this study four aims were considered important:
milk should • The composition of the propionibacterial flora in Swiss raw be investigated. New and promising methods are necessary for the differentiation and classification of species and strains. Preferably, a rapid and reliable method to identify propionibacteria on genus and species level should be developed.
used in Emmen¬ • The propionibacterial cultures P-culture and Prop 96 tal cheese manufacture in Switzerland should be analysed by molecu¬ lar-biological methods and their presence in premium cheese investi¬ gated.
of • More knowledge about the involvement propionibacterial species and strains in the formation of brown spots and the occurrence of split
defect in cheese should be elaborated. 1 Introduction ^
• Selected wild Propionibacterium strains should be tested in the manufacture of model Emmental cheese and conclusions from the quality of the cheeses should be drawn for their suitability as alterna¬ tive to the existing cultures. Some wild propionibacterial strains should finally be tested in the manufacture of Swiss-type cheese on large scale. -.iQz 2, Literature
2. Literature
2.1. History and taxonomy of the genus Propionibacterium
As early as 1906, Von Freudenreich and Orla-Jensen isolated various bacteria from cheese, among them also bacteria producing propionic acid and proposed the designation Propionibacterium. Orla-Jensen (1909) isolated later propionibacteria from milk and described them in more detail.
The morphology of propionibacteria cells varies. They can be coccoid, oval or long. Cells may be single, in pairs forming a V, in chains, aggre¬ gates or various other forms. According to Cummins and Johnson (1986) they are nonmotile, nonsporing, gram-positive, generally catalase- positive and anaerobic to aerotolerant. Fermentation products are mainly propionic acid, acetic acid and C02. The G+C content of their DNA ranges from 65-67mol%. Fig. 1 shows the major lines of descent among the bacteria, with propionibacteria belonging to the gram-positive bacte¬ ria with a high DNA G+C content. Fig. 2 gives a view of phylogenetic groups of lactic acid bacteria and their relatives.
The size of the Propionibacterium genome varies from 2410 to 3060 kb (kilobases) depending on the species (Gautier et al., 1992). Also within the species, P. freudenreichii seems to vary to a large extent, as Rehberger (1993) found genomes with sizes varying up to 30%.
In 1930 the genus Propionibacterium comprising nine species emerged for the first time in „Bergey's manual of determinative bacteriology" (Bergey, 1930), later up to eleven different species of propionibacteria, all isolated from milk or milk products, were recognised. Today only four dairy species are recognised. In 1972 Johnson and Cummins reclassified the species by DNA-DNA hybridisation. This classification still exists to¬ day in the latest edition of „Bergey's manual of determinative bacteriol- ogy"(Cummins and Johnson, 1986). 2 Literature -11-
Gram-posithe bacteria low DNA G+C
Fiaolniclci utni C\U)pha Gram-positi\e bacteria high DNA G+C Nm ospua Deuwtota Green nonsulfur bacteria llicimotogales Aquife\ Fig.1: Major lines of descent among the Bacteria based upon compara¬ 16S rRNA se¬ tive analysis of about 3700 at least 90% complete quences (Schleifer and Ludwig, 1995). (Bar indicates 10% estimated sequence divergence.) J2= 2, Litergture OCIHHOU US Bifiilnhaaci mm Fig. 2: Major phylogenetic groups of lactic acid bacteria and their relatives among the gram-positive with low (upper part) and high (lower part) molar DNA G+C content based on at least 90% complete 16S rRNA sequences (Schleifer and Ludwig, 1995). (Bar indicates 10% estimated sequence divergence.) P. freudenreichii comprises the previous species P. globosum, P. orien- tum and P. coloratum, P. freudenreichii and P. shermanii. The debate as to whether or not to divide the species P. freudenreichii into two subspe¬ cies P. freudenreichii subsp. freudenreichii and P. freudenreichii subsp. shermanii is still under discussion, even though these subspecies are acknowledged by "Bergey's manual of determinative bacteriology" (Cummins and Johnson, 1986). Some authors are willing to accept a differentiation on the level of the subspecies (De Carvalho et al., 1994) and are suggesting the reitroduction of an additional P. globosum as a subspecies. Riedel et al. (1994) argue that DNA-DNA homology within the species P. freudenreichii is too high to justify a separation into sub¬ species. 2. Literature J£ The differentiation between the two or the three subspecies is entirely based on two criteria. P. freudenreichii subsp. freudenreichii is able to reduce nitrate and does not produce acid from lactose, whereas P. freudenreichii subsp. shermanii is unable to reduce nitrate and produces acid from lactose, whereas P. freudenreichii subsp. globosum possesses a nitrate reductase and has also the ability to ferment lactose. Colonies of P. freudenreichii are grey to white, but may be cream, tan or pink (Cummins and Johnson, 1986). P. jensenii includes the older species P. raffinosaceum, P. technicum, P. peterssonii and P. zeae. Colonies are white, cream or pink (Cummins and Johnson, 1986). P. thoenii comprises the previous P. rubrum and- P. thoenii. P. thoenii colonies vary in colour from brownish red to orange. In recent years the classification of P. rubrum has become controversial. It was placed under P. thoenii in 1972 by Johnson and Cummins, which was later confirmed by Baer (1987). Malik et al. (1968), Britz and Steyn (1980) and Britz and Riedel (1991) classified P. rubrum as P. jensenii. This classification was confirmed by De Carvalho et al. (1995), Riedel et al. (1994) and recently by Riedel and Britz (1996). P. acidipropionici comprises the older P. pentosaceum and P. arabino- sum. Colonies are grey or white (Cummins and Johnson, 1986). According to Charfreitag and Stackebrandt (1989) P. jensenii, P. thoenii and P. acidipropionici form a tight phylogenetical cluster, P. freuden¬ reichii is more remote. Propionibacteria isolated from human skin are classified into the species P. acnes, P. avidum, P. granulosum, and P. lymphophilum. These spe¬ cies do not occur in milk or milk products and will not be considered fur¬ ther. -AAz 2. Literature 2.2. Factors affecting growth of propionibacteria 2.2.1. Temperature In YEL (yeast extract lactate broth), which is generally used to culture propionibacteria, the optimal growth temperature is between 30°C and and 32°C et al In a (Hettinga Rembold, 1972) (Malik , 1968) study by Park et al (1967a), 80% of P freudenreichii, but only 40% of P thoenii, and 25% of P acidipropionici strains tested were able to grow at 7 2°C in YEL None of P jensenii strains used in this study grew at 7 2°C P freudenreichii seems to grow better at low temperatures than the other species, and this has an impact on cheese quality, as will be discussed later P freudenreichii also proved to be more heat resistant than the other three species According to Malik et al (1968) P freudenreichii strains were inactivated at a temperature of 62 8°C for 30min This is an impor¬ tant characteristic with regard to the utilisation of propionibacteria for hard and semi-hard cheese manufacture, where heating temperatures can exceed 50°C 2.2.2. pH Optimal pH for growth of propionibacteria is between 6 5 and 7 0, but they may survive at pH between 5 1 and 8 5 2.2.3. Effect of sodium chloride Propionibacteria survive a maximal concentration of 4 5% NaCI in YEL (Hettinga and Rembold, 1972), whereas concentrations below 1 5% NaCI had no effect on the growth or fermentation (Langsrud and Rembold, 1973a) In cheese the salt concentration is an important factor influenc¬ ing the growth of propionibacteria 2. Literature -15- 2.3. Metabolism Before 1970 propionibacteria were considered as strict anaerobes. Later, however, De Vries et al. (1972) demonstrated that they were also able to grow under aerobic conditions. The nutritional demands of propionibacteria are rather simple. As carbon source they can utilise different sugars. In milk the energy source is lac¬ tose, whereas in cheese lactate (Von Freudenreich and Jensen, 1906) and aspartate (Crow, 1986) are metabolised. 2.3.1. Propionic acid fermentation Propionibacteria convert lactate in cheese into propionic acid, acetic acid and C02, usually with a limited formation of succinate. L-(+)lactate is used preferably (Crow, 1986a). Under aerobic conditions P. freuden¬ reichii and P acidipropionici do not produce propionic acid from lactate, but only acetic acid (Pritchard et al., 1977). The main steps of propionic acid fermentation are the conversion of py¬ ruvate to oxaloacetate by carboxylation and then through succinate and succinyl-CoA to methylmalonyl-CoA and propionyl-CoA. From 3 moles lactate, two moles propioniate, and one mole of each acetate, C02, and HzO is formed. The net gain is one molecule of ATP (Metzler, 1977). 2.3.2. Metabolism of aspartate The fermentation balance is of more theoretical interest, because in an environment such as cheese other energy sources like aspartate are also available. Propionibacteria use both lactate and aspartate (Crow, 1986a, 1986b), and in the presence of aspartate the fermentation of lactate to propioniate is coupled with the fermentation of aspartate to succinate. According to Crow (1986b) aspartate is rapidly metabolised during the ripening of Emmental cheese. The theoretical propionate/acetate ratio established in the propionic acid fermentation is thus modified to the dis¬ advantage of propionate. The production of acetate, carbon dioxide and GLUCOSE ACETOINE Dl ACETYLS ACETOLACTATE CETALDEHYDE TPP^f NAD CoAeh CO .TPP 2 PYRUVATE .'pi^CQz oX. S^AMP^Ppl/ {£ f C7 S *)A0» !^NADH2 /—*PPI MALATE NAD MErHn.AMi.OWyi. C \. * |H20 (B12) PROPIONYbCoA Erylftllol CO . FUMARATE 2 / ADP'PI » FADH2 SUCCINYL-CoA O-arablnose / c p A-Qluconlque / NAD *ATP I FAD . <31P / SUCCINATE Glycerol --^^ GLUCOSE / Gal ou Lac NH3 COAah ASPARTATE (QTP) Fig. 3: Metabolism of propionibacteriaaccording to Dupuis (1994) 2. Literature J£ succinate is enhanced. Dupuis (1994) published the above fermentation scheme, which should reflect the conditions in cheese (Fig.3). 2.3.3. Proteinase and peptidase activity Propionibacteria are generally considered to have low proteolytic activity. Perez Chaia et al. (1988a) found the highest proteinase activity at the optimal growth temperature of 30°C. Lower activity was at 15°C for all Propionibacterium species and at 45°C for P. jensenii and P. acidipropi¬ onici. Optimum pH for activity was between 5.7 and 6.1, although P. acidipropionici strains were able to degrade casein in milk at pH 5.1. Du¬ puis et al. (1995) found a proteolytic system in all species of propionibac¬ teria. They measured two types of activity, one acting on B-casein and the other on Peptidase activity makes an important contribution to the flavour forma¬ tion of Emmental cheese. The production of a minimum amount of proline is needed for the typical nutty compound of the Emmental flavour (Hettinga and Reinbold, 1972; Langsrud and Reinbold, 1973c). Proline release by P. freudenreichii depends on pH and temperature (Langsrud et al., 1977). According to 0stlie et al. (1995), the proline iminopeptidase activity increase is correlated with autolysis of propionibacteria. In addi¬ tion propionibacteria harbour several other peptidases than the above mentioned proline iminopeptidase. Various amino acids can be de¬ graded, especially aspartate, alanine, serine and glycine depending on species and strain, and this results in additional C02 in cheese (Langsrud et al., 1995; Tobiassen et al., 1996. 2.3.4. Lipase and esterase activity Knowledge of the lipase and esterase activity of propionibacteria is rather limited. Oterholm (1967) measured a hundred fold higher lipolytic activity of P. freudenreichii subsp. shermanii compared to that of lactic acid bacteria. According to Dupuis et al. (1993) the highest lipolytic ac¬ tivity was found at 45°C and at pH 6.8. Between three to six different esterase activities were detected, two of them common to various strains of P. freudenreichii subsp. freudenreichii (Dupuis et al., 1993). Lipases -J3z 2. Literature and esterases may play a considerable role in fat hydrolysis and flavour composition of Emmental cheese. 2.4. Some other features of propionibacteria 2.4.1. Plasmids The presence of plasmids in propionibacteria has been described by dif¬ ferent researchers (Panon, 1988; Perez Chaia et al., 1988b; Rehberger and Glatz, 1990). The occurrence of plasmids in Propionibacterium strains varies between 25 and 38% depending upon the authors. Also the sizes of plasmids vary from 3kb to over 150kb. Not much is known about the characteristics coded on these plasmids. Rehberger and Glatz (1987) found some evidence, that the ability to ferment lactose of P. freudenreichii subsp. globosum strains might be plasmid-coded. Fermen¬ tation characteristics, pigment production, and growth characteristics do not seem to be plasmid-borne (Lyon and Glatz, 1993). 2.4.2. Bacteriophages The first description of a bacteriophage active on P. freudenreichii dates from 1992 (Gautier et al.) and suggests, that the phage might be used as a vector for DNA transfer within propionibacteria. Although only 18% of P. freudenreichii strains analysed were sensitive to phages, Gautier et al. (1995) found bacteriophages active against propionibacteria in at least 50% of Swiss-type cheese samples. All of the investigated cheeses made from raw milk and ripened under conditions similar to Emmental, contained bacteriophages. Phages occurred in propionibacteria selected as starters as well as in native strains from milk. They seem to proliferate during cheese ripening, and could be detected only, when propionibacte¬ ria reached 108 cfu/g. The authors showed, however, that in spite of the presence of bacteriophages in Swiss-type cheese, the coexistence with phage-sensitive propionibacteria hardly has an impact on the quality of the cheese. 2. Literature J& 2.4.3. Bacteriocins Bacteriocins are defined as proteins or peptides which are bactericidal to other, usually closely related bacteria (Klaenhammer, 1993). They have a narrow spectrum of activity and are mostly plasmid-borne. Lyon and Glatz (1991) purified and characterised a bacteriocin, PLG-1, from P. thoenii. This propionicin was active against P. thoenii and P jensenii, but showed no inhibitory effect against either subspecies of P. freudenreichii. PLG-1 was heat labile, sensitive to different proteolytic enzymes, and stable between pH 3 and 9. The bacteriocin showed inhibitory activity also against several gram-positive bacteria, including lactic acid bacteria, some gram-negative bacteria, as well as yeasts and moulds. The same authors characterised PLG-1 further (Lyon and Glatz, 1993) as not being inhibitory against the producer strain. The main accumulation of propi¬ onicin PLG-1 occurred during the stationary growth phase, and the pro¬ duction was not plasmid-coded. Another bacteriocin, jenseniin G, was isolated from P. jensenii (Grinstead and Barefoot, 1992), and was active against P. acidipropionici, P. jensenii, and L. delbrueckii subsp. lactis. Jenseniin G was active at pH 7, sensitive to proteolytic enzymes, heat stable, and resistant to freezing and cold storage. Similarly to PLG-1 it seems to be coded on the chro¬ mosome. Analysis of Microgard, a metabolite of P. freudenreichii subsp. sher¬ manii, inhibiting gram-negative food spoilage organisms and pathogenic bacteria suggested, that the active component is a bacteriocin (Al-Zoreky et al., 1993). It was heat stable and sensitive to proteolytic enzymes. Mi¬ crogard has found practical significance as conservation agent in cot¬ tage cheese (Daeschel, 1989). 2.4.4. Interactions with lactic acid bacteria The interaction between lactic acid bacteria and propionibacteria is an important feature in the manufacture of Swiss-type cheese. Controversial results of favourable and unfavourable impact on the two bacterial spe¬ cies involved have been published. Perez Chaia et al. (1994) reported that the rapid pH reduction by the growth of L. helveticus, the slow lac¬ tate utilisation by P. acidipropionici and, consequently, lactate accumula- -2Qz 2. Literature tion are the main reasons for the inhibition of propionibacteria in mixed cultures with lacobacilli. When propagating L helveticus and P. freuden¬ reichii subsp. shermanii together, a lower cell yield of lactobacilli but an increase in the fermentation activities of both microorganisms has been observed (Perez Chaia et al., 1987) and propionic acid production was increased. A mixed cultivation of L. acidophilus and P. freudenreichii subsp. shermanii also seems to be beneficial for the proliferation of both bacteria (Liu and Moon, 1982). All these observations were made in liq¬ uid cultures. Biede et al. (1977) reported that during cheese manufacture the addition of increasing amounts of L. bulgaricus led to a lower pH, lower levels of acetyl production and proteolysis and a decrease in the number of propi¬ onibacteria. Jimeno et al. (1995) showed that L rhamnosus and L. casei inhibited the growth of P. freudenreichii in cheese. The authors sug¬ gested that the relative amounts of copper and citrate in cheese could play an important role in this inhibition. The inhibition of P. freudenreichii growth could not be reproduced in cultures with the usual media. 2.5. Propionibacteria in practice 2.5.1. Natural habitat Few investigations on the natural habitat of dairy propionibacteria, ex¬ cept for milk and dairy products, have been made. All species have been isolated from soil and silage (Beerens et al., 1986), some species from fodder and dung (Mantere-Alhonen, 1977), but also from anaerobic fer- mentors (Riedel and Britz, 1993). Plastourgos and Vaughn (1957) iso¬ lated P. acidipropionici (P. pentosaceum) and P. jensenii (P. zeae) strains from spoiled olives. In the rumen Propionibacterium sp. are among other species responsible for urea breakdown and ammonia re¬ lease (Wallace and Cotta, 1988). The contamination of milk with propionibacteria occurs according to the experiences of the FAM (Schaeren, 1997) generally as follows: ru¬ men/intestine -> faeces -> milking machine -> insufficient cleaning and/or faulty installation of milking machine -> possible growth of propi¬ onibacteria in the milking machine between milking times -> milk. Thus, 2, Literature -21z the main focus of milk infection with propionibacteria seems to be the milking machine (Sollberger, 1996). An additional contamination of the milk with propionibacteria can according to the FAM (Thurlemann et al., 1991; Hani, 1991) happen in the cheese factory through insufficiently cleaned equipment and through the water supply. The average contami¬ nation was in Switzerland 250cfu/mL in supplier milk and 650cfu/mL in cheese milk (Thurlemann et al., 1991). In Italy the average contamination in supplier milk for Grana cheese manufacture was 720cfu/mL (Carcano etal., 1995). In the past, when cows and pigs were kept together in the same shed, the cheese defect referred to as „red spots" was more widespread. The P. rubrum strains responsible for the defect were according to Amrein (1997) mainly translated through skin contact between cows and pigs. It has not yet been investigated which role propionibacteria from others than animal sources play in the contamination of milk. 2.5.2. Use of propionibacteria Propionibacteria are important in Swiss-type cheese, where they produce the characteristic eyes and flavour (Langsrud and Reinbold, 1973a and 1973b). For a long time propionibacteria were used to produce vitamin B12 (Youngsmith and Apiraktivongse, 1983; Hatanaka et al., 1988; Ye et al., 1996). Pseudomonas strains have replaced propionibacteria in com¬ mercial vitamin B12 production, because they grow faster and yields are higher. The antimicrobial activity of the metabolites acetic and propionic acid, particularly of the later, is of interest in the conservation of food (Sobczak and Konieczna, E., 1987; Colomban, 1993). Other metabolites such as diacetyl and bacteriocins have also been implicated in the inhi¬ bition of Gram negative bacteria and pathogens (Al-Zoreky et al., 1993). Propionibacteria can also be used directly as food preservatives (Glatz, 1992), because the production of bacteriocins (Mantere-Alhonen, 1983; Grinstead and Barefoot, 1992) may play an important role. The use of propionibacteria cultures as natural preservatives has been suggested for the storage of corn with a high moisture content and in the prepara¬ tion of silage. -32z 2. Literature 2.5.3. Propionibacteria as probiotics Today, various bacteria are proposed as probiotics. Probiotics are de¬ fined as a live microbial feed supplement which beneficially affects the host animal by improving its intestinal microbial balance (Fuller, 1989). In order to exert a probiotic impact in the intestine, bacteria have to survive the passage through the stomach and duodenum and adhere to the in¬ testinal mucosae. The probiotic effect is based on the production of beneficial metabolites and antimicrobial compounds. As consequence, higher resistance to harmful bacteria responsible for intestinal disorders, stabilisation of the intestinal microflora, anticarcinogenic and anticholes- terolemic influences have been postulated. Propionibacteria were sug¬ gested as probiotics for humans and as probiotic growth promotors for pigs (Mantere-Alhonen, 1995). Propionibacteria were tested in combination with other bacteria such as lactobacilli, lactococci and enterococci for increasing weight gain of calves (Cerna et al., 1991) and curing intestinal disorders in calves, pig¬ lets and hens (Vladimirov et al., 1977; Antipov and Subbotin, 1980; Tuikov et al., 1980). According to Mantere-Alhonen (1982) P. freuden¬ reichii subsp. shermanii proved to be growth stimulating in piglets. How¬ ever, no colonisation or adhesion of propionibacteria to the intestinal mucosae of the piglets could be proven, although the resistance of P. freudenreichii to in vitro digestion was demonstrated (Mantere-Alhonen, 1983). It was suggested, that the high mineral contents (Mn, Zn, Cu, Fe) in propionibacteria could be partly responsible for the shown probiotical effect (Mantere-Alhonen, 1982; Mantere-Alhonen and Vuorinen, 1989). Perez Chaia et al. (1995) reported that dairy propionibacteria in the gut of mice favourably affected lipid metabolism and the immune system. Probiotic food products in which propionibacteria were combined with bifidobacteria or lactic acid bacteria, have also been described. A sour milk product was developed by Mantere-Alhonen and MSkinen (1987) containing P. freudenreichii subsp. shermanii and Lactobacillus acidophi¬ lus. The product showed favourable sensory quality, mild flavour, good consistency and storage properties. According to Mantere-Alhonen (1995) the effect of the product was tested with promising results. How¬ ever, technological problems may arise because propionibacteria need an incubation time of four to five days before the final inoculation into milk. 2. Literature £3: All the species of propionibacteria for the above trials were isolated from Jiving food". Living food is defined as fresh uncooked food, including vegetables, fruit, nuts, germinated seeds, beans and certain fermented products (Mantere-Alhonen and Ryhanen, 1994). 2.5.4. Propionibacteria in cheese manufacture 2.5.4.1. Swiss-type cheese Already in 1906 Von Freudenreich and Jensen recognised the impor¬ tance of propionibacteria in Emmental cheese. They thought, however, that a slight production of propionic acid and the formation of eyes in cheese might be possible without propionibacteria. Today, selected propionibacteria cultures of the species P. freudenreichii are added to milk in order to guarantee a controlled propionic acid fermentation and to obtain the desired eyes and flavour. With technical improvement of cheese manufacture the role of the natural propionic acid bacteria flora of raw milk seems to have become less important (Merilainen and Antila, 1976). Propionibacteria are also added in the manufacture of Leerdam- mer cheese (Britz and Riedel, 1994), but also occur in some other cheese types such as Appenzell, Tilsit and Sbrinz (Baer et al., 1993), Parmiggiano Reggiano and Grana Padano (Thompson and Marth, 1985; Carcano et al., 1995) and Mozzarella (Champagne and Lange, 1990). In the latter cheeses, however, the conditions for their proliferation are not favoured, thus propionibacteria do not contribute to ripening. 2.5.4.2. "Brown spots" defect A cheese defect called "brown spots" has been described in Swiss hard and semi-hard cheeses made from raw milk (Baer et al., 1993) and is caused by the formation of large colonies of P. freudenreichii. The de¬ velopment of this defect in Emmental cheese can be suppressed by adding increased amounts of propionibacteria to the milk. In other hard or semi-hard cheese such as Tilsit, Sbrinz and Appenzell, where propio¬ nibacteria are present in the raw milk, development of brown spots can be influenced by technological parameters. More intensive salting, in¬ creasing the copper or lactate content in young cheese, and addition of facultative heterofermentative lactobacilli can partially prevent the occur- -2£z 2, Literature rence of brown spots (Zaugg, 1995). Complete inhibition, however, re¬ quires salt concentrations of 15-20g/kg cheese (Sollberger, 1996), which diminishes the sensorial properties of cheese seriously. Brown spots ap¬ pear after three months of ripening, first in the peripheral area than in the whole cheese loaf. Brown-spotted cheese can not be sold as premium grade product, because consumers are unsure about the origin of brown spots and find them unappetising (Zaugg, 1995). "Red spots", attributed to red-pigmented P. rubrum colonies have also been found in Swiss cheeses, the defect, however, is very rare (Baer et al., 1993; Baer and Ryba, 1992). 2.5.4.3. Split defect A more commonly acknowledged defect caused by propionibacteria is the so-called "split defect" (Park et al., 1967b) or late fermentation in Emmental and other hard and semi-hard cheeses produced with raw milk. The height of the cheese loaf is increased and fissures or cracks in the cheese body may appear, while the cheese is being ripened in the cold cellar at approximately 10°C (Park et al, 1967b; Steffen, 1979). Propionibacteria strains able to grow at low temperature and to produce large amounts of C02 are held responsible for the split defect (Hettinga et al., 1974). The strains of propionibacteria involved have been reported to have different metabolic characteristics compared with strains not pro¬ ducing splits, in particular higher activity of various enzymes at lower temperatures and pH (Hettinga and Reinbold, 1975). Park et al. (1967b) reported that P freudenreichii subsp. shermanii strains show a greater tendency to split cheese than P. acidipropionici strains. They emphasised the importance of propionibacteria selection for cheese manufacture, although they stated that other factors than the Propionibacterium strain may play a role in late fermentation. Steffen (1979) found that Emmental with late fermentation differed from Emmental without split defects in a more intense proteolysis, and lower molecular weight compounds, a stronger propionic acid fermentation and a more rigid cheese body. The author concludes that the native micro¬ flora of raw milk and technological parameters have a share in the de¬ velopment of late fermentation. Hettinga et al. (1974) also observed a less elastic cheese body. Brendehaug and Langsrud (1985) blamed the 2. Literature 25; amino acid metabolism of propionibacteria as a possible cause for split defect and Sebastiani and Tschager (1993) held the fermentation of as¬ partate to succinate responsible. They as well as Crow and Turner (1986) saw a possible solution in selecting propionibacteria preferably utilising lactate. According to Reinbold (1978) some Streptococcus, Lactobacillus and Mi¬ crococcus strains may enhance propionibacterial growth through asso¬ ciative or symbiotic effects leading to secondary fermentation. Propionibacteria can also cause quality problems in Italian cheeses as reported by Carcano et al. (1993) studying late blowing of Grana cheese. Formaldehyde is commonly used in Grana cheese to suppress the un- desired cheese flora. This strategy, however, is not sufficient to eliminate split defect. Champagne and Lange (1990) isolated propionibacteria from Mozzarella, where they were presumably responsible for excessive gas production and a change in texture and flavour. 2.6. Methods for enumeration, identification, differentiation and classification of propionibacteria 2.6.1. Isolation and enumeration Foschino et al. (1988) stated, based on studies of the literature and their own observations, that propionibacteria are a very heterogenous group, especially regarding their adaptability to different cultural conditions and to pH variations. This adaptability seems to depend on the species, but may also vary within the same species. Consequently, the development of a selective medium for propionibacteria isolation and enumeration ap¬ pears to be difficult. Propionibacteria must be incubated anaerobically and have an excep¬ tionally long generation time. Von Freudenreich and Jensen (1906) cul¬ tured propionibacteria in a receptacle similar to a beer bottle containing a broth of peptone, potassium phosphate, sodium chloride and lactate lime. Kamber, Reinbold and Hussong (1952) proposed the use of sodium -2Sz 2. Literature thioglycollate to enhance selectivity. Malik et al. (1968) proposed a me¬ dium called yeast extract lactate agar (YELA). This medium is still com¬ monly used to isolate propionibacteria, although many other naturally occurring bacteria in milk are also able to grow on it. (Britz and Holzapfel, 1973; Drinan and Cogan, 1992; Thierry et al., 1994). Various researchers tried to increase selectivity of YELA by adding inhibitory compounds (Kurmann, 1962; Hettinga et al., 1968; Britz and Holzapfel, 1973; Peberdy and Fryer, 1976; Drinan and Cogan, 1992). These at¬ tempts, however, have not been very successful. The most recent sug¬ gestion for a selective medium for isolation and enumeration of propioni¬ bacteria comes from Thierry et al. (1994) and Thierry and Madec (1995). Their medium is, contrary to other media, based on glycerol and contains lithium and antibiotics. 2.6.2. Identification and classification based on biochemical methods Classical biochemical differentiation of propionibacteria is reported in „Bergey's manual of determinative bacteriology" by Cummins and John¬ son (1986). Propionibacteria are gram-positive and generally catalase- positive. They produce large amounts of propionic and acetic acids. Three main criteria are used to differentiate the four species as is shown in Table 1. Table 1: Main criteria for biochemical differentiation of lactic propioni¬ bacteria according to Cummins and Johnson (1986). Organism Fermentation of sucrose Reduction of B-Hemolysis and maltose nitrate P. freudenreichii - ± - P. jensenii + - - P. thoenii + - + P. acidipropionici + ± - A comprehensive table containing information on fermentations and other metabolic reactions useful to classify propionibacteria is also given by Cummins and Johnson (1986). These criteria are partially used in the 2. Literature -2L commercially available kit API 50 CH (API System S.A., France), which is generally recognised to be an efficient system for the determination of the carbohydrate utilisation patterns of propionibacteria (Britz and Riedel, 1991). The API 50 CH, however, does not have a key for identifi¬ cation of propionibacteria. De Carvalho (1994) reported difficulties in differentiating between P. jensenii and P. thoenii with the API kit. The cost of the kit should also be taken into account when analysing large numbers of strains. Kurmann (1960), Malik et al. (1968) and De Carvalho (1994) revealed serious inconsistencies in comparative taxonomy based on biochemical criteria. Numerical taxonomic studies, where each bio¬ chemical criterion has the same weight, seem to give better results (Britz and Riedel, 1991; De Carvalho, 1994). Nevertheless, classification of propionibacteria based on biochemical criteria does not permit a clear identification of strains (Britz and Riedel, 1991) and, in addition, is both time consuming and costly. 2.6.3. Identification and classification based on genomic criteria There is no official definition of the term .species". In „Bergey's manual of determinative bacteriology" (Staley and Krieg, 1986) a bacterial spe¬ cies is described as a collection of strains that share many features in common and differ considerably from other strains. Wayne et al. (1987) stated that DNA-DNA hybridisation was the best method to determine the species. This method gives an average result of similarity between bac¬ teria based on the comparison of the entire genome of two bacterial strains. In 1972, Johnson and Cummins regrouped dairy propionibacteria by DNA-DNA hybridisation into four groups. Bacteria with a DNA-DNA homology of 70% or more should be considered as belonging to the same species. De Carvalho (1994) differentiated propionibacteria using among other criteria also DNA-DNA hybridisation. With this procedure he was able to classify the studied strains in four distinct groups. Charfreitag and Stackebrandt (1989) determined the inter- and in- trageneric relationships of the genus Propionibacterium by analysis of the 16S rRNA sequences. They found that P. jensenii, P. thoenii and P. acidipropionici are closely related, while P. freudenreichii stands more apart. According to these authors the genus Propionibacterium repre¬ sents a well-defined taxon that stands isolated among other major groups of the actinomycetes. The phylogenetic neighbours of Propioni- -28z 2. Literature bacterium were determined in this study as Nocardioides and Terrabac- ter. The rRNA gene restriction patterns are used in the identification of bac¬ teria by ribotyping. De Carvalho et al. (1994) found that the four different species of dairy propionibacteria yielded different restriction patterns with species-specific bands. They were even able to differentiate P. freuden¬ reichii subsp. freudenreichii from P. freudenreichii subsp. shermanii. The patterns from propionibacteria were different from those of closely re¬ lated bacteria and other dairy bacteria. They judged ribotyping to be an excellent tool to identify propionibacteria. Riedel and Britz (1996) used the same technique to classify propionibacteria. They found the P. freudenreichii species to be homogeneous with both subspecies showing identical profiles. P. acidipropionici strains could be separated into two groups, older P. pentosaceum and P. arabinosum. A larger rRNA restric¬ tion patterns heterogeneity was observed for P. jensenii and P. thoenii strains. De Carvalho (1994) and Gautier et al. (1996) used DNA fingerprinting of propionibacteria by pulsed-field gel electrophoresis. This method is sensitive enough to differentiate between bacterial strains. RAPD-PCR (Random Amplified Polymorphic DNA-Polymerase Chain Re¬ action) has also been used to differentiate many species of gram-positive and gram-negative bacteria on an intraspecific or even strain level. RAPD-PCR is today currently used by many researchers and is an impor¬ tant tool in clinical epidemic research. De Carvalho (1994) showed RAPD-PCR to be an excellent tool to differentiate between propionibac¬ teria strains. Tailliez and Matte (1995) and Proudy et al. (1995) proved the RAPD technique to be discriminative on the species level. 2.6.4. Identification and classification by other methods Patterns of protein electrophoresis have been utilised to identify and classify various bacterial species. Baer (1987) and Baer and Ryba (1988) applied SDS-PAGE (SDS-polyacrylamide gel electrophoresis) to propio¬ nibacteria strains. It was possible to classify strains into the four species, and the method was proved to be more accurate than the classical bio¬ chemical determination. These findings were supported by Riedel and 2. Literature £3z Britz (1992) and Fessler and Puhan (1995). SDS-PAGE has been applied with success to other dairy bacteria (Tsakalidou et al, 1992; Zourari et al., 1992; Hertel et al., 1993; Descheemaeker et al., 1994; Patarata et al., 1994; Tsakalidou et al, 1994; Samalis et al., 1995). SDS-PAGE was enhanced by Baer and Ryba (1991) by immunoblotting of the gels. They were then able to differentiate between various strains of P. freuden¬ reichii. Britz and Steyn (1979) used pyrolysis-gas-liquid chromatography to identify propionibacteria. They were able to distinguish between the cu¬ taneous and the dairy species. However, the species and subspecies of the genus Propionibacterium could not be distinguished from each other by using pyrograms. Jakob (1995) classified cutaneous and dairy Propi¬ onibacterium species according to their methylated lipid components by gas chromatography. Differentiation on strain level was not possible. Baer and Ryba (1992) developed a serological method of agglutination and used it to identify the four propionibacteria species in Swiss dairy products. =2&: 3. Material and methods 3. Material and methods 3.1. Growth media for propionibacteria 3.1.1. Yeast extract lactate broth (YEL) and yeast extract lactate agar (YELA) Both media were proposed by Malik et al. (1968), and consisted of 10g/L Trypticase, 10g/L yeast extract, 10mL/L sodium lactate, 0.25g/L KH2P04, and 5mg/L MnS04, pH 7.0. YEL was modified to 24mL/L sodium lactate (50% v/v), 30g/L casein peptone, and 30g/L yeast extract, pH 6.8; YELA is YEL with additional 15g/L agar. 3.1.2. Minimum growth medium (MGM) MGM was prepared according to Crow (1986a) with modifications accord¬ ing to Baer (1996). 1L MGM/lactate consists of 788mL solution A, 10mL solution B, 2mL solution C, 100mL solution D, and 100mL solution E. In MGM/lactose solution E is replaced by solution F. Solution A is 0.5g KH2P04, 0.15g K2HP04, 0.15g MgS04-7H20, 0.015g MnCI2-4H20, 0.01g CoCI2, 0.01g FeS04-7H20, 0.01g NaCI, 0.1g L- tryptophan, 2.0mg panthotenic acid, 1.0mg p-aminobenzoic acid, 2.0mg nicotinic acid ad 800mL dist. H20, pH 6.8. Solution B is 100mg adenine, 100mg guanine, 100mg uracil, 100mg thiamine-HCI ad 100mL 0.05M HCI, sterilised by filtration. Solution C is 2mg riboflavin ad 100mL dist. H20, sterilised by filtration. Solution D is 5g casein peptone ad 500mL dist. H20. Solution E is 120mL sodium lactate (50% v/v) ad 500mL dist. H20, pH 6.8. Solution F is 25g lactose ad 500mL dist. H20. 3.1.3. MF95C MF95C was prepared according to Jimeno (1996). 1.4mL K2HP04 1mol/L, 3.5mL DL-lactic acid (88%), 1mL citric acid 1mol/L, and 0.37g CaC03 are made up to approximately 20mL with dist. water; 5.2mL KOH 1mol/l are added and the pH is adjusted to 5.0 with NaOH 5mol/L; 2g 3. Material and methods -31- tryptone and 1g yeast extract are added and the pH adjusted to 5.6 with NaOH 5mol/L; 2mL MgS04 1mol/L, 0.17mL ZnCI2, and 58uL CuCI2 are added and the pH adjusted with NaOH 5mol/L. The content of Na is cal¬ culated and adjusted to a total of 30mmol Na with NaCI. The volume is adjusted to 100mL and the solution sterilised by filtration. 3.1.4. Peptone whey Peptone whey (PS) is a medium developed by the microbiological de¬ partment of the Federal Dairy Research Institute, Switzerland. It consists of 10g/L casein peptone, 3g/L yeast extract, 143mL/L Rivelia concentrate (Miroma AG), and 0.5mL/L anti foam agent (Dow Corning 1510, food grade, Pluss-Stauffer AG). The pH is adjusted to 6.8 ± 0.2. 3.1.5. Dilution solution Dilutions of liquid cultures were made in tubes containing 9mL of the fol¬ lowing solution: 8g/L NaCI, 1g/L casein peptone, pH 7.2±0.2. 3.2. Media for other bacteria RCM broth, MRS and MRS-agar, M17 and M17 with 0.05M lactose, MS- and KM-agar are described by Atlas (1993). LS5 and LS5 with 0.05M lactose were developed by Casey and Meyer (1985) and FH-agar by Isolini et al. (1990). 3.3. Microorganisms 3.3.1. Propionibacteria Most of the as propionibacteria isolated strains from raw milk were pro¬ vided by MIBD/SICL (Milchwirtschaftliche Inspektions- und Beratungsdi¬ enste, Dairy Inspection and Consulting Services)-laboratories on YELA. The number of strains and the laboratories were: :22: 3. Matenal and methods MIBD Aargau (AG) with 72 strains MIBD Bern (BE) with 31 strains SICL Fribourg (FR) with 99 strains MIBD Nordostschweiz (NO) with 36 strains MIBD Nordwestschweiz (NW) with 48 strains MIBD St.Gallen-Appenzell AR (SG, AR) with 33 strains MIBD Thurgau (TG) with 45 strains SICL Vaud-Geneve (VD) and SICL Neuchatel (NE) with 17 strains MIBD Zentralschweiz (ZS) with 72 strains (Fig. 6). Reference strains (20)1 from ATCC (American Type Culture Collection, Rockville, Md.), DSM (Deutsche Sammlung von Mikroorganismen und Zellkulturen GmbH) and the Collection of Microorganisms of the FAM (Federal Dairy Research Institute, Switzerland) and wild strains from the Collection of Microorganisms of the FAM (27) were received freeze dried. Other supposed propionibacteria were isolated from alpine milk (21) and Emmental cheese with and without brown spots (39), Appenzell (43), Raclette (10), and Sbrinz (72) with brown spots and Gruyere (46) and Sbrinz (50) with splitting defect. Propionibacteria from cheeses with brown spots were isolated from the spots by inoculating as much of each spot as possible without including the surrounding region in YEL under sterile conditions. From Emmental without fault and from Gruyere and Sbrinz with splitting defect 10g of cheese were sampled under sterile conditions and mixed thoroughly with 90mL peptone water (consisting of 10g/L casein peptone, 5g/L NaCI, and 20g/L tri-sodium citrate di-hydrate, pH 7.3 ± 0.2). 0.2mL of this mixture were spread on YELA. All strains were isolated in three subsequent steps on YELA inoculated for 10 days at 30°C under anaerobic conditions and subcultured subse¬ quently in YEL for 5 to 7 days at 30°C. Strains were stored for short-term utilisation in YEL at 4°C, and for long-term utilisation in sterilised recon¬ stituted powder milk at -70°C. The aspect of each colony of all supposed propionibacterial strains was described, their morphology observed under the microscope (Wild M20, 1 number of strains is given in brackets 3, Material and method? ^ Heerbrugg) at 100x magnification and their catalase activity was tested with 3% (v/v) H202. The name, origin and description of supposed propionibacteria strains are given in Tables 1-17, Annex. 3.3.2 Other bacteria Other bacteria were provided by CNRZ (Centre National de recherches Zootechniques, Jouy-en-Josas, France), DSM and FAMb (Biochemistry, Federal Dairy Research Institute, Switzerland) (Table 18, Annex). 3.4. Sugar fermentation The differentiation of Propionibacterium sp. was carried out according to Cummins and Johnson (1986) with medium MGM. Sodium lactate solu¬ tion E was replaced by either a solution of 5g/L trehalose, 5g/L mannitol or 5g/L arabinose. 3.5. Growth curves Growth curves of propionibacteria were determined on microtiter plates. 17u.L of liquid bacterial culture were mixed with 170u,L MF95C in each well of the plate, so that the absorbance at 650nm was approximately 0.2. The plates were kept in an anaerobic jar at 30°C and the absor¬ bance was measured in an Uvikon photometer at 650nm once a day for 13 days. The repeatability for 18 growth curves of the P-culture was 94%. Bacteria were allocated according to their growth rate and their growth maximum into seven groups: zero, hardly, slow/weak, slow/medium, slow/high, fast/weak, and fast/high (Table 19, Annex). 3.6. Survival rate at cheese manufacture conditions Propionibacteria strains intended for model cheese manufacture were submitted to heating conditions similar to those encountered in cheese =34= 3 Material and methods manufacture: - 32°C for 30min - scalding from 32°C to 53°C over 30min - 53°C for 45min - scalding from 53°C to 32°C over 24h. Propionibacteria were grown in YEL for 5 days and plated on YELA for counting. They were either incubated in YEL or in sterile milk during the heating process and then again plated on YELA for counting. 3.7. Analysis of the soluble cell free protein extracts 3.7.1. Preparation of protein extracts 10mL of a liquid culture were centrifuged for 10min at 4°C and 9990g in a Sorvall centrifuge and the sediment suspended in 150uL TES-buffer (10mM Tris-CI, pH7.5, 1mM EDTA, 100mM NaCI). Approximately 300mg glass beads (BioSpec Products, 11079101) were added and the sus¬ pension was vortexed for five times one minute with a cooling period of one minute on ice between each vortexing period. Glass beads and in¬ tact bacterial cells were removed by centrifugation for 10min at 450g on a MSE centrifuge. The protein concentration of the supernatant was de¬ termined according to the method of Lowry et al. (1951). 3.7.2. Protein gel electrophoresis and staining of gels The resolving gel contained 33% acrylamide stock solution (36% (w/v) acrylamide, 0.8% (w/v) N,N'-methylene bisacrylamide), 50% Tris stock solution (0.75M Tris, 0.2% SDS (w/v), pH 8.8), 0.05% ammonium persul¬ phate (w/v), and 0.05% TEMED. The stacking gel contained 10% acry- lamid stock solution, 50% Tris stock solution (0.25M Tris, 0.2% SDS, pH 6.8), 0.05% ammonium persulphate (w/v), and 0.1% TEMED. The final acrylamide concentration of the gel was 12%. Protein extracts were di¬ luted to 2.13mg/mL in sample buffer (12.5mL Tris stock solution pH 6.8, 2.5g SDS, 5.0mL glycerine, 5.0mL DTT 100mM, 0.5mL bromphenol blue 1% ad 50mL dist. water) and heated at 100°C for two minutes. 10u.L of this solution were loaded into the wells. Electrophoresis was run in a Protean II Slab Cell purchased from BioRad at 50mA for the stacking gel 3. Material and methods ^ and 70mA for the resolving gel at 15°C in Tris glycine buffer (0.025M Tris, 0.192M glycine, 0.1% (w/v) SDS. Gels were fixed overnight in a solution of 10% acetic acid, and 30% ethanol followed by 3 times 20min in 30% ethanol. Gels were washed twice for 15min in distilled water, dipped in 0.025% sodium dithionite for 1min, and washed twice for 1min in distilled water. Gels were stained for 20 to 30min in silver nitrate solution (0.2% AgN03, 1mM HCHO) and washed for 1min in destilled water. The gels were developed in sodium carbonate solution (6% Na2C03 6mM HCHO, 20u.M Na2S203. Develop¬ ment was stopped by adding 4% acetic acid to the sodium carbonate solution. Gels were finally washed 4 times 30min in distilled water and dried in a 2003 Slab Gel Dryer from LKB. The gels were photographed with a Sony CCD camera and the patterns analysed with GelCompar from Kortrijk, Belgium. Profils were classified using UPGMA (unweighted pair group method using arithmetic averages) (Sokal and Michener, 1958). 3.8. Purification of genomic DNA Centrifugation of 10mL liquid culture, 10min at 9990g (Sorvall centrifuge) ; Suspension of sediment in 9mL buffer A (100mM sodium borate, 10mM EDTA, 25% saccharose, pH 8.0) 4- Addition of 1mL lysozyme-solution, (20mg/mL in buffer A), mixing 4- Incubation, 1h at 37°C ; Centrifugation, 10min at4°C and 4430g (Sorvall centrifuge) I Suspension of sediment in 4.75mL 20mM EDTA, pH 7.0 I Addition of 0.25mL 20% SDS I Incubation, 15min at room temperature I Addition of 1.25mL 5.0M sodium perchlorate, mixing thoroughly =3& 3 Matenal and methods I Addition of 3.0mL chloroform-solution (1 part amyl-alcohol, 24 parts chloroform), mixing thoroughly I Incubation, 15mm at room temperature Centrifugation, 10min at 9990g (Sorvall centrifuge) I Addition of 3mL propanol to 5mL of the upper phase I Incubation, 15min at room temperature I Centrifugation, 10min at 9990g (Sorvall centrifuge) I Dissolving sediment in 200ul TE buffer (10mM Tris-CI, 1mM EDTA, pH 8.0) J, Addition of 5u.L RNase (DNase-free)1 J, Incubation, 30min at 37°C I Addition of 1u.L Proteinase K (Merck, 1.07393) I Incubation, 30min at 37°C J, Centrifugation, 10min at 9990g (Sorvall centrifuge) I Dissolving sediment in 200u,L TE buffer 3.9. Restriction enzyme analysis 17u.L DNA-solution were mixed with 153u.L TE buffer and absorption was measured at 260nm (absorption 1.0 corresponds to 50u.g/mL DNA) and DNA concentration was adjusted to 100u.g/mL. 14u.L dist. water, 4u.L 1 dissolving 10mg RNase A (Serva, 34390) in 1mL 20mM EDTA, pH 7 0 I Incubation, 15mm at 100°C 4 Cooling slowly down to room temperature, stocking at -20°C 3. Material and methods ^Iz buffer (10xconcentrated, Boehringer Mannheim), 20uL DNA 100u,g/mL, 1uL BSA 20mg/mL (bovine serum albumin, Boehringer Mannheim, 711454), and 1uL restriction enzyme (all purchased from Boehringer Mannheim) were mixed in a Eppendorff tube and incubated for 2h at 37°C. 22uL of this solution were run overnight on an 0.8% agarose gel (Bio-Rad, 162-0126) at 50V in TBE-buffer (90mM Tris, 90mM borate, 2mM EDTA, pH 8.4-8.5. The restriction fragments were visualised under UV light in the presence of ethidium bromide buffer (10mM sodium bo¬ rate, 1mg/L ethidium bromide, pH 8.0). Photographs were taken with a Polaroid MP4 Land Camera. 3.10. Ribotyping DNA was extracted as described in chapter 3.6., and the DNA concen¬ tration adjusted to 1000u;g/mL. The preparation of the probe and the whole ribotyping procedure including blotting, hybridisation and colouring were carried out according to Hawcroft and Geary (1996). 3.11. Restriction analysis of 23S rRNA 3.11.1. Preparation of DNA extracts The extraction of purified DNA was performed as summarised in 3.8.. For the preparation of crude DNA extracts the following method was devel¬ oped: From liquid cultures: 0.4mL of a liquid bacterial culture were centrifuged for 10min at 7240g in a Hettich Mikroliter centrifuge under sterile condi¬ tions. The sediment was suspended in 100u.L sterile destilled water. The solution was boiled for 5min and then centrifuged for 15min at 7240g. 2ul of the supernatant were used for the PCR-reaction. From colonies: Bacterial material (1 colony) was suspended in 100u.L sterile distilled water. The solution was boiled for 5min and then centri¬ fuged for 15min at 7240g. 2jjL of the supernatant were used for the PCR- reaction. :2S: 3. Material and methods 3.11.2. Polymerase chain reaction and restriction analysis Taq polymerase, PCR buffer and dNTP mix were purchased from Boe¬ hringer Mannheim (1578 553) and the primers from MWG-Biotech (Ebersberg, Germany). The PCR-reaction mix contained 83.5u.L sterile distilled water, 10u.L PCR buffer, 2ul dNTP mix, 1uL each of primers 5'- MADGCGTAGNCGAWGG-3' and 5'-GTGWCGGTTTNBGGTA-3' (Roller et al., 1992) 100nmol/mL, 0.5ul Taq DNA polymerase and 2u.L crude or purified DNA. PCR amplification was performed on a thermo-cycler from Inotech as follows: - 25 cycles: 95°C for 1min 52°C for 2min 72°C for 3min - one final step at 72°C for 5min 1u.L Msp I (Boehringer Mannheim, 633 526) and 1u.L BSA (bovine serum albumin, Boehringer Mannheim, 711 454) were added to 20u,L of the PCR amplification product and incubated for 2h at 37°C. Both restricted and untreated PCR amplification products were separated on a 4% aga¬ rose gel (NuSieve 3:1 agarose, FMC, 50092) and visualised under UV light in the presence of ethidium bromide buffer (10mM sodium borate, 1mg/L ethidium bromide, pH 8.0). DNA size standard was 50-2000bp ladder from Bio-Rad with bands of 2000, 1500, 1000, 700, 500, 400, 300, 200, 100 and 50bp. 3.12. RAPD RAPD (Random Amplified Polymorphic DNA) with primer SK2: The ex¬ traction of DNA was as summarised under 3.6.1. Taq polymerase, PCR buffer and dNTP mix were purchased from Boehringer Mannheim (1578 553) and the primer SK2 5'-GCC GCC GCC GCC-3' (De Carvalho, 1994) from Microsynth (Balgach, Switzerland). The PCR-reaction mix contained 83.5u.L sterile distilled water, 10u.L PCR buffer, 2^L dNTP mix, 2u.L each of primer SK2 100nmol/mL, 0.5u.L Taq DNA polymerase and 2u,L purified DNA. 3. Material and methods -M: PCR amplification was performed on a thermo-cycler from Inotech (Dottikon, Switzerland) as follows: - one cycle of denaturation at 94°C for 4min - 45 cycles: 94°C for 1min 43°Cfor 1min 72°C for 2min - one cycle of elongation at 72°C for 10min. RAPD with primer DF4: The extraction of DNA was as summarised under 3.6.1.. RAPD analysis beads were purchased from Pharmacia Biotech (27-9500-01) and the primer DF4 5'-CGC CGC CGT CGC-3' from MWG Biotech. The PCR-reaction mix contained one analysis bead, 24u.L sterile distilled water, 0.6u.L of primer DF4 50nmol/mL, and 0.4u.L purified DNA. PCR amplification was performed on a thermo-cycler from Inotech as follows: - one cycle of denaturation at 94°C for 4min - 25 cycles: 94°C for 1min 43°Cfor 1min 72°C for 2min - one cycle of elongation at 72°C for 10min. Five volumes of the PCR amplification products of both primers were mixed with one volume of loading buffer (0.25% bromphenol blue, 15% Pharmacia Ficoll 400), separated on a 4% agarose gel (NuSieve 3:1 aga¬ rose, FMC, 50092) and visualised under UV light in the presence of ethidium bromide buffer (10mM sodium borate, 1mg/L ethidium bromide, pH 8.0). DNA size standard was 50-2000bp ladder from Bio-Rad with bands of 2000, 1500, 1000, 700, 500, 400, 300, 200, 100 and 50bp. Photos were taken with a Sony CCD camera and the patterns analysed with GelCompar from Kortrijk, Belgium. Profils were classified using UPGMA (unweighted pair group method using arithmetic averages) (Sokal and Michener, 1958). -40- 3 Material and methods 3.13. Plasmids The method was carried out according to Casey (1994): Centrifugation of 10mL liquid culture, 10min at 9990g (Sorvall centrifuge) I Suspension of sediment in 9mL buffer A (100mM sodium borate, 10mM EDTA, 25% saccharose, pH 8.0) 4- Addition of 1mL lysozyme-solution, (20mg/mL) and 100uL mutanolysin (2000u/mL, Sigma 9901), mixing i Incubation, 1-2h at 37°C i Centrifugation, 10min at 4430g (Sorvall centrifuge) 4- Suspension of sediment in 0.5mL distilled water 4- Addition of 0.5mL sodium borate buffer (0.1M sodium borate, 10mM EDTA, 1% SDS, pH 12.6), J, Mixing gently and thoroughly, adjusting at pH 12.4 4- Incubation, 15min on ice i Addition of 0.5mL 5.0M sodium acetate buffer of 4°C, pH 4.7, mixing thoroughly J, Incubation, at least 1h on ice 4- Centrifugation, 20min at 4°C and 9990g (Sorvall centrifuge) I Addition of 0.72mL propanol to 1.2mL of the upper phase (in an Eppendorff tube) J, Incubation, 15min at room temperature 4- Centrifugation, 30min at 7240g (Hettich Mikroliter centrifuge) Washing sediment with 70% ethanol 3. Matenal and methods -£\- I Dissolving sediment in 79uL distilled water 4- Addition of 10u.L buffer (330mM Tris acetate, 660mM potassium acetate, 100mM magnesium acetate, 5mM DTT, 0 1% Triton X-100, pH 7 8) I Centrifugation, 15mm at 7240g (Hettich Mikroliter centrifuge) I Addition of 1ul Plasmid-Safe (Epicentre Technologies E3101K), 10u.L ATP 10mM and 5u.L RNAse (DNase-free)1 4- Incubation, 30mm at 37°C 4- Addition of 20uL loading buffer (0 25% bromphenol blue, 15% Pharmacia Ficoll 400) 45u.L of this solution were run overnight on an 0 8% agarose gel (Bio- Rad, 162-0126) at 50V in TBE-buffer (90mM Tris, 90mM borate, 2mM EDTA, pH 8 4-8 5 The plasmids were visualised under UV light in the presence of ethidium bromide buffer (10mM sodium borate, 1mg/L ethidium bromide, pH 8 0) DNA size standard was supercoiled DNA lad¬ der from Sigma (D5292) with bands of 16 210, 14 174, 12 138, 10 102, 8 066, 7 048, 6 030, 5 012, 3 990, 2.972, and 2 067kb Photographs were taken with a Polaroid MP4 Land Camera 3.14. Emmental cheese 3.14.1. Model Emmental cheese manufacture Small sized Emmental cheeses from pasteurised milk were produced in the pilot plant manufactory of the FAM Plan for manufacture of model Emmental 1 dissolving 10mg RNase A (Serva, 34390) in 1mL 20mM EDTA, pH 7 0 Incubation, 15mm at 100°C i Cooling slowly down to room temperature stocking at -20°C -42- 3 Matenal and methods milk quantity 120L (pasteurised at 72°C for 15s) H20 10L copper sulphate 13mL perscalding of kettle milk 31°Cfor 30min desired acidity of cultures 30°SH/45°SH quantity of cultures 100mL/30mL coagulation 32°C for 35mm desired acidity of cultures 30°SH/45°SH quantity of cultures 30mL/ 100mL rennet extract 19mL propionibacteria 0 25 drops (final concentration approxi¬ mately 1 5x103cfu/mL milk grown in peptone whey) H20 2L precheesing 35mm grain size 4-8mm scalding 53°C for 30mm (+3L H20) stirring 40mm at 53°C pressing room 50°C for 2h, ramping from 50°C to 35°C over 5h, 35°C for 10h, ramping from 35°C to25°Cover1h, 25°Cfor4h pressure 8bar for 0 25h, 10bar for 5h, then 4bar cellar salting for 1d, cellar 12-14°C for 14d, cellar1 21-23X for 61-80d, cellar2 11-14°C for 26-45d Eight cheeses were produced in parallel per day, one cheese being the reference cheese containing the P-culture3 with the strains P1409, P1410, P1411, P1412, P1413, and P1414 (Tables 1 and 19, Annex) The 1 at 80% relative humidity 2 at 85-90% relative humidity 3 one of the two cultures used all over Switzerland for Emmental manufacture 3. Material and methods ^2z other cheeses were produced with a single propionibacterial strain. Three cheeses were produced with: AG53.1, AG73.1 (Tables 2 and 20, Annex) BE17.2 (Tables 3 and 21, Annex) FR4.1, FR21.2, FR22.2 (Tables 4 and 22, Annex) N023.1, N024.1 (Tables 5 and 23, Annex) NW8.1 (Tables 6 and 24, Annex) SG24.1, SG31.1 (Tables 7 and 25, Annex) TG38.1, TG41.1 (Tables 8 and 26, Annex) VD9.1 (Tables 9 and 27, Annex) ZS26.1, ZS29.1 (Tables 10 and 28, Annex) 209.1 (Tables 11 and 29, Annex). Two cheeses were produced with: FR26.1 (Tables 4 and 22, Annex) NW2.1, NW5.1 (Tables 6 and 24, Annex). Nine cheeses were manufactured with P-culture. A total of 66 model Emmental cheeses were manufactured. 3.14.2. Analysis and sensoric tests performed on model cheeses Cheeses were analysed chemically for their water content, pH, and L- and D-lactic acid on day 1 and 120; for their fat, NaCI, formic acid, vola¬ tile fatty acids, acetic acid, propionic acid, iso-butyric acid, n-butyric acid, iso-valerian acid, iso-capronic acid, n-capronic acid, succinic acid, and amino acids content on day 120. All these analyses were performed ac¬ cording to the Quality Security norms of the FAM. Microbial analyses included counting of enterococci on KM-agar, of salt tolerant germs on MS-agar, and of propionibacteria on YELA. These counts were performed on day 1, 30, 60, 80, and 120 after cheese manufacture. Exterior and sensory quality of the cheeses were judged 120 days after manufacture according to their exterior and sensory quality by a panel of three people. _=44L 3. Material and methods The best cheeses were submitted to triangle test according to DIN 10951. Approximately 12 people of the accredited sensory panel of the FAM attempted to differentiate the taste of cheese produced with a sin¬ gle strain or that of the reference cheese produced on the same day. They were also requested to state their personal preference for either the single or the pair of cheeses. 3.14.3. Emmental cheese manufacture Normally sized Emmental cheeses were manufactured from raw milk in the manufactory Uettligen. Plan for manufacture of Emmental: evening milk temperature 12-14°C milk quantity 1000L(raw) H20 10% prescalding of kettle milk 32°C for 30min desired acidity of cultures 30°SH/50°SH quantity of cultures 1-1.5%o (quantity of facultative (0.2%o) heterofermentative culture) coagulation 32°C for 54-58r quantity of cultures 1.5-2%o rennet extract 140mL propionibacteria 5 drops (final © 3x10 cfu/mL milk, grown in peptone whey) H,0 10% precheesing 30min grain size 4-6mm scalding 53°C for 30min (+3L H20) stirring 50min at 53°C pressure ramping from 0.1 to 0.5bar for 19-21h 3. Material and methods' ^ cellar salting for 3d, cellar 13-16°C for 14-16d, cellar1 22-24°C for 1-2months cellar2 12-14°C for 2-3months Four cheeses were produced in parallel in a day, one cheese being the reference cheese containing the P-culture with the strains P1409, P1410, P1411, P1412, P1413, and P1414 (annex 1 and 2, table 1). The other cheeses were produced with a single propionibacterial strain, either NW8.1 (annex 1 and 2, table 6), TG38.1 (annex 1 and 2, table 8), or ZS26.1 (annex 1 and 2, table 10). The four different propionibacteria cultures were or were not combined with facultative heterofermentative culture (MK 3008, provided from the FAM). Each variant was produced twice; there were a total of 16 cheeses. Cheeses were analysed chemically for their water content and pH on day 1 and 150; fat, free fatty acids, L- and D-lactic acid, and L-leucin amin- opeptidase on day 150 according to the Quality Security norms of the FAM. The exterior and sensory quality of the cheeses were judged 5 months after manufacture by a panel of four experts. 1 at 80% relative humidity 2 at 85-90% relative humidity 4§z 4 Results and discussion 4. Results and discussion 4.1. Propionibacterial flora in Swiss raw milk from the lowlands and the alps 4.1.1. Introduction Different methods were used to identify the 453 supposed propionibacte¬ ria strains isolated from raw milk from the Swiss lowlands by MIBD labo¬ ratories and the 21 strains from alpine milk isolated during this work. Identification of species was performed by protein profile analysis and confirmed by restriction analysis of 23S rRNA. Further classification of P. freudenreichii strains into the subspecies P. freudenreichii subsp. freudenreichii or P. freudenreichii subsp. shermanii was performed by determining growth on MGM with lactose. P. freudenreichii subsp. freudenreichii only utilises lactate, whereas P. freudenreichii subsp. shermanii utilises both lactate and lactose. For identification at the strain level 23S rRNA restriction analysis proved to be insufficient. Protein profile analysis show slight differences visually, which, however, can not be evaluated by the GelCompar program. Therefore, ribotyping, total genomic DNA restriction analysis, plasmid content of strains, and RAPD were tested as methods for strain identification. 4.1.2. Reproducibility of protein profiles and restriction analysis profiles No systematic analysis of reproducibility of protein profiles was per¬ formed, because other researchers using SDS-PAGE reported a repro¬ ducibility of this technique at 92-98% (Costas, 1990; Costas et al., 1990), 95.9 ± 1.7% (Costas et al., 1994), and 93-97% level (Tsakalidou et al., 1994). With 23S rRNA restriction analysis a reproducibility of 100% was achieved. This method has the advantage that with propionibacteria only three or four bands appeared and the species were, therefore, easily recognised. As shown in the Annex (Tables 19 and 29), of the total of 225 strains submitted to 23S rRNA restriction analysis from Swiss raw milk (including strains from alpine milk and type strains), only six (2.7%) were classified differently by 23S rRNA restriction analysis than by pro- 4. Results and discussion ^LL tein profile analysis. Three strains (FR3.1, FR407.1, ZS47.1) were not recognised as Propionibacterium sp. by protein profile analysis, even though they could be classified by 23S rRNA restriction analysis as P. acidipropionici ( FR3.1, FR407.1) and P. freudenreichii (ZS47.1). Three other strains (FR51.1, ZS14.1, ZS16.1) were identified as P. thoenii by protein profile analysis, but as P. jensenii by 23S rRNA restriction analy¬ sis. Because the evaluation of 23S rRNA restriction patterns is easier and the reproducibility better, the results obtained by restriction analysis of the 23S rRNA were preferred to those obtained by protein profile analysis. 4.1.3. Propionibacterium species in lowland milk Of the total 453 strains received from MIBD laboratories, 74 were identi¬ fied as not belonging to propionibacteria, 267 were identified as P. freudenreichii, 32 as P. acidipropionici, 72 as P. jensenii and 8 as P. thoenii (Tables 20-28, Annex). The identification was based on compar¬ ing the protein profiles with profiles of the type strains. For P. jensenii, P. acidipropionici, P. thoenii, and bacteria other than propionibacteria the results were confirmed with restriction analysis of the 23S rRNA gene. Fig. 4a shows the protein profiles of the four Propionibacterium type strains, which were used for comparison, Fig. 4b shows an example of a dendrogramm of 45 protein patterns from the region AG and four type strains. Fig. 5a shows the amplified fragments of the 23S rRNA gene of the four Propionibacterium type strains and Fig. 5b the patterns gener¬ ated after restriction with Msp I. In cases where the protein profile analysis was not conclusive enough for the classification of an isolated strain, restriction analysis of the 23S rRNA gene was performed. 16% of the strains originally identified as propionibacteria (based on morphology and/or catalase activity, Tables 2-10, Annex) were found not to belong to this genus when analysed by protein profile analysis and restriction analysis of their 23S rRNA. The test based on restriction analysis of 23S rRNA performed on crude bacterial extracts and devel¬ oped during this work is helpful for rapid identification of propionibacteria (Fessleret al., 1998). -48- 4 Results and discussion wiiiua. strains Fig. 4a: Protein profiles of Propionibacterium type ATCC 4874, (1=P. acidipropionici ATCC 25562, 2=P. thoenii 3=P. freudenreichii ATCC 6207, 4=P. jensenii ATCC 4868). from the areas covered by the Fig. 6 shows the distribution of the strains P. MIBD laboratories. Of the strains identified as propionibacteria P. freudenreichii forms the largest group with 71%, followed by jensenii with 2%. P. freuden¬ with 19%, P. acidipropionici with 8%, and P. thoenii of Swit¬ reichii and P. jensenii are encountered in all examined regions or SG/AR, zerland. P. acidipropionici was not found in VD/NE, NW, NO, and due and P. thoenii was absent in VD/NE, FR, TG, SG/AR, probably three FR, AG, to the restricted number of analysed strains. Only regions, The reason for the and ZS harboured all four species of propionibacteria. in Switzerland can only be diversity of Propionibacterium sp. distribution as or feeding, so¬ guessed at, since different factors such grass silage flora on the journ of cows on the alps during summer, propionibacterial role farm and in the cheese factory could play a 4. Results and discussion -49- 10 20 30 40 50 60 70 80 90 100 lllH..nll|lillllllll.l.i..l..t .! 11, i 11 I .ml..., I e17gel2 5 P freudenreichii AG 25 2 e17gel26 P freudenreichii AG 25 3 e29gel12 P freudenreichii AG 72 1 e17gel2 1 P freudenreichii AG 22 1 e23gel2 6 P freudenreichii AG 521 e23gel2 10 P freudenreichii AG 58 1 e23gel2 7 P freudenreichii AG 53 1 e8gel2 7 P freudenreichii ATCC 9614 e19gel2 9 P freudenreichii AG 40 1 e19gel2 11 P freudenreichii AG 42 2 elSgell 7 P freudenreichii AG 31 1 e17gel2 7 P freudenreichii AG 261 e17gel2 8 P freudenreichii AG 271 e29geM 3 P freudenreichii AG 73 1 e23gel2 9 P acidipropion AG 56 1 e18gel1 1 P acidipropion AG 21 1 e17gel2 4 P acidipropion AG 25 1 e17geE 10 P acidipropion AG 282 e23gel2 4 P acidipropion AG 50 1 e17gel2 2 P acidipropion AG 222 e26gel13 P acidipropion AG 661 e19gel2 5 P acidipropion AG 34 1 e26gel12 P acidipropion AG 63 t e29gel1 1 P acidipropion ATCC 25562 e37ge!2 9 P jensenii AG 71 1 e23gel1 11 P jensenii AG 471 e19gsl2 7 P jensenii AG 381 e19ge!2 8 P jensenii AG 39 1 61998(26 P jensenii AG 361 e19gel2 3 P jensenii AG 32 1 e23gel2 2 P jensenii AG 48 2 e26gel1 4 P jensenii AG 67 1 e17gel2 3 P jensenii AG 24 1 e9gel1 10 P jensenii ATCC 4868 e17gel2 9 P thoenii AG 28 1 e11gel2 1 P thoenii ATCC 4874 e26gel1 1 11 AG 62 1 e17gel2 11 ? ? AG 29 1 e17gel2 12 o •> AG 30 1 e14gel2 7 T> AG 17 1 e14gel2 8 ? 7 AG 181 e12gel1 1 >•> AG 131 e12ge(1 3 o •> AG 153 e7gel1 1 11 AG 112 e9gel2 1 ? 9 AG 133 e7gel1 2 ?•> AG 11 4 e9gel2 3 ?> AG 152 e8ge!2 10 ? •> AG 3.1 e8gel2 11 ? •> AG 3.2 Fig. 4b: Dendrogramm for some strains isolated from raw milk of the re¬ gion AG and type strains based on protein patterns. -50- 4 Results and discussion Fig. 5a: Amplified fragments of Fig. 5b: Patterns generated after 23S rRNA of Propioni¬ restriction with Msp I of bacterium type strains the amplified fragment of (1=P. acidipropionici Propionibacterium type ATCC 25562, 2=P. thoe¬ strains (1=P. acidipropi¬ nii ATCC 4874, 3=P. jen¬ onici ATCC 25562, 2=P. senii ATCC 4868, 4=P. thoenii ATCC 4874, 3=P. freudenreichii subsp. jensenii ATCC 4868, freudenreichii ATCC 4=P. freudenreichii 6207, 5=50-2000bp lad¬ subsp. freudenreichii der). ATCC 6207, 5=50-2000 bp ladder). Details of results for the identification of all propionibacteria strains of Swiss raw milk by both protein profile analysis and restriction analysis of 23S rRNA are given in the Annex, Tables 20-28. The only previous study performed on the population of propionibacteria in Swiss raw milk (Baer and Ryba, 1992) showed, that P. freudenreichii was almost exclusively present, whereas P. acidipropionici and P. jensenii were rarely detected and P. thoenii was not found at all. The authors attributed the predominance of P. freudenreichii to the intensive of milk. It effort over the last years to improve the hygienic quality is, however, not possible to prove this assumption as no systematic analy¬ was made before 1992 ses of propionibacteria in Swiss raw milk 4. Results and discussion -51- -=52^. 4. Results and discussion 4.1.4. Propionibacteria from alps From 38 milks originating from three different alpine regions where Gru¬ not be¬ yere cheese is produced 21 strains were isolated. One strain did long to propionibacteria, whereas of the remaining 20 strains, 11 were P. freudenreichii, 3 P. jensenii, and 6 P. thoenii. P. acidipropionici was not found. Fig. 7 shows the distribution of propionibacteria in alpine milk and in lowland raw milk from the region FR. The alps where the milk came from are also situated in this region. In both regions P. freudenreichii was the predominant species, with 62% in lowland raw milk and 55% in alpine milk. In lowland raw milk 21% of the strains were P. jensenii and in al¬ pine milk 15%. Of the alpine milk strains 30% were classified as P. thoenii compared to only 2% in lowland milk. The large proportion of P. thoenii in alpine milk could be explained by different hygienic conditions as well as difference in vegetation. Detailed classification of propionibacteria strains from alpine milk are given in Table 29, Annex. 70 60 50 - lowland raw milk Dalpine raw milk c 40 o I 30 20 10 P. thoenii P. freudenreichii P. jensenii P. acidipropionici milk Fig. 7: Distribution of propionibacteria in alpine and lowland raw from the region FR. 4. Results and discussion ^ 4.1.5. Classification of P. rubrum P. rubrum has for a long time been accepted as a separate species be¬ cause of its red-coloured colonies, but was later included into P. thoenii (Cummins and Johnson, 1986). Of the 14 P. thoenii strains isolated from Swiss raw milk including alpine milk, 9 had red or red-brown colonies. From the total of 474 investigated strains, 42 were red, red-brown or or¬ ange-brown in colour. Of these strains 4 were not propionibacteria, 4 P. freudenreichii, 2 P. acidipropionici, 9 P. thoenii and 24 P. jensenii. Con¬ sequently the colony colour appears to be an insufficient criterion to classify strains into species. Strain DSM 20275, previously classified as P. rubrum and now as P. thoenii, could not be assigned to either P. jensenii or P. thoenii by pro¬ tein profile analysis. By 23s rRNA restriction profiles analysis, however, DSM 20275 could clearly be classified as P. jensenii, confirming the re¬ classification of P. rubrum to P. jensenii by Malik et al. (1968), Britz and Steyn (1980), Britz and Riedel (1991), De Carvalho et al. (1995) who considered P. rubrum as a p-hemolytic biovar of P. jensenii, Riedel et al. (1994), and Riedel and Britz (1996). Also, none of the investigated strains showed an identical protein profile with DSM 20275. The strains with the most similar profiles were our strain FR413.1 (72.7% similarity) and DSM 20274 (73.6% similarity). By 23S rRNA restriction analysis both DSM 20274 and FR413.1 as well as DSM 20275 were classified as P. jensenii. 4.1.6. P. freudenreichii subspecies The two subspecies P. freudenreichii subsp. freudenreichii and P. freudenreichii subsp. shermanii could not be distinguished by the mo¬ lecular-biological methods used in this work. Protein profiles and restric¬ tion profiles with Msp I of 23S rRNA did not show any difference. In order to distinguish the two subspecies, the utilisation of lactose by P. freuden¬ reichii strains was determined. Of the total 278 P. freudenreichii strains 49.3% did not grow on lactose and were classified as P. freudenreichii subsp. freudenreichii and the other 50.7% showing growth on lactose as P. freudenreichii subsp. shermanii (Tables 20-29, Annex). zSAz 4. Results and discussion The existence of the two P freudenreichii subspecies is disputed among experts In the present work it was not possible to identify these subspe¬ cies with such powerful tools as SDS-PAGE and RAPD Johnson and Cummins (1972) reported a high DNA homology between the subspe¬ cies The differentiation of P freudenreichii subsp freudenreichii and P freudenreichii subsp shermanii on the basis of nitrate reduction and lactose fermentation seems to be questionable, because it has not yet been confirmed by genetic methods as shown also in this work De Car¬ valho (1994) considered the subspecies as biovars of P freudenreichii using DNA-DNA hybridisation 4.1.7. Differentiation between strains Several methods were used to investigate the propionibacterial strain variety in Swiss raw milk Ribotyping is a labour-intensive method and proved after the testing of a dozen strains and ten different restriction enzymes to be insufficient for the detection of strain variations, even if it permitted differentiation between the species Restriction analysis of to¬ tal DNA is simple to perform, but the large number and poor resolution of bands made clear distinction impossible Twenty strains and a dozen re¬ striction enzymes were tested before discarding this method Plasmid analysis and RAPD proved promising and were, therefore, used in order to investigate the diversity of propionibacteria in Swiss raw milk Until now no investigations regarding strain variety of propionibacteria in Swiss raw milk exist Thus, the results should contribute to more knowl¬ edge about the richness of the propionibacterial flora 4.1.7.1. Plasmid profile Plasmid analysis was carried out only on some propionibacteria from lowland milk Of a total of 446 propionibacterial strains, only 30 3% con¬ tained plasmids Of the plasmid-containing strains 82 3% had one, 15 (13 3%) two and 4 4% more than two plasmids Three strains, FR12 2, SG27 1, and ZS48 1 carried five plasmids Fig 8 shows an example of plasmid electrophoresis on agarose 4 Results and discussion -55 Of P freudenreichii strains 29 4% carried plasmids of P jensenii 36 6% and of P acidipropionici 30 0% All eight analysed P thoenii strains were lacking plasmids 2kb to more than 16 2kb The size of plasmids ranged from as little as 1 Of all plasmid-carrymg strains 89 0% carried plasmids larger than 16 2kb Plasmids with a size between 6-8kb seem to be relatively frequent observed with one P (61 5%) in P jensenii strains This was only acidipropionici and two P freudenreichii strains The number and size of 20-29 Annex plasmids in individual strains are given in Tables Fig. 8: Plasmid detection by agarose electrophoresis (1,2 4,7 10=no plasmid 5 8=one plasmid 3, 9=two plasmids, 6=kb ladder) 4, Results and discussion -Mz Perez Chaia et al. (1988b) found plasmids in 25% of the analysed propi¬ onibacteria. Their study, however, covered only 30 strains compared with the 373 strains in this study, 27 P. freudenreichii and 3 P. acidipropionici. P. jensenii, the species with the highest percentage of plasmid-carrying strains in the present study, was not included. The size of plasmids var¬ ied between 3.2 and 47kb; smaller plasmids were not observed. Panon (1988) detected plasmids in 38% of a total of 53 strains, including all propionibacteria species. The functional properties of Propionibacterium plasmids are not known, yet. Evidence, that lactose utilisation in P. freudenreichii might be plas- mid-linked (Rehberger and Glatz, 1987), could not be supported by the present work by accepting the subdivision of Johnson and Cummins (1986). Of the 78 plasmid-carrying P. freudenreichii strains, 38 were able to ferment lactose and were, consequently, classified as P. freudenreichii subsp. shermanii, and 40 were identified as P. freudenreichii subsp. freudenreichii lor not being able to ferment lactose. Since only 30.3% of all strains carried plasmids, this type of analysis could, as could also been foreseen from previous studies, not be used for systematic identification of propionibacteria strains. 4.1.7.2. RAPD The method of choice for investigating strain variety in Swiss raw milk proved to be RAPD. Several primers were studied (Table 2) and the two, which gave consistent numbers as well as clearest bands were chosen. Both were 12-mers with 100% G+C. Primer SK2 was used for all propio¬ nibacteria strains, whereas primer DF4 was used to identify P. freuden¬ reichii strains. The average intra-gel reproducibility with 15 strains was 94 ± 2% and inter-gel reproducibility with 12 strains was 90 ± 2%. The total of 278 P. freudenreichii strains yielded, at the 88%-similarity level, 154 different profiles with DF4 and 112 profiles with SK2. The identification of the strains is primer dependent, because with RAPD only a small part of the total DNA can be amplified. The RAPD profiles for two strains may, thus, be identical with one primer, but vary with an¬ other primer. The larger the number of primers used, the more evidence 4. Results and discussion -57- for the identity of a strain can be obtained. In this study, strains were considered as identical, when with each of the two primers the same re¬ sult was obtained. Table 2: Primers investigated for RAPD of propionibacteria (selected primers are in bold type). Primer Sequence B01 5'-GTT TCG CTC C-3' B02 5'-TGA TCC CTG G-3' B03 5'-CAT CCC CCT G-3' B04 5'-GGA CTG GAG T-3' B05 5'-TGC GCC CTT C-3' B06 5'-TGC TCT GCC C-3' B07 5'-GGT GAC GCA G-3' B08 5'-GTC CAC ACG G-3' B09 5'-TGG GGG ACT C-3' B10 5'-CTG CTG GGA C-3' B11 5'-GTA GAC CCG T-3' B12 5'-CCT TGA CGC A-3' B13 5'-TTC CCC CGC T-3' B14 5'-TCC GCT CTG G-3' B15 5'-GGA GGG TGT T-3' B16 5'-TTT GCC CGG A-3' B17 5'-AGG GAA CGA G-3' B18 5'-CCA CAG CAG T-3' B19 5'-ACC CCC GAA G-3' B20 5'-GGA CCC TTA C-3' DF1 5'-GCC GCC GCC GCC-3' DF2 5'-GCG CGC GCG CGC-3' DF3 5'-GCG GCA GCG GCG-3' DF4 5'-CCG CCG CCG CCG-3' DF5 5'-CGC CGC CGT CGC-3' DF6 5'-CGG CGG CGG CGG-3' DF7 5'-GGC GGC GGC GGC-3' SK21 5'-GCC GCC GCC GCC-3' 1 Meunier and Grimont (1993) -J08z 4. Results and discussion Only a few strains could be classified as being identical with both prim¬ ers, so that finally 219 different P. freudenreichii strain types were ob¬ tained. Identical strains usually came from the same region and are listed in Table 3. The P. freudenreichii strain diversity in Swiss raw milk is im¬ portant to note, since it offers a large reservoir of strains for future use in the dairy industry. Table 3: Groups of P. freudenreichii strains showing identical profiles with SK2 and DF4. Identical strains AG25.2, AG25.3, AG33.1 AG27.1, AG40.1 BE2.1, BE9.1, BE12.1, BE22.2, BE23.1, BE26.1 BE3.1, BE17.1 BE13.1, BE18.1 BE17.2, BE19.2, BE21.1, BE22.1 BE30.1, FR4.1 FR31.1, FR45.1 FR47.1, FR48.1 N02.1, N02.2, N03.2, N06.1, N06.2 N013.1, N016.1 N021.1.TG6.1 SG6.1, SG7.1 SG28.1, FR24.1 TG1.1.TG1.2, TG3.2, TG21.1 TG15.1, TG24.1, TG38.1.SG24.1 VD7.1, 8.1, 9.1 ZS1.1.ZS7.1 ZS19.1.ZS37.2 ZS19.2, ZS20.1 ZS36.1, N018.2 304.1, 307.1 The use of the primer SK2 on ist own permitted classification of the 32 P. acidipropionici, 14 P. thoenii, and 75 P. jensenii strains into 30, 12 and 50 different profile groups respectively. Because of the smaller signifi¬ cance of these species compared to P. freudenreichii, only one primer was used to differentiate the strains. It was possible to differentiate the 4. Results and discussion zSSr various strains of P jensenii, P acidipropionici and P thoenii by RAPD with the single primer SK2 The strain diversity of propionibacteria in Swiss raw milk is extraordinary Almost 80% of the P freudenreichii strains isolated from raw milk dif¬ fered from one another For P jensenii, P acidipropionici and P thoenii strains the variety seems even to be higher It is also interesting to note that of the four couples of strains (AG25 2/ AG25 3, N02 1/N02 2, N06 1/N06 2, TG1 1/TG1 2, Table 3) having identical RAPD patterns and originating from the same milk, two did not have the same plasmid profile This result may indicate horizontal trans¬ fer of plasmids between bacteria and/or spontaneous loss of plasmids -3Qz 4. Results and discussion 4.2. Propionibacteria used for Emmental cheese manufacture in Switzerland 4.2.1. Introduction Today, two different propionibacteria cultures are used in manufacture of Emmental cheese in Switzerland. Both are supplied and controlled by the Department of Microbiology, FAM and used by all Emmental producers in Switzerland. One of these cultures, called P-culture, is a mixture of six P. freudenreichii subsp. shermanii strains, P1409, P1410, P1411, P1412, P1413, and P1414 (Table 1, Annex). The strains were chosen at the be¬ ginning of this century from six different Emmental factories producing premium grade cheese. Unfortunately, no records from this period, indi¬ cating the accurate origin of the strains, have survived. The six strains have been and are still individually subcultured and are mixed and propagated in a fermenter, when commercial batches are produced. In recent times, FAM looked for an alternative culture with slower fermen¬ tation and less intensive C02 production. Extensive studies with wild propionibacteria strains were performed. Finally, two strains, both P. freudenreichii subsp. shermanii and isolated from the same Appenzell cheese, were retained. The mixture of these two strains, P111 and P112 (Table 1, Annex) is called Prop 96 and is since 1996 recommended by the FAM for Emmental manufacture as an alternative to the P-culture. It has proven successful in practice. Protein profiles and restriction analysis of the 23S rRNA gene were used to confirm the species, and growth on MGM with lactose to confirm the subspecies. Plasmid analysis was performed and RAPD with SK2 and DF4 was applied to differentiate on the strain level as well as to detect, which of the strains of the commercially used cultures can be found in Emmental cheese. 4.2.2. Protein profiles, plasmids and RAPD of commercial strains The classification of P1409, P1410, P1411, P1412, P1413, P1414, P111, and P112 as P. freudenreichii by the FAM with biochemical methods could be confirmed by protein profile analysis as well as by 23S rRNA 4 Results and discussion -61- analysis The growth on MGM with lactose confirmed the further distinc¬ tion as P. freudenreichii subsp. shermanii. Although the protein profile differences analysis can not differentiate between the subspecies, small between the six P-culture strains could be seen as is shown in Fig. 9. P1410, P1411, P1412, and P1414 showed practically identical profiles, whereas P1409 and P1413 were slightly different. RAPD with SK2 and DF4 of the strains in P-culture confirmed the differences between the strains obtained by protein profile analysis (Fig. 10a and 10b). • >"«"jf ••"""I *~"i Fig. 9: Protein profiles of P. freudenreichii strains in P-culture (arrows indicate strain specific bands). -62- 4 Results and discussion Fig. 10a: RAPD profiles of P-cul- Fig 10b: RAPD profiles of P-cul¬ ture strains with SK2 ture strains with DF4 (50-2000bp ladder at (50-2000bp ladder at the right) the right) As shown in Fig 10a and 10b P1410, P1411, P1412 and P1414 are identical strains (the variation of intensity of one band in the pattern of for a difference between strains et al P1410 is no sign (Davin-Regli , Jutras et al This is because in Swiss 1995, Maura , 1995)) remarkable, raw milk very few strains were found to have identical RAPD profiles Because the six strains of the P-culture can not be traced back to par¬ ticular Emmental cheese factories, it can not be excluded, that several of the strains were isolated from cheeses manufactured in the same place P1413 contains a plasmid larger than 16 2kb, the other strains carry no plasmids According to our results P-culture consists of only three, and not six different strains It can, therefore, be assumed that eliminating three of the identical strains would possibly not change cheese quality 4 Results and discussion -63 from the strains in Strains P111 and P112 in Prop 96 are clearly different P-culture No differences could be detected by protein profile analysis between P111 and P112 They could also not been differentiated by RAPD with either SK2 and DF4 Their profiles were identical as is shown P111 and P112 are almost in Fig 11a and 11b Consequently certainly were isolated the same strain This does not come as a surprise, as both nor P112 contain form the same Appenzell cheese Neither P111 plas¬ mids Fig. 11a: RAPD profiles of Prop Fig. 11b: RAPD profiles of Prop 96 strains with SK2 96 strains with DF4 (50-2000bp ladder (50-2000bp ladder at the right) at the right) 4 2.3. Detection of propionibacteria of commercial cultures Emmental Propionibacteria were isolated from three premium grade cheeses and compared by RAPD with strains of the P-culture to deter¬ cheeses Ell Elll and ElV mine which of the strains were present in the when their (Table 13, Annex) Strains were considered as identical pro¬ files had a similarity of at least 88% with both primers -34z 4. Results and discussion Not all of the P-culture strains were found in the three analysed Emmental cheeses. P1413 was completely absent and P1409 detected in only one cheese. All isolated strains were classified by growth on MGM with lactose as P. freudenreichii subsp. shermanii. Nine of the ten propionibacteria strains isolated from Ell were identical with P1410 (identical strains P1411, P1412, P1414) and one strain was identical with P1409. No wild propionibacteria strain was found in cheese Ell. Of the ten propionibacteria strains isolated from Elll, only strain Elll.3 was identical to P1410 and all other strains were wild strains. Only two of the wild strains gave identical profiles with both primers, meaning that a total of eight different wild strains was present in cheese Elll. The quality of the cheese had not been affected, although 90% of the isolated strains originated from raw milk. All ten propionibacteria strains isolated from cheese ElV were identical with P1410 (identical with P1411, P1412, P1414). We conclude, that premium grade Emmental cheese can possibly be manufactured by the addition of only P1410 (identical with P1411, P1412, P1414) to the kettle milk, because it was the only strain detected in premium grade Emmental cheese ElV. P1409 seems to contribute only slightly to the ripening of Emmental cheese, and P1413 does not seem to be able to grow at all. Also it is possible to produce premium grade Emmental cheese with naturally occuring propionibacteria in raw milk. Unfortunately Prop 96 culture could not be studied, because cheeses manufactured with this culture were not yet available. The question, as to why not all P-culture strains were found in Emmental cheese should be considered more closely. One possibility is, that strains P1409, P1410 (identical strains P1411, P1412, P1414) and P1413 have different heat resistance. This hypothesis, however, was not supported by the determination of survival rates under cheese manufac¬ ture conditions, since P1413 was reduced a hundred fold by scalding to 53"C, P1409 a thousand fold, and the other strains of the P-culture even more than a thousand fold. Another reason for the absence of P1413 in all cheeses and low contents of P1409 may be their low salt tolerance, so that the more halophilic P1410 and wild strains become dominant. In 4 Results and discussion \ ^ order to obtain more knowledge concerning this behaviour of the P- culture strains further studies should be carried out concerning growth in media with varying salt concentration and influence of temperature dur¬ ing ripening of Emmental cheese. It can also be concluded that the wild flora seems not to have been af¬ fected by the commercially distributed P-culture and Prop 96 of the FAM. P-culture consists of six and Prop 96 of only two P. freudenreichii subsp. shermanii strains. None of the wild strains showed with both primers an identical profile at the 88%-similarity with any strain of the commercial cultures. ^6: 4, Results and discussion 4.3. Propionibacteria and cheese faults 4.3.1. introduction Propionibacteria have been reported to be responsible for the cheese faults called "brown spots" (Baer et al., 1993) and "split defect" (Park et al., 1967; Hettinga et al., 1974; Steffen, 1979). In order to gain knowl¬ edge about the species and subspecies involved in these faults, propio¬ nibacteria isolated from the brown spots of various cheese types as well as from cheese with split defect were submitted to protein profile analy¬ sis and were cultured in MGM with lactose or lactate. The isolated strains were also analysed by RAPD with primers SK2 and DF4. 4.3.2. Brown spots 4.3.2.1. Emmental Only one Emmental cheese with brown spots could be analysed, be¬ cause this defect has become rarer in recent years due to addition of higher amount of P-culture to cheese milk as proposed by Baer et al. (1993). Ten strains isolated from the brown spots from cheese El were by protein analysis classified as P. freudenreichii. Out of these, nine strains belonged, according to their growth in MGM with lactose or lac¬ tate, to the subspecies shermanii and only one was found to be P. freudenreichii subsp. freudenreichii (Table 30, Annex). Comparison of the RAPD profiles with those of the P-culture showed that in cheese El the four strains El.2, El.3, El.5 and El.10 were identical to the one of P1410 from the P-culture. The six other strains were wild strains. None of the wild strains were identical with any P. freudenreichii strains isolated from raw milk of the region BE, where Emmental is pro¬ duced. However, this is not surprising since it was very rare to find the same strain of Propionibacterium in two different milks. According to Baer et al. (1993) brown spots occur in Emmental cheese, when the number of propionibacteria added to the raw milk is too low. The cheese manufacturer obviously added the P-culture to the milk, since four strains of P1410 were found. Nevertheless, the proliferation of wild strains of propionibacteria in this cheese may indicate, that the amount of P-culture added was too low. See paragraph 4.4.5. for a further discussion on this 4. Results and discussion £Zi subject. Several, or, at least one of the wild individual strains are pre¬ sumably responsible for the occurrence of brown spots. Nevertheless, it cannot be excluded, that other factors during cheese manufacture may promote the formation of brown spots in Emmental. 4.3.2.2. Sbrinz From a total of eight Sbrinz cheeses with brown spots 72 strains were isolated (Tables 15 and 33, Annex). Two of the isolated strains (SI.3 and SI.7) were identified by protein profile analysis and 23S rRNA restriction analysis as not being propionibacteria, all other strains were according to their protein profiles classified as P. freudenreichii. It was, therefore, concluded that only P. freudenreichii strains are able to survive the scalding temperature of 56-58°C during Sbrinz manufacture and to grow during the ripening. Of the 70 P. freudenreichii strains 17 were found to belong to subspecies freudenreichii and the remaining 53 strains were P. freudenreichii subsp. shermanii. Except for cheese Sll, all cheeses har¬ boured P. freudenreichii subsp. freudenreichii strains. It was not possible to detect a specific strain common to all Sbrinz cheeses with brown spots by RAPD neither with primer SK2 nor with primer DF4. Analysis of the RAPD profiles allowed the distinction of dif¬ ferent groups of strains as shown in Table 4. Ten groups could be formed with SK2, three strains did not belong to any of these groups. With DF4 14 groups were found and 16 strains could not be grouped. Using SK2 as a primer the largest group comprised 29 strains isolated from all cheeses except from cheese SI and had identical RAPD profiles with strain P1409 from the P-culture. This finding, however, could not be confirmed when the strains were analysed with primer DF4, as none of the strains showed identical profiles with P1409. Thus, none of the strains of the commercial cultures P-culture and Prop 96 were found in Sbrinz cheeses with brown spots. Of all the P. freudenreichii strains isolated from raw milk of the region ZS, where Sbrinz cheese is produced, only ZS34.1 was found to be identical to strains isolated from brown spots of Sbrinz, namely to SI.1, SI.2 and SI.6. -68- 4. Results and discussion Table 4: RAPD profiles of propionibacteria strains isolated from Sbrinz cheeses with brown spots. Strain Identical strains with SK2 and DF4 Found in cheeses SI.1 SI.2, SI.6 SI SII.1 SII.2, SII.3, SII.4, SII.5, SII.6, Sll, SIM SII.7, SII.8, SII.9, SII.10, SHI.3, SIII.4, SIII.6, SIII.7 SIII.1 SIII.2 Sill SMI.5 none Sill SMI.8 SMI.9, Sill.10 Sill SIV.2 none SIV SIV.3 none SIV SIV.4 none SIV SIV.5 SIV.6 SIV SIV.7 SIV.8, SIV.9 SIV SIV.10 none SIV SV.1 SV.2, SV.3 SV SV.4 none SV SV.5 none SV SV.6 SV.7, SV.8, SV.9 SV SV.10 none SV SVI.2 SVI.3 SVI SVI.4 none SVI SVI.5 none SVI SVI.6 SVI.10 SVI SVI.7 none SVI SVI.8 none SVI SVI.9 none SVI SVII.2 none SVI I SVII.3 none SVI I SVII.4 none SVI I SVII.5 SVII.6, SVII.7, SVII.10 SVI I SVII.9 none SVI 11 SVIII.1 none SVI 11 SVIII.2 none SVI 11 SVIII.3 none SVI 11 SVIII.4 SVIII.9 SVI 11 svm SVIII.5 none SVIII.6 none SVI 11 4. Results and discussion £& cont. SVIII.7 none SVIII SVIII.8 SVIII.10 SVIII Different P. freudenreichii strains belonging to both subspecies were in¬ volved in the formation of brown spots in Sbrinz. When combining the two primers SK2 and DF4, a total of 37 different strains could be identi¬ fied. Only one strain was found in two different Sbrinz cheeses. It can therefore be concluded that no individual P. freudenreichii strain was re¬ sponsible for the brown spots defect in Sbrinz. According to the experi¬ ence of Amrein et al. (1993) the formation of brown spots in Sbrinz cheese can be prevented by more intensive salting of the cheese. 4.3.2.3. Appenzell From four Appenzell cheeses with brown spots 43 strains were isolated (Tables 13 and 31, Annex). Protein profiles and analysis of the 23S rRNA gene showed, that one of the strains did not belong to the genus Propi¬ onibacterium. Eight strains were P. acidipropionici, two P. jensenii, and the remaining P. freudenreichii. Of the 32 P. freudenreichii strains 22 were able to ferment lactose and were thus classified as P. freudenreichii subsp. shermanii, 10 did not grow on MGM with lactose and were grouped into P. freudenreichii subsp. freudenreichii. It is interesting to note, that all of the four Appenzell cheeses contained either strains of P. acidipropionici or P. jensenii or both (Table 31, An¬ nex). Thus, all Propionibacterium sp. are able to survive the scalding temperature of 44°C applied in the manufacture of Appenzell cheese. The growth of the eight isolated P. acidipropionici strains and of one of the P. jensenii strains on the medium MF95C was considerably slower than of P. freudenreichii (Table 31, Annex). These slow growing strains were isolated from all of the four Appenzell cheeses. Although Park et al. (1967a) showed, that most P. acidipropionici and P. jensenii strains grow slowly at 7°C, the ripening temperature of 14-16°C for Appenzell cheese apparently does not prevent their growth. A previous study (Bachmann and Isolini, 1995) showed, that P. acidipropionici added to milk in Emmental cheese manufacture showed an insufficient propionic acid -ISk 4. Results and discussion fermentation and caused brown spots. According to the same authors the slow fermentation process of propionibacteria increases the risk for the formation of brown spots. Consequently, brown spot formation in Appen¬ zell cheese could probably be caused in addition to P. freudenreichii also by P. jensenii and P. acidipropionici. RAPD profiles of the P. freudenreichii strains obtained with primers SK2 and DF4 could be divided into twenty different patterns at the 88%- similarity level. None of the strains were found in two different cheeses. Also, none of the strains had with both primers profiles similar to the P. freudenreichii strains isolated of raw milk from the regions AR, SG and TG, where Appenzell cheese is produced. 4.3.2.4. Raclette Ten strains isolated from the brown spots of one Raclette cheese were analysed (Tables 14 and 32, Annex). Eight of the strains were not propi¬ onibacteria, and the remaining two were, according to protein profiles, 23S rRNA restriction profiles and growth on MGM with lactose, classified as P. freudenreichii subsp. freudenreichii. Both strains showed different profiles with SK2 and DF4. The high portion of unknown bacteria in the brown spots of this Raclette cheese is notable, but the evidence not suf¬ ficient for drawing final conclusions. 4.3.2.5. Conclusions Strains of both subspecies of P. freudenreichii seem to be involved above all in the formation of brown spots in Emmental, Raclette and Sbrinz. P. freudenreichii seems to be the only Propionibacterium sp. able to survive scalding of more than 50°C. In all analysed Appenzell cheeses, in addition to P. freudenreichii, slow growing P. acidipropionici and P. jensenii strains, which are known to cause brown spots, have been found. No specific strain is responsible for the brown spots defect. More intensive salting of cheeses to above 1.5% of dry matter could ac¬ cording to the experience of the FAM prevent the formation of brown spots. Nevertheless it would be a far better approach to prevent con¬ tamination of milk with propionibacteria. The contamination seems to 4. Results and discussion -J^ happen in the cheese factory and not at the farm, since with one excep¬ tion none of the strains found in cheeses with brown spots were similar to P. freudenreichii strains isolated from raw milk from the regions, where the cheeses are produced. Consequently, more knowledge about the contamination sources in the cheese factory should be collected. 4.3.3. Split defect 4.3.3.1. Introduction The investigation dealing with propionibacteria involved in split defect concerns Gruyere and Sbrinz cheese. In these cheese types, propioni¬ bacteria are not desirable and the ripening temperature is not adapted to suppress secondary fermentation. The investigation of propionibacteria involved in split defect did not cover Emmental cheese, where a selected propionibacterial culture is added. The added propionibacteria strains most probably suppress the propagation of propionibacteria strains origi¬ nating from the raw milk. 4.3.3.2. Sbrinz From five Sbrinz cheeses with split defect 49 strains were isolated (Table 17, Annex). All strains were classified by protein profile analysis as P. freudenreichii. Growth on MGM with lactose showed, that 71.4% were P. freudenreichii subsp. shermanii, and 28.6% P. freudenreichii subsp. freudenreichii (Table 35, Annex). The combination of the RAPD results obtained with both primers SK2 and DF4 allowed by RAPD the distinction of 29 different strains. None of the strains were found in all five Sbrinz cheeses and only two strains were found in two different cheeses. None of the RAPD profiles of the strains isolated from Sbrinz with split defect showed similarity to any of the strains isolated from raw milk of the re¬ gion ZS, where Sbrinz cheese is produced. Like in Sbrinz with brown spots, strains isolated from Sbrinz with split defect were identified as P. freudenreichii. Thus, this species seems to be the most heat resistant withstanding the scalding of up to 58°C. This result is in agreement with the results of Malik et al. (1968). An other J2i 4. Results and discussion explanation could be that the salt tolerance of P freudenreichii helps them to survive the long ripening period of over 15 months 4.3.3.3. Gruyere All of the 45 strains isolated from five Gruyere cheeses with split defect were classified by protein profile analysis as P freudenreichii (Tables 16 and 34, Annex) The ability to ferment lactose identified 53 3% as P freudenreichii subsp shermanii and 46 7% as subsp freudenreichii RAPD with primers SK2 distinguished at the 88%-similarity level 18 dif¬ ferent strains None of the strains were found in two or more cheeses and strains with identical profiles with both primers could only be ob¬ served within the same cheese Strains showing identical profiles to one of the P freudenreichii strains isolated from raw milk of the region FR, where Gruyere is produced were not detected 4.3.3.4. Conclusions In Gruyere and Sbrinz cheese with split defect only P freudenreichii subsp freudenreichii and P freudenreichii subsp shermanii were found The strains of P-culture and Prop 96 are not involved in this defect in Sbrinz and Gruyere Several P freudenreichii seem to have the capacity to cause splitting in these cheese types, because the variety of isolated strains is extraordinary In order to prevent split defect in Sbrinz and Gruyere it would be necessary to trace the origin of the propionibacteria in these cheeses 4. Results and discussion -J2z 4.4. Model Emmental cheese manufacture with selected wild Propionibacterium strains 4.4.1. Introduction Selected propionibacteria strains were tested in model Emmental cheese manufacture. The cheeses were produced at the FAM and were made from 120L pasteurised milk instead of the 1000L used for a commercially produced Emmental cheese. One of the aims of model cheese produc¬ tion was to determine, whether with some of the strains cheese defects can be provoked. Proliferation of propionibacteria in the cheese was fol¬ lowed, in order to study correlations between growth characteristics and cheese defects. The results are of interest for developing alternative propionibacteria cultures for Emmental cheese manufacture in addition to the presently available P-culture and Prop 96. 4.4.2. Selection of strains Growth rate studies in MF95C, a medium similar to the water phase of cheese, were performed on all strains. As shown in Tables 19-35, Annex, 88% of the P. acidipropionici isolated from Swiss raw milk, 96% of P. jensenii, and 100% of P. thoenii showed less intensive growth on MF95C compared with most P. freudenreichii strains (Tables 20-29, Annex) and were, thus, placed into the appropriate growth groups "zero", "hardly", "slow/weak", "slow/medium", and "fast/weak" (footnote, Table 19, An¬ nex). Of the P. freudenreichii strains only 24% showed reduced growth. For model Emmental cheese manufacture P. freudenreichii strains from the growth groups "fast/high" and "slow/high" were considered, because reduced fermentation process of propionibacteria increases the risk for brown spots (Bachmann and Isolini, 1995). In addition one strain from each of the groups "hardly" and "fast/weak" was taken. Protein and RAPD profiles were compared with the profiles of P-culture, Prop 96 and earlier tested wild strains of propionibacteria (Bachmann and Isolini, 1995). The goal was to select strains from various regions of Switzerland different from those in commercially used cultures. The re- latedness between strains was investigated by protein profiles. Strains with varied protein profiles different from P1410 were chosen. None of the selected strains were identical when analysed by RAPD with DF4. -Mz 4. Results and discussion Two of the chosen strains carried plasmids larger than 16.2kb. The se¬ lected strains (Table 5) represent only a small part of the richness of P. freudenreichii in Swiss raw milk. Both subsp of P. freudenreichii were included. The reduction of the se¬ lected strains at cheese manufacture conditions was also analysed The reduction factor for P1410 was 10000 in YEL and for P1410 as well as the selected wild strains 10 in milk. Table 5: P. freudenreichii strains selected for Emmental cheese manu¬ facture. strain subsp. growth plasmids protein profile reduction factor identity with at cheese P1410[%]1 manufacture conditions in YEL AG53.1 freud. fast/high 7.0kb 66.1 25 BE17.2 freud. fast/high negative 52.6 100000 FR22.2 freud. fast/high negative 61.3 5 N023.1 freud. fast/high negative 70.9 2 NW2.1 freud. fast/weak negative 70 8 100000 SG24.1 freud. fast/high negative 71 4 150 TG38.1 freud. fast/high > 16.2kb 64.7 30 209.1 sherm. fast/high — 60.2 15 AG73.1 sherm. fast/high negative 58.2 15 FR4.1 sherm. fast/high negative 61.9 10 FR21.2 sherm. fast/high negative 61.5 100 FR26 1 sherm. hardly negative 55.8 25 N024.1 sherm. fast/high negative 66.0 600 NW5.1 sherm. slow/high negative 48.3 5 NW8.1 sherm. fast/high negative 63.2 30 SG31.1 sherm fast/high negative 71.7 70 TG41.1 sherm. fast/high negative 60.0 30 VD9.1 sherm. fast/high > 16.2kb 70.5 350 ZS26.1 sherm. fast/high negative 42.2 10 ZS29.1 sherm. fast/high > 16.2kb 74.8 15 1 Pearson correlation 4. Results and discussion -75- 4.4.3. Quality of model Emmental cheese A total of 66 cheeses were produced at a rate of six or eight per day, in¬ cluding a reference cheese with the P-culture With each of the chosen propionibacteria strains two or three cheeses were produced on separate days No wild propionibacteria could be detected after pasteurisation of the milk Various microbiological and chemical analyses were made dur¬ ing the ripening of the cheeses At the end of the ripening period (120 days) cheeses were judged for their quality by a small panel of three people Table 6 summarises the results of the sensory analyses Fig 12 shows some of the model Emmental cheeses Table 6: Quality of manufactured model Emmental cheeses (notes mean values) Strain(s) Cheese Score1 Remarks number Taste Total P-culture 1 50 21 7 9 47 20 7 17 50 21 0 slightly extended eyes 25 53 21 6 33 50 21 0 41 53 21 3 49 40 20 4 55 47 190 61 50 20 0 209 1 6 47 190 small splits, extended eyes 29 50 22 0 small splits 57 40 16 6 small splits, extended eyes AG53 1 2 40 17 6 brown spots 38 1 7 130 brown spots, no eyes 60 20 120 brown spots, sparse eyes AG73 1 15 50 20 7 sparse eyes 28 30 16 0 sparse eyes 64 37 170 sparse eyes BE17 2 4 1 7 14 0 no eyes, no typical Emmental flavour 31 1 7 14 7 no eyes, no typical Emmental flavour 65 2 7 15 0 no eyes, no typical Emmental flavour JSz. 4. Results and discussion cont FR4 1 8 23 14 3 brown spots, very small eyes 47 23 14 3 brown spots, no eyes 59 23 11 0 brown spots, no eyes FR21 2 24 53 197 splits, extended eyes 39 50 20 3 brown spots, extended eyes 58 37 184 FR22 2 13 27 14 3 brown spots, sparse eyes 27 27 150 brown spots 54 33 153 brown spots, sparse eyes FR26 1 18 50 193 sparse eyes 34 47 20 1 sparse eyes N023 1 3 33 183 sparse black spots, red periphery 43 30 150 red periphery 56 30 150 red perphery N024 1 11 30 157 brown spots, small eyes 44 23 133 brown spots, no eyes 53 37 157 brown spots, sparse eyes NW2 1 14 20 137 sparse black spots, sparse eyes 26 1 7 14 7 no eyes NW5 1 20 47 187 42 53 21 7 NW8 1 12 53 23 0 46 50 20 3 SG24 1 23 57 20 7 extended eyes 35 53 20 6 extended eyes 66 50 20 0 extended eyes SG31 1 7 50 187 brown spots, sparse big black spots 36 47 20 1 brown spots 52 43 20 3 brown spots TG38 1 10 57 23 1 21 53 22 7 40 47 20 7 TG41 1 5 53 20 3 30 47 20 4 62 47 197 4 Results and discussion -ZL. COnt. VD9 1 16 50 21 0 extended eyes 37 43 193 slightly extended eyes 63 33 186 ZS26 1 22 50 21 3 32 43 20 0 48 50 20 4 ZS29 1 19 40 20 0 45 47 194 51 53 21 9 1 Maximum 6 0 points for each aspect (appearance, eyes, texture and taste), in total 24 0 points As shown in Table 6, the defects brown spots, red peripheral area and split defect could be provoked by the same strain The only exception was cheese 39, where brown spots were detected, whereas in cheeses 24 and 58 manufactured with the same strain FR21 2, no brown spots were found Plating of propionibacteria from cheese 39 on YELA showed two distinct colony forms, one smaller and brown, the other larger and beige in colour This finding suggests, that either the milk or the cheese vat were contaminated with a propionibacterial strain causing brown spots Consequently, brown spots and split defect in Emmental cheeses are clearly strain-related The reddish-orange colour on the peripheral area just under the surface observed in cheeses manufactured with strain N023 1 has previously been observed by FAM experts, but this defect seems to be very rare and its causes were not followed further The defect of black spots found in cheeses 3, 7 and 14 was never ob¬ served before by the panel experts This phenomenon does not seem to be caused by propionibacteria alone, because different bacteria were detected in the spots under the microscope With six of the tested strains, model Emmental cheeses of good quality were produced (Table 6) Three of the best were selected to produce commercial Emmental cheese in order to find an alternative to P-culture and Prop 96 J3z. 4 Results and discussion with different wild Fig. 12: Model Emmental cheeses produced propioni¬ bacteria strains 12; bottom: (from left to right: top: 1, 2, 3, 4; middle 9, 10, 11, 17, 18, 19, 20). 4.4.4. Influence of subspecies strains selected, only Of the seven P. freudenreichii subsp. freudenreichii SG24.1 satis¬ cheeses made with TG38.1 were of good quality. yielded Cheeses manufactured with factory cheeses with a tendency to split. with BE17.2 and NW2.1 AG53.1 and FR22.2 showed brown spots, those N023.1 the red lacked typical Emmental flavour and those with produced peripheral area. cheeses Of the 13 P. freudenreichii subsp. shermanii strains, produced ZS29.1, ZS26.1) with six strains yielded good (TG41.1, NW8.1, FR26.1, Cheeses with four and with one strain (AG73.1) satisfactory quality. a to split defect strains (209 1, VD9.1, NW5.1, FR21.2) showed tendency FR4.1 and N024.1) and cheeses produced with three strains (SG31.1, had brown spots. shermanii strains seem It can be concluded, that P freudenreichii subsp. cheese quality than P freuden¬ generally to produce better Emmental reichii subsp. freudenreichii strains. 4. Results and discussion JSz. 4.4.5. Microbiological and chemical analysis Propionibacteria, enterococci and halophilic bacteria were determined on days 1, 30, 60, 80 and 120. No enterococci were found in any of the cheeses made from pasteurised milk and halophilic bacteria were de¬ tected only up to 600cfu/g. These bacteria, however, did not play any role in quality, since no differences between cheeses with and cheeses without halophilic bacteria were detected. Fig. 13 shows propionibacteria growth curves with reference to model Emmental cheese quality. 5.00E+09 4.50E+09 Good quality 4.00E+09 3.50E+09 Satisfactory quality 3.00E+09 Brown spots 5> 2.50E+09 "G 2.00E+09 Tendency to split 1.50E+09 defect Red area 1.00E+09 peripheral 5.00E+08 No flavour O.OOE+00 -5.00E+08 -i Day Fig. 13: Growth of propionibacteria in model Emmental cheeses of differ¬ ent quality (good quality: NW5.1, NW8.1, TG38.1, TG41.1, ZS26.1, ZS29.1, P-culture; satisfactory quality: AG73.1, FR26.1; brown spots: AG53.1, FR4.1, FR22.2, N024.1, SG31.1; tendency to split de¬ 209.1 FR21.2 SG24.1 VD9.1 red area: fect: , , , ; peripheral N023.1; no flavour: BE17.2, NW2.1). Strains yielding cheeses with good quality show the typical growth curve of propionibacteria in cheese with a rapid increase until day 60 and a -Mz 4. Results and discussion slow decrease after day 90. The two strains, giving cheeses of satisfac¬ tory quality did not show a decrease in cfu/g after 90d. Growth curves of strains giving cheeses with a tendency to split defect showed a very rapid increase up to 4x109cfu/g followed by decrease and another in¬ crease after day 90. This second increase in proliferation is most proba¬ bly responsible for the split defect due to the additional production of C02 at this stage of ripening. The five strains giving cheeses with brown spots showed lower proliferation of propionibacteria and for four of the five strains a very low level (in most cases not detectable) on day 1. Counts of cfu/g were only diminished by half on day 120 compared to good cheeses. These results seem for the exception of strain SG31.1 to confirm the results of previous investigations by Baer et al. (1993). Strains involved in the production of model Emmental cheeses with red peripheral area and no typical Emmental flavour showed reduced growth compared to all other strains. The following chemical analyses were carried out: water content, pH, L- and D-lactic acid on day 1 and 120; fat, NaCI, volatile fatty acids, formic acid, acetic acid, propionic acid, iso-butyric acid, n-butyric acid, iso- valerian acid, iso-capronic acid, n-capronic acid, succinic acid, and amino acids content on day 120. Results of interest and of day 120 only (mean values) for cheeses produced with each strain are listed in Table 7. Cheeses produced with BE17.2 and NW2.1 had a very high lactate con¬ tent, because lactate was only insufficiently utilised by the slowly prolif¬ erating strains. These cheeses had no typical Emmental flavour nor eyes (see also Fig. 12) caused by slow growth of propionibacteria. The two strains were significantly less heat resistant than the other strains (Table 5). The cheeses with brown spots, produced with strains AG53.1, FR4.1, FR22.2, N024.1 and SG31.1 generally had a higher lactate content and rather low propioniate, acetate and volatile fatty acids content due to the reduced growth and metabolism of propionibacteria. Cheeses showing tendency to split defect had no lactate. The cheeses produced with N023.1 showing the red peripheral area did not show any particular chemical characteristics. Table 7: Chemical analyses for Model Emmental cheeses produced with each strain (mean values and standard deviations). Acetate Volatile acids Strain(s) D- and L-lactate Succinate Propionate Aspartate Asparagin fatty [mmol/kg] [mmol/kg] [mmol/kg] [mg/kg] [mg/kg] [mmol/kg] [mmol/kg] P-culture 5.2±7.0 11.4±1.5 75.1128.7 0.010.0 0.010.0 44.4+2.9 134.819.2 AG53.1 77.0±60.6 6.7+2.5 13.516.0 231.5 143.7 9.412.6 23.618.8 N023.1 67.5+9.1 4.7±0.6 58.611.7 213.4 309.0 27.310.6 87.012.3 BE17.2 142.7±7.3 0.5+0.1 3.111.0 264.8 544.7 4.2H.0 8.212.0 TG41.1 2.8+2.4 12.1±0.9 67.4130.1 78.4 0.0 42.9+0.0 139.019.2 209.1 0.0±0.0 6.4±0.6 93.213.0 241.7 220.5 44.211.8 138.415.4 SG31.1 18.4±12.0 11.9±1.6 73.211.0 0.0 0.0 36.610.0 112.011.1 FR4.1 116.0+19.1 1.8±0.4 27.1+12.1 264.7 816.6 14.214.3 42.0116.3 TG38.1 0.0±0.0 12.4±0.7 99.3 0.010.0 0.010.0 47.110.0 147.6 N024.1 105.6±15.5 7.3+2.8 27.7113.4 97.7 559.5 17.516.4 46.9118.6 NW8.1 9.5±11.7 4.9±0.1 87.9110.0 215.6 467.3 37.315.9 127.8113.6 FR22.2 102.6±12.1 2.7+1.1 32.3113.7 205.0 796.9 16.316.3 51.5117.5 NW2.1 143.3±3.9 0.5±0.1 3.610.0 241.6 863.5 510.0 9.7 AG73.1 51.6+6.3 4.7±0.1 66.710.6 248.8 486.7 31.910.6 100.410.6 VD9.1 0.0±0.0 10.2±0.4 94.912.7 0.0 0.0 44.211 140.211.6 FR26.1 18.5±14.4 5.9±0.4 78.210.0 177.1 404.0 36.510.0 115.6 ZS29.1 7.8±9.7 4.9+0.3 92.911.8 265.0 401.6 41.310.6 137.011.2 NW5.1 0.0±0.0 11.4±0.0 100.4 0.0 0.0 44.610.0 146.3 ZS26.1 28.1±2.3 4.510.6 80.610.9 264.4 549.9 37.210.4 119.2+0.5 SG24.1 0.9±2.3 11.610.7 92.911.1 0.0 0.0 43.210.0 139.312.3 FR21.2 0.0±0.0 10.411.3 95.312.4 0.0 0.0 45.610.2 142.012.5 -82- 4. Results and discussion Most of the cheeses of favourable quality had a high propionate content, sometimes even higher than the reference cheeses. They also generally showed a higher acetate and volatile fatty acid content as well as a ten¬ dency towards higher succinate content. The results of chemical analy¬ ses are in agreement with the good proliferation of the involved propioni¬ bacterial strains. Aspartate and asparagin content was generally, but not exclusively lower as in good quality cheeses. Cheeses of satisfactory quality differed only slightly from those of good quality. Cheeses pro¬ duced with propionibacteria strains FR26.1 and NW2.1 showing atypical growth on MF95C different from one another. The "fast/weak" strain NW2.1 showed only slow proliferation in model cheeses, yielding to cheeses with no typical Emmental flavour. The other strain FR26.1 classified in the growth category "hardly" yielded cheeses of satisfactory quality. 4.4.6. Sensorics of model Emmental cheeses From a total of 66 model cheeses samples with favourable quality, or with a slight tendency to splitting, were submitted to triangle tests by the accredited sensory panel of the FAM consisting of 12 experts. Cheeses having an equivalent or higher score for taste compared to the reference cheese were selected. The members of the panel were asked to distin¬ guish between the taste of cheese produced with a selected strain and the reference cheese as well as to state their personal preference for either a single or a pair of cheeses. No significant differences in taste between the cheese produced with any of the strains and the reference cheese was found with exception of strain FR21.2 (Table 8). This evaluation seems to contradict the results in Table 6, in which the judgement regarding the taste of a three-member panel is summarised. The first panel of three persons was particularly looking for the typical taste of an Emmental cheese, which led to the re¬ sult, that a number of experimental cheeses were found to be superior to the reference cheese produced with P-culture on the same day with the same milk. The second panel of 12 people was less specific in looking for the particular taste of the Emmental cheese, but rather to determine, whether a difference could be recognised in the triangle test. Thus, we may conclude, that Emmental can be produced with different individual wild strains without affecting the sensory quality of the cheese. 4. Results and discussion ^&. Table 8: Results of triangle test. cheese strain reference cheese1 difference 29 209.1 (2x) 25 not significant 24 FR21.2(2x) 17 significant at a=0.05 42 NW5.1 (2x) 41 not significant 12 NW8.1 (2x) 9 not significant 46 NW8.1 (2x) 48 not significant 46 NW8.1 48 (2x) not significant 23 SG24.1 (2x) 17 not significant 10 TG38.1 (2x)2 9 not significant 21 TG38.1 (2x) 17 not significant 5 TG41.1 1(2x) not significant 16 VD9.1 (2x) 9 not significant 22 ZS26.1 (2x) 17 not significant 51 ZS29.1 49 (2x) not significant 1 manufactured on the same day with the P-culture 2 (2x) means two samples of this cheese are compared with one sample of the other one 4.4.7. Raster electron microscopy Images were taken with a raster electron microscope (REM) at the Swiss Federal Institute of Technology Zurich of a model Emmental cheese with brown spots produced with strain FR22.2. At a magnification of 6000 of a section through one of the brown spots many bacteria forming a dark contrast were seen in Fig. 14. Figs. 15a and 15b show the same bacteria at a magnification of 75000. The slightly thicker cell wall in Fig. 15b is due to the difference in section. In the zone around brown spots other bacteria are also visible (Fig. 16). These are longer and have a different cell wall as noticeable by the difference in contrast. None of the darker bacteria were found in zones volatile of brown spots. An image of propi¬ onibacterial strain FR22.2 (Fig. 17) in YEL taken at a magnification of 75000 shows the same cell wall as the bacterium in the brown spots of Fig. 16. Consequently, propionibacteria, visualised under the REM with a dark contrast with and a slightly lighter coloured cell wall, were found in cheeses with brown spots and only in the spots themselves. M. 4 Results and discussion cheese Fig. 14: REM photo of brown spot of model Emmental produced with FR22 2 magnification 6000 **--^^ggS«fc*' .«£ a 4 Results and discussion 85- I of mo Figs. 15a and 15b REM photo of P freudenreichii in brown spot del Emmental cheese produced with FR22 2 magni¬ fication 75000 (a longitudinal section b cross-section) of model Fig. 16. REM photo of Lactobacillum in brown spot free zone Emmental cheese produced with FR22 2 magnification 27000 -86- 4 Results and discussion ,- yf^fir- ^ / / u " f w ^ 3 .•^ ik Fig. 17: REM photo of P freudenreichii FR22 2 in YEL, magnification 75000 4. Results and discussion -82- 4.5. Emmental cheese manufacture with three wild Propio¬ nibacterium strains 4.5.1. Introduction Of the 20 strains used for model Emmental cheese manufacture, three strains TG38.1, ZS26.1 and NW8.1 which yielded cheeses of favourable quality were selected for Emmental cheese manufacture at a commercial cheese factory. For comparison reference cheeses were produced with the P-culture. A total of 16 cheeses, including reference cheeses were produced. Four cheeses were produced with each of the selected strains and with the P-culture. Of the four cheeses per strain two were produced with the addition of MK 3008, which consists of Lb. casei subsp. casei. This bacterium is added during the manufacture of Emmental in Switzer¬ land to prevent an excessive propionic acid fermentation and, conse¬ quently, to reduce the risk of split defect. The interaction between P. freudenreichii and Lb. casei subsp. casei in cheese has been tested empirically by the FAM. Investigations on the exact biochemical proc¬ esses are currently being performed. 4.5.2. Quality of cheeses The quality of cheeses was examined after five months of ripening by a panel of four experts. The results of the sensory analyses are summa¬ rised in Table 9. Emmental cheeses with Lb. casei (MK 3008) added and therefore with smaller eyes were found to be superior by the experts, because the risk of crumbling during cutting the cheese was reduced. Figs. 18 and 19 show Emmental cheeses produced with each strain and P-culture with and without MK 3008 added. The taste of cheeses was also influenced by the addition of MK 3008. Cheeses with P-culture and strain ZS26.1 lost in taste, whereas cheeses with TG38.1 and NW8.1 were found to be better. The total score of cheeses produced with wild strains was signifi¬ cantly higher than that of cheeses manufactured with P-culture. These results are in agreement with the results in Table 6 with model Emmental cheeses. -88- 4 Results and discussion Table 9: Quality of manufactured Emmental cheeses (mean values for grading two cheeses in each group by four panel members) Stram(s) Score1 without MK 3008 with MK 3008 Taste Total Taste Total P-culture 4 7 164 44 17 7 TG38 1 44 196 50 21 9 NW8 1 4 5 183 49 20 4 ZS26 1 4 9 20 5 4 8 20 7 1 Maximum 6 0 points for each aspect (outward appearance, eyes tex¬ ture and taste), in total 24 0 points Fig. 18: Emmental cheeses produced with three wild propionibacteria strains and P-culture (1 ZS26 1 with MK 3008, 2 NW8 1 without MK 3008, 3 TG38 1 without MK 3008, 4 P-culture with MK 3008) 4. Results and discussion -89- Fig. 19: Emmental cheeses produced with three wild propionibacteria strains and P-culture MK 3: (1: ZS26.1 without MK 3008, 2: P-culture without 3008, TG38.1 with MK 3008, 4: NW8.1 with MK 3008). manufactured with the It can be concluded that Emmental cheese can be of dif¬ addition of various wild propionibacterial strains. A mixed culture individual strains ferent strains seems not to be necessary, because with be The three the same or even better quality of cheese can produced. alternative wild strains NW8.1, TG38.1 and ZS26.1 would be valuable cultures and should be tested further. -SQz 5. Conclusion 5. Conclusion 5.1. Methods for identification and classification Of the various methods tested in this study, three proved to be especially useful for the differentiation of propionibacteria. Electrophoresis of pro¬ teins with the help of computer imaging is a tool with good reproducibility when classifying propionibacteria into species. Nevertheless, restriction profile analysis of the 23S rRNA gene is the method of choice. It is fast, reliable, absolutely reproducible and it can be performed on crude bacte¬ rial extract of a liquid culture or a colony as well as on purified DNA. Some but not all strain differences may be recognised when comparing protein profiles. But for the differentiation of propionibacterial strains RAPD is a more informative method. Inter-gel reproducibility of 88% of different DNA extractions was attained in this study. RAPD is, however, more time consuming than often mentioned in the literature, because various primers have to be tested and cycling conditions adapted to them. To detect strain identities two or more primers should be used, be¬ cause the larger the numbers of primers used, the more evidence for the identity of strains is obtained. RAPD could in the future serve to detect contamination sources of milk as well as to follow the proliferation of a strain during ripening. 5.2. Propionibacterial flora in Swiss raw milk The propionibacterial flora in Swiss raw milk is extremely diverse and rich. In raw milk from the lowlands all four dairy species were found, in raw milk from the alps P. acidipropionici was absent and P. thoenii made up 30% compared to that of only 2% found in lowland raw milk. The rea¬ son for this difference and its consequence for cheese manufacture would be an interesting field for further studies. Strain diversity of propionibacteria in Swiss raw milk is extraordinary. Most of the strains were only detected once by RAPD. The natural propi¬ onibacterial diversity has not been influenced by the commercially used P. freudenreichii strains, as none of these strains were found in raw milk. As only 30% of all analysed propionibacteria carried plasmids, plasmid content did not serve for further identification of strains. The plasmids 5. Conclusion ^ might however be interesting for future studies, as the characteristics coded on them still remain unknown and there is a possibility of horizon¬ tal transfer and/or spontaneous loss. The cheese industry has in raw milk a large reservoir of Propionibacterium strains for future applications and developments. 5.3. P. rubrum and P. freudenreichii subsp. P. rubrum, which is today included in the P. thoenii species could by protein electrophoresis not be classified. Restriction analysis of the 23S rRNA gene, however, showed that P. rubrum must be classified as P. jensenii. This finding is of taxonomic importance and supports other re¬ searchers, who have recently worked on this subject (Britz and Riedel, 1991; Riedel et al., 1994; De Carvalho et al., 1995); Riedel and Britz, 1996). The existence of the two subspecies P. freudenreichii subsp. freuden¬ reichii and P. freudenreichii subsp. shermanii is disputed among experts. Johnson and Cummins (1972) reported a high DNA homology between the subspecies, which was indirectly confirmed in the present work, in that methods such as SDS-PAGE and restriction analysis of the 23S rRNA gene were not able to distinguish between subspecies. The differ¬ entiation of a subspecies on the basis of nitrate reduction and lactose fermentation remains questionable, as long as it has not been confirmed genetically. The present results support the opinion, that subdivision of P. freudenreichii should be abandoned. 5.4. Prop 96 and P-culture RAPD profiles as well as protein profiles of the two P. freudenreichii strains comprised in the Prop 96 culture showed, that these two strains are identical. The preparation of the culture would be simplified by re¬ moving one of the "strains" without any risk for the quality of cheese. Protein profiles and RAPD profiles of the six strains included in the P- culture showed clearly, that four strains are identical and in fact only three different strains make up this culture. Analysis of strains isolated from three premium grade Swiss-type cheeses proved, that one of the z92z S. Conclusion three strains was not found in any of the cheeses and another strain only in one cheese. It may be that of the six original strains in the P- culture only one is responsible for cheese quality. 5.5. Cheese defects In Emmental and Sbrinz with brown spots as well as in Gruyere and Sbrinz with split defect only P. freudenreichii strains were found. In each analysed cheese various wild propionibacteria were found, and conse¬ quently different strains may provoke these cheese defects. It is impor¬ tant to note, that only one strain was identical to strains isolated from raw milk. In Appenzell with brown spots P. freudenreichii, P. acidipropionici and P. jensenii strains were detected. In addition to P. freudenreichii, present in all cheeses, either P. acidipropionici or P. jensenii were isolated. It is known, that the slow growing P. acidipropionici strains provoke the for¬ mation of brown spots in cheese. Also P. jensenii strains grow rather slowly and could therefore cause this defect. It is possible that the for¬ mation of brown spots in Appenzell is caused by other Propionibacterium sp. than in Emmental and Sbrinz. It would be of great interest to know more about the contamination sources of milk with propionibacteria and the consequences for cheese quality. With the help of SDS-PAGE and/or restriction analysis of the 23S rRNA gene differentiation on species level is possible. RAPD allows the distinction of contaminating strains and the investigation of their role in cheese ripening. 5.6. Emmental cheese manufacture The manufacture of model Swiss-type cheeses with different individual wild P. freudenreichii strains showed that cheese defects are strain de¬ pendent. The strains producing cheeses with a tendency towards split defect could not be characterised by typical protein or RAPD profiles. Cheeses with brown spots were chemically different. The most striking difference, however, was observed between the growth of strains produc¬ ing cheeses with brown spots and those producing cheeses of good 5. Conclusion £2: quality. The growth of strains, which provoked the formation of brown spots was slower, but at the end of the ripening the cfu/g were approxi¬ mately equal to those of propionibacteria in cheeses without brown spots. However, the cultivation of these strains alone in broth MF95C, a medium similar to cheese, does not allow a prediction about their per¬ formance in cheese. With the manufacture of model and commercial Emmental it was also proved, that cheese of favourable sensory and overall quality may be manufactured with different single wild strains. The extraordinary rich¬ ness of propionibacteria in Swiss raw milk is a good basis for further tri¬ als. 4±_ 6, Bibliography 6. Bibliography R M Handbook of media L C ed Atlas, , (1993) microbiological (Parks, , ) 465, 531, 532, 542, 621, 765, CRC Press, Florida N J W E Charactenzahon of Al-Zoreky, , Ayres, W, Sandine, (1993) propionibacterial growth metabolites for Gram Cult Prod J 28 4-12 inhibitory negative bactena, Dairy , (2) Amrein, R (1997) Personal communication, Cheese Technology, FAM R H - Verursacher der Amrein, , Rentsch, F, Sollberger, (1993) Propionsaurebakterien wichtig- sten Sbnnzfehler, FAM discussion group Sbrinz, 2-12 Antipov, V A, Subbotin, V M (1980) Use of probiotics - efficacy and prospects, Dairy Sci Abstr 1982, 44, 8567 H P D von wilden im Em¬ Bachmann, , Isolini, (1995) Prufung Propionsaurebaktenen-Stammen mentaler Modellkase, Schweiz Milchw Forschung, 24(2) 19-22 Baer, A (1987) Identification and differentiation of propionibacteria by electrophoresis of their Milchwiss 42 431-433 proteins, , (7) Baer, A, Ryba, I (1988) Identification rapide des bacteries propioniques, Schweiz Milchw For¬ schung, 17 (3) 56-57 Baer, A, Ryba, I (1992) Serological identification of propionibacteria in milk and cheese sam¬ Int J 299-310 ples, Dairy, , 2, Baer, A, Ryba, I, Grand, M (1993) Ursachen der Entstehung von braunen Tupfen im Kase, Schweiz Milchw Forschung, 22(1) 3-7 Baer, A (1996) Personal communication, Biochemistry, FAM H C C in the differentiation of Beerens, , Neut, , Romond, (1986) Important properties gram-po- sitive non-sponng rods in the genera Propionibacterium, Actinomyces and Bifidobacte¬ in other man E M G rium, In Anaerobic bacteria habitats than (Barnes, , Mead, C, eds), Vol 2,1346-1353, Blackwell Scientific Publications, Great Britain Bergey, D H (1930) Bergey's manual of determinative bactenology, 3rded, The Williams & Wilkins Co, Baltimore, 299-302 SL G EG Influence of Lactobacillus on Biede, , Reinbold, W, Hammond, (1977) bulgancus microbiology and chemistry of Swiss cheese, J Dairy Sci, 59 (5) 854-858 J T Ammo acid metabolism in cells ex¬ Brendehaug, , Langsrud, (1985) propionibacteria resting periments with 4 strains, J Dairy Sci, 68 (2) 281-289 T J W H The of different media for the isolation and enu¬ Bntz, , Holzapfel, (1973) suitability meration of propionibacteria from dairy products, S Afr J Dairy Technol, 5 (4) 213-216 T J P L studies on acid bactena, Bntz, , Steyn, (1980) Comparative propionic Phytophylactica, 12, 89-103 T J K -H J A numerical taxonomic of strains from Bntz, , Riedel, (1991) study Propiombactenum dairy sources, J Appl Bactenol, 71,407-416 6. Bibliography -95- T J K-H J in Leerdammer Bntz, , Riedel, (1994) Propiombactenum species diversity cheese, Int J Food Microbiol, 22,257-267 M R R M in Italian hard Carcano, , Todesco, , Lodi, , Brasca, (1995) Propionibacteria cheese, Lait, 75(4-5)415-426 M G J Presence of in lactic acid bactena J Casey, , Meyer, (1985) x-prolyl-dipeptidyl-peptidase Dairy Sci, 68 (12)3212-3215 Casey, M G (1994) Personal communication, Biochemistry, FAM B M H M I Effect of the Proma Cerna, , Cerny, , Betkova, , Patncny, P, Soch, , Opatrna, (1991) probiotics on calves, Dairy Sci Abstr 55,1735 C P M sur Mozzarella de amencain Champagne, , Lange, (1990) Analyses fromages type ayant montre une anormale de Sci Alim 10 43-55 production gaz, , (1) Charfreitag, O, Stackebrandt, E (1989) Inter- and intragenenc relationships of the genus Propi¬ ombactenum as determined by 16S rRNA sequences, J Gen Microbiol, 135, 2065- 2070 Colomban, A, Roger, L, Boyaval, O (1993) Production of propionic acid from whey permeate by sequential fermentation, ultrafiltration, and cell recycling, Biotechnol Bioeng, 42, 1091-1093 Costas, M (1990) Numencal analysis of sodium dodecyl sulphate-polyacrylamide gel electro- phoretic protein patterns for the classification, identification and typing of medically im¬ portant bactena, Electrophoresis, 11, 382-391 M B P R J L R Costas, , Pot, , Vandamme, , Kersters, K., Owen, , Hill, (1990) Interlaboratory comparative study of the numencal analysis of one-dimensional sodium dodecyl sul¬ phate-polyacrylamide gel electrophoretic protein patterns of Campylobacter strains, Electrophoresis, 11 (6)467-474 Costas, M, Holmes, B, Ganner, M, On, S L W, Hoffman, P N, Worsley, M A, Pamgrahi, H (1994) Identification of outbreak-associated and other strains of Clostndium difficile by numencal analysis of SDS-PAGE protein patterns, Epidemiol Infect, 113,1-12 Crow, VL (1986a) Utilization of lactate isomers by Propiombactenum freudenreichii subsp shermanii regulatory role for intracellular pyruvate, Appl Environ Microbiol, 52 (2) 352- 358 Crow, V L (1986b) Metabolism of aspartate by Propiombactenum freudenreichii subsp sher- mamr effect on lactate fermentation, Appl Environ Microbiol, 52 (2) 359-365 Crow, V L, Turner, K W (1986) The effect of succinate production on other fermentation prod¬ ucts in Swiss-type cheese, N Z J Dairy Sci Technol, 21, 217-227 Cummins, C S, Johnson, J L (1986) Genus I Propiombactenum Orla-Jensen 1909, In Bergey's P H N S M J G manual of systematic bactenology (Sneath, A, Mair, , Sharpe, E, Holt, , eds), Vol 2,1346-1353, The Williams & Wilkins Co, Baltimore Daeschel, MA (1989) Antimicrobial substances from lactic acid bactena for use as food pre¬ servatives, Food Technol, 43 (1) 164-167 -96- 6. Bibliography Y R N C de P Variations in DNA con¬ Davm-Regli, A, Abed, , Charrel, , Bollet, , Micco, (1995) centrations affect the reproducibility of RAPD fingerprint patterns, Res Microbiol, 146, 561-568 De Carvalho, A F (1994) Systematique des bacteries propioniques laitieres classification, no¬ menclature et identification Dissertation 9429, 69, E N S A de Rennes De A F M F Identification of Carvalho, , Gautier, , Gnmont, (1994) dairy Propiombactenum spe¬ cies by rRNA gene restriction patterns, Res Microbiol, 145,667-676 De A F S M PAD Reclassification of Carvalho, , Guezenec, , Gautier, , Gnmont, (1995) Propi¬ ombactenum rubrum" as P jensenii, Res Microbiol, 146, 51-58 P B A M K of the Descheemaeker, , Pot, , Ledeboer, , Vemps, T, Kersters, (1994) Companson Lactococcus lactis differential medium (DCL) and SDS-PAGE of whole-cell proteins for the identification of lactococci to subspecies level, Syst Appl Microbiol, 17 (3) 459-466 Van W M C A H Influence of on DeVnes, W, Wijck-Kapteyn, , Stouthamer, (1972) oxygen growth, cytochrome synthesis and fermentation pattern in propionibacteria, J Gen Mi¬ crobiol 31-41 , 76, F D T M Detection of in J Res 1-5 Drinan, , Cogan, (1992) propionibactena cheese, Dairy , 59, C C P and esterase activities of Dupuis, , Corre, , Boyaval, (1993) Lipase Propiombactenum freudenreichii subsp freudenreichii, Appl Environ Microbiol, 59(12)4004-4009 C C P Proteinase of Dupuis, , Corre, , Boyaval, (1995) activity dairy Propiombactenum, Appl Microbiol Biotechnol, 42, 750-755 D Z Identification et caractensation de bactenes dans le Fessler, , Puhan, (1995) propioniques lait cru suisse, Poster D2, First International Symposium on Dairy Propionibactena, Rennes, France R G G bactena in different Foschino, , Galli, A, Ponticelli, , Volonteno, (1988) Propionic activity culture conditions, Ann Microbiol, 38, 207-222 R Probiotics in man and J Bacterid 365-378 Fuller, (1989) animals, Appl , 66, N P Determination of size of Gautier, M, Mouchel, , Rouault, A, Sanseau, (1992) genome Propiombactenum species by pulsed-field gel electrophoresis, Lait, 72,421-426 M S D Charactenzation of a Gautier, , Rouault, A, Lortal, , Leioir, Y, Patureau, (1992) phage infecting Propiombactenum freudenreichii, Lait, 72,431-435 Gnnstead, D A, Barefoot, S F (1992) Jenseniin G, a heat-stable bactenocin produced by Propi¬ ombactenum jensenii P126, Appl Environ Microbiol, 58 (1) 215-220 Hani, J P (1991) Etude concernant la teneur en bacilles butynqueset bacteries propioniques du lait destine a la fabrication de gruyere, FAM internal report, 3,1-6 H S S T Production of vitamin Hatanaka, , Wang, E, Taniguchi, , lijima, , Kobayashi (1988) B12 by a fermentor with a hollow-fiber module, Appl Environ Microbiol, 56, 215-220 D M C The use of a labelled in an Hawcroft, , Geary, (1996) nonradioactively probe system electrophoretic ribotyping method for the differentiation of strains of coagulase-negative staphylococci, Electrophoresis, 17, 55-57 6. Bibliography ££l C K H Differentiation of lactobacilli Hertel, , Ludwig, W, Pot, B, Kersters, K, Schleifer, (1993) occunng in fermented milk products by using oligonucleotid probes and electrophoretic protein patterns, Syst Appl Microbiol, 16 (3) 463-467 D H G W The bactena - a review J Milk Hettinga, , Reinbold, (1972) propionic-acid l-Growth, Food Technol, 35 (5) 295-301 D H G E R defect of Swiss cheese I Effect of Hettinga, , Rembold, W, Vedamuthu, (1974) Split strain of Propiombactenum and wrapping matenal, J Milk Food Technol, 37 (6) 322-328 D H G W defect of Swiss cheese II Effect of low Hettinga, , Reinbold, (1975) Split tempera¬ tures on the metabolic activity of Propiombactenum, J Milk Food Technol, 38 (1) 31-35 D M H Selektivmedien zum Nachweis von und fakultativ Isolini, , Grand, , Glattli, (1990) obligat heterofermentativen Laktobazillen, Schweiz Milchw Forschung, 19 (3) 57-59 J M J H interactions between acid Jimeno, , Lazaro, , Sollberger, (1995) Antagonistic propionic bacteria and non-starter lactic acid bactena, Lait, 75 (4-5) 401-414 Jimeno, J (1996) Personal communication, Biochemistry, FAM C S Cell wall and acid similanties Johnson, J L , Cummins, (1972) composition deoxyribonucleic among the anaerobic coryneforms, classical propionibactena, and strains of Arachma propionica, J Bactenol, 109,1047-1066 G RV A method for the isolation and Kamber, C S , Reinbold, W, Hussong, (1952) plating enumeration of propionibactena, J Dairy Sci, 35, 915-919 Klaenhammer, T R (1993) Genetics of bactenocins produced by lactic acid bactena, FEMS Mic¬ robiol Rev, 12, 39-86 Kurmann, J (1962) Eine methodische Verbesserung zur Bestimmung des Propionsaurebakte- nengehaltes der Emmentalerkasereimilch, XVI Int Dairy Congr C, 377-382 Langsrud, T, Rembold, G W (1973a) Flavor development and microbiology of Swiss cheese I Milk quality and treatments, J Milk Food Technol, 36 (10) 487-490 Langsrud, T, Reinbold, G W (1973b) Flavor development and microbiology of Swiss cheese II Starters, manufacturing processes and procedures, J Milk Food Technol, 36 (11), 531- 542 Langsrud, T, Reinbold, G W (1973c) Flavor development and microbiology of Swiss cheese III Ripening and flavor production, J Milk Food Technol, 36 (12) 593-609 Langsrud, T, Reinbold, G W, Hammond, E A (1977) Proline production by Propiombactenum shermanii P 59, J Dairy Sci, 60 (1) 16-23 Langsrud, T, Sorhaug, T, Vegarud, G E (1995) Protein degradation and ammo acid metabo¬ lism by propionibactena, Lait, 75 (4-5) 325-330 Liu, J AP, Moon, N J (1982) Commensahstic interaction between Lactobacillus acidophilus and Propiombactenum shermanii, Appl Environ Microbiol ,44, 715-722 O H N J A R J Protein measurement with the Lowry, , Rosebrough, , Farr, L, Randall, (1951) folin phenol reagent J Biol Chem, 193,265-275 zSSz 6. Bibliography W J B A Partial and characterization of a bacteriocin Lyon, , Glatz, (1991) purification produced by Propiombactenum thoenii, Appl Environ Microbiol, 57(3)701-706 W J B A Isolation and of a Lyon, , Glatz, (1993) purification propiomcin PLG-1, bactenocin pro¬ duced by a strain of Propionibacfenum thoenii, Appl Environ Microbiol, 59 (1) 83-88 A C G E R of Malik, , Reinbold, W, Vedamuthu, (1968) Evaluation the taxonomy of the Propio¬ mbactenum, Can J Microbiol, 14(11) 1185-1191 Mantere-Alhonen, S (1977) Uber neuere Untersuchungen der Propionsaurebaktenen Eine Ubersicht aus der Literatur, Meijeritiet Aikak, 35, 88-99 Mantere-Alhonen S (1982) Die Propionibaktenen der Molkerenndustne als Darmkanalmikro- Aikak 95-96 ben, Meijeritiet , 40, Mantere-Alhonen, S (1983) On the survival of Propiombactenum freudenreichii culture during in vitro gastric digestion, Meijeritiet Aikak, 41,19-23 S A new of sour milk with 45 Mantere-Alhonen, (1987) type propionibactena, Meijeritiet Aikak , (1) 49-61 Mantere-Alhonen, S (1989) Mineral elements in the cell mass of lactic acid and bifidobactena, Milchwiss, 44, 758-760 S E L Lactobacilli and in Mi¬ Mantere-Alhonen, , Ryhanen, (1994) propionibactena living food, crobiol Ahm Nutr, 12, 399-405 Mantere-Alhonen, S (1995) Propionibactena used as propiotics - a review, Lait, 75(4-5)447- 452 Maura E R M I L of PCR the Jutras, , Miller, , Pepper, (1995) Optimization arbitranly pnmed for identification of bactenal isolates, J Microbiol Methods, 24, 55-63 Menlainen, V, Anttla, M (1976) Uber die Propionsaurebaktenen im finnischen Emmentalerkase, Meijeritiet Aikak, 34,107-115 Metzler, DE (1977) Biochemistry, the chemical reactions of living cells, 553, Academic Press, New York Meunier, JR, Gnmont, PAD (1993) Factors affecting reproducibility of random amplified polymorphic DNA fingerpnnting Res Microbiol, 144, 373-379 Orla-Jensen, S (1909) Die Hauptlinien des nattlriichen Baktenensystems, Zentralblatt fur Bak- tenologie, Parasitenkunde und Infektionskrankheiten, Abteilung 2,22, 305-346 Oterholm, A (1967) Glycerol ester hydrolases of some lactic and propionic acid bactena, PhD Thesis, University of Illinois, Urbana USA Panon, G (1988) Presence of plasmids in propionic acid bactena Lait, 68 (1) 103-107 H S G E G WS Growth of at Park, , Reinbold, W, Hammond, , Clark, (1967a) propionibactena low temperatures, J Dairy Sci, 50 (4) 589-591 H S G E G Role of in defect of Park, , Reinbold, W, Hammond, (1967b) propionibactena split Swiss cheese, J Dairy Sci, 50 (6) 820-823 6. Bibliography zSSz. M S B A M Identification of lactic acid Patarata, L, Pimental, , Pot, , Kersters, K, Faia, (1994) bactena isolated from Portuguese wines and musts by SDS-page, J Appl Bactenol, 76 (3) 288-293 Paul, G E, Boothe, S J (1988) Properties and characteristics of a bactenocin-like substance produced by Propiombactenum acnes isolated from dental plaque, Can J Microbiol, 34, 1344-1347 Pesce de Ruiz A P G Interaction between Lactobacil¬ Perez Chaia, A, Holgado, , Oliver, (1982) lus helveticus and Propiombactenum freudenreichii subsp shermanii, Microbiol Ahm Nutr, 5, 325-331 Pesce de Ruiz A P G Effect of and Perez Chaia, A, Holgado, , Oliver, (1988a) pH temperature on the proteolytic acitvity of propionibactena, Microbiol Alim Nutr, 6, 91-94 Perez Chaia, A, Sesma, F, Pesce de Ruiz Holgado, A, Oliver, G (1988b) Screening of plas¬ mids in strains of Propiombactenum and mesophihc lactobacilli isolated from Swiss-type cheeses, Microbiol Alim Nutr, 6,171-174 Perez Chaia, A, Strasser de Saad, AM, Pesce de Ruiz Holgado, A, Oliver, G (1994) Com¬ petitive inhibition of Propiombactenum acidipropionici by mixed cultunng with Lactobacil lus helveticus, J Food Protect, 56 (4) 341-344 M F T F selective media for the enumeration of Peberdy, , Fryer, (1976) Improved propionibac¬ tena from cheese, N Z J Dairy Sci Technol, 11,10-15 Plastourgos, S, Vaughn, R H (1957) Species of Propiombactenum associated with zapatera spoilage of olive, Appl Microbiol, 5, 267-271 G J W H M W D E Effect of Pntchard, G , Wimponny, T, Moms, A, Lewis, A, Hughes, (1977) oxygen on Propiombactenum shermanii grown in continuous culture, J Gen Microbiol, 102(1)223-233 M De A F V Differentiation of Proudy, I, Gautier, , Carvalho, , Daval, S, Rioux, (1995) dairy Propiombactenum strains by Random Amplified Polymorphic DNA (RAPD), Poster A2, First International Symposium on Dairy Propionibactena, Rennes, France T G B A Characterization of DNA in Rehberger, , Glatz, (1987) plasmid Propiombactenum freudenreichii subsp globosum P93 evidence for plasmid-linked lactose utilization, J Dairy Sci, 70, suppl 1,80 T G B A Characterization of Envi¬ Rehberger, , Glatz, (1990) Propiombactenum plasmids, Appl ron Microbiol, 56 (4) 864-871 Rehberger, T G (1993) Genome analysis of Propiombactenum freudenreichii by pulsed-field gel electrophoresis, Curr Microbiol, 27,21-25 K -H T J Differentiation of "classical" nu¬ Riedel, J , Bntz, (1992) Propiombactenum species by mencal analysis of electrophoretic protein profiles, System Appl Microbiol, 15 (4) 567- 572 K -H J T J in anaerobic Riedel, , Bntz, (1993) Propiombactenum species diversity digesters, Biodiv Conserv, 2,400-411 K -H J B D T J Justification of the "classical" Riedel, , Wingfield, , Bntz, (1994) Propiombacte¬ num species concept by restriction analysis of the 16S nbosomal RNA genes, System Appl Microbiol, 17, 536-542 JOJL 6. Bibliography K -H J T J Justification of the Riedel, , Bntz, (1996) "classical" Propiombactenum species con cept by ribotyping, System Appl Microbiol, 19, 370-380 C K-H bactena with a DNA Roller, , Ludwig, W, Schleifer, (1992) Gram-positive high G+C content are charactenzed by a common insertion within their 23S rRNA genes J Gen Microbiol, 138, 1167-1175 J E J G Differentiation of Samelis, , Tsakalidou, , Metaxopoulos, , Kalantzopoulos, (1995) Lac¬ tobacillus sake and Lactobacillus curvatus isolated from naturally fermented Greek dry salami by SDS-page of whole-cell proteins, J Appl Bactenol, 78 (2) 157-163 Schaeren, W (1997) Personal communication, Milk Production, FAM K H W of the Lactobacillus and related Schleifer, , Ludwig, (1995) Phylogeny genus genera, System Appl Microbiol, 18 (4), 461-467 H E durch - Eine Ursa- Sebastiani, , Tschager, (1993) Succmatbildung Propionsaurebaktenen che der Nachgarung von Emmentaler', dmz, 4, 76-80 E E Possibilities of bactena in the of Sobczak, , Konieczna, (1987) using propionic production acetic acid from whey, Acta Aliment Pol, 13, 351-357 R R C D A statistical method for Sokal, , Michener, (1958) evaluating systematic relationships Univ Kansas Sci Bull, 38, 1409-1438 Sollberger, H (1996) Einflussfaktoren auf das Verhalten von Propionsaurebaktenen in Kase Si- tuationsanalyse zu Forschungsschwerpunkt 1996 der Sektion Kasetechnologie, FAM internal report 11,7-17 J N R Classification of an In Ber- Staley, T, Kneg , (1986) procaryotic organisms, overview, manual of P H N S M E gey's systematic bacteriology (Sneath, A, Mair, , Sharpe, , Holt, J G Vol The Williams & Wilkins , eds), 1,1-4, Co, Baltimore Steffen, C (1979) Vergleichende Untersuchungen in Emmentalerkasen mit und ohne Nachga¬ rung, Schweiz Milchw Forschung, 8 (3) 44-48 P O Evaluation of the RAPD for the classification of Tailhez, , Matte, (1995) technique dairy propionibactena strains, Poster A1, First International Symposium on Dairy Propionibac¬ tena, Rennes, France M N R Croissance des bactenes dans le Thierry, A, Madec, , Richoux, (1994) propioniques fromage comparaison de 2 milieux de denombrement, Lait 74 (2), 161-171 Thierry, A, Madec, M N (1995) Propionibactena count in raw milk, Lait, 75 (4-5), 315-323 Thompson, T.L, Marth, E H (1985) Changes in Parmesan cheese dunng npening microflora- coliforms, enterococci, anaerobes, propionibactena and staphylococci, Milchwiss, 41 (4) 201-205 P R P J H in der ThQrlemann, , Amrein, , Meier, , Hani, P, Schar, (1991) Propionsaurebaktenen Kasereimilch, FAM internal report, 15,1-16 R A H L T Punfication and charactenzaton Tobiassen, O, Pnpp, , Stepaniak, , Sorhaug, (1996) of an endopeptidase from Pmpiombactenum freudenreichii, J Dairy Sci, 79 (12) 2129- 2136 6, Bibliography -101- E E G of Tsakalidou, , Zoidou, , Kalantzopoulos, (1992) SDS-polyacrylamide gel electrophoresis cell proteins from Lactobacillus delbrueckn subsp bulgancus and Streptococcus salivar- ius strains isolated from and Milchwiss 47 subsp thermophilus yoghurt cheese, , (5) 296-298 E E E Tsakalidou, , Manolopoulou, E, Kabaraki, , Zoidou, , Pot, B, Kersters, K, Kalantzopoulos, G (1994) The combined use of whole-cell protein extracts for the identification (SDS- PAGE) and enzyme activity screening of lactic acid bactena isolated from traditional Greek dairy products, System Appl Microbiol, 17 (3) 444-448 R M I calves a milk Tuikov, , Knstova, , Rizvanov, K, Vladimirov, (1980) Early weaning using additive and a preparation of freeze-dned bacteria, Dairy Sci Abstr 1982,44,1689 Vladimirov, Rizvanov, K, Vasilev, A, Kenov, P (1977) Tests on replacement of antibiotics in sucking calf feeding by a freeze-dned preparation of bactena, Dairy Sci Abstr 40, 2118 Von E 0 Recherches surla fermentation dansle Freudenreich, , Jensen, (1906) propionique fromage d'Emmental, Tirage speciale de I'Annuaire agncole de la Suisse, 1-22 R J MA Metabolism of In The rumen Wallace, , Cotta, (1988) nitrogen-containing compounds, microbial P N Elsevier London ecosystem (Hobson, , ed), 185-216, Applied Science, K M S K Efficient of vitamin from Ye, M, Shijo, , Jin, , Shimizu, (1996) production B,2 propionic acid under vanation of dissolved J Ferment 82 bactena oxygen concentration, Bioeng , (5) 484-491 B P Vitamin from curd with Youngsmith, , Apiraktivongse, (1983) B,2 production soybean whey Propiombactenum freudenreichii, J Ferment Technol, 61,105-107 Zaugg, E (1995) Braune Tupfen im Halbhartkase harmlos aber unerwunscht, Schweiz Milchz, 121 (33) 1-9 J M J SDS-solubilized whole-cell Zouran, A, Commissaire, , Desmazeaud, (1992) protein pat¬ terns of Streptococcus salwanus subsp thermophilus and Lactobacillus delbruecku subsp bulgancus isolated from Greek yoghurts, J Dairy Res, 59 (1) 105-109 -102- 7. Annex 7. Annex Table 1: Propionibactena reference strains - colony description and morphology Strain Species Colony Morphology ATCC 6207 P freud subsp freud T beige single, oval or long ATCC 9614 P freud subsp sherm T beige single or in pairs, coccoid or oval DSM 20270 P freud subsp she/777 beige single, coccoid or oval P1409 P freud subsp shemj beige single, oval, partly club shaped P1410 P freud subsp sherm beige single, oval, partly club shaped P1411 P freud subsp sherm beige single, oval, partly club shaped P1412 P freud subsp sherm beige single, oval, partly club shaped P1413 P freud subsp sherm beige in pairs or groups, oval, partly club shaped P1414 P freud subsp shemj beige single or in pairs, oval P111 P freud subsp sherm beige single, oval, partly club shaped P112 P freud subsp sherm beige single, oval, partly club shaped ATCC 4868 P jensenii T beige single or in groups, coccoid or oval DSM 20274 P jensenii (zeae) white-beige in twos or chains, oval DSM 20278 P jensenii white-beige single or in pairs or groups, coc¬ coid or oval, DSM 20279 P jensenii (peterssonu) beige, single or in pairs, oval or long, catalase negative partly club shaped ATCC 4874 P thoenii J red-brown single or in pairs or groups, coc¬ coid or oval DSM 20275 P thoenii (rubrum) orange-brown single or in pairs, coccoid DSM 20277 P thoenii orange-brown, single or in pairs, coccoid or oval catalase negative ATCC P acidipropionici T white-beige, ca¬ single or in pairs, long, partly club 25562 talase negative shaped DSM 20272 P acidipropionici white-beige, ca¬ single or groups, oval or long, (pentosaceum) talase negative partly club shaped DSM 20273 P acidipropionici white-beige single or groups, long or very (arabinosum) long, partly club shaped Table 2: "Propionibactena" strains from MIBD Aargau (AG) - colony descnption and morpholo¬ gy Strain Date Ongin Colony Morphology AG2~1 03 01 94 OberrUti-Wyden, supplier 1 brownish in groups, oval, irregular AG3 1 03 01 94 OberrUti-Wyden, supplier 3 white-brownish single, coccoid or oval AG3 2 03 01 94 OberrUti-Wyden, supplier 3 red-brownish in groups, oval AG6 1 03 01 94 Oberruti-Wyden, supplier 11 white single, oval AG6 2 03 01 94 OberrOti-Wyden, supplier 11 yellow single, coccoid or oval AG7 1 03 01 94 Oberruti-Wyden, supplier 11 yellow with depressi¬ single, coccoid on 7. Annex cont AG8 1 03 01 94 Oberruti-Wyden, supplier 12 brown, catalase ne¬ single, coccoid or gative oval AG8 2 03 01 94 Oberruti-Wyden, supplier 12 light brown in groups, oval AG9 1 03 01 94 Oberruti-Wyden, supplier 13 dark brown with light single or in pairs, edge oval AG11 1 03 01 94 Oberruti-Wyden, supplier 15 white-brownish in groups, coccoid AG112 03 01 94 Oberruti-Wyden, supplier 15 yellowish single, coccoid AG131 26 01 94 Geltwil, supplier 1 white with depression single or in groups, coccoid AG13 3 26 01 94 Geltwil, supplier 1 white single or in pairs, coccoid AG14 1 26 01 94 Geltwil, supplier 2 beige with depression single or in pairs, coccoid AG15 2 26 01 94 Geltwil, supplier 9 yellowish with de¬ single, coccoid or pression oval AG15 3 26 01 94 Geltwil, supplier 9 white with depression single, coccoid, various sizes AG171 24 03 94 Hendschiken, supplier 5 beige, irregular single or in pairs or groups,coccoid or oval AG181 24 03 94 Hendschiken, supplier 6 yellow single or in pairs, coccoid or oval AG18 2 24 03 94 Hendschiken, supplier 6 beige, irregular single or in pairs, coccoid AG21 1 24 03 94 Hendschiken, supplier 13 beige single or in pairs or groups, coccoid or oval AG22 1 30 05 94 Reussegg, supplier 7 brown with light edge single or in pairs, oval, club shaped AG22 2 30 05 94 Reussegg, supplier 7 beige single or in groups, oval, club shaped AG241 30 05 94 Reussegg, supplier 11 yellow single or in pairs or groups, oval AG25 1 30 05 94 Reussegg, supplier 15 beige, irregular single, oval or long AG25 2 30 05 94 Reussegg, supplier 15 light brown single, oval AG25 3 30 05 94 Reussegg, supplier 15 brown-yellowish single or in pairs or groups, coccoid or oval AG25 4 30 05 94 Reussegg, supplier 15 beige single or in pairs or groups, coccoid AG261 30 05 94 Oberruti-Wyden, supplier 2 light brown single or in pairs, coccoid or oval AG271 30 05 94 Oberruti-Wyden, supplier 4 light brown single, oval AG281 30 05 94 Oberruti-Wyden, supplier 7 red single or in pairs or groups, oval AG28 2 30 05 94 Oberruti-Wyden, supplier 7 beige single or in pairs or groups, oval or long AG29 1 30 05 94 Oberruti-Wyden, supplier 11 beige-yellowish single, oval or long AG301 30 05 94 Oberruti-Wyden, supplier 12 beige-brownish single or in groups, oval or long, club shaped AG31 1 30 05 94 Oberruti-Wyden, supplier 13 brown with light edge single, oval AG32 1 30 06 94 Muhlau, kettle milk light brown-orange single or in pairs or groups, oval ziMz 7, Annex cffloL AG33 1 30 06 94 Reussegg, tank brown with light edge single, oval AG34 1 30 06 94 MUhlau, supplier 17 light brown-orange single, oval AG361 30 06 94 Reussegg, supplier 7 light brown-orange single or in pairs or groups, coccoid AG371 30 06 94 Reussegg, supplier 8 brown with light edge single or in groups long AG381 30 06 94 Reussegg, supplier 11 light brown-orange single or in pairs, oval AG391 30 06 94 Oberruti-Wyden, kettle milk red-brown single or in pairs or groups, coccoid AG40 1 30 06 94 Oberruti-Wyden, supplier 4 brown with light edge single, oval AG42 1 30 06 94 Oberruti-Wyden, supplier 13 yellow single or in pairs or groups, coccoid AG42 2 30 06 94 Oberruti-Wyden, supplier 13 brown with light edge single or in pairs or groups, coccoid AG451 26 07 94 MUhlau, supplier 17 beige single or in pairs or groups, coccoid AG46 1 26 07 94 MUhlau, supplier 18 beige, catalase nega¬ single or in groups, tive oval AG46 2 26 07 94 MUhlau, supplier 18 brown single or in pairs, oval AG47 1 26 07 94 Muhlau, supplier 26 beige single or in groups, coccoid or oval AG48 1 26 07 94 Muhlau, supplier 27 light brown single, long AG48 2 26 07 94 Muhlau, supplier 27 light brown-yellowish single or in pairs, coccoid AG491 26 07 94 Reussegg, supplier 5 beige-yellowish single or in groups, coccoid AG501 26 07 94 Reussegg, supplier 7 beige single or in pairs or groups, coccoid or oval AG51 1 26 07 94 Reussegg, supplier 8 light brown single or in groups, oval AG521 26 07 94 Reussegg, supplier 15 brown single, oval or long AG53 1 26 07 94 Oberruti-Wyden, supplier 13 brown with light edge single or in pairs, oval AG541 22 08 94 MUhlau, supplier 18 beige single or in pairs, coccoid or oval AG56 1 22 08 94 MUhlau, supplier 27 beige, catalase nega¬ single or in pairs, tive coccoid AG58 1 22 08 94 Reussegg, supplier 15 matt brown single or in groups, oval AG591 22 08 94 Reussegg, tank beige single or in groups, coccoid or oval AG601 22 08 94 MUhlau, kettle milk light brown single, oval AG62 1 07 10 94 MUhlau, supplier 18 white-beige single, oval or very long AG63 1 0710 94 Muhlau, supplier 20 matt brown, catalase single, long, partly negative club shaped AG64 1 07 10 94 Reussegg, supplier 8 brown with light edge single, long or very long AG65 1 07 10 94 Reussegg, tank beige-yellowish in pairs or groups, coccoid AG66 1 0710 94 MUhlau, kettle milk beige, catalase nega¬ single or in pairs, tive oval 7 Annex zlQSz SSUL AG671 07 10 94 Oberruti-Wyden, supplier 12 light brown single, oval or long AG68 1 07 10 94 Oberruti-Wyden, supplier 13 brown with light edge single, oval AG691 07 10 94 Oberruti-Wyden, kettle milk brown single, long AG71 1 17 11 94 OberrUti-Wyden, supplier 2 white-beige single or in pairs or groups, oval AG72 1 17 1194 OberrUti-Wyden, supplier 3 brown with light edge single, oval AG731 17 11 94 OberrUti-Wyden, supplier 4 brown with light edge single or in pairs, oval AG751 17 1194 Oberruti-Wyden, supplier 12 brown with light edge single or in pairs or groups, oval Table 3: "Propionibactena" strains from MIBD Bern (BE) - colony description and morphology Strain Date Origin Colony Morphology BE1 1 11 01 94 Schwendimatte, kettle milk brown with ligth edge single, oval or very long BE21 10 01 94 Schwendimatte, kettle milk matt brown with de¬ sinlge, long pression BE31 10 01 94 Schwendimatte, whey matt brown single, oval BE4 1 04 01 94 Nods, kettle milk 2 matt brown single, oval or long BE61 25 01 94 Lignieres, supplier 40 beige, catalase nega¬ single, coccoid tive BE71 25 01 94 Lignieres, supplier 2 white-beige, catalase single or in pairs, negative coccoid or oval BE81 25 01 94 Nods, kettle milk 1 beige single, oval BE8 3 25 01 94 Nods, kettle milk 1 light brown single, coccoid or oval BE9 1 25 01 94 Schwendimatte, kettle milk brown with light edge single, oval or long BE9 2 25 01 94 Schwendimatte, kettle milk light brown in groups, oval BE101 22 02 94 Schwendimatte, delivery brown with light edge single, oval or long milk BE111 08 03 94 Schwendimatte, kettle milk white single or in pairs or groups, oval BE12 1 11 04 94 Schwendimatte, kettle milk brown with light edge single or in groups, long or very long BE12 2 1104 94 Schwendimatte, kettle milk light brown single or in groups, long BE131 26 04 94 Schwendimatte, delivery matt brown single or in pairs or milk groups, long, cur¬ ved BE161 25 05 94 Schwendimatte, kettle milk brown with yellowish single, oval edge BE171 2106 94 Schwendimatte, kettle milk brown single or in pairs, long BE17 2 21 06 94 Schwendimatte, kettle milk light brown in groups, very long, club shaped BE181 04 07 94 Schwendimatte, delivery brown single or in groups, milk long or very long BE191 08 08 94 Schwendimatte, kettle milk brown single, oval or long BE201 08 08 94 Schwendimatte, delivery brown smgl, oval milk BE21 1 09 08 94 Schwendimatte, kettle milk brown single or in groups, long or very long -106- 7, Annex cont BE22 1 09 08 94 Schwendimatte, delivery beige in groups, long or milk very long BE22 2 09 08 94 Schwendimatte, delivery brown single, coccoid or milk oval BE23 1 06 09 94 Schwendimatte, kettle milk beige-yellowish single or in pairs, oval BE25 1 22 11 94 Burgen, supplier 18 matt brown single or in groups, oval BE26 1 22 11 94 Schwendimatte, supplier 6 brown with light edge single, long, cur¬ ved BE27 1 22 1194 Niederstocken, supplier milk beige single or in pairs, coccoid or oval BE28 1 22 11 94 Niederstocken, supplier milk beige single or in pairs, oval BE29 1 22 1194 Niederstocken, supplier milk brown single, long, cur¬ ved BE30 1 22 11 94 Schwendimatte, whey white-beige single, oval Table 4: "Propionibactena" strains from MIBD Fnbourg (FR) - colony description and morpholo¬ gy Strain Date Origin Colony Morphology FR1 1 21 01 94 945, supplier 14 matt brown single or in groups, coccoid FR21 21 01 94 945, supplier 1 matt brown, catalase single or in groups, oval negative FR3 1 21 01 94 945, supplier 12 red-brown single or in groups, long FR4 1 27 01 94 861, supplier 6 beige single, coccoid FR51 27 01 94 1018, supplier 3 matt brown single, coccoid or oval FR5 2 27 01 94 1018, supplier 3 red-brownish in groups, oval FR7 1 03 02 94 861, kettle 2 beige single or in pairs, oval or long FR7 2 03 02 94 861, kettle 2 brown with light edge single or in pairs, oval FR81 03 02 94 849, supplier 9 brown with light edge single or in pairs, oval FR91 03 02 94 889, supplier 5 beige in groups, coccoid FR11 1 08 02 94 852, grains kettle 1 brown in pairs or groups, oval, irre¬ gular FR112 08 02 94 852, grams kettle 1 red-brownish single or in pairs or groups, oval, irregular FR12 1 08 02 94 847, supplier 8 brownish-yellowish in groups, coccoid FR12 2 08 02 94 847, supplier 8 brown single or in pairs, oval, club shaped FE131 16 02 94 889, supplier 5 brown with light edge single, oval FR15 1 24 02 94 967, supplier 8 brown with light edge single, coccoid or oval FR161 24 02 94 887, supplier 11 brown with light edge single, coccoid FR171 02 03 94 848, supplier 4 red-brown in groups, coccoid or oval FR17 2 02 03 94 848, supplier 4 white single or in chains, coccoid FR17 4 02 03 94 848, supplier 4 brownish-yellowish single, oval with white edge FR17 5 02 03 94 848 supplier 4 white, catalase nega¬ single or in groups, coccoid tive FR18 1 02 03 94 889, supplier 5 yellow in groups, coccoid or oval FR18 2 02 03 94 889, supplier 5 brown with light edge single coccoid or oval 7. Annex zWz. cont FR18 3 02 03 94 889, supplier 5 light brown single, long or very long, partly curved FR19 2 02 03 94 931, supplier 1 brown, irregular single, coccoid or oval FR20 1 02 03 94 917, supplier 25 brown single or in groups, oval FR20 2 02 03 94 917, supplier 25 light brown single, long FR20 3 02 03 94 917, supplier 25 beige-yellowish single or in groups, coccoid or oval FR21 1 02 03 94 1043, supplier 14 beige single, coccoid or long FR212 02 03 94 1043, supplier 14 brown with light edge single, oval, club shaped FR21 3 02 03 94 1043, supplier 14 yellow single or in groups, long FR22 1 02 03 94 847, supplier 3 beige single, coccoid or oval FR22 2 02 03 94 847, supplier 3 brown with light edge single, oval FR22 3 02 03 94 847, supplier 3 yellow single or in pairs, oval FR23 1 22 03 94 999, supplier 14 brown with light edge single, oval FR23 2 22 03 94 999, supplier 14 beige, catalase nega¬ single or in pairs, oval tive FR23 3 22 03 94 999, supplier 14 brown-yellowish single or in pairs or groups, oval FR24 1 22 03 94 861, supplier 15 brown with light edge single or in pairs, oval FR24a 2 07 06 94 966, kettle milk 3 beige single or in groups, coccoid or oval FR251 07 06 94 966, supplier 20 brown, catalase ne¬ single, oval gative FR261 07 06 94 880, centrifuged brown with light edge single, oval milk FR27 1 07 06 94 924, kettle milk 4 brown with light edge single, oval or long FR28 1 13 06 94 847, kettle milk 1 beige-brownish in pairs or groups, coccoid or oval FR28 2 13 06 94 847, kettle milk 1 brown with light edge single, coccoid or oval FR29 1 13 06 94 912, supplier 2 beige single or in pairs or groups, coccoid FR29 2 13 06 94 912, supplier 2 brown with light edge single, coccoid or oval FR30 1 21 06 94 917, supplier 17 brown with light edge single, oval FR31 1 21 06 94 917, supplier 18 brown with light edge single, oval or long FR31 2 21 06 94 917, supplier 18 beige single or in pairs or groups, long, club shaped FR32 1 12 07 94 895, supplier 1 beige single or in groups, coccoid FR33 1 12 07 94 917, supplier 18 beige single, oval FR341 06 09 94 949, supplier 22 beige-yellowish single or in pairs or groups, oval FR35 1 06 09 94 999, supplier 9 red-brown single or in pairs or groups, oval FR352 06 09 94 999, supplier 9 matt brown single or in groups, oval FR361 06 09 94 839, skimmed milk matt brown single or in pairs or groups, oval FR37 1 12 09 94 1059, supplier 6 brown with light edge single or in pairs, oval FR38 1 12 09 94 847, supplier 5 matt brown single or in groups, oval FR38 2 12 09 94 847, supplier 5 beige-yellowish single or in pairs, coccoid FR38 3 12 09 94 847, supplier 5 brown with light edge single or in pairs or groups, long FR391 12 09 94 1059, supplier 5 light brown single or in pairs, oval or long FR40 1 23 09 94 840, supplier 12 beige-yellowish single or in pairs or groups, coccoid or oval zMz. 7. Annex cont. FR40 2 23 09 94 840, supplier 12 red-brown single or in pairs or groups, coccoid or oval FR41 1 23 09 94 840, supplier 4 brown with light edge single, oval, partly club sha¬ ped FR421 04 10 94 929, supplier 20 dark red-brown single, oval FR43 1 04 10 94 895, kettle milk 2 matt brown single, long FR43 2 04 10 94 895, kettle milk 2 brown with light edge single, oval FR44 1 04 10 94 861, kettle milk brown with light edge single, oval FR45 1 11 1094 986, supplier 12 light brown single or in pairs or groups, coccoid FR46 1 11 10 94 948, kettle milk 2 white-beige, catalase single or in pairs or groups, negative oval FR47 1 1910 94 880, skimmed milk brown with light edge single, oval or long FR48 1 1910 94 880, supplier 6 light brown single, coccoid or oval FR49 1 24 10 94 862, supplier 5 beige, catalase nega¬ single, coccoid tive FR50 1 24 10 94 906, supplier 29 beige-brownish single or in groups, coccoid FR51 1 24 10 94 930, supplier 1 red-brown single or in pairs or groups, long or very long FR52 1 1011 94 887, kettle milk brown with light edge single, oval FR53 1 10 11 94 966, kettle milk 2 red-brown single or in pairs or groups, coccoid FR53 2 10 11 94 966, kettle milk 2 white, catalase nega¬ single or in pairs or groups, tive coccoid FR53 3 10 1194 966, kettle milk 2 brown with light edge single or in pairs, coccoid FR55 1 25 11 94 966, supplier 28 white, catalase nega¬ single or in pairs, oval tive FR55 2 25 11 94 966, supplier 28 matt brown single or in pairs, oval or long, club shaped FR56 1 25 1194 966, supplier 23 brown with light edge single or in pairs, oval FR57 1 29 1194 846, kettle milk 1 matt brown single, oval FR581 291194 880, supplier 6 matt brown single or in pairs, oval FR591 29 11 94 880, supplier 5 brown single, oval or long FR60 1 0612 94 986, supplier 23 red-brown single or in pairs, coccoid or oval FR60 2 0612 94 986, supplier 23 matt brown single, coccoid or oval FR61 1 0612 94 986, supplier 19 brown with light edge single, oval or long FR621 06 12 94 986, supplier 9 red-brown single, oval FR631 28 12 94 887, supplier 7 brown with light edge single or in pairs, oval FR64 1 28 12 94 851, supplier 16 brown with light edge single or in pairs, oval, cur¬ ved FR64 2 2812 94 851, supplier 16 beige-yellowish single or in pairs, oval, large FR651 28 12 94 851, supplier 17 brown with light edge single or in pairs, oval FR401 1 06 06 95 Grangeneuve beige, catalase nega¬ single, oval tive FR403 1 27 06 95 Grangeneuve red-brown single, very long, club sha¬ ped, curved FR4051 1107 95 Grangeneuve red-brown single or in groups, coccoid FR405 2 11 07 95 Grangeneuve beige-brownish single or in groups, coccoid FR407 1 02 08 95 Grangeneuve beige, catalase nega¬ single or in pairs or groups, tive coccoid or oval FR4131 30 08 95 Grangeneuve red single, coccoid FR4141 05 09 95 Grangeneuve brown single, oval or long 7. Annex -109- Table 5: "Propionibactena" strains from MIBD Nordostschweiz (NO) - colony description and morphology Strain Date Origin Colony Morphology N01 1 28 01 94 SchUbelbach yellow single or in pairs or groups, coccoid N02 1 15 02 94 Seelmatten brown single, oval or long N02 2 15 02 94 Seelmatten beige single, long or very long N03 1 09 02 94 ? beige, irregular single or in groups, coccoid or oval N03 2 09 02 94 Seelmatten light brown single, coccoid or oval N03 3 09 02 94 Seelmatten light brown single or in pairs, coccoid or oval N05 1 07 03 94 Steinenbrucke light brown with light single, oval or long edge N06 1 07 03 94 Ringwil light brown single, oval or long N06 2 07 03 94 Ringwil brownish with light single or in pairs, long edge N07 1 07 03 94 Vicosoprano, kettle white single, oval, club shaped milk N081 07 03 94 Savognin, tank light brown with light single, oval edge N091 07 03 94 Tinzen, tank light brown single, oval N09 2 07 03 94 Tinzen, tank brown with light edge single, long or very long N09 3 07 03 94 Tinzen, tank white single, coccoid or oval NO10 1 26 03 94 Bettswil light yellow single or in pairs or groups, oval NO10 2 26 03 94 Bettswil light brown single, oval N011 1 1104 94 Sternenberg light brown with light single, oval or long edge N011 2 11 04 94 Stemenberg beige single or in pairs or groups, oval N011 3 1104 94 Sternenberg red-brown single or in pairs or groups, coccoid N012 1 15 04 94 Rieti light brown single, oval N0131 02 05 94 Gebertingen beige-brownish single, coccoid or oval N0151 24 05 94 Matten, kettle milk brown single, oval 2 N0161 06 06 94 Gebertingen, kettle red-brown single, coccoid or oval milk N0181 1107 94 Herrgass, supplier beige-brownish single, oval 2 N0191 22 07 94 Alp Madan, kettle light brown single, coccoid or oval milk N021 1 17 08 94 Goldau brown with light edge single, oval N022 1 07 09 94 Herschmettlen matt brown single, oval N0231 0310 94 Bauma matt brown single, oval or long N0241 14 1194 Matten brown single or in pairs, oval N0251 21 11 94 Unteragen, kettle light brown single or in pairs or groups, milk 2 coccoid or oval N0261 11 1194 Warthausen, kettle white-brownish single or in pairs or groups, milk oval N0271 22 1194 Seelmatten, kettle brown single or in pairs, coccoid milk N028 1 12 12 94 unknown beige single or in pairs, oval N029 1 19 12 94 Wih beige single or in pairs, coccoid or oval -110- 7 Annex cont NO30 1 19 12 94 Rieden brown with light edge single or in pairs, coccoid or oval NO30 2 1912 94 Rieden matt brown single or in pairs, oval Table 6: "Propionibactena" strains from MIBD Nordwestschweiz (NW) - colony description and morphology Strain Date Origin Colony Morphology NW1 1 20 01 94 unknown matt brown single, coccoid or oval NW12 20 01 94 unknown red-brown single or in pairs or groups, coccoid NW21 20 01 94 unknown light brown single, oval or long NW3 1 24 03 94 unknown light brown single or in pairs, oval or long NW4 1 24 03 94 unknown brown with light edge single, oval NW4 2 24 03 94 unknown beige-yellowish single or in pairs or groups, coccoid or oval NW5 1 24 03 94 unknown brown single or in pans, coccoid or oval NW61 24 03 94 unknown brown with light edge single, coccoid or oval NW71 28 03 94 263, kettle milk 1 beige single or in pairs or groups, oval NW8 1 28 03 94 296, kettle milk 2 light brown single or in pairs, coccoid or oval NW9 1 28 03 94 255, kettle milk 2 red-brown single or in pairs or groups, coccoid or oval NW9 2 28 03 94 255, kettle milk 2 white single or in pairs or groups, oval NW101 28 03 94 263, kettle milk 2 white single or in pairs or groups, coccoid NW11 1 28 03 94 263, kettle milk 3 light brown single or in pairs or groups, coccoid NW121 31 03 94 214, kettle milk 1 light brown with light single or in pairs, oval or edge long NW131 31 03 94 217, kettle milk 1 brown with light edge single, oval NW13 2 31 03 94 217, kettle milk 1 light brown single or in pairs, oval NW13 3 31 03 94 217, kettle milk 1 beige single or in pairs, long or very long NW14 1 31 03 94 217, kettle milk 2 light brown with light single or in pairs, oval edge NW16 1 10 06 94 255, kettle milk 1 brown with light edge single or in pairs, oval or long NW171 10 06 94 263, kettle milk 3 yellow single or in pairs or groups, coccoid or oval NW17 2 10 06 94 263, kettle milk 3 light brown single, oval NW181 10 06 94 255, kettle milk 2 beige, catalase nega¬ single or in pairs, oval or tive long NW18 2 10 06 94 255, kettle milk 2 brown with light edge single, oval NW191 10 06 94 296, kettle milk 3 beige-yellowish single or in groups, oval NW201 10 06 94 296, kettle milk 2 beige single or in pairs or groups, oval NW20 2 10 06 94 296, kettle milk 2 brown single, oval NW20 3 10 06 94 296, kettle milk 2 brown with light edge single, oval 7. Annex -111- cont. NW221 10 06 94 263, kettle milk 1 beige single, coccoid or oval NW231 10 06 94 263, kettle milk 2 yellow single or in groups, oval or very long NW23 2 10 06 94 263, kettle milk 2 white-beige single or in groups, oval NW24 1 10 06 94 217, kettle milk 2 brown with light edge single or in pairs, oval NW251 10 06 94 214 kettle milk 2 light brown single or in pairs or groups, oval or long NW26 1 10 06 94 270, supplier 16 light brown-orange single, oval NW271 10 06 94 270, supplier 14 red-brown single, coccoid or oval NW27 2 10 06 94 270, supplier 14 light brown single or in pairs or groups, oval NW28 1 10 06 94 270, supplier 11 brown single, oval NW301 10 06 94 278, kettle milk 1 brown with light edge single or in groups, oval NW31 1 10 06 94 451, kettle milk 2 light brown single, oval NW32 2 10 06 94 451, kettle milk 1 beige single or in pairs or groups, oval or long, club shaped NW331 10 06 94 270, supplier 17 red single, oval or long, partly club shaped NW33 2 10 06 94 270, supplier 17 beige-brownish single, coccoid or oval NW33 3 10 06 94 270, supplier 17 brown with light edge single, coccoid or oval NW342 10 06 94 278, kettle milk 2 brown single, oval NW361 10 06 94 270, kettle milk 1 red-brown single or in pairs, coccoid NW371 10 06 94 270, kettle milk 2 brown with light edge single or in groups, oval NW381 23 06 94 unknown brown with light edge single, cocoid or oval NW391 23 06 94 unknown brown with light edge single, oval Table 7: "Propionibactena" strains from MIBD St Gallen-Appenzell AR (SG, AR) - colony descnption and morphology Strain Date Origin Colony Morphology SG1 1 1512 93 supplier 4 white-yellowish, irre¬ single or in pairs, oval gular SG12 15 12 93 supplier 4 white, catalase nega¬ single, coccoid tive SG2 1 15 12 93 supplier 15 brownish in pairs or groups, oval SG2 2 15 12 93 supplier 15 brownish in pairs or groups, oval SG3 2 1512 93 supplier 18 yellow-brown in groups, coccoid SG4 1 13 0194 supplier 53 beige single, oval SG51 13 0194 supplier 60 yellow-brown in groups, coccoid SG61 18 02 94 supplier 77 light brown single, oval SG71 18 02 94 supplier 77 brown single, coccoid or oval SG9 1 03 06 94 supplier 51 beige-brown single, coccoid or oval SG101 03 06 94 supplier 55 red-brown single or in pairs, coccoid or oval SG10 2 03 06 94 supplier 55 beige single, coccoid or oval SG11 1 28 06 94 supplier 21 brown with light edge single, oval SG121 03 06 94 supplier 29 brown single or in pairs, oval SG131 05 07 94 supplier 106 beige single or in pairs or groups, coccoid SG141 05 07 94 supplier 101 beige-yellow single or in pairs, coccoid or SG161 28 07 94 kettle milk 12 red-brown single, oval SG171 28 07 94 supplier 43 brown with light edge single, oval -112- 7. Annex cont. SG18 1 13 08 94 supplier 95 beige single, long or very long SG191 13 08 94 supplier 92 white-beige single or in pairs or groups, coccoid or oval SG201 07 09 94 supplier 74 matt brown single, long SG22 1 07 09 94 supplier 83 brown with light edge single or in pairs or groups, oval SG231 07 09 94 supplier 94 light brown-yellow in pairs or groups, coccoid SG24 1 20 09 94 supplier 36 brown with light edge single, oval SG25 1 20 09 94 supplier 32 brown single, oval SG27 1 05 10 94 kettle milk 67 matt brown single or in pairs, oval SG281 05 10 94 supplier 106 brown with light edge single or in pairs, oval or long SG291 17 11 94 supplier 102 brown with light edge single, oval or long SG30 1 17 11 94 supplier 44 light brown single, oval SG31 1 17 11 94 supplier 100 light brown, catalase single, oval negative SG331 22 11 94 supplier 2 brown with light edge single, oval AR1 1 26 01 94 supplier 48-33-G brown with light edge single, coccoid or oval AR12 26 01 94 supplier 48-33-G beige single or in pairs, coccoid Table 8: "Propionibactena" strains MIBD Thurgau (TG) - colony description and morphology Strain Date Origin Colony Morphology TG1 1 08 12 93 Herdern, grains brown with light edge single or in pairs, oval TG12 08 12 93 Herdern, grains yellowish in groups, coccoid TG2 1 13 12 93 Sirnach, kettle milk brownish in pairs, oval, partly club shaped TG3 1 15 12 93 Herdern, grains brown with light edge single, very long, club sha¬ ped, curved TG3 2 15 12 93 Herdern, grains brown with light edge single, coccoid TG4 1 03 01 94 Hosenruck, deli¬ white with beige edge single or in pairs, coccoid very milk with depression TG4 2 03 01 94 Hosenruck, deli¬ matt brown single, coccoid or long, very milk partly club shaped TG4 3 03 01 94 Hosenruck, deli¬ beige single, oval or long partly very milk club shaped TG51 110194 Lanzenneunforn beige single, oval TG6 1 17 01 94 Stelzenhof, kettle brown with light edge in pairs, oval or long milk 1 TG8 1 27 01 94 Holzhof, tank beige single, oval or very long Weiningen TG9 1 03 02 94 Leutenegg, kettle beige single, coccoid or oval milk TG101 10 02 94 Buch, grains beige in groups, coccoid or oval TG11 1 23 02 94 Herdern, evening beige, irregular single, oval milk TG121 24 02 94 Sonnental, whey 2 beige single, oval or long, club shaped TG131 07 03 94 Fruthwilen light brown with light single or in pairs or groups, edge oval or long TG14 1 14 03 94 Fnltschen, before beige single or in pairs, oval termization 7. Annex -113- cont TG151 28 03 94 Happerswil, after light brown with light single, coccoid or oval termrzation edge TG16 1 07 04 94 Stelzenhof, grains beige single or in pairs or groups, 2 coccoid TG17 1 12 04 94 Hagenbuch, before yellowish with white single or in pairs, coccoid termization edge TG17 2 12 04 94 Hagenbuch, before red-brown single or in groups, coccoid termization TG18 1 21 04 94 Sirnach, grains light brown single, oval TG19 1 05 05 94 Engishofen light brown single or in pairs, oval TG20 1 04 05 94 Herdern, grains brown with light edge smlge, oval TG21 1 18 05 94 Herdern, grains brown with light edge single, oval TG22 1 19 05 94 wangi beige single, oval TG23 1 13 06 94 Loh, kettle milk beige single or in groups, oval TG23 2 13 06 94 Loh, kettle milk beige single or in pairs or groups.coccoid TG24 1 16 06 94 Happerswil, after brown with light edge single, coccoid or oval termization TG25 1 27 06 94 Bichelsee, before brown with light edge single or in groups, coccoid termization or oval TG261 04 07 94 Riethof white single or in pairs, oval or long TG27 1 07 07 94 Dozwil, skimmed brown single, oval milk TG301 16 08 94 Fnltschen, tank brown single, coccoid or oval TG31 1 18 08 94 Happerswil, supp¬ beige single, oval lier 10 TG321 25 08 94 Leutenegg, whey 3 red-brown single or in pairs or groups, oval TG34 1 01 09 94 wangi, tank 1 red-brown single or in pairs or groups, oval TG35 1 0109 94 Wangi, light brown single, oval TG361 07 09 94 Herdern, after brown with light edge single or in groups, oval or termization very long TG37 1 13 09 94 Rothenhausen light brown single or in pairs, oval TG381 21 09 94 Dozwil light brown single, oval TG39 2 21 09 94 Herdern, after brown single or in pairs, oval termization TG401 27 09 94 Happerswil, grains brown with light edge single, oval, partly club sha¬ ped TG41 1 1310 94 Herder, supplier 1 brown with light edge single, oval, club shaped TG421 101194 Altishausen, kettle brown with light edge single, oval, club shaped milk 2 TG431 1212 94 Bettwiesen, grains brown single, oval, irregular -114- 7, Annex Table 9: "Propionibactena" strains from MIBD Vaud-Geneve (VD) and Neuchatel (NE) - colony descnption and morphology Strain Date Origin Colony Morphology VD1 1 10 12 93 Provence brown single, oval VD2 1 24 01 94 Villars-Burquin, brown with light edge single or m pairs, long kettle milk VD3 1 28 02 94 Chattilens brownish-greenish single or in pairs or groups, coccoid or oval VD3 2 28 02 94 Chattilens white single, coccoid or oval VD4 1 09 03 94 Villars-Burquin red-brown in pairs, coccoid or oval VD5 1 15 03 94 Montncher brown, irregular single, oval VD5 2 15 03 94 Montncher light brown single or in groups, oval VD7 1 08 06 94 Ballens brown with light edge single, oval VD81 06 07 94 Ballens brown with light edge single, oval VD9 1 19 07 94 Ballens brown with light edge single, oval VD10 1 11 10 94 Peney-le-Forat brown with light edge single, oval VD10 2 11 10 94 Peney-le-Forat beige single, oval VD11 1 26 10 94 Ropraz beige-yellowish single or in groups, oval VD112 26 10 94 Ropraz brown single, oval VD12 1 1612 94 Bussy sur Moudon light brown single or in pairs, oval or long NE1 1 08 01 94 unknown brown with light edge single or in groups, coccoid NE2 1 12 01 94 unknown matt brown single, long or very long, partly club shaped Table 10: "Propionibactena" strains from MIBD Zenfralschweiz (ZS) - colony descnption and morphology Strains Date Origin Colony Morphology ZS1 1 20 12 93 Alpnach-Dorf, deli¬ light brown single or in pairs, oval very milk ZS2 1 20 12 93 Alpnach-Dorf, brown with light edge single, oval evening milk ZS2 2 20 12 93 Alpnach-Dorf, yellowish single, long or very long evening milk ZS2 3 20 12 93 Alpnach-Dorf, yellow-brown in pans or groups, coccoid evening milk ZS2 4 20 12 03 Alpnach-Dorf, brownish single or in groups, oval evening milk ZS2 5 2012 93 Alpnach-Dorf, red-brown in pairs or groups, oval evening milk ZS3 1 07 01 94 Schurfmoos yellowish single, oval ZS4 1 07 01 94 Kerns-Dorf, kettle brown single, oval milk ZS51 07 01 94 Kerns-Dorf, kettle brown single or in pairs, oval or milk long ZS5 2 07 01 94 Kerns-Dorf, kettle white single, coccoid or oval milk ZS5 3 07 01 94 Kerns-Dorf, kettle brownish single or in pairs, oval milk ZS61 11 01 94 Alpnachdorf, deli¬ red-brown with light single or in groups, oval very milk edge ZS7 1 11 01 94 Alpnach-Dorf, brown with light edge single, oval or long, partly kettle milk 1 club shaped 7. Annex -115:. cont. ZS8 1 03 01 94 St Niklausen, light brown with light single, oval or long, partly supplier milk edge club shaped ZS91 08 02 94 Schintmoos beige single, coccod or oval, partly club shaped ZS101 08 02 94 Schintmoos white in pairs or groups, oval ZS11 1 08 02 94 Sifelen red-brown in groups, coccoid ZS131 10 03 94 Kerns-Dorf yellow single or in groups, oval ZS13 2 10 03 94 Kerns-Dorf light brown single, long or very long, partly curved ZS141 10 03 94 Alpnach-Dorf beige single, oval ZS15 1 10 03 94 Schintmoos white single or in groups, coccoid ZS15 2 10 03 94 Schintmoos light brown single, long ZS15 3 10 03 94 Schintmoos brown-greenish with single, oval light edge ZS16 1 10 03 94 Alpnach-Dorf, beige-yellowish in pairs or groups, coccoid evening milk ZS16 2 10 03 94 Alpnach-Dorf, light brown single, long or very long, evening milk curved ZS16 3 10 03 94 Alpnach-Dorf, red-brown single or in pairs, coccoid evening milk ZS16 4 10 03 94 Alpnach-Dorf, brown-greenish with single, long or very long, evening milk light edge partly club shaped ZS17 1 28 03 94 Rotmoos, supplier beige single or in pairs, coccoid or milk oval ZS18 1 28 04 94 Bodenberg, milk brown with light edge single, long sieve ZS191 2804 94 Hildisneden beige single or in pairs, oval ZS19 2 28 04 94 Hildisneden brown with light edge single or in pairs, oval ZS20 1 28 04 94 Hildisneden light brown with light single or in pairs, oval cur¬ edge ved ZS21 1 28 04 94 Herbng, evening brown with light edge single, oval or long milk ZS212 28 04 94 Herbng, evening beige single or in pairs or groups, milk coccoid or oval ZS22 1 28 04 94 Menzberg, supplier brown-yellowish single or in groups, coccoid milk ZS231 28 04 94 Menzberg, supplier light brown single or in pairs or groups, milk coccoid ZS24 1 28 04 94 Bodenberg, milk brown single or in pairs, long sieve ZS251 16 05 94 Rain-Dorf, evening brown single, long, curved milk ZS261 16 05 94 Herbng, supplier matt brown single, oval milk ZS26 2 16 05 94 Herbng, supplier yellowish-brown single or in pairs, coccoid milk ZS271 07 06 94 Gerschnialp, kettle beige-brownish single or in pairs or groups, milk oval ZS281 07 06 94 Twerenegg, supp¬ yellow single or in pairs or groups, lier milk coccoid or oval ZS29 1 23 06 94 Gerschnialp light brown single or in pairs or groups, coccoid or oval ZS321 28 06 94 Alpnach-Dorf, beige single or in goups, oval evening milk -116- 7. Annex cont. ZS33 1 11 07 94 supplier milk red-brown single or in pairs or groups, oval ZS33 2 11 07 94 supplier milk brown with light edge single or in pairs or groups, oval ZS34 1 11 07 94 supplier milk brown with light edge single or in pairs, oval, club shaped ZS34 2 11 07 94 supplier milk beige-brownish single or in groups, long ZS35 1 11 07 94 supplier milk red-brown single, coccoid or oval ZS361 27 07 94 Gundolingen, brown single, oval kettle milk 1 ZS36 2 27 07 94 Gundolingen, light brown-yellowish single or in pairs, coccoid kettle milk 1 ZS37 1 27 07 94 Wegmatten brown with light edge single or in groups, oval ZS37 2 27 07 94 Wegmatten beige single, long or very long ZS38 2 06 09 94 Udligenswil, deli- red-brown single, coccoid or oval very milk ZS39 1 06 09 94 Rotmoos, kettle beige single or in pairs or groups, milk 1 oval ZS39 2 06 09 94 Rotmoos, kettle brown with light edge single, oval or long, club milk 1 shaped ZS40 1 06 09 94 Rotmoos, kettle brown single, coccoid or oval milk 2 ZS41 1 06 09 94 Udligenswil, supp- brown with light edge single, coccoid or oval Iier milk ZS42 1 06 09 94 Alpnach-Dorf, deli- brown with light edge single or in pairs or groups, very milk oval ZS42 2 06 09 94 Alpnach-Dorf, deli- white-beige single, oval or long very milk ZS43 1 20 10 94 Simisberg, supplier beige-brownish single, coccoid or oval milk ZS43 2 20 10 94 Simisberg, supplier red-brown single or in pairs or groups, milk oval or long ZS44 1 2010 94 Hmterberg, supp- red-brown single or in pairs or groups, Iier milk oval ZS44 2 20 10 94 Hmterberg, supp- beige, catalase nega- single, coccoid her milk tive ZS45 1 10 11 94 Hmterberg, supp- matt brown single, oval or long her milk ZS46 1 101194 Greppen, supplier brown with light edge single, oval milk ZS47 1 10 11 94 SchQIen, supplier brown with light edge single or in pairs, oval milk ZS48 1 10 11 94 Henzberg, supplier brown with light edge single or in chains, oval milk ZS49 1 10 1194 Holz, evening milk red-brown single or in groups, oval ZS49 2 10 11 94 Holz, evening milk light brown single or in pairs, oval ZS50 1 10 12 94 Margel, supplier brown with light edge single or in groups, oval or milk long ZS51 1 1012 94 BUrg, supplier 7 beige-brownish single or in pairs, long 7. Annex -117- Table 11: "Propionibactena" strains from the alps - colony descnption and morphology Strain Date Origin Colony Morphology 1061 1107 95 Le Van, Les Cles beige single or in pairs or groups, oval or long, partly club sha¬ ped 106 2 1107 95 Le Van, Les Cles red-brown single or in pairs or groups or chains, coccoid 106 3 1107 95 Le Van, Les Cles orange-brown in groups, oval 1101 07 08 95 Le Van, Les Cles light brown single, oval 1141 04 09 95 Le Van, Les Cles brown single, long or very long, partly curved 201 1 06 06 95 Pa M, Arpilles red, catalase negative single or in pairs, coccoid 204 1 27 06 95 Pa M, Arpilles beige, catalase nega¬ single or in pairs or groups tive or chains, coccoid 206 1 12 07 95 Pa M, Arpilles red-brown single, coccoid or oval 209 1 02 08 95 Pa M, Arpilles light brown single, oval or very long, partly curved 2101 07 08 95 Pa M, Arpilles light brown single or in chains, oval or long, partly curved 2121 22 08 95 Pa M,Arpilles light brown, irregular single or in pairs, oval 2131 30 08 95 Pa M, Arpilles red, catalase negative single or in pairs, coccoid 2132 30 08 95 Pa M, Arpilles brown single or in pairs, oval 213 3 30 08 95 Pa M, Arpilles beige in pairs or groups, long, curved 301 1 06 06 95 Montbovon beige in pairs or groups, coccoid 3012 06 06 96 Montbovon brown single, oval 3041 27 06 95 Montbovon brown single, oval 304 2 27 06 95 Montbovon beige single or in pairs or chains, coccoid 307 1 25 07 95 Montbovon beige-yellowish single, oval or long, partly curved 3081 02 08 95 Montbovon brown single, oval 311 1 15 08 95 Montbovon light brown single or in pairs, long partly curved -118- 7 Annex Table 12: "Propionibactena" strains from Emmental cheese with and without brown spots -col¬ ony description and morphology Cheese Origin Strains Colony Morphology Emmental I Uettligen, 5 months, El 1 - El 10 brown single, oval with brown spots Emmental II 1749, 5 months Ell 1 - Ell 10 brown single, oval Emmental III 1177, 8 months Elll 1-Elll 10 brown single, oval Emmental IV Uettligen, 6 months ElV 1-ElV 5, ElV 7- brown single, oval ElV 10 Table 13: "Propionibactena" strains from Appenzell cheese with brown spots - colony descnp¬ tion and morphology Cheese Origin Strain Colony Morphology Appenzell Ebersol, AM beige, catalase negative single or in pairs, oval I 6 months AI3 brown single, coccoid or oval AI4 beige, catalase negative single or in pairs, oval AI5 brown single or in pairs or groups, oval AI5 1 beige, catalase negative single or in pairs, oval AI6 beige, catalase negative single or in pairs, oval AI7 brown single or in pairs, oval AI8 beige, catalase negative single or in pairs, oval AI9 brown single or in pairs, oval Al 10 brown single or in pairs, oval Appenzell Dicken, AIM beige, catalase negative single or in pairs, oval II 6 months All 2 brown single or in pairs or in groups, oval All 3 white, catalase negative single or in pairs, oval All 4 brown single, oval All 5 brown single or in pairs, oval All 6 brown single or in pairs or in groups, oval All 7 brown single or in pairs or in groups, oval All 8 brown single or in pairs or in groups, oval All 9 brown single or in pairs or in groups, oval All 10 red-brown single or in pairs, coccoid or oval Appenzell Kohlbrunnen, AIIM brown single or in pairs, oval III 6 months All! 2 brown single or in pairs or in groups, oval AIM 3 brown single or in pairs, oval AMI -4 light brown single or in pairs or in groups, oval AIII5 brown single or in pairs or in groups, oval AIM 6 brown single or in pairs or in groups, oval 7. Annex -life. cont. AIII7 brown single or in pairs or in groups, coccoid or oval AIII8 brown single or in pairs, oval AIM 9 red single or in pairs or in groups, coccoid or oval Alll 9 1 brown single or in pairs, oval Alll 10 white, catalase negative single, coccoid or oval Appenzell Rossfallen, AlV 1 brown single or in pairs, oval IV 6 months AlV 2 brown single or in pairs, oval AlV 3 brown single or in groups, oval AlV 4 brown single or in pairs, oval AlV 5 brown single or in pairs, oval AlV 6 brown single or in pairs, oval AlV 7 brown single or in pairs, oval AlV 7 1 beige, catalase negative single, coccoid or oval AlV 8 brown single or in pairs, oval AlV 9 brown single or in pairs, oval AlV 9 1 beige, catalase negative single, oval AlV 10 brown single or in pairs, oval Table 14: "Propionibactena" strains from Raclette cheese with brown spots - colony description and morphology Cheese Ongin Strain Colony Morphology Raclette Val R1 white-beige, catalase negative single or in pairs, coccoid d'alpage d'Annwiers R2 brown single or in pairs or groups, oval R3 white-beige, catalase negative single or in pairs, coccoid R4 white-beige, catalase negative single or in pairs, coccoid R5 white-beige, catalase negative single or in pairs, coccoid R6 brown single or in pairs or groups, oval R7 white-beige, catalase negative single or in pairs, coccoid R8 white-beige, catalase negative single or in pairs, coccoid R9 white-beige, catalase negative single or in pairs, coccoid R10 brown single or in pairs or groups, oval Table 15: "Propionibactena" from Sbnnz cheese with brown spots - colony descnption and mor¬ phology Cheese Origin Strain Colony Morphology SbnnzI Udligenswil, SM light brown single, oval 30 months SI 2 light brown single, coccoid or oval SI 3 white, catalase single or in pairs, coccoid negative SI 6 light brown single, coccoid or oval SI 7 beige, catalase single or in pairs, coccoid negative -120- 7. Annex cont. Sbnnz II Alpnach-Dorf, SIM light brown single or in pairs or groups, coccoid 6 months SM 2 light brown single or in pairs or groups, coccoid SM 3 light brown single or in pairs or groups, coccoid SII4 light brown single or in pairs or groups, coccoid SM 5 light brown single or in pairs or groups, coccoid SM 6 light brown single or in pairs or groups, coccoid SM 7 light brown single or in pairs or groups, oval SM 8 light brown single or in pairs or groups, coccoid or oval SM 9 light brown single or in pairs or groups, coccoid SIMO light brown single or in pairs or groups, coccoid Sbrinz I Alpnach-Dorf, SUM light brown single or in pairs, coccoid or oval 6 months SMI 2 light brown single or in pairs, oval SIM 3 light brown single or in pairs, oval SIM 4 light brown single or in pairs, coccoid or oval SMI 5 light brown single or in pairs, oval SMI 6 light brown single or in pairs, oval Sill 7 light brown single or in pairs, coccoid SHI 8 light brown single or in pairs, oval Sill 9 light brown single or in pairs, oval Sill 10 light brown single or in pairs, oval Sbrinz IV Twerenegg, SIV 2 light brown single or in pairs or groups, coccoid 19 months or oval SIV 3 light brown single or in pairs, oval SIV 4 light brown single or in pairs or groups, oval SIV 5 light brown single or in pairs, coccoid or oval SIV 6 light brown single or in pairs, coccoid or oval SIV 7 light brown, single or in pairs, oval catalase negative SIV 8 light brown single, oval or long SIV 9 light brown single or in pairs, oval SIV 10 light brown single, oval Sbnnz V Twerenegg, SV1 light brown single or in groups, oval 6 months SV 2 light brown single or in pairs or groups, coccoid or oval SV3 light brown single, coccoid or oval SV4 light brown single, oval SV5 light brown single, oval SV6 light brown single, coccoid or oval SV7 light brown single, coccoid or oval SV8 light brown single, coccoid or oval SV9 light brown single or in groups, oval SV10 light brown single, oval Sbnnz VI Schintmoos, SVI 2 light brown single or in pairs, coccoid or oval 15 months SVI 3 light brown single or in pairs, coccoid or oval SVI 4 light brown single, coccoid or oval SVI 5 white-beige single or in pairs, coccoid or oval SVI 6 beige, catalase single or in pairs, coccoid negative SVI 7 beige single, long, partly club shaped SVI 8 light brown single, oval 7. Annex -121- cont. SVI 9 light brown single or in pairs, oval SV110 light brown single or in pairs, oval or long Sbnnz VII Hildisneden, SVII2 light brown single or in pairs or groups, coccoid 12 months or oval SVII3 light brown single, coccoid or oval SVII4 light brown single, oval SVII5 light brown angle or oval, coccoid SVII6 light brown single or n pairs, coccoid or oval SVII7 light brown single, oval SVII8 light brown single or in pairs, coccoid or oval SVI19 beige single or in groups, coccoid or oval SVI110 light brown single, coccoid or oval Sbrinz VIII Schoned, SVII11 light brown single or in groups, oval 20 months SVIII 2 light brown single or in pairs, oval SVIII 3 light brown single or in pairs, oval SVIII 4 light brown single or in pairs or groups, oval SVIII 5 light brown single or in pairs, oval SVIII 6 light brown single or in pairs, oval SVIII 7 light brown single or in groups, oval SVIII 8 light brown single or in pairs, oval SVIII 9 light brown single or in pairs, oval SVII110 light brown single or in pairs, oval Table 16: "Propionibactena" strains from Gruyere cheese with splitting defect - colony descnp¬ tion and morphology Cheese Origin Strain Colony Morphology Gruyere I 1005/9504 Gl 1 -Gl 5, Gl 7-G110 brown single or in pairs, oval or long GruyereII 2660/9418 Gil 1-Gil 10 brown single, oval Gruyere III 2660/9505 Gill 2-Gill 10 brown single, oval, partly club shaped Gruyere IV 894/9420 GIV1.GIV6 brown with light single, oval edge GIV2-GIV5, GIV7, GIV10 brown single, oval Gruyere V 2544/9501 GV 1 - GV 7, GV 9, GV 10 brown with ligth single, oval edge Table 17: "Propionibactena" strains from Sbnnz cheese with splitting defect - colony descnption and morphology Cheese Origin Strain Colony Morphology Sbnnz I Udligenswil, GS11 - GSM 0 brown with light edge single or in pairs, 14 months oval Sbnnz II Udligenswil, GSI11 - GSI110 brown single or in pairs, 17 months oval Sbnnz III Udligenswil, GSII11 - GSIII 9 brown single or in pairs, 16 months oval GSII110 brown with light edge single, coccoid Sbnnz IV Schoned, 13 GSIV1 - GSIV 10 brown single or in pairs, months oval zi22z 7. Annex cont. Sbrinz V Alpnach-Dorf, GSV1.GSV2, brown with depression single or in pairs, 18 months GSV 4, GSV 7 - oval GSV10 GSV 3, GSV 6 brown single, oval Table 18: Bacterial strains (except propionibactena) used in this work Number strain and biochemical identification source 2492 Lactobacillus delbrQcku ssp bulgancus T FAMb 877 Lactobacillus delbrucku ssp lactis 1 FAMb 144 Lactobacillus delbrucku ssp lactis FAMb 145 Lactobacillus delbrucku ssp lactis FAMb 169 Lactobacillus delbrQcku ssp lactis FAMb 170 Lactobacillus delbrucku ssp lactis FAMb 278 Lactobacillus delbrucku ssp lactis FAMb 286 Lactobacillus delbrucku ssp lactis FAMb 933 Lactobacillus delbrOcku ssp lactis FAMb 1050 Lactobacillus delbrQcku ssp lactis FAMb 1051 Lactobacillus delbrQcku ssp lactis FAMb 1052 Lactobacillus delbrQcku ssp lactis FAMb 1086 Lactobacillus rhamnosus T FAMb 1211 Lactobacillus rhamnosus FAMb 1212 Lactobacillus mamnosus FAMb 1213 Lactobacillus mamnosus FAMb 1214 Lactobacillus mamnosus FAMb 1215 Lactobacillus rhamnosus FAMb 1216 Lactobacillus rhamnosus FAMb 1217 Lactobacillus rhamnosus FAMb 1218 Lactobacillus rhamnosus FAMb 1219 Lactobacillus rhamnosus FAMb 1220/23 Lactobacillus mamnosus FAMb 1116 Lactobacillus casei T FAMb 1227 Lactobacillus casei FAMb 1228 Lactobacillus casei FAMb 1229 Lactobacillus casei FAMb 1087 Lactobacillus plantarum T FAMb 2051 Lactobacillus plantanim FAMb 2053 Lactobacillus plantanim FAMb 2057 Lactobacillus plantarum FAMb 2062 Lactobacillus plantarum FAMb 2063 Lactobacillus plantarum FAMb 2064 Lactobacillus plantarum FAMb 2065 Lactobacillus plantarum FAMb 2066 Lactobacillus plantarum FAMb 2069 Lactobacillus plantarum FAMb 2071 Lactobacillus plantarum FAMb 2075 Lactobacillus plantarum FAMb 2076 Lactobacillus plantarum FAMb 1084 Lactobacillus zeae T FAMb 7. Annex zi2&. cont 932 Streptococcus thermophilus T FAMb 879 Enferococcus faeca/is T FAMb 2645 Enterococcus faecium T FAMb 3271 Enferococcus sp FAMb 3272 Enterococcus sp FAMb 3273 Enterococcus sp FAMb 3274 Enterococcus sp FAMb 3275 Enterococcus sp FAMb 3276 Enterococcus sp FAMb 3277 Enterococcus sp FAMb 2493 Lactococcus lactis J FAMb 3009 Lactococcus lactis FAMb 3010 Lactococcus lactis FAMb 3011 Lactococcus lactis FAMb 3013 Lactococcus lactis FAMb 3018 Lactococcus lactis FAMb 3233 Lactococcus lactis FAMb 552 Clostndium butyncum T DSM 1322 Clostridium butyncum FAM 795 Clostridium sporogenes T DSM 1755 Clostndium sporogenes T FAM 2637 Clostndium tyrobutyncum T DSM 608 Clostndium tyrobutyncum CNRZ 1 according to Cummins and Johnson (1986) -124- 7. Annex Table 19: Classification of propionibactena reference strains according to different charactens- tica strain species according to subspecies plasmids, size growth on MF95C1 protein profile 23S rRNA ATCC 6207 P freud P freud freud negative slow/high ATCC 9614 P freud P freud sherm negative fast/high DSM 20270 P freud P freud sherm negative fast/high P1409 P freud P freud shenn negative slow/high P1410 P freud P freud sherm negative fast/high P1411 P freud P freud shenn negative fast/high P1412 P freud P freud shenn negative fast/high P1413 P freud P freud sherm >16 2kb fast/high P1414 P freud P freud shenv negative slow/high P111 P freud P freud sherm negative fast/high P112 P freud P freud sherm negative fast/high ATCC 4868 P jensenii P jensenii negative slow/weak DSM 20274 P jensenii P jensenii _. negative slow/medium DSM 20278 P jensenii P jensenii _ negative slow/weak DSM 20279 P jensenii P jensenii — negative slow/medium ATCC 4874 P thoenii P thoenii — negative slow/weak DSM 20275 P thoenii P thoenii _. negative slow/medium DSM 20277 P thoenii P thoenii -- negative slow/medium ATCC P acidipropi¬ P acidipropionici — negative hardly 25562 onici DSM 20272 P acidipropi¬ P acidipropionici — negative slow/weak onici DSM 20273 P acidipropi¬ P acidipropionici — negative fast/weak onici time needed to reach maximum turbidity at 650nm zero <0 030 at days 1-13 hardly 0 030-0 059 at days 1-13 slow/weak 0 060-0 269 at days 11-13 slow/medium 0 270-0 499 at days 11-13 slow/high >0 500 at days 11-13 fast/weak 0 060-0 269 at days 3-8 fast/high >0 500 at days 3-8 Table 20: Classification of propionibactena strains from MIBD Aargau (AG) according to differ¬ ent characteristics strain species according to subspecies plasmids, size growth on MF95C1 protein profile 23S rRNA AG2 1 noP noP — >16 2kb slow/high AG3 1 noP noP — negative slow/high AG3 2 noP noP — negative zero AG61 noP noP — negative hardly AG6 2 noP noP — negative hardly AG7 1 noP noP — >16 2kb, 2 5kb zero 7, Annex zi25z. cc-nt, AG81 P acidipropionici P acidipropionici — negative slow/medium AG8 2 P freud _ shenn negative slow/high AG91 P freud — fraud negative fast/high AG11 1 noP noP _ negative hardly AG112 noP noP — 4 9kb, 3 Okb hardly AG131 noP noP _ negative hardly AG13 3 noP noP — negative hardly AG14 1 noP noP — 4 7kb,3 7kb zero AG15 2 noP noP -- 3 1kb, 2 4kb hardly AG15 3 noP noP — >16 2kb, 8 2kb, hardly 4 7kb AG17 1 noP noP — negative hardly AG18 1 noP noP — negative hardly AG18 2 noP noP — negative hardly AG21 1 P acidipropionici P acidipropionici — 6 2kb slow/medium AG22 1 P freud — freud >16 2kb fast/high AG22 2 P acidipropionici P acidipropionici _ >16 2kb fast/weak AG24 1 P jensenii P jensenii — negative slow/weak AG251 P acidipropionici P acidipropionici — >16 2kb slow/medium AG25 2 P freud — sherm negative fast/high AG25 3 P freud — sherm >16 2kb fast/high AG25 4 noP noP — negative hardly AG26 1 P freud — sherm negative fast/high AG271 P freud — sherm negative fast/high AG281 P thoenii P thoenii — negative fast/weak AG28 2 P acidipropionici P acidipropionici — 6 6kb slow/medium AG29 1 noP noP — 8 5kb, 6 2kb, hardly 4 8kb AG30 1 noP noP — negative fast/weak AG31 1 P freud — freud >16 2kb slow/high AG32 1 P jensenii P jensenii _ negative slow/weak AG33 1 P freud — sherm >16 2kb fast/high AG34 1 P acidipropionici P acidipropionici — negative slow/medium AG36 1 P jensenii P jensenii — 7 2kb slow/weak AG37 1 P freud — sherm negative fast/high AG381 P jensenii P jensenii — negative slow/medium AG39 1 P jensenii P jensenii — 8 Okb fast/weak AG40 1 P freud — freud >16 2kb fast/high AG42 1 P jensenii P jensenii — negative slow/weak AG42 2 P freud — freud >16 2kb slow/high AG451 noP noP — negative slow/medium AG461 P acidipropionici P acidipropionici — negative slow/medium AG46 2 P freud — sherm negative hardly AG47 1 P jensenii P jensenii — negative slow/weak AG481 P freud — shemj negative slow/high AG48 2 P jensenii P jensenii — >16 2kb slow/medium AG491 P jensenii P jensenii — negative slow/weak AG501 P acidipropionici P acidipropionici — >16 2kb slow/medium AG51 1 P freud — sherm negative slow/medium AG521 P freud _ sherm negative fast/high AG53 1 P freud — freud 7 Okb fast/high AG541 noP noP — negative slow/medium AG56 1 P acidipropionici P acidipropionici — negative slow/medium AG58 1 P freud — shenn >16 2kb fast/high AG591 noP noP — negative hardly -126- 7. Annex cont AG601 P freud — shenn negative fast/high AG621 noP noP — >16 2kb, hardly »16 3kb, 7 8kb AG63 1 P acidipropionici P acidipropionici -- negative slow/medium AG64 1 P freud — sherm negative slow/high AG65 1 noP noP ... negative zero AG661 P acidipropionici P acidipropionici ... negative fast/weak AG67 1 P jensenii P jensenii ... negative slow/medium AG681 P freud — freud negative slow/high AG691 P freud — freud negative slow/high AG71 1 P jensenii P jensenii ._ negative slow/medium AG72 1 P freud — freud >16 2kb fast/high AG73 1 P freud — sherm negative fast/high AG75 1 P freud ... freud negative slow/high 1 see Table 19 Table 21: Classification of propionibacteria strains from MIBD Bern (BE) according to different characteristics strain species according to subspecies plasmids, size growth on MF95C1 protein profile 23S rRNA BE1 1 P freud _. shenn negative fast/high BE21 P freud — sherm negative slow/high BE3 1 P freud — freud negative fast/high BE4 1 P freud — sherm negative fast/high BE61 P acidipropionici P acidipropionici — negative fast/weak BE71 P acidipropionici P acidipropionici — negative fast/weak BE81 noP noP — negative hardly BE8 3 P freud — freud >16 2kb slow/high BE91 P freud — shenn negative fast/high BE9 2 noP noP _. negative fast/weak BE10 1 P freud _. sherm negative fast/weak BE111 P acidipropionici P acidipropionici — negative slow/weak BE121 P freud — sherm negative fast/high BE12 2 P freud _. freud negative fast/high BE131 P freud — freud negative slow/high BE161 P freud ._ shenv negative fast/high BE171 P freud — freud negative fast/high BE17 2 P freud — freud negative fast/high BE181 P freud — freud negative fast/high BE191 P freud — freud negative slow/high BE201 P jensenii P jensenii — negative slow/medium BE21 1 P freud _ freud negative fast/high BE221 P freud ... freud negative fast/high BE22 2 P freud — sherm negative fast/high BE231 P freud — freud negative slow/high BE25 1 P freud — freud >16 2kb fast/high BE261 P freud — sherm negative slow/weak BE27 1 P jensenii P jensenii — >16 2kb, 6 6kb slow/weak 7. Annex -127- cont BE281 P jensenii P jensenii — negative slow/weak BE291 P freud — freud negative fast/high BE30 1 P freud ~- shemj negative fast/high 1 see Table 19 Table 22: Classification of propionibactena strains from MIBD Fnbourg (FR) according to differ¬ ent characteristics strain species according to subspecies plasmids, size growth on MF95C1 protein profile 23S rRNA FR1 1 noP noP _. >16 2kb fast/weak FR21 P freud — freud «17 5kb, fast/high 11 2kb, 7 8kb FR31 noP P acidipropiomci _ negative fast/weak FR4 1 P freud — sherm negative fast/high FR51 P freud ._ sherni 14kb fast/high FR5 2 P jensenii P jensenii _ 6 9kb fast/weak FR71 P freud — freud negative fast/high FR7 2 P freud — freud 4 1kb, 1 7kb fast/high FR8 1 P freud — shenn >16 2kb fast/high FR91 noP noP — negative hardly FR11 1 noP noP — negative fast/weak FR112 P jensenii P jensenii — negative fast/weak FR121 noP noP — 34kb, 2 Okb hardly FR12 2 P freud sherm >16 2kb,4 4kb, fast/high 3 9kb, 2 6kb, 16kb FE131 P freud — sherni negative fast/high FR151 P freud — sherni negative hardly FR161 noP noP _ negative zero FR17 1 noP noP — negative zero FR17 2 P acidipropionici P acidipropiomci — >16 2kb fast/weak FR17 4 P freud — shemj negative hardly FR17 5 P acidipropionici P acidipropiomci — >16 2kb zero FR18 1 noP noP _. negative hardly FR18 2 P freud _ freud negative hardly FR18 3 P freud _ sherm >16 2kb hardly FR19 2 P freud — freud 3 3kb, 1 8kb fast/high FR201 P freud — freud negative zero FR20 2 P freud — fraud negative zero FR20 3 P jensenii P jensenii — >16 2kb hardly FR21 1 P freud — sherni negative fast/high FR212 P freud — shenn negative fast/high FR213 P jensenii P jensenii — >16 2kb hardly FR221 P freud — freud >16 2kb fast/high FR22 2 P freud — freud negative fast/high FR22 3 P jensenii P jensenii — 6 8kb fast/weak FR231 P freud — freud negative hardly FR23 2 P acidipropiomci P acidipropiomci — negative slow/medium FR23 3 P jensenii P jensenii — negative hardly zl23z. 7. Annex cont FR24 1 P freud sherm negative zero FR24a 2 P acidipropiomci P acidipropiomci — negative slow/medium FR251 noP noP — negative slow/weak FR26 1 P freud __ sherm negative hardly FR27 1 P freud — sherm >16 2kb fast/high FR28 1 P jensenii P jensenii — negative slow/medium FR28 2 P freud — freud >16 2kb fast/high FR291 noP noP — negative hardly FR29 2 P freud — freud negative fast/high FR301 P freud — freud negative fast/high FR31 1 P freud _ sherm negative fast/weak FR312 P jensenii P jensenii ... negative slow/medium FR32 1 noP noP ... negative zero FR33 1 P freud ... sherm negative fast/high FR341 P jensenii P jensenii — >16 2kb slow/medium FR35 1 P jensenii P jensenii — negative fast/weak FR35 2 P freud ... sherm negative fast/high FR36 1 P freud — sherm negative fast/high FR37 1 P freud ~- freud negative fast/high FR38 1 P freud -- freud >16 2kb fast/high FR38 2 P thoenii P thoenii — negative fast/weak FR38 3 P freud _ freud >16 2kb fast/high FR391 P freud — freud >16 2kb fast/high FR40 1 P jensenii P jensenii — 14 Okb, 2 8kb fast/weak FR40 2 P jensenii P jensenii -- 6 8kb slow/high FR41 1 P freud — freud negative slow/high FR42 1 P jensenii P jensenii — 8 7kb, 6 7kb slow/medium FR431 P freud — freud negative fast/high FR43 2 P freud — freud negative fast/high FR44 1 P freud — freud negative fast/high FR45 1 P freud — freud negative fast/weak FR46 1 P jensenii P jensenii — negative fast/weak FR47 1 P freud P freud freud negative fast/high FR48 1 P freud P freud freud >16 2kb fast/high FR49 1 P acidipropiomci P acidipropiomci — negative slow/high FR501 P thoenii P thoenii — negative fast/weak FR511 P thoenii P jensenii — 7 Okb slow/medium FR52 1 P freud — freud negative fast/high FR53 1 P jensenii P jensenii — >16 2kb slow/high FR53 2 P acidipropiomci P acidipropiomci — negative stow/medium FR53 3 P freud — freud >16 2kb fast/weak FR55 1 P acidipropiomci P acidipropiomci — negative slow/weak FR55 2 P acidipropiomci P acidipropiomci — 6 6kb slow/weak FR561 P freud — sherm >16 2kb slow/high FR57 1 P freud — freud negative zero FR581 P freud — freud negative fast/high FR591 P freud — sherm negative slow/weak FR601 P jensenii P jensenii — negative fast/weak FR60 2 noP noP — negative fast/weak FR61 1 P freud — freud negative fast/high FR621 noP noP — negative zero FR631 P freud — sherm >16 2kb fast/high FR64 1 P freud — freud >16 2kb fast/high FR64 2 noP noP — negative slow/weak FR65 1 P freud — freud >16 2kb fast/high 7 Annex -129- cont, FR401 1 P acidipropiomci P acidipropiomci slow/weak FR403 1 P acidipropiomci P acidipropiomci fast/weak FR4051 P freud freud slow/weak FR405 2 P freud freud slow/weak FR407 1 noP P acidipropiomci fast/weak FR4131 P jensenii P jensenii slow/medium FR4141 P acidipropionici P acidipropiomci hardly 1 see Table 19 Table 23: Classification of propionibactena strains from MIBD Nordostschweiz (NO) according to different charactenstica strain species according to subspecies plasmids, size growth on MF95C1 protein profile 23S rRNA N01 1 noP noP — 4 1kb 3 , 9kb, slow/high 3 5kb,1 15 kb N021 P fraud — sherni >16 2kb fast/high N02 2 P freud — shem? >16 2kb fast/high N03 1 noP noP — >16 2kb, hardly 14 8kb N03 2 P freud — sherni >16 2kb fast/high N03 3 noP noP — negative hardly N051 P freud _ sherni negative fast/high N061 P freud — sherni negative fast/high N06 2 P freud _. sherni negative fast/high N071 noP noP — negative hardly N081 P freud — freud 1 1kb fast/high N091 P freud — sherm negative fast/high N09 2 noP noP — 9 1kb, 8 6kb, hardly 52kb, 4 9kb N09 3 noP noP — negative hardly NO101 noP noP — negative slow/weak NO10 2 P freud — shemj negative fast/high N011 1 P freud — sherm >16 2kb fast/high N0112 P freud — sherm >16 2kb fast/high N0113 P thoenii P fhoenw — negative zero N0121 P freud — freud negative slow/high N013 1 P freud — freud negative slow/high N0151 P freud — freud negative fast/weak N0161 P freud _ freud >16 2Kb fast/high N0181 P freud P freud — negative zero N0191 P freud — freud >16 2kb zero N021 1 P freud — freud negative slow/high N022 1 P freud — sherm negative fast/high N0231 P freud — freud negative fast/high N0241 P freud — sherm negative fast/high N0251 P freud — freud negative fast/high N026 1 P jensenii P jensenii — 6 6kb slow/weak N0271 P freud — freud negative fast/high N028 1 P freud — sherm negative fast/high N0291 P freud — freud negative slow/medium -130- 7. Annex cont. NO30 1 no P noP 9 3kb,5 9kb hardly NO30 2 P freud freud >16 2kb fast/weak 1 see Table 19 Table 24: Classification of propionibactena strains from MIBD Nordwestschweiz (NW) accord¬ ing to different charactenstica strain species according subspecies plasmids, size growth on MF95C1 protein profile 23S rRNA NW1 1 P freud — sherm negative fast/high NW12 noP noP -- negative hardly NW2 1 P freud — freud negative fast/weak NW3 1 P freud — freud negative fast/high NW4 1 P freud — sherm negative zero NW4 2 noP noP — 9 2kb, 5 8kb hardly NW51 P freud _. sherm negative slow/high NW61 P freud ... freud negative hardly NW7 1 P freud — freud negative zero NW81 P freud ... shenn negative fast/high NW9 1 P jensenii P jensenn -- negative hardly NW9 2 P jensenii P jensenii -- negative hardly NW101 P jensenii P jensenii — >16 2kb slow/weak NW11 1 P freud — sherm negative fast/high NW12 1 P freud -- freud negative fast/high NW131 P freud — freud negative hardly NW13 2 P freud -- freud negative fast/weak NW13 3 P freud — freud negative fast/high NW14 1 P freud — freud negative fast/high NW16 1 P freud — freud negative fast/high NW17 1 P jensenii P jensenn — negative fast/weak NW172 P freud — sherm >16 2kb hardly NW18 1 noP noP ... 9 4kb, 5 3kb zero NW18 2 P freud — freud negative slow/weak NW19 1 noP noP — negative hardly NW201 noP noP ._ negative zero NW20 2 P freud — freud negative slow/high NW20 3 P freud — shenn negative fast/high NW22 1 P freud — freud negative fast/high NW231 P jensenii P jensenn _. >16 2kb fast/weak NW23 2 P freud __ freud negative fast/high NW24 1 P freud _. shenn >16 2kb fast/high NW25 1 P freud — sherm >16 2kb fast/high NW261 P jensenii P jensenn — negative slow/medium NW27 1 P thoenii P thoenii — negative slow/weak NW27 2 P freud ... sherni negative fast/high NW281 P freud — freud negative fast/high NW30 1 P fnsud — freud negative zero NW31 1 P freud — freud negative fast/high NW32 2 P jensenii P jensenn — >16 2kb, 7 Okb slow/medium NW33 1 P freud — shenn >16 2kb, 8 7kb hardly NW33 2 P freud — sherm negative fast/high 7. Annex -131- cont. NW33 3 P freud _. shenn negative fast/high NW34 2 P freud ._ freud negative fast/high NW36 1 P thoenii P fhoenw — negative slow/weak NW37 1 P jensenn P jensenn _ negative slow/medium NW38 1 P freod — shenn >16 2kb fast/high NW39 1 no P noP — negative zero 1 see Table 19 Table 25: Classification of propionibactena strains from MIBD St Gallen-Appenzell AR (SG.AR) according to different charactenstica strain species according to subspecies plasmids, size growth on MF95C1 protein profile 23S rRNA SG1 1 noP noP — «20 5kb, zero >16 2Kb SG12 noP noP — >16 2kb, 9 5kb, hardly 5 4kb, 3 3kb SG2 1 P jensenn P jensenn — 13 0kb,8 8kb zero SG2 2 P jensenii P jensenn — 13 Okb, 8 8kb zero SG3 2 noP noP — negative hardly SG4 1 P freud — freud negative zero SG5 1 noP noP ._ >16 2kb hardly SG61 P freud — freod negative hardly SG71 P freud — freud negative fast/weak SG91 P freud — sherm negative fast/weak SG101 P jensenn P jensenn — negative hardly SG10 2 P freud — freud >16 2kb slow/high SG11 1 P freud — sherm negative slow/high SG121 P freud — sherm negative slow/high SG131 noP noP __ negative zero SG14 1 P jensenn P jensenn — >16 2Kb, 6 8kb slow/weak SG161 P jensenn P jensenn — negative slow/high SG17 1 P freud — sherni negative fast/high SG181 P freud — freud >16 2kb slow/weak SG191 P jensenn P jensenn — negative slow/medium SG201 noP noP — >16 2Kb slow/high SG221 P freud — shemj negative fast/high SG231 noP noP — negative zero SG24 1 P freud — freud negative fast/high SG251 noP noP — 12 Okb, 8 5kb, zero 5 3kb SG271 P freud shemj 3 8kb, 3 2kb, fast/high 2 8kb, 2 4kb, 18kb SG281 P freud — sherni negative hardly SG291 P freud _ sherni negative fast/high SG30 1 P freud — freud negative fast/high SG31 1 P fraud — shenn negative fast/high SG331 P freud — freud negative fast/high -132- 7 Annex cont AR1 1 P freud P freud sherm negative slow/high AR1 2 P freud P freud sherm negative fast/high 1 see Table 19 Table 26: Classification of propionibactena strains MIBD Thurgau (TG) according to different charactenstica Strain species according to subspecies plasmids, size growth on MF95C1 protein profile 23S rRNA TG1 1 P freud — freud negative slow/high TG12 P freud _ freud >16 2kb fast/high TG2 1 P freud — sherm negative fast/high TG3 1 noP noP — >16 2kb 9 5kb, hardly 5 4Kb TG3 2 P freud — freud >16 2kb slow/high TG4 1 noP noP ... negative hardly TG4 2 P freud — shenn negative fast/weak TG4 3 noP noP ... negative fast/weak TG51 P freud ... shenn >16 2kb fast/high TG6 1 P freud ._ shenn negative hardly TG8 1 P freud — sherm >16 2kb fast/high TG9 1 P freud _. sherm >16 2kb hardly TG10 1 Pjensenn Pjensenn — negative slow/weak TG11 1 noP noP — negative zero TG12 1 P freud — freud negative fast/high TG13 1 P freud _ sherm negative hardly TG14 1 P freud _. freud negative fast/high TG15 1 P freud — freud negative zero TG161 P acidipropiomci P acidipropiomci — negative slow/weak TG171 P acidipropiomci P acidipropiomci — negative fast/weak TG17 2 P acidipropiomci P acidipropionici _. negative fast/weak TG18 1 P freud — shenn negative fast/high TG19 1 P freud — freud 12kb fast/high TG20 1 P freud ._ freud negative fast/high TG21 1 P freud — freud >16 2kb fast/high TG22 1 P freud — shenn negative fast/high — TG231 P freud . freud negative fast/high TG23 2 noP noP — 8 1kb,4 6kb zero TG24 1 P freud — sherm negative fast/high TG251 P freud _. freud >16 2kb zero TG261 P freud — freud negative fast/high TG27 1 P freud -- freud >16 2kb fast/high TG301 P freud — freud >16 2kb fast/high TG31 1 P freud — sherm negative fast/high TG321 Pjensenn Pjensenn _ >16 2kb, 7 Okb slow/weak TG34 1 noP noP — negative slow/weak TG35 1 P freud — freud negative slow/high TG36 1 noP noP — >16 2kb zero TG37 1 P freud — freud >16 2kb slow/high TG38 1 P freud — freud >16 2kb fast/high 7. Annex zi22z. cont. TG39 2 P freud fraud negative slow/high TG401 P freud freud negative fast/high TG41 1 P freud shenn negative fast/high TG42 1 P freud shenn >16 2Kb fast/high TG43 1 P freud shenn >16 2kb, *20kb fast/high 1 see Table 19 Table 27: Classification of propionibactena strains from MIBD Vaud-Geneve (VD) and Neucha- tel (NE) according to different charactenstica strain species according to subspecies plasmids, size growth on MF95C1 protein profile 23S rRNA VD1 1 P freud — shemj >16 2kb fast/high VD2 1 P freud — sherm negative zero VD31 P freud — sherni negative slow/high VD3 2 P jensenn P jensenn — negative fast/weak VD4 1 P jensenn P jensenn — negative fast/weak VD5 1 P freud — sherm negative fast/high VD5 2 P jensenn P jensenn — negative fast/weak VD71 P freud — sherm >16 2kb fast/high VD8 1 P freud — shenn >16 2kb fast/high VD9 1 P freud — sherni >16 2kb fast/high VD101 P freud — sherm negative slow/high VD10 2 P freud _ freud negative hardly VD11 1 P jensenn P jensenn — negative slow/medium VD112 P fraud — freud negative fast/high VD12 1 P freud — freud negative fast/high NE1 1 P freud — shenn negative hardly NE2 1 P freud — sherm >16 2kb fast/high 1 see Table 19 Table 28: Classification of propionibactena strains from MIBD Zentralschweiz (ZS) according to different charactenstica strains species according to subspecies plasmids, size growth on MF95C' protein profile 23S rRNA ZS1 1 P freud — sherni negative fast/high ZS21 P freud — shenn negative hardly ZS2 2 noP nop — negative zero ZS2 3 noP nop — negative zero ZS2 4 P jensenn P jensenn _ negative fast/weak ZS2 5 P jensenn P jensenn — negative zero ZS31 P freud — shemj negative fast/high ZS4 1 P freud — freud >16 2kb slow/high ZS51 P freud — sherm negative fast/high ZS5 2 noP nop — negative zero ZS5 3 P jensenn P jensenii — negative hardly ZS61 P jensenn P jensenn — negative zero -134- 7. Annex cont ZS7 1 P freud sherm negative fast/high ZS81 P freud ... sherm negative fast/high ZS9 1 P freud — sherm negative fast/high ZS101 P freud — sherm >16 2kb slow/high ZS11 1 P jensenn P jensenn — negative zero ZS13 1 noP noP ... negative hardly ZS13 2 P freud _ freud negative fast/high ZS14 1 P thoenii P jensenn _. negative hardly ZS15 1 P jensenn P jensenn _. negative hardly ZS15 2 P freud _. freud negative slow/high ZS15 3 P freud — sherm negative fast/high ZS161 P thoenii P jensenn ... 6 9kb hardly ZS16 2 P freud — sherm >16 2kb fast/high ZS16 3 P jensenn P jensenn .- negative hardly ZS16 4 P freud — sherm >16 2kb hardly ZS17 1 P acidipropiomci P acidipropiomci _ >16 2kb fast/weak ZS181 P freud ... freud negative fast/high ZS191 P freud ... freud negative fast/high ZS19 2 P freud ... sherm negative slow/high ZS20 1 P freud ... sherm negative slow/high ZS21 1 P freud — freud negative zero ZS212 P freud — sherm negative fast/high ZS22 1 P thoenn P thoenii _ negative fast/weak ZS23 1 P jensenn P jensenn ... 6 7kb hardly ZS24 1 P freud ... sherm negative fast/high ZS25 1 P freud P freud freud >16 2kb fast/weak ZS26 1 P freud P freud sherm negative fast/high — slow/weak ZS26 2 P jensenn P jensenn negative ZS271 P jensenn P jensenn ... negative slow/high ZS28 1 P jensenn P jensenn — negative hardly ZS29 1 P freud — sherm >16 2kb fast/high ZS32 1 P freud _. shenn negative fast/high — fast/weak ZS33 1 P jensenn P jensenn negative ZS33 2 P freud — sherm negative slow/weak ZS34 1 P freud ... sherm negative hardly ZS34 2 P freud — freud >16 2kb slow/high ZS35 1 P thoenii P thoenn — negative fast/weak ZS36 1 P freud _. freud >16 2kb hardly ZS36 2 noP noP _ >16 2kb, 5 3kb hardly ZS37 1 P freud — sherm negative slow/high ZS37 2 P freud ._ freud negative fast/high -- 7 slow/medium ZS38 2 P jensenn P jensenn >16 2kb, Okb — slow/weak ZS391 P jensenn P jensenn negative ZS39 2 P freud _ freud negative zero ZS401 P freud — freud negative zero ZS41 1 P freud — freud negative zero ZS421 P freud — sherm negative fast/high — 13 9 4kb slow/medium ZS42 2 P jensenn P jensenn 4kb, ZS43 1 P freud _ sherm negative fast/high _ slow/medium ZS43 2 P jensenn P jensenn negative — slow/medium ZS44 1 P jensenn P jensenn negative ZS44 2 P acidipropiomci P acidipropiomci — negative slow/weak ZS45 1 P freud — sherm negative fast/high ZS461 P freud — freud negative fast/high ZS47 1 noP P freud — >16 2kb slow/high 7, Annex -J35z. cont. ZS48 1 P freud freud >16 2kb,64kb, fast/high 3 6kb, 2 4Kb, 1 5kb ZS49 1 P freud freud negative slow/weak ZS49 2 P freud sherm 16kb fast/high ZS50 1 P freud sherni 3 6kb, 2 3kb, fast/high 1 5kb ZS51 1 P freud freud negative fast/high 1 see Table 19 Table 29: Classification of propionibactena from the alps according to different charac- tenstica strain species according to subspecies growth onMF95C protein profile 23S rRNA 1061 P jensenn P jensenn — slow/medium 1062 P thoenii P thoenn — slow/weak 106 3 P jensenn P jensenn — slow/medium 1101 P freud P freud sherni fast/high 114 1 P freud — freud hardly 201 1 P thoenn P fhoenii — slow/weak 204 1 P thoenii P thoenii — fast/weak 206 1 P thoenn P thoenii — slow/weak 209 1 P freud — shenn fast/high 2101 P freud — freud slow/weak 212 1 P freud — sherm fast/high 213 1 P thoenii P thoenn — fast/weak 213 2 P freud — sherm fast/high 213 3 P jensenn P jensenii — fast/weak 301 1 noP noP — hardly 3012 P freud — sherm slow/medium 3041 P freud — sherni slow/medium 304 2 P thoenii P thoenii — slow/weak 307 1 P freud — sherm slow/weak 3081 P freud — freud fast/high 311 1 P fraud — freud slow/medium ' see Table 19 c < r£ £r ££ re rr r £ r r £^,r r r r c j: rrr^rrrr^jirnrrg)g>o>o>g)g>g)D>g)g)g)a}0)D>0)0)S'o>S,o>g)0)0>0)0)g>o>g>0)SI0)0>0)0)a>g> SiJaJaiiisiJialalaSaia'SIaSaJaJsSsgSalEiSaJsSiSaSaS.SsSS&Ss&SSiS iSc m tS aj q> © S ro cococo(o Q."H „ O o 8 c <° o « 1% Ol)«jl)»jll«pP)p«lll«0J«)ljlll)«»»««l)l)«ll)0)0l«»««»III))« o OJO £ a. o. 0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0,0,0.0.0.0.0.0.0.0.0.0.0.0.0.0.0. iiJu]i5EEi5iDEi5EDwiDujijjn]EDiIJuJuJwwuJUjiDiDLuiD LAnnex JSi. Table 31: Classification of propionibactena strains from Appenzell cheese with brown spots according to different charactenstica strain species according to subspecies plasmids, size growth on protein profile 23S rRNA MF95C1 AM P acidipropionici P acidipropiomci ... negative slow/medium AI3 noP P freud sherm negative fast/high AI4 P acidipropiomci P acidipropiomci — negative slow/medium AI5 P freud — sherm negative fast/high AI51 P acidipropiomci P acidipropiomci ... negative slow/medium AI6 P acidipropiomci P acidipropiomci — negative slow/medium AI7 P freud — sherni negative fast/high AI8 P acidipropiomci P acidipropiomci — negative slow/medium AI9 P freud — sherni negative fast/high Al 10 P freud — fraud negative fast/high AIM P acidipropiomci P acidipropiomci — negative slow/medium All 2 P freud — fraud negative fast/high All 3 P freud — sherm >16 2kb, 7 4kb zero All 4 P freud _ shenn >16 2kb slow/high All 5 P freud _ freud negative hardly All 6 P freud — shenn >16 2kb fast/high All 7 P freud — sherm >16 2kb fast/high All 8 P freud — sherm negative fast/high All 9 P freud -- shenn >16 2kb fast/high All 10 noP P jensenn — negative fast/high AIII1 P freud _ sherm negative fast/high Alll 2 P freud — freud negative fast/high Alll 3 P freud — sherm negative fast/high Alll 4 P freud — freud >16 2kb fast/high Alll 5 P freud _ shenn negative fast/high Alll 6 P freud _ sherm >16 2kb fast/high Alll 7 P freud — freud >16 2kb hardly Alll 8 P freud — sherm negative fast/high Alll 9 noP P jensenn — >16 2Kb, 6 7kb slow/medium Alll 9 1 P freud — shenn negative fast/high Alll 10 noP noP — negative hardly AIV1 P freud — sherm negative fast/high AlV 2 P freud _ freud negative fast/high AlV 3 P freud _ sherm negative fast/high AlV 4 P freud — shenn negative fast/high AlV 5 P freud — freud negative fast/high AlV 6 P freud — sherm negative fast/high AlV 7 P freud — freud negative fast/high AlV 7 1 P acidipropiomci P acidipropiomci — negative slow/medium AlV 8 P freud — freud negative fast/high AlV 9 P freud — sherm negative fast/high AlV 91 P acidipropiomci P acidipropiomci — negative slow/high AlV 10 P freud — shenn negative fast/high 1 see Table 19 -138- 7. Annex Table 32: Classification of propionibactena strains from Raclette cheese with brown spots ac¬ cording to different charactenstica strain species according to subspecies growth on MF95C' protein profile 23S rRNA R1 noP noP — slow/weak R2 P freud P freud freud slow/high R3 noP noP _. fast/weak R4 noP noP ... slow/weak R5 noP noP — slow/weak R6 P freud P freud freud slow/high R7 noP noP ._ slow/weak R8 noP noP ... fast/weak R9 noP noP ... hardly R10 noP noP — hardly ' see Table 19 Table 33: Classification of propionibactena strains from Sbrinz cheese with brown spots accord¬ ing to different charactenstica strain species according to subspecies plasmids, size growth on protein profile 23S rRNA MF95C1 St 1 P freud — sherm negative fast/high SI 2 P freud — freud >16 2kb fast/high SI 3 noP noP — >16 2kb, fast/weak 12 8kb, 5 5kb SI 6 P freud — freud >16 2kb fast/high SI 7 noP noP — >16 2kb, 8 5kb, fast/weak 4 6kb SIM P freud — sherm >16 2kb fast/high SII2 P freud — shenn >16 2kb fast/high SII3 P freud — sherm >16 2kb fast/high SII4 P freud — sherni negative fast/high SII5 P freud -- sherm negative fast/high SII6 P freud — sherm >16 2kb fast/high SII7 P freud — sherm negative fast/high SII8 P freud — sherm negative fast/high SM 9 P freud -- sherm negative fast/high SII10 P freud _. sherm >16 2kb fast/high sun P freud — sherm negative fast/high Sill 2 P freud — freud negative slow/high Sill 3 P freud — sherm negative fast/high Sill 4 P freud — sherm negative fast/high SHI 5 P freud — fraud negative slow/high Sill 6 P freud — sherm negative fast/high SMI 7 P freud — sherm negative fast/high Sill 8 P freud — shenn negative fast/high Sill 9 P freud — sherm negative fast/high SHI 10 P freud — sherm negative slow/high SIV 2 P freud — freud negative slow/high SIV 3 P freud -- freud negative fast/high SIV 4 P freud — sherm negative fast/high SIV 5 P freud — freud negative fast/high SIV 6 P freud — freud negative fast/high i .2 M S & I I II I I I T3T3T3T3T3-0-D'OX|-a-DT3t3T3T3T3T3T)^>T3T3T3^>T3T3T3T3T3TlT3T3X3T3T3^T3T3T5-OT3^l^333333333333333333333333333333333333333333 O.O.Q.Q.O.Q.O.O.Q.Q.Q.O.O.Q.Q.Q.Q.O.Q.O.O.Q.Q.Q.Q.O.Q.O.O.O.O.O.O.Q.O.Q.Q.O.Q.O.Q.O. O a o(un»«iinM0i»Srf'n"l>,1""l,ll'r h-0OO>^-T_cs|(O^.inu,^aJO>°2;21"!£,~'f^r- ======S ======>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>OTcocowOTwwOTOTcowwOTWOTcowwwoiOTWOTOTOTwcnOTWwcowcotowwwOTcocoOTco -140- 7 Annex Table 34: Classification of propionibactena strains from Gruyere cheese with splitting defect according to different charactenstica strain species according to protein profile subspecies growth on MF95C GM P freud freud fast/high GI2 P freud freud fast/high GI3 P freud freud fast/high GI4 P freud freud fast/high GI5 P freud sherm fast/high GI7 P freud freud fast/high Gl8 P freud freud fast/high GI9 P freud freud fast/high Gl 10 P freud freud fast/high GII1 P freud freud fast/high Gil 2 P freud freud fast/high Gil 3 P freud freud fast/high Gil 4 P freud freud fast/high Gil 5 P freud freud fast/high Gil 6 P freud freud fast/high Gil 7 P freud sherm fast/high Gil 8 P freud freud fast/high Gil 9 P freud freud fast/high GIMO P freud freud fast/high Gill 2 P freud shenn fast/high GUI 3 P freud sherm slow/high Gill 4 P freud shenn fast/high Gill 5 P freud sherm fast/high Gill 6 P freud shenn fast/high Gill 7 P freud sherm fast/high Gill 8 P freud sherm fast/high Gill 9 P freud sherm fast/high Gill 10 P freud sherm fast/high GIV1 P freud sherm fast/high GIV2 P freud freud hardly GIV3 P freud freud fast/high GIV4 P freud freud fast/high GIV5 P freud shenn fast/high GIV6 P freud sherm fast/high GIV7 P freud sherm fast/high GIV10 P freud freud fast/high GV1 P freud shenn fast/high GV2 P freud sherm fast/high GV3 P freud sherm fast/high GV4 P freud shenn hardly GV5 P fraud shenn fast/high GV6 P freud sherm fast/high GV7 P freud sherm fast/high GV9 P freud shenn fast/high GV10 P freud sherm fast/high 'see Table 19 7. Annex -141- Table 35: Classification of propionibactena strains from Sbnnz cheese with splitting defect ac¬ cording to different charactenstica strain species according to protein profile subspecies growth on MF95C1 GSI1 P freud shenn fast/high GSI2 P fraud sherm fast/high GSI3 P freud sherm fast/high GSI4 P freud sherm fast/high GSI5 P freud sherm fast/high GSI6 P freud sherm fast/high GSI7 P freud sherni fast/high GSI8 P freud sherni fast/high GSI9 P freud sherm fast/high GSI10 P freud sherm fast/high GSII1 P freud shenn fast/high GSII2 P freud shenn fast/high GSM 3 P freud sherm fast/high GSII4 P freud freud fast/high GSII5 P freud sherm fast/high GSII6 P freud freud slow/high GSII7 P freud freud fast/high GSII8 P freud freud fast/high GSII9 P freud shenn fast/high GSI110 P freud freud fast/high GSII11 P freud shenn fast/high GS1II2 P freud sherm fast/high GSIII 3 P freud freud fast/high GSIII4 P freud sherm fast/high GSIII 5 P freud shenn fasMiigh GSIII 6 P freud sherm fast/high GSIII 7 P freud sherm fast/high GSIII 8 P freud sherm fast/high GSIII 9 P freud freud fast/high GSII110 P freud sherm fast/high GSIV1 P freud sherm fast/high GSIV 2 P freud sherm fast/high GSIV 3 P freud sherm fast/high GSIV 4 P freud sherm fast/high GSIV 5 P freud sherm fast/high GSIV 6 P freud sherm fast/high GSIV 7 P freud shenn fast/high GSIV 8 P freud sherni fast/high GSIV 9 P freud shenn fast/high GSIV 10 P freud sherm fast/high GSV1 P freud fraud hardly GSV 2 P freud freud hardly GSV 3 P freud sherni fast/high GSV 4 P freud freud hardly GSV 6 P freud sherm fast/high GSV 7 P freud freud hardly GSV 8 P freud freud hardly GSV 9 P freud freud hardly GSV 10 P freud freud hardly 'see Table 19 Curriculum vitae 1969 born September 25th in St. Gallen 1975-1983 primary and secondary education in St. Gallen 1983-1988 high school in St. Gallen, diploma type B 1988-1993 studies at the Swiss Federal Institute of Technology (ETH), Zurich 1993 diploma as dipl. Lm-lng ETH 1993-1997 assistent at the Laboratory of Dairy Science, Swiss Federal Institute of Technology Zurich, thesis at the Federal Dairy Research Institute, Berne-Liebefeld from 1997 Junior Consultant at McKinsey & Company, Inc., Zu¬ rich