INFORMATION TO USERS
The most advanced technology has been used to photograph and reproduce this manuscript from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer.
The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction.
In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion.
Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand corner and continuing from left to right in equal sections with small overlaps. Each original is also photographed in one exposure and is included in reduced form at the back of the book.
Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6" x 9" black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order.
University Microfilms International A Bell & Howell Information Company 300 North Zeeb Road, Ann Arbor, Ml 48106-1346 USA 313/761-4700 800/521-0600 Order Number 9031070
Hyperexpression ofBacillus a thuringiensis delta-endotoxin gene in Escherichia coli and localization of its specificity domain
Ge, Zhixing Albert, Ph.D.
The Ohio State University, 1990
UMI 300 N. Zeeb Rd. Ann Arbor, MI 48106 HYPEREXPRESSION OF A Bacillus thuringiensis DELTA-ENDOTOXIN GENE IN Escherichia coli AND LOCALIZATION OF ITS SPECIFICITY DOMAIN
DISSERTATION
Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of the Ohio State University
By Albert Zhixing Ge, B.S., M.S.
*****
The Ohio State University 1990
Dissertation Committee: Approved by Charles J. Daniels Donald H. Dean
Robert M. Pfister ^ Q v ijx J L ^ GP.-dyK^ William R. Strohl Advisor
Department of Microbiology To The Memory Of My Father To My Mother ACKNOWLEDGMENTS
I wish to express my sincere appreciation to my advisor, Dr. Donald H. Dean, for taking me as his student, and for his continuous support, encouragement, and patience throughout the course of this study. It was under his guidance and care that I evolved to become a scientist with some business sense, and I don’t know where else I could have spent my past five years that were more enjoyable and worthwhile.
I would also like to thank Drs. Charles J. Daniels, Robert M. Pfister, and William R. Strohl for their valuable advice and serving on my Dissertation Committee.
Thanks also go to the past and present members of Dr. Dean’s lab: Brian Almond, Thomas Boyle, Karen Butsch, Chih-I Cheng, George Chen, April Curtiss, Ching Dai, Mary Beth Dunn, Michael Grouse, John Hopper, In-Seok Kwak, Mi Lee, Jim McLinden, Dave Rivers, Josanne Sabourin, and Comrade Dan Zeigler for their assistance, ideas, understanding, and friendship.
Finally, I want to thank my brother and sister for their encouragement and support whenever I needed them, and from which I gained confidence and comfort. VITA
November 7,1956 Born - Shanghai, China
1978 -1980 ...... Shanghai No.2 Medical University Shanghai, China
1981 -1982 ...... B.S., Microbiology, Eastern Michigan University, Ypsilanti, Michigan
1983-1984 ...... Graduate Research Associate, Departments of Chemical Engineering, and Molecular and Cellular Biology The University of Michigan Ann Arbor, Michigan
1982-1984 ...... M.S., Bioengineering, The University of Michigan, Ann Arbor, Michigan
1985-198 6 ...... Graduate Teaching Assistant, Department of Microbiology, The Ohio State University Columbus, Ohio
1986-198 8 ...... M.B.A., Marketing, College of Business The Ohio State University Columbus, Ohio
1986-1989 ...... Graduate Research Associate Department of Microbiology The Ohio State University Columbus, Ohio PUBLICATIONS
Ge, A.Z., Shivarova, N.I., and Dean, D.H.:Location of the Bombyxmori Specificity Domain on a Bacillus thuringiensis delta-endotoxin Protein. Proc. Natl. Acad. Sci. USA 86 (1989) 4037-4041.
Ge, A.Z., Pfister, P.M. and Dean, D.H.: Hyperexpression of a Bacillus thuringiensis delta-endotoxin-encoding gene in Escherichia coli : properties of the product. Gene, (1990).
Milne, R., Ge, A.Z., Rivers, D., and Dean, D.H.: Specificity of Insecticidal Crystal Proteins, Implications For Industrial Standardization. L. Hickle and W. Fitch (eds), American Chemical Society Symposium (1990).
FIELD OF STUDY
Microbiology
v TABLE OF CONTENTS
Dedication ii Acknowledgments iii Vita iv List of Tables viii List of Figures ix
Introduction 1
Literature Review 4
B. thuringiensis and its insecticidal crystal protein 4 The diversity of B. thuringiensis crystal proteins 6 Classification of B. thuringiensis and its crystal protein genes 7 Extrinsic factors affecting the specificity of crystal proteins 10 Intrinsic factors affecting the specificity of crystal proteins 16 Mode of action of B. thuringiensis delta-endotoxin 21 References 24
Part One 31
Introduction 32 Material and methods 34 Strain and growth condition 34 Delta-endotoxin gene, plasmids and restriction enzyme 34 Cloning, transformation and electrophoresis 36 Crystal toxin purification 36 Protein samples preparation 37 Enzyme-linked immunosorbant assay (ELISA) 37 Site-directed mutagenesis and DNA sequencing 38 Results 39 Effects of promoters on expression of cry gene 39 Effect of E. coli host strains on expression 45 Effect of other genetic elements on expression 45 Physical and biological properties of crystals made in E. coli 52 Discussion 52 Conclusion 60 References 61
vi Part Two 64
Introduction 65 Material and Methods 72 Original of delta-endotoxin genes, host and vectors 72 Site-directed mutagenesis and other molecular genetics techniques 72 Construction of site-directed mutants, substitution mutants and overexpression mutants 73 Purification of overexpressed crystal proteins 76 Bioassays 77 Results 78 Specificity of delta-endotoxin 78 Location of specificity domain for B. mori 80 Discussion 84 Conclusion 92 References 93
vii LIST OF TABLES
TABLE PAGE
1.1 Classification of crystal protein toxin genes of B. thuringiensis 9
1.2 Midgut pH and host plant for lepidopteran larvae 11
2.1 Escherichia coli strains and genotypes 35
2.2 Expression of crylA(c)73 in Escherichia coli 44
2.3 Expression of crylA(c)73 by Ptac/pKK223-3 in various host strains 46
2.4 Percentage of plasmid (pUC & pBR) retention in Escherichia coli strain JM103 53
3.1 Specificity of insecticidal crystal proteins on selected insects 79
viii LIST OF FIGURES
FIGURES PAGE
1.0 Positions of five conserved amino acids blocks on Bacillus thuringiensis delta-endotoxins 17
2.1 Cloning scheme of cry1A(c)73 into pKK223-3, pUC19 and pET-3a 40
2.2 Polyacrylamide gel electrophoresis of overexpressed proteins 42
2.3 Microscopic pictures of E. coli strain overexpressing crylA(c)73 and B. thuringiensis HD-73 47
2.4 The mutated 5’DNA sequence of crylA(c)73 50
2.5 Escherichia coli cell expressing crylA(c)73 54
3.1 Comparison of amino acid sequences of cryiA(a), crylA(b), and crylA(c) 67
3.2 Amino acid sequences of the five conserved blocks on B. thuringiensis delta-endotoxins 69
3.3 Construction of substitution mutations and expression plasmid of cry genes 74
3.4 Creation of different hybrid delta-endotoxins between CrylA(a)1 and CrylA(c)73 81
3.5 Computer based plots of protein secondary structures 87
ix INTRODUCTION
Bacillus thuringiensis is a microorganism of considerable economic importance. The key commercial value of B. thuringiensis centers on the delta-endotoxin produced by the cells during their life cycle. B. thuringiensis delta-endotoxin can be crystallized within the sporangium and has been shown to be toxic to many damaging insects such as Manduca sexta (tobacco horn worm), Trichoplusia ni (cabbage looper), Heliothis zea (corn ear worm), Aedes aegypti
(mosquitoes), and Leptinosarsa decemlineata (Colorado potato beetle) (Hall et al., 1977; de barjac, H., 1978; Dulmage, et al., 1981; Aronson, et al., 1986; Andrews, et al., 1987; McPherson, et al., 1988). With its natural protein nature and broad insecticidal spectrum, B. thuringiensis delta-endotoxin has demonstrated a great potential to replace some synthetic, chemical insecticides that may have long term detrimental consequence to our environment.
The mode of action of delta-endotoxin (crystal protein) is an intriguing problem yet to be solved. Crystal proteins from various strains of B. thuringiensis have molecular weight ranging from 25,000 to 130,000 daltons, and possess different host specificities against insects. Current studies on crystal proteins reveal that despite size differences, there are conserved and variable domains on different crystal proteins. The functional significance of the conserved and variable domains is not clear, and is a subject of research pursued by many groups
1 2
around the world. It is speculated that the variable domains on crystal proteins may associate with protein-receptor interaction and as a result, affect the specificity of crystal proteins. As a first attempt to understand the functional domains of crystal proteins, it is desirable to know what differentiates the specificities among crystal proteins. One of the goals of this research, therefore, is to define the specificity domains of crystal proteins. To accomplish this goal, two B. thuringiensis crystal proteins (CrylA(a) and CrylA(c)) were chosen as model proteins. CrylA(a) and CrylA(c) are both lepidoptera specific, and they share more than 80% overall sequence homology. Both toxins have about the same activity against M. sexta. However CrylA(a) is more than 100 times active than CrylA(c) against Bombyx mori (silk worm). CrylA(c), on the other hand, is about 55 times more effective toward Heliothis virescens (tobacco bud worm) than CrylA(a) (Ge et al., 1989).
The strategy devised to carry out this research consists of two components. The first component was to overexpress both genes ( crylA(a), and crylA(c)) so that it would not be of difficult to obtain either gene product for bioassay; the second component was to create several chimeric proteins so that both conserved and variable regions on these two genes could be examined for their roles in crystal protein specificity.
By localizing the specificity domains of B. thuringiensis crystal proteins, it makes it possible to analyze the composition of amino acids in the specificity domain, to understand the functional significance of these amino acids, and to gain insight into the importance of individual amino acids. If the specificity domain 3 is closely related to the binding domain of the crystal protein, then through protein engineering, we could change amino acids in this domain to increase the fidelity between the crystal proteins and receptors and to improve the potency of insecticidal crystal proteins, and create novel insecticidal proteins with desired specificity and better efficacy. LITERATURE REVIEW
B. thuringiensis And Its Insecticidal Crystal Proteins Bacillus thuringiensis is a soil-born, crystalliferous microorganism closely related to B. cerus. In the later growth and stationary phases, B. thuringiensis sporulates, and accompanying sporulation, it produces distinctive parasporal inclusions within the sporangium. The first report of the isolation of this bacterium came in 1902 from Ishiwata (1902) in Japan, as he identified the microorganism as the etiological agent to the Japanese silk worm, Bombyx mori. in 1915, a similar bacterium was also isolated and identified by Berliner (1911) as a causative agent of disease to Anagasta kuehniella (Mediterranean flour moth). Subsequently, Berliner named the microorganism Bacillus thuringiensis after the name of the German province of Thuringia. Berliner has also been credited for the initial description of the presence of an inclusion body inside the sporulating cell.
After the discovery of B. thuringiensis as an insect pathogen a few attempts were made to explore the possibility of using the organism as an insect control agent. Husz (1927) and Chorine (1930) used B. thuringiensis as an insecticide for the control of corn borer. Nevertheless, due to the lack of basic knowledge, early attempts to use B. thuringiensis as a pest control agent were regarded
4 5 inconclusive as summarized by Heimpel and Angus (1963): "In retrospect, all work prior to 1950 indicated that S. thuringiensis var. thuringiensis and related varieties were pathogenic under certain conditions for a number of Lepidoptera larvae, but without yielding sufficient knowledge for rational exploitation."
It was not until early 1950s that the renaissance of research of B. thuringiensis began. In 1951, about 25 years after Berliner (1911) and Mattes (1927) described the inclusion body of B. thuringiensis, Steinhaus initiated a new wave of research on this microorganism. He reported the isolation of a new B. thuringiensis strain, B. thuringiensis var. entomocidus, from diseased larvae of Paralipsa gularis. He attempted to use this microorganism to control the alfalfa caterpillar, Colias eurytheme (Steinhaus, 1951). Despite his effort, however, he failed to recognize the important role played by inclusion bodies within the bacteria. In his paper, Steinhaus showed a micrograph of the diamond-shaped crystals, but only referred them as remains of vegetative cells. In the meantime, working with a culture from Steinhaus, C.L. Hannay (1953) was able to demonstrate the highly retractile, characteristic parasporal inclusions within cells of B. thuringiensis using a new technique. This finding, later, led Hannay (1955) to the discovery of the proteinaceous nature of the parasporal inclusions. Following Hannay’s discovery, Angus (1956) subsequently proved that the insecticidal toxicity of B. thuringiensis was associated with parasporal inclusion. From 1950 to the mid-1970s a number of B. thuringiensis strains were isolated and these strains were believed to be active only on lepidoptera (Andrews et al., 1987). However, some new isolates were soon found to be highly toxic to larvae of certain dipteran insects, especially 6
those hemotagenous insects such as mosquitoes and black flies (Hall et al., 1977; Aronson et al., 1986; Andrews and Bulla, 1982), and coleopteran insects (Sekar et al., 1987; McPherson etal., 1988).
The Diversity OfB. thuringiensis Crystal Proteins In 1980, advancement in genetic engineering initiated a revolution in research on B. thuringiensis insecticidal proteins. Schnepf and Whiteley (1981) were the first to clone a B. thuringiensis crystal protein gene and expressed it in E. coli. The protein extracts from E. coli were demonstrated to be toxic to tobacco hornworm larvae. Since then, many other B. thuringiensis crystal protein toxin genes have been cloned. To date, 42 B. thuringiensis crystal protein gene sequences have been reported. Some of these sequences are identical or very similar to each other. Taking this into account, 14 distinct crystal protein genes can be identified (Hofte and Whiteley, 1989).
The crystal protein production of a B. thuringiensis strain seems to be related to its plasmid profile. Stahly et al. (1978) studied the correlation of the presence of plasmids with crystal production. They reported that crystal production was absent in mutants where plasmids were lost as a result of heat shock treatment. This suggested that plasmids were involved in crystal production. A more detailed investigation of the plasmid content and toxicity of B. thuringiensis strains were reported by Gonzalez et al. (1980; 1981; 1982; 1984). In these studies they showed that the 75-Md plasmid in B. thuringiensis strain HD-2, and the 50-Md plasmid in strain HD-73 were involved in crystal production. Loss of these plasmids resulted in the loss of crystal proteins and toxicity. Therefore, B. 7 thuringiensis delta-endotoxin genes appear to be plasmid-born. Kronstad et al (1983) demonstrated that some B. thuringiensis strains can carry more than one copy of crystal protein genes. They used an internal EcoR\ fragment (ca. 700bp) from crylA(a) gene as a probe, and hybridized it with DNA isolated from 22 B. thuringiensis strains. Hybridization was found to occur in a single plasmid in eight strains, to more than one plasmid in seven strains, and to one or both of large plasmids in two strains. In one case, hybridization was only limited to linear DNA fragments, suggesting that crystal protein genes also may reside on chromosomal DNA. This study also showed that there was considerable sequence homology among different crystal protein genes.
Classifications OfB. thuringiensis And Its Crystal Protein Genes Since the discovery of B. thuringiensis, many strains have been isolated. They seem to have different host specificity. In 1962 deBarjac and Bonnefoi (1962) made an initial attempt to classify different B. thuringiensis strains based on serological tests. They divided the bacteria into six serotypes based upon the H antigen of the flagella. Some of the serotypes were further divided into subserotypes according to the different biochemical test results (deBarjac and Thompson, 1970; deBarjac and Bonnefoi, 1972). This method showed some usefulness of the serotype classification for B. thuringiensis strains isolated in nature. B. thuringiensis strains of the same serotype seem to have a general correspondence in host specificity. However, it became inadequate to define some mutant strains whose delta-endotoxin genes had been altered. It is unlikely that these mutant strains still retain a relationship between specificity and flagella 8
antigen. Currently, there are 22 defined B. thuringiensis serotypes. The host specificity varies among different serotypes. Sometimes individual strains within the same serotype have a distinct spectrum of insect toxicity.
In the 1980s, when gene cloning came into fashion, new terms were used loosely, to classify different crystal protein toxin genes. The first three crystal protein genes cloned from B. thuringiensis were toxic against Lepidoptera, and had Hindlll fragments of 4.5Kb, 5.3Kb, and 6.6 Kb respectively. The Hind\\\ fragment in each case contained the whole 5’ portion of each gene. As a result, these genes were referred to as 4.5Kb gene, 5.3Kb gene, and 6.6Kb gene (Schnepf and Whiteley, 1981; Klier et al., 1982; Adang et al., 1985; Kronstad and Whiteley, 1986; Hofte and Whiteley, 1989). Genes cloned from other strains sometimes were simply referred to as Type B and Type C genes (Hofte et al.,
1988). This certainly created a state of confusion in the B. thuringiensis literature.
Recently, Hofte and Whiteley (1989) have proposed an uniform nomenclature and classification scheme for B. thuringiensis crystal proteins. This system is based on protein structure and host specificity. According to this system, B. thuringiensis crystal proteins can be divided into four major groups: Cryl, Cryll, Crylll, and CryIV (Table 1.1). Cryl group proteins are specific against lepidopteran insects and have molecular weights of about 135-kDa. Cryll group proteins are active against both Lepidoptera and Diptera. Their molecular weight is about 70-kDa. Crylll type proteins are toxic toward coleopteran insects. CryIV group protein toxins are specific against Diptera, especially mosquitoes and black flies. Molecular weight of proteins in this group varies from 64-kDa to 120-kDa. 9 TABLE 1.1
Classification Of Crystal Protein Toxin Genes Of B. thuringiensis
Bacillus Genea thuringiensis Specificity Size(kDa) Other Normenclature subspecies crylA(a) kurstaki HD-1, aizawai, sotto Lepidoptera 133.2 4.5-kb gene, cryl-l crylA(b) berliner 1715, kurstaki HD-1 Lepidoptera 131.0 5.3-kb gene,cryl-2,bt2 aizawai IPL-7, aizawai IC-1 kurstaki NRD-12 crylA(c) kurstaki HD-73, Lepidoptera 133.3 6.6-kb gene kurstaki HD-244 crylB thuringiensis HD-2 Lepidoptera 138.0 cryA4, Type B entomocidus HD-110 crylC entomocidus 601 Lepidoptera 132.5 Type C, BTVLI, Bta cryiD aizawai HD-68 Lepidoptera 132.5 cryllA kurstaki HD-263 Lepidoptera 70.9 P2 gene, cryBI kurstaki HD-1 & Diptera cryllB kurstaki HD-1 Lepidoptera 70.8 cryB2 crylllA san diego, tenebrionis Coleoptera 73.1 cryC crylVA israelensis Diptera 134.4 130-kDa protein gene 125-kDa protein gene, ISRH3 crylVB israelensis Diptera 127.8 130-kDa endotoxin, ISRH4,Bf8,135-kDa protein gene crylVC israelensis Diptera 77.8 ORF1 crylVD israelensis Diptera 72.4 cryD cytA israelensis, morrisoni PG-14 Diptera & 27.4 27-kDa toxin gene Cytolytic
a Classification proposed by Hofte and Whiteley
b crylA genes have been divided into three subclass, (a), (b), and (c) 10
Extrinsic Factors Affecting The Specificity Of Crystal Proteins Even though the mode of action of B. thuringiensis crystal proteins is not fully understood, some extrinsic factors are known to play important roles in the action of crystal proteins against insects. The route of entry of B. thuringiensis crystal protein is through the insect digestive tract. Neither the crystal protein nor the protoxin subunit is toxic upon the injection into hemocoel of host insects or when tested in insect tissue cultures (Luthy and Ebersold, 1981). Therefore, the crystal protein’s solubilization, activation, and interaction with receptors on insect cells are critical to the susceptibility of insects to crystal proteins.
After entering insects, B. thuringiensis crystal proteins first need to be solubilized in the insect midgut before their toxicity can be demonstrated. Crystal proteins are made from individual protein molecules cross linked by disulfide bonds. In vitro experiments have indicated that solubilization of crystal proteins requires either a reducing agent or an alkaline environment. The insect midgut pH is important in the regulation of enzymatic reactions in digestion, protein degradation, the control of solubility and toxicity of stomach poisons, and the determination of the gut flora (House, 1974). Berenbaum (1979) conducted a survey of published midgut pH values for 60 species in 20 families. Even though there was no clear correlation between gut pH and insect feeding habits, he found in his survey that there was a close relationship between alkalinity of the midgut and the growth form of the host plant (Table 1.2). The mean midgut pH of caterpillars which feed exclusively on herbs and forbs is about 8.29 and that of insects which feed only on trees is 8.67 (Berenbaum, 1979). Insects that feed on trees that contain a high degree of condensed tannins tend to have an alkaline 11
TABLE 1.2
Midgut pH and Host Plant For Lepidopteran Larvae
Species Midgut pH Host Plant
Trichoplusia ni 7.0-7.6 Forbs:Gramineae Heliothis obsoleta 8.0 Herbs:90 spp. in 18 herbaceous families including: Gramineae, Polygonaceae, etc. Pierls brassicae 8.0 Herbs:Cruciferae Spodoptera litura 8.2-Q.5 Herbs:114 spp. in 25 herbaceous families including: Solanaceae, Gramineae, etc. Pieris nap/' 9.3 Herbs:Cruciferae Bombyxmori 9.4-10.3 Trees:Moraceae Manduca sexta 9.6 Herbs:Solanaceae Heliothis armlgera 7.8-8.0 Herbs:Amaranthaceae, etc. 12
midgut pH. Condensed tannins are characteristic of many woody families. They can form a complex with proteins in a living system, and as a result, limit the supply of nitrogen. In an alkaline environment the tannin-protein complex is less stable and tends to dissociate. An alkaline environment of insect midgut can help insects to acquire enough nitrogen supply for their growth, but it also facilitates and enhances the action of B. thuringiensis crystal proteins. Faust et al. (1967) showed that initial dissolution of crystals is affected only by the nonenzymatic alkaline agent in the insect midgut. Presolubilized crystal proteins give lower LD50 value than native crystal proteins against S. mori, H. virescens, P. brassicae (Tojo and Aizawa, 1983; Hofte et al., 1986; Jaquet et al., 1987). Crystal proteins that are not readily dissolved in the midgut of Anagasta kuehniella due to its low midgut pH are not effective against the insect. However, the insect may become susceptible to crystal proteins if the crystals are dissolved and activated prior to ingestion (Yamvrias, 1962).
Solubilized S. thuringiensis crystal proteins have to be activated to display insecticidal activity. Experiments carried out with 5’ and 3’ deletion of cry genes showed that the toxic fragment resided within NH2-terminal half of the crystal protein. More precisely, the toxic peptide stretches from codon 29 to codon 607, 615, or 623, dependent upon the gene (Adang et al., 1985; Schnepf and Whiteley, 1985; Hofte et al., 1986; Sanchis et al., 1989; Bietlot et al., 1989). The activation of crystal protein is carried out by insect midgut proteases. Tojo and Aizawa (1983) reported that gut juice proteases from Silkworm, Bombyx mori, were able to degrade the 130,000 Da crystal protein from B. thuringiensis subsp. kurstaki strain HD-1 to a 59,000 Da fragment. The 59,000 Da peptide retained the toxicity 13 to B. mori. Similar findings on the activation of crystal proteins by gut proteases were also reported by other researchers (Luthy and Ebersold, 1981). Andrews et al.(1985) showed that the crystal protein could also be activated in vitro by trypsin. Tryptic digestion of 135,000 Da crystal protein produced a 68,000 Da peptide, which was toxic to both Trichoplusia ni and Aedes aegypti. Since insect midgut tends to have a high pH, insect gut proteases are generally considered alkaline proteases. Ishaaya et al. (1971) showed that the optimum pH for proteases activity in Spodoptera littoralis was about 11. The protein content in the diet also strongly stimulated the proteolytic activity, especially in the midgut wall.
The susceptibility of insects to B. thuringiensis crystal protein may also depend on the presence of the right proteases inside insect midgut. Haider et al. (1986) reported that treatment of the native crystal protein toxin from B. thuringiensis var. colmeri with Aedes aegypti gut extract gave a 52,000 Da protein. The protein was fully toxic to all mosquitoes, but only to one lepidopteran insect, Spodoptera frugiperda. On the other hand, a 55,000 Da toxic fragment produced from the treatment with Pieris brassicae midgut enzyme or trypsin was specific only to lepidopteran cell lines. In vivo bioassay data were in agreement with these findings. A two-step activation of the 130,000 Da protein by successive treatment with trypsin and A. aegypti gut enzyme suggested that the 52,000 Da dipteran toxin was derived from the 55,000 Da lepidopteran toxin.
The display of activity by the activated B. thuringiensis delta-endotoxin may be, to a large extent, dependent on the interaction between the toxin and the toxin-binding sites on the insect midgut epithelial cells. Experiments carried out 14 by Hofmann et al (1988) demonstrated that high affinity binding sites were present on insect midgut brush border membrane. In one experiment, crylA(b) and crylB toxins were labeled with I125 and tested on brush border membrane vesicles (BBMV) prepared from Pierce brassicae and Manduca sexta. CrylA(b) is toxic against both insects, and CrylB is active only against P. brassicae. The binding test revealed that the activated CrylA(b) bound with high affinity to BBMV from both insects, while CrylB toxin only bound to BBMV from P. brassicae. This result was further strengthened by subsequent heterologous competition experiments with other toxins. CrylA(a) and CrylA(c) toxins were about equally toxic against M. sexta, and CrylVB, a diptera specific toxin, had no activity toward the same insect. Using I125 labeled CrylA(b) toxin, competitive binding experiments showed that both CrylA(a) and CrylA(c) competed with CrylA(b) for the same binding sites, whereas the diptera specific CrylVB did not. Interestingly, the binding experiment also revealed that CrylA(b) and CrylB toxins had different binding sites on the same BBMV from P. brassicae.
Similar binding experiment was also performed by Van Rie et al (Van Rie et al., 1990). CrylA(a), CrylA(b), and CrylA(c) protein toxins have about the same activity against M. sexta, and CrylA(c), as well as CrylA(b), are much more toxic against H. virescens than CrylA(a). I125 labeled toxins were used in the binding assay of BBMV from M. sexta and H. virescens. It was found that the binding constants of these three toxins were not significantly different. However, the concentrations of binding sites to these toxins on BBMV from H. virescens were distinctly different. The concentration of binding sites for CrylA(a), CrylA(b), and 15
CrylA(c) were 3.7,9.7, and 19.5 pmol/mg BBMV respectively. The difference in binding sites concentration reflects the difference of in vivo toxicity of these three toxins towards H. virescens.
The nature of binding sites is still a matter of active research. Treatment of binding sites with various proteases showed that at least part of the binding sites is proteinaceous (Hofmann et al., 1988). Knowles and Ellar (1986) earlier suggested that a 146 kDa glycoprotein from CF-1 cell membrane may be a possible target for toxin binding. They also reported that A/-acetylgalactosamine could block the binding of toxins from B. thuringiensis var. kurstaki HD-1 to this putative receptor on the CF-1 cell membrane (Knowles et al., 1984). However, A/-acetylgalactosamine did not show any interference with the binding of I125 labeled CrylA(b) to BBMV from M. sexta and CrylB to BBMV from P. brassicae (Hofmann et al., 1988). Since the CF-1 cell line was established from the ovary cell of Choristoneura fumiferana, it may not be representative of insect midgut cells.
The insecticidal activity of B. thuringiensis crystal proteins may be enhanced by B. thuringiensis spores. Mohd-Salleh et al. (1982; 1983) tested B. thuringiensis var. kenyae, B. thuringiensis var. galleriae, and B. thuringiensis var. kurstaki against 6-day old European cornborer larvae, Ostrinia nubilalis. Although pure spores were not and pure crystals were toxic to the larvae, maximum larval mortality was achieved only when they were fed with a combination of spores and crystals. Prasertphon and Tanada (1973) reported that sporulation of B. thuringiensis was found in three insects they studied. When third-instar larvae of 16
Heliothis armigera were fed with spores at different concentrations, all were dead. Seven of 12 dead insects examined had spores and crystals. Similar findings were also observed in the second-instar larvae Prodenia litura, and the third- and forth-instar larvae of Spodoptera exigua. Certainly B. thuringiensis crystal proteins are mainly responsible for the insecticidal activity of B. thuringiensis. After the initial action by crystal proteins, bacterial spores or vegetative cells may penetrate to the hemolymph of insects and germinate. Cells could enter vegetative stage and sporulate using insect derived nutrients. Burges et al also suggested that a 130-140 kDa spore coat protein similar to crystal protein may be toxic to insects (Burges et al., 1976). However, this evidence is not conclusive.
Intrinsic Factors Affecting The Specificity Of Crystal Proteins Probably because of the lack of understanding of the tertiary structure of B. thuringiensis insecticidal crystal proteins, most published research on the specificity of crystal proteins tends to be limited on processes which act upon crystal proteins. Only a few papers in the literature address the topic of crystal protein specificity per se.
Although B. thuringiensis produces a number of insecticidal crystal proteins, and the specificities of these proteins vary, these proteins do share a certain degree of homology. From an alignment of crystal protein sequences from each group it is possible to identify five conserved regions (Figure 1.0). These regions were separated by highly variable sequences of various lengths for the different crystal proteins. The Cryll proteins and CrylVD protein are little different. These proteins show significant sequence homology to other cry proteins only in the Figure 1.0 Amino acid sequence comparison of delta-endotoxins from different groups. The five highly conserved blocks along proteins are marked by dark boxes and numbered accordingly.
17 AMINO ACID SEQUENCE COMPARISON BETWEEN LEPIDOPTERAN TOXINS (Cry IA 100 200 300 400 500 600 700 800 900 1000 1100 * * * +*** + **» conserved I hypervariable partially conserved cry IA(a)1 cry IA(b)1 cry IA(c)73 cry III A cry IV B [ X W MXJOOZW» -00 cry IV C ( I 1 300 cry II A C X 1 3 4 5 . TOXIC FRACTION ^ “ ■ . P BLOCK OF SIMILAR AN AMINO ACIDS = ■ Figure 1.0 19 block 1 region. Analysis of many 130-kDa proteins reveals that these proteins have a very high content of glutamic and aspartic acid which account for more than 20% of the total amino acids. These two amino acids are responsible for the low isoelectric point of 4.4 (Cooksey, 1971). The 130-kDa proteins can be divided into the NH2-terminal half (amino acids 1 to 618) and COOH-terminal half. The COOH-terminal half is highly conserved among 130-kDa proteins. It contains the majority of cysteine residues, and presumably is involved in the formation of the crystalline structure. The NH2-half of the protein, which is the toxic fragment of the protein, can be further divided into the conserved domain (amino acids 1 to 280) and variable domain (amino acids 280-623). In the conserved domain, especially within the first 120 amino acids there is a section of amino acids which demonstrate hydrophobic characters and are representative of trans-membrane structure. Interestingly, although amino acids composition may vary in this region among different delta-endotoxins the hydrophobic character is conserved, suggesting a functional significance. Recent experiments (Convents et al., 1990) suggest that there are two structural domains on the activated toxins specific against lepidopteran larvae (Cryl type). The evidence came from physical chemistry study of CrylA(b) proteins from B. thuringiensis var. berliner 1715. The activated B. thuringiensis toxin was examined by fluorescence spectroscope at different pH (4, 8, and 11) in both native and denatured conditions. It was observed that the protein unfolding pattern at pH4 and pH11 were biphasic suggesting a two domain structure. In addition, circular dichroism spectra at 222nm revealed an unfolding of a protein region rich in alpha-helix structure. This part of the protein was less stable at 20 pH11 than at pH8. This evidence was further supported by electrophoretic pattern of protease digestion. Four proteases (chymotrypsin, thermolysin, pronase, and papain) were employed to digest CrylA(b) toxin. Results from proteases digestions showed that two protease resistant peptides were generated. According to these findings it was proposed that CrylA(b) toxin was composed of two protease resistant structural entities linked by a flexible region. These data are in good agreement with the secondary structure prediction. The N-terminal half of the toxin contains several helical stretches, whereas the C-terminal half of the toxin contains alternating beta-sheet structures. The N-terminal half of the toxin may be involved in perturbation of cell membrane. The interaction between toxin and cell membrane was simulated in vitro by using activated crystal proteins and phospholipid liposomes (Haider and Ellar, 1989). Liposomes were loaded with tritium labeled uracil or tritium labeled alanine. Addition of protein toxins into the liposome suspension caused immediate decrease in light transmittance and leakage of tritium labeled molecules into solution. This suggests that protein toxin bind to phospholipid vesicles and cause reorganization of lipid assemblies, which includes vesicularization and aggregation and/or fusion. Deletion study found that only the first 240 amino acids of the activated toxin were required to interact with liposome (Haider and Ellar, 1989). This probably reflects the functional importance of alpha-helical structures in this region. 21 If the two domain model for lepidoptera-specific protein (130-140kDa) stands, then a reasonable guess is that the C-terminal half of the toxin should contain the binding domain of the toxin, and therefore, is the intrinsic factor that determines the fidelity between the toxin and receptors, and as a result, the protein toxin specificity. Unfortunately, there has been very little study on the C-terminal half of the toxin, where amino acid sequences vary greatly among different toxins. Haider et al. (1989) have shown that the change of the individual amino acid (Ile568 to Thr) at the C-terminal half of CrylA(b) toxin was able to alter the specificity of the toxin. The single change of amino acid transformed a lepidoptera specific toxin into a diptera specific toxin. This probably was accomplished by altering the protease sensitivity of arginyl peptide bond at Arg567. However, there is no report in the literature that present direct evidence to account for the difference of specificities among lepidoptera specific toxins because of differential proteolytic action. Mode Of Action Of B. thuringiensis Delta-endotoxin The mechanism by which delta-endotoxin acts upon insect midgut epithelial cells may start to be clear after years of controversy. Travers et al. (1976) reported that delta-endotoxin had stimulatory effect on oxygen uptake by mitochondria isolated from B. mori midgut epithelium. And at the sametime it inhibited ATP production. This indicates that delta-endotoxin may act as an uncoupler of oxidative phosphorylation. However, this view is not shared by others. Tojo (1986) showed that the activated delta-endotoxin did not enter midgut epithelial cells and the target of delta-endotoxin seemed to be cell membrane rather than mitochondria. Activated toxin could cause physical 22 damage to cell membrane prepared from B. mori midgut epithelial cells. Angus reported that delta-endotoxin could cause similar symptoms in B. mori as the antibiotic valinomycin (Angus, 1968). This suggested that the toxin might act as an ionophore, probably affecting K+ permeability. Probably the most detailed study on the mode of action of B. thuringiensis delta-endotoxin was from Knowles et al. (1987). In their investigation it was found that delta-endotoxin could form small pores on the cell membrane, either by directly inserting itself into cell membrane or by perturbing cell membrane structure. The formation of small pores on cell membrane leads to a net ion inflow and an equilibration of ion concentration across the cell membrane. As a result of a net ion inflow an influx of water causes cell swelling, leakage of large cellular molecules, and eventual lysis. Large molecules in the medium such as raffinose and sucrose can serve as osmotic protectants and prevent this colloid osmotic lysis. Using osmotic protectants pore size was estimated about 1-2 nm in diameter. A number of events could occur as a result of colloid osmotic lysis. Some of these events may also account for phenomena people observed earlier. In normal physiological condition apical membranes of insect midgut epithelial cells (columnar cells and goblet cells) form an electroosmotic barrier separating the midgut lumen from cell cytoplasm. A K+-ATPase system located on the goblet cell apical membrane is involved in generating three gradients across this barrier (Gupta et al., 1985). An electrical voltage gradient more than 180 mV with lumen side positive is responsible for amino acid uptake via columnar cells through 23 voltage-dependent K+ symports. An over 100-fold H+ gradient with lumen side alkaline and a 10 fold K+ gradient with lumen side concentrated are adaptations to the high tannin and high K+ content in dietary plant material. Upon the action by delta-endotoxin, because of formation of small pores on cell membrane, the K+ gradient starts to dissipate. However, initially the active, ATP dependent K+ pump does not seem to be affected. It continues to pump K+ from hemolymph-side to lumen-side. The K+ gradient dissipates as a result of the increased passive leak-back of the K+ from lumen to hemolymph (Harvey, and Wolfersberger, 1979). The disruption of a steady state K+ gradient across midgut epithelial cells can lead to a rapid membrane depolarization. The loss of a K+ gradient will affect insect nutrient uptake as some amino acids uptake is K+ dependent (Sacchi, et al., 1986). The increase of permeability of cell membrane also eliminates the trans-membrane pH gradient and raises the intracellular pH. As a result, the ATP level would drop because the oxidative phosphorylation cannot occur properly in an alkaline environment. Finally, the disruption of cellular metabolism would activate lysosomes and lead to permenant structural damage (Harvey et al., 1986) REFERENCES Adang, M. J., Staver, M. J., Rocheleau, T. A., Leighton, J., Barker, R. F. and Thompson, D. V.: Characterized full-length and truncated plasmid clones of the crystal protein of Bacillus thuringiensis subsp. kurstaki HD-73 and their toxicity to Manduca sexta. Gene 36 (1985) 289-300. Andrews, R.E., Jr. and Bulla, L.A., Jr. Toxins of sporeforming bacteria. In: Spores VIII, edited by Levinson, H.S., Sonenshein, A.L. and Tipper, D.J. Washington,D.C.: American Society for Microbiology, 1982, p. 57. Andrews, R.E., Jr., Bibilos, M.M. and Bulla, L.A.:Protease activation of the entomocidal protoxin of Bacillus thuringiensis subsp. kurstaki. Appl. Environ. Microbiol. 50 (1985) 737-742. Andrews, R. E ., Jr., Faust, R. M., Wabiko, H., Raymond, K. C. and Bulla, L. A ., Jr.: The biotechnology of Bacillus thuringiensis. CRC Critical Reviews in Biotechnology 6 (1987) 163-232. Angus, T. A.: Association of toxicity with protein-crystalline inclusions of Bacillus sotto Ishiwata. Can. J. Microbiol. 2 (1956) 122-131. Angus, T. A.: Similarity of effect of valinomycin and Bacillus thuringiensis parasporal protein in larvae of Bombyx mori. J. Invert. Pathol. 11 (1968) 145-146. Aronson, A. I., Beckman, W. and Dunn, P .'.Bacillus thuringiensis and related insect pathogens. Microbiol. Rev. 50 (1986) 1-24. Berenbaum, M.: Adaptive significance of midgut pH in larval lepidoptera. The American Naturalist (1979) 138-146. Berliner, E.: Uber de schlaffsucht der Mehlmottenraupe. Zeitschrift fur das Gesamstadt 252 (1911) 3160-3162. Bietlot, H., Carey, P. R., Choma, C., Kaplan, H., Lessard, T. and Pozsgay, M.: Facile preparation and characterization of the toxin from Bacillus thuringiensis var. kurstaki. Biochem. J. 260 (1989) 87-91. 24 25 Burges, H. D., Thomson, E. M. and Latchford, R. A.: Importance of spores and delta-endotoxin protein crystals of Bacillus thuringiensis in Galleria mellonella. J. Invert. Pathol. 27 (1976) 87-94. Chorine, V.: De I’ntitisation des microbes entomophytes dans la lutte contre les insects unisibles et de la destruction par ces microbes des chenilles de la pyrale du Mais. Acta. Bot. Inst. R. Univ. Zagrebensis, 5 (1930) 7-17. Convents, D., Houssier, C., Lasters, I. and Lauwereys, M.: The Bacillus thuringiensis delta-endotoxin Evidence fro a two domain structure of the minimal toxic fragment. J. Biol. Chem. 265 (1990). Cooksey, K.E. The protein crystal toxin of Bacillus thuringiensis: Biochemistry and mode of action. In: Microbial control of insects and mites, edited by Burges, H.D. and Hussey, N.W. London: Academic press, 1971, p. 247-274. de Barjac, H. and Bonnefoi, A.: Essai de classification biochimique et serologique de 24 souches de Bacillus du type B. thuringiensis. Entomophaga 7 (1962) 5-31. de Barjac, H. and Thompson, D. V.: A new serotype of Bacillus thuringiensis var. thompsoni (Serotype 11). J. Invert. Pathol. 15 (1970) 141-144. de Barjac, H. and Bonnefoi, A.: Presence of H antigenic subfactors in serotype V of Bacillus thuringiensis with the description of a new type: Bacillus thuringiensis var. canadensis. J. Invert. Pathol. 20 (1972) 212-213. de Barjac, H.: Un nouveau candidat a la lutte biologique contre les moustiques: Bacillus thuringiensis var. israeiensis. Entomophaga. 23 (1978) 309-319. Dulmage, H.T. and Cooperators: Insecticidal activity of isolates of Bacillus thuringiensis and their potential for pest control. In: Microbial Control of Pests and Plant Diseases 1970-1980, edited by Burges, H.D. London: Academic Press, 1981, p. 193-222. Faust, R. M., Adams, J. R. and Heimpel, A. M.: Dissolution of the toxic parasporal crystals from Bacillus thuringiensis var. pacificus by the gut secretions of the silkworm, Bombyx mori. J. Invert. Pathol. 9 (1967) 488-499. 26 Ge, A.Z., Shivariva, N.I. and Dean, H.D.: Location of the Bomby mori specificity domain on a Bacillus thuringiensis delta-endotoxin protein. Proc. Natl. Acad. Sci. USA 86 (1989) 4037-4041. Gonzalez, J. M ., Jr. and Carlton, B. C.: Patterns of plasmid DNA in crystalliferous and acrystalliferous strains of Bacillus thuringiensis. Plasmid 3 (1980) 92-98. Gonzalez, J. M ., Jr., Dulmage, H. T. and Carlton, B. C.: Correlation between specific plasmids and endotoxin production in Bacillus thuringiensis. Plasmid 5 (1981)351-365. Gonzalez, J. M ., Jr., Brown, B. J. and Carlton, B. C.: Transfer of Bacillus thuringiensis plasmids coding for delta-endotoxin among strains of B. thuringiensis and B. cereus. Proc. Natl. Acad. Sci. USA 79 (1982) 6951-6955. Gonzalez, J. M ., Jr. and Carlton, B. C.: A large transmissible plasmid is required for crystal toxin production in Bacillus thuringiensis variety israelensis. Plasmid 11 (1984) 28-38. Gupta, B. L., Dow, J. A. T., Hall, T. A. and Harvey, W. R.: Electron probe X-ray microanalysis of the effects of Bacillus thuringiensis var kurstaki crystal protein insecticide on ions in an electrogenic K+-transporting epithelium of the larval midgut in the lepidopteran, Manduca sexta in vitro. J. Cell Sci. 74 (1985) 137-152. Haider, M. Z., Knowles, B. H. and Ellar, D. J.: Specificity of Bacillus thuringiensis var. colmeri insecticidal delta-endotoxin is determined by differential proteolytic processing of the protoxin by larval gut proteases. Eur. J. Biochem. 156 (1986) 531-540. Haider, M. Z. and Ellar, D. J.: Functional mapping of an entomocidal delta-endotoxin. Single amino acid changes produced by site-directed mutagenesis influence toxicity and specificity of the protein. J. Mol. Biol. 208 (1989) 183-194. Haider, M. Z. and Ellar, D. J.: Mechanism of action of Bacillus thuringiensis insecticidal delta-endotoxin: Interaction with phospholipid vesicles. Biochim. Biophys. Acta 978 (1989) 216-222. 27 Hall, I.M., Arakawa, K.Y., Dulmage, H.T. and Correa, J.A.: The pathogenicity of strains of Bacillus thuringiensis to larval of Aedes and to Culex mosquitoes. Mosq. News 37 (1977) 246-251. Hannay, C.L.: Crystalline inclusion of Bacillus thuringiensis Berliner. Nature 172 (1953) 1004-1004. Hannay, C. L ., and Fitz-James, P.: The protein crystals of Bacillus thuringiensis Berliner. Can. J. Microbiol. 1 (1955) 694-710. Harvey, W.R. and Wolfersberger, M.G.: Mechanism of inhibition of active potassium transport in isolated midgut of Manduca sexta by Bacillus thuringiensis endotoxin. J. Exp. Biol. 83 (1979) 293-304. Harvey, W.R., Cioffi, M. and Wolfersberger, M.G.: Transport physiology of lepidopteran midgut in relation to the action of Bacillus thuringiensis delta-endotoxin. In: Fundamental and applied aspects of invertebrate pathology, edited by Samson, J.M., Vlak, J.M. and Peters, D. Wageningen, The Netherlands.: Society of Invertebrate Pathology, 1986, p.11-13. Heimpel, A.M. and Angus, T.A. Diseases caused by certain sporeforming bacteria. In:Insect pathology and advanced treaties, Vol 2. edited by Steinhaus, E.A.: Academic press, 1963, p21- Hofmann, C., Luthy, P., Hutter, R. and Pliska, V.: Binding of the delta-endotoxin from Bacillus thuringiensis to brush-border membrane vesicles of the cabbage butterfly (Pieris brassicae). Eur. J. Biochem. 173 (1988) 85-91. Hofmann, C., Vanderbruggen, H., Hofte, H., Van Rie, J., Jansens, S. and Van Mellaert, H.: Specificity of Bacillus thuringiensis delta-endotoxin is correlated with the presence of high-affinity binding sites in the brush border membrane of target insect midguts. Proc. Natl. Acad. Sci. USA 85 (1988) 7844-7848. Hofte, H., deGreve, H., Seurinck, J., Jansens, S., Mahillon, J., Ampe, C., Vandekerckhove, J., vanMontagu, M., Zabeau, M. and Vaeck, M.: Structural and functional analysis of a cloned delta endotoxin of Bacillus thuringiensis beriiner 1715. Eur. J. Biochem. 161 (1986) 273-280. 28 Hofte, H., Van Rie, J., Jansens, S., Van Houtven, A., Vanderbruggen, H. and Vaeck, M.: Monoclonal antibody analysis and insecticidal spectrum of three types of lepidopteran-specific insecticidal crystal proteins of Bacillus thuringiensis. Appl. Environ. Microbiol. 54 (1988) 2010-2017. Hofte, H. and Whiteley, H. R.: Insecticidal crystal proteins of Bacillus thuringiensis. Microbiological Reviews 53 (1989) 242-255. House, H.L. Digestion. In: Physiology of Insects, edited by Rockstein, M. New York: Academic Press, 1974, p. 63-117. Husz, B.: Bacillus thuringiensis beriiner, a bacterium pathogenic to corn borer larvae. A preliminary report. International Corn borer Invest. Sci. Rpts., 1927, 1927-1928, p. 191-193. Ishaaya, I., Moore, I. and Joseph, D.: Protease and amylase activity in larvae of the Egyptian cotton worm, Spodoptera littoralis. J. Insect. Physiol. 17 (1971)945-953. Ishiwata, S.: One kind of severe flacherie (sotto disease) Daihihon Sanshi Keiko 9 (1902) 1-5. Jaquet, F., Hutter, R. and Luthy, P.: Specificity of B. thuringiensis delta-endotoxin. Appl. Environ. Microbiol. 53 (1987) 500-504. Klier, A., Fargette, F., Ribier, J. and Rapoport, G.: Cloning and expression of the crystal protein genes from Bacillus thuringiensis strain beriiner 1715. EMBOJ. 1 (1982) 791-799. Knowles, B. H., Thomas, W. E. and Ellar, D. J.: Lectin-like binding of Bacillus thuringiensis var. kurstaki lepidopteran-specific toxin is an initial step in insecticidal action. FEBS Lett. 168 (1984) 197-202. Knowles, B. H. and Ellar, D. J.: Characterization and partial purification of a plasma membrane receptor for Bacillus thuringiensis var kurstaki Lepidopteran-specific delta-endotoxin. J. Cell Sci. 83 (1986) 89-101. Knowles, B. H. and Ellar, D. J.: Colloid-osmotic lysis is a general feature of the mechanism of action of Bacillus thuringiensis delta-endotoxin with different insect specificity. Biochim. Biophys. Acta 924 (1987) 509-518. 29 Kronstad, J. W., Schnepf, H. E. and Whiteley, H. R.: Diversity of locations for Bacillus thuringiensis crystal protein genes. J. Bacteriol. 154(1) (1983) 419-428. Kronstad, J. W. and Whiteley, H. R.: Three classes of homologous Bacillus thuringiensis crystal-protein genes. Gene 43 (1986) 29-40. Luthy, P. and Ebersold, H.R. Bacillus thuringiensis delta-endotoxin: histopathology and molecular mode of action. In: Pathogenesis of Invertebrate Microbial Diseases, edited by Davidson, E.W. Totowa, NJ: Allanheld Osmun, 1981, p. 235-268. Mattes, O.: Parasitare krankheiten der meulmotten larven und versuche uber ihre verwendbarkeit als biologisches bekampfungsmittel. Sitzber. Ges. Befoerder. Ges. Naturw. Marburg, 62 (1927) 381-417. McPherson, S. A., Perlak, F. J., Fuchs, R. L., Marrone, P. G., Lavrik, P. B. and Fischhoff, D. A.: Characterization of the coleopteran-specific protein gene of Bacillus thuringiensis var. tenebrionis. Bio/Technology 6 (1988) 61-66. Mohd-Salleh, M. B. and Lewis, L. C.: Toxic effects of spore/crystal ratios of Bacillus thuringiensis on European corn borer larvae. J. Insect. Pathol. 39 (1982) 290-297. Mohd-Salleh, M. B. and Lewis, L. C.: Comparative effects of spore-crystal complex and thermostable exotoxin of six subspecies of Bacillus thuringiensis on Ostrinia nubilalis (Lepidoptera: Pyralidae). J. Invert. Pathol. 41 (1983) 336-340. Prasertphon, S., Areekul, P. and Tanada, Y.: Sporulation of Bacillus thuringiensis in host cadavers. J. Invert. Pathol. 21 (1973) 205-207. Sacchi, V.F., Parenti, P., Hanozet, G.M., Giordana, B., Luthy, P. and Wolfersberger, M.G.: Bacillus thuringiensis toxin inhibits K+ gradient-dependent amino acid transport across the brush border membrane of Pieris brassicae midgut cell. FEBS Lett. 204 (1986) 213-218. Sanchis, V., Lereclus, D., Menou, G., Chaufaux, J., Guo, S. and Lecadet, M. -M: Nucleotide sequence and analysis of the N-terminal coding region of the Spodoptera active delta-endotoxin gene of Bacillus thuringiensis aizawai 7.29. Mol. Microbiol. 3 (1989) 229-238. 30 Schnepf, H. E. and Whiteley, H. R.: Cloning and expression of the Bacillus thuringiensis crystal protein gene in Escherichia coli. Proc. Natl. Acad. Sci. USA 78(5) (1981) 2893-2897. Schnepf, H. E. and Whiteley, H. R.: Delineation of a toxin-encoding segment of a Bacillus thuringiensis crystal protein gene. J. Biol. Chem. 260(10) (1985) 6273-6280. Sekar, V., Thompson, D. V., Maroney, M. J., Bookland, R. G. and Adang, M. J.: Molecular cloning and characterization of the insecticidal crystal protein gene of Bacillus thuringiensis var. tenebrionis. Proc. Natl. Acad. Sci. USA 84 (1987) 7036-7040. Stahly, D. P., Dingman, D. W., Bulla, L. A ., Jr. and Aronson, A. I.: Possible origin and functions of the parasporal crystals in Bacillus thuringiensis. Biochem. Biophys. Res. Commun. 84 (1978) 581-588. Steinhaus, E. A.: Possible use of B.t. beriiner as an aid in the control of alfalfa caterpillar. Hilgardia 20 (1951) 359-381. Tojo, A. and Aizawa, K.: Dissolution and degradation of Bacillus thuringiensis delta-endotoxin by gut juice protease of the silkworm Bombyxmori. Appl. Environ. Microbiol. 45 (1983) 576-580. Tojo, A.: Mode of action of bipyramidal delta-endotoxin of Bacillus thuringiensis subsp. kurstaki HD-1. Appl. Environ. Microbiol. 51 (1986) 630-633. Travers, R. S., Faust, R. M. and Reichelderfer, C. F.: Effects of Bacillus thuringiensis var. kurstaki delta-endotoxin on isolated lepidopteran mitochondria. J. Invert. Pathol. 28 (1976) 351. Van Rie, J., Jansens, S., Hofte, H., Degheele, D. and Van Mellaert, H.: Specificity of Bacillus thuringiensis delta-endotoxin: Importance of specific receptors on the brush border membrane of the mid-gut of target insects. Eur. J. Biochem. (1990). Yamvrias, C.: Contribution of I’etude du mode d’action de Bacillus thuringiensis vis-a-vis de la teigne de la farine, Anagasta kuehniella Zell. (Lepidoptera). Entomophaga 7 (1962) 101-159. PART ONE Hyper-expression and Properties of Bacillus thuringiensis delta-endotoxin in E. coi'r. 48% of Total Cellular Protein 31 INTRODUCTION Among biological insect control agents, Bacillus thuringiensis is one of the most commonly used microbes. The insecticidal ability of B. thuringiensis is mainly attributed to the presence of its insecticidal crystal protein (Cry) toxins synthesized when cells undergo sporulation. To date, about 14 distinct S. thuringiensis crystal proteins and their genes have been identified and studied (Hofte and Whiteley, 1989). They constitute a broad spectrum of insecticides against insects of Lepidoptera, Diptera, and Coleoptera. In B. thuringiensis, cry genes are expressed differentially. Thus, the specificity of B. thuringiensis against insect hosts varies, depending not only upon the presence or absence of a particular cry gene, but also upon the level of expression of a cry gene. For instance, crylA(c) gene, although present, is believed to be poorly expressed in strain HD-1 (Yamamoto et al., 1988; Rothnie and Geiser, 1989); the content of Cryll(A,B) proteins also varies among those strains that carry its gene (Yamamoto and lizuka, 1983). The mechanisms which control this variable expression are not known. Because the B. thuringiensis Cry toxin is a biological pesticide of considerable economic importance, it is desirable to study factors that influence the expression of these genes. An effective expression system for cry genes in E. coii, where the parameters of expression are better understood, will allow selective expression of cry genes of particular host specificity and facilitate the study of their gene 33 products. It can also serve as an effective vehicle for industrial production to deliver genetically modified and functionally improved crystal proteins as a result of protein engineering. A number of factors can be altered to improve the expression of the cry genes. Selection of a strong E. coli promoter to improve the expression of cry genes has been reported by Oeda, et al. (1987). By introducing the tac promoter upstream of a 130kDa gene (crylA(b)) from B. thuringiensis aizawai IPL7, they achieved an enhanced expression. However, they did not measure the level of expression. In this study E. coli expression systems for B. thuringiensis cry genes and the influence of other factors, including different E. coli host strains, high and low copy number plasmids, the second codon of cry gene, and extra terminators were investigated. At optimal conditions, the level of expression of B. thuringiensis crylA(c) gene can amount to about 50% of total cellular proteins, and can produce up to 248 mg/l in shake flasks. 34 MATERIALS AND METHODS (a) Strains and growth conditions. Strains used in this study are shown in Table 2.1. E. coli strains JM101, JM103, and JM105 were obtained from J. Messing (Rutgers, The State University, New Brunswick, NJ). SG 21166 and SG 1117, both Ion' strains, were obtained from Susan Gottesman. DE3, host strain of pET-3a, was obtained from F.W. Studier. All transformants were grown in LB with ampicillin (50 ug/ml) in shake flasks at 37°C in a water bath shaker at a speed of 275 rpm. Cell growth was monitored by a spectrophotometer (Beckman Model 25) at OD=600. DE3 strain was induced by IPTG at a final concentration of 1 mm at OD600= 1.0. (b) Delta-endotoxin gene, plasmids and restriction enzymes. Plasmid pKK223-3 was obtained from Pharmacia; pET-3a was obtained from F.W. Studier, and pUC19 was obtained from J. Messing. Plasmid pOS1002, described previously (McLinden et al., 1985), was used as a source of delta-endotoxin gene crylA(c)73 from B. thuringiensis var. kurstaki, HD 244. The gene nomenclature suggested by Hofte and Whiteley (1989) is employed. Plasmids were purified by cesium chloride gradient method (Maniatis et al., 1982). All enzymes were purchased from BRL except Ndel which was from New England Biolabs. 35 TABLE 2.1 Escherichia coli strains and genotypes Strain Genotype JM101 supE, thi,A(Jac-proAB), [F’ traD36, proA+B +, /ac/qZAM15] JM103 supE, thi, rpsL, endAI, sbcB 15, hsdR4,&(lac-proAB) [F’ traDZQ, proA+B +, /ac/qZAM15] JM105 thi, rpsL, endA'\, sbcB15, hsdR4, a (lac-proAB) [F’ traD36, proA+B +, /ac/qZAM15] SG21166 lon-510,Alac,supF15 SG1117 Ion- 146::Tn10 DE3 F, hsdR, gal 36 (c) Cloning, transformation and electrophoresis. Plasmid pOS1002 was digested with Ndel, and the 3.7 kb fragment which carried the crylA(c) gene was separated by electrophoresis in 1% agarose gels. The desired fragment was purified by electroelution followed by twice phenol extraction, chloroform extraction, and ethanol precipitation. The cohesive ends of the purified Ndel fragment were filled in with dNTPs (Sigma) with the Klenow fragment of DNA Polymerase I to obtain blunt ends (Maniatis et al., 1982). Subsequently, this fragment was cloned into the Smal site of pKK223-3 and pUC19. The Ndel fragment was also cloned into the Ndel site of pET-3a. SG21166 and SG1117 were transformed according to the method of Chung et al. (1989). Other strains were transformed by the CaCI2 method (Maniatis et al., 1982). (d) Crystal toxin purification. Plasmid pOS4201 (pKK223-3 with crylA(c)) was transformed into JM103. The culture was grown at 37°C in 500 ml LB with ampicillin (50 ug/ml) in a shake flask. After 48 hours, the cells were centrifuged (10 min, 6000xg). The cell pellet was resuspended in 50 ml buffer (50 mM tris pH8, 50 mM EDTA, 15% sucrose, 200 ug/ml lysozyme), and subsequently sonicated on ice (3x5 min, large tip, Fisher Sonic Dismembrator Model 300). The suspension was centrifuged (10 min, 15000xg), and the supernatant was discarded. The pellet was washed twice with 0.5M NaCI and 2% Triton X-100, incubated in the same buffer on ice for 30 min, and washed twice with 0.5M NaCI. Finally, the crystal protein was solubilized in 8M urea and further purified through a gel filtration column (Sephacryl 300, 37 Pharmacia). The purified crystal protein toxin was examined by PAGE and silver staining. The protein concentration was determined by Coomassie Protein Assay Reagent (Pierce). (e) Protein samples preparation. Volumes of 1 ml culture were collected in microfuge tubes and cell pellets were washed twice in a TEN buffer (50mM Tris pH8,150mM NaCI, 1mM EDTA). Each cell pellet was resuspended in 500 ul 8M urea and briefly sonicated with a small tip. Afterward, 500 ul dH20 was added to each sample. The total protein concentration in each sample was determined by Coomassie Protein Assay Reagent. (f) Enzyme-linked immunosorbant assay (ELISA). Anti-CrylA(C)73 antibody was prepared as described earlier (McLinden et al., 1985). Purified CrylA(C)73 toxin solution was diluted to 4 ng/ul in PBS (40.5 ml of 0.2M Na2HP04, 9.5 ml of 0.2M NaH2P04, 0.15M NaCI in 1 liter H20, pH7.4). 100 , \ ul of the solution was added into each well of flat bottom microassay plates (lmmulon-2, Dynatech) and incubated at 4°C overnight. Plates were washed twice with 0.05% BSA in PBS. 100 ul of blocking solution (1% BSA in PBS) was added into each well and incubated at room temperature for 1 hour. Plates were washed twice with washing buffer (0.05% BSA, 0.05% tween-20 in PBS). In addition, 100 ul of each sample, as well as standard, were added in duplicate into wells of flexible assay plates (MicroTest III, Falcon). A two-fold dilution was done with washing buffer. 100 ul of diluted anti-CrylA(C)73 antibody solution (1:4000) was added into each well and incubated at 37°C for 1 hour. 100 ul of the Ag-Ab 38 mixture from each well was then added into corresponding wells of flat bottom microassay plates and incubated for another hour at 37°C. Plates were washed 5x in washing buffer, and 100 ul of diluted (1:4000 in 1% BSA) goat anti-rabbit-HRP (BioRad) was added into each well and incubated at room temperature for 1 hour. Plates were washed again 5x in washing buffer. 100 ul of TMB microwell peroxidase substrate (Kirkegaard & Perry) was added into each well. After proper color intensity was reached, 100 ul of 0.1 N HCI was added per well to stop further color development. The absorbance was immediately recorded at OD405 by EIA Reader (Model EL-307, BioTek Instrument, Inc.) (g) Site-directed mutagenesis and DNA sequencing. An oligonucleotide 5’-GGTAACATATGAAAAACAATCCG-3’ was synthesized with an Applied Biosystems Model 380B DNA synthesizer at the Biochemistry Instrumentation Center, The Ohio State University. The oligonucleotide changes the second codon of the crylA(c)73 gene from GAU to AAA and introduces a Ndel site (CATATG). Site-directed mutagenesis was carried out with a Muta-Gene kit from BioRad following the method of Kunkel (1985). DNA sequencing was accomplished with Sequenase (United States Biochemical Corporation) following the method of Sanger et al. (1977). 39 RESULTS (a) Effects of promoters on expressioncry of gene. The crylA(c) was cloned from B. thuringiensis var. kurstaki, HD 244 (McLinden et al., 1985), and have subsequently determined (Sabourin, McLinden and Dean, unpublished observations) that the DNA sequence is identical to that of B. thuringiensis var. kurstaki, HD 73 (Adang et al., 1985). By the recently revised nomenclature (Hofte and Whiteley, 1989), this gene should be called crylAfc)73. The expression of this gene from its own promoter in E. cofi (pOS1002) was measured to be about 0.24% of total cellular protein (data not shown). To examine the effects of some of the strongest reported promoters in E. coli, the crylA(c)73 gene was cloned into pKK223-3, pET-3a, and pUC19, as illustrated in Figure 2.1. In pOS4201, the Ptac promoter was placed in front of the toxin gene; in pOS9300 phi10, aT7 promoter was in front of the gene; in pOS8204, the lacZ promoter was used. In all three constructions, the protein toxin gene also carried its own promoter and terminator. pOS4201 and pOS8204 were transformed into JM103, and pOS9300 was transformed into DE3. The expression of crylA(c)73 in both JM103 and DE3 was evident from SDS-PAGE (Figure 2.2) and western blots (data not shown). At 24 and 48 hours, total protein assay and ELISA were performed, and the ratio of toxin to total protein was determined (Table 2.2). The expression of the gene in 4201 was much better than that in 9300 and 8204. The 4201 culture entered the stationary phase with about 8x10s cells/ml. After about 48 hours, the expression of crylA(c)73 reached a level of 284 mg/l, which was Figure 2.1 The Ndel fragment containing crylA(c) gene was cloned into the Smal site on pKK223-3, and Ndel site on pET-3ato generate pOS4201 and pOS9300 respectively. The Dral-BamHI fragment from pOS4201 was cloned into Hindi and BamHI sites on pUC19 following two-step subcloning to generate pOS8204. D, Dral; H, Hindi; B, BamHI; N, Ndel. 40 r r n B ( T ) POS4201 r IA(c) cry PKK223-3 ND pET-3a Figure 2.1 POS9300 r IA(c) cry r IA(c) cry * U 19 pUC Ndei N m POS8204 r IA(c) cry Figure 2.2 7% polyacrylamide gel of overproduced proteins from pOS4201 /JM103, and pOS9300/DE3. The arrow indicates the position of the CrylA(C) protein toxin. pKK223-3 and pET-3a are negative controls. HD-73 is a B. thuringiensis strain, HD-73, which carries only the crylA(c) toxin gene. 42 CO I o CO o CO o CM CO co co CM "M" O) N- CM I I X. CO H CO * o LU o a a a a a I 205,000 “ 116,0 0 0 - 94.000 - 66.000 - Figure 2.2 44 TABLE 2.2 Expression of crylA(c)73 in Escherichia coli CrylA(c)73 Strain# Plasmid construction M g /m l (percent of total protein) 24 h 48 h 4201 crylA(c), pKK223-3/Ptac 172 (39%) 284 (48%) 9300/DE-3 crylA(c), pET-3a/PT7 17 (11%) 30 (16%) 8204 crylA(c), pUC19/Plac 47(8%) 67(7%) 8208 crylA(c), pBR322 8(1% ) 9(1% ) 8201 crylA(c),pUC19/Ptac/rnB ^ T2 83 (25%) 86 (15%) 8002 crylA(c), pU C19/Ptac 21 (9%) 22 (5%) 8209 crylA(c), pBR322,AAA* 15(2%) 18(3%) 8207 cry/A('c;,pKK223-3/Ptac,AAA 163 (33%) 215 (36%) 8205 crylA(c), pUC19/P,ac,AAA* 30(9%) 35(6%) # The host strain for all recombinant plasmids is JM103 except pOS9300 whose host is DE-3 * The changed second codon of crylA(c)73 gene 45 about 50% of total cellular proteins. Longer incubation further increased the amount of toxin (data not shown). The final amount of crystal proteins produced after 48 hours is much higher than that produced by B. thuringiensis (Table 2.3). The expression of the gene in 9300 and 8204 was relatively weak. The 9300 culture was induced by IPTG. After about 30 hours, the gene expression reached its peak which was about 34 ug/ml and 15% of total cellular proteins. In all cases, the expressed toxin protein formed distinguishable crystals inside cells. In 4201, at the end of 60 hours of growth, the aggregate could occupy as much as 2/3 of the cytoplasm, as seen under the electron microscope (Figure 2.3B). The aggregates in 9300 an 8204 were noticeably smaller. (b) Effect of E. coli host strains on expression. Five E. coli strains were examined to study host-dependent gene expression under the control of the Ptac promoter (pOS4201). The level of expression of the crylA(c)73 varied among different strains (Table 2.3). JM103 was the best host among the strains tested. The potential of the Ion mutation was also examined which eliminates the La protease to enhance expression. Despite the Ion' genotype of strains SG 21166 and SG 1117, expression was very poor. It was also observed that crylA(c) 73 expression was the same with or without IPTG. (c) Effect of other genetic elements on expression. It has been noted (Minton et al., 1988) that the pUC vectors maintain a higher copy number (500-700 copies per cell) than their parental plasmid, pBR322 (150-200 copies per cell), because of a single base change in the origin of replication. Since the pKK223-3 expression vector replicates by the pBR322 46 TABLE 2.3 Expression of crylA(c)73 by Ptac/pKK223-3 in various host strains CrylA(c)73; ng/m\ (percent of total protein) Strains: 24 h 48 h B. thuringiensis HD 73 22.0 (10%) 58.0 (51%) Escherichia coli JM101 94.0 (25%) 71.0 (23%) JM103 172.0 (39%) 284.0 (48%) JM105 13.0(4%) 8.0 (3%) SG21166 47.0 (14%) 27.0 (9%) SG1117 1.0 (0.002%) 0.6 (0.002%) Figure 2.3 Electronmicrogrph of B. thuringiensis and E. coli overexpressing crylA(c). (A). Scanning electron microscopic picture showing E. coli strain 4201 (pOS4201 /JM103) expressing crylA(c)73. (B). Thin-section electron microscopic picture showing strain 4201, and lattice structure of crystals of crylA(c)73 in JM103. (C). Thin-section electron microscopic picture of B. thuringiensis strain HD-73. Arrows point to position of crystal proteins. 47 Figure 2.3 00 49 origin, it is interesting to observe the effects of increasing the copy number on expression. The ^tac promoter and crylA(c)73 gene on a BamHI fragment from pOS4201, and the rrnB TVT2 terminators from pKK223-3 on a Hindi fragment, were transfered respectively into the BamHI and Smal sites of pUC19. As shown in Table 2.2, the expression of crylA(c) on the higher copy plasmid, pOS8201, was less than that of the lower copy plasmid, pOS4201. The effect of multiple terminators on cry gene expression in E. coli was also studied. The native terminator of the crylA(c)73 gene is extremely strong (aG = -30.4). The rrnB TVT2 terminators on the pKK223-3 vector are also considered as effective terminators (Brosius, et al., 1981). The rrnB TVT2 were cloned downstream from crylA(c)73 gene terminator (pOS8201). It was found (Table 2.2) that with extra terminators, crylA(c) gene is expressed better than without them (pOS8002). The second codon has been implicated to play an important role in gene expression in E. coli (Looman et al., 1987). This effect is unrelated to codon preference. Using site-directed mutagenesis, AAA, the strongest reported second codon, was substituted for the relatively weak codon, GAT, of the crylA(c)73 gene. The mutant was verified by sequencing and Ndel restriction enzyme analysis (Figure 2.4). As shown in Table 2.2, in a native promoter background, the change of the second codon (pOS8209) can increase the level of expression by 2- to 3-fold. However, in a strong promoter background (pOS8205 and pOS8207), no detectable increase was observed over the maximum obtained level of expression in JM103 (pOS4201). Figure 2.4 DNA sequence of the starting codon and the second codon of crylA(c) gene. The second codon has been mutated from GAU to AAA. To facilitate the screening process a Ndel site (CATATG) was also created in the 5’ non-coding region. 50 51 G A T C G Figure 2.4 52 The stability of some of recombinant plasmids in E. coli strain JM103 was also studied. In the study, it was found that pBR322 and its derivatives (pKK223-3) were much more stable than pUC19 and its derivatives (Table 2.4). (d) Physical and Biological properties of crystals made E.coli.in The crystals from 4201 and 9300 were examined in thin sections with the electron microscope. Crystals formed in E. coli had the same lattice structure as native crystal proteins from B. thuringiensis (Figure 2.5). Although crystals usually were egg-shaped or convoluted, occasionally bipyramid structure could also be seen. Scanning electron microscopy revealed that sections of E. coli cells bulged out due to the overexpression of the crylA(c)73 gene (Figure 2.3A). Bioassays using crystals from E. coli against Bombyx mori, Manduca sexta, and Trichoplusia ni showed compatible toxicity as those from B. thuringiensis (data not shown, Ge et al., 1989). DISCUSSION Overexpression of foreign genes in E. coli is a powerful technique to prepare copious amounts of proteins for analysis. Two of the problems which are often encountered during overexpression are that the protein is deposited as an insoluble aggregate, or that the reported reducing environment of the E. coli cytoplasm renders the protein inactive (Goeddel, et al., 1979; Lim et al., 1989). Two E. coli expression systems on three expression vectors for B. thuringiensis crylA(c)73 were studied in this investigation. The intracellular crystals in E. coli were not random protein aggregates. Electron microscope analysis revealed that 53 TABLE 2.4 Percentage of plasmid (pUC & pBR) retention in Escherichia coli strain JM 103 Strain Plasmid Construction 24 hours 48 hours 8002 crylA(c), p UC19/Ptac 27% 11% 8201 crylA(c),pUCI9/Ptac,rrnB T^T2 95% 34% 8204 c/y/Afc) ,pUC19 / P|ac 85% 42% 8205 crylA(c),pUC\9/P\ac,AAA* 80% 11% 4201 crylA(c), pKK223-3/Ptac 75% 67% 8207 crylA(c), pKK223-3/Ptac, AAA* 92% 81% * The changed second codon of crylA(c)73 gene Figure 2.5 Hyperexpressed Cry!A(c) protein molecules form crystals in Escherichia coli. The crystals bear the same lattice structure as those formed inside Bacillus thuringiensis. (A) Crystal lattice structure from E. coli cells. (B) Crystal lattice structure from B. thuringiensis cells. 54 Figure 2.5 56 they have a lattice structure arranged in an orderly fashion similar to those in B. thuringiensis. They can be readily purified by sonication and several washings. There is very little contamination with the troublesome proteases which are ubiquitous in B. thuringiensis spp. and reduce the yield of gene product. The degree of expression of crylA(c)73 in the two systems was, however, markedly different, suggesting that the efficiency of gene expression depends on many genetic factors. Both pKK223-3 and pET-3a are derivatives of pBR322, and the tac and phi10 promoters located on these two vectors, respectively, are considered two of the strongest promoters known in E. coli. The tac promoter is a composite promoter whose -35 region is from the tryptophan operon and -10 region from lacuv5 (Amann et al., 1983). It has been used to overproduce human growth hormone, cl repressor, and human proinsulin (Reznikoff and Gold, 1986). Amann et al. (1983) reported the overproduction of cl repressor at about 30% of total cellular proteins. The phi10 is a promoter recognized only by T7 RNA polymerase. Transcription from T7 promoters is reportedly capable of rapidly saturating the translational machinery of E. coli and, as a result, a target protein can be accumulated to more than 50% of total cellular protein in a short time (Studier, F.W. and Moffatt, B.A., 1986). The expression of crylA(c)73 is also host strain-dependent. The level of expression of crylA(c)73 varied among the five E. coli strains we tested. JM103 was the best host strain followed by JM101 and JM105. At optimal conditions, more than 300 mg of crylA(c)73, which amounted to 50% of total cellular protein, 57 could be obtained from a liter of JM103 culture in shake flask. Experiments are in progress to determine the maximum yield under fermentation conditions. The expression of crylA(c)73 gene in JM101 and JM105 was significantly less than that in JM103. It is interesting to note that the only known genetic difference between JM103 and JM105 is that JM103 is a supE strain. There has been no report in the literature to support the notion that supE has a positive influence upon the expression of crylA(c)73. Even though the tac promoter is supposed to be under the control of lac repressor, the expression of crylA(c)73 in JM101, JM103, and JM105 was constitutive. Since the culture (4201) was able to grow on minimal media, F’ factor was presumably retained by the cells. The lac repressor gene somehow may have been defunctionalized on F’ factor and as a result, it made induction of expression unnecessary. Achievement of a high level of expression of a foreign gene also depends on the stability of the gene product inside the host strain. In E. coli, the Ion gene product has been associated with the degradation of abnormal proteins (Goff et al., 1984). Mutations in the Ion gene have increased the half-lives of many cloned gene products (Buell et al., 1985). Two Ion' strains were tested in the experiment. However, in both strains, the crylA(c)73 gene was not expressed as well as in JM101 or JM103. It is possible that crystals of the crylA(c)73 gene product are relatively resistant to attack by the Ion gene product as well as other E. coli proteases, and form "protein warehouses" that maintain the viability of the cell during overexpression. 58 According to Looman et al. (1987), the second codon of a gene may play an important role in gene expression in E. coli. The lacZ gene with AAA as its second codon has much higher level of gene expression in E. coli than GAU. When the second codon GAU was changed to AAA on the crylA(c)73 gene, a 2- to 3-fold increase in gene expression in E. coli was observed in a native promoter background (pOS8209). With a stronger promoter background (pOS8105, pOS8207), however, there is no increase in crylA(c)73 gene expression. It is likely that the expression level of 50% of the total cellular protein is approaching the maximum expression of the crylA(c)73 in E. coli. The plasmid stability is another critical factor affecting gene expression. Even though pBR322 derivatives have lower copy number than pUC19 derivatives, pBR322 derivatives are more stable than pUC19 derivatives. Stability of pBR322 derivatives may explain why expression of crylA(c) in pOS4201 is far better than that in pOS8201. The stability of pUC19 derivatives also seems to be influenced by the compatibility between a cloned strong promoter (Ptac) and strong downstream terminators. The successful cloning of a strong promoter requires a downstream terminator of comparable strength (Stuber and Bujard, 1981; Gentz et al., 1981). The plasmid pOS8002, which contains Ptac, crylA(c)73, is very unstable. However, when rrnB T1 ,T2 are inserted at the end of crylA(c)73, the stability of pOS8002 is greatly improved. The crylA(c)73 gene terminator may not be strong enough to stop the RNA polymerase read-through, which may interfere with plasmid replication. 59 Crystals purified from E. coli cells have comparable insecticidal activity to those from B. thuringiensis. CrylA(c)73 and CrylA(a) were tested on Manduca sexta, Trichoplusia ni, and Bombyxmori (Ge et al., 1989). Each toxin showed expected activity against these insects. Similar results were also reported by Oeda et al. (1987) and Shimizu et al. (1988) using B. thuringiensis crystal toxins from E. coli against Plutella xylostella larvae. Synthesis of toxin proteins in E. coli seems to be higher than it is in B. thuringiensis (Table 2.3). One major reason is that unlike B. thuringiensis, E. coli cells do not lyse in stationary phase. This provides E. coli cells a longer time to synthesize toxin protein. It is also found that E. coli cells carrying B. thuringiensis insecticidal crystal proteins can be readily killed by heat at 70°C without any negative effect on toxin proteins (data not shown). This indicates a promising potential for reducing the cost of production and for the production of modified B. thuringiensis insecticidal protein toxins. 60 CONCLUSION B. thuringiensis crystal protein is a biological pesticide of considerable economic importance. Therefore it is desirable to study factors influencing the expression of delta-endotoxin (cry) genes in E. co//, an organism whose expression system has been well characterized. Overexpression of cry genes in £. coli would facilitate the study of genetically engineered delta-endotoxins or those with novel host specificities. To improve cry gene expression in E. coli I tested three different E. coli promoters (Ptac, Pphii 0, P)ac), changed the second codon of cry, examined five E. coli host strains, and investigated the effect of plasmid copy number and extra terminators on expression. The level of expression was measured by ELISA. Maximal expression was obtained with the Ptac promoter and host strain JM103. With the native promoter, the change of the cry gene’s second codon from GAU to AAA increased the expression by two- to three-fold. Plasmids with the pBR322 origin gave higher levels of expression than those with the higher copy-number pUC19 origin. Extra transcription terminators improved plasmid retention in E. coli and as a result increased the level of cry gene expression. Under optimal conditions the cry gene product accounted for up to 50% of total cellular protein in 48 hour cultures. Cells in these cultures produced crystals with an identical lattice structure as those seen in B. thuringiensis. REFERENCES Adang, M. J., Staver, M. J., Rocheleau, T. A., Leighton, J., Barker, R.F. and Thompson, D.V.: Characterized full-length and truncated plasmid clones of the crystal protein of Bacillus thuringiensis subsp. kurstaki HD-73 and their toxicity to Manduca sexta. Gene 36 (1985) 289-300. Amann, E., Brosius, J. and Ptashne, M.: Vectors bearing a hybrid trp-lac promoter useful for regulated expression of cloned genes in Escherichia coli. Gene 25 (1983) 167-178. Brosius, J., Dull, T.J., Sleeter, D.D. and Noller, H.F.: Gene organization and primary structure of a ribosomal RNA operon from Escherichia coli. J. Mol. Biol. 148 (1981) 107-127. Buell, G., Schulz, M. F., Selzer, G., Chollet, A., Mowa, N.R., Semon, D., Escanez, S. and Kawashima, E.: Optimizing the expression in E.coli of a synthetic gene encoding somatomedin-C (IGF-I). Nucl. Acid. Res. 13 (1985) 1923-1938. Chung, C.T., Niemela, S.L. and Miller, R.H.: One-step preparation of competent Escherichia coli: Transformation and storage of bacterial cells in the same solution. Proc. Natl. Acad. Sci. USA 86 (1989) 2172-2175. Ge, A.Z., Shivarova, N.I. and Dean, D.H.: Location of the Bombyxmori specificity domain on a Bacillus thuringiensis delta-endotoxin protein. Proc. Natl. Acad. Sci. USA 86 (1989) 4037-4041. Gentz, R., Langner, A., Chang, A.C.Y., Cohen, S.N. and Bujard, H.: Cloning and analysis of strong promoters is made possible by the downstream placement of a RNA termination signal. Proc. Natl. Acad. Sci. USA 78 (1981) 4936-4940. Goeddel, D. V, Kleid, D. G., Bolivar, F., Heyneker, H. L., Yansura, D. G., Crea, R., Hirose, T., Kraszewski, A., Itakura, K. and Riggs, A. D.: Expression in Escherichia coli of chemically synthesized genes for human insulin. Proc. Natl. Acad. Sci. USA 76 (1979) 106-110. 61 62 Goff, S. A., Casson, L. P. and Goldberg, A. L.: Heat shock regulatory gene htpR influences rates of protein degradation and expression of the Ion gene in Escherichia coli. Proc. Natl. Acad. Sci. USA 81 (1984) 6647-6651. Hofte, H. and Whiteley, H. R.: Insecticidal crystal proteins of Bacillus thuringiensis. Microbiol. Rev. 53 (1989) 242-255. Kunkel, T.A.: Rapid and efficient site-specific mutagenesis without phenotypic selection. Proc. Natl. Acad. Sci. USA 82 (1985) 488-492. Lim, W. K., Smith-Somerville, H. E. and Hardman, J. K.: Solubilization and renaturation of overexpressed aggregates of mutant tryptophan synthase -subunits. Appl. Environ. Microbiol. 55 (1989) 1106-1111. Looman, A. C., Bodlaender, J., Comstock, L. J., Eaton, D., Jhurani, P., deBoer, H. A. and van Knippenberg, P. H.: Influence of the codon following the AUG initiation codon on the expression of a modified lacZ gene in Escherichia coli. EMBO J. 6 (1987) 2489-2492. Maniatis, T., Fritsch, E.F. and Sambrook, J.: Molecular Cloning. A Laboratory Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, 1982. McLinden, J. H., Sabourin, J. R., Clark, B. D., Gensler, D. R., Workman, W. E. and Dean, D. H.: Cloning and expression of an insecticidal k-73 type crystal protein gene from Bacillus thuringiensis var. kurstaki into Escherichia coli. Appl. Environ. Microbiol. 50 (1985) 623-628. Minton, N. P., Chambers, S. P., Prior, S. E., Cole, S. T. and Garnier, T.: Copy number and mobilization properties of pUC plasmids. Focus 10 (1988) 56-56. Oeda, K., Oshie, K., Shimizu, M., Nakamura, K., Yamamoto, H., Nakayama, I. and Ohkawa, H.: Nucleotide sequence of the insecticidal protein gene of Bacillus thuringiensis strain aizawai IPL7 and its high-level expression in Escherichia coli. Gene 53 (1987) 113-119. Reznikoff, W. and Gold, L.: Maximizing gene expression, Butterworths, Stoneham, MA, 1986. 63 Rothnie, H. and Geiser, M.: Differential Expression of the 3 delta-endotoxin Genes in Bacillus thuringiensis subsp. kurstaki HD-1. In Baker, R. and Dunn, P. (Eds.), New Directions In Biological Control. UCLA Symposia on Molecular and Cellular Biology 112. Alan R. Liss, Inc., New York, NY, 1989. Sanger, F.A., Nicklen, A. and Coulson, A.R.: DNA sequencing with chain-terminating inhibitors. Proc. Natl. Acad. Sci. USA 74 (1977) 5463-5467. Shimizu, M., Oshie, K., Nakamura, K., Takada, Y., Oeda, K. and Ohkawa, H.: Cloning and expression in Escherichia coli of the 135-kDa insecticidal protein gene from Bacillus thuringiensis subsp. aizawai IPL7. Agric. Biol. Chem. 52 (1988) 1565-1573. Stuber, D. and Bujard, H.: Organization of transcriptional signals in plasmids pBR322 and pACYC184. Proc. Natl. Acad. Sci. USA 78 (1981) 167-171. Studier, F.W. and Moffatt, B.A.: Use of bacteriophage T7 RNA polymerase to direct selective high-level expression of cloned genes. J. Mol. Biol. 189 (1986) 113-130. Yamamoto, T. and lizuka, T.: Two types of entomocidal toxins in the parasporal crystals of Bacillus thuringiensis kurstaki. Arch. Biochem. Biophys. 277 (1983) 223-241. Yamamoto, T., Ehmann, A., Gonzalez, J. M. and Carlton, B. C.: Expression of three genes coding for 135-kilodalton entomocidal protein in Bacillus thuringiensis kurstaki. Curr. Microbiol. 17 (1988) 5-12. PART TWO Localization of the Bombyx mori Specificity Domain on a Bacillus thuringiensis delta-endotoxin Protein 64 65 INTRODUCTION Bacillus thuringiensis is a microorganism of considerable importance. The economic value of B. thuringiensis has been attributed to its toxicity against insects in agriculture, forestry, and health care. B. thuringiensis has a very narrow insecticidal spectrum (Dulmage and Cooperator, 1981; Yamamoto et al., 1983; Jaquet et al., 1987), and is not toxic toward mammalians and other animals. Because of its biodegradable nature and proven safe usage in the environment, B. thuringiensis has a great potential to replace many chemical pesticides. The biochemical basis of this insecticides is the crystal proteins (delta-endotoxins) produced by B. thuringiensis during its growth. Most lepidopteran-specific strains of B. thuringiensis synthesize a 135 kDa protein that crystallize inside bacteria and requires proteolytic cleavage to become active against insects. To date, more than 40 B. thuringiensis crystal protein genes have been cloned and sequenced. Some of these genes are similar or identical. Nevertheless, 14 unique delta-endotoxin genes can be identified (Hofte and Whiteley, 1989). Many strains of B. thuringiensis can carry more than one of these crystal protein genes (Kronstad et al., 1983; Yamamoto and lizuka, 1983; Yamamoto et al., 1988). B. thuringiensis crystal proteins enters its hosts through their digestive system. As a result, the solubilization of crystals, the activation of the crystal protein, the susceptibility of insects, as well as the nature of the toxin jointly contribute to the specificity of the delta-endotoxin. Wilcox et al. (1989) were the first to recognize 66 the role of a cloned crystal protein gene (cry) in the specific action against a particular insect by noting that a strain lacking a "5.3 kb" gene (crylA(b)) showed less activity against Spodoptera exigua. Kondo et al. (1987) demonstrated the specificity of different cloned delta-endotoxin genes (crylA(a) and crylA(b)) from B. thuringiensis HD-1. In qualitative bioassays with Escherichia coli clones of these genes, they observed a significant difference in activity against Bombyx mori (Silkworm). Hofte et al. (1988) have investigated five lepidopteran-active genes which differ in specificity against several insects. Analysis of amino acid sequences from different crystal proteins derived from corresponding DNA sequences reveals that variations of specificities of different delta-endotoxins can be traced to the difference of amino acid sequences on these proteins (Geiser et al., 1986). Among three best studied crystal proteins, crylA(a), cryiA(b) and crylA(c), the majority of amino acid differences is located somewhere between amino acid 283 to 612 (Figure 3.1). An alignment of proteins from each group (cryl, cryll, crylll and cryVI, Hofte and Whiteley, 1989), however, shows five conserved regions on these crystal proteins (Figure 3.2). The functional significance of these conserved regions is yet to be revealed. The difference in amino acids among distinct crystal proteins seems to contribute to the differences of tertiary structure and, as a result, specificity of these proteins. Hofte et al. (1988) and Huber-Lukac (1983 and 1986) demonstrated that monoclonal antibodies were able to differentiate crystal proteins from different groups and those within the same group. This indicates that there is an unambiguous correlation between antigenic structure and insect Figure 3.1 Comparison of amino acid sequences of CrylA(a), CrylA(b), and CrylA(c). The numbering system is relative to the CrylA(c)73 protein. The letters of the sequence are in the single letter amino acid codes: A, Ala; C, Cys; D, Asp; E, Glu; F, Phe; G, Gly; H, His; I, lie; K, Lys; L, Leu; M, Met; N, Asn; P, Pro; Q, Gin; R, Arg; S, Ser; T, Thr; V, Val; W, Trp; Y, Tyr. The dark boxes mark amino acids which are different and non-conserved, and the gray boxes mark amino acids which are dissimilar but conserved at the same position comparing CrylA(a), CrylA(b), and CrylA(c). Some restriction sites on the corresponding DNA sequence are shown here as landmark. < > indicates restriction sites would be added, and () indicates restrictions sites would be removed by site-specific mutagenesis. 67 COMPARISON OF AMINO ACIO SEOUENCES OF LEPIDOPTERA 6-ENDOTOXINS FROM crvlA (o). ervIA(b) AND crvlA fc) GENES N s il 101 20 30 "lo so 70 100 120 130 140 150 I A l MONNPNlNECIPYNCLSPEEVEVLCGERIETGVTPlDISlSlTOFLLSEFVPGAGFVLGLVDIIUGIFGPSOUDAFpVQIEQLINORlEEFARNQAISRLEGlSNLYQlYAESFREVEAOPTNPALREEMRlQFNOMNSAt.TTAIPLlAV 8171SHONNPNINECIPYNCLSNPEVEVLGG£RIETGYTPiD!SLSLTQFLLSEFVPGAGFVLGLVDIIUGIFGPSQUDAFLVOIEQl.lNORIEEFARNQAISRLEGLSNLYOIYAESFREUEAOPINPAlREEMRIQFNOMNSALTTAIPLFAV C73 MOMNPN«NECIPYNa.SPEEVEVlGGERIETGYTPlDISlSlTQFllSEFVPGAGFVlGlVDIlUGIFGPSQUOAFlVQIEQUNQRlEEFARNQAlSRlEGlSNUQirAESFREUEAOPTNPALRE£HRlOFNDHNSALTTAlPlFAV I I I N s il C la l. EcoRI A s u lI , 160 180 190 200 210 220 230 240 25012 6 0 ' 270 280 A1 GNYQVPLLSVYVQAANlHl$Vl>RDVSVFGQRWGFDAATlKSRYNDlTRLIGNYTDYAVRVYKTGL£RWGF>OSROUVRYNQFRRELTLTVLDiVAt.FsNYDSRRYPIRTVSQI.TREIYTNPVLENfOGSFRGmAQrlEqnlRqPHlKOIL B17150NYQVPLLSVYVQAANt.HLSVLROVSVFGORUGFDAAYINSRYNDLTRLIGNYTOhAVRUYNTGLERVUGPDSROUiRrNQFRRElTlTVLOIVsLFPNrDSRtYPIRTVSaLTREIYINPVLENFOGSFRGlAollEgilRlpHLHOIL C73 ONYQVPLLSVYVOAANLHLSVLROVSVFGORUGFDAAYINSRYNOLTRLIGNYIDYAVRUYNIGLERVUGPOSROWRYNOFRRELTLTVLDIVALFPNYOSRRYPIRTVSOUREIYTNPVLENFOGSFRGSAOGIERSIRSPHLKOIL I I A s u ll. EcoRI. Haelll EcoRI. « P s tI> B$tKI (ASUll) A1 PtFSUqHRSAEFNNIlpSsqlTOlPltKstnlgsGtSWkGPGFTGGOilRrtSpGq-IstlRvnltaPl ...... SqRYRVRlRYAStTnlqfhtsidgrplnqgnfsATnsSgsNLOSgsFrtvgfttpFnfsngSSvftlsahvfnSG-neVyl B1715PMFSU1HRSAEFNN1IpSsqlTOIPltlCstnLgsGTSWkGPGFTGGOiIRrtSpGq’ Is tlR v n ita p l ...... SqRYRVRiRYAStYntqfhtsidgrplnggnfsATnsSgsMLQSgsFrtvgfttpFnfsngSSvftlsahvfnSG-neVyl ■ I ■ _ i n ni C73 PAFSUlHRSAEFHAllASDSnQlPAYKGNFLFKG'SVISGPGFTGGOLVRLNSSGXNION'RGTIEVPiHFPSFSFRYRVRVRYASVFPlHLHVMUGNSSIFSMFVPAFAFSlDMLQSSDFGYFESANAF- -TSSLGNIVGVRMFSGFAGVIt ■ ■ I I . i in ti i i i ■ ...... P s t l, A l DRiEFvPaevTfEAEYdLERAOKAVNeLFTSsNQiGLKTcfVTOYHIDQVSNLVecLSOEFCLOEtqELSEKVKHAKRLSOERMLLOOpNFrglNRQldRGUrGSTdlTIOGGODVFICENYVTLlGTFDECYPTYLYOXIDESKLXAyTRY B1715DRiEFvPacvTfEAEYdlERAQKAVNeLFTSsNQiGlKTdVTDYHIDQVSNLVecLSQEFClDEKkELSEKVKHAICRLSOERMLLOOpNFrglNRQldRGUrGSTd[Ti mi i ■ i i i i i I0GG00VFKENYVTLlGTFDECYlTYLYQK10ESKLKAyTRY i C73 ORFEFIPVTATLEAEYNLERAOKAVNALFTSTNQLGLKTNVTOYHIOOVSNLVTYLSOEFCl.DEKRELSEICVKHAtCRLSOERMLLOOSNFICDtNRQPERGUGGSTGITIQGGOOVFKEKYVTLSGTFOECYPTYLYOKIOESXlXAFTRY I EH I I { HI I I I I I Xhol/Aval AccI Bell 760 770 780 790 800 810 820 830 850 870 890 A l QlRGYlEDSOOlEIYLIRYNAKHETVNVPGYGSLWPLSAQSPIGKCGEPNRCAPHlEUNPOlDCSCRDGEKCAHHSHHFSlOIDVGCTDlNEDtCVVVIFKIKTOOGF'ARLGNlEFlEEKPLVGEALARVKRAEttVRDICREKLEUETNI B17150lRGYlE0$O0lElYllRYNAKHETVNVPGTGSLUPlSApSPIGJCC...... AHHSHHFSLD JDVGCTDINEDLGWVIFKIKTQOGHARLGNIEFLEEKPLVGEALARVKRAEKKURDKREKLEVETNI C73 QLRGYIEOSOOIEIYLIRYNAXHETVNVPGTGSLUPLSAQSPIGKCGEPNRCAPHLEUNPOLDCSCROGEKCAHHSKHFSLOIDVGCTOLNEOLGWVIFKDCTOOGHARLGNLEFLEEKPLVGEALARVKRAEICXUROKREKLEUETNI 910 920 930 940 9S0 960 970 980 990 1000 1010 1020 1030 1040 1050 A l mEAKESVOAlFVNSQYOQLOADtNlAMIHAADKRVHStREAVLPELSViPGVNAAlFEELEGRIFTAFSLYDARNVlKKGOFNNGLSCUNVXGHVDVEEQNNQRSVLVlPEWEAEVSQEVRVCPGRGYlLRVTAYKEGYGEGCVTIHE B1715VYKEAKESVDAlFVNSQYOrlQADTHlAHlHAADKRVHStREAYLPELSVlPCVKAAlFEElEGRlFTAFSLYDARHVlKNGDFKHGlSCVUVKGKYOVEEONKQR$VLVVPEWEAEVSQEYRVCPGRGYIl.RVYAYKEGYGEGCVTlHE C73 VnrKEAXESVDAlFVNSOYOOLQADTNIAMINAADICRVHSIREAYLPELSVIPGVNAAIFEELEGRIFTAFSLYOARNVItCNGOFNNGLSCUNVKGKVDVEEONNORSVLWPEUEAEVSOEVRVCPGRGYILRVTAYICEGYGEGCVTIHE A1 IENNTDELKFSNCVEEEIYPNNTVTCNOYTVNQEEYGGAYTSRNRGYNEA------PSVPADYASVYEEKSYtDGRRENPCEFNRGYROYTPLPVGYVTKElEYFPETDKVUlElGETEGTFIVDSVEUlMEE B1715IENHTO£t.KFSNCVEEEvYPNNTV7CNOYTatOEEYeGtYTSRNRGYdgAyesnsSVPAOYASaYEEICaYtOGRRdNPCEsNRGYgOYTPLPaGYVTKELEYFPETOI(VUIEIGETEGTFIVOSVELLLHEE C73 1ENMTDEUFSKCVEEEIYPNNTVTCNDYTVNQEEYGGAYTSRNRGYNEA------PSVPADYASVYEEKSYROGRRENPCEFNRGYRDYTPLPVGYVTICELETFPETOXVWIEIGETEGTFIVDSVELLLHEE Figure 3.1 Figure 3.2 Five conserved amino acid sequences identified by an alignment of delta-endotoxins from each group. Identical and similar amino acids are boxed with gray background. 69 Block 1 crylA (b) V Q U PL L S U V U Q fi R N LHLSU L R 0 crylC F E U P L LS U VR 0 R fl NLHLR1 L R 0 cry 32 A V N 1 L U L s S V R 0 R RHLH L T U LN Q cry32B V ELLLI p i V R Q U RH F N LLL1 R 0 cryHA V E U L F L TT Y fl 0 RR N T H LFLL K D cry 31A V Q L L L L P L F fl Q R fl N H H L SF 1 K D Block 2 crylA (b) U 1 R Y N g F R R E L T LT U L 0 I U SL F P N Y D S R T Y P 1 R T U S g l ------T R E 1 Y crylC U 1 TY N R L R R D LTL T U L D 1 R R F F P N Y D N R R Y P i g p U G g l ------T R E V Y cry H A U NT Y N T Y R T Kn T TR U L D L U R L F P H YD U GK Y P 1 G u g sEL ------T R E 1 Y cryH B U 1 TFND Y K REH T 1 QU L D 1 L RL F R S Y 0 P Rfl Y P fl D K 1 D N T K L S K T E F T R E 1 Y cryHA UUNFNR Y R RE tl T L T U L 0 L 1 RL F P LY 0 UR L V P K E UK T EL ------T R D U L cryniA [FjH [U]Tl [L]D F R T Y n F L N U [ F ] E [ g U S 1 U S L F - - - K Y g S [L ]n U S S G fl N [L ]Y fl S G S G P G G Block 3 crylA (b) U 1 H R S RE F N N 1 P S s g 1 T Q l P L T K S TN L GS G T S 1 U K G P G F T G G 0 1 L crylC U T H R SR T LTH W<. D P E R 1 H Q 1 P L U K GF R U U G GT S U 1 T G P G F T G G D 1 L cry 3 2 A U T HSS U D PK H T Y T HL T T Q I P R U KR N S L G T R SK U u g G P G H T G G D L 1 cry 3 2 B U T H R 1 U 0 P HN Q Y T 0 R I T Q U PR U K S H F L N R T R K U 1 K G P GHT G G 0 L U cryHA U TH K S U D F F N MD S K K I T 0 L P L U < RY < L 0 s G R S U u R G PR FT GG D 1 I Block 4 Block 5 crylA (b) S g R Y R U R V R Y R ST crylA (b) U Y 0 R m F UP R crylC T g R Y RL R F R Y R S S crylC L Y itD W K ! sfE; 1 1 L R cryH A Q g s y F 1 R 1 R Y R SN cryH A U L I 0 K m F L P 1 cry 32 B T R S Y G L R L R YRRH cryH B U 1 0 R 1 1 P 1 cry3IIA s g tc v R RR 1 H Y RST cryHA U Y D K MB F 1 P U o Figure 3.2 71 specificity. The variable region of a crystal protein is speculated, on a reasonable ground, to be responsible for the specificity of a protein (Geiser et al., 1986; Andrews et al., 1987). Geiser et al.(1986) suggested that the hypervariable regions on CrylA(a), CrylA(b), and CrylA(c) might be responsible for the biological activity of delta-endotoxin proteins. However no direct evidence has ever been presented to support this view. Understanding the biochemical and genetic mechanisms for insecticidal action and specificity of insecticidal proteins would be richly rewarded by revealing a means of synthesizing new biorational insecticides. In this part the insect specificities of crylA(a) and crylA(c) gene products against the insects B. mori, Trichoplusia ni, and Manduca sexta were demonstrated. By exchanging regions of these two genes and bioassays of these hybrid proteins, the B. mori specificity domain on crylA(a) gene product was located. 72 MATERIALS AND METHODS Origins of delta-endotoxin genes, hosts and vectors E. coli containing the crylA(a)'\ gene, ES1 (pES1) (Schnepf and Whiteley, 1981), was obtained from the American Type Culture Collection as ATCC 31995. The isolation of the crylA(c)73 gene from B. thuringiensis strain HD-244 has been previously reported (McLinden et al., 1985) as E. coli 87-22 (pOS1002). This gene has been sequenced (Sabourin, McLinden, and Dean, unpublished) and found to have exactly the same sequence as the crylA(c)73 gene from B. thuringiensis var. kurstaki, HD-73 (Adang et al., 1985). The E. coli host, JM103, was obtained from Pharmacia, and cloning vectors M13mp18 (BRL), pUC19 and pKK223-3 (Pharmacia) were obtained from the suppliers. Site-directed mutagenesis and other molecular genetic techniques Oligonucleotides were synthesized with an Applied Biosystems Model 380 B DNA Synthesizer in the Biochemistry instrumentation Center of the Department of Biochemistry, The Ohio State University. Single primer site-directed mutagenesis followed the method of Gillam and Smith (1979) and selection of mutants was by the method of Kunkel (1985) and screening by restriction enzyme analysis. Other molecular genetic techniques are as described in Maniatis, Fritsch, and Sambrook (1982),and Perbal (1984). 73 Construction of site-directed mutants, substitution mutants, and overexpression mutants Figure 3.3 illustrates the overall strategy used in this study. Steps in the construction were given as follows: (A) To overexpress crylA(a), crylA(c), and other hybrid proteins as a result of site-directed mutations and reciprocal switching mutations, the crylA(a) 1 and crylA(c)73 genes were excised from their original plasmids with Ndel, and purified by electroelution. The cohesive ends of the fragments were filled-in with dNTPs and the Klenow fragment. Subsequently they were cloned into the Sma I site of a modified pKK223-3, which had its EcoRI site removed (by filling-in with dNTPs and the Klenow fragment). (B) To prepare for site-directed mutagenesis, the Hincll-Kpnl fragments of crylA(a) and crylA(c) which contained the 5’ portion of each gene was cloned into M13mp18. (C) The following oligonucleotides were synthesized and used in site-directed mutagenesis to eliminate and/or add desired restriction enzyme sites: 1. 5’-GTAAGAATACGCTAC-3’ was used to remove the third EcoRI (E) site from crylA(a) 1, creating cryfA(a) 1 E3(-). This also converted the amino acid lie 523 of crylA(a) 1 to Val 523 found in crylA(c) 73. 2. 5’-GAATTTGCTAGG-3’ was used to remove the first EcoRI site from the previously constructed mutant, creating crylA(a) 1 E^-), E3(-). This oligonucleotide was also used to remove the first EcoRI site from crylA(c)73, creating crylA(c)73 E1 (-). The removal of the first EcoRI site did not alter the amino acids. 3. 5’-GTAACCCTCGAGGCA-3’ was used to add an Xhol (X) site to the crylA(a) 1 E3(-) mutant creating crylA(a)1 X(+), E3(-), and crylA(a) 1 X(+), Figure 3.3 Construction of substitution mutations and expression plasmid of cry genes. C, C/ai; E, EcoRI; He, Hinc\l, K, KpnV, Nd, A/del; Ns, Nsil, RV, EcoRV; S, SstI; X, Xho\. < E> indicates that EcoRI site was removed. 74 8-endotoxin genes, crylA(a) or cry 1A(c) Nd Ns E E E (x) K Ns Hc"“(m13mp18 (Nd) 7~(Nd) pUC8 r-K (Nd)^ RI(-) T-(Nd) c s RV x (Nd)JI\PK^2I23-3 f-(Nd) Figure 3.3 76 E1 (-), Eg(-). The addition of an Xhol site changed amino acid 610 from Phe to Leu, which was not a conservative change, but it was shown below that this change alone did not alter insect toxicity. (D) Modified crylA(a) and crylA(c) genes were put into pKK223-3 EcoRI(-), in preparation for bioassay, by replacing the Nsil-Kpnl portions of the parental genes, cloned in pKK223-3 EcoRI(-), with Nsil-Kpnl portions of the site-directed mutant genes. (E) Exchange of the constant regions flanked by EcoRI., and EcoRI2 on crylA(a) 1 and crylA(c)73 genes was done by exchanging the EcoRI.,-EcoRI2 fragments between crylA(a) 1 E3(-) and crylA(c)73. (F) The EcoRI2-Kpnl fragments from modified crylA(a) and crylA(c) were subcloned into the equivalent sites in the polylinker on pUC19. (G) Small blocks of DNA sequence of the variable region flanked by EcoRI2 and Xhol were exchanged between the crylA(a) 1 genes and the crylA(c)73 gene by the scheme shown in Figure 3.4A and 3.4B. (H) Hybrid DNA sequences were transferred back into parental genes on expression vectors by cloning the EcoRI2-Xhol portions of the DNA fragment on pUC19 into corresponding crylA(a) 1 (E.,, E3(-), X(+)) or crylA(c)73 (E,(-)) genes previously cloned in pKK223-3 EcoRI(-), yielding the hybrid mutants shown in Figure 3.4A and 3.4B. Purification of overexpressed crystal proteins The hybrid crystal proteins were expressed to the level of approximately 35-50% of total cellular protein in JM103 using the modified expression vector pKK223-3 as judged by scanning total protein displayed by PAGE with a laser scanner. Crystal protein was deposited as large irregular or diamond-shaped crystals appearing similar in lattice arrangement as those made in S. 77 thuringiensis. Crystal proteins were purified from overexpressing cells after 48 hours of growth in LB with 50 ug/ml ampicillin. Cultures were centrifuged at 6000xg for 10 minutes. Then cell pellet was resuspended in 50 ml suspension buffer (lizuka et al., 1983), and subsequently was sonicated on ice (large tip, Fisher Sonic dismembrator Model 300). After several burst, cell suspension was checked under a phase contrast microscope. Then the suspension was centrifuged (Green et al., 1989), and the supernatant was discarded. The pellet was washed at least twice with 0.5 M NaCI and 2% Triton X-100, incubated in the same buffer on ice for 30 minutes, and finally washed twice with 0.5 M NaCI. The purified crystals were solubilized in 8 M urea and diluted to 4 M with H20. The purity of crystal protein preparation was examined by PAGE and silver staining. Usually ICP amounts to more than 90% of the total protein in each preparation. The total protein concentration was determined by Coomassie Protein Assay Reagent (Pierce) with bovine serum albumin (Sigma) as a standard. Bioassays Where LD50 values are reported, 10 to 20 insects were used for each point on the probit plot, and 4 to 5 points were used in determination of the slope. LD50 values are reported as ug protein applied to artificial diet or leaf disks. LD50 values and 95% fiducial limits were calculated with the PROBIT.SAS program. Controls consisted of E. coli containing only the expression vector, pKK223-3. Bioassays of both M. sexta and T. ni were conducted as described in McLinden, et al. (1985). Bioassays of B. mori were conducted by pipetting 50 ul of diluted ICP on a mulberry leaf disk (1.5 cm diameter) and adding 1 larvae per cup. Insects were inspected in 24 hours. Ten larvae were used per concentration of 78 delta-endotoxin in reporting insect mortality. Each data point (percentage of mortality) was an average of four or five tries. M. sexta eggs were obtained from Michael Jackson, USDA South Atlantic Area Tobacco Research Laboratory, Oxford, NC. T. ni eggs were obtained from Lillian Moug, USDA Western Cotton Research laboratory, Phoenix, AZ. B. mori eggs were obtained from Dr. Y. Tanada (Univ. California, Riverside) and/or purchased from Carolina Biologicals and reared on mulberry leaves. RESULTS Specificity of delta-endotoxins In order to determine if insecticidal crystal proteins (ICP) exert selective toxic activity toward particular insects, individual genes were cloned, overexpressed and partially purified. Quantitative data was sought on the insecticidal specificity of crylA(a) and crylA(c) ICPs. Table 3.1 shows the toxicity of purified ICPs from OSU 4101 and OSU 4201 which overexpress the wild type crylA(a) 1 and crylA(c)73 genes respectively. The CrylA(a) toxin shows greater activity toward B. mori, while the CrylA(c) toxin shows greater activity toward T. ni. Both toxins show equally strong activity against M. sexta. The clearest difference between the activities of the two ICPs was on B. mori and this insect was chosen for further work. 79 TABLE 3.1 Specificity of Insecticidal Crystal Proteins On Selected Insects LD50 microgram Lepidoptera Insects CrylA(a) CrylA(c) Manduca sexta 0.077 (0.024-0.135)a 0.072 (0.062-0.082) Trichoplusia ni 2.88 (1.34-5.67) 0.32(0.12-0.81) Bombyx mori 0.37 (0.0007-0.792) >150b a Numbers in parentheses are upper and lower confidence limits b Insects showed no signs of toxicity at this amount 80 Location of the specificity domain forS. mori Effort was made to locate the regions of the delta-endotoxin genes which are responsible for differences in larvicidal specificity by exchanging segments of the crylA(a)1 gene with corresponding segments of crylA(c)73. The resultant substitution mutants are illustrated in Figure 3.4A. Line 1 in both Figure 3.4A and Figure 3.4B illustrate the native crylA(a)l and crylA(c)73 genes, respectively, cloned into the modified expression vector, pKK223-3 (EcoRI'). Site-directed mutant forms of these genes, wherein certain restriction enzyme sites were removed or added for purposes of constructing hybrid genes, are illustrated in line 2 and 3 of Figure 3.4A and line 2 of Figure 3.4B. Exchange of the "conserved region" (Line 4 of Figure 3.4A; line 3 of Figure 3.4B), resulted in no measurable effect on the toxicity. Exchange of the complete "hypervariable region" (Figure 3.4A, line 5; and Figure 3.4B, line 4) resulted in complete transfer of toxic activity. Subdivision of the "hypervariable region" into an amino-terminal half (EcoRI2-Sstl, yielding OSU 4110 and OSU 4210) and a carboxy-terminal half (Sstl-Xhol, yielding OSU 4109 and OSU 4209) yielded interesting results. Transfer of the amino-terminal half of the hypervariable from crylA(a) 1 to crylA(c)73 transferred virtually all of the toxic activity (Line 9 of Figure 3.4B); while the reciprocal exchange resulted in complete loss of activity (Line 10 of Figure 3.4A). This region is therefore referred to as the B. mori "specificity domain" (EcoRI2-Sstl). Exchange of the carboxy-terminal half of the hypervariable region, in toto, had no effect on the toxicity of either gene product (Line 9 of Figure 3.4A; line 8 of Figure 3.4B). Figure 3.4 Hybrid mutants cloned into the expression vector pKK223-3 (modified to remove its EcoRI site). The diagram at the top illustrates the toxicity region of the protein (Tox. Domain) and the conserved (Conser.) or and hypervariable (Var.) domains. (A) Of OSU 4100 series, line: 1, OSU 4101, the original crylA(a) gene; 2, OSU 4102, site-directed mutant of crylA(a) gene with the EcoRI, and EcoRIg sites removed and a Xhol site added to facilitate exchanges shown in lines 5-12; 3, OSU 4103, site-directed mutant with the EcoRIg site removed to facilitate the exchange shown in line 4; 4, OSU 4104, EcoRI fragment from pOS4103 exchanged with corresponding fragments from crylA(c), 5-12, OSU 4105-4112, different DNA blocks from pOS 4102 exchanged with corresponding blocks from crylA(c). (B) Of OSU 4200 series, line: 1, OSU 4201, the original crylA(c) gene; 2, OSU 4202, site-directed mutant with EcoRI., site removed to ease exchanges shown in lines 4-11; 3, OSU 4103, EcoRI fragment from pOS4201 exchanged with corresponding fragment of crylA(a); 4-11, OSU 4205-4212, DNA fragments from pOS4202 exchanged with corresponding fragments of crylA(a). MORT, percent of insect mortality with a dose of 4ug protein applied to food disks (mean of three assays, 10-20 insects per assay). 81 k-Tox. Domain-*! ______I Conser. I Var. (Sm/Nd) CSRV (Nd/Sm) I Ys fl f2 If3 f I Ns EgCSRVX K 2 _ i...... Ns E1 E2CSRV K Ns E! Eg CSRV K 4 — 1 o 0 Ns E?CSRV X K 1 111 1 5 im m k ...... - 0 Ns Eg CSRV X K ...... 1...... 6 y 1 1 m m L " — 6 8 Ns Eg CSRV X K ..... INI .... J... 7r __ 1 1■ysy.-Vsy.-yss.-yssM-:-y.-: ^ Ns Eg CSRV X K ____ ^ m m m m 8 1 11 „ -- Q Ns e2c s r v X K __ 00 Ns Eg CSRV X K _ 1 .... n j j ....j...... J... 10 , _ o Ns Eg CSRV X K _ 1 i y, i 1 11 -37 Ns Eg CSRV X K ____ 1 ...... yy... j.... J 12 ...... • ' - 0 Figure 3.4 83 k-Tox. Domain-* B 1 Conser. 1 Var. Mort (Sm/Nd) CS RV (Nd/Sm) I Ns PX Ns Ep CS RV X K Ns E, E2CSRVX K EgCSRVX K EpCS RV X K EgCSRVX K EgCSRVX K Eg CS RV X K EgCSRVX K EgCSRVX K EgCSRV X K Figure 3.4 84 Further exchanges subdividing the specificity domain (EcoRI2-Sstl) into subsets (EcoRI2-Clal, yielding OSU 4106 and OSU 4206; and Clal-Sstl, yielding OSU 4111 and OSU 4211) resulted in reduction of toxicity to B. mori with gene products from the crylA(a) 1 background (line 6 and 11 of Figure 3.4A), while transferring no activity to the crylA(c)73 background (line 5 and 10 of Figure 3.4B). Subdivision of the carboxy-terminal portion of the hypervariable region (Sstl-Xhol) had a drastic effect on toxicity to B. mori. No toxicity was transferred or retained in either gene background when this region was split (line 7,8, and 12 of Figure 3.4A; line 6,7, and 11 of Figure 3.4B), even if the specificity domain from crylA(a) 1 remained intact (line 11 of Figure 3.4B; line 8 of Figure 3.4A). DISCUSSION Since late 1950s B. thuringiensis has been used world-wide as an alternative to some petrochemical based insecticides. After more than 30 years field application it has been demonstrated as an environmentally safe and effective biopesticide, but its narrow insecticidal specificity and short-lived effectiveness may have limited its broader usage. Thus, understanding the mechanism and the insecticidal nature of B. thuringiensis delta-endotoxin become increasingly urgent. Only after we understand the mode of action of the delta-endotoxin can we overcome these limitations. 85 The insecticidal specificity of B. thuringiensis has been known to exist for years (Dulmage, 1981). In some cases it has been shown that the specific activity could be broadly different, affecting different orders of insects such as B. thuringiensis var. israelensis which affects Diptera, and B. thuringiensis var. tenebrionis which affects Coleoptera. However many past reports about the specificity of B. thuringiensis can hardly be correlated to individual delta-endotoxins as most of these discussions failed to recognize the heterogeneous pool of delta-endotoxin that could exist in a single strain. It was not until recently that the specificity of pure ICPs have been quantitatively demonstrated for different species of lepidoptera. In the above experiment the specificities of crylA(a) and crylA(c) have been demonstrated (Table 3.1) against M. sexta, T. ni, and B. mori. Despite more than 80% sequence homology, crylA(a) and crylA(c) have distinct activity against B. mori. CrylA(a) is about 400 times more active than CrylA(c). It may be assumed that differences of specificity between these two proteins are due to amino acid differences in proteins. Therefore, it is reasonable to suggest that the specificity domain is less likely to lie in the conserved regions of a delta-endotoxin gene. CrylA(a) and CrylA(c) proteins have 1176 and 1178 amino acids respectively. After enzymatic cleavage at amino acid 28 (Arg) and 623 (Lys). The activated toxin has a molecular size of about 65 kDa. Amino acids before glutamine 331 on CrylA(a) and CrylA(c) consist of the conserved domain of the toxin. Within this domain amino Acids 29 (lie) to 90 (Glu) between the two proteins are identical. The sequence from amino acids 91 (Phe) to 332 (Glu) are almost identical except 11 amino acids differences. The 86 DNA sequence corresponding to this region (AA91 to AA331) is flanked by EcoRI sites. To prove that the conserved domain does not contribute to toxin specificity a switch of the region between two EcoRI sites is performed. Through bioassays of genetically engineered hybrid toxins (4104 and 4204) it is shown that despite 11 amino acids differences, indeed the conserved domain does not measurably affect the specificity of the CrylA(a) toxin against B. mori. Amino acids 332 (Phe) to 623 (Lys) define the variable domain on CrylA(a) and CrylA(c) toxins. There are total 131 different amino acids scattering in the variable domain. A hybrid protein approach is used to localize the B. mori specificity domain on CrylA(a) toxin. It is evident from bioassays of hybrid proteins 4210 and 4110 that amino acids 332 (Phe) to 450 (Ala) on CrylA(a) toxin are involved in the specificity against B. mori. Comparing amino acid sequences between CrylA(a) and CrylA(c) in this region there are 52 amino acid differences, 24 of which are considered non-conserved changes, 23 conserved and 5 mis-alignments, according to the Dayhoff, et al. categories of conservative amino acids (1987). The predicted secondary structure of S. mori specificity domain of CrylA(a)1 and the corresponding sequences of CrylA(c)73 reveals a potential structural basis for the specificity differences. The algorithms of Kyte and Doolittle (1982) and Chou and Fasman (1978) have been used and it was observed from the computer based analysis that the CrylA(a)1 protein has two hydrophilic peaks (Figure 3.5a) which are superimposed on two beta-sheets (Figure 3.5b) at Figure 3.5 Predicted protein secondary structure of the Bombyx mori specificity domain of crylA(a) (a-c) and crylA(c) (d-f). Numbering of this domain, 0-120, corresponds to amino acids 332-452 of CrylA(a) protein, (a and d) Hydrophobicity plots (Kyte and Doolittle, 1982). (b and e) Predicted amphiphilic beta-sheet (Chou and Fasman, 1978). (c and f) Predicted amphiphilic alpha-helical structure (Chou and Fasman, 1978). Arrows marks hydrophilic peaks (a and d) and beta-sheet (b and e). In a and d, the hydrophobic value of each amino acid is plotted. 87 88 100 TV I T“ V 1 T"V | r- i i i i I i i i i I i i i i i i i i i I 0 . 8 - 0 .8 - 0 .6 - 1 0.6 - i 1 0.4- 0.4- 0 .2 - 0 . 2 - 0 .0 - 0.0- ■ i i i i i i i 11 i ■ i i i i i i i i I i I I I I I I I I T " I I I T I I 7 1 I I I I I 0.8 - 0 .8 - f 0 .6 - 0 .6 - 0.4- 0.4- 0. 2- 0.2 0.0 - 0.0 H ■ i i » i i i 11 i i I I I I 1 I I I I I I 100 100 Figure 3.5 89 positions around 70 and 90 in the domain (residues 403 and 423 on the toxin). The CrylA(c)73 has only one hydrophilic peak (Figure 3.5d) which is superimposed on a single beta-sheet about position 90 of the corresponding domain (Figure 3.5e). Both proteins are predicted to have alpha-helix structures around position 30. The presence of hydrophilic beta-sheets suggests surface structures that could interact with midgut cell receptors. A recent report revealed that ICP specificity is correlated with high affinity cell binding (Hofte et al., 1986). This would suggest that the region defined as the B. mori specificity domain is involved in receptor binding. There is an increasing evidence that B. thuringiensis delta-endotoxin consists of two structural domains, one for receptor binding and the other for membrane-insertion (Hofmann et al., 1988; Haider and Ellar, 1989; Choma et al., 1990b; Choma et al., 1990a; Convents et al., 1990; Van Rie et al., 1990). The predicted secondary structure and physical chemistry analysis indicate that the N-terminal portion of the toxin is rich in alpha-helix structure which have distinct trans-membrane characteristics. The C-terminal portion, on the other hand, has alternating beta-sheet structure which is uniquely hydrophilic. Protease digestion also supports a two-domain model. Chymotrypsin, thermolysin, and pronase are able to cleave the native toxin into two fragments of 30 and 36 kDa respectively (Convents et al., 1990). Papain, on the other hand, is able to cleave the toxin in a denaturing condition where the toxin begins unfolding (Choma et al., 1990b). These evidences suggest that the toxin is structured as two distinct domains connected by an interdomain region. As the toxin unfolds this interdomain region is exposed and subsequently cleaved. Analysis of the N-terminal portions of the 90 two fragments resulting from papain digestion shows that the papain cleavage site is at glycine 327. This gives a 34.5 kDa N-terminal fragment and a 32.3 kDa C-terminal fragment. Interestingly, the papain cleavage site is almost superimposed on the EcoRI2 site of the corresponding DNA sequence which is mentioned above as a landmark dividing the conserved and variable domains on the toxin (Figure 3.1). The binding of delta-endotoxin to receptors on insect midgut epithelial cells may be initiated by charge-charge interaction on the surface of proteins. Negatively charged cAMP seems to be able to block the toxicity of delta-endotoxin (Himeno et al., 1985). A number of other negatively charged molecules are also reportedly capable to inhibit B. thuringiensis delta-endotoxins non-specifically (Knowles and Ellar, 1987). These molecules may be able to interfere with access of the toxin to their receptors on midgut cell membrane. Among the 48 amino acids in the B. mori specificity domain on CrylA(a) toxin two amino acids (Glu 378 and Asp 417) are negatively charged; one is positively charged (Arg 404) and 16 are polar amino acids. It will be interesting to see if changes of a single or multiple amino acids in the domain will weaken or enhance the toxicity of delta-endotoxin. A few hybrid protein toxins, particularly those whose creation involves the EcoRV site near the carboxyl terminal portion of the hypervariable region, demonstrate no toxicity against B. mori (4107,4207,4108,4208,4112, and 4212). Preliminary data (R. Milne, personal communication) indicate that these protoxin proteins are not stable. They are completely degraded by B. mori gut juices 91 rather than being activated properly. It is highly possible that exchanges involving this EcoRV site could disrupt the structural integrity of the delta-endotoxin, and therefore, eliminate the toxicity. The EcoRV site is inside one of the highly conserved domains preceding an alternating arginine sequence (Arg-Tyr-Arg-Val-Arg-lle-Arg-Tyr). The significance of this region is unknown. However, since this region is extremely conserved among most B. thuringiensis delta-endotoxins, it may well be responsible for structural stability. 92 CONCLUSION Accompanying sporulation, different Bacillus thuringiensis strains produce various crystal proteins (delta-endotoxin) which are toxic to a number of insects belonging to Lepidoptera, Diptera, and Coleoptera. In an effort to identify the insect specificity of crystal protein toxins (Cry), two cry genes were cloned into the Escherichia coli expression vector pKK223-3, and bioassays were performed with purified crystals. The CrylA(a) protein (from a crylA(a), or "4.5 kb" gene, from B. thuringiensis var. kurstaki, HD-1) was found to be 400X more active against Bombyxmori than CrylA(c) protein (from an crylA(c)73, or "6.6 kb" gene, from B. thuringiensis var. kurstaki, HD-244). The CrylA(c) protein was 9X more active against Trichoplusia ni than the CrylA(a) protein, while both have similar activity against Manduca sexta. To locate the specificity domain of the CrylA(a) protein for B. mori, site-directed mutagenesis was used to introduce or remove restriction enzyme sites, facilitating the exchange of regions of the two genes. The hybrid genes were overexpressed, and purified hybrid proteins were used in bioassays. The B. mori specificity domain of CrylA(a) toxin is located in the amino-terminal portion of the "hypervariable region" between amino acids 332 and 450. Computer based analysis reveals that the predicted secondary structure of this region contains hydrophilic beta-sheet structures. This may suggest that amino acids in this region constitute a receptor binding domain of the crystal protein. References Adang, M. J., Staver, M. J., Rocheleau, T. A., Leighton, J., Barker, R. F. and Thompson, D. V.: Characterized full-length and truncated plasmid clones of the crystal protein of Bacillus thuringiensis subsp. kurstaki HD-73 and their toxicity to Manduca sexta. Gene 36 (1985) 289-300. Andrews, R. E ., Jr., Faust, R. M., Wabiko, H., Raymond, K. C. and Bulla, L. A ., Jr.: The biotechnology of Bacillus thuringiensis. CRC Critical Reviews in Biotechnology 6 (1987) 163-232. Choma, C. T., Surewicz, W. K., Mantsch, H. H., Carey, P. R., Pozsgay, M. and Kaplan, H.: Secondary structure of the entomocidal toxin from Bacillus thuringiensis subsp. kurstaki HD-73. J. Prot. Chem. (1990a). Choma, C. T., Surewicz, W. K., Carey, P. R., Pozsgay, M., Raynor, T. and Kaplan, H.: Unusual proteolysis of the protoxin and toxin from Bacillus thuringiensis: Structural implications. Eur. J. Biochem. (1990b). Chou, P.Y. and Fasman, R.F.: Prediction of the secondary structure of proteins from their amino acid sequence. Adv. Enzymol. 47 (1978) 45-148. Convents, D., Houssier, C., Lasters, I. and Lauwereys, M.: The Bacillus thuringiensis delta-endotoxin Evidence fro a two domain structure of the minimal toxic fragment. J. Biol. Chem. 265 (1990). Dulmage, H.T. and Cooperators: Insecticidal activity of isolates of Bacillus thuringiensis and their potential for pest control. In: Microbial Control of Pests and Plant Diseases 1970-1980, edited by Burges, H.D. London: Academic Press, 1981, p. 193-222. 93 94 Geiser, M., Schweitzer, S. and Grimm, C.: The hypervariable region in the genes coding for entomopathogenic crystal proteins of Bacillus thuringiensis: nucleotide sequence of the kurhdl gene of subsp. kurstaki HD1. Gene 48 (1986) 109-118. Gillam, S. and Smith, M.: Site-specific mutagenesis using synthetic oilgodeoxyribonucleotide primers: I. optimum conditions and minimum oligodeoxyribonucleotide length. Gene 8 (1979) 81-97. Haider, M. Z. and Ellar, D. J.: Mechanism of action of Bacillus thuringiensis insecticidal delta-endotoxin: Interaction with phospholipid vesicles. Biochim. Biophys. Acta 978 (1989) 216-222. Himeno, M., Koyama, N., Funato, T. and Koyama, T.: Mechanism of action of Bacillus thuringiensis insecticidal delta-endotoxin on insect cells in vitro. Agric. Biol. Chem. 49 (1985) 1461-1468. Hofmann, C., Vanderbruggen, H., Hofte, H., Van Rie, J., Jansens, S. and Van Mellaert, H.: Specificity of Bacillus thuringiensis delta-endotoxin is correlated with the presence of high-affinity binding sites in the brush border membrane of target insect midguts. Proc. Natl. Acad. Sci. USA 85 (1988) 7844-7848. Hofte, H., deGreve, H., Seurinck, J., Jansens, S., Mahillon, J., Ampe, C., Vandekerckhove, J., vanMontagu, M., Zabeau, M. and Vaeck, M.: Structural and functional analysis of a cloned delta endotoxin of Bacillus thuringiensis berliner 1715. Eur. J. Biochem. 161 (1986) 273-280. Hofte, H., Van Rie, J., Jansens, S., Van Houtven, A., Vanderbruggen, H. and Vaeck, M.: Monoclonal antibody analysis and insecticidal spectrum of three types of lepidopteran-specific insecticidal crystal proteins of Bacillus thuringiensis. Appl. Environ. Microbiol. 54 (1988) 2010-2017. 95 Hofte, H. and Whiteley, H. R.: Insecticidal crystal proteins of Bacillus thuringiensis. Microbiological Reviews 53 (1989) 242-255. Huber-Lukac, M., Luthy, P. and Braun, D. G.: Specificities of monoclonal antibodies against the activated delta-endotoxin of Bacillus thuringiensis var. thuringiensis. Infection and Immunity 40 (1983) 608-612. Huber-Lukac, M., Jaquet, F., Luthy, P., Huetter, R. and Braun, D. G.: Characterization of monoclonal antibodies to a crystal protein of Bacillus thuringiensis subsp. kurstaki. Infection and Immunity 54 (1986) 228-232. lizuka, T., Faust, R. M. and Ohba, M.: Comparative profiles of plasmid DNA and morphology of parasporal crystals in four strains of Bacillus thuringiensis subsp. darmstadiensis. Appl. Ent. Zool. 18(4) (1983) 486-494. Jaquet, F., Hutter, R. and Luthy, P.: Specificity of B. thuringiensis delta-endotoxin. Appl. Environ. Microbiol. 53 (1987) 500-504. Knowles, B. H. and Ellar, D. J.: Colloid-osmotic lysis is a general feature of the mechanism of action of Bacillus thuringiensis delta-endotoxin with different insect specificity. Biochim. Biophys. Acta 924 (1987) 509-518. Kondo, S., Tamura, N., Kunitate, A., Hattori, M., Akashi, A. and Ohmori, I.: Cloning and nucleotide sequencing of two insecticidal delta-endotoxin genes from Bacillus thuringiensis var. kurstaki HD-1 DNA. Agric. Biol. Chem. 51 (1987) 455-463. Kronstad, J. W., Schnepf, H. E. and Whiteley, H. R.: Diversity of locations for Bacillus thuringiensis crystal protein genes. J. Bacteriol. 154(1) (1983) 419-428. Kunkel, T. A.: Rapid and efficient site-specific mutagenesis without phenotypic selection. Proc. Natl. Acad. Sci. USA 82 (1985) 488-492. 96 Kyte, J. and Doolittle, R.F.: A simple method for displaying the hydropathic character of a protein. J. Mol. Biol. 157 (1982) 105-132. Maniatis, T., Fritsch, E.F. and Sambrook, J. Molecular Cloning, A Laboratory Manual :Cold Spring Harbor Laboratory, 1982. McLinden, J. H., Sabourin, J. R., Clark, B. D., Gensler, D. R., Workman, W. E. and Dean, D. H.: Cloning and expression of an insecticidal k-73 type crystal protein gene from Bacillus thuringiensis var. kurstaki into Escherichia coli. Appl. Environ. Microbiol. 50 (1985) 623-628. Perbal, B. A practical guide to molecular cloning: John Wiley & Sons, New York, 1984. Schnepf, H. E. and Whiteley, H. R.: Cloning and expression of the Bacillus thuringiensis crystal protein gene in Escherichia coli. Proc. Natl. Acad. Sci. USA 78 (1981) 2893-2897. Thorne, L., Garduno, F., Thompson, T., Decker, D., Zounes, M. A., Wild, M., Walfield, A. M. and Pollock, T. J.: Structural similarity between the lepidoptera- and diptera-specific insecticidal endotoxin genes of Bacillus thuringiensis subsp. kurstaki and israelensis. J. Bacteriol. 166 (1986) 801-811. Van Rie, J., Jansens, S., Hofte, H., Degheele, D. and Van Mellaert, H.: Specificity of Bacillus thuringiensis delta-endotoxin: Importance of specific receptors on the brush border membrane of the mid-gut of target insects. Eur. J. Biochem. (1990). Wilcox, D.R., Shivakumar, A.G., Melin, B.E., et al. Genetic engineering of bio-insecticides. In: Protein Engineering: Applications in Sciences, Industry and Medicine. (1990). 97 Yamamoto, T. and iizuka, T.: Two types of entomocidal toxins in the parasporal crystals of Bacillus thuringiensis kurstaki. Arch. Biochem. Biophys. 277 (1983) 223-241. Yamamoto, T., Garcia, J. A. and Dulmage, H. T.: Immunological properties on entomocidal protein of Bacillus thuringiensis and its insecticidal activity. J. Invert. Pathol. 41 (1983) 122-130. Yamamoto, T., Ehmann, A., Gonzalez, J. M. and Carlton, B. C.: Expression of three genes coding for 135-kilodalton entomocidal protein in Bacillus thuringiensis kurstaki. Curr. Microbiol. 17 (1988) 5-12.