MINIREVIEW. How Does Bacillus thuringiensis Produce So Much Insecticidal Crystal Protein?

Transcription

1 JOURNAL OF BACTERIOLOGY, Nov. 1995, p Vol. 177, No /95/$ Copyright 1995, American Society for Microbiology MINIREVIEW How Does Bacillus thuringiensis Produce So Much Insecticidal Crystal Protein? HERVÉ AGAISSE AND DIDIER LERECLUS* Unité de Biochimie Microbienne, Centre National de la Recherche Scientifique URA 1300, Institut Pasteur, 25 rue du Docteur Roux, Paris cedex 15, France Members of the genus Bacillus are widely used as sources of industrial enzymes, fine biochemicals, antibiotics, and insecticides (for a review, see reference 27). One of these species, Bacillus thuringiensis, accounts for more than 90% of the biopesticides used today (for recent reviews on B. thuringiensis and its toxins, see references 6, 33, and 38). The entomopathogenic properties of this bacterium are due at least in part to the production of -endotoxins that make up the crystalline inclusions characteristic of B. thuringiensis strains. In 1989, Höfte and Whiteley proposed a classification for -endotoxins (30). They distinguished four major classes of -endotoxins (CryI, -II, -III and -IV) and cytolysins (Cyt), found in the crystals of the mosquitocidal strains, on the basis of their insecticidal and molecular properties. The -endotoxins belonging to each of these classes were grouped in subclasses (A, B, C... and a, b, c...) according to sequence. Generally, these proteins are toxic for lepidoptera (CryI), both lepidoptera and diptera (CryII), coleoptera (CryIII), and diptera (CryIV). These various insecticidal proteins are synthesized during the stationary phase and accumulate in the mother cell as a crystal inclusion which can account for up to 25% of the dry weight of the sporulated cells (Fig. 1). The amount of crystal protein produced by a B. thuringiensis culture in laboratory conditions (about 0.5 mg of protein per ml) and the size of the crystals (24) indicate that each cell has to synthesize 10 6 to endotoxin molecules during the stationary phase to form a crystal. This is a massive production of protein and presumably occupies a large proportion of the cell machinery. Nevertheless, sporulation and the associated physiological changes proceed in parallel with -endotoxin production. The aim of this minireview is to analyze the various mechanisms by which B. thuringiensis accumulates large quantities of toxins as biologically active protein crystals. TRANSCRIPTIONAL MECHANISMS Crystal formation involves accumulation of toxin proteins. One method of accumulating a protein is to express the corresponding gene from a strong promoter in a nondividing cell, thus avoiding protein dilution by cell division. This is indeed the case in B. thuringiensis, where Cry protein inclusions are formed in stationary-phase cultures. Bacillus strains have evolved at least two different pathways in response to nutrient stress: nondividing semiquiescent period and a developmental program of spore formation. It is * Corresponding author. Mailing address: Unité de Biochimie Microbienne, Institut Pasteur, 25 rue du Docteur Roux, Paris cedex 15, France. Phone: Fax: Electronic mail address: lereclus@pasteur.fr. therefore convenient to distinguish between stationary phase genes that are dependent on sporulation and those that are not. Sporulation-dependent cry genes. In Bacillus species, the endospore develops in a sporangium consisting of two cellular compartments known as the mother cell and the forespore. In B. subtilis, the development process is temporally regulated at the transcriptional level by the successive activation of six sigma factors which, by binding to RNA polymerase, determine which gene promoters are recognized (28). These factors are the primary sigma factor of vegetative cells, A, and five factors that are activated during development and are called H, F, E, G, and K in order of their appearance during sporulation (43). The A and H factors are active in the predivisional cell, E and K are active in the mother cell, and F and G are active in the forespore. The cryia gene, encoding toxins active against lepidoptera, is a typical example of a sporulation-dependent cry gene that is only expressed in the mother cell compartment of B. thuringiensis. Two overlapping transcription start sites have been mapped, defining two overlapping promoters (BtI and BtII) used sequentially (55). BtI is active between about t 2 and t 6 of sporulation, and BtII is active from about t 5 onwards (t n is n hours after the end of the exponential phase). In vitro experiments have shown that transcription from BtI is initiated by a form of RNA polymerase containing an alternative sigma factor, 35 (12), whereas transcription from BtII is initiated by a form of RNA polymerase containing another alternative sigma factor, 28 (13). The genes encoding 35 and 28 have been cloned and sequenced, and their deduced amino acid sequences show 88 and 85% identity with E and K of B. subtilis, respectively (1). E and K mutants of B. thuringiensis have been constructed, and the cryiaa promoter activity in such mutants was monitored with a cryiaa - lacz transcriptional fusion (10). The K mutant displayed lower -galactosidase activity (about 50%) than the wild-type strain, whereas -galactosidase synthesis was abolished in the E mutant. Thus, these two sigma factors are involved in cryiaa transcription in vivo. The -galactosidase activity due to a cryiaa - lacz fusion (about 15,000 Miller units on average during sporulation) in B. thuringiensis is very high, suggesting that the promoters involved in cryiaa expression are very strong (Fig. 2). From in vitro transcription experiments it appears that at least three other cry genes (cryib, cryiia, and cyta) contain either BtI alone or both BtI and BtII (12). Consensus sequences for promoters recognized by B. thuringiensis RNA polymerase containing sporulation-specific sigma factor have been deduced from alignment of the promoter regions of these genes (Table 1). Many other cry genes (e.g., cryiva, cryivb, 6027

2 6028 MINIREVIEW J. BACTERIOL. FIG. 1. Electron micrograph of B. thuringiensis subsp. thuringiensis berliner 1715 during sporulation. The dark parasporal inclusion is the insecticidal crystal. and cryivd genes from subsp. israelensis, and cry34 and cry40 genes from subsp. thompsoni) are considered to be sporulation-stage-specific genes because their promoter regions contain these consensus sequences (11, 20, 58, 59). However, there has been neither in vitro transcription analysis nor comprehensive genetic analysis in E and K mutants of B. thuringiensis. Non-sporulation-dependent cry genes. The cryiiia gene, found in the coleopteran-active B. thuringiensis subsp. tenebrionis (49), is a typical example of a non-sporulation-dependent cry gene. Analysis with transcriptional fusions to the lacz reporter gene indicates that the kinetics of cryiiia and cryiaa expression are different (Fig. 2). The cryiiia promoter is weakly but significantly expressed during the vegetative phase of growth (unlike the cryiaa promoters), is activated at the end of exponential growth, and remains active only until about t 8 in sporulation medium. Maximum expression from the cryiiia promoter is very high as assessed from the -galactosidase activity obtained (15,000 Miller units during stationary phase). Unlike BtI and BtII, the cryiiia promoter is similar to promoters recognized by the primary sigma factor of vegetative cells, A (see Table 1 and reference 5). The expression of cryiiia is not dependent on sporulation-specific sigma factors either in B. subtilis (4) or in B. thuringiensis (48). Moreover, cryiiia expression is increased and prolonged in mutant strains which are unable to initiate sporulation (e.g., spo0f [41] or spo0a [4, 36, 48]). Thus, cryiiia activation is not regulated by the genes regulating sporulation initiation but rather by some regulator(s) affecting gene expression during the transition from exponential growth to the stationary phase. Furthermore, kinetic studies with cryiiia - lacz fusions suggest that in both B. subtilis and B. thuringiensis there is an event during sporulation which turns off cryiiia expression (see references 4 and 48 and Fig. 2). Two types of regulators acting on gene expression during stationary phase have been identified in B. subtilis (51). One type actively prevents gene expression during vegetative growth. The effects of these repressors are abolished at the onset of stationary phase (e.g., AbrB, Hpr, Sin, Pai proteins, etc.). The second type of regulator activates expression of previously silent genes (e.g., Deg proteins, ComP/ComA, SenS, Ten proteins, etc.). To date, the gene involved in cryiiia regulation has not been identified. TABLE 1. Nucleotide sequence of the cry gene promoters Promoter 35 region Spacer 10 region Reference FIG. 2. -Galactosidase activities of B. thuringiensis strains carrying the cryiaa - lacz and the cryiiia - lacz fusions. B. thuringiensis cells were grown in SP medium at 30 C. Time zero (t 0 ) indicates the onset of the stationary phase. t n is the number of hours before ( ) or after time zero. Consensus E GCATNT N 14 or 15 CATANNNT 12 (B. thuringiensis) cryia(a) BtI GCATTT N 15 CATATGTTT 12 Consensus K ND a ND TNATANNNTG 13 (B. thuringiensis) cryia(a) BtII ND ND TCATAAGATG 13 Consensus A TTGACA N 17 or 18 TATAAT 44 (B. subtilis) cryiiia TTGCAA N 18 TAAGCT 5 a ND, not determined.

3 VOL. 177, 1995 MINIREVIEW 6029 THE cry GENE COPY NUMBER The expression level of a gene may be influenced by its copy number, and gene amplification has frequently been used to overproduce proteins of industrial interest. The cry genes are located on plasmids, and many B. thuringiensis strains carry different cry genes (for a review, see reference 38). This natural amplification of the cry genes might therefore contribute to the high production of toxins in the different B. thuringiensis strains. However, B. thuringiensis strains which harbor a single cryi gene (e.g., subsp. kurstaki HD73) synthesize bipyramidal crystals which are not significantly smaller than those produced by strains harboring three or four different cryi genes (e.g., subsp. kurstaki HD1 or subsp. aizawai 7.29). Thus, assuming that the copy numbers of the large plasmids carrying the cry genes are similar in these B. thuringiensis strains, it appears that the production of toxins in B. thuringiensis is not strictly proportional to the copy number of the cry genes. This suggests that the capacity of the B. thuringiensis strains to produce the crystal proteins is limited (although at a high level) and reaches a maximum at a certain number of cry gene copies in the cell, above which there is no further increase in synthesis. Various experimental data support this idea. When a cryiac gene is cloned in a strain harboring other cryi genes, significantly less CryIAc protein is produced than when it is cloned into an acrystalliferous strain of B. thuringiensis (7). However, the reduced expression of cryia genes is not observed when they are introduced into a strain harboring a cryiiia gene (35, 39). The expression of these two genes is summed, presumably because they do not share rate-limiting elements of the expression systems. These effects may arise from titration of the specific sigma factors recruited for cry gene transcription and are consistent with saturation of cry gene expression being primarily at a transcriptional level. In agreement with this hypothesis, cloning cry genes in Bacillus species by using high-copy-number plasmids frequently disturbs the physiological equilibrium of the cells and prevents their sporulation (21, 22). mrna STABILITY The rate of mrna degradation has substantial consequences on gene expression. An important element in maximizing gene expression is the production of stable mrna. It is obviously expected that highly expressed proteins are encoded by stable mrnas. For example, ompa mrna, which encodes the major Escherichia coli outer membrane protein, displays an unusually long half-life (i.e., 20 min versus 2 to 3 min on average for E. coli mrnas) (45). Therefore, it is likely that the high level of -endotoxin production is due in part to the mrnas being stable. Glatron and Rapoport demonstrated that the mrnas encoding the crystal protein are among the relatively long-lived mrnas present in B. thuringiensis during the stationary phase and have an average half-life of 10 min (26). Studies of the degradation of unusually stable mrnas both in E. coli and in B. subtilis have led to the identification of a number of cis-elements that act as mrna stabilizers. These determinants of mrna stability are generally part of untranslated regions and are classified into two groups, according to their location in the mrna: (i) 3 terminal structures and (ii) 5 mrna stabilizers. Both of these types of determinant have been found in cry gene mrnas. 3 terminal structures. Wong and Chang first showed that the 3 terminal fragment of the cryiaa gene from B. thuringiensis subsp. kurstaki HD1 acts as a positive retroregulator (54). The fusion of this fragment with the 3 end of heterologous genes increases the half-life of their transcripts and consequently their expression level. Deletion analysis of this fragment showed that the region conferring the enhancing activity coincided with the potential transcriptional terminator sequence. This region harbors inverted repeat sequences with the potential to form stable stem-loop structures. 3-5 exoribonucleases are sensitive to RNA secondary structure. In particular, their processive activities are impeded by 3 stem-loop structures (for a review, see reference 29). Therefore, it is likely that the cry terminator is involved in mrna stability by protecting the cry mrna from exonucleolytic degradation from the 3 end. However, the involvement of the terminator in cryia mrna stability has not been analyzed by deleting this structure from the corresponding gene and it has therefore not been demonstrated that the 3 stem-loop structure is important for cryia gene expression. It is increasingly apparent that, for the majority of messages, the 3 stem-loop structure blocks exonuclease digestion for a sufficiently long time that degradation is initiated by upstream endonucleolytic cleavage. Therefore, the stability of cryia gene mrna may also be due to the absence of internal endonuclease cleavage sites. The putative terminator sequences downstream from various cry genes are widely conserved. Recently, a stem-loop binding protein that impedes the progress of the E. coli exoribonuclease PNPase in vitro has been identified (15). Thus, the binding of an additional factor to the 3 stem-loop structure of the cry mrna may be involved in the stabilizing effect of this region. However, whether such a mechanism exists in B. thuringiensis remains to be established. 5 mrna stabilizer. The promoter region of the cryiiia gene is complex: full expression requires the presence of a large DNA region extending up to 600 bp upstream of the translational start site (5, 19). Structural and functional analysis using transcriptional fusions to the lacz reporter gene has revealed that two distinct regions are involved in cryiiia expression (Fig. 3A). The upstream region is involved in transcription and harbors the cryiiia promoter sensu stricto. The downstream region is involved in cryiiia expression at a posttranscriptional level. This region is responsible for the accumulation of cryiiia mrna as a relatively stable transcript with a5 extremity corresponding to nucleotide position 129 (with reference to the translational start codon of cryiiia). Fusion of this region to the 5 region of the lacz gene transcribed from a promoter inducible in B. subtilis increases the stability of the lacz fusion mrna and results in a 10-fold increase of both mrna steady-state level and -galactosidase synthesis (3). In conferring stability to a heterologous sequence to which it is fused (e.g., the lacz gene), the downstream region has all the characteristics of a 5 mrna stabilizer (8). The determinant of stability appears to be a consensus Shine-Dalgarno (SD) sequence present close to the 5 extremity of the cryiiia mrna (Fig. 3B). This SD sequence does not direct translation initiation. However, analysis of mutations introduced into this region strongly suggests that the SD sequence is involved in stability through interaction with the 3 end of 16S rrna. Therefore, the binding of a 30S ribosomal subunit to the SD sequence located in the 5 untranslated region of cryiiia may stabilize the corresponding transcript. It is noteworthy that such SD sequences are also present in similar positions in at least two other members of the cryiii gene family, cryiiib and cryiiib2 (Fig. 3B), suggesting that the corresponding transcripts are probably stabilized by a similar mechanism. Recently, it has been shown that a segment of early RNA from B. subtilis bacteriophage SP82 could act as a

4 6030 MINIREVIEW J. BACTERIOL. FIG. 3. The cryiiia promoter region. (A) Model for cryiiia gene expression. Transcription of cryiiia is initiated at nucleotide position 558, giving the transcript T-558. This transcript is processed to give a stable form (T-129) which accumulates in the cells and appears as the major cryiiia mrna (5). (B) The 5 stabilizer downstream region. The DNA region downstream from nucleotide position 129 of three cryiii genes is indicated. The nucleotide sequences matching with the 3 end of the B. subtilis 16S rrna are boxed. Asterisks (nucleotide positions 129 and 128) represent the 5 ends of the major cryiii transcripts as previously described (5, 23). 5 mrna stabilizer (31). Similarly, the determinant of stability appears to be a polypurine sequence that resembles a ribosome binding site. Therefore, SD sequence may be a general determinant of mrna stability in Bacillus species. CRYSTALLIZATION OF THE INSECTICIDAL PROTEINS The obtention of high concentrations of proteins requires production of stable proteins. The biological context in which B. thuringiensis has to resolve this problem is that Bacillus species produce large amounts of a variety of proteolytic enzymes, especially during the stationary phase, and B. thuringiensis does not escape this generic characteristic. The bacterium must therefore have mechanisms to prevent premature proteolysis of the toxins. The proteolytic degradation of proteins is a common problem for the bioprocessing industry, and biotechnology currently uses several approaches to minimize this effect: secretion, utilization of proteolysis inhibitors or mutant strains with a reduced protease activity, construction of protein fusions for inclusion body formation, etc. (for a review, see reference 25). According to the environmental and biological constraints to which B. thuringiensis is naturally subjected (e.g., the necessity of proteases for its nutrient requirements), most of these strategies are excluded. The bioprocess used by B. thuringiensis is the formation of protease-resistant inclusion bodies. However, these protein inclusions have to be easily solubilized in the gut of insect larvae if they are to be toxic. The formation of the crystal structure and its solubility characteristics presumably depend on a variety of factors, including the secondary structure of the Cry proteins, the energy of the disulfide bonds, and the presence of additional components. Expression of several cryi genes in E. coli and B. subtilis leads to the synthesis of biologically active inclusions (14, 16, 46, 50), suggesting that the 130- to 140-kDa CryI proteins can spontaneously form crystals independent of the host bacterium. These results appear to agree with the principle of protein self-assembly, as first suggested by Lecadet s observation (34) that bipyramidal crystals denatured in 8 M urea reverted to their original shape when the urea was removed by dialysis. However, Du et al. (24) have recently shown that, after denaturation in 8 M urea and recrystallization, the bipyramidal crystals become soluble only at ph values of 12, instead of the ph 10 for the untreated protein. This loss of solubility correlated with a loss of toxicity. Thus, particular physicochemical conditions in the bacterial cells (E. coli, B. subtilis, or B. thuringiensis) are required to ensure correct assembly of the CryI molecules to form a soluble and active inclusion. Since the C-terminal regions of CryI, CryIVA, and CryIVB proteins are structurally similar (30, 37), their crystallization is presumably also similar. Indeed, the C-terminal halves of these molecules are cysteine rich and are not involved in toxicity. It is generally assumed that these regions contribute to the stability of the crystals through the formation of disulfide bonds. The crystallization of the 73-kDa CryIIIA protein may be slightly different: this protein forms a flat rectangular crystal inclusion in both B. subtilis and B. thuringiensis (4), but crystals generated by recrystallization from NaBr solutions and dialysis remain soluble and active against coleopteran larvae (9). The difference between the solubility characteristics of CryI and CryIIIA crystals is probably due to the absence of the cysteinerich C-terminal region from the CryIIIA protoxins. The CryIIIA crystals are composed of polypeptides which appear not to be linked by disulfide bridges (9). Analysis of the threedimensional structure of the CryIIIA protein suggests that four intermolecular salt bridges might be implicated in the formation of the crystal inclusion (40). There is various evidence for crystallization of CryIIA (71 kda), Cry IVD (72 kda), and CytA (27 kda) proteins proceeding via a different process which requires the presence of accessory proteins. The genes encoding CryIVD and CryIIA toxins are organized in an operon and are cotranscribed with genes not involved in toxicity; the gene cyta is transcribed divergently to cryivd (Fig. 4). Helen Whiteley s group has shown that a 20-kDa protein (P20 in Fig. 4) is required for stabilization of CytA and CryIVD in E. coli (2, 42, 52). The 20-kDa protein forms a complex with CytA which may act as a scaffold to facilitate (or stabilize) assembly of CytA into protease-resistant oligomers. This hypothesis is supported by the observation that the 20-kDa protein is not required for highlevel CytA production in E. coli strains carrying mutations which affect their proteolytic activity (52). Moreover, the 20- kda protein is found in the CytA crystals in B. thuringiensis, although not in stochiometric amounts with respect to CytA (2). The role of the 20-kDa protein in CytA and CryIVD crystallization in B. thuringiensis is less clear. The results of three studies disagree on the importance of its role (17, 56, 57).

5 VOL. 177, 1995 MINIREVIEW 6031 FIG. 4. Structural organization of the cryiia, cryivd, and cyta genes. The genes encoding CytA, CryIVD, P19, and P20 proteins are indicated as described by Adams et al. (2) and Dervyn et al. (20). The genes encoding ORF1, ORF2, and CryIIA are indicated as previously described by Widner and Whiteley (53). Arrows represent the transcripts and the direction of transcription. However, the discrepancy between these results could arise from the presence or absence of the P19 protein, which may also be involved in crystallization but which was not considered in these studies. The gene encoding P19 was recently mapped upstream of cryivd (Fig. 4; 20) and might be involved in the crystallization process of CytA and CryIVD. It slightly increases the synthesis of CryIVD in a B. subtilis wild-type strain, whereas it has no effect in a protease-deficient B. subtilis mutant strain (47). P19 is very similar to ORF1 (33% identity) encoded by the first gene of the cryiia operon (Fig. 4; 53). However, ORF1 has no detectable effect on the synthesis of CryIIA in B. thuringiensis (18). The protein which is involved in the efficient expression and presumably in the crystallization of CryIIA is the orf2 gene product (18), which was previously found as a component of the CryIIA crystals (53). These studies implicate a variety of accessory proteins (P19, P20, ORF1, and ORF2) in the synthesis or in the stabilization of the crystal proteins. According to the literal sense of the word, these proteins could be designated chaperones; however, they do not fit the current scientific usage of the term. Their designation as helper proteins (32), or more precisely as crystallization proteins (Crz proteins) may be thus more appropriate. CONCLUDING REMARKS B. thuringiensis produces a large variety of insecticidal crystal proteins, which reflects its remarkable adaptative capacity to settle in the large ecological niche constituted by insects. The biological mechanisms used by the bacterium to make enough toxin to kill insects are a striking example of bacterial evolution. From the plasmid localization of cry genes to the involvement of accessory proteins in the crystallization process, B. thuringiensis has developed a number of sophisticated systems entirely devoted to massive production of the toxins in the mother cell compartment during the stationary phase. The reason for the variety of toxins appears evident: expansion of the host spectrum. The reason for the maintenance of various different expression systems, rather than a single, powerful one, is less clear. It would be interesting to examine this question in the context of the ecology of the bacterium by considering its growth and the expression of a given cry gene in various environmental conditions, including the gut of its potential target insects. It is tempting to speculate that the gut environment of a given insect species might be favorable to one expression system rather than another (i.e., sporulation dependent versus sporulation independent). Similarly, the crystallization characteristics of a toxin appear to correlate with its insecticidal activity spectrum. An example is the CryIII toxin active against coleopteran larvae. CryIII lacks the C-terminal segment found in other Cry proteins. This causes the CryIIIbased crystals to be soluble in the acidic guts of the coleopteran larvae. Crystal proteins (as CryI, CryIVA, and CryIVB) that contain this C-terminal region are solubilized only at a high ph, which is characteristic of the lepidopteran and dipteran gut environment. Whatever the biological significance of their differences, the various expression systems of the cry genes provide a series of models to study genetic expression during the stationary phase of gram-positive sporulating bacteria. Further studies of the cry gene transcripts will undoubtedly lead to an understanding of the structural and functional basis of mrna stability in Bacillus species. This has large biotechnological implications. Indeed, the transcriptional and posttranscriptional mechanisms described in this review could be used to overproduce proteins of industrial interest. A better knowledge of the crystallization process used by B. thuringiensis might also provide information allowing heterologous proteins to be stabilized as biologically active crystal inclusions. ACKNOWLEDGMENTS We thank A. Bravo, E. Dervyn, S. Poncet, and S. Salamitou for sharing information prior to publication and J. Ribier for kindly providing the electron micrograph. We are grateful to A. Edelman for English corrections. H.A. was supported by a grant from Roussel-Uclaf. D.L. was supported by the Institut National de la Recherche Agronomique. REFERENCES 1. Adams, L. F., K. L. Brown, and H. R. Whiteley Molecular cloning and characterization of two genes encoding sigma factors that direct transcription from a Bacillus thuringiensis crystal protein gene promoter. J. Bacteriol. 173: Adams, L. F., J. E. Visick, and H. R. Whiteley A 20-kilodalton protein is required for efficient production of the Bacillus thuringiensis subsp. israelensis 27-kilodalton crystal protein in Escherichia coli. J. Bacteriol. 171: Agaisse, H., and D. Lereclus. Unpublished data. 4. Agaisse, H., and D. Lereclus Expression in Bacillus subtilis of the Bacillus thuringiensis cryiiia toxin gene is not dependent on a sporulationspecific sigma factor and is increased in a spo0a mutant. J. Bacteriol. 176: Agaisse, H., and D. Lereclus Structural and functional analysis of the promoter region involved in full expression of the cryiiia toxin gene of Bacillus thuringiensis. Mol. Microbiol. 13: Aronson, A. I The two faces of Bacillus thuringiensis: insecticidal proteins and post-exponential survival. Mol. Microbiol. 7: Baum, J. A., D. M. Coyle, M. P. Gilbert, C. S. Jany, and C. Gawron-Burke Novel cloning vectors for Bacillus thuringiensis. Appl. Environ. Microbiol. 56: Bechhoffer, D. H mrna stabilizers, p In J. G. Belasco and G. Brawerman (ed.), Control of messenger RNA stability, Academic Press, Inc., San Diego, Calif. 9. Bernhard, K Studies on the delta-endotoxin of Bacillus thuringiensis var. tenebrionis. FEMS Microbiol. Lett. 33: Bravo, A., and D. Lereclus. Unpublished data. 11. Brown, K. L Transcriptional regulation of the Bacillus thuringiensis subsp. thompsoni crystal protein gene operon. J. Bacteriol. 175: Brown, K. L., and H. R. Whiteley Isolation of a Bacillus thuringiensis RNA polymerase capable of transcribing crystal protein genes. Proc. Natl. Acad. Sci. USA 85: Brown, K. L., and H. R. Whiteley Isolation of the second Bacillus thuringiensis RNA polymerase that transcribes from a crystal protein gene promoter. J. Bacteriol. 172: Calogero, S., A. M. Albertini, C. Fogher, R. Marzari, and A. Galizzi Expression of a cloned Bacillus thuringiensis delta-endotoxin gene in Bacillus subtilis. Appl. Environ. Microbiol. 55: Causton, H. C., B. Py, R. S. McLaren, and C. H. Higgins mrna degradation in Escherichia coli: a novel factor which impedes the exoribonucleolytic activity of PNPase at stem-loop structures. Mol. Microbiol. 14: Chak, K.-F., and D. J. Ellar Cloning and expression in Escherichia coli of an insecticidal crystal protein gene from Bacillus thuringiensis var. aizawai HD-133. J. Gen. Microbiol. 133: Chang, C., Y.-M. Yu, S.-M. Dai, S. K. Law, and S. S. Gill High-level cryivd and cyta gene expression in Bacillus thuringiensis does not require the 20-kilodalton protein, and the coexpressed gene products are synergistic in their toxicity to mosquitoes. Appl. Environ. Microbiol. 59:

6 6032 MINIREVIEW J. BACTERIOL. 18. Crickmore, N., and D. J. Ellar Involvement of a possible chaperonin in the efficient expression of a cloned CryIIA -endotoxin gene in Bacillus thuringiensis. Mol. Microbiol. 6: De Souza, M. T., M.-M. Lecadet, and D. Lereclus Full expression of the cryiiia toxin gene of Bacillus thuringiensis requires a distant upstream DNA sequence affecting transcription. J. Bacteriol. 175: Dervyn, E., S. Poncet, A. Klier, and G. Rapoport Transcriptional regulation of the cryivd gene operon from Bacillus thuringiensis subsp. israelensis. J. Bacteriol. 177: Donovan, W. P., C. Dankocsik, and M. P. Gilbert Molecular characterization of a gene encoding a 72-kilodalton mosquito-toxic crystal protein from Bacillus thuringiensis subsp. israelensis. J. Bacteriol. 170: Donovan, W. P., J. M. González, Jr., M. P. Gilbert, and C. Dankocsik Isolation and characterisation of EG2158, a new strain of Bacillus thuringiensis toxic to coleopteran larvae, and nucleotide sequence of the toxin gene. Mol. Gen. Genet. 214: Donovan, W. P., M. J. Rupar, A. C. Slaney, T. Malvar, M. C. Gawron-Burke, and T. B. Johnson Characterization of two genes encoding Bacillus thuringiensis insecticidal crystal proteins toxic to coleoptera species. Appl. Environ. Microbiol. 58: Du, C., P. A. W. Martin, and K. W. Nickerson Comparison of disulfide contents and solubility at alkaline ph of insecticidal and noninsecticidal Bacillus thuringiensis protein crystals. Appl. Environ. Microbiol. 60: Enfors, S.-O Control of in vivo proteolysis in the production of recombinant proteins. Trends Biotechnol. 10: Glatron, M. F., and G. Rapoport Biosynthesis of the parasporal inclusion of Bacillus thuringiensis: half-life of its corresponding messenger RNA. Biochimie 54: Harwood, C. R Bacillus subtilis and its relatives: molecular biological and industrial workhorses. Trends Biotechnol. 10: Helmann, J. D., and M. J. Chamberlin Structure and function of bacterial sigma factors. Annu. Rev. Biochem. 57: Higgins, C. F., H. C. Causton, G. S. C. Dance, and E. A. Mudd The role of the 3 end in mrna stability and decay, p In J. G. Belasco and G. Brawerman (ed.), Control of messenger RNA stability. Academic Press, Inc., San Diego, Calif. 30. Höfte, H., and H. R. Whiteley Insecticidal crystal proteins of Bacillus thuringiensis. Microbiol. Rev. 53: Hue, K. K., S. D. Cohen, and D. H. Bechhofer A polypurine sequence that acts as a 5 mrna stabilizer in Bacillus subtilis. J. Bacteriol. 177: Koni, P. A., and D. J. Ellar Cloning and characterization of a novel Bacillus thuringiensis cytolytic delta-endotoxin. J. Mol. Biol. 229: Lambert, B., and M. Peferoen Insecticidal promise of Bacillus thuringiensis. BioScience 42: Lecadet, M.-M Action comparée de l urée et du thioglycolate sur la toxine figurée de Bacillus thuringiensis. C. R. Acad. Sci. Ser. D 264: Lecadet, M.-M., J. Chaufaux, J. Ribier, and D. Lereclus Construction of novel Bacillus thuringiensis strains with different insecticidal specificities by transduction and by transformation. Appl. Environ. Microbiol. 58: Lereclus, D., H. Agaisse, M. Gominet, and J. Chaufaux Overproduction of encapsulated insecticidal crystal proteins in a Bacillus thuringiensis spo0a mutant. Bio/Technology 13: Lereclus, D., C. Bourgouin, M.-M. Lecadet, A. Klier, and G. Rapoport Role, structure, and molecular organization of the genes coding for the parasporal -endotoxins of Bacillus thuringiensis, p In I. Smith, R. A. Slepecky, and P. Setlow (ed.), Regulation of procaryotic development. American Society for Microbiology, Washington, D.C. 38. Lereclus, D., A. Delécluse, and M.-M. Lecadet Diversity of Bacillus thuringiensis toxins and genes, p In P. Entwistle, J. S. Cory, M. J. Bailey, and S. Higgs (ed.), Bacillus thuringiensis, an environmental biopesticide: theory and practice, John Wiley & Sons, Ltd., Chichester, U.K. 39. Lereclus, D., M. Vallade, J. Chaufaux, O. Arantes, and S. Rambaud Expansion of insecticidal host range of Bacillus thuringiensis by in vivo genetic recombination. Bio/Technology 10: Li, J., J. Carroll, and D. J. Ellar Crystal structure of insecticidal -endotoxin from Bacillus thuringiensis at 2.5 Å resolution. Nature (London) 353: Malvar, T., and J. Baum Tn5401 disruption of the spo0f gene, identified by direct chromosomal sequencing, results in cryiiia overproduction in Bacillus thuringiensis. J. Bacteriol. 176: McLean, K. M., and H. R. Whiteley Expression in Escherichia coli of a cloned crystal protein gene of Bacillus thuringiensis subsp. israelensis. J. Bacteriol. 169: Moran, C. P RNA polymerase and transcription factors, p In A. L. Sonenshein, J. A. Hoch, and R. Losick (ed.), Bacillus subtilis and other gram-positive bacteria. American Society for Microbiology, Washington, D.C. 44. Moran, C. P., N. Lang, S. F. J. Legrice, G. Lee, M. Stephens, A. L. Sonnenshein, J. Pero, and R. Losick Nucleotide sequences that signal the initiation of transcription and translation in Bacillus subtilis. Mol. Gen. Genet. 186: Nilsson, G., J. G. Belasco, S. N. Cohen, and A. von Gabain Growth rate dependent regulation of mrna stability in E. coli. Nature (London) 312: Oeda, K., K. Oshie, M. Shimizu, K. Nakamura, H. Yamamoto, I. Nakayama, and H. Ohkawa Nucleotide sequence of the insecticidal protein gene of Bacillus thuringiensis strain aizawai IPL-7 and its high-level expression in Escherichia coli. Gene 53: Poncet, S Ph.D. thesis. Université Paris VII. 48. Salamitou, S., H. Agaisse, A. Bravo, and D. Lereclus. Unpublished data. 49. Sekar, V., D. V. Thompson, M. J. Maroney, R. G. Bookland, and M. J. Adang Molecular cloning and characterization of the insecticidal crystal protein gene of Bacillus thuringiensis var. tenebrionis. Proc. Natl. Acad. Sci. USA 84: Shivakumar, A. G., G. J. Gundling, T. A. Benson, D. Casuto, M. F. Miller, and B. B. Spear Vegetative expression of the -endotoxin genes of Bacillus thuringiensis subsp. kurstaki in Bacillus subtilis. J. Bacteriol. 166: Smith, I Regulatory proteins that control late-growth development, p In A. L. Sonenshein, J. A. Hoch, and R. Losick (ed.), Bacillus subtilis and other gram-positive bacteria. American Society for Microbiology, Washington, D.C. 52. Visick, J. E., and H. R. Whiteley Effect of a 20-kilodalton protein from Bacillus thuringiensis subsp. israelensis on production of the CytA protein by Escherichia coli. J. Bacteriol. 173: Widner, W. R., and H. R. Whiteley Two highly related insecticidal crystal proteins of Bacillus thuringiensis subsp. kurstaki possess different host range specificities. J. Bacteriol. 171: Wong, H. C., and S. Chang Identification of a positive retroregulator that stabilizes mrnas in bacteria. Proc. Natl. Acad. Sci. USA 83: Wong, H. C., H. E. Schnepf, and H. R. Whiteley Transcriptional and translational start sites for the Bacillus thuringiensis crystal protein gene. J. Biol. Chem. 258: Wu, D., and B. A. Federici A 20-kilodalton protein preserves cell viability and promotes CytA crystal formation during sporulation in Bacillus thuringiensis. J. Bacteriol. 175: Wu, D., and B. A. Federici Improved production of the insecticidal CryIVD protein in Bacillus thuringiensis using cryia(c) promoters to express the gene for an associated 20-kDa protein. Appl. Microbiol. Biotechnol. 42: Yoshisue, H., T. Fukada, K.-I. Yoshida, K. Sen, S.-I. Kurosawa, H. Sakai, and T. Komano Transcriptional regulation of Bacillus thuringiensis subsp. israelensis mosquito larvicidal crystal protein gene cryiva. J. Bacteriol. 175: Yoshisue, H., T. Nishimoto, and T. Komano Identification of a promoter for the crystal protein-encoding gene cryivb from Bacillus thuringiensis subsp. israelensis. Gene 137: