expression bacillus thuringiensis require the 20
TRANSCRIPT
APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Mar. 1993, p. 815-821
0099-2240/93/030815-07$02.00/0 Copyright © 1993, American Society
for Microbiology
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 CHENG CHANG,1 YONG-MAN yU,2 SHU-MEI DAI,2 SARA K. LAW,2
AND SARJEET S. GILL1' 2*
Department of Entomology2 and Interdepartmental Graduate Program in Environmental Toxicology, 1 University of California, Riverside, California 92521
Received 31 August 1992/Accepted 8 December 1992
Interactions among the 20-kDa protein gene and the cyt4 and cryIVD genes located in a 9.4-kb HindIll fragment were studied. A series of plasmids containing a combination of these different genes was constructed by using the Escherichia colil/Bacilus thuringiensis shuttle vector pHT3101. The plasmids were then used to transform an acrystalliferous strain, cryB, derived from B. thuringiensis subsp. kurstaki. The results from sodium dodecyl sulfate-polyacrylamide gel electrophoresis and immunoblot analyses suggest that although the 20-kDa protein is required for the efficient CytA protein production in E. coli, it is not required in B. thuringiensis. With or without the truncated 20-kDa protein gene, the CytA and/or Cry1VD proteins are
produced and form parasporal inclusions in B. thuringiensis cells. However, more-efficient expression is obtained when a second protein, probably acting as a chaperonin, is present. In addition, the time course
studies show that the CytA and CrylVD proteins are coordinately produced. Both the crude B. thuringiensis culture and purified inclusions from each recombinant B. thuringiensis strain are toxic to Culex quinquefas- ciatus larvae. The parasporal inclusions formed in B. thuringiensis cells are mosquitocidal, with CytA synergizing CryIVD toxicity.
Bacillus thuringiensis subsp. israelensis and B. thurin- giensis subsp. morrisoni (PG-14) both produce spherical parasporal inclusions that are highly toxic to dipteran larvae, such as mosquitoes and black flies (18, 27). The parasporal inclusions from both subspecies produce 27-, 72-, 125-, and 135-kDa polypeptides, the CytA, CryIVD, CryIVB, and CryIVA proteins, respectively (12-15, 19). The B. thurin- giensis subsp. morrisoni parasporal inclusions also contain a
144-kDa protein not found in B. thuringiensis subsp. israe- lensis (26). The genes encoding the CytA, CryIVA, CryIVB, and CryIVD proteins have been cloned and expressed in both Escherichia coli and B. thuringiensis (1, 4, 8, 9, 13, 24, 33, 35). All proteins common to both subspecies are mos-
quitocidal, with the Cry proteins having the greatest insec- ticidal activity (2, 4, 8, 10, 33). The CytA protein, in addition to its mosquitocidal activity (34), is also hemolytic and cytolytic (17, 31). The insecticidal activity of B. thuringiensis subsp. israe-
lensis is a complex interaction of the four inclusion body proteins, CryIVA, CryIVB, CryIVD, and CytA (15). Among these proteins, CytA exclusion from inclusion bodies has relatively little effect on insecticidal activity (9). Although each protein is mosquitocidal, none is as active as the intact parasporal inclusion. The higher insecticidal activity of the intact parasporal inclusions is primarily due to the synergis- tic interaction between the proteins present in these inclu- sions (2, 5, 8, 20, 38, 39).
In addition to the synergistic interaction between CytA and CryIVD, these proteins are encoded by the same
HindIII fragment which also encodes a 20-kDa protein. In
* Corresponding author.
E. coli, CytA levels are increased because of posttransla- tional stabilization by this 20-kDa protein (1, 32). It is not known, however, whether this 20-kDa protein plays any role in CytA or CryIVD levels in B. thuingiensis. To investigate the role of the 20-kDa protein, a recombinant system in B. thuringiensis (4) was used. We report here that the 20-kDa protein is not required for the production of high levels of either CytA or CryIVD in B. thurngiensis. However, the presence of an additional protein increases the CytA and CryIVD levels in an acrystalliferous strain of B. thuringien- sis. Mosquitocidal activity evaluations also demonstrate that CytA synergizes the insecticidal activity of CryIVD to Culex quinquefasciatus.
MATERIALS AND METHODS
Bacterial strains, plasmids, and general methods. Bacterial strains used in this study were E. coli JM101, JM109, and XL-1 (Stratagene, La Jolla, Calif.), BMH 71-18 mutS (Promega, Madison, Wis.), and the acrystalliferous strain cryB, derived from B. thuringiensis subsp. kurstaki HD1, obtained from A. Aronson, Department of Biology, Purdue University. The plasmid pMl, which encodes the B. thur- ingiensis subsp. morrisoni (PG-14) CytA, CryIVD, and 20-kDa proteins, was obtained from B. Federici and S. Sivasubramanian of our department (14). The E. coliIB. thuringiensis shuttle vector pHT3101 (22) was obtained from D. Lereclus, Institut Pasteur, Paris, France. Site-directed mutagenesis was performed with pSELECT-1 by using the Altered Sites in vitro mutagenesis system (Promega) follow- ing the manufacturer's instructions while all other cloning was performed with pBluescriptIl SK+ (Stratagene). Stan- dard protocols were used for restriction enzyme digestion,
815
http://aem .asm
.org/ D
ligation, and E. coli transformation (29). B. thuringiensis cryB cells were transformed by electroporation essentially as described previously (22).
Construction of the pCG plasmid series. The construction of pCG1 has been described previously (4). For pCG2 construction, pM1 was restricted with EcoRI. The 5.7-kb fragment, which contains cytA and a truncated cryIVD gene, was separated by agarose gel electrophoresis and purified by Geneclean (Bio 101, La Jolla, Calif.). The purified fragment was ligated into the unique EcoRI site of pHT3101 and then used for transformation in E. coli JM101. Transformants, which were selected on Luria-Bertani agar containing ampi- cillin (50 ,ug/ml), 20 mM isopropyl-p-D-thiogalactopyrano- side (IPTG), and 80 mg of 5-bromo-4-chloro-3-indolyl-p3-D- galactopyranoside (X-Gal) per ml were isolated, and the sizes and orientations of the EcoRI inserts were determined. To construct pCG4, the cryIVD EcoRI site was deleted by
site-directed mutagenesis with the Altered Sites in vitro mutagenesis system. Briefly, the 5.3-kb BamHI-PstI frag- ment from pCG1 which contains the cryIVD gene was inserted into the BamHI-PstI sites of pSELECT-1. The cryIVD internal EcoRI site was then deleted by primer- directed mutagenesis. The mutant, identified by an EcoRI site deletion, was then restricted with EcoRI. The 4.4-kb EcoRI fragment which contains cryIVID, a truncated cytA gene, and a truncated 20-kDa protein gene was then inserted into the unique EcoRI site of pHT3101, resulting in pCG4. For pCG5 construction, the 4.9-kb pMl BamHI-ClaI frag- ment was subcloned into pBluescriptII SK+. The pCG2 2.0-kb SacI-BamHI fragment was then inserted into SacI- BamHI sites, giving pCGB12. The 6.9-kb pCGB12 KpnI- SacI fragment which contains the cryIID, cytA, and 20-kDa protein genes was then inserted into the shuttle vector pHT3101, resulting in the plasmid pCG5. The 4.4-kb pM1 EcoRV-PstI fragment was isolated and
ligated with a SacI linker, and the resulting fragment was inserted into the unique SacI-PstI sites of pHT3101, result- ing in the construction of pCG6. To obtain pCG8, the 6.9-kb SacI-Clal fragment from pCG5 was ligated with the 7.2-kb SacI-ClaI fragment of pCG1. To construct pCG10, the 4.9-kb pCG6 SpeI-BamHI fragment, which lacks the B. thuringiensis origin of replication but retains the E. coli replication origin, was isolated. This fragment was then ligated with the 1.2-kb SpeI-BamHI fragment of pCGE1, which contains the 3' cryIVD fragment but lacks the 20-kDa protein gene, resulting in the formation of pCG9; the BamHI site in pCGE1 was introduced by the polymerase chain reaction. For expression in B. thunngiensis, the 3.6-kb pHT3101 BamHI fragment containing the B. thunngiensis replication origin was inserted into the unique pCG9 BamHI site; this new construct, pCG10, contains only the cryIVD gene. The plasmid pCG12 was constructed by replacing the
6.9-kb SacI-ClaI fragment of pCG5, which contains all three genes, with the 4.0-kb SacI-ClaI fragment of pCG6, which contains cryIVD and the gene encoding the 20-kDa protein. For construction of pCG13, the 2.9-kb SacI-EcoRV frag- ment of pCG5, which contains cytA, was isolated from the agarose gel, and XbaI 10-mer linkers were used to modify the blunt end. After restriction, the SacI-XbaI fragment was cloned into the corresponding restriction sites of pHT3101. For construction of pCG17, the 0.93-kb pCGB12 HaeIII-
ClaI fragment, containing the 20-kDa protein gene and a 32-bp 3' end of cryIVD, was used to replace the 3.98-kb fragment containing the 20-kDa protein gene and cryIVD. The resulting construct was then cloned into pHT3101 to
obtain pCG13, which contains cytA, 32 bp of cryIVD, and the 20-kDa protein gene.
cyt4 and cryIVD gene expression. B. thunngiensis subsp. kurstaki recombinant CG strains containing the pCG plas- mids were analyzed for the CytA and CryIVD proteins. These B. thuningiensis subsp. kurstaki transformants were grown on nutrient agar plates containing 50 p,g of erythro- mycin per ml for 3 days at 30°C. The bacterial cultures were isolated by washing the culture plates with deionized water. Crude culture aliquots were boiled in sample treatment buffer and analyzed by discontinuous sodium dodecyl sulfate polyacrylamide gel electrophoresis, (SDS-PAGE) with 4.5% acrylamide (pH 6.8) and 10% acrylamide (pH 8.8) as the stacking and separating gels, respectively (21). The gels were then stained with 0.1% Coomassie blue R-250. Alternatively, the proteins resolved by SDS-PAGE were transferred to nitrocellulose for 16 h at a constant current of 250 mA, and the nitrocellulose was then probed with rabbit antibody developed against either the purified whole parasporal inclu- sion or the purified 72-kDa toxin of B. thuringiensis subsp. israelensis by using methods described previously (16). Goat anti-rabbit immunoglobulin G-alkaline phosphatase was used as the second antibody, and chromogenic development was then achieved with nitroblue tetrazolium chloride (1 mg/ml in H20) and 5-bromo-4-chloro-3-indolyl phosphate (5 mg/ml in dimethylformamide). For time course studies, cultures were terminated at 12,
24, 36, 48, and 72 h. The plate bacterial cultures were isolated with 10 mM EDTA and sedimented by centrifuga- tion at 15,000 x g for 10 min. After the protein concentration was estimated (23), the cultures were analyzed by discontin- uous SDS-PAGE and immunoblotting as described above.
Purification of parasporal inclusions. B. thunngiensis subsp. kurstaki CG strains were cultured on nutrient agar plates containing 50 ,ug of erythromycin per ml for 5 days at 30°C to ensure sporulation and complete autolysis. The spore-parasporal inclusion mixture was thoroughly washed with 1 M NaCl-10 mM EDTA and sedimented by centrifu- gation at 15,000 x g for 10 min. The pellet was resuspended in water, sonicated, loaded onto a continuous 40 to 70% Renografin density gradient, and centrifuged at 15,000 rpm for 30 min in an SW28 rotor as described previously (40). The parasporal inclusions were then subjected to a second centrifugation. The purified parasporal inclusions were washed three times with distilled water, the protein concen- tration was measured (23), and the inclusions were stored at 40C.
Insect bioassays. The larval mosquitocidal activity of each B. thuringiensis CG strain was determined. For determina- tion of crude bacterial culture toxicity, the cultures were serially diluted and added to distilled water (total volume, 10 ml) containing 10 fourth-instar C. quinquefasciatus larvae. The 24-h mortality was determined by counting the number of surviving larvae. Bioassays were performed in triplicate. The mosquitocidal activity of purified inclusions from the
B. thuringiensis subsp. kurstaki CG series was also assessed. Briefly, 0.1 ml of parasporal inclusion dilutions was added to 99.9 ml of distilled water containing 20 fourth-instar C. quinquefasciatus larvae. To determine a 50% lethal concen- tration (LC50), 10 different inclusion concentrations (1 to 1,000 ng/ml) were used, with at least four replicates per concentration. The 24-h mortality was determined, and LC50 and 95% lethal concentration (LC95) values were calculated by probit analyses (28). Controls for the bioassays utilized mosquito larvae reared similarly but not exposed to either
816 CHANG ET AL.
http://aem .asm
.org/ D
cytA ciyIVD
E
FIG. 1. Construction of the pCG plasmid series and determination of cryIVD and cytA gene expression in B. thuningiensis (Bt) and the mosquitocidal activity of the gene products obtained. B, BamHI; C, ClaI; E, EcoRI; EV, EcoRV; H, HindIII; P, PstI; Sa, SacI; X, XbaI. + + +, + +, and + indicate that 100, >50, or <50% mortality, respectively, was observed in the mosquito larva bioassay with crude bacterial cultures.
the crude bacterial cultures or the purified parasporal inclu- sions.
RESULTS
To determine the interactions between the CytA, CryIVD, and the 20-kDa proteins or their genes, several DNA frag- ments derived from the 9.4-kb HindIII fragment ofpMl were introduced into the shuttle vector pHT3101 for expression in B. thuringiensis (Fig. 1). The CytA and CryIVD protein levels were detected by SDS-PAGE analysis (Fig. 2). Immu- noblot analysis confirmed the identity of the CytA and CryIVD proteins (data not shown). In all experiments, the 20-kDa protein was not observed in the acrystalliferous strain cryB.
Role of the 20-kDa protein in cyt4 gene expression. The 20-kDa protein is reportedly required for efficient CytA production in E. coli (1, 32). The plasmids pCG2 and pCG13 were therefore constructed to determine whether the 20-kDa protein is required for cytA gene expression in the strain cryB derived from B. thunngiensis subsp. kurstaki. Our results show that CG2 and CG13, cryB strains transformed with pCG2 and pCG13, respectively, expressed significant amounts of the 27-kDa CytA protein at levels readily detect- able by SDS-PAGE (Fig. 2). High CytA levels similar to those in CG2 were also observed in CG17 (data not shown). However, the CytA protein in CG13 is synthesized in lower
kDa
72
27
1 2 3 4 58a 7 9 10 81112131415161718
FIG. 2. SDS-PAGE analysis of cryIVD and cytA gene expression in B. thuningiensis subsp. kurstaki (cryB). Lanes 1 to 10 were loaded
with 10 pLg of total protein from each B. thuningiensis crude culture
(3 days at 30'C). Lanes 11 to 18 were loaded with 5 p.g of purified inclusions from each indicated B. thuringiensis strain. Lanes: 1 and
11, CG1; 2 and 12, CG2; 3 and 13, CG4; 4 and 14, CG5; 5 and 15,
CG6; 6 and 16, CG8; 7 and 17, CG10; 8 and 18, CG12; 9, CG13; 10,
cryB; S, Standard proteins. Molecular size markers are indicated on
the right.
http://aem .asm
.org/ D
APPL. ENVIRON. MICROBIOL.
amounts than in strains CG2 and CG17, and in addition, the CG13 CytA protein is proteolytically cleaved to a 24-kDa protein. Thus, the 27-kDa protein is less stable in CG13 than in CG2 and CG17. After 5 days of culture on nutrient agar plates at 30°C, strain CG2 formed large ovoid parasporal inclusions during sporulation. These ovoid inclusions were readily observed under phase-contrast microscopy and in quantities sufficient to form a distinct band when they were purified by Renografin gradients.
Since the EcoRI fragment used to construct the pCG2 contains a portion of the cryIVD gene and its promoter, immunoblot analyses were performed to detect the presence of the CryIVD (72-kDa) protein by using antibodies raised against the intact B. thuringiensis subsp. israelensis paras- poral inclusions. The truncated cryIVD gene in this con- struct does not have a translation stop codon, and the cryIVD and P-galactosidase genes are in frame; therefore, the fusion protein formed includes part of the ,B-galactosi- dase protein. A 74-kDa CryIVD-13-galactosidase fusion pro- tein containing a C-terminal deletion of 77 amino acids from the CryIVD protein but containing 98 amino acid residues from the P-galactosidase was produced and confirmed by immunoblot analysis (data not shown). Although strain CG13 can form inclusions, spore formation in this strain is aberrant and the bacterial cells do not autolyze after 6 days of culture at 30°C. The bacterial strain CG13 produces the 27-kDa protein and a 24-kDa proteolytic product (Fig. 2). However, because of the lack of autolysis in the CG13 strain, parasporal inclusions are difficult to purify from this strain. Consequently, strain CG17 was constructed and produced CytA levels comparable to the CytA levels in CG2. Our observations show that the 20-kDa gene product is not required for efficient CytA protein production in B. thurin- giensis, in contrast to that observed in E. coli.
B. thuringiensis CryIYD production in the absence of the 20-kDa protein. We previously obtained high cryIVD gene expression in B. thuringiensis by using the shuttle vector pHT3101 (4). However, this construct contained, in addition to cryl'D, a truncated cytA gene and the 20-kDa protein gene. Therefore, to determine whether the 20-kDa protein is involved in the CryIVD protein production, the plasmids pCG4, pCG6, and pCG1O were constructed (Fig. 1). Plasmid pCG4 contains cryIVD, a truncated cytA gene, and a trun- cated 20-kDa protein gene. The plasmid pCG6 has only cryIVD and a 20-kDa protein gene, while pCG10 has only cryIVD but lacks the invert repeats (10). These plasmids were then used for transformation of the strain cryB derived from B. thunngiensis subsp. kurstaki. The results of SDS-PAGE (Fig. 2) and immunoblot anal-
yses show that the cryIVD gene is expressed in all three bacterial strains, i.e., CG4, CG6, and CG10; but bacterial strain CG1O produces lower CryIVD levels than do strains CG4 and CG6 (Fig. 3 and 4). During SDS-PAGE and immunoblotting analyses, most of the CryIVD products were observed in the pellet fraction following centrifugation at 15,000 x g for 10 min, demonstrating that the CrylVD protein is mostly in the inclusion and that only a small portion is released into the supernatant after cell lysis. Time course studies of CG6 and CG10 strains showed that the CryIVD protein production can be observed within 24 h (Fig. 5, lanes 4 and 9).
IS231-like transposase role in CryIVD and CytA protein production. An open reading frame encoding an IS231-like transposase downstream of the 20-kDa protein gene was reported in B. thunngiensis subsp. israelensis (1). The potential role of the transposase-like fragment in B. thuiin-
- 7 2 1 t D E~7
S1 2 05X03L6 7i=00 FIG. 3. SDS-PAGE analysis of the 20-kDa protein influence on
cryIVD gene expression. Each B. thuringiensis strain was incubated at 30°C for 3 days, and the lysates were then centrifuge at 15,000 x g for 10 min. The pellets were brought to the original volume with H20. Samples (10 p1) from either the supernatant or pellet fractions were used for electrophoresis. Lanes 1 to 4 were loaded with samples from the pellets, and lanes 5 to 8 were loaded samples from the supernatant fractions. Lanes: 1 and 5, CG4; 2 and 6, CG6; 3 and 7, CG10; 4 and 8, cryB; S, standard proteins. The protein band in lane 8 is near the dye front and has a molecular size of <20 kDa. The 72-kDa marker is indicated on the right.
giensis subsp. mornisoni PG-14 was evaluated by using the B. thuringiensis expression system because in some crude bacterial cultures increased CryIVD levels were observed. Two plasmids, pCG5 and pCG8, were constructed, with the latter having the DNA encoding the putative transposase fragment; SDS-PAGE and immunoblot analyses showed that there was no difference in CryIVD and CytA levels in these two constructs (Fig. 2, lanes 4 and 6). Temporal expression of the cryIVD and cytA genes. Expres-
sion of the cryIVD and cytA genes was monitored over a 72-h period. Low-level expression was observed within 24 h, with high levels apparent within 48 h (Fig. 6). CryIVD and CytA protein levels did not increase further after a 96-h incuba- tion. The time course studies showed that CryIVD and CytA are synthesized simultaneously. However, since each pro- tein is synthesized independently (Fig. 2 and 4), it appears that their genes are independently regulated.
Mosquitocidal activity and inclusion shape. All B. thunn- giensis subsp. kurstaki strains transformed with the pCG series form parasporal inclusions during sporulation. Under light microscopy, the inclusions formed by CG4 and CG6 were similar to those observed for CG1 (4). However, the inclusions formed by CG2, CG8, and CG17 are large and ovoid in shape. Thus, high CytA levels result in an inclusion formation whose shape is different from that observed in wild-type B. thuringiensis subsp. israelensis. Both the crude bacterial cultures and the purified inclu-
sions from these strains are toxic to C. quinquefasciatus larvae (Fig. 1; Table 1). Parasporal inclusions obtained from CG strains (CG1, CG4, and CG6), which produced only the CryIVD protein, all had similar LC50 values (Table 1).
818 CHANG ET AL.
http://aem .asm
.org/ D
kDa Q-7
1 234 56 788
FIG. 4. Immunoblot analysis of the 20-kDa protein influence on cryBIVD gene expression. B. thuringiensis strains were incubated at 30'C for 3 days, and the lysates were then centrifuge at 15,000 x g for 10 min. The pellets were brought to the original volume with H20. Supernatants (5 p.1) and pellet fractions (2 p.l) were used. Rabbit antiserum developed against B. thuringiensis subsp. israe- lensis CrylVD protein was used for detection. Lanes 1 to 4 were loaded with samples from the pellets and lanes 5-8 were loaded with samples from the supernatant fractions. Lanes: 1 and 5, CG4; 2 and 6, CG6; 3 and 7, CG10; 4 and 8, cryB; S, standard proteins. The prominent 30- to 34-kDa bands in lanes 5 and 6 are CryIVD proteolytic products (5). Molecular size markers are indicated on the right.
Strains containing both CytA and CryIVD have the highest mosquitocidal activity, while strains containing only CytA are the least mosquitocidal (Table 1). A densitometer scan of the proteins produced in the CG8 strain (Fig. 2) showed that the CytA and CryIVD proteins were produced in a ratio of 7:3. Since the CytA protein has little mosquitocidal activity, the LC50 value that is observed for pCG8 parasporal inclu- sions therefore results from ca. 11 ng of the CrylVD protein per ml.
DISCUSSION
Adams et al. (1) reported that the 20-kDa protein is essential for efficient CytA production in E. coli. No CytA is detected in the absence of the 20-kDa protein, and CytA accumulation in E. coli is due to its stabilization by the 20-kDa protein (32). Stabilization apparently occurs because the CytA is protected from proteolysis in E. coli. In the CG13 strain, which contains only cytA4, CytA inclusions were formed. Some of the CytA protein produced in CG13 is, however,, processed to a 24-kDa peptide. Similarly, in the CG2 strain, which contains cytA and a 3' truncated cryIVD but lacks the 20-kDa protein gene,, large CytA oval inclu- sions are detected and high CytA levels are observed. Therefore, the 20-kDa protein is not essential for CytA protein production in B. thuringiensis. However, the pres- ence of a second protein, either the 20-kDa protein or CrylVD, enhances the stability of CytA in an acrystallifer- ous B. thwingiensis subsp. kurstaki strain,, cryB. Armstrong et al. (3) reported that the 24-kDa peptide is the
- 72 kDa
FIG. 5. Results of time course studies showing the 20-kDa pro- tein influence on cryIVD gene expression by immunoblot analyses. Rabbit antiserum developed against the B. thuringiensis subsp. israelensis CryIVD protein was used for detection. Lanes: 1, 250 ng of pure B. thuringiensis subsp. israelensis CryIVD protein; 2 to 6, cryIVD gene expression in strain CG6 at 6, 12, 24, 48, and 72 h, respectively; 7 to 11, cryIVD gene expression in strain CGlO at 6, 12, 24, 48, and 72 h, respectively; 12, cryB. One microgram of total protein was loaded in each lane. S, standard proteins. The promi- nent 30- to 34-kDa bands are CryIVD proteolytic products. The 72-kDa marker is indicated on the right.
active form of the CytA toxin. Further, Ward et al. (34) showed that the residues Arg-25 and Arg-30 are important for the CytA inclusion formation in E. coli. These two residues are removed when the 27-kDa CytA protein is proteolytically cleaved to a 24-kDa protein (17). Conse- quently, the 24-kDa protein produced in CG13 is probably not readily packaged into the inclusion body, and hence, the soluble 24-kDa protein could be cytotoxic to the bacterial cell (11); thus, CytA levels are lower. However, if the CytA protein is continuously packaged, this protein synthesis continues, resulting in high cytA gene expression as ob- served in CG2. Strain CG13 does not autolyze; after 6 days of incubation at 30°C, the inclusion body is still maintained in the cell. This lack of autolysis could be due to the cleaved CytA protein cytotoxicity, or, alternatively, the presence of either CryIVD or the 20-kDa protein is required for autolysis because all strains that contained either gene encoding these proteins autolyzed after sporulation. CryIVD levels are low in the CG1O strain, which lacks the
20-kDa gene and the inverted repeats normally present 3' to the cryIVD coding region. In B. thuringiensis, the presence of inverted repeats following a structural gene increases the stability of the mRNA (10, 25, 33, 37). The high CryIVD levels in CG4, which contains only a truncated 20-kDa gene and cryIVD, are comparable to those observed in the wild- type bacterial strain and support the possible role of the invert repeats. Alternatively, the absence of an additional protein, such as the 20-kDa protein, contributes to the low CryIVD yield in CG10. Hence, in CG4 either the 20-kDa protein is not required for CryIVD production or the 20-kDa truncated gene product maintains its activity even though ca. 50% of the 20-kDa gene was deleted, although the mainte- nance of its biological function is unlikely.
Therefore, it appears that in acrystalliferous B. thuring-
VOL. 59, 1993
http://aem .asm
.org/ D
$ 1 2 3 4 5 6 7 8 9 101112
FIG. 6. Temporal expression of the cytA and cryID genes. Lanes 1 to 6 were probed with rabbit antiserum developed against the intact B. thuringiensis subsp. israelensis parasporal inclusion, while lanes 7 to 12 were probed with rabbit antiserum developed against the CryIVD protein of B. thuringiensis subsp. israelensis. Lanes: 1 and 7, 24-h crude culture from strain CG6; 2 and 8, 48-h culture from strain CG6; 3 and 9, 96-h culture from strain CG6; 4 and 10, 24-h culture from strain CG8; 5 and 11, 48-h culture from strain CG8; 6 and 12, 96-h culture from strain CG8. In each well, 2 ,ul of crude culture was loaded. S, standard proteins. Molecular size markers are indicated on the right.
iensis subsp. kurstaki, the presence of an additional protein, whether it is CryIVD or the 20-kDa protein, results in the stabilization of CytA. Similarly, higher CryIVD levels are observed in the presence of a 20-kDa protein or truncated 20-kDa and CytA proteins. In the absence of a second protein, the protein levels observed are lower. Likewise, the presence of an additional protein, the ORF2 gene product, also appears to be critical for high-level stable expression of CryIIA (7, 36). The role of a second protein in high-level cry and cyt gene
expression is uncertain. Potentially, these proteins, even
TABLE 1. Mosquitocidal toxicity of the CytA and CryIVD purified parasporal inclusions to fourth-instar
C quinquefasciatus larvae
Strain Mosquitocidal toxicity' LC5o (ng/ml) LC95 (ng/ml)
CG1 39.7 (33.9-46.5) 184 (140-266) CG2 301 (227-443) 3,980 (1,960-12,600) CG4 37.3 (31.2-44.2) 150 (115-216) CG6 36.5 (29.4-44.5) 204 (151-308) CG8 25.6 (21.6-30.2) 91.9 (71.3-132) CG13 > 1,000 NDb
a Values in parentheses represent the fiducial limits at LC50 and LC95 levels.
b ND, not determined.
small ones as observed in CG4, can apparently protect the CytA, CryIVD, and CryIIA proteins from protease cleavage prior to parasporal inclusion body formation. Alternatively, this additional protein could provide the matrix or scaffold- ing for CytA, CryIIA, and CryIVD packaging and parasporal inclusion formation. Time course studies show that both CytA and CryIVD are
detected by 24 h of incubation at 30°C. The two toxin genes encoding these proteins are thus expressed at almost the same time; however, it is not known whether these two genes are coordinately regulated. Further, CytA levels higher than CryIVD levels are observed. The relative amounts of these two proteins are similar to those found in the intact inclusion body of B. thuringiensis subsp. israelen- sis. A large ovoid inclusion is formed by bacterial strains
producing CytA. This shape differs from that found in B. thuringiensis subsp. israelensis, suggesting that the large parasporal inclusion in B. thuringiensis subsp. israelensis is an aggregate of the CryIVA, CryIVB, and CytA proteins, while CryIVD is packaged separately (4). Inclusions from each B. thuringiensis strain except CG13 are highly toxic to C. quinquefasciatus larvae. A synergistic effect was ob- served for B. thuringiensis strains which produced both CytA and CryIVD. CytA, which has low mosquitocidal activity, synergizes CryIVD toxicity by about four- to five- fold (30). This study demonstrates that if both CytA and CryIVD are present in the same inclusion a synergistic interaction is observed. These results are in agreement with published reports which state that purified CytA synergizes CryIVD mosquitocidal activity (5, 20, 38, 39). The mecha- nism by which CytA synergizes CryIVD mosquitocidal activity is unknown. Potentially, the cell membrane aggre- gates formed by CytA (6) could facilitate increased CryIVD interaction with target cell membranes or, alternatively, expedite translocation or transportation of an active CryIVD moiety to its target. However, the mosquitocidal activity of CytA and CryIVD
together, with an LC50 of 26 ng/ml to C. quinquefasciatus, is an order of magnitude lower than that observed with intact parasporal inclusions from B. thuringiensis subsp. israelen- sis, with an LC50 of 3.8 ng/ml (4). Consequently, the other CryIV proteins also contribute significantly to the intact B. thuringiensis subsp. israelensis parasporal inclusion toxicity to Culex species.
ACKNOWLEDGMENTS
This work was supported in part by grants from NIH ES03298, the University of California System-wide Biotechnology Research and Education Program, and the University of California Mosquito Control Research Program.
REFERENCES 1. Adams, L. F., J. E. Visick, and H. R. Whiteley. 1989. A
20-kilodalton protein is required for efficient production of the Bacillus thuringiensis subsp. israelensis 27-kilodalton crystal protein in Escherichia coli. J. Bacteriol. 171:521-530.
2. Angsuthanasombat, C., N. Crickmore, and D. J. Ellar. 1992. Comparison of Bacillus thuringiensis subsp. israelensis CryIVA and CryIVB cloned toxins reveals synergism in vivo. FEMS Microbiol. Lett. 94:63-68.
3. Armstrong, J. L., G. F. Rohrmann, and G. S. Beaudreau. 1985. Delta-endotoxin of Bacillus thuringiensis subsp. israelensis. J. Bacteriol. 161:39-46.
4. Chang, C., S. M. Dai, R. Frutos, B. A. Federici, and S. S. Gill. 1992. Properties of a 72-kilodalton mosquitocidal protein from Bacillus thuringiensis subsp. morrisoni PG-14 expressed in B.
820 CHANG ET AL.
http://aem .asm
.org/ D
thuringiensis subsp. kurstaki by using the shuttle vector pHT3101. Appl. Environ. Microbiol. 58:507-512.
5. Chilcott, C. N., and D. J. Ellar. 1988. Comparative toxicity of Bacillus thuningiensis var. israelensis crystal proteins in vivo and in vitro. J. Gen. Microbiol. 134:2551-2558.
6. Chow, E., G. J. P. Singh, and S. S. Gill. 1989. Binding and aggregation of the 25-kilodalton toxin of Bacillus thuingiensis subsp. israelensis to cell membranes and alteration by mono- clonal antibodies and amino acid modifiers. Appl. Environ. Microbiol. 55:2779-2788.
7. Crickmore, N., and D. J. Ellar. 1992. Involvement of a possible chaperonin in the efficient expression of a cloned CryIIA f-endotoxin gene in Bacillus thuningiensis. Mol. Microbiol. 6:1533-1537.
8. Delecluse, A., C. Bourgouin, A. Klier, and G. Rapoport. 1988. Specificity of action on mosquito larvae of Bacillus thuringiensis israelensis toxins encoded by two different genes. Mol. Gen. Genet. 214:42-47.
9. Delecluse, A., J.-K. Charles, A. Klier, and G. Rapoport. 1991. Deletion by in vivo recombination shows that the 28-kilodalton cytolytic polypeptide from Bacillus thuringiensis subsp. israe- lensis is not essential for mosquitocidal activity. J. Bacteriol. 173:3374-3381.
10. Donovan, W. P., C. Dankocsik, and M. Pearce-Gilbert. 1988. Molecular characterization of a gene encoding a 72-kilodalton mosquito-toxic crystal protein from Bacillus thuringiensis subsp. israelensis. J. Bacteriol. 170:4732-4738.
11. Douek, J., M. Einav, and A. Zaritsky. 1992. Sensitivity to plating of Escherichia coli cells expressing the cryA gene from Bacillus thuningiensis var. israelensis. Mol. Gen. Genet. 232: 162-165.
12. Earp, D. J., E. S. Ward, and D. J. Ellar. 1987. Investigation of possible homologies between crystal proteins of three mosqui- tocidal strains of Bacillus thuringiensis. FEMS Microbiol. Lett. 42:195-199.
13. Frutos, R., C. Chang, S. S. Gill, and B. A. Federici. 1991. Nucleotide sequence of genes encoding a 72-kDa mosquitocidal protein and an associated 20-kDa protein are highly conserved in subspecies of Bacillus thuringiensis from Israel and The Phillipines. Biochem. Syst. Ecol. 19:599-609.
14. Galjart, N. J., N. Sivasubramanian, and B. A. Federici. 1987. Plasmid location, cloning, and sequence analysis of the gene encoding a 27.3-kilodalton cytolytic protein from Bacillus thur- ingiensis subsp. momrisoni (PG-14). Curr. Microbiol. 16:171-177.
15. Gill, S. S., E. A. Cowles, and P. Pietrantonio. 1992. The mode of action of Bacillus thuingiensis endotoxins. Annu. Rev. Ento- mol. 37:615-636.
16. Gill, S. S., J. M. Hornung, J. E. Ibarra, G. J. P. Singh, and B. A. Federici. 1987. Cytolytic activity and immunological similarity of the Bacillus thuringiensis subsp. israelensis and Bacillus thuringiensis subsp. morrisoni isolate PG-14 toxins. Appl. En- viron. Microbiol. 53:1251-1256.
17. Gill, S. S., G. J. P. Singh, and J. M. Hornung. 1987. Cell membrane interaction of Bacillus thuringiensis subsp. israelen- sis cytolytic toxins. Infect. Immun. 55:1300-1308.
18. Goldberg, L. J., and J. A. Margalit. 1977. Bacterial spore demonstrating rapid larvicidal activity against Anopheles ser- gentii, Uranotaenia unguiculata, Culex univitattus, Aedes ae- gypti and Culexpipiens. Mosq. News 37:355-358.
19. Hofte, H., and H. R. Whiteley. 1989. Insecticidal crystal proteins of Bacillus thuringiensis. Microbiol. Rev. 53:242-255.
20. Ibarra, J. E., and B. A. Federici. 1986. Isolation of a relatively nontoxic 65-kilodalton inclusion from the parasporal body of Bacillus thuwingiensis subsp. israelensis. J. Bacteriol. 165:257-533.
21. Laemmli, U. K. 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature (London) 227:680-685.
22. Lereclus, D., S. 0. Arant, J. Chaufaux, and M.-M. Lecadet. 1989. Transformation and expression of a cloned delta-endo- toxin gene in Bacillus thuringiensis. FEMS Microbiol. Lett. 60:211-217.
23. Lowry, 0. H., N. J. Rosebrough, A. L. Farr, and R. J. Randall. 1951. Protein measurement with the Folin phenol regent. J. Biol. Chem. 193:265-275.
24. McLean, K. M., and H. R. Whiteley. 1987. Expression in Escherichia coli of a cloned crystal protein gene of Bacillus thunngiensis subsp. israelensis. J. Bacteriol. 169:1017-1023.
25. Newbury, S. F., N. H. Smith, E. C. Robinson, I. D. Hiles, and C. F. Higgins. 1987. Stabilization of translationally active mRNA by prokaryotic REP sequences. Cell 48:297-310.
26. Padua, L. E., and B. A. Federci. 1990. Development of mutants of the mosquitocidal bacterium Bacillus thuringiensis subsp. morrisoni (PG-14) toxic to lepidopterous or dipterous insects. FEMS Microbiol. Lett. 66:257-262.
27. Padua, L. E., M. Ohba, and K. Aizawa. 1984. Isolation of a Bacillus thuringiensis strain (serotype 8a: 8b) highly and selec- tively toxic against mosquito larvae. J. Invertebr. Pathol. 44: 12-17.
28. Raymond, M. 1985. Presentation d'un programme basic d'anal- yse log-probit pour micro-ordinateur. Cah. ORSTROM Ser. Entomol. Med. Parasitol. 23:117-121.
29. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: a laboratory manual, 2nd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y.
30. Tabashnik, B. E. 1992. Evaluation of synergism among Bacillus thuringiensis toxins. Appl. Environ. Microbiol. 58:3343-3346.
31. Thomas, W. E., and D. J. Ellar. 1983. Mechanism of action of Bacillus thuringiensis var israelensis insecticidal delta-endo- toxin. FEBS Lett. 154:362-368.
32. Visick, J. E., and H. R. Whiteley. 1991. Effect of a 20-kilodalton protein from Bacillus thuringiensis subsp. israelensis on pro- duction of the CytA protein by Escherichia coli. J. Bacteriol. 173:1748-1756.
33. Ward, E. S., and D. J. Ellar. 1988. Cloning and expression of two homologous genes of Bacillus thuringiensis subsp. israelen- sis which encode 130-kilodalton mosquitocidal proteins. J. Bac- teriol. 170:727-735.
34. Ward, E. S., D. J. Ellar, and C. N. Chilcott. 1988. Single amino acid changes in the Bacillus thuringiensis var. israelensis delta- endotoxin affect the toxicity and expression of the protein. J. Mol. Biol. 202:527-535.
35. Ward, E. S., A. R. Ridley, D. J. Ellar, and J. A. Todd. 1986. Bacillus thuringiensis var. israelensis delta-endotoxin: cloning and expression of the toxin in sporogenic and asporogenic strains of Bacillus subtilis. J. Mol. Biol. 191:13-22.
36. Widner, W. R., and H. R. Whiteley. 1989. Two highly related insecticidal crystal proteins of Bacillus thuringiensis subsp. kurstaAi possess different host range specificities. J. Bacteriol. 171:965-974.
37. Wong, H. C., and S. Chang. 1986. Identification of a positive retroregulator that stabilizes mRNAs in bacteria. Proc. Natl. Acad. Sci. USA 83:3233-3237.
38. Wu, D., and F. N. Chang. 1985. Synergism in mosquitocidal activity of 26 and 65 kDa proteins from Bacillus thuringiensis subsp. israelensis crystal. FEBS Lett. 190:232-236.
39. Yu, Y.-M., M. Ohba, and K. Aizawa. 1987. Synergistic effects of the 65- and 25-kilodalton proteins of Bacillus thuringiensis strain PG-14 (serotype 8A:8B) in mosquito larvicidal activity. J. Gen. Appl. Microbiol. 33:459-462.
40. Yu, Y.-M., M. Ohba, and S. S. Gill. 1991. Characterization of mosquitocidal activity of Bacillus thuringiensis subsp. fukuo- kaensis crystal proteins. Appl. Environ. Microbiol. 57:1075- 1081.
VOL. 59, 1993
http://aem .asm
.org/ D
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 CHENG CHANG,1 YONG-MAN yU,2 SHU-MEI DAI,2 SARA K. LAW,2
AND SARJEET S. GILL1' 2*
Department of Entomology2 and Interdepartmental Graduate Program in Environmental Toxicology, 1 University of California, Riverside, California 92521
Received 31 August 1992/Accepted 8 December 1992
Interactions among the 20-kDa protein gene and the cyt4 and cryIVD genes located in a 9.4-kb HindIll fragment were studied. A series of plasmids containing a combination of these different genes was constructed by using the Escherichia colil/Bacilus thuringiensis shuttle vector pHT3101. The plasmids were then used to transform an acrystalliferous strain, cryB, derived from B. thuringiensis subsp. kurstaki. The results from sodium dodecyl sulfate-polyacrylamide gel electrophoresis and immunoblot analyses suggest that although the 20-kDa protein is required for the efficient CytA protein production in E. coli, it is not required in B. thuringiensis. With or without the truncated 20-kDa protein gene, the CytA and/or Cry1VD proteins are
produced and form parasporal inclusions in B. thuringiensis cells. However, more-efficient expression is obtained when a second protein, probably acting as a chaperonin, is present. In addition, the time course
studies show that the CytA and CrylVD proteins are coordinately produced. Both the crude B. thuringiensis culture and purified inclusions from each recombinant B. thuringiensis strain are toxic to Culex quinquefas- ciatus larvae. The parasporal inclusions formed in B. thuringiensis cells are mosquitocidal, with CytA synergizing CryIVD toxicity.
Bacillus thuringiensis subsp. israelensis and B. thurin- giensis subsp. morrisoni (PG-14) both produce spherical parasporal inclusions that are highly toxic to dipteran larvae, such as mosquitoes and black flies (18, 27). The parasporal inclusions from both subspecies produce 27-, 72-, 125-, and 135-kDa polypeptides, the CytA, CryIVD, CryIVB, and CryIVA proteins, respectively (12-15, 19). The B. thurin- giensis subsp. morrisoni parasporal inclusions also contain a
144-kDa protein not found in B. thuringiensis subsp. israe- lensis (26). The genes encoding the CytA, CryIVA, CryIVB, and CryIVD proteins have been cloned and expressed in both Escherichia coli and B. thuringiensis (1, 4, 8, 9, 13, 24, 33, 35). All proteins common to both subspecies are mos-
quitocidal, with the Cry proteins having the greatest insec- ticidal activity (2, 4, 8, 10, 33). The CytA protein, in addition to its mosquitocidal activity (34), is also hemolytic and cytolytic (17, 31). The insecticidal activity of B. thuringiensis subsp. israe-
lensis is a complex interaction of the four inclusion body proteins, CryIVA, CryIVB, CryIVD, and CytA (15). Among these proteins, CytA exclusion from inclusion bodies has relatively little effect on insecticidal activity (9). Although each protein is mosquitocidal, none is as active as the intact parasporal inclusion. The higher insecticidal activity of the intact parasporal inclusions is primarily due to the synergis- tic interaction between the proteins present in these inclu- sions (2, 5, 8, 20, 38, 39).
In addition to the synergistic interaction between CytA and CryIVD, these proteins are encoded by the same
HindIII fragment which also encodes a 20-kDa protein. In
* Corresponding author.
E. coli, CytA levels are increased because of posttransla- tional stabilization by this 20-kDa protein (1, 32). It is not known, however, whether this 20-kDa protein plays any role in CytA or CryIVD levels in B. thuingiensis. To investigate the role of the 20-kDa protein, a recombinant system in B. thuringiensis (4) was used. We report here that the 20-kDa protein is not required for the production of high levels of either CytA or CryIVD in B. thurngiensis. However, the presence of an additional protein increases the CytA and CryIVD levels in an acrystalliferous strain of B. thuringien- sis. Mosquitocidal activity evaluations also demonstrate that CytA synergizes the insecticidal activity of CryIVD to Culex quinquefasciatus.
MATERIALS AND METHODS
Bacterial strains, plasmids, and general methods. Bacterial strains used in this study were E. coli JM101, JM109, and XL-1 (Stratagene, La Jolla, Calif.), BMH 71-18 mutS (Promega, Madison, Wis.), and the acrystalliferous strain cryB, derived from B. thuringiensis subsp. kurstaki HD1, obtained from A. Aronson, Department of Biology, Purdue University. The plasmid pMl, which encodes the B. thur- ingiensis subsp. morrisoni (PG-14) CytA, CryIVD, and 20-kDa proteins, was obtained from B. Federici and S. Sivasubramanian of our department (14). The E. coliIB. thuringiensis shuttle vector pHT3101 (22) was obtained from D. Lereclus, Institut Pasteur, Paris, France. Site-directed mutagenesis was performed with pSELECT-1 by using the Altered Sites in vitro mutagenesis system (Promega) follow- ing the manufacturer's instructions while all other cloning was performed with pBluescriptIl SK+ (Stratagene). Stan- dard protocols were used for restriction enzyme digestion,
815
http://aem .asm
.org/ D
ligation, and E. coli transformation (29). B. thuringiensis cryB cells were transformed by electroporation essentially as described previously (22).
Construction of the pCG plasmid series. The construction of pCG1 has been described previously (4). For pCG2 construction, pM1 was restricted with EcoRI. The 5.7-kb fragment, which contains cytA and a truncated cryIVD gene, was separated by agarose gel electrophoresis and purified by Geneclean (Bio 101, La Jolla, Calif.). The purified fragment was ligated into the unique EcoRI site of pHT3101 and then used for transformation in E. coli JM101. Transformants, which were selected on Luria-Bertani agar containing ampi- cillin (50 ,ug/ml), 20 mM isopropyl-p-D-thiogalactopyrano- side (IPTG), and 80 mg of 5-bromo-4-chloro-3-indolyl-p3-D- galactopyranoside (X-Gal) per ml were isolated, and the sizes and orientations of the EcoRI inserts were determined. To construct pCG4, the cryIVD EcoRI site was deleted by
site-directed mutagenesis with the Altered Sites in vitro mutagenesis system. Briefly, the 5.3-kb BamHI-PstI frag- ment from pCG1 which contains the cryIVD gene was inserted into the BamHI-PstI sites of pSELECT-1. The cryIVD internal EcoRI site was then deleted by primer- directed mutagenesis. The mutant, identified by an EcoRI site deletion, was then restricted with EcoRI. The 4.4-kb EcoRI fragment which contains cryIVID, a truncated cytA gene, and a truncated 20-kDa protein gene was then inserted into the unique EcoRI site of pHT3101, resulting in pCG4. For pCG5 construction, the 4.9-kb pMl BamHI-ClaI frag- ment was subcloned into pBluescriptII SK+. The pCG2 2.0-kb SacI-BamHI fragment was then inserted into SacI- BamHI sites, giving pCGB12. The 6.9-kb pCGB12 KpnI- SacI fragment which contains the cryIID, cytA, and 20-kDa protein genes was then inserted into the shuttle vector pHT3101, resulting in the plasmid pCG5. The 4.4-kb pM1 EcoRV-PstI fragment was isolated and
ligated with a SacI linker, and the resulting fragment was inserted into the unique SacI-PstI sites of pHT3101, result- ing in the construction of pCG6. To obtain pCG8, the 6.9-kb SacI-Clal fragment from pCG5 was ligated with the 7.2-kb SacI-ClaI fragment of pCG1. To construct pCG10, the 4.9-kb pCG6 SpeI-BamHI fragment, which lacks the B. thuringiensis origin of replication but retains the E. coli replication origin, was isolated. This fragment was then ligated with the 1.2-kb SpeI-BamHI fragment of pCGE1, which contains the 3' cryIVD fragment but lacks the 20-kDa protein gene, resulting in the formation of pCG9; the BamHI site in pCGE1 was introduced by the polymerase chain reaction. For expression in B. thunngiensis, the 3.6-kb pHT3101 BamHI fragment containing the B. thunngiensis replication origin was inserted into the unique pCG9 BamHI site; this new construct, pCG10, contains only the cryIVD gene. The plasmid pCG12 was constructed by replacing the
6.9-kb SacI-ClaI fragment of pCG5, which contains all three genes, with the 4.0-kb SacI-ClaI fragment of pCG6, which contains cryIVD and the gene encoding the 20-kDa protein. For construction of pCG13, the 2.9-kb SacI-EcoRV frag- ment of pCG5, which contains cytA, was isolated from the agarose gel, and XbaI 10-mer linkers were used to modify the blunt end. After restriction, the SacI-XbaI fragment was cloned into the corresponding restriction sites of pHT3101. For construction of pCG17, the 0.93-kb pCGB12 HaeIII-
ClaI fragment, containing the 20-kDa protein gene and a 32-bp 3' end of cryIVD, was used to replace the 3.98-kb fragment containing the 20-kDa protein gene and cryIVD. The resulting construct was then cloned into pHT3101 to
obtain pCG13, which contains cytA, 32 bp of cryIVD, and the 20-kDa protein gene.
cyt4 and cryIVD gene expression. B. thunngiensis subsp. kurstaki recombinant CG strains containing the pCG plas- mids were analyzed for the CytA and CryIVD proteins. These B. thuningiensis subsp. kurstaki transformants were grown on nutrient agar plates containing 50 p,g of erythro- mycin per ml for 3 days at 30°C. The bacterial cultures were isolated by washing the culture plates with deionized water. Crude culture aliquots were boiled in sample treatment buffer and analyzed by discontinuous sodium dodecyl sulfate polyacrylamide gel electrophoresis, (SDS-PAGE) with 4.5% acrylamide (pH 6.8) and 10% acrylamide (pH 8.8) as the stacking and separating gels, respectively (21). The gels were then stained with 0.1% Coomassie blue R-250. Alternatively, the proteins resolved by SDS-PAGE were transferred to nitrocellulose for 16 h at a constant current of 250 mA, and the nitrocellulose was then probed with rabbit antibody developed against either the purified whole parasporal inclu- sion or the purified 72-kDa toxin of B. thuringiensis subsp. israelensis by using methods described previously (16). Goat anti-rabbit immunoglobulin G-alkaline phosphatase was used as the second antibody, and chromogenic development was then achieved with nitroblue tetrazolium chloride (1 mg/ml in H20) and 5-bromo-4-chloro-3-indolyl phosphate (5 mg/ml in dimethylformamide). For time course studies, cultures were terminated at 12,
24, 36, 48, and 72 h. The plate bacterial cultures were isolated with 10 mM EDTA and sedimented by centrifuga- tion at 15,000 x g for 10 min. After the protein concentration was estimated (23), the cultures were analyzed by discontin- uous SDS-PAGE and immunoblotting as described above.
Purification of parasporal inclusions. B. thunngiensis subsp. kurstaki CG strains were cultured on nutrient agar plates containing 50 ,ug of erythromycin per ml for 5 days at 30°C to ensure sporulation and complete autolysis. The spore-parasporal inclusion mixture was thoroughly washed with 1 M NaCl-10 mM EDTA and sedimented by centrifu- gation at 15,000 x g for 10 min. The pellet was resuspended in water, sonicated, loaded onto a continuous 40 to 70% Renografin density gradient, and centrifuged at 15,000 rpm for 30 min in an SW28 rotor as described previously (40). The parasporal inclusions were then subjected to a second centrifugation. The purified parasporal inclusions were washed three times with distilled water, the protein concen- tration was measured (23), and the inclusions were stored at 40C.
Insect bioassays. The larval mosquitocidal activity of each B. thuringiensis CG strain was determined. For determina- tion of crude bacterial culture toxicity, the cultures were serially diluted and added to distilled water (total volume, 10 ml) containing 10 fourth-instar C. quinquefasciatus larvae. The 24-h mortality was determined by counting the number of surviving larvae. Bioassays were performed in triplicate. The mosquitocidal activity of purified inclusions from the
B. thuringiensis subsp. kurstaki CG series was also assessed. Briefly, 0.1 ml of parasporal inclusion dilutions was added to 99.9 ml of distilled water containing 20 fourth-instar C. quinquefasciatus larvae. To determine a 50% lethal concen- tration (LC50), 10 different inclusion concentrations (1 to 1,000 ng/ml) were used, with at least four replicates per concentration. The 24-h mortality was determined, and LC50 and 95% lethal concentration (LC95) values were calculated by probit analyses (28). Controls for the bioassays utilized mosquito larvae reared similarly but not exposed to either
816 CHANG ET AL.
http://aem .asm
.org/ D
cytA ciyIVD
E
FIG. 1. Construction of the pCG plasmid series and determination of cryIVD and cytA gene expression in B. thuningiensis (Bt) and the mosquitocidal activity of the gene products obtained. B, BamHI; C, ClaI; E, EcoRI; EV, EcoRV; H, HindIII; P, PstI; Sa, SacI; X, XbaI. + + +, + +, and + indicate that 100, >50, or <50% mortality, respectively, was observed in the mosquito larva bioassay with crude bacterial cultures.
the crude bacterial cultures or the purified parasporal inclu- sions.
RESULTS
To determine the interactions between the CytA, CryIVD, and the 20-kDa proteins or their genes, several DNA frag- ments derived from the 9.4-kb HindIII fragment ofpMl were introduced into the shuttle vector pHT3101 for expression in B. thuringiensis (Fig. 1). The CytA and CryIVD protein levels were detected by SDS-PAGE analysis (Fig. 2). Immu- noblot analysis confirmed the identity of the CytA and CryIVD proteins (data not shown). In all experiments, the 20-kDa protein was not observed in the acrystalliferous strain cryB.
Role of the 20-kDa protein in cyt4 gene expression. The 20-kDa protein is reportedly required for efficient CytA production in E. coli (1, 32). The plasmids pCG2 and pCG13 were therefore constructed to determine whether the 20-kDa protein is required for cytA gene expression in the strain cryB derived from B. thunngiensis subsp. kurstaki. Our results show that CG2 and CG13, cryB strains transformed with pCG2 and pCG13, respectively, expressed significant amounts of the 27-kDa CytA protein at levels readily detect- able by SDS-PAGE (Fig. 2). High CytA levels similar to those in CG2 were also observed in CG17 (data not shown). However, the CytA protein in CG13 is synthesized in lower
kDa
72
27
1 2 3 4 58a 7 9 10 81112131415161718
FIG. 2. SDS-PAGE analysis of cryIVD and cytA gene expression in B. thuningiensis subsp. kurstaki (cryB). Lanes 1 to 10 were loaded
with 10 pLg of total protein from each B. thuningiensis crude culture
(3 days at 30'C). Lanes 11 to 18 were loaded with 5 p.g of purified inclusions from each indicated B. thuringiensis strain. Lanes: 1 and
11, CG1; 2 and 12, CG2; 3 and 13, CG4; 4 and 14, CG5; 5 and 15,
CG6; 6 and 16, CG8; 7 and 17, CG10; 8 and 18, CG12; 9, CG13; 10,
cryB; S, Standard proteins. Molecular size markers are indicated on
the right.
http://aem .asm
.org/ D
APPL. ENVIRON. MICROBIOL.
amounts than in strains CG2 and CG17, and in addition, the CG13 CytA protein is proteolytically cleaved to a 24-kDa protein. Thus, the 27-kDa protein is less stable in CG13 than in CG2 and CG17. After 5 days of culture on nutrient agar plates at 30°C, strain CG2 formed large ovoid parasporal inclusions during sporulation. These ovoid inclusions were readily observed under phase-contrast microscopy and in quantities sufficient to form a distinct band when they were purified by Renografin gradients.
Since the EcoRI fragment used to construct the pCG2 contains a portion of the cryIVD gene and its promoter, immunoblot analyses were performed to detect the presence of the CryIVD (72-kDa) protein by using antibodies raised against the intact B. thuringiensis subsp. israelensis paras- poral inclusions. The truncated cryIVD gene in this con- struct does not have a translation stop codon, and the cryIVD and P-galactosidase genes are in frame; therefore, the fusion protein formed includes part of the ,B-galactosi- dase protein. A 74-kDa CryIVD-13-galactosidase fusion pro- tein containing a C-terminal deletion of 77 amino acids from the CryIVD protein but containing 98 amino acid residues from the P-galactosidase was produced and confirmed by immunoblot analysis (data not shown). Although strain CG13 can form inclusions, spore formation in this strain is aberrant and the bacterial cells do not autolyze after 6 days of culture at 30°C. The bacterial strain CG13 produces the 27-kDa protein and a 24-kDa proteolytic product (Fig. 2). However, because of the lack of autolysis in the CG13 strain, parasporal inclusions are difficult to purify from this strain. Consequently, strain CG17 was constructed and produced CytA levels comparable to the CytA levels in CG2. Our observations show that the 20-kDa gene product is not required for efficient CytA protein production in B. thurin- giensis, in contrast to that observed in E. coli.
B. thuringiensis CryIYD production in the absence of the 20-kDa protein. We previously obtained high cryIVD gene expression in B. thuringiensis by using the shuttle vector pHT3101 (4). However, this construct contained, in addition to cryl'D, a truncated cytA gene and the 20-kDa protein gene. Therefore, to determine whether the 20-kDa protein is involved in the CryIVD protein production, the plasmids pCG4, pCG6, and pCG1O were constructed (Fig. 1). Plasmid pCG4 contains cryIVD, a truncated cytA gene, and a trun- cated 20-kDa protein gene. The plasmid pCG6 has only cryIVD and a 20-kDa protein gene, while pCG10 has only cryIVD but lacks the invert repeats (10). These plasmids were then used for transformation of the strain cryB derived from B. thunngiensis subsp. kurstaki. The results of SDS-PAGE (Fig. 2) and immunoblot anal-
yses show that the cryIVD gene is expressed in all three bacterial strains, i.e., CG4, CG6, and CG10; but bacterial strain CG1O produces lower CryIVD levels than do strains CG4 and CG6 (Fig. 3 and 4). During SDS-PAGE and immunoblotting analyses, most of the CryIVD products were observed in the pellet fraction following centrifugation at 15,000 x g for 10 min, demonstrating that the CrylVD protein is mostly in the inclusion and that only a small portion is released into the supernatant after cell lysis. Time course studies of CG6 and CG10 strains showed that the CryIVD protein production can be observed within 24 h (Fig. 5, lanes 4 and 9).
IS231-like transposase role in CryIVD and CytA protein production. An open reading frame encoding an IS231-like transposase downstream of the 20-kDa protein gene was reported in B. thunngiensis subsp. israelensis (1). The potential role of the transposase-like fragment in B. thuiin-
- 7 2 1 t D E~7
S1 2 05X03L6 7i=00 FIG. 3. SDS-PAGE analysis of the 20-kDa protein influence on
cryIVD gene expression. Each B. thuringiensis strain was incubated at 30°C for 3 days, and the lysates were then centrifuge at 15,000 x g for 10 min. The pellets were brought to the original volume with H20. Samples (10 p1) from either the supernatant or pellet fractions were used for electrophoresis. Lanes 1 to 4 were loaded with samples from the pellets, and lanes 5 to 8 were loaded samples from the supernatant fractions. Lanes: 1 and 5, CG4; 2 and 6, CG6; 3 and 7, CG10; 4 and 8, cryB; S, standard proteins. The protein band in lane 8 is near the dye front and has a molecular size of <20 kDa. The 72-kDa marker is indicated on the right.
giensis subsp. mornisoni PG-14 was evaluated by using the B. thuringiensis expression system because in some crude bacterial cultures increased CryIVD levels were observed. Two plasmids, pCG5 and pCG8, were constructed, with the latter having the DNA encoding the putative transposase fragment; SDS-PAGE and immunoblot analyses showed that there was no difference in CryIVD and CytA levels in these two constructs (Fig. 2, lanes 4 and 6). Temporal expression of the cryIVD and cytA genes. Expres-
sion of the cryIVD and cytA genes was monitored over a 72-h period. Low-level expression was observed within 24 h, with high levels apparent within 48 h (Fig. 6). CryIVD and CytA protein levels did not increase further after a 96-h incuba- tion. The time course studies showed that CryIVD and CytA are synthesized simultaneously. However, since each pro- tein is synthesized independently (Fig. 2 and 4), it appears that their genes are independently regulated.
Mosquitocidal activity and inclusion shape. All B. thunn- giensis subsp. kurstaki strains transformed with the pCG series form parasporal inclusions during sporulation. Under light microscopy, the inclusions formed by CG4 and CG6 were similar to those observed for CG1 (4). However, the inclusions formed by CG2, CG8, and CG17 are large and ovoid in shape. Thus, high CytA levels result in an inclusion formation whose shape is different from that observed in wild-type B. thuringiensis subsp. israelensis. Both the crude bacterial cultures and the purified inclu-
sions from these strains are toxic to C. quinquefasciatus larvae (Fig. 1; Table 1). Parasporal inclusions obtained from CG strains (CG1, CG4, and CG6), which produced only the CryIVD protein, all had similar LC50 values (Table 1).
818 CHANG ET AL.
http://aem .asm
.org/ D
kDa Q-7
1 234 56 788
FIG. 4. Immunoblot analysis of the 20-kDa protein influence on cryBIVD gene expression. B. thuringiensis strains were incubated at 30'C for 3 days, and the lysates were then centrifuge at 15,000 x g for 10 min. The pellets were brought to the original volume with H20. Supernatants (5 p.1) and pellet fractions (2 p.l) were used. Rabbit antiserum developed against B. thuringiensis subsp. israe- lensis CrylVD protein was used for detection. Lanes 1 to 4 were loaded with samples from the pellets and lanes 5-8 were loaded with samples from the supernatant fractions. Lanes: 1 and 5, CG4; 2 and 6, CG6; 3 and 7, CG10; 4 and 8, cryB; S, standard proteins. The prominent 30- to 34-kDa bands in lanes 5 and 6 are CryIVD proteolytic products (5). Molecular size markers are indicated on the right.
Strains containing both CytA and CryIVD have the highest mosquitocidal activity, while strains containing only CytA are the least mosquitocidal (Table 1). A densitometer scan of the proteins produced in the CG8 strain (Fig. 2) showed that the CytA and CryIVD proteins were produced in a ratio of 7:3. Since the CytA protein has little mosquitocidal activity, the LC50 value that is observed for pCG8 parasporal inclu- sions therefore results from ca. 11 ng of the CrylVD protein per ml.
DISCUSSION
Adams et al. (1) reported that the 20-kDa protein is essential for efficient CytA production in E. coli. No CytA is detected in the absence of the 20-kDa protein, and CytA accumulation in E. coli is due to its stabilization by the 20-kDa protein (32). Stabilization apparently occurs because the CytA is protected from proteolysis in E. coli. In the CG13 strain, which contains only cytA4, CytA inclusions were formed. Some of the CytA protein produced in CG13 is, however,, processed to a 24-kDa peptide. Similarly, in the CG2 strain, which contains cytA and a 3' truncated cryIVD but lacks the 20-kDa protein gene,, large CytA oval inclu- sions are detected and high CytA levels are observed. Therefore, the 20-kDa protein is not essential for CytA protein production in B. thuringiensis. However, the pres- ence of a second protein, either the 20-kDa protein or CrylVD, enhances the stability of CytA in an acrystallifer- ous B. thwingiensis subsp. kurstaki strain,, cryB. Armstrong et al. (3) reported that the 24-kDa peptide is the
- 72 kDa
FIG. 5. Results of time course studies showing the 20-kDa pro- tein influence on cryIVD gene expression by immunoblot analyses. Rabbit antiserum developed against the B. thuringiensis subsp. israelensis CryIVD protein was used for detection. Lanes: 1, 250 ng of pure B. thuringiensis subsp. israelensis CryIVD protein; 2 to 6, cryIVD gene expression in strain CG6 at 6, 12, 24, 48, and 72 h, respectively; 7 to 11, cryIVD gene expression in strain CGlO at 6, 12, 24, 48, and 72 h, respectively; 12, cryB. One microgram of total protein was loaded in each lane. S, standard proteins. The promi- nent 30- to 34-kDa bands are CryIVD proteolytic products. The 72-kDa marker is indicated on the right.
active form of the CytA toxin. Further, Ward et al. (34) showed that the residues Arg-25 and Arg-30 are important for the CytA inclusion formation in E. coli. These two residues are removed when the 27-kDa CytA protein is proteolytically cleaved to a 24-kDa protein (17). Conse- quently, the 24-kDa protein produced in CG13 is probably not readily packaged into the inclusion body, and hence, the soluble 24-kDa protein could be cytotoxic to the bacterial cell (11); thus, CytA levels are lower. However, if the CytA protein is continuously packaged, this protein synthesis continues, resulting in high cytA gene expression as ob- served in CG2. Strain CG13 does not autolyze; after 6 days of incubation at 30°C, the inclusion body is still maintained in the cell. This lack of autolysis could be due to the cleaved CytA protein cytotoxicity, or, alternatively, the presence of either CryIVD or the 20-kDa protein is required for autolysis because all strains that contained either gene encoding these proteins autolyzed after sporulation. CryIVD levels are low in the CG1O strain, which lacks the
20-kDa gene and the inverted repeats normally present 3' to the cryIVD coding region. In B. thuringiensis, the presence of inverted repeats following a structural gene increases the stability of the mRNA (10, 25, 33, 37). The high CryIVD levels in CG4, which contains only a truncated 20-kDa gene and cryIVD, are comparable to those observed in the wild- type bacterial strain and support the possible role of the invert repeats. Alternatively, the absence of an additional protein, such as the 20-kDa protein, contributes to the low CryIVD yield in CG10. Hence, in CG4 either the 20-kDa protein is not required for CryIVD production or the 20-kDa truncated gene product maintains its activity even though ca. 50% of the 20-kDa gene was deleted, although the mainte- nance of its biological function is unlikely.
Therefore, it appears that in acrystalliferous B. thuring-
VOL. 59, 1993
http://aem .asm
.org/ D
$ 1 2 3 4 5 6 7 8 9 101112
FIG. 6. Temporal expression of the cytA and cryID genes. Lanes 1 to 6 were probed with rabbit antiserum developed against the intact B. thuringiensis subsp. israelensis parasporal inclusion, while lanes 7 to 12 were probed with rabbit antiserum developed against the CryIVD protein of B. thuringiensis subsp. israelensis. Lanes: 1 and 7, 24-h crude culture from strain CG6; 2 and 8, 48-h culture from strain CG6; 3 and 9, 96-h culture from strain CG6; 4 and 10, 24-h culture from strain CG8; 5 and 11, 48-h culture from strain CG8; 6 and 12, 96-h culture from strain CG8. In each well, 2 ,ul of crude culture was loaded. S, standard proteins. Molecular size markers are indicated on the right.
iensis subsp. kurstaki, the presence of an additional protein, whether it is CryIVD or the 20-kDa protein, results in the stabilization of CytA. Similarly, higher CryIVD levels are observed in the presence of a 20-kDa protein or truncated 20-kDa and CytA proteins. In the absence of a second protein, the protein levels observed are lower. Likewise, the presence of an additional protein, the ORF2 gene product, also appears to be critical for high-level stable expression of CryIIA (7, 36). The role of a second protein in high-level cry and cyt gene
expression is uncertain. Potentially, these proteins, even
TABLE 1. Mosquitocidal toxicity of the CytA and CryIVD purified parasporal inclusions to fourth-instar
C quinquefasciatus larvae
Strain Mosquitocidal toxicity' LC5o (ng/ml) LC95 (ng/ml)
CG1 39.7 (33.9-46.5) 184 (140-266) CG2 301 (227-443) 3,980 (1,960-12,600) CG4 37.3 (31.2-44.2) 150 (115-216) CG6 36.5 (29.4-44.5) 204 (151-308) CG8 25.6 (21.6-30.2) 91.9 (71.3-132) CG13 > 1,000 NDb
a Values in parentheses represent the fiducial limits at LC50 and LC95 levels.
b ND, not determined.
small ones as observed in CG4, can apparently protect the CytA, CryIVD, and CryIIA proteins from protease cleavage prior to parasporal inclusion body formation. Alternatively, this additional protein could provide the matrix or scaffold- ing for CytA, CryIIA, and CryIVD packaging and parasporal inclusion formation. Time course studies show that both CytA and CryIVD are
detected by 24 h of incubation at 30°C. The two toxin genes encoding these proteins are thus expressed at almost the same time; however, it is not known whether these two genes are coordinately regulated. Further, CytA levels higher than CryIVD levels are observed. The relative amounts of these two proteins are similar to those found in the intact inclusion body of B. thuringiensis subsp. israelen- sis. A large ovoid inclusion is formed by bacterial strains
producing CytA. This shape differs from that found in B. thuringiensis subsp. israelensis, suggesting that the large parasporal inclusion in B. thuringiensis subsp. israelensis is an aggregate of the CryIVA, CryIVB, and CytA proteins, while CryIVD is packaged separately (4). Inclusions from each B. thuringiensis strain except CG13 are highly toxic to C. quinquefasciatus larvae. A synergistic effect was ob- served for B. thuringiensis strains which produced both CytA and CryIVD. CytA, which has low mosquitocidal activity, synergizes CryIVD toxicity by about four- to five- fold (30). This study demonstrates that if both CytA and CryIVD are present in the same inclusion a synergistic interaction is observed. These results are in agreement with published reports which state that purified CytA synergizes CryIVD mosquitocidal activity (5, 20, 38, 39). The mecha- nism by which CytA synergizes CryIVD mosquitocidal activity is unknown. Potentially, the cell membrane aggre- gates formed by CytA (6) could facilitate increased CryIVD interaction with target cell membranes or, alternatively, expedite translocation or transportation of an active CryIVD moiety to its target. However, the mosquitocidal activity of CytA and CryIVD
together, with an LC50 of 26 ng/ml to C. quinquefasciatus, is an order of magnitude lower than that observed with intact parasporal inclusions from B. thuringiensis subsp. israelen- sis, with an LC50 of 3.8 ng/ml (4). Consequently, the other CryIV proteins also contribute significantly to the intact B. thuringiensis subsp. israelensis parasporal inclusion toxicity to Culex species.
ACKNOWLEDGMENTS
This work was supported in part by grants from NIH ES03298, the University of California System-wide Biotechnology Research and Education Program, and the University of California Mosquito Control Research Program.
REFERENCES 1. Adams, L. F., J. E. Visick, and H. R. Whiteley. 1989. A
20-kilodalton protein is required for efficient production of the Bacillus thuringiensis subsp. israelensis 27-kilodalton crystal protein in Escherichia coli. J. Bacteriol. 171:521-530.
2. Angsuthanasombat, C., N. Crickmore, and D. J. Ellar. 1992. Comparison of Bacillus thuringiensis subsp. israelensis CryIVA and CryIVB cloned toxins reveals synergism in vivo. FEMS Microbiol. Lett. 94:63-68.
3. Armstrong, J. L., G. F. Rohrmann, and G. S. Beaudreau. 1985. Delta-endotoxin of Bacillus thuringiensis subsp. israelensis. J. Bacteriol. 161:39-46.
4. Chang, C., S. M. Dai, R. Frutos, B. A. Federici, and S. S. Gill. 1992. Properties of a 72-kilodalton mosquitocidal protein from Bacillus thuringiensis subsp. morrisoni PG-14 expressed in B.
820 CHANG ET AL.
http://aem .asm
.org/ D
thuringiensis subsp. kurstaki by using the shuttle vector pHT3101. Appl. Environ. Microbiol. 58:507-512.
5. Chilcott, C. N., and D. J. Ellar. 1988. Comparative toxicity of Bacillus thuningiensis var. israelensis crystal proteins in vivo and in vitro. J. Gen. Microbiol. 134:2551-2558.
6. Chow, E., G. J. P. Singh, and S. S. Gill. 1989. Binding and aggregation of the 25-kilodalton toxin of Bacillus thuingiensis subsp. israelensis to cell membranes and alteration by mono- clonal antibodies and amino acid modifiers. Appl. Environ. Microbiol. 55:2779-2788.
7. Crickmore, N., and D. J. Ellar. 1992. Involvement of a possible chaperonin in the efficient expression of a cloned CryIIA f-endotoxin gene in Bacillus thuningiensis. Mol. Microbiol. 6:1533-1537.
8. Delecluse, A., C. Bourgouin, A. Klier, and G. Rapoport. 1988. Specificity of action on mosquito larvae of Bacillus thuringiensis israelensis toxins encoded by two different genes. Mol. Gen. Genet. 214:42-47.
9. Delecluse, A., J.-K. Charles, A. Klier, and G. Rapoport. 1991. Deletion by in vivo recombination shows that the 28-kilodalton cytolytic polypeptide from Bacillus thuringiensis subsp. israe- lensis is not essential for mosquitocidal activity. J. Bacteriol. 173:3374-3381.
10. Donovan, W. P., C. Dankocsik, and M. Pearce-Gilbert. 1988. Molecular characterization of a gene encoding a 72-kilodalton mosquito-toxic crystal protein from Bacillus thuringiensis subsp. israelensis. J. Bacteriol. 170:4732-4738.
11. Douek, J., M. Einav, and A. Zaritsky. 1992. Sensitivity to plating of Escherichia coli cells expressing the cryA gene from Bacillus thuningiensis var. israelensis. Mol. Gen. Genet. 232: 162-165.
12. Earp, D. J., E. S. Ward, and D. J. Ellar. 1987. Investigation of possible homologies between crystal proteins of three mosqui- tocidal strains of Bacillus thuringiensis. FEMS Microbiol. Lett. 42:195-199.
13. Frutos, R., C. Chang, S. S. Gill, and B. A. Federici. 1991. Nucleotide sequence of genes encoding a 72-kDa mosquitocidal protein and an associated 20-kDa protein are highly conserved in subspecies of Bacillus thuringiensis from Israel and The Phillipines. Biochem. Syst. Ecol. 19:599-609.
14. Galjart, N. J., N. Sivasubramanian, and B. A. Federici. 1987. Plasmid location, cloning, and sequence analysis of the gene encoding a 27.3-kilodalton cytolytic protein from Bacillus thur- ingiensis subsp. momrisoni (PG-14). Curr. Microbiol. 16:171-177.
15. Gill, S. S., E. A. Cowles, and P. Pietrantonio. 1992. The mode of action of Bacillus thuingiensis endotoxins. Annu. Rev. Ento- mol. 37:615-636.
16. Gill, S. S., J. M. Hornung, J. E. Ibarra, G. J. P. Singh, and B. A. Federici. 1987. Cytolytic activity and immunological similarity of the Bacillus thuringiensis subsp. israelensis and Bacillus thuringiensis subsp. morrisoni isolate PG-14 toxins. Appl. En- viron. Microbiol. 53:1251-1256.
17. Gill, S. S., G. J. P. Singh, and J. M. Hornung. 1987. Cell membrane interaction of Bacillus thuringiensis subsp. israelen- sis cytolytic toxins. Infect. Immun. 55:1300-1308.
18. Goldberg, L. J., and J. A. Margalit. 1977. Bacterial spore demonstrating rapid larvicidal activity against Anopheles ser- gentii, Uranotaenia unguiculata, Culex univitattus, Aedes ae- gypti and Culexpipiens. Mosq. News 37:355-358.
19. Hofte, H., and H. R. Whiteley. 1989. Insecticidal crystal proteins of Bacillus thuringiensis. Microbiol. Rev. 53:242-255.
20. Ibarra, J. E., and B. A. Federici. 1986. Isolation of a relatively nontoxic 65-kilodalton inclusion from the parasporal body of Bacillus thuwingiensis subsp. israelensis. J. Bacteriol. 165:257-533.
21. Laemmli, U. K. 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature (London) 227:680-685.
22. Lereclus, D., S. 0. Arant, J. Chaufaux, and M.-M. Lecadet. 1989. Transformation and expression of a cloned delta-endo- toxin gene in Bacillus thuringiensis. FEMS Microbiol. Lett. 60:211-217.
23. Lowry, 0. H., N. J. Rosebrough, A. L. Farr, and R. J. Randall. 1951. Protein measurement with the Folin phenol regent. J. Biol. Chem. 193:265-275.
24. McLean, K. M., and H. R. Whiteley. 1987. Expression in Escherichia coli of a cloned crystal protein gene of Bacillus thunngiensis subsp. israelensis. J. Bacteriol. 169:1017-1023.
25. Newbury, S. F., N. H. Smith, E. C. Robinson, I. D. Hiles, and C. F. Higgins. 1987. Stabilization of translationally active mRNA by prokaryotic REP sequences. Cell 48:297-310.
26. Padua, L. E., and B. A. Federci. 1990. Development of mutants of the mosquitocidal bacterium Bacillus thuringiensis subsp. morrisoni (PG-14) toxic to lepidopterous or dipterous insects. FEMS Microbiol. Lett. 66:257-262.
27. Padua, L. E., M. Ohba, and K. Aizawa. 1984. Isolation of a Bacillus thuringiensis strain (serotype 8a: 8b) highly and selec- tively toxic against mosquito larvae. J. Invertebr. Pathol. 44: 12-17.
28. Raymond, M. 1985. Presentation d'un programme basic d'anal- yse log-probit pour micro-ordinateur. Cah. ORSTROM Ser. Entomol. Med. Parasitol. 23:117-121.
29. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: a laboratory manual, 2nd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y.
30. Tabashnik, B. E. 1992. Evaluation of synergism among Bacillus thuringiensis toxins. Appl. Environ. Microbiol. 58:3343-3346.
31. Thomas, W. E., and D. J. Ellar. 1983. Mechanism of action of Bacillus thuringiensis var israelensis insecticidal delta-endo- toxin. FEBS Lett. 154:362-368.
32. Visick, J. E., and H. R. Whiteley. 1991. Effect of a 20-kilodalton protein from Bacillus thuringiensis subsp. israelensis on pro- duction of the CytA protein by Escherichia coli. J. Bacteriol. 173:1748-1756.
33. Ward, E. S., and D. J. Ellar. 1988. Cloning and expression of two homologous genes of Bacillus thuringiensis subsp. israelen- sis which encode 130-kilodalton mosquitocidal proteins. J. Bac- teriol. 170:727-735.
34. Ward, E. S., D. J. Ellar, and C. N. Chilcott. 1988. Single amino acid changes in the Bacillus thuringiensis var. israelensis delta- endotoxin affect the toxicity and expression of the protein. J. Mol. Biol. 202:527-535.
35. Ward, E. S., A. R. Ridley, D. J. Ellar, and J. A. Todd. 1986. Bacillus thuringiensis var. israelensis delta-endotoxin: cloning and expression of the toxin in sporogenic and asporogenic strains of Bacillus subtilis. J. Mol. Biol. 191:13-22.
36. Widner, W. R., and H. R. Whiteley. 1989. Two highly related insecticidal crystal proteins of Bacillus thuringiensis subsp. kurstaAi possess different host range specificities. J. Bacteriol. 171:965-974.
37. Wong, H. C., and S. Chang. 1986. Identification of a positive retroregulator that stabilizes mRNAs in bacteria. Proc. Natl. Acad. Sci. USA 83:3233-3237.
38. Wu, D., and F. N. Chang. 1985. Synergism in mosquitocidal activity of 26 and 65 kDa proteins from Bacillus thuringiensis subsp. israelensis crystal. FEBS Lett. 190:232-236.
39. Yu, Y.-M., M. Ohba, and K. Aizawa. 1987. Synergistic effects of the 65- and 25-kilodalton proteins of Bacillus thuringiensis strain PG-14 (serotype 8A:8B) in mosquito larvicidal activity. J. Gen. Appl. Microbiol. 33:459-462.
40. Yu, Y.-M., M. Ohba, and S. S. Gill. 1991. Characterization of mosquitocidal activity of Bacillus thuringiensis subsp. fukuo- kaensis crystal proteins. Appl. Environ. Microbiol. 57:1075- 1081.
VOL. 59, 1993
http://aem .asm
.org/ D