Identification and Characterization of a Previously Undescribed cyt ...

APPLIED AND ENVIRONMENTAL MICROBIOLOGY, July 1997, p. 2716–2721 0099-2240/97/$04.0010 Copyright © 1997, American Society for Microbiology

Vol. 63, No. 7

Identification and Characterization of a Previously Undescribed cyt Gene in Bacillus thuringiensis subsp. israelensis ALEJANDRA GUERCHICOFF, RODOLFO A. UGALDE,

AND

CLARA P. RUBINSTEIN*

Instituto de Investigaciones Bioquı´micas F. Leloir, Fundacio ń Campomar, 1405 Capital Federal, Argentina Received 21 November 1996/Accepted 20 April 1997

Mosquitocidal Bacillus thuringiensis strains show as a common feature the presence of toxic proteins with cytolytic and hemolytic activities, Cyt1Aa1 being the characteristic cytolytic toxin of Bacillus thuringiensis subsp. israelensis. We have detected the presence of another cyt gene in this subspecies, highly homologous to cyt2Aa1, coding for the 29-kDa cytolytic toxin from B. thuringiensis subsp. kyushuensis. This gene, designated cyt2Ba1, maps upstream of cry4B coding for the 130-kDa crystal toxin, on the 72-MDa plasmid of strain 4Q2-72. Sequence analysis revealed, as a remarkable feature, a 5* mRNA stabilizing region similar to those described for some cry genes. PCR amplification and Southern analysis confirmed the presence of this gene in other mosquitocidal subspecies. Interestingly, anticoleopteran B. thuringiensis subsp. tenebrionis belonging to the morrisoni serovar also showed this gene. On the other hand, negative results were obtained with the antilepidopteran strains B. thuringiensis subsp. kurstaki HD-1 and subsp. aizawai HD-137. Western analysis failed to reveal Cyt2A-related polypeptides in B. thuringiensis subsp. israelensis 4Q2-72. However, B. thuringiensis subsp. israelensis 1884 and B. thuringiensis subsp. tenebrionis did show cross-reactive products, although in very small amounts.

PG14 have the highest similarity between their toxin complements. In fact, their Cyt1Aa hemolysins were found to differ only in one amino acid residue (14). Although lower homologies were detected with other mosquitocidal subspecies, Cyt2Aa1 from B. thuringiensis subsp. kyushuensis showed 39% identity (and 70% functional homology) with Cyt1Aa1 from B. thuringiensis subsp. israelensis (22). A new variant (Cyt1Aab1) was recently described for B. thuringiensis subsp. medellin (serotype H30) which is 86% identical to Cyt1Aa1 from B. thuringiensis subsp. israelensis (31). We report the detection and characterization of a cyt2 variant in B. thuringiensis subsp. israelensis that is highly homologous to cyt2Aa1 from B. thuringiensis subsp. kyushuensis.

Bacillus thuringiensis mosquitocidal strains produce parasporal inclusions that are composed of several toxic polypeptides which fall into two classes: Cry and Cyt, that act together to effect the larvicidal activity of the parasporal crystal (23). In B. thuringiensis subsp. israelensis, three major antidipteran toxins ranging from 68 to 135 kDa (Cry4A, Cry4B, and Cry11A in the new nomenclature) contribute to the overall toxicity of intact crystals in a synergistic manner (5, 9, 25). Cyt toxins are hemolytic and cytolytic in vitro and are specifically active against dipteran larvae in vivo (20, 35); they are smaller than Cry polypeptides (25 to 28 kDa), and three types have been defined according to immunoreactivity. Cyt1 (former CytA) hemolysins are characteristic of B. thuringiensis subsp. israelensis and subsp. morrisoni PG14 (16, 33). Cyt2 (former CytB) is found in the inclusions of B. thuringiensis subsp. darmstadiensis 73-E10-2 (20) and subsp. kyushuensis (21), and CytC is representative of subsp. fukuokaensis (35). A new type, CytD, has been recently proposed for B. thuringiensis subsp. jegathesan (18). They do not share homology with Cry toxins, but both types would share a common cytolytic mechanism involving colloid-osmotic lysis (19); however, the actual mechanism of pore formation between the Cry and Cyt toxins might be different (9). Both Cyt1A and Cyt2A have been shown to form cation-selective channels, and although they have a broad cytolytic activity in vitro, it has been suggested that an insect-specific receptor may be essential for these toxins to be active in vivo (22, 32). This could explain the high specificity of Cyt toxins against dipteran larvae. Molecular homologies between different mosquitocidal B. thuringiensis strains have been established at the protein as well as the DNA level (29). According to these studies, B. thuringiensis subsp. israelensis and B. thuringiensis subsp. morrisoni

MATERIALS AND METHODS Strains, plasmids, and media. B. thuringiensis strains and plasmids used in this work are listed in Table 1. Escherichia coli DH5a was used for plasmid construction and propagation (15). B. thuringiensis strains were maintained in sporulation Schaeffer’s agar medium (28) and grown in Luria-Bertani medium (24) for DNA isolation. Liquid cultures were grown with aeration (shaking) at 37°C (E. coli) or 30°C (B. thuringiensis). When appropriate, ampicillin was added to autoclaved media at 100 mg/ml. DNA manipulations. Restriction enzymes and T4 DNA ligase (Gibco BRL) were used as recommended by the manufacturers. DNA restriction fragments or PCR-amplified fragments were purified from agarose gels with a Gene Clean kit (Bio 101). Plasmids from E. coli were prepared as described by Birnboim and Doly (6). Plasmid DNA was isolated from B. thuringiensis strains as described in reference 7 and further purified by Qiagen columns (DIAGEN GmbH, QIAGEN Inc.). DNA sequencing. DNA sequences were determined on double-stranded DNA by the chain termination method (27) after subcloning of the 1.4-kb HindIIIEcoRI insert from pRX80 into pUC19 and also directly from pRX80. The Sequenase version 2.0 DNA sequencing kit (Amersham, Cleveland, Ohio) was used with the pUC commercial primers M13/pUC sequencing primer (240) and M13/pUC reverse sequencing primer (224) (New England Biolabs, Inc). Internal primers were designed according to the cyt2Ba1 gene (Gibco BRL). PCR conditions. Twenty to 50 ng of purified plasmid DNA was added to the PCR mixtures (0.2 mM deoxynucleoside triphosphates, 2 mM MgCl2, 0.5 U of Taq polymerase [Promega], 100 ng of PCR primers) in a final volume of 50 ml. The oligonucleotide primers were as follows: upper, 59AATACATTTCAAGG AGCTA39; lower, 59TTTCATTTTAACTTCATATC39. Amplification was performed in a thermal cycler (M.J. Research Minicycler PTC100) by using a single denaturation step (3 min at 94°C), followed by a 35-cycle program, with each

* Corresponding author. Mailing address: Instituto de Investigaciones Bioquı´micas F. Leloir, Fundacıón Campomar, Av. Patricias Argentinas 435, 1405 Capital Federal, Argentina. Phone: 54-1-8634015. Fax: 54-1-8652246. 2716

NEW cyt VARIANT IN B. THURINGIENSIS SUBSP. ISRAELENSIS

VOL. 63, 1997

2717

TABLE 1. B. thuringiensis strains and plasmid used in this study Strain or plasmid

Strains B. thuringiensis B. thuringiensis B. thuringiensis B. thuringiensis B. thuringiensis B. thuringiensis B. thuringiensis B. thuringiensis B. thuringiensis Plasmid; pRX80 a

subsp. subsp. subsp. subsp. subsp. subsp. subsp. subsp. subsp.

kyushuensis 74F6-18 fukuokaensis israelensis 1884 darmstadiensis 73-E10-2 tenebrionis (morrisoni) kurstaki HD73 aizawai HD137 morrisoni PG14 israelensis 4Q2-72

Serotype

H11a, -11c H3a, -3d, -3e H14 H10a, -10b H8a, -8b 3a, 3b 7 H8a, -8b H14

Relevant phenotype

Source

Toxic for dipteran larvae Toxic for dipteran larvae Toxic for dipteran larvae Toxic for dipteran larvae Toxic for coleopteran larvae Toxic for lepidopteran larvae Toxic for lepidopteran larvae Toxic for dipteran larvae Toxic for dipteran larvae; bears only the 72-MDa plasmid

BGSCa (4U1) BGSC (4AP1) Pasteur Institute (France) BGSC (4M1) BGSC BGSC (4D4) BGSC (4J5) Pasteur Institute (France) BGSC (4Q5)

Apr; SstI fragment carrying cry4B from B. thuringiensis subsp. israelensis cloned into pUC18

A. Delećluse, Pasteur Institute, France

BGSC, Bacillus Genetic Stock Center, The Ohio State University, Columbus.

cycle consisting of denaturation at 94°C for 45 s, annealing at 42°C for 45 s, and extension at 72°C for 1 min; a final extension step of 72°C for 5 min was also included. Twenty-microliter samples from each PCR mixture were electrophoresed on 1.5% agarose gels in 0.53 Tris-acetate buffer at 100 V for 30 to 35 min and stained with ethidium bromide. Southern blotting. B. thuringiensis plasmid DNAs digested with EcoRI or HindIII were separated by electrophoresis in a 0.8% agarose gel and transferred by capillarity to nitrocellulose filters (24). Hybridization was performed overnight with a radiolabelled probe at 56°C in a solution containing 0.5% sodium dodecyl sulfate (SDS), 53 Denhart’s solution, 100 mg of salmon sperm DNA per ml, and 63 SSC (13 SSC is 150 mM NaCl plus 15 mM sodium citrate [pH 7.0]). Filters were washed at room temperature with two changes of 0.2% SDS and 33 SSC and then once with 0.5% SDS–13 SSC, before exposure to Fuji XR film at 270°C with an X-ray intensifier. Protein analysis. B. thuringiensis strains were grown in Schaeffer’s liquid sporulation medium (28) until lysis. Spore-crystal mixtures from 10-ml culture samples were harvested by centrifugation at 12,000 3 g for 20 min and then washed once in 1 M NaCl–2 mM phenylmethylsulfonyl fluoride–10 mM EDTA. Pellets were resuspended in sample buffer (24) supplemented with phenylmethylsulfonyl fluoride and EDTA as described before, boiled for 10 min, and subjected to SDS–15% polyacrylamide gel electrophoresis. Protein concentrations were determined by the Bradford assay (Bio-Rad) on solubilized samples (26). Proteins were electrotransferred to nitrocellulose membranes and detected immunologically according to the method of Koni and Ellar (22), with the following modifications: incubation with the anti-Cyt2 antiserum (kindly provided by David Ellar, University of Cambridge) was performed at room temperature for 1 h and then overnight at 4°C. The antiserum was added at a 1:500 dilution. The Gibco BRL detection system (biotinylated second antibody, streptavidin-alkaline phosphatase, nitroblue tetrazolium–5-bromo-4-chloro-3-indolylphosphate toluidinium) was used as recommended by the manufacturer. RNA extraction and RT-PCR. B. thuringiensis strains were cultured in Schaeffer’s medium at 30°C with shaking. Samples were taken at around t3 (t0 is defined as the onset of sporulation, and tn indicates the number of hours after t0). Cells were harvested by centrifugation and then resuspended in 10 ml of protoplasting buffer (15 mM Tris-HCl [pH 8], 8 mM EDTA, 0.45 mM sucrose) and 0.4 mg of lysozyme per ml. The homogenized samples were incubated for 15 min in ice, centrifuged for 5 min at 7,000 rpm at 4°C, and resuspended in 0.5 ml of lysis buffer (10 mM Tris-HCl [pH 8.0], 10 mM NaCl, 1 mM sodium citrate, 1.5% SDS) in the presence of diethylpyrocarbonate. After 5 min in ice, 55 ml of 2 M sodium acetate (pH 4.0) was added before the addition of 500 ml of water-equilibrated phenol-chloroform (1:1). Phases were separated by centrifugation, and the aqueous phase was recovered. RNAs were precipitated from the aqueous phase by mixing with isopropyl alcohol, followed by incubation at 220°C for 30 min and centrifugation at no more than 12,000 3 g for 20 min at 4°C. Single-stranded cDNA synthesis was performed as described in reference 13 with the lower PCR primer for the reverse transcription (RT) step. Aliquots (1/10 of the singlestranded DNA-cDNA mixture) were used for PCRs without further purification. The amplification conditions were as described above. Computer analysis. DNA sequences were analyzed by using the National Center for Biotechnology Information’s BLAST WWW Server and with the MegAlign program (Macintosh 3.03; DNASTAR Inc). Nucleotide sequence accession number. The nucleotide sequence data reported here have been submitted to GenBank and assigned accession no. U52043.

RESULTS Sequence analysis. Sequencing of the upstream region of the cry4B gene from B. thuringiensis subsp. israelensis 4Q2-72 revealed a long open reading frame on the complementary strand. This sequence was found at about 1 kb from cry4B, on the same SstI fragment of the 72-MDa megaplasmid (11). When compared against the database sequence bank with the BLAST WWW Server (National Center for Biotechnology Information), the sequence showed a high degree of homology with cyt2Aa, the gene coding for the cytolytic endotoxin from B. thuringiensis subsp. kyushuensis. This sequence, which appears to constitute a monocistronic transcript, showed consensus sequences for the 235 and 210 midsporulation promoters previously reported for several B. thuringiensis toxins, suggesting a sE-dependent transcription for this gene (3, 8). Some other interesting features were also found in this upstream region. (i) A typical Bacillus ribosome binding site (RBS), GGAGG (34), was found at position 125, followed by two potential stem-loop structures that might form at the inverted repeats between nucleotides 165 and 196 or 168 and 193, respectively. These sequences resemble the 59 mRNA stabilizer region described for cry3 B. thuringiensis genes (1, 2). In fact, this RBS is included in a longer stretch of nucleotides which is identical to the stabilizing region from cry3A (GAAA GGAGGGA [enclosed in an open box in Fig. 1]). (ii) A second, nontypical RBS (GGGGG) was found at position 271; 11 bases downstream lies the ATG start codon for the open reading frame corresponding to this gene (260 codons long). (iii) At the 39 end, this sequence revealed a possible terminator secondary structure constituted by two partially overlapping hairpin loops (from nucleotides 1065 to 1082) that include the stop codon TAA. The presence of 39 terminator-stabilizing sequences is highly conserved among B. thuringiensis toxin genes (1). The low G1C content of the entire sequence (27%) is in full agreement with the cloned cyt2Aa1 gene and is typical of B. thuringiensis endotoxin genes. Sequence alignments. DNA sequence comparisons made with the Blastn and the MegAlign programs revealed a high degree of homology between this sequence and the cyt2Aa1 gene from B. thuringiensis subsp. kyushuensis (22). The highest similarity (80%) was found from nucleotides 344 to 1054. Sequence comparisons at the (predicted) protein level confirmed that Cyt2 from B. thuringiensis subsp. israelensis was highly similar to Cyt proteins in general and, in particular (67.6%), to

2718

GUERCHICOFF ET AL.

APPL. ENVIRON. MICROBIOL.

FIG. 2. Amino acid sequence alignment between Cyt2Ba1 from B. thuringiensis subsp. israelensis and the reported Cyt2Aa1 from B. thuringiensis subsp. kyushuensis. Solid and broken boxes, predicted conserved b strands and a helices, respectively.

FIG. 1. (A) Localization of cyt2Ba in B. thuringiensis subsp. israelensis 4Q272. An SstI fragment from pRX80 carrying cry4B and part of cry10A is shown. This fragment comes from the 72-MDa megaplasmid of B. thuringiensis subsp. israelensis 4Q2-72. Arrows indicate transcriptional direction. S, SstI; H1 and H2, HindIII; EI, EcoRI; EV, EcoRV. (B) Nucleotide sequence of the cyt2Ba1 gene from B. thuringiensis subsp. israelensis. Promoter 235 and 210 consensus sequences are boxed. RBSs are printed in outlined letters, and start (ATG) and stop (TAA) codons are underlined. Potential hairpin structures are indicated by solid arrows. Primers used for PCR amplification are indicated by broken arrows. The predicted size of the PCR product is 469 bp.

1884 and subsp. morrisoni PG14 DNAs, although a weaker signal was observed in B. thuringiensis subsp. israelensis 1884. Even when similar amounts of these DNAs were loaded on the agarose gel, they were visually estimated, and therefore no quantitative conclusions can be drawn. In the case of B. thuringiensis subsp. tenebrionis, hybridization to a 4.2-kb EcoRI fragment was observed. A HindIII fragment of around 6 kb was observed in B. thuringiensis subsp. kyushuensis, whereas under these conditions, no signal was found in B. thuringiensis subsp. darmstadiensis and subsp. fukuokaensis. Expression of the cyt2-like genes. Western blot analysis of spore-crystal extracts was performed with B. thuringiensis subsp. israelensis 4Q2-72 and 1884, as well as in B. thuringiensis subsp. tenebrionis, in order to determine whether these genes

Cyt2Aa. In fact, as shown in Fig. 2, predicted a helices and b strand domains (22) are fairly well conserved throughout the entire sequence. PCR and Southern analysis. A pair of oligonucleotide primers was designed from two highly conserved regions between B. thuringiensis subsp. israelensis 4Q2-72 and B. thuringiensis subsp. kyushuensis cyt2 genes (Fig. 1). In fact, the upper and lower primers are 85 and 100% homologous, respectively. This primer set was used in PCR amplification experiments in order to search for the presence of this gene in other B. thuringiensis strains. As shown in Fig. 3, amplification products of the expected size (469 bp) were obtained with purified plasmid DNAs as templates in all of the mosquitocidal strains, whereas no product was observed for antilepidopteran strains. Interestingly, B. thuringiensis subsp. tenebrionis with anticoleopteran activity showed the 469-bp fragment. It is noteworthy that when the annealing temperature was raised from 42°C to 45°C, no amplification products were obtained for B. thuringiensis subsp. darmstadiensis or subsp. fukuokaensis, suggesting a lower homology with the primers. Southern analysis of the same DNAs digested with EcoRI or HindIII (for B. thuringiensis subsp. kyushuensis) was also performed in order to confirm the presence of cyt2-related sequences in the different strains (Fig. 4). The 469-bp amplification product from pRX80 was used as a probe. At stringency conditions allowing around a 30% mismatch, a single EcoRI hybridization band of 4.7 kb was observed with B. thuringiensis subsp. israelensis 4Q2-72 and

FIG. 3. PCR amplifications. Plasmid DNA preparations from different B. thuringiensis strains were subjected to PCR amplification as described in Materials and Methods. Lanes: 1, B. thuringiensis subsp. aizawai; 2, B. thuringiensis subsp. kyushuensis; 3, B. thuringiensis subsp. morrisoni PG14; 4, B. thuringiensis subsp. fukuokaensis; 5, B. thuringiensis subsp. tenebrionis; m, 100-bp DNA ladder; 6, B. thuringiensis subsp. israelensis 1884; 7, B. thuringiensis subsp. darmstadiensis; 8, B. thuringiensis subsp. kurstaki; 9, B. thuringiensis subsp. israelensis 4Q2; 10, PCR control reaction without DNA template.

VOL. 63, 1997


FIG. 4. (A) Southern blotting analysis. DNA transfer and hybridization were carried out as described in Materials and Methods. The 469-bp PCR amplification product from pRX80 was used as a probe. Lanes: 1, B. thuringiensis subsp. morrisoni PG14; 2, B. thuringiensis subsp. israelensis 1884; 3, B. thuringiensis subsp. israelensis 4Q2-72; 4, B. thuringiensis subsp. fukuokaensis; 5, B. thuringiensis subsp. kyushuensis; 6, B. thuringiensis subsp. darmstadiensis; 7, B. thuringiensis subsp. tenebrionis.

were expressed in these strains. As can be seen in Fig. 5, B. thuringiensis subsp. israelensis 4Q2-72 did not show any reactive polypeptide of the expected size. However, B. thuringiensis subsp. israelensis 1884 revealed a band of the same size as that of B. thuringiensis subsp. kyushuensis Cyt2Aa. A larger polypeptide (around 28 to 29 kDa) cross-reacted with the antiCyt2Aa antiserum in B. thuringiensis subsp. tenebrionis. DISCUSSION B. thuringiensis subsp. israelensis is the mosquitocidal subspecies most thoroughly studied. One of the major components of its toxin crystal is the cytolytic endotoxin Cyt1Aa1. We have found in this subspecies another gene coding for a cytolytic toxin, highly homologous to the reported cyt2Aa1 from B. thuringiensis subsp. kyushuensis (22). The physical map of the megaplasmid carrying the B. thuringiensis subsp. israelensis toxin genes has very recently been reported (4). Our findings add a new component to this map, contributing to the knowledge of the coding information available for this megaplasmid (which amounts to less than 20% up to date). This variant, which has been designated cyt2Ba1 (10), enters a phylogram of cry and cyt genes (not shown) at a node of about 67% identity with the reported Cyt2Aa1, constituting a new member of the Cyt2 family. B. thuringiensis subsp. darmstadiensis Cyt2 might also be a different variant. The lack of hybridization observed in this subspecies under our stringency conditions might reflect this difference. In the case of B. thuringiensis subsp. fukuokaensis, the lack of signal in the Southern experiment but the presence of the expected PCR amplification product could be the result of homology with cytC (coding for the Cyt toxin described for this subspecies) and not with a cyt2 variant. Cloning and sequencing of all of the amplification products are currently being done to clarify these possibilities. Western blotting experiments done so far have failed to reveal the presence of the expected polypeptide in B. thuringiensis subsp. israelensis 4Q2-72, whereas a small amount of a polypeptide similar in size to B. thuringiensis subsp. kyushuensis

2719

Cyt2Aa was present in B. thuringiensis subsp. israelensis 1884 (both strains derive from the original B. thuringiensis subsp. israelensis isolate). Interestingly, B. thuringiensis subsp. tenebrionis also showed a polypeptide that cross-reacted with the antiserum, although it appears to be larger than Cyt2. B. thuringiensis subsp. tenebrionis does not show low-molecular-mass polypeptides as crystal components in addition to the 68- to 73-kDa Cry3A toxin (30). However, isolate EG2158 has been described to synthesize the major protein of 68 kDa and two minor components of approximately 29 and 30 kDa. The latter, is part of a diamond-shaped inclusion found in this isolate, not present in B. thuringiensis subsp. tenebrionis, and not essential for toxic activity (12). In fact, in both B. thuringiensis subsp. israelensis 1884 and subsp. tenebrionis, the immunoreactive polypeptides are minor components, present in small amounts when compared with the other components of the spore-crystal preparations (not shown). This low level of expression could explain why the toxin might have not been detected in 4Q2-72 at our resolution level. This could also imply that these genes may be functionally “cryptic,” due to insufficient expression, although being transcriptionally active. In fact, RT-PCR studies carried out with B. thuringiensis subsp. israelensis confirmed that cyt2Ba is normally transcribed. When total RNAs from B. thuringiensis subsp. israelensis 4Q272 and 1884 were reverse transcribed and then amplified with our PCR primer set, the expected 469-bp amplification product could be observed (Fig. 6). This result shows that the promoter sequences found upstream of the open reading frame are functional and suggests that a full-length message is produced in these strains. With the information at hand, we can conclude that a new cyt2 variant is present in B. thuringiensis subsp. israelensis 4Q272, of which we know the sequence and localization. Interestingly, two strains belonging to serotype H8a8b and to two different pathotypes (PG14 and a strain of B. thuringiensis subsp. tenebrionis), also carry this type of gene. This is also true for the two other members of this serotype, with antilepidopt-

FIG. 5. Western blotting analysis. B. thuringiensis spore-crystal preparations were processed for immunodetection with an anti-Cyt2Aa1 specific antiserum as described in Materials and Methods. Around 40 mg of total protein was originally loaded in each lane. Lanes: 1, B. thuringiensis subsp. tenebrionis; 2, B. thuringiensis subsp. israelensis 4Q2-72; 3, B. thuringiensis subsp. israelensis 1884; 4, B. thuringiensis subsp. kyushuensis. Molecular mass markers are indicated on the left in kilodaltons.

2720

GUERCHICOFF ET AL.

APPL. ENVIRON. MICROBIOL. ACKNOWLEDGMENTS We thank Neil Crickmore and Daniel Zeigler of the Bacillus thuringiensis Toxin Gene Nomenclature Committee and the BGSC for helpful discussion and phylogenetic protein comparisons and David Ellar for the anti-Cyt2A1 antiserum. We are grateful to Carmen Sanchez Rivas for helpful discussion and valuable material and A. Delećluse for providing us with pRX80. We also thank Guido Pollevick for training A.G. in DNA sequencing. A.G. is a predoctoral fellow of the National Council for Scientific and Technological Research (CONICET).

FIG. 6. RT-PCR of total RNAs from B. thuringiensis subsp. israelensis 1884 (lanes 1 and 2) and 4Q2-72 (lanes 3 and 4). Lane 5 contains a 100-bp DNA ladder. Even lanes, control reactions (without reverse transcriptase).

eran specificity, strains HD517 and 518, as we have very recently found (data not shown). This may have evolutionary implications, suggesting an ancestral character for the morrisoni serovar, from which the others might have diverged. Expression of the corresponding polypeptides is presently being investigated for all of the subspecies found to bear cyt2-related sequences. The presence of a 59 mRNA stabilizing sequence is another interesting feature which, described for the first time in a cyt gene, seems to reinforce the connection between cyt-bearing subspecies and the anticoleopteran pathotypes. These sequences appear to be consensus RBSs located in the upstream untranslated region and which can stabilize the transcript through the interaction with the 39 end of the 16S rRNA. They are found in similar positions in the three types of cry3 genes described so far (1). We have also found a potential stem-loop structure downstream of the RBS, although the secondary structures associated with the cry3 stabilizers are located upstream of these sequences (3). The stabilizing region described in the early messenger from B. subtilis bacteriophage SP82 (also a polypurine sequence) is also preceded by a secondary structure, although it does not seem to be essential for stabilization (17). The constant presence of Cyt polypeptides (26) among mosquitocidal B. thuringiensis strains suggests that a selective advantage might exist for their expression. In the case of the Cyt2-related toxins here described, their role (if any) in the toxicity of the different strains remains to be established. In the case of B. thuringiensis subsp. israelensis, the low level of expression and the similar larvicidal activities of strains 4Q2-72 and 1884 toward mosquito larvae suggest that these cytolysins (which coexist with Cyt1Aa) are dispensable for full activity. In fact, Cyt1Aa has been shown to be more active than Cyt2 or CytC (26, 35). Nevertheless, the high degree of conservation found among the different isolates is extremely significant. Because of their apparently ubiquitous nature among the mosquitocidal pathotype and the morrisoni serotype, these cyt genes might also constitute markers that could be exploited for screening purposes once all versions are sequenced and good sets of PCR primers are established from highly conserved regions.

REFERENCES 1. Agaisse, H., and D. Lereclus. 1994. Structural and functional analysis of the promoter region involved in full expression of the cryIIIA toxin gene of Bacillus thuringiensis. Mol. Microbiol. 13:97–107. 2. Agaisse, H., and D. Lereclus. 1995. How does Bacillus thuringiensis produce so much insecticidal crystal protein? J. Bacteriol. 177:6027–6032. 3. Baum, J. A., and T. Malvar. 1995. Regulation of insecticidal crystal protein production in Bacillus thuringiensis. Mol. Microbiol. 18:1–12. 4. Ben-Dov, E., M. Einav, N. Peleg, S. Boussiba, and A. Zaritsky. 1996. Restriction map of the 125-kilobase plasmid of Bacillus thuringiensis subsp. israelensis carrying the genes that encode delta-endotoxins active against mosquito larvae. Appl. Environ. Microbiol. 62:3140–3145. 5. Ben-Dov, E., S. Boussiba, and A. Zaritsky. 1995. Mosquito larvicidal activity of Escherichia coli with combinations of genes from Bacillus thuringiensis subsp. israelensis. J. Bacteriol. 177:2851–2857. 6. Birnboim, H., and J. Doly. 1979. A rapid alkaline extraction procedure for screening recombinant plasmid DNA. Nucleic Acids Res. 7:1513–1523. 7. Bourgouin, C., A. Klier, and G. Rapoport. 1986. Characterization of the genes encoding the haemolytic toxin and the mosquitocidal d-endotoxins of Bacillus thuringiensis var israelensis. Mol. Gen. Genet. 205:390–397. 8. Brown, K. L. 1993. Transcriptional regulation of the Bacillus thuringiensis subsp. thompsoni crystal protein gene operon. J. Bacteriol. 175:7951–7957. 9. Crickmore, N., E. J. Bone, J. A. Williams, and D. J. Ellar. 1995. Contribution of the individual components of the d-endotoxin crystal to the mosquitocidal activity of Bacillus thuringiensis subsp. israelensis. FEMS Microbiol. Lett. 131:249–254. 10. Crickmore, N., and D. Zeigler. 1996. Personal communication. 11. Delećluse, A., S. Poncet, A. Klier, and G. Rapoport. 1993. Expression of cryIVA and cryIVB genes, independently or in combination, in a crystalnegative strain of Bacillus thuringiensis subsp. israelensis. Appl. Environ. Microbiol. 59:3922–3927. 12. Donovan, W. P., J. M. Gonzalez, Jr., M. P. Gilbert, and C. Dankocsik. 1988. Isolation and characterization of EG2158, a new strain of Bacillus thuringiensis toxic coleopteran larvae, and nucleotide sequence of the toxin gene. Mol. Gen. Genet. 214:365–372. 13. Dumas Milne Edwards, J. B., P. Ravassard, C. Icard-Liepkalns, and J. Mallet. 1995. cDNA cloning by RT-PCR, p. 96–98. In M. J. McPherson, B. D. Hames, and G. R. Taylor (ed.), PCR2, a practical approach. Oxford University Press, Inc., New York, N.Y. 14. Earp, D. J., E. S. Ward, and D. J. Ellar. 1987. Investigation of possible homologies between crystal proteins of three mosquitocidal strains of Bacillus thuringiensis. FEMS Microbiol. Lett. 42:195–199. 15. Hanahan, D. 1983. Studies on the transformation of Escherichia coli with plasmids. J. Mol. Biol. 166:557–580. 16. Hofte, H., and H. R. Whiteley. 1989. Insecticidal crystal proteins of Bacillus thuringiensis. Microbiol. Rev. 53:242–255. 17. Hue, K. K., S. D. Cohen, and D. H. Bechhofer. 1995. A polypurine sequence that acts as a 59 mRNA stabilizer in Bacillus subtilis. J. Bacteriol. 177:3465– 3471. 18. Kawalek, M. D., S. Benjamin, H. L. Lee, and S. S. Gill. 1995. Isolation and identification of novel toxins from a new mosquitocidal isolate from Malaysia, Bacillus thuringiensis subsp. jegathesan. Appl. Environ. Microbiol. 61: 2965–2969. 19. Knowles, B. H., and D. J. Ellar. 1987. Colloid-osmotic lysis is a general feature of the mechanism of action of Bacillus thuringiensis d-endotoxins with different specificity. Biochim. Biophys. Acta 924:509–518. 20. Knowles, B. H., M. R. Blatt, M. Tester, J. M. Horsnell, J. Carroll, G. Menestrina, and D. J. Ellar. 1989. A cytolytic d-endotoxin from Bacillus thuringiensis var israelensis forms cation-selective channels in planar lipid bilayers. FEBS Lett. 244:259–262. 21. Knowles, B. H., P. J. White, C. N. Nicholls, and D. J. Ellar. 1992. A broad spectrum cytolytic toxin from Bacillus thuringiensis var kyushuensis. Proc. Roy. Soc. Lond. Ser. B 248:1–7. 22. Koni, P. A., and D. J. Ellar. 1993. Cloning and characterization of a novel Bacillus thuringiensis cytolytic delta-endotoxin. J. Mol. Biol. 229:319–327. 23. Lereclus, D., A. Delećluse, and M.-M. Lecadet. 1993. Diversity of Bacillus thuringiensis toxins and genes, p. 37–69. In P. F. Entwistle, J. S. Cory, M. J. Bailey, and S. Higgs (ed.), Bacillus thuringiensis, an environmental biopesticide: theory and practice. John Wiley & Sons, Ltd., Chichester, United Kingdom.

VOL. 63, 1997


24. Maniatis, T., E. Fritsch, and J. Sambrook. 1982. Molecular cloning: a laboratory manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. 25. Park, H., H. Kim, D. Lee, Y. Yu, B. Jin, and S. Kang. 1995. Expression and synergistic effect of three types of crystal protein genes in Bacillus thuringiensis. Biochem. Biophys. Res. Commun. 214:602–607. 26. Ragni, A., I. Thie´ry, and A. Delećluse. 1996. Characterization of six highly mosquitocidal Bacillus thuringiensis strains that do not belong to H-14 serotype. Curr. Microbiol. 32:48–54. 27. Sanger, F., S. Nicken, and A. R. Coulson. 1977. DNA sequencing with chain-terminating inhibitors. Proc. Natl. Acad. Sci. USA 74:5463–5467. 28. Schaeffer, P., J. Millet, and J. Aubert. 1965. Catabolic repression of bacterial sporulation. Proc. Natl. Acad. Sci. USA 554:701–711. 29. Sekar, V. 1987. DNA homology between the crystal toxin genes of several mosquito pathogenic Bacillus thuringiensis strains. Curr. Microbiol. 15:1–3. 30. Sekar, V., D. V. Thompson, M. J. Maroney, R. G. Bookland, and M. J. Adang. 1987. Molecular cloning and characterization of the insecticidal crystal protein of Bacillus thuringiensis var tenebrionis. Proc. Natl. Acad. Sci. USA 84:7036–7040.

2721

31. Thiery, I., A. Delećluse, M. C. Tamayo, and S. Orduz. 1996. Identification of a gene for Cyt1A-like hemolysin from Bacillus thuringiensis subsp. medellin and expression in a crystal-negative B. thuringiensis strain. Appl. Environ. Microbiol. 63:468–473. 32. Thomas, W. E., and D. J. Ellar. 1983. Bacillus thuringiensis var israelensis crystal d-endotoxins: effects on insect and mammalian cells in vitro and in vivo. J. Cell Sci. 60:181–197. 33. Ward, E. S., D. J. Earp, B. H. Knowles, S. M. Pinnavaia, C. N. Nicholls, T. J. Sawyer, P. E. Granum, and D. J. Ellar. 1988. Diversity of structure and specificity of mosquitocidal d-endotoxins. Zentralbl. Bakteriol. Suppl. 17:67– 68. 34. Wong, H. C., H. E. Schnepf, and H. R. Whiteley. 1983. Transcriptional and translational start sites for the Bacillus thuringiensis crystal protein gene. J. Biol. Chem. 258:1960–1967. 35. Yu, Y.-M., M. Ohba, and S. S. Gill. 1991. Characterization of mosquitocidal activity of Bacillus thuringiensis subsp. fukuokaensis crystal proteins. Appl. Environ. Microbiol. 57:1075–1081.