Identification and Characterization of Three Previously Undescribed Crystal Proteins from Bacillus thuringiensis subsp. jegathesan

Abstract

The total protoxin complement in the parasporal body of mosquitocidal strain, Bacillus thuringiensis subsp. jegathesan 367, was determined by use of a polyacrylamide gel block coupled to mass spectrometry. A total of eight protoxins were identified from this strain, including five reported protoxins (Cry11Ba, Cry19Aa, Cry24Aa, Cry25Aa, and Cyt2Bb), as well as three previously undescribed (Cry30Ca, Cry60Aa, and Cry60Ba) in this isolate. It was interesting that the encoding genes of three new protoxins existed as cry30Ca-gap-orf2 and cry60Ba-gap-cry60Aa. The cry30Ca and a downstream orf2 gene were oriented in the same direction and separated by 114 bp, and cry60Ba was located 156 bp upstream from and in the same orientation to cry60Aa. The three new protoxin genes were cloned from B. thuringiensis subsp. jegathesan and expressed in an acrystalliferous strain under the control of cyt1A gene promoters and the STAB-SD stabilizer sequence. Recombinant strain containing only cry30Ca did not produce visible inclusion under microscope observation, while that containing both cry30Ca and orf2 could produce large inclusions. Cry60Aa and Cry60Ba synthesized either alone or together in the acrystalliferous host could yield large inclusions. In bioassays using the fourth-instar larvae of Culex quinquefasciatus, Cry60Aa and Cry60Ba alone or together had estimated 50% lethal concentrations of 2.9 to 7.9 μg/ml; however, Cry30Ca with or without ORF2 was not toxic to this mosquito.

INTRODUCTION

Products based on Bacillus thuringiensis subsp. israelensis have been widely used worldwide as a bacterial insecticide for controlling mosquito larvae (1). The high toxicity of this bacterium is mainly attributed to an arsenal of protoxins, such as Cry4Aa, Cry4Ba, Cry11Aa, and Cyt1Aa, harbored in its parasporal crystals, and their synergistic interaction. Such features also effectively suppressed resistance toward B. thuringiensis subsp. israelensis in mosquito populations (2–5). Although no sustained resistance to B. thuringiensis subsp. israelensis in the field has been observed yet, laboratory selection using its individual protoxins showed that resistance can evolve under heavy continuous applications (6, 7). Consequently, there is a good deal of interest in the isolation of supplemental mosquitocidal strains of B. thuringiensis whose toxic properties differ from those of B. thuringiensis subsp. israelensis. To date, a number of other mosquitocidal B. thuringiensis strains of various subspecies, such as morrisoni, canadensis, medellin, entomocidus, and jegathesan, have been reported (8–10).

Among these isolates, B. thuringiensis subsp. jegathesan 367 has received much attention since it is nearly as toxic to a variety of mosquito species as B. thuringiensis subsp. israelensis but produces different protoxins. For example, B. thuringiensis subsp. jegathesan is as toxic as B. thuringiensis subsp. israelensis to Anopheles stephensi and is only slightly less toxic to Aedes aegypti and Culex pipiens but is more toxic when tested against Culex quinquefasciatus (11, 12). To date, five protoxins, Cry11Ba (11), Cry19Aa (13), Cry24Aa (14), Cry25Aa (14), and Cyt2Bb (15), have been identified and characterized from B. thuringiensis subsp. jegathesan. The following methods were used to identify these protoxins and their encoding genes. The amino-terminal sequence of the target protein band on SDS-PAGE analysis was determined by N-terminal sequencing, which was then used to generate proper hybridization probes for selecting the positive clones containing the corresponding gene from total DNA libraries.

It was reported that mass spectrometry (MS) techniques could be used to identify B. thuringiensis protoxins (16, 17). Recently, we reported a convenient and straightforward method for analyzing the overall protoxin composition in B. thuringiensis strains by using a polyacrylamide gel block coupled to liquid chromatography-tandem mass spectrometry (LC-MS/MS) (18). This improved method had advantages in terms of accuracy and efficiency over traditional protein-based means, as well as one-dimensional SDS-PAGE coupled to an MS technique, when analyzing protoxins with high sequence homology produced in a single strain (18–20). During the search for new toxins from various B. thuringiensis strains using this improved MS method, we analyzed the protoxin complement in the parasporal body of B. thuringiensis subsp. jegathesan 367. Three protoxins (Cry30Ca, Cry60Aa, and Cry60Ba), which have not been described in this strain, were identified utilizing this improved MS method. These protoxin genes were cloned and expressed in the crystal-negative strain under the control of cyt1A gene promoters and the STAB-SD stabilizer sequence. The toxicities of the expressed products from recombinant strains were tested with C. quinquefasciatus.

MATERIALS AND METHODS

Bacterial strains and plasmids.

B. thuringiensis subsp. jegathesan 367 was obtained from the International Entomopathogenic Bacillus Center at the Pasteur Institute, France. Escherichia coli DH5α and acrystalliferous B. thuringiensis subsp. israelensis 4Q7 were used as host strains for transformation. The plasmid pSTAB was the source of cyt1A promoters and the STAB-SD sequence (21, 22), and the shuttle vector pHT315 (23) was used as a cloning vector.

Incorporation of protein samples into polyacrylamide gel block.

B. thuringiensis subsp. jegathesan 367 was grown to sporulation in GYS medium (22) at 30°C with shaking. The crystal-spore mixture was collected by centrifugation and resuspended in 0.5 M NaCl, sonicated for 5 min on ice, and then washed three times with deionized water. Subsequently, the crystal-spore mixture was suspended in the solubilization buffer (50 mM Tris-HCl [pH 6.8], 2.0% SDS, 5% β-mercaptoethanol) and boiled for 5 min. The protein solution was directly incorporated into a 10% polyacrylamide gel block as described previously (18) with minor modifications. The protein solution (5 μg of protein), 30% acrylamide/bisacrylamide, deionized water, 10% ammonium persulfate, and TEMED (N,N,N′,N′-tetramethylethylenediamine) were mixed in a 0.5-ml Eppendorf tube. After polymerizing for 1 h at room temperature, the proteins embedded in the gel block were fixed with 50% methanol–10% acetic acid–40% H₂O.

In-gel digestion and LC-MS/MS analysis.

The gel block embedding protein mixture was washed with 25 mM NH₄HCO₃, dehydrated with acetonitrile, and then dried in a vacuum centrifuge. The gel-bound proteins were reduced in 10 mM dithiothreitol and then alkylated in 55 mM iodoacetamide in the dark at room temperature for 45 min. After washing with acetonitrile (ACN), the gels were dried in a vacuum centrifuge. The dried gel pieces were reswollen in 50 μl of 25 mM NH₄HCO₃ containing 0.25 μg of proteomics-grade trypsin (Sigma-Aldrich). After an incubation of 16 h at 37°C, the released peptides were extracted twice with a 50% acetonitrile solution. The supernatants were pooled and lyophilized in a vacuum centrifuge.

LC-MS/MS experiments were performed on an LTQ XL mass spectrometer (Thermo Fisher Scientific, USA) coupled to a Finnigan LC system (Thermo Fisher Scientific). The digested peptides dissolved in an aqueous solution of 0.1% formic acid (FA) were loaded and desalted online on a reverse-phase precolumn (C18 PepMap column; LC Packings), and then resolved on a nanoscale C18 PepMap capillary column (LC Packings) at a flow rate of 200 nl/min with a gradient of ACN–0.1% FA before injection into the ion trap mass spectrometer. Peptides were separated using a 65-min linear gradient from 0 to 100% ACN in 0.1% FA. Spectra were obtained in a data-dependent acquisition mode, which consisted of a survey scan over the m/z range of 300 to 2,000, followed by five scans on the most intense precursor ions activated for 30 ms by the excitation energy level of 35%.

Data analysis.

With the SEQUEST search engine, the raw MS/MS data were searched against the in-house database, which were downloaded as a FASTA formatted sequences, including all B. thuringiensis subspecies from NCBI. The search parameters were as follows: tryptic specificity allowed for up to two missed cleavage, the mass tolerance of precursor and fragment ions were set at 1.5 and 1.0 Da, and carbamidomethylation of cysteines and methionine oxidation were set as fixed and variable modifications, respectively. The target-decoy search strategy was applied by searching the MS/MS spectra against the reversed and randomized B. thuringiensis proteome sequences to assess the rate of false-positive identifications. All of the peptide matches were filtered on the basis of their false-positive rate (<1%). Proteins with more than two unique peptides or a single unique peptide that had at least seven amino acids and had a high-quality spectra containing at least three consecutive b or y ions were considered reliably identified.

Cloning of the B. thuringiensis subsp. jegathesan toxin genes.

Three protoxins (Cry30Ca, Cry60Aa, and Cry60Ba) which have not been described in B. thuringiensis subsp. jegathesan were identified by our MS analysis. Their encoding genes were then cloned from B. thuringiensis subsp. jegathesan 367 and expressed in the acrystalliferous B. thuringiensis subsp. israelensis 4Q7. In order to render these genes to be expressed under the control of the strong chimeric cyt1A-p/STAB-SD promoter, the 0.6-kb cyt1A promoter and STAB-SD combination sequence was first obtained by PCR from plasmid pSTAB using the primers STAB-F and STAB-R (Table 1) and then ligated into plasmid pHT315 between HindIII and SalI, producing recombinant plasmid pSTAB-J.

Table 1.

Primers used for PCR amplification in this study

Primer	Sequence (5′–3′)a
STAB-F	GCGCAAGCTTGATTTCAAATTTTCCAAACTTAAA
STAB-R	GCGCGTCGACCTTTCTTATCATAATACATAATTTTCA
11Bter-F	AAAAGGATCCGAAAACAATGAAAAAGCATTCCCCTC
11Bter-R	GCGCGAATTCTTGTATGCCATCAAGAAAAAATATTATGG
30C-F	AAGCGTCGACAAAGAAGAGGGGGCATGTTTTAAATGAATCTTTATGGGAATAAGAATG
30C-R	GACGGGATCCTTAGTTTACTGTACAAGGTACTACACCTTG
30CORF2-R	GGTTGGATCCTTATCCTTTCGCGCAAAGTAATTCGATGCTTTCTACATA
60B-F	AAGCGTCGACAGGGAGGCATTTTAATGACAATT
60B-R	GACGGGATCCCATAATAACTATTTTTGTTATAAG
60A-F	AAGCGTCGACTGCGAGGGGGAATATCTTTATGG
60A-R	GACGGGATCCTTCACCATTAGGCCATAAACACAT

^aRestriction endonuclease cleavage sites for HindIII, SalI, BamHI, and EcoRI are underlined. ATG codons are shown in boldface. The ribosome-binding site for cry30Ca expression is derived from cry11Ba1 and is shaded.

For Cry30Ca synthesis, PCR primers (Table 1) were designed based on the published cry30Ca1-s2orf2 sequence in B. thuringiensis subsp. sotto (GenBank accession no. AB193814). Because the promoter and terminator sequence of cry30Ca1-s2orf2 were not available, the putative ribosome-binding site and terminator sequence from cry11Ba were utilized for toxin expression. The 202-bp terminator sequence was obtained from strain 367 by PCR with the primers 11Bter-F and 11Bter-R and then cloned into pSTAB-J between BamHI and EcoRI, producing plasmid pSTAB-JT. Primers 30C-F (containing the putative ribosome-binding site of cry11Ba) and 30C-R were used to amplify cry30Ca, while the primers 30C-F and 30CORF2-R were used to amplify cry30Ca-orf2 from strain 367. The obtained 2.1- and 3.6-kb PCR fragments were cloned into pSTAB-JT between SalI and BamHI, respectively, producing expression plasmids pST30C and pST30CORF2.

All primers used for Cry60 synthesis were designed based on the published sequence in B. thuringiensis subsp. israelensis ATCC 35646 (GenBank accession no. AAJM01000002, positions 25287 to 27600), in which cry60Aa is located downstream of cry60Ba. Three DNA fragments harboring cry60Ba and cry60Aa, as well as cry60Ba and cry60Aa, together were obtained from strain 367 by PCR using the primers 60B-F and 60B-R and the primers 60A-F and 60A-R, as well as the primers 60B-F and 60A-R, respectively. The three PCR products, with sizes of 1.1, 1.2, and 2.3 kb, were cloned into pSTAB-J between SalI and BamHI, producing expression plasmids pPFT60B, pPFT60A, and pPFT60AB. All expression plasmids were transformed into the acrystalliferous strain 4Q7 by electroporation (22).

Protein analysis and microscopic observation.

B. thuringiensis recombinants were grown to sporulation in GYS medium containing 25 μg of erythromycin/ml at 30°C with shaking. The crystal-spore mixture was obtained by centrifugation and resuspended in 0.5 M NaCl, sonicated for 5 min on ice, and then washed three times with deionized water. The crystal-spore mixture was then subjected to SDS–12% PAGE analysis. Different samples were treated in parallel for the quantification of relative toxin yields by using the densitometry software. In order to confirm the expressed proteins, specific bands on the SDS-PAGE gel were sliced and digested by trypsin and then subjected to LC-MS/MS analysis individually. Observation of the sporulating cultures at the sporangium stage was conducted with phase-contrast microscope (Zeiss Photomicroscope III).

Bioassays.

Lyophilized crystal-spore mixtures were suspended in deionized water. Suspensions were diluted to six or seven different concentrations in 6-oz. cups in a final volume of 100 ml. Bioassays were replicated three times using 30 fourth-instar larvae of C. quinquefasciatus (S-Lab strain) per concentration. After 48 h of exposure at 28°C, dead larvae were counted, and the median (50%) lethal concentration (LC₅₀) and the LC₉₅ were calculated by Probit analysis.

Nucleotide sequence accession numbers.

The nucleotide sequences for cry30Ca and cry60Aa in strain 367 have been deposited in the GenBank database (accession no. GQ368655 and GQ398500) and designated cry30Ca2 and cry60Aa1 by the Bacillus thuringiensis Toxin Nomenclature Committee.

RESULTS

Assessment of protoxin composition.

By utilizing a polyacrylamide gel block coupled to LC-MS/MS analysis, we determined the amino acid sequences of a number of tryptic peptides. According to the SEQUEST search results, different proteins were suggested and sorted by the total score based on the identified peptides. Since Cry proteins often share high sequence homology, several Cry proteins might be suggested by SEQUEST software based on an identified conserved peptide. Thus, manual inspection was conducted for ensuring the certainty of suggested protoxins. The toxins with lower scores should have several discriminating peptides that did not appear in those with a higher score. Based on this principle, we detected five reported protoxins (Cry11Ba, Cry19Aa, Cry24Aa, Cry25Aa, and Cyt2Bb), as well as three previously undescribed (Cry30Ca, Cry60Aa, and Cry60Ba), in the parasporal body of B. thuringiensis subsp. jegathesan 367 (Table 2).

Table 2.

Summary of results from the protein profiling experiment in B. thuringiensis subsp. jegathesan 367 by using a gel block coupled to LC-MS/MS^a

Accession no.	Protein	Mol mass (Da)	Total SEQUEST score	No. of unique peptides	Total no. of queries matched	Sequence coverage (%)
gi|8928012	Cry24Aa	75,911.5	885.52	16	147	36.80
gi|8928023	Cry11Ba	81,293.2	582.23	18	92	41.99
gi|54695305	Cry30Ca	77,389.6	497.77	6	88	15.12
gi|8469212	Cyt2Bb	30,040.7	346.95	3	56	26.24
gi|8928013	Cry25Aa	75,595.8	194.63	7	26	19.85
gi|8928007	Cry19Aa	74,696.1	149.81	7	17	22.69
gi|75758299	Cry60Ba	34,967.7	121.52	6	18	36.36
gi|75758300	Cry60Aa	33,834.2	87.69	6	11	27.39
gi|75362008	ORF2-30C	55,184.8	330.84	9	45	31.28
gi|75490097	ORF2-19A	60,023.6	202.47	6	27	26.81

^aProtoxins identified in strain 367 were sorted by the total SEQUEST score.

Detection of ORF2 proteins.

Besides eight protoxins mentioned above, we also detected two proteins (gi numbers 75362008 and 75490097) belonging to open reading frame 2 (ORF2), which was featured by high identity with the C-terminal half of 130-kDa Cry toxin. The two ORF2 proteins were associated with Cry30Ca and Cry19Aa, respectively, and were designated ORF2-30C and ORF2-19A (Table 2). The amino acid sequence alignment showed that ORF2-30C and ORF2-19A were 63 and 66% identical to the C-terminal half of Cry4Aa. As supporting information, all protoxins and other proteins identified in strain 367 by use of a gel block coupled to LC-MS/MS, along with the peptides on the basis of which they are identified, was provided in Table S1 in the supplemental material.

Sequence analysis.

The 3.6-kb PCR fragment amplified from strain 367 by primers 30C-F and 30CORF2-R contained two ORFs oriented in the same direction and separated by 114 bp: cry30Ca with 2067 bp and orf2-30C with 1,458 bp. The predicted masses of Cry30Ca and ORF2-30C were 77.3 and 55.1 kDa. The Cry30Ca protein contained the five conserved blocks present in most of Cry toxins. Analysis of 2.3-kb PCR product indicated that the cry60Ba was located 156 bp upstream from and in the same orientation to cry60Aa in strain 367. The cry60Aa and cry60Ba were 912 and 960 bp long, encoding polypeptides with predicted masses of 33.8 and 34.9 kDa. No conserved blocks were found in these two toxins. Amino acid sequence alignment showed that Cry30Ca and ORF2-30C of strain 367 were 97.0% identical to the putative Cry30Ca1 and its associated S2ORF2 of B. thuringiensis subsp. sotto, respectively. Cry60Aa and Cry60Ba were 99 and 100% identical to the putative products of cry60Aa and cry60Ba of B. thuringiensis subsp. israelensis ATCC 35646. Meanwhile, Cry60Aa and Cry60Ba of B. thuringiensis subsp. jegathesan were 43% identical to each other. No significant sequence similarity was observed between Cry60 and other known Cry and Cyt protoxins.

Expression of Cry30 and Cry60 proteins.

All recombinant strains containing cry30 and cry60 sporulated normally in GYS medium. No obvious inclusion was visible under phase-contrast microscope for the recombinant strain 4Q7(pST30C) harboring only cry30Ca, while large inclusion was formed in 4Q7(pST30CORF2) which contained both cry30Ca and orf2-30C genes (Fig. 1A). SDS-PAGE analysis showed that 4Q7(pST30C) expressed two relatively faint bands at around 60 and 77 kDa, which were both identified as Cry30Ca by MS analysis. The band of 60 kDa was apparently derived from degraded product of intact Cry30Ca. Two major products of 77 and 55 kDa were detected in inclusions obtained from 4Q7(pST30CORF2), which were verified by MS analysis to be the products of cry30Ca and orf2-30C, respectively (Fig. 1B; see Table S2 in the supplemental material). Analysis of protein bands with densitometry software showed that the amount of Cry30Ca synthesized by 4Q7(pST30CORF2) was 5.9-fold greater than that produced by 4Q7(pST30C). These results suggested that ORF2 could assist in stability and crystallization of Cry30Ca.

Fig 1.

Expression of cry30Ca with or without orf2-30C. In both constructs, expression was under the control of the cyt1A-p/STAB-SD sequence. (A) Phase-contrast micrographs of recombinant B. thuringiensis 4Q7 strains producing Cry30Ca. (a) Wild-type strain 367; (b) 4Q7(pST30C); (c) 4Q7(pST30CORF2). Arrows indicate crystal inclusions. (B) SDS-PAGE analysis of the recombinant B. thuringiensis 4Q7 strains. Lane M, molecular mass marker; lane 1, wild-type strain 367; lane 2, 4Q7(pST30C); lane 3, 4Q7(pST30CORF2). Cry30Ca and ORF2-30Ca are indicated by arrows.

For the three recombinant strains harboring cry60, a large inclusion was observed under a phase-contrast microscope (Fig. 2A), and the expected protein band of 33 to 35 kDa was demonstrated by SDS-PAGE (Fig. 2B). MS analysis of the target protein band confirmed that 4Q7(pPFT60A) expressed only Cry60Aa toxin and that 4Q7(pPFT60B) synthesized only Cry60Ba toxin. Both toxins were expressed in recombinant strain carrying pPFT60AB (see Table S2 in the supplemental material). Densitometry scanning of the SDS-PAGE gel revealed that the amounts of Cry60 toxin synthesized by the recombinant strains 4Q7(pPFT60A), 4Q7(pPFT60B), and 4Q7(pPFT60AB) were 3.1-, 9.6-, and 7.0-fold greater, respectively, than that produced by the wild-type strain 367.

Fig 2.

Expression of cry60Aa and cry60Ba alone or together. In all constructs, expression was under the control of the cyt1A-p/STAB-SD sequence. (A) Phase-contrast micrographs of recombinant B. thuringiensis 4Q7 strains producing Cry60. (a) 4Q7(pPFT60A); (b) 4Q7(pPFT60B); (c) 4Q7(pPFT60AB). Arrows indicate crystal inclusions. (B) SDS-PAGE analysis of the recombinant B. thuringiensis 4Q7 strains. Lane M, molecular mass marker; lane 1, wild-type strain 367; lane 2, 4Q7(pPFT60A); lane 3, 4Q7(pPFT60B); lane 4, 4Q7(pPFT60AB); lane 5, B. thuringiensis 4Q7 transformed with pHT315. Cry60 toxins are indicated by arrows.

Mosquito larvicidal activity.

Mosquitocidal activity was determined by using fourth-instar larvae of C. quinquefasciatus. Crystal-spore preparations obtained from lysed cultures were used for all strains in the bioassays (Table 3). The recombinant strains expressing Cry60 showed moderate toxicity against C. quinquefasciatus with LC₅₀s of 7.9 μg/ml for 4Q7(pPFT60A), 5.5 μg/ml for 4Q7(pPFT60B), and 2.9 μg/ml for 4Q7(pPFT60AB). However, confidence limits of the median lethal concentrations for the three recombinant strains were overlapped, which indicated that there was no statistically significant synergism between Cry60Aa and Cry60Ba. No toxicity was observed with recombinant strains expressing Cry30Ca alone or both Cry30Ca and ORF2, even at a concentration of 200 μg/ml.

Table 3.

Mosquitocidal activities of the recombinant B. thuringiensis strains producing Cry30 and Cry60 protoxins

Strain	Toxin composition	LC (μg/ml)a		Mean slope ± SEM
		LC50	LC95
4Q7(pPFT60A)	Cry60Aa	7.9 (4.3–24.8)	569.5 (104.9–29,850.5)	0.9 ± 0.2
4Q7(pPFT60B)	Cry60Ba	5.5 (3.8–9.3)	70.3 (30.1–358.1)	1.5 ± 0.3
4Q7(pPFT60AB)	Cry60Aa+Cry60Ba	2.9 (1.8–5.6)	166.6 (46.9–2,226.5)	0.9 ± 0.2
4Q7(pST30C)	Cry30Ca	NTb	NT	NT
4Q7(pST30CORF2)	Cry30Ca+ORF2	NT	NT	NT

^aValues are shown in μg of spore-crystal mixture per ml. Confidence limits are indicated in parentheses.

^bNT, not toxic at 200 μg/ml.

DISCUSSION

Our previous report indicated that combining a polyacrylamide gel block and LC-MS/MS could be viably applied for the analysis of protoxins, as shown by the identification of the expected protoxins from B. thuringiensis subsp. kurstaki HD1 and subsp. israelensis 4Q2-72. In the present study, we adopted such a method to evaluate protoxin complement in the mosquitocidal B. thuringiensis subsp. jegathesan 367. Three protoxins (Cry30Ca, Cry60Aa, and Cry60Ba) were newly found in strain 367 in addition to the five previously reported protoxins. The discovery of additional protoxins showed that the parasporal body of B. thuringiensis subsp. jegathesan was more complex than previously understood.

The cry30Ca1 and its associated s2orf2 gene were found in B. thuringiensis subsp. sotto, but there was no report about their encoded products. In our study, the 77-kDa Cry30Ca2 and 55-kDa ORF2-30C were identified in the parasporal body of B. thuringiensis subsp. jegathesan 367, and their encoding genes were found to be separated by a 114-bp interval. The similar gene structure (ORF1-gap-ORF2), in which a typical 130-kDa toxin gene was split into two separate ORFs, has been found in some other instances, such as cry10A (24), cry19A (13, 25), cry39A (26), cry40A (GenBank accession no. AB074414), and cry44A (8). The products of the two ORFs resembled the N-terminal half and the C-terminal half, respectively, of a 130-kDa toxin. At present, this gene configuration is mainly limited to diptera-specific B. thuringiensis strains, and the events leading to this arrangement are unknown. It may have evolved through the insertion of a DNA fragment into a gene that would otherwise encode a 130-kDa protein or mutations accrued in the intergenic region between the two ORFs.

Cry30Ca was poorly expressed on its own and appeared degraded in SDS-PAGE analysis. More Cry30Ca was produced and obvious inclusion was formed when ORF2-30C was synthesized simultaneously. ORF2-30C was 92% identical to the ORF2-19A, which is associated with Cry19Aa. The report by Barboza-Corona et al. suggested that ORF2-19A assisted in the synthesis and crystallization of Cry19Aa by functioning like a C-terminal crystallization domain (25). Thus, we speculated that the ORF2-30C might also act like a C-terminal region for facilitating inclusion body formation of Cry30Ca.

The two ORFs for cry60Aa and cry60Ba were first found in B. thuringiensis subsp. israelensis ATCC 35646 by genome sequencing and originally proposed to encode Cry15Aa-like toxins (27). However, their expressed products have not been reported previously. In the present study the 33- and 35-kDa products of the two ORFs were detected in the parasporal body of B. thuringiensis subsp. jegathesan 367 through proteomic technique and then named Cry60Aa and Cry60Ba by the Bacillus thuringiensis Toxin Nomenclature Committee. In addition, cry60Aa and cry60Ba and their products were also detected in B. thuringiensis subsp. malayensis 4AV1 (obtained from the Bacillus Genetic Stock Center, Ohio State University, Columbus) by our study (data not shown). It is interesting that cry60Ba-gap-cry60Aa gene structure existed in these three strains belonging to different subspecies. In B. thuringiensis subsp. israelensis ATCC 35646, an insertion element, IS240, is located upstream of the cry60Ba gene, and a transposable element, Tn5401, is in the downstream of cry60Aa gene. These elements may be related to the distribution of toxin genes in various B. thuringiensis strains.

The cry60Ba-gap-cry60Aa gene structure seems different from those reported previously. The 33-kDa Cry60Aa and the 35-kDa Cry60Ba synthesized either alone or together in the B. thuringiensis acrystalliferous host could yield large inclusion bodies. However, B. thuringiensis transformants containing cry19Aa or orf2-19A alone could not produce visible inclusions (25). Similar results were also observed for the 34- and 40-kDa proteins from B. thuringiensis subsp. thompsoni (28) and the 32- and 40-kDa proteins from B. thuringiensis subsp. dakota (29). However, Cry60Aa and Cry60Ba should not be referred to as binary toxins (30) since both proteins were not required simultaneously for mosquitocidal activity. The regulation of cry60Ba-gap-cry60Aa expression in B. thuringiensis subsp. jegathesan is not known at this stage, and further investigations are needed.

Bioassay results showed that Cry60 toxins only showed moderate toxicity and that Cry30Ca had no activity against C. quinquefasciatus, although these protoxins were derived from B. thuringiensis subsp. jegathesan, which is highly toxic to this mosquito, with a reported LC₅₀ of 0.03395 μg/ml in our previous study (31). This finding strongly confirmed that the other toxins coexisting in the parasporal body, such as Cry11Ba, were responsible for its high toxicity against C. quinquefasciatus (11, 14). A previous study indicated that the cloned Cry19Aa was toxic to A. stephensi and C. pipiens but not to A. aegypti, even though B. thuringiensis subsp. jegathesan showed high toxicity to A. aegypti (13). Thus, we intend to evaluate the toxicity of Cry30 and Cry60 against other mosquitoes. Meanwhile, the synergism between these protoxins and the coexisting ones in B. thuringiensis subsp. jegathesan deserved to be determined. Indeed, it was reported that Cry29A from B. thuringiensis subsp. medellin strain 161-131 was not toxic to the tested dipteran larvae but showed a synergistic activity with the coexisting mosquitocidal Cry11Bb (32).

Protoxin components and their relative abundance are both helpful for knowing about the feature of parasporal body of B. thuringiensis. Even though MS analysis alone is not a quantitative method, there are several empirical indications that can help to estimate the relative abundance of a protein in a mixture. Among the reported indicators, the protein abundance index (PAI), which represents the number of identified peptides divided by the number of theoretically observable tryptic peptides per protein, is considered a direct parameter to describe the abundance and has been applied to LC-MS/MS analysis of human spliceosome complex (33). Meanwhile, total score of the identified protein in MS analysis could also reflect its abundance in a mixture. PAI value and total score work fairly well as long as proteins in the mixture are of similar size (34). For example, Cry24Aa, Cry30Ca, Cry25Aa, and Cry19Aa in strain 367 were in the 74- to 77-kDa mass range. We obtained the following PAI values and total SEQUEST scores for the four protoxins: Cry24Aa (0.13, 885.52), Cry30Ca (0.10, 497.77), Cry25Aa (0.06, 194.63), and Cry19Aa (0.05, 149.81). Thus, it is fair to assume that the relative proportion of Cry24Aa was higher than those of the other three toxins. We noted that the GTG start codon was used in the cry24Aa gene, while the other three toxin genes used the conventional ATG start codon. It seemed that the expression level of cry24Aa was not influenced by its rare start codon.