A Novel Bacillus thuringiensis Cry-Like Protein from a Rare Filamentous Strain Promotes Crystal Localization within the Exosporium

Abstract

Mutation of a novel cry-like gene (cry256) from Bacillus thuringiensis resulted in a protein crystal, normally located within the spore's exosporium, being found predominately outside the exosporium. The cry256 gene codes for a 3-domain Cry-like protein that does not correspond to any of the known Cry protein holotypes.

TEXT

During sporulation, the bacterium Bacillus thuringiensis produces a protein crystal comprised of one or more members of the Cry family of protein toxins. The crystal is normally located outside the exosporium, which surrounds the spore (1); however, relatively few strains of B. thuringiensis have been shown to localize their crystals within the exosporium (2, 3). With the aim of understanding how some crystals are localized within the exosporium, Debro et al. showed that in B. thuringiensis subsp. finitimus, the genes necessary for localizing crystals within the exosporium where located on two plasmids (4). Zhu et al. later identified these genes (5) and showed that not all B. thuringiensis organisms used these genes to localize crystals within their exosporium. Building on this previous work, we undertook a study to identify genes responsible for crystal localization in a novel filamentous strain of B. thuringiensis called Bt2-56 (Fig. 1A and A′) (6, 14).

Fig 1.

(A) B. thuringiensis strain Bt2-56 stained with Coomassie blue; (A′) transmission electron micrograph of negatively stained BT2-56; (B) M1-11 stained with Coomassie blue; (B′) scanning electron micrograph of M1-11; (C) 15-54 stained with Coomassie blue; (C′) scanning electron micrograph of 15-54. Physical structures are the exosporium (e), spore (s), crystal (c), and filament bundle (f). nm, not mutated.

Inactivation of gene cry256 causes loss of crystals from the exosporium.

Genes in Bt2-56 were mutated by random transposon insertion using plasmid pMarA (7). Mutants with crystals located outside the exosporium were identified by viewing stained preparations of sporulated mutants (8). Approximately 84% of the crystals from a mutant called M1-11 were observed outside the exosporium (Fig. 1B and B′).

Inverse PCR using primers left-tn1 and oIPCR2a (Table 1) was used to amplify M1-11 genomic DNA flanking the transposable element, indicating that transposons had integrated into both a histidine kinase gene and a cry-like gene that we call cry256. The DNA sequence of both strands of the entire cry256 gene and its immediate flanking regions was then determined by “walking” the gene, using primers designed from previously sequenced DNA regions (GenBank accession number JQ670887). The DNA sequence of cry256 from Bt2-56 (the unmutated parent strain of M1-11) was also determined, using an Illumina GAIIx sequencer at the University of Houston's IMD Sequencing Center. The 36-bp sequencing reads were assembled using Velvet (9) and visualized with the software program Tablet (10). DNA translations were performed using a Web-based DNA-to-protein tool (11; the Web application utilized is located at http://insilico.ehu.es/translate/). The transposon was found to have inserted into a region that separated the last 49 amino acids of the protein encoded by cry256 from the rest of protein (see file S1 in the supplemental material [the insertion site highlighted in red at position 1528]). Aside from the transposon, the DNA sequence of cry256 in mutant M1-11 was identical to that obtained for the wild-type cry256 gene in Bt2-56.

Table 1.

Primers used in this study

Primer	Sequencea	Use
left-tn1	CGCAACTGTCCATACTCTGAT	Transposon-specific primer
oIPCR2a	GGGAATCATTTGAAGGTTGG	Transposon-specific primer
left-tn2	AGTTCGCTAGATAGGGGT	Transposon-specific primer
oIPCR3	See reference 7	Transposon-specific primer
256e-F	AAAAGAATTCCAAATGAGGGATTTATGTGG	cry256 amplification
256e-R	TGGAGAATTCTAATATCCAGTCGTTTCCTGC	cry256 amplification
m1-11rc6	AAAAAGTCCATAGGAGTCATTGTT	Confirm cry256 disruption
Ltest	ACGCTTTCTATCGACCTTCTGGAC	Confirm cry256 disruption

^aUnderlining identifies the EcoRI recognition sequence.

Since the cause-and-effect relationship of mutating cry256 and crystal localization in M1-11 was complicated by the insertion of a second transposon into a histidine kinase gene, we confirmed the role of cry256 in crystal localization by independently inactivating cry256 via homologous recombination. Primers 256e-F and 256e-R were used to amplify a subregion of cry256 that was cloned into a unique EcoRI site in pKO1 (a derivative of pMarA that lacks the transposon), creating pKO1-256. The cry256 gene in Bt2-56 was inactivated by inserting pKO1-256 into cry256 via a single-crossover event, creating mutant 15-54 (Fig. 2), which had a phenotype similar to that of mutant M1-11 (Fig. 1C and C′). Disruption of cry256 in mutant 15-54 was confirmed by PCR using primers m1-11rc6 and Ltest, which amplify a specific-size fragment only if pK01-256 crosses into cry256 (Fig. 2).

Fig 2.

Inactivation of cry256 gene in Bt2-56 via homologous recombination. (I) Plasmid pKO1-256; (II) cry256 gene in Bt2-56; (III) strain 15-54. The checkered area denotes cry256 DNA, the thin solid-black line denotes the pKO1-256 plasmid backbone, and the thick solid-black line denotes genomic DNA flanking cry256. Region IIIA is the cry256 5′ flanking region and a partial cry256 gene, region IIIB is the plasmid pKO1-256 backbone, and region IIIC is the entire cry256 gene along with 16 bp of its 5′ flanking sequence and 3′ flanking sequence. (IV) PCR confirmation of mutant 15-54 using primers 1 (m1-11rc6) and 2 (Ltest) with different templates. Lane 1, DNA markers (asterisks denote 1-kb and 1.5-kb bands); lane 2, Bt2-56 genomic DNA; lane 3, M1-11 genomic DNA; lane 4, 15-54 genomic DNA; lane 5, an equimolar mix of Bt2-56 genomic DNA and plasmid pKO1-256. Only the 15-54 template (lane 4) produced the 1,055-bp amplicon indicative of recombination (III). Lanes 6 through 8 contained the same templates as lanes 1 through 4, respectively, but were amplified with primers 256e-F and 256e-R, which amplified a 994-bp region of cry256 and indicated that all the templates were amplifiable.

The cry256 gene codes for a Cry-like protein.

The cry256 gene codes for P256, a 1,098-amino-acid protein of approximately 127 kDa, and differs from genes involved with crystal localization in B. thuringiensis subsp. finitimus (5). The NCBI's homology search tool BLASTP (12) indicated that the B. thuringiensis Cry protein Cry21Ba1 (GenBank accession number BAC06484), a nematicidal protein, had the highest amino acid homology to P256, followed by three other Cry proteins also associated with nematicidal activity (Cry5, Cry12, and Cry14). An amino acid comparison of P256 with these other nematicidal genes (see file S1 in the supplemental material) revealed a strongly conserved carboxyl region corresponding to the Cry protein crystallization domain, the highly divergent “linker” region commonly located between domain III of a 3-domain toxin and the crystallization domain, with noticeable conservation of the 5-amino-acid blocks that broadly show conservation among Cry proteins exhibiting the 3-domain structure (3, 13), other conserved protein blocks showing conservation among nematicidal proteins and P256 (blocks 7 to 9), and a highly truncated domain II region. Domain II of P256 is 83 amino acids in length, while those of the other nematicidal proteins are 189, 213, 215, 204, and 205 amino acids in length.

cry256 is probably not found in the exosporium-localized crystal.

Loss of cry256 did not result in an observable loss in crystal size or shape (data not shown). Furthermore, SDS-PAGE analysis of gradient-purified crystals from the unmutated strain Bt2-56 showed only one band in the 100-kDa range. Automated Edman degradation-based N-terminal protein sequencing of this protein (Protein Analysis Facility, Institute for Cellular and Molecular Biology, University of Texas, Austin, TX, USA) showed that its amino-terminal sequence (MVQLDDLLPNYNNVLANP) differs from that calculated for cry256 (MDNSSNSSSISNNVLITP).

Cry256 appears to promote, but is not required, for crystal localization.

Both mutants M1-11 and 15-54 showed that even with a disrupted cry256 gene, some bacteria were still able to localize their crystals within the exosporium (Fig. 1B and C). This result indicates that cry256 may act to increase the efficiency of the cell's crystal localization machinery, as opposed to being an indispensable component. We are presently working to determine the identities of other genes involved with crystal localization in Bt2-56 and to determine how P256 promotes crystal localization.