Distribution, Diversity, and Potential Mobility of Extrachromosomal Elements Related to the Bacillus anthracis pXO1 and pXO2 Virulence Plasmids (original) (raw)

Abstract

The presence of a pXO1- and/or pXO2-like plasmid(s) in clinical isolates of Bacillus cereus sensu stricto and in strains of the biopesticide Bacillus thuringiensis has been reported recently, and the pXO2-like plasmid pBT9727 and another pXO2-like plasmid, pAW63, were found to be conjugative. In this study, a total of 1,000 B. cereus group isolates were analyzed for the presence of pXO1- and pXO2-like replicons and for the presence of pXO2-related conjugative modules. pXO1- and pXO2-like replicons were present in ca. 6.6% and 7.7% of random environmental samples, respectively, and ca. 1.54% of the strains were positive for pXO2-like transfer module genes. Only the strains harboring a pXO2-like replicon also contained the corresponding transfer genes. For the strains which contained a pXO1- and/or pXO2-like replicon(s), a large plasmid(s) whose size was similar to that of pXO1-like and/or pXO2-like plasmids was also observed, but none of these isolates were found to carry the Bacillus anthracis toxin or capsule virulence genes. Furthermore, 17 of 22 pXO2-like plasmids containing the transfer modules were able to self-transfer and to mobilize small plasmids. No pXO1- or pXO2-like plasmid lacking the cognate transfer modules has been found to have transfer potential. In the strains possessing the putative pXO2-like conjugative apparatus, variations in the presence of the group II introns B.th.I.1 and B.th.I.2 were observed, suggesting that there is important flexibility in the conjugation modules and their regulation. There was no consistent correlation between a pXO2-like repA dendrogram and the presence of the tra region or between a virB4 dendrogram and transfer ability. Discrepancies between pXO2-like repA and virB4 dendrograms were also observed, indicating that the evolution of pXO2 is an active process.

The Bacillus cereus group includes Bacillus anthracis, B. cereus sensu stricto, Bacillus thuringiensis, Bacillus mycoides, Bacillus pseudomycoides, and Bacillus weihenstephanensis. All members of this group are genetically closely related and could be considered members of a single species or taxon (15, 16, 24, 25). Several of their major phenotypic features, including their virulence spectra, are directly associated with large plasmids. In B. anthracis, the anthrax toxin and capsule genes responsible for anthrax disease are located on the 182-kb pXO1 and 95-kb pXO2 plasmids, respectively (26, 27). These plasmids are not self-transmissible but can be mobilized by conjugative plasmids, such as pXO14 from B. thuringiensis (29). The biopesticide properties that are the main characteristics of B. thuringiensis and distinguish it from B. cereus are due to large plasmids carrying cry genes. These genes are responsible for the production of different Cry toxins that form insecticidal crystal protein inclusions. Some of these large plasmids have been found to have conjugative transfer capability (12, 20). For instance, plasmid pHT73 from B. thuringiensis subsp. kurstaki carries cry toxin genes and is conjugative (41). Another cry toxin-encoding element, pBtoxis from B. thuringiensis subsp. israelensis, is a plasmid that carries cry toxin genes, and while it is not self-transmissible, it can be mobilized by other plasmids from B. thuringiensis, such as pAW63 or pXO16 (11, 21). The host strains of pHT73 and pBtoxis are two of the most widely used commercial B. thuringiensis biopesticide strains. In contrast to B. thuringiensis and B. anthracis, B. cereus sensu stricto is known mainly as an opportunistic pathogen of mammals, causing food-associated intoxications manifested by either emetic or diarrheal syndromes. The remaining members of the B. cereus group, B. mycoides, B. pseudomycoides, and B. weihenstephanensis, are distinguished on the basis of morphology (rhizoidal growth) (B. mycoides and B. pseudomycoides) and physiology (psychrotolerance) (B. weihenstephanensis), which seem to be determined mainly by chromosomal genetic determinants (24, 25). They may also have enteropathogenic potential (19a, 35, 36).

Recent studies have reported the presence of pXO1-like plasmids in clinical isolates of B. cereus, as well as the presence of pXO2-like plasmids in both a clinical strain and a commercial biopesticide strain of B. thuringiensis (8, 17, 28, 38). The pXO1-like plasmids pCER270, pPER272, pBC10987, and pBCXO1 were found to display a high degree of sequence similarity and synteny to pXO1, especially in the putative replication region and in certain segments that constitute the shared backbone of this plasmid group (28). pBCXO1 was the plasmid most similar to pXO1, with 99.6% nucleotide identity and full conservation of the anthrax pathogenicity island region. It was identified in a clinical isolate of B. cereus designated BC9241 that caused a severe case of pneumonia resembling pulmonary anthrax (18). Also of considerable interest is pCER270 (also designated pCERE01), a plasmid isolated from an emetic B. cereus strain that shares less sequence similarity with pXO1 but is based on the same backbone. This plasmid harbors a nonribosomal peptide synthetase gene cluster that includes the genetic determinants of cereulide, the cyclic dodecadepsipeptide responsible for the emetic syndrome type of _B. cereus_-induced food intoxication (8, 19, 28). The pXO2-like plasmids pAW63 (∼72 kb) and pBT9727 (∼77 kb) display a very high level of sequence similarity and synteny to pXO2 (17, 37, 38, 42). The only sizable region that is unique to pXO2 is a ∼30-kb pathogenicity island that contains the anthrax-specific capsule genes. Besides having similar replication genes, all three of these plasmids possess an approximately 40-kb transfer region containing genes encoding homologs of several key components of conjugative type IV secretion systems (T4SS), which mediate the delivery of protein and DNA substrates from donor cells to recipient cells during conjugative transfer (31).

The T4SS was originally characterized for the Ti (tumor-inducing) plasmid of Agrobacterium tumefaciens (33) and is the archetype of conjugative transfer complexes in gram-negative bacteria. The T4SS of A. tumefaciens is formed by 11 VirB proteins (VirB1 to VirB11) encoded by a single operon, and its operation involves an additional component, VirD4. VirB4 and VirB11 play a role in providing the energy for transfer, while VirD4 is postulated to be the coupling protein mediating the interaction between the relaxosome and mating pair formation systems (3, 4, 6, 43). All three proteins display ATPase activity and have been found to be necessary to mediate early steps of the DNA translocation pathway (3). Interestingly, homologs of the VirB4, VirB11, and VirD4 proteins appear to be ubiquitous not only in gram-negative bacteria but also in the known conjugative transfer systems of gram-positive bacteria termed type IV secretion-like systems (for a recent review, see reference 7). These observations suggest that these three proteins form a fundamental core of conjugative systems. Accordingly, homologs of VirD4, VirB11, and VirB4 were found in the transfer regions of pAW63, pBT9727, and pXO2.

Although the transfer regions of pAW63, pBT9727, and pXO2 are extensively conserved, several predicted transfer genes on pXO2, including the VirD4 gene homolog, are interrupted by what appear to be nonsense mutations or punctual frameshifts. These mutations or frameshifts may explain at least in part why pXO2 is not capable of self-transfer, yet can be mobilized by a helper plasmid (29). In contrast, both pAW63 and pBT9727 have been shown to be capable of promoting their own transfer, as well as that of small mobilizable plasmids, although the transfer frequencies of pAW63 were typically higher than those of pBT9727 (40).

In addition, two group II introns (B.th.I.1 and B.th.I.2) were identified in the transfer genes on pAW63. Group II introns are self-splicing, mobile retroelements, some of which have been shown experimentally to be able to invade new DNA sites and transfer between species (9, 10). Group II introns have been found in some gram-positive conjugative elements; e.g., the Ll.LtrB intron has been found in the relaxase gene of three Lactococcus lactis conjugative elements (34). B.th.I.1 was found in a coding sequence upstream of the virB4 homolog on pAW63. This intron contains a gene encoding an endonuclease protein (_i_ntron-_e_ncoded _p_rotein [IEP]) that is presumed to be responsible for its splicing (i.e., excision of the intron RNA from the transcribed exon gene mRNA). B.th.I.2 was found in the virD4 gene homolog of pAW63, but it does not possess an IEP gene. Neither B.th.I.1 nor B.th.I.2 was found in the corresponding region of either pBT9729 or pXO2. Considering the key role of the VirD4 homolog in the conjugative process, it is expected that excision of B.th.I.2 is required for expression of functional VirD4 by pAW63 and, consequently, for plasmid transfer to proceed. A recent study confirmed, by using reverse transcription-PCR, that B.th.I.2 was indeed spliced in vivo from pAW63 (37), although the necessity of this processing for effective plasmid transfer remains to be demonstrated.

Incremental discoveries, mostly in clinical settings, of B. cereus and B. thuringiensis strains that are “close neighbors” of B. anthracis are leading to a greater understanding of the genomic complexity of the B. cereus group, particularly with regard to clinically important pathotypes. Considering the close association of virulence determinants with the pXO1-like and pXO2-like plasmid gene pool of the group, it is reasonable to expect that assessing the prevalence, diversity, and transfer capabilities of such plasmids may provide useful information that complements clinical studies. In this study, a PCR-based strategy was developed to screen for the presence of pXO1- and/or pXO2-like replicons, as well as for the presence of transfer modules related to those of the pXO2-like conjugative plasmid pAW63. In addition, newly identified plasmids were tested to determine their abilities to promote their own transfer and the transfer of small mobilizable plasmids. The genetic relationship among newly identified plasmids was also investigated.

MATERIALS AND METHODS

Bacterial strains and culture conditions.

One thousand strains belonging to the B. cereus group were analyzed for the presence of pXO1- and pXO2-like replicons and for the presence of pXO2-related conjugative modules (Table 1). The B. mycoides and B. pseudomycoides strains could be recognized easily by their rhizoid growth, but the B. cereus, B. thuringiensis, and B. weihenstephanensis strains used were limited to strains received from other researchers, mostly from collections. For the other isolates used, the name _B. cereus_-B. thuringiensis was used in order to avoid possible misidentification and, hence, misinterpretation. Characteristics of a subset of strains studied further in this work are shown in Table 4 (see below). For the other strains, details can be obtained from us upon request.

TABLE 1.

Origins and distribution of strains tested for the presence of pXO1-like repX, pXO2-like repA, and T4SS-like transfer genes

TABLE 4.

Genotypic characterization and potential plasmid mobilization capability of a subset of B. cereus strains containing pXO2-like repA and/or pXO1-like repX

B. thuringiensis subsp. israelensis GBJ001(pUB110) or GBJ001(pBC16) and GBJ002 were used as helper and recipient strains, respectively, in triparental mating. GBJ001 and GBJ002 are derivatives of 4Q2 (42), cured of all plasmids, with chromosomal resistance to streptomycin (Sm) and nalidixic acid (Nal), respectively. The mobilizable plasmid pUB110 (originating from Staphylococcus aureus) and pBC16 (from B. cereus) carry kanamycin (Km) and tetracycline (Tet) resistance genes, respectively. Another derivative of 4Q2, 4Q7-Rif, carries a selection marker for rifampin (Rif) and was used as a recipient in biparental mating. Antibiotics were used at the following concentrations: Sm, 100 μg/ml; Nal, 15 μg/ml; Km, 50 μg/ml; Tet, 5 μg/ml; and Rif, 50 μg/ml.

PCR screening.

An alignment of replication genes of pXO1-like plasmids, including pBCXO1, pCER270, pPER272, pBC10987, and pXO1, and an alignment of replication genes and transfer genes of pXO2-like plasmids, including pAW63, pBT9727, and pXO2, were used to design the “consensus” primers listed in Table 2. Figure 1 shows the positions of the primers used for detection of vir genes. The primer pairs used for detection of group II introns were designed to target sequence regions that are conserved well in pXO2, pBT9727, and pAW63 and flank the B.th.I.1 and B.th.I.2 inserts on pAW63 (Fig. 1). For each pXO2-like replicon known to carry the transfer genes, amplification with each primer pair yielded either a short amplicon corresponding to the native intronless gene sequence or a long amplicon corresponding to the gene sequence containing an intron. “Long” amplicons produced by the primer pair targeting the original locus of the B.th.I1 intron were defined as amplicons containing a B.th.I1-type insertion, and the amplicons produced by the pair targeting B.th.I2 were defined as amplicons containing a B.th.I2-type insertion. In order to verify whether “long” amplicons contain a B.th.I1- or B.th.I2-type insertion, new primer pairs designed using the B.th.I.1 and B.th.I.2 sequences directly were used to investigate the presence of introns in the amplicons (data not shown).

TABLE 2.

Primers designed in this study

FIG. 1.

Positions of the conjugation-related primers designed for the pXO2-like tra region. The numbers indicate positions in the pAW63 sequence (accession number DQ025752). The positions of three major coding sequences encoding T4SS-like proteins (VirB4, D4, and B11) are indicated, as is the position of the sequence encoding the IEP associated with the group II intron B.th.I.1. Three primer pairs, MB4_F2/MB4_R2, MD4_F2/MD4_F2R2, and MB11_F1/MB11_R1, were used to detect the occurrence of the T4SS-like genes virB4, virD4, and virB11, respectively. Primer pairs B.th.I1_F1/B.th.I1_R2 and MD4_F1/MD4_R1 were used to detect the occurrence of the introns B.th.I.1 and B.th.I.2, respectively. The solid bars indicate predicted PCR products; the broken lines indicate introns.

The repX-F4/repX-R2 and repA-F1/repA-R3 primer pairs were used to detect the presence of pXO1- and pXO2-like replicons, respectively. Primer pairs lef-1F/lef-1R, cya-1F/cya-2R, pa-1F/pa-1R, and capA-1F/capA-1R were used for detection of anthrax toxin and capsule genes as described previously (22). Furthermore, two primer pairs, pUB110_F/pUB110_R and pBC16_F/pBC16_R, designed using the sequences of mobilizable plasmids pUB110 and pBC16, respectively, were used to detect the presence of these plasmids in transconjugants (Table 2). Total DNA was prepared by the method of Hansen and Hendriksen (14). The PCRs were carried out as described previously (32). Strains AW06 and Bt9727 were used as positive controls for detecting pXO2-like plasmids, and strain F4810/72 was used as the control for detecting pXO1-like plasmids.

Sequence analysis.

PCR products obtained using primer pairs repA-F1/repA-R2 and MB4-F2/MB4-R2 were purified and sequenced. Sequence alignment was performed with the DS Gene package (Accelrys Inc.). Aligned sequences were analyzed using the Molecular Evolutionary Genetics Analysis package, version 3.1 (http://www.megasoftware.net). Phylogenetic trees were constructed by the neighbor-joining method (30) based on the DNA sequences of pXO2-like repA and virB4 by using the Kimura 2 parameter genetic distance model. Bootstrap confidence values were generated using 500 permutations of the data set.

Conjugation experiments.

Plasmid conjugation transfer was performed by using previously described protocols (39). Triparental mating was performed using B. cereus group isolates thought to contain a pXO1- and/or pXO2-like plasmid(s) (PCR positive for repX and/or repA), a helper strain possessing a mobilizable plasmid [GBJ001(pUB110) (Smr Kmr) or GBJ001(pBC16) (Smr Tetr)], and B. thuringiensis subsp. israelensis strain GBJ002 (Nalr) as the recipient. In parallel, strains AW06 and Bt9727, which contain conjugative plasmids pAW63 and pBT9727, respectively, were used as donors in control experiments.

Transconjugants containing mobilizable plasmids were first detected based on their resistance to Nal and Km (pUB110) or Nal and Tet (pBC16) and their sensitivity to Sm. In parallel, controls with donor, helper, and recipient strains, grown separately, were plated on selective media to assess the occurrence of spontaneous resistant mutants. Primer pairs pUB110-F/pUB110-R and pBC16-F/pBC16-R were then used to confirm the presence of pUB110 or pBC16. Finally, the repX-F4/repX-R2 and repA-F1/repA-R3 primer pairs were used to detect the potential cotransfer of pXO1- and pXO2-like plasmids, respectively, into the transconjugants containing the mobilizable plasmid. Biparental mating was then performed using the transconjugant containing both the conjugative plasmid (pXO1- or pXO2-like) and the small mobilizable plasmid (pUB110 or pBC16) as the donor and B. thuringiensis 4Q-7 (Rif mutant) as the recipient. The resulting transconjugants were used for evaluation of the antibiotic resistance and PCR detection as described above. Figure 2 shows the conjugational transfer strategy used for triparental mating and biparental mating.

FIG. 2.

Strategy used for conjugation in triparental mating and biparental mating. (A) Triparental mating was performed using a donor strain containing a pXO1- and/or pXO2-like plasmid(s), the recipient strain B. thuringiensis subsp. israelensis GBJ002, and the helper strain B. thuringiensis subsp. israelensis GBJ001 containing a mobilizable plasmid (pUB110 or pBC16). The antibiotic markers carried by the host strain and plasmids are indicated. Two configurations of potential transconjugants were screened by using their resistance to Nal and Km (pUB110) or to Nal and Tet (pBC16) and by using PCR (PCR positive for pUB110 or pBC16 and then PCR performed to detect the presence of pXO1- or pXO2-like replication genes in transconjugants). (B) Biparental mating was performed using the transconjugants from the triparental mating experiments, which contained both the mobilizable plasmid and pXO1- or pXO2-like plasmid, and the recipient strain B. thuringiensis subsp. israelensis 4Q7-Rif. Bacteria are indicated by thin, medium, and thick lines to distinguish different host strains. Curved arrows indicate the possible transfers.

Plasmid profile and Southern blotting.

Large plasmids were extracted using the method described by Andrup et al. (2). Plasmid DNA samples were run on a gel and transferred to Hybond N+ (Amersham). Probe labeling was performed using the repA-F2/repA-R3 (pXO2-like) and repX-F1/repX-R1 (pXO1-like) primer pairs according to the protocol of the manufacturer of the kit (Roche). Hybridization was performed as described previously (32).

RESULTS

Detection of replicons related to the pXO1 and pXO2 plasmids.

A total of 1,000 B. cereus group isolates were screened by PCR for the presence of pXO1-like repX and pXO2-like repA replicons. Of the strains tested, 649 were random environmental isolates and 351 were nonrandom isolates obtained from different laboratories and international collections (Table 1). Isolates were defined as “random” if they were obtained from environmental niches without selection for specific characteristics and as “nonrandom” if they originated in studies performed with specific aims. In many cases, the name _B. cereus_-B. thuringiensis was used since the actual identity of a strain was unknown. The strains originated from natural niches (soil, water, and air), from organisms (plants, insects, and mammals), or from food matrices, and they came from a broad spectrum of geographic locations (Asia, America, Africa, and Europe). It should be noted that strains from the same location that displayed the same plasmid pattern were represented by a single isolate.

Of the 1,000 B. cereus group isolates in the collection (Table 1), 160 contained a pXO1-like replicon, 100 contained a pXO2-like replicon, and 22 contained both a pXO1-like replicon and a pXO2-like replicon. These replicons were found mainly in B. cereus and B. thuringiensis isolates, and they were not present in the six B. weihenstephanensis strains tested. Interestingly, 7 of the 111 environmental B. mycoides isolates were found to harbor a pXO2-like replicon, and of the 12 B. mycoides isolates from laboratory collections, 3 contained a pXO1-like replicon and 5 contained a pXO2-like replicon (Table 1). To our knowledge, this is the first time that pXO1- and/or pXO2-like replicons have been found in B. mycoides.

Compared to strains from the other niches sampled (air, water, or plants), the B. cereus group strains originating from soil had a greater tendency to harbor pXO1- and/or pXO2-like replicons. Of the 136 lab collection isolates that originated from soil in China, 41 were found to contain a pXO1-like replicon and 23 were found to contain a pXO2-like replicon, while 2 contained both a pXO1-like replicon and a pXO2-like replicon. A high prevalence of the replicons was also observed in lab collection isolates originating from food. Of the 139 food-associated isolates, 67 contained a pXO1-like replicon and 7 contained a pXO2-like replicon, while 10 isolates contained both a pXO1-like replicon and a pXO2-like replicon (Table 1).

The percentages of random environmental isolates that contained only a pXO1-like replicon and only a pXO2-like replicon were ca. 6.6% and 7.7%, respectively. Only 3 of 649 (0.46%) random isolates tested contained both types of replicons. In comparison, the percentages of nonrandom isolates (from different laboratories and international collections) that contained both types of replicons were much higher. About 33.3% and 14.2% of the 351 lab isolates harbored only a pXO1-like and only a pXO2-like replicon, respectively, while 5.4% contained both types of replicons (Table 1 and Fig. 3).

FIG. 3.

Percentages of the B. cereus group strains with pXO1- and pXO2-like plasmids and T4SS-like transfer modules for environmental and lab collection isolates. Results are shown for environmental isolates, lab collection isolates, and isolates containing the pXO2-like tra region, based on the presence of the virB4 and virD4 transfer genes, which encode the core apparatus of the T4SS.

Plasmid DNA was extracted from isolates containing pXO1- and/or pXO2-like replicons and visualized on agarose gels (Fig. 4). All profiles obtained included at least one large plasmid, and some profiles contained up to six large plasmids and several smaller plasmids (<20 kb). For the strains that were PCR positive for the pXO1-like replicon and/or the pXO2-like replicon, at least one large-plasmid band at an approximate size similar to that of known pXO1-like plasmids (180 to 280 kb) and/or pXO2-like plasmids (70 to 100 kb) was observed. Southern blot hybridization was performed to verify that the positive results for the PCR assay corresponded to results for the large plasmids. For instance, VD142 and ISP2954 both contained a 75-kb plasmid that hybridized to the pXO2 probe (Fig. 4A and B, lanes 2, 3, and 7). IS075, which was PCR positive for both pXO1- and pXO2-like replicons, contained one 75-kb element and one >200-kb element that hybridized to the pXO2 and pXO1 probes, respectively (Fig. 4, lanes 6). VD022, VD023, and Schrouff had two plasmids, a ca. 100-kb plasmid and a >200-kb plasmid that hybridized to the pXO2 and pXO1 probes, respectively (Fig. 4, lanes 4, 5, and 8). It was interesting that the sizes of the pXO2-like plasmids of VD022, VD023, and Schrouff were all estimated to be about 100 kb, which is greater than the size of pAW63 (72 kb); thus, these plasmids might be more similar to pXO2 (Fig. 4B, lanes 2, 4, 5, and 8). In addition, different intensities of hybridization signals were observed, which may have been the result of sequence variation among the plasmids or of variations in the amount of plasmid DNA in the samples or both.

FIG. 4.

Large-plasmid and Southern blotting of strains containing pXO1- and/or pXO2-like plasmid(s). (A) Plasmid profiles; (B) Southern blot patterns obtained with the partial pXO2-like repA fragment as the probe; (C) Southern blot patterns obtained with the partial pXO1-like repX fragment as the probe. Lane 1, AND508 (1) containing four large plasmids (used as references; the position of the second large plasmid, pBtoxis, is indicated by an arrow); lane 2, AW06 harboring pAW63 (72 kb; position indicated by an arrow); lane 3, VD142; lane 4, VD022; lane 5, VD023; lane 6, IS075; lane 7, ISP2954; lane 8, Schrouff. VD142 and ISP2954 are PCR positive for pXO2-like repA; VD022, VD023, IS075, and Schrouff are PCR positive for both pXO2-like repA and pXO1-like repX. chr, chromosome. The scale to the left of each blot was used as a ladder to compare the positions of plasmids and hybridization signals. The arrowheads and arrows indicate the positions of pXO2- and pXO1-like plasmids, respectively.

Furthermore, all strains containing a pXO1- and/or pXO2-like replication gene(s) were screened by PCR for the anthrax toxin genes and for the capsule genes. However, none of the 282 strains tested were found to carry the B. anthracis virulence genes.

Occurrence of key conjugative components related to T4SS.

A subset of 493 isolates, including 177 random and 316 nonrandom strains, was screened for the presence of homologs of the conjugative T4SS genes virB11, virB4, and virD4. Interestingly, it was observed from the PCR results that only the isolates harboring a pXO2-like replicon were positive for the target genes (data not shown). Therefore, of the 507 isolates that were positive, only those that were positive for the pXO2-like replicon were screened for the presence of the pXO2-like transfer genes.

Only a minority of the environmental isolates (∼1.54%) were PCR positive for pXO2-like transfer genes, and none of the strains harboring both a pXO1-like replicon and a pXO2-like replicon appeared to contain the transfer genes. In contrast, 8.8% of the lab collection isolates that had nonrandom origins and contained the pXO2-like replicon contained pXO2-like transfer genes, and 4.8% of the strains harboring both a pXO1-like replicon and a pXO2-like replicon also contained the transfer module genes (Fig. 3). A total of 58 of the 122 strains that possessed a pXO2-like replicon also contained homologs of virB4 and virD4. Only 25 of these strains were PCR positive for the virB11 homolog, however, which may reflect greater sequence variation in virB11 gene homologs (Table 1).

Occurrence of group II introns in the transfer region.

Group II introns B.th.I.1 and B.th.I.2 have been identified in the transfer region of pAW63 but were not present in pXO2 or pBT9727 (38). A PCR-based assay was performed using primers designed to target regions flanking B.th.I.1 and B.th.I.2 inserts on pAW63. Hence, the presence of the introns could be determined by the sizes of the different amplicons (for details, see Materials and Methods). It should also be noted that the primers were based on conserved sequences flanking B.th.I.1 and B.th.I.2, because it was impractical to target the introns themselves since nearly identical copies have been found in the chromosome of the pAW63 host strain, as well as in other strains of the B. cereus group (37).

The 58 isolates known to contain a pXO2-like replicon and the transfer genes were tested for the presence of introns. All four possible configurations were observed (Table 3). The pXO2-like replicons of nine isolates had both B.th.I.1- and B.th.I.2-type insertions, the pXO2-like replicon of one isolate had only a B.th.I.1-type insertion (Schrouff), the pXO2-like replicon of one isolate had only a B.th.I.2-type insertion (VD148), and the pXO2-like replicons of 15 isolates had no insertion. PCRs performed with 32 isolates did not yield an amplicon despite repeated attempts, presumably due to sequence variability.

TABLE 3.

Occurrence of group II introns B.th.I.1 and B.th.I.2 in 58 B. cereus strains containing pXO2-like repA and type IV-like secretion system genes

Conjugative transfer and mobilization capabilities of pXO2- and pXO1-like plasmids.

Previous studies have shown that triparental mating is a suitable method for assaying the transfer of conjugative plasmids for which there is no direct means of selection (39). Indeed, many conjugative plasmids not only are self-transmissible but are also able to promote the transfer of other small mobilizable plasmids. Fifty-one B. cereus group isolates were selected for testing based on their plasmid content (4, 39, and 8 isolates contained a pXO1-like plasmid, a pXO2-like plasmid, and both types of plasmids, respectively) and the presence of a putative tra region. They were all analyzed individually to determine their abilities to promote the transfer of other small plasmids.

All but three (015, Bc4-4, and Bc27-1) of the B. cereus group isolates containing only a pXO2-like replicon and the transfer module were capable of mobilizing pUB110 and/or pBC16. In contrast, none of the 22 B. cereus group isolates that contained a pXO2-like replicon lacking the cognate transfer module was found to promote plasmid transfer. Interestingly, three isolates that contained both a pXO1-like replicon and a pXO2-like replicon but apparently lacked the cognate transfer module (VD022, VD023, and B5-2) were also found to be capable of mobilizing small plasmids. No isolate containing only a pXO1-like replicon was able to promote plasmid transfer under the conditions used for the assay (Table 4).

Of the 20 isolates with the ability to mobilize small plasmids, 12 displayed a transfer efficiency similar to that of B. thuringiensis subsp. kurstaki AW06 containing pAW63 (Table 4). These strains included the three strains containing both a pXO1-like replicon and a pXO2-like replicon but apparently lacking the transfer module. For VD045, VD142, VD148, DBt012, and DBt065, the measured transfer efficiencies were ca. 102 to 103 times lower than that of AW06 and similar to that of B. thuringiensis subsp. konkukian strain Bt9727, which contains pBT9727 (Table 4).

PCR-based testing showed that most transconjugants obtained in these triparental mating experiments contained only the small mobilizable plasmid (pUB110 or pBC16), while a few transconjugants contained both the pXO2-like plasmid and the small plasmid. No transconjugant was found to contain the pXO1-like plasmid (data not shown). For instance, selected transconjugants obtained from triparental mating involving either T03001 or VD148 as the donor (designated TT03001-1 and TVD148-1, respectively) were PCR positive for the pXO2-like replicon and contained a large plasmid whose size was similar to that of pAW63 and a small plasmid whose size was similar to that of pBC16 (Fig. 5, lanes 3 and 9). Although the original donor, T03001, contained a pXO1-like replicon, transconjugant T03001-1 was PCR negative for the pXO1-like replicon, suggesting that it was the pXO2-like plasmid rather than the pXO1-like plasmid that had conjugated and mobilized pBC16.

FIG. 5.

Plasmid profiles of strains involved in tri- and biparental mating. Lane 1, AND508 containing four large plasmids (the position of the reference element, pBtoxis, is indicated by an arrow); lane 2, T03001; lane 3, TT03001-1 [transconjugant from triparental mating of T03001, GBJ001(pBC16), and GBJ002]; lane 4, TT03001-2 (transconjugant from biparental mating of TT03001-1 and 4Q7-Rif); lane5, VD023; lane 6, TK023-1 [transconjugant from triparental mating of VD023, GBJ001(pUB110), and GBJ002]; lane 7, TT023-1 [transconjugant from triparental mating of VD023, GBJ001(pBC16), and GBJ002]; lane 8, VD148; lane 9, TVD148-1 [transconjugant from triparental mating of VD148, GBJ001(pBC16), and GBJ002]; lane 10, TVD148-2 (transconjugant from biparental mating of TVD148-1 and 4Q7-Rif); lane 11, AW06 harboring pAW63 (72 kb; position indicated by an arrow); lanes 12 and 13, GBJ001(pUB110) and GBJ001(pBC16) containing the mobilizable plasmids pUB110 and pBC16, used as helper strains in triparental mating. Chr., chromosome. The arrowheads and arrows indicate the positions of pXO2- and pXO1-like plasmids, respectively, as identified by Southern blot hybridization using partial pXO2-like repA and pXO1-like repX fragments as probes (data not shown).

Many of the isolates used as donors in triparental mating contained large plasmids besides pXO1- or pXO2-like plasmids that might have been responsible for promoting the observed plasmid transfer. Biparental mating was performed with the transconjugants from the triparental mating experiments, which contained both the mobilizable small plasmid and the pXO2-like plasmid, and the recipient strain B. thuringiensis subsp. israelensis 4Q7-Rif (Fig. 2). The results indicated that the pXO2-like plasmids were indeed capable of promoting transfer to a new recipient strain. For instance, the plasmid profiles and Southern blots obtained for transconjugants TT03001-2 (transconjugant from biparental mating of TT03001-1 and 4Q7-Rif) and TVD148-2 (transconjugant from biparental mating of TVD148-1 and 4Q7-Rif) confirmed that pAW63 and pBC16 had been transferred to the recipient strain in both cases (Fig. 5, lanes 4 and 10).

In the case of the isolates possessing both a pXO1-like replicon and a pXO2-like replicon but apparently lacking the transfer module (i.e., VD022, VD023, and B5-2), the PCR and plasmid patterns of the transconjugants indicated that the mobilized plasmid pUB110 or pBC16 was present but the pXO1- or pXO2-like plasmid was not present. For instance, transconjugants TK023-1 [from triparental mating of VD023, GBJ001(pUB110), and GBJ002] and TT023-1 [from triparental mating of VD023, GBJ001(pBC16), and GBJ002] contained only pUB110 and pBC16, respectively (Fig. 5, lanes 6 and 7). This suggests that it was not the pXO1- or pXO2-like plasmid but rather some other unknown plasmid(s) or conjugative transposon(s) that had promoted mobilization of the small plasmid. Alternatively, the pXO1- or pXO2-like plasmid may have been able to transfer the small plasmid but not itself, although this is less likely.

Phylogeny of pXO2-like replicons.

In order to investigate the phylogeny of pXO2-like replicons identified in this study, a set of 33 isolates was chosen to represent the diverse combinations of replicons, transfer modules, and transfer capabilities. For each isolate, the pXO2-like repA PCR amplicon was sequenced and compared with all other amplicons, including the corresponding sequences from pXO2, pAW63, and pBT9727.

The dendrogram based on this set of sequences showed that the pXO2-like replicons clustered into two major groups (Fig. 6A). Group I comprised 22 replicons, including pAW63 and pXO2, while group II comprised 14 replicons, including pBT9727. Remarkably, these two groups did not correlate with the presence of a transfer module. Nevertheless, it is interesting that plasmids bearing the putative intron(s) (B.th.I1 and/or B.th.I2) were found to cluster with pAW63 in group I (Fig. 6A).

FIG. 6.

pXO2-like repA dendrogram (A) and alignment of partial repA sequences (B). The levels of identity of the 36 sequences ranged from 77% to 100% based on DNA sequences (and from 79% to 100% based on amino acid sequences). The pBT9727, pAW63, and pXO2 reference plasmids are enclosed in boxes. Filled triangles indicate strains containing transfer modules, while open triangles indicate strains missing the cognate transfer modules. Stars indicate strains carrying a B.th.I1 and/or B.th.I2 insertion(s). ++, good mobilization efficiency (similar to that of pAW63); +, low mobilization efficiency (102 to 103 times lower than that of pAW63); −, no mobilization. The scale bar indicates genetic distance (0.02 nucleotide substitution per site). The boxes in panel B indicate the two indels consisting of seven and eight nucleotides present in the repA sequences. All bootstrap support values of >70% are indicated at the appropriate nodes.

Nucleotide sequence analysis of all _repA_-derived amplicons revealed the presence of two short insertions/deletions (indels) composed of seven and eight nucleotides. The two indels were located centrally in the gene fragment, on either side of a 7-nucleotide region. The presence of the indels matched the top-level phylogenetic clustering of group I (with indels) and group II (without indels) (Fig. 6B).

Although some pXO2-like replicons from isolates originating from the same location were observed to have identical sequences for the replication gene fragment, their hosts were not considered to be clonal since their plasmid patterns were not identical; two examples are the replicons from VD022 and VD023 (isolated from the same pond) (Fig. 4A, lanes 4 and 5) and the replicons from IS195 and IS075 (isolated from a small mammal) (data not shown).

Phylogenetic discrepancies between pXO2-like replicons and virB4 transfer genes.

The sequences of repA and the virB4 homolog (the best-conserved T4SS-like gene in pXO2-related plasmids) of 22 isolates containing pXO2-like plasmids were compared with the sequences of all other strains and with the corresponding sequences of pXO2, pAW63, and pBT9727 (Fig. 7). The comparative analysis showed that both _repA_- and _virB4_-based trees contained two distinct clusters. Although most close associations were the same in both clusters, for four isolates (VD115, VD142, VD148, and Bc27-1) there were discrepancies in their relative repA and virB4 relationships (Fig. 7A and B). In a similar way, there was little correlation between genetic relationships based on _virB4_-like sequences and transfer capabilities measured in triparental mating experiments.

FIG. 7.

Comparison of pXO2-like repA (A) and virB4 (B) dendrograms. The levels of identity for the 22 sequences ranged from 77% to 100% based on repA sequences (and from 80% to 100% based on amino acid sequences) and from 73% to 100% based on virB4 sequences (and from 72% to 100% based on amino acid sequences). For an explanation of the symbols and scale bar see the legend to Fig. 6. All bootstrap support values of >70% are indicated at the appropriate nodes.

DISCUSSION

A key member of the B. cereus group, B. anthracis, is distinguished on the basis of phenotypic and molecular characteristics that are conferred by two large plasmids, pXO1 and pXO2. However, the concept of B. anthracis as a separate species may be called into question in light of recent discoveries of “close-neighbor” strains of B. cereus and B. thuringiensis that possess unusual features, including large plasmids that share a common backbone with pXO1 and/or pXO2 (8, 28, 38). In this study, we developed a screen for pXO1- and pXO2-like plasmids in order to investigate their prevalence and diversity in 1,000 isolates of the B. cereus group originating from various environmental samples and laboratory collections. These 1,000 isolates covered a broad spectrum of different geographic and substrate origins, serotypes, and virulence characteristics. Isolates were categorized either as “random,” if they originated from samples obtained from ecological niches to assess aspects of biodiversity, or as “nonrandom,” if they originated from laboratory collections that had been assembled for clinical or agricultural purposes. Besides these isolates, 100 additional strains of other Bacillus species were also screened, but none of them was found to contain pXO1- or pXO2-like plasmids (data not shown).

Identification of pXO1- and pXO2-like plasmids was performed by PCR detection of their replicons. The replication genes were chosen as the diagnostic characteristics for the plasmid groups because their cognate origins are usually under stringent selective constraints that limit the sequence drift in these regions. Accordingly, repX (the replication gene in pXO1), repA (the replication gene in pXO2), and similar genes (28, 38) were used as a basis to develop the PCR assay that was then used for the 1,000 isolates in the collection. The results of the PCR detection assay were further supported by plasmid profiling and Southern blot hybridization for a subset of isolates. The results confirmed that the isolates reported to harbor pXO1- and/or pXO2-like replicons do contain plasmids that are the corresponding sizes and display strong hybridization to a probe targeting the corresponding replication genes (Fig. 4 and Fig. 5).

Overall, the results of the screening revealed that pXO1- and pXO2-like plasmids are widely distributed across a variety of environmental niches, especially soil, and in geographical locations as far apart as Scotland and China. The prevalence of pXO1-like plasmids appears to be similar to the prevalence of pXO2-like plasmids in random environmental isolates. The much higher prevalence of pXO1-like plasmids in the nonrandom isolates seems to reflect a strong bias in the contents of some of the lab collections that were used. For instance, the 139 B. cereus lab collection strains isolated from food were obtained mainly from clinical cases. Since the genetic determinants of cereulide, the emetic toxin which causes one of the two food poisoning syndromes attributed to B. cereus, are located in pXO1-like plasmids (8, 19, 28), it is logical that B. cereus strains isolated from food poisoning cases more likely carry pXO1-like plasmids.

A natural concern in studies of these plasmids is their potential for crossing species boundaries and driving horizontal gene transfer within the B. cereus group, where horizontal gene transfer is particularly relevant given the amount of debate that surrounds the current species definitions and their consequences for public health, as well as agro-industrial interests. In order to assess the basic transfer potential of the pXO1- and pXO2-like plasmids, their host isolates were screened for homologs of the T4SS genes virB4, virB11, and virD4 that were originally described to be in the transfer region of the pXO2-like conjugative plasmid pAW63 (38). There were no positive results for isolates containing pXO1-like replicons, which was not surprising since the targeted genes are not part of the regions conserved among known pXO1-like plasmids. While pXO1 has recently been reported to encode some T4SS-like components (13), the T4SS gene homologs on pXO1 proposed by Grynberg and coworkers display low levels of homology to known T4SS genes and are physically scattered throughout the plasmid, making it doubtful that they could function as the concerted secretion machinery required for conjugation. With regard to the isolates containing pXO2-like replicons, screening for the T4SS homologs revealed that about one-half of them possess the transfer region. Experimental assays performed with a subset of these isolates showed that most of them were capable of promoting their own transfer, as well as that of small mobilizable plasmids, albeit with various levels of efficiency. The failure of some plasmids to promote transfer may have been due to the detection limit of the assay (<10−7 transconjugant per recipient) or to genetic deficiency in the transfer region, as is thought to be the case for pXO2 (40).

Interestingly, a phylogenetic analysis based on the pXO2-like repA sequences showed only a low-level correlation between genetic clustering and the presence of the transfer genes (Fig. 6). One interpretation of this is that the ancestral form of the plasmid was conjugative and descendent lineages underwent genetic drift, leading to loss of transfer capability. The genetic analysis also showed some discrepancies between repA and virB homolog trees (Fig. 7). Three nonmutually exclusive explanations for this observation are that (i) the virB4 homolog may not be a good representative of the transfer module, (ii) the replication and transfer modules are subject to different selective constraints, and (iii) the plasmid backbones may have undergone various recombination and gene transfer events.

In contrast, the phylogenetic analysis showed that there was a strong correlation between the phylogeny of the plasmids and the presence of B.th.I1-type and B.th.I2-type insertions. Plasmids bearing the putative introns were found to cluster with pAW63, and most plasmids lacking these introns in the cluster were among the plasmids that appear to have lost the transfer genes. This suggests that the putative introns were transmitted vertically within the cluster rather than by multiple intron invasion events. What the presence of introns may mean for this cluster of plasmids is an interesting question. Splicing of the L. lactis intron Ll.LtrB, which is located in a relaxase gene, is known to be required for the transfer of its host element (5, 23), and it has been suggested that the rate of intron splicing may influence the regulation of conjugative behavior. Splicing of B.th.I.2 and restoration of the _virD4_-like gene exon mRNA have been demonstrated in pAW63 (37), but the regulatory implications of the presence of introns interrupting the transfer genes are still under investigation.

In addition to the basic modules that allow them to replicate or to promote conjugative transfer, many plasmids contain passenger genes that confer adaptive functions to the host cell, and the classic example is antibiotic resistance genes. The main payload of pXO1 is a pathogenicity island that contains the anthrax toxin genes and genes encoding several regulators that are crucial to the development of the disease. Similarly, pXO2 carries a pathogenicity island that contains the gene cluster responsible for producing the protective capsule that protects the bacterium during infection. The isolates containing either pXO1- or pXO2-like replicons were therefore screened by PCR for the presence of the anthrax toxin and capsule genes, but all of them were negative (data not shown). Nevertheless, considering the wide range of plasmid sizes that were observed in the plasmid profiles, it is very probable that many plasmids carry passenger genes, and it would be extremely interesting to explore their functional diversity. The trend observed in known pXO1-like plasmids is an association with virulence determinants, such as the gene cluster encoding nonribosomal peptide synthetase that leads to production of the emetic toxin cereulide (8). If this trend occurs, there may be a great clinical advantage in characterizing the payloads of the pXO1-like plasmids identified in this study. In contrast, the known pXO2-like plasmids pAW63 and pBT9727 are both cryptic, apart from their conjugative properties. It may be that conjugative plasmids with efficient transfer capabilities are under less selective pressure to encode functions that are beneficial to the host as “payment” for their residence in the cell. In this sense the transfer module could be seen as a postpartitioning stability system, with plasmid-bearing cells “reinfecting” siblings that lose the plasmid. Assessing the prevalence of passenger genes in the newly identified pXO2-like plasmids should shed some light on whether carrying a payload is the exception or the rule in this family, and comparing the distributions of the genes may reveal whether there is a correlation between the inability to transfer and the presence of “useful” accessory genes, as one might speculate in the case of pXO2.

Finally, there is considerable debate about how much natural plasmid “spread” (i.e., representation of related plasmids in a given range of bacterial populations) is truly due to horizontal transfer and how much is due simply to vertical inheritance followed by differentiation within progeny. In this context, it seems particularly significant that pXO2-like plasmids were found in four B. mycoides or B. pseudomycoides strains and, perhaps most importantly, that these plasmids do not cluster together but instead are scattered throughout the replicon-based dendrogram among B. cereus and B. thuringiensis plasmids. While the distinction between the latter two species is fuzzy to the point that it may not constitute a sound basis for the comparison of host genome background and plasmid phylogeny, there is much greater confidence in the phylogenetic separation between these taxa and the species B. mycoides and B. pseudomycoides. Therefore, the observations made concerning the pXO2-like plasmids from B. mycoides and B. pseudomycoides make a solid case for the conclusion that there have been multiple plasmid transfers among different species of the B. cereus group. In future investigations, analyzing plasmid and host genomes by multilocus sequence typing should allow greater characterization of the scale and implications of this phenomenon.

Acknowledgments

We gratefully acknowledge Valérie Duprez, Florence Hoton, and Olivier Minet for their contributions to this work.

This project was supported by grants from the European Space Agency (MISSEX, AO-2004, and PRODEX C90255), the National Fund for Scientific Research, and the Université catholique de Louvain. Part of this work was also supported by short-term fellowships from EMBO and National Fund for Scientific Research to G.V.D.A.

Footnotes

▿

Published ahead of print on 20 March 2009.

REFERENCES

• 1.Andrup, L., J. Damgaard, and K. Wassermann. 1993. Mobilization of small plasmids in Bacillus thuringiensis subsp. israelensis is accompanied by specific aggregation. J. Bacteriol. 175**:**6530-6536. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 2.Andrup, L., K. K. Barfod, G. B. Jensen, and L. Smidt. 2008. Detection of large plasmids from the Bacillus cereus group. Plasmid 59**:**139-143. [DOI] [PubMed] [Google Scholar]
• 3.Atmakuri, K., E. Cascale, and P. J. Christie. 2004. Energetic components VirD4, VirB11 and VirB4 mediate early DNA transfer reactions required for bacterial type IV secretion. Mol. Microbiol. 54**:**1199-1211. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 4.Backert, S., and T. F. Meyer. 2006. Type IV secretion systems and their effectors in bacterial pathogenesis. Curr. Opin. Microbiol. 9**:**207-217. [DOI] [PubMed] [Google Scholar]
• 5.Belhocine, K., A. B. Mak, and B. Cousineau. 2007. Trans-splicing of the Ll.LtrB group II intron in Lactococcus lactis. Nucleic Acids Res. 35**:**2257-2268. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 6.Christie, P. J. 2004. Type IV secretion: the Agrobacterium VirB/D4 and related conjugation systems. Biochim. Biophys. Acta 1694**:**219-234. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 7.Christie, P. J., K. Atmakuri, V. Krishnamoorthy, S. Jakubowski, and E. Cascales. 2005. Biogenesis, architecture, and function of bacterial type IV secretion systems. Annu. Rev. Microbiol. 59**:**451-485. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 8.Ehling-Schulz, M., M. Fricker, H. Grallert, P. Riek, M. Wagner, and S. Scherer. 2006. Cereulide synthetase gene cluster from emetic Bacillus cereus: structure and location on a mega virulence plasmid related to Bacillus anthracis toxin plasmid pXO1. BMC Microbiol. 6**:**20. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 9.Eickbush, T. H. 1999. Mobile introns: retrohoming by complete reverse splicing. Curr. Biol. 9**:**R11-R14. [DOI] [PubMed] [Google Scholar]
• 10.Ferat, J. L., and F. Michel. 1993. Group II self-splicing introns in bacteria. Nature 364**:**358-361. [DOI] [PubMed] [Google Scholar]
• 11.Gammon, K., G. W. Jones, S. J. Hope, C. M. de Oliveira, L. Regis, M. H. Silva Filha, B. N. Dancer, and C. Berry. 2006. Conjugal transfer of a toxin-coding megaplasmid from Bacillus thuringiensis subsp. israelensis to mosquitocidal strains of Bacillus sphaericus. Appl. Environ. Microbiol. 72**:**1766-1770. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 12.González, J. M., B. J. Brown, and B. C. Carlton. 1982. Transfer of Bacillus thuringiensis plasmids among strains of B. thuringiensis and B. cereus. Proc. Natl. Acad. Sci. USA 79**:**6951-6955. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 13.Grynberg, M., Z. Li, E. Szczurek, and A. Godzik. 2007. Putative type IV secretion genes in Bacillus anthracis. Trends Microbiol. 15**:**191-195. [DOI] [PubMed] [Google Scholar]
• 14.Hansen, B. M., and N. B. Hendriksen. 2001. Detection of enterotoxic Bacillus cereus and Bacillus thuringiensis strains by PCR analysis. Appl. Environ. Microbiol. 67**:**185-189. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 15.Helgason, E., N. J. Tourasse, R. Meisal, D. A. Caugant, and A. B. Kolstø. 2004. Multilocus sequence typing scheme for bacteria of the Bacillus cereus group. Appl. Environ. Microbiol. 70**:**191-201. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 16.Helgason, E., O. A. Økstad, D. A. Caugant, H. A. Johansen, A. Fouet, M. Mock, I. Hegna, and A. B. Kolstø. 2000. Bacillus anthracis, Bacillus cereus, and Bacillus thuringiensis: one species on the basis of genetic evidence. Appl. Environ. Microbiol. 66**:**2627-2630. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 17.Hernandez, E., F. Ramisse, J. P. Ducoureau, T. Cruel, and J. D. Cavallo. 1998. Bacillus thuringiensis subsp. konkukian (serotype H34) superinfection: case report and experimental evidence of pathogenicity in immunosuppressed mice. J. Clin. Microbiol. 36**:**2138-2139. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 18.Hoffmaster, A. R., J. Ravel, D. A. Rasko, G. D. Chapman, M. D. Chute, C. K. Marston, B. K. De, C. T. Sacchi, C. Fitzgerald, L. W. Mayer, M. C. Maiden, F. G. Priest, M. Barker, L. Jiang, R. Z. Cer, J. Rilstone, S. N. Peterson, R. S. Weyant, D. R. Galloway, T. D. Read, T. Popovic, and C. M. Fraser. 2004. Identification of anthrax toxin genes in a Bacillus cereus associated with an illness resembling inhalation anthrax. Proc. Natl. Acad. Sci. USA 101**:**8449-8454. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 19.Hoton, F. M., L. Andrup, I. Swiecicka, and J. Mahillon. 2005. The cereulide genetic determinants of emetic Bacillus cereus are plasmid-borne. Microbiology 151**:**2121-2124. [DOI] [PubMed] [Google Scholar]
• 19a.Hoton, F. M., N. Fornelos, E. N'Guessan, X. Hu, I. Swiecicka, K. Dierick, E. Jääskeläinen, M. Salkinoja-Salonen, and J. Mahillon. 2009. Family portrait of Bacillus cerus and Bacillus weihenstephanensis cereulide-producing strains. Environ. Microbiol. Rep., in press. [DOI] [PubMed]
• 20.Hu, X., B. M. Hansen, J. Eilenberg, N. B. Hendriksen, L. Smidt, Z. Yuan, and G. B. Jensen. 2004. Conjugative transfer, stability and expression of a plasmid encoding a cry1Ac gene in Bacillus cereus group strains. FEMS Microbiol. Lett. 231**:**45-52. [DOI] [PubMed] [Google Scholar]
• 21.Hu, X., B. M. Hansen, Z. Yuan, J. E. Johansen, J. Eilenberg, N. B. Hendriksen, L. Smidt, and G. B. Jensen. 2005. Transfer and expression of the mosquitocidal plasmid pBtoxis in Bacillus cereus group strains. FEMS Microbiol. Lett. 245**:**239-247. [DOI] [PubMed] [Google Scholar]
• 22.Jackson, P. J., M. E. Hugh-Jones, D. M. Adair, G. Green, K. K. Hill, C. R. Kuske, L. M. Grinberg, F. A. Abramova, and P. Keim. 1998. PCR analysis of tissue samples from the 1979 Sverdlovsk anthrax victims: the presence of multiple Bacillus anthracis strains in different victims. Proc. Natl. Acad. Sci. USA 95**:**1224-1229. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 23.Klein, J. R., Y. Chen, D. A. Manias, J. Zhuo, L. Zhou, C. L. Peebles, and G. M. Dunny. 2004. A conjugation-based system for genetic analysis of group II intron splicing in Lactococcus lactis. J. Bacteriol. 186**:**1991-1998. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 24.Lechner, S., R. Mayr, K. P. Francis, B. M. Prüss, T. Kaplan, E. Wiessner-Gunkel, G. S. A. B. Stewartz, and S. Scherer. 1998. Bacillus weihenstephanensis sp. nov. is a new psychrotolerant species of the Bacillus cereus group. Int. J. Syst. Bacteriol. 48**:**1373-1382. [DOI] [PubMed] [Google Scholar]
• 25.Nakamura, L. K. 1998. Bacillus pseudomycoides sp. nov. Int. J. Syst. Bacteriol. 48**:**1031-1035. [DOI] [PubMed] [Google Scholar]
• 26.Okinaka, R. T., K. Cloud, O. Hampton, A. R. Hoffmaster, K. K. Hill, P. Keim, T. M. Koehler, G. Lamke, S. Kumano, D. Manter, Y. Martinez, D. Ricke, R. Svensson, and P. J. Jackson. 1999. Sequence, assembly, and analysis of pXO1 and pXO2. J. Appl. Microbiol. 87**:**261-262. [DOI] [PubMed] [Google Scholar]
• 27.Okinaka, R. T., K. Cloud, O. Hampton, A. R. Hoffmaster, K. K. Hill, P. Keim, T. M. Koehler, G. Lamke, S. Kumano, J. Mahillon, D. Manter, Y. Martinez, D. Ricke, R. Svensson, and P. J. Jackson. 1999. Sequence and organization of pXO1, the large Bacillus anthracis plasmid harboring the anthrax toxin genes. J. Bacteriol. 181**:**6509-6515. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 28.Rasko, D. A., M. J. Rosovitz, O. A. Økstad, D. E. Fouts, L. Jiang, R. Z. Cer, A. B. Kolstø, S. R. Gill, and J. Ravel. 2007. Complete sequence analysis of novel plasmids from emetic and periodontal Bacillus cereus isolates reveals a common evolutionary history among the _B. cereus_-group plasmids, including Bacillus anthracis pXO1. J. Bacteriol. 189**:**52-64. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 29.Reddy, A., L. Battisti, and C. B. Thorne. 1987. Identification of self-transmissible plasmids in four Bacillus thuringiensis subspecies. J. Bacteriol. 169**:**5263-5270. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 30.Saitou, N., and M. Lei. 1987. The neighbor-joining method: a new method for reconstructing phylogenetic trees. Mol. Biol. Evol. 4**:**406-425. [DOI] [PubMed] [Google Scholar]
• 31.Salmond, G. P. C. 1994. Secretion of extracellular virulence factors by plant pathogenic bacteria. Annu. Rev. Phytopathol. 32**:**181-200. [Google Scholar]
• 32.Sambrook, J., and D. W. Russell. 2001. Molecular cloning: a laboratory manual, 3rd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.
• 33.Stachel, S. E., and E. W. Nester. 1986. The genetic and transcriptional organization of the vir region of the A6 Ti plasmid of Agrobacterium tumefaciens. EMBO J. 5**:**1445-1454. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 34.Staddon, J. H., E. M. Bryan, D. A. Manias, and G. M. Dunny. 2004. Conserved target for group II intron insertion in relaxase genes of conjugative elements of gram-positive bacteria. J. Bacteriol. 186**:**2393-2401. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 35.Stenfors, L. P., R. Mayr, S. Scherer, and P. E. Granum. 2002. Pathogenic potential of fifty Bacillus weihenstephanensis strains. FEMS Microbial. Lett. 215**:**47-51. [DOI] [PubMed] [Google Scholar]
• 36.Thorsen, L., B. M. Hansen, K. F. Nielsen, N. B. Hendriksen, R. K. Phipps, and B. B. Budde. 2006. Characterization of emetic Bacillus weihenstephanensis, a new cereulide-producing bacterium. Appl. Environ. Microbiol. 72**:**5118-5121. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 37.Van der Auwera, G. A., and J. Mahillon. 2008. Transcriptional analysis of the conjugative plasmid pAW63 from Bacillus thuringiensis. Plasmid 60**:**190-199. [DOI] [PubMed] [Google Scholar]
• 38.Van der Auwera, G. A., L. Andrup, and J. Mahillon. 2005. Conjugative plasmid pAW63 brings new insights into the genesis of the Bacillus anthracis virulence plasmid pXO2 and of the Bacillus thuringiensis plasmid pBT9727. BMC Genomics 6**:**103. [DOI] [PMC free article] [PubMed] [Google Scholar]
• 39.Van der Auwera, G. A., S. Timmery, F. Hoton, and J. Mahillon. 2007. Plasmid exchanges among members of the Bacillus cereus group in foodstuffs. Int. J. Food Microbiol. 113**:**164-172. [DOI] [PubMed] [Google Scholar]
• 40.Van der Auwera, G. A., S. Timmery, and J. Mahillon. 2008. Self-transfer and mobilisation capabilities of the pXO2-like plasmid pBT9727 from Bacillus thuringiensis subsp. konkukian 97-27. Plasmid 59**:**134-138. [DOI] [PubMed] [Google Scholar]
• 41.Vilas-Bôas, G., L. A. Vilas-Bôas, D. Lereclus, and O. M. N. Arantes. 1998. Bacillus thuringiensis conjugation under environmental conditions. FEMS Microbiol. Ecol. 25**:**369-374. [DOI] [PubMed] [Google Scholar]
• 42.Wilcks, A., N. Jayaswal, D. Lereclus, and L. Andrup. 1998. Characterization of plasmid pAW63, a second self-transmissible plasmid in Bacillus thuringiensis subsp. kurstaki HD73. Microbiology 144**:**1263-1270. [DOI] [PubMed] [Google Scholar]
• 43.Yeo, H. J., and G. Waksman. 2004. Unveiling molecular scaffolds of the type IV secretion system. J. Bacteriol. 186**:**1919-1926. [DOI] [PMC free article] [PubMed] [Google Scholar]