Multilocus sequence typing scheme for bacteria of the Bacillus cereus group

Abstract

In this study we developed a multilocus sequence typing (MLST) scheme for bacteria of the Bacillus cereus group. This group, which includes the species B. cereus, B. thuringiensis, B. weihenstephanensis, and B. anthracis, is known to be genetically very diverse. It is also very important because it comprises pathogenic organisms as well as bacteria with industrial applications. The MLST system was established by using 77 strains having various origins, including humans, animals, food, and soil. A total of 67 of these strains had been analyzed previously by multilocus enzyme electrophoresis, and they were selected to represent the genetic diversity of this group of bacteria. Primers were designed for conserved regions of housekeeping genes, and 330- to 504-bp internal fragments of seven such genes, adk, ccpA, ftsA, glpT, pyrE, recF, and sucC, were sequenced for all strains. The number of alleles at individual loci ranged from 25 to 40, and a total of 53 allelic profiles or sequence types (STs) were distinguished. Analysis of the sequence data showed that the population structure of the B. cereus group is weakly clonal. In particular, all five B. anthracis isolates analyzed had the same ST. The MLST scheme which we developed has a high level of resolution and should be an excellent tool for studying the population structure and epidemiology of the B. cereus group.

FIG. 1.

Genetic relationships among 77 isolates of the B. cereus group. Strains were clustered by UPGMA applied to a distance matrix of pairwise differences between allelic profiles. Alleles were identified either for entire gene fragments (Table 4) (a) or after fragments were split into three equal-length parts (Table 5) (a). The scale bars indicate the percentages of mismatches between allelic profiles.

FIG. 2.

Phylogenetic relationships among 77 B. cereus group isolates inferred from individual genes. Trees were constructed by using the NJ method applied to pairwise distances among strains, computed by using Kimura's (19) nucleotide substitution model. Five groups of strains that were recovered for each tree were designated I, IIa, IIb, IIc, and IId. For the sake of simplicity, strain designations and detailed relationships among isolates within the groups are not shown (detailed trees and compositions of the clusters are available from us). For each of the five groups, the numbers in parentheses indicate the number of missing isolates/number of additional isolates compared to the number of isolates in the adk tree, and the numbers above the branches indicate bootstrap values, expressed as percentages (based on 1,000 replicates). In the adk tree, clusters I, IIa, IIb, IIc, and IId contain 18, 8, 12, 13, and 14 strains, respectively. B. anthracis belongs to cluster IIc. Isolates AH 1272 and AH 1273 (ST 52) and AH 1247 (ST 49), which cluster within cluster I, within cluster II, or between the two groups depending on the gene analyzed, are indicated by an asterisk and a number sign, respectively. The scale bars indicate 0.005 nucleotide substitution per site.

FIG. 3.

Comparison of MEE and MLST data for a subset of 67 B. cereus group isolates (out of the 77 strains used in this study). Both dendrograms were generated by the UPGMA method from a matrix of coefficients of genetic distances. The scale bars indicate the percentages of allelic differences between electrophoretic types for MEE and between STs for MLST. The MEE dendrogram was based on 13 enzyme loci as described previously (11-13).