The genes required for PGA synthesis in B. anthracis

are part of the cap operon (capBCADE) located on the pX02 plasmid (Makino et al., 1989; Uchida et al., 1993; Candela et al., 2005). The capD gene encodes a γ-glutamyltranspeptidase, or PGA depolymerase, which was found to be required for the covalent anchoring of PGA to the peptidoglycan in B. anthracis (Candela & Fouet, 2005). PGA producers that lack capD produce a loose slime layer instead of a covalently linked capsule (Candela & Fouet, 2005). While all Group I and II isolates in this study tested negative for the four cap genes tested (including capD), they were still able to produce a covalently linked PGA capsule. Homologs to several of the cap genes AC220 have been found in other bacteria, such as Bacillus subtilis, Bacillus licheniformis, and Staphylococcus epidermidis (Ashiuchi & Misono, 2002; Urushibata et al., 2002; Candela & Fouet, 2005, 2006; Candela et al., 2005; Kocianova Selleckchem Verteporfin et al., 2005). It is possible that the Group I and II isolates contain PGA synthesis genes that more closely resemble these homologs than those in the cap operon of B. anthracis and are thus too divergent to amplify using the cap PCR primers used in this study. A clear difference in colony morphology on bicarbonate

agar was evident for about half of the isolates. These isolates, as well as the B. cereus G9241-positive control, produced dull or dry colonies, whereas virulent strains of B. anthracis produce the shiny or the mucoid morphology. Thus, although commonly used for capsule expression and observation in B. anthracis, the colony morphology of other species on bicarbonate agar alone is not an accurate indicator of capsule production. The d-PGA capsule of B. anthracis is required for virulence, in addition to Phospholipase D1 the tripartite toxin encoded by the pX01 plasmid. The role in virulence, if any, of the B. anthracis-like PGA capsules produced by

Group I and II isolates, however, is unknown. We cannot be certain that these isolates were responsible for the infections, as Bacillus spp. are frequent contaminates of clinical samples, and no further epidemiological data were available (Farrar & Reboli, 2006). However, during the course of this study, 20 clinical isolates from CDC’s Special Bacteriology Reference Laboratory’s (SBRL) culture collection were identified as having a high degree of 16S rRNA gene sequence similarity (99.87–100%) to one or more of the Group I or II isolates, and were subsequently tested for capsule production. These isolates had been sent by various states or territories (AK, CA, CO, MO, OH, PA, PR, TN, VA, and VT) for identification from 2001 to 2008, and were ultimately identified by SBRL as either Bacillus spp. or Brevibacterium spp. Nineteen of the 20 isolates tested positive for capsule production, detected by both methods described above.