Search for Bacillus anthracis potential vaccine candidates by a functional genomic-serologic screen

Abstract

Bacillus anthracis proteins that possess antigenic properties and are able to evoke an immune response were identified by a reductive genomic-serologic screen of a set of in silico-preselected open reading frames (ORFs). The screen included in vitro expression of the selected ORFs by coupled transcription and translation of linear PCR-generated DNA fragments, followed by immunoprecipitation with antisera from B. anthracis-infected animals. Of the 197 selected ORFs, 161 were chromosomal and 36 were on plasmids pXO1 and pXO2, and 138 of the 197 ORFs had putative functional annotations (known ORFs) and 59 had no assigned functions (unknown ORFs). A total of 129 of the known ORFs (93%) could be expressed, whereas only 38 (64%) of the unknown ORFs were successfully expressed. All 167 expressed polypeptides were subjected to immunoprecipitation with the anti-B. anthracis antisera, which revealed 52 seroreactive immunogens, only 1 of which was encoded by an unknown ORF. The high percentage of seroreactive ORFs among the functionally annotated ORFs (37%; 51/129) attests to the predictive value of the bioinformatic strategy used for vaccine candidate selection. Furthermore, the experimental findings suggest that surface-anchored proteins and adhesins or transporters, such as cell wall hydrolases, proteins involved in iron acquisition, and amino acid and oligopeptide transporters, have great potential to be immunogenic. Most of the seroreactive ORFs that were tested as DNA vaccines indeed appeared to induce a humoral response in mice. We list more than 30 novel B. anthracis immunoreactive virulence-related proteins which could be useful in diagnosis, pathogenesis studies, and future anthrax vaccine development.

FIG. 1.

Transcription-translation and immunoreactivity profiles of selected ORFs. [³⁵S]methionine-labeled transcription and translation products (lanes T&T) were analyzed by SDS-PAGE, followed by autoradiography. The labeled T&T products were immunoprecipitated with different B. anthracis antisera (lanes R-1, R-2, R-3, R-4, G-1, and G-2). NRG was a mixture of naïve rabbit and guinea pig sera. The positions of size markers (in kDa) are indicated on the left. Note the seronegative responses for pXO₁-12, pXO₁-66, and BA3854.

FIG. 2.

Determination of IP titers of selected T&T products. Titrations were performed with a constant volume of the T&T reaction mixture and with threefold serial dilutions of the different sera. (A) Precipitation of PA with the R-1 and G-1 antisera. (B and C) Immunoprecipitation profiles for selected ORF products with the R-1 antiserum. The histograms show the radioactivity counts for the IP profiles of PA in panel A and of DppA and HtrA in panel C. End points for IP titers are indicated by arrows.

FIG. 3.

Transcription-translation and immunoreactivity profiles of selected pXO1 and pXO2 ORFs. (A) T&T and IP analysis of PA, LF, and EF, performed with the R-1, G-1, and R-2 antisera. (B) T&T reactions of the seven pXO1 ORFs with unknown functions that were successfully expressed. (C) Detection of radioactivity for IP reactions of pXO1 ORF products with the R-1, R-2, G-1, G-2, and R-3 antisera. (D) T&T and IP analysis of pXO2 ORFs with the R-2 and G-1 antisera. pXO₂-8 reacted only with the G-1 antiserum, while pXO₂-42 reacted only with the R-2 antiserum. All other pXO2 ORF products were negative with all of the antisera tested (Table 1).