To the Editor: Yersinia pestis causes plague, which primarily affects rodents, but is an invasive and virulent pathogen among humans. Y. pestis infection is endemic in small rodent populations in different parts of the world, and the bacterium is considered a potential bioweapon because it can be easily isolated, produced, dried, and dispersed as an aerosol. Antimicrobial drug treatment can be lifesaving during the early stages of illness; hence, rapid and sensitive methods for Y. pestis detection in environmental and clinical samples are required. Multiple PCR assays for Y. pestis detection that primarily detect markers located on plasmids have been developed (1–6). The plasminogen activator/coagulase (pla) gene, located on plasmid pPCP1, is incorporated into most Y. pestis PCR assays, and in several studies it was the prime or sole marker (1,2,5,7–9). Reasons for including pla in these assays are its occurrence in multiple copies, its absence from closely related Yersinia species, and its role in Y. pestis virulence (1,4,5).

While validating the specificity of a multiplex qPCR assay for the detection of Y. pestis (6), we obtained DNA from the dissected peritoneum of a black laboratory rat (Rattus rattus), which tested positive for the pla gene. Two other Y. pestis signature sequences were not amplified. Additional samples were analyzed from black (n = 11) and brown (Rattus norvegicus [n = 4]) rats that had been caught on poultry and pig farms in the southeastern region of the Netherlands during 2008. Positive indicators for pla were found in samples from 8 of these black rats and in samples from 2 of the brown rats. Samples from 2 laboratory rats tested negative for pla. Inferences of the incidence of pla-positive rats cannot be made because of low sample numbers and potential bias in capturing rats that had putative infections.

To exclude the possibility of contamination of host DNA with DNA from intestinal flora during isolation of the peritoneum, we examined the occurrence of pla in other tissues. Lung and liver samples were available from all 17 rats, and leg tissue samples were available from 7 rats, 5 of which had positive peritoneal tissue test results. The leg and lung tissues of 1 rat and the leg tissue of another rat tested positive, albeit at considerably lower quantities (higher quantification cycles) than pla values measured in peritoneal samples. These results did not support the likelihood of contamination during sampling or the occurrence of local infections; they did support the hypothesis of a systemic infection in the rats. To investigate whether the presence of the pla gene sequences indicated the presence of the carrier pPCP1 plasmid in Y. pestis, we designed PCR assays for the amplification of 3 conserved regions of this plasmid (Technical Appendix). Each assay produced PCR products from Y. pestis; only the transposase gene was amplified from Y. pseudotuberculosis. None of the PCR assays amplified DNA from samples collected from rats.

Pla genes obtained from 2 of the peritoneum samples collected from black rats were sequenced and appeared to be identical (GenBank accession no. HQ606074). Alignment with Y. pestis pla genes, which are highly conserved among Y. pestis isolates, revealed 11 nt differences in 880 bp (98.8% similarity). A BLAST (http://blast.ncbi.nlm.nih.gov/Blast.cgi) search retrieved such highly similar genes only from Y. pestis sequences; the next most similar sequences of other Enterobacteriaceae were at 78% similarity.

The genewalking PCR procedure was used to explore sequences adjacent to the pla gene (Technical Appendix). One PCR product was sequenced and appeared to be in part homologous to the pla gene, but the adjacent sequence displayed high homology to genes coding for replicon (rep) proteins in several bacterial genera in the family Enterobacteriaceae, e.g., Escherichia, Shigella, and Salmonella. The existence of a concatenated pla-rep sequence in rat tissue samples was confirmed by amplification of a PCR product from a primer targeting the pla gene, combined with a primer targeting the rep gene sequence that was acquired by using the genewalking procedure. The resulting 223-bp PCR product (GenBank accession no. JQ756394) consisted of a 141-bp sequence identical to the Y. pestis pla gene, linked to a 72-bp sequence that was 97% similar to enterobacterial rep protein genes. Attempts to obtain more sequence information from rep sequences by using primers derived from conserved domains in enterobacterial rep genes were unsuccessful. This suggests that the pla-rep sequence is derived from uncharacterized bacteria. Rep proteins function as replication activators of their carrier plasmids.

The rep sequences identified in this study were most similar to those of plasmids involved in bactericidal activity, a function that is also ascribed to the bacteriocin pesticin gene clusters of Y. pestis pPCP1 plasmids. The occurrence in unknown organisms that have pla genes that are similar to Y. pestis pla genes has consequences for the detection of Y. pestis. To prevent false positive results, detection protocols should include at least 1 supplemental target to confirm the presence of Y. pestis (6). In addition, investigators using pla gene analysis, for instance, while reconstructing ancient plague epidemics (10), should be aware of the occurrence of these homologs.