Detection of salmonellae in chicken feces by a combination of tetrathionate broth enrichment, capillary PCR, and capillary gel electrophoresis.

This report describes a rapid detection procedure for salmonellae from chicken feces by the combination of tetrathionate primary enrichment (preenrichment [PE])-bacterial lysis-capillary PCR and capillary gel electrophoresis. Pure Salmonella enterica serovar Enteritidis 64K was reisolated and detected by capillary PCR after buffered peptone water and nutrient broth, tetrathionate broth base Hajna (TTBH), and tetrathionate broth (TTB) preenrichments. When the same culture was mixed with intestinal homogenate, bacteriological reisolation and capillary PCR detection was achieved only by TTBH and TTB preenrichments. Capillary gel electrophoresis revealed that a Salmonella genus-specific 281-bp PCR product was detected when Salmonella strains but not non-Salmonella strains were tested. The detection limit of capillary PCR with whole-cell DNA extracted from pure Salmonella enterica serovars Enteritidis 64K, Typhimurium LT2-CIP60-62, and Gallinarum 64K was 3, 3, and 9 CFU ml(-1), respectively. The detection limit of capillary PCR from whole-cell DNA extracted from intestinal homogenate artificially contaminated with the same three strains was 3, 3, and 7 CFU ml(-1), respectively. We compared the results of the capillary PCR and bacteriological examination from the natural samples. Thirty-five of 53 naturally contaminated samples produced a specific PCR product. In 9 of the 35 PCR-positive samples, Salmonella could not be detected bacteriologically either by PE or a primary and delayed secondary enrichment (DSE) combination. In the 18 PCR-negative samples, 4 samples were found to harbor Salmonella by both PE and DSE and 14 samples were positive after DSE. Fifty-three additional intestinal homogenate samples, which were negative by their PE and DSE in bacteriological examination, were found to be also negative by their PCRs. The total time required to detect Salmonella with the capillary PCR method we used was approximately 20 h. If samples are from clinically diseased birds, the total time for PCR and detection is reduced to 2 h since the 18-h PE is not required. These results indicate that TTB enrichment, bacterial lysis, and genus-specific capillary PCR combined with capillary gel electrophoresis constitute a sensitive and selective procedure which has the potential to rapidly identify Salmonella-infected flocks.