P3-183 Comparative Evaluation of Two Multiplex Real-time Quantitative PCR (qPCR) for Detection of Salmonella spp. and S. Enteritidis in Pooled Egg Preenrichment Samples

Wednesday, July 12, 2017
Exhibit Hall (Tampa Convention Center)
Gina Ryan , U.S. Food and Drug Administration , College Park , MD
Melanie Butler , U.S. Food and Drug Administration , College Park , MD
Anna Laasri , U.S. Food and Drug Administration , College Park , MD
Thomas Hammack , U.S. Food and Drug Administration , College Park , MD
Introduction: Foodborne pathogens, Salmonella entericaserotypes Enteritidis (SE) is a major problem in shell egg contamination. Current standards for validating qPCR assays for SE inclusivity require testing 100 strains at high concentration levels (LOD + 2 logs).

Purpose:  Comparative evaluation of two FDA-developed multiplex SE qPCR assays for performance under challenging conditions (e.g., low contamination levels, egg-preenrichment background, divergent SE strains).

Methods: Two previously validated SE assays were evaluated: prot6E-qPCR (targets invA and prot6E genes) and safA-qPCR (targets ttrA and safA genes). A total of 350 Salmonella strains, including 100 diverse serotypes (FDA SAFE collection) and 250 S. Enteritidis from egg and poultry-associated sources were tested for inclusivity at three levels (LOD, LOD+1 log and LOD+2 logs) in pure isolate broth culture or egg preenrichment samples.

Results: At the highest level tested, all strains were correctly detected with safA-qPCR (both probes) and prot6E-qPCR (invA probe), while false negative results (29/250 SE) were obtained with prot6E-qPCR (prot6e probe). Assay performance for inclusivity was significantly reduced at lower contamination levels and in matrix background, though the extent varied widely for both assays and individual probes. Overall, differences in SE-qPCR assays ability to detect diverse Salmonella strains under challenging conditions may not be apparent using current validation guidelines.

Significance: Our data showed current validation standards (i.e., detection at LOD+2 logs) may not be sufficient to accurately predict assay performance under sub-optimal conditions (i.e., near the LOD, diverse strains, food matrix background) encountered in food environments.