Nadine Yossa, Sonny Huang, Travis Canida, Rachel Binet, Dumitru Macarisin, Rebecca Bell, Sandra Tallent, Eric Brown, Thomas Hammack
Reference methods for microbiological safety assessments of cosmetics rely on culture methods that reveal colonies of live microorganisms on growth media. Rapid molecular technologies, such as qPCR, detects the presence of target DNA in samples from dead and viable cells. DNA intercalating dyes, such as propidium monoazide (PMAxx), are capable of restricting PCR amplification to viable microbial cells. Here we developed singleplex and multiplex real time (qPCR) assays for the detection of Bacillus cereus (B. cereus) using 16S rRNA and phosphatidylcholine-specific phospholipase C (PLC) gene specific sequences coupled with PMAxx. The limit of detection was determined to be ~ 1 log CFU/ml for 16S rRNA and 3 log CFU/ml for PLC detection in pure culture using an eye shadow isolate, B. cereus 3A. We assessed the inclusivity and exclusivity of our qPCR assays using 212 strains, including 143 members of B. cereus, 38 non- B. cereus. and 31 non-Bacillus species; inclusivity was 100% for the 16S rRNA and 97.9% for the PLC targets; the exclusivity was 100% for 16S rRNA and 98.6% for PLC targets. These qPCR assays were then used to assess samples of commercial cosmetics: one set of liquid face toners (N = 3), artificially contaminated with B. cereus 3A, and one set of powdered cosmetics (N = 8), previously determined to be contaminated with B. cereus. For some samples, test portions were analyzed by qPCR in parallel, with and without PMAxx treatment. All test portions were simultaneously streaked on BACARA plates to confirm viable cells of B. cereus, according to the culture method. We found no difference in sensitivity between the singleplex and the multiplex qPCR assays (P > 0.05). Inoculated samples that did not recover B. cereus on plates still showed amplification of the DNA targets. However, that amplification was significantly delayed in PMAxx -treated samples (P < 0.0001) with C value differences of 7.82 for 16S rRNA and 7.22 for PLC. Likewise, amplification delay was significant (P < 0.0001) with inoculated samples that recovered B. cereus on plates with C value differences of 2.96 and 2.36 for 16S rRNA and PLC, respectively, demonstrating the presence of dead cells in the samples. All our qPCR results correlated with detection on BACARA plates (kappa, k = 0.99), independently of the presence of PMAxx in the PCR assays. Nevertheless, the amplification threshold with PMAxx dyes was significantly higher than the non-PMAxx dyes. Our findings confirm qPCR can be used for more rapid detection of microorganisms in cosmetics, including B. cereus, and selective detection of viable cells can be improved using PMAxx dyes.
