Validation of a locally-developed DNA amplification system (DAS) for the detection of Staphylococcus aureus and Escherichia coli 0157:H7 in meat, milk and their products

The protocol for detection of Staphylococcus aureus from milk and milk products was optimized. Samples were enriched in nutrient-rich (N-R) broth supplemented with sodium chloride to inhibit the growth of other contaminating microorganisms. N-R broth + 2.5% was used only in the enrichment of UHT[ultra-high temperature]-processed fresh cow's milk, while N-R broth + 7% NaCl was used for the enrichment of the rest of the samples analyzed. The crude DNA from the enriched samples was extracted by boiling with 0.12 N NaOH for 10 min. PCR [polymerase chain reaction] was performed using the developed S. aureus DASsupTM kit and the result was compared with the culture method of plating in Baird Parker Agar (BPA). The presence of S. aureus contamination was determined from 70 swab samples (41-food contact surfaces, 12 cow's udders, 12 food handlers, and 5 carcasses), 47 raw materials (23 raw cow's milk, one raw carabao's milk, 6 liquid milk blends, 2 lecithin, 4 natural casings, and 11 raw meat), and 21 finished products (8 soft white cheese, 4 powdered filled milk, 3 mortadella, and one each of fresh cow's milk, chocolate milk, yogurt, pork sausage, beef sausage, and hotdog). All these samples were obtained from one meat and three dairy processing plants in the Philippines. Results showed that S. aureus was detected in 32% (N=138) of all the samples analyzed by both PCR and cultural methods. The incidence of the microbe was detected most often from raw materials which were composed of 18 raw cow's milk, 10 raw meat, 4 natural casings, and one raw carabao's milk. Only 7 out of 70 total swab samples were positive for S. aureus. The least incidence of the microbe was detected from the finished products which were composed of one soft white cheese, one breakfast pork sausage, and 2 smoked mortadella. A perfect agreement was obtained between the PCR and cultural methods in the analysis of all samples (N=138). A total of 33 presumptive S. aureus isolates were obtained and purified from all the samples analyzed and their identities were confirmed by ancillary tests such as coagulase production, gram reaction, DNAse test and mannitol utilization. The kit for PCR-based detection of E. coli O157:H7 was validated using as samples unspiked and artificially-spiked liquid milk, soft cheese and frozen beef burgers, unspiked raw meat ingredients and processed meats. A total of 15 cheese and 15 milk samples were analyzed by conventional culture method and by PCR method using the DASsupTM kit.The percent agreement between conventional culture method and the DASsupTM was 86.67% for the 15 cheese samples and 100% for the fifteen milk samples. In the analysis of five different brands of frozen hamburger patties unspiked and spiked with EHEC 0157:H7 (B-10084) at low (50 cfu/g) and high (1000 cfu/g) levels, all samples, including the uninoculated matrices, gave positive reactions with the EHEC 0157:H7 DAS kit. Results of the batch 2 testing, wherein nine uninoculated beef burger patties were analyzed, showed that one sample was found to be positive for EHEC O157:H7 by PCR detection. In batch 3 analysis, 15 out of the total number of 23 samples were positive by PCR analysis. The positive samples were two swabs of carcasses, raw meats, most of the meat ingredients and some of the finished products. Out of the 45 presumptive isolates from cheese, only one isolate, CS-9b, showed typical E. coli reactions by conventional tests and by colony PCR. In the case of the batch 1 meat isolates, none of the picked colonies were confirmed as E. coli by either the biochemical tests nor by PCR. For batch 2, four, isolates from sample A exhibited possible E. coli O157 identification as observed in its IMViC reactions and MUG-negative characteristic in Fluorocult 0157 agar. However, PCR result of the four presumptives A-4, A-5, A-7, and A-10 did not amplify the target 0.3 kb amplicon instead produced a multi-band pattern. Screening of colonies included sorbitol-fermenting isolates four of which namely A-22, A-76, A-77 and A-80 were found to be possibly variant 0157 strains. In batch 3, positive isolates were only retrieved in raw meat ingredient samples P3 (fat of pork) and P4 (ground fat of pork) and in finished product FP3 (smoked mortadella A). Isolates P3a and P4b were characterized along with three isolates obtained from FP3. The percentage agreement between the two methods employed in the analyses of uninoculated meat samples could not be computed because conventional confirmatory test were lacking. Likewise, the performance parameters such as sensitivity, specificity, false positive rate and false negative rate could not be determined. Improvement in the conventional tests and increasing the number of colonies sampled for confirmation would have to be undertaken. Other colonies belonging to other species or other strains should also be tested for amplification of the 0.3 kb (300 bp) target band to rule out false PCR positive reactions.