Comparison of the isolation rates and characteristics of Salmonella isolated from antibiotic-free and conventional chicken meat samples.

Salmonella contamination in chicken samples can cause major health problems in humans. However, not only the effects of antibiotic treatment during growth but also the impacts of the poultry slaughter line on the prevalence of Salmonellae in final chicken meat sold to consumers are unknown. In this study, we compared the isolation rates and antimicrobial resistance of Salmonellae among antibiotic-free, conventional, conventional Korean native retail chicken meat samples, and clonal divergence of Salmonella isolates by multilocus sequence typing. In addition, the distribution of extended-spectrum ß-lactamase (ESBL) genes in ESBL-producing Salmonella isolates was analyzed. A total of 72 retail chicken meat samples (n = 24 antibiotic-free broiler [AFB] chickens, n = 24 conventional broiler [CB] chickens, and n = 24 conventional Korean native [CK] chickens) was collected from local retail markets in Seoul, South Korea. The isolation rates of Salmonellae were 66.6% in AFB chickens, 45.8% in CB chickens, and 25% in CK chickens. By analyzing the minimum inhibitory concentrations of ß-lactam antibiotics with the disc-diffusion test, we found that 81.2% of Salmonella isolates from AFB chickens, 63.6% of isolates from CB chickens, and 50% of isolates from CK chickens were ESBL producers; all ESBL-positive isolates had the CTX-M-15 genotype. Interestingly, all ESBL-producing Salmonellae were revealed as ST16 by multilocus sequence typing and had the genetic platform of blaCTX-M gene (IS26-ISEcp1-blaCTX-M-15-IS903), which was first reported in Salmonellae around the world. The Salmonella ST33 strain (S. Hadar) isolated in this study has never been reported in South Korea. In conclusion, our findings showed that antibiotic-free retail chicken meat products were also largely contaminated with ESBL-producing Salmonellae and that their ESBL genes and genetic platforms were the same as those isolated from conventional retail chicken meat products.

Comparison of 2 types of broths and 3 selective agars for the detection of Campylobacter species in whole-chicken carcass-rinse samples.

In this study, we compared the effectiveness of 2 types of Bolton broths and 3 selective media for isolating Campylobacter spp. from naturally contaminated whole-chicken carcass-rinse samples. One hundred chickens were rinsed with buffered peptone water, and the rinses were added to 2× Bolton broth (with or without blood supplementation). The samples were incubated and then streaked onto Preston agar, modified cefoperazone charcoal deoxycholate agar (mCCDA), and Campy-Cefex agar, which was followed by incubation under microaerobic conditions. No statistical differences were observed (P > 0.05) in isolation rate and selectivity between the 2 types of Bolton broths. Among the 3 selective agars, Preston agar yielded a significantly (P < 0.05) better isolation rate and selectivity. The Campy-Cefex agar, which is recommended by many food authorities for its high quantitative detection ability, showed extensive contamination with competing microorganisms and exhibited the lowest isolation rate and selectivity.

Evaluation of selective enrichment broths and chromogenic media for Salmonella detection in highly contaminated chicken carcasses.

We evaluated the effectiveness of 2 selective enrichment broths, Rappaport-Vassiliadis Soy (RVS) and Muller-Kauffmann tetrathionate with novobiocin (MKTTn), for the isolation of Salmonella from chicken carcasses obtained from 3 different types of retail markets. We also compared a chromogenic agar, chromID Salmonella agar (SM-ID 2), with a classic plating medium, xylose lysine deoxycholate agar (XLD). Salmonella were isolated from 118 of the 180 samples (65.5%). Salmonella were detected in 105 samples (88%) plated on XLD and 111 samples (94%) plated on SM-ID 2 when RVS broth was used for enrichment, and 43 samples (36.4%) plated on XLD and 67 samples (56.8%) plated on SM-ID 2 when the MKTTn broth was used. The highest sensitivity was found in the RVS-XLD combination (0.99), followed by RVS-SM-ID 2 (0.97). The specificity of the RVS-SM-ID 2 combination was the highest (0.89), but that of the MKTTn-XLD combination was zero. The results of this study indicate that the selective enrichment broths had a great effect on the sensitivity and specificity of plating media, and our study confirms that the RVS broth is the most suitable enrichment for the investigation of Salmonella in chicken carcasses. This observation suggests that use of RVS broth for selective enrichment and SM-ID 2 for selective isolation may be the best combination to determine the presence of Salmonella in chicken carcasses.

Prevalence, phenotypic traits and molecular characterization of emetic toxin-producing Bacillus cereus strains isolated from human stools in Korea.

AIMS: To investigate the prevalence and genotypic/phenotypic characters of emetic toxin-producing Bacillus cereus strains isolated from sporadic food poisoning cases in Korea. METHODS AND RESULTS: The prevalence of emetic B. cereus was determined in 56 899 stool samples from sporadic food poisoning cases in Korea between 2004 and 2006. We assessed toxin profiles, phenotypic traits and antibiotic resistance. The molecular subtyping was ascertained using an automated repetitive sequence-based PCR (rep-PCR) system, DiversiLab™, with these emetic strains isolated from sporadic food poisoning cases and other emetic strains isolated from an outbreak and food samples. Emetic B. cereus was present in 0·012% of sporadic food poisoning cases. The prevalence of nheABC, hblCDA, cytK and entFM enterotoxin genes among emetic strains was 100, 14·3, 14·3 and 100%, respectively. Most emetic strains were negative for salicin hydrolysis (100%), starch fermentation (85·7%) and haemolysis (85·7%). One emetic isolate, VK7, exhibited several unique traits, such as harbouring the hbl gene and ability to hydrolyse starch. All isolated strains were highly resistant to ß-lactam antibiotics. All emetic strains except VK7 exhibited an identical rep-PCR banding pattern, while nonemetic strains were classified into various pulsotypes. CONCLUSIONS: Most emetic strains except one isolate exhibited similar genotypic/phenotypic traits and subtyping pattern. Automatic rep-PCR (DiversiLab™) may be used to discriminate emetic strains from nonemetic strains, although we could not distinguish between most emetic strains using that. SIGNIFICANCE AND IMPACT OF THE STUDY: Result of this study may contribute an extended database on the prevalence and toxigenic traits of emetic B. cereus strains isolated from Korea.