Polymerase Chain Reaction Detection of Enterotoxins Genes in Coagulase-Negative Staphylococci Isolated from Brazilian Minas Cheese

Introduction

E nterotoxigenic Staphylococcus aureus is one of the major pathogens causing food poisoning worldwide (Dinges et al., 2000). It has been considered the only pathogenic species among Staphylococcus, while coagulase-negative staphylococci (CNS) have been classified just as contaminant agents (Kloos and Bannerman, 1994). Recently, several authors have suggested that this group of microorganisms should be studied regarding the presence of enterotoxin genes (Rodriguez et al., 1996; Veras et al., 2008; Rall et al., 2010).

In Brazil, Minas cheese is the most popular cheese consumed all over the country. It is a soft and fresh white cheese with high pH, 55% moisture content, and low percentage of salt (1.4%–1.6%). It is produced on both industrial and domestic scales by adding lactic cultures or by direct acidification of milk (Carvalho et al., 2007). In a minor-scale production, small producers use raw milk in nonhygienic conditions. Besides the nonpasteurized milk, Staphylococcus genus from the food handler may contaminate Minas cheese (Carmo et al., 2004). The present study aimed to analyze the frequency of the sea, seb, sec, and sed genes in CNS strains isolated from samples of Minas soft cheese and to evaluate their ability to produce enterotoxins in vitro.

Materials and Methods

Results

CNS were found in 65 of 90 samples (72.2%): 23 (35.4%) were Staphylococcus saprophyticus, 16 (24.6%) Staphylococcus warneri, 10 (15.4%) Staphylococcus epidermidis, 9 (13.8%) Staphylococcus xylosus, 3 (4.6%) Staphylococcus haemolyticus, 2 (3.1%) Staphylococcus schleiferi subsp. schleiferi, and each one of (1.5%) Staphylococcus capitis subsp. urealyticus and Staphylococcus caprae. Simultaneous presence of two species in the same sample was also observed: one of them showing S. saprophyticus and S. warneri, and the other S. warneri and S. epidermidis.

Table 1 shows the results of molecular tests for the sea, seb, sec, and sed genes detection, and sea was the most frequent gene (18.5%). S. warneri was the only species presenting strains with the sea, seb, and sec genes.

Despite the presence of the enterotoxin genes, these strains did not produce enterotoxins in vitro.

Discussion

S. aureus is one of the most prevalent pathogens in food, causing several outbreaks (Veras et al., 2008), but the potential of CNS to cause foodborne disease is not clearly elucidated.

The enterotoxin genes in CNS have been described using polymerase chain reaction techniques. Herein, 17 (26.2%) CNS strains presented genes for classical enterotoxin production. A similar percentage was observed by Rall et al. (2010), with 19.4% of CNS isolated from food handlers. A lower prevalence was observed by Vernozy-Rozand et al. (1996), with 5.3% of the microorganisms isolated from goat milk and cheese presenting enterotoxin genes, and by Rosec et al. (1997), who found that all of 264 CNS strains were negative for these genes. On the other hand, sec and sed were present in 57.1% of CNS isolated by Rodriguez et al. (1996). Veras et al. (2008) found five out of eight (62.5%) of CNS encoding such genes.

The sea gene was found more frequently (18.5%), while sed was not observed. The SE genes seb and sec were found in similar percentages (3.1% and 4.6%, respectively). Rall et al. (2010) observed that sea also occurred more frequently among classical enterotoxin genes (24.1% of CNS).

Despite the presence of classical enterotoxin genes in 26.2% of CNS strains, none of them produced enterotoxins in vitro as evaluated by RPLA method. This result is in agreement with Harvey and Gilmour (1988, 1990), who showed that none of 353 and 384 strains of CNS isolated from goats milk and milk powders produced enterotoxins, respectively. One may speculate that it might be due to the low amounts of SEs production, which is not detectable by the method. Enterotoxin was produced using sac culture methods, which are considered to be the most efficient techniques (Robbins et al., 1974), and according to the manufacturer's instructions, the sensitivity of the Kit has been reported to be 0.5 ng/mL. Besides, all ATCC controls produced enterotoxins. Thus, probably these isolates were not expressing the enterotoxin genes. According to Robbins et al. (1974), CNS poorly produce enterotoxins even in optimal conditions. Besides, defects in toxin expression may occur due to point mutations that convert the toxin genes to silent genes (Okoji et al., 1993).

However, several authors have shown that enterotoxin production may be highly variable. Crass and Bergdoll (1986) observed that 10% of CNS isolated from food and animals were enterotoxin producers. A percentage of 5.3% was reported by Vernozy-Rozand et al. (1996), whereas Zell et al. (2008) reported that 45.7% of CNS isolated from food produced enterotoxins. Veras et al. (2008) observed that 62.5% of these microorganisms isolated from outbreaks were enterotoxin producers as well. Taken together, our data and those from literature point out that CNS should not be ignored in epidemiological studies of foodborne outbreaks, because the presence of genes for enterotoxins indicates the potential ability of those species to synthesize the proteins. Therefore, further investigations on the role of such genes are still required.