Occurrence and molecular characterization of Staphylococcus aureus strains isolated from meat and dairy products by PCR-RFLP

Abstract

A total of 100 Staphylococcus aureus strains were isolated from 1,047 food samples. In addition to biotyping all isolates, the occurrence of the tst gene and protein were evaluated by PCR and the RPLA test, respectively. Moreover, polymorphism of the X-region of the protein A gene was analyzed by PCR-RFLP. TSST₁ was detected in 12 strains and production of TSST₁ in all strains was confirmed by RPLA assay. It was noteworthy that 66.7% of TSST₁-producing S. aureus strains belonged to the human biotype. The result of genotyping amplification of the spa gene displayed size polymorphisms and revealed seven different clusters ranging from 1,200 bp to 1,600 bp. Digestion of amplicons with HindIII and HaeII resulted in six distinct patterns for each of them. An amplicon of 1,450 bp in size was the predominant type of Spa-PCR product. The accuracy of the Spa typing system by PCR-RFLP was 100%, thus confirming this method as a reliable tool in routine epidemiological investigation for S. aureus.

Introduction

Foodborne illness is a serious public health problem, and is associated with reduced economic growth. Staphylococcus aureus is the second or third most important cause of these illnesses throughout the world (Normanno et al. 2005). S. aureus strains produce several toxins and have virulence factors that are responsible for pathogenicity. Enterotoxins and toxic shock syndrome toxin 1 (TSST₁) are members of the pyrogenic toxin superantigen (PTSAg) family. SAgs act through major histocompatibility complex (MHC) class II molecules and the specific V_B chain of the T cell receptor (TCR), which leads to extensive T cell proliferation and release of proinflammatory cytokines (Normanno et al. 2007). TSST₁ is the causative agent in the severe and life threatening toxic shock syndrome (TSS) in humans. Increasing rates of non-menstrual TSS, and horizontal transfer of these strains with mobile genetic elements (Hwang et al. 2007), especially in high risk people such as the elderly, young children and immunocompromised people, are among the worrying issues surrounding this toxin. Typing of S. aureus is an important tool in the evaluation of strain origin, infection rout, reservoir detection, surveillance of major infections, epidemiological outbreak investigation and setting up preventive strategies (Manfreda et al. 2005). During the last few years, phenotypic methods have been replaced with molecular approaches for typing of S. aureus. However, biotyping using the simplified system published by Devriese (1984), indicates sources of S. aureus strains in animal foods. These methods could complement genotyping methods (Hata et al. 2005). Numerous molecular typing methods can be used for characterization of S. aureus strains. Among these techniques, pulse field gel electrophoresis (PFGF) became the gold standard, with excellent discriminatory power for genotyping of S. aureus strains (Strommenger et al. 2008). However, it has disadvantages, e.g., it is time-consuming, labor intensive and difficult to interpret and standardize (Hallin et al. 2007). spa typing based on the polymorphic X region of the protein A gene (spa), provides an easy, inexpensive, rapid, simple to interpret, accurate and reproducible type analysis for both local and global epidemiological studies (Wichelhaus et al. 2001). Food safety rules in the industrialized world differ from those in developing countries. Furthermore, the findings of epidemiological investigation varies from one locality to another. To date, there is no data published on the characterization of S. aureus strains in foods in Iran. The aims of the current study were: (1) to evaluate the prevalence of S. aureus in dairy and meat products in Iran; (2) to characterize the isolated strains based on production of TSST₁ using PCR and immunoassay methods; and (3) to perform spa typing of the isolated strains by determination of different patterns of the variable regions of the protein A gene based on PCR followed by restriction fragment length polymorphism (RFLP) patterns.

Materials and methods

Sampling

A total of 1,047 food samples including 455 dairy products, 458 meat products and 134 other products were purchased and were collected from July 2006 to November 2007, in Tehran, Iran (Table 1).

Table 1 Occurrence of Staphylococcus aureus strains in analyzed samples

Identification of Staphylococcus isolates

Baird Parker agar containing egg-yolk tellurite emulsion (Merck, Darmstadt, Germany) was used for isolation. Isolates were identified using the following criteria: production of coagulase, DNase, catalase, mannitol fermentation, hemolytic zone on 5% sheep blood agar (Merck), VP test and Gram staining (Hata et al. 2005).

Detection of tst gene by PCR and immunoassay

For each strain, an overnight culture in Brain heart infusion Agar (Merck) was centrifuged and treated with 2 µl lysostaphin (10 mg/ml; Sigma, St. Louis, MO) for 1 h at 37°C (Normanno et al. 2007). DNA extraction was carried out by using the QIA amp DNA mini kit (Qiagen, Hilden, Germany).

The presence of the Staphylococcal tst gene was assessed by PCR. The sequence of the primers used was: 5′-CATCTACAAACGATAATATAAAGG-3′, and 5′-CATTGTTATTTTCCAATAACCACCCG-3′ forward and reverse, respectively (Fueyo et al. 2005b). PCR reactions were performed by using HotStar Taq plus Master Mix Kit (Qiagen). For amplification, the reaction mixture (25 µl) contained 1 µl of each primer (20 pmol/µl), 12.5 µl Hotstar Taq Plus master mix (DNA polymerase, MgCl₂,dNTP), 2.5 µl 10x Coralload buffer (Buffer, dyes and gel loading reagent), 5 µl RNase-free water and 3 µl DNA (50 ng/µl). Reactions cycles were are follows: 94°C for 3 min, followed by denaturation at 94°C for 45 s, annealing at 55°C for 30 s, then extension at 72°C for 1.5 min. After 35 cycles, a final 10 min extension at 72°C was performed. Amplicons were assessed by electrophoresis in 1% agarose gels. Staphylococcus aureus RN 8465 and Staphylococcus epidermidis ATCC 12228 were used as positive and negative controls, respectively.

The production of TSST₁ was determined using a commercial reverse passive latex agglutination kit (TST-RPLA, Oxoid, Wesel, Germany) according to the manufacturer's instructions.

Biotyping

Biotyping was performed according to the simplified system described by Devriese (1984), using four tests including staphylokinase production with bovine fibrinogen (Sigma), β-haemolysis on sheep blood agar (Merck), coagulation of bovine plasma for a duration of 6 h, and crystal violet growth types on Tryptose agar (Merck). All strains were classified to host specific ecovars and non-host specific (NHS) biotypes (Table 2).

Table 2 Classification of Staphylococcus aureus non-host specific (NHS) biotypes (Devriese 1984)

Molecular typing by spa gene restriction profile analysis

The X region of the spa gene was amplified by PCR using the primers SPA1 (5′-ATCTGGTGGCGTAACACCTG-3′) and SPA2 (5′-CGCTGCACCTAACGCTAATG-3′; Wichelhaus et al. 2001). All PCR series included S. aureus RN 8465 and S. epidermidis ATCC 12228 as positive and negative amplification controls, respectively. Prior to RFLP, a visible amount of PCR products of the spa gene were separated completely on a 2% agarose gel and DNA was purified thoroughly using a DNA extraction kit (Fermentas, St. Leon-Rot , Germany) according to the manufacturer's instructions. Purified amplicons were digested with 50 U HindIII (Fermentas) at 37°C for 16 h and 5 U HaeII (Roche, Mannheim, Germany) at 37°C for 4 h, in separate reactions. All digested fragments were assessed by electrophoretic separation in 2% (w/v) agarose gels.

Typeability is defined as the number of strains that are located in a type by a typing system (Hallin et al. 2007).

Statistical analysis

Data were analyzed by means of chi-square (χ ²) test and Fisher's exact test.

Results

On the basis of the results obtained from the cultural and biochemical properties of the 1,047 food samples analyzed, 100 (9.5%) were contaminated with S. aureus, of which 45 isolates showed β-haemolysis, 30 α-hemolysis and 25 non-haemolysis. Sixteen (3.5%) of the 458 meat product samples, and 77 (17.1%) of the 455 dairy products were contaminated with S. aureus. The contamination rate of dairy products was significantly higher than that of other foods (P < 0.001; Table 1). Of the 100 strains isolated, 12 (12%) isolates harbored the tst gene as shown by PCR amplification. Eight (10.4%) of the 77 dairy products and 3 (18.8%) of the 16 meat products were contaminated with TSST₁-producing S. aureus strains. Figure 1 shows the presence of the tst gene.

Fig. 1

PCR amplification products for the Staphylococcus aureus tst gene. Lanes: M DNA marker (100 bp ladder); 1 negative control; 2, 4 negative products; 3 positive control S. aureus strain RN8465; 5 positive product

All of the tst-containing S. aureus strains detected by PCR produced TSST₁ by RPLA immunoassay detection. Conversely, RPLA confirmed the PCR results for all 12 strains. Of the 100 analyzed strains by the biotyping method, 29 belonged to the human ecovar, 47 to the NHS ecovar, 11 to the poultry ecovar, 9 to the bovine ecovar and 4 were not allocated (Table 3). Among the NHS strains, 8 (8%) were K-β + CV:A, 10 (10%) K + β-CV:A, 22 (22%) K + β + CV:A and 7 (7%) K-β-CV: C. Non ovine ecovar was found (Table 4). Notably, 66.7% of TSST₁-producing S. aureus strains belonged to the human biotype.

Table 3 Distribution of the tst gene and biotypes among the S. aureus strains isolated from different foods

Table 4 Biotypes of S. aureus isolated from foods

Amplification of the X region of the spa gene revealed a single amplicon for each isolate. The sizes of the PCR products ranged from approximately 1,200 to 1,600 bp. The 100 strains were divided into seven different clusters (A–G) with seven different sizes of approximately 1,200, 1,250, 1,300, 1,450, 1,500, 1,550 and 1,600 bp. Gene polymorphisms of the X region of protein A are shown in Fig. 2. Among the seven types, type D with an amplicon of 1,450 bp was predominant (27%) followed by type C (18%), type G (17%), type E (16%), type F (14%), type A (5%) and type B (3%). Restriction with the enzyme HindIII resulted in six distinct RFLP patterns among the 100 strains, which were classified as HindS₁–HindS₆ (Fig. 3). The results revealed that RFLP-type HinSd₄ predominated, with 28% isolates (Table 4). All the types obtained from PCR-RFLP with HindIII subdivide into several subtypes. Thus, 22 genotypes in total were identified (Table 5). Genotypes C₁ and D₄ were most prevalent (13%). Furthermore, the 100 isolates generated six RFLP different restriction patterns (HaeS1–HaeS6) using the restriction enzyme HaeII (Fig. 4). Sixteen genotypes were obtained with HaeII (Table 5). RFLP type HaeS₃ and genotype D₁ were predominant using HaeII digestion (31% and 21%, respectively). Typeability of spa typing using the PCR-RFLP system on the 100 strains analyzed was 100%, and all of the 100 strains could be typed using this method.

Fig. 2

Polymorphisms of the gene encoding the X region of protein A in S. aureus strains isolated from foods. Lanes: M 100 bp plus DNA marker, 1 positive control S. aureus COL strain, 2–9 PCR products, 10 negative control S. epidermidis ATCC12228

Fig. 3

PCR amplicons of the X region of protein A were digested with HindIII. Lanes: M ₁ 100 bp plus DNA marker; 1 S. aureus COL strain; 2–7 different restriction patterns; M ₂ 100 bp DNA marker

Table 5 Occurrence of S. aureus types and subtypes based on the X region of the protein A gene

Fig. 4

PCR amplicons of the spa gene were digested with HaeII. Lane M ₁ 100 bp plus DNA marker; Lane M ₂ , 100 bp DNA ladder; Lane 1, S. aureus COL strain; Lane 11, negative control; Lanes 2–10 and 12–15, different restriction patterns

Discussion

Foodborne illnesses are caused by consuming contaminated foods. Staphylococcus aureus can grow easily in every environment and can contaminate different foods (Oh et al. 2007). In our survey, the rate of occurrence of S. aureus in the foods analyzed was 9.5%, with a significant higher rate in dairy products (17.1%) vs meat product (3%). In comparison with other studies throughout the world, the presence of S. aureus is lower than on meat and dairy products in Italy (12.5%; Normanno et al. 2007), on bovine and caprine bulk milk and in raw milk products (17%; Jorgensen et al. 2005), and on raw and cooked meat products (57.1%; Atanassova et al. 2001). Nevertheless, in their study on ready-to-eat food in Korea, Oh et al. (2007) reported that 8.6% of samples were contaminated by S. aureus. Thus, the higher contamination rate of dairy products in our study is agreement with some other reports.

We found also the presence of TSST₁-producing S. aureus in foods. The results showed that 12% of strains were carriers of the tst gene. Jung et al. (2005) suggested a need for TSST₁ testing in food following the isolation of S. aureus from handled foods. The present study found a good correlation between RPLA and PCR results for detection of TSST₁-producing S. aureus. However, PCR detects only the presence of the genes, whereas monitoring production of TSST₁ is dependent on immunoassay methods (Zschock et al. 2000). The prevalence of TSST₁ by S. aureus strain (12%) was lower than that associated with foods and bovine mastitis (42.2, 15.5, 27.1, 13.5, 58.4, and 34.8%) reported in previous studies by Chapaval et al. (2006), Akineden et al. (2001), Adesiyun et al. (1992), Oh et al. (2007), Hata et al. (2005) and Nagase et al. (2002), respectively. However, in some other reports this rate was lower than in the present study (Fueyo et al. 2005a; El-Ghodban et al. 2006).

Numerous typing techniques to characterize S. aureus strains include genotypic and phenotypic methods (Wu and Della-Latta 2002). We chose the biotyping technique established by Devriese (1984) as a conventional typing method. This simplified biotyping system has been useful in tracing the origin of S. aureus strains in animal feed and in the food industry (Kitai et al. 2005). In the present study, of the 100 strains biotyped, 47% belonged to the NHS biotype, followed by 29% to human, 11% to poultry and 9% to bovine biotypes. In dairy products the NHS biotype was predominant (42%), while in meat products the human biotype had a higher rate (43.8%). In our study, the human biotype was more prevalent in comparison with some other studies on biotyping of S. aureus strains recovered from meat products (Capita et al. 2002; Rodríguez-Calleja et al. 2006). This supports a positive correlation between human biotype strains and handled meat foods and suggests that human contamination of meat products can be primary source of staphylococcal food poisoning, especially considering that 66.7% of TSST₁-producing S. aureus strains also belonged to the human biotype. Casciano et al. (2003) found that, in biotyping of S. aureus strains from raw milk, NHS biotype is the predominant type (66.6%). These results confirm our findings. In two studies carried out on the characterization of S. aureus strains isolated from meat and dairy product, in total, the human biotype occurred more frequently than other biotypes (Kérouanton et al. 2007; Normanno et al. 2007). These findings are in disagreement with the present study. However, strains that are NHS and human biotype may pose a potential risk for public health (Capita et al. 2002).

From among the available molecular typing methods for S. aureus strains, we performed spa typing by using PCR-RFLP. The polymorphic X region of the protein A (spa) contains a variable number of repeats of a 24-bp region that provides a suitable tool for epidemiological investigation (Hallin et al. 2007). In our study, amplification of the spa gene yielded a single amplicon with seven different sizes ranging from approximately 1,200 to 1,600 bp. An amplicon with a size of 1,450 bp was the most prevalent. Many reports using spa typing of S. aureus strains isolated from food as a genotyping method have been published (Annemüller et al. 1999; Jakubczak et al. 2007). Annemüller et al. (1999) reported that spa typing results corresponded well to DNA fingerprinting patterns. Six different RFLP patterns (HindS₁–HindS₆) were observed among the 100 strains analyzed here after HindIII digestion. These strains were analyzed further with the restriction enzyme HaeII, yielding six distinct patterns. Unlike the coa gene, which has commonly been used for typing, no reports are available on spa typing of S. aureus strains isolated from food by PCR-RFLP. Only one study has been carried out on this subject in S. aureus strains isolated from bovine mastitis milk in Germany (Strommenger et al. 2008). Wichelhaus et al. (2001) identified PCR-RFLP as a new typing system for S. aureus with a discriminatory power corresponding effectively with PFGE. This method represents a powerful, rapid, reliable, easy and inexpensive system for epidemiological surveillance (Hallin et al. 2007). In our study, typeability using the PCR-RFLP system was 100%. We subdivided the types obtained from each digestion into several subtypes (Table 5). These types aimed for genotype definition similar to that already carried out for coagulase genotypes (Katsuda et al. 2005). Finally, 22 genotypes with HindIII and 16 genotypes with HaeII digestion were obtained.

In summary, the prevalence of S. aureus strains in different foods in Iran is no higher than in other countries. The NHS biotype is the predominant ecovar in our strains, although the pattern in meat and dairy products is different. Type D spa products, HindS₄ of PCR-RFLP and subsequently genotypes C₁ and D₄ with HindIII digestion and types HaeS₃ of PCR-RFLP and genotype D₁ with HaeII digestion were the most prevalent, respectively. Spa typing by PCR-RFLP was a reliable system for genotyping of S. aureus strains with 100% typeability in our study. This paper is the first report on characterization of S. aureus isolates in meat and dairy products in Iran.