1. Introduction
Staphylococcus aureus is one of the most common foodborne pathogens causing intoxication. Staphylococci can be introduced in the environment of food processing installations through various routes, as raw materials, food handlers or poor hygiene in food processing equipment. They are resistant to desiccation and can survive on different surfaces and resist sanitation forming biofilms [
1].
S. aureus can also contaminate foods during preparation and processing, and temperature abuse conditions during transport and/or storage can allow bacterial growth and enterotoxin production.
S. aureus is able to tolerate pH ranges from 4.5 to 9.0 and NaCl concentrations up to 9% and can grow and express virulence in a wide range of environmental conditions [
2]. The pathogen causes food poisoning through the ingestion of heat stable staphylococcal enterotoxins (SE) preformed in food. Meat and meat products, milk and dairy products, bakery products, salads, etc., are commonly involved in staphylococcal food poisoning (SFP). Bacterial toxins represented the second most common causative agent of outbreaks in the European Union (EU) with 19.3% foodborne outbreaks in 2019 [
3].
Staphylococcal enterotoxins are the main virulence factors associated with
S. aureus and the primary cause of staphylococcal food poisoning. Most of the staphylococcal food poisoning outbreaks are classified as weak-evidence outbreaks, as only the classical enterotoxins SEA, SEB, SEC, SED and SEE can be detected commercially [
4]. Together with these five SEs, new described enterotoxins and staphylococcal enterotoxin-like proteins have been characterized [
5,
6]. A high number of strains harbor the enterotoxin gene cluster (
egc) [
7], containing newer enterotoxin genes (
seg,
sei,
sem,
sen,
seo and
seu), and widely distributed in
S. aureus isolated from foods and food handlers [
8]. Genes encoding SE are located on mobile elements as plasmids, bacteriophages and pathogenicity islands, representing an additional risk factor in food intoxications due to the posible horizontal gene transfer. Additionally,
S. aureus strains usually carry more than one SE gene [
9]. Besides enterotoxins,
S. aureus produce other virulence factors such as exfoliative toxins, toxic shock syndrome toxin, or the Panton-Valentine leukocidin (PVL) [
10]. Staphylococcal enterotoxin A accounts for 80% of reported SFP cases, followed by enterotoxin B. Its worldwide predominance has been extensively documented [
11].
Multidrug resistant strains have been found in SFP cases and isolated from foods. Multiple antibiotic-resistant strains of
S. aureus are spreading rapidly around the world, which raises serious health concerns [
12]. Methicillin resistant
S. aureus (MRSA) is a major nosocomial emerging pathogen with increasing concern on the livestock industry. Livestock associated MRSA (LA-MRSA) are common colonizers of swine, and could be transmitted from production animals to humans [
13]. MRSA ST398, the most prevalent lineage in Europe [
14], has been increasingly isolated from meat and dairy products [
15,
16,
17].
The incidence of S. aureus in the food processing environment and the characterization of the isolated strains will provide useful information in the control of SFP and contribute to improve strategies to eliminate the pathogen. The aim of this work was to investigate the prevalence of S. aureus from the environment and different products in six dry-cured meat processing facilities. Further, toxigenicity and antimicrobial resistance of isolated S. aureus was examined in order to evaluate the potential risk associated with the presence of this pathogen.
4. Discussion
S. aureus is frequently isolated from meat processing facilities, from contact and non-contact surfaces and from raw materials and different product categories. Contamination of meat products results from poor hygienic practices during processing and storage. In this study, overall prevalence of S. aureus in the environment and different categories of products from five out of six dry-cured meats processing facilities was low (4.7%). Prevalence was higher during processing than after cleaning and disinfection, although the pathogen was detected in clean surfaces (1.4%). S. aureus contamination was higher in batter and casings (13.4%) than in equipment surfaces, whereas the pathogen was not detected in final products.
Our prevalence results are in concordance with previous studies conducted in Spain, with a 3.2% in disinfected surfaces from different meat industries [
21]. Higher incidence was observed in a cutting room with coagulase-positive
S. aureus in 15.5% of equipment samples during cutting operation, and 31.8% of meat samples for dry-cured sausages [
22]. Incidence reported by Gounadaki et al. [
23] in food contact and non-contact surfaces of three of seven processing plants producing traditional fermented and/or dry sausages was 11.7%, whereas the pathogen was not detected in batter or final products. According to several authors [
24,
25], raw materials or ingredients are one of the main sources of
S. aureus contamination in meat processing plants, data in agreement with our results on higher contamination in batters and casings. In our work, MRSA presence was relatively high (13.2%), although the pathogen was detected only in one of the industries investigated.
In general, MRSA contamination in meat is lower in Europe (3.2%) compared to other continents [
26]. MRSA was detected in a 5% of RTE food samples positive for
S. aureus [
16], a value higher that the 1.3% in retail foods reported by Yang et al. [
27]. Average incidence of
S. aureus in retail meat including pork in China was 35% [
17], similar to pork products in Spain (33.6%), with a high rate of MRSA found in 21.8% of samples, mainly in meat products with skin (ears and snout) [
13].
The
S. aureus isolates from the processing facilities investigated presented, in general, a high diversity of genotypes by PFGE. A large diversity has been already reported for
S. aureus from the environment and food [
28], and from clinical isolates [
29]. In general, an association of pulsotypes (or clusters of pulsotypes) with the production plant was observed. For example, PT8C and PT9C with more than 90% similarity were isolated only in plant C.
MLST allowed the definition of 11 different sequence types (STs). As shown in
Figure 2, the most abundant was ST30 (CC30) (32% of isolates), followed by ST12 (CC12) (24%). Most of the STs detected in our study have been previously characterized in in pork products from other countries: ST398 (35%), ST1, ST30, ST45, ST15 and ST9 in Denmark [
30], and ST1 was predominant in samples from pork meat in US [
31]. CC45 and CC1 predominated among MSSA isolates from pork meat samples [
13]. On the other hand, MRSA ST398 is the major sequence type colonizing pigs in Europe [
32], and predominated in pork meat samples in Spain [
13]. In the present work, MRSA ST398 was not the dominant ST, but it was the majority in facility A. Different studies have found that
S. aureus CC30 is predominant among human nasal carriers of the bacterium [
33], and in cases of bacteremia in Denmark [
30]. Although it has also been isolated from RTE food [
4], it is not one of the major clonal complexes of porcine origin. We cannot rule out a human origin of those isolates.
The combination of PFGE and MLST revealed association between PTs and CCs with processing plants. Thus, PT1 belonged to CC45, only in plant C, PT10 with CC15 in plant B, PT12 with CC30 and plant B, among others. Taking together PFGE and MLST, we might ascertain that, in general, 1-3 clones are isolated at each plant.
Regarding the presence of enterotoxin genes, in the present work, 55% of
S. aureus isolates carried one or more SE genes. Higher percentages (66%) were reported from industry surfaces [
21], different food products (69%) [
34], and from fermented pork sausages (60%) [
35]. According to our results, SE genes were not detected in all ST398 and ST12 strains. SEA is the enterotoxin most frequently involved in SFP cases [
11], while a lower number of cases are attributed to SEB, SEC, and SED. In our study, all the strains harboring classical SE genes (
sea,
seb and
sec) effectively produced the corresponding toxin (SEA, SEB and SEC) as detected by agglutination test. This data confirms the virulence of these strains. Among the classic enterotoxins, SEB and SEE have been associated with infectious strains of bovine origin, mainly of ST188 [
36,
37]. Therefore it can be expected a low prevalence of those toxins amongst our isolates. In fact, only one ST1 SEB-producing isolate was characterized. Moreover, SED and SEE toxins or their genes were not detected.
The enterotoxin gene cluster
egc (
seg,
sei,
sem,
sen,
seo,
seu/
sew) was found in 37% of the isolates from CC30, CC45 and one ST9 isolate. This percentage is higher than previously reported from different sources, as 14.1% (food strains) [
38] or 18.7% (RTE foods) [
39]. However, higher percentages (50-70%) of
egc in samples from healthy human carriers has been previously reported [
40]. According to our results, the
egc cluster was, in most cases, associated with classical SE genes,
sea or
sec, as previously observed [
41,
42]. Genes from
egc as
sei or
seg have been linked to outbreaks [
41]. Moreover, Schwendimann et al. [
4] demonstrated that 75% egc positive strains expressed SEG and 100% SEI, indicating that these
egc enterotoxins are involved in SFP. Therefore, these isolates are potentially pathogenic. Dicks et al. [
20] found association between the
egc type and clonal complexes of
S. aureus. Thus, OMIWNG variant was present in CC1, CC5 and CC22. In our study, this variant was found in ST9, which belongs to CC1. We also found OMIWNG variant in ST45 (CC45), not described in the mentioned paper. The investigation of a higher number of strains worldwide could ascertain the relationship between
egc variants and clonal complexes of
S. aureus. Recombination between genes inside
egc have been described (i.e.
sel33 is a recombination between
sew and
sen [
20]. Also, incomplete egc variants lacking any of the genes have been found. Thus, Song et al. [
37] reported a 39.5% of
egc strains lacking
seu. In another study, most of the isolates of swine origin possessed an incomplete
egc, lacking two of the genes [
43]. Although our isolates possessed a complete egc, the absence of genes in egc regarding the pathogenicity of the strains needs to be further investigated.
Variations in the amino acid sequence of seo, sei, sem, sen and seg might constitute new genes. For example, a variant of seu firstly named seu2 is now considered sew.
The
seh encoding SEH toxin was found in three (7.8%) CC1 isolates from three different industries. The
seh gene seems to be restricted to CC1 isolates [
29], and has been found in
S. aureus from RTE foods [
39,
44]. SEH has been reported as cause of SFP cases, highlighting the importance of the detection of
seh gene in foodborne outbreaks.
The toxin shock staphylococcal toxin gene
tst, was found at high prevalence in the present study (63.2%). Xie et al. [
29] observed the presence of
tst gene in 48% clinical isolates from China, and by Argudín et al. [
8] in 25.8% of isolates from food and food handlers in Spain. On the contrary, lower prevalence rates have been reported, with values of 2.1% [
39], 7.2% [
44] or 17% [
16]. In another study, the detection rate of
tst was high in MRSA ST9 strains from swine and human clinical isolates [
45]. In our study, the presence of
tst gene was detected in almost all CCs, not restricted to any specific ST. This high proportion of isolates expressing
tst will need to be further investigated.
Concerning Panton-Valentine leukocidin, with the published primers [
46], a length compatible amplicon (180 bp) was not obtained. However, the inespecific amplicons were sequenced, and a sequence compatible with
lukED was observed in five ST12 isolates. This is probably a consequence of the similarity of the different leucocidin genes. In contrast, previous studies conducted in Spain have found high proportions of PVL, both in clinical and food isolates [
8,
47].
The intercellular gene cluster adhesion (
ica) operon is one of the main factors involved in biofilm production by
S. aureus [
48]. Biofilm formation is a well-known mechanism for survival to disinfectants in the food industry [
49]. The
ica operon plays an important role in biofilm formation, specially through the exposition to NaCl [
50], used as ingredient and preservative in dry-cured meat products. High salt concentrations might select the isolates with presence and activity of
ica operon.
Notably, all MRSA ST398 isolates were negative for all tested enterotoxin,
tst and
pvl genes, in agreement with other MRSA results from slaughter pigs [
51]. In the scientific literature, the detection of toxin encoding genes in MRSA CC398 is low, although they have been found colonizing or causing infections in humans [
52,
53,
54].
The ability of
S. aureus to acquire and develop resistance to multiple antibiotics that can be transmitted to humans by ingestion of contaminated food products is recognized worldwide. In the present study, the percentage of
S. aureus resistant to antibiotics was high (71%), although higher percentages in meat have been recorded [
17]. Values of resistance to three or more antibiotics (15.8%) were similar to 16.7% reported for
S. aureus from RTE foods [
40]. All ST12, ST5554, ST7 and ST433 isolates were susceptible to all tested antibiotics. Higher values were reported by Gutierrez et al. [
21], with 70% of strains from food industry surfaces susceptible to 10 antibiotics tested. In our study, resistance to penicillin was observed in 68% of the isolates. Similar percentages have been reported in
S. aureus from food or associated to food poisoning [
34,
42]. The presence of MDR strains is common among
S. aureus isolates from meat and poultry samples [
13]. Multiresistance to several classes of antimicrobial agents is also common in MRSA ST398 isolates and has been reported worldwide [
55,
56,
57]. In our study, MRSA ST398 strains were resistant to two or more antibiotics, and only MRSA isolates were resistant to cefoxitin and ciprofloxacin. MRSA isolates from the present work showed also resistance to tetracycline, a common trait in
S. aureus of animal origin [
58]. This resistance seems to be acquired by livestock associated (LA)-MRSA CC398 after the introduction in livestock from human MSSA [
59]. Presence of MDR is a matter of concern for the food industry, although in our study they were not detected in clean surfaces or final products.
In the present work, the pathogen was detected in a low number of samples after cleaning and disinfection. Our results showed a high variability in the environment, but in general, the cleaning and disinfection procedures were efficient. The highest contamination was recorded on meat batters that could contaminate surfaces during processing. Some points are critical for S. aureus presence and this knowledge is important for the improvement of hygiene control procedures. The presence of S. aureus with virulence determinants and resistance to antimicrobials represent a potential health hazard for consumers. In addition, multidrug resistant MRSA ST398 strains increases the risk for the spread of this pathogen.