Enterotoxin- and Antibiotic-Resistance-Encoding Genes Are Present in Both Coagulase-Positive and Coagulase-Negative Foodborne Staphylococcus Strains

Food poisoning by staphylococcal enterotoxins (SE) is a major cause of foodborne illness, often associated with coagulase-positive staphylococci (CPS). The increase in the number of methicillin-resistant Staphylococcus aureus (MRSA) strains is another major problem associated with CPS. However, reports of the association of SE and methicillin-resistant Staphylococcus with coagulasenegative staphylococci (CNS) are beginning to re-emerge. In this context, the aim of this study is to investigate the presence of staphylococcal enterotoxin genes and to characterize the phenotypic and genotypic antimicrobial resistance in 66 isolates of Staphylococcus spp. (47 CNS and 19 CPS) recovered from ready-to-eat (RTE) street food sold in Maputo, Mozambique. Seven virulence genes encoding SE (sea, seb, sec, sed and see) and two toxins (hlb and sak) were screened by multiplex PCR (MPCR). Antimicrobial resistance against 12 antibiotics was evaluated by the disk diffusion method. The presence of genes encoding resistance to penicillin, methicillin, vancomycin and erythromycin (blaZ, mecA, vancA, vancB, ermA, ermB and ermC) were also screened by PCR. At least one of the seven virulence genes assessed in this study was detected in 57.9% and 51% of CPS and CNS isolates, respectively. In CPS isolates, the most frequent gene was hlb (47.4%), followed by sec (15.8%) and sea, seb and sed genes with 5.3% each. In CNS isolates, the most frequent gene was sec (36.2%) followed by sak (17%), hlb (14.9%), sed (12.8%) and seb (6.4%). Five of the twelve CPS in which virulence genes were detected were also antibiotic-resistant. All the CNS isolates harboring virulence genes (n = 27, 57.4%) were antimicrobial-resistant. The prevalence of multidrug resistance was higher (59.6%) in CNS than in CPS (26.3%) isolates. Regarding the presence of antibiotic-resistance genes, blaZ (penicillin-resistant) was the most frequent in both CPS (42.1%) and CNS (87.2%), followed by the mecA (encoding methicillin resistance) and vancA genes (vancomycin-resistant), which represented 36.8% and 31.6% in CPS isolates and 46.8% in CNS isolates, respectively. The prevalence of vancomycin-resistant staphylococci has been increasing worldwide and, to our knowledge, this is the first study to report the occurrence of vancomycin-resistant staphylococci in Mozambique. These results emphasize the need to investigate CNS isolates in parallel with CPS, as both constitute public health hazards, given their potential to produce SE and spread antimicrobial resistance genes.


Introduction
Staphylococcus species are frequently found on the human skin and nasal mucosa, but can also be found in soil, water and food products, among others [1][2][3][4]. The contamination of the food matrix by Staphylococcus spp. can occur via contaminated raw materials, such as meat and non-pasteurized milk. Additionally, cross-contamination from utensils and equipment as well as from other contaminated products, particularly those subjected to handling [5,6], are also relevant transmission routes. The pathogenicity of S. aureus is related to the presence of several virulence factors, including S. aureus enterotoxins (SE), hemolysins encoded by hla and hlb genes and staphylokinase encoded by the sak gene [6][7][8][9]. Hemolysins are involved in the tissue adhesion of the pathogens, colonization and tissue invasiveness, thus promoting pathogenicity [6,7]. Staphylokinase is a bacteriophage-encoded protein expressed by lysogenic strains of S. aureus [10]. It is a protein participating in the disintegration of fibrin and considered one of the S. aureus virulence factors [10,11].
Food poisoning by Staphylococcal enterotoxin is one of the most prevalent foodborne illnesses, often associated with coagulase-positive Staphylococcus (CPS) [2]. Currently, more than 23 staphylococcal or staphylococcal-related enterotoxins are known, and their ingestion in amounts ranging from 20 ng to 1 mg can cause severe symptoms in humans [6,12]. These proteins are resistant to proteolysis and are thermostable [13][14][15]. Although the precise mechanisms of action of SE are still unclear, it is known that foodborne SE directly affects the intestinal epithelium and the vagus nerve (pneumogastric nerve), causing the stimulation of the emetic center [16]. Classical enterotoxins A-E have been associated with more than 90% of staphylococcal food poisoning cases worldwide [6] and often linked with CPS [17,18]. However, reports of the association of enterotoxins from food poisoning with coagulase-negative staphylococcus (CNS) are beginning to re-emerge [17,[19][20][21][22].
Antibiotic resistance, particularly methicillin-resistant Staphylococcus aureus (MRSA), is another problem associated with staphylococci [23,24]. In recent decades, the incidence of MRSA strains has increased worldwide [12], which has compromised the treatment of nosocomial infections, especially in low-income countries [25]. In 2019, an estimated 1.2 million deaths were directly related to antibiotic-resistant strains, and S. aureus was the second most important attributable or associated antibiotic-resistant species [26]. However, antibiotic resistance has also been reported in coagulase-negative staphylococcal strains [27,28]. Furthermore, several studies showed that CNS may be reservoirs for the mecA genes encoding penicillin-binding protein (PBP2a) [27][28][29][30], thus emphasizing the importance of CNS in public health.
When associated with poor hygiene practices by handlers and poor hygienic conditions of the exposure environment, ready-to-eat (RTE) street food presents a potential risk of transmission of staphylococcal strains [31][32][33][34][35][36]. In Maputo, Mozambique, street foods are sold in environments with poor sanitation conditions and lack of potable water, which increases the risk of street food contamination [33,34,37,38]. This study is a follow-up to previous research by   [31], and aims to evaluate the antimicrobial resistance profile and the occurrence of both virulence-and antibiotic-resistance genes in 66 coagulase-positive and coagulase-negative Staphylococcus isolates previously recovered from RTE foods sold on the streets of Maputo, Mozambique.

Material and Methods
This study used 70 isolates of presumptive Staphylococcus previously recovered from 81 RTE street food sold in Maputo, described in   [31]. In general, all the food samples were considered unsatisfactory for consumption [31]. Staphylococcus spp. presumptive colonies were obtained after incubation for 48 h at 37 • C onto Chromogenic Baird Parker Agar-CBPA (Biokar Diagnostics, Beauvais, France) according to ISO 6888 [39], as described in   [31]. From each plate, one CFU that presented typical characteristics of Staphylococcus was selected.

Species Identification and Phylogenetic Analysis
The sequenced 16S rRNA amplicons were submitted to blastn using megablast algorithm against Reference RNA sequences database (RefSeq RNA) from the National Center of Biotechnology Information (NCBI) for identification. The reference 16S rRNA sequences of the twelve species identified by BLAST search and the outgroup (Macrococcus caseolyticus) were retrieved from NCBI RefSeq RNA database (Table S1). The multiple alignment of the reference 16S rRNA sequences and those of the 70 presumptive Staphylococcus isolates from this study was performed in MEGA X version 10.2.6 (PA, USA) using MUSCLE algorithm (v10.0.5). The alignment was truncated to achieve a core alignment of 1 kb. For phylogenetic analysis, the core alignment was used to construct a maximum-likelihood tree with 100 replicates, using General Time Reversible substitution model and Gamma distributed with invariant sites rate, in MEGA X (v10.0.5). Tree editing and annotation was performed using iTol (v6.5.4) [41].

PCR for the Identification of Virulence and Antibiotic Resistance Genes
The presence of seven genes encoding toxin hemolysin (hlb), staphylokinase (sak), and five enterotoxins (sea, seb, sec, sed, and see) was investigated. For DNA extraction, six bacterial colonies grown on TSA plates for 18 ± 2 h at 37 • C were suspended in 300 µL of sterile ultrapure water and incubated in a boiling bath for 10 min [42]. Subsequently, the tubes were centrifuged at 16,000× g for 12 min, and the lysate supernatants (lysates) were stored at −20 • C. The master mix for MPCR was prepared with 12.5 µL of Taq DNA Polymerase NZYTaq II2x Colourless Master Mix (MZYTech, Lisboa, Portugal), 1 µL of each forward and reverse primer indicated in Table 1 (final concentration 0.3 µM), 2 µL of template DNA and sterile ultrapure water to fill 25 µL of total volume. Based on the differences in the hybridization temperatures, thermocycling reactions were performed in two different MPCRs, one for the hlb, sea and see genes and the other for the sak, seb, sec and sed genes, according to Adame-Gómez et al. (2020) [2]. The screening for the antibiotic-resistance genes used seven primer pairs for the detection of genes encoding resistance to four antimicrobials ( Table 2). The master mix for MPCR was prepared with 12.5 µL of Taq DNA Polymerase NZYTaq II2x Colourless Master Mix (MZTech, Lisboa, Portugal), 1 µL of each forward and reverse primer indicated in Table 2 (final concentration 0.3 µM), 2 µL of DNA template and sterile ultrapure water to complete 25 µL of the total volume. For mecA detection, the reaction mixtures were subjected to the following amplification conditions: initial denaturation, 94 • C for 5 min, followed by 30 cycles of denaturation 94 • C for 30 s, annealing 52 • C for 30 s, elongation 72 • C for 30 s, and a final elongation at 72 • C for 5 min. The detection of blaZ was performed with an initial denaturation at 94 • C for 5 min, followed by 30 cycles of denaturation 94 • C for 20 s, annealing 60 • C for 30 s, elongation 72 • C for 90 s, and a final elongation at 72 • C for 5 min. For the detection of erythromycin-(ermA, ermB and ermC) and vancomycin (vancA and vancB)-resistance genes, monoplex PCRs were performed with the pairs of primers indicated in Table 2, as described by Prunier et al. (2003) [44]. For the detection of vancomycin (vancB)-resistance genes, monoplex PCRs were performed as described by Al-Amery [45] (Table 2). Table 2. Primers used for the detection of antimicrobial-resistance genes and the respective amplicon product sizes.

Primer for the Target Gene Primer Sequence (5 -3 ) Amplicon Size (bp) References
All PCR reactions were run in a thermocycler GeneAmp ® PCR System 9700, Applied Biosystems (Bio-Rad Laboratories, Segrate, Milan, Italy). The resulting PCR products were resolved on 2% (m/v) agarose gels in 1× TAE buffer, in a EC330 Thermo Fisher Scientific tank (Atlanta, GA, USA) at 8 V/cm for 60 min. The gels were stained with GelRed (Frilabo, Maia, Portugal) and analyzed using a Gel Doc™ EZ System (Bio-Rad Laboratories, Segrate, Milan, Italy). For calculating the size of the PCR products, the molecular marker 100 bp DNA Ladder (Invitrogen, CA, USA) was used.

Data Interpretation
For the evaluation of the antimicrobial resistance profile, the inhibition halos were measured (millimeter) and compared to those described in the CLSI (2021) [50]. Isolates were considered non-susceptible to a given antibiotic when they showed intermediate or full resistance according to the CLSI clinical breakpoints. Multidrug resistance was considered as non-susceptibility to at least one agent in three or more antimicrobial categories and/or resistance to methicillin [51].

Results
The 70 presumptive staphylococcal isolates used in this study were identified based on 16S rRNA gene sequence analysis (Figure 1). Staphylococcus aureus represent 27% of the 70 isolates under study and correspond to all CPS isolates. The most representative CNS species are S. warneri (22.9%), S. saprophyticus (18.6%), S. xylosus (10%) and S. pasteuri (5.6%). Four isolates initially classified by biochemical methods as belonging to the Staphylococcus genus were identified as belonging to the species Mammaliicoccus sciuri (Figure 1). These four isolates did not show in vitro methicillin or erythromycin resistance. Two of the three Mammaliicoccus sciuri isolates presented virulence genes and one was resistant to vancomycin and to penicillin G.

Detection of Virulence Genes
Virulence genes were detected among the 39 identified Staphylococcus isolates (Figure 1). CPS (63.2%) and CNS (57.4%) showed at least one of the seven genes assessed in this study (Figure 1). The hlb gene was detected in 47.4% of the CPS and 14.9% of the CNS isolates. In CPS, hlb was the most frequent virulence gene, followed by sec 15.8% (Table 3). In the CNS, sec was the most frequent gene (36.2%), followed by sak (17%) and hlb (14.9%) ( Table 3).   In the CPS, except for the seb gene that did not co-occur with the other screened genes, all genes co-occurred with hlb (hlb and sea in one isolate; hlb and sec, and hlb and sed in two isolates, each) ( Figure 1). In the CNS, the coexistence of hlb and seb, and hlb and sec was detected in two isolates, respectively: sed and sak in one isolate, and the co-occurrence of seb, sec and sed genes in one isolate ( Figure 1).
The see gene was not detected in any of the isolates. The sak gene was detected exclusively in CNS isolates (8/47) ( Table 3).
Twenty-eight (59.6%) CNS isolates and five (26.3%) CPS isolates were multidrugresistant (Table S2). On the other hand, 6.4% of the CNS isolates and 21.1% of the CPS isolates showed no resistance to any of the antibiotics tested.

Antimicrobial-Resistance Genes
Among the 66 CPS and CNS isolates, the blaZ gene encoding penicillin resistance was the most frequently detected resistance gene (74.2%). However, among the CNS group, this percentage (87.2%) almost doubled the corresponding percentage in the CPS isolates (42.1%) ( Table 5). The mecA gene encoding methicillin resistance was the second most frequently detected (43.9%) in all isolates (Table 5). This percentage was also higher in CNS isolates (46.8%) than in CPS (36.8%) ( Table 5). The mecA gene was detected in all cefoxitin-resistant isolates and in 6.1% of non-cefoxitin-resistant isolates. Likewise, the co-occurrence of mecA and blaZ was detected in CPS (26.3%) and CNS (29.4%) isolates. To investigate the presence of vancomycin-resistance genes, vancA and vancB were used ( Table 5). The results obtained showed that 42% of the isolates showed the vancA gene (26.3% among CPS and 46.8% among CNS). However, the vancB gene was detected only in the CNS (14.9%) ( Table 5). In this group, one of the isolates that tested positive for vancB also tested positive for vancA. Three target genes were used to search for erythromycin resistance, ermA, ermB and ermC. The presence of erythromycin resistance genes was detected in 42.1% of the CPS isolates and in 18.2% of the CNS isolates. The ermC gene was the most frequent in CPS isolates (26.3%), while the ermB gene was the most frequent in CNS isolates (14.9%) ( Table 5). Four CPS isolates (15.8%) showed the presence of ermA (Table 5).

Discussion
Staphylococcal enterotoxins are one of the most important causes of food poisoning. For this reason, estimating the prevalence of their coding genes in Staphylococcus isolates recovered from food enables risk assessment for food poisoning [52]. In this study, 32 (48.5%) of the 66 Staphylococcus spp. isolates analyzed harbored genes encoding enterotoxins (Figure 1). In fact, the occurrence of these genes was even similar in the CNS (33.3%) and in the CPS (31.6%). This high prevalence of SE in the CNS is very relevant and contrasts with several studies that pointed out CPS as the main source of SE in food [53]. Although the presence of enterotoxins in CNS strains has been described by other authors [53,54], few studies have been conducted to assess the presence of these enterotoxins in foodborne CNS. In one of these few studies, Moura et al. (2012) [53] reported that the occurrence of SEs in black pudding isolates was even higher in the CNS (67.5%) compared to CPS isolates (32.5%). Previous studies list the staphylococcal enterotoxins sea, seb and sec as the most frequent in foodborne outbreaks [54]. Sec and sed are often associated with contamination by animals, and sea and seb are associated with human contamination, through food handlers [35]. The higher prevalence of the sec gene (15.8% in CPS and 36.2% in CNS) in this study, followed by sed (12.8% in the CNS), suggests that the main source of contamination of these RTE food was of animal origin. In fact, these RTE foods are prepared and/or sold in Maputo markets where fish and raw meat are simultaneously sold, as well as living animals, such as chickens and goats [32,33]. So, taking all results together, we may speculate that these secand sed-positive Staphylococcus spp. isolates are likely associated with cross-contamination, either indirectly through mechanical vectors or via the food matrix of animal origin [31].
The hlb gene encoding hemolysin was detected in nine CPS isolates (47.4%) and in seven (14.9%) CNS isolates. In CPS, 50% of the isolates that tested positive for hlb also tested positive for some SE-encoding genes. In the CNS, this percentage was 75%. Isolates simultaneously harboring SE-and hemolysin-encoding genes can potentially be considered more pathogenic than those positive for the SE genes but negative for hlb, since hemolysin holds a transversal role in pathogenicity, immune system evasion, and often in the sequestration of nutrients transported by erythrocytes [55].
Five of the twelve CPS that presented virulence genes were also resistant to antibiotics. Of these five, two were multidrug-resistant. In CNS, all the 57.4% isolates with virulence genes were antibiotic-resistant. In Mozambique, as in many other sub-Saharan African countries, there is lack of antibiotic-resistance data for Staphylococcus originating from food [56,57]. The results obtained in the present study show a high prevalence of resistance to β-lactams, mainly to penicillin (83.3%), methicillin (37.9%) and vancomycin (27.7%). Similar results regarding penicillin resistance (76%) were obtained for S. aureus isolates from raw milk in Maputo province [58]. Another study also carried out in Maputo [56] reported 89% of penicillin resistance among S. aureus from hospital origin. These authors suggested a selective pressure towards the emergence of resistance due to the overuse of this antibiotic. In Poland, Pyzik et al. (2019) [21] found a high resistance to penicillin G (54.3%), amoxicillin (49.6) and methicillin (16.5%) in CNS isolated from poultry. In Greece, Sergelidis et al. (2014) [59] evaluated methicillin resistance in Staphylococcus spp. recovered from ready-to-eat fish products and found a high resistance to penicillin and ampicillin (85.7%), erythromycin and clindamycin (71.4%), amoxicillin/clavulanic acid, gentamicin and tobramycin (57.1%) and oxacillin (28.5%) in S. aureus.
Among the penicillin-resistant isolates (83.3%), the blaZ gene was detected in 8 CPS and 41 CNS. A low percentage of methicillin resistance was found in CPS isolates (26.3%). Nhatsave et al. (2021) [58] did not find methicillin resistance in S. aureus isolates from raw milk in Maputo. Similar results have been reported in several southern African countries [60][61][62]. In South Africa, Govender et al. (2019) [63] reported 21% methicillin resistance in S. aureus isolated from poultry meat products, and Oosthuysen et al. (2014) [60] reported 15.3% of MRSA among S. aureus isolates of hospital origin. In Tanzania [61] and Uganda [62], a low prevalence of MRSA (5.4-5.7%) in hospital-origin S. aureus was also reported. In contrast, a high frequency of penicillin (88.2%) and methicillin (57.9%) resistance was found in CNS isolates. These results corroborate other studies that consider the CNS as a reservoir of antimicrobial-resistance genes [64][65][66][67]. In South Africa, Asante et al. [67] also reported a high occurrence of resistance to penicillin (100%) and methicillin (76.4%) in CNS isolates. In Greece, Sergelidis et al. (2014) [59] reported high prevalence of methicillin resistance (41.7%) in CNS isolated from RTE foods. In Poland, 27.6% of CNS isolated from poultry were methicillin-resistant [21]. These antibiotic-resistant CNS strains may constitute a major cause of infectious diseases, because of their potential ability to form biofilms and thus colonize community or hospital settings [64]. In addition to penicillin and methicillin resistance, a high percentage of vancomycin resistance was observed in 26.3% of the CPS isolates and in 21.3% of the CNS isolates. Among these vancomycin-resistant isolates, 33.3% of CPS and 100% of CNS were also methicillin-resistant. This vancomycin-and methicillinresistance profile has also been reported in South Africa [67], where 46.2% of CNS isolates that were vancomycin-resistant were also methicillin-resistant. All CPS isolates exhibiting vancomycin resistance were also positive for the vancA gene. However, the percentage of CNS isolates with vancA was more than double the percentage of vancomycin-resistant isolates, suggesting the existence of non-expressed resistance genes (genotypic resistance).
While in Egypt, Al-Amery [45] reported 27% to 54% vancomycin resistance in CPS isolates from camel meat and slaughterhouse workers' hands, to our knowledge, in Mozambique, this is the first study to report the occurrence of vancomycin-resistant Staphylococci strains. According to Wu et al. [68], the prevalence of vancomycin-resistant staphylococcal strains has been increasing worldwide, from 2% in 2005 to a prevalence of 7% in 2020. In Africa, an estimated 16% of S. aureus isolates are vancomycin-resistant [68].
A high percentage of multidrug resistance was detected in this study (26.3% in CPS and 59.6% in CNS). In CPS, apart from the macrolide ERY, with a high percentage of resistant isolates, the relatively low-multidrug-resistance phenotype can be explained by the high susceptibility to non-ß-lactam antibiotics. Different results were reported by Mandomando et al. [69] [67] may be influenced by the clinical origin of the isolates. Normally, isolates of hospital origin are more resistant than those of environmental origin due to regular exposure to various antibiotics [70,71].

Conclusions
The isolates analyzed in this study were recovered from street foods with a high burden of Staphylococcus (potentially hazardous). The sec and sed coding genes were the most frequent virulence genes. Since sec and sed are associated with contamination of animal origin, these results highlight the need to control products of animal origin, especially at the level of pre-and post-slaughter inspection of animals. In both CPS and CNS isolates, a high percentage of penicillin resistance was observed. However, the CNS isolates showed an even higher percentage of methicillin resistance than the CPS. These results highlight the importance of CNS in parallel with CPS isolates, as both can be potentially dangerous when it comes to the possibility of causing food poisoning. In addition, CNS isolates can potentially spread antibiotic-resistance genes. To the best of our knowledge, this is the first report of vancomycin-resistant staphylococcal strains in Mozambique. These data are extremely important given the current increase in cases of vancomycin resistance, worldwide and in Africa in particular.

Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.

Data Availability Statement:
The authors confirm that the data supporting the findings of this study are available within the article and its Supplementary Information.