Antimicrobial Resistance, Enterotoxin and mec Gene Profiles of Staphylococcus aureus Associated with Beef-Based Protein Sources from KwaZulu-Natal Province, South Africa

Annually, approximately 23,000 cases of food poisoning by Staphylococcus aureus enterotoxins are reported worldwide. The aim of this study was to determine the occurrence and characterize S. aureus on beef and beef products in South Africa. Organ meats (n = 169), raw processed meat (n = 110), raw intact (n = 53), and ready-to-eat meats (n = 68) were obtained from 25 retail outlets. S. aureus was isolated and enumerated according to the ISO 6888-1 method. Identification of the strains was performed by MALDI-TOF MS. The antimicrobial resistance was determined using the disc diffusion test. The presence of methicillin-resistance genes and the staphylococcal enterotoxin genes was determined by PCR. Prevalence was low (13/400; CI 1.7–5) and all but one positive sample were from organ meats. Eight isolates were resistant to at least one antibiotic. Two isolates carried the mecC gene. All the isolates tested positive for seg, seh, sei, and sep, whilst 53.8% were positive for sea. None of the isolates was positive for ser, sej, seb, sec, or sed. The prevalence of S. aureus was low, with organ meats being the most contaminated. The presence of mecC-positive MRSA and of enterotoxins warrants further investigation and risk assessment.


Introduction
Beef is known for its role in supplying protein, minerals, and vitamins in human nutrition [1]. Due to its high nutritional content, beef is an excellent substrate for the growth of microorganisms, of which some are leading causes of meat spoilage [2]. Spoilage of meat is enhanced by inadequately stored or packed meat [3]. Different storage conditions, such as cold and gaseous composition, on packed meat are most likely to suppress the microflora, among them S. aureus [4].
S. aureus causes staphylococcal food poisoning (SFP) through the ingestion of food contaminated with staphylococcal enterotoxins [5]. This enterotoxaemia is characterized by diarrhea, nausea, abdominal cramping, and vomiting within 24 h of eating [6]. Contamination of food by S. aureus may originate from the animal, the food handlers, and the environment. It may be a consequence of poor hygiene during processing from slaughter to final product or inappropriate storage and household manipulations; however, contamination of meat is a complicated process which may occur well before the meat reaches retail to final product or inappropriate storage and household manipulations; however, contamination of meat is a complicated process which may occur well before the meat reaches retail outlets [7]. In addition to toxins encoded by the seb, sec, sed, and see genes, in particular, strains that produce the Staphylococcal Enterotoxin A (SEA), encoded by the sea gene, have caused a large number of outbreaks [8,9].
Antimicrobial resistance (AMR) is an increasing global challenge mainly driven by the overall use of antimicrobials [10]. In certain S. aureus clones, AMR is a major problem, especially in methicillin-resistant S. aureus (MRSA), of which the prevalence increases globally [11]. In the last 15 years, MRSA clonal complex 398 was discovered in food-producing animals, while other sequence types (ST), such as ST1, ST5, ST9, ST97, ST130, and ST433, have been reported to a lesser extent [12]. These strains were subsequently named Livestock Associated (LA)-MRSA. LA-MRSA Clonal complex 398 (CC 398) is mainly prevalent in Europe and North America; however, it has been reported in Asia as well as in Africa [12]. To a lesser extent, LA-MRSA CC398 has been associated with different infections in humans, including skin and soft tissue illnesses, ventilator-associated pneumonia, and septicemia [13].
There are few studies that have been conducted on meat and meat products in Africa so far [14,15]; most studies did not type the isolates, and only few of the studies have identified AMR genes [14][15][16][17]. Therefore, this study aimed to determine the occurrence, AMR, and virulence genes of S. aureus isolated from beef and beef products in retail outlets of the KwaZulu-Natal province, South Africa.

Ethical Approval
Ethical approval for this study was obtained from the University of Zululand with certificate number UZREC 171110-030 PMG 2019/112.

Study Design
The cross-sectional study involved the collection and microbiological analysis of meat and meat products from retail outlets and butcheries from King Cetshwayo and iLembe districts in the KwaZulu-Natal province of South Africa. King Cetshwayo district covers an area of 8213 km 2 from the north coast region, whereas iLembe covers an area of 3269 km 2 from the south coast region with a population of approximately 885,944 and 606,809, respectively [18]. The two districts contribute to about a quarter of the total population in KwaZulu-Natal KZN (Figures 1 and 2).  //municipalities.co.za/map/124/king-cetshwayo-district-municipality (accessed on 1 November 2020) [19].

Sample Size Determination
There are important statistical variables to consider when determining the sample size for a surveillance study [20]. These include z, α, p, and d, where z (1.96) is the normal deviate for two-tailed alternative hypotheses [21]. Alpha (α) is the level of significance and it is usually 5%, which implies that having a 5% probability of incorrectly rejecting a null hypothesis is acceptable [22]. The p-value is the expected prevalence proportion, and a prevalence of 50% (0.50) was assumed in this study based on a national surveillance of foodborne pathogens in South Africa by [23], which detected about 56% S. aureus in diverse meat and meat products from various establishments. The value of d is precision, and at the confidence interval of 95%, d is 0.05. In this study, the following formula was used to calculate the sample size of the surveillance study: . ( . ) However, 400 samples were collected in this study for robust results.

Sample Collection
A total of 400 samples were collected during the cross-sectional study. Twenty-five retail outlets and butcheries from King Cetshwayo and iLembe districts were included in this study. The beef samples were ready-to-eat beef products (n = 68), raw processed beef (n = 110), raw intact beef (n = 53), and organ meats (n = 169). Samples were packed into sampling bags using strict aseptic techniques, and placed in cooler bags containing ice packs to maintain a temperature of approximately 4 °C. The packaged samples were transported immediately to the Microbiology laboratory at the University of Zululand, Department of Biochemistry and Microbiology, for further bacteriological examination. Samples were analyzed immediately after arrival.

Sample Size Determination
There are important statistical variables to consider when determining the sample size for a surveillance study [20]. These include z, α, p, and d, where z (1.96) is the normal deviate for two-tailed alternative hypotheses [21]. Alpha (α) is the level of significance and it is usually 5%, which implies that having a 5% probability of incorrectly rejecting a null hypothesis is acceptable [22]. The p-value is the expected prevalence proportion, and a prevalence of 50% (0.50) was assumed in this study based on a national surveillance of foodborne pathogens in South Africa by [23], which detected about 56% S. aureus in diverse meat and meat products from various establishments. The value of d is precision, and at the confidence interval of 95%, d is 0.05. In this study, the following formula was used to calculate the sample size of the surveillance study: However, 400 samples were collected in this study for robust results.

Sample Collection
A total of 400 samples were collected during the cross-sectional study. Twenty-five retail outlets and butcheries from King Cetshwayo and iLembe districts were included in this study. The beef samples were ready-to-eat beef products (n = 68), raw processed beef (n = 110), raw intact beef (n = 53), and organ meats (n = 169). Samples were packed into sampling bags using strict aseptic techniques, and placed in cooler bags containing ice packs to maintain a temperature of approximately 4 • C. The packaged samples were transported immediately to the Microbiology laboratory at the University of Zululand, Department of Biochemistry and Microbiology, for further bacteriological examination. Samples were analyzed immediately after arrival.

Microbiological Analysis Control Strains for Quality Control
The S. aureus ATCC 25,923 (Microbiologics, MN, USA) and field strains (positive for tested virulence factors) were included in all laboratory experiments as positive control strains. ATCC 25,922 was used as negative control.

Detection, Enumeration, and Isolation and Identification of S. aureus
The detection, enumeration, and isolation of S. aureus was performed according to the ISO 6888-1:1999 AMD 2018 standard method [24]. Briefly, each sample was analyzed for the presence of S. aureus by weighing 25 g, followed by addition of 225 mL of buffered peptone water. The samples and buffered peptone water were thoroughly mixed in a homogenizer (Bagmixer 400 cc, Interscience, France) for 2 min at 10 stroke/s. Subsequently, ten-fold serial dilutions were made using sterile pipettes [25]. From these dilutions, 0.1 mL was inoculated in duplicate onto Baird Parker agar plates (Oxoid, UK) using the spread plate technique, as described by Goja et al. [25]. Plates were incubated at 37 • C for 24 to 48 h. After incubation, the typical colonies were counted. Typical Staphylococcus spp. appeared as shiny black colonies [26]. To calculate the number of colony-forming units per gram (CFU/g), the colonies on the countable plate were multiplied by final dilution factor. The presumptive colonies were purified three times through sub-culturing on nutrient agar (Oxoid, UK) and incubation at 37 • C for 24 h. The presumptive S. aureus colonies were subjected to Gram staining, catalase test, mannitol salt, and free and bound coagulase tests (Oxoid, UK) [27]. Gram-positive cocci that appeared purple with grape-like shape were catalase-positive, appearing yellow on mannitol salt agar due to mannitol fermentation, and where coagulase-positive, were considered to be presumptive S. aureus and the colonies were subjected to further tests. Identification of S. aureus was confirmed using MALDI-TOF MS, according to the manufactures instructions for the MALDI Biotyper®(Bruker Daltonics, Germany) [28]. All confirmed S. aureus isolates were streaked on 5% sheep blood agar plates to identify the type of hemolysin they produce [29]. The plates were incubated at 37 • C for 24 h ± 2 [30].

Antimicrobial Susceptibility Testing
For the antimicrobial susceptibility test, the Kirby Bauer disk diffusion method according to Clinical Laboratory Standard Institute guidelines was applied [31,32]. Briefly, from a pure bacterial culture, 2-5 colonies were suspended in 5 mL sterile saline solution. The bacterial concentration was adjusted to an optical density of 0.5 on the McFarland scale [31,33]. A sterile cotton swab was dipped into the suspension and excess fluid was removed by squeezing the swab at the top of the bijou bottles. The bacteria were inoculated onto Mueller Hinton agar (Thermofisher, Waltham, MA, USA) by streaking in three different directions to obtain confluent bacterial growth. The medium surface was allowed to dry, followed by placing the following antimicrobial disks: ciprofloxacin (5 µg  The enterotoxin genes (sea, seb, sed, sec, she, seg, ser, sei, sep, sej) and mec genes (mecA and mecC) were assessed by PCR using the primers listed in Table 1. The 20 µL PCR reaction mixtures contained 10-30 ng of template DNA (in 1 µL), NEB one Taq 2× master mix with standard buffer (10 µL), forward primer (1 µL), reverse primer (1 µL), and nuclease Microorganisms 2022, 10, 1211 5 of 13 free water (7 µL). PCR amplifications for sea, seb, sec, sed, and ser were carried out in a thermal cycler with the following thermal conditions: initial denaturation for 5 min at 95 • C; 35 cycles of 30 s at 94 • C, 40 s at 56 • C, and 1 min at 68 • C; final extension for 5 min at 68 • C. The PCR conditions for seg, sei, sep, sej and she were similar to the above, except that the annealing stage was performed at 53 • C for 40 s. 2.6.2. Agarose Gel Electrophoresis PCR products were subjected to electrophoresis in 1.5% agarose gels stained with ethidium bromide at 3 volts/cm for approximately 30 min. The 50 bp and 100 bp DNA ladders were used to estimate the size of PCR amplicons. The PCR amplicons were visualized under ultraviolet light and the gel images were documented using a gel documentation system (E-Box).

Prevalence of S. Aureus in Meat
Out of the 400 beef and beef products that were analyzed, 3.25% (n = 13; CI 1.7-5) tested positive for S. aureus (Table 2). From each of the 13 positive samples, one isolate was retained for further investigation. The S. aureus-positive samples were predominantly organ meats (n = 10/13; CI 46.2-95), followed by raw intact beef (n = 2/13; 1.9-45). Only one of the 13 S. aureus-positive samples was from ready-to-eat beef. No S. aureus was detected in raw processed beef. The 13 S. aureus from 13 positive samples showed alpha, beta, and gamma hemolysis reactions on 5% sheep blood agar. Table 3 shows the results of Staphylococcus aureus enumeration of the 13 positive samples. The S. aureus counts from beef-based products ranged from 2.65 log 10 CFU/g to Microorganisms 2022, 10, 1211 6 of 13 4.1 log 10 CFU/g (Table 4). S. aureus from 1 of the 13 positive samples (ox kidneys) were too numerous to count.

Methicillin-Resistant Determinants
All isolates were tested for the presence of mecA and mecC genes. None of the isolates were positive for mecA genes. Two isolates (15.4%; CI 1.9-45) tested positive for mecC gene (Figure 3).

Methicillin-Resistant Determinants
All isolates were tested for the presence of mecA and mecC genes. None of the isolates were positive for mecA genes. Two isolates (15.4%; CI 1.9-45) tested positive for mecC gene (Figure 3).

S. aureus Enterotoxin Genes
In Table 5 and Figures 4 and 5, the results of the virulence genes are shown. Out of eleven enterotoxin genes that were tested, five (sep, seh, sei, sej, sea) were detected. All S. aureus strains were positive for seg, seh, and sei, sep. The sea gene was detected in 7 of the 13 S. aureus (53.84%).

S. aureus Enterotoxin Genes
In Table 5, Figures 4 and 5, the results of the virulence genes are shown. Out of eleven enterotoxin genes that were tested, five (sep, seh, sei, sej, sea) were detected. All S. aureus strains were positive for seg, seh, and sei, sep. The sea gene was detected in 7 of the 13 S. aureus (53.84%).

Discussion
The aim of this study was to determine the occurrence, AMR, and virulence characteristics of S. aureus from products of bovine origin in retail outlets of selected municipalities. Though most studies in Africa found prevalences of between 15 and 40%, with the exception of 55% S. aureus detection in Algeria, the overall occurrence of 3.25% from this study is lower compared to most of the previous African studies [41][42][43][44][45]. Higher preva-

Discussion
The aim of this study was to determine the occurrence, AMR, and virulence characteristics of S. aureus from products of bovine origin in retail outlets of selected municipalities. Though most studies in Africa found prevalences of between 15 and 40%, with the exception of 55% S. aureus detection in Algeria, the overall occurrence of 3.25% from this study is lower compared to most of the previous African studies [41][42][43][44][45]. Higher prevalences of up to 65% have also been found in non-African countries such as Turkey, Jordan, United States of America (USA), and several countries in Europe [46][47][48][49][50][51][52]. It is important to note that differences in methodology and sample size should be taken into account and may explain, in part, the differences seen [53,54]. It is important to apply a system such as the ISO standard, used in this study, to allow direct comparison of the prevalence between different studies.
The relatively high S. aureus prevalence in organ meats (kidneys, livers, lungs) compared to raw intact beef meat and ready to-eat meat was conspicuous in this study (though not to a significant extent). Probably the organ meat may be more prone to cross contamination compared to other meat types, and S. aureus can be found in the intestines [55]. S. aureus occurrence was also observed in ready to-eat meats and this may be attributed to cross contamination and growth due to further preparation [56,57]. The differences in preparation and the type of preparation of the ready-to-eat (RTE) beef, as well as conservation of the product, may play a large influence.
The average counts of the S. aureus-positive samples in this study ranged from 2.65 log 10 and 4.07 log 10 per gram in organ meat. These S. aureus counts are lower than those that were previously observed for organ beef meat in South Africa (5.1 log-log 5.6) [14].
When considering the S. aureus limit of 100 CFU/g in RTE, proposed by the guidelines for environmental health officers on the interpretation of microbiological analysis data of food [58], the 13 positive samples were not within the compliance limits, though they were sold at retail level. The situation is concerning for RTE biltong, which is not processed further prior to consumption. The contaminated samples were mainly plucked meats that might not be subjected to similar strict hygiene scrutiny as beef cuts. It is possible that the S. aureus counts may have increased in the pluck meats during transportation, probably due to an inadequate cold chain, or the level of preservation at the retail level may have contributed to an increase in bacterial numbers.
As we found only few strains, comparing with other studies is difficult. The isolates from this study were, in general, more susceptible to antimicrobials than those from other studies on beef in South Africa [15] and other African countries [42,44], but are similar to studies in Europe [50], and higher than what has been detected in the United States of America [59].
Interestingly, two isolates were MRSA, with only one detected phenotypically. Phenotypic methicillin resistance should always be confirmed by PCR, as false positive and false negative results may be obtained by the phenotypic tests. While this may not seem a lot, it may have a significant public health impact. The MRSA isolates from this study were mecC-positive. While this resistance gene has not been associated extensively with MRSA either in humans or animals, it has, however, been isolated mainly from animals, including wildlife [60]. In most countries, the mecA gene is mostly found in MRSA [15,[61][62][63][64][65]. However, in South Africa, the mecC gene has also been shown as the sole methicillin-resistance gene in strains from different food-producing animal species as well as wild birds [17]. This might indicate a very specific and unique situation in South Africa and urges for a more large-scale study of MRSA on food-producing animals, wild animals, and foods derived from animals in South Africa so as to determine the human health hazard. These studies should include whole-genome sequencing to determine their true epidemiology.
Staphylococcal enterotoxins (SEs) types SEA to SEE have been reported to account for approximately 95% of food poisoning outbreaks caused by staphylococci [66], whilst the remainder may be due to the other SE types, including SEG, SEH, SEI, SEJ, SEK, SEL, SEM, SEN, and SEO [5]. Based on the positive enterotoxin genes, it is clear that many S. aureus isolates from this study are enterotoxigenic. Some of the genes found in this study, such as sea, seh, seg, and sei, have been associated with outbreaks of food poisoning in different parts of the world [5,[67][68][69][70]. However, seg and sei have not frequently been isolated from food isolates and are, rather, associated with staphylococcal scarlet fever and toxic shock syndrome [71]. The seg, sei, and seh have also been identified in patients with other S. aureus-associated infections [72].
In the current study, all the 13 S. aureus isolates tested positive for seg, sei, and seh genes. The seg and sei genes are components of the egc operon, together with sem, sen, and seo enterotoxin genes [70]. The egc operon is located on a mobile genetic element (MGE) [70] and can thus be transferred to non-pathogenic S. aureus [60]. This combination is, however, rarely found in strains involved in toxi-infections [5,38,46,71].

Conclusions
In conclusion, the current study contributes to the knowledge about S. aureus on beef in South African markets. While the overall prevalence was relatively low, care should, however, be taken when handling pluck meats to avoid cross contamination with utensils, working surfaces, and RTE. Few S. aureus isolates exhibited antimicrobial resistance; however, the presence of mecC-positive S. aureus strains is worrisome. Five classical staphylococcal enterotoxin genes were identified from these isolates, which indicate a health risk to the consumers. The observation of mecC-positive MRSA that are present on food and have been reported also in food-producing animals warrants a One Health study on MRSA in food-producing animals, pet animals, wildlife, and foods in South Africa. These studies should include whole-genome sequencing so as to determine the epidemiology and origins of mecC-positive MRSA.