Prevalence and Characteristics of Staphylococcus aureus Associated with Meat and Meat Products in African Countries: A Review

Antimicrobial resistance has been increasing globally, which negatively affects food safety, veterinary, and human medicine. Ineffective antibiotics may cause treatment failure, which results in prolonged hospitalisation, increased mortality, and consequently, increased health care costs. Staphylococcus aureus causes a diverse range of infections including septicaemia and endocarditis. However, in food, it mainly causes food poisoning by the production of enterotoxins. With the discovery of methicillin-resistant S. aureus strains that have a separate reservoir in livestock animals, which were termed as livestock-associated methicillin-resistant S. aureus (LA-MRSA) in 2005, it became clear that animals may pose another health risk. Though LA-MRSA is mainly transferred by direct contact, food transmission cannot be excluded. While the current strains are not very pathogenic, mitigation is advisable, as they may acquire new virulence genes, becoming more pathogenic, and may transfer their resistance genes. Control of LA-MRSA poses significant problems, and only Norway has an active mitigation strategy. There is limited information about LA-MRSA, MRSA in general, and other S. aureus infections from African countries. In this review, we discuss the prevalence and characteristics of antimicrobial susceptible and resistant S. aureus (with a focus on MRSA) from meat and meat products in African countries and compare it to the situation in the rest of the world.


Introduction
The increase in antimicrobial-resistant zoonotic bacterial pathogens has become a significant public health challenge. There are diverse drivers of antimicrobial resistance, with the use of antimicrobial agents in food-producing animals as a major driver of resistance in animal bacteria [1]. Ineffective antimicrobials may cause treatment failure, which results in prolonged hospitalisation and increased mortality, and consequently, increased health care costs.
Staphylococcus aureus causes a diverse range of infections in humans. These include severe diseases such as septicaemia and endocarditis, which are frequently associated with high mortality if not treated properly [2]. This bacterium has been reported in the last few decades to be resistant to many of the available antimicrobial agents, and recently, it became resistant to one of the last lines of treatment daptomycin and linezolid [3,4]. However, the most common cause of nosocomial and community-associated staphylococcal infections is methicillin-resistant S. aureus (MRSA) [5].
S. aureus has been documented as one of the most significant causes of hospitalacquired infections in the past decades. It belongs to the ESKAPE group of bacteria

Prevalence of S. aureus in Raw Unprocessed Red Meat
In Table 1, a summary of the studies performed on raw unprocessed red meat in Africa is listed. * Confidence interval: the confidence interval (CI) was set at 95% and calculated based on the exact binomial method [35].
We could find data on a total of 2853 samples from all over Africa. Overall, 699 were contaminated with S. aureus. Positive samples included raw unprocessed beef, pork, goat meat, camel meat, lamb/sheep, and unspecified red meat products. The total prevalence of S. aureus in meat products was 24.5%, and of all red meats, beef samples showed the highest prevalence with 33.08% of the samples being positive. These findings indicate a higher potential risk of beef for human infections. Compared to data from other countries, the results are similar [36][37][38][39], though in China, a higher percentage of contamination was found, with a little over 50% of the raw beef samples found positive [40], as well as in a specific study from the US where a prevalence of 65.6% was found in retail beef [14], while in another US study, the overall positivity S. aureus was 16.4% [41]. This large variation in both Africa and worldwide may be attributed to several factors, including sampling method, isolation methods, sampling site on the carcass, different cuts of meat, contamination during and after the slaughter, storage of the meat, as well as the processing of the meat [42]. A total of 691 pork samples were analysed in Africa (Table 1), and almost 14% of them were found positive for S. aureus. These findings are similar to what was found in the US [38], while much higher percentages were found in other studies from the US and Brazil [39,[43][44][45]. In Asia, a lower prevalence was found [40].
There was quite some variation in the prevalence between the African countries, even between studies from the same country as exemplified by the Nigerian studies. In Nigeria, there was an apparent increase in prevalence [27,46], and the reasons for this apparent increase remain unclear. In South Africa, two studies performed in the same year showed also substantial differences [25,26]. However, the studies used a different type of sample and a different way of collecting the sample, different isolation methods, and were conducted in a different geographical location. In Ivory Coast, similar to beef, the prevalence was lowest [24]. Additionally, here it is clear that the few data available do not allow a broad conclusion on the prevalence nor evolution of the contamination of meat products in Africa. Hence, it is necessary to conduct more research on the subject in Africa.
Only one study reported the prevalence of S. aureus from goat meat [27]. The prevalence was on the lower side. Similarly, only two studies were conducted on the prevalence of S. aureus in sheep meat. Both were conducted in South Africa, and the results showed major differences [25,26]. Only Egypt reported the prevalence of S. aureus from camel meat [32]. Those studies do not allow to have a good view on the prevalence of S. aureus in those meats; however, it indicates that also those meat products can substantially be contaminated. Surprisingly, on unspecified raw red meat in Nigeria, nearly all products were contaminated [34], while a former study showed a low level of contamination [33].
The distribution of Staphylococcus aureus in red meat differs, thus widely based on different studies, and it seems to be dependent on both the country and the meat product, though other factors cannot be excluded. No firm conclusion on differences between countries, in time or red meat type, can be made, as the studies are too limited and too different. It is, however, clear that S. aureus contamination of red meat products is, and remains, a problem that needs mitigation.

Prevalence of S. aureus in Raw Unprocessed Poultry
Compared to red meats, quite a lot more samples from poultry have been analysed, though the numbers are still limited relative to the vastness of the African continent. Moreover, the studies have been performed in a limited number of counties (Table 2). We could find data from 11,080 samples of chicken meat, and of those, a little over 20% were found to be contaminated with S. aureus ( Table 2). The prevalence of S. aureus in poultry meat seems to be lower than in beef. The oldest study is from Nigeria and was performed in 2009 on a relatively small number of samples. The prevalence was very high, with 80% of the samples being positive [27]. A second larger study performed in 2011 from Tunisia showed, however, only 51.19% of the samples were positive [47]. Due to differences in sample size, it is difficult to assess whether there was a potential evolution. Similarly, in South Africa, the prevalence seems to drop over time, but it remained a high prevalence [25,26]. Other studies in African countries were limited, though one very large study in Algeria showed a prevalence of close to 20%, which is close to the average prevalence measured for all of Africa [28].
The overall prevalence found in Africa is similar to what has been found in China and Brazil [45,48]. In the US, a higher prevalence was found in most of the studies, though one study showed a prevalence as low as 0.3% [39,43,49]. There are few data available from Asiatic countries, and the prevalence rates were overall lower than in most other countries [48,50]; however, a conflicting study reported a higher prevalence of 70% [20]. From the data published, it is clear that poultry in Africa is also contaminated with S. aureus, though to a lesser extent than most other studies. Table 3 shows the studies that have determined the prevalence of S. aureus in raw processed meats. All were beef samples. Out of 247 tested samples, 31 (13%) were positive for S. aureus. One would expect higher levels of contamination, compared to unprocessed meat, as processed meat is more prone to microbial contamination. However, we could not confirm this with the published data. It should be noted also that some processes such as smoking and drying may reduce contamination, and from the above studies, the meat used was dried and smoked. As no ample data are present, no real conclusions can be drawn for this type of food.   Table 4 indicates the prevalence of S. aureus from ready-to-eat meat, and only one study was found and few samples were included. This study showed a low prevalence of S. aureus. The contamination of meat by S. aureus across the food chain is a complicated process. The contamination may originate from animals, as well as from humans. It has been shown before that in humans, the main source of contamination is food handlers that carry S. aureus in their noses or hands [54]. Improper hygiene at that level should be avoided to reduce the odds of meat contamination and food poisoning. However, due to the multitude of small slaughter and meat processing operations in Africa, this may prove to be difficult.

Prevalence of S. aureus in Ready-to-Eat Meat
The other source of contamination is the animals themselves. Food-producing animals carry S. aureus on their skin, nose, as well as in their intestine. The main factors that influence the level of contamination are the length at which animals are transported and the methods which are used to move animals from one place to another, holding conditions, geographic location, as well as climate changes [55]. Next to that, it is always important to follow proper slaughter and food handling protocols to minimise contamination with pathogenic microorganisms. The growth of S. aureus and the production of enterotoxins in food is caused by improper handling of foods and the improper storage conditions that support the growth of this pathogen [56]. In developing countries, pathogens are being transmitted from livestock to humans through contact or contaminated meat or meat products [57]. Only typing of the strains can bring clarity in the origin of the contamination and the public health risk associated with the contamination.
In Africa, red meat has the highest prevalence of S. aureus, compared to other types of meat. S. aureus has been reported to be one of the pathogens causing foodborne infections from different regions worldwide. Findings indicate a specific problem on the African continent in beef production that needs further investigation.
Few studies in Africa have assessed contamination sites across the food chain. However, in Ethiopia, S. aureus was found on all equipment that was used in the abattoirs [58].
From the results obtained in our review, more research needs to be conducted in Africa to fully understand the prevalence and way of spread of S. aureus via the food chain. It is also important that African countries should improve the hygiene conditions, especially in abattoirs, butcheries, and retail shops to improve public health. S. aureus has been associated with foodborne diseases in different parts of the world, and infections caused by this pathogen are difficult to treat because some strains are resistant to antibiotics.

Antimicrobial Resistance in S. aureus from Meat and Meat Products in Africa
The data for antimicrobial resistance in different types of meat from various African countries comprised 20 antibiotics that were tested, representing 10 different classes of antibiotics (Table 5). Not all studies used the same antibiotics, hence the difference in the number of strains tested. Studies originated from Nigeria, South Africa, Cameroon, Algeria, Tunisia, and Egypt, and the cumulated results are represented in Table 5. Due to the limited number of studies and the limited number of strains involved, few conclusions can be drawn on differences between regions. Nevertheless, an overall view may help in what antimicrobial resistance is present in Africa. Due to the differences in the numbers of strains tested, the resistance percentages within an antibiotic class may vary. Therefore, the data presented only give an impression of the resistance problem.
Upon examination of the data in more detail, we observe that the resistance against beta-lactam antibiotics is in general quite high. It is surprising to see such high resistance against the penicillinase-resistant β-lactam antibiotics, as the isolation methods that were used were not specifically targeting MRSA. It should be noted, however, that in the concerned studies, the resistance was not confirmed by PCR and may thus be lower [28,32,34,54,59]. Nevertheless, it is an indication that the levels of MRSA on meat samples may be quite high. It has indeed been shown that meat can be contaminated by LA-MRSA, especially pork, as the prevalence of LA-MRSA is high in pigs, as well as by human MRSA strains [60]. Nevertheless, MRSA should be confirmed by PCR, as the disk diffusion test is not very accurate for the determination of MRSA [61]. Thus, the true prevalence of MRSA in meat remains obscure. Interestingly, one group tested for ceftaroline, an anti-MRSA antibiotic, and the resistance is high [26]. Vancomycin (0.7%) and gentamicin (1%) showed very low resistance, and only against amikacin, there was quite some resistance. Resistance against vancomycin in S. aureus, as determined by the disk diffusion test, is not very reliable [62], and the presence of vancomycin resistance in S. aureus may thus be questionable. Resistance against the tetracyclines is somewhat varying between studies, as was the case with the difference in the prevalence of resistance against the different antibiotics in this class. A similar observation can be made for fluoroquinolones. Resistance against other antibiotics was medium to low. Apart from the presence of MRSA, there does not seem to be a very large problem with resistance in this bacterium on meat and meat products in Africa.
Compared to other parts in the world where antimicrobial resistance has been assessed on strains from food, similar resistance percentages were noted such as in China, [63,64], the USA [39], Greece [65], India [66], Italy [37,67,68]. Few other studies showed differences, including a German study, which showed high resistance in tetracycline, oxacillin, erythromycin, clindamycin, kanamycin, gentamicin, and ciprofloxacin in S. aureus from turkey meat [69].

Antimicrobial Resistance Genes Detected in S. aureus from Meat in Africa
Only two studies reported the resistance genes present in S. aureus (MSSA) from meat products in Africa ( Table 6). The genes identified are commonly identified worldwide [76]. However, since the data come only from two countries, with each one being a limited study (Table 6), conclusions are difficult to make. Table 6. Antimicrobial resistance genes in S. aureus from different meat products in African countries.

Genotyping of Different S. aureus Meat Isolates from Africa
The genetic background of only a few strains has been determined, and all came from a single study in Tunisia (Table 7) [47]. Four strains were MSSA ST398, which is surprising as most ST398 strains are MRSA and came from chicken and veal. MSSA CC398 has been mainly associated with humans [77], where the human evasion cluster genes IEC are, in general, present [78]. Other clonal complexes detected included ST22 which is also mainly associated with human infections [79], though it has also been associated with infections in animals including outbreaks in veterinary clinics [80][81][82]. The ST8 clone is a typical community-acquired clone also associated with hospital outbreaks, usually being an MRSA as the USA300 clone [83], which has been reported worldwide, including sub-Saharan Africa [84,85]. The presence of MSSA ST398, ST8, and ST22 indicates that they might be more associated with humans contaminating the meat, either during slaughter, processing, or storage. We separated the studies dealing with the detection of MRSA from the other studies, as they applied a very different methodology. Using selective isolation, the studies were focusing only on the presence of a specific subtype of S. aureus and thus not giving a full picture of the presence of S. aureus. Moreover, these MRSA have never been implicated in food poisoning and have thus a more zoonotic aspect. These strains can be of animal origin (the LA-MRSA) or they can originate from humans. The LA-MRSA are specific types of MRSA that can be distinguished by molecular typing. Therefore, the prevalence of LA-MRSA is mainly related to carcass contamination from animal origin, rather than other origins. Therefore, the prevalence of LA-MRSA is in direct relation with the prevalence in the animal itself, although not excluding cross-contamination during slaughter and processing.
We found data on the selective isolation of MRSA from a total of 3746 meat samples which included beef, pork, and poultry ( Table 8). The numbers should be interpreted with care, however, as not all MRSA cases were confirmed, and false-positive results are possible. This is also evident in Table 6, where not all strains are resistant to all β-lactam antibiotics. The highest prevalence of MRSA was observed in pork (12%), which is the animal species with the highest prevalence of LA-MRSA worldwide. However, the carriage differs across the countries, and the prevalence at pig meat level is, in general, lower than the prevalence at farm level [86]. The prevalence rates in poultry and beef were 6.76% and 6.12%, respectively. The prevalence of LA-MRSA in these latter species has been shown to be lower than in pigs except for veal calves [86]. The prevalence of MRSA in pork in Africa is, however, much lower, compared to what has been found in countries with a high prevalence of LA-MRSA in pigs [45,67,[87][88][89]. More research is required in order to determine whether the prevalence of LA-MRSA in animals is indeed lower in Africa. Nevertheless, the prevalence of MRSA in pork from countries such as China and the USA, where the prevalence in pigs is moderate to high, was lower [36,41,43,64]. The prevalence of MRSA in chicken meat from Africa (6.76%) was lower or similar, compared to most other studies [20,41,45,48,64,90,91].
Prevalence in beef was 6.12%, which was higher than 4%, 0.8%, and 1.7% found in beef meat from the USA [36,38,49] and also 4.4% found in Hong Kong [91], while in Brazil, a very high prevalence (23.3%) was found [45]. However, more research is required in order to ascertain the extent of MRSA contamination of meat in Africa.

Antimicrobial Resistance Profile of MRSA in African Countries
From Table 9, it is clear that not all tests were performed accurately, as MRSA is resistant to all beta-lactam antibiotics that were included in testing (there were no anti-MRSA β-lactams included). Nevertheless, several strains were not confirmed by PCR to be MRSA, which might affect the final results. Resistance was commonly found against tetracyclines, which is indicative of the presence of LA-MRSA ST398. These strains are classically nearly 100% resistant to tetracyclines. LA-MRSA is also typically multidrug resistant, with high resistance percentages against aminoglycosides, macrolides, and lincosamides, as well as fluoroquinolones, though this may vary a bit according to the studies [88,[94][95][96][97]. Thus, the data strongly indicate that many MRSA could be LA-MRSA. The high prevalence of vancomycin resistance has to be interpreted with care, as the studies were carried out using the disk diffusion test, and this is not very reliable [98,99]. Vancomycin resistance has never been reported before in LA-MRSA and is extremely rare in other MRSA.
Unfortunately, there are little data from Africa on the prevalence of LA-MRSA [42], though it is present in animals. It has indeed been shown that meat can be contaminated by LA-MRSA, especially pork since the prevalence of LA-MRSA is, in general, highest in pigs [60]. Nevertheless, MRSA should be confirmed by PCR, as the disk diffusion test is not very accurate for the determination of MRSA [61]. The true prevalence of MRSA in meat remains thus unclear.  References: [26,30,46,47,53,[70][71][72][73][74][75]92,93].

Antimicrobial Resistance Genes Detected in MRSA from Meat in Africa
In few studies, a selection of resistance genes in MRSA have been determined, which makes the data rather incomplete ( Table 10). The genes found in Africa are the classical genes found worldwide, though one big exception is the presence of vanA and vanB in Egyptian camel and beef isolates [30,32]. Vancomycin resistance mediated by the vanA or vanB gene in staphylococci has rarely been described in staphylococci and is mainly associated with vancomycin resistance in enterococci [10]. The reason for this presence of vancomycin could be the use of avoparcin, another glycopeptide antibiotic, which is frequently used in Egypt for the prevention of necrotic enteritis and as a growth promoter in the poultry industry [100]. Another description of vancomycin resistance in Egypt was from ready-to-eat food (beef burgers and hot dogs); however, in these four strains, no van gene-mediated resistance could be detected [101]. Unfortunately, these strains have not been typed, and therefore, it is unclear whether these might be strains of human or animal origin. A survey conducted from Egypt hospital on dairy food, food handlers, and patients indicated the presence of methicillin-resistant, coagulase-positive, and vancomycinresistant staphylococci but only from humans (food handlers and patients) [102], whereas coagulase-negative staphylococci (both MR and VR) were identified from food, food handlers, and patients; however, antimicrobial resistance genes were not studied [102]. Hospital infections with VRSA have been described on occasions in Egypt [103]. The presence of vancomycin resistance in MRSA is highly concerning, as it renders an infection with these bacteria very hard to treat. The situation seems unique to Egypt.

Genotyping of MRSA Meat Isolates from Africa
In South Africa, ST612 (a CC8 strain) has been associated with human infections and has been identified as a dominant clone in hospitals in Cape Town and other provinces [71,104]. However, this clone has also been frequently isolated from animals in other continents such as horses in Australia [105], leading to the detection of the strains also in veterinarians working with horses and other people in contact with horses [67,[106][107][108], which indicates the possible transmission from humans to animals. This clone was also multidrug resistant, hence the importance of its spread through the food chain [109].
The presence of ST398 (CC398) indicates that the situation in North Africa might be more similar to Europe, as this is the most detected LA-MRSA clone in animals. Probably this is due to its closer geographical location. The ST398 clonal complex is associated with multi-resistance, including heavy metal resistance, which highlights the need for more studies in Africa on its prevalence. There is thus a need for more research on the prevalence and characteristics of LA-MRSA in African countries [29].
The current data are not representative of what types of MRSA can be found in meat from Africa, nor even of the two countries on which we have information regarding typed strains. The typed MRSA isolates are derived from a single study in Tunisia and two studies from South Africa (Table 11) [47,71,110]. Only in the study by Chairat, typical LA-MRSA ST398 were detected, next to an ST30 strain, which is known to be human associated [47]. In South Africa, one ST239 was found, which is peculiar, as ST239 has been found in animals only on rare occasions and only in Belgium [111][112][113]. ST239 has been mainly reported as an MRSA clone spreading in Asia, though it has also been detected in South Africa and other African countries [114][115][116][117]. It is thus not clear whether this strain might be of animal or human origin, as there are not enough data on the types of MRSA in animals in South Africa.

Conclusions
Unfortunately, there is little information on S. aureus from meat products in Africa, and there is even less on the types and their antimicrobial resistance. Nevertheless, the observation is that meats can be heavily contaminated with S. aureus, including MRSA, and this represents a potential public health threat. There is thus an urgent need for more studies on the subject to be able to estimate the public health burden. The current studies are fragmentary using different methodologies, which makes it difficult to compare studies. While the MRSA clones on meat detected in few studies seem to be mainly of human origin, the presence of LA-MRSA CC398 warrants further investigation, as does the presence of VISA.  Acknowledgments: The authors express gratitude to Sidney Maliehe for assisting the first author with formatting the manuscript.

Conflicts of Interest:
The authors declare no conflict of interest. The fundrs had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.