Genotypic and Phenotypic Characterizations of Methicillin-Resistant Staphylococcus aureus (MRSA) on Frequently Touched Sites from Public Hospitals in South Africa

The hospital environment acts as a reservoir in the transmission of pathogens, such as MRSA, which may cause hospital-acquired infections. This study aimed to ascertain the prevalence, genetic relatedness, antibiotic resistance, and virulence profile of MRSA on some frequently touched hospital sites in South Africa. A total of 777 swabs were randomly collected from 11 frequently touched sites in the hospital environment of three wards of four public hospitals in the KwaZulu-Natal province of South Africa. Isolation of S. aureus and confirmation were done using genotypic and phenotypic methods. Antibiotic susceptibility testing was performed using the Kirby–Bauer disk-diffusion method. MRSA isolates were determined by the presence of the mecA gene. Virulence and resistance genes were detected using a standard monoplex PCR assay. ERIC-PCR was conducted to evaluate the genetic relatedness. An overall prevalence of 12.7% for S. aureus isolates was obtained. Out of these, 89.9% (89/99) were confirmed to be MRSA. The sites with the highest prevalence were the occupied beds (16.2% (16/99)), unoccupied beds (16.2% (16/99)), patient files (14.1% (14/99)), ward phones (13.1% (13/99)), and nurses' tables (14.1% (14/99)). The virulence genes with the highest observed frequency were hld (87 (87.9%)) and LukS/F-PV (53 (53.5%)). The resistance genes with the highest frequency were the tetM and tetK genes detected in 60 (60.6%) and 57 (57.6%) isolates, respectively. The ERIC-PCR results obtained indicated a high level of genetic diversity; however, intraclonal (within a hospital) and interclonal (between hospitals) clusters of MRSA were observed. The study showed that MRSA can contaminate various surfaces, and this persistence allows for the dissemination of bacteria within the hospital environment. This highlights the need for improved infection prevention and control (IPC) strategies in public hospitals in the country to curb their potential transmission risks.


Introduction
A hospital-acquired infection (HAI) or nosocomial infection develops during hospitalisation or within 48 hours after the patient has been discharged. In most cases, this is not the initial cause of hospital admission [1]. Staphylococcus aureus (S. aureus) is considered one of the most important [2] pathogens responsible for HAIs. HAIs are a financial burden in developed and developing countries causing significant strain on the economy due to the high cost of treatments and increased mortality and morbidity rates that are associated with these types of infections [3][4][5]. In addition, S. aureus is one of eight significant pathogens listed by the Global Antimicrobial Resistance Surveillance System (GLASS) alongside Shigella spp., Salmonella spp., Streptococcus pneumoniae, Klebsiella pneumoniae, Neisseria gonorrhoeae, Acinetobacter spp., and Escherichia coli [6].

Sample
Site. Swab samples were collected over three months from September to November 2017. Four provincial public hospitals are classified under the South African National Health Act of 2003 [17] according to their different levels of healthcare viz. central (hospital A), tertiary (hospital B), regional (hospital C), and district (hospital D). Central/ specialized hospitals, with bed sizes between 800 and 1200, accept referrals from both district and regional hospitals. Tertiary hospitals have 400 to 800 beds available and provide specialised services such as those offered by a regional hospital. ey also receive referrals from regional hospitals; however, referrals are not limited to provincial boundaries. Regional hospitals which have between 200 and 800 beds receive referrals from several district hospitals and provide services to a specific regional population. District hospitals have between 50 and 600 beds, depending on their classification of either being small, medium, or large. ese hospitals receive outreach and support from specialists based at regional hospitals. Samples were collected from three wards viz. intensive care unit (ICU) and paediatric and general wards of the studied hospitals.

Sample Collection.
A total of 777 swabs were collected from eleven predetermined and frequently touched surfaces in the hospital environment viz. ward phones, nurses' tables, drip stands, sinks, blood pressure (bp) machines, patient files, occupied and unoccupied beds, ventilators, mops, and the utility room door handles (Supplementary Table S1). e stratified random sampling method was used. Samples were collected by gently swabbing approximately a 5 cm circumference of the surfaces using capped Amies Agar swabs in translucent transport media ( ermo Fisher Scientific, Waltham, MA, USA). e swabs were then placed in a cooler box with ice for transportation to the laboratory and processed within four hours upon arrival.

Phenotypic Identification and Isolation of Staphylococcus aureus.
For enrichment purposes, swabs were broken off into labelled 50 mL blue-capped centrifuge tubes containing 20 mL of tryptic soy broth (TSB) (Sigma-Aldrich, Germany) and incubated at 37°C for ± 24 hrs while stirred (Steyn Scientific, USA). After incubation, a loopful of broth was aseptically inoculated and streaked onto plates containing chromogenic agar, HiChrome TM Aureus Agar Base (HiMedia, India), and supplemented with 2% egg-yolk tellurite emulsion (HiMedia, India). e plates were incubated (Shel Lab, Sheldon Manufacturing Inc., USA) at 37°C for ± 24 hrs. e presumptive S. aureus colonies were subsequently stored in cryovials containing TSB supplemented with 10% glycerol (VWR International Life Sciences, Amresco, Parkway) until further analyses.

Deoxyribonucleic Acid (DNA) Extraction.
Total genomic DNA was extracted using the conventional boiling method [18]. Briefly, colonies from pure S. aureus cultures were suspended in 300 μL of TE (Tris-EDTA) (10 mM Tris-HCl pH 8.0 with 1 mM EDTA) buffer and vortexed to homogenise the cells. e suspension was boiled in a heating block at 100°C for 10 minutes with immediate cooling on ice for 5 minutes.
e tubes were centrifuged (Beckman Coulter Microfuge 16, United States) at 14000 ×g for 5 minutes. e supernatant was transferred into a new sterile tube and stored at −20°C for PCR use. Nanodrop readings were taken to measure the concentration and purity of isolated DNA using the ermo Scientific Nanodrop 2000, UV-Vis spectrophotometer (Wilmington, Delaware, USA). Readings within 1.8-1.9 at the ratio of 260/280 were regarded as pure DNA. Nuclease-free water ( ermo Fisher Scientific, Waltham, MA USA) was used to adjust the DNA concentration accordingly for PCR.

Molecular Confirmation of S. aureus.
Primers for the polymerase chain reaction (PCR) were synthesised at Inqaba Biotechnology (South Africa). e primer sequences, annealing temperatures, and product sizes are indicated in Table 1. e species-specific thermonuclease nuc gene [19] was used for the identification of S. aureus. e positive control S. aureus ATCC 25913 and negative control (nuclease-free water) were used for all reactions.
e Bio-Rad T100 TM ermal Cycler (Singapore) was used to carry out the PCR assays. A 15 μL reaction contained 7 μL of DreamTaq Green PCR Master Mix (2X) ( ermo Fisher Scientific, Waltham, MA, USA), 0.5 μL of both reverse and forward primers of a 20 μM primer concentration, 3 μL of template DNA, and 4 μL of nuclease-free water making a total reaction volume of 15 μL. Gel electrophoresis was carried out by loading the PCR products into a 1.5% (w/vol) agarose gel with 5 μL of 10 mg/ml ethidium bromide ( ermo Fisher Scientific, Waltham, MA, USA) and run at 100 volts for 30 minutes in a tank containing Tris-borate-EDTA (pH 8.3, 1X). A 100 bp DNA ladder (New England Biolabs, Ipswich, USA) was used as the molecular weight marker. Gels were viewed on the ChemiDoc TM Imaging System (Bio-Rad Laboratories Inc., USA).

Genotypic Identification of Resistance and Virulence
Genes. DNA extraction and volumes used to conduct singleplex reactions for the resistance and virulence genes were as described previously (Table 1). e PCR primers and conditions followed are indicated in Table 1. e resistance genes, aac (6′)-aph (2″), blaZ, tetK, tetM, and ermC, which confer resistance to aminoglycoside, macrolide-lincosamide-streptogramin B (MLS B ), tetracycline, and erythromycin, respectively, were identified. Virulence genes for Panton-Valentine leucocidin (LukS/F-PV) and alpha and delta haemolysin (hla and hld) were also identified. e positive control S. aureus ATCC 25913 and negative control (nuclease-free water) were used for all the reactions conducted.

Analysis of the Genetic Diversity of S. aureus Using Enterobacterial Repetitive Intergenic Consensus (ERIC) PCR.
Genomic DNA was extracted using the Quick-DNA TM Miniprep Plus kit (Zymo Research, USA) according to the manufacturer's instructions. e purity and concentration were measured using the ermo Scientific Nanodrop 2000, UV-Vis spectrophotometer (Wilmington, Delaware, USA). A reaction volume of 25 μL was set up: 12.5 μL of DreamTaq Green PCR Master Mix 2X ( ermo Fisher Scientific, Waltham, MA, USA), 9.3 μL of nuclease-free water, 3 μL of template DNA, and 0.1 μL of 100 mM ERIC 1 (5′-CACT-TAGGGGTCCTCGAATGTA-3′) and ERIC 2 (5′-AAG-TAAGTGATGGGGTGAGCG-3′) primers (Inqaba Biotechnology, South Africa). e PCR conditions were as follows: 95°C for 2 minutes, 35 cycles of 30 s of denaturation at 90°C, 1 min of annealing at 52°C, 8 min of extension at 65°C, and final elongation at 65°C for 16 min. Gel electrophoresis was performed at 70 V for 75 min. e ERIC-PCR gels were captured (ChemiDoc TM Imaging System (Bio-Rad Laboratories Inc., USA) and analysed using BioNumerics software version 6.6 Applied Maths NV (bioMérieux, Sint-Martens-Latem, Belgium). A dendrogram was produced using an unweighted pair group with arithmetic mean (UPGMA) method and Dice coefficient parameters of 1% tolerance and 0.5% optimisation.

Statistical Analysis. SPSS software version 25 (IBM SPSS
Statistics) was used for statistical analyses using a probability value of less than 0.05 (p < 0.05) for significance. Fisher's exact and Pearson's chi-square tests were used to determine the significance of the relationship between the genes detected and the hospital/site/wards from which the samples originated. For each statistical model, the dependent variable was the virulence gene presence (0 � absent; 1 � present).
Pearson's correlation analyses (Table S3) (Table S3). e relationships between the genes and site, genes and ward, and genes and hospital were examined. Pearson's chisquare and Fischer's exact test indicated that there was a significant relationship (p < 0.05) between the mecA gene and the site. A significant relationship (p < 0.05) was identified between the hospital and the tetK, ermC, aac (6′)aph (2″), and LukS/F-PV genes (Table S2).

Enterobacterial Repetitive Intergenic Consensus Sequence-PCR (ERIC-PCR) of MRSA.
e enterobacterial repetitive intergenic consensus (ERIC-PCR) was used to determine the genetic diversity of MRSA isolates. Bands were produced for 87 of the isolates recovered; 2 of the isolates were nontypeable (these isolates showed no bands under ERIC-PCR conditions). e 87 isolates were assigned to 54 different ERIC types, namely, A-BC, based on a similarity index of ≥60% (shown by the solid red line) (Figure 2). e results obtained indicated a high level of genetic diversity in the study isolates. However, it was observed that 21% (18/87) of the S. aureus isolates were grouped into six major ERIC types: F (n � 3), G (n � 3), P (n � 3), W n � 3), AC (n � 3), and AG (n � 3) (Figure 2). Of these six main ERIC types, three major ERIC types (P, W, and AC) were shared between different two hospitals, viz. district hospital (D) and regional hospital (C), indicating a potential interclonal spread of MRSA. More so, the AG (n � 3) cluster indicated that isolates from the ward telephone (n � 2) and unoccupied bed (n � 1) within the paediatrics ward of the tertiary hospital (B) belonged to the same ERIC type, also suggesting a possible intraclonal spread of MRSA.

Discussion
e sanitation of a hospital environment plays a crucial role in spreading pathogenic organisms such as MRSA [25]. MRSA can be transferred from person to person or from person to frequently touched objects in the hospital environment, and vice versa [15]. e overall prevalence of S. aureus obtained was 12.7% (99/777) which comprised 89 isolates of MRSA (methicillin-resistant S. aureus) and ten isolates of MSSA (methicillin-susceptible S. aureus). e result obtained indicated a low prevalence of MSSA. Although the reported MSSA prevalence was low, MRSA and MSSA do not differ in the diseases they cause [26]. In Africa, a lower prevalence rate of 2.7% (1/37) was reported by Adekunle et al. [27], who had studied environmental isolates collected from a general hospital in Nigeria. In addition, a prevalence of 17% (8/47), mainly from door handles, was reported in a study conducted in three government hospitals in Ghana [28]. e prevalence rate obtained in this study was higher than a survey conducted by Mukhiya et al. [29], who obtained 40.7% (11/27) from environmental isolates collected from hospitals in Nepal, a developing country. e level of healthcare provided by these hospitals had not been specified. e prevalence in our study was also higher than that of Ekrami et al. [30], who obtained an MRSA prevalence of 60.0% from hospital environmental isolates collected from hospitals in Iran which is also a developing country. South African public hospitals lack funding, have a shortage of resources, are understaffed, and are often overcrowded [31]. ese factors greatly affect IPC implementation and may contribute to a lack of hospital hygiene management. e regional hospital (hospital C) had the highest number of isolates compared to the other three hospitals. However, patients may act as vectors that translocate hospital-acquired pathogens between hospitals [32].
is was evident in a study conducted by Donker et al. [32], who reported a positive correlation (33.0%) between patient referrals and the incidence of hospital-acquired pathogens such as MRSA in hospitals in England and the Netherlands.

International Journal of Microbiology
One of the transmission routes of MRSA is through direct skin contact and shedding of epidermal skin cells [33]. Infected patients or patients that are carriers of MRSA may shed their skin onto the hospital beds.
e results also revealed that there was a significant relationship (p < 0.05) between the mecA gene and the site (shown in Table S2).
ese results are similar to those of Adwan et al. [34], who also reported a significant relationship. It is interesting to note that the highest prevalence of MRSA was obtained from the unoccupied beds compared to other sites in the study (Figure 1). ese results are an indication that IPC protocols pertaining to the laundry or hospital bed disinfection were unsatisfactory as the presence of MRSA may have emanated from a previously admitted patient. e results showed that the occupied and unoccupied beds accounted for 32.4% of the total isolates. ese results indicate substandard cleaning agents or improper execution of IPC protocols. Contamination of unoccupied beds may occur before the patient has had direct contact with the site. ese results are consistent with the results obtained by Pinon et al. [14], who conducted a study in a hospital in France as S. aureus has been reported to contaminate bed linen even though the sheets and pillowcases had been washed [14]. A possible reason may be attributed to the survival abilities of S. aureus for one to 90 days or more on fabrics and materials such as cotton, cotton terry, cotton-polyester blend, polyester, and polypropylene plastic [35].
S. aureus has been associated with a low infectious dosage indicating that S. aureus is highly contagious even in small amounts; only 15 cells of S. aureus introduced into experimental lesions were enough to result in infection [36]. erefore, patients through direct contact with the surface may be exposed to the pathogen through open wounds or postsurgical procedures, thus suggesting that contaminated surfaces may be an essential and underappreciated source of MRSA transmission [33]. Transmission of pathogens is dependent on a range of factors which are but not limited to the viability of the pathogen on that environmental site, relative humidity, the frequency of contact between patients, healthcare workers, and contaminated surfaces, ambient temperature, and the dose of the transmitted pathogen [34]. e antibiotic susceptibility testing results showed that only 46% of the isolates were resistant to cefoxitin, the antibiotic used as an indication for methicillin resistance of S. aureus (MRSA) isolates, according to CLSI recommendations (M100-S27, 2017). e results obtained were contrary when compared to the high mecA gene presence of 89% indicated by the PCR results. S. aureus strains that are methicillin resistant carry either the mecA, mecB, or mecC genes. ese genes are acquired genetic determinants that encode for PBP2a or PBP2a ' , low-affinity penicillin-binding proteins [26]. PBP2a has a very low affinity for most β-lactam antibiotics [26]. e level of mecA transcription or presence in isolates does not predict the level of phenotypic methicillin resistance. A possible explanation is found in the work of Lee et al. [37], who identified three attributing factors. e first factor is stringent stress response (the bacteria's reaction to different stress conditions, such as amino acid, fatty acid, S. aureus  100  90  80  70  60  50  40  30  20 Key iron limitation, and heat shock) [37]. e work of Boylevavra et al. [38] presented the second factor. e inactivation of VraS was shown to have induced the transcription of mecA but did not increase the level of PBP2a activity [38].
VraS is a part of the regulatory system made up of the sensor protein VraS and response regulator protein VraR. ese proteins are involved in controlling the cell wall peptidoglycan biosynthesis [37]. e third factor is the chaperone foldase protein, PrsA, which changes the amount of correctly folded PBP2a that is found in the membrane. As a result, this would in turn affect the methicillin resistance without hindering the transcription of the mecA gene. us, these factors are an indication that the gene presence does not determine antibiotic resistance [37]. e presence of the LukS/F-PV gene is commonly associated with strains of community-acquired MRSA; however, this varies based on geographic locations [39]. e virulence gene for Panton-Valentine leucocidin, LukS/F-PV, was detected in 53.5% of the isolates. is percentage was higher than that previously reported by Adwan et al. [34], who obtained an incidence of 14.3% in a similar study. e presence of this gene indicates the possible production of a toxin, which induces the formation of virulence pores in leukocytes [34]. If patients are exposed to these strains, this may result in severe chronic skin infections or necrotising pneumonia with an extremely high mortality rate even in young and healthy patients. e presence of the LukS/F-PV gene obtained in this study is contrary to the results of Bhatta et al. [39], who reported that there was no presence of this gene among hospital environment isolates collected in Nepal, which may have been an indication of the gene not being associated with isolates from the hospital environment.
ere are limited studies that have implemented ERIC-PCR to evaluate the clonality of S. aureus isolates from the hospital environment. e ERIC-PCR results obtained in this study showed a high diversity between the MRSA isolates collected. e results showed that 2 of the isolates were nontypeable with ERIC-PCR. ese results were comparable to the study conducted by Adwan et al. [34] whose results indicated that two isolates out of 265 swabs collected from two hospitals in Iran were nontypeable using ERIC-PCR. e ERIC-PCR results indicated that three of the six major ERIC types were shared between the district hospital (D) and regional hospital (C) indicating a potential interclonal spread of MRSA clones as district hospitals fall under level 1 of the referral system of South African hospitals [40]. Patients are referred from the district hospitals to the local and regional hospital whenever the correct health service cannot be offered. ese results further highlight the transmission risk of pathogens between hospitals due to referrals. e spread of MRSA was also observed between two sites of the same hospital ward (Figure 2) suggesting an intraclonal spread of MRSA. e failure to adhere to IPC measures is a likely cause. However, the study was limited by the lack of information on the infection control practices at these hospitals that precluded the comparative analyses of the IPC practices between different hospital settings, warranting further studies to ascertain this.

Conclusion
In summary, the study highlights the prevalence and phenotypic and genotypic characterizations of S. aureus indicating that the hospital environments could act as potential reservoirs in the transmission of MRSA encoding diverse antibiotic resistance and virulence genes in South African public hospitals.
is was further supported by ERIC-PCR typing showed by the recovery of genetically similar MRSA isolates from different surfaces within the same hospital (intraclonal spread) and between different hospitals (interclonal spread), which can be further disseminated to other sites if IPC measures are suboptimal. We conceive that our work can be used as a framework for future surveillance initiatives to improve hospital hygiene through IPC-centred strategies to minimize the presence of drug-resistant and pathogenic microorganisms that are present in the hospital environment.
is would reduce hospital-acquired infections, due to contaminated sites, and provide a safe environment for patients and healthcare workers.

Data Availability
e generated data used to support the findings of this study are included within the article.
software. College of Health Sciences Young Researchers Grant Competition (LA Bester) of the University of Kwa-Zulu-Natal funded this study. Table S1: description of the samples collected from hospitals, wards, and sampling sites. Table S2: Fischer's exact test and Pearson's chi-square test for virulence and resistance genes from eleven frequently touched sites, wards, and hospitals.