Whole Genome Sequencing and Antimicrobial Resistance of Staphylococcus aureus from Surgical Site Infections in Ghana

Staphylococcus aureus (S. aureus) is a common cause of surgical site infections (SSIs) globally. Data on the occurrence of methicillin-susceptible S. aureus (MSSA) as well as methicillin-resistant S. aureus (MRSA) among patients with surgical site infections (SSIs) in sub-Saharan African are scarce. We characterized S. aureus from SSIs in Ghana using molecular methods and antimicrobial susceptibility testing (AST). Wound swabs or aspirate samples were collected from subjects with SSIs. S. aureus was identified by matrix-assisted laser desorption ionization–time of flight mass spectrometry (MALDI-TOF-MS); AST was performed by Kirby-Bauer disk diffusion, and results were interpreted according to the Clinical and Laboratory Standards Institute (CLSI) guideline. Detection of spa, mecA, and pvl genes was performed by polymerase chain reaction (PCR). Whole-genome sequencing (WGS) was done using the Illumina MiSeq platform. Samples were collected from 112 subjects, with 13 S. aureus isolates recovered. Of these, 92% were sensitive to co-trimoxazole, 77% to clindamycin, and 54% to erythromycin. Multi-drug resistance was detected in 5 (38%) isolates. The four mecA gene-positive MRSA isolates detected belonged to ST152 (n = 3) and ST5 (n = 1). In total, 62% of the isolates were positive for the Panton-Valentine leukocidin (pvl) toxin gene. This study reports, for the first time, a pvl-positive ST152-t355 MRSA clone from SSIs in Ghana. The occurrence of multi-drug-resistant S. aureus epidemic clones suggests that continuous surveillance is required to monitor the spread and resistance trends of S. aureus in hospital settings in the country.

Staphylococcus aureus (S. aureus) is a frequent cause of SSI [11][12][13]. Among S. aureus strains, the prevalence of methicillin resistant S. aureus (MRSA) is rising on the African continent [14][15][16][17]. This is of major concern due to the multi-drug-resistant nature of MRSA and the limited therapeutic options available to treat infected patients.
In Ghana, SSIs are the most common healthcare-associated infections and account for up to 33% of all hospital-acquired infections [5]. Although S. aureus, particularly MRSA, in SSI can have severe adverse prognostic implications, scant epidemiologic information exists, including information on its occurrence, antimicrobial resistance and molecular characteristics. In this study, we describe the antimicrobial resistance patterns and molecular characteristics of S. aureus recovered from subjects with SSIs at two hospitals in Accra, Ghana.

Results
Of the 112 subjects, 56 (50%) and 56 (50%) were enrolled at the 37 Military Hospital (37-MH) and Korle-Bu Teaching Hospital (KBTH), respectively. Subjects recruited at the 37-MH consisted of 28 (50%) males; those from KBTH comprised 50 (89.3%) females. Table 1 describes the population sampled at the two hospitals. Data collected on antibiotic use before and after surgery indicate that metronidazole, amoxicillin/clavulanic acid, clindamycin, and ciprofloxacin were frequently administered to subjects in both hospitals.

Proportions of Subjects Positive for Staphylococcus aureus and Antimicrobial Resistance
Culturing of one sample per patient showed bacterial growth (i.e., S. aureus, K. pneumoniae, E. coli, and P. aeruginosa) in 70 (62.5%) of the 112 samples. S. aureus was identified in 13 of 70 culture-positive samples (18.6%). Of the 13 S. aureus positive patients, two were co-infected with E.coli, two with P. aeruginosa, and five with K. pneumoniae.

Discussion
This study provides insights into the phenotypic and molecular characteristics of S. aureus and MRSA recovered from subjects with SSIs in Ghana. The proportion (11.6%) of SSIs positive for S. aureus found in this study is similar to what has been found in other countries [18,19]. On the contrary, a previous study conducted in a private hospital in Ghana recorded a 54% S. aureus prevalence [20]. Of note, the CDC criteria used in the selection of the subjects in this study were not used in the previous studies.
S. aureus isolates in Africa are known to have high rates of resistance to erythromycin, tetracycline, and co-trimoxazole due to frequent prescription of these drugs [21]. Therefore, the high level of resistance to tetracycline was not surprising since its use is still high in Ghana at the primary healthcare level. On the contrary, sensitivity to clindamycin and co-trimoxazole was high, probably because newer drugs have led to less use of those drugs for common indications. Additionally, the majority of the isolates were susceptible to clindamycin; this indicates that clindamycin is still useful for staphylococcal treatment in Ghana. The presence of three lincosamide-resistant isolates, likely due to frequent use in hospitals, suggests that resistance is emerging, and prudent use of this antibiotic is imperative.
Vancomycin is the drug of choice for treating MRSA infections [22], although reduced susceptibility of MRSA to vancomycin has been reported [17,23]. In this study, all MRSA isolates were fully susceptible to vancomycin, which is consistent with Ghanaian reports from healthy carriers and other types of infections [24][25][26][27], as well as in other parts of Africa [28,29]. These findings are most likely related to the low availability of the drug and the subsequent scant use in Ghana [30]. We observed that one MRSA isolate resistant to amoxicillin clavulanic acid and another resistant to ciprofloxacin were recovered from two different patients who were given the same antimicrobial agent for post-surgical antimicrobial therapy (Table 2). There is, therefore, a possibility that the treatment may be ineffective in such patients.
Previous studies in Ghana used spa and MLST typing of S. aureus isolates [24][25][26][27], while in this study, WGS was used to characterize the isolates. The use of WGS has many advantages in light of giving precise data on antimicrobial resistance, typing for outbreak control, and surveillance, as well as toxin gene profiling. In this study, we also found an isolate resistant to chloramphenicol, a drug that was not tested by disk diffusion (Table 2). These findings will help improve infection control within the hospital setting in the future.
We found PVL-positive ST152 (t355) as the dominant clone in the collection of S. aureus recovered from SSI isolates. This clone is known to be circulating in Ghana and widely distributed in many African countries as well as in Europe [21,26,[31][32][33][34]. In Ghana, PVLpositive t355 (ST152) MSSA is a common clone and has been previously reported [26]. However, ST152-MRSA has not been identified despite examining over 500 S. aureus isolates originating from more than 2000 carriage and clinical human samples [24][25][26][27]. Consequently, the novel MRSA clones identified in this study demonstrate the shifting resistance patterns and potential selective pressure due to antibiotic use. PVL-positive ST152-MRSA is a pandemic clone that has been associated with community-acquired MRSA in Central Europe and the Balkans [34][35][36]. Since PVL-MRSA is associated with significant infections in patients without any risk factors [37], this finding is of concern and could inform the prognosis of SSIs in Ghana.
Our study also identified other sequence types, including ST5, ST30, ST3249, and ST1. The sequence type ST5 has been previously reported in Ghana [26,33], and one of the two ST5-MSSA in this study carried the pvl gene, consistent with previous reports [26]. It has been documented that ST30 is the common sequence type in Australia, but it has also been described in Ghana [24][25][26][27]. In addition, studies in other countries like South Africa, USA, and Germany [38] have also reported ST30-MSSA. Although ST30 commonly carries the PVL toxin [21,[24][25][26], the ST30-MSSA in this study did not carry the pvl gene. This finding contrasts with what has been previously reported in Ghana [24][25][26]. The present study also identified a sequence type, ST3249, which has only been previously identified in a burn unit in a hospital in Ghana [39].
Various virulence factors allow S. aureus to colonize its host and cause disease. The enterotoxin genes (sea, sei, sem, sen, seo, sek, sep), leukocidin genes (E and D), and γ toxin genes (hlgA, hlgB and hlgC) identified in this study allow invasion and damage of host tissues [40]. PVL, leukocidin (E and D), and γ toxins (hlgA, hlgB and hlgC) have been suggested to contribute to enhanced virulence, suggesting worse clinical outcomes. This makes it necessary to properly identify genes for the purpose of controlling and monitoring their spread and informing infection control measures. Additionally, some studies have reported the tst gene associated with certain MSSA lineages [41,42]. Similarly, the only positive tst gene isolate was MSSA, belonging to t127 (ST1). The tst gene encodes the toxic shock syndrome toxin (TSST-1). This toxin is known to be secreted by some S. aureus isolates and causes toxic shock syndrome, which is a life-threatening condition.

Study Design, Site, and Sampling Procedures
A hospital-based cross-sectional study of SSI was conducted between June and November 2018 at the 37 Military Hospital (37-MH) and Korle-Bu Teaching Hospital (KBTH) in Accra, Ghana. Subjects from 37-MH were enrolled from the Surgical Outpatient Department and inpatient General Surgery, Obstetrics and Gynaecology, Orthopaedic, and Trauma wards. Subjects from KBTH were recruited from Surgical, Orthopaedics, Maternity, Obstetrics and Gynaecology, Neurosurgery, and Paediatric Surgery wards.
In this study, patients who had undergone surgery, had developed infections, and met the CDC criteria for classification as SSI [1] were recruited into the study. Demographic data, such as age, sex, ward of admission, operation type, period of hospitalization, and antimicrobial therapy, were collected from the medical records of each patient.
Using a sterile cotton-tipped applicator or a syringe, a wound swab, fluid, or aspirate was aseptically collected at the time of recruitment from subjects who gave informed consent. The samples were transported to the Noguchi Memorial Institute for Medical Research, University of Ghana, for phenotypic and molecular analysis.

Isolation and Identification of S. aureus
Samples were cultured on blood agar (Oxoid, Basingstoke, Hants, UK) and mannitol salt agar (Oxoid, Basingstoke, Hants, UK) and incubated for 18-24 h at 37 • C. S. aureus was identified by colonial morphology, Gram stain characteristics, catalase test, and coagulase test (Rabbit plasma) and confirmed using MALDI-TOF-MS (Bruker, Billerica, MA, USA).

PCR Detection of spa, pvl, and mecA genes
Crude DNA was extracted from S. aureus isolates as described previously [44]. A multiplex PCR was performed to detect the spa, pvl, and mecA genes as described previously [45], with slight modifications. Cycling conditions for PCR and primers for all genes detected are listed in Table 4. Each PCR tube contained 12.5 µL of multiplex PCR Mastermix (Qiagen, Hilden, Germany), 2.5 µL of RNase-free water (Qiagen, Hilden, Germany), 8 µL of primer mix, and 2 µL of DNA template. DNA amplification was performed using an Eppendorf Mastercycler (Eppendorf, Hamburg, Germany). A 2% w/v agarose gel was used to analyze the amplified products. The expected band sizes for the various genes were spa (variable region: 200-600 bp), mecA (162 bp), and pvl (80 bp). Table 4. Primer sequences used in the study.

Whole-Genome Sequencing
Whole-genome sequencing was performed with the Illumina MiSeq sequencer (Illumina, San Diego, CA, USA). The extracted DNA was quantified using a Qubit doublestrand (ds) high-sensitivity (HS) kit (Thermo Fisher Scientific, Waltham, MA, USA). A DNA concentration between 100 and 500 ng in 30 µL of the extract was used as starting material for library preparation. Libraries were prepared according to the manufacturer's instruction using the Nextera DNA Flex Library preparation kit (Illumina, San Diego, CA, USA). Amplified libraries were quantified with the 2100 bioanalyzer system (Agilent, Santa Clara, CA, USA) and subsequently by quantitative PCR (qPCR) using the Kapa Sybr Fast qPCR kit (Kapa Biosystems, Wilmington, MA, USA). Using the fragment sizes generated by the bioanalyzer and the concentration of individual libraries from the qPCR, the libraries were normalized, pooled, and loaded into the MiSeq.The MiSeq sequencer generated 250 paired-end reads with barcoding.
The raw fastq files were trimmed using BBDuk trimmer (v 1.0) (http://sourceforge. net/projects/bbmap/) at a Phred quality score ≥20. An in-house pipeline by US Army Medical Research Institute for Infectious Diseases (USAMRIID) was used to blast the trimmed reads to the NCBI non-redundant database. The best-matching sequences were selected and used as references to perform reference-based assembly. Assembly was done using Geneious Prime (v 2019.2) (www.geneious.com) while maintaining default settings with a minimum of 3 × read-depth coverage required for consensus calling.
The fasta files generated were then uploaded in the Center for Genomic Epidemiology webpage (https://cge.cbs.dtu.dk/services/ (accessed on 4 June 2019)) to determine the spa types, multi-locus sequence types, resistance, and virulence genes present. For phylogenetic analysis, whole-genome alignment was done using Mauve (v 1.1.1) [46]. The maximum likelihood phylogeny of S. aureus isolates was constructed based on the GTR GAMMA nucleotide substitution model [47] using RAxML (v 4.0) implemented in Geneious Prime (v 2020.0.4) (www.geneious.com).

Data Analysis
Demographic data and laboratory results were entered into Access 2007 and exported to STATA software version 13 for analysis. Descriptive statistical analysis was conducted and associations were determined using Fisher's exact test with significant p-values < 0.05.

Conclusions
The epidemiology of S. aureus SSIs remains a public health concern in hospitals in Ghana. Novel molecular methods were utilized in Ghana to identify and characterize multidrug-resistant S. aureus clones, including the pandemic pvl-positive ST152-t355 MRSA. These findings suggest that continuous surveillance is required to monitor the spread and resistance trends of these and other clones in hospital settings in Ghana and within the West African region.

Limitations
The study was limited by the number of S. aureus isolates, and it did not follow subjects prospectively to assess clinical outcomes. Additional data will allow an assessment of antimicrobial resistance over time and will also detect the emergence of resistant clones.  Data Availability Statement: The data used and/or analyzed during the current study are available from the corresponding author on reasonable request.