Diversity of SCCmec Elements in Methicillin-Resistant Staphylococcus Aureus (MRSA) Recovered from a Healthy Student Population in Kenya

Background: Methicillin-resistant S. aureus continues to be a concern for public health systems, particularly due to infections emerging in non-hospital settings. Resistance to methicillin is presently classied as a serious phenomenon because the majority of methicillin-resistant strains are also multi-drug resistant. The genetic determinant of resistance to methicillin and other β-lactam antibiotics is the mec-A gene, which lies in the SCCmec resistance island. In Kenya, studies done previously have shown the existence of SCCmec types in clinical isolates, but similar information on isolates recovered from healthy populations is scanty. Methods: A cross-sectional study was conducted on healthy university students residing within the university residence halls to determine the carriage of MRSA. MRSA was detected using Cefoxitin (30µg), and mec-A gene and Sccmec elements were detected using conventional PCR methods. A total of 237 students were recruited, and 657 swabs were collected using standards methods for recovering S. aureus. Results: A total of 231 S. aureus isolates were recovered, out of which 26 (11.3%) were MRSA. Out of the 26 MRSA strains, 17 carried the mecA gene in their gene cassettes. SCCmecV was the most prevalent (61.5%), followed by SCCmecII (53.9%) among the MRSA strains. SCCmecIVa, SCCmecIVb, SCCmecIVc and SCCmecIVd were absent in all the isolates. SCCmecV was found to be highly prevalent (64.7%) followed by SCCmecII, 8 (47.1%) among the mecA-positive MRSA strains. On the other hand, small proportions of mecA-negative isolates harbored SCCmecI (0.9%), SCCmecII (3.3%), SCCmec III (0.5%) and SCCmecV (2.3%). study strains from

not to be native to S. aureus, but rather, it has been acquired via the mecA gene for more than 40 years. The mecA gene encodes a protein known as penicillin-binding protein (PBP), which is designated as PBP2a. S. aureus produces four PBPs, namely, PBP1, PBP2 PBP3 and PBP4 that are anchored on the cytoplasmic membrane (Navratna et al., 2010). Penicillin-binding proteins function in the assembly and regulation of stages of the synthesis of a bacterial cell wall. Whereas the four PBPs are susceptible to alteration by β-lactam antibiotics resulting in the death of bacterial cells, PBB2a is refractory to the action of all presently used β-lactam antibiotics. The PBP2a can take over the functions of the four staphylococcal PBPs during exposure to β-lactam antibiotics (Kondo et al., 2007).
The mecA gene is known to be carried on a peculiar type of mobile genetic element inserted into the staphylococcal chromosome, known as the staphylococcal cassette chromosome mec (SCCmec) element (Katayama et al., 2000). The SCCmec elements share four characteristics. First, they carry the mec gene complex, which consists of a methicillin-resistance determinant gene (mecA), its regulatory genes and insertion sequences; second, they carry the ccr gene complex responsible for the mobility of the element and its associated sequences; third, they have distinct directly repeated nucleotide sequences and inverted complementary sequences at each end; and fourth, they integrate into the 3′ end of an open reading frame (ORF), orfX. Despite the similarities identi ed above, structures of SCCmec elements are also divergent (Kondo et al., 2007).
In Kenya, the few studies which sort to examine the diversities of SCCmec elements have revealed the existence of different types of SCCmec elements. In a recent study that analyzed isolates obtained from a government hospital and a private referral hospital in Nairobi County, SCCmecIII was found to be more prevalent, followed by SCCmecIV. The predominance of these two SCCmec elements was attributed to their small sizes, which facilitated their transmission in both community and hospital settings (Omuse et al., 2016). Another study that focused on isolates recovered from a government hospital setting also revealed the existence of SCCmecIII among all the six MRSA isolates obtained from inpatient screens (Aiken et al., 2014). In a study done by Maina et al., (2013)

Recruitment of participants
Participants were recruited from the hostels. Students who resided at the university and agreed to provide written consent for participation by providing samples and lling the questionnaire were included in the study. Students who were on antibiotics and those who consented, but did not provide a nasal swab were excluded from the study.
This study was conducted in accordance with the regulations of the scienti c and ethics review unit (SERU) of the Kenya Medical Research Institute (KEMRI) (Protocol No. KEMRI/SERU/CMR/0024/3123).

Study design
This cross-sectional study was conducted in 2016 on healthy students residing in a university's residence halls. A systematic random sampling procedure was used to select the students' hostel rooms where all occupants were recruited. A total of 237 healthy students were included in the study.

Collection of swabs
Swabs were collected from the nostrils, mobile phones, and pens used by participants using sterile swab sticks pre-moistened with sterile normal saline 0.85% (NaCl). All swabs were put in Amies transport media for transportation to the laboratory within a maximum of 4 hours. For enrichment, the swabs were transferred to trypticase soy broth (TSB) and incubated for 18 to 24 hours (Coia et al., 2006). Isolation and identi cation of S. aureus isolates A loopful of inoculum from the TSB cultures was inoculated in on Mannitol Salt Agar (Oxoid) plates and incubated aerobically at 37°C for 24 hours. Standard methods of identifying S. aureus, as described by Baron (1996), were employed. The plates were assessed for colony morphology and yellow pigment formation associated with S. aureus. Suspect colonies were subjected to Gram staining, coagulase, and catalase tests. Spa typing was also done for con rmation of all S. aureus isolates.

Extraction of DNA
The extraction of DNA was done using a 10% Chelex solution prepared using 1X TE buffer following a modi cation of the procedure used by (HwangBo et al., 2010). A 300µl volume of 10% Chelex solution was added to fresh sterile Eppendorf tubes. To each Eppendorf tube lled with 300µl volume of 10% Chelex solution, a loopful of 24 hours old bacterial colonies were emulsi ed. The Chelex/sample tubes were then vortexed for 5-10 seconds. The tubes were then transferred to a heating block set at 95 ºC for 20 minutes. After boiling for 20 minutes, the tubes were left to cool for 10 minutes before centrifugation at 14700 rpm for 10 minutes. After that, 50µl of the supernatant was transferred to fresh Eppendorf tubes pre lled with 450µl of RNase DNase free PCR water. All the tubes were labeled carefully using the respective isolate codes.

Spa typing
Ampli cation and detection of mecA gene were done on all S. aureus isolates using conventional PCR methods. The primers used are listed in Table 1. In every PCR tube, 4µl of the ready to mix 5x FIREPol ® Master Mix (Solis Biodyne, UK), 0.4µl of both forward and reverse primers at a concentration of 10pmol of each primer, an aliquot of 1µl of BSA, 12.2 µl of RNase DNase free PCR water and 2µl of sample DNA was added. A single control strain of S. aureus was incorporated in the DNA ampli cation process. PCR reactions were set for 5 minutes at 80°C; 35 cycles of 45 seconds at 94°C, 45 seconds at 60°C, and 90 seconds at 72°C; a single nal extension at 72°C for 10 minutes and a holding temperature at 4°C. The excepted band sizes were to be variable (Strommenger et al., 2008).

Ampli cation and detection of mecA gene
Ampli cation and detection of mecA gene were done on all S. aureus isolates using conventional PCR methods. The primers used are listed in Table 1. In every PCR tube, 4µl of the ready to mix 5x FIREPol ® Master Mix (Solis Biodyne, UK), 0.4µl of both forward and reverse primers, 1µl of BSA, 12.2µl of RNase DNase free PCR water and 2µl of sample DNA was added. Ampli cation reactions were set at an initial denaturation temperature of 95 ºC for 5 minutes. After that, the preparation was subjected to 30 cycles of denaturation at 94 ºC for 60 seconds, annealing at 56 ºC for 60 seconds, extension at 72 ºC for 60 seconds and a nal extension at 72 ºC for 7 minutes. A holding temperature was set at 4°C. One control strain was incorporated into the DNA ampli cation process. The excepted PCR product encoding for mecA gene was 533 bp in size (Kondo et al., 2013).

SCCmec typing
SCCmec typing was done for all isolates using PCR. Primers speci c for SCCmecI, SCCmecII, SCCmecIII, SCCmecIVa, SCCmecIVb, SCCmecIVc, SCCmecIVd and SCCmecV elements were used as published previously (Zhang et al., 2005). The primer sequences are provided in Table 1. PCR procedures were similar to those for the mecA gene, but the annealing temperature was set at 65°C for 45 seconds. A control strain of S. aureus known to be positive for the gene was incorporated in all the PCR sets.  (20,76.9%), followed by mobile phones (7, 26.9%) and pens (5,19.2%). In several cases, MRSA was also found to be harbored in the nostrils, mobile phones and pens used by the same student at the same time, as shown in Table 2. Antimicrobial resistance phenotypes All 231 isolates of S. aureus were tested against a panel of 11 antimicrobials. Resistance to ampicillin was very high (84%). Lower resistances of less than 20% were recorded on other isolates. Precisely, resistances to amoxicillin, chloramphenicol, gentamicin, cipro oxacin, nitrofurantoin and nor oxacin were below 10%. None of the isolates was resistant to linezolid, and only a single isolate showed intermediate resistance to nitrofurantoin, as shown in Figure 1.
Following the screening of the 231 isolates for antimicrobial resistance, 26 (11.3%) were found to be MRSA, while 16 (9.9%) were MDRs. Out of the 26 MRSA strains, 13 (50%) were also MDRs. All the MRSA and MDR strains showed total resistance to ampicillin. Resistances above 40% were observed for erythromycin, nor oxacin and trimethoprim-sulfamethoxazole among the MRSAs. Similar observations were made for the MDRs, which also showed over 40% resistance levels to cipro oxacin (50%) and nor oxacin (75%). For both MRSA and MDRs, less than 20% of resistance levels were recorded for amoxicillin-clavulanic acid, chloramphenicol and gentamicin, as shown in Figure 2. Statistical analysis revealed that MRSA strains were more likely to display multi-drug resistance phenotypes compared to the MSSA strains (P-value = 0.001).

Multiple SCCmec elements
This study revealed four different combinations, each consisting of two distinct SCCmec elements. These included; SCCmecI and SCCmecIII, SCCmecI and SCCmecV, SCCmecII and SCCmecV and SCCmecII and SCCmecV. Almost half of the MRSA strains, 11 (42.3%), carried SCCmecII and SCCmecV compared to those that harbored SCCmecI and SCCmecV, 3 (11.5%), SCCmecI and SCCmecIII, 2 (7.7%), and SCCmecIII and SCCmecV, 1 (3.9%). Only two combinations of the SCCmec elements were seen among the MDR strains, where 8 (50%) of these strains carried SCCmecII and SCCmecV, and only a single MDR varied SCCmecI and SCCmecV. See Table 6 below. Carriage of S. aureus in a population that is otherwise considered as healthy has shown global variation. Previous studies have revealed colonization prevalence ranging between 20%-40%, which is reported as the prevalence limit in literature, in different study populations (Chambers & Deleo, 2009;Sivaraman et al., 2009). In this study, the prevalence of nasal, phone and pen colonization fell within the range of 20-40%.
Similar results have also been reported by other studies which focused on college students in Tanzania  (Okamo et al In the present study, a prevalence of MRSA (11.3%) was found among a healthy population of university students aged between 18-30 years in Central Kenya. The MRSA strains were detected using cefoxitin (30 μg) discs, which, according to Rasheed & Ahmed (2010), is an alternative method to PCR in the detection of MRSA strains. So far in Kenya, this is the rst study to present results on the prevalence of MRSA among healthy university students residing within institution-based residence halls. Other recent studies like (Zakai, 2015) have reported a higher MRSA prevalence (

Antimicrobial resistance phenotypes
This study established that MRSA strains were more likely to display multi-drug resistance compared to MSSA strains, a nding which has also been reported by Gupta et al. (2013) in India. Among the published studies done in Kenya (all of which focused on clinical isolates), none reported this property among the MRSA strains. Therefore, this study highlights crucial ndings regarding the circulation of multi-drug resistant MRSA and MSSA strains among a healthy student population. Given the fact that S. aureus is transmitted via direct or indirect person-to-person contact, there is the possibility of an increased prevalence of multi-drug resistant MRSA and MSSA strains among the student population, which is a worrisome scenario. It is also worth noting that the multi-drug resistance property renders multi-drug resistant strains as one of the most intractable pathogenic organisms in the history of antibiotic therapy. Therefore, the circulation of these multi-drug resistant S. aureus among students presents a challenge to the management of Staphylococcal infections that may develop, particularly in regards to acquiring effective antibiotics, which may be expensive.
Carriage of mecA gene Molecular analysis in this study showed that the S. aureus isolates were genetically diverse. A majority of the MRSA strains carried the mecA gene, while none of the MSSA strains carried this gene (Table 4-6).
Among the studies done in Kenya, carriage of the mecA gene has been reported among the MRSA strains.
In their study, Maina et al. (2013) reported that all the MRSA strains harbored the mecA gene, and similar to this study, none of the MSSA strains were found to harbor this gene. In a recent study conducted in Zambia, all the MRSA were found to harbor the mecA gene (Samutela et al., 2017). Similar to this study also, a majority of MRSA (90.2%) strains carried mecA gene, while 9.8% were negative for the gene in a study conducted in Sudan (Elhassan etal., 2015). According to available literature, the presence of the mecA gene forms the basis of resistance to beta-lactam antibiotics (Murakami et al., 1991;Sharma et al., 1998;Wongwanich et al., 2000). Therefore, basing on suggestions by the majority of researchers in this eld, the mecA gene was linked to cefoxitin resistance among the mecA-positive isolates recovered in this study. Other genetic elements may also be considered for the explanation of cefoxitin-resistance mechanisms (Elhassan et al., 2015). Hence, the basis of cefoxitin-resistance among the mecA-negative MRSA isolates characterized in this study could be attributed to two possible reasons. Firstly, the hyperproduction of β-lactamases, as described by Olanyika et al., (2009) and secondly, development of speci c alterations in the variable amino acids found in the PBPs cascades as highlighted by (Ba et al., 2014).

Diversity of SCCmec elements
Screening of all the S. aureus isolates for the presence of SCCmec elements revealed the existence of SCCmecV as the most prevalent mec type, followed closely by SCCmecII, then SCCmecI and lastly, SCCmecIII, which was only present in three isolates. Contrary to the clinical isolates recovered in Kenya, the non-clinical isolates examined in this study did not harbor any of the SCCmecIV types. SCCmecV has been shown to be of small molecular size, and this could be the possible reason for its high prevalence among isolates recovered in this study (Omuse et al., 2016). No other published study done in Kenya with similar ndings was found. Previously, SCCmecIV has been reported to circulate highly in community settings, possibly because of its small size, which renders it more mobile and thus can be inserted into multiple S. aureus lineages (Omuse et al., 2016;Tong et al., 2012). It was unclear why none of the isolates recovered in this study harbored SCCmecIV. More studies could also be done to ascertain the prevalence of the existence of this combination of SCCmec elements among MRSA strains that are also MDRs.

Variants of SCCmec elements
In the present study, two MSSA strains were also found to carry SCCmecI. Additionally, among the mecA-

Conclusions
In summary, the present study demonstrated the existence of varied SCCmec types among S. aureus strains circulating in the studied student population. The SCCmecV was the most prevalent. Some strains carried multiple SCCmec elements in the gene cassettes, while other strains lacked the mecA gene but still carried the SCCmec elements in their gene cassettes.

Consent for publication
Authors have approved the review and publication of this manuscript.

Availability of data and materials
Data collected in this study is available for perusal

Competing interests
Authors have to competing interests to declare.

Funding
This work was funded partly by the Social Ecology of Human Infectious Diseases Initiative (ESEI), Grant Reference: G1100783/1.

Authors' contributions
Osborn Khasabuli performed the laboratory work and wrote the manuscript. Dr. Caroline Ngugi and Dr. John Kiiru provided expert advice in designing the study and manuscript writing process.