Molecular Characterization of Community-Associated Methicillin-Resistant Staphylococcus aureus in Iranian Burn Patients

Background & Objective: Methicillin-resistant Staphylococcus aureus (MRSA) is reported as one of the important bacterial causes of burn wound infections. This study was carried out to investigate molecular characterization of community-associated MRSA (CA-MRSA) isolated from Iranian burn patients. Methods: A total of 31 isolates of S. aureus were collected from the Motahari Burns Hospital (Tehran, Iran) in 2016. All isolates were collected from outpatients and inpatients within 48 hours of admission. The mecA, pvl, tsst-1, hla-α, and psmα genes detecting, SCCmec, agr and PFGE typing were done. Results: A total of 13 (41.9%) isolates were cefoxitin-resistant and mecA-positive, which were considered as MRSA. The SCCmec typing MRSA strains revealed type II in 1 (7.7%), type III in 9 (69.2%), and other types in 3 isolates (23.7%) cases. The agr typing of all 31 isolates showed that 14 (45.2%), 1 (3.2%), 6 (19.4%), and 10 (32.3%) strains belonged to agr groups 1, 3, 4, and unknown type, respectively. The pvl, tsst-1, hla-α, and psmα genes were positive in 3 (9.7%), 4 (12.9%), 21 (67.7%), and 31 (100%) isolates, respectively. Considering the cut-off values of ≥50%, 3 groups of related isolates (cluster A1, B1, and C1) in PFGE study were observed. Conclusion: The MRSA strains of this study were initially isolated as Community-associated S. aureus (CA-MRSA); however molecular characterization showed that a significant proportion of them had hospital-associated MRSA (HA-MRSA) features. Therefore, it is likely that the HA-MRSA strains are spread among the community.


Introduction
Due to the skin damaged and weakened innate immunity, burn patients become susceptible to various infections (1, 2), among which bacterial infections are the main problems in hospitalized burn patients (3,4).
Staphylococcus aureus is one of the most common bacteria isolated from burn infections (5). Methicillinresistant S. aureus (MRSA) including hospitalassociated MRSA (HA-MRSA) (3,6) and communityassociated MRSA (CA-MRSA) (7) are reported as the important bacterial causes of burn wound infections. The centers for disease control (CDC) defined a CA-MRSA infection as any MRSA infection diagnosed for an outpatient or within 48 hours of hospitalization if the patient lacks such healthcare-associated MRSA risk factors as dialysis or use of invasive medical devices, surgery, and residence in a long-term care facility or hospitalization during the previous year; all other MRSA infections were considered to be HA-MRSA (8)(9)(10).
Cefoxitin is a potential inducer of the mecA regulatory system and is used for screening MRSA strains (11). Staphylococcal cassette chromosome mec (SCCmec) element of MRSA strains carry mecA gene that mediates resistance to methicillin (10,12). HA-MRSA strains carry large SCCmec types II and III, which are responsible for resistance to other antibiotics. However, SCCmec types IV and V elements are found in CA-MRSA strains (10,13).
It should be noted that there are several problems in terms of using CA-MRSA and HA-MRSA. First, both strains can be circulated in the community; second, it is possible that CA-MRSA strains are gradually converted to nosocomial pathogens (10,14).
These virulence factors can make extensive infection and severe damage (7,12,15,16). The agr regulatory system has a crucial role in the development of skin infection cases by most CA-MRSA USA300 strains (17). The PVL toxin can be found in both MSSA and MRSA strains (18) but typically is associated with CA-MRSA strains (7). TSST-1 is a superantigen that is absorbed from cutaneous infections and it localizes the kidneys (19). It is reported that the pore-forming αhemolysin of S. aureus plays a key role in soft tissue infection (20). The psmα may play a role in biofilms development and lysis of human cell types such as leukocytes and erythrocytes; also, it stimulates inflammation response (21).
Phenotypic and genotypic characterization of CA-MRSA and HA-MRSA can be helpful in the treatment, infection control, and prevention of the spread of antibiotic-resistant bacterial strains (1,22,23). This study was carried out to investigate molecular characterization of CA-MRSA isolated from Iranian burn outpatients and hospitalized patients within 48 hours of admission.

Bacterial Isolates
A total of 31 isolates of S. aureus were collected from Shahid Motahari Burns Hospital (Tehran, Iran) in the time period of May to October 2016. All isolates were collected from outpatients and inpatients within 48 hours of admission. Bacterial morphology study and biochemical tests including catalase, tube and slide coagulase, DNase, and mannitol-fermenting were done to identify S. aureus. After identification, all strains were preserved in trypticase soy broth medium containing 15% glycerol at -70°C.

Molecular Detection and Typing DNA Extraction
Four to five fresh colonies of bacteria were suspended in 400 μL of lysis solution buffer containing lysostaphin (200μg/mL) (Sigma-Aldrich Co, Germany). Then, DNA was extracted by phenol/chloroform /isoamyl alcohol (Merck Co).

PCR Amplification
The mecA was used as target gene for the detection of MRSA strains (24). The primers described by Zhang et al.
(24) and Shopsin et al. (25) were used for SCCmec typing and agr typing, respectively. RN6390, RN6607, RN8465, and RN4550 strains were used as positive control for agr typing. The pvl, tsst-1, hla-α (26), and psmα (27) genes were used as target gene for virulotyping. The list of primers sequences of this study, target genes, and the length of PCR products are presented in Table 1.

Ref
Pulsed-field Gel Electrophoresis (PFGE) PFGE analysis based on SmaI restriction enzyme (Takara) digestion was carried out using the PulseNet standard protocol on 13 SCCmec-positive S. aureus strains (PulseNet, www.cdc.gov/pulsenet). After optimizing conditions, PFGE was run on CHEF MAPPER XA (BIO RAD) device as follows: initial switch 5 s, final switch 40 s, run time 21 h, voltage 200 (6v/cm). 1% agarose gel (intron biotechnology) in TBE 0.5 X was used for electrophoresis. According to the PulseNet protocol, Salmonella enterica serotype Braenderup (H9812) was used as a control. After gel staining with ethidium bromide (10 mg/mL), imaging was performed by UV light gel documentation system (BioDocAnalyze). GelCompar II version 6.6.11 software was used for drawing dendrogram graph and interpretation of results. Cut-off values of ≥50% were considered to define clusters.

Patients
A total of 31 non-duplicate S. aureus isolates were investigated in this study. All isolates were collected from burn patients within 48 hours of admission. 20 (64.5%) of these patients were inpatients and 11 (35.5%) outpatients.
According to the antibiotic susceptibility testing of all 31 strains, the highest resistance was related to erythromycin (48.4%) compared to 58.2% (2), 77.8% (30), and 90.6% (16) in Iran. All the 13 MRSA strains in this study were sensitive to linezolid, which was similar to the findings of the studies conducted by Montazeri (31,32), and Goudarzi (16) from Iran and Murray (7) from the US.
Based on the CDC definition of CA-MRSA and HA-MRSA strains (8)(9)(10) and according to other studies (10,13), it seems that most of these SCCmec types belonged to HA-MRSA strains. Since sampling was done within 48 hours after admission, the strains were expected to be CA-MRSA. According to a hypothesis, this has been affected by various factors such as having the same source of infection or spread of infected patients with HA-MRSA strains among the community. Since checking the patients' referral records during the last year was not possible in this study, it is possible that most of these strains were HA-MRSA. Furthermore, having high relative resistance to selected antibiotics including clindamycin in most strains also confirms this probability. Eight (61.5%) of MRSA strains in this study and 75.5% in Goudarzi's study (16) were resistant to clindamycin, which indicates these strains had high resistance to other antibiotics. Also, 5 (38.5%) clindamycin-susceptible strains in this study had a lower resistance to other antibiotics. Therefore, it can be concluded that 8 (61.5%) of these strains correspond to HA-MRSA and 5 (38.5%) to CA-MRSA.
The PFGE cluster results with cut-off values of ≥50% indicated that MRSA strains in this study, to some extent, had CA-MRSA characteristics. Also, the 3 strains in clusters A1, 3 in B1, and 2 in C1 have the same pattern. This similarity suggests that the source of infection has been the same in each cluster.
The PVL toxin typically is associated with CA-MRSA strains (7). In the present study, 3 (9.7%) MRSA strains were pvl-positive. In contrast to the study by Goudarzi that reported the positive pvl strains belonged to SCCmec type IV (16), in this study 2 of 3 strains belonged to the SCCmec type III. Although all MSSA strains in this study were pvl-negative, some other studies in Iran reported that 7.23% (18), 15.1% (16), and 33.3% (18) of MSSA cases were positive.
The agr typing assay of all 31 isolates showed that the highest rate belonged to agr type 1 (45.1%) compared to 84.9% in Iran (16) and 96.5% in China (3); meanwhile, the lowest rate belonged to agr type 3 (3.2%) compared to 15.1% in Iran (16). In this study, 32.3% strains were not typeable, which might be related to the presence of different mutations, sensitivity, and specificity of primers.
The tsst-1 gene detection showed that 12.9% of the isolates in this study belonged to MSSA strains, but this gene was not seen in MRSA strains. In contrast to this study, in a different study in Iran 18.9% of MRSA isolates were tsst-positive which were distributed among SCCmec type III (16).
In the present study, 67.7% strains were hla-positive compared to another study in Iran which reported 51.8% (15). Data analysis showed that 32.2% belonged to MRSA and 35.5% to MSSA.
Our results showed that 41.9% of isolates were MRSA, which indicates the presence of antibiotic resistance elements in a significant number of strains and necessity of performing antibiogram before starting treatment. Furthermore, the MRSA strains of this study were initially isolated as CA-MRSA, however molecular characterization showed that a significant proportion of them had HA-MRSA features. Therefore, it is likely that the HA-MRSA strains are spread among the community. Finally, the investigation of virulence factors in this study showed the presence of various toxin genes in isolates that can enhance their ecological niche ability and also make treatment more complicated, especially in burn patients.