Genetic relatedness of methicillin resistant Staphylococcus aureus isolates from Steve Biko hospital

Abstract

Methicillin resistant Staphylococcus aureus (MRSA) causes severe infections in humans. An estimated two billion people carry S. aureus worldwide, of which fifty-three million harbour MRSA strains. The ability of MRSA to become multiresistant poses tremendous challenges to the healthcare system and has emerged as a major concern in the community setting worldwide. Consequently, it is important to distinguish MRSA infections because infections caused by different strains of MRSA, such as healthcare-associated MRSA (HA-MRSA) and community-associated MRSA (CA-MRSA) require different antibiotic regimens. Another concern in the hospital setting is the emergence of some of the CA-MRSA strains, which are more virulent than HA-MRSA due to the expression of the Panton-Valentine leukocidin (PVL) gene, which causes severe skin infections, sepsis and necrotising pneumonia. These current situations indicate a need for continued surveillance and infection control programmes using molecular methods to monitor MRSA strains in hospital settings so that appropriate therapy may be initiated and MRSA outbreaks may be prevented. The relatedness of MRSA isolates, using various molecular methods, has been reported in a previous study conducted in the department. However, pulsed-field gel electrophoresis (PFGE), which is the gold standard for MRSA typing, was not included. A total of 194 clinical MRSA isolates were collected from the Steve Biko Academic Hospital in the Gauteng province (South Africa) from April 2010 to August 2011 and analysed using six multiplex-PCR (M-PCR) assays. Following identification, detection and characterisation using the M-PCR assays, PFGE was used to determine the genetic relatedness of these MRSA isolates. The first M-PCR assay showed a prevalence of 100% for the 16S rRNA and mecA genes. A single isolate (0.5%) carried the PVL gene. Five staphylococcal cassette chromosome mec (SCCmec) typing M-PCR assays were used to distinguish between HA-MRSA and CA-MRSA isolates. Inconclusive results were obtained for 26.3% (51/194) of the MRSA isolates, which showed bands for either SCCmec type II or SCCmercury. The remaining 143 MRSA isolates showed a prevalence of 64.3% (92/143) for HA-MRSA and 35.7% (51/143) for CA-MRSA. The SCCmec types for the HA-MRSA isolates were: SCCmercury [74% (68/92)], SCCmec type II [19.5% (18/92)] and SCCmec type I [6.5% (6/92)]; while the SCCmec types and subtypes belonging to the CA-MRSA isolates were: subtypes IVd [92.1% (47/51)], IVa [3.9% (2/51)], IVb [2% (1/51)] and SCCmec type V [2% (1/51)]. No SCCmec type III or VIII was detected in the MRSA isolates. The PFGE typing method clustered the 191 MRSA isolates into eleven pulsotypes designated pulsotype A to K. Pulsotype A was the dominant pulsotype, including 66% (127/191) of the HA-MRSA and 19% (36/191) of the CA-MRSA isolates. Fifteen percent (28/191) of the MRSA isolates were unrelated to pulsotype A, which included 7% (13/191) of the HA-MRSA and 8% (15/191) of the CAMRSA isolates. Multiplex-PCR SCCmec typing assays and PFGE typing remain important tools for the characterisation of MRSA strains. A standardised SCCmec M-PCR assay can provide more accurate and reliable results. The results indicated that the HA-MRSA and CA-MRSA strains analysed in this study were closely related in this hospital setting, which necessitates continuous monitoring and surveillance to ensure and guide infection control policies.