Abstract
Methicillin-resistant S. aureus (MRSA) are a major cause of nosocomial infections requiring accurate and rapid molecular techniques for epidemiological analysis. Increased genomic sequence information has resulted in the development of several sequence-based epidemiological approaches. However, current efforts, such as multi-locus sequence typing (MLST), have been found to be insufficiently sensitive for hospital epidemiology of MRSA. In addition, direct sequence analysis of multiple loci is associated with issues of cost, time, and complexity of data analysis. Therefore, the analysis, either directly or indirectly, of fewer genomic sequences without sacrificing epidemiological discrimination is a desirable goal. It has been hypothesized that the Staphylococcal genome contains sequences that mutate at a rate which would provide epidemiologically-relevant information. The goal of this thesis was to identify and evaluate target sequences and techniques for use in sequence-based epidemiology. Existing Staphylococcal sequence databases were queried for the recognition of hypervariable genetic regions, which resulted in the identification of target sequences within yo/C, gyrA, qgrB, ISR, and r/rn. The genetic variability and discrimination contained within each of these genetic regions was assessed by analysis of a panel of 20 previously characterized MRS A strains. While none of the target sequences were as discriminatory as the current molecular epidemiological 'gold standard' pulsed-field gel electrophoresis (PFGE), several of the regions, alone or in combination, produced similar or slightly better results than MLST. In addition, the analysis of these target sequences was investigated against several mutation detection assays performed on the Transgenomics WAVE or the Nanogen* electronic microarray platform. While not applicable to all of the genetic sequences, the mutation detection methods investigated were found to be relatively rapid, accurate, and easy to perform. Finally, the application of selected target sequences and assays, was found to be unable to differentiate many of the unrelated isolates from a panel of 9 clinical MRSA isolates. Since the mutation detection assays were found to be quite accurate, this lack of discrimination is attributed to insufficient variability within the target sequences. The results of this study suggest that epidemiologically-relevant target sequences remain to be optimized and that alternative mutation detection assays represent a potential platform for efficient and cost- effective sequence-based epidemiological analysis.