AbstractDevice-related infection is a major complication of orthopaedic surgery and treatment can be complex and costly, requiring both antibiotic administration and surgery. Due to its ubiquity on the human skin Staphylococcus epidermidis is a common cause of orthopaedic device-related infection (ODRI). Confusingly, for the same reason, S. epidermidisis also a common contaminant of microbiological specimens and cultures. The culture of indistinguishable isolates from multiple specimens of peri-prosthetic tissue or fluid is a diagnostic indicator of ODRI. However, determining the relatedness of S. epidermidis isolates is difficult using conventional microbiological methods. Whole-genome sequencing provides a definitive measure of the relatedness of bacterial isolates and has the potential to be used to discriminate between S. epidermidis ODRI and contamination. Furthermore, the same assay allows detection of genetic elements associated with antimicrobial resistance and can be used to guide antibiotic therapy. This thesis aims to enhance our understanding of S. epidermidis diversity and provide the basis for a tool that can aid ODRI diagnosis and clinical management. This thesis is comprised of three chapters studying the genomics of S. epidermidis relating to population structure, the population dynamics of infection and antimicrobial resistance. The first study characterises the population structure of 192 colonies of S. epidermidis isolated from carriage sites of five healthy individuals. Multi-locus sequence typing was performed and compared with the genomic method of pairwise distance measurements. Pairwise distance measurements between colonies of different sequence types and from different individuals were used to define subtypes. The phylogeny of all isolates was established using ClonalFrameML and compared with sequence type and subtype data. A threshold of 100 single nucleotide variants (SNVs) was determined as a predictor of relatedness with a positive predictive value of 95%. Finally, the application of MLST, maximum-likelihood method, and subtyping were compared to characterise the population structure of the isolates.
The second study applied pairwise distance measurements to discriminate between 325 clinically defined infecting and contaminating isolates of S. epidermidis from peri-prosthetic specimen cultures. Haplotrees were then constructed to visualise the population structure of S. epidermidis associated with ODRI and specimen contamination. This study showed that population structures of S. epidermidis ODRI and S. epidermidis contaminating strains differ in ways that can be measured and visualised.
The third study genes and mutations associated with resistance to 13 antimicrobials were compiled into FASTA files. The FASTA files were used with the Basic Local Alignment Search Tool to interrogate the genomes of 87 isolates of S. epidermidis. Results were compared with conventional antimicrobial susceptibility testing methods. Discordant results were further investigated with minimum inhibitory concentration testing, biochemical testing and interrogation of resistance genes. The genotyping methods yielded a sensitivity of 100% and specificity of 98% in correlating with the phenotypic methods. These three studies demonstrate that whole-genome sequencing (WGS) can be used to enhance our understanding of bacterial populations in normal human carriage, infection and contamination. The methods used could be developed to provide a tool to aid diagnosis of S. epidermidis infections whilst also giving accurate information regarding the antimicrobial resistance of clinically important isolates.
|Date of Award||2022|
|Supervisor||Martin Llewelyn (Supervisor) & Dr Daniel J Wilson (Supervisor)|