Pathogenomic analysis and molecular diagnosis of microbial keratitis

Abstract

Introduction: Microbial keratitis (MK) is a serious infection of the cornea and a leading cause of monocular blindness globally. Incidence rates of MK are increasing due to a rise in associated risk factors such as contact lens use. To guide selection of the most appropriate antimicrobial therapy and ensure the best possible visual outcome for the patient, it is critical for ophthalmologists to rapidly and accurately identify the specific causal pathogen. This is a challenging prospect due to the nonspecific symptoms shared between aetiologies and limitations of the current culture-based ‘gold-standard’ diagnostic techniques. Molecular-based techniques such as PCR and DNA sequencing can offer greater sensitivity and more information about infectious pathogens. Bacterial whole genome sequencing offers the possibility to better understand the genetic basis of pathogenicity. This thesis will assess the feasibility of molecular-based approaches to both improve the diagnostic of MK and expand knowledge of one of the most frequently isolated corneal pathogens.

Methods: Whole-genome sequencing (WGS) was performed upon a cohort of 84 clinical Staphylococcus aureus bacterial isolates using a hybrid approach incorporating both Illumina and Nanopore reads. Population analysis was performed using multilocous sequence typing, a phylogenetic tree was constructed using a multiple core genome SNP alignment, virulence and antibiotic resistance determinants were identified with the use of curated databases and a pangenome analysis was performed using a non-ocular dataset for comparison.

A targeted assay utilizing rapid LAMP technology was developed for the detection of Acanthamoeba species. Candidate gene targets for an original pan Acanthamoeba assay were determined by identifying sequences with a sufficient level of conservation using systematic pairwise BLAST alignments upon publicly available draft genomes. LAMP primers for these candidates were assessed against a previously published 18S Acanthamoeba assay, with the strongest performing assay selected for lyophilization into a final commercial format.

Full-length 16S amplicon profiling was performed upon a cohort of 61 clinical corneal specimens using Nanopore sequencing. Bacterial taxonomic classification was performed using minimap2 to align reads to the NCBI 16S database. The Phyloseq R package was used for downstream diversity analysis.

Results: WGS generated 84 high quality S. aureus draft genomes, with these high levels of completeness facilitated high resolution analysis. CC45, CC30 & CC22 were determined to be the predominant clonal lineages. Examination of a recurrent isolate subgroup revealed that relapse was caused by the exact same strain as the initial infection, with the exception of one pair which had significantly longer time between patient sampling. Virulence gene profiling defined a core set of 118 genes, including adherence factors clfA, clfB, fnbA, fnbB. Virulence gene hierarchical clustering revealed four distinct subgroups with a tendency to group according to clonal lineage. Complete genetic resistome profiles were constructed for all samples, with resistance to fluoroquinolones, chloramphenicol & gentamicin detected. Predicted resistance to the key antimicrobial classes including fluoroquinolones, chloramphenicol and gentamicin was phenotypically validated.
A lyophilized LAMP assay targeting the 18S rRNA Acanthamoeba gene was developed with a detection limit of 1pg of purified DNA & 100 plasmid target copies. 16S amplicon profiling revealed that bacteriology culture result was supported by sequencing data in only 17/37 (46%) of cases. Examination of culture negative taxonomic profiles provided insight into likely causation with several cases indicating predominance of potential keratitis pathogens.

Conclusion: In conclusion, this work has shown the potential of molecular-based techniques to enhance both pathogen detection and understanding of the genetic basis of MK. WGS of a S. aureus ocular cohort facilitated characterization of this important human pathogen in the context of corneal infection. The development of a rapid LAMP assay for Acanthamoeba detection - just one step to commercialization - adds to the current molecular toolkit with the potential to act as a diagnostic adjunct. This work has also demonstrated the feasibility of performing 16S profiling upon clinical corneal samples and highlighted its potential to become a more regular component of routine MK diagnosis. The disparity between culture results and 16S profiling underscores the issue of relying solely upon culture and highlights the need for further research into determining causative organisms. These findings contribute to the ongoing efforts to improve diagnostic strategies for MK and, in turn, therapeutic outcomes.

Thesis is embargoed until 31 December 2027.

Date of Award	Dec 2024
Original language	English
Awarding Institution	Queen's University Belfast
Sponsors	Medical Research Council
Supervisor	Gary Keating (Supervisor), Stephen Kaye (Supervisor) & David Simpson (Supervisor)