Using genomic epidemiology to improve molecular diagnostics for Streptococcus pneumoniae and to study streptococcal population structure

Thesis

Various Streptococcus species, including Streptococcus pneumoniae (pneumococcus), can asymptomatically colonise the human upper respiratory tract. Pneumococci can also cause life-threatening infections such as meningitis and pneumonia, constituting a large burden on global health. Using large bacterial genome datasets, this thesis aimed to improve nucleic acid amplification tests used to identify pneumococci and to enhance our understanding of the streptococcal population structure.

First, I assessed the in silico performance of existing PCR-based diagnostics in 7,547 pneumococcal genomes and 1,825 genomes of 55 non-pneumococcal Streptococcus (NPS) species. I designed a new multiplex quantitative PCR (qPCR) assay targeting Xisco and SP2020 and validated this assay in vitro, finding high specificity, sensitivity, and accurate performance across a range of pneumococcal amplicons.

Building on these results, I assessed the in silico performance of existing pneumococcal loop-mediated isothermal amplification (LAMP) assays, revealing cross-reactivity with up to ten NPS species. I designed and validated in vitro a highly sensitive, specific and fast new real-time LAMP assay targeting SP2020, which could be particularly useful in low-resource settings.

Finally, I compiled a diverse Streptococcus genome dataset to investigate in silico differences across 105 species, including guanine-cytosine content and total assembly length. Using two different methods, I studied the size and function of the streptococcal core genome. Whilst seven single locus typing targets were unable to accurately stratify S. pneumoniae from closely related viridans streptococci, the phylogeny based on 412 streptococcal core genes provided a robust framework for mapping the in silico presence of antimicrobial resistance genes and virulence factors across the genus.

This thesis harnessed the power of genomic data to improve pneumococcal diagnostic assays and to study differences between streptococcal species. The new multiplex qPCR and real-time LAMP assays can facilitate improved pneumococcal identification in clinical and research microbiology laboratories and the in silico plus in vitro methodology can be applied to improve diagnostics for other pathogens. The insights gained into streptococcal population structure contribute to an enhanced understanding of streptococcal evolution, epidemiology and virulence factor distribution.

Authors: Ahlers, FM

• DOI: 10.5287/ora-8g0pyyedy
• Type of award: DPhil
• Level of award: Doctoral
• Awarding institution: University of Oxford
• Language: English
• Keywords: population health; bacterial genomics; epidemiology; infectious diseases; in vitro diagnostics
• Subjects: Public health surveillance; Genomics; Communicable diseases; Bacterial genetics; Epidemiology