Using whole genome sequencing to accurately predict β-lactam antimicrobial susceptibility profiles of invasive Streptococcus pneumoniae isolated in Canada
| dc.contributor.author | Hink, Rachel Kathleen | |
| dc.contributor.examiningcommittee | Bay, Denice (Medical Microbiology and Infectious Diseases) Mulvey, Michael (Medical Microbiology and Infectious Diseases) Zelenitsky, Sheryl (Pharmacy) | en_US |
| dc.contributor.supervisor | Zhanel ,George (Medical Microbiology and Infectious Diseases) Adam, Heather (Medical Microbiology and Infectious Diseases) | en_US |
| dc.date.accessioned | 2020-04-03T13:08:30Z | |
| dc.date.available | 2020-04-03T13:08:30Z | |
| dc.date.copyright | 2020-03-20 | |
| dc.date.issued | 2020-03-16 | en_US |
| dc.date.submitted | 2020-03-20T11:27:29Z | en_US |
| dc.degree.discipline | Medical Microbiology and Infectious Diseases | en_US |
| dc.degree.level | Master of Science (M.Sc.) | en_US |
| dc.description.abstract | Streptococcus pneumoniae is a common etiologic agent of invasive bacterial disease worldwide. A proportion of invasive S. pneumoniae isolates demonstrate resistance to β-lactam antibiotics through the mechanism of target site modification. β-lactams are frequently used to treat invasive pneumococcal disease (IPD), therefore necessitating prompt detection of resistance. Whole-genome sequencing (WGS) potentially offers reduced time to results compared to phenotypic antimicrobial susceptibility testing (AST) methods. The purpose of this thesis was to develop three WGS-based AST models predicting S. pneumoniae β-lactam MICs. We hypothesized that WGS cannot be used exclusively in the determination of S. pneumoniae β-lactam MIC values. Invasive S. pneumoniae isolates were selected from the SAVE study for WGS. Sequencing was performed through the SAVE partnership between the Canadian Antimicrobial Resistance Alliance and the National Microbiology Laboratory. Genetic analyses included the characterization of PBP amino acid sequences, phylogeny and amino acid conservation analysis. AST of study isolates was performed in triplicate for three β-lactams of interest. Multiple linear regression models were developed and validated for the three β-lactams, generating predicted MICs that were compared to phenotypic values. Of the six pneumococcal PBPs, PBP2x, PBP2b and PBP1a were the primary focus of this study. PBP amino acid sequences were shown to be highly mutated in β-lactam-nonsusceptible isolates, whereby the level of sequence variation ranged from 0-13.8%, 0-7.1% and 0-12.4% for PBP2x, PBP2b and PBP1a, respectively. Sequence variability was correlated with serotype and the level of resistance; PBP sequence clusters were comprised of predominant serotype/MIC combinations in each of the three phylogenies. Distinct regions of variability in the 3D structures of the PBPs were observed by conservation analysis, commonly within the transpeptidase domain active site. Several known resistance-conferring substitutions were selected as predictors of MIC. Variables were weighted using forward regression, generating models for penicillin, ceftriaxone and meropenem with adjusted R2 values of 0.964, 0.936, and 0.971, respectively. All three models demonstrated EA (≥95%) and CA (≥90%) rates within CLSI acceptable performance specifications. Model rates of error were also at acceptable levels, except for VME rates (3.6%) for penicillin by CLSI IV meningitis breakpoints and mE rates (10.5%) for ceftriaxone. | en_US |
| dc.description.note | May 2020 | en_US |
| dc.identifier.uri | http://hdl.handle.net/1993/34631 | |
| dc.language.iso | eng | en_US |
| dc.rights | open access | en_US |
| dc.subject | Streptococcus pneumoniae | en_US |
| dc.subject | Invasive | en_US |
| dc.subject | Whole genome sequencing | en_US |
| dc.subject | Antibiotic resistance | en_US |
| dc.title | Using whole genome sequencing to accurately predict β-lactam antimicrobial susceptibility profiles of invasive Streptococcus pneumoniae isolated in Canada | en_US |
| dc.type | master thesis | en_US |