Evaluating whole-genome sequencing strategies in AMR research through characterisation of mobile genetic elements in wildlife-origin Escherichia coli

Background Resistance against antimicrobials critically important to human medicine has been frequently observed in livestock and clinical settings where the selection pressure imposed by antimicrobial use is high. Today, it is detected in wildlife even in the absence of a direct selection pressure....

Full description

Bibliographic Details
Main Author: Gunasekera, Samantha Thilini
Other Authors: O'Dea, Mark, Abraham, Sam, Wang, Penghao
Format: Thesis
Language:English
Published: 2019
Subjects:
Online Access:https://researchrepository.murdoch.edu.au/id/eprint/50510/
Description
Summary:Background Resistance against antimicrobials critically important to human medicine has been frequently observed in livestock and clinical settings where the selection pressure imposed by antimicrobial use is high. Today, it is detected in wildlife even in the absence of a direct selection pressure. On Penguin Island, 660 m from the coast of Western Australia, Escherichia coli isolated from little penguin and feral pigeon faecal samples (n = 20) was found to be phenotypically resistant to extended-spectrum cephalosporins and fluoroquinolones. It was hypothesised that the frequent observations of resistance against CIAs in this group of birds were a result of horizontal gene transfer. However, mobile genetic elements have historically been challenging to characterise using second-generation sequencing technology alone. These difficulties have been linked to coverage bias caused by library preparation, and to insufficient read length to assemble long, repetitive genomic regions. To address the limitations of read length, second-generation sequencing technologies were used in conjunction with third-generation sequencing technologies to produce more contiguous assemblies while maintaining accuracy. To address coverage bias, the performance of the Nextera XT and Nextera Flex library preparation kits were evaluated to assess whether either library preparation kit was associated with reduced sequencing bias. Results The results of the library preparation kit comparison found that the Nextera Flex library preparation kit produced more even coverage than the Nextera XT kit, however tagmentation bias, GC content bias, de novo assembly quality and antimicrobial resistance (AMR) gene detection was either unchanged or poorer than Nextera XT. Resistance to extended-spectrum cephalosporins observed in the E. coli isolates was mediated mainly by blaCTX-M-15 and was shown to have circulated through the different isolates via mobile genetic elements such as ISEcp1, IS26 and Tn2. Resistance against fluoroquinolones occurred primarily through mutations in quinolone resistance-determining regions, with only a minority of isolates harbouring plasmid-mediated quinolone resistance genes. Using a combination of second and third generation sequencing technology, a composite transposon conferring multidrug-resistance to macrolides, folate pathway inhibitors and aminoglycosides was found in varying conformations in over half of the E. coli isolates. Conclusions This study was the first to assess the Nextera Flex library preparation kit in the context of AMR research and found that coverage bias was markedly improved. However, this did not impact the practical applications of whole-genome sequencing data such as de novo assembly or AMR gene detection. Using a combination of second and third-generation sequencing technologies, this study established the presence of multiple mobile genetic elements conferring resistance to critically important antimicrobials in E. coli isolated from avian wildlife on Penguin Island, Western Australia.