The role of mobile genetic elements for the transmission of antimicrobial resistance worldwide

The increasing prevalence of antimicrobial-resistant (AMR) bacteria is a major threat to global public health. It undermines the treatment of infectious diseases, thus leading to increased cost and complexities. Mobile genetic elements (MGE) have been vital for recruiting and disseminating resistance genes to pathogenic bacteria. MGEs are a diverse group of genomic elements that can transpose independently within or between DNA molecules. There are several types of MGEs with different properties. For example, some MGE types can only transpose within a bacterial cell, whilst others can conjugate between bacteria. Likewise can some carry accessory genes, while others can modulate host gene expression. The interplay of different MGEs forms a complex transposition network that can recruit and spread rapidly between bacteria. Therefore, identifying and characterizing MGEs are important to understanding AMR epidemiology.

The rapid development of massive parallel sequencing has made genomic analysis highly available, and many datasets are available in public repositories. Despite this, few studies of the global mobilome and its association with AMRs exist. A possible bottleneck limiting more researchers from analyzing MGEs could be the lack of user-friendly tools and available comprehensive MGE databases. During this PhD, we aimed to study the importance of MGEs for spreading antibiotic resistance in bacterial populations worldwide. We also wanted to create an MGE prediction tool to enable more researchers to analyze MGEs from sequence data. The three projects enclosed in this thesis are presented below.

In Manuscript I, we developed the MobileElementFinder tool for identifying MGEs in assembled sequence data. The tool includes a database that contains annotation and sequences of 6,208 known integrated MGEs. There are two versions of MobileElementFinder, an online service optimized for exploration and a local install with a command line interface used for bulk analysis. We used MobileElementFinder to study the MGE population of zoonotic Salmonella enterica from Europe and the USA using 1,725 single isolates already published in the European Nucleotide Archive (ENA). We found Salmonella to harbour a great diversity of MGEs where samples of the same lineage tended to carry similar elements. Geography
and host animals seemed to have little influence on the carried MGEs.

In Manuscript II, we studied the global population of MGEs in human-associated bacteria through metagenomic analysis of sewage. We used sewage as it approximates human gut bacteria from a geographic area. We aligned the reads to three databases, MobileElementFinder, ResFinder, and Silva, to determine the abundance of MGEs, AMRs, and bacterial taxa. We found systematic differences in the abundance of MGEs, where some elements tended to be more abundant in discrete geographic regions while others were prevalent in all continents. Some MGEs tended to be more prevalent in the temperate zone and others in the tropical zone, suggesting that climate could be a factor shaping the MGE population. We investigated MGEs whose abundance correlated with the abundance of bacterial genera, where we saw that MGEs tended to be associated with multiple closely related genera. Our findings suggest that most MGEs are more prone to be transposed to related species than distant ones, but some elements seemed to have a broader host range.

In Manuscript III, following our study of the global diversity of MGE in Manuscript II, we wanted to investigate if the observed regional differences were also present in clinical pathogens. We analyzed MGEs and ARGs in 3,087 sequenced isolates from 35 countries, representing 115 species, of which we primarily focused on Staphylococcus aureus, Enterococcus faecalis, Escherichia coli, and Klebsiella pneumoniae. We found clear species-specific differences in the number and diversity of harboured MGEs and their association with resistance genes. We found that isolates of the same lineage tended to carry similar MGEs, with no major difference between geographic regions. This was similar to the findings in Manuscript I. We found host phylogeny to be an important factor influencing the carried MGEs, confirming our observations in Manuscript II. We identified MGEs that seemed to have a broad host range with the ability to be transposed to bacteria of different phyla. Several broad host range MGEs were also associated with AMRs, suggesting that these elements could be of greater importance for disseminating AMRs.

Understanding how ARGs can be transmitted between bacteria could be important for assessing the risk of resistance genes spreading to human pathogens and hopefully be used to improve surveillance and preventive strategies. The results of this thesis provide novel insight into the dynamics of integrated MGEs and show how multiple MGEs interplay to disseminate resistance genes.