dc.description.abstract
The presence of resistance determinants in livestock and food and the possible AMR dissemination is a global threat. It can result in treatment failure when trying to treat infections caused by no longer susceptible microorganisms. Especially, resistance development and spread of resistance to highest priority critically important antimicrobial agents as quinolones is highly undesirable. The pentapeptide repeat proteins encoded by qnr genes is a PMQR leading to an increased MIC against quinolones and fluoroquinolones. Tackling the spread of PMQRs needs in-depth investigation of its prevalence, of the vector characteristics and of the main dissemination paths. Therefore, the establishing of appropriate protocols for characterizing the plasmids carrying the PMQR as qnr with e.g. different sequencing techniques is highly desirable. The OHEJP-ARDIG project targets the international and integrative examination of the topics evolved around resistance development. Located within ARDIG, this thesis aims to understand the prevalence and characteristics of quinolone- and fluoroquinolone-resistant commensal E. coli and their mobile genetic elements. A focus lies on qnr-carrying plasmids, and their characterization, using an optimized sequencing and assembling approach.
Therefore, we used different sequencing and assembling strategies for assessing their reliability for AMR monitoring in commensal E. coli. Isolates were subjected to WGS with Illumina NextSeq, PacBio and ONT for an in-depth characterization of their plasmid content. We further assembled the generated raw reads with different techniques, including long-read only, short-read only and hybrid approach. The established data was compared for validity with data from laboratory-generated experiments. We found long-read sequencing resulting in error prone prediction of the plasmid genome, while short-read sequencing was rather insufficient for linking AMR genes to specific plasmids. Only a hybrid approach allowed for an overall analysis of the whole plasmid genome and its characteristics. With the establishing of the most reliable sequencing technique for detecting plasmids, we scrutinized the prevalence of qnr on MGEs in E. coli from German livestock and food, as understanding the pathways of PMQR spread begins with monitoring the presence of e.g. qnr genes on plasmids. Thus, we investigated the prevalence of the qnr-carrying plasmids in commensal E. coli. We aimed to detect the common characteristics of qnr-carrying plasmids and E. coli as well as their association to other risk factors as e.g. virulence genes. We found qnr to be widely spread over different livestock and food matrices, detected in different ST of E. coli. We frequently detected qnr and qac co-existing on the same plasmid and in association to other resistance genes like cephalosporin determinants. In addition, most of the investigated isolates had point mutations in the QRDR, leading to even higher MIC values. Thus, qnr-carrying E. coli often harboured multiple risk factors that need to be considered when evaluating their impact on resistance development and spread in livestock and food. As we detected qnrS and qnrB to be the most frequent variants in German livestock and food in E. coli, we investigated the plasmids, carrying these resistance genes. We found qnrS1 to be highly prevalent in the analysed samples, located mainly on IncX plasmids. All here investigated IncX plasmids carried a bla resistance determinant and were recognized as transmissible. Thus, it seemed that IncX plasmids are the main vector for the dissemination of qnrS resistance genes. While qnrS is frequently detected in livestock and food samples, qnrB was often reported in samples, isolated from humans, associated with ESBL E. coli. This combination of resistance against two important antimicrobial agents is highly undesirable from a clinical point of view. Therefore, we further examined the presence and characteristics of qnrB-carrying ESBL E. coli. Here, we found a small Col-plasmid to be the main vector of qnrB19. In addition, larger IncH and IncN plasmids were detected as carriers for qnrB. While the Col-plasmid did not carry any other resistance genes, the other prevalent plasmid types were responsible for a multi-resistance phenotype. In addition, all plasmids were characterized as transmissible. Thus, another vector was characterized, presumably responsible for the spread of qnrB in ESBL E. coli. However, the E. coli harboruing the qnrB or qnrS genes were highly heterogenic in their STs and O:H-types.
Overall, we found qnr genes frequently in combination with other resistance determinants, virulence factors and quaternary ammonium compounds. Moreover, known and unknown point mutations within the chromosome increased the MIC against quinolones and fluoroquinolones. All these factors demonstrate the importance of the resistance determinant qnr. As the general characteristics of the E. coli, like the resistome, the carried virulence genes or the ability of pathogenicity, was highly diverse, the general risk outgoing from the qnr-carrying E. coli is also broad. However, we detected prevalent plasmid types carrying qnrB and qnrS, recognized as a probable driver of qnr spread. Furthermore, we have shown that the choice of sequencing and assembly methods is of high importance when investigating MGEs. Only with the correct sequencing and assembly approach, a reliable risk assessment can be ensured. With this study, we contributed to the understanding of the influence of MGEs for the dissemination and dynamics of important resistance determinants in commensal E. coli.
en