Selection for AMR in complex microbial communities at sub-therapeutic antibiotic concentrations

Antimicrobial resistance (AMR) is an increasing problem in human and animal pathogens, and has been highlighted as a serious threat to public health by the Chief Medical Officer. There is increasing evidence that antibiotic usage in agriculture may be contributing to the emergence of AMR in the clinic and the government's 5 year AMR strategy highlights improved knowledge and understanding of AMR as a key priority. This project will investigate the evolution of AMR in complex microbial communities using laboratory evolution experiments. For the first time, selection for AMR in a clinically important opportunistic pathogen, E. coli, will be studied in the presence and absence of a complex community of bacteria. The community will be from pig faeces, and will be incubated in anaerobic fermenters that have previously been used as simple gut models.

Traditional approaches to studying AMR in bacteria often consider only one species in isolation. This approach can be useful in studying mutation based adaptation to antibiotics; however it does not consider the impact of horizontal gene transfer where resistance genes are acquired from the microbial community. Single species experiments also fail to consider competition between different species of bacteria that have differing intrinsic resistance to antibiotics.

Competition experiments in the presence and absence of a complex microbial community will help us understand how evolution of AMR occurs. We will expose experimental microbial communities to different concentrations of antibiotic representing the sub-therapeutic concentrations found in the animal and human gut during oral antibiotic therapy. This will give insights into the way complex interactions that occur within microbial populations affect evolution of AMR, relative to selection for AMR in single species experiments. Exposure to low concentrations of antibiotics was traditionally thought to be unimportant in selection for AMR. However recent data suggests that selection occurs at much lower concentrations than previously thought, and preliminary data in Gaze's lab suggests that selection for AMR gram-negative opportunistic pathogens can be greater at lower sub-therapeutic concentrations than at higher concentrations closer to those used to treat infections.

We will also investigate the effects of selection by a single antibiotic on relative abundance and diversity of all known resistance genes. This will provide data on indirect co-selection for AMR genes due to genetic linkages in the genomes of bacteria in the complex community and on mobile genetic elements capable of transferring multiple genes between bacteria. Cell sorting and next generation sequencing techniques will identify all known AMR genes transferred to a genetically tagged E. coli under antibiotic selection. Conversely, we will also investigate AMR gene transfer from a tagged E. coli to all other bacteria within the complex community.

Evolution of AMR will be studied using a genetically tagged focal strain and AMR plasmid in the presence and absence of a complex microbial community using next generation sequencing and bioinformatic approaches.

In this instance we have chosen to use a pig derived community under selection with oxytetracycline, as selection for AMR within the pig microbiome is an important issue given the size of the global pig population and the amounts of antibiotics used in pig husbandry. However, this model system will be used to study phenomena that are likely to occur in all complex microbial communities under antibiotic selection.

We will use genetically tagged host strains and AMR plasmids, allowing quantitation using real-time PCR and separation of host strain and plasmid bearing bacteria using fluorescently activated cell sorting. The latter approach has been validated by Barth Smets who has kindly agreed to supply strains. Selection for the resistance plasmids will be quantified as a function of antibiotic concentration from the minimum inhibitory concentration of the susceptible focal strain to zero, in the presence and absence of the complex microbial community. Gene transfer between the focal strain and the community will be investigated using cell sorting to separate the focal strain and plasmid recipients from the community. The extent of gene transfer from focal strain to the community and vice versa will be assessed using metagenomic methods. Changes in community structure and AMR gene abundance and diversity during selection experiments will be assessed using amplicon and shotgun sequencing respectively followed by bioinformatic analyses using Qiime and analysis using a structured antimicrobial resistance gene database.

This project is innovative in its combination of selection experiments using genetically tagged strains and plasmids in the presence and absence of a complex microbial community, and the use of next generation sequencing approaches.

Beneficiaries of the research will include academics (see academic beneficiaries) and key stakeholders from government agencies including Defra, the Food Standards Agency (FSA), the Veterinary Medicines Directorate (VMD) and members of the Defra Antimicrobial Resistance Committee (DARC) and Advisory Committee on Antimicrobial Resistance and Healthcare Associated Infections (ARHAI). Individuals working in areas of veterinary and clinical medicine concerned with antibiotic therapy and AMR will also benefit including vets, pharmacists, infection control practitioners and clinical microbiologists. Private sector stakeholders within the animal and human biopharma industry would also benefit from a deeper understanding of the way AMR evolves and the role sub-therapeutic concentrations of antibiotics play in selection for resistance.

There is a fundamental lack of knowledge regarding the relationship between antibiotic exposure and selection for resistance in complex communities such as those within the human and animal microbiomes. Whilst patterns of resistance in different microbial communities under different selective regimens will vary the proposed research represents one of the first attempts to investigate selection for AMR in complex microbial communities under controlled replicated conditions using cutting edge molecular tools. Previous research using animal models is subject to lack of true controls and adequate replication to account for confounding variables. Working with genetically tagged strains in animals is also logistically difficult and costly. In vitro work to date largely relies on single or dual species experiments which do not account for horizontal transfer of resistance genes or competition between taxa with differing intrinsic levels of resistance. Where mixed communities are studied, the system is usually regarded as a "black box" with limited ability to disentangle changes within the community.

Our innovative approach combining molecular microbial ecology and experimental evolutionary biology utilising next generation sequencing and bioinformatic approaches is well suited to tackling this complex problem.

Increased understanding of how antibiotics select for resistance in complex microbial communities is crucial in developing best practice in terms of prescribing and prophylaxis in animals. Interventions that reduce the rate at which AMR evolves have the potential to improve the nation's health and reduce social and economic costs associated with AMR. Our data will strengthen the evidence base relating to the impacts on microbial communities of exposure to antibiotics at a range of concentrations encompassing those likely to be present in the human and animal gut.

It is likely that novel insights into the evolution of AMR will be made within the lifetime of the grant, potentially informing policy on antibiotic usage and identifying key questions for further research.

THE RF and technician will develop understanding of the evolution of AMR and will receive training in the use of molecular tools and experimental approaches required to investigate the problem. This expertise will be valuable within the academic, public and private sectors.