The use of next generation sequencing to measure the effects of antibiotics on resistance in the host microbiota

Antimicrobial resistance is increasing worldwide, causing more serious and difficult to treat infections, and is of sufficient severity that the Chief Medical Officer has tabled it to be added to the national risk register. We face a future where infections become untreatable and previously routine procedures such as hip replacements and chemotherapy become too high risk to perform.

Resistance is increasing both in the bacteria causing serious infection, but also in the 'good' bacteria in our gut which form part of our 'microbiome' - the community of microorganisms which share our body and are essential for daily function.

The gut microbiome forms a reservoir of resistance genes which can be transmitted between different species of bacteria - including disease-causing strains, and cause them to become resistant too. It has been shown that resistance in the gut bacteria is associated with an increased likelihood of developing a resistant infection.
Resistant bacteria can be acquired- from food sources (where there is high antibiotic use in farmed meat and aquaculture) from countries of high resistance prevalence (predominantly low-resource countries), and from healthcare institutions. By using antibiotics to treat infection, we are selecting for resistant organisms and increasing antimicrobial resistance.

It has been shown that antibiotic use correlates with resistance on a population level, and based on this it is thought that by reducing overall antibiotic use we can reduce resistance. However reduction in antibiotic use has to be achieved safely and without putting patients at undue risk from undertreated infection, and there is not yet robust evidence that a reduction in antibiotic use in this way can provide a reduction in resistance.

The long term aim of the project is to develop antibiotic treatment strategies able to reduce the development of resistance. To do this, we need to be able to measure and compare how much a particular antibiotic increases resistance. We also need to be able to understand what other factors lead to resistance. The bacteria colonising the gut are present in large numbers in faecal samples. Due to rapid advances in DNA sequencing technology we are able to measure resistance genes these bacteria by extracting bacterial DNA from faecal samples, and sequencing it in days, rather than months.

I propose to pioneer and develop this Next-Generation Sequencing (NGS) technology to detect antibiotic resistance genes in faecal samples (measuring the gut ' resistome'). It has already been used to detect resistance in samples from the Human Microbiome Project, and the pattern of resistance was shown to follow country-wide antibiotic use. I will use it to assess the factors which lead to increased resistance in the gut bacteria, to try and find out where this resistance is being acquired from (food, travel, healthcare institutions). I will also use it to study the association between prior antibiotic use and resistance, including mode of delivery - there is evidence in mouse models that giving antibiotics orally may increase resistance in the gut compared to giving the same dose intravenously. I will then look at the dynamic effect of antibiotics by recruiting patients who will receive antibiotic therapy and analysing multiple faecal samples to see whether resistance returns to prior levels or is maintained.

In this project I aim to identify which antibiotic strategies cause the least selection for antimicrobial resistance in the gut. This information will be used by clinicians to choose antibiotic strategies to take to clinical trials to show a reduction in resistance without risking undertreating infection, and to help decide which antibiotics to use to minimise the risk of future resistant infection to patients.

Aims:

I will develop a semi-quantatitive assay of antibiotic resistance genes (resistome) using b Next Generation Sequencing (NGS) and bioinformatic analysis of faecal samples.

1) Samples from 20 patients will be used to develop the assays to compare against conventional culture-based resistance assessment.
2) A Cross-Sectional Study of 100 patients will investigate:
The extent to which environmental exposure to resistant bacteria affects the resistance profile, the baseline distribution of the resistome, and whether the duration and route of antibiotic use affects resistance in the faecal flora. 100 participants will allow the standard deviation of the 'resistome' to be estimated within 14% with 95% confidence (2-sided alpha=0.05)

Methods

1) Appropriate ethical approval and patient consent will be obtained.
2)Conventional microbiology performed using culture on selective medium and qualitative assessment of resistance.
3) Genomic analysis : DNA extraction using mechanical disruption and Qiagen faecal DNA kit, Illumina Miseq and Hiseq sequencing platforms. Bio-informatics analysis using the Oxford MMMC pipeline, including BLAST searches and de-novo assembly against reference genes.
4) Epidemiological analysis using conventional statistical analysis.

Opportunities arising from this study:

1) Translational: this study will provide direct evidence on antibiotic selection pressure to support choice of antibiotic in guidelines and evidence to persuade clinicians to reduce antibiotic use.
2) Scientific: the assay would potentially form a primary superiority endpoint for an intervention trial comparing antibiotic strategies. Clinical outcomes and antibiotic consumption are expected to form secondary non-inferiority and superiority endpoints respectively. It will also expand the existing knowledge on mechanisms of resistance selection in the host microbiome.

The range of beneficiaries from this research project:

1) Academics
a) This will be of benefit to scientists investigating the mechanisms of resistance selection in enterobacteriaciae, and using next-generation sequencing to analyse complex microenvironments.
Timescale: 1-3 years
b) It will be of benefit to clinical researchers to inform choice of antibiotic treatment strategies to take to clinical trials with increased likelihood of showing reduction in resistance.
c) It will also provide them with a potential outcome measure by which to conduct these clinical trials.
Timescale 3-5 years

2) Practitioners

a) This will provide evidence to enable clinicians to compare and weigh up the benefits of different antibiotic treatment strategies in terms of benefit to patient and harm in terms of resistance in their microflora and risk of future resistant infection.
Timescale: 3-5 years

3) Decision making bodies

a) Information on the association of resistance with environmental exposure factors will be of interest to public health bodies to inform interventions to reduce resistance.
b) Local and National bodies leading policy and direction of research on antibiotic use will be be able to use the information on to direct future strategies for research.
Timescale - 5 years
c) At a local level, information on resistance prevalence in the faecal microbiota and enterobacteriaciae in the region will inform changes in antibiotic treatment strategies and guidelines.
Timescale - 1-3 years

4) Public

a) This research project will provide evidence to enable us to select antibiotic treatment strategies able to minimise resistance selection, reducing the risk to treated patients of drug resistant infection.
Timescale- 5-10 years
b) It will also identify environmental factors which affect resistance in the gut microflora enabling interventions at a policy and also personal level designed to reduce exposure to resistance.
Timescale: 3-5 years