The UK's current national antimicrobial resistance (AMR) action plan recognises that there is not enough understanding of AMR transmission between animals and the environment.  BioSS is working with SRUC scientists to assess and monitor the presence and potential spread of bacterial AMR genes in farms with pigs, poultry, cattle and sheep, on which different antibiotic treatments are administered.  By establishing routes of AMR transmission and persistence from farm animals to the local soil, vegetation and wildlife environment, this project aims to identify mitigation and containment measures with potential to reduce the spread of AMR.

Antimicrobial resistance (AMR) is a significant public health challenge that results in increased morbidity, mortality, and healthcare costs worldwide.  The emergence of AMR is driven by the widespread use and misuse of antimicrobial drugs in humans, animals, and agriculture. The SRUC farms at Easter Bush are a useful exemplar to understand the presence and potential spread of bacterial AMR genes because of the close proximity of four livestock farms (pig, cattle, sheep and poultry), interspersed with fields where manure application records are available.  Semi-wild environments are also present, which might help establish a “natural” AMR baseline, and/or allow detection of transmission pathways beyond the farm system.

Samples have already been collected from farm and wild animal faeces, and from soil and vegetation from surrounding fields.  Faecal and environmental samples are currently being sequenced using a metagenomics approach, whereby DNA is extracted and sequenced, allowing the microbiome (bacterial community) in the sample to be reconstructed.  Individual bacterial isolates will also be grown from a subset of microbiome samples, and sequencing will be performed on these isolates, to allow assignment of AMR genes to a given species.

BioSS’s key contribution will be to provide an appropriate bioinformatics analysis of these samples.  In recent years, many genomic tools have been developed to detect AMR genes in sequencing data, and these can be broadly divided into alignment-based methods and assembly-based methods.  Alignment-based methods compare sequencing reads directly to a reference database of known AMR genes.  Assembly-based methods also query AMR reference databases, but after sequencing reads are assembled into ‘contigs’: sets of DNA segments or sequences that overlap and hence provide a contiguous representation of a genomic region.  For genomes obtained with the standard procedures used in microbiological laboratories (typically cloning of individual isolates), assembly-based methods are more effective in identifying AMR genes due to their high sensitivity and specificity.  In microbiome or metagenomics studies, alignment-based methods might be more effective due to the high complexity of the data and the low abundance of AMR genes in the sample.  Both these approaches will be adopted when analysing the data, in order to maximise sensitivity.  The two data sets will then be cross-referenced, with the aim to extrapolate species-level AMR information from isolates to microbiome samples.  The combined AMR data set will allow for the spatial distribution of AMR genes to be established, together with potential dispersal patterns.  This analysis will be used to guide a second round of more targeted sampling and sequencing, where initial findings and predictions can be further tested and validated.

This work was done in collaboration with Arianne Lowe, Lesley Smith, and Mike Hutchings at SRUC and Alexander Corbishley at The Roslin Institute and was funded under the Scottish Government's Strategic Research Programme for environment, agriculture and food. Other BioSS staff involved: Paolo Ribeca, Sonia Mitchell, Glenn Marion