Using metagenomics to investigate the impact of hospital stay and the ARK intervention on the human gut resistome

Abstract

Antimicrobials are vital for modern medicine. Antimicrobial use selects for antimicrobial resistant bacteria, particularly among the gut microflora. Minimising antimicrobial resistance (AMR) selection by avoiding unnecessary antibiotic use helps combat AMR. Metagenomic analyses have the potential to provide accurate detection and quantification of AMR genes within an individual’s gut microbiome (gut ‘resistome’), allowing the impact of different types of antibiotic exposures to be evaluated and guide interventions to reduce AMR. We have developed a short-read sequencing approach to characterise the gut ‘resistome’ and piloted this in two clinical sample sets. The first consisted of paired stool samples from older hospitalised adults. 25 pairs of samples (1 to 50 days apart) showed a median AMR gene reads/kb/million total reads (RPKM) of 1841 (124 to 17,832), and a median AMR gene count of 55 (2 to 101). The second consisted of faecal discards from Clostridium difficile testing at a hospital eleven months apart. In these samples (n=21) the median AMR gene read was 923 RPKM (240 to 19,475), and the median AMR gene count was 36 (9 to 82). In both sample sets there was a trend towards an increase in AMR gene RPKM and number between the time points. dfrF, ErmB and tetW were the commonest AMR genes in both sample sets. This approach is being applied to analyse the impact of an intervention (ARK-Hospital) designed to change antibiotic prescribing behaviour. The data will allow us to determine the patient-level impact of reduced antibiotic exposure on carried AMR.