Use of genomics to explore AMR persistence in an outdoor pig farm with low antimicrobial usage.

Food animals may be reservoirs of antimicrobial resistance (AMR) passing through the food chain, but little is known about AMR prevalence in bacteria when selective pressure from antimicrobials is low or absent. We monitored antimicrobial-resistant Escherichia coli over 1 year in a UK outdoor pig farm with low antimicrobial usage (AMU) compared to conventional pig farms in the United Kingdom. Short and selected long-read whole-genome sequencing (WGS) was performed to identify AMR genes, phylogeny and mobile elements in 385 E. coli isolates purified mainly from pig and some seagull faeces. Generally, low levels of antimicrobial-resistant E. coli were present, probably due to low AMU. Those present were likely to be multi-drug resistant (MDR) and belonging to particular Sequence Types (STs) such as ST744, ST88 or ST44, with shared clones (<14 Single Nucleotide Polymorphisms (SNPs) apart) isolated from different time points indicating epidemiological linkage within pigs of different ages, and between pig and the wild bird faeces. Although importance of horizontal transmission of AMR is well established, there was limited evidence of plasmid-mediated dissemination between different STs. Non-conjugable MDR plasmids or large AMR gene-bearing transposons were stably integrated within the chromosome and remained associated with particular STs/clones over the time period sampled. Heavy metal resistance genes were also detected within some genetic elements. This study highlights that although low levels of antimicrobial-resistant E. coli correlates with low AMU, a basal level of MDR E. coli can still persist on farm potentially due to transmission and recycling of particular clones within different pig groups. Environmental factors such as wild birds and heavy metal contaminants may also play important roles in the recycling and dissemination, and hence enabling persistence of MDR E. coli. All such factors need to be considered as any rise in AMU on low usage farms, could in future, result in a significant increase in their AMR burden.

The importance of using whole genome sequencing and extended spectrum beta-lactamase selective media when monitoring antimicrobial resistance.

To tackle the problem of antimicrobial resistance (AMR) surveillance programmes are in place within Europe applying phenotypic methods, but there are plans for implementing whole genome sequencing (WGS). We tested the benefits of WGS using Escherichia coli collected from pig surveillance performed between 2013 to 2017. WGS was performed on 498 E. coli producing ESBL and AmpC enzymes, recovered from pig caeca on MacConkey + cefotaxime (McC + CTX) agar, as recommended by the European Commission, or ESBL agar, used additionally by United Kingdom. Our results indicated WGS was extremely useful for monitoring trends for specific ESBL genes, as well as a plethora of AMR genotypes, helping to establish their prevalence and co-linkage to certain plasmids. Recovery of isolates with multi-drug resistance (MDR) genotypes was lower from McC + CTX than ESBL agar. The most widespread ESBL genes belonged to the blaCTX-M family. blaCTX-M-1 dominated all years, and was common in two highly stable IncI1 MDR plasmids harbouring (blaCTX-M-1,sul2, tetA) or (blaCTX-M-1, aadA5, sul2, dfrA17), in isolates which were phylogenetically dissimilar, suggesting plasmid transmission. Therefore, WGS provided a wealth of data on prevalence of AMR genotypes and plasmid persistence absent from phenotypic data and, also, demonstrated the importance of culture media for detecting ESBL E. coli.

Molecular epidemiology of isolates with multiple mcr plasmids from a pig farm in Great Britain: the effects of colistin withdrawal in the short and long term.

Background: The environment, including farms, might act as a reservoir for mobile colistin resistance (mcr) genes, which has led to calls for reduction of usage in livestock of colistin, an antibiotic of last resort for humans. Objectives: To establish the molecular epidemiology of mcr Enterobacteriaceae from faeces of two cohorts of pigs, where one group had initially been treated with colistin and the other not, over a 5 month period following stoppage of colistin usage on a farm in Great Britain; faecal samples were also taken at ∼20 months. Methods: mcr-1 Enterobacteriaceae were isolated from positive faeces and was WGS performed; conjugation was performed on selected Escherichia coli and colistin MICs were determined. Results: E. coli of diverse ST harbouring mcr-1 and multiple resistance genes were isolated over 5 months from both cohorts. Two STs, from treated cohorts, contained both mcr-1 and mcr-3 plasmids, with some isolates also harbouring multiple copies of mcr-1 on different plasmids. The mcr-1 plasmids grouped into four Inc types (X4, pO111, I2 and HI2), with mcr-3 found in IncP. Multiple copies of mcr plasmids did not have a noticeable effect on colistin MIC, but they could be transferred simultaneously to a Salmonella host in vitro. Neither mcr-1 nor mcr-3 was detected in samples collected ∼20 months after colistin cessation. Conclusions: We report for the first known time on the presence in Great Britain of mcr-3 from MDR Enterobacteriaceae, which might concurrently harbour multiple copies of mcr-1 on different plasmids. However, control measures, including stoppage of colistin, can successfully mitigate long-term on-farm persistence.

mcr-1 and mcr-2 variant genes identified in Moraxella species isolated from pigs in Great Britain from 2014 to 2015.

Objectives: To determine the occurrence of mcr-1 and mcr-2 genes in Gram-negative bacteria isolated from healthy pigs in Great Britain. Methods: Gram-negative bacteria (n = 657) isolated from pigs between 2014 and 2015 were examined by WGS. Results: Variants of mcr-1 and mcr-2 were identified in Moraxella spp. isolated from pooled caecal contents of healthy pigs at slaughter collected from six farms in Great Britain. Other bacteria, including Escherichia coli from the same farms, were not detected harbouring mcr-1 or mcr-2. A Moraxella porci-like isolate, MSG13-C03, harboured MCR-1.10 with 98.7% identity to MCR-1, and a Moraxella pluranimalium-like isolate, MSG47-C17, harboured an MCR-2.2 variant with 87.9% identity to MCR-2, from E. coli; the isolates had colistin MICs of 1-2 mg/L. No intact insertion elements were identified in either MSG13-C03 or MSG47-C17, although MSG13-C03 harboured the conserved nucleotides abutting the ISApl1 composite transposon found in E. coli plasmids and the intervening ∼2.6 kb fragment showed 97% identity. Six Moraxella osloensis isolates were positive for phosphoethanolamine transferase (EptA). They shared 62%-64.5% identity to MCR-1 and MCR-2, with colistin MICs from 2 to 4 mg/L. Phylogenetic analysis indicated that MCR and EptA have evolved from a common ancestor. In addition to mcr, the ß-lactamase gene, blaBRO-1, was found in both isolates, whilst the tetracycline resistance gene, tetL, was found in MSG47-C17. Conclusions: Our results add further evidence for the mobilization of the mcr-pap2 unit from Moraxella via composite transposons leading to its global dissemination. The presence of mcr-pap2 from recent Moraxella isolates indicates they may comprise a reservoir for mcr.

In vitro investigations into the use of antimicrobials in combination to maintain efficacy of fluoroquinolones in poultry.

The aim of this study was to determine if apramycin, colistin or lincomycin-spectinomycin, in combination with enrofloxacin, was able prevent the emergence of mutants with reduced susceptibility to fluoroquinolone antibiotics in vitro. MICs were determined for enrofloxacin alone and in combination for panels of Campylobacter (n=37), Escherichia coli (n=52) and Salmonella (n=52) isolates. MIC results suggested that apramycin, colistin and lincomycin-spectinomycin worked in an additive/indifferent way when each was combined with enrofloxacin. Apramycin was considered the most promising antibiotic for combination-therapy in conjunction with enrofloxacin, and further evaluations (MBCs, MPCs and time-kill-curves) were performed for this combination for selected isolates. Results suggest combination-therapy of enrofloxacin with apramycin increases the efficacy, as well as decreasing the emergence and survival of bacteria with mutational resistance to fluoroquinolone antibiotics. Such combination-therapy, minimising the development of mutational resistance, may have relevance for Campylobacter, E. coli and Salmonella infections in poultry.

Evaluation of MALDI-ToF as a method for the identification of bacteria in the veterinary diagnostic laboratory.

Matrix-Assisted Laser Desorption Ionisation-Time of Flight (MALDI-ToF) Mass Spectrometry with Bruker MALDI Biotyper software was evaluated as a method for identifying veterinary bacteria. For 620 isolates (~100 bacterial species), identification by MALDI-ToF and non-16S rDNA methods (mainly phenotypic/biochemical) agreed to species-level (95.3%) and to species/genus-level (100%), but in the absence of 16S rDNA as a gold standard. For a further panel of 107 anaerobes and 234 aerobes (~100 bacteria species) using 16S rDNA results as the gold standard, MALDI-ToF/biochemical tests showed 97.8/96.6% species-level and 99.6/93.5% genus-level agreement for aerobes and 95.3/93.6% species-level and 100/95.3% genus-level agreement for anaerobes compared to the gold standard. Where results were obtained from direct spots, direct spots overlaid with formic acid and extracts, 89.4% of 180 aerobes and 90.1% of 152 anaerobes were identified by MALDI-ToF. MALDI-ToF was shown to be a rapid and reliable method to identify veterinary bacteria.

Development and evaluation of internal amplification controls for use in a real-time duplex PCR assay for detection of Campylobacter coli and Campylobacter jejuni.

A common problem of both conventional and real-time PCR assays is failure of DNA amplification due to the presence of inhibitory substances in samples. In view of this, our aim was to develop and evaluate internal amplification controls (IACs) for use with an existing duplex real-time PCR assay for Campylobacter coli and Campylobacter jejuni. Both competitive and non-competitive IACs were developed and evaluated. The competitive approach involved a DNA fragment of the coding region of the fish viral haemorrhagic septicaemia virus, flanked by the mapA PCR primers, whilst the non-competitive approach utilized an extra set of universal 16S rDNA primers. Both IAC-PCR assay types were evaluated using cultures of Campylobacter and chicken caecal content samples. Both IACs were sensitive to caecal inhibitors, making them suitable for detecting inhibition which could lead to false-negatives. Results showed that both IACs at optimum concentrations worked well without reducing the overall sensitivity of the PCR assay. Compared to culture, the optimized competitive IAC-PCR assay detected 45/47 positives (sensitivity 93.6 %, specificity 80.1 %); however, it had the advantage over culture in that it could detect mixed infections of C. coli and C. jejuni and was capable of giving a result for a sample within a day.