Canine Saliva as a Possible Source of Antimicrobial Resistance Genes

Abstract

While the One Health issues of intensive animal farming are commonly discussed, keeping companion animals is less associated with the interspecies headway of antimicrobial resistance. With the constant advance in veterinary standards, antibiotics are regularly applied in companion animal medicine. Due to the close coexistence of dogs and humans, dog bites and other casual encounters with dog saliva (e.g., licking the owner) are common. According to our metagenome study, based on 26 new generation sequencing canine saliva datasets from 2020 and 2021 reposited in NCBI SRA by The 10,000 Dog Genome Consortium and the Broad Institute within Darwin's Ark project, canine saliva is rich in bacteria with predictably transferable antimicrobial resistance genes (ARGs). In the genome of potentially pathogenic Bacteroides, Capnocytophaga, Corynebacterium, Fusobacterium, Pasteurella, Porphyromonas, Staphylococcus and Streptococcus species, which are some of the most relevant bacteria in dog bite infections, ARGs against aminoglycosides, carbapenems, cephalosporins, glycylcyclines, lincosamides, macrolides, oxazolidinone, penams, phenicols, pleuromutilins, streptogramins, sulfonamides and tetracyclines could be identified. Several ARGs, including ones against amoxicillin-clavulanate, the most commonly applied antimicrobial agent for dog bites, were predicted to be potentially transferable based on their association with mobile genetic elements (e.g., plasmids, prophages and integrated mobile genetic elements). According to our findings, canine saliva may be a source of transfer for ARG-rich bacteria that can either colonize the human body or transport ARGs to the host bacteriota, and thus can be considered as a risk in the spread of antimicrobial resistance.

Keywords: antimicrobial resistance; bacteriome; dog saliva; mobilome; resistome.

Figure 1

Saliva core bacteriome. The relative abundances of genera that achieved more than 1% of the bacterial hits in any of the samples. In the sample No. 20, no reads were classified to bacteria.

Figure 2

Identifed ARGs by samples. For each sample–ARG combination, only the best finding is plotted. The size and the color of the dots correspond to the coverage and the sequence identity of hits on reference genes, respectively. In sample No. 20, there was no identifiable ARG. The gene names that are too long are abbreviated (acrA: E. coli acrA; emrE: E. coli emrE).

Figure 3

Mobile ARG frequency by bacteria of origin. The size of the dots indicates the occurrence frequency of the given gene flanked by iMGE, positioned in plasmid or prophage.