Acinetobacter baumannii from grass: novel but non-resistant clones

Abstract

Acinetobacter baumannii is one the most worrisome nosocomial pathogens, which has long been considered almost mainly as a hospital-associated bacterium. There have been some studies about animal and environmental isolates over the last decade. However, little effort has been made to determine if this pathogen dwells in the grass. Here, we aim to determine the evolutionary relationships and antibiotic resistance of clones of A. baumannii sampled from grass to the major human international clones and animal clones. Two hundred and forty genomes were considered in total from four different sources for this study. Our core and accessory genomic epidemiology analyses showed that grass isolates cluster in seven groups well differentiated from one another and from the major human and animal isolates. Furthermore, we found new sequence types under both multilocus sequence typing schemes: two under the Pasteur scheme and seven for the Oxford scheme. The grass isolates contained fewer antibiotic-resistance genes and were not resistant to the antibiotics tested. Our results demonstrate that these novel clones appear to have limited antibiotic resistance potential. Given our findings, we propose that genomic epidemiology and surveillance of A. baumannii should go beyond the hospital settings and consider the environment in an explicit One Health approach.

Keywords: Acinetobacter baumannii; Genomic epidemiology; One Health; antibiotic resistance; grass; international clones; non-clinical environments.

Fig. 1.

Maximum-likelihood phylogeny based on the single gene families without recombination. The ICs are highlighted in different colours. Animal isolates are marked with dark violet and blue circles. Light blue taxa describe the new STs as per Pasteur scheme. The first ring shows the isolation source, and the most external ring shows the grass groups identified. The violet dots in the internal nodes show bootstrap values≥80 %. The scale bar denotes the substitutions per site.

Fig. 2.

Correlation matrix of gene content. Heat map showing the correlation matrix of the gene content variation among the isolates. The first bottom row depicts the individual grass groups; whereas the second bottom row gives the source (human ICs, animal, grass) for all the isolates.

Fig. 3.

Heat map showing the antibiotic resistance genes per grass isolate. On the right antibiotic-resistance genes with significant matches are shown. Perfect matches are in blue squares, whereas no perfect matches are in violet. Isolate names are provided at the bottom. Drug classes are coloured-coded. For the OXAs and ADCs variants, we put in parentheses the gene families they belong to. It is also highlighted which database identified which ARG: AMRFinderPlus (letter A), CARD (letter C) or identified by both (letter B).