Comment on: Transferable resistance to colistin: a new but old threat

Sir,

We would like to comment on the leading article published in JAC Advance Access on 24 June 2016, regarding the recently discovered transferable resistance to colistin and its observation that ‘this new resistance had spread from the veterinary to the human domain’.¹

Most Escherichia coli and other bacteria (including Salmonella) containing plasmids with colistin resistance genes mcr-1 and mcr-2 have been isolated from terrestrial animals and/or food produced from them.^1,2 However, we would like to call your readership's attention to the fact that these genes and the various plasmids harbouring them have been circulating in animal and human populations for quite some time. In some cases, the original genetic events responsible for mobilization of the mcr genes may thus have happened relatively long ago and may quite plausibly have resulted from colistin use in agriculture and aquaculture.^1,3,4 This latter activity is growing exponentially, is associated with excessive antimicrobial use and provides multiple opportunities for generating linkage between the resistomes of aquatic bacteria and those of human pathogens.³

Published information is consistent with the suggestion that aquaculture could be involved in the origin and selection of some of these mcr genes. First, Vibrio spp. and Photobacterium damselae, potential fish and human pathogens, are resistant to colistin.⁵ Second, an ethanolamine phosphotransferase (EptA, PmrC), the type of enzyme encoded by mcr-1, was detected by functional genomics in Shewanella algae, an aquatic bacterium belonging to a genus that also contains opportunistic fish and human pathogens.⁶ This cloned Shewanella gene encoded colistin resistance when expressed in E. coli and was highly similar in sequence to genes in the opportunistic fish and human pathogens Plesiomonas shigelloides and P. damselae subsp. piscicida.⁶ Third, the deduced amino acid sequence of MCR-1 protein displays important similarity to the phosphoethanolamine transferase of Enhydrobacter aerosaccus, an aquatic bacterium.⁷ Fourth, the MCR-2 protein has important sequence similarity to the phosphoethanolamine phosphotransferase protein of the aquatic bacteria Vibrio halioticoli and Psychrobacter spp.² Fifth, mcr-1 genes have been found in E. coli isolated from birds of the marine environment (herring gulls in Europe, kelp gulls in South America).¹ Sixth, water in aquacultural environments in developing countries is often contaminated with animal and human pathogens.³ In addition, mcr-1 is widely distributed in Asia where colistin is extensively used in animal husbandry and where fish under integrated aquaculture are often fed animal manure containing antimicrobials, including colistin.^4,8,9

These findings suggest that the potential exists in the aquaculture environment for selection of colistin-resistant bacteria whose resistance is the result of an LPS modification encoded by a chromosomal ethanolamine transferase gene and the passing of this gene to terrestrial bacteria by horizontal gene transfer to yield colistin-resistant human pathogens.^7–9 This occurrence might also be aided by the presence of other antimicrobial genes in the mcr plasmids and the presence of these antimicrobials in the aquaculture environment. The mcr genes are not the first example of chromosomally located antimicrobial resistance genes in aquatic environmental bacteria to become located in plasmids and thus transmissible. A similar situation has been previously described with the plasmid-mediated quinolone resistance genes qnrA and qnrS.³ Mobilization of mcr-1 genes between the chromosome of aquatic bacteria and plasmids could have been carried out by ISs flanking some mcr-1 genes, i.e. ISApl1, IS1294.¹⁰ Because mcr-1 has been recently found to be located in multiple antimicrobial resistance IncFII plasmids that have the ability to be chromosomally integrated,¹¹ it could be also postulated that a plasmid of this type could integrate into the chromosome of a strain containing the mcr-1 gene, and then be excised as an IncFII-mcr replicon able to transfer mcr at high frequency. The negative potential of this series of events on human health can serve as a cautionary tale regarding excessive use of antimicrobials in aquaculture. As indicated above, such use could lead to multiple opportunities for linking the resistomes of aquatic bacteria and human pathogens.

Transparency declarations

None to declare.