Lateral Antimicrobial Resistance Genetic Transfer is active in the open environment

Abstract

Historically, the environment has been viewed as a passive deposit of antimicrobial resistance mechanisms, where bacteria show biological cost for maintenance of these genes. Thus, in the absence of antimicrobial pressure, it is expected that they disappear from environmental bacterial communities. To test this scenario, we studied native IntI1 functionality of 11 class 1 integron-positive environmental strains of distant genera collected in cold and subtropical forests of Argentina. We found natural competence and successful site-specific insertion with no significant fitness cost of both aadB and bla _VIM-2 antimicrobial resistance gene cassettes, in a model system without antibiotic pressure. A bidirectional flow of antimicrobial resistance gene cassettes between natural and nosocomial habitats is proposed, which implies an active role of the open environment as a reservoir, recipient and source of antimicrobial resistance mechanisms, outlining an environmental threat where novel concepts of rational use of antibiotics are extremely urgent and mandatory.

Figure 1

Two model systems of dissemination of antimicrobial resistance gene cassettes. In the source-sink model system there is a flow (grey arrow) from the clinic (source) to the open environment (sink). However, few events of recombination linked to antimicrobial resistance can be maintained in the environment (W < 1). In the bidirectional model system, clinical and environmental bacteria disseminate in both habitats. These bacteria from different niches can exchange and maintain antimicrobial resistance gene cassettes (ARGCs) (W ≥ 1) acquiring new phenotypes or not (Supplementary Table S1). Both the open environment and hospitals act as a reservoir and source of ARGCs able to deal with antibiotic pressure, while integrons modulate this exchange in part; arrows suggest possible ways of bacterial dissemination between settings. With the bidirectional model, the results are (i) a greater presence of ARGCs in the environment and in the clinical habitat, and (ii) the possibility that strains with new ARGCs penetrate in the hospital environment. Different shapes and letters represent distinct bacterial species or strains that can inhabit diverse micro-habitats. Red and green represent strains with or without ARGCs, respectively. Arrow width represents the amount of flux of resistant strains between habitats.

Figure 2

Model system of environmental intI1-positive strains used for ARGC acquisition. (a) Origin of environmental strains. Escherichia coli 4IgSN1, Enterobacter sp. 1IgSLAM2, Acinetobacter sp. 1IgSLAM1 and Acinetobacter sp. 1IgSN3 were isolated in Iguazú National Park in Misiones Province, whereas, Pseudomonas sp. 1SL5, Enterobacter sp. 10AL1, Aranicola sp. 9AL34, Pseudomonas sp. 7AN1, Aeromonas media 1AC2, Vibrio sp. 1AC4 and Pantoea dispersa 10FZSS14 were obtained from Tierra del Fuego Island, Argentina. South America map was modified from public domain artwork (https://openclipart.org/detail/181050/argentina-location). (b) Phylogenetic tree of environmental intI1 alleles. A total of 36 alleles of the intI1 gene were identified from a BLASTn query using AF313471 as a reference (April 2015). Black dotted circles correspond to environmental strains used in this study. The inner lane corresponds to the provenance of the intI1-positive strain: blue for Europe, yellow for Asia, light blue for Oceania, light green for America and orange for pandemic dissemination. The medium lane corresponds to the type of sample: water drop for water sample, brown circle for soil, monkey for monkey faeces sample, pig for pig faeces sample and fox for fox faeces sample. The outer lane corresponds to the clinical (red circle) or environmental (green circle) source of the strain. (c and d) Steps of ARGC acquisition without antibiotic selection. Environmental isolates were subjected to artificial and natural transformation with plasmids paadB (c), carrying the aadB gene cassette, and pVIM2 (d), carrying the bla _VIM-2 gene cassette. After transformation, functional native IntI1 excised the ARGC and inserted it in the native attI1 site. Plasmid maintenance during 40 generations was observed in environmental strains that are in blue letters. Environmental strains able to perform ARGCs acquisition by both natural and artificial transformation are in green letters, while isolates that only insert the ARGC as a result of an artificial transformation are in black letters.

Figure 3

Pair-wise mixed culture competition experiments. Competence assays of environmental strains Pseudomonas sp. 1SL5 and E. coli 4IgSN1, against the corresponding ARGC (aadB or bla _VIM-2) inserted at the native attI1 site containing strain were conducted in order to determine the relative fitness. Error bars indicate 95% confidence intervals. By definition, a relative fitness (w) of 1.00 indicates no difference in relative fitness. Relative fitness values: Pseudomonas sp. 1SL5::aadB W = 1.03 (0.88–1.18), E. coli 4IgSN1::aadB W = 1.02 (0.86–1.17), E. coli 4IgSN1::bla _VIM-2 W = 0.97 (0.92–1.01). CI is confidence interval.