Multidrug Resistant Acinetobacter Isolates Release Resistance Determinants Through Contact-Dependent Killing and Bacteriophage Lysis

Abstract

Antimicrobial resistance is an ancient bacterial defense mechanism that has rapidly spread due to the frequent use of antibiotics for disease treatment and livestock growth promotion. We are becoming increasingly aware that pathogens, such as members of the genus Acinetobacter, are precipitously evolving drug resistances through multiple mechanisms, including the acquisition of antibiotic resistance genes. In this study, we isolated three multidrug resistant Acinetobacter species from birds on a free-range farm. Acinetobacter radioresistens, Acinetobacter lwoffii, and Acinetobacter johnsonii were isolated from hens, turkeys and ducks and were resistant to 14 clinically relevant antibiotics, including several listed by the World Health Organization as essential medicines. Co-culturing any of the three Acinetobacter species with Acinetobacter baumannii resulted in contact-dependent release of intact resistance determinants. We also isolated several lytic bacteriophages and selected two of these phages to be included in this study based on differences in plaquing characteristics, nucleic acid content and viral morphology. Both phages released host DNA, including antibiotic resistance genes during cell lysis and we demonstrated that these resistance determinants were transferable to a naïve strain of Escherichia coli. This study demonstrates that contact-dependent competition between bacterial species can readily contribute to DNA release into the environment, including antibiotic resistance determinants. We also highlight that the constant lysis and turnover of bacterial populations during the natural lifecycle of a lytic bacteriophage is an underappreciated mechanism for the liberation of DNA and subsequent genetic exchange.

Keywords: Acinetobacter; bacteriophages; contact-dependent killing; environmental isolates; gene transfer; multidrug resistance.

FIGURE 1

Bacterial killing assay. Acinetobacter contact-dependent killing. (A) A. baumannii kills all isolated Acinetobacter species (LH2, LH6, and D16). The surviving isolates are shown after 4 h of contact. A gfp-encoding plasmid was introduced into the prey species to distinguish them from A. baumannii strain 19606. The A. baumannii ATCC 19606 cdi1 and cdi2 mutant strains were compared to WT and no competition controls to assess the activity of CDI on the isolated strains. (B) The role of T6SS-dependent killing was tested for the isolated strains (LH2, LH6, and D16). The A. baumannii strain 17978 hcp mutant strain, deficient in T6SS activity, was compared to WT and no competition. The surviving isolates are shown after 4 h of contact. All assays were performed at least in triplicate.

FIGURE 2

Contact-dependent DNA release. (A) Schematic showing assay concept. The predator strain (Ab19) liberates the pBAV1K-T5-gfp plasmid from the prey strains being tested. Then the presence of the plasmid is detected by amplification of the KmR gene on the plasmid. (B) Schematic detailing the different incubation conditions being used. (i) Prey strain and Ab19 are mixed 1:1 and spotted directly onto the agar plate; (ii) prey strain and Ab19 are separated by a membrane; (iii) controls include incubations of each strain on the plates individually. (C) A 1.0% agarose gel depicting the PCR-based KmR gene (450 bp) detection of released DNA mediated by contact dependent killing representative of triplicate experiments shown in Supplementary Figure S1. Co-cultures separated by a membrane to inhibit T6SS are denoted with parentheses. (D) Relative intensities of PCR products in each lane from Supplementary Figure S1 were measured by densitometry using Image Lab^TM. The band produced by amplification of pBAV1K-T5-gfp was used as the standard (Rel. Intensity = 1.0). The averages are represented, and error bars represent the standard error of the mean. Student’s paired T-tests were performed for each co-incubation compared to its corresponding membrane-separated control incubation. The p-values for the LH2g, LH6g, and D16g datasets are indicated.

FIGURE 3

Bacteriophage-mediated DNA release. (A) Analyses of extracellular DNA from cultures with and without phage infection. Agarose gel electrophoresis showing intact DNA release from phage-infected cultures, including cellular RNA at approximately 1 and 1.7 kb (see Supplementary Figure S2). (B) Agarose gel electrophoresis of CAP1 and CAP3 genomes. (C) Relative intensities of PCR products from Supplementary Figure S3 were averaged for each treatment and are represented from triplicate experiments. Standard error of the mean is represented by the error bars, and Student’s paired T-tests were performed with p-values indicated. (D) Chemically competent E. coli were transformed with equal amounts of isolated DNA from sample sets from Supplementary Figures S3A,B and the number of kanamycin resistant colonies are indicated.