Plasmid Carriage of bla_NDM-1 in Clinical Acinetobacter baumannii Isolates from India

Abstract

NDM-1 probably emerged in Acinetobacter species prior to its dissemination among Enterobacteriaceae, and NDM-1-like enzymes are increasingly reported in Acinetobacter species. Here, we report on the genetic context of bla_NDM-1 in the earliest known NDM-1-producing organisms, clinical isolates of Acinetobacter from India in 2005. These strains harbor bla_NDM-1 plasmids of different sizes. The gene is associated with the remnants of the Tn125 transposon normally associated with bla_NDM-1 in Acinetobacter spp. The transposon has been disrupted by the IS26 insertion and subsequent movement events.

TEXT

Multidrug-resistant Enterobacteriaceae producing the New Delhi metallo-β-lactamase-1 (NDM-1) enzyme have been isolated around the world, with many cases linked to travel, especially to South Asia (1–3). Carbapenem resistance in the opportunistic pathogen, Acinetobacter baumannii, is mostly associated with OXA-type β-lactamases (4), but NDM-1-like enzymes have been increasingly reported (5–10). In all bacterial species, the bla_NDM-1 gene is associated with at least one copy of a complete or partial ISAba125 gene, which provides bla_NDM-1 with a strong promoter (11–13). ISAba125 almost certainly originates from Acinetobacter spp., and there is evidence that bla_NDM-1 was probably formed by a fusion event between the aminoglycoside resistance gene, aphA6, and an ancestral carbapenemase in an Acinetobacter background (13), thus implying that bla_NDM-1 spread to Enterobacteriaceae from Acinetobacter spp.

A small number of studies have reported a relatively high prevalence of NDM-1-producing A. baumannii, causing infections in intensive care patients in Indian hospitals (10, 14). There are also several case reports of colonization and infection with these organisms from European countries, many with epidemiological links with travel to North Africa or the Balkans (5, 6, 11), and from the Middle East (9). In China, bla_NDM-1 has been frequently identified in Acinetobacter spp., including A. baumannii, both in clinical cases and from environmental sources (7, 15–17). The genetic context of bla_NDM-1-like genes in Acinetobacter spp. shows less variation than that observed in other genera. In most strains, bla_NDM-1 is bracketed by two copies of ISAba125 to form a Tn125 transposon. The content of Tn125 is usually conserved, with the occasional exceptions resulting from truncation by insertion of other IS elements. Tn125 is found inserted in several different gene locations, with direct repeats at either end indicating movement by transposition. All isolates from Europe and North Africa are chromosomally located, but in China, bla_NDM-1 is reportedly found mostly on plasmids in Acinetobacter spp. other than A. baumannii (15–18).

We received 9 Acinetobacter isolates that were collected in 2005 from a hospital in Tamil Nadu, India. Most isolates were from patients receiving intensive care and were isolated from blood, pus, and respiratory secretions. The initial identification and sensitivity testing were performed using a BD Phoenix system (Becton, Dickinson, Franklin Lakes, NJ, USA), with supplementary sensitivity testing performed using the Etest (bioMérieux, LaPlane, France) or MIC test strip (Liofilchem, Roseto degli Abruzzi, Italy) method (Table 1). The confirmation of A. baumannii identification was performed by PCRs for bla_OXA-51-like genes. All isolates were bla_OXA-51-like positive by PCRs except CHI-40-1, and all were extensively drug resistant (19), with the A. baumannii isolates being resistant to all drugs tested except colistin and in some cases amikacin and tobramycin. Seven isolates were positive for bla_NDM-1 by PCRs, with CHI-41 and CHI-44 being the only bla_NDM-1-negative isolates. All of the A. baumannii isolates carried bla_OXA-23-like genes, confirmed to be associated with ISAba1, and thus having a strong upstream promoter, by PCR. All PCR amplicons were confirmed by sequencing.

TABLE 1.

Study isolates, specimen types, and antimicrobial MICs for Acinetobacter isolates

Isolate	Specimen type	MIC (μg/ml) fora:
		Azt	Caz	Taz	Imp	Mem	Ami	Gent	Tob	Cip	Col	Rif	Tig
A. baumannii CHI-16	Blood	>16	≥256	>16/4	≥32	≥32	≤4	>4	2	≥32	≤1	6	2
A. baumannii CHI-18	Blood	>16	≥256	>16/4	≥32	≥32	>16	>4	>4	≥32	≤1	4	3
A. baumannii CHI-32	Blood	>16	≥256	>16/4	≥32	≥32	≤4	>4	2	≥32	≤1	6	2
A. baumannii CHI-34	Sputum	>16	≥256	>16/4	≥32	≥32	≤4	>4	>4	≥32	≤1	4	2
Acinetobacter sp CHI-40-1	Pus	>16	≥256	>16/4	≥32	≥32	>16	>4	4	≥32	≤1	≥256	0.75
A. baumannii CHI-40-2	Pus	>16	≥256	>16/4	≥32	≥32	>16	>4	>4	≥32	≤1	6	2
A. baumannii CHI-41	Sputum	>16	≥256	>16/4	≥32	≥32	>16	>4	>4	≥32	≤1	6	1.5
A. baumannii CHI-44	Endotracheal aspirate	>16	≥256	>16/4	≥32	≥32	>16	>4	>4	≥32	≤1	6	2
A. baumannii CHI-45-1	Endotracheal aspirate	>16	≥256	>16/4	≥32	≥32	>16	>4	>4	≥32	≤1	3	2

^aMICs were determined by the BD Phoenix system except for ceftazidime, imipenem, meropenem, ciprofloxacin, rifampin, and tigecycline, which were determined by an Etest or a MIC test strip. Ami, amikacin; Azt, aztreonam; Caz, ceftazidime; Cip, ciprofloxacin; Col, colistin; Gent, gentamicin; Imp, imipenem; Mem, meropenem; Rif, rifampin; Taz, piperacillin-tazobactam; Tig, tigecycline; Tob, tobramycin.

A. baumannii isolates were then typed by the multilocus sequence typing (MLST) method described by Turton et al. (20) and by pulsed-field gel electrophoresis (PFGE) of ApaI-digested genomic DNA. All A. baumannii isolates producing NDM-1 by MLST were within group II, which corresponds with worldwide clone 1. The 2 bla_NDM-1-negative isolates were within group I. ApaI profiles were similar but not identical for all group II A. baumannii isolates and differed substantially from those for group I isolates (Fig. 1a) and CHI-40-1.

FIG 1.

(a) PFGE of ApaI-digested genomic DNA. Lane 1, λ ladder (48.5-kb concatemers); lane 2, CHI-16; lane 3, CHI-18; lane 4, CHI-32; lane 5, CHI-34; lane 6, CHI-40-1; lane 7, CHI-40-2; lane 8, CHI-41; lane 9, CHI-44; lane 10, CHI-45-1; lane 11, λ ladder. (b) PFGE of S1 endonuclease-digested genomic DNA; (c) in-gel hybridization with ³²P-labeled bla_NDM-1 gene probe. Lanes 1, λ ladder; lanes 2, Klebsiella pneumoniae KP506 (NDM positive control); lanes 3, Escherichia coli UAB190 (NDM negative control); lanes 4, CHI-16; lanes 5, CHI-18; lanes 6, CHI-32; lanes 7, CHI-34; lanes 8, CHI-40-1; lanes 9, CHI-40-2; lanes 10, CHI-45-1; lanes 11, λ.

To investigate the genetic context of bla_NDM-1, we performed sequencing by primer walking in CHI-32, CHI-34, and CHI-45-1 using primers designed against the context in A. baumannii 161/07 (21). Gene probing was performed by in-gel hybridization of S1 nuclease and NotI PFGE gels. The gene probes were made using a random priming method with bla_NDM-1, ISAba125, IS26, and ISCR27 PCR products labeled with [³²P]CTP. As the primer walking PCR results were consistent for all three isolates studied, products were only fully sequenced for CHI-45-1. The full Tn125 structure was present; however, ISCR27 contains an IS26 insertion (Fig. 2). The sequence analysis of PCR amplicons revealed that the fragment of ISCR27 downstream of the IS26 insertion is present both in its normal position and upstream of bla_NDM-1. The results of probing NotI gels with bla_NDM-1, IS26, and ISCR27 suggest that it is the latter context which is more common. S1 gels demonstrated multiple plasmids in all A. baumannii isolates. The probing showed that bla_NDM-1 was on multiple bands, ranging in size from ∼45 kb to ∼300 kb. We believe most of the larger bands represent cointegrate formation rather than the presence of multiple plasmids carrying bla_NDM-1, since the bands increase in size by intervals of approximately the size of the smallest band for each isolate (Fig. 1b). Mating experiments were performed on plates and in broth. A. baumannii CHI-32, CHI-34, and CHI-45-1 were used as donors, and Escherichia coli UAB190 and Acinetobacter pittii AG3528 were used as recipients (both rifampin resistant). No transconjugants were obtained after multiple mating experiments.

FIG 2.

Gene maps of the genetic context of bla_NDM-1 in Acinetobacter species. aphA6 codes for 3′ phosphotransferase VI aminoglycoside-modifying enzyme, ble for bleomycin resistance protein, trpF for phosphoribosylanthranilate isomerase, tat for twin-arginine translocation pathway signal sequence domain protein, cutA1 for periplasmic divalent cation tolerance protein, groES for cochaperonin, groEL for chaperonin, ISCR27 for insertion sequence common repeat 27 transposase, oriIS for origin of insertion of ISCR27, and Δmfs for major facilitator superfamily (MFS) metabolite/H⁺ symporter. Arrows indicate the direction of the transcription of genes. The genes tat, cutA1, groES, and groEL are shaded in the same color because they are believed to be from a common source, with similar genes found in synteny in both Xanthomonas and Pseudoxanthomonas spp.

The A. baumannii isolates analyzed in this study are among the earliest found to produce NDM-1, having been initially identified in 2005. This is the first time that the genetic context of A. baumannii isolates from the Indian subcontinent has been analyzed. The findings are compatible with the hypothesis that bla_NDM-1 might have been disseminated from Acinetobacter to Enterobacteriaceae in South Asia. In these isolates, Tn125 has been disrupted by IS26, and subsequent rearrangement has resulted in bla_NDM-1 being within an IS26 composite transposon, which might potentially mobilize bla_NDM-1. Otherwise, the genes usually found on Tn125 are conserved, and so the genetic context is compatible with being the progenitor of the bla_NDM-1 in many of the Enterobacteriaceae for which sequences are available. bla_NDM-1 is located on plasmids in these isolates, which could facilitate mobilization of the gene to other bacterial species. The mating experiments suggest that the plasmids are nonconjugative, but they may be mobilizable with a helper plasmid. That these clinical isolates of A. baumannii producing NDM-1 were clonally related demonstrates the potential for bla_NDM-1 establishing itself in successful strain backgrounds capable of being disseminated in the hospital environment and further compromising therapeutic options in the treatment of significant bacterial pathogens.