Presence of VIM-Positive Pseudomonas Species in Chickens and Their Surrounding Environment

ABSTRACT

Metallo-β-lactamase gene bla_VIM was identified on the chromosome of four Pseudomonas sp. isolates from a chicken farm, including one Pseudomonas aeruginosa isolate from a swallow (Yanornis martini), one Pseudomonas putida isolate from a fly, and two P. putida isolates from chickens. The four isolates shared two variants of bla_VIM-carrying genomic contexts that resemble the corresponding regions of clinical metallo-β-lactamase-producing Pseudomonas spp. Our study suggests that the surveillance of carbapenemase-producing bacteria in livestock and their surrounding environment is urgently needed.

TEXT

Carbapenems are critically important antimicrobials as a last line of defense against multidrug-resistant Gram-negative bacterial infections (1, 2). As such, the increasing prevalence of carbapenemase-producing isolates in animal husbandry is of great concern. While the metallo-β-lactamase (MBL)-producing bacteria have been commonly identified from food animals (3–7), bla_MBL-carrying Pseudomonas spp. are rarely reported in animal husbandry or the surrounding environment. Although we have reported the high prevalence of NDM in Enterobacteriaceae isolates from poultry production in Shandong Province (7), carbapenemase-producing non-Enterobacteriaceae isolates have not been identified in the same region. Here, we report four chromosome-borne VIM-positive Pseudomonas isolates: one Pseudomonas aeruginosa isolate from a swallow (Yanornis martini), one Pseudomonas putida isolate from a fly, and two P. putida isolates from chickens. The bla_VIM-2 gene was identified in the P. aeruginosa isolate, but 27 bp at the 3′-terminal region of bla_VIM-2 was truncated by an IS6100 element in three P. putida isolates, resulting in a new variant of bla_VIM gene, which was designated bla_VIM-48; the bla_VIM-carrying genomic regions in these Pseudomonas sp. isolates closely resembled the corresponding regions of clinical MBL-producing Pseudomonas spp.

Ninety-eight nonduplicated samples were randomly collected, with an informed consent form, from a commercial chicken farm in Shandong Province, China (chicken cloaca swabs, n = 30; flies, n = 30; dog anal swabs, n = 17; swallow fecal swabs, n = 10; farmer fecal swabs, n = 6; sewage, n = 5); the procedures for the collection of all samples were consistent with those previously reported (7). All samples were plated on CHROMagar Pseudomonas (CHROMagar, Paris, France) containing 8 μg/ml meropenem (Ouhe Technology Company, Beijing, China). Putative Pseudomonas colonies that turned blue when tested were recovered from 17 samples, and one colony from each sample was identified to the species level by 16S rRNA sequencing (8). Of these, four isolates were bla_VIM positive, confirmed by PCR and sequencing (9). Further matrix-assisted laser desorption ionization–time of flight mass spectrometry (MALDI-TOF MS; Bruker Daltonik GmbH, Bremen, Germany) analysis confirmed that the positive isolates consisted of one P. aeruginosa isolate from swallow fecal swabs, including DZ-B1; one P. putida isolate from a fly, DZ-F23; and two P. putida isolates, DZ-C20 and DZ-C18, from chicken cloaca swabs.

MIC analysis showed that the four Pseudomonas sp. isolates were almost resistant to all of the β-lactam antibiotics tested, including meropenem, imipenem, aztreonam, and ceftazidime, with only isolate DZ-B1 showing susceptibility to aztreonam (4 μg/ml) (Table 1) (10, 11). To further investigate the genetics background of the four isolates, whole-genome sequencing was conducted as described previously (7), and draft assembly sequences were searched against the antibiotic resistance gene database (https://cge.cbs.dtu.dk/services/ResFinder/), which confirmed the presence of bla_VIM and other antibiotic resistance genes (Table 1). Additionally, to investigate the location of the resistance element, the genomic DNA of four VIM-producing isolates was digested with I-Ceu1 and separated by pulsed-field gel electrophoresis (PFGE). The result and Southern blot analysis showed that bla_VIM was located on the chromosomes of all four Pseudomonas isolates (see Fig. S1 in the supplemental material). Moreover, the core genome phylogenetic analysis was conducted to reveal the relationship between the four isolates and other known Pseudomonas isolates carrying bla_VIM in the NCBI database (see Table S1 in the supplemental material) using the Parsnp program (12). The phylogenetic tree revealed that three bla_VIM-carrying P. putida isolates had distinct genomic heterogeneity, which is consistent with PFGE analysis of SpeI patterns (Fig. S1). The ST385 of the P. aeruginosa isolate DZ-B1 from our study was confirmed by whole-genome sequencing. This sequence type has been associated with clinical P. aeruginosa isolates from India (13). The genome context of DZ-B1 was closely related to that of seven bla_VIM-2-carrying clinical P. aeruginosa isolates from GenBank (see Fig. S2 in the supplemental material).

TABLE 1.

Antimicrobial susceptibility profiles of Pseudomonas putida isolates DZ-C20, DZ-C18, and DZ-F23; Pseudomonas aeruginosa DZ-B1; and recombinant E. coli strains HSG398-V01 and HSG398-DH5α

Bacterial isolate	Resistance gene(s)	MIC (mg/liter) ofa:
		MEM	IMI	ATM	CAZ	GEN	CIP	COL	TGC	TMP
DZ-C18	blaCARB-4, blaVIM48, strA, strB, aacA4, aadA13, QnrVC1, mph(A), cmlA1, floR, tet(A), tet(G), sul1	128	128	256	128	64	32	2	8	≥256
DZ-C20	blaCARB-4, blaVIM-48, strA, strB, aacA4, aphA6, QnrVC1, mph(A), floR, catB, tet(G), sul1	64	32	128	128	128	32	4	2	≥256
DZ-F23	blaCARB-4, blaVIM-48, strA, strB, aacA4, aphA6, QnrVC1, mph(A), floR, tet(G), sul1	64	32	32	32	16	32	1	8	≥256
DZ-B1	blaVIM-2, blaOXA-50, ampC, strA, strB, aacA4, aph(3′)-IIa, aph(3′)-IIb, aac(6′)-IIa, catB7, floR, tet(G), sul1, dfrB1, fosA	32	32	4	64	256	256	2	8	≥256
HSG398-V01	blaVIM-48	0.125	1	0.032	2
HSG398-DH5α		0.0625	0.25	0.032	0.125

^aAntimicrobial agents are abbreviated as follows: MEM, meropenem; IMI, imipenem; ATM, aztreonam; CAZ, ceftazidime; GEN, gentamicin; CIP, ciprofloxacin; COL, colistin; TGC, tigecycline; TMP, trimethoprim.

Four bla_VIM-carrying contigs were identified from four isolates and confirmed by PCR and sequencing (primers listed in Fig. S3 in the supplemental material). Analysis of the flanking regions of bla_VIM-2 on the chromosome of DZ-B1 revealed that it was located in a Tn5090-like transposon bracketed by two 25-bp inverted repeats, IRi and IRt (Fig. 1A), suggesting that the whole region can be mobilized with the tni machinery (14, 15). This complete transposon shared 92.3% (7,296/7,907) nucleotide sequence identity to a P. aeruginosa Tn5090-like transposon, also containing bla_VIM-2 gene, isolated from a Chinese patient (GenBank accession no. AM993098.1), and even greater sequence identity (99.4%, 4,147/4,171) was found within the corresponding tniC-bla_VIM-2-aacA4-dhfr2 gene array located at the 3′ end of the Tn5090-like transposon. The Tn5090-like transposon contained a conserved segment with an integrase gene, intl1, at the 3′ end. Intl1 is associated with the integration of resistance gene cassette array bla_VIM-2-aacA4-dhfr2, which is in the opposite orientation and confers resistance to carbapenems, some aminoglycosides, and trimethoprim, respectively. Another conserved gene cluster, tniA-tniB-hp-tniC, was located at the 5′ end of the transposon (14). tniA and tniB were predicted to be involved in the transposition, while tniC codes for a recombination protein that differs from those encoded by intl1, tniA, and tniB (Fig. 1A) (14). This gene arrangement, in combination with intl1, is involved in the transfer of a bla_VIM-2-carrying cassette in clinical Pseudomonas spp. (16–19).

FIG 1.

(A) Genetic context of bla_VIM in Pseudomonas sp. isolates from chickens, a fly, and a swallow and structural comparison with the corresponding genetic regions in Tn5090-like and Tn6217 transposons. Arrows indicate the positions and directions of transcription of the genes. Different genes are indicated by different types of shading. (B) Nucleotide sequence of the region encompassing the 3′ ends of IS6100 and bla_VIM-48. The coding regions of the two open reading frames are marked by gray shading, and the right inverted repeat (IRR) of IS6100 is underlined.

A 9,875-bp fragment containing the bla_VIM-2-like gene was observed in three P. putida strains that shared 99.9% sequence identity. Comparison analysis revealed that the bla_VIM-2-like gene had a 27-bp deletion at the 3′-terminal region, including its original stop codon, versus the 801-bp bla_VIM-2 in DZ-B1 and the previously reported P. aeruginosa isolates. Immediately downstream of the 27-bp deletion was an intact insertion element, IS6100, and the truncated bla_VIM-2 gene appeared to be terminated at the TAG stop codon located in the right inverted repeat of IS6100. This resulted in an 810-bp open reading frame, designated VIM-48 (GenBank accession no. KY362199), that was 9 bp longer than the previously reported gene (Fig. 1B). Further analysis of the flanking regions revealed that it was very similar to the Tn6217 region in the bla_IMP-9-carrying plasmid pOZ176 from a clinical P. aeruginosa isolate from Guangzhou, China (20). The upstream and downstream regions of bla_VIM-48 contained the aacA4-intI1-Tn1403-like and sul1-hp-IS6100 gene clusters, respectively. These two regions shared 99.9% (2,792/2,794) and 99.9% (5,726/5,733) nucleotide sequence identity, respectively, to the corresponding regions of Tn6217; however, the downstream fragment was inverted, perhaps leading to the missing 27-bp segment at the 3′ end of bla_VIM-2 (Fig. 1A).

Similar to the Tn5090-like transposon in P. aeruginosa isolate DZ-B1, a 3,801-bp segment harboring the IS6100-bla_VIM-48-aacA4-intI1 gene cluster in the three P. putida isolates was bracketed by two 25-bp inverted repeats, IRi and IRt, suggesting that it reached its current location by transposition (Fig. 1A) (21). Moreover, compared to a 6,942-bp fragment in the 3′ region of the bla_VIM-2-carrying segment in P. aeruginosa DZ-B1, only a 405-bp segment containing dhfr2 was absent from the P. putida strains. The remaining two nucleotide fragments were highly similar, sharing 100% (1,444/1,444) and 99.9% (5,091/5,097) identity, respectively (Fig. 1A).

The 810-bp bla_VIM-48 gene was amplified by PCR and confirmed by sequencing using the primers listed in Fig. S3 and then cloned into vector pHSG398 (TaKaRa, Dalian, China), which resulted in a recombinant plasmid pHSG398-V01. The vector and recombinant plasmid were introduced into Escherichia coli DH5α (TaKaRa, Japan) by electroporation, resulting in two recombinant E. coli strains, designated HSG398-V01 and HSG398-DH5α. MIC analysis showed that HSG398-V01 was more active than HSG398-DH5α against imipenem and ceftazidime, with 4- and 8-fold increases, respectively (Table 1), which suggested that this bla_VIM-48 gene is functional in the three P. putida isolates. Phenotypic detection of metallo-β-lactamases was performed on the three P. putida isolates according to the previous report (22), and synergy between imipenem and EDTA was observed for all three strains (data not shown).

To date, 47 variants of the bla_VIM gene have been reported (23), and all of them were 801 bp long and differed in several amino acids, whereas the bla_VIM48 gene in this study was 810 bp long, possibly due to the inverted insertion of a IS6100-contig-containing fragment, revealing a new generation mechanism for the variant of bla_VIM gene. Similar to the spread of bla_VIM-2 in bacteria of clinical origin, the mobile genetic elements, such as Tn5090, Tn1403, intl1, and IS6100, may play an important role in the dissemination of the bla_VIM gene and its conserved flanking regions in bacteria among food animals (16, 17, 24). Our study has found MBL-producing Pseudomonas isolates not only in the livestock but also in their surrounding environment, suggesting that surveillance of carbapenemase-producing bacteria in livestock and their surrounding environment is urgently needed.

Accession number(s).

All genome assemblies of 4 strains were deposited in GenBank and are registered under BioProject accession number PRJNA381373, and the new VIM enzyme was designated VIM-48 and deposited in GenBank under accession number KY362199.