Molecular Characterization of Fluoroquinolone-Resistant Aeromonas spp. Isolated from Imported Shrimp

Abstract

Sixty-three nalidixic acid-resistant Aeromonas sp. isolates were obtained from imported shrimp. Phylogenetic analysis of gyrB sequences indicated that 18 were A. enteropelogenes, 26 were A. caviae, and 19 were A. sobria. Double missense mutations in the quinolone resistance-determining region (QRDR) of gyrA at codon 83 (Ser→Val/Ile) and codon 92 (Leu→Met) coupled with a point mutation of parC at codon 80 (Ser→Ile/Phe) conferred high levels of quinolone resistance in the isolates. A majority of A. enteropelogenes and A. caviae strains harbored toxin genes, whereas only a few A. sobria strains harbored these genes. The fluoroquinolone-resistant Aeromonas spp. exhibited higher cytotoxicity than fluoroquinolone-sensitive, virulent Aeromonas spp. to rat epithelial cells.

TEXT

The United States in 2009 imported 589,670 metric tons of farmed shrimp worth more than $6 billion from Asia (27). Copious amounts of antibiotics are used in the shrimp ponds to stimulate growth and to retard the incidence of diseases caused by overcrowded, factory farm conditions (4, 5, 9). The indiscriminate use of these antibiotics may select bacteria resistant to multiple antibiotics, and such bacteria may transfer their antibiotic resistance determinants to pathogenic bacteria (14, 16).

Aeromonads are ubiquitous, psychrophilic, Gram-negative microbes commonly found in fresh water, estuaries, and other coastal waters (6, 7, 8). They are known to cause hemorrhagic septicemia in aquatic organisms. These microbes also cause several diseases in penaeid shrimp (5). Antibiotics, such as fluoroquinolones, are used to curtail infections (5, 9). However, prolonged abuse of antibiotics could result in the selection of fluoroquinolone-resistant Aeromonas spp. (20). The presence of fluoroquinolone-resistant aeromonads in shrimp could pose public health concerns because these bacteria are associated with outbreaks of human infectious diseases in immunocompromised patients (9, 29). Quinolones are the drugs of choice for the treatment of Aeromonas-induced infections (11). Thus, U.S. regulatory agencies, such as the FDA and CDC (28), want to limit the prevalence of fluoroquinolone-resistant bacteria in food-producing animals to protect the efficacy of the drugs. Since little information is available on the prevalence of fluoroquinolone-resistant aeromonads in imported shrimp, we decided to investigate the prevalence of fluoroquinolone-resistant Aeromonas spp. and the mechanism of resistance in these bacteria to the antibiotic. We report here for the first time that imported shrimp may be a reservoir of virulent, fluoroquinolone-resistant strains of A. enteropelogenes, A. caviae, and A. sobria.

Isolation, characterization, and identification of bacteria.

Bacteria were isolated from frozen shrimp (Penaeus monodon) imported from Thailand. Typically, 2 g of thawed shrimp sample was enriched for 6 h in alkaline peptone medium (pH 8.0) supplemented with 30 μg/ml of ampicillin and nalidixic acid. Enriched samples were streaked on MacConkey agar and incubated at 37°C overnight. Presumptive positive colonies were biochemically characterized and identified at the genus level by using the Vitek GNI⁺ card with VTK-R07-01 software (bioMérieux Vitek, Hazelwood, MO) and by fatty acid methyl ester analysis (MIDI, Newark, DE).

Determination of MICs.

The MICs of nalidixic acid and ciprofloxacin were determined by the broth dilution method using Mueller-Hinton broth (15).

PCR amplification and sequencing of gyrB and phylogenetic data analysis.

The gyrB gene was amplified from the template DNA of all fluoroquinolone-resistant aeromonads by PCR (25, 30). Sequencing reactions were performed on both strands of DNA after purification of PCR amplicons. After the assembly and alignment of the contigs, the sequences were finalized and compared with those available in the GenBank database by using NCBI BLAST to identify the alignment with closely related aeromonads. The nucleotide sequences were aligned by using ClustalX 2.1, and the phylogenetic evolutionary tree was constructed by the neighbor-joining method (24) with the MEGA 5.1 program (12).

Primer design and detection of fluoroquinolone resistance genes by PCR.

The presence of qnrAB, gyrAB, and parC was detected in the template DNA by PCR as detailed elsewhere (18).

Detection of mutations in the QRDRs.

For detection of mutations in the quinolone resistance-determining regions (QRDRs), purified target genes (gyrAB and parC) were sequenced with the primers used for amplification (18).

Detection of toxin genes by PCR.

PCR assays for the amplification of the aerolysin (aer), cytotoxic enterotoxin (act), and cytotonic enterotoxin (ast and alt) genes were performed with the template DNA of the isolates (17).

Epithelial cell cytotoxicity assay.

The cytotoxicities of fluoroquinolone-resistant A. enteropelogenes, A. caviae, and A. sobria strains were compared with those of fluoroquinolone-sensitive but pathogenic strains of A. hydrophila (ATCC 7966), A. veronii (ATCC 9071), and A. sobria AS3 (in-house culture). Cytotoxic activities were determined with the lactate dehydrogenase (LDH) assay using the CytoTox 96 nonradioactive cytotoxicity assay (Promega, Madison, WI) according to the manufacturer's instructions. Statistical significance was calculated using the unpaired t test in GraphPad software.

Isolation, characterization, and identification of fluoroquinolone-resistant Aeromonas spp.

Approximately 317 bacterial colonies exhibiting typical aeromonad morphological characteristics were isolated from more than 364 shrimp samples. Sixty-three of the 317 (ca. 20%) isolates were resistant to nalidixic acid. Data from the GNI⁺ Vitek system indicated that all 63 isolates were Aeromonas spp.

PCR amplification, sequencing of gyrB gene, and phylogenetic analysis.

A 1.2-kb region of the gyrB gene was amplified and sequenced. The gyrB sequences derived were aligned with those from reference strains, including A. enteropelogenes AN-35 (AY987508), A. caviae B14 (JQ234886.1), A. hydrophila 2WCL102 (JQ085479), A. veronii PY50 (HQ540320), A. allosaccharophila JA07 (GU205210), A. jandaei 4pM29 (FJ940783), A. media JHS07 (GU205212), A. schubertii HYB2 (HQ731459), A. popoffii JW08 (GU205208), A. sobria JY081016-1 (GQ232760), A. salmonicida subsp. smithia JF4097 (FN394064), and A. sharmana DSM 17445T (AM490259), and phylogenetic analysis was undertaken (25, 30). The gyrB sequences of 18/63 isolates (for example, strains AH811 and AH519) showed maximum sequence similarity with those of A. enteropelogenes AN35, and the gyrB sequences of 26/63 (for example, strains AV810 and AV1) had maximum sequence similarity with a reference strain of A. caviae B14 (Fig. 1). The gyrB sequences of 19/63 isolates (for example, strain AS110) had the maximum sequence similarity with reference strain A. sobria JY081016-1 (Fig. 1). All strains were resistant to 64 μg/ml of nalidixic acid, and eight strains were resistant to >256 μg. The MIC values of ciprofloxacin (0.5 to 128 μg/ml) for these eight strains were determined. All strains were resistant to 4 μg, 4/8 strains were resistant to 8 μg, and 2/8 strains were resistant to 128 μg of the antibiotic.

Fig 1.

Phylogenetic tree based on the gyrB sequence, showing the relationships of select aeromonads. Sequences were finalized and compared with those available in GenBank database by using NCBI BLAST. Phylogenetic tree was constructed by the neighbor-joining method with the MEGA 5.1 program. Numbers at branching points represent bootstrap values from 1,000 replicates (only values of >50%).

Amplification of gyrA, gyrB, parC, qnrA, and qnrB and analysis of gene sequences.

Synthetic oligonucleotide primers (Table 1) specific for the amplification of the 462-bp gyrA, 276-bp gyrB, and 232-bp parC genes were used to amplify the respective PCR amplicons from the template DNA of the isolates. Sequence analysis of the QRDRs of the gyrA and parC amplicons indicated 3 different types of mutations that confer high levels of quinolone resistance in these isolates (Fig. 2). These three types were (i) strains of A. enteropelogenes and A. caviae having double point mutations exclusively in the QRDR of gyrA, (ii) strains of A. enteropelogenes and A. caviae having double mutations in gyrA accompanied by a point mutation in parC, and (iii) strains of A. sobria having single point mutations in the QRDRs of gyrA and parC.

Table 1.

Sequences of oligonucleotide primers used for the detection of virulence genes and fluoroquinolone resistance genes in Aeromonas spp. isolated from shrimp

Gene	Product size (bp)	Primer direction,a sequence	Tm (°C)b
aer	431	F, CCTATGGCCTGAGCGAGAAG	63
		R, CCAGTTCCAGTCCCACCACT
act	231	F, AGAAGGTGACCACCACCAAGAACA	65
		R, AACTGACATCGGCCTTGAACTC
ast	331	F, TCTCCATGCTTCCCTTCCACT	63
		R, GTGTAGGGATTGAAGAAGCCG
alt	442	F, TGACCCAGTCCTGGCACGGC	64
		R, GGTGATCGATCACCACCAGC
gyrA	462	F, CGACCTTGCGAGAGAAAT	59
		R, GTTCCATCAGCCCTTCAA
gyrB	276	F, GCGCGTGAGATGACCCGCCGT	56
		R, CTGGCGGTAGAAGAAGGTCAG
parC	232	F, CTTTGCGCTACATGAATTTA	64
		R, CAGGTTATGCGGTGGAATAT
qnrA	580	F, AGAGGATTTCTCACGCCAGG	62
		R, TGCCAGGCACAGATCTTGAC
qnrB	264	F, GGMATHGAAATTCGCCACTG	60
		R, TTTGCYGYYCGCCAGTCGAA

^aF, forward; R, reverse.

^bT_m, melting temperature.

Fig 2.

Sequence analysis of mutations in quinolone resistance-determining regions (QRDRs) of Aeromonas spp.

Analysis of the presence of double mutations exclusively in gyrA indicated a serine residue (S) at position 83 replaced by a valine (V) and a leucine residue (L) at position 92 replaced by a methionine (M) (Fig. 2A). In addition, another double mutation variant in the QRDR of gyrA was also detected. Mutation at position 83 resulted in serine replacement by isoleucine, and mutation at position 92 resulted in the replacement of leucine by methionine (Fig. 2B). Strains of A. enteropelogenes and A. caviae that harbored the above-described double mutations also contained a point mutation in parC. The point mutation resulted in the replacement of serine by isoleucine at position 80 in parC amplicons (Fig. 2D).

Amplification and analysis of gyrA from strains of A. sobria indicated only a single point mutation. The mutation resulted in the replacement of serine by isoleucine at position 83 (Fig. 2C). These strains also had a point mutation at position 80 (Ser→Phe) in parC amplicons (Fig. 2E). No mutations were detected in gyrB, qnrA, and qnrB.

Amplification of toxin genes (aer, act, ast, and alt).

The PCR protocol (Table 1) unambiguously amplified the aer gene (Fig. 3A) from the template DNA of 17/18 (94.0%) A. enteropelogenes strains, 24/26 (92.0%) A. caviae strains, and 3/19 (15.0%) A. sobria strains. A pair of synthetic act-specific primers (Table 1) amplified a 231-bp PCR amplicon from a majority of A. enteropelogenes strains (16/18, 88.0%), 92.0% of A. caviae isolates (Fig. 3B), and 7/19 (36.0%) A. sobria strains. Template DNA from all aer-positive A. enteropelogenes also contained the act gene. However, only 14 of the 26 strains (54.0%) of A. caviae harbored the act gene. None of the A. sobria isolates contained both aer and act genes. Oligonucleotide primers failed to amplify the two cytotoxic enterotoxin genes (331-bp ast and 442-bp alt) from the template DNA of any of the 63 isolates.

Fig 3.

PCR amplification of virulence genes (aer and act) from the template DNA of fluoroquinolone-resistant aeromonads isolated from imported shrimp. (A) Lane 1, 100-bp molecular size marker; lane 2, PCR negative; lanes 3 to 7, 431-bp aerolysin (aer) amplified from the template DNA of the isolates. (B) PCR amplification of the cytotoxic enterotoxin (act). Lane 1, 100-bp molecular size marker; lane 3, PCR negative; lanes 2 and 4 to 7, 231-bp act gene amplified from the template DNA.

Determination of cytotoxicity of fluoroquinolone-resistant aeromonads.

Data from our cytotoxicity assays indicate that both fluoroquinolone-resistant and fluoroquinolone-sensitive strains of Aeromonas spp. caused an increase in LDH release in the culture supernatant after 24 h postinfection, indicating that they were cytotoxic to the rat epithelial cells (Fig. 4). However, the cytotoxicity of the fluoroquinolone-resistant Aeromonas spp. showed statistically significant higher LDH activity (P < 0.03) than the fluoroquinolone-sensitive but virulent strains of A. veronii ATCC 9071, A. hydrophila ATCC 7966, and A. sobria AS3 (Fig. 4). Furthermore, fluoroquinolone-resistant A. caviae AV14 had the highest cytotoxicity among all the strains that were investigated.

Fig 4.

Cytotoxic activities of Aeromonas spp. isolated from imported shrimp. LDH activities were estimated for the fluoroquinolone-resistant Aeromonas spp. isolated from imported shrimp and compared with the activities of fluoroquinolone-sensitive but virulent strains A. veronii ATCC 9071, A. hydrophila ATCC 7966, and A. sobria AS3 (in-house strain). The LDH assay was performed with a CytoTox 96 nonradioactive kit. Error bars show standard deviations.

The characterization and identification of the members of the genus Aeromonas by conventional biochemical methods is often inaccurate and controversial, warranting the use of molecular techniques that provide unambiguous, precise diagnosis (25). Recently, molecular biology methods have been used in the identification of aeromonads. The amplification and sequencing of the gyrB gene (encoding the B-subunit of DNA gyrase, a type II DNA topoisomerase) has proved to be an accurate, unequivocal molecular chronometer in the identification, strain differentiation, and phylogenetic analysis of the genus Aeromonas and other Gram-negative bacteria (25, 30, 31). Three different species of aeromonads were identified among 63 fluoroquinolone-resistant bacteria by the gyrB gene sequence analysis. Eighteen of the 63 isolates were identified as A. enteropelogenes, 26/63 as A. caviae, and 19/63 as A. sobria. To the best of our knowledge, little information is available on the isolation and identification of A. enteropelogenes from food-producing ecosystems. The identification of this rarely reported strain indicates the usefulness of gyrB sequencing in unequivocal identification of aeromonads.

A majority of mutations described to date in Gram-negative bacteria have been found within the N termini of the gyrA, gyrB, and parC proteins, between codons 83 and 87 in the QRDRs (1, 3, 10, 18, 19, 22, 23). Mutations at these locations alter the binding of quinolones to the active site and lead to reduced susceptibility (22, 23). However, sequence analysis of all highly quinolone-resistant strains of Aeromonas spp. in the present study indicated amino acid changes in the GyrA subunit, at positions 83 and 92. The exclusive double missense mutations in the gyrA QRDR, independent of mutation in parC, may be an alternate mechanism involved in the regulation of high-level quinolone resistance in aeromonads. Double missense mutations in the gyrA QRDR of Pseudomonas (1) and Escherichia coli (3, 10), particularly at positions 83 and 87, are associated with increased levels of quinolone resistance.

Reports have indicated the necessity of concurrent point mutations in both gyrA and parC for high-level quinolone resistance in several clinical and environmental strains of aeromonads (19, 23). These isolates had double mutations at codon 83 (Ser→Ile/Arg) of the gyrA QRDR, coupled with a missense mutation at position 80 (Ser→Ile/Phe) in the parC QRDR. Contrarily, our results indicate unique double missense mutations in gyrA at codon 83 (Ser→Ile) and codon 92 (Leu→Met) coupled with a point mutation in parC at position 80 (Ser→Ile/Phe). It is possible that these mutations could also be a vital, alternate mechanism involved in regulating high levels of quinolone resistance in aeromonads. In addition, to our knowledge, the point mutation at position 92 (Leu→Met) is a novel, unique mutation recorded for the first time in aeromonads. It is possible that mutations at codon Ser83 may confer high levels of quinolone resistance without the necessity of concurrent point mutations in parC in aeromonads isolated from imported shrimp.

The pathogenicity of Aeromonas spp. is attributed to various putative virulence genes (6, 7, 21). The mechanisms of pathogenesis are complex and multifactorial. Aeromonads are known to produce at least four different kinds of enterotoxins (6, 7, 17), and many of these putative virulence genes exhibit hemolytic, cytotoxic, and enterotoxic activities (7). However, the occurrence and prevalence of these toxin genes in aeromonads isolated from different ecosystems are ambiguous. Chacon et al. (6) indicated that 65.0% of the aeromonads isolated from clinical and environmental samples harbored the aer and act genes. All freshwater fish isolates of A. hydrophila harbor the aer gene, but strains of A. veronii do not contain the gene (8). However, Nawaz et al. (17) reported that template DNA from 96.0% of the A. veronii strains isolated from catfish harbored the putative aerolysin gene. However, very little is known about the virulence mechanisms in A. enteropelogenes and A. caviae. Our results indicate that the majority of the fluoroquinolone-resistant A. enteropelogenes and A. caviae strains and a few of the A. sobria isolates from imported shrimp are putatively virulent because they harbored aer and act genes.

The importance of various types of secretion systems, as well as virulence factors, in the cytotoxicity caused by Aeromonas spp. has been the focus of research for the development of new therapeutic agents (2). Earlier studies have demonstrated that a type 3 secretion system (T3SS) is required for A. hydrophila pathogenesis. Yu et al. showed that insertional inactivation of two of the T3SS genes (aopB and aopD) led to decreased cytotoxicity in carp epithelial cells (32). Another mechanism of pathogenesis includes a requirement of bacterium-host cell interaction for the type 6 secretion system (T6SS) to induce cytotoxicity in eukaryotic cells. Suarez et al. delineated the importance of a T6SS effector protein, VgrG1, from A. hydrophila that induces host cell toxicity by ADP ribosylation of actin (26). Another study demonstrated the importance of T2SS in A. veronii (13). Results from our study indicate that the mutations in the gyrA and parC genes that are components of T3SS may play a vital role in increased cytotoxicity of the bacteria. However, further studies are needed to confirm the role of these mutations in increased cytotoxicity of fluoroquinolone-resistant strains.