Isolation,Molecular Characterization,and Antibiotic Susceptibility of Vibrio parahaemolyticus in Korean Seafood

Introduction

V ibrio parahaemolyticus is one of the most important causes of gastroenteritis (Daniels et al., 2000). It is associated exclusively with the consumption of raw or improperly cooked contaminated seafood, especially oysters (Daniels et al., 2000). V. parahaemolyticus is recognized as an important human pathogen globally (Joseph et al., 1982; Daniels et al., 2000; Ottaviani et al., 2008). According to the official statistics issued by the Korea Food and Drug Administration (KFDA, 2010), V. parahaemolyticus is one of the most common causes of foodborne disease in Korea, and caused 17 outbreaks with a total of 663 patients in 2005, 25 outbreaks (547 patients) in 2006, 33 outbreaks (634 patients) in 2007, 24 outbreaks (329 patients) in 2008, and 12 outbreaks (106 patients) in 2009.

East Asians, especially Koreans and Japanese, consume a unique diet. Koreans enjoy a wide variety of both raw finfish and shellfish. Although raw oysters have such high densities of V. parahaemolyticus that the consumption of raw oysters is known to cause illness in humans (Daniels et al., 2000), almost all Koreans prefer raw oysters to already cooked oysters because of their fresh taste and high nutritional value. Also, seafood cross-contaminated with raw oyster can cause high risk for V. parahaemolyticus infections in the United States (Daniels et al., 2000). In Korea especially, all seafood in fishery markets, including oysters, is sold in the same seawater and tanks, causing cross-contamination.

Virulence of V. parahaemolyticus is commonly associated with the tdh and trh genes, encoding thermostable direct haemolysin (TDH) and tdh-related haemolysin (TRH), respectively (Lee et al., 2008). Open reading frame 8 (ORF8) is considered as a potential factor causing epidemics and a genetic marker for V. parahaemolyticus O3:K6 strains (Parvathi et al., 2006). V. parahaemolyticus pandemic strains such as O3:K6 strains exhibit a unique toxRS sequence responsible for the current pandemic in many countries (Matsumoto et al., 2000).

Increasingly, there have been more reports of antibiotic resistance in Vibrio species. Emergence of microbial resistance to multiple drugs is a serious clinical problem in the treatment, increasing the fatality rate (Okoh and Igbinosa, 2010).

In this study, we evaluated the incidence of V. parahaemolyticus in Korean seafood, related seawater, and clinical isolates from May to December of 2009, except between July and September. To carry out risk assessment for pathogenic V. parahaemolyticus, we conducted PCR assays targeting several virulence genes on isolates from seafood and seawater samples. Additionally, we evaluated the susceptibility of isolates to 22 commercial antibiotics and also evaluated the presence of antibiotic-resistant genes via PCR assays. Finally, we conducted repetitive extragenic palindromic sequence polymerase chain reaction (REP-PCR) analysis for the subtyping of V. parahaemolyticus strains.

Methods

Antibiotic susceptibility profile

The antibiotic susceptibility of V. parahaemolyticus strains was determined via the standard disk diffusion method (Bauer et al., 1966). The 22 antibiotic disks (Oxoid, Cambridge, United Kingdom) used in this study were selected using guidelines recommended by the Clinical and Laboratory Standards Institute (CLSI), formerly the National Committee for Clinical Laboratory Standards (NCCLS). Antibiotic disks used in this study were shown in Figure 1. Escherichia coli ATCC 25922 was employed as a bacterial strain for quality control. The sensitivity and resistance of the strains and the zone diameter interpretive standards were assessed in accordance with the CLSI criteria (CLSI, 2005). PCR amplification was carried out to detect antibiotic-resistant genes in the strains using specific primers for the detection of SXT integrase, sul2, floR, dfr18, TetA, strB, and dfrA1 (Table 1).

OPEN IN VIEWER

FIG. 1.

Antibiotic susceptibility test of isolated strains in Korea was performed by disk diffusion method. Antibiotic resistance patterns of eight groups (Pen and β-lac, penicillins and β-lactam/β-lactamase inhibitor combinations; Ceph, cephems; Carb, carbapenems; Amino, aminoglycosides; Tet, tetracycline; Quin, quinolones; Phe, phenicol; FPI, folate pathway inhibitors), 22 commercial antibiotics (AMP, ampicillin; PRL, piperacillin; AMC, amoxicillin-clavulanic acid; SAM, ampicillin-sulbactam; TZP, piperacillintazobactam; KZ, cefazolin; FEP, cefepime; CTX, cefotaxime; FOX, cefoxitin; CAZ, ceftazidime; CXM, cefuroxime sodium; KF, cephalothin; IPM, imipenem; MEM, meropenem; AK, amikacin; CN, gentamicin; TE, tetracycline; CIP, ciprofloxacin; LEV, levofloxacin; OFX, ofloxacin; C, chloramphenicol; SXT, trimethoprim-sulfa) were presented. Antibiotic sensitivity of isolated strains was decided by zone diameter interpretive standards (CLSI, 2005) and percentages of isolates exhibiting sensitive (S), intermediate (I), and resistance (R) against various antibiotics were indicated.

Results

Genotyping using REP-PCR

In REP-PCR, 11–21 amplified bands of size 200–2000 bp were recognizable in the V. parahaemolyticus strains (Fig. 2). Eight amplified bands with approximate molecular sizes of 200, 270, 300, 350, 400, 570, 700, and 1000 bp were present in all V. parahaemolyticus strains (Fig. 2). For the rest, three bands (240, 740, and 1400 bp) were found to be common in most strains (Fig. 2). The relationships among V. parahaemolyticus strains were evaluated via cluster analysis of the REP-PCR generated patterns (Fig. 2). The V. parahaemolyticus strains were clustered into three groups (Fig. 2). Group A, the predominant group (42.8% of the total number of strains), contains ATCC 27969, ATCC 17802, and seafood isolates. Group B (32.1%) contains ATCC 33844, clinical isolates (CRS 09-17, CRS 09-72), seafood, and seawater isolates. Group C (21.4%) contains seafood isolates. According to REP-PCR analysis results, as compared with the interspecies reference strain, all the V. parahaemolyticus strains were closely related (discrimination index for V. parahaemolyticus, 0.97) (Marshall et al., 1999); they differed significantly from the reference strain of V. vulnificus ATCC 33148, having a dissimilarity value of 66.3 (Fig. 2).

OPEN IN VIEWER

FIG. 2.

Dendrogram illustrating the clustering of amplification patterns of Vibrio parahaemolyticus with repetitive extragenic palindromic sequence polymerase chain reaction (REP-PCR). The dendrogram was produced using Bionumerics software, and the clusters were determined by the unweighted pair group method, arithmetic mean (UPGMA) algorithm. The similarity units are arbitrary.

Discussion

In Korea, the demand for seafood has increased profoundly with improving income levels. As raw oysters constitute the most important source of V. parahaemolyticus and the number of V. parahaemolyticus increases in summer, purchasing oysters from retail outlets in the summer is prohibited in Korea.

The number of V. parahaemolyticus increases in summer and peaks from July to September in oysters in Korea (Lee et al., 2008). Although the pathogen is abundant during that time, samplings were not conducted because of the lack of supply. Two selective mediums (TCBS agar and CHROMagar^™ Vibrio) were used for the isolation of V. parahaemolyticus to improve accuracy and specificity. Pinto et al. (2011) previously conducted a comparison between TCBS agar and CHROMagar^™ Vibrio for the isolation of V. parahaemolyticus; CHROMagar^™ Vibrio was more accurate and specific than TCBS agar.

Fabbro et al. (2010) previously reported the sensitivity and specificity of biochemical identification for the confirmation of V. parahaemolyticus. In this study, biochemical methods were rejected due to their poor accuracy. V. parahaemolyticus evidenced a huge variability in diagnostic features among species, and thus it was difficult to identify V. parahaemolyticus via biochemical methods (Croci et al., 2007). The results of the biochemical tests evidenced lower sensitivity (48%) and specificity (18%) in cases of resistance to Vibriostatic O/129 (10 μg) (Fabbro et al., 2010). All strains employed in our study proved resistant to Vibriostatic O/129 (data not shown). Molecular confirmation was conducted via three PCR assays for high accuracy. PCR identification was conducted using specific primers, Vp. flaE (Tarr et al., 2007). In our study, the accuracy of that PCR identification was relatively low (50%) in contrast to that reported by Tarr et al. (2007), who noted that PCR identification rapidly and accurately discriminated four Vibrio species. PCR amplification for the detection of the species-specific tl gene was considered a sufficient method for the confirmation of V. parahaemolyticus (Baker-Austin et al., 2008; Lee et al., 2008). In this study, tl gene detection yielded false-positive identifications (two strains, 7.69%), as previously reported by another author (Croci et al., 2007). Molecular confirmation by PCR assays for the detection of both the toxR and tl genes is recommended for high accuracy, although Fabbro et al. (2010) noted that PCR assays for the toxR and tl genes generated the same results. In our study, 16S rRNA gene analysis produced misidentifications, owing to the strictly genetic similarity among Vibrio species, as described previously by Fabbro et al. (2010). According to Vongxay et al. (2008), the virulence of V. parahaemolyticus was correlated with the presence of the tdh and trh genes, and the tdh-positive isolates were more virulent than the trh-positive isolates; additionally, the clinical isolates evidenced higher cytotoxicity than was detected in the seafood samples (Vongxay et al., 2008). The V. parahaemolyticus O3:K6 serotype is responsible for many recent V. parahaemolyticus outbreaks, including epidemics in India, Russia, Southeast Asia, Japan, North America, and Chile (Myers et al., 2003; Fuenzalida et al., 2006). Recent outbreaks in the United States associated with V. parahaemolyticus such as the 1997 Pacific Northwest outbreak (one death, 209 patients) and the Galveston Bay outbreak (416 patients, the largest outbreak in the United States) were caused by the O3:K6 serotype (Myers et al., 2003). The V. parahaemolyticus outbreak in Korea in 1998 was also caused by the V. parahaemolyticus O3:K6 serotype (Matsumoto et al., 2000). ORF8 may contribute to the virulence of V. parahaemolyticus O3:K6 serovar strains (Okuda et al., 1997). In this study, two clinical isolates were tdh and ORF8 positive, and one of the two clinical isolates (CRS 09-17) was toxRS/new positive. Two clinical isolates were pandemic strains, according to the report of Okura et al. (2003), who reported that the detection of ORF8 is sufficient for the identification of pandemic strains. Moreover, Matsumoto et al. (2000) developed the toxRS/new PCR technique to detect the toxRS sequence of the new O3:K6 clone (since 1996) GS-PCR, group-specific PCR, and reported that GS-PCR can be used to distinguish the new O3:K6 clone from the old O3:K6 strains. It was confirmed that the clinical isolate (CRS 09-17; tdh+, ORF8+, toxRS/new+) that was toxRS/new positive was the new O3:K6 strain and the other clinical isolate (CRS 09-72; tdh+, ORF8+, toxRS/new-), which was toxRS/new negative, was the old O3:K6 strain.

Okoh and Igbinosa (2010) reported that V. parahaemolyticus evidenced resistance to ampicillin, streptomycin, tetracycline, and polymyxin B. 100% ampicillin resistance was exhibited by all V. parahaemolyticus strains. In this study, we noted high resistance in penicillins & β-lactam/β-lactamase inhibitor combinations and cephems; 100% of cefazolin resistance as well as ampicillin resistance was noted. Analysis of antibiotic resistance characteristics showed differences between the clinical isolates and the other isolates (seafood and its related environmental isolates). Although all of the isolates in this study evidenced multiple resistance, the clinical isolates were resistant to more than four antibiotics, and the other isolates were resistant to more than 11 antibiotics. The seafood and environmental isolates (22 isolates in total) evidenced an average of 16.45 antibiotic resistances per isolate compared with the clinical isolates (two isolates in total), which evidenced an average of six antibiotic resistances. Compared with a previous report regarding the antibiotic resistance of environment isolates from coastal areas in the United States (Baker-Austin et al., 2008), the number of antibiotic resistances of environment isolates in this study was greater (17.33 on average) than that noted in a previous report (7.47 on average) (Baker-Austin et al., 2008). The PCR results for the detection of antibiotic resistance genes demonstrated that TetA (25%) and strB (4.16%) were detected, which confirms resistance to tetracycline and streptomycin, respectively. The high resistance (75%) to tetracycline was observed in the antibiotic resistance patterns and high percentage (25%) of strains contained TetA. In addition, all isolates that had TetA gene showed resistance to tetracycline. This differs from the report of Okoh and Igbinosa (2010), who noted that some V. parahaemolyticus contained between one and six antibiotic resistance genes.

In this study, REP-PCR was selected as the molecular typing method because of its good reproducibility and the high discriminative ability of its fingerprints, as opposed to pulsed-field gel electrophoresis (PFGE), which is very labor-intensive and time-consuming (Wong and Lin, 2001). The REP-PCR method described herein proved to be a suitable method for the typing of V. parahaemolyticus, and could differentiate V. parahaemolyticus from other species and differentiate intraspecific strains, as reported previously by Wong and Lin (2001). Clustering based on REP-PCR did not coincide with the isolation sources or patterns of antibiotic resistance. Two clinical isolates harboring virulence genes were clustered into the same group, but the other seafood isolate clustered within the same group harbored no virulence gene. In this study, genotyping via REP-PCR yielded independent results with isolation sources, antibiotic resistance, and virulence.

Conclusion

According to a previous report (Lee et al., 2008), the number of V. parahaemolyticus was undetectable in December. In this study, V. parahaemolyticus was isolated in December, albeit at significantly reduced frequency and toxicity. In Korea, it is generally recommended that raw seafood should not be eaten from August to October, but V. parahaemolyticus outbreaks cannot be completely prevented by this practice. Therefore, it is recommended that seafood be properly cooked all year round, and this is particularly the case for children, the elderly, or adults with impaired immune systems. In this study, all strains evidenced multiple antibiotic resistance, and the seafood isolates exhibited resistance to a broad variety of commercial antibiotics (16.32 on average). Based on these results, a large-scale future study and effective prevention strategies are required; additionally, alternatives to conventional antibiotics should be developed.