Prevalence,Antibiotic-Resistance,and Virulence Characteristics of Vibrio parahaemolyticus in Restaurant Fish Tanks in Seoul,South Korea

Introduction

Vibrio parahaemolyticus is a gram-negative, halophilic, rod-shaped bacterium with flagella that prefers warm temperatures, and that occurs in marine or estuarine environments, but that is sometimes even found in freshwater environments (Broberg et al., 2011; Ceccarelli et al., 2013; Haley et al., 2014; Dong et al., 2016). Since it was first isolated in Japan in 1950, V. parahaemolyticus (especially the O3:K7 serotype) has been recognized as a major source of foodborne infectious disease in many countries, mainly via ingestion in raw, undercooked, or inappropriately handled seafood (Nair et al., 2007; Velazquez-Roman et al., 2014). During the last decade, V. parahaemolyticus has been reported to be the third most common foodborne disease-causing bacteria in Korea (FoodSafetyKorea, 2017).

A rare large-scale foodborne V. parahaemolyticus outbreak arose due to cross-contamination during the preparation and processing of squid in Seoul, Korea in 2017 (Jung, 2018). A previous study by Hara-Kudo et al. (2013) showed V. parahaemolyticus-inoculated seawater to be capable of causing cross-contamination of fish surfaces, viscera, gills, and cutting boards. Nevertheless, many seafood restaurants in Korea still store live fish and shellfish in seawater-filled fish tanks for raw consumption. Moreover, these are stocked from live-fish transportation units that are highly likely to be contaminated with V. parahaemolyticus (Park and Kim, 2018).

The three major syndromes caused by V. parahaemolyticus infection are gastroenteritis, wound infections, and septicemia. Gastroenteritis is often self-resolving, such that patients are only treated to maintain hydration; however, immunocompromised patients sometimes require treatment with antibiotics such as tetracycline, doxycycline, and/or ciprofloxacin (Broberg et al., 2011; Velazquez-Roman et al., 2014). Unfortunately, the overuse of antibiotic agents in human medicine and veterinary/aquaculture applications has led to the emergence of antibiotic resistance in many bacteria, including ampicillin resistance in V. parahaemolyticus (Shaw et al., 2014; Letchumanan et al., 2015; Elmahdi et al., 2016).

Although most environmentally occurring V. parahaemolyticus strains do not produce thermostable direct hemolysin (TDH) and/or TDH-related hemolysin (TRH), which are mainly pathogenic, all V. parahaemolyticus strains possess a range of virulence factors, including adhesins, toxins, and secreted effectors, most and increasingly often including type III secretion systems (T3SSs) and type VI secretion systems (T6SSs), respectively (Ceccarelli et al., 2013). TDH and TRH are considered to be major V. parahaemolyticus virulence factors that cause hemolysis, cytotoxicity, and enterotoxicity by forming tetrameric pores on the host cell surface (Broberg et al., 2011; Zhang and Orth, 2013). T3SS1 effectors occur in environmental V. parahaemolyticus isolates, and enable bacteria to both avoid host immune responses, and induce host-cell autolysis. Conversely, T3SS2 effectors are produced by clinical V. parahaemolyticus isolates, and in contrast to the T3SSs of other bacteria, enable V. parahaemolyticus bacteria to invade, survive, and replicate in host cells (except phagocytic cells); thus, they function with hemolysin to incur enterotoxicity (Ham and Orth, 2012; Ceccarelli et al., 2013; Zhang and Orth, 2013). Quorum sensing negative and positive regulates T6SS1 (produced by clinical V. parahaemolyticus isolates) and T6SS2 (produced by both clinical and environmental V. parahaemolyticus isolates) activity, respectively, while surface sensing produces the opposite effect. Similarly, warm-marine and low-salt conditions promote T6SS1 and T6SS2 activity, respectively (Ceccarelli et al., 2013; Salomon et al., 2013; O'Boyle and Boyd, 2014). Urease encoded by ureC appears to facilitate bacterial colonization of the gastric mucosa by enabling V. parahaemolyticus bacteria to protect themselves against stomach acid. Furthermore, urease detection in V. parahaemolyticus is often associated with the presence of the trh gene (Costa et al., 2013).

To date, studies investigating V. parahaemolyticus in Korea have analyzed strains isolated from patient, marine-food, and seawater samples, and have focused on the production of hemolysin, but have not investigated the production of virulence-related genes (Ryu et al., 2010; Cho, 2014; Park et al., 2016, 2018). Thus, the present study aimed to evaluate the prevalence, antibiotic-resistance, and virulence characteristics of V. parahaemolyticus isolated from restaurant fish-tank water in Seoul, South Korea.

Materials and Methods

Results

Discussion

The incidence of V. parahaemolyticus in the analyzed fish-tank water samples was similar to that previously detected in fish-tank water samples collected from urban seafood restaurants in Busan, Korea (Cho, 2014). Most notably, the urban-based prevalence reported in that study (43.8%) (Cho, 2014) and our present study (50.7%) were both markedly higher than the 20% range detected in coastal areas, and environmental sea- and estuary-water samples (Velazquez-Roman et al., 2012; Haley et al., 2014; Silva et al., 2018). It is likely that the higher prevalence reported in the urban fish-tank water samples is the result of an accumulation effect caused by the standard method of replacing tank water, whereby two-thirds of the water is replaced, and one-third is retained. In contrast, the continuous circulation of seawater in the environment likely reduces bacterial accumulation. Notably, temperature has been suggested to modulate bacterial levels in both fish-tank water (Song et al., 2017) and seawater (Haley et al., 2014), consistent with the fact that a higher prevalence of V. parahaemolyticus isolates was detected in September than October in the present study, and that previous studies have likewise reported that V. parahaemolyticus is predominantly detected in seawater in Korea between July and September (Park et al., 1993, 2018). In fact, it is likely that the herein observed increased prevalence of V. parahaemolyticus in the fish tanks sampled in October reflect lower seasonal environmental seawater contamination rates. This is probably because the average and standard deviation of seawater temperature in Korea differed by 23.95°C and 1.17 in September and 21.38°C and 1.53 in October, respectively (KMA, 2016).

Previous studies have reported that ampicillin resistance in V. parahaemolyticus strains has increased worldwide (Elmahdi et al., 2016). Notably, V. parahaemolyticus isolates collected from seawater have been thus far found to have little resistance to antibiotics other than β-lactams such as ampicillin (Shaw et al., 2014; Silva et al., 2018), while isolates collected from commercial marine foods in Seoul (Ryu et al., 2010) and oysters in Korea (Kang et al., 2017) have been shown to be highly resistant to both streptomycin and vancomycin/streptomycin, and 03B2-lactam antibiotics. V. parahaemolyticus isolated from patients in China (Xie et al., 2017) and India (Pazhani et al., 2014) have similarly been shown to be resistant to ampicillin and streptomycin. Although the detected antibiotic resistance was relatively low in the present study, the observed trend was similar to those described by these previous reports. Currently, the antibiotic drug recommended for the first-line treatment of V. parahaemolyticus infections is tetracycline, followed, if necessary, by treatment with a combination of the broad-spectrum antibiotics cephalosporin and doxycycline (Elmahdi et al., 2016). Notably, some V. parahaemolyticus isolates screened in the present study were shown to be resistant to tetracycline, ceftazidime, and cefotaxime; thus, an accurate and up-to-date understanding of antibiotic-resistance trends is essential in the clinical setting to effectively prevent and/or treat V. parahaemolyticus-induced disease.

To date, virulence-related genes have been rarely detected in V. parahaemolyticus isolates collected from seawater, but are often identified in clinical isolates (Ceccarelli et al., 2013). Interestingly, one previous study investigating V. parahaemolyticus in samples of seawater collected from Black sea reported results similar to those presented here, in that all analyzed V. parahaemolyticus isolates had T3SS1, but often lacked other secretion systems (Haley et al., 2014). Conversely, a separate study detected T3SS1 in all analyzed environmental and clinical V. parahaemolyticus isolates, along with a high prevalence of T3SS2, and a limited number (one and two) of isolates that were tdh positive (tdh ⁺)/trh ⁺ and ureC ⁺, respectively (Tsai et al., 2013). Further studies reported that 6% of analyzed isolates collected from seawater, plankton, and sediment samples were trh ⁺ (Caburlotto et al., 2009), and that 1.4% of analyzed isolates collected from commercial crustacean samples in Italy were detected to be trh ⁺, while none were tdh ⁺ (Caburlotto et al., 2016). The results of studies comparing V. parahaemolyticus isolates from fish or shrimp with clinical isolates have suggested that tdh and trh are more regularly harbored by the latter (Xu et al., 2014; Xie et al., 2017), while in contrast, a similar proportion of V. parahaemolyticus isolates collected from oysters and clinical settings are tdh ⁺ and/or trh ⁺ (Jones et al., 2012). Finally, the results of the study by Tsai et al. (2013) revealed that T3SS2 ⁺/tdh ⁺ isolates occur more often in the clinical setting than the environment (in seawater). Since its effect on bacterial virulence was poorly understood until recently, few studies to date have investigated the T6SS-genes in V. parahaemolyticus; however, one study did report the incidence of T6SS1 (39.6%) and T6SS2 (100%) among V. parahaemolyticus isolates collected from crayfish (Dong et al., 2016). T6SS2 and T3SS2 effectors are known to cooperate during infection (Ceccarelli et al., 2013); however, in none of the isolates in the present study were detected T3SS2 nor hemolysin genes, suggesting that they may have had limited or no enterotoxicity.

Conclusions

This study detected V. parahaemolyticus in ∼50% of the analyzed restaurant fish-tank water samples. Moreover, these displayed antibiotic resistance, and carried some virulence-related genes, but not tdh or trh. These findings are alarming, not only because T3SS1 alone is capable of incurring cytotoxicity, but also because they reveal the possibility of conferred and thereby increased antibiotic resistance in V. parahaemolyticus isolates as a result of horizontal gene transfer (Ceccarelli et al., 2013). While additional studies conducted in all seasons and with a greater number of samples are needed to confirm these results, our findings support that restaurant fish tanks require strict management to prevent V. parahaemolyticus infections via effective cleaning, temperature control, and seawater replacement/circulation, particularly between July and September.