V. cholerae, a Gram-negative motile rod causes massive cholera outbreaks such as the one following the 2010 earthquake in Haiti (Sack et al., 2004; Chin et al., 2011; Katz et al., 2013). Cholera is a global threat to public health and it was estimated that between 2008 and 2012 cholera caused an annual average of 2.9 million cases, and 95,000 deaths, worldwide (Ali et al., 2015). Particular serogroups (O1 and O139) of this bacterium are responsible for cholera epidemics and pandemics. Human infection with V. cholerae begins with ingestion of contaminated food or water containing the bacterium. V. cholerae colonizes the small intestine and secretes cholera enterotoxin (CT) into the host cells resulting in rapid efflux of chloride ions and water into the lumen of the intestine, leading to profuse diarrhea and severe dehydration (Kaper et al., 1995).

Non-O1/non-O139 V. cholerae serogroups are also linked to V. cholerae gastroenteritis as well as to wound infections and bacteremia (Deshayes et al., 2015). V. cholerae O1, O139 and non-O1/O139 comprise a single taxonomic species and their habitats attributes are similar (Lewin, 1996), however, recently it has been suggested that not all strains of V. cholerae species share the same niche (Kirchberger et al., 2016). The role of CT in the environment is not understood.

V. cholerae is commonly associated with chitin-containing zooplankton, particularly copepods (Huq et al., 1983) and chironomids (Broza and Halpern, 2001; Halpern et al., 2004). Recent evidence supports the hypothesis that fish and waterbirds may also be intermediate reservoirs and vectors of V. cholerae (Halpern et al., 2008; Halpern and Izhaki, 2010).

Evidence from the literature correlated fish with a few cholera cases or epidemics. The first records date back over more than 50 years. It was postulated that cholera endemicity in India was due to hilsa fish (Pandit and Hora, 1951) (Table 2). Morgan et al. (1960), suggested that the origin of El Tor vibrios outbreaks in Thailand might have been fish that are often eaten raw in the Pacific area. Cholera was associated with eating salted fish, sardines and other fish from an atoll lagoon in the Pacific Ocean (Merson et al., 1977; Kuberski et al., 1979; McIntyre et al., 1979). A cholera outbreak in Tanzania (67 patients, including 11 deaths) was correlated with handling and eating fish at social gatherings (Killewo et al., 1989). Out of 12 cholera cases caused by V. cholerae O1, serotype Ogawa, biotype El Tor, in the southern Italian region of Puglia, in 1994, three patients reported consumption of raw fish (Maggi et al., 1997) (Table 2). Consumption of dried fish correlated significantly with cholera risk in Tanzania (Acosta et al., 2001). In July 2001, a case of cholera, caused by V. cholerae O1, serovar Inaba, biovar El Tor, was reported in Berlin. Interestingly, the patient had most likely been infected while handling and preparing fish imported from Nigeria (Schürmann et al., 2002). A food trace-back investigation following three cases of cholera in Sydney, Australia, found that the only exposure common to all cases was consumption of raw whitebait imported from Indonesia (Forssman et al., 2007). V. cholerae O1 serovar Ogawa was identified as the causative agent in all three cases. V. cholerae non-O1 was isolated from stools of a fisherman who had fished, cooked and eaten a lake fish in Italy (Piantieri et al., 1982). The source of Vibrio cholerae non-O1 that was found in a wound, was linked with a tropical fish tank (Booth et al., 1990) (Table 2).

A few publications (mentioned above and in Table 1), correlated the presence of V. cholerae with a disease in fish (Yamanoi et al., 1980; Kiiyukia et al., 1992; Dong et al., 2015; Rehulka et al., 2015; Kiani et al., 2016). Kiiyukia et al. (1992), who isolated V. cholerae non-O1/O139 from diseased ayu fish in Japan, emphasized that healthy ayu fish caught in Lake Biwa, Japan in the rivers running into this lake, also harbored V. cholerae but without showing any signs of a disease. Kiani et al. (2016) isolated V. cholerae along with other pathogens from diseased Nile tilapia that were cultured in floating cages in Thailand. However, it was not proven that V. cholerae was indeed the causative agent of the disease (Table 1). All the above studies simply assumed that because they isolated V. cholerae from the diseased fish, this species was responsible for the disease. Rehulka et al. (2015) injected a fish with relatively large dose of bacteria (e.g., 2 × 10⁸ cells) to obtain fish mortality but without following all Koch postulates rules. Hence we argue that at least for some cases other bacterial species or viruses and not V. cholerae were probably responsible for the fish disease (Table 1).

Senderovich et al. (2010) isolated V. cholerae from 15 different heathy fish species. They found 5 × 10³ and 1.4 × 10² colony forming units (cfu) of V. cholerae per gr intestine content in Sarotherodon galilaeus (Galilee St. Peter's fish) and in Mugil cephalus (Flathead gray mullet), respectively. None of these fish showed any signs of disease. Nevertheless, there is a scarcity of quantitative studies of V. cholerae in fish. Many other studies reported the presence of V. cholerae in different healthy fish species that were sampled from both marine and freshwater habitats (listed in Table 1) but these studies did not quantify the numbers of V. cholerae in the fish.

Not all the fish species are inhabited by V. cholerae. For example, Jones et al. (2013) detected V. cholerae only in 4 out of 10 fish species sampled in the Gulf of Mexico (estuarine habitat). Similarly, Senderovich et al. (2010) did not detect V. cholerae in 4 out of 14 freshwater and in 43 out of 44 marine fish species. Scrutiny of the list of the fish species found to host V. cholerae revealed that all belonged to Actinopterygii class (Table S1). V. cholerae was identified from 30 species belonging to 9 different orders within this class.

Fish may actually benefit from V. cholerae that inhabit their intestine. Strains of V. cholerae secrete extracellular enzymes such as proteases (Halpern et al., 2003) and chitinases (Pruzzo et al., 2008; Senderovich et al., 2010). These enzymes may have a role in the digestion of macromolecules like proteins and chitin in the fish gut. Chitin, a polymer of β 1,4 N-acetylglucoseamine, is the main component of crustaceans' (copepods) and insects' (chironomids) exoskeletons. This insoluble polymer is a source of carbon and nitrogen (Cohen-Kupiec and Chet, 1998; Laviad et al., 2016). Senderovich et al. (2010) found that all V. cholerae strains isolated from 15 different fish species were able to degrade chitin. Thus, it is possible that the fish intestine serves as hosts for V. cholerae while the bacteria may play a role in helping the fish digest its chitinous zooplankton prey. As the fish that carry the bacteria swim from one location to another (some fish species move from rivers to lakes or sea and the reverse), they serve as vectors for V. cholerae. Nevertheless, fish are consumed by waterfowls, which disseminate the bacteria on a global scale (Halpern et al., 2008; Halpern and Izhaki, 2010).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.