Occurrence of Virulence Genes Associated with Human Pathogenic Vibrios Isolated from Two Commercial Dusky Kob (Argyrosmus japonicus) Farms and Kareiga Estuary in the Eastern Cape Province, South Africa

Background: Seafood-borne Vibrio infections, often linked to contaminated seafood and water, are of increasing global public health concern. The aim of this study was to evaluate the prevalence of human pathogenic vibrios and their associated virulence genes isolated from fish and water samples from 2 commercial dusky kob farms and Kareiga estuary, South Africa. Methods: A total of 200 samples including dusky kob fish (n = 120) and seawater (n = 80) were subjected to Vibrio screening on thiosulfate-citrate-bile salts-sucrose agar (TCBS). Presumptive isolates were confirmed and delineated to V. cholerae, V. parahaemolyticus, V. vulnificus, and V. fluvialis by PCR. Various pathogenic gene markers were screened: V. parahaemolyticus (trh and tdh), V. vulnificus (vcgE and vcgC) and V. fluvialis (stn, vfh, hupO, vfpA). Restriction Fragment Length Polymorphism (RFLP) of the vvhA gene of V. vulnificus strains was performed to determine the associated biotypes. Results: Total Vibrio prevalence was 59.4% (606/1020) of which V. fluvialis was the most predominant 193 (31.85%), followed by Vibrio vulnificus 74 (12.21%) and V. parahaemolyticus 33 (5.45%). No V. cholerae strain was detected. One of the V. parahaemolyticus strains possessed the trh gene 7 (9.46%) while most (91.9%; 68/74) V. vulnificus isolates were of the E-type genotype. V. fluvialis virulence genes detected were stn (13.5%), hupO (10.4%) and vfpA (1.0%). 12.16% (9/74) of V. vulnificus strains exhibited a biotype 3 RFLP pattern. Conclusions: This is the first report of potentially pathogenic vibrios from healthy marine fish in the study area, and therefore a public health concern.


Introduction
In recent years, there have been concerns about the microbiological safety of fish and other seafood due to increased outbreaks of seafood-borne pathogens. Members of the genus Vibrio are among the bacterial pathogens increasingly implicated in seafood-associated infections [1]. Although not all species of the genus are pathogenic to humans, incidences of human Vibrio-related illnesses over the last few years have been on the rise, particularly in developing countries, with Vibrio cholerae, V. vulnificus, V. parahaemolyticus and V. fluvialis being the most important [2,3]. Only toxigenic strains of O1 and O139 V. cholerae are associated with cholera illness; however, other strains belonging to non-O1/non-O139 have been associated with sporadic outbreaks of diarrhoea through the ingestion of contaminated seafood [4]. V. parahaemolyticus is one of the leading causes of seafood-associated diarrhoeal infection in many countries including the United States, Asian and other developing countries [5][6][7][8][9]. It is one of the frequent causative agents of gastroenteritis, caused by the consumption of raw, undercooked seafood. Although it is best known for causing gastroenteritis, cases of wound infections and septicaemia caused by V. parahaemolyticus have been reported [6,10]. Vibrio vulnificus, on the other hand, is a notorious pathogen in aquaculture and marine environments. It is second to V. parahaemolyticus as the most frequently encountered bacterial species that cause infection of food-borne origin and has received attention due to the high mortality rate of infection, making it a significant health risk [11]. Infection with this pathogen is invasive, and it causes either gastroenteritis or necrotizing wound infections, both which may lead to septicaemia [5]. Three V. vulnificus biotypes have been described. Biotype 1 is the most common worldwide, is primarily a human pathogen, and is responsible for numerous clinical cases of the disease [12]. Biotype 2 strains are primarily eel pathogens, which cause eel vibriosis or death and may be opportunistic for humans [13]. Biotype 3, is the most recently discovered, is known to be geographically restricted to Israel where it caused an outbreak of disease in fish farmers and consumers of Tilipia fish [12,14]. Vibrio fluvialis has been termed an "emerging" pathogen, due to the increased numbers of cholera-like diarrheal outbreaks and sporadic extra-intestinal cases attributed to it [15]. In the last decade, it has been reported in a variety of sources including aquaculture [16], final effluents of waste water treatment plants [17], fish and other seafood [18].
Human pathogenic vibrios produce a wide array of virulence factors which contribute to mild to fatal illnesses [2]. Largely, V. cholerae's ability to cause disease is dependent on the production of two major virulence determinants, cholera toxin (CT) and the toxin-coregulated pilus (TCP). The CT is encoded by the ctxA and ctxB genes found on the intergrated prophage CTXϕ and it is responsible for the manifestation of diarrhoea with severe water and electrolyte loss [19]. The TCP on the other hand, encoded by tcpA, is required for V. cholerae colonisation of the small intestinal epithelium [19]. The expression of these genes is controlled by the toxR regulon, in response to in vivo stimuli [20]. The thermostable direct hemolysin (tdh), in particular, as well as the thermostable direct hemolysin-related gene (trh) are major contributors to the pathogenicity of V. parahaemolyticus [8,21], owing to its biological activities, including its cytotoxic and enterotoxic effects [10]. The virulence of V. vulnificus is mainly coded for by the virulence-correlated gene (vcg). Differences in virulence degrees have been found in two major genotypes, with the degree of virulence being related to the origin of the strain, with clinical strains displaying higher virulence than environmental isolates [22]. V. fluvialis, on the other hand is known to produce several potent toxins including the heat stable enterotoxin (stn), although their roles in pathogenesis are not well established [23]. However, as is the case with V. cholerae, the toxR plays an important role in the bile resistance of V. fluvialis, an initial phase in the establishment of the disease. Other virulence factors include the heme utilization protein gene (hupO), extracellular haemolysin gene (vfh), and the V. fluvialis protease gene (vfp) [24].
A few reports on pathogenic vibrios in South Africa have been made, mostly from final effluents of waste water treatment plants [25][26][27]. However, recent epidemiological data is lacking. There is also little attention being paid to the occurrence of these pathogens in seafood or fish of commercial interest, which could be major source of seafood infection. We therefore investigated the occurrence of human pathogenic vibrios in dusky kob fish as well as water samples from two commercial farms and the Kareiga Estuary and investigated the presence of virulent associated genes that may aid in pathogenicity, which presents a guide to the risk they may pose to the community.

Sampling
An ethical clearance for the study was obtained from the University of Fort Hare Research Ethics Committee (UREC) (CLA011SFRI01). Water and fish samples were harvested from two aquaculture dusky kob farms: Farm 1, with coordinates 13.0670 • N, 59 Samples from the wild were purchased directly from fishermen as they came ashore or captured by fishing experts from SAIAB (South African Association for Aquatic Biodiversity). Upon capture, all fish were inspected for signs of any disease or abnormality, of which none were evident. Therefore, only healthy juveniles of consumption size (body mass of ≥0.6 Kg), approximately ≥7 months old were included in the study. Sampling was spread between November 2014 and October 2015, spanning the four seasons of the year. A total of 120 dusky kob fish (100 from fish farms and 20 from the wild) and 80 water samples were included in the study. The fish were ice killed and placed immediately in sterile zip lock bags post-harvest. Water was aseptically collected in sterile 2 L bottles from fish tanks at harvest as well as from the Kariega estuary during each sampling. Fish and water samples were transported in cool chain to the laboratory and processed within 4 h of collection.

Bacteriological Analyses
The outer surfaces of the fish were rinsed with sterile distilled water prior to bacteriological analysis. The skin, gill and gut were aseptically excised with a scalpel and scissors and homogenised. A 1:10 broth dilution enrichment of homogenate into alkaline peptone water (Oxoid, Basingstoke, UK), pH 8.6, was carried out and this was incubated at 37 • C for 4-6 h, and further sub-cultured on thiosulfate-citrate-bile salts-sucrose (TCBS) agar (Oxoid, Basingstoke, UK). To enhance detection of Vibrio species from water samples, analysis was done according to American Public Health Association (APHA) [28], where 100 mL each of water samples were filtered through a 0.45-µm membrane, with the membrane inoculated on TCBS (Oxoid, Basingstoke, UK) as well as pre-enrichment of water samples in APW prior to culture on TCBS. Following 24-48 h incubation, 2-5 distinctive presumptive Vibrio colonies per TCBS plate were inoculated in brain heart infusion (BHI) agar (Oxoid, Basingstoke, UK) for purity and incubated at 37 • C for 24 h. Preliminary tests on presumptive isolates included gram staining, oxidase (using of oxidase test strips (MB0266A, MICROBACT™ Oxoid, Basingstoke, UK) and catalase tests.

Molecular Confirmation of Vibrio Species
Presumptive Vibrio isolates were subjected to a polymerase chain reaction (PCR) assay for molecular identification. DNA was isolated by the boiling method where pellets from 18-24 h broth cultures were suspended in 200 µL of sterile distilled water, and cells lysed by boiling (100 • C for 15 min) in a digital Accu dri-block (Lasec, Capetown, SA). Cell debris was removed by centrifugation at 13,000× g for 5 min and the supernatants were used directly as templates in PCR reactions or stored in 50 µL aliquots at −20 • C until use. Primers targeting the 16S rRNA gene variable region that spans between 700 bp and 1325 bp were used to confirm whether isolates belonged to the genus Vibrio [29]. Confirmed isolates were further delineated to four human pathogenic species: V. cholerae, V. paraheaemolyticus, V. vulnificus and V. fluvialis. Vibrio cholerae, V. paraheaemolyticus and V. vulnificus were detected by multiplex PCR using species-specific primers as reported earlier [30], while single PCRs were performed to detect V. fluvialis [23]. Vibrio cholerae was also detected by targeting the ompW gene [31]. The various primer sets used for PCR confirmation and species differentiation are shown on Table 1. V. vulnifius ATCC 27562, V. parahaemolyticus ATCC 17802 and V. fluvialis ATCC 33809, as well as extracted DNA from a confirmed O139 V. cholerae strain (obtained from the Microbiology laboratory of the Natural Resources and the Environment (NRE), Council for Scientific Research (CSIR)) were used as positive controls.

Detection of Virulence Genes
PCR confirmed V. parahaemolyticus isolates were screened for the presence of tdh and trh virulence genes using previously reported primers [32], while V. vulnificus virulence genes (vcgE and vcgC) were detected as previously described [33]. Four gene-specific primer sets (Table 2) were used in a touch-down multiplex PCR amplification of V. fluvialis virulence genes, heat stable enterotoxin (stn), heme utilization protein gene (hupO), extracellular haemolysin gene (vfh) and V. fluvialis protease gene (vfp) [24]. Table 2 indicates the target genes and primer sets used for detection of Vibrio virulence genes. All amplifications were performed in 25 µL reactions, each consisting of 12.5 µL of 2× master mix (Thermo scientific, Johannesburg, South Africa), 0.5 µL of each oligonucleotide (Inqaba Biotec, Pretoria, South Africa), and an appropriate volume of nuclease free water (Thermo scientific, Waltham, MA, USA). PCR products were separated by electrophoresis in 1-2% agarose gels containing 5 µg/mL ethidium bromide (Sigma-Aldrich, St. Louis, MO, USA) run at 100 V for 45 min and visualised under a UV transilluminator (Alliance 4.7 XD-79, Uvitec, Cambridge, UK).

Biotyping of V. Vulnificus
The restriction fragments of the vvhA gene obtained using KpnI or PstI can be used to differentiate biotype 3 from biotypes 1 and 2 [34]. The vvhA nucleotide sequence of biotype 3 strains contain restriction sites for these enzymes while those of biotypes 1 and 2 do not. Biotyping of the confirmed V. vulnificus strains was therefore carried out as earlier reported [34]; the V. vulnificus cytotoxin-hemolysin gene (vvhA) was amplified using the primer pair Fw 5 CAGCTCCAGCCGTTAACCGAACCACCCGC-3 and Rv 5 -TTCCAACTTCAAACCGAACTATGAC-3 . Amplicons were resolved on 1.5% agarose gel followed by restriction fragment length polymorphism (RFLP). The two restriction enzymes, (KpnI and PstI), were used for restriction digest following the manufacturer's instructions (New England Biolabs, Ipswich, MA, USA). The restricted digests were separated by electrophoresis in 2% agarose gels and visualised in a UV transilluminator (Alliance 4.7 XD-79, Uvitec, Cambridge, UK).

Biotyping of V. Vulnificus
The restriction fragments of the vvhA gene obtained using KpnI or PstI can be used to differentiate biotype 3 from biotypes 1 and 2 [34]. The vvhA nucleotide sequence of biotype 3 strains contain restriction sites for these enzymes while those of biotypes 1 and 2 do not. Biotyping of the confirmed V. vulnificus strains was therefore carried out as earlier reported [34]; the V. vulnificus cytotoxin-hemolysin gene (vvhA) was amplified using the primer pair Fw 5′ CAGCTCCAGCCGTTAACCGAA CCACCCGC-3′ and Rv 5′-TTCCAACTTCAAACCGAACTATGAC-3′. Amplicons were resolved on 1.5% agarose gel followed by restriction fragment length polymorphism (RFLP). The two restriction enzymes, (KpnI and PstI), were used for restriction digest following the manufacturer's instructions (New England Biolabs, Massachusetts, USA). The restricted digests were separated by electrophoresis in 2% agarose gels and visualised in a UV transilluminator (Alliance 4.7 XD-79, Uvitec, Cambridge, UK).

Prevalence of Vibrio Species
A total of 1020 presumptive isolates were obtained by culture, of which 606 (59.4%) were positive by PCR as belonging to the genus Vibrio ( Figure 1). Vibrio fluvialis was the most predominant, 193 (31.85%), followed by V. vulnificus, 74 (12.21%) and V. parahaemolyticus 33 (5.45%) ( Table 3, Figures 2 and 3). No V. cholerae was detected. The rest of the isolates, 306 (50.5%) were assumed to belong to other species of the genus not considered in this study.

Biotyping of V. Vulnificus
The restriction fragments of the vvhA gene obtained using KpnI or PstI can be used to differentiate biotype 3 from biotypes 1 and 2 [34]. The vvhA nucleotide sequence of biotype 3 strains contain restriction sites for these enzymes while those of biotypes 1 and 2 do not. Biotyping of the confirmed V. vulnificus strains was therefore carried out as earlier reported [34]; the V. vulnificus cytotoxin-hemolysin gene (vvhA) was amplified using the primer pair Fw 5′ CAGCTCCAGCCGTTAACCGAA CCACCCGC-3′ and Rv 5′-TTCCAACTTCAAACCGAACTATGAC-3′. Amplicons were resolved on 1.5% agarose gel followed by restriction fragment length polymorphism (RFLP). The two restriction enzymes, (KpnI and PstI), were used for restriction digest following the manufacturer's instructions (New England Biolabs, Massachusetts, USA). The restricted digests were separated by electrophoresis in 2% agarose gels and visualised in a UV transilluminator (Alliance 4.7 XD-79, Uvitec, Cambridge, UK).

Prevalence of Vibrio Species
A total of 1020 presumptive isolates were obtained by culture, of which 606 (59.4%) were positive by PCR as belonging to the genus Vibrio ( Figure 1). Vibrio fluvialis was the most predominant, 193 (31.85%), followed by V. vulnificus, 74 (12.21%) and V. parahaemolyticus 33 (5.45%) ( Table 3, Figures 2 and 3). No V. cholerae was detected. The rest of the isolates, 306 (50.5%) were assumed to belong to other species of the genus not considered in this study.

Discussion
Marine filter feeders are highly susceptible to surface or tissue contamination of marine origin, which may include potential human pathogens. Other marine foods are not exempt from bacterial contamination either. Members of the genus Vibrio are ubiquitous in marine and estuarine waters and therefore the high total prevalence (59.4%) recorded in our study was not surprising. What concerned us was the presence of virulence traits among potentially pathogenic vibrios.
Three (V. parahaemolyticus, V. fluvialis, and V. vulnificus) of the four human pathogens considered in this study were detected from the fish and water samples, with some strains harboring one or more virulence determinants. This is similar to an earlier study [35] reporting a number of Vibrio species which are indigenous aquatic bacteria that contaminate fish, including V. cholerae, V. parahaemolyticus, and V. vulnificus, mostly present in warm (>15 °C) saline waters, and are implicated in many food-borne illnesses. The most frequently isolated non-cholera-causing human pathogenic Vibrio species was V. fluvialis, 31.85% (193/606). This result is similar to others previously reported where V. fluvialis was the most frequently occurring species isolated from fish ponds [36], as well as in a variety of seafood [37]. The high frequency of occurrence of this pathogen in marine reservoirs could point towards why an increasing number of cases in the last decade have been reported terming it "emerging". First reported in 1975 in a patient with diarrhoea in Bahrain [38], V. fluvialis has since become distributed worldwide, and is associated with numerous reports of food poisoning [39,40], as well as gastroenteritis with diarrheal illness connected with the consumption of raw or improperly cooked seafood [41,42]. We also detected V. vulnificus (12.21%) and V. parahaemolyticus (5.45%) although at lower frequencies than those recorded in waste water treatment facilities [25][26][27]. The presence of this pathogen in marine fish and aquaculture could be a source of infection to humans, especially to the immunocompromised. Earlier reports in Israel in 1996 and 1997 [12], implicated V. vulnificus in an outbreak of invasive infection in handlers of whole fresh fish purchased from fish ponds. V. vulnificus was also the source of necrotizing fasciitis that led to the amputation of the lower limb of a 17 year old boy after minor trauma during exposure to a contaminated fish tank [43]. Although cases of V. vulnificus infection may be generally low, it has received attention due to high case fatalities related to consumption of contaminated seafood [44,45].
Bacterial pathogenicity depends on the production of virulence factors, with virulence genes acting as major orchestrators. The dominant virulence factors of V. parahaemolyticus with regard to its pathogenicity are tdh and trh [10]. All V. parahemolyticus strains detected in this study were negative for the tdh gene, while trh was detected in one isolate (1.4% (1/33). These results are, however, not surprising, as previous reports have indicated the presence of these virulent signatures in only low numbers (1-7%) in environmental samples [46]. Similarly, Vibrio isolates recovered from West Sumatran rivers and shrimp farms in Sri Lanka were void of these virulence genes [47,48]. A low prevalence (0% tdh and 4.2% trh) was also recorded from 96 seafood and water samples in Turkey [8].
Although clinical cases of V. parahaemolyticus are associated with increased detection of these genes, the continuous report of pathogenic V. parahaemolyticus in the last few years and the fact that it is the most commonly encountered Vibrio species implicated in seafood infection is an indication that other

Discussion
Marine filter feeders are highly susceptible to surface or tissue contamination of marine origin, which may include potential human pathogens. Other marine foods are not exempt from bacterial contamination either. Members of the genus Vibrio are ubiquitous in marine and estuarine waters and therefore the high total prevalence (59.4%) recorded in our study was not surprising. What concerned us was the presence of virulence traits among potentially pathogenic vibrios.
Three (V. parahaemolyticus, V. fluvialis, and V. vulnificus) of the four human pathogens considered in this study were detected from the fish and water samples, with some strains harboring one or more virulence determinants. This is similar to an earlier study [35] reporting a number of Vibrio species which are indigenous aquatic bacteria that contaminate fish, including V. cholerae, V. parahaemolyticus, and V. vulnificus, mostly present in warm (>15 • C) saline waters, and are implicated in many food-borne illnesses. The most frequently isolated non-cholera-causing human pathogenic Vibrio species was V. fluvialis, 31.85% (193/606). This result is similar to others previously reported where V.'fluvialis was the most frequently occurring species isolated from fish ponds [36], as well as in a variety of seafood [37]. The high frequency of occurrence of this pathogen in marine reservoirs could point towards why an increasing number of cases in the last decade have been reported terming it "emerging". First reported in 1975 in a patient with diarrhoea in Bahrain [38], V. fluvialis has since become distributed worldwide, and is associated with numerous reports of food poisoning [39,40], as well as gastroenteritis with diarrheal illness connected with the consumption of raw or improperly cooked seafood [41,42]. We also detected V. vulnificus (12.21%) and V. parahaemolyticus (5.45%) although at lower frequencies than those recorded in waste water treatment facilities [25][26][27]. The presence of this pathogen in marine fish and aquaculture could be a source of infection to humans, especially to the immunocompromised. Earlier reports in Israel in 1996 and 1997 [12], implicated V. vulnificus in an outbreak of invasive infection in handlers of whole fresh fish purchased from fish ponds. V. vulnificus was also the source of necrotizing fasciitis that led to the amputation of the lower limb of a 17 year old boy after minor trauma during exposure to a contaminated fish tank [43]. Although cases of V. vulnificus infection may be generally low, it has received attention due to high case fatalities related to consumption of contaminated seafood [44,45].
Bacterial pathogenicity depends on the production of virulence factors, with virulence genes acting as major orchestrators. The dominant virulence factors of V. parahaemolyticus with regard to its pathogenicity are tdh and trh [10]. All V. parahemolyticus strains detected in this study were negative for the tdh gene, while trh was detected in one isolate (1.4% (1/33). These results are, however, not surprising, as previous reports have indicated the presence of these virulent signatures in only low numbers (1-7%) in environmental samples [46]. Similarly, Vibrio isolates recovered from West Sumatran rivers and shrimp farms in Sri Lanka were void of these virulence genes [47,48]. A low prevalence (0% tdh and 4.2% trh) was also recorded from 96 seafood and water samples in Turkey [8].
Although clinical cases of V. parahaemolyticus are associated with increased detection of these genes, the continuous report of pathogenic V. parahaemolyticus in the last few years and the fact that it is the most commonly encountered Vibrio species implicated in seafood infection is an indication that other virulence factors may exist [49][50][51]. Mahoney et al. [52] reported environmental isolates of V. parahaemolyticus lacking the tdh and trh but producing putative virulence factors such as extracellular proteases, biofilm and siderophores. Recently, Raghunath reported finding environmental isolates of V. parahaemolyticus lacking the tdh and trh genes but highly cytotoxic to human gastrointestinal cells [53]. Some Vibrio fluvialis strains in this study were found to possess three of the four targeted virulent genes, stn 13.5%, hupO 10.4% and vfh 6.2%. Although prevalence of these genes may be low, it should not be ignored as there are increased reports of this pathogen in the last decade [36,37,40]. Moreover, some of these virulent genes may be transferred to similar or other bacterial species via horizontal gene transfer or mobile genetic elements, leading to virulent human infections. The exotoxin-hemolysin vvhA virulence protein produced by V. vulnificus facilitates the release of iron from haemoglobin. It also contributes to the bacteria's virulence by producing cytotoxic effects [54]. The virulence correlated gene vcg, however, has been one of the major virulence factors of V. vulnificus and can be used to distinguish potentially virulent from avirulent strains. The vcgC gene is linked mainly with clinical isolates while the vcgE is linked mostly with environmental isolates [55]. In this study, over 90% of V. vulnificus isolates were of the vcgE genotype, an indication of less virulent strains. This result is similar to that of a study recording a significantly higher percent of the E-genotype of V. vulnificus isolates recovered from oysters [56]. Although such a result is expected from environmental samples, it is, however, contrary to a report where nearly an equal percent of vcgE (47%) and vcgC (53%) were detected from oyster isolates [57] and water areas surrounding oyster harvest [56]. Continuous monitoring of seafood and fish products remains a necessity as infection with V. vulnificus, although rare, is usually fatal, particularly in immunocompromised persons.
Restriction digest of V. vulnificus vvhA gene revealed that most of the isolates belonged to biotypes 1 and 2, while a small percentage (12.16%), belonged to the biotype 3. Biotype 1 is primarily a human pathogen, and the leading cause of seafood-related deaths due to primary septicaemia [54], while biotype 2 mainly causes eel vibriosis [13]. However, Amaro and Biosca [58], reported a clinical V. vulnificus strain isolated from a human wound which belonged to biotype 2, suggesting that biotype 2 pathogenic for eels are also opportunistic for humans. In addition, some biotype 2 strains isolated from human samples could not be definitively linked to eel exposure or manipulation despite earlier reports to the contrary [13]. Seawater and fish species other than eels have been suggested, therefore to be putative reservoirs of biotype 2 [59]. Results from our study indicate that dusky kob and its surrounding waters are likely to be reservoirs for V. vulnificus, which could either cause infection or be opportunistic for humans particularly in persons with underlying disease. The first reported case of V. vulnificus biotype 3 determined to be the cause of severe soft tissue infections, and associated with exposure to contaminated fish, (especially tilapia or common carp), was described in 1996-1997 [34]. This biotype is thought to be geographically restricted to Israel. Our study is the first report of V. vulnificus biotype 3 isolated from a marine source in South Africa.

Conclusions
To the best of our knowledge, our findings represent the first report on the prevalence of human pathogenic Vibrio species with some virulent signatures isolated from healthy marine fish in South Africa. Although no Vibrio cholerae was detected, the occurrence of other human pathogenic vibrios, including V. parahaemolyticus, V. fluvialis and V. vulnificus, (all of which can cause sporadic and epidemic gastroenteritis as well as other serious diseases), is of public concern. Detection of V. vulnificus biotype 3 indicates its circulation in the region. Even though transmission of fish pathogens to humans may be relatively low, it is a health risk that needs to be considered by fish farmers, fishermen and others who handle or consume fish products, particularly in conjunction with the increased rate of seafood-associated Vibrio infections. Consumption of properly cooked fish is recommended, while frequent epidemiological studies could be carried out to better understand the associated health risk and implement control strategies for reduction of possible human infections.

Acknowledgments:
The authors are grateful to the South African Institute for Aquatic Biodiversity (SAIAB) and NRF for funding this study. They are also grateful to the directors of the fish farms for provision of samples for the study.
Author Contributions: Justine Fri designed the study, carried out sample collection, laboratory and data analysis, and drafting of manuscript. Henry Akum Njom contributed expert judgments in the laboratory analysis, data interpretation and proof reading of manuscript. Anna Maria Clarke participated in study design, organization of sampling and supervised the experiments. Roland Ndip Ndip participated in study design, and added expert knowledge to the technicalities of the study.

Conflicts of Interest:
The authors declare no conflict of interest.