Serological and Antibiotic Resistance Patterns As Well As Molecular Characterization of Vibrio parahaemolyticus Isolated from Coastal Waters in the Eastern Province of Saudi Arabia

Vibrio parahaemolyticus belongs to the halophilic genus of Vibrionaceae family that inhabits coastal and marine environments and is a major food-borne pathogen. In the Gulf Cooperation Council (GCC) countries and Saudi Arabia in particular, there is a lack of information regarding the detection of pandemic clone or serovariants of V. parahaemolyticus pandemic clones. Here, 400 seawater samples were collected and examined for the presence of V. parahaemolyticus from 10 locations along the coast of Eastern Province in Saudi Arabia. The recovered isolates were serotyped, and studied for antimicrobial resistance, virulence genes, and markers of pandemicity using PCR and Arbitrarily primed (AP)-PCR typing patterns. All 40 isolates were tested negative for tdh, trh, and toxRS genes. Six serotypes were identified and three were clinically significant. Antibiotic susceptibility testing of isolates revealed high resistance towards penicillins, cephalosporins, and polymyxin; 60% of isolates were multi-drug resistant, whereas all isolates were susceptible to quinolones, carbapenems, sulfonamides, and tetracycline. The multiple antibiotic resistance (MAR) index among antibiotic resistance patterns of isolates revealed that 12 (30%) isolates had recorded significant MAR index higher than 0.2. AP-PCR fingerprinting could group all isolates into five distinct and identical pattern clusters with more than 85% similarity. Our findings demonstrate that pandemic serovariants of pandemic clones were not exclusively limited to strains isolated from fecal specimens of infected patients. Nine environmental strains of serotype O1:KUT, O1: K25, and O5:K17 were isolated from costal seawater, and thus the spread of these serovariants strains of pandemic clone of V. parahaemolyticus in the environment is to avoid any kind of threat to public health.


Introduction
Vibrio parahaemolyticus is a medically important organism listed under the family of Vibrionaceae of Vibrio species; it lives in marine environments [1,2]. This bacterium was reported for the first time in 1950 due to consumption of contaminated shirasu, which led to an outbreak of food poisoning ultimately leading to 20 deaths and 272 illnesses [3]. Among the Vibrio species, V. parahaemolyticus is a premier causative agent of foodborne illness [4]. In humans, infection with this bacterium may lead to gastroenteritis after ingestion of contaminated seafood products [5]. V. parahaemolyticus is spread worldwide along coastal environments and accumulates in the digestive tract of filter feeders such as molluscan bivalves [6]. For the pathogenic strains of V. parahaemolyticus to be considered toxigenic, they should be encoded with either thermostable direct hemolysin (TDH) genes or encoded with thermostable-related hemolysin (TRH) [7]. The clinical strains of V. parahaemolyticus isolated from fecal specimens nearly all harbor trh and/or tdh genes, but the detection of these genes in environmental isolates is usually rare [6,8].
In 1996, the V. parahaemolyticus serotype O3:K6 emerged in Kolkata in Eastern India and the Bay of Bengal. It was responsible for an outbreak of acute gastroenteritis [9]. Several cases of O3:K6 pandemic clone strains were associated with foodborne outbreaks in subsequent years; sporadic cases have been reported in the Gulf coast and Atlantic coast of the United States as well as Southeast Asia [10,11]. Similar cases were of a new serotype O3:K6 were reported in Europe, Africa, South America, and Mexico [12][13][14][15]. The spread of these remarkable and unique serotypes suggested that this pathogen was a pandemic clone, and thus universal public health measurements were needed especially with regard to the consumption of seafood [8]. Therefore, group-specific (GS) PCR-based method was developed targeted to toxRS gene. This gene encoding the transmembrane protein is a molecular marker for the detection of pandemic strains of this organism [11].
Strains of V. parahaemolyticus harboring virulence genes were responsible for most foodborne gastroenteritis outbreaks after consumption of contaminated seafood. They have been reported in the United States [16], many Asian countries [17], and South America [18]. There have been several studies on the prevalence and occurrence of pandemic clones of V. parahaemolyticus in coastal areas and cases have been reported among infected humans worldwide. However, there are few reports in Gulf cooperation council (GCC) countries. Relative to Asia and most of North America, V. parahaemolyticus cases are rarely associated with the consumption of infected seafood in Saudi Arabia and other Middle Eastern countries. This probably due to a lack of systems for monitoring foodborne pathogens and it might also be due to not reported cases. To the best of our knowledge, this is our first report on pandemic clones of V. parahaemolyticus serovariants in environmental seawater samples in the Eastern Province of Saudi Arabia.

Sampling Sites
The Arabian Gulf is a semi-enclosed marine that covers an area of about 240,000 km 2 . It is characterized by conspicuous fluctuations in water temperatures and high levels of salinity [19]. A significant percentage of the global sea-transported oil is shipped through the Gulf, and thus its ecosystem is under stress from continuous discharge of hydrocarbon pollutants and crude oil spills [20]. Its coastline has seen drastic economic and social development [21]. Industrial and sewage discharge combined with low water exchange rates has caused the Arabian Gulf to be one of the highest anthropogenically impacted regions in the world [21]. Ten different locations were chosen for seawater collection alongside the coastline of the Arabian Gulf. These included beaches for public use, fishing areas, and recreational water sources. The locations are Fanateer corniche (FNC), Dammam corniche (DMC), Dammam marina front (DMF), Tarout corniche (TRC), Sayhat corniche (SEC), Qatif corniche (QTC), Palm Beach Jubail (PBJ), and Almorjan Island (MOI); Fig. 1. Some locations were divided into two sites that were coordinated using Garmin GPS during sample collection (Table 1).

Sample Collection and Water Parameters
The physical water parameters (temperature and salinity) were measured during samples collection via a multi-parameter water quality meter (YSI-50 series, Horiba, Japan) in accordance with the manufacturer's instructions. Four hundred seawater samples were collected just below the surface in duplicate sterile 500 ml screw-cap bottles (Fischer, UK) from different sampling sites between April 2015 and January 2016 (Table 1). The seawater samples were collected from 10 sampling sites as illustrated in Fig. 1. The collected samples were kept in insulated coolers without ice and transferred to the laboratory within three to five hours and processed immediately upon arrival.

Isolation and Biochemical Identification
The strains of V. parahaemolyticus in this study were isolated and identified according to the bacteriological analysis manual of the Food and Drug Administration [22]. All commercial culture media and reagents used in this study were purchased from Oxoid unless indicated otherwise. In accordance with the FDA method for V. parahaemolyticus isolation, seawater samples enriched in alkaline peptone water (APW) were adjusted to 3% sodium chloride (NaCl) concentrations [22]. The preparation of 3% NaCl-APW was performed as described in the bacteriological  [22]. Peptone and NaCl were dissolved in sterilized distilled water, dispensed into screw-cap bottles, and autoclaved. Their pH was adjusted to 8.5 ± 0.2. Following that, 25 ml of seawater samples were added to 225 ml of APW adjusted to 3% NaCl. As soon as the enrichment was completed at 35 ± 2 °C for 18-24 h, a loop full of the enriched sample in APW was streaked onto CHROMagar Vibrio (CHROM, France) and plates were incubated overnight at 37 °C. Three to five mauve color colonies on the CHROM Vibrio agar plates were then presumptively identified as V. parahaemolyticus and inoculated into Kliger iron medium with 1% NaCl, tryptic soy agar with 3% NaCl, and nutrient broth with 8% NaCl. Isolates positive for oxidase tests with alkaline slant and acid butt reactions in the Kliger iron agar medium were further analyzed. In nutrient broth, no growth without and growth with 8% NaCl were identified as V. parahaemolyticus as described elsewhere [23]. Further confirmation of V. parahaemolyticus used API 20E strips (Biomeruex, France). Prior to performing the API test, an inoculum of 2% saline was prepared from tryptic soy agar (TSA) colonies. ATCC 17,802 strain of V. parahaemolyticus was used as a positive control. Colonies of V. parahaemolyticus were identified presumptively and were inoculated into tryptic soy broth (TSB) supplemented with 3% sodium chloride and stored with 30% glycerol at -80 °C.

Nucleic Acid Extraction
The presumptive colonies of V. parahaemolyticus strains were sub-cultured into Luria-Bertani (LB) agar plates adjusted to 1% NaCl and incubated overnight at 37 °C. At least five colonies were boiled in 500 µl of sterile distilled water for 15 min to extract the nucleic acid [24].

Confirmation of V. parahaemolyticus to Species
Level Using PCR Targeted to the toxR Gene All of the presumptively identified isolates of V. parahaemolyticus isolated via CHROMagar Vibrio and biochemical tests were further confirmed using PCR targeted to the toxR gene as described elsewhere [25]. The PCR amplification and primers are presented in

Detection of Virulence Genes
The 40 isolates that were toxR positive were further amplified using tdh and trh gene primer sets [26] as described in Table 2

Group-Specific (GS) PCR
A GS-PCR method to specifically detect the toxRS sequence of the new O3:K6 clone of pandemic V. parahaemolyticus strains was performed as described [11]; see

Serotyping of V. parahemolyticus
Serotyping of all 40 isolates of V. parahaemolyticus used the slide agglutination technique via the V. parahaemolyticus antiserum kit. The method was performed according to the manufacturer's instructions (Denka Seiken, Tokyo, Japan). The antiserum test kit consists of 11 V. parahaemolyticus O and 71 K antisera. Serotyping followed the manufacturer's instructions. Briefly, an aliquot of the bacterial cell suspension in normal saline (3% NaCl) was prepared from overnight test culture grown in TSA containing 3% NaCl. For the O serotype, the fresh prepared suspension of bacterial cells was boiled for 2 h and the boiled cells were agglutinated with specific anti-O antibodies. The remaining nonboiled bacterial cell was used for the determination of the K serotype and subjected to agglutination with specific anti-K antibodies.

Antibiotic Resistance Pattern
The next step was according to a previously described protocol following the Kirby-Bauer disk diffusion susceptibility test [27]. An overnight culture of V. parahaemolyticus was suspended in 5 ml sterile aliquots of normal saline adjusted to 2% NaCl. Sterile cottonwool swabs were used for each test suspension and were inoculated onto Muller-Hinton Agar supplemented with 2% NaCl. Twenty-eight Oxoid antibiotic discs (Oxoid, England) were commercially available and represented 10 classes of antibiotics. Antibiotic disks were placed onto Muller-Hinton Agar using an automated disk dispenser (Oxoid, UK). The diameter of the zone of inhibition was measured in mm using Vernier calipers, and interpretation of the results was recorded as sensitive (   while performing antimicrobial susceptibility tests. The calculation of multiple antibiotics resistance (MAR) index was performed for V. parahaemolyticus isolates as described elsewhere [28]. Briefly, the MAR index was calculated by applying the formula MAR = a/b, where "a" represents the antibiotic to which the test isolate showed resistance, and "b" is the total number of antibiotics. A value greater than 0.2 indicates that the isolates were isolated from high-risk sources [29].

DNA Fingerprinting Analysis
The analysis of DNA fingerprints generated by AP-PCR were processed using GelJ software to analyze the DNA fingerprint [30]. The unweighted average pair group method (UPGMA) and a Dice similarity coefficient were used to investigate the genetic relationship among V. parahaemolyticus isolates. The analysis of DNA fingerprints was performed at position tolerance and optimization value of 1%. Discrimination of AP fingerprints were implemented with a cut-off of 85% to differentiate the number of AP fingerprint types among the genotyped isolates of V. parahaemolyticus.

Statistical Analysis
Descriptive statistics were used to summarize the distributions of temperature and salinity for the seawater parameters at the sampling sites. The Spearman's rank correlation was used to assess the association between the two parameters.
The MAR values were compared with 0.2 (as a mean value) using the one sample t test and compared with 0.2 again (as a median value) using the Wilcoxon test. PAST software was used for these comparisons [31].

Water Physical Parameters
The

Screening of Potential Virulence Gene Markers
In this study, none of V. parahaemolyticus were reported positive for major virulence genes (tdh and/or trh) genes. Moreover, none of the isolates tested positive for the pandemic clone using the GSPCR method targeted to the toxRS gene (Table 4).

Serotyping of V. parahaemolyticus
For epidemiological purposes, serotyping was performed on all 40 isolates using O and K antisera. The distribution of V. parahaemolyticus serovars according to sample locations is presented in Fig. 2. Thirteen (32.5%) defined serotypes were identified and 27 (67.5%) isolates underwent O antisera serotyping; however, they did not react with K antisera and resulted in untypeable (UT) serovars ( Table 4). The most prevalent serovar reported in this study was O10:KUT with 26 isolates (Fig. 2), six isolates were O5:K17, four isolates were O5:K30, two isolates were O1:K25 and the least prevalent were single isolates from O1:KUT and O2:K28, respectively. O1:K25 and O5:K17 serovars were recovered from seawater samples collected during the month of April 2015, whereas O1:KUT and O2:K28 were isolated from seawater samples collected during the month of October 2015 (Fig. 3). One and three isolates of O5:K30 serovar were recovered only from QTC locations during October 2015 and January 2016, respectively (Fig. 3). Furthermore, the O10:KUT serovar was isolated over all months of sampling except April 2015. This was recovered from all sampling locations except FNC and PBJ sites (Figs. 2 and 3).

Molecular Subtyping Using AP-PCR
AP-PCR could type all 40 isolates investigated here. The fingerprints were achieved using the AP-2 arbitrarily primer sequence, which revealed 3 to 5 bands ranging in size from 300 to 1000 bp and approximately 670 bp DNA fragments. This was shared between electrophoresed AP-PCR amplicons for all isolates (Fig. 4). The AP-PCR fingerprints grouped the 40 isolates into five clusters (I to V) at a cut-off value of 85% to assign the AP fingerprint types (A1, A2, A3, A4, and A5) as presented in Table 4 and Fig. 5. Among all clusters, cluster III (A3) had the largest number of isolates and comprised 4 and 9 isolates belonging to serovar O5:K30 and O10:KUT; these were isolated from QTC and MOI sites, respectively (Fig. 5). Four isolates (VP785, VP851, VP852, and VP853) of O5:K30 serovar isolated from QTC location during October 2015 and January 2016 were grouped together in cluster III (Fig. 5). Similarly, five of six isolates of O5:K17 serovar isolated from FNC site during April 2015 were grouped together in cluster IV with nine isolates of O10:KUT serovars recovered from different geographical sites and time intervals (Fig. 5). and was isolated from the seawater surface and sediment. The presence and survival of this bacterium in the marine environment are related and connected to environmental parameters such as sea surface temperature and salinity [33][34][35]. The overall isolation rate of V. parahaemolyticus was 6.8% in the coastal water. The highest number of positive water samples for V. parahaemolyticus was collected from MOI and the recorded water salinity and surface water temperature during the month of December 2015 was 43.6 ppt and 17.4 °C, respectively ( Table 1). The spread and epidemiology of infections are related to water temperatures, and most outbreaks of V. parahaemolyticus occur during the warmer months. The reported appropriate temperature for the proliferation of V. parahaemolyticus is 30 °C [37]. Recently, several reports from new coastal areas worldwide have documented the occurrence of V. parahaemolyticus; these incidences are related to an increase in seawater temperatures [5,10,[38][39][40]. This climatic change in temperature is responsible for outbreaks of diarrhea caused by V. parahaemolyticus in places such as Alaska where seawater temperatures are normally low and V. parahaemolyticus infection is very rare [41]. Outbreaks were also seen in southern Chile at Puerto Montt between 2004 and 2007. However, in this study we found that all 40 isolates of V. parahaemolyticus strains tested negative for tdh and trh genes and no isolates tested positive for the pandemic clone of V. parahaemolyticus O3:K6 while using group-specific PCR (GS-PCR). Strains of V. parahaemolyticus harboring tdh or trh genes can cause gastroenteritis in humans [6]. Our study is consistent with several studies showing that environmental strains rarely possess the tdh and/or trh virulence genes [1,6,26,42,43]. However, the presence of these were found in up to 90% of clinical strains of V. parahaemolyticus [5,44,45].
Recent studies have shown that from 1996 to date, at least 21 different serotypes of V. parahaemolyticus appeared to have identical genotypes and molecular profiles to those of O3:K6; these were collectively described as "serovariants" of O3:K6 isolates in which the widespread serotypes were O4:K68, O1:K25, and O1:KUT "untypeable" [1,11]. A  possible theory is divergence of these serotypes from the O3:K6 isolates by alteration of O and/or K antigens [46]. A recent study has suggested that the O/K antigen modification is a biological characteristic of the pandemic V. parahaemolyticus needed for survival in the face of host immunological resistance and varying external environments [47]. In our present study, we detected three serotypes (O1: KUT, O1:K25 and O5:K17) that were included in the pandemic group of O3:K6 clonal serovariants; all of these lacked genetic markers of the pandemic clone (tdh − /toxRS − ) ( Table 4). In general, an isolate of V. parahaemolyticus will not be considered to be a pandemic clone unless it harbors tdh and toxRS (tdh + /toxRS + ) [48].
Recently, a pandemic clone carrying tdh + and toxRS + was reported in Europe, the United States, Mexico, Bangladesh, and China [8,49]. However, our current findings contradict these studies. Our findings agree with the concerns raised by Jones and colleagues regarding the reliability of the tdh and trh genes as virulence markers-their studies reported negative strains for tdh and trh [50]. Furthermore, the same study indicated that both tdh, trh, and T3SS2 genes are not necessarily predictive of pathogenic potential; their study highlighted the need for more-detailed pathogenicity investigations of V. parahaemolyticus [50]. A recent study from India investigated the prevalence of V. parahaemolyticus in retail seafood in Kerala. That study reported that none of the isolates harbored toxRS although 14 out of 29 isolates were positive for the tdh gene [51]. A study conducted in southern Thailand investigated 865 clinical strains of V. parahaemolyticus isolated from patients at Hat Yai hospitals between 2000 and 2005; the study results revealed that there was a significant decrease in the proportion of infections by pandemic strain of V. parahaemolyticus [52]. Therefore, all strains isolated from patients from 2003 to 2004 were obtained after screening for major virulence gene markers, the O:K serotype, and pandemic clones using GSPCR defined as GSPCRpositive tdh + trh − . These remained stable at percentages of 64.1, 67.5, 69.7, and 67.7% of the total isolates each year. In 2004 to 2005, there was a decrease of pandemic clone (GSPCR + tdh + trh − ) percentages from 56.1 to 55.5%, respectively [52]. Similarly, two studies conducted in the southern region of Thailand reported variability properties that remained consistent with the V. parahaemolyticus strains isolated from individual patients; these data indicated that some patients were infected with unique strains of V. parahaemolyticus (GSPCR + tdh − trh − ) suggesting that in vivo changes might have occurred in certain individuals leading to a deleted tdh gene [53,54].
A study from Italy analyzed two V. parahaemolyticus strains isolated in May 2007 from Northern Italy seawater and plankton samples. The results showed the presence of the virulence genes tdh and orf8 as well as pandemic-specific markers. Interestingly, the two strains showed serotypes not included in the 'pandemic group' [34]. From the literature, not all strains of V. parahaemolyticus have the same pathogenic potential, but infections in humans are usually caused by diverse serotypes. To date, 21 serotypes are known as 'pandemic group' or serovariants of O3:K6 isolates based on nearly identical genotypes and molecular profiles [1]. However, as documented in literature, there is no complete agreement on a single marker for identification of pandemic clone. Such identification has been based on detection of toxRS gene using GSPCR and the presence of tdh gene (toxRS + /tdh + ) [34,48,55]. Several reports on the pandemic clone of O3:K6 have emerged after genetic elements were transferred to the pathogenic strains to increase their robustness and ability to cause infection in human [56,57]. Here, positive strains for toxRS were not detected, but constant surveillance is highlighted to detect the emergence of any pandemic clones. Of note, comparatively clinical strains were reported and did not contain virulence gene markers of tdh and trh [4]. While most hemolysins were absent, V. parahaemolyticus remained as toxigenic and elucidated the expression of other virulence activities [58][59][60][61]. Recent studies have indicated the detection of MDR among strains of V. parahaemolyticus isolated from coastal water and fisheries products versus clinical isolates [29,[62][63][64][65][66]. Here, 60% of V. parahaemolyticus isolates were MDR (Table 6). Our study is consistent with other reports showing that V. parahaemolyticus is increasingly resistant towards cephalosporins and penicillins [66][67][68]. In this study, the antibiotic resistance patterns among V. parahaemolyticus indicated that 12 (30%) of isolates had significant MAR index values above 0.2, while the highest MAR index value above 0.3 indices occurred among two strains of O1:K25 and O10:KUT serovars isolated from DMF and MOI sites, respectively. This agrees with other studies elsewhere that detected an MAR index above 0.2 among 45% of pandemic and non-pandemic V. parahaemolyticus isolated from seafood [69]. The MAR indices that exceed 0.2 indicated that these isolates originated from high-risk contamination and may pose human risks [70,71].
In the case of V. parahaemolyticus, this may be due to overuse of antibiotics by fisheries and aquaculture [72]. Although the availability of those antibiotics is important for the productivity and food security, their inappropriate use undermines their benefits [73]. Recently, the FAO of the United Nations established an action plan to improve awareness of antimicrobial resistance to prevent the excessive use of antibiotics (FAO, 2016). In addition, seawater is becoming contaminated by disposal of medical waste as well as sewage-these all contain antibiotics from human and animals. These can exacerbate resistance in pathogenic bacteria [72]. The transmission of multidrug resistance genes can undermine the antibiotic used for treatment of vibrio infection. The resistance gene might complicate the treatment of severe vibriosis infections [74,75]. Moreover, the presence of drug resistance genes in marine environments can lead to potential reservoirs that might play role in transferring these resistance genes to pathogenic bacteria through horizontal gene transfer via conjunction, transformation, or transduction [63]. Consequently, the environment has a remarkable role in the global spread of the clinically relevant antibiotic resistance and therefore imposes significant human health risks [76].
The analysis of AP-PCR fingerprints revealed that there is no site or location that influences the clustering of the isolates except for two isolates of serotype O1:K25 isolated from DMF; these were found in a single cluster and exhibited an identical pattern (Fig. 5). Also, six isolates of serotype O5:K17 recovered from FNC location were clustered with other isolates recovered from different sampling sources. Our findings are consistent with a recent study from Kerala in India, which investigated genetic relatedness on V. parahaemolyticus seafood isolates. This study revealed that the impact of geographical factors could be excluded because all the isolates were collected from a single location and did not group into one cluster [51]. Interestingly, we observed that all isolates genotyped by AP-PCR shared approximately 670 bp DNA fragment between electrophoresed AP-PCR amplicons thus suggesting that these isolates are clonally related. The results of the AP-PCR analysis were also consistent-the results support the view of other AP-PCR experiments that those pandemic serotypes and other emerged serovars show almost identical fragment patterns suggesting that these strains in the pandemic group might have originated from the same clone [5,11,48,77]. The AP-PCR method is frequently used to investigate the genetic relationship among V. parahaemolyticus strains isolated from different sources including clinical samples [23,48,53].

Conclusion
This study reported the first investigation and detection of V. parahaemolyticus O3:K6 pandemic clone serovariants of O1: KUT, O1:K25, and O5:K17 included within the 'pandemic group' recovered from the coastal environment in Saudi Arabia. The isolation of V. parahaemolyticus serotypes within the 'pandemic group' in the marine environment might constitute a public health concern if consumed in contaminated seafood. However, the genetic pandemic marker and virulence genes are usually associated with clinical strains isolated from stool specimens and, to date, these are hardly detected in seafood and environmental samples. The isolation of V. parahaemolyticus serovariants occurs in the marine environment, and none of the isolates positive for tdh and trh can pose a public health concern. Our study along with a similar study conducted in the Georgian coast of the Black Sea-found that none of V. parahaemolyticus isolates were positive for tdh and trh genes [39]. This study demonstrated the global spread and dissemination of V. parahaemolyticus serovariants in the marine environment. Its presence in the coastal environment in the Eastern Province of Saudi Arabia requires long-term monitoring consideration.

Consent for publication Not applicable.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.