GENES OF VIRULENCE AND ANTIMICROBIAL RESISTANCE IN Vibrio parahaemolyticus PREVALENT IN AREAS OF OSTREICULTURE

This study aimed to quantify Vibrio spp. and evaluate the profile of antimicrobial susceptibility and virulence factors in Vibrio parahaemolyticus strains in the samples of oysters collected from two estuaries of the Baixo Sul, Bahia. The samples were collected between June 2015 and January 2016 from natural banks (E1) (oyster) and from cultivation areas (E2) (water and oyster). For the quantification of Vibrio spp., the most probable number (MPN) was determined and the characterization of V. parahaemolyticus marker (species-specific) gene tl was performed. Pathogenicity was observed with the Kanagawa test, and the presence of the tdh, trh and ure genes were tested. The antimicrobial sensitivity test included disc diffusion method with 15 antimicrobial discs, β-lactamase enzyme production, and the presence of blaTEM, blaSHV, and blaCTX-M. The maximum density of Vibrio spp. in the samples from cultivation was 4.70 log MPN g-1 and extractivism was 6.10 log MPN g-1, with the temperature being the most influencing variable on the presence of microorganisms in the cultivation area (E2). The tl gene was detected in 71% of the isolates, without the presence of tdh, trh and ure genes. All strains of V. parahaemolyticus were Kanagawa negative. High antimicrobial resistance was observed in the β-lactam antibiotics (cephalothin 72%, ampicillin 60%) and aminoglycosides (amikacin 64%), with multi-resistance in 88% of the strains of V. parahaemolyticus, of which 68% of the resistance was mediated by plasmids. Phenotypically, no production of β-lactamase enzymes was observed, but the presence of blaTEM genes was observed. The multiresistant character of plasmids reported in V. parahaemolyticus strains increases the concern about native bacteria in the marine environment since it can potentially compromise the control of this bacterium infection in bivalve mollusks.


INTRODUCTION
The Baixo Sul da Bahia occupies an area of 6.451 km 2 , corresponding to about 1.14% of the entire state.It comprises of 11 municipalities and 7 coastal areas (Cairu, Taperoá, Camamú, Igrapiúna, Ituberá, Nilo Peçanha, and Valença).It is a region of great environmental and landscape diversity, cut off by islands, estuaries, and deep bays (FISCHER et al., 2007); the extension of mangroves (120,000 hectares area) is significant, as >20% of the local population survives directly or indirectly from artisanal fishing (IBAM, 2010).
In the Northeast region of Brazil, oyster breeding has gradually been consolidated with the cultivation of the native species Crassostrea rhizophorae.Oysters are filtering organisms, with a water filtration capacity of 19-50 L h -1 , and, through a low-selective capacity, they can accumulate biological pollutants such as bacteria in their tissues (ORBAN et al., 2007;BALLESTEROS et al., 2016).These pollutants when ingested in natura or poorly cooked, which is a common occurrence in the coastal regions, can cause significant food disease outbreaks, mainly through Vibrio strains (YU et al., 2016).The genus Vibrio includes native species of marine environments, estuaries, and coastal waters (NASCIMENTO et al., 2011).
The pathogenesis of V. parahaemolyticus can mainly be attributed to the presence of two virulence factors, the direct thermostable hemolysin (TDH) and the related thermostable hemolysin (TRH).The virulence factor TDH is considered as an important enterotoxin and has been identified as the main virulence factor for the species (XIE et al., 2017).Thus, phenotypical tests for the purpose of rapid identification and effective monitoring of oyster culture or extractive environments are essential to avoid and control the outbreaks of food diseases.
In the outbreaks of alimentary gastroenteritis involving V. parahaemolyticus, the treatment is conducted with chemotherapeutics.However, the transmission Vibrio strains resistant to the various drugs have increased in recent years (MANJUSHA and SARITA, 2013;LOU et al., 2016), because the aquatic environment is continuously receiving human and veterinary antimicrobial residues via domestic and hospitals sewage (BOUKI et al., 2013).Bacteria produce and secrete antibiotics into the environment for signaling and regulatory purposes as well as to protect themselves from antimicrobial toxicity through the acquisition and expression of resistance genes in their respective microbial communities (KANG et al., 2017).
Considering the epidemiological importance of Vibrio, as well as the increase in antimicrobial-resistant bacteria in the food of marine origin due to the increase in anthropic action in estuarine systems, where the aquaculture activities or extraction of bivalve mollusks are carried out, this study aimed to quantify Vibrio spp.and evaluate the profile of antimicrobial susceptibility and virulence factors in V. parahaemolyticus strains in water samples, cultured oysters (C.rhizophorae), and oysters extractivism, in two estuaries of the Southern Bahia Lowlands region, Brazil.

Study area
The estuary of Taperoá (E1), which houses a natural oyster bank, is located in the municipality of Taperoá, Baixo Sul da Bahia, and is bathed by the Almas river (13º32'17"S 39º05'55"W).The estuary (E2) that bathes the locality of Torrinhas receives several arms of rivers, of which the river Una (13°33'54.84"S39°0'29.99"W) is the main one.In this area, oyster cultivation is conducted in a flashlight system (Figure 1).Cultivation in lanterns is recommended for deep waters in the absence of strong currents, as the lanterns stand vertically (Figure 2).The cultivation practice consists of collecting oyster seeds directly from nature and fattening them in lanterns through the filtration of organic matter suspended in the water (SILVA and SILVA, 2007).The extraction areas E1 and E2 are approximately 6-km apart (Figure 1).Six oyster samples were collected from natural banks (E1-13°32'43.8"S39°05'09.1"W)(Figure 1) between June 2015 and January 2016, totaling 720 oysters.The oysters were taken from the women engages in culturing oyster due to the difficulties in a direct access to the place.
-Area of cultivation (E2) Oyster collections in the E2 area were conducted on the same day as that of sample collection from the E1 area.Oyster collection was done randomly.At each harvest, about 36 oysters were taken from the two lanterns, totaling 216 oysters.The water samples collected for each point (point P1 [13°34'39.576"S 39°1'22.800"W]after cultivation (peck-crop direction), point P2 [13°34'23.196"S 39°0'41.652"W] in cultivation, and point P3 [13°34'23.664"S39°1'2.640"W],prior to cultivation (Figure 1) consist of a composite sample (1 L of water at five equidistant points), homogenates (5 L), and a 1-L aliquot was withdrawn for analysis.In each collection, the parameters of temperature, pH, and salinity were determined by means of a HORIBA multiparameter probe.
All the collected materials were conditioned in a thermal box containing ice and sent for immediate analysis.The interval between collection and analysis was less than 4 h.The oysters were washed under running water, dried on a paper towel, and opened under aseptic conditions for the withdrawal of the liquid and intervalvar muscle.
For the quantification of Vibrio spp., 50 g of oysters and 50 mL of water were aseptically transferred to an Erlenmeyer flask containing 450 mL of phosphate-buffered saline (PBS), with decimal dilutions up to 10 -4 .Aliquots of 1 mL were transferred to a series of 5 tubes (oysters and water) containing alkaline peptone water (APA) + 3% NaCl, incubated at 37 °C for 18-24 h.Each microbial growth tube was seeded on citrate bile sucrose (TCBS) thiosulfate agar plates and incubated at 37 °C for 18-24 h.From each TCBS agar plate, 5 colonies characteristic of V. parahaemolyticus (sucrose negative, round and opaque blue-green) were selected and transferred to 3% NaCl Tryptone Soya Agar (TSA), incubated at 37 °C for 24 h, and submitted to Gram-staining for the verification of the morpho-tinctorial characteristics.The microbial density of Vibrio spp. was determined from the positive APA tubes [sucrose negative strains and confirmed according to the identification code of NOGUEROLA and BLANCH (2008)] in Most Probable Number (MPN) per gram or milliliter (SILVA et al., 2010).

Pheno-genotypic characterization and molecular identification of V. parahaemolyticus
A total of 78 strains were analyzed with specific molecular markers for the identification of Vibrio tl species.From the positive results, virulence tests were performed on Petri dishes containing Wagatsuma® agar incubated at 37 °C for 18 h.The colonies with transparent halos around them were considered as positive (SILVA et al., 2010).These strains were submitted to the tests for genes of virulence and pathogenicity.
The strains positive for the phenogenotypic tests were submitted to molecular identification from the 16S rDNA region with the primers: 8FN:5'-AGAGTTTGATCMTGGCTCAG-3' and 1492R':5'-TACGGYTACCTTGTTACGACTT-3' as described by WEISBURG et al. (1991).The amplicons were purified using the ILLUSTRA® GFX PCR DNA and Gel Band Purification Kit (GE Healthcare Life Sciences) for further nucleotide identification using the ABI-Prism 3500 Genetic Analyzer (Applied Biosystems) automated sequencer.Sequence editing and assembly was performed with the Sequencher 4.1.4program (Gene Code Corporation).The BLAST program was used to compare the sequences of each isolate with those found in the public databases NCBI (2018) and ENSEMBL BACTERIA (2018).
For the calculation of the index of multiple antimicrobial resistance (MAR index), the ratio between the numbers of antimicrobials to which the isolate was resistant to the total number of antimicrobials tested was performed.Index >0.13 was characterized as a multidrug-resistant strain (KRUMPERMAN, 1983).
Plasmid-R-mediated resistance was evaluated for the strains showing an antimicrobial resistance profile.The strains were picked in nutrient broth at 37 °C for 24 h.Subsequently, 200 μL aliquots were transferred to tubes containing Luria Bertani (LB) 3% NaCl (control) and 3% LB broth NaCl + acridine orange (AO) at the concentration of 100 μg mL -1 as a curing agent.After incubation at 37 °C for 24 h, the cultures were again submitted to the antibiogram (MOLINA-AJA et al., 2002).
The phenotypic screening for Extended Spectrum Betalactamases (ESBL) producers was conducted using the disk approximation method in compliance with the standards of the CLSI (2009).A disc containing a third generation cephalosporin and another, 20 mm apart, containing the beta-lactamase inhibitor (amoxicillin + clavulanic acid) was used.The positivity of test was verified by the appearance of a "ghost zone" or widening of inhibition halo of cephalosporin.

Statistical analysis
The values of the MPN g -1 or mL -1 variables of the oyster and water samples were transformed into Log (x + 1).The averages values obtained were submitted to the Bartlett test for homoscedasticity and Shapiro-Wilk test for normality.To perform the tests and confection of the Boxplot graphics and contour area, the software R 3.4.4(R CORE TEAM, 2016) was used.

Microbiological analysis of Vibrio spp.
The variation in the amount of Vibrio spp.(log MPN mL -1 and log MPN g -1 ) in the samples of oysters and water was verified during the collection period, with oysters exhibiting greater variation in their bacterial microbiota (Figure 3A).
For the environmental parameters of the water, the average temperature at the collection points (P1, P2, and P3) was 27 °C, pH 7.5, and a salinity of 20 ppm, showing interactions between the Figure 3. Boxplot of the density of Vibrio spp. in water samples (log MPN mL -1 ) and oysters (log MPN g -1 ) from the farming area (A), and cultivation oysters and extractivism oysters samples (B) in two estuaries of the Baixo Sul, Bahia.

5/10
quantification of Vibrio spp.and the physicochemical parameters, that is, the highest concentration of Vibrio spp.occurred at 29 °C with a salinity of 19 ppm (Figure 4A) and a pH of 7.1 and 8.0 (Figure 4B).
The estimation of the density variation of Vibrio spp.among the oyster samples (cultivation × extractivism) verified that the variation of Vibrio spp.(E2) was lower than that in extractive oysters (E1) (Figure 3B).
Of the 78 isolates of Vibrio spp., 71% (55) presented the tl gene, and amplification of 16S rDNA sequences identified it as V. parahaemolyticus.All strains of V. parahaemolyticus were negative for the Kanagawa test (β-hemolysis), with the absence of the tdh, trh, and ure genes.

Antimicrobial susceptibility
As shown in Figure 5, a high antimicrobial resistance for the antimicrobials amikacin (aminoglycoside), cephalothin, ampicillin and aztreonam (β-lactam) was observed.Moreover, 100% efficiency was observed for β-lactam cefoxitin in 68% of V. parahaemolyticus strains.For other antimicrobials belonging to these two families, a high intermediate resistance was observed, mainly to aztreonam (68%) and imipenem (52%).
Despite the high resistance to the wide-spectrum β-lactam aztreonam, no production of ESBL enzymes was observed in V. parahaemolyticus strains.Of the three types of genes (blaTEM, blaSHV, and blaCTX-M) studied, only the blaTEM gene was detected.

DISCUSSION
In the oysters from cultivation samples, a greater variation of bacteria between the samplings (p ≤ 0.05) (Figure 3A) is attributable to its filtration feed process, which contributes to the accumulation of microorganisms in the tissues, where the gills, glands, and digestive tissues are the best site for detection of V. parahaemolyticus (YU et al., 2016).The microbiological analysis in oysters has been shown to be safer than that in water in relation to the sanitary quality of the food, since water represents the environmental condition at the time of collection and, in oyster samples, these retain the microorganisms longer (SANDE et al., 2010).
The occurrence of Vibrio genus is not always associated with the presence of indicator microorganisms, rather with the physicochemical alterations of the environment.The temperature (29 °C) was the parameter that most influenced the density of Vibrio spp.when the salinity was approximately 19 ppm (Figure 4A) and the pH was around 8.0 (Figure 4B).The Northeastern coast favors the proliferation of Vibrio throughout the year due to the optimal environmental conditions of bacterial growth, with salinity varying from 15 to 25 ppm, pH of 5-11, and temperature > 21 °C (ANACLETO et al., 2013).In Bahia, in the summer months when the temperatures reach 35 °C, the supply of raw oysters increases on beaches and the lack of notification of gastroenteritis data does not allow consumers to be aware of the actual number of cases involving these foods.According to the data from the Ministry of Health in 2017, between 2000 and 2017, of the 12,503 cases of outbreaks of food-borne diseases, 46.82% were not identified (BRASIL, 2018).

7/10
Brazil has not established microbiological parameters for the presence of Vibrio spp. or V. parahaemolyticus in oysters, although, in countries such as New Zealand and Japan, the maximum level allowed for V. parahaemolyticus in raw consumed shellfish is 2 log CFU g -1 or 100 CFU g -1 (LEE et al., 2008), and in the United States and Canada, oysters may only be marketed if they contain <4 CFU g -1 or 10,000 CFU g -1 of the bacteria (CDC, 1998).The Portaria SVS/MS nº 451 of 1997 (BRASIL, 1997) established limits of 3.70 log CFU g -1 or 5,000 CFU g -1 for V. parahaemolyticus in raw fresh fish, which was revoked by the resolution RDC nº 12 of 2001 (BRASIL, 2001).The absence of V. parahaemolyticus as an indicator of quality in bivalve mollusks was also observed in the Program of Hygienic-Sanitary Control of Bivalve Mollusks (PNCMB) which, when establishing the monitoring of contaminating microorganisms, contemplates only E. coli and marine biotoxins as a preventive measure of harmful effects on the consumer health (BRASIL, 2012).It is believed that the low incidence of cases involving V. parahaemolyticus justifies the absence of bacterium in Brazilian food regulations.However, in a country where under-reporting of food outbreaks cases is high and where the involvement of an etiologic agent in 79.3% of food outbreaks cases remains unknown (BRASIL, 2018), it is necessary that public health institutions reconsider the inclusion of V. parahaemolyticus in food from fisheries, mainly in raw bivalve mollusks.KLEIN et al. (2014) found that the wide distribution of Vibrio spp. in aquatic environments tends to increase the risk to humans as agents in food outbreaks, especially when harboring pathogenic species such as V. parahaemolyticus, V. vulnificus, V. mimicus and V. cholerae.
The filtering characteristic of bivalve mollusks makes the choice of cultivation area essential for the health of cultivated organisms.In the present study, on when comparing the risk of oyster intake by Vibrio spp.obtained in natural banks with oysters obtained in the culture system, a higher density of bacteria in oysters were obtained from extractivism (E1) (Figure 3).This is due to the proximity of natural banks in the town of Taperoá, a municipality with a population 40-times greater than that of the town of Torrinhas (E2) -a quilombola community composed of 450 inhabitants (IBGE, 2015).The lack of a sewage treatment system in most Brazilian municipalities directly affects the environmental quality of the estuaries due to the release of domestic wastes, as observed in the estuary of the river Una (SANTOS et al., 2015;BONDIOLI et al., 2017), which contributes to the proliferation and maintenance of Vibrio in the water due to the increase of debris and nutrients (COSTA et al., 2009).
The gene tl, in addition to being a specific marker for the identification of V. parahaemolyticus, codes for thermolabile hemolysin (TLH).Although the contribution of hemolysin to the pathogenicity of this species remains unknown, its expression is regulated positively under conditions that mimic the human intestine (KLEIN et al., 2014).Correlation with the gene tl and the V. parahaemolyticus species was also reported by LEAL et al. (2008) in clinical isolates from patients in Pernambuco, Ceará, and Alagoas when observing the amplification of gene in all strains of V. parahaemolyticus.
Other species of Vibrio may also be able to amplify the gene tl, although these species present a smaller amplicon and different phenotypic characteristics of V. parahaemolyticus, such as positive sucrose colonies on TCBS agar and positivity for the Voges-Proskauer test (YÁÑEZ et al., 2015), which was not observed in the strains in the present study during the additional tests.
The absence of the virulence genes tdh and trh and the absence of the Kanagawa phenomenon (β-hemolysis) in the strains of V. parahaemolyticus allowed the establishment of a phenogenotypic relation of the isolates.In environmental samples, the detection of the tdh and trh genes was low (only 1-3% of the isolates), whereas, in clinical strains, the detection percent was 90% (VONGXAY et al., 2008).Similar results were also reported by COSTA SOBRINHO et al. (2011), who observed high density of V. parahaemolyticus in oysters commercialized in the retail trade in São Paulo and did not detect strains positive for the tdh and trh genes.However, GUTIERREZ-WEST et al. (2013) reported a high frequency of virulence genes tdh (48%) and trh (8.3%) in estuarine V. parahaemolyticus strains.For the authors, although this is a poorly impacted estuary, this fact demonstrates the existence of other ecologically relevant functions for these genes, including perhaps infection by other hosts, where the products of these genes are important.
The absence of the gene ure in the V. parahaemolyticus strains demonstrates a correlation between the presence of the gene trh and gene ure encoding the urease enzyme, which has been proposed as an additional virulence marker for some strains of V. parahaemolyticus (YEUNG and BOOR, 2004).Even the pathogenicity of V. parahaemolyticus strains showing no virulence with the genes studied, should not be ruled out, since other virulence factors may be present in this and other Vibrio species not studied in the present study, such as the secretion of type 3 (T3SSs), which contains pathogenicity factors that causes lysis of the infected host cell (CECCARELLI et al., 2013).Thus, continuous monitoring of V. parahaemolyticus in fishing grounds is necessary for the better understanding of the ecological behavior of the species, since these strains act as reservoirs of known or undiscovered virulence genes, but involved in human pathogenicity, that can be disseminated in the marine environment (CABURLOTTO et al., 2009).
The risk of seafood outbreaks involving V. parahaemolyticus is aggravated when the selective pressure of antimicrobial residues and the release of resistant Enterobacteriaceae family bacteria into the rivers and estuaries directly affect natural microbial populations such as vibrios.The high antimicrobial resistance observed in aquatic isolates of V. parahaemolyticus is due to their ability to exchange genetic determinants (XIE et al., 2017).This fact is aggravated by the high intermediate resistance index for β-lactams, such as imipenem (52%), aztreonam (68%) (monobactam), and ceftriaxone of the third-generation, ceftriaxone (12%) (Figure 5).Most of the antimicrobials tested in the present study are recommended for the treatment of infections caused by Vibrio spp.Third-generation cephalosporins are considered to provide some of the best defenses against serious infections by these microorganisms, and the presence of resistant isolates, even 8/10 in small percents, puts future treatments for vibrios infections at risk (SHAW et al., 2014).
One of the factors that contributed to the reduction in the efficacy of treatment with first-generation antimicrobials such as ampicillin and cephalothin, has been its wide use in aquaculture (ROCHA et al., 2016).High antimicrobial resistance to ampicillin and cephalothin has also been reported by LOU et al. (2016) and ROCHA et al. (2016), respectively. Already WANG et al. (2015) warned against the use of antimicrobials in cultivated areas, reporting that more >40% of V. parahaemolyticus strains isolated from abalones showed antimicrobial resistance, especially in areas of shrimp farming.
For TANIL et al. (2005), MAR index of >0.2 is due to contamination from high-risk sources, which puts human health at risk.The high index of multiple antimicrobial resistance with MAR >0.20 in 88% of V. parahaemolyticus strains demonstrates the anthropic influence in the study areas (Table 1), which has contributed to the fact that non-pathogenic environmental bacteria are becoming the largest reservoir of resistance to broad-spectrum antimicrobials (MANJUSHA and SARITA, 2013) along with becoming a risk to the new-generation antimicrobial agents for controlling the pathogenicity of V. parahaemolyticus (YU et al., 2016), since, after the death of the bacterium and the consequent rupture of the cell wall and membranes, genetic elements like plasmids and integrons are disseminated in the aquatic environment for incorporation in other bacteria (ROCHA et al., 2016).
Since V. parahaemolyticus strains show high intermediate resistance or resistance to the two aminoglycosides amikacin and gentamicin and to cephalosporins (antimicrobials usually used in the treatment of vibrios infections), it is recommended that, in cases of infections caused by these microorganisms, treatment with tetracycline, quinolones (nalidixic acid), and fenicol (chloramphenicol) is the best chemotherapeutic option in this region.
Although the production of β-lactamase enzymes (ESBL) in the strains of V. parahaemolyticus is still not frequent, the high resistance to second-and third-generation β-lactams observed in the present study demonstrates that antimicrobial resistance in the environmental strains needs to be monitored.In the present study, it was observed that there were no significant differences between the two strains of V. parahaemolyticus tdh and trh-negative and ESBL-producing strains in Hong Kong (WONG et al., 2012).The production of β-lactamase enzymes has been a feature observed mainly in enterobacteria (MEYER and PICOLI, 2011); however, self-transmissible plasmids from Gram-negative enterobacteria are responsible for the production of ESBL in strains of V. parahaemolyticus isolated from fish in Hong Kong (WONG et al., 2012).
In the microorganisms with >300 ESBL subtypes, the genes that encode TEM, CTX, or SHV are most commonly studied (BUSH and JACOBY, 2010).The presence of the blaTEM gene in the strains of V. parahaemolyticus was satisfactory because it is a narrow-spectrum β-lactamase, which hydrolyzes first-generation penicillin and cephalosporin (BUSH and JACOBY, 2010).The strains of V. parahaemolyticus also bearing blaTEM gene were reported by ROJAS et al. (2011) in the samples of bivalve mollusks at different points of sale in the State of São Paulo.According to the authors, it is necessary that continuous monitoring of halophilic vibrios in hospital areas near the coastal regions and in places of consumption of seafood be conducted to notify the incidence of vibrioses in the population.

CONCLUSION
Anthropogenic action in the Baixo Sul da Bahia estuaries has affected the resident microbiota, contributing thereby to the increase in multidrug resistance mediated by plasmids in V. parahaemolyticus strains.This fact represents a risk for raw oyster ingestion in the region with the transfer of resistant plasmids to the intestinal microbiota of the consumers.Thus, it is necessary to create a regulation for food and environment with specified limits to the presence of V. parahaemolyticus in seafood as well as guidelines aimed at the commercialization of these organisms from extractivism, since what is currently more focused on the production areas.It is believed that the adoption of Good Manipulation Practices together with educational actions promoted by the Sanitary and Epidemiological Surveillance can help regulate the quality of the products marketed by the fishing production chain in order to minimize the risks of food outbreaks.

Figure 1 .
Figure 1.Map of the oyster and water collection points (P1, P2 and P3) in the E1 and E2 estuaries, Baixo Sul da Bahia.

Figure 4 .
Figure 4.The plot between the environmental parameters of water and the density of Vibrio spp. in the area of cultivation (E2), Baixo Sul da Bahia.(A) Salinity x temperature and (B) pH x temperature.

Table 1 .
Multi-resistance profile, MAR index, and plasmid resistance of V. parahaemolyticus in samples of water and cultivation oysters and extractivism oysters, in two estuaries of Baixo Sul, Bahia.