Vibrio parahaemolyticus Is Associated with Diarrhea Cases in Mexico, with a Dominance of Pandemic O3:K6 Clones

In the present study, we conducted surveillance of the V. parahaemolyticus strains present in clinical samples from six geographical regions of Mexico (22 states) from 2004 to 2011. The serotype dominance, virulence genes, presence of pandemic O3:K6 strains, and antibiotic resistance of the isolates were investigated. In total, 144 strains were isolated from the clinical samples. Seven different O serogroups and twenty-five serovars were identified. Most clinical isolates (66%, 95/144) belonged to the pandemic clone O3:K6 (tdh+, toxRS/new+ and/or orf8+) and were detected in 20 of the 22 states. Among the pandemic clones, approximately 17.8% (17/95) of the strains cross-reacted with the antisera for the K6 and K59 antigens (O3:K6, K59 serotype). Other pathogenic strains (tdh+ and/or trh+, toxRS/new−, orf8−) accounted for 26.3%, and the nonpathogenic strains (tdh− and/or trh−) accounted for 7.6%. Antimicrobial susceptibility testing showed that most of the strains were resistant to ampicillin (99.3%) but were sensitive to most tested antibiotics. The level of multidrug resistance was 1.3%. Our results indicate that pandemic O3:K6 is present in most Mexican states, thus, constant surveillance of V. parahaemolyticus strains in diarrhea patients is a public health priority and is useful for conducting risk assessments of foodborne illnesses to prevent V. parahaemolyticus outbreaks. Overall, our observations indicate that the pandemic O3:K6 clone of V. parahaemolyticus has become a relatively stable subpopulation and may be endemically established in Mexico; therefore, constant surveillance is needed to avoid new outbreaks of this pathogen.


Introduction
Vibrio parahaemolyticus (V. parahaemolyticus) is a facultative anaerobic, Gram-negative, curved rod-shaped bacterium that is commonly found worldwide in marine and estuary environments [1]. It causes approximately 50% of total bacterial food poisoning outbreaks, which primarily result from the consumption of raw, undercooked, or mishandled seafood and marine products [2,3]. Although not all strains of V. parahaemolyticus are considered to be pathogenic, the potentially virulent strains are commonly differentiated from the likely avirulent strains by the presence of thermostable direct (tdh) and/or tdh-related (trh) hemolysin genes, or both [4][5][6]. The main disease that is produced by these toxins is acute gastroenteritis, and its symptoms include diarrhea with abdominal cramps, nausea, best of our knowledge, this is the first report to describe the presence and distribution of V. parahaemolyticus strains in clinical isolates for six geographical regions (in 22 Mexican states) of Mexico, and the pandemic O3:K6 clone is the predominant clone causing gastrointestinal infections in this country.

Area of Study and Bacterial Strains
The strains used in this study were derived via the InDRE biosurveillance strategy to identify cases of diarrhea caused by V. parahaemolyticus. This biosurveillance is made possible by a national network of public health laboratories throughout Mexico; each laboratory takes samples based on convenience sampling of patients with suspected diarrhea associated with V. parahaemolyticus by rectal or fecal swabs, and then the bacterium is isolated following InDRE guidelines [33]. In total, 144 strains of V. parahaemolyticus from 22 Mexican states were isolated between 2004 and 2011 in Mexico. These 22 Mexican states were distributed in 6 different regions of Mexico (Table 1), according to the Diario Oficial de la Federación (www.dof.gob.mx, accessed on 30 June 2022) and the Instituto Nacional de Estadística y Geografía (www.inegi.org.mx, accessed on 30 June 2022). Briefly, non-bloody stool samples (from clinical cases who had consumed raw, undercooked, or poorly handled seafood and/or seafood products) were collected in Cary-Blair transport medium (Thermo Fisher, Waltham, MA, USA) and transported at room temperature (RT) to the State Public Health Laboratory within 2 h. These samples were also spiked in sterile alkaline peptone water (APW) (pH 8.6) for 6-8 h at 37 • C. After incubation, the enrichment broths (APW) were spread on thiosulfate-citrate-bile salts-sucrose agar (TCBS) plates and/or on CHROMagar Vibrio (CV) medium (CHROMagar, Paris, France), and subsequently incubated at 37 • C for 18-24 h [34]. At least three randomly selected typical V. parahaemolyticus colonies were isolated from each plate and identified by biochemical and PCR tests, as described below. Isolated bacteria were stored at -80 • C in 20% (v/v) glycerol for further analysis (i.e., serotyping and virulence genes).

PCR Assays
The DNA of V. parahaemolyticus strains was extracted using the Wizard Genomic DNA Purification Kit (Promega Corp., Madison, WI, USA), according to the manufacturer's instructions. The polymerase chain reaction amplifications were performed in 25 µL reaction volumes consisting of 1 × GoTaq Green Master Mix (Promega), primers targeting either the tl gene [4], the pR72H plasmid [35,36], or the tdh, trh [4], toxRS/new [37], and orf8 genes [38], and 0.5 µL of purified genomic DNA template, with the remaining volume consisting of molecular-biology-grade water. PCR was carried out in a C1000 Thermal Cycler (Bio-Rad Laboratories, Hercules, CA, USA). Ten-microliter aliquots of each amplification product were separated by electrophoresis in 2% agarose gels. Ethidium bromide was added at a concentration of 0.5 mg/mL to enable visualization of the DNA fragments with a digital imaging system (Gel Doc EZ imager, Bio-Rad, Hercules, CA, USA). The sizes of the PCR fragments were compared against a 50 bp DNA ladder (Promega DNA Step Ladder).

Determination of O:K Serotypes
The O (somatic) and K (capsular) serotypes of the strains were determined using a commercially available V. parahaemolyticus antiserum test kit with O1-O11 antisera and 71 K antisera (Denka Seiken, Tokyo, Japan), which was used according to the manufacturer's instructions. Briefly, the strains were first grown overnight at 37 • C on LB agar containing 3% NaCl. Later, a pool of colonies was collected and suspended in 1 mL of saline solution and then split into two aliquots (0.5 mL). For O serotyping, one aliquot was boiled at 121 • C for 2 h. The remaining cell suspension, which was not boiled, was used for serotyping based on the K antigen.

Statistical Analysis
All statistical analyses were performed using SPSS v.20.0 (IBM Corp., Armonk, NY, USA). Chi-squared tests were conducted to evaluate significance; p-values of ≤0.05 were considered to be statistically significant.

Geographical Distributions of the V. parahaemolyticus Strains Isolated from Diarrhea Cases in Mexico
From 2004 to 2011, a total of 144 V. parahaemolyticus strains were isolated from stool specimens or rectal swabs that were collected from persons with gastroenteritis who had eaten seafood or marine products in 22 Table 1). Regarding the distribution of V. parahaemolyticus by year, 2011 was the year in which the most cases of diarrhea caused by this bacterium were found (63/144)-mainly in the North Pacific region (29/63); the remaining distribution by year is shown in Table S1.

Distributions of Pathogenic and Pandemic Vibrio parahaemolyticus Serotypes in Mexico
In the present study, O3:K6 represented the predominant serovar (66%; 95/144) in most regions, whereas the cross-reacted serovars O3:K6 and K59 represented 17.8% (17/95) of the pandemic clones in the North Pacific region (60.3%), North Gulf region (77.2%), Central Pacific region (76.4%), Central Gulf region (75%), South Pacific region (100%), and South Gulf region (42.8%) ( Table 4). Only in two states in the Central Pacific region (Queretaro and Guanajuato) were the pandemic O3:K6 clones of V. parahaemolyticus not detected (Table 4 and Figure 1). It is noteworthy that in the clinical V. parahaemolyticus strains, pathogenic versions (with tdh and/or trh) were detected at levels between 17.6% and 35.8% in the geographical regions analyzed (except in the South Pacific region and in the states of Aguascalientes, Michoacán, Puebla, Tlaxcala, and Tabasco), where pathogenic strains were not detected (Table 4 and Figure 1). In contrast, the nonpathogenic group (7.6%; 11/144) was detected in only three different regions and four states: in the North Pacific region (3.7%; Sinaloa and Nayarit), Central Pacific region (11.7%; Ciudad de Mexico), and South Gulf region (33.3%; Campeche) ( Table 4 and Figure 1).  Table 5 summarizes the percentages of the antibiotic resistance profiles of V. parahaemolyticus that were isolated from the clinical samples from 2004 to 2011. The largest proportion of V. parahaemolyticus strains was resistant to ampicillin (99.3%), followed by gentamicin (15.9%), and the smallest proportions were resistant to ceftazidime (2.7%), sulfamethoxazole-trimethoprim (0.6%), and cefotaxime (0.6%). Of the 144 V. parahaemolyticus isolates, 99.3% showed some degree of resistance to at least one antibiotic, and 1.3% showed multidrug resistance (i.e., resistance to two or three antibiotics). Based on these results, the resistance rates of the pandemic O3:K6 and nonpathogenic strains of V. parahaemolyticus to ampicillin in our study were 100%, with 97.3% resistance for the pathogenic strains, but low resistance levels were determined for gentamicin (15.7%), ceftazidime (4.2%), sulfamethoxazole-trimethoprim, and cefotaxime (1%). Additionally, low resistance levels were determined for gentamicin (21%) in the pathogenic strains. In contrast, with regard to the overall levels of antibiotic resistance, only 2.1% of the pandemic strains exhibited multidrug resistance (Table 5). Furthermore, the mean antibiotic inhibition diameters of the 144 V. parahaemolyticus strains are shown in Table S2.

Discussion
In Mexico, the first outbreak (1230 cases) that was caused by the pandemic O3:K6 clone occurred in 2004 and was associated with the consumption of raw or undercooked shrimp in Sinaloa-a state in northwestern Mexico [22,27]. Since then, there have been many reports of V. parahaemolyticus-associated gastroenteritis cases in a relatively small geographical area of Sinaloa [22]; therefore, the pandemic O3:K6 clone has become established endemically on the Pacific coast of Mexico [20]. In recent years, new cases have arisen in different areas of Mexico. The present study provides an overview of the presence of the pandemic isolates of V. parahaemolyticus in clinical samples that were collected over eight years from 2004 to 2011, along with their distributions in Mexico across six regions and 22 states (out of the 32 states of Mexico), indicating that the pandemic O3:K6 clone (with the tdh and toxRS/new genes, and with or without the orf8 gene) has been disseminated and endemically established in Mexico, in addition to the appearance of clinical cases that are associated with a pandemic strain that cross-reacts with the antisera for the K6 and K59 antigens (O3:K6,59), which suggests the emergence of new clinical strains with pandemic traits. In addition, we detected high serotypic and genetic diversity in both the pathogenic and nonpathogenic strains. To the best of our knowledge, these observations represent the first report that documents eight years of pandemic O3:K6 clone persistence in six geographic regions (in 22 Mexican states) of Mexico.
As in our previous studies in northwestern Mexico, where most strains that were isolated from 2004 to 2013 belonged to the O3:K6 serotype [22], in this study, covering different regions of Mexico, most of the isolated strains consisted of serotype O3:K6. A regional predominance of O3:K6 pandemic strains has been reported in different geographical areas of the world. For example, O3:K6 was the predominant serovar in studies conducted in China [41,42], India [43], Thailand [44], and other Asian countries [12,15,45], as well as in Peru in 2007 [37], in Brazil, and in Chile (2004)(2005)(2006)(2007)(2008)(2009) [21,46]. Furthermore, Harth et al. [47] made interesting observations regarding serotype replacement in Chile and reported a decrease in the number of outbreaks that were caused by the O3:K6 pandemic and an increase in the number of cases that were caused by the pandemic isolates belonging to serotype O3:K59, with 25% cross-reacting with the antisera for the K6 and K59 antigens. In our study, from 2004 to 2011, we determined that 17.8% (17/95) of strains belonged to the cross-reacted serovars O3:K6 and K59, carrying pandemic genes (e.g., tdh, toxRS/new, and orf8). Certain mechanisms have been proposed for the serovar changes-for example, mutation and lateral gene transfer in the genes for the biosynthesis of capsular polysaccharides (K antigen) and the somatic O antigen, which may be among the methods for bacteria to adapt to environmental changes and human defense responses [48]. The rapid emergence of the non-O3:K6 serotypes carrying pandemic markers provides another example of the predisposition of V. parahaemolyticus to genetic change [13,48].
In addition, the recent temperature increases could facilitate the establishment and spread of deliberately or accidentally introduced species and increase the rates of Vibrioassociated illness in other parts of the world [49][50][51][52][53]. It has been reported that elevated temperatures facilitate the proliferation of this pathogen [54]; however, in this study, pandemic V. parahaemolyticus strains were detected in six geographical regions of Mexico with different environmental conditions and, thus, temperatures. As such, it is highly likely that the warming of our coastal oceans (e.g., the Pacific Ocean and the Gulf of Mexico) will accelerate the spread of Vibrio parahaemolyticus to the southern latitudes, from the North Pacific region to the South Gulf region, and possibly further than 2700 km, where the warming effect is projected to be most pronounced.
The origins and dissemination routes of pandemic V. parahaemolyticus from its arrival in Mexico remain unknown. However, we speculate that the increased seawater temperatures and consumption of raw or undercooked seafood (e.g., oysters and shrimp) represent factors that are extremely favorable for the dissemination of this microorganism. In Mexico, shrimp is the most important seafood export, as either farm-raised or wildcaught shrimp [55,56]. Cases of gastroenteritis, which have been attributed to raw shrimp consumption in southern Sinaloa and northern Nayarit, have been documented [20,22,27]. All of these factors-including climate change [51,57], discharge of ballast waters from ships traveling from areas of V. parahaemolyticus endemicity [58], geographical location and local eating habits, sample selection and laboratory testing, and the human activities of seafood distribution chains-may have influenced and facilitated the dissemination of the O3:K6 pandemic clone and its serovariants in the country.
Interestingly, in this study (2004-2011), we did not identify new serovars or differences from those isolated in our previous investigations that were conducted from 2004 to 2013 in northwestern Mexico [20,22]. We presume that this finding indicates that the serovars from Sinaloa were also present in other regions of Mexico during the same time period. It has been proposed that stereotyping cannot differentiate all strains that are isolated from different regions or sources [14]. It is well known that the presence of the tdh and/or trh genes in a strain has been demonstrated to rapidly induce inflammatory gastroenteritis [5], and this trait is routinely used to determine the pathogenicity of V. parahaemolyticus strains [59]. Several studies have demonstrated that up to 90% of all clinical isolates possess the tdh and/or trh genes (i.e., the pandemic serotype O3:K6 strain has increasing prominence) [11,22,46]. The data obtained in the present study from 2004 to 2011 are in accordance with these observations, as 92.2% of the clinical strains carried the tdh and/or trh genes (66% were in the pandemic group and 26.3% were in the pathogenic group, with several serotypes). On the other hand, and most importantly, 7.6% (11/144) of the clinical isolates were classified in the nonpathogenic group (e.g., tdh− and trh-negative), which is a greater prevalence than that reported in our previous studies, with 6.5% and 2.8% from 2004 to 2011 [22] and 2011 to 2013 [20], respectively. However, this rate was lower than that reported by Chao et al. [14] in Jiangsu, China, where approximately 12% of the clinical isolates were both tdhand trh-negative.
Another contribution of our study is the investigation of the susceptibility of the isolated V. parahaemolyticus strains to the first-line antibiotics that are utilized in Mexico. The frequent phenomenon of multidrug resistance in V. parahaemolyticus directly affects the application of antibiotics and the prevention and treatment of bacterial infectious diseases [60]. We determined that most V. parahaemolyticus isolates were resistant to ampicillin (99.3%), independent of their pathogenic potential and geographical origin, based on data previously reported in Sinaloa [20]. Interestingly, in southeastern China from 2013 to 2017, Chen et al. [61] reported 88.5% resistance to ampicillin, and Ottaviani et al. [62] also reported high resistance to ampicillin (100%) from shellfish and clinical sources in Italy. This resistance to ampicillin around the world is very common in V. parahaemolyticus strains that have been isolated from environmental and clinical samples [11,12,63] and suggests that these drugs have an ineffective and invalid role in the treatment of V. parahaemolyticus [64,65]. In contrast, most isolates are sensitive to ceftazidime, cefotaxime, chloramphenicol, ciprofloxacin, nalidixic acid, sulfamethoxazole-trimethoprim, and tetracycline, which can be used as alternative antibiotic therapies, in accordance with the data presented by Hernandez et al. [20]. In Mexico, and possibly in other countries, patients suffering from disease caused by V. parahaemolyticus are treated with empirical antibiotic therapy, which could generate resistance to first-line antibiotics.
The arrival and massive proliferation of the V. parahaemolyticus pandemic strain in these six regions (22 states) of Mexico offered us an exceptional opportunity to study the evolution of a clonal strain in its natural environment. The results presented here complement previous findings [20,22] and offer new observations that increase our knowledge of the V. parahaemolyticus outbreaks in Mexico. Overall, we show that the pandemic O3:K6 strain has become a relatively stable bacterial subpopulation of the diverse V. parahaemolyticus population that is present in clinical samples in Mexico.
To the best of our knowledge, this is the first paper to report the presence of V. parahaemolyticus in 22 Mexican states; moreover, we demonstrated that pandemic O3:K6 clone is the dominant etiological agent of diarrhea caused by V. parahaemolyticus in Mexico. The limit of this study was that we could not demonstrate the prevalence of V. parahaemolyticus in Mexico because this work was derived from a biosurveillance strategy in which a representative sample from the national network of public health laboratories was analyzed based on convenience sampling.

Conclusions
The results of this study provide evidence that the pandemic O3:K6 clone is dominant among diarrhea cases caused by V. parahaemolyticus in Mexico. The pandemic strains were found from northern to southern Mexico, in 20 of the 22 studied Mexican states across the six Mexican regions. Furthermore, this is the first study to detect the cross-reacted serovars O3:K6 and K59 carrying pandemic genes (e.g., tdh, toxRS/new, and orf8), guaranteeing continuous monitoring of V. parahaemolyticus strains and allowing public health authorities to provide the best education for emergency care physicians and general practitioners in the event of local or multistate foodborne outbreaks of V. parahaemolyticus-associated gastroenteritis. Finally, the data presented in this work indicate that the pandemic O3:K6 clone of V. parahaemolyticus has become a relatively stable subpopulation and may be endemically established in Mexico.