Spreading of Pandemic Vibrio parahaemolyticus O3:K6 and Its Serovariants: A Re-analysis of Strains Isolated from Multiple Studies

In China, V. parahaemolyticus has been a leading cause of foodborne outbreaks and bacterial infectious diarrhea since the 1990s, and most infections have been associated with the pandemic V. parahaemolyticus O3:K6 and its serovariants. However, a comprehensive overview of the sero-prevalence and genetic diversity of the pandemic V. parahaemolyticus clone in China is lacking. To compensate for this deficiency, pandemic isolates in both clinical and environmental Chinese samples collected from multiple studies were analyzed in this study. Surprisingly, as many as 27 clinical pandemic serovariants were identified and were widely distributed across nine coastal provinces and two inland provinces (Beijing and Sichuan). O3:K6, O4:K68, and O1:KUT represented the predominant clinical serovars. Only four environmental pandemic serovariants had previously been reported, and they were spread throughout Shanghai (O1:KUT, O3:K6), Jiangsu (O3:K6, O4:K48), Zhejiang (O3:K6), and Guangdong (O4:K9). Notably, 24 pandemic serovariants were detected within a short time frame (from 2006 to 2012). The pandemic isolates were divided into 15 sequence types (STs), 10 of which fell within clonal complex (CC) 3. Only three STs (ST3, ST192, and ST305) were identified in environmental isolates. Substantial serotypic diversity was mainly observed among isolates within pandemic ST3, which comprised 21 combinations of O/K antigens. The pandemic O3:K6 serotype showed a high level of sequence diversity, which was shared by eight different STs (ST3, ST227, ST431, ST435, ST487, ST489, ST526, and ST672). Antimicrobial susceptibility testing revealed that most isolates shared similar antibiotic susceptibility profiles. They were resistant to ampicillin but sensitive to most other drugs that were tested. In conclusion, the high levels of serotypic and genetic diversity of the pandemic clone suggest that the involved regions are becoming important reservoirs for the emergence of novel pandemic strains. We underscore the need for routine monitoring to prevent pandemic V. parahaemolyticus infection, which includes monitoring antimicrobial responses to avoid excessive misuse of antibiotics. Further investigations are also needed to delineate the specific mechanisms underlying the possible seroconversion of pandemic isolates.

In China, V. parahaemolyticus has been a leading cause of foodborne outbreaks and bacterial infectious diarrhea since the 1990s, and most infections have been associated with the pandemic V. parahaemolyticus O3:K6 and its serovariants. However, a comprehensive overview of the sero-prevalence and genetic diversity of the pandemic V. parahaemolyticus clone in China is lacking. To compensate for this deficiency, pandemic isolates in both clinical and environmental Chinese samples collected from multiple studies were analyzed in this study. Surprisingly, as many as 27 clinical pandemic serovariants were identified and were widely distributed across nine coastal provinces and two inland provinces (Beijing and Sichuan). O3:K6, O4:K68, and O1:KUT represented the predominant clinical serovars. Only four environmental pandemic serovariants had previously been reported, and they were spread throughout Shanghai (O1:KUT, O3:K6), Jiangsu (O3:K6, O4:K48), Zhejiang (O3:K6), and Guangdong (O4:K9). Notably, 24 pandemic serovariants were detected within a short time frame (from 2006 to 2012). The pandemic isolates were divided into 15 sequence types (STs), 10 of which fell within clonal complex (CC) 3. Only three STs (ST3, ST192, and ST305) were identified in environmental isolates. Substantial serotypic diversity was mainly observed among isolates within pandemic ST3, which comprised 21 combinations of O/K antigens. The pandemic O3:K6 serotype showed a high level of sequence diversity, which was shared by eight different STs (ST3, ST227, ST431, ST435, ST487, ST489, ST526, and ST672). Antimicrobial susceptibility testing revealed that most isolates shared similar antibiotic susceptibility profiles. They were resistant to ampicillin but sensitive to most other drugs that were tested. In conclusion, the high levels of serotypic and genetic diversity of the pandemic clone suggest that the involved regions are becoming important reservoirs for the emergence of novel pandemic strains. We underscore the need for routine monitoring to prevent pandemic V. parahaemolyticus infection, which includes monitoring antimicrobial responses to avoid excessive misuse of antibiotics. Further investigations are also needed to delineate the specific mechanisms underlying the possible seroconversion of pandemic isolates.
Keywords: Vibrio parahaemolyticus, multilocus sequence typing, pandemic clone, gastroenteritis, genetic diversity INTRODUCTION Vibrio parahaemolyticus, a Gram-negative bacterium, is a natural inhabitant of estuarine and coastal environments. In humans, this pathogen is a globally important cause of acute gastroenteritis. It's a multi-serotype pathogenic bacteria, and can be classified into 13 O serotypes and 71 K serotypes based on somatic (O) antigens and capsular (K) antigens (Han et al., 2008). Since 1997, the pandemic isolates, including serotype O3:K6 and its serovariants, have spread globally by either sporadic diarrhea or contaminated food-related outbreaks (Nair et al., 2007;Chowdhury et al., 2013).
All the pandemic isolates share the following specific genetic markers: the presence of the thermostable direct hemolysin(tdh) gene, the absence of the TDH-related hemolysin(trh) gene, and a distinctive toxRS sequence (toxRS/new), which can be amplified by a group-specific PCR (GS-PCR) (Matsumoto et al., 2000). Until 2007, 22 serotypes had been traced to the pandemic clone based on these characteristics (Nair et al., 2007). In recent years, an increasing number of pandemic serotypes have been reported to be widely distributed in countries across four continents (Asia, Europe, the Americas and Africa) (Ansaruzzaman et al., 2008;Ottaviani et al., 2010;Powell et al., 2013;Li et al., 2014;Velazquez-Roman et al., 2014;Guerrero et al., 2017), suggesting that these pandemic serotypes pose a mounting public health threat. This threat calls for higher surveillance of the pandemic clone to reduce illnesses.
In China, V. parahaemolyticus has been the leading cause of foodborne outbreaks and bacterial infectious diarrhea since the 1990s, especially in coastal regions (Gao et al., 2016;Li et al., 2016). The pandemic serotype O3:K6 was first documented as the dominant serotype in 2002 and was proven to be a pandemic clone in 2008 (Vongxay et al., 2008). During the period of 2007-2012, approximately 56% of the clinical isolates in southern coastal areas of China had pandemic characteristics (Li et al., 2014). In one of our multi-center active surveillance programs, we found that 63.3% of the tested isolates were pandemic isolates in southeastern China from 2009 to 2013 .
The above findings indicate that the pandemic clone of V. parahaemolyticus plays an important role in causing infectious diseases in China. However, although laboratory-based surveillance for acute infectious diarrhea has been established in several coastal regions (Yu et al., 2011;Zhang et al., 2014;Li et al., 2015), a nationwide study or review of the distribution of infections caused by the pandemic isolates of V. parahaemolyticus has not been conducted in this country. Thus, a full understanding of the spread of this unique clone is needed to prevent outbreaks and sporadic illnesses in China.
In this study, we identified Chinese pandemic isolates of V. parahaemolyticus mostly from published literatures, and reanalyzed the sero-prevalence and genetic diversity of these pandemic isolates as a whole. We isolated some of these pandemic isolates in our active studies of diarrheal infection, and mainly focused on their antimicrobial responses in this study, which has not been shown previously. Overall, our intention is to generate a comprehensive overview of the spread of pandemic Vibrio parahaemolyticus O3:K6 and its serovariants in China since the emergence of this clone.

Datasets Utilized in the Present Study
To identify as many available pandemic isolates of V. parahaemolyticus as possible, we conducted a comprehensive search of several databases, including the US National Library of Medicine, PubMed, Elsevier, Springer, and China National Knowledge Infrastructure. We searched for all relevant Chinese studies using combinations of the following terms (through October 1, 2016): "Vibrio parahaemolyticus," "pandemic clone, " "pandemic strains, " "pandemic isolates, " "O3:K6 serotype, " and "O3:K6 clone." Additional eligible studies were identified from references cited in the relevant articles. The full text of each potentially relevant paper was scrutinized, and a total of 290 representative clinical and environmental V. parahaemolyticus isolates (toxRS/new+, tdh+ and trh−) were extracted from 16 studies and selected as the research subjects of this investigation.
Among the 290 pandemic isolates, 120 are from our laboratory, including 98 ST3 isolates, 21 ST88 isolates, and one ST672 isolate. Most of the isolates include information about the sampling area, year of isolation, source, serotype and multilocus sequence typing. Details on the individual isolates are summarized in Additional file 1: Table S1.
We grouped these 290 pandemic isolates according to the integrity of background information (Figure 1), and then carried out the re-analysis of the spreading of pandemic V. parahaemolyticus O3:K6 and its serovariants.

Assignment to Clonal Complexes
Various typing methods have been used to distinguish V. parahaemolyticus isolates for epidemiological investigations (Marshall et al., 1999;Gonzalez-Escalona et al., 2008;Lüdeke et al., 2015). The high accuracy and repeatability of current sequencing technology, the ability to compare DNA sequences universally and the ability to share data among laboratories make multi-locus sequence typing (MLST) a complete, robust, and reliable typing method. In recent years, scientists have begun using whole genome sequencing (WGS) to analyze historical collections of isolates, providing new insights for understanding population dynamics among different V. parahaemolyticus isolates. However, many laboratories are not currently using WGS, as data are not easily shared among laboratories. For this reason, we used MLST to assign clonal complexes for the pandemic isolates collected in this study.
The MLST scheme used internal fragments of the seven housekeeping genes [recA (729 bp) , dnaE (557 bp) , gyrB (592 bp) , dtdS (458 bp) , pntA (430 bp) , pyrC (493 bp) , and tnaA (423 bp) ]. The standard amplification protocol was published on the V. parahaemolyticus MLST web site (http://pubmlst.org/vparahaemolyticus/). The allele designations and sequence types (STs) of all the selected isolates had been determined based on the variation of the seven genes. Based on the defined STs, all the pandemic isolates were compared using global optimal eBURST analysis (goeBURST) version 1.2.1 (http://www.phyloviz.net/goeburst/). Clonal complexes were conservatively defined as a cluster of STs, in which all STs were linked as single-locus variants (SLVs, two STs differing from each other at a single locus) to at least one other ST (Feil et al., 2004). The singleton STs corresponded to STs differing from the others by three or more of the seven loci (Esteves et al., 2015).

Antimicrobial Susceptibility Testing
We selected the pandemic isolates [ST3 (n = 98) and ST88 (n = 21)] detected in our laboratory from the 290 isolates to be analyzed in this study and other 33 non-pandemic isolates identified in our laboratory to conduct antimicrobial susceptibility testing with 20 antimicrobial agents (Figure 1). The testing was performed using the disk diffusion method according to the Clinical and Laboratory Standards Institute (CLSI, 2010). E. coli ATCC25922 and Staphylococcus aureus ATCC25923 were employed as bacterial strains for quality control. Characterization of the resistance and susceptibility profile of the isolates was determined by measuring inhibitory zone, and then compared with the interpretative chart ( Table 1).
When sorting the isolates by the year of isolation, we found that only the O3:K6, O1:K25, O4:K68, and O1:KUT4 serotypes had been reported prior to 2005. However, from 2006 to 2012, as many as 24 more pandemic serotypes were detected ( Table 3). This may be a significant period for the spreading of pandemic serotypes in China, but underreporting and possible study deviation should also be considered. As these factors directly affect the accuracy of our conjecture. O3:K6, first discovered in 1996, was detected each year from 2002 to 2012.

Genetic Differentiation of the Pandemic Isolates
Among the 290 collected pandemic isolates, 28 isolates was not typed by MLST analysis in their original studies, the other 262 isolates were involved the MLST analysis in the current study (Figure 1). These pandemic isolates exhibited 15 STs, revealing high genetic diversity ( Table 4). The sequence variation of the isolates in Guangdong (11 STs) was significantly higher than that in other provinces (Figure 1). ST3 was the only sequence type that caused a wide range of infections in as many as ten provinces. Only three sequence types (ST3, ST192, and ST305) had ever been identified in environmental isolates.
Ten of the 15 pandemic STs could be classified in CC3 (Figure 3). ST305 and ST672 were DLVs of ST3. Pandemic ST88 was found only in Zhejiang province during 2010-2012 and it belongs to CC345, not CC3. ST302 originated from in the Shanghai province and was identified as a singleton with no relationship to other STs.

The Association between Pandemic STs and Serotypes
The pandemic isolates within ST3 comprised 21 clinical and two environmental serotypes (O3:K6 and O4:K48), thus exhibiting high serotypic diversity ( Other serotypes were clustered in no more than two different pandemic STs.

Antibiotic Susceptibility Profile of Different Pandemic Isolates
The results of antimicrobial susceptibilities of the 98 pandemic ST3 isolates, 21 pandemic ST88 isolates and 33 non-pandemic isolates are shown in Table S2. The results indicate similar antimicrobial profiles between isolates within different STs. In other words, there were not obvious differences in their resistance spectrums (Figure 4). Specifically, the isolates were mostly resistant to ampicillin (AMP) (95.2% of ST88, 85.7% of

DISCUSSION
As an emerging public health concern, pandemic V. parahaemolyticus infection has attracted wide attention from scientists (Jun et al., 2014;Velazquez-Roman et al., 2014;de Jesús Hernández-Díaz et al., 2015). The present study has provided an overview of the prevalence of pandemic isolates of V. parahaemolyticus in both clinical and environmental samples collected from multiple Chinese studies. We demonstrated that these pandemic isolates showed high serotypic and genetic diversity. The O3:K6 pandemic isolates (persistent for 11 years from 2002 to 2012) spread across 11 provinces indicate that the pandemic clone has been endemically established in China. Continued monitoring of antibiotic resistance patterns in pandemic isolates is urgently needed to avoid the excessive misuse of antibiotics, although most of the isolates tested in this study only showed high resistance to ampicillin. The serovariants of this pandemic clone were abundant and variable. From 1996 to 2007, up to 22 pandemic serovariants were identified around the world (Nair et al., 2007), suggesting that the pandemic isolates could easily adapt new serotypes to survive in new environments. In this study, we found that 27 pandemic clinical serotypes and four environmental serotypes have been identified in the collected Chinese isolates. This is a significant number when compared to the number of serotypes found in other countries or regions (Nair et al., 2007;Pazhani et al., 2014;Velazquez-Roman et al., 2014), although the regional persistence of O3:K6 pandemic isolates has also been discovered in many other countries, such as Peru (Gil et al., 2007), Chile   (Ansede-Bermejo et al., 2010), Japan (Okuda et al., 1997), India (Pazhani et al., 2014), and Thailand (Mala et al., 2016). To the best of our knowledge, no study has explained why serotypes of the pandemic clone are so diverse in China. Researchers around the world have made some progress in finding the cause of such serodiversity of the pandemic clone. Such research has determined that the O-and K-antigens are mostly mutated concurrently by horizontal gene transfer (HGT) (Okura et al., 2008), and it is speculated that new serovariants have emerged from the pandemic O3:K6 strains via replacement of the putative O and K antigen gene clusters (Okura et al., 2008;Harth et al., 2009;Chen et al., 2010). This is important for the survival of the pandemic clone in the face of changing external environments and host immunological resistance.
The analyzed isolates in this study were distributed in regions where the differences of temperature and other environmental factors are vast. Therefore, we suspect that the strains must adapt to different living environments in the transmission process by altering their biological properties more frequently. Making serological changes may be one of the most efficient ways for this to be accomplished. However, the specific mechanism of serotype conversion is not yet known. The highest priority is currently to continuously monitor the emergence of new serovariants to prevent the pandemic strains from causing outbreaks along the coastline and spreading to other countries and regions.
MLST is known to provide greater resolution than stereotyping. In a previous study, we provided an extensive review of isolates found in Chinese patients, and the results indicated a high degree of genetic diversity and a complicated population structure of V. parahaemolyticus (Han et al., 2015). With the present MLST study, we intended to demonstrate genetic similarities or differences among the pandemic isolates identified from different sources. As expected, we found that most of the pandemic STs were classified into the same clonal complex (CC3). However, some molecular divergence was also found. All the pandemic ST88 in Zhejiang province were assigned to CC345 (not CC3), and ST302, which originated in Shanghai province, was identified as a singleton with no relationship to other pandemic STs. Singletons do not seem to belong to the same clone as other pandemic isolates, but they do share the pandemic traits (toxRS/new+, tdh+, and trh−).
Chen et al. demonstrated the isolates of pandemic ST302 were clustered with other pandemic isolates based on other molecular typing methods, such as enterobacterial repetitive intergenic consensus sequence PCR (ERIC-PCR) (Chen et al., 2012). Similarly, in our pulsed field gel electrophoresis (PFGE) analysis (data not shown), the pandemic isolates of ST88 and ST3 shared the same PFGE profile. Thus, we observed that current typing and clustering methods may lead to controversial results, making it difficult to draw conclusions.
Therefore, a combined application of several molecular typing techniques should be considered in epidemiological investigations of V. parahaemolyticus pandemic isolates. As explained before (in Materials and Methods section), WGS is the best way to accurately portray the evolution and population structure of V. parahaemolyticus isolates at the molecular level (Cui et al., 2015;Haendiges et al., 2015), but the high cost limits its popularity in the analysis of large quantities of specimens.
Another important aspect of this study was the investigation of the antimicrobial susceptibility of different pandemic isolates. Our results revealed similar antibiotic susceptibility profiles in pandemic ST3, ST88, and non-pandemic isolates. This finding was similar to that in the work of Elmahdi et al. (2016). They concluded that the sampling location or month in which the samples were collected did not significantly impact V. parahaemolyticus resistance patterns because isolates from both environmental and clinical sources shared similar antibiotic resistance profiles. Unsurprisingly, the majority of the isolates tested in this study showed ampicillin resistance, which is very common in V. parahaemolyticus isolates recovered from different sources (Sun et al., 2013;de Jesús Hernández-Díaz et al., 2015;Elmahdi et al., 2016;Mala et al., 2016). This result suggests that ampicillin should have a negligible role in the treatment of V. parahaemolyticus infection. In fact, a survey conducted in the United States showed that very high V. parahaemolyticus ampicillin resistance could be traced as far back as 1978 (Blake et al., 1979).
In contrast, most of the isolates tested were sensitive to the majority of antibiotics tested, and all isolates were susceptible to LVX, MEM, IPM, FEP, and CAZ. This result suggests that these drugs can be used as an alternative antibiotic therapy. It must be noted that recently isolated V. parahaemolyticus strains, including pandemic strains, have displayed resistance to multiple antibiotics (Jun et al., 2012(Jun et al., , 2014, which increases concerns about possible antibiotic treatment failure. Although we did not discover any multidrug resistant bacteria (MDR) isolates in this study, continued monitoring of pandemic strain susceptibility to antibiotic resistance is urgently needed to avoid the excessive misuse of antibiotics used to treat infections that pose threats to public health.
Our findings represent a comprehensive review of the pandemic V. parahaemolyticus O3:K6 and its serovariants by thoroughly assessing an extensive collection of clinical and environmental pandemic isolates from multiple Chinese studies. High levels of serotypic and genetic diversity in the pandemic clone are found, which suggests that the involved regions are becoming important reservoirs for the emergence of novel pandemic strains, which makes the clinical management of the infection and its prevention potentially challenging. Thus, we underscore the need for routine clinical and environmental monitoring to prevent pandemic V. parahaemolyticus infection and dissemination, including monitoring of antimicrobial response even though most current antimicrobial agents in routine use are effective. The mechanism in which the isolates undergo seroconversion with pandemic genetic marks warrants extended investigation.

AUTHOR CONTRIBUTIONS
Conceived and designed the experiments: DH and CH. Performed the experiments: DH, FY, and HT. Analyzed the data: DH, CW, and PZ. Contributed reagents/materials/analysis tools: DH and CR. Wrote the paper: DH.