Molecular Typing And Genetic Variation Analysis Of Vibrio Parahaemolyticus Isolates From Different Sources And Different Regions

Vibrio parahaemolyticus, a Gram-negative halophilic bacterium, is one of themajor casues of foodborne diseases. Food poisoning caused by V. parahaemolyticusare annually reported worldwide, especially in Asia. The Chinese NationalInfectious Disease Internet Reporting System (http://www.moh.gov.cn) showed thatV. parahaemolyticus has become the leading cause of microbial food poisoningincidents in China during2009to2011. Since1996, a new serovar O3:K6emergedas a major cause of illness because of its stronger ability to spread and pathogenicity.In addition to O3:K6, others, such as O4:K8, O4:K68, O1:K25, O3:K29and O1:K56,are also the common serotypes in clinical. In Mainland China, we do not fullyunderstand the serodiversity, virulence, genotype and evolution of the V.parahaemolyticus strains isolated from different regions and sources.In this work, we collected206V. parahaemolyticus strains from Shanghai,Shenyang and Beijing respectively between2009and2011. By using biochemicalidentification, serotyping and PCR techniques, we analysed the distribution ofserotype and virulence genes for the isolates from different regions and sources.Then we subtyped the strains by using the methods of pulsed-field gelelectrophoresis (PFGE), multilocus sequence typing (MLST) and multilocusvariable number tandem repeat analysis (MLVA) in order to explore the geneticrelationships and evolution, and provide insight into the genotypic characteristics ofthe new serotypes.Among the206strains,51were isolated from Shenyang city,15from Beijingand140from Shanghai, and155clinical isolates and51environmental isolates wereincluded in this study. It is noteworth that we've also monitored3food poisoning outbreaks in Shanghai, and32isolates were screened among them.V. parahaemolyticus has sero-diversity, only a few serotypes can causeinfections. In this study206strains were serotyped into42different combinations ofthe somatic (O) and capsular (K) antigens, of which14were firstly reportedworldwide. O3:K6clone was the predominant serovar, and it caused three foodpoisoning outbreaks in Shanghai. Twenty five new serotypes were identified amongthe155clinical isolates, and of them seven (O3:K8, O3:K9, O2:K53,O3:K13,O5:K6, O2:KUT and O4:KUT) were reported for the first time. The51environmental isolates were typed into25serotypes, of which10were firstlyreported. Interestingly,10serotypes were shared by the environmental and clinicalisolates, and this result provides an indirect evidence for the correlation of the V.parahaemolyticus infections and environment or food contamination.Thermostable direct hemolysin (tdh), thermostable direct hemolysin-relatedhemolysin (trh) and thermolabilehemolysin (tl) genes are the main virulence factorsof V. parahaemolyticus. Here we screened the tdh, tl, trh, toxR and flaE genes fromthe206strains by using PCR assays. All the strains contained the tl gene, and all butone had the toxR and flaE genes. Only2isolates were positive for trh, and70.9%ofthe isolates possessed the tdh gene. All of the O3:K6strains contained the toxR, tdhand tl genes. Furthermore, among the serotypes shared by environmental andclinical isolates, e.g. O1:KUT, O3:K7, O3:K8, O3:K13, O3:KUT and O4:KUT, theclinical isolates had different tdh amplification result from the environmentalisolates. The above results showed that tdh gene can be used as a marker fordifferentiating the virulent and avirulent strains. In this study several isolates did notcontain the tdh gene, the tdh gene may have been lost during subculture.In this study, we used the standard PFGE method recommended by PulseNet,the international food-borne monitoring network, to analyze the genotypes of191V.parahaemolyticus strains. Cluster analysis was performed by using the BioNumericeVersion6.0software with the UPGMA method and Dice coefficient. One hundredand ninety-one strains were divided into111PFGE patterns, and10major clusters(one to ten) were obtained according to the cutoff value of62%. Cluster Three wasthen divided into two sub-clusters (subcluster A and B) at the cutoff value of64%.One hundred and four strains were located in subcluster A, accounting for54.5%ofthe strains, and most of them were clinical isolates and there were only two environmental isolates. In this subcluster (A), isolates belonging to O3accounted for94%, and they included the O3:K6, O4:K68, O1:KUT and O3:K25isolates. Thisresult revealed that the O4:K68, O1:KUT and O3:K25serotypes may be evolvedfrom O3:K6. Furthermore, the PFGE patterns of O3:K6in this subcluster (A) hadobvious regional variation except for several isolates which shared the same patterns.In cluster Five, most of them were environmental isolates and they showed differentpatterns. But in this cluster one clinical isolate belonging to the new serotype O5:K6shared the same pattern with two environmental isolates belonging to O5:KUT andone environmental isolate belonging to O5:K17. This result indicated that thesethree serotypes had close genetic relationship.The MLST method described in PubMLST database (http://pubmlst.org/vparahaemolyticus/) was used to subtype the201V. parahaemolyticus strains. Theywere divided into63Sequence types (STs), and90new allele numbers and48newSTs were identified. The above result indicated that V. parahaemolyticus strains hadconsiderable genetic diversity and heterogeneity. In this study the serotypes O3:K6,O3:KUT, O4:K68, O3:K25, O1:K6and O1:KUT were probably originated fromO3:K6as they all belonged to ST3. Interestingly,14isolates of serotype O4:K8andone O3:K8isolate had the same PFGE pattern and belonged to a new ST, ST478which experienced recombination in the recA housekeeping gene. Phylogeneticanalysis generated by eBURST software showed that ST3was the founder, whichhad12single locus variants and5double locus variants, and they formed a bigclonal complex. Among the48new STs,12STs including ST331,332,333,433,434,435,444,447,448,452,472and478had SLVs or DLVs, the rest weresingletons surrounding the ST3clonal complex.We selected8loci with high polymorphisms from the14VNTR loci, and usedthese loci to analyze the204V. parahaemolyticus strains. They were typed into156VNTR genotypes, and59VNTR genotypes were obtained from the92strains ofserotype O3:K6. This result also showed the genetic diversity of V.parahaemolyticus, and proved that the MLVA method developed in this study hadhigher resolution that PFGE and MLST. Cluster analysis showed that the isolateswith the same serotypes generally had the same genotypes or clustered together.There are exceptions that many isolates belonging to the same serotype revealeddistantly genetic relationship. The isolates of serotypes O4:K68, O3:K25, O3:KUT, O1:KUT and O1:K33, as well as O3:K6,are clustered together, revealing that O3:K6maybe the ancestor of other serotypes. Moreover, the VNTR genotypes of serotypeO3:K6and O4:K8had clearly regional variation. The above results provide the basisfor the source-tracking of V. parahaemolyticus infections or food poisonings.Phylogenetic analysis constructed by minimum spanning tree showed that otherserotypes may be evolved from O3:K6.Put VNTR results import BioNumerics software for cluster analysis,204Vpstrains were divided into156VNTR types.The polymorphism of VP2-07, VPTR1and VPTR6were significantly higher than that of the other five sites, respectively,47,32,26different alleles.The92strains of03: K6serotypes ware divided into59types, which shows the Vp strain has a genetic polymorphism.The same serotypestrains have the same VNTR type or got together, and some strains have the sameserotype, but their MLVA types ware quite different, which may have differentepidemic clones or with different evolutionary origins.In the sub-cluster A1, theserotypes04: K68,03: K25,03: KUT,01: KUT and01: K33were got together with03: K6serotype strains. Which shows these types may be evolved from the03: K6strains. In this study, the environmental isolates and clinical isolates have greatergenetic differences, they may have different genetic origins. MLVA typingandserotyping, PFGE typing have a good correlation, three methods not only canreflect the slight variation of the genetic level between strains, the methods are alsosimple, have short experimental period and low cost, the results are digitized, we canexchange the information through the BioNumerics database. MLVA has a higherresolution was compared with MLST and PFGE. Typing techniques and analysismethods established in this study laid the technical foundation for the future ofinfectious diseases and infectious diseases network monitoring, early warning,outbreak traceable. The new pathogen monitored by this study also provided thebasic data for the prevention and treatment of infectious diseases and otherapplications.In conclusion, we obtained the following meaningful results:①Wepreliminary illustrated the serotype distribution of V. parahaemolyticus strains fromdifferent sources and regions, and14new serotypes were identified for the first time.②We analyzed the distribution of main virulence genes from V. parahaemolyticusstrains of different sources and regions, and considered that the tdh gene can be used as the marker for differentiating the virulent and avirulent strains.③We subtypedthe V. parahaemolyticus strains by using PFGE, MLST and MLVA, and90newallele numbers and48new STs were obtained, and we found the recombination ofrecA housekeeping gene. This result revealed the highf heterogeneity for V.parahaemolyticus.④V. parahaemolyticus showed considerable genetic diversity,and the environmental isolates had higher genetic diversity than clinical isolates.Phylogenetic analysis indicated that other serotypes such as O4:K68, O3:K25,O3:KUT and O1:KUT may have originated from O3:K6.⑤This study providessome data support for the correlation of V. parahaemolyticus infections andenvironmental or food contamination, and provides technique support for thesource-tracking, confirmation of new emergent serotypes, epidemiologicalinvestigation, genetic variation and population evolution of foodborne pathogens.