Epidemiology of Vibrio parahaemolyticus Outbreaks, Southern Chile

One-sentence summary for table of contents: Outbreaks are decreasing and the O3:K6 pandemic strain is being replaced by a new serotype and new strains.

Until 2006, ≈100% of the cases analyzed were caused by a clonal group originally observed in Southeast Asia in 1996 (4,5,7). This group was known as the V. parahaemolyticus pandemic strain because it had reached coastal environments worldwide and caused outbreaks. It belongs to the O3:K6 serovar, although at least 21 serovariants have emerged since 1996 (8). These serovariants also have specifi c sequences corresponding to genes such as toxRS/new (9), open reading frame 8 (orf8) (10), and tdh, but lack others such as trh, which is found in other pathogenic strains.
Genome sequencing of the RIMD2210633 pandemic strain showed that it has 2 sets of gene clusters that encode the type III secretion system (TTSS) apparatus (11). This apparatus is used by several gram-negative pathogenic bacteria to secrete and translocate virulence factor proteins into the cytosol of eukaryotic cells (12). TTSS1 is involved in cytotoxicity against HeLa cells, and TTSS2 is involved in enterotoxic activity in a rabbit ileal loop test (13). The fi rst cluster is located on the large chromosome, and the second is located on the small chromosome. The second cluster contains 2 copies of the tdh gene and is located on a pathogenicity island (a discrete genetic unit that contains genes responsible for pathogenicity and virulence) probably obtained by recent lateral transfer (11). TTSS2 has been found only in strains showing β-type hemolysis on a specialized blood agar medium known as Wagatsuma agar (11). This hemolysis is called the Kanagawa phenomenon and is considered a useful marker for identifi cation of pathogenic strains. Recently, TTSS genes related to the TTSS2 cluster were reported in clinical and environmental non-O1, non-O139 V. cholerae strains (14).
The clonal nature of the pandemic V. parahaemolyticus isolates was ascertained by the similarity of patterns obtained by genome restriction fragment length polymorphism-pulsed-fi eld gel electrophoresis (15), arbitrarily primed PCR (7,9), direct genome restriction enzyme analysis (DGREA) (4), and multilocus sequence typing (MLST) (16,17). However, the pandemic strain is a minor fraction of a diverse and shifting V. parahaemolyticus population found in shellfi sh in Puerto Montt (2). In an effort to understand the epidemiology of these outbreaks, we studied V. parahaemolyticus isolates obtained from human cases and shellfi sh during the summer of 2007. Our results indicate replacement of the O3:K6 pandemic strain by new serotype O3:K59 and new pathogenic V. parahaemolyticus groups. The decrease in the number of clinical cases may have been caused by diminution of the V. parahaemolyticus O3:K6 pandemic group in regional seafood.

Strains
V. parahaemolyticus RIMD 2210633 (VpKX) and RIMD 2210086 (VpI) were obtained from the Research Institute for Microbial Diseases, Osaka University, Osaka, Japan. The Chilean environmental strains, identifi ed as PMA with a number according to origin and year of isolation, were obtained from shellfi sh samples obtained during outbreaks from 2004 though 2007. Most strains have been described (4). PMC38.7, PMC60.7, PMC53.7, and PMC75.7 are isolates from clinical samples obtained in 2007. Each of these isolates corresponds to the isolate type of the 23 groups differentiated by DGREA as described (4) and reported in this article.
PCR products were purifi ed by using either the Wizard SV Gel or PCR Clean-Up Systems (Promega, Madison, WI, USA) and sequenced in both directions by Macrogen (Seoul, South Korea) or by McLAB (South San Francisco, CA, USA) by using forward and reverse amplifi cation primers or primers M13F and M13R (MLST loci). DNA sequences were analyzed individually and manually assembled. Alignments and sequence similarities were obtained by using BioEdit software (22). Sequences obtained were deposited in GenBank under accession nos. EU185060-EU185092.
Amplifi cation and sequencing of the variable region of the 16S rRNA (rrs) between nucleotides 357 and 518 (Escherichia coli numbering) were performed as described (23). This analysis consisted of separation of rrs alleles in PMC38.7 by pulsed-fi eld gel electrophoresis and PCR amplifi cation of the variable region in excised bands as described (24).

V. parahaemolyticus Associated with Human Cases in 2007
V. parahaemolyticus isolates from 37 human casepatients with diarrhea from the summer of 2007 in the Puerto Montt region were analyzed and grouped according to serotype, presence of genetic markers (orf8, toxRS/new, tlh, tdh, trh), and distinctiveness of their DGREA patterns (Table 1; Figure 1). One isolate from each patient was characterized. On the basis of genetic markers and DGREA pattern, isolates from 27 patients corresponded to the pandemic clonal group. However, 40% of the 20 serotyped strains of this group contained a K 59 capsular antigen instead of the characteristic K 6 antigen, and 25% cross-reacted with antisera for K 6 and K 59 antigens. Another difference from strains of previous years was that the relative number of cases associated with the pandemic strain (73%) was lower than the 100% observed in previous summers (4). Isolates from the other 27% of patients with clinical cases lacked the characteristic markers of the pandemic strains, i.e., orf8 and toxRS/new, and had 4 new DGREA groups. One of these groups contained 4 tdh-positive isolates (11% of cases). A second group contained 4 isolates positive for tdh and trh genes. The other 2 groups contained 1 isolate each, and both were negative for these 2 markers (Table 1). One isolate chosen as the type strain of each group was typed by MLST (17). MLST sequence type corresponded with groups determined by other analyzed genetic properties ( Table 1).

Characterization of Nonpandemic Strains
Strains positive for tdh, other than the pandemic strain, had not been isolated from patients with clinical cases in Puerto Montt during the 3 previous summer outbreaks. It has been reported that this gene may be spread by insertion sequence-like elements (25,26). The possibility that tdh found in nonpandemic strains was derived from the pandemic strain was explored. PCR amplicons of the tdh gene of the 2 nonpandemic groups was sequenced in isolates designated as type strains for each group: PMC60.7 for the group containing tdh and trh and PMC38.7 for the group containing only tdh. The amplicon of isolate PMC60.7 had an identical sequence to that reported for tdhA except for 1 nucleotide (11). Conversely, PMC38.7 had an identical sequence to that expected for a mixture of the 2 tdh genes in VpKX. These 2 genes differ slightly and the mixture of their PCR products should show polymorphisms in specifi c sites (11).
This observation suggested the presence of tdhS and tdhA genes in PMC38.7 with identical sequences to those found in the pandemic strain. Because these 2 genes are located close to each end of the pathogenicity island in chromosome 2 of the pandemic V. parahaemolyticus (11), the presence of the entire island was explored by PCR amplifi cation of 11 genes of TTSS2 located in the island and of genes VPA1321 and VPA1376 located at the extremes of the island near tdhA and tdhS, respectively. Each tested gene was found in the PMC38.7 strain, and sequences of their PCR products were identical to those reported for the pandemic strain genes, except for 1 nucleotide in VPA1342. Serotyping of PMC38.7 indicated an O10:K20 serovar. Because MC38.7 is genetically different from the pandemic strain (Table 1, Figure 1), the high degree of homology of these genes suggested that the entire pathogenicity island had recently been transferred from the pandemic strain. PMC38.7 also differs from the pandemic strain and most clinical isolates by the presence of intragenomic heterogeneity among its multiple 16S rRNA genes, a feature seldom observed in clinical isolates but frequently observed among environmental isolates. This fi nding is probably caused by lateral transfer of rrs (23). Three rrs genes, with sequences corresponding to V. parahaemolyticus groups VpD1-B4, ATA65-B2, and VpKX-AB (23), were observed in PMC38.7.

V. parahaemolyticus Associated with Shellfi sh
There are a large number of V. parahaemolyticus strains in the environment in the Puerto Montt region. Only the pandemic strain was isolated from clinical samples before and during the summer of 2006, but 20 different  63.7, 64.7, 65.7, 66.7, 20.7 Figure 2). This fi nding increases the number of different strains found in shellfi sh of the region to 23. Figure 3 shows the number of shellfi sh samples containing each of the 23 DGREA groups found during the past 4 years, including the summer of 2007 (2; this study). Only 4 of the 20 shellfi sh samples examined produced tdh-positive enrichment cultures, which is indicative of pathogenic strains. However, no tdhpositive isolates were obtained from these enrichments after plating on thiosulfate citrate bile salts sucrose agar.

T3SS2 Genes in Other Nonpandemic V. parahaemolyticus Strains
Because the presence of TTSS2 genes is not exclusive to the pandemic strain (20), their occurrence in the other environmental and clinical V. parahaemolyticus DGREA groups found in Puerto Montt was explored. We analyzed by PCR amplifi cation 20 strains corresponding to 18 environmental DGREA groups (PMA79, 112, 118, 189, 337, 339, 3316, 1.5, 19.5, 22.5, 27.5, 13.6, 34.6, 36.6, 40.6, 1.7, 11.7, and 21.7) and 1 strain from each of 3 new clinical groups (PMC60.7, 53.7, and 75.7) found in Puerto Montt in 2007 for the VPA1335 gene (found in T3SS2). Among these strains, only PMA339, isolated from shellfi sh in the summer of 2004, was positive. PCR amplifi cation of PMA339 for the other genes in the pathogenicity island showed positive reactions for all the genes tested in PMC38.7. Nevertheless, sequences of amplicons showed strong differences from those found in PMC38.7 and those reported for the pandemic strain genome. Similarity ranged from 99.4% for VPA1362 to 93.8% for VPA1346; the average for all genes tested was 97.7%. These differences indicate a much larger evolutionary distance between the T3SS2 genes in PMA339 and the pandemic strain than between PMC38.7 and the pandemic strain.

Discussion
The epidemiology of outbreaks caused by V. parahaemolyticus in the Puerto Montt region is changing. The number of clinical cases caused by the pandemic strain has decreased, accompanied by a change of serotype from O3:K6 to O3:K59. The changing serotype of the pandemic strain has been recently reviewed by Nair et al. (8). These authors showed that the more recent serotypes do not have the propensity for increasing hospital admissions observed with O3:K6, and some type of change in the epidemic process seems to be evident. Genes for the biosynthesis of capsular polysaccharides, which are major antigens (K) of V. parahaemolyticus, are probably encoded in a gene cluster characterized by variability that may occur through lateral gene transfer (27). Isolates reacting with antisera for K 6 and K 59 antigens may have changed part of the K epitopes reacting with the commercial polyclonal antiserum.  7, 5.7, 6.7, 7.7, 8.7,15.7, 16.7, 17.7, 19.7, 20.7, 22.7, 23.7, 24.7, 25.7, 26.7, 27.7, 28.7, 33.7, 37.7, 41.7, 42.7, 43.7, 47.7, 48.7 demic strains. Pathogenicity island genes identical to those in the pandemic strain in this bacterial group and differences in housekeeping genes and DGREA patterns are best explained by transfer of the pathogenicity island from the pandemic strain to an indigenous strain. The indigenous V. parahaemolyticus population in shellfi sh is diverse, and the predominant strains seem to change every year. A detailed examination of the putative genomic island in PMC38.7, its integration site, and its fl anking regions, will probably help differentiate among possible mechanisms of DNA transfer. The presence of TTSS2 genes is not exclusive of the pandemic and PMC38.7 strain; they were also found in an environmental isolate of V. parahaemolyticus (PMA339). However, PMA339 has not been observed among clinical isolates. TTSS2 genes have been found in other V. parahaemolyticus clinical strains (20); however, we identifi ed them in environmental strains. On the basis of sequences obtained from their amplicons, TTSS2 genes in PMA339 seem to have independently evolved from the pandemic strain over a considerable time. Nevertheless, they are still more closely related to the TTSS2 of V. parahaemolyticus than to those recently found in non-O1 or non-O139 V. cholerae (14).
Little is known of the origin of the other 3 clinical groups (60.7, 1.5, and 75.7). Group 60.7 contains tdh and trh, but this tdh does not seem to be derived from the pandemic strain. Strains of groups 1.5 and 75.7 lack both pathogenicity-associated genes, but fi nding these isolates in patients is not unusual (28). PMC75.7, a clinical strain, contains a recA gene that is closely related to that of PMA339, the environmental isolate containing TTSS2 genes (17 and http://pubmlst.org/vparahaemolyticus). However, in view of a recent report (29), one should consider that the 2 human isolates lacking tdh or trh genes may correspond to nonvirulent strains that proliferate during infection with a virulent strain.