Genetic Analysis of Vibrio parahaemolyticus O3:K6 Strains That Have Been Isolated in Mexico Since 1998.

Vibrio parahaemolyticus is an important human pathogen that has been isolated worldwide from clinical cases, most of which have been associated with seafood consumption. Environmental and clinical toxigenic strains of V. parahaemolyticus that were isolated in Mexico from 1998 to 2012, including those from the only outbreak that has been reported in this country, were characterized genetically to assess the presence of the O3:K6 pandemic clone, and their genetic relationship to strains that are related to the pandemic clonal complex (CC3). Pathogenic tdh+ and tdh+/trh+ strains were analyzed by pulsed-field gel electrophoresis (PFGE) and multilocus sequence typing (MLST). Also, the entire genome of a Mexican O3:K6 strain was sequenced. Most of the strains were tdh/ORF8-positive and corresponded to the O3:K6 serotype. By PFGE and MLST, there was very close genetic relationship between ORF8/O3:K6 strains, and very high genetic diversities from non-pandemic strains. The genetic relationship is very close among O3:K6 strains that were isolated in Mexico and sequences that were available for strains in the CC3, based on the PubMLST database. The whole-genome sequence of CICESE-170 strain had high similarity with that of the reference RIMD 2210633 strain, and harbored 7 pathogenicity islands, including the 4 that denote O3:K6 pandemic strains. These results indicate that pandemic strains that have been isolated in Mexico show very close genetic relationship among them and with those isolated worldwide.

Introduction

Vibrio parahaemolyticus is a bacterium that inhabits warm marine environments and are able to causes gastroenteritis, wound infections and septicemia, that are associated with the ingestion of raw seafood, primarily filter-feeding bivalves mollusks [1]. The pathogenicity of this bacterium has been correlated with several virulent factors, including thermostable direct hemolysin (TDH), TDH-related hemolysin (TRH) and the type III secretion system (T3SS) [2-5]. V. parahaemolyticus strains that contain tdh or trh, are considered to be pathogenic [6].

The first clinical cases of V. parahaemolyticus were reported in Japan in 1950 [7], although in 1996, an atypically rise in infections occurred in India due to tdh+/O3:K6 isolates [8], spreading to southeast Asia in just a few months [9]. Subsequently, this clone has been associated with most V. parahaemolyticus clinical cases worldwide [10-12]. Genetic studies have demonstrated that all the post-1996 O3:K6 strains, are identical but differ significantly from earlier isolates, also known as old-O3:K6 (tdh/trh+) strains [13]. Post-1996 tdh/O3:K6 strains have the same MLST sequence type (ST-3) and are considered to be the ancestors of the pandemic serovariants, which now constitute the pandemic clonal complex (CC3) [14].

In the Americas, the pandemic strain O3:K6 was first isolated from a clinical case in Peru in 1996 [15], after which outbreaks of this strain were reported in Peru in 1998 [15]; Chile in 1997,1998, 2004 and 2005 [16]; the US in 1998 [17] and Brazil in 2002 [18]. In Mexico, 1 localized outbreak (2003 to 2004) of V. parahaemolyticus O3:K6 strains was reported on its northwest Pacific coast, where more than 1200 clinical cases were registered after ingestion of raw shrimps [19]. Velazquez-Roman et al. [20], isolated the pandemic clone O3:K6 from clinical and environmental sources, in the northwest state of Sinaloa, Mexico from 2004 to 2010. Recently, Hernández-Díaz et al. [21] indicated that the pandemic clone O3:K6 has established itself endemically in the northwest Pacific coast of Mexico.

The surveillance programs that have been implemented by the Mexican Diagnostics and Epidemiological Reference Institute (InDRE), since 1991 focus on pathogenic vibrios, especially toxigenic V. cholerae [22]. Nevertheless, suspicious strains of V. parahaemolyticus have been isolated and maintained in the InDRE strain collection and donated to CICESE for future studies. These strains, those isolated in 2004 during the outbreak in Mexico, and environmental strains that have been isolated from shrimps, oysters and biofouling of ships hulls, were examined in this study (Table 1).

Table 1

Vibrio parahaemolyticus strains from CICESE and CAIM culture collections.

Strain	Collection origin	Year	Source	Location	tdh	trh	orf8	Serotype
CAIM 728*	T3979	1951	Clinical	Jap	+	-	-	O5:K12
CAIM 729T*	TX2103T	1998	Clinical	Tex	+	-	+	O3:K6
CAIM 1400T*	CAIM	2004	Clinical	Sin	+	-	+	O3:K6
CAIM 1435*	CAIM	2004	Clinical	Sin	+	-	-	NA
CAIM 1474*	CAIM	2004	Clinical	Sin	+	-	+	O3:K6
CAIM 1477*	CAIM	2004	Clinical	Sin	+	-	+	O3:K6
CAIM 1490*	CAIM	2004	Clinical	Sin	+	-	+	O3:K6
CAIM 1693*	CAIM	2004	Clinical	Sin	+	-	+	O3:K6
CAIM 1772	CAIM	2005	Shrimp	Sin	+	+	-	O5:K17
CICESE-170*	InDRE	1998	Clinical	Hgo	+	-	+	O3:K6
CICESE-171*	InDRE	1998	Clinical	Hgo	+	-	+	O3:K6
CICESE-172*	InDRE	1998	Clinical	CDMX	+	-	-	NA
CICESE-173*	InDRE	1998	Clinical	Hgo	+	-	+	O3:K6
CICESE-174*	InDRE	1998	Clinical	CDMX	+	-	+	O3:K6
CICESE-175*	InDRE	1998	Clinical	Hgo	+	-	+	O3:K6
CICESE-176*	InDRE	1998	Clinical	Hgo	+	-	+	O3:K6
CICESE-177*	InDRE	1998	Clinical	Mich	+	-	+	O3:K6
CICESE-178*	InDRE	1998	Clinical	Hgo	+	-	+	O3:K6
CICESE-179*	InDRE	1998	Clinical	Hgo	+	-	+	O3:K6
CICESE-180*	InDRE	1998	Clinical	Hgo	+	-	+	O3:K6
CICESE-181*	InDRE	1998	Clinical	Hgo	+	-	+	O3:K6
CICESE-182*	InDRE	1998	Clinical	Hgo	+	-	+	O3:K6
CICESE-183*	InDRE	1998	Clinical	Hgo	+	-	+	O3:K6
CICESE-184*	InDRE	1998	Clinical	Hgo	+	-	+	O3:K6
CICESE-185*	InDRE	1998	Clinical	Hgo	+	-	+	O3:K6
CICESE-186*	InDRE	1999	Clinical	Hgo	+	-	+	O3:K6
CICESE-187*	InDRE	2000	Clinical	Tamps	+	-	+	O3:K6
CICESE-188*	InDRE	2009	Clinical	NL	+	-	+	O3:K6
CICESE-250*	CICESE	2012	Biofouling	BC	+	-	-	NA
CICESE-251*	CICESE	2012	Biofouling	BC	+	-	-	NA
CICESE-253	CICESE	2012	Biofouling	BC	+	-	-	NA
CICESE-254	CICESE	2012	Biofouling	BC	+	-	-	NA
CICESE-273*	CICESE	2012	Biofouling	BC	+	-	+	O3:K6
CICESE-274	CICESE	2012	Biofouling	BC	+	-	-	NA
CICESE-374*	CICESE	2011	Oyster	CDMX	+	+	-	NA
CICESE-375*	CICESE	2011	Oyster	CDMX	+	+	-	NA
CICESE-376	CICESE	2011	Oyster	CDMX	+	+	-	NA
CICESE-390	CICESE	2011	Oyster	CDMX	+	+	-	NA

(Jap) Japan, (Tex) Texas, USA), Mexican states: Sin (Sinaloa), Hgo (Hidalgo) CDMX (Ciudad de México), Mich (Michoacán), Tamps (Tamaulipas), NL (Nuevo León).

(*) Strains selected for MLST analysis. NA: not analyzed.

The aim of this study was to evaluate the presence of pandemic V. parahaemolyticus O3:K6 strains in Mexico, isolated from the environment and clinical cases, from 1998 to 2012 and determine their similarity with O3:K6 strains that have been isolated worldwide. Our results confirm the presence of the pandemic strain since 1998.

Material and Methods

Bacterial isolates

Thirty-eighth toxigenic V. parahaemolyticus strains were selected. Nineteen clinical strains were isolated by the InDRE from stool samples in hospitals from various Mexican states and years. Nine strains were obtained from the Collection of Aquatic Important Microorganisms (www.ciad.mx/caim), including strain CAIM 729 (TX2103T), which was isolated during the 1998 Texas (USA) outbreak [23,24], and strain CAIM 728, isolated in Japan. Most CAIM strains were isolated in 2004 during the outbreak in Mexico [19]. Four strains were isolated from oysters (Crassostrea spp.) that were collected at the La Nueva Viga seafood market in Mexico City in 2011, and 6 strains were isolated from the biofouling of commercial ships hulls in 2012 [25]. Strains that were isolated from oysters were examined per the Bacteriological Analytical Manual [26]. Strains were identified as V. parahaemolyticus, using the species-specific molecular markers tlh [27] and pR72H [28], homolysins were detected with tdh and trh specific primers [27]. Only those strains that were positive for the tdh and/or trh hemolysin genes were included (Table 1).

DNA extraction and PCR amplification

Genomic DNA from overnight Luria-Bertani (LB) broth cultures of suspicious V. parahaemolyticus strains was purified with the WizardTM Genomic Kit (Promega) per the manufacturer. Purified DNA was diluted in RNAse/DNAse-free water (~50 ng μl-1). The strains were tested by PCR for presence of the pR72H, tlh, tdh, trh and ORF8 genes. The PCR was performed using the conditions and primers as reported in the literature (Table 2). The PCR products were amplified with GoTaqTM DNA polymerase (Promega), and amplicons were resolved by gel electrophoresis (1.5% agarose) in 1X TBE (45 mM Tris, 45 mM boric acid, 10 mM EDTA, pH 8.0) and visualized under UV light. The strains CAIM 1400T (tdh+/ORF8+) and CAIM 1772 (tdh+/trh+) were used as controls for the PCR analysis.

Table 2

Primers used for the molecular characterization of V. parahaemolyticus.

Gene	Primer	Amplicon	Sequence size	Reference
tlh	tlh-F AAAGCGGATTATGCAGAAGCACTG	450	-	27
	tlh-R GCTACTTTCTAGCATTTTCTCTGC
pR72H	VP33-F TGCGAATTCGATAGGGTGTTAACC	387	-	28
	VP32-R CGAATCCTTGAACATACGCAGC
trh	trh-F TTGGCTTCGATATTTTCAGTATCT	500	-	27
	trh-R CATAACAAACATATGCCCATTTCCG
tdh	tdh-F GTAAAGGTCTCTGACTTTTGGAC	269	-	27
	tdh-R TGGAATAGAACCTTCATCTTCACC
ORF8	O3MM824-F AGGACGCAGTTACGCTTGATG	369	-	24
	O3MM1192-R CTAACGCATTGTCCCTTTGTAG
	MLST
	CHROMOSOME I
dnaE	dnaE-F tgtaaaacgacggccagtCGRATMACCGCTTTCGCCG	596	557	PubMLST
	dnaE-R caggaaacagctatgaccGAKATGTGTGAGCTGTTTGC
gyrB	gyrB-F tgtaaaacgacggccagtGAAGGBGGTATTCAAGC	629	592	PubMLST
	gyrB-R caggaaacagctatgaccGAGTCACCCTCCACWATGTA
recA	recA-F CGCCGCTGCGCTAGGTCAAA	773	731	This study
	recA-R CGCGATTTCTGGATGCTCACGC
	CHROMOSOME II
dtdS	dtdS-F tgtaaaacgacggccagtTGGCCATAACGACATTCTGA	497	458	PubMLST
	dtdS-R caggaaacagctatgaccGAGCACCAACGTGTTTAGC
pntA	pntA-F- GGTGCGGAGTTCCTGACCGTG	470	430	This study
	pntA-R AACGCAGCAGGTGCTACCGA
pyrC	pyrC-F tgtaaaacgacggccagtAGCAACCGGTAAAATTGTCG	533	493	PubMLST
	pyrC-R caggaaacagctatgaccCAGTGTAAGAACCGGCACAA
tnaA	tnaA-F tgtaaaacgacggccagtTGTACGAAATTGCCACCAAA	463	423	PubMLST
	tnaA-R caggaaacagctatgaccAATATTTTCGCCGCATCAAC

Serotyping

A commercial V. parahaemolyticus antiserum kit (Denka Seiken, Tokyo, Japan), was used for serological typing per the manufacturer’s instructions (www.abcam.com/technical). CAIM 1400T was used as a positive control for O3:K6, and CAIM 1772 (O5:K17), was included as a negative control.

PFGE analysis

Thirty-eight V. parahaemolyticus and 2 V. cholerae strains (Table 1), were analyzed by PFGE with the standardized CDC protocol for V. cholerae [29], with minor modifications. Briefly, genomic DNA was digested with 20 U of Not-I (New England BioLabs), and restriction fragments were resolved on a CHEF MapperTM PFGE system (BioRad). The run time was divided into 2 blocks, 13 h (2–10-s pulse time) and 12 h (20–25-s pulse time) at 6 V cm-1 with an angle of 120°. Lambda ladder for PFGE (New England BioLabs) was used as the molecular size marker. Thiourea (50 μM) was added during the electrophoresis. Gels were stained with ethidium bromide and visualized under UV light. Images of the gels were taken after electrophoresis and were used to generate a dendrogram with GelCompar II (Applied Maths, Kortrijk, Belgium) by Dice coefficient-unweighted pair group method with arithmetic averages (UPGMAs).

MLST analysis

Thirty-two strains that were classified as tdh and tdh/trh were selected from the PFGE clusters (Table 1) for analysis by MLST. Seven loci from both chromosomes (Table 2) were chosen; 5 loci were amplified using primers in the PubMLST database for V. parahaemolyticus (http://pubmlst.org/vparahaemolyticus/) by PCR per González-Escalona et al. [14]. Each locus, observed as a single band after electrophoresis, was purified with the AxyPrep-PCR Kit (AxygenTM) and sequenced by SeqXcel Inc. (CA, USA) using the M13 universal primers (Forward-TGTAAAACGACGGCCAGT and Reverse-CAGGAAACAGCTATGACC, 5’-3’). For loci recA and pntA, new primers were designed (Table 2), and used to amplify fragments as follows: 10 min at 96°C for initial denaturalization, followed by 35 cycles (1 min at 96°C, 1 min at 59°C, and 1 min at 72°C), 10 min at 72°C for the final extension, and maintenance at 4°C. The amplicons were then sequenced using the designed primers for recA and pntA (Table 2).

The sequences for each locus, were queried against the V. parahaemolyticus PubMLST database, to determine the allelic profile (AP) and sequence type (ST) (S1 Table). Novel alleles were submitted to PubMLST; the nucleotide sequences data that were reported for the MLST are available in the GenBank database under accession numbers KP455743-KP455966.

Haplotype diversity (Hd), pairwise nucleotide diversity (π) genetic variability (θ) and number of polymorphic sites (PSs), were determined for individual loci (S2 Table) using DnaSP, version 5.10.1 [30]. A neighbor-joining tree was constructed with concatenated sequences of these loci, with a final sequence of 3682 bp (in the following order: dnaE, gyrB, recA, dtdS, pntA, pyrC and tnaA), using Mega 5.2.2. [31], Kimura’s 2-parameter model and bootstrapping methods (1000 replicates).

The APs that were generated by MLST, were used to assign the strains to a clonal complex (S1 Table) with goeBURST, V1.2.1 (www.phyloviz.net/goeburst, [32]. The APs of the 32 strains were compared using Kimura’s 2-parameter model, against the database for V. parahaemolyticus strains in PubMLST, including within the CC3 [14, 33].

Genomic analysis

Genomic DNA of an overnight LB culture of a single colony of the O3:K6 V. parahaemolyticus strain (CICESE-170), isolated in 1998, was extracted using the WizardTM Genomic kit (Promega). The quantity and quality of the DNA were determined on a NanoDropTM 2000 (Thermo Scientific, Wilmington, DE), and the DNA was diluted to ~1000 ng/μl in DNAse/RNAse-free water. Whole-genome sequencing was performed using Illumina Myseq™ genome analyzer (Illumina Inc., USA) per the manufacturer’s instructions.

The reads that were obtained from the CICESE-170 strain were assembled into contigs using CAP3 [34] and Vague 1.0.5 [35]. The synteny of contigs was obtained with Mauve 2.3.1 [36], using both chromosomes of the V. parahaemolyticus strain RIMD 2210633 (GenBank accession number: BA000031 and BA000032) as the reference genome [37].

Comparative analyses were performed between the whole-genome sequencing results on CICESE-170 and RIMD 2210633 as the reference strain. Contigs that were obtained from CICESE-170 were submitted and inspected with Rapid Annotation using Subsystem Technology (RAST) server (http://rast.nmpdr.org), and comparisons were made, based on the functions and sequences of both strains. The analyses were evaluated in the Seed Viewer, focusing on virulence, disease, defense, phages, prophages, transposables elements, plasmids, iron acquisition and stress response. Contigs from CICESE-170 were also inspected by alignment using Geneious 4.8 [38] and BLAST analysis to compare them with the reference strain. Whole-genome sequences of CICESE-170 were submitted to GenBank under accession number MIEM01000000.

Results

Species identification

In this study, the same V. parahaemolyticus strains were identified at the species level using the tlh or the pR72H primers; thus, both sets of primers could be used to determine species. Of the InDRE V. parahaemolyticus collection, 19 isolates were tdh+/trh and 18 were ORF8+. The 6 environmental strains that were isolated from biofouling were tdh+/trh, 1 of which was confirmed to be ORF8+ (CICESE-273). The 4 isolates from the oyster samples were tdh+/trh+. From the CAIM collection, 8 strains were tdh+/trh, 1 was tdh+/trh and 6 were ORF8+ as reported (Table 1).

Twenty-five ORF8+ strains were positive for the O3 and K6 antisera, including those that have been isolated by InDRE since 1998. The strains that were isolated during the 2004 outbreak in Mexico (CAIM collection) and 1 environmental strain from biofouling of a ship hull from 2012 (Table 1) were also ORF8+ and O3:K6+.

Restriction fragments that were over 48.5 kb were examined, generating 2 branches (Fig 1A). Branch I contained all V. parahaemolyticus strains and branch II contained the out-group of V. cholerae strains. V. parahaemolyticus strains had high diversity, 31 patterns were obtained from 38 strains. At 50% of similarity, branch I split into 8 clusters (A to H), according to their genetic and serological characteristics (tdh/trh, tdh or tdh/ORF8/O3:K6). The lowest similarity among the 8 clusters was <24%.

Fig 1

Dendrograms of CICESE and CAIM V. parahaemolyticus O3:K6 strains for: a) the neighbor-joining tree (UPGMA) by PFGE analysis after digestion with NotI and b) the neighbor-joining tree (Kimura’s 2 parameters) by MLST of the concatenated sequences of 7 loci.

The twenty-five O3:K6 strains, formed a single cluster (E), with >50% similarity and 21 PFGE patterns. Two clinical strains that were isolated in 1998 (CICESE-171 and CICESE-177) showed >88% and >93% of similarity with CAIM 729T and CAIM 1400T, strains that were isolated during the Texas (USA) and Sinaloa (northwestern Mexico) outbreaks, respectively. Cluster E had low similarity (<37%) with the non-pandemic strains (clusters A, B, C, D, F, G, and H). Cluster A (n = 4) presented 2 PFGE patterns at >94% similarity; these strains (tdh+/trh-) were isolated from environmental samples (biofouling). Clusters B, D, G, and H contained 1 strain (tdh+ trh-), with less than a 29% similarity between clusters. Clusters C (n = 1) and F (n = 4) typified as tdh+/trh+, were isolated from environmental samples (shrimp and oysters).

The concatenated sequences of 7 loci that were analyzed from 32 selected strains of V. parahaemolyticus based on PFGE analysis, were separated them into 3 clusters according to their genetic and serological characteristics (Fig 1B). Cluster I comprised all the pandemic strains (tdh/ORF8/O3:K6), cluster II contained tdh/trh strains, and cluster III was composed of tdh isolates. Cluster I included the strains that have been isolated by InDRE since 1998, those during the Mexican outbreak, a strain from the biofouling of a ship hull in provenance from Fukuyama, Japan (CICESE-273), and the pandemic strain (CAIM 729T), that was isolated during the Texas outbreak. With the exception of the CICESE-185 strain, all ORF8/O3:K6 strains in cluster I had 100% similarity.

The 7 loci generated 43 alleles. Most loci had 6 alleles; the exception was gyrB, which had 7 (S1 Table). The dnaE-3, recA-19, dtdS-4, pntA-29, pyrC-4, and tnaA-22 alleles were present in 25 strains, and gyrB-4 was presented in 24. Three loci had novel alleles due to the presence of a 1–2 bp difference and were assigned in the PubMLST database as gyrB-415, recA-295, and tnaA-231 (S1 Table).

The haplotype diversity (Hd) was 0.391 for most of the alleles; gyrB had an Hd of 0.44. The highest value for nucleotide diversity was observed for recA (π = 0.01054), whereas pyrC had the lowest value (π = 0.00378). The recA locus had the highest genetic variability (θ = 0.01345); dnaE had the lowest (θ = 0.00535). The number of polymorphic sites varied per locus, ranging from 11 for dnaE to 34 for recA (S2 Table).

The 5 novel STs that we obtained were submitted to the PubMLST database, assigned as ST-1137 (CAIM 1435), ST-1138 (CICESE-172), ST-1139 (CICESE-185), ST-1140 (CICESE-250 and 251), and ST-1141 (CICESE-374 and 375). All ORF8/O3:K6 strains contained ST-3, with the exception of strain CICESE-185, which harbored ST-1139; these strains were assigned to CC3 (S1 Table). The sequences that were obtained with the recA and pntA primers that we designed, had the same alleles (19 and 29) for all the O3:K6 strains, including the reference strain CAIM 729 (TX22103T), as previously reported with different primers [14].

The results on the 7 loci of concatenated sequences for 32 CICESE strains and 25 strains from pubMLST, are shown in Fig 2. Thirty-one strains from CICESE/CAIM had 100% similarity with the 9 strains from pubMLST that have been isolated worldwide, reported as O3:K6. Non-O3:K6 CICESE/CAIM and reference strains were grouped separately (Fig 2).

Fig 2

Assignment of CICESE/CAIM isolates and reference pubMLST sequences of O3:K6 and non-O3:K6 V. parahaemolyticus strains to a clonal complex.

Whole-genome analysis

Whole-genome sequencing of the CICESE-170 strain generated 1,970,230 (2 x 150 bp) paired-end reads that were assembled into 695 contings (N50 = 12,542 bp) at 52x coverage. The genome was 5,029,544 bp with a GC content of 45.3%. Analysis with the RAST server identified 4513 coding sequences (CDSs), 49 predicted RNAs, and 143 possibly missing genes. These CDSs were distributed into 26 categories, 104 subcategories, and 528 subsystems (RAST Job: 304198). The categories included virulence, disease, and defense (94 CDSs); iron acquisition system and metabolism (65 CDSs); stress response (173 CDSs); and phages, prophages (subsystem) (5 CDSs).

A comparative analysis of concatenated contigs showed that the CICESE-170 and RIMD 2210663 strains are >99.3% similar (Fig 3).

Fig 3

Whole-genome comparison of CICESE-170 and RIMD 2210633, implemented in BlastDotPlot.

The CICESE-170 strain contained 7 pathogenic islands (VPals), commonly described for V. parahaemolyticus O3:K6 pandemic strains (Table 3); 5 VPals (VPal-1 to VPal-5) lay in contigs that were associated with chromosome I, and 2 VPals (VPal-6 and VPal-7) resided in contigs of chromosome II. Most VPals (VPal-1 to -6) had higher that 99.7% similarity with RIMD 2210633 and had a GC content of 38.3% to 44.3%. VPal-7 contained the virulence factors, T3SS2 and the tdh gene, with 96.9% similarity to the reference sequence. Other elements in CICESE-170, commonly linked to the pandemic group, were the O- and K-antigen genes (VP0190-VP0238) and the phage f237 (VP1549-VP1562) with its corresponding ORF8 (VP1561). These elements were 100% similar to the reference sequence.

Table 3

Comparative results between CICESE-170 and RIMD 2210633.

	RIMD 2210633	CICESE-170
Elements	*Location	CDS	bp	%GC	*similarity %
Vpal-1	VP0380-VP0403	24	22,380	41.5	100.0
Vpal-2	VP0635-VP0643	9	8286	44.3	99.9
Vpal-3	VP1071-VP1095	24	30,913	42.2	99.9
Vpal-4	VP2131-VP2144	14	16,160	38.9	99.9
Vpal-5	VP2900-VP2910	11	15,134	38.3	99.9
Vpal-6	VPA1253-VPA1270	18	26,750	43.4	99.7
Vpal-7 (T3SS2)	VPA1310-VPA1398	87	75,384	38.6	96.9
Phage f237	VP1549-1590	15	8,755	45.1	100
O3 and K6	VP0190-VP0238	49	51,483	40.4	100

* Based on results from V. parahaemolyticus RIMD 2210633.

Discussion

Our results indicate that V. parahaemolyticus O3:K6 (tdh+/ORF8+) pandemic strains have been linked to clinical cases since 1998 in 6 Mexican states, particularly in the northwestern state of Sinaloa, most of which failed to develop into an outbreak [20]. Watery diarrhea, which is the most common symptom of V. parahaemolyticus infection, is self-limiting resolving in 3 days; thus, it is possible that most of O3:K6 infections were not reported or detected by the Mexican health system. Further, these reports might have been masked by the Mexican cholera epidemic from 1991 to 2000 [22].

In this study, by PFGE analysis, all strains with the characteristics of pandemic strains (tdh+/ORF8+/O3:K6) were grouped into a single cluster (E), with 21 patterns (Fig 1A). Wong et al. [9] and Yeung et al. [39] also assembled the pandemic strains into a single cluster with fewer PFGE patterns (8 and 7, respectively) between their O3:K6 isolates. Even if PFGE analysis reveals relative low similarity between pandemic strains, they are clearly separated from non-O3:K6 strains. PFGE results contrast with the high similarity among O3:K6 strains that is observed with other molecular approaches, such as MLST [14] and toxRS sequencing [40]. PFGE has been used frequently to discriminate between strains from various regions [41], but no such differentiation was noted in our study.

By MLST, using 7 housekeeping genes, Mexican V. parahaemolyticus isolates showed high genetic similarity between them. Most of the tdh/ORF8/O3:K6 strains isolated in Mexico, contained ST-3 and were thus associated with reference strains of the pandemic clone CC3 (Fig 2; S1 Table), which have been isolated worldwide since 1996, as did the reference strain CAIM 729 (TX22103T) [14, 33].

Notably, in 1998, seafood-associated clinical cases of V. parahaemolyticus O3:K6 were reported in Peru, Chile, and the US [15-17], the same year in which most InDRE strains were isolated. Thus, it appears that the O3:K6 pandemic clone encountered favorable conditions for its dispersion in these countries, including Mexico, likely due to the strong El Niño event that was registered in 1997–1998, as suggested [42].

Whole-genome sequencing of the CICESE-170 strain indicated a high genetic similarity with the reference sequence of RIMD 2210663 (>99.3%). Both strains shared mobile genetic elements, such as VPals (Table 3). RAST server analysis showed that CICESE-170 contained 4513 CDSs, 319 fewer than RIMD 2210663 (4832 CDSs). These differences might be associated with the adaptation to local environmental conditions, because similar disparities have been reported among O3:K6 strains from various countries [43].

All O3:K6 strains were positive for the ORF8 region by PCR, a region that is present in the phage f237 and was detected in CICESE-170, showing 100% similarity compared with RIMD 2210633 (Table 3). This phage has been used as a molecular marker due to its high specificity for the pandemic group [44] and according to Nasu et al. [45], it is involved in the epidemic potential of V. parahaemolyticus O3:K6.

Whole-genome sequencing showed that CICESE-170 contained 7 VPals that are common to V. parahaemolyticus O3:K6, most of which had high genetic similarity with the reference pandemic strain RIMD 2210633 (Table 3). The 7 VPals in CICESE-170 had a low GC content (38.3% to 44.3%), versus those of the overall genome (45.3%) and an earlier report (45.4%) [37, 46].

The four VPals that characterize the post-1996 pandemic strains were also detected in CICESE-170, confirming that this strain, with its high similarity to other Mexican O3:K6 isolates by MLST, is associated with pandemic V. parahaemolyticus O3:K6, that has been isolated worldwide [43]. VPals 1, 4, 5 and 6, which have been identified as a distinctive characteristic of O3:K6 pandemic strains and their serovariants, might have been acquired by horizontal gene transfer and might provide the pandemic strains with the ability to infect humans or to adapt to several environments [46]. VPal-7, which harbors the main pathogenic elements (tdh and T3SS2) of V. parahaemolyticus, was also detected in CICESE-170, but it had the lowest genetic similarity (96.9%) to RIMD 2210633.

Conclusions

The V. parahaemolyticus O3:K6 strains that were isolated in Mexico, were grouped into a single cluster by PFGE and MLST (Fig 1A and 1B). They were associated with O3:K6 sequences in the PubMLST database and were related to CC3 (Fig 2). Bioinformatics analysis of the CICESE-170 genome, demonstrated high genetic similarity with the reference sequence of the RIMD 2210633 strain (Fig 3) and the characteristic elements of pandemic O3:K6 strains. These findings show that V. parahaemolyticus O3:K6 pandemic strains, have been detected in Mexico since 1998 and confirm their persistence in this country from 1998 to 2012, probably without undergoing significant genetic changes. Whole genome analysis of additional Mexican strains need to be performed to confirm this statement.

Allelic profiles of 7 loci by MLST analysis.

ST (sequence type), CC (clonal complex), D (double) and S (singleton).

(DOCX)

Click here for additional data file.

Statistics of the 7 loci employed in the MLST analysis of V. parahaemolyticus.

Hd (haplotype diversity), π (pairwise nucleotide diversity), Θ (genetic variability), PSs (polymorphic sites).

(DOCX)

Click here for additional data file.