Presence of genes for type III secretion system 2 in Vibrio mimicus strains

Background Vibrios, which include more than 100 species, are ubiquitous in marine and estuarine environments, and several of them e.g. Vibrio cholerae, V. parahaemolyticus, V. vulnificus and V. mimicus, are pathogens for humans. Pathogenic V. parahaemolyticus strains possess two sets of genes for type III secretion system (T3SS), T3SS1 and T3SS2. The latter are critical for virulence of the organism and be classified into two distinct phylogroups, T3SS2α and T3SS2β, which are reportedly also found in pathogenic V. cholerae non-O1/non-O139 serogroup strains. However, whether T3SS2-related genes are present in other Vibrio species remains unclear. Results We therefore examined the distribution of the genes for T3SS2 in vibrios other than V. parahaemolyticus by using a PCR assay targeting both T3SS2α and T3SS2β genes. Among the 32 Vibrio species tested in our study, several T3SS2-related genes were detected in three species, V. cholerae, V. mimicus and V. hollisae, and most of the essential genes for type III secretion were present in T3SS2-positive V. cholerae and V. mimicus strains. Moreover, both V. mimicus strains possessing T3SS2α and T3SS2β were identified. The gene organization of the T3SS2 gene clusters in V. mimicus strains was fundamentally similar to that of V. parahaemolyticus and V. cholerae in both T3SS2α- and T3SS2β-possessing strains. Conclusions This study is the first reported evidence of the presence of T3SS2 gene clusters in V. mimicus strains. This finding thus provides a new insight into the pathogenicity of the V. mimicus species.


Background
The type III secretion system (T3SS) is possessed by gram-negative bacteria, especially those occurring in animal and plant pathogens, e.g. Yersinia, Shigella, Salmonella, Pseudomonas and Escherichia species [1][2][3]. The T3SS secretes and translocates effector proteins into the cytosol of eukaryotic cells, thus contributing to bacterial virulence against the host [1]. While the T3SS apparatus is well conserved in these bacteria, the specific properties of the effectors which are secreted via T3SS and symptomatic effects caused by the effectors on the host organism vary widely [1].
Vibrios are gram-negative γ-proteobacteria which are ubiquitous in marine and estuarine environments [4,5]. Several of the more than 100 Vibrio species are pathogens for fish, shellfish, coral, and mammals [6], and Vibrio parahaemolyticus was the first species in which the presence of T3SS was reported [7].
V. parahaemolyticus is a cause of food-borne gastroenteritis in humans, and almost all strains isolated from diarrheal patients produce the thermostable direct hemolysin (TDH) and/or the TDH-related hemolysin (TRH), which are encoded by the tdh and trh genes, respectively [8][9][10]. V. parahaemolyticus strains, which exhibit the Kanagawa phenomenon (KP), a beta-hemolysis detectable on a special blood agar (Wagatsuma agar) [11], possess two tdh genes, tdhA and tdhS, but not the trh gene [10,12,13]. In contrast, KP-negative clinical V. parahaemolyticus strains possess the trh gene only or both the trh and tdh genes. Genome sequencing of the KP-positive V. parahaemolyticus strain RIMD2210633 demonstrated that it possesses two sets of the genes for T3SS on chromosomes 1 and 2 (T3SS1 and T3SS2, respectively) [7]. It has further been demonstrated that T3SS2 is involved in enterotoxicity of the organism, and is considered to be an important factor in the pathogenicity of diarrheal illness [14]. The T3SS2 genes are located on a pathogenicity island (PAI) known as Vp-PAI on chromosome 2, and the genes were found in KP-positive, but not in KP-negative strains [7,14,15].
After the discovery of the T3SS genes in V. parahaemolyticus, other vibrios such as V. alginolyticus, V. harveyi, V. tubiashii and V. cholerae were also found to possess the genes for T3SS [14,[16][17][18]. While the T3SSs of V. alginolyticus, V. harveyi and V. tubiashii, are more closely related to T3SS1 of V. parahaemolyticus [14], that of V. cholerae is similar to T3SS2 of V. parahaemolyticus [17]. In addition, several studies have demonstrated that some V. cholerae non-O1/non-O139 serogroup strains, which do not possess the cholera toxin gene, do possess a set of T3SS genes in a PAI (VPI-2) on their chromosome [17,19]. It has further been suggested that the T3SS of non-O1/non-O139 V. cholerae is also involved in the pathogenicity of the bacterium [17].
In our most recently reported study, we used the sequencing and PCR assay of the genomic DNA of the TH3996 strain to detect the presence of a novel PAI (Vp-PAI TH3996 ) in trh-positive (KP-negative) V. parahaemolyticus strains [20]. The Vp-PAI TH3996 was found to contain a set of genes for T3SS, and the T3SS of the TH3996 strain to be essential for the enterotoxicity of this strain [20]. Phylogenetic analysis indicated that the T3SS genes of TH3996 are related to that of RIMD2210633, but belong to distinct lineage, with the former known as T3SS2β and the latter as T3SS2α [20]. Subsequent studies showed that T3SS2α and T3SS2β are present in, respectively, KP-positive and trh-positive V. parahaemolyticus strains and are also distributed among pathogenic V. cholerae non-O1/non-O139 serogroup strains [20].
A previous study examined the distribution of the T3SS2-related genes in Vibrio species, but tested only for the presence of the T3SS2α genes and in a limited number of strains from each species [14]. In this study, we re-investigated the distribution of the genes for T3SS2 in various Vibrio species and targeted both the T3SS2α and T3SS2β genes.
The assay demonstrated that, in addition to part of the V. cholerae strains, as previously reported, amplicons of the expected size of at least several of the T3SS2 genes were obtained from all of the V. hollisae strains and some of the V. mimicus strains. However, none of the genes tested in any of the remaining 29 species could be amplified (see Additional file 2).
Among the 46 non-O1/non-O139 V. cholerae strains isolated from patients (28 strains) or environments (18 strains), we obtained the amplicons of at least one gene encoding the apparatus protein of the T3SS2α genes from 10 strains (see below). In two V. cholerae strains, which constitute the PCR products of T3SS2β genes, at least six genes for the apparatus and two genes for the translocons could be amplified (see Additional file 2). We therefore concluded that the aforementioned 10 V. cholerae strains were T3SS2α-positive and the two were T3SS2β-positive. Of these 12 T3SS2-positive strains, only one, the V. cholerae strain RIMD2214415, which possesses T3SS2α genes, was isolated from the environment. Therefore, as far as we could determine in this study, T3SS2 genes of V. cholerae tend to be found in clinical strains rather than in environmental isolates.
In all of the five V. hollisae strains tested, the amplicons for three genes of T3SS2α, vscN2, vscR2 and vscT2, were obtained with the PCR assay, but no other T3SS2α genes or any T3SS2β genes could be amplified. The PCR products for vscN2R2T2 could be partially sequenced, which confirmed that the amplicons that could be obtained are more closely related to the T3SS2α than to the T3SS2β genes (data not shown).
The PCR products of the genes for T3SS2 were detected in nine of 15 clinical or environmental V. mimicus strains. The genes encoding the apparatus proteins of T3SS2, vscN2C2R2T2U2 and vcrD2, were amplified by PCR in all the T3SS2-positive V. mimicus strains, although the amplicons for the genes encoding effector proteins, i.e., vopCLP, could not be obtained in a few of these strains (see Additional file 2). Of the nine T3SS2positive strains, at least six genes for the apparatus proteins and two genes for the translocons of T3SS2α genes could be amplified from eight strains, while PCR amplification led to the detection in a V. mimicus strain of the amplicons of the T3SS2β genes, i.e., six genes encoding the apparatus proteins vscN2C2R2T2U2 and vcrD2, two genes encoding the translocons vopB2D2, and two genes for the regulators vtrAB. In the other six V. mimicus strains, no amplicons of the genes for either type of T3SS2 could be obtained (see Additional file 2). Of the nine T3SS2-positive V. mimicus strains, eight were therefore identified as T3SS2α-positive, and one as T3SS2β-positive. These findings suggest that, in addition to their distribution in V. parahaemolyticus and V. cholerae strains, the genes for T3SS2 are found in V. hollisae and V. mimicus strains, although only three T3SS2 genes were detected in V. hollisae strains.
Distribution of tdh or trh in T3SS2-positive V. cholerae and V. mimicus strains In V. parahaemolyticus, the strains which possess the T3SS2 gene cluster also possess the tdh and/or trh genes [7,20]. To examine whether the tdh or trh genes also coexist in T3SS2-possessing V. cholerae and V. mimicus strains, we performed a PCR assay using the pair of primers (see Additional file 1) which target the tdh or trh genes of V. parahaemolyticus RIMD2210633 or TH3996 strains, respectively. Among the T3SS2-possessing V. mimicus strains, the tdh gene could be detected in all T3SS2α-positive strains, although no amplicons of the trh gene could be obtained in V. mimicus strains, while neither of the amplicons could be obtained in the T3SS2positive V. cholerae strains. To analyze the distribution of the tdh or trh genes in 12 T3SS2-positive V. cholerae strains, we performed an additional PCR assay using the primer set (see Additional file 1) targeting the region between A33_1702 and the downstream region of the V. cholerae AM-19226 strain which is homologous with tdh of V. parahaemolyticus. PCR products could be obtained for both the T3SS2α-and T3SS2β-positive V. cholerae strains, except for the RIMD2214321 (T3SS2α-possessing) strain (data not shown). These results suggest that the tdh gene may be related to the presence of the T3SS2 gene cluster in V. cholerae and V. mimicus strains.

Gene organization of the T3SS2 gene cluster in V. mimicus
The results presented here demonstrated that some V. mimicus strains possess the genes for T3SS2α or T3SS2β. Since the gene organizations of the T3SS2 gene cluster in the organism were completely unknown, we attempted to analyze the gene organization of the T3SS2 region in V. mimicus strains. To this end, we performed PCR scanning against the genomic DNA of T3SS2-positive V. mimicus strains RIMD2218080 (T3SS2α) and RIMD2218067 (T3SS2β) by using six PCR primer pairs for each strain (see Additional file 3).
In the T3SS2α-positive strain RIMD2218080, PCR products of the expected size were detected for all primer pairs (data not shown), thus suggesting that the gene organization of the T3SS2 gene cluster in strain RIMD2218080 is similar to that of the T3SS2α gene cluster in V. parahaemolyticus strain RIMD2210633. In the T3SS2β-positive strain RIMD2218067, amplicons of the expected size were obtained for five primer pairs, although the size of the product obtained for a primer pair between the vopD2 and vopC genes was notably larger, by approximately 5 kb, than that of the region of the V. parahaemolyticus TH3996 (data not shown). However, the PCR product which was amplified between vopD2 and vopP was the same size as that of the V. parahaemolyticus TH3996 strain. This suggested that the gene organization of T3SS2 of V. mimicus RIMD2218067 is similar to that of T3SS2β of V. parahaemolyticus, while a genetic element of approximately 5 kb, which could not be detected in the region of V. parahaemolyticus strain TH3996, may exist in the region between the vopP and vopC genes in V. mimicus strain RIMD2218067.
These findings suggest that the gene organization of the T3SS2 gene clusters, both T3SS2α and T3SS2β, in V. mimicus strains are basically similar to those of the V. parahaemolyticus and V. cholerae strains.
Phylogenetic analysis of the T3SS2-related genes in V. mimicus Next, we analyzed the phylogeny of the T3SS2 genes identified in V. mimicus strains. The purified amplicons of the genes for vscN2R2T2 in the T3SS2-positive V. mimicus strains were sequenced and the nucleotide sequences thus obtained were used for phylogenetic analysis. In addition, we used the nucleotide sequences of the three T3SS2 genes of the two V. parahaemolyticus strains RIMD2210633 and TH3996, and the four V. cholerae strains, AM-19226, 1587 and 623-39, as well as V51, identified to date. Phylogenetic trees for each of the genes were constructed with the Neighbor-Joining (NJ) method.
The analysis demonstrated that the PCR products of the T3SS2 genes in V. mimicus strains RIMD2218022, 2218042, 2218069, 2218070, 2218080, 2218081, 2218082 and 2218083 belong to the cluster containing the T3SS2α genes of V. parahaemolyticus strain RIMD2210633 and that of V. cholerae strains AM-19226 and V51 (Figure 1). In contrast, the amplicons obtained from the T3SS2 genes in the V. mimicus strain RIMD2218067 were found to be closely related to the T3SS2β genes in the V. parahaemolyticus TH3996 strain and V. cholerae strains 1587 and 623-39 ( Figure 1). These findings confirmed that, similar to the findings for V. parahaemolyticus and V. cholerae strains, the T3SS2 of V. mimicus strains could be classified into two phylogroups, T3SS2α and T3SS2β.

Presence and absence of the genes in VPI-2 and Vp-PAI
Both the T3SS2 gene cluster of V. parahaemolyticus and the T3SS gene cluster of V. cholerae can be found on PAIs [7,19,20]. In V. cholerae non-O1/non-O139 strains AM-19226, NRT36S, V51, 1587 and 623-39, the PAI (VPI-2) contains the genes for sialic acid metabolism in addition to the genes for T3SS (Figure 2) [19]. In V. parahaemolyticus strains RIMD2210633 and TH3996, on the other hand, the homologues for sialic acid metabolism were not found in the Vp-PAI (Figure 2  may be more closely related to that of V. cholerae VPI-2 than of V. parahaemolyticus Vp-PAI ( Figure 2).

Cytotoxicity assay of mutant strains
Previous studies have demonstrated that T3SS2s of V. parahaemolyticus RIMD2210633 and TH3996 as well as V. cholerae AM-19226 contribute to the pathogenicity of these organisms [14,17,20,[22][23][24]. To determine the possible contribution of T3SS2 to the pathogenicity of V. mimicus strains, we compared the cytotoxicity of the wild-type and mutant strains for cultured cell lines. T3SS-deficient mutants were constructed by disruption of the homologue of the vscN2 gene, which encodes an ATPase of T3SS2, in V. mimicus RIMD2218042 (α type) and RIMD2218067 (β type) strains. To confirm the deletion of the vscN2 gene, PCR amplification using oligonucleotide primer pairs was performed (see Additional file 1 and 8). The growth of the mutant strains in LB medium (1% NaCl) was indistinguishable from that of the parental strains (data not shown). Both V. mimicus RIMD2218042 and RIMD2218067 strains were cytotoxic for Caco-2 cells at 3 h post-infection. The cytotoxicity of both the T3SS2α-and T3SS2β-deficient mutant strains tended to decrease, but there were no significant differences between T3SS2α-and T3SS2β-deficient mutant strains and their parental strains (see Additional file 9).

Discussion
A recent study of ours demonstrated that two lines of distinct lineage of the T3SS2 gene cluster, T3SS2α and T3SS2β, are present in the KP-positive and trh-positive V. parahaemolyticus strains, respectively [20]. Although a previously reported study using dot blot analysis could not detect the genes for T3SS2 in 16 Vibrio species, the probes and PCR primers used in previous studies were designed based on the sequence information of the T3SS2α genes in V. parahaemolyticus strain RIMD2210633 [7,14]. Since the T3SS2β genes cannot be detected by either PCR amplifications or comparative genomic hybridization analysis targeting the T3SS2α genes [7,15], we re-investigated the distribution of the T3SS2 genes, both T3SS2α and T3SS2β, in Vibrio species.
To examine the distribution of the genes for T3SS2 in vibrios other than V. parahaemolyticus, we performed a PCR assay using PCR primer pairs targeting both the T3SS2α and T3SS2β genes. Of the 32 Vibrio species tested, the T3SS2-related genes were detected in three species, V. cholerae, which was previously reported, as well as V. hollisae and V. mimicus.
In V. hollisae strains, only three genes for T3SS2α, vscN2, vscR2, and vscT2, were detected. Nevertheless, the fact that the PCR reactions for these three genes were positive in all the five V. hollisae strains tested is intriguing. We speculate that the other genes for T3SS2α might be absent in these particular V. hollisae strains, or that the sequences of the other genes included variations that would make PCR amplification with the primer pairs used in this assay difficult. These possibilities should be examined in the future by more detailed genetic analyses, e.g. sequencing of the region flanking the T3SS2-related genes.
A previous study showed that the T3SS2-related genes are present in V. mimicus strains [25]. In our study, the PCR assay also demonstrated the presence of the T3SS2 genes in V. mimicus strains. Of the 15 V. mimicus strains tested in this study, 12 strains were isolated from patients and three from the environment, with all T3SS2-positive strains isolated from patients. There were eight T3SS2α-positive and one T3SS2β-positive strain among the T3SS2-positive V. mimicus strains. The gene organization of the T3SS2 gene cluster in the V. mimicus strains containing T3SS2α or T3SS2β, was basically similar to that of the V. parahaemolyticus and V. cholerae strains. Ours is thus the first study to demonstrate that the two distinct types of T3SS2 gene clusters, T3SS2α and T3SS2β, are present not only in V. parahaemolyticus and V. cholerae but also in V. mimicus strains. Furthermore, we could show that the structures of the V. mimicus PAIs containing the T3SS genes may be more closely related to those of V. cholerae VPI-2 than of V. parahaemolyticus Vp-PAI ( Figure 2). In contrast, the ORFs in VPI-2 were not detected in any of the T3SS-negative V. mimicus strains. This implies, therefore, that the similar PAI cassettes containing the T3SS2 gene cluster were acquired through horizontal gene transfer in V. cholerae and V. mimicus (Figure 3).
The PCR primer pairs used in this study were found to be useful for detecting as well as distinguishing the genes for T3SS2α and T3SS2β in Vibrio species. In particular, the PCR assays targeting the three genes, vscN2, vscR2 and vscT2, produced stable and reliable results for detection of T3SS2-related genes. We therefore consider that, for determining the presence or absence of these genes, PCR amplification using the primer pairs for the vscN2R2T2 genes of T3SS2α or T3SS2β is effective and rapid. Although only a limited number of strains of the non-human pathogenic Vibrio species was examined in this study, more extensive studies of those species using more strains may well reveal the presence of the T3SS2 genes in vibrios other than the ones reported here.
Previous studies showed that the T3SSs of V. parahaemolyticus and V. cholerae contribute to their pathogenicity for humans [14,17,20,[22][23][24]. In V. mimicus, a bacterium which is known to be a causative agent of gastroenteritis in humans, the hemolysin was previously reported as a major virulence factor [26]. To assess the function of T3SS of V. mimicus in pathogenicity in our study, we evaluated the cytotoxicity of V. mimicus for Caco-2 cells because V. parahaemolyticus had a T3SS2dependent cytotoxic effect on Caco-2 cells [22]. The results indicated that both T3SS2α-possessing and T3SS2β-possessing V. mimicus strains showed the cytotoxic activity on Caco-2 cells in this assay. Although we could not detect statistically significant differences between T3SS-deficient mutants and parental strains, there was a tendency for the cytotoxicity of T3SS-deficient mutants to diminish than that of the parental mutants. A previous report showed that the deletion of the hemolysin gene in V. mimicus significantly reduced fluid accumulation in rabbit ileal loop tests, but the mutant partially retained this action, which suggests that, besides the hemolysin, V. mimicus may contain an additional virulence determinant(s) [26]. It is therefore possible that T3SS is a candidate for the previously unidentified virulence determinant in pathogenic V. mimicus strains for humans. The observed ambiguous differences in cytotoxicity between the mutants and the parental strains may be due to insufficient expression of T3SS of V. mimicus under the culturing conditions used in this study, because it is still unclear what the optimal conditions are for inducing T3SS of V. mimicus. This possibility needs to be examined in future studies.

Conclusions
This study demonstrated the presence of the gene cluster for T3SS2α or T3SS2β in V. mimicus, a bacterium which is known to be a causative agent of gastroenteritis in humans. Since it was reported that the T3SSs of V. parahaemolyticus and V. cholerae contribute to their pathogenicity for humans, the T3SS in V. mimicus identified in this study also might be a candidate virulence factor of this organism for humans. This possibility needs to be examined in future studies.  yeast extract, 0.5%) with 3% NaCl for V. parahemolyticus and in Difco marine broth 2216 for V. nigripulchritudo, V. pectenicida and V. halioticoli. Other bacteria were grown in LB broth with 1% NaCl.

Oligonucleotide primers and PCR conditions
Additional file 1 shows the oligonucleotide primers used in this study. Chromosomal DNA from Vibrio species strains was extracted for PCR as previously described [20]. For detection of the presence of the T3SS2 genes in related Vibrio species, PCR using the EX-PCR Kit (Takara Shuzo, Kyoto, Japan) was performed. The PCR conditions were as follows: after initial denaturation at 94°C for 3 min, a cycle of 94°C for 30 s, 55°C for 30 s, and 72°C for 30 s, 45 s, 1 min or 1.5 min was repeated 30 times. PCR scanning of the V. mimicus was performed using genomic DNA as a template and with a long accurate (LA) PCR kit (Takara Shuzo) as previously described [20]. Custom-synthesized oligonucleotides for the PCR were purchased from GeneDesign (Osaka, Japan).

DNA sequencing and informatic analysis
To sequence the DNA fragments amplified by PCR, the fragments were purified with the PCR Gel Extraction Kit (QIAGEN, Valencia, CA) according to the manufacturer's protocol. DNA sequencing was performed with the ABI PRISM 3130 (Applied Biosystems, Foster City, CA) and the BigDye v3.1 cycle sequencing kit (Applied Biosystems). The Genetyx sequence analysis program (Software Development, Tokyo, Japan) was used for computer analysis of DNA sequences. Homology searches against deposited sequences were performed with the aid of data from the National Center for Biotechnology Information (NCBI) using the BLAST network service http://www.ncbi.nlm.nih.gov and the BLAST service at the Genome Information Research Center http://genome.naist.jp/bacteria/vpara/. Sequence information was obtained from the NCBI. The computer program CLUSTAL W was used for the nucleotide sequence alignment and phylogenetic analysis.

Construction of vscN2 deletion mutant strains of V. mimicus
A four-primer PCR technique was used to engineer an in-frame deletion mutation as described previously [14]. Briefly, the upstream and downstream sequences of vscN2 of T3SS2α or T3SS2β were amplified using the pairs listed in Additional file 1. The two fragments, amplified with primers 1 and 3, and 2 and 4, respectively, were used as templates for a second PCR using primers 1 and 4, which generated a PCR product containing the desired deletion. The amplified fragments were then sequenced and subcloned into an R6K-ori suicide vector pYAK1 and transformed into E. coli SM10λpir.

Cytotoxicity assays
For cytotoxicity assays, eukaryotic cells were seeded at 3 × 10 4 cells well -1 in 96-well plates and cultured for 48 h to confluency. The cells were co-cultured with PBS-washed bacteria at a multiplicity of infection (moi) of 10 for 2-6 h. The release of lactate dehydrogenase (LDH) into the medium was quantified by using a CytoTox96 non-radioactive cytotoxicity kit (Promega) according to the manufacturer's instructions. The LDH release (per cent cytotoxicity) was calculated with the following equation: [optical density at 492 nm (OD 492 ) of experimental release -OD 492 of spontaneous release]/(OD 492 of maximum release -OD 492 of spontaneous release) × 100. Spontaneous release is defined as the amount of LDH released from the cytoplasm of uninfected cells, and maximum release as the total amount of LDH released after the complete lysis of uninfected cells.

Statistical analysis
Statistical significance was determined with the t test. A P value of < 0.05 was considered statistically significant.

Additional material
Additional file 1: Table S1. The primer sequences used in this study. Oligonucleotide primers sequences used in this study are listed in this table.
Additional file 2: Table S2. Distribution of the T3SS2a or T3SS2b genes on Vp-PAI in 32 Vibrio species. The species and strain ID of Vibrio strains used in this study are listed in this table.
Additional file 3: Figure S1. Gene organization of the T3SS2a and T3SS2b gene clusters in V. parahaemolyticus strains. Genetic organization of T3SS2 in V. parahaemolyticus TH3996 (β type) and RIMD2210633 (α type) strains. Genes are indicated by arrows, with red arrows indicating the genes encoding putative apparatus proteins of T3SS2, blue arrows the genes encoding putative regulatory and effector proteins of T3SS2, and gray arrows the genes encoding hypothetical proteins. The colors of the arrows are identical to those used in a previous report of ours [20]. The 12 lines with arrowheads at both ends, representing PCR-Vpa1-Vpa6 and PCR-Vpb1-Vpb6, designate the regions that were amplified for PCR scanning.