Introduction

Members of the family Vibrionaceae are Gram-negative Gammaproteobacteria and are found in a wide variety of aquatic biotopes, including extremophile species that can live at great depths such as Photobacterium profundum or symbionts such as Aliivibrio fischeri (Thompson and Swings 2006).Most members of the family Vibrionaceae are free-living marine bacteria but many can also form biofilms on marine invertebrate exoskeletons or other surfaces (Thompson and Swings 2006). The members of this family have also an important impact on aquaculture; Aliivibrio salmonicida, Vibrio anguillarum and Vibrio vulnificus are major fish pathogens, while Vibrio harveyi and Vibrio parahaemolyticus are the most important pathogens of shrimp (Austin and Austin 1999; Gomez-Gil et al. 1998, 2003; Tran et al. 2013; Soto-Rodriguez et al. 2015).

Notably, the genus Vibrio consists of over 90 species (http://www.bacterio.net/vibrio/html). However, within this genus there are closely related species that are difficult to identify. The genus is divided in many clades according to their phylogenetic relationships established by Multilocus Sequence Analyses (MLSA, Sawabe et al. 2007, 2013). In particular, the so-called Marisflavi clade (Lucena et al. 2012) consists of three species, Vibrio aestivus, Vibrio marisflavi and Vibrio stylophorae based on the 16S rRNA gene sequences. These species were recently isolated from seawater and corals (Wang et al. 2011; Sheu et al. 2011, Lucena et al. 2012). However, the robustness of this clade has not yet been supported by MLSA. Furthermore, the closely related Mediterranei clade consists of five species, Vibrio maritimus, Vibrio variabilis, Vibrio mediterranei, Vibrio madracius and Vibrio thalassae (Sawabe et al. 2013; Moreira et al. 2014; Tarazona et al. 2014).

DNA–DNA hybridization (DDH) is still considered a gold standard for species delineation, despite that other methods have greater resolving power and are easier to perform. Such methods comprise of genomic taxonomy techniques such as genome-to-genome distance (GGD), average nucleotide identity (ANI), and genotype-to-phenotype on the basis of whole genome sequences (Amaral et al. 2014; Thompson et al. 2015). In this study, a genomic taxonomy study was performed to classify a novel Vibrio strain, CAIM 1540T.

Materials and methods

Bacterial strain and growth conditions

The strain CAIM 1540T was isolated on Marine Agar from a cultured oyster (C. corteziensis) in La Cruz, Sinaloa state, México (23°55′05.9″N 106°53′24.7″W) on February 13th, 2004. The type strains of V. aestivus CAIM 1861T and V. marisflavi CAIM 1886T were obtained from the Collection of Aquatic Important Microorganisms (CAIM). The strains were grown on Tryptone Soy Agar (TSA; Oxoid) + 2 % NaCl (w/v) and incubated at 30 °C for 24 h. Cultures were maintained frozen at −80 °C in Tryptone Soy Broth (TSB) supplemented with 2 % NaCl (w/v) and 15 % (v/v) glycerol as preservative.

Phenotypic analyses

The following phenotypic tests were performed with the strains (MacFaddin 1993; Noguerola and Blanch 2008): Gram staining, catalase and oxidase activities, cell morphology, motility, oxidation/fermentation test, methyl red test, Voges–Proskauer test, utilisation of citrate, arginine dihydrolase, lysine and ornithine decarboxylation, and nitrate reduction. Further characterisation was done using API 20E test strips (bioMérieux) and Biolog GN MicroPlate™. The strains were grown onTSB to determine salt tolerance (0–10 % NaCl), growth at different temperatures (8, 20, 30, 37, 40 °C) and pH (2–14). The sensitivity to the vibriostatic agent O/129 (2,4-diamino-6, 7- diisopropylpteridine, 150 µg per disc) and the use of 44 substrates as sole carbon and energy sources were determined as described previously (Macián et al. 2001).

Analysis of fatty acid methyl esters (FAMEs) was performed according to the Microbial Identifications Systems (MSI) (MIDI, Newark, DE, USA) protocol as described by Sasser (1990). The cells were grown on TSA supplemented with 2.0 % NaCl (w/v) and incubated at 25 °C for 24 h.

DNA isolation, amplification, sequencing, and sequence analysis

Genomic DNA was extracted from 24 h cultures on TSA supplemented with 2 % NaCl (w/v) using a Promega kit (Wizard® Genomic DNA Purification Kit). The amplification and sequencing of the 16S rRNA gene, and of the genes ftsZ (cell division protein), gapA (glyceraldehyde 3-phosphate dehydrogenase), recA (recombinase A gene) and topA (topoisomerase I) were performed as previously described (Pascual et al. 2010; Cano-Gomez et al. 2010; Yoshizawa et al. 2010). Sequence data analysis was carried out with the DNASTAR Lasergene SEQMAN program. Sequence similarity of the 16S rRNA was determined using the EzTaxon-e server (Kim et al. 2012). Phylogenetic trees were reconstructed using the neighbour-joining, maximum-likelihood, and maximum parsimony algorithms with MEGA ver. 5.05 (Tamura et al. 2011). Sequences of phylogenetically closely related species were obtained from GenBank/EMBL/DDBJ (Supplementary Table 1).

The draft genome of CAIM 1540T was sequenced by means of the Ion Torrent PGM platform as described earlier (Quail et al. 2012; Moreira et al. 2014) with minor modifications as follows. Library preparation was carried out using the Ion Plus Fragment Library Kit, with 1 μg DNA (in Low TE, 50 μL). DNA was fragmented using the BioRuptor®Sonication System as described in the Ion Plus Fragment Library Kit protocol. End repair, adapter ligation, nick repair, and amplification (10 cycles) were also performed as described in the Ion Plus Fragment Library protocol. 300 and 350 bp fragments were selected through agarose gel (2 % m/v) electrophoresis (E-Gel SizeSelect, Life technologies). Quality and concentration of the libraries were determined using an Agilent 2100 Bioanalyzer with the associated High Sensitivity DNA kit (Agilent Technologies), as well as with an Ion library Quantization kit using TaqMan® in a CFX96™ Real-Time PCR System (Bio-Rad). The amount of library required for template preparation was calculated using the Template Dilution Factor calculation described in the manufacturers protocol. Emulsion PCR and enrichment steps were carried out in the Ion OneTouchTM 200 Template Kit v2. Ion Sphere Particle quality assessment was carried out as outlined in this protocol. Sequencing was done using a 318 chip with barcoding. The Ion PGM™ 200 Sequencing Kit was used for sequencing following the recommended protocol and Torrent Suite 1.5 was used for analyses. The reads were assembled de novo with Newbler (RunAssembly ver. 2.3).

The DNA G+C mol% was estimated using a method described previously (Moreira et al. 2011) and from the draft genome sequence.

Accession numbers

The GenBank/EMBL/DDBJ accession numbers for the 16S rRNA, ftsZ, gapA, recA, and topA gene sequences of strain CAIM 1540T are JQ434105, KP698215, KP698212, KP698213, and KP698214 respectively; all other gene sequences used in this study are listed in the supplementary Table 1. The genome accession number of the type strain CAIM 1540T is JYJP00000000.

Result and discussion

Bacterial characterisation

Strain CAIM 1540T was isolated from a cultured oyster (C. corteziensis) in La Cruz, Sinaloa state, México. The strain showed phenotypic characteristics that place it clearly asa member of the genus Vibrio: the cells were observed to be motile small rods, Gram-negative, facultatively anaerobic, oxidase and catalase positive. The strain was found to require sodiumions for growth; the optimal salinity range obtained was between 3 and 5 % (Fig. S1). No growth was observed at NaCl concentrations of 7 % or higher. Optimal pH was established at a range of 6–8 (Fig. S2). The optimal temperature was found to be 30 °C (Fig. S3) and no growth was observed at 40 °C. The strain was found to grow on thiosulfate-citrate-bile-sucrose agar (TCBS) agar (Difco) as green colonies. Strain CAIM 1540T was also found to be able to reduce nitrates to nitrites, to be positive for methyl red test, urea hydrolysis, l-tryptophan deaminase, indole production, and bovine gelatin hydrolysis. Strain CAIM 1540T was found to be negative for citrate utilisation, the Voges–Proskauer test, arginine dihydrolase and lysine and ornithine decarboxylase. Several phenotypic tests could be selected to clearly differentiate CAIM 1540T from the closest related species Vibrio species (Table 1).

Table 1 Phenotypic characteristics that distinguish V. mexicanus sp. nov. from related Vibrio species
Full size table

The fatty acid profile of strain CAIM 1540T showed the main features (Table 2) of the members of the genus Vibrio.

Table 2 Total fatty acid content (%) of V. mexicanus sp. nov. (CAIM 1540T) and of related Vibrio species
Full size table

Phylogenetic analysis

The closest related species identified by 16S rRNA gene sequence were found to be V. aestivus and V. marisflavi, with similarity values of 99.02–97.05 % respectively (Fig. 1); the third member of the proposed Marisflavi clade (Lucena et al. 2012), V. stylophorae, was found to show a more distant relationship (95.04 %) and this species was well separated in the phylogenetic tree (Fig. 1). Similarities with members of the Mediterranei clade were 96.7 % with V. maritimus and 96.6 % with V. variabilis. The 16S rRNA gene sequence analysis clearly separates CAIM 1540T from V. aestivus forming a separate branch, supported with a high bootstrap value(Fig. 1).

Fig. 1
figure 1

Phylogenetic tree based on partial 16S rRNA gene sequences obtained by the neighbour joining method based on the Jukes–Cantor model. GenBank sequence accession numbers are given in parentheses. Numbers at nodes denote the level of bootstrap based on 1000 replicates; only values greater than 50 % are shown. Vibrio cholerae was used as bacterial out-group. Bar 0.5 % estimated sequenced divergence. Scale bar, base substitutions per site

Full size image

MLSA has been proposed as a valuable technique for the identification and classification of vibrios (Cano-Gomez et al. 2010). In this study, sequences of the following four housekeeping genes were obtained: cell division protein (ftsZ, 432 pb), glyceraldehyde-3-phosphate dehydrogenase A (gapA,602 bp), protein recombinase A ( recA, 494 bp), and DNA topoisomerase 1 (topA, 402 bp); these sequences were compared with those of related species. The phylogenetic trees based on the housekeeping genes ftsZ (Fig. S4), gapA (Fig. S5), recA (Figs. S6, S10), topA (Fig. S7) and the 16S rRNA gene (Fig. 1), and on the concatenated sequences of these five genes (Figs. 2, S8, S9), confirmed the clustering of the proposed novel Vibrio species as a independent branch with high bootstrap values, and its distinction from the closest phylogenetic neighbours. In all individual trees the isolate formed a monophyletic group with V. aestivus. Notably these analyses suggest that this lineage is well separated from V. marisflavi, suggesting that neither V. aestivus nor CAIM 1540T should be considered members of the Marisflavi clade.

Fig. 2
figure 2

Phylogenetic tree based on concatenated sequences (3131 bp) of the four housekeeping genes ftsZ (432 bp), gapA (602 bp), recA (494 bp) and topA (402 bp), and the 16S rRNA gene (1261 bp) sequences available from the GenBank (accession numbers are listed in Supplementary Table S1) Neighbour Joining. Numbers at nodes denote the level of bootstrap based on 1000 replicates; only values greater than 50 % are shown. V. cholerae was used as bacterial out-group. Scale bar, base substitutions per site

Full size image

Each MLSA gene was analysed for recombination events with the program SplitsTree v4 (Huson and Bryant 2006) (Figs. S11–S18).Recombination analyses are useful to detect recombination events that may affect the topology of phylogenetic trees; a recombination event increases the distance between species. Several recombination events were observed, especially in recA as seen for other vibrios (González-Castillo et al. 2014). Nevertheless, the analysis of concatenated gene sequences provides more informative data and minimizes the weight of recombination events, making it a tool that increases the quality of phylogenetic analyses and provides a greater power of taxonomic resolution (Pascual et al. 2010). In this study recombination events were minimized by using concatenated gene sequences and, therefore, phylogenetic tree topology was unaffected.

Genomic analysis

DDH has been the gold standard for prokaryotic classification at the genomic level as it offers a numerical and relatively stable limit, but in the era of genomics this method seems to be out-dated and could be replaced by a comparison between sequenced genomes. ANI with at least 20 % of the genome of the query strains rather than its complete sequence, is enough to clearly differentiate species (Richter and Rosselló-Móra 2009).ANI is calculated with two algorithms, BLAST and MUMmer. The proposed limits for the definition of species are set at 95–96 % of ANI values (Lucena et al. 2012). Therefore, shotgun genome sequencing was performed on strain CAIM 1540T. The subsequent assembly produced 167 contigs (N 50 = 88,690 bp, G+C 43.7 %) for a 5.4 Mb genome.

Comparison of the draft genome of strain CAIM 1540T yielded ANI values (Table 3) of 89.6 % (ANIb) and 90.6 % (ANIm) with the closest related species V. aestivus, 71.5 % (ANIb) and 85.5 % (ANIm) with V. marisflavi, 72.6 % (ANIb) and 85.7 % (ANIm) with V. maritimus, and 72.6 % (ANIb) and 85.8 % (ANIm) with V. variabilis. These values are clearly below the species-delineating threshold of 96 %, indicating that strain CAIM 1540T does not belong to these previously described species.

Table 3 Results of ANI calculations (%) using JS pecies software
Full size table

Digital DDH calculations were also performed for these genome sequences. The genomic distance was calculated using the Genome-To-Genome Distance Calculator (GGDC) (Auch et al. 2010; Meier-Kolthoff et al. 2013). In silico DDH is calculated by the Genome Blast Distance Phylogeny (GBDP), which was devised as an approach for the inference of phylogenetic trees or networks from a given set of wholly (or even incompletely) sequenced genomes (Henz et al. 2005) and was subsequently revisited and enhanced (Auch et al. 2010). Strains from the same prokaryotic species share >70 % in silico GGDC (Thompson et al. 2013). Comparisons with the draft genome of strain CAIM 1540T yielded GGDC values as low as 39.5 ± 2.5 % with V. aestivus and 20–23 % with other closely related species (Table 4), confirming that the strain does not belong to any of these species.

Table 4 Results of GGDC calculations using BLAST+
Full size table

Another tool with a high resolving power, which uses genome sequences is vibrio phenotyping, this program is based on the search for those enzymes related to the phenotype of interest. It uses BLAST to assign a positive match if the identity is greater than 40 % with a sequence query length greater than 70 %. It is an alternative for the phenotypic identification of vibrios (Amaral et al et al. 2014; Thompson et al 2015). CAIM 1540T showed similar results to those obtained with commercial systems (Table 5).

Table 5 In silico phenotypic characteristics that distinguish V. mexicanussp. nov.from related Vibrio species
Full size table

In addition to our genomic studies of strain CAIM 1540T, we found that the recently described species V. thalassae, which was assigned to the Mediterranei clade (Tarazona et al. 2014; Validation List 160, Oren and Garrity 2014) and V. madracius (Moreira et al. 2014; Validation List 161, Oren and Garrity 2015) comprise a single species because the values of ANIb 95.8 %, ANIm 96.6 % and GGDC 70.2 ± 2.9 % (Tables 3, 4) are above the accepted thresholds for delineating prokaryotic species. Thus, V. madraciuscan likely be considered a later heterotypic synonym of V. thalassae.

The DNA G+C mol% range reported for members of the genus Vibrio is between 38 and 51 % (Dieguez et al. 2011).The precise DNA G+C mol% for strain CAIM 1540T was obtained from the draft genome sequence to be 43.7 %; this value was also estimated by a real time-PCR method (Moreira et al. 2011) to be 44.3 mol%.

The polyphasic taxonomic study which involved phenotypic, genotypic, genomic and phylogenetic analyses support the proposal of a novel species, for which the name Vibrio mexicanus sp. nov. is proposed, with CAIM 1540T as the type strain.

Description of Vibrio mexicanus sp. nov

Vibrio mexicanus (me.xi.ca´nus N.L. masc. adj. mexicanus from México).

Gram-negative, curved bacilli that grow as green colonies on TCBS agar, motile and facultatively anaerobic, not luminescent and do not swarm on marine agar or on TSA with 2.0 % NaCl. Growth occurs at 1–6 % NaCl (optimally in 4 % NaCl), no growth without NaCl or with more than 7 % NaCl. Grows at 20, 30, 37 °C (optimum 30 °C). Grows at 4–11 pH (optimally at pH 6–8). Sensitive to the vibriostatic agent O/129 (150 µl per disc). Oxidase and catalase positive. Negative for arginine dihydrolase, lysine and ornithine decarboxylases, positive for nitrate reduction, methyl red test, urea, l-tryptophan, indole, and bovine gelatin; negative reaction for utilisation of citrate and the Voges–Proskauer test. Ferments d-glucose and amygdalin but not d-mannitol, l- rhamnose, inositol, d-sorbitol, d-melibiose, l-arabinose and d-sucrose (in API 20E tests). Utilises the following substrates as sole sources of carbon: 2-ketoglutarate, l-arabinose, l-aspartate, d-cellobiose, d-glucosamine, d-mannitol, d-fructose, glycerol, d-glucose, l-alanine, d-galactose, l-glutamate, lactose, d,l-lactate, malate, maltose, d-mannose, l-leucine, d-melibiose, N-acetyl-d-glucosamine, l-ornithine, l-rhamnose, d-ribose, succinate, sucrose, d-xylose and d-trehalose. Negative for utilisation ofacetate, citrate, d-galacturonate, ϒ-aminobutyrate, d-gluconate, pyruvate, d-glucuronate, salicin, glycine, p-hydroxybenzoate, l-lysine, l-histidine, m-inositol, propionate, putrescine, l-threonine and tyrosine. Using Biolog GN2 MicroPlates, oxidizes the following substrates: d-melibiose, glycerol, d,l, α glycerolphosphate, d-fructose, d-cellobiose, glucose-1-phosphate, glucose-6-phosphate, p-hydroxyphenylacetic acid, bromosuccinic acid, β-methyl-d-glucoside, itaconic acid, succinamic acid, hydroxy-l-proline, l-fucose, glucuronamide, l-leucine, l-rhamnose, glycogen, d-galactose, formic acid, α-ketoglutaric acid, l-alaninamide, l-ornithine, Tween 40, Tween 80, α-d-glucose, d-sorbitol, d-galacturonic acid, d,l-lactic acid, l-alanine, sucrose, d-gluconic acid, malonic acid, N-acetyl-d-glucosamine, α-d-lactose, d-trehalose, propionic acid, adonitol, lactulose, turanose, d-glucuronic acid, quinic acid, l-aspartic acid, l-arabinose, maltose, d-saccharic acid, l-glutamic acid, d-mannitol, sebacic acid and d-mannose; weak positive reactions for inosine, thymidine and phenylethylamine. Negative for I-erythritol, d-alanine, m-inositol, putrescine, 2-aminoethanol, acetic acid, l-histidine, α-cyclodextrin, cis-aconitic acid, dextrin, d-psicose, citric acid, α-ketobutyric acid, d-raffinose, gentiobiose, d-galactonic acid lactone, α-ketovaleric acid, l-phenylalanine, l-proline, N-acetyl-d-galactosamine, l-alanyl-glycine, l-pyroglutamic acid, d-glucosaminic acid, l-asparagine, d-serine, l-serine, 2,3-butanediol, urocanic acid, uridine, xylitol, α-hydroxybutyric acid, l-threonine, d-arabitol, methyl pyruvate, β-hydroxybutyric acid, glycyl-l-aspartic acid, d,l-carnitine, mono-methyl-succinate, γ-hydroxybutyric acid, succinic acid, glycyl-l-glutamic acid, and γ-aminobutyric acid. The major fatty acids are summed feature 3 (comprising C16:1 w7c and/or C16:1 w6c and/or C15:0 iso 2–OH), C16:0, summed feature 8 (C18:1 w6c and/or C18:1 w7c) and C14:0. The following fatty acids are present in small amounts: C15:0 iso, C17:0 iso, C12:0, summed feature 2 (C14:0 3OH and/orC16:1iso) and C12:0 3OH.

The type strain is CAIM 1540T. The type strain was isolated from a cultured oyster, C. corteziensis, in La Cruz, Sinaloa state, México and deposited as CAIM 1540T and as CECT 8828T. The GenBank/EMBL/DDBJ accession numbers for the 16S rRNA, ftsZ, gapA, recA, and topA gene sequences of strain CAIM 1540T are JQ434105, KP698215, KP698212, KP698213, and KP698214 respectively. The genome accession number of the type strain CAIM 1540T is JYJP00000000.