Genome analysis of clinical genotype Vibrio vulnificus isolated from seafood in Mangaluru Coast, India provides insights into its pathogenicity

Abstract

Vibrio vulnificus an opportunistic human pathogen native to marine/estuarine environment, is one of the leading causes of death due to seafood consumption and exposure of wounds to seawater worldwide. The present study involves the whole genome sequence analysis of an environmental strain of V. vulnificus (clinical genotype) isolated from seafood along the Mangaluru coast of India. The sequenced genome data was subjected to in-silico analysis of phylogeny, virulence genes, antimicrobial resistance determinants, and secretary proteins using suitable bioinformatics tools. The sequenced isolate had an overall genome length of 4.8 Mb and GC content of 46% with 4400 coding DNA sequences. The sequenced strain belongs to a new sequence type (Multilocus sequence typing) and was also found to branch with a phylogenetic lineage that groups the most infectious strains of V. vulnificus. The seafood isolate had complete genes involved in conferring serum resistance yet showed limited serum resistance. The study identified several genes against the antibiotics that are commonly used in their treatment, highlighting the need for alternative treatments. Also, the secretory protein analysis revealed genes associated with major pathways like ABC transporters, two-component systems, quorum sensing, biofilm formation, cationic antimicrobial peptide (CAMP) resistance, and others that play a critical role in the pathogenesis of the V. vulnificus. To the best of our knowledge, this is the first report of a detailed analysis of the genomic information of a V. vulnificus isolated from the Indian subcontinent and provides evidence that raises public health concerns about the safety of seafood.

Keywords:

• Vibrio vulnificus
• whole genome sequence
• public health
• India

1. Introduction

Vibrio vulnificus is an opportunistic Gram-negative human pathogen that can cause potentially fatal infections in vulnerable or immune-compromised individuals (Oliver 2013). Despite early diagnosis and aggressive antibiotic treatment, the infection progresses rapidly most often into primary septicemia resulting in death within 24 h (Baker‐Austin and Oliver 2018). Infections with V. vulnificus have been reported from all over the world. Every year, ∼100 cases of V. vulnificus-related primary septicemia are reported in the USA, with almost all cases linked to raw oysters’ consumption in the Gulf Coast region (Oliver 2013; Heng et al. 2017). The density of V. vulnificus is strongly influenced by salinity and water temperature with nearly 85% of such infections being reported during the warmer months and could also affect favorably by global warming and frequently encountered in the molluscan seafood from different parts of the world (Parvathi et al. 2005; Kumar et al. 2006; Givens et al. 2014). Despite extensive research into the occurrence and ecology of V. vulnificus in coastal waters of India, infections caused by this organism are rarely reported with only five clinical reports of V. vulnificus infections documented till date (Saraswathi et al. 1989; De and Mathur 2011; Madiyal et al. 2016; D’Souza et al. 2018; Bhat et al. 2019).

Human infections due to V. vulnificus occur either through consumption of molluscan seafood (gastroenteritis), or by a portal of entry provided through any abrasion or open wounds exposed to seawater (wound infections, necrotizing fasciitis, and primary septicemia). It accounts for more than 95% of deaths in USA due to contaminated seafood consumption and ranking third of all seafood borne infections with a fatality rate of 31% (Ralston et al. 2011; Oliver 2013) and has one of the highest fatality rates among the many food-borne pathogens with fatality rates comparable to pathogens of Biosafety Level (BSL) 3 and 4 (Baker‐Austin and Oliver 2018). Earlier studies on V. vulnificus have identified two genotypes—genotype C (C clade) and genotype E (E clade) based on a dimorphism in the vcg (virulence correlated gene) allele (Rosche et al. 2010). However, in due course it was observed the classification was inconsistent with reference to their sample source (Guerrero et al. 2019). Further V. vulnificus were grouped into three biotypes namely, biotype 1, shellfish associated, most abundant accounting for most of the clinical cases; biotype 2 mostly associated with diseased eels and rarely found to infect humans; biotype 3 earlier thought to be geographically restricted to Israel but also reported from Japan (Bisharat et al. 1999; Hori et al. 2017). Roig et al. (2017) described the phylogenetic diversity into five lineages namely, L1—mixture of clinical and environmental biotype 1 isolates; L2—mixture of biotype 1 and 2 isolates from various sources; L3—exclusively biotype 3 isolates; L4 and L5—biotype 1 isolates from Spanish and Israel region, respectively. López-Pérez et al. (2019) also described the population structure into four ecotypes based on genome analysis of 113  V. vulnificus genomes. Recently, Multilocus sequence typing based phylogenetic analysis has further described a common ancestor originating from East Asia, evolving into two main lineages which further evolved into distinct geographical populations (Bisharat et al. 2020). A study from India has also shown V. vulnificus infection in genetically improved farmed Tilapia (Sumithra et al. 2019).

Geographical and epidemiological comparisons of V. vulnificus have not been possible due to a lack of data from all over the world. Several studies have demonstrated the presence of pathogenic genotypes among environmental isolates of V. vulnificus, demonstrating the insufficiency of genotypic characteristics to predict the isolation source or virulence of whether a given isolate is of clinical significance or not. It is therefore important to know from the public health perspective whether the V. vulnificus population found in Indian marine environments can cause human infections. During 2017–2018, about 70 seafood samples were collected along Mangaluru coast, India was subjected to bacterial isolation using conventional and molecular characterization using species specific marker gyrB gene and virulence markers, vvhA and rtxA as reported previously (Kumar et al. 2006). All the isolates obtained in this study had similar biochemical traits and enzymatic activity and hence representative isolates were selected for the characterization of virulence properties (D’Souza et al. 2020). The present study entails whole genome sequencing and comparative genome analysis of our seafood associated isolate of V. vulnificus and its comparison with 23 isolates of V. vulnificus retrieved from NCBI GenBank. Additionally, a secretome analysis of the V. vulnificus E4010 sequences was carried out to find genes involved in pathogenesis and virulence utilizing systems biology approaches. Overall, the goal of the current research was to provide new insights into the genome of V. vulnificus E4010 and to better understand the interactions between secretory proteins, which can aid in developing disease-management methods for conditions including severe septicemia and necrotizing fasciitis.

2. Methods

2.1. Bacterial strain

V. vulnificus were isolated from seafood samples collected in and around Mangaluru coast of South India. The isolation of V. vulnificus bacteria from the samples were performed following the U.S. Food and Drug Administration (FDA’s) Bacteriological Analytical Manual (BAM) (Kaysner and DePaola 2004) with brief modifications. Altogether 21 isolates were obtained from seafood samples and were confirmed phenotypically and genotypically (D’Souza et al. 2020). Further four representative isolates were subjected to characterization and results revealed no considerable difference among these isolates. Hence one isolate V. vulnificus E4010 was studied in detail for its ability to withstand gastric acidity by activating lysine-dependent acid response system (D’Souza et al. 2019) and contact-mediated rtxA1 expression phenomenon which was shown to correlate with actin disintegration and cytotoxicity upon infection of human cells in-vitro (D’Souza et al. 2020). The isolate E4010 was also tested for antibiotic susceptibility following CLSI guidelines. Briefly, the culture of 0.6 OD₆₀₀ grown in Mueller-Hinton broth (HiMedia Laboratories Pvt. Ltd., Mumbai, India) was spread on Mueller-Hinton agar plates (HiMedia Laboratories Pvt. Ltd., Mumbai, India). The antibiotic discs (HiMedia Laboratories Pvt. Ltd., Mumbai, India) were placed on the medium, and plates were incubated for 16–18 h at 37 °C. Further, to gain deeper insights into the genome composition, V. vulnificus E4010 was subjected to whole genome sequence analysis.

2.2. Genome sequencing and annotation of V. vulnificus E4010

The genomic DNA was extracted using the CTAB-proteinase K method. Sequencing libraries were prepared from 1 ng of genomic DNA using the Nextera XT DNA library preparation kit according to the manufacturer’s instructions (Illumina, San Diego, CA, USA) and sequenced at the Commercial Genome Sequencing facility in Bengaluru, India, using an Illumina MiSeq platform generating 300 base paired-end reads. The prepared libraries’ quantity and size distribution were determined using a Qubit fluorometer and an Agilent Tape Station, respectively. The resulting FASTQ files were quality trimmed before being assembled with SPAdes (Bankevich et al. 2012) to generate contigs without gaps. To assess the quality of each genome assembly, the assembled sequences were subjected to QUAST v4.5 (Gurevich et al. 2013), which measured the total length and N50 value which is the length of the shortest contig that covers 50% of the overall assembly. The structural gene prediction and functional annotations were performed using the Prokaryotic Genomes Automatic Annotation Pipeline (PGAAP) (www.ncbi.nlm.nih.gov/genomes/static/Pipeline). Using JSpecies version 1.2.1 (Richter et al. 2016) the pairwise average nucleotide identity (ANI) of the assembled genomes with the standard V. vulnificus CMCP6 and V. vulnificus YJ016 was calculated for the species identity. Digital DNA-DNA hybridization (dDDH) was also calculated using the GGDC 2.0 server (http://ggdc.dsmz.de/distcalc.php) for V. vulnificus E4010, V. vulnificus CMCP6, and V. vulnificus YJ016. To investigate the genomic relatedness of V. vulnificus E4010, average nucleotide identity (ANI) was calculated between genome sequences of V. vulnificus E4010 and 14  V. vulnificus genomes available in the NCBI database (Table 1) using JSpecies software version 1.2.1 (Richter et al. 2016).

Table 1. Summary of Vibrio vulnificus genomes used in the study.

Sl. No.	Strain name	Country of origin	Year of isolation	Source of isolation	GenBank Accession No.
1	93U204	Taiwan	2004	Tilapia, diseased	GCA_000746665
2	95-8-6	Denmark	1995	Eel	PDNE00000000
3	95-8-7	Denmark	1995	Eel	PDGM00000000
4	95-8-161	Denmark	1995	Eel	PDGN00000000
5	960426-1_4C	Denmark	1996	Eel	PDGL00000000
6	ATCC 27562	USA	1979	Estuarine	GCA_002224265
7	BAA87	Israel	1996	Wound infection	JDSE00000000
8	CG64	Taiwan	2011	Seawater	JZEU00000000
9	CICESE 316	Mexico	2011	Oyster	QKYO00000000
10	CICESE 325	Mexico	2011	Oyster	QKYP00000000
11	CMCP6	South Korea	2016	Clinical	GCA_004355205
12	E64MW	UK	–	Environmental	AFSX00000000
13	FORC_017	South Korea	2018	Clinical—blood	GCA_001675245
14	JY1305	USA	–	Oyster	AFSW00000000
15	JY1701	UK	–	Environmental	AFSY00000000
16	LSU1015	USA	–	Clinical—wound	PDNQ00000000
17	LSU1657	USA	–	Clinical—wound	PDFZ00000000
18	LSU2098	USA	–	Clinical—wound	PDFY00000000
19	MO6-24/O	South Korea	–	Human	GCA_000186585
20	SC9629	South Korea	2011	Seafood	JYGG00000000
21	SC9740	South Korea	2011	Seawater	JZEP00000000
22	SC9761	South Korea	2011	Seafood—oyster	JZEO00000000
23	YJ016	Taiwan	–	Clinical	GCA_000009745
24	E4010 (This study)	India	2018	Oyster	WRPB00000000

2.3. MLST and phylogenetic analysis

The sequence type (ST) of the isolate was investigated based on allelic profiles of ten housekeeping genes, using the MLST 2.0 server (https:/cge.cbs.dtu.dk/services/MLST/) (Larsen et al. 2012). The concatenated sequence alignments of 10 housekeeping genes were used to construct a phylogenetic Neighbor-Joining tree using default parameters with tree visualization by iTOL v3 (Letunic and Bork 2016) for the E4010 genome and 23 global sequences of isolates retrieved from NCBI GenBank (Table 1). The core genome-based phylogenetic analysis was constructed using Bacterial Pan Genome Analysis (BPGA) pipeline where clustering was performed by the USEARCH clustering algorithm v10.0.240_win32 with a 95% cut-off identity. A neighbour-joining core genome phylogenetic tree was constructed with default parameters (Chaudhari et al. 2016) for V. vulnificus E4010, and genomes retrieved from NCBI.

2.4. In-silico virulotyping

Putative virulence factors in V. vulnificus E4010 was investigated using VF analyzer and was compared to the reference strain of V. vulnificus CMCP6. The presence of antibiotic-resistance genes in the genome of V. vulnificus E4010 was determined using PATRIC’s Genome Annotation Service, which employs k-mer-based AMR gene prediction (Wattam et al. 2017). CONTIGuator, a bacterial genome finishing tool (Galardini et al. 2011), was used to reduce the draft genome to chromosome 1 and 2 scaffolds. The pathogenicity islands on two chromosomes off V. vulnificus E4010 were visualized by mapping BLAST Ring Image Generator (BRIG) against reference genome V. vulnificus CMCP6 (Alikhan et al. 2011) and putative pathogenicity islands were predicted using the Island Viewer (Bertelli et al. 2017). In addition, the presence of plasmid sequences was checked using the PlasmidFinder-2.0 Server (Carattoli et al. 2014), phage sequences using PHASTER (Arndt et al. 2016), and CRISPR sequences using CRISPRCasFinder (Couvin et al. 2018), respectively. The rtxA domain search was performed using SMART tool (https://smart.embl.de) and compared to previously predicted rtxA domains of V. vulnificus reference genomes (Kwak et al. 2011; Kim 2018).

2.5. Secretome prediction and analysis

TargetP version 2.0 (Almagro Armenteros et al. 2019) and SignalP version 6.0 (Teufel et al. 2022) were used to examine a set of 4527 proteins from V. vulnificus to predict the secretory signal peptide. Proteins having a SignalP D-Score of Y, a cutoff value of 0.45 for 0 Tm/0.50 for 0.50 Tm, and a TargetPLoc = S were initially merged. The protein sets were then examined using DeepTMHMM to determine whether they included transmembrane domains (Hallgren et al. 2022). To generate the corresponding per-residue sequence of labels, DeepTMHMM uses a deep learning encoder-decoder sequence-to-sequence model that accepts a protein sequence as input. The labels used for each residue are signal peptide (S), inside cell/cytosol (I), alpha membrane (M), beta membrane (B), periplasm (P), and outside cell/lumen of ER/Golgi apparatus/lysosomes (O). The protein’s architecture is determined by the labelling order of its residues. Transmembrane regions that are <10 amino acids from the expected cleavage site in mature peptides were considered for further study, and peptides with 0 or 1 transmembrane regions were retained.

PSORTb v.3.0 was used to predict the subcellular localization of bacterial proteins (Yu et al. 2010). Since a protein’s subcellular localization might give clues about its function in an organism, computational prediction of the subcellular localization of proteins is a useful tool for genome analysis and annotation. The prediction of proteins on the cell surface is particularly interesting for bacterial infections since these proteins may serve as the main targets for drugs or vaccines. Several characteristics of a protein’s basic structure, such as the presence of a signal peptide or membrane-spanning alpha-helices, affect the protein’s subcellular localization.

Functional Annotation and Classification of Proteins of Prokaryotes (FACoP) is used to classify genes for Gene Set Enrichment Analysis. Supported classes are: GO, InterPro (IPR), KEGG-orthology KO, KEGG-pathways, eggNOG, COG, PFAM, and KEYWORDS (de Jong et al. 2022). All the above-mentioned functional classes were assigned using UniProt protein annotation, where over 55,000 bacterial genomes’ specified protein sequences were all mapped to the Swiss-Prot prokaryote database of UniProt (The UniProt Consortium 2021). The best hit for this mapping was DIAMOND (Buchfink et al. 2015), with an e-value cutoff of 0.01. By comparing the descriptions of the COG functional categories and the InterPro annotation, proteins were categorized into COG functional categories. Based on the ‘KW’ field in the UniProt protein database, a unique class of keywords was created that excluded overused and underused terms (de Jong et al. 2022).

2.6. Interactome construction and analysis of secretory proteins

Protein sequence similarity and knowledge of previously identified PPIs serve as the foundation for predicting PPI networks. The secretory protein sequences were selected using the procedures mentioned in the above section. We constructed the PPI network of V. vulnificus secretory proteins using the Search Tool for the Retrieval of Interacting Genes/Proteins database (STRING v11.5) (Szklarczyk et al. 2021). The STRING database contains known and anticipated PPIs, including direct (physical) and indirect (functional) interactions derived primarily from genomic context, high-throughput studies, co-expression, and computer prediction methods. When given a list of proteins as input, STRING will search for their neighbor’s interactor—proteins with whom the input proteins directly interact and then construct a PPI network that includes all of these proteins and their interactions. We first constructed the PPI network of secretory proteins from V. vulnificus using the seed proteins as input, which included the seed proteins and their neighbors. At a medium level of confidence, all of their interactions were generated from high-throughput lab experiments and prior information in curated databases (sources: experiments, databases; score ≥0.40). Further, the Cytoscape 3.9.1 tool was used to visualize these PPI networks, and default parameters were utilized to compute the node attributes of the network (Su et al. 2014).

2.7. In-silico PPI network validation

As a preliminary sanity check of our hypothetical PPI network, we intended to demonstrate that the predicted networks exhibit realistic and usual network features. As a result, we calculated some network statistics and contrasted them to those of established biological networks. The Shapiro–Wilk normality test (Shapiro and Wilk 1965) and the shortest path (Wang et al. 2010) were computed using the Cytoscape plugin Network Analyzer (Assenov et al. 2008), along with the degree distribution (Barabasi and Oltvai 2004), network topology, and the Shapiro–Wilk normality test. According to network theory, the three key indices of connection degree (k), betweenness centrality (BC), and closeness centrality (CC) are typically used to rank the nodes in a network (Raman 2010). Other topological properties include the total number of nodes, each node’s linking components, the network width, radius, density, the number of neighbors for each node, the clustering coefficient, and the average shortest path length (Assenov et al. 2008).

Furthermore, we investigated the inherent network structure by performing cluster analysis. A highly complex biological process is aided by several subnetworks or functional modules (clusters) of proteins intricately connected to a larger biological network. No matter how these modules impact the core network, they also influence each participating node that performs a specific role in the network (Chen et al. 2016). The aforementioned drug-associated/actionable genes were located in clusters obtained with strict parameters and filtered for noise in MCODE, a CytoScape plugin/application that uses an automated clustering algorithm to extract/identify densely connected regions or protein complexes in a PPI network (Bader and Hogue 2003). The degree cutoff value was 2.0, the minimum number of edges required for a node to be scored. The node score threshold, which controls how new nodes are added to the cluster, was set to 0.2, implying that the new node’s score must be at least 80% higher than the cluster’s seed node. The K-Core value, used to rule out clusters without a maximally interconnected core, was given for four edges. MCODE won’t be negatively impacted by the massive size of the interaction dataset and the estimated high false-positive rate over the whole network.

3. Results

3.1. Genomic features of V. vulnificus E4010

V. vulnificus E4010 genome was sequenced using the Illumina MiSeq platform resulting in 2,701,430 reads assembled using SPAdes. The final assemblage comprised 87 contigs with a length of 4,875,404 bp and a GC content of 46%. The largest contig was 429,948 bp in length with N50 size being 250,841 bp. The predicted coding DNA sequences (CDS) in the V. vulnificus E4010 genome was 4400 with 4186 of them being protein-coding genes. Additionally, 29 rRNA genes (nine 5S, ten 16S and ten 23S), 102 tRNA genes, and four non-coding RNAs were found in its genome. A summary of genomic features of V. vulnificus E4010 is presented in Table 2(A). ANI and dDDH analysis of the assembled genome of V. vulnificus E4010 demonstrated >98% pairwise ANI and >89% dDDH values to the reference genome. In addition, Euclidean distance matrix calculated from the percentage of ANI values derived for sequenced and NCBI retrieved strains was used to generate a heatmap to examine the nucleotide similarity (Figure 1(A)) confirmed the V. vulnificus E4010 belonged to clinical genotype though it is isolated from Oyster.

Figure 1. (A) Heat-map of Euclidean distance matric calculated based on percentage values of average nucleotide identity (ANI) values of different strains of Vibrio vulnificus reveals two major groups. The strain sequenced in this study is highlighted in the red box. Details of genomes retrieved from NCBI GenBank are provided in Table 1. The values indicate the extent of dissimilarity from dark pink to green color; (B) MLST based phylogenetic analysis of V. vulnificus genomes and details of their country, source, and ST shows differential branching pattern based on vcg alleles. (vcg-C—red; vcg-E—green). the newly sequenced isolate is marked with ‘a star’; (C) Core genome-based phylogenetic analysis of V. vulnificus genomes showing two distinct lineages based on vcg alleles (vcg-C marked in red and vcg-C marked in green color).

Table 2. Summary of the (A) assembly and annotation; (B) virulence genes identified by VFDB; (C) antimicrobial resistance genes identified by PATRIC in the genome of Vibrio vulnificus E4010 genome.

(A)
Sl.No	Attributes             Values
1.	Genome size	4,875,404 bp
2.	Total number of reads	2,701,430
3.	Contigs	87
4.	Maximum contig size	429,948 bp
5.	N50	250,841 bp
6.	G + C content (%)	46
7.	Coding sequences	4400
8.	tRNA	102
9.	rRNA	29
10.	ncRNA	4
11.	NCBI accession number	WRPB01000000
12.	MLST sequence type	ST-570
13.	ANIb % (against CMCP6)	98.17
14.	ANIb % (against YJ016)	98.00
15.	dDDH % (against CMCP6)	90.5
16.	dDDH % (against YJ016)	89.4
(B)
VF class	Virulence factors	Related genes identified in E4010
Adherence	Mannose-sensitive hemagglutinin (MSHA type IV pilus), Type IV pilus	mshA–mshN, pilB–pilD
Antiphagocytosis	Capsular polysaccharide	cpsA–cpsD, cpsF, cpsH–cpsJ, hp1, wbfV/wcvB, wecA, wza–wzc
Chemotaxis and motility	Flagella	cheA, cheB, cheR, cheV, cheW, cheY, cheZ, lag, flaA–flaC, flaE, lag, flaI, flgA–flgN, flhA, flhB, flhF, flhG, fliA, fliD–fliL, fliN–fliS, flrA–flrC, motA, motB, motX, motY
Enzymes	Metalloproteases	hap/vvp
Iron uptake	Heme receptors, periplasmic binding protein-dependent ABC transport systems, vibriobactin biosynthesis, vibriobactin utilization	hutA, hutR, vctC, vctD, vctG, vctP, viuC, viuD, viuG, viuP, vibA–vibF, vibH, viuA, viuB
Quorum sensing	Autoinducer-2	luxS
Secretion system	EPS type II secretion system	epsC, epsE–epsN, gspD
Toxin	Hemolysin/cytolysin, RTX toxin, Thermolabile hemolysin	vvhA, rtxA–rtxD, tlh
(C)
AMR mechanism	Related genes identified in E4010
Antibiotic activation enzyme	katG
Antibiotic inactivation enzyme	catB family
Antibiotic target in susceptible species	alr, ddl, dxr, EF-G, EF-Tu, folA, dfr, folP, gyrA, gyrB, Iso-tRNA, kasA, murA, rho, rpoB, rpoC, s10p, s12p
Antibiotic target protection protein	qnrB family
Antibiotic target replacement protein	fabV
Efflux pump conferring antibiotic resistance	mdtL, Tet (35), tolC/opmH
Gene conferring resistance via absence	gidB
Protein altering cell wall charge conferring antibiotic resistance	gdpD, pgsA
Regulator modulating expression of antibiotic resistance genes	H-NS, oxyR

3.2. Multi locus sequence typing (MLST) and phylogenetic analysis

Based on allelic profiles of ten housekeeping genes glp, gyrB, mdh, metG, purM, dtdS, lysA, pntA, pyrC, and tnaA, a new sequence type was identified for V. vulnificus E4010 which was submitted to the pubMLST database and assigned ST-570 by the curator. The MLST based phylogeny of V. vulnificus E4010 and 23 global isolates (only from Lineage 1 and 2) showed vcg-C genotypes emerging from a single lineage while vcg-E genotypes branched polyphyletic (Figure 1(B)). The studied strain V. vulnificus E4010 was branched with C-genotype isolates. A core genome-based phylogenetic tree was constructed (Figure 1(C)) and also showed a distinct branching pattern among the two genotypes and V. vulnificus E4010 branched with C-genotype.

3.3. Virulotyping

The in-silico analysis of the genome of V. vulnificus E4010 using VF analyzer for potential virulence factors revealed 119 putative virulence determinants possibly involved in the pathogenesis (Table 2(B)). The comparative overview of virulence factors identified in E4010 with the type strains CMCP6 and YJ016 is depicted in Supplementary Figure S1. Seventeen out of eighteen features that code for mannose-sensitive hemagglutinin (mshA; mshN) and type IV pilus (pilD) which are essential for adhesion and biofilm formation were seen in E4010 along with fibronectin-binding outer membrane protein (OmpU). In the V. vulnificus E4010 genome, the genes coding for two-component system such as, the sensor kinase and response regulators, were also detected. Of the several mechanisms of acid response, the amino acid decarboxylase system coded by cadBA operon, cadA (Lysine decarboxylase- Lysine/Cadaverine antiporter), along with its positive regulator (cadC) and transcriptional regulator (aphB) is identified to be intact in the genome of V. vulnificus E4010 (Figure 2). The virulence factors attributed to phagocytic activity like capsular polysaccharides (cpsA–cpsD, cpsF, cpsH–cpsJ, hp1, wbfV/wcvB, wecA, wza-wzc) were present in the isolate E4010. Also, the investigation of the representative environmental isolate V. vulnificus E4010 identified all those genes that were involved in conferring resistance against human serum by a clinical isolate YJ016. The study also revealed the presence of genes encoding for extracellular cytolysin (vvhA) and multifunctional auto-processing repeats-in-toxin (MARTX) in the genome of E4010. The VFDB analysis performed on this isolate confirmed the existence of genes coding for various iron acquisition systems (hutA, hutR, vctC, vctD, vctG, vctP, viuC, viuD, viuG, viuP, vibA–vibF, vibH, viuA, viuB) that may contribute to the iron uptake and utilization and contribute in establishing infection. Other virulence factors detected in E4010 are genes coding for metalloproteases (hap/vvp), Autoinducer-2 involved in quorum sensing (luxS), and EPS Type II Secretion System (T2SS) (epsC, epsE–epsN, gspD).

Figure 2. Schematic representation of virulence and fitness genes identified in Vibrio vulnificus E4010.

3.4. Pathogenicity islands and multifunctional autoprocessing repeats-in-toxin (MARTX) gene clusters

The reference-based genome assembly using CONTIGuator produced two scaffolds of each chromosome of E4010: chromosome 1 (3,007,903 bp) and chromosome 2 (1,770,110 bp). The BLAST atlas generated against the CMCP6 reference sequence reveals a high degree of sequence similarity between the two (Figure 3(A)). The MARTX toxin gene clusters were present in the small chromosome showing similarity to reference genomes. T6SS-encoding genes were also discovered on the small chromosome which is primarily responsible for toxin secretion and DNA transport. The results show that the V. vulnificus genomic islands are present entirely (VVI-1, VVI-2) or partially (VVI-3, VVI-4, and VVI-5) in the genome of E4010. Furthermore, the Island viewer tool was used to predict putative genomic islands by com­bining three prediction methods: IslandPath-DIMOB, SIGI-HMM, and IslandPick (Supplementary Figure S2). Fourteen and nine putative genomic islands were discovered on chromosomes 1 and 2, respectively. Transposons, prophage sequences, hemolysin precursors, secretion systems, outer membrane proteins, and hypothetical proteins are among the genes encoded by putative genomic islands. In the genome of V. vulnificus E4010, no plasmids, complete prophage sequences, or CRISPR sequences were identified. The domain prediction of the MARTX in sequenced strain V. vulnificus E4010 identified four of the effector domains (Figure 3(B)) along with N—terminal and C—terminal conserved regions.

Figure 3. (A) BLAST atlas for chromosome 1 and 2 of Vibrio vulnificus E4010 generated using BRIG with CMCP6 as reference; (B) Schematic representation of Multifunctional autoprocessing repeats-in-toxin (MARTX) gene clusters in V. vulnificus.

3.5. Antibiotic resistance determinants

The presence of antibiotic resistance genes (ARGs) in V. vulnificus E4010 was identified through PATRIC analysis. Results revealed the presence of several genetic elements that matched with well-characterized ARGs listed in the database. Table 2(C) summarizes the annotated ARG classes and their corresponding mechanisms that were identified in the genome of V. vulnificus E4010. However, antibiotic sensitivity testing of sequenced isolate showed it to be sensitive to all antibiotics tested (Amikacin—30 µg; Azithromycin—15 µg; Cefotaxime—30 µg; Ceftazidime—30 µg; Ceftriaxone—30 µg; Chloramphenicol—30 µg; Ciprofloxacin—5 µg; Co-trimoxazole—25 µg; Doxycycline—30 µg; Imipenem—10 µg; Piperacillin-tazobactam—100/10 µg; Tigecycline—15 µg; Trimethoprim—5 µg) despite the presence of ARGs.

3.6. Secreted protein prediction and analysis of the V. vulnificus proteome

Of the 4527 proteins, 460 proteins could be annotated as classical secretory proteins by SignalP version 6.0, while TargetP version 2.0 classified 819 proteins as secretory proteins. After merging the filtered sets (see Supplementary File S1 for SignalP and TargetP proteins) and removing duplicate segments, 825 proteins were scanned using TMHMM software. A total of 1691 sequences were predicted as secretory proteins, after the removal of 577 transmembrane proteins from the protein data set (Supplementary File S2). Secretory proteins predicted in the previous step were further screened using PSORTb v.3 for predicting the subcellular localization. The results revealed that secretory proteins were distributed as 27 extracellular, 40 outer membranes, 94 cytoplasmic membranes, 74 periplasmic, 22 cytoplasmic, and 322 unknown localizations (Supplementary File S3). The multiple functional classifications of these predicted secretory proteins were performed by using a comprehensive database FUNAGE-Pro. Here, the most common functional classes, such as COG and KEGG (Supplementary Figures S3(A,B)), EggNOG, and operon classes were analyzed and reported in Supplementary File S4. Additionally, for analyzing predicted V. vulnificus secretome, GO terms were assigned to 577 putative secretory proteins in three GO categories namely, molecular function (12), biological process (14), and cellular component (8) (Supplementary File S5).

3.7. The architecture of the secretory proteins interaction network

To find interactions between 577 secretory proteins that have been experimentally validated, the Search Tool for the Retrieval of Interacting Genes (STRING) database was searched. Here, we concentrated on investigating the core interactome of the V. vulnificus E4010, and to better understand pathogenicity in general and identify potentially promising targets for further wet-lab research. A total of 375 nodes and 1606 edges were part of the basic PPI network constructed by STRING (Figure 4 and Supplementary Figure S4). The results of the topological analysis of each node are listed in Supplementary File S6, including the degree, BC, eccentricity, CC, EC, clustering coefficient, etc. For a random network of the same size, the number of edges is significantly higher (p-value = 1.00–16); the nodes were more interconnected than anticipated. The PPI network was shown to be relatively small compared to a random graph, and the proteins may have biological importance. In a biological network, the interacting nodes representing the importance of active genes were integrated using centrality metrics (de Jong et al. 2022). The degree distribution shows a scale-free topology and measures up to a power law (Figure 5(A)). We chose the nodes with BC and/or degree values more than the mean plus standard deviation, and the majority of the proteins/nodes in the secretome network have <17 interactions. Forty-nine nodes, as shown in Table 3, correspondingly, have large connections. A total of 13% of the links in the network are made up of these 49 secretory proteins. In the enormous network, the large node degree is discovered to be 46, with an average degree value of 6.36 and an anticipated number of edges of 743. The network’s average betweenness centrality was 0.0077, while the shortest path’s average length was 3.905. There were 12.58 neighbors on average for all 375 nodes. The average path length distribution (APL = 3.905) of the V. vulnificus E4010 secretory network is shown in Figure 5(B). Fewer path lengths fall into the extreme top categories (path lengths 8 and 9) compared to the lower extremities (path lengths 1 and 2). This suggests that the majority of the proteins in the graph can be related to one another by a small number of paths.

Figure 4. Basic PPI network of Vibrio vulnificus E4010 constructed using STRING consisting of 375 nodes and 1606 edges with clusters highlighted.

Figure 5. (A) the degree distribution shows a scale-free topology and measures up to a power law; (B) the average path length distribution (APL = 3.905) of the Vibrio vulnificus E4010 secretory network.

Table 3. The nodes with BC and/or degree values more than the mean plus standard deviation, and the majority of the proteins/nodes in the secretome network of Vibrio vulnificus E4010.

Name	BetweennessCentrality	ClosenessCentrality	ClusteringCoefficient	Degree
VV93_v1c37740	0.052259	0.370297	0.418357	46
VV93_v1c08230	0.031516	0.364522	0.440244	41
VV93_v1c15230	0.016923	0.350187	0.479082	39
VV93_v1c38150	0.012309	0.350187	0.489879	39
VV93_v1c00420	0.011222	0.354167	0.508772	39
VV93_v1c34010	0.033399	0.3692	0.422475	38
VV93_v1c14430	0.030783	0.358238	0.41394	38
VV93_v1c29010	0.025644	0.350187	0.442943	37
VV93_v1c13620	0.008062	0.347261	0.539039	37
VV93_v1c28640	0.007351	0.345656	0.522523	37
VV93_v1c45320	0.127174	0.373253	0.106349	36
VV93_v1c41410	0.027762	0.359962	0.418487	35
VV93_v1c11280	0.003939	0.338156	0.604278	34
VV93_v1c09620	0.054645	0.369931	0.4375	33
bamC	0.048895	0.332149	0.387097	31
VV93_v1c21290	0.045545	0.362403	0.367742	31
VV93_v1c23110	0.024793	0.341865	0.6	31
VV93_v1c20740	0.015004	0.332444	0.524631	29
VV93_v1c13550	0.011469	0.327496	0.583744	29
VV93_v1c30830	0.018287	0.329225	0.560847	28
VV93_v1c09230	0.006439	0.330973	0.571429	28
AIL70004.1	0.002905	0.329515	0.701058	28
VV93_v1c27660	0.070245	0.346296	0.169231	26
VV93_v1c15720	0.008373	0.320205	0.658462	26
VV93_v1c20800	0.006278	0.315346	0.649231	26
VV93_v1c28020	0.073653	0.338768	0.16	25
surA	0.038047	0.321306	0.237154	23
VV93_v1c05840	0.010784	0.330973	0.604743	23
VV93_v1c00830	0.007517	0.335727	0.675889	23
VV93_v1c22970	0.007349	0.328936	0.687747	23
VV93_v1c10670	0.007108	0.337545	0.494071	23
VV93_v1c22490	0.006563	0.335125	0.506494	22
luxP	0.016389	0.327209	0.657143	21
VV93_v1c32790	0.005421	0.33156	0.6	21
VV93_v1c33530	0.047597	0.348555	0.126316	20
VV93_v1c33510	0.030682	0.353832	0.242105	20
VV93_v1c04260	0.027533	0.293103	0.163158	20
VV93_v1c31910	0.018138	0.321306	0.231579	20
lolB	0.014632	0.290147	0.257895	20
AIL70622.1	0.00737	0.332149	0.567251	19
flgT	0.007103	0.325784	0.48538	19
lolA	0.006986	0.286371	0.374269	19
VV93_v1c20810	0.004725	0.2992	0.699346	18
VV93_v1c22540	0.034587	0.318298	0.345588	17
VV93_v1c10080	0.02802	0.330681	0.213235	17
VV93_v1c30240	0.011091	0.337545	0.5	17
VV93_v1c06430	0.008978	0.291051	0.397059	17
VV93_v1c11600	0.002581	0.303079	0.838235	17
VV93_v1c06340	3.81E-04	0.309859	0.830882	17

3.8. Clustering and modularity within V. vulnificus secretome

The global clustering coefficient for the V. vulnificus secretome graph is 0.365. In contrast, a randomly generated network created from the same set of vertices has a clustering coefficient of 0.146 ± 0.146 (z = −561.21, p < 0.0001) proving that the clustering in the secretome network was not the result of random chance. The secretome network of V. vulnificus contains ten clusters, ranked in size (number of proteins) and density (interconnectivity). Sixty-three (≈18%) of the network’s 575 total proteins are found in clusters (Figure 6). Based on terminology from Gene Ontology and STRING, functional descriptions are used. Cluster 1 (C1) has no specific enrichment found for the set of secretory proteins identified. Most of these proteins are uncharacterized, with only few associated with cyclic-di-gmp binding (VV93_v1c00420, VV93_v1c08230, VV93_v1c09620, VV93_v1c13620, VV93_v1c14430, VV93_v1c15230, VV93_v1c21290, VV93_v1c22970, VV93_v1c28640, VV93_v1c34010, VV93_v1c37740, VV93_v1c41410), sporulation-like domain, and lipoate-protein ligase activity (VV93_v1c00420, VV93_v1c08230, VV93_v1c13620, VV93_v1c15230, VV93_v1c22970, VV93_v1c28640, VV93_v1c34010, VV93_v1c37740), AsmA, and leucine-rich repeat domain superfamily (VV93_v1c08230, VV93_v1c34010, VV93_v1c37740), thiamine kinase, and ycfl-like superfamily (VV93_v1c20800 and VV93_v1c20810). The majority of the proteins in cluster 2 (C2) are strongly associated with amino acid ­(VV93_v1c32290, VV93_v1c37060, VV93_v1c42090, VV93_v1c42110, VV93_v1c42780) and nitrogen compound transport (VV93_v1c30270, VV93_v1c32290, VV93_v1c37060, VV93_v1c42090, VV93_v1c42110, VV93_v1c42780). The Pfam domain enriched includes Bacterial extracellular solute-binding proteins, family 3 (VV93_v1c17920, VV93_v1c32290, VV93_v1c37060, VV93_v1c42090, VV93_v1c42110, VV93_v1c42780).

Figure 6. The secretome network of Vibrio vulnificus E4010 contains 10 clusters, ranked in size (number of proteins) and density (interconnectivity). Sixty-three (about 18%) of the network’s 575 total proteins are found in clusters.

Cluster C3 includes the secretory proteins associated with the biological process like peptidoglycan metabolic process (GO:0000270), cell wall organization or biogenesis (GO:0071554), cell wall organization (GO:0071555), cellular component organization or biogenesis (GO:0071840), carbohydrate derivative metabolic process (GO:1901135), cellular component organization (GO:0016043) and peptidoglycan catabolic process (GO:0009253). The molecular functions enriched by proteins of C3 cluster include peptidoglycan muralytic activity (GO:0061783) and lytic transglycosylase activity (GO:0008933). Further, it was observed that the Transglycosylase SLT domain (PF01464), Pfam domain were enriched by the secretory proteins of C3 cluster like VV93_v1c06430, mltF, and VV93_v1c22540. Finally, cluster C4 proteins metQ, VV93_v1c15790, PotD, speE, and VV93_v1c32680 were typically enriched in the biological process like Polyamine transport (GO:0015846) and Nitrogen compound transport (GO:0071705). These proteins were also enriched with molecular function Polyamine binding (GO:0019808). Further, four cluster proteins (metQ, PotD, VV93_v1c15790, VV93_v1c32680) were enriched with KEGG pathway ABC transporters (vvl02010).

4. Discussion

In this study, V. vulnificus E4010 isolated from seafood harvested from Mangaluru coast of southern India was sequenced to better characterize genetic attributes involved in pathogenesis and virulence utilizing systems biology approaches. The assembled genome had a length of 4,875,404 bp and a GC content of 46%. The comparison of the assembled genome of V. vulnificus E4010 to the reference genomes V. vulnificus CMCP6 and V. vulnificus YJ016 demonstrated >98% pairwise ANI and >89% dDDH values, thereby supporting the taxonomic validity of the sequenced data. Recently, the four-lineage classification of V. vulnificus was proposed based on MLST phylogeny by Bisharat et al. (2020). According to the proposed classification, V. vulnificus has a common ancestor that possibly originated in Asia, with ­lineages 1 and 2 consisting of C and E-clade, respectively. Lineage 3 comprised mostly of strains from Israel and East Asia, and lineage 4 of strains from Western Europe. MLST analysis of the V. vulnificus E4010 assigned an unique sequence type ST-570 by the curator based on distinct allelic profiles. The newly identified ST-570 was nearest to ST-409, an ST identified in Vv063, an environmental isolate from China (https://pubmlst.org/). The evolutionary relationships inferred based on the MLST phylogeny confirmed that V. vulnificus E4010, though isolated from shellfish (hence considered an environmental isolate) is more closely related to the isolates obtained from clinical cases. Our findings are consistent with a previous report which stated that the C-genotype of V. vulnificus strains does not always correlate with their isolation source (Guerrero et al. 2019).

V. vulnificus possesses a diverse set of virulence factors that enable the pathogen to infect and cause pathological changes in the infected host (Oliver 2013). The ability of V. vulnificus to adhere to the host and then infiltrate the tissue barrier is critical for establishing infection and systemic disease. Studies have shown that most of the highly pathogenic strains of V. vulnificus possess pilus, and their transcriptional level increases significantly during their adhesion to the host (Paranjpye and Strom 2005; D’Souza et al. 2020). Their study further showed that the inactivation of pilus reduces their ability to form biofilms and its adherence to Hep-2 cells. The attachment of the isolate of interest, V. vulnificus E4010, to human cells has previously been studied and its adherence capacity was comparable to clinical strains, resulting in F-actin disintegration and cytotoxicity (D’Souza et al. 2020). The adherence of the pathogen is also facilitated by a fibronectin-binding outer membrane protein (OmpU) (Goo et al. 2006). Flagellar-based motility and chemotaxis were identified as one of the contributing factors to establish infection by V. vulnificus. The gene’s coding for two-component system, the sensor kinase and response regulators would allow the bacterium to sense and respond to specific environmental cues by regulating the direction of flagellar rotation (Wadhams and Armitage 2004). The ability to tolerate and/or resist gastric acidity and counteracting the initial resistance is an essential virulence determinant vital for V. vulnificus to establish and cause food-borne infections in humans.

The amino acid decarboxylase system involved in acid stress response coded by cadBA operon, and its transcriptional regulator (aphB) is essential to counteract external acidification (Rhee et al. 2006). The previous study by D’Souza et al. (2019) demonstrated the acid tolerance response in V. vulnificus E4010, which was comparable to the acid stress response of clinical strains. It is well-known that upon colonization, V. vulnificus evades epithelial layers to enter the bloodstream and its survival in the systemic circulation depends on its ability to tolerate complement proteins and phagocytosis (Oliver 2013). The virulence factors attributed to phagocytic activity like capsular polysaccharides present in the isolate E4010 are found to be responsible for conferring an opaque colony morphology (D’Souza et al. 2020), whose expression is a known mechanism among pathogenic bacteria to overcome phagocytosis and complement-mediated cell lysis (Pettis and Mukerji 2020). Starks et al. (2006) demonstrated that clinically derived isolates of V. vulnificus are more resistant towards serum compliment proteins compared to their environmental counterparts and are also enabled to resist the phagocytosis. D’Souza et al. (2020) showed that seafood isolated V. vulnificus had a limited serum resistance compared to clinical isolate. It is an interesting observation that the genes responsible for serum resistance were intact in the genome of seafood isolate similar to a blood isolate YJ016 (Carda-Diéguez et al. 2018) yet the isolate was unable to survive in the presence of serum. However, the survival of seafood isolated V. vulnificus in the heat inactivated serum (D’Souza et al. 2020) indicates that the growth of V. vulnificus E4010 could be inhibited by heat-labile antimicrobial molecules that are present in the serum. Further studies are essential to understand the failure of seafood isolates to resist serum when the genes responsible for resistance are present in their genome.

D’Souza et al. (2020) also showed V. vulnificus E4010 exhibited a high level of cytotoxicity and actin perturbation post-infection with HeLa cells. This experimental evidence is further supported by the presence of vvhA and MARTX genes in the genome of E4010, being key virulence factors responsible for cytopathic and cytotoxic effects (Satchell 2015). Infection by V. vulnificus is commonly seen in patients with chronic liver disease, immunodeficiencies, and/or hematological disorders characterized by elevated levels of iron (Horseman and Surani 2011). The ability to acquire iron from the host facilitates the infection by enhancing its growth as well as reducing the activity of neutrophils. The analysis provides an insight into the presence of several structural and extracellular components possibly involved in the pathogenesis. According to the results obtained in this study in combination with previous experiments to elucidate the virulence potential, environmental isolates may have the attributes that induce the disease process; however, the precise mechanism of action remains to be elucidated.

Previous research revealed the presence of strain-specific genomic islands, in contrast to V. cholerae and V. parahaemolyticus, where known pathogenicity islands are said to be present in virulent strains (Quirke et al. 2006). Pathogenicity islands in V. vulnificus isolates show a high level of genome plasticity (Pérez-Duque et al. 2021). Further analysis of these putative islands could help understand the pathogenicity and genome plasticity among the V. vulnificus compared to other vibrios. The Multifunctional autoprocessing repeats-in-toxin (MARTX) is one of the prominent effector delivery platforms in the Gram-negative bacteria (Kim 2018). Studies have shown that rtxA1 gene encoding MARTX toxin in V. vulnificus has enhanced the pathogenic potential of the bacterium towards both human and animal hosts (Lee et al. 2007, 2013). Though this gene was first identified in V. cholerae, it has been identified in other ­vibrio members and has been observed to show differences in effector domains due to genetic variations including in the V. vulnificus (Kwak et al. 2011; Kim 2018). The MARTX domain identified in sequenced strain V. vulnificus E4010, the cysteine protease domain (CPD) is a conserved domain that involves in the processing of inter effector domain regions in the host cell. A Rho-GTPase—Inactivation Domain (RID) depolymerizes the actin filaments indirectly by inactivating the Rho—family GTPases, a cytoskeleton disrupting Actin Cross-Linking Domain (ACD) by covalently cross-linking the monomeric actins in cytosol and Alpha/Beta Hydrolase Domain (ABH) that inhibits the endosomal/autophagic trafficking which could also affect the cell signaling (Kim 2018).

Members of the genus Vibrio are mostly susceptible to antibiotics of veterinary and clinical importance (Oliver 2013). Treatment of V. vulnificus infections with antibiotics as soon as possible has been shown to reduce the organism’s lethality (Kitamura et al. 2016; Heng et al. 2017). The anti­biotics of choice for V. vulnificus infection in adults are doxycycline and ceftazidime, while the CDC ­recommends trimethoprim, sulfamethoxazole (co-­trimoxazole), and an aminoglycoside for children. The Infectious Diseases Society of America (IDSA) recommends a combination of doxycycline and ceftriaxone or cefotaxime (Horseman and Surani 2011; Emamifar et al. 2015). This treatment regimen is significantly effective in treating V. vulnificus induced septicemia, according to a study conducted to determine its efficacy (Oliver 2013; Huang et al. 2016; Trinh et al. 2017). Apart from the combination of doxycycline and ceftazidime, ciprofloxacin and cephalosporin, or tigecycline (TGC), was also found to be a better potential therapeutic option (Yu et al. 2017; Kim et al. 2019). Cephalosporin use, on the other hand, has been discouraged due to rising antibiotic resistance (Lee et al. 2019). Since the resistant pattern of these antibiotics has shown different profiles in different countries, clinicians and microbiologists have been tasked with deciding on treatment protocols (Heng et al. 2017). In the last few decades, the indiscriminate and increased use of antibiotics has resulted in the evolution of resistance among vibrios to penicillins, cephems, aminoglycosides, carbapenems, tetracyclines, quinolones, polymyxins, and monobactams (Heng et al. 2017). The resistance genes identified in V. vulnificus E4010 could provide resistance to several antibiotic classes, such as aminocoumarins, aminoglycosides, macrolides, nitroimidazoles, quinolones, tetracyclines, and peptide-based antibiotics (Van Hoek et al. 2011). The katG resistance gene-mediated catalase-peroxidase activity modifies the antibiotics thereby reducing the susceptibility of the bacterium (Loewen et al. 2018). gyrA, gyrB genes are responsible for quinolone resistance which occur in a specific region of protein known as quinolone resistance-determining region (QRDR) in V. vulnificus (Roig et al. 2009). Similarly, the qnrB family resistance genes protect DNA gyrase and topoisomerase IV from quinolone inhibition (Gil-Marqués et al. 2021). A previous study on antibiotic resistance in V. vulnificus isolated from India’s west coast found multidrug resistance to ampicillin, amoxicillin, carbenicillin, ceftazidime, cephalothin, colistin, and streptomycin (Vaseeharan et al. 2005). However, pathogens that have ARGs but do not show phenotypic resistance to the antibiotics could be due to silent resistance genes. Most commonly the resistance patterns are often focused on strains exhibiting the phenotypic resistance. However, the presence of silent genes makes it complicated to assess the resistance capability of the bacterium as these genes could revert back to phenotypic expression during suitable conditions in due course of time (Stasiak et al. 2021; Shetty et al. 2023). The systems biology approach to characterize the V. vulnificus E4010 genome revealed that various secretory proteins from V. vulnificus are associated with major pathways like ABC transporters, two-component systems, quorum sensing, biofilm formation, cationic antimicrobial peptide (CAMP) resistance and others that play a critical role in the pathogenesis of the V. vulnificus.

5. Conclusion

Finding the virulence potential of seafood isolate of V. vulnificus is of importance from the public health concern. The present study helps in understanding the genotypic profile of the isolate and gives a viewpoint on its pathogenicity. The secreted protein pathway analysis showed that several V. vulnificus secretory proteins are linked to essential pathways including ABC transporters, two-component systems, quorum sensing, biofilm formation, and cationic antimicrobial peptide (CAMP) resistance, among others, and are necessary for the pathogenesis of the V. vulnificus. The disparity in the genotypic and phenotypic attributes related to fitness and antimicrobial resistance in the environmental isolates of V. vulnificus invites further investigation in this area. A thorough characterization of the various genotypes and biotypes of V. vulnificus is essential for understanding the evolution and diversity at the genomic level.

Acknowledgements

The authors thank Nitte (Deemed to be University) for providing computational infrastructural facilities for the bioinformatics support. The authors also thank the data submitters and curators of the public databases from which the data has been retrieved for analysis purposes.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The draft genome of Vibrio vulnificus E4010 was earlier deposited in DDBJ/ENA/GenBank and given the accession WRPB00000000. The version described in this paper is version WRPB01000000. The new MLST profiles of V. vulnificus E4010 (pubMLST id: 775) have also been submitted to the pubMLST database (https://pubmlst.org).

Additional information

Funding

This work was supported by the Science and Engineering Research Board, Department of Science and Technology, Government of India (ECR/2017/000721), Indian Council of Medical Research, Government of India (FBO/Adhoc/1/2022-ECD-IIe-136272), Nitte (Deemed to be University) Faculty Startup Research Grant (NUFR1/2016/20-04) and the Swedish Research Council (VR 2016-05655).