Comparative Genomics Analysis of Vibrio anguillarum Isolated from Lumpfish (Cyclopterus lumpus) in Newfoundland Reveal Novel Chromosomal Organizations

Vibrio anguillarum is a Gram-negative marine pathogen causative agent of vibriosis in a wide range of hosts, including invertebrates and teleosts. Lumpfish (Cyclopterus lumpus), a native fish of the North Atlantic Ocean, is utilized as cleaner fish to control sea lice (Lepeophtheirus salmonis) infestations in the Atlantic salmon (Salmo salar) aquaculture industry. V. anguillarum is one of the most frequent bacterial pathogens affecting lumpfish. Here, we described the phenotype and genomic characteristics of V. anguillarum strain J360 isolated from infected cultured lumpfish in Newfoundland, Canada. Koch’s postulates determined in naïve lumpfish showed lethal acute vibriosis in lumpfish. The V. anguillarum J360 genome was shown to be composed of two chromosomes and two plasmids with a total genome size of 4.56 Mb with 44.85% G + C content. Phylogenetic and comparative analyses showed that V. anguillarum J360 is closely related to V. anguillarum strain VIB43, isolated in Scotland, with a 99.8% genome identity. Differences in the genomic organization were identified and associated with insertion sequence elements (ISs). Additionally, V. anguillarum J360 does not possess a pJM1-like plasmid, typically present in virulent isolates from the Pacific Ocean, suggesting that acquisition of this extrachromosomal element and the virulence of V. anguillarum J360 or other Atlantic isolates could increase.


Introduction
Vibrio spp. is naturally ubiquitous in aquatic and marine environments [1]. Some members of this genus can cause infections in humans after exposition to contaminated water, such as Vibrio cholerae, the causative agent of cholera, and after consumption of raw contaminated seafood, such as V. parahaemolitycus, V. alginolyticus, and V. vulnificus [1][2][3]. Other members of Vibrio spp. such as V. splendidus, V. nereis, V. harveyi, V. damsela, V. tubiashi, and V. anguillarum are pathogens of aquatic organisms including cultured fish species [4,5]. V. anguillarum is a Gram-negative marine pathogen,

V. anguillarum Isolation
V. anguillarum strain J360 was isolated from the head kidney of infected cultured lumpfish in Newfoundland, Canada. Fish with classic vibriosis symptoms were netted and immediately euthanized with an overdose of MS222 (400 mg/L) (Syndel Laboratories, Nanaimo, BC, Canada). Tissue samples (e.g., head kidney, liver, and spleen) were collected and placed into sterile homogenized bags (Nasco whirl-pak ® , Fort Atkinson, WI, USA). The infected tissues were weighed and homogenized in PBS up to a final volume of 1 mL. One hundred microliters of the homogenized tissue suspension were plated onto Trypticase soy agar (TSA) plates and incubated at 15 • C for 48 h. Isolated colonies were selected and purified for further analysis. Bacteria stock was preserved at −80 • C in 10% glycerol and 1% peptone solution.

Matrix-Assisted Laser Desorption/Ionization-Time-of-Flight (MALDI-TOF) Mass Spectrometry and Serotypification Analysis
Serotypification and MALDI-TOF were conducted commercially at the University of Prince Edward Island, Canada. The MALDI Biotyper RTC was conducted according to the MALDI Biotyper 3.1 user manual and parameter settings as previously published [23].

Biochemical, Enzymatic, and Physiological Characterization
V. angillarum J360 growth curves were conducted in triplicates at 15 • C, 28 • C, and under rich and iron-limited conditions according to established protocols [24]. Briefly, a single colony of V. angillarum J360 was inoculated in 3 mL of TSB and incubated in a roller for 24 h at 15 • C. An inoculum of 300 µL of cells at mid-log phase (OD 600 ≈ 0.7) was added to 30 mL of fresh TSB into 250 mL flasks and incubated for 48 h at 15 • C and 28 • C with aeration (180 rpm) in an orbital shaker. Bacterial growth was monitored spectrophotometrically until OD 600 ≈ 2.0 ± 0.3 (≈8 × 10 8 CFU/mL) using a Genesys 10 UV spectrophotometer (Thermo Spectronic, Thermo Fischer Scientific, USA). Growth curves under iron-limited conditions were determined using three different 2,2-dipyridyl concentrations (100, 150, 250 µM). Controls consisted of nonsupplemented TSB. The doubling time was estimated using the optical density (OD) values, g = b − B, where b is the OD value at the end of the time interval and B is the OD value at the beginning of the time interval.

Siderophores Synthesis
V. anguillarum J360 was grown under conditions previously described. Bacterial cells were harvested at mid-log phase, at an optical density (OD 600 ) of ≈0.7 (≈4.1 × 10 8 CFU/mL), washed three times with phosphate-buffered saline (PBS; 136 mM NaCl, 2.7 mM KCl, 10.1 mM Na 2 HPO 4 , 1.5 mM KH 2 PO 4 (pH 7.2)) [22] at 4200× g for 10 min at room temperature, and resuspended in 1 mL of PBS. An inoculum of 300 µL of the bacterial suspension was added to 3 mL TSB medium and TSB medium supplemented with 100 µM of FeCl 3 or 100 µM of 2,2-dipyridyl. V. anguillarum J360 was grown at 15 • C for 24 h with aeriation. Following the incubation period, the cells were harvested at mid-log phase, washed twice with PBS, and resuspended in 100 µL of PBS. Five microliters of the concentrated bacterial pellet was inoculated onto CAS agar plates [21] and incubated at 15 • C and 28 • C for 48 h.

Infection Assay in Lumpfish
Naive lumpfish (weight ≈ 55 g) were transferred from the JBARB to the AQ3 biocontainment Cold-Ocean Deep-Sea Research Facility (CDRF) for infection assays. Fish were separated into three 500 L tanks containing 60 fish per dose and acclimated for 2 weeks under previously described conditions. The infection procedures were conducted according to established protocols [27]. Briefly, fish were anesthetized with 40 mg of MS222 (Syndel Laboratories, BC, Canada) per liter of sea water and intraperitoneally infected with 100 µL of 10 6 , or 10 7 CFU per dose of V. anguillarum J360. The Control group (n = 60) was mock-injected with PBS. Mortality was monitored daily until 30 days post-infection. Samples of liver, spleen, and head kidney were taken from moribund fish to re-isolate the pathogen.

DNA Extraction and Sequencing
V. anguillarum J360 was grown under conditions previously described. Bacterial cells were harvested at mid-log phase, at an optical density OD 600 of ≈0.7 (≈4.1 × 10 8 CFU/mL) and washed three times with PBS. DNA extraction was conducted using a Wizard DNA extraction High Molecular Weight Kit (Promega, Madison, WI, USA). The DNA integrity and purity were evaluated by gel electrophoresis (agarose gel 0.8%) [28] and spectrophotometry (Genova-Nano Spectrophotometer, Jenway, UK). Libraries and sequencing were conducted commercially at Genome Quebec (Montreal, QC, Canada) using PacBio RS II and Miseq Illumina sequencers.

Genome Assembly, Annotation, and Mapping
PacBio reads were assembled at Genome Quebec using Celera Assembler (August 2013 version). Annotation was completed using the Rapid Annotation Subsystem Technology pipeline (RAST) (http://rast.nmpdr.org/) [29]. The two V. anguillarum J360 chromosomes and large plasmid were submitted to the National Center for Biotechnology Information (NCBI) and re-annotated using the NCBI Prokaryotic Genome Annotation Pipeline.
To detect small plasmids, the Illumina reads were trimmed using CLC Genomics Workbench v20.0 (CLC Bio) and examined for quality using FastQC version 12 [30]. High-quality Illumina reads were assembled using the CLC Genomics Workbench de novo tool and aligned to the reference V. anguillarum J360 chromosomes and large plasmid using the genome finishing module tools with default parameters. Illumina sequences that did not align with the chromosomes or large plasmid were analyzed and annotated using the previously described methods. The V. anguillarum J360 whole genome was mapped by using DNA plotter software [31].

Whole Genome Comparison and Phylogeny Analysis
The genomes utilized are listed in Table 1. Whole genomes were aligned to calculate the average nucleotide identity (ANI) using the CLC Genomic workbench v20 (CLC Bio) whole genome analysis tool with default parameters (Min. initial seed length = 15; Allow mismatches = yes; Min. alignment block = 100). A minimum similarity (0.8) and a minimum length (0.8) were used as parameters for CDS identity. A comparative heat map was constructed using the heat map tool with default parameters (Euclidean distance method and complete cluster linkages). Phylogenetic analysis was performed in two different software, CLC Genomic workbench v20.0 and MEGA X [32], with the same parameters for robustness comparison purposes, using the extracted alignment from the ANI analysis. Evolutionary history was calculated using the neighbor-joining method [33] with a bootstrap consensus of 500 replicates, and evolutionary distance was computed using the Jukes-Cantor method [34]. Photobacterium damselae 91-197 (AP018045/6) chromosomes were utilized as outgroups [35]. Whole genome dot plots between closely related V. anguillarum strains were constructed using the CLC Genomic workbench v20.0, a whole genome analysis tool to visualize and further analyze genomic differences. Comparative alignment analysis represents homologous regions, translocations, and inversions within the two strain genomes for chromosome-I and for chromosome-II. Homologous regions were identified as locally colinear blocks (LCBs), which represent conserved regions that do not present rearrangements, and genomic gaps (GGs) were identified as unmatched regions. Analysis was performed using the progressive Mauve v20150206 [36].

Multi-Locus Sequence Analysis of V. anguillarum Housekeeping Genes
Multi-locus analysis (MLSA) was used to infer the phylogeny history of V. anguillarum strains using reference genes including 16S ribosomal RNA subunit (rrn), cell-division protein (ftsZ), glyceraldehyde-3-phosphate dehydrogenase (gapA), gyrase beta subunit (gyrB), rod shape-determining protein (mreB), uridine monophosphate (UMP) kinase or uridylate kinase (pyrH), recombinase A (recA), RNA polymerase alpha subunit (rpoA), and topoisomerase I (topA) gene sequences. Only genes from complete genomes were considered for the MLSA. Sequences were aligned using CLC Genomic workbench v20.0 (CLC Bio). Concatenation of locus sequences was made using Sequence Matrix software v1.7.8 [42]. Phylogenetic analysis was performed using the two software mentioned above with the same parameters. Evolutionary history was estimated using the neighbor-joining method [33] with a bootstrap consensus of 500 replicates, and evolutionary distance was computed using the Jukes-Cantor method [34]. The gene loci and accession numbers are listed in Table S1.

Comparative Analysis of V. aguillarum J360 Large Plasmid
A genomic comparison of virulent and nonvirulent plasmids between V. anguillaurm species was performed using the whole genome alignment tool of CLC Genomics workbench v20.0 with default parameters. Plasmids used in this analysis were: p292-VIB12 (CP023312); p15-VIB43 (CP023056); pVaM3 (CP006701); pJM1 (AY312585); p67vangNB10 (LK021128); and p65-ATCC 68554 (CP023210). Plasmids were aligned to calculate the ANI. A comparative heat map was constructed using the heat map tool with default parameters (Euclidean distance method and complete cluster linkages).

Statistical Analysis
Fish survival percentages were transformed to arc-sin ( √ survival rate ratio). One-way ANOVA was used to determine significance, followed by Tukey's post-hoc test, to determine significant differences (p < 0.05). All statistical analyses were performed using GraphPad Prism 7 (GraphPad Software, California, CA, USA).

Ethics Statement
All animal protocols required for this research were approved by the Institutional Animal Care Committee and the Biosafety Committee at Memorial University of Newfoundland (MUN). Animal assays were conducted under protocols #16-92-KG, #18-01-JS #18-03-JS, and biohazard license L-01, approved on 26-05-2020.

Phenotypic, Biochemical, and Enzymatic Characterization
V. anguillarum strain J360 was capable of growing in Tryptic soy broth (TSB) and Luria Bertani (LB) medium up to 30 • C ( Table 2). The V. anguillarum J360 doubling time in TSB supplemented with 2% NaCl at 15 • C was 2 h ( Figure 1A) and 1 h at 28 • C ( Figure 1B). V. anguillarum J360 did not grow at 37 • C, in TCBS selective media at 15 and 28 • C, or in the absence of NaCl. V. anguillarum J360 was shown to be motile and capable of synthesizing type I fimbria, oxidase, and catalase ( Table 2). The antibiogram analysis showed that V. anguillarum J360 is ampicillin-resistant and susceptible to tetracycline, oxytetracycline, sulfamethoxazole, chloramphenicol, colistin sulphate, oxalinic acid, and O-129 (Table 2). Table 2. Phenotypic characteristics of V. anguillarum J360.

Characteristic
Vibrio anguillarum J360 Growth at:    V. anguillarum J360 growth under iron-limited conditions was evaluated at 15 • C on TSB media with different 2,2-dipyridyl concentrations (100, 150, 250 µM). V. anguillarum J360 was able to grow in the presence of high concentrations of 2,2-dipyridyl. The doubling time at 100 µM of 2,2-dipyridyl was 3 h ( Figure 1C); however, it was increased to 4 h (OD 600 ≈ 0.7) and 5 h (OD 600 ≈ 0.2) at 150 and 250 µM of 2,2-dipyridyl, respectively ( Figure 1C). No siderophores secretion was observed in the cells grown under iron-enriched conditions. Additionally, there was no significant differences in the size of the halo for siderophores secretion between nonsupplemented TSB and supplemented TSB with 100 µM of 2,2-dipyridyl at 28 • C ( Figure 1D). Nonetheless, a small difference in the halo increased diameter was observed for siderophores secretion under iron-limited conditions at 15 • C ( Figure 1D). Hemolytic activity was evaluated on blood agar plates at 28 • C and 15 • C. V. anguillarum J360 hemolytic activity was observed only at 28 • C ( Figure 1E).

MALDI-TOF and Agglutination Analysis
The MALDI-TOF mass spectrometry score for V. anguillarum was 1.96, indicating that there was a low confidence for identification. The V. anguillarum agglutination test was positive for the O2 serotype and negative for the O1 serotype.

Infection Assay in Specific Pathogen-Free Lumpfish
Naïve cultured lumpfish (≈55 g) were intraperitoneally (ip) infected with V. anguillarum J360 to evaluate its virulence (Figure 2A,B). Two groups of 60 lumpfish were injected with 1 × 10 6 and 1 × 10 7 CFU/dose, respectively. The control group was mock-infected with PBS, and mortality was monitored until 30 days post-infection (dpi). Mortality began at 2 dpi and reached 100% in both doses at 10 dpi ( Figure 2C). Vibriosis clinical signs were observed at 5 dpi, including hemorrhage over the lateral lines, dorsal and/or caudal fins, ventral sucker, vent, mouth, and the operculum. Additionally, infected fish exhibited exophthalmia ( Figure 2B).

V. anguillarum J360 Genome Sequencing and Annotation
V. anguillarum genomic DNA sequenced by PacBio resulted in five contiguous sequences (contigs). The larger contigs corresponded to circularized chromosome-I (3,320,860 bp), chromosome-II (1,171,281 bp), and a large plasmid pVaJ360-I (56,630 bp) with coverage assemblies of 211, 167, and 36 times, respectively. The plasmid profile of V. anguillarum J360 indicated that there was also a small plasmid ( Figure 3A). Using Illumina reads, we were able to assemble pVaJ306-II (11,995 bp) with a coverage of 226 times. The V. anguillarum J360 genome was submitted to NCBI under the BioProject (PRJNA485045) and BioSample (SAMN09781303). The complete genome of V. angullarum J360 possesses two chromosomes [chromosome-I (NZ_CP034672) and chromosome-II (NZ_CP034673)], a large plasmid pVaJ360-I (NZ_CP034674), and a small plasmid pVaJ360-II (MT050454), and has an estimated total length of 4.55 Mb and a G + C content of 44.6% ( Figure 3B, Table 3). RAST pipeline annotation predicted a total of 441 subsystems and 3149 coding sequences (CDS) for chromosome-I;  Figure 3A). Using Illumina reads, we were able to assemble pVaJ306-II (11,995 bp) with a coverage of 226 times. The V. anguillarum J360 genome was submitted to NCBI under the BioProject (PRJNA485045) and BioSample (SAMN09781303). The complete genome of V. angullarum J360 possesses two chromosomes [chromosome-I (NZ_CP034672) and chromosome-II (NZ_CP034673)], a large plasmid pVaJ360-I (NZ_CP034674), and a small plasmid pVaJ360-II (MT050454), and has an estimated total length of 4.55 Mb and a G + C content of 44.6% ( Figure 3B, Table 3). RAST pipeline annotation predicted a total of 441 subsystems and 3149 coding sequences (CDS) for chromosome-I; a total of 88 subsystems and 1143 CDSs for chromosome-II; a total of 2 subsystems and 96 CDSs for the large plasmid pVaJ360-I; and a total of 24 CDS in a single subsystem and 1 total RNAs sequence for the small plasmid pVaJ360-II (Table 4). The NCBI Prokaryote Genome Annotation pipeline showed a total of 4371 genes predicted, a total of 10 (5S), 9 (16S), and 9 (23S) rRNAs, 105 tRNAs, and 4 non-coding RNAs(ncRNAs) for the whole genome (Table 5). , and small plasmid pVaJ360-II (~10 MDa) were mapped using DNA plotter. Blue bars represent forward genes; light blue bars represent reverse genes; green bars represent pseudogenes and miscellaneous features. Gold (high) and violet (low) represent G + C content percent; green (high G + C) and black (low G + C) represent skew.  Table 5. NCBI prokaryotic Genome Annotation pipeline V. anguillarum J360 genome annotation summary. , and small plasmid pVaJ360-II (~10 MDa) were mapped using DNA plotter. Blue bars represent forward genes; light blue bars represent reverse genes; green bars represent pseudogenes and miscellaneous features. Gold (high) and violet (low) represent G + C content percent; green (high G + C) and black (low G + C) represent skew.

Whole Genome Alignment, Phylogeny, and Synteny
Phylogenetic analysis of Vibrio spp. was performed using CLC Bio with only complete genome sequences (Table 1). Phylogenetic analysis of chromosome-I showed that there are three clusters with three or more strains, whereas V. tasmaniensis, V. parahaemolyticus, and V. campbellii clustered together as one and V. fluvialis clustered separately. By contrast, V. anguillarum species were represented by two clusters plus V. anguillarum strains CNEVA, MHK3, VIB12, NB10, and 8-9-116 that clustered separately. V. anguillarum J360 was closely related to V. anguillarum VIB43 isolated from Scotland, UK ( Figure 4A, Table 1). Phylogenetic analysis of chromosome-II indicated four clusters, whereas non-V. anguillarum species clustered together, and V. fluvialis clustered separately ( Figure 4B). Similar to chromosome-I, V. anguillarum J360 chromosome-II was closely related to V. anguillarum VIB43 ( Figure 4B). The same results were observed in the phylogenetic analysis using MEGA X for chromosome-I ( Figure S1A) and chromosome-II ( Figure S1B). The heat map indicated that there was a high identity between V. anguillarum J360 and V. anguillarum VIB43 alignments of chromosome-I ( Figure 4C) and chromosome-II ( Figure 4D), respectively. The ANI analysis for the whole genome alignment indicates a 99.95% for chromosome-I ( Figure S2A) and 99.93% for chromosome-II of genome identity ( Figure S2B) that support previous observed results. The dot plot visualization matches between V. anguillarum J360, and the closest related strain, V. anguillarum VIB43, showed that there was a high similarity within the genome. However, two inversion events and genomic gaps (GGs) were identified ( Figure 5A,B). The whole genome alignment identified several locally collinear blocks (LCBs), described as conserved segments free from genomic rearrangements [36]. The comparative alignment analysis of each chromosome showed five LCBs in chromosome-I ( Figure 5C) and two LCBs in chromosome-II ( Figure 5D). Additionally, the LCBs identified in both chromosomes are conserved, and in agreement with the reversion events and GGs identified in the dot plot analysis ( Figure 5). identified ( Figure 5A,B). The whole genome alignment identified several locally collinear blocks (LCBs), described as conserved segments free from genomic rearrangements [36]. The comparative alignment analysis of each chromosome showed five LCBs in chromosome-I ( Figure 5C) and two LCBs in chromosome-II ( Figure 5D). Additionally, the LCBs identified in both chromosomes are conserved, and in agreement with the reversion events and GGs identified in the dot plot analysis ( Figure 5).    Table 1. Photobacterium damselae 91-197 as an outgroup. Analysis was performed using CLC workbench v.20 (CLC Bio). Black arrows represent V. anguillarum J360 genome and V. anguillarum VIB43 as the closest related strain. Light blue square represents the percentage of identity between strains.

Multi-Locus Sequence Analysis (MLSA) and Phylogeny
We also utilized MLSA to contrast these results with the whole genome analyses. MLSA was computed using the nine housekeeping genes listed in Table S1. Gene sequences were aligned, concatenated, and analyzed. The phylogenetic analysis performed in CLC Bio showed that there were five clusters with at least two strains, whereas V. anguillarum J360 clustered alone. This analysis indicated that V. anguillarum J360 is closely related to V. anguillarum NB10, which clustered with V. campbelli and V. tasmaniensis, and is distantly related to V. anguillarum 775 and V. anguillarum M3 ( Figure S3A). The same results were observed using MEGA X software ( Figure S3B).
Metalloprotease coding genes were found in both chromosomes, including a CPBP family intermembrane metalloprotease (DYL72_00295) and a sprT family zinc-dependent metalloprotease (DYL72_09830) gene, and three metalloprotease genes (pmbA, tldD, and ftsH) are present in chromosome-I. Additionally, a M6 family domain that possesses metallopeptidase activity was found in chromosome-I. Motility genes were found in chromosome-I, except for flagellar motor brake proteins that were found in chromosome-II. Chemotaxis genes were found on both chromosomes, including 17 genes in chromosome-I and 16 genes in chromosome-II that encode for methyl-accepting chemotaxis proteins. Four copies of cheV and five copies of cheW were also found, and these are involved in phosphorylation-dependent excitation and methylation-dependent adaptation, respectively. We identified five type IV pilus-associated genes in chromosome-I, and type VI secretion system-related genes in both chromosomes, including six copies of tssI, and two operons encoding tssBCEFG and tssKJHFE2. Two secretion system families were identified, including the Hcp type VI secretion system family effector and DotU family type IV and type VI secretion system-related proteins. A single quorum-sensing-associated gene was found in chromosome-II, which is a quorum-sensing autoinducer synthase. Transcriptional regulators were found in chromosome-I, such as lysR, cysB, nhaR, and hfq, and luxR was present in both chromosomes.

Genomic Islands (GIs)
Twenty-one putative GIs were identified within the chromosomes; 15 GIs in chromosome-I and 6 GIs in chromosome-II, respectively ( Figure 6; Supplementary file 1 and 2). The GIs size ranged from 5 to 73.4 kb with a total of 724 genes. The largest genomic island (GI15) (Figure 6) consisted of 147 genes (predicted by at least one of the three software used, see Method section), and was flanked by two site-specific integrase genes and a zinc ribbon-domain protein.  Genes encoding for integrases, porins, transposases, and iron transport were found among the GIs. Genes such as phosphonate C-P lyases, associated with the cleavage of carbon-phosphorus compounds for organic phosphorus reservoir in marine bacteria [44], zinc ribbon domain proteins, AbrB/MazE/SpoVT DNA-binding domains, a ParA domain, for a GCN5-related N-acetyltransferase (GNAT) associated with regulatory post-transcriptional acetylation [45], two site-specific integrases, and a MasF transcriptional regulator related to toxin/anti-toxin systems (Supplementary file 1) were identified in these GIs. GI5 is the smallest genomic island identified in V. anguillarum J360 (Figure 6), which possesses five unique genes that encode for acetyltransferase, acyltransferase, formyltransferase, asparagine synthase, and one hypothetical protein (Supplementary File 1). Genes encoding for integrases, porins, transposases, and iron transport were found among the GIs. Genes such as phosphonate C-P lyases, associated with the cleavage of carbon-phosphorus compounds for organic phosphorus reservoir in marine bacteria [44], zinc ribbon domain proteins, AbrB/MazE/SpoVT DNA-binding domains, a ParA domain, for a GCN5-related N-acetyltransferase (GNAT) associated with regulatory post-transcriptional acetylation [45], two site-specific integrases, and a MasF transcriptional regulator related to toxin/anti-toxin systems (Supplementary file 1) were identified in these GIs. GI5 is the smallest genomic island identified in V. anguillarum J360 (Figure 6), which possesses five unique genes that encode for acetyltransferase, acyltransferase, formyltransferase, asparagine synthase, and one hypothetical protein (Supplementary File 1).

V. anguillarum Large Plasmids Analysis
The plasmid profile indicates that V. anguillarum J360 harbors one large plasmid and one small plasmid ( Figure 3A). The large plasmid pVaJ360-I (~60 kb) has genes that encode for integrases, DNA-binding proteins, peptidases, site-specific integrases, resolvase, mobile elements, and pro-phages. Comparative analysis showed that pVaJ360-I is not related to V. anguillarum virulent plasmids such as pJM1 (strain 775); p65 (strain ATCC-68554); pJM1-like plasmid, p67 (strain NB10); and p15 (strain VIB43) ( Figure S5A). The ANI analysis of these plasmids demonstrated that pVaJ360-I does not have a percentage of identity with pJM1 or pJM1-like plasmids ( Figure S5B), suggesting that pVaJ360 is not a virulent plasmid.
The small plasmid pVaJ360-II (~12 kb) has only 10 CDSs that encode for hypothetical proteins, transposases, mobile elements, a transcriptional regulator LysR, a tRNA Glu, and additionally 14 miscellaneous features. BLASTn analysis indicates that this plasmid has no similarities with the pJM1 or pJM1-like plasmids described above.

Discussion
Lumpfish across hatcheries and deployment sites frequently show signs of systemic bacterial infection, including skin lesions, gill hemorrhages, and bacterial aggregations in lymphoid organs (i.e., spleen, liver, head kidney) [10]. In the United Kingdom, Iceland, and Norway, several bacterial outbreaks have been reported in lumpfish hatcheries and at cage sites, and the most frequent pathogen detected is V. anguillarum [46]. Thus, it is not surprising that this pathogen was found to be present in Atlantic Canada.
V. anguillarum serotypes O1, O2, and O3 are the most prevalent strains among the 23 serotypes currently described [18,19]. Agglutination assays indicated that V. anguillarum J360 is O2, similar to other V. anguillarum strains isolated from lumpfish infections in the North Atlantic [46]. The biochemical profile obtained using API20NE showed that V. anguillarum J360 was able to reduce sugars, urea, and produce indole, suggesting a 99% possibility for V. fluvialis (Table S3). Although the biochemical profile did not identify V. anguillarum J360, its phenotypic characterization is consistent with other V. anguillarum isolates [25,26], except that V. anguillarum J360 was positive for urease (Table S4). This result is coincident with the presence of the urease encoding gene in chromosome-II (DYL2_19555). In addition, V. anguillarum J360 was not able to grow in TCBS-selective media, suggesting susceptibility to bile salts or a poor adaptation to the culture medium (Table 2) [47].
V. anguillarum J360 showed a thermo-inducible α-hemolysin activity at 28 • C, but no hemolytic activity was observed at temperatures below 15 • C ( Figure 1E). Hemolytic activity is an important virulence factor for V. anguillarum [48]. For instance, severe hemorrhages are a typical clinical sign of V. anguillarum infection in fish, including lumpfish [46]. Coincidently, in the current study, V. anguillarum J360 was shown to be highly virulent in lumpfish, and infected lumpfish displayed severe hemorrhagic symptoms at 5 dpi ( Figure 2B), similar to other strains described in Marco-Lopez et al. 2013. Koch's postulates for V. anguillarum J360 showed that lumpfish infected with 1 × 10 6 and 1 × 10 7 CFU/dose reached 100% mortality within 10 dpi at 10 • C ( Figure 2C). In addition, V. anguillarum was re-isolated from the spleen, liver, and head-kidney, confirming Koch's postulates. The original V. anguillarum outbreak in cultured lumpfish and the infection assays in the current study showed similar clinical signs ( Figure 2B). V. anguillarum hemolytic activity was evident during infection. However, the lumpfish is a cold-water fish typically cultured between 6 and 12 • C [49]. These results contradicted with the V. anguillarum thermo-inducible hemolytic activity at 28 • C ( Figure 1E). Perhaps, internal fish conditions (e.g., innate and adaptive immunity) triggered V. anguillarum hemolytic activity. These results suggest that its regulatory mechanisms need further analysis.
V. anguillarum J360 possesses two chromosomes, a large plasmid and a small plasmid ( Figure 3 and Table 4). Vibrio spp. and V. anguillarum genomes selected for phylogenetic and comparative genomics analysis possess two chromosomes and one large plasmid (Table 1). Typically, serotype O1 harbors a virulent plasmid, called pJM1 or pJM1-like, and serotypes O2 and O3 strains possess a nonvirulent large plasmid [50,51] or they do not harbor large plasmids [20,37]. V. anguillarum J360 is a O2 serotype that does not harbor a virulence plasmid ( Figure S5), suggesting that this strain could increase its virulence if a virulence plasmid is acquired.
The total genome size of V. anguillarum J360 is 4,561,566 bp (Table 4), which is larger than the currently available V. anguillarum genomes (Table 1). This may suggest that V. anguillarum J360 may have acquired genetic material through horizontal gene transfer and/or adapted to its lumpfish host or to environmental conditions in Atlantic Canada.
Phylogenetic distance based on the whole genome alignment analysis of chromosome-I and chromosome-II showed that V. anguillarum J360 is closely related to V. anguillarum VIB43, and distantly related to V. anguillarum VIB12 ( Figure 4A,B). Interestingly, V. anguillarum VIB43 and VIB12 were isolated from the same host species, sea bass (Dicentrarchus labrax), but from different locations. V. anguillarum VIB43 was isolated in Scotland and V. anguillarum VIB12 was isolated in the Mediterranean Sea [20]. The ANI analysis between V. anguillarum J360 and V. anguillarum VIB43 showed a 99.93% identity for chromosome-I and 99.95% for chromosome-II ( Figure 4C,D and Figure S2A,B), which suggest that these two strains share a common ancestor.
Additionally, the whole genome phylogenetic analysis showed that V. anguillarum J360 and VIB43 are not closely related to V. anguillarum 775, M3, and NB10, which is contradictory to previous MLSA studies that indicated that V. anguillarum VIB43 and VIB12 are closely related to those V. anguillarum strains [20,50]. The MLSA computed the phylogenetic stress based on concatenated conserved sequences [51,52]. In this study, we used nine conserved housekeeping genes, including 16S rRNA, ftsZ, gapA, gyrB, mreB, pyrH, recA, rpoA, and topA (Table S1). In contrast to the whole genome phylogenetic analysis, the MLSA showed that V. anguillarum J360 clusters alone, and the closest related strain is V. anguillarum NB10 isolated from rainbow trout (Oncorhynchus mykiss) in the Gulf of Bothnia, Sweden [6], which is distantly related to V. anguillarum 775 and M3 strains ( Figure S3, Table 1). V. anguillarum M3 was isolated from Japanese flounder (Paralichthys olivaceus) in Shandong, China, and classified as closely related to V. anguillarum 775 isolated from Coho salmon (Oncorhynchus kisutch) in the Pacific coast of USA [7,8]. By contrast, MLSA phylogenetic analysis based only on the 16S rRNA gene showed that V. anguillarum NB10 is closely related to V. anguillarum M3 [6].
In contrast to the MLSA analysis, the whole genome phylogenetic analysis of V. anguillarum strains is in concordance with the geographic origin of the strain isolation. For instance, according to the whole genome phylogenetic analysis, V. anguillarum J360 and V. anguillarum VIB43, both isolated in the North Atlantic Ocean, are highly related, and closely related to V. anguillarum strains 90-11-286, S3, and JLL237 isolated in Finland (Table 1; Figure 4). Actually, these geographic locations are a natural habitat for lumpfish populations [53].
In contrast to the whole genome phylogenetic analysis, the MLSA uses protein-coding genes, which evolved at a slow but constant rate, and it could have better resolution, especially at the species level [51,52]. However, the selection and number of coding genes, and alignment method, are variable for MLSA. We found that the phylogenetic analysis using whole genomes is more reliable than MLSA, and the robustness of our analyses showed to be consistent with two different software. In addition, whole genome analyses allow homologous regions, deletion, translocation, and inversion events to be identified.
Genome alignment and synteny analysis between V. anguillarum J360 and V. anguillarum VIB43 showed a high similarity within the chromosome sequences, but inversions and unmatched regions were also observed ( Figure 5A,B). This suggests that homologous recombination events play an important role in V. anguillarum evolution, perhaps influenced by insertion sequence (IS) elements such as chromosomal integrons or "super integrons" (SIs) describing Vibrio spp. and several Gram-negative species [20]. We determined that there are five LCBs in chromosome-I ( Figure 5C) and two LCBs in chromosome-II ( Figure 5D) shared between V. anguillarum J360 and VIB43. Further analysis revealed that all the LCBs present in chromosome-I have small inversion events ( Figure 5A,C). In addition, we found that LCBs-1 and -4 have genome gaps (GGs) or unmatching regions in both strains ( Figure 5A and Figure S4A). The GGs identified in LCB-1 of V. anguillarum J360 chromosome-I are not present in V. anguillarum VIB43 LCB-1 ( Figure S4A). These identified GGs possess several genes that encode for IS families transposases (IS66, ISL3, IS3, IS5) and site-specific integrases previously described in the V. anguillarum VIB43 genome, and with high similarity to V. anguillarum NB10, 775, and ATCC-6855 genomes [6,20]. We found that the unique GGs in LCB-1 of V. anguillarum J360 possess genes related to iron uptake and iron homeostasis ( Figure S4A), suggesting that these genes could be acquired by horizontal gene transfer.
In V. anguillarum J360 chromosome-II, two GGs were identified in LCB-1 and LCB-2 ( Figure S4B), and both GGs have an IS630-like element belonging to the ISVa15 transposase family. This IS630-like element is not present in V. anguillarum VIB43 LCBs. According to the description of Holm et al. (2018), ISVa3-ISVa20 are new insertion sequence (IS) elements in the V. anguillarum genomic repertory that are responsible for the divergency within strains. This suggests that V. anguillarum J360 and V. anguillarum VIB43 could be derived from a common ancestor and adapted to local environmental conditions and host species.
Pathogenesis-associated genes were found in both chromosomes, but chromosome-I harbors most of the virulence genes and their respective transcriptional regulators (Table 6). No virulence-associated genes were found in the large plasmid pVaJ360-I or in the small plasmid pVaJ360-II. The V. anguillarum virulence plasmid pJM1 possesses intrinsic virulence genes associated with iron uptake, like anguibactin biosynthesis (angA-angE, vabA-E) and anguibactin transport (fatA-fatD) [8,54,55]. By contrast, all the V. anguillarum J360 iron homeostasis-related genes are in its chromosomes. For instance, genes related to ferric-anguiobactin and siderophore uptake (e.g., exbB and exbD2, respectively) are present in chromosome-I. Comparative genomic analyses showed that the large virulent plasmids pJM1, P67-NB10, and p65-ATCC have high similarity ( Figure S5A,B). However, the large plasmid pVaJ360-I and the small plasmid pVaJ360-II of V. anguillarum J360 do not present similarity ( Figure S5A) or identity ( Figure S5B) with other reported plasmid sequences, nor possess virulence-associated genes. This suggests that V. anguillarum J360 does not harbor plasmids previously described in V. anguillarum, including serotype O2 strains [50].
Hemolysins are important virulence factors for V. anguillarum species, and contribute to its attachment, tissue colonization, and iron homeostasis, thus increasing its pathogenicity [56,57]. V. anguillarum J360 has four hemolysin genes, and these are consistent with the hemorrhagic clinical signs observed in lumpfish during the infection assays ( Figure 2B). In addition, a thermolabile hemolysin gene is present in chromosome-II, which can be related to the thermo-inducible hemolytic phenotype of V. anguillarum J360 ( Figure 1E).
The resistance of V. anguillarum J360 to ampicillin (Table 2) relates to the presence of a class C beta-lactamase (ampC)-encoding gene in chromosome-II. Metalloproteases such as pmbA, tldD, and ftsH genes, which are associated with carbon storage, hydrolysis of peptide bonds, and virulence, were also identified.
V. anguillarum J360 does not possess some of the virulence genes present in V. anguillarum strains isolated from the Pacific coasts, including the metalloproteases empA and prtV [8,58]. Nonetheless, V. anguillarum J360 harbors a M6 family metalloprotease (DYL72_17780) similar to the prtV gene, associated to gelatinase activity (Table S3).
In addition, genes that encoded for secreted enzymes such as phospholipase and lipases were found in chromosome-II, which correlates with the lipase (C 14 )-positive phenotype observed in the enzymatic profile (Table S3). V. anguillarum J360 possesses several genes associated with flagella and motility, such as the operons fliRQPONMLKJIHGFE (DYL72_03140-DYL72_03205), flgLKJIHGFEDCB (DYL72_03685-DYL72_03735), and motYBA (DYL72_12660, DYL72_00275, DYL72_12090) located in chromosome-I, which is consistent with the mot + phenotype of V. anguillarum J360 (Table 2).
In concordance with the locations of the virulence factors, transcriptional regulators such as luxR, lysR, cysB, nhaR, and hfq were mostly located in chromosome-I, and an additional copy of luxR was found in chromosome-II. LuxR belongs to a transcriptional activators family, that together with an N-(3-oxodecanoyl)-L-homoserine lactone (ODHL), mediates the signal transduction mechanisms of quorum-sensing genes such as luxICDABE operon [59]. The duplication of luxR in V. anguillarum J360 suggests that quorum-sensing plays an important role in the biology of this strain.
In contrast, 19 GIs were identified in V. anguillarum VIB43 ( Figure files 2 and 4), which indicates that these three GIs are genomic pathogenic islands, perhaps acquired through horizontal gene transference [60]. These results support the hypothesis that these IS elements (IS66, ISL3, IS3, IS5) 6 are responsible for the genomic gaps (GGs) and genomic rearrangements previously mentioned, as well as support the 0.05-8% of genomic differences observed at the identity analyses.

Conclusions
In this study, the complete genome of V. anguillarum J360 serotype O2 isolated from infected cultured lumpfish in Newfoundland, Canada was reported. V. anguillarum J360 has a larger genome size (4,549,571 bp) as compared to other available V. anguillarum genomes. The V. anguillarum J360 genome has genes related to antibiotic resistance, hemolysin activity, gelatinase, and lipases that play a major role in virulence. V. anguillarum J360 was shown to be closely related to the V. anguillarum VIB43 strain isolated in Scotland, UK, from sea bass. Comparative genomics revealed that five LCBs are shared between V. anguillarum J360 and V. anguillarum VIB43 chromosome-I, and two LCBs are shared in chromosome-II. Twenty-one genomic islands (GIs) were identified within the chromosomes of V. anguillarum J360. Similar GIs identified in V. anguillarum J360 were found in V. anguillarum VIB43 chromosomes, and this is consistent with the whole genome phylogenetic analysis. V. anguillarum J360 has virulence-associated genes in both chromosomes but does not harbor a virulent plasmid.