Comparative Genomic Analysis of a Novel Vibrio sp. Isolated from an Ulcer Disease Event in Atlantic Salmon (Salmo salar)

Ulcer diseases are a recalcitrant issue at Atlantic salmon (Salmo salar) aquaculture cage-sites across the North Atlantic region. Classical ulcerative outbreaks (also called winter ulcer disease) refer to a skin infection caused by Moritella viscosa. However, several bacterial species are frequently isolated from ulcer disease events, and it is unclear if other undescribed pathogens are implicated in ulcer disease in Atlantic salmon. Although different polyvalent vaccines are used against M. viscosa, ulcerative outbreaks are continuously reported in Atlantic salmon in Canada. This study analyzed the phenotypical and genomic characteristics of Vibrio sp. J383 isolated from internal organs of vaccinated farmed Atlantic salmon displaying clinical signs of ulcer disease. Infection assays conducted on vaccinated farmed Atlantic salmon and revealed that Vibrio sp. J383 causes a low level of mortalities when administered intracelomic at doses ranging from 107–108 CFU/dose. Vibrio sp. J383 persisted in the blood of infected fish for at least 8 weeks at 10 and 12 °C. Clinical signs of this disease were greatest 12 °C, but no mortality and bacteremia were observed at 16 °C. The Vibrio sp. J383 genome (5,902,734 bp) has two chromosomes of 3,633,265 bp and 2,068,312 bp, respectively, and one large plasmid of 201,166 bp. Phylogenetic and comparative analyses indicated that Vibrio sp. J383 is related to V. splendidus, with 93% identity. Furthermore, the phenotypic analysis showed that there were significant differences between Vibrio sp. J383 and other Vibrio spp, suggesting J383 is a novel Vibrio species adapted to cold temperatures.


Introduction
Ulcerative diseases in Atlantic salmon (Salmo salar) aquaculture were first reported in Norway in 1980, and still are a significant health/economic issue for the North Atlantic region [1]. The Gram-negative marine pathogen Moritella viscosa is typically described as the etiological agent of the classic ulcer disease (also called winter ulcer disease) in European farmed fish [2,3]. Although M. viscosa is the primary pathogen associated with ulcer disease, it is not the only bacterium isolated from ulcers and lesions in Atlantic salmon. The isolation and identification of various bacterial species from ulcer-disease cases have been reported, including Aliivibrio wodanis (formerly Vibrio wodanis) and Tenacibaculum sp. [4,5]. In fact, Tenacibaculum might target scarified skin, and co-infect wounds with M. viscosa [5]. Also, M. viscosa might co-infect with A. wodanis, which may limit the growth of M. viscosa [6]. In addition, A. wodanis has the ability to solo infect Atlantic salmon [7].
Since the first documented outbreak of ulcerative disease in Eastern Canada (summer 1999) caused by A. wodanis in Atlantic salmon [8], several pathogens including M.

Biochemical, Enzymatic and Physiological Characterization
The biochemical profile of Vibrio sp. J383 was characterized using API 20E, API 20NE, and API ZYM according to the manufacturer's (BioMerieux, Marcy-l'Etoile, France) instructions. The strips were incubated at 15 °C for 48 h, and the results were analyzed using API web (BioMerieux). The primary characterization of Vibrio sp. J383 was performed based on the Gram stain, capsule stain, and morphological and cultural characteristics [16]. Vibrio sp. J383 growth rate was determined in TSB with 2% NaCl at 4, 15, 28, and 37 °C. Halophilic growth was evaluated in LB supplemented with 0.5 and 2% of NaCl. Also, catalase and oxidase activity were measured according to standard protocols [16,19]. Hemolytic activity was assessed in TSA with 5% salmon blood and sheep blood agar at 15 °C [19,20]. The bacteria's liposaccharide (LPS) profile was examined based on previous protocols [21,22]. Vibrio sp. J383 growth curves were determined in triplicate at 15 °C according to established protocols [18].

Siderophore Synthesis
A siderophores secretion assay was performed using CAS plates according to standard procedure [24]. Briefly, previously mentioned conditions were used to cultivate Vibrio sp. J383, and mid-log phase bacteria with an optical density (O.D.) at 600 nm of 0.7 were harvested, washed three times with PBS at 6000 rpm for 10 min, and then resuspended in 1 mL of PBS. This bacterial culture was used to inoculate TSB with 2% NaCl, and TSB with 2% NaCl supplemented with 100 µM of FeCl3 or 100 µM of 2,2 dipyridyl in a ratio 1:10 (bacteria: culture media). Vibrio sp. J383 was cultured under aeration for 24 h at 15 °C. The cells were collected at the mid-log phase after the incubation time, washed twice with PBS, and resuspended in 100 µL of PBS. After that, CAS agar plates were inoculated with 5 µL of the concentrated bacterial pellet and incubated at 15 °C for 48 h [24].

Fish Origin and Holding Conditions
Farmed Atlantic salmon (~200-250 g) of New Brunswick (Saint John River) origin that had been vaccinated with ALPHA JECT micro IV (Pharmaq, Overhalla, Norway) were

Biochemical, Enzymatic and Physiological Characterization
The biochemical profile of Vibrio sp. J383 was characterized using API 20E, API 20NE, and API ZYM according to the manufacturer's (BioMerieux, Marcy-l'Etoile, France) instructions. The strips were incubated at 15 • C for 48 h, and the results were analyzed using API web (BioMerieux). The primary characterization of Vibrio sp. J383 was performed based on the Gram stain, capsule stain, and morphological and cultural characteristics [16]. Vibrio sp. J383 growth rate was determined in TSB with 2% NaCl at 4, 15, 28, and 37 • C. Halophilic growth was evaluated in LB supplemented with 0.5 and 2% of NaCl. Also, catalase and oxidase activity were measured according to standard protocols [16,19]. Hemolytic activity was assessed in TSA with 5% salmon blood and sheep blood agar at 15 • C [19,20]. The bacteria's liposaccharide (LPS) profile was examined based on previous protocols [21,22]. Vibrio sp. J383 growth curves were determined in triplicate at 15 • C according to established protocols [18].

Siderophore Synthesis
A siderophores secretion assay was performed using CAS plates according to standard procedure [24]. Briefly, previously mentioned conditions were used to cultivate Vibrio sp. J383, and mid-log phase bacteria with an optical density (O.D.) at 600 nm of 0.7 were harvested, washed three times with PBS at 6000 rpm for 10 min, and then resuspended in 1 mL of PBS. This bacterial culture was used to inoculate TSB with 2% NaCl, and TSB with 2% NaCl supplemented with 100 µM of FeCl 3 or 100 µM of 2,2 dipyridyl in a ratio 1:10 (bacteria: culture media). Vibrio sp. J383 was cultured under aeration for 24 h at 15 • C. The cells were collected at the mid-log phase after the incubation time, washed twice with PBS, and resuspended in 100 µL of PBS. After that, CAS agar plates were inoculated with 5 µL of the concentrated bacterial pellet and incubated at 15 •  University; MUN) in 3800 L tanks supplied with 95-100% air saturated, and UV-treated, flow through seawater at 10-12 • C, and an ambient photoperiod (spring-summer). The fish were fed three days per week at 1% body weight with a commercial dry pellet (Skretting, BC, Canada; 50% protein, 18% fat, 1.5% carbohydrate, 3% calcium, 1.4% phosphorus). All experiments were conducted under approved institutional animal ethics protocols (#18-1-JS and #18-03-JS).

Bacterial Inoculum Preparation
Isolated strains were grown, harvested, and used to infect the Atlantic salmon. Briefly, bacterial cells were harvested at the mid-log phase, at an O.D. at 600 nm of~0.7, and washed three times with PBS at 6000 rpm for 10 min. Bacterial O.D. was monitored using a Genesys 10 UV spectrophotometer (Thermo Spectronic, Thermo Fischer Scientific, MA, USA) and by plating to determine the colony forming units (CFU/mL) [25,26].

Koch's Postulates
The infection procedures were conducted in the AQ2 biocontainment unit at the Cold-Ocean Deep-Sea Research Facility (CDRF), MUN. Fish were transferred to the AQ2-CDRF unit and acclimated for one week at 10 • C before infection. An initial infection screening assay was conducted in Atlantic salmon (200 g) intraperitoneally (ip) injected with a high dose (1 × 10 8 CFU/dose) of each isolated strain. Each infected group consisted of 5 fish in individual tanks under optimal conditions. Fish ip injected with PBS were used as a negative control, and fish ip injected with M. viscosa J311 (ATCC BAA-105) were used as a positive control.
A total of 135 Atlantic salmon (~250 g) were used to evaluate Koch's postulates for Vibrio sp. J383 [15,26,27]. The infection procedures were conducted according to established protocols [25,28]. Briefly, fish were anesthetized with 0.05 g/L MS-222 and individually injected with 100 µL of the respective bacterial inoculum. Fish were divided into three 500 L tanks containing 45 Atlantic salmon each, and intracelomic (ic) infected with 10 6 , 10 7 , and 10 8 CFU/dose, respectively. Mortality was monitored until 12 weeks post-infection (wpi). The water temperature was increased during the experiment, starting at 10 • C for 4 wpi, then raised to 12 • C at 5 wpi, and finally increased to 16 • C at 9 wpi until 12 wpi. Tissue samples (e.g., head kidney, liver, spleen, and blood) from 6 fish of each dose were aseptically taken at 2 wpi. Bacterial loads were determined by established protocols [25,28]. Briefly, liver, spleen and head kidney were aseptically removed, and sections of the collected tissues were placed into sterile homogenizer bags (Nasco whirl-pak ® , Fort Atkinson, WI, USA), weighed, and PBS was added to a final volume of 1 mL. Then, the tissues were homogenized, the suspensions were serially diluted (1:10), and then plated onto TSA supplemented with 2% NaCl. To determine the number of bacteria CFU per g of tissue, the plates were incubated at 15 • C for at least 5 days. The total bacterial count was normalized to 1 g of tissue according to the initial weight of the tissue as previously described [25]. Additionally, blood samples were taken from 6 fish per dose every two weeks until the end of the experiment (i.e., at 2, 4, 6, 8, 10, and 12 wpi). Heparin (100 mg/mL) was added to the blood samples, serially diluted in filtered sterilized seawater (1:10) and plated in TSA with 2% NaCl.
Since mortality was evident at 12 • C, a second infection trial was conducted only at 12 • C. A total of 160 Atlantic salmon (~250 g) were equally distributed in four 500 L tanks containing 40 fish each and acclimated at 12 • C for one week before infection. One group of fish was not infected and used as a negative control. Fish in the other group were ic injected with 10 8 CFU/dose of Vibrio sp. J383. Blood samples were collected randomly from 9 fish every two weeks to determine the bacterial load until 14 wpi. Also, tissue samples (e.g., head kidney, liver, and spleen) were collected at 12 and 14 wpi from 9 fish. This experiment was conducted according to established protocols [25,28], and mortality was monitored daily until 12 wpi. Vibrio sp. J383 was grown, harvested, and washed as previously described. According to the manufacturer's instructions (Promega, Madison, WI, USA), the Wizard Genomic DNA Purification Kit was used to extract the genomic DNA (gDNA) of Vibrio sp. J383. The gDNA was quantified by spectrophotometer using a Genova Nano Micro-Spectrophotometer (Jenway, UK) and evaluated for purity and integrity by electrophoresis (0.8% agarose gel) [29]. Libraries and sequencing were conducted at Genome Quebec (Canada) using PacBio and Miseq Illumina sequencers.

Comparative Genomics Analysis
Average nucleotide identity (ANI) was calculated by whole genome alignments using the CLC genomic workbench whole genome analysis tool with default parameters.
(Min. initial seed length = 15; Allow mismatches = yes; Min. alignment block = 100). A minimum similarity of 0.8 and a minimum length of 0.8 were used as parameters for CDS identity. A comparative heat map was made using CLC. Phylogenetic analysis was performed using two different software packages, CLC Genomic workbench v22.0 and MEGA11 [33]. Evolutionary history was estimated using the Neighbor-Joining method with a bootstrap consensus of 500 replicates [34], and evolutionary distance was computed using the Jukes-Cantor method [35]. Photobacterium damselae 91-197 (AP018045/6) chromosomes were utilized as an outgroup [36]. Whole genome dot plots between closely related strains were constructed using the whole genome analysis tool to visualize and analyze genomic differences. Comparative alignment analysis was conducted using the CLC genomic workbench (Qiagen, v22.0). This analysis was used to identify homologous regions (locally collinear blocks), translocations, and inversions within the two bacterial genomes for chromosomes 1 and 2.

Genomic Islands
The detection of genomic islands (GIs) was conducted using the Island Viewer v.4 pipeline (https://www.pathogenomics.sfu.ca/islandviewer/browse/; accessed on 1 February 2023), which integrates Island Path-DIMOB, SIGH-HMM, and Island Pick analysis tools into a single platform [37]. Analysis was performed for both chromosomes and the plasmid. SecReT6 v3 web server was utilized to identify and annotate the type VI secretion system (T6SS) genes that share sequence homology with characterized T6SSs [38,39].

Statistical Analysis
Fish survival rates were transformed using an arc-sin (survival rate ratio) function. One-way ANOVAs were utilized to identify significant differences. GraphPad Prism 9 was used to conduct all statistical analyses (GraphPad Software, California, CA, USA).

Phenotypic Characterization
Vibrio sp. J383 displayed substantial growth in TSB with 2% NaCl between 15 • C and 4 • C (Table 1). However, it did not grow well at 28 • C and did not grow at 37 • C. Vibrio sp. J383 grew well in LB with 1% and 2% NaCl at 15 • C. However, it did not grow in LB supplemented with 0 and 0.5% NaCl at 15 • C. These results indicated that Vibrio sp. J383 is psychotropic and halophilic. Vibrio sp. J383 showed hemolytic activity in sheep and salmon blood agar at 15 • C ( Figure S1G,I). Vibrio sp. J383 was shown to be motile, oxidase and catalase-positive, and type I-fimbria-negative (Table 1). Table 1. Phenotypic characteristics of Vibrio sp. J383.

Characteristics (Growth at)
Vibrio J383 Gram Stain Gram-Negative Capsule stain + Hemolysin in Salmon blood agar(15 • C) + Hemolysin in Sheep blood agar (15 • C) + Hemolysin in Salmon blood agar (28 •  To evaluate siderophore synthesis, Vibrio sp. J383 was grown under iron-enriched and iron-limited conditions, inoculated onto CAS agar plates, and incubated at 15 • C for 48 h. Siderophore secretion was observed under iron-enriched (100 µM of FeCl 3 ), ironlimited (100 µM of 2,2 dipyridyl), and control conditions (TSB). Furthermore, there were no noticeable differences in the size of the halo for siderophore secretion between different groups ( Figure S1B).

Infection Trials in Atlantic Salmon
We isolated a total of five strains, three from the head kidney, one from the liver, and one from the spleen, from different infected fish. An initial screening to determine the virulence of the five isolates was conducted in Atlantic salmon (200 g). The fish were transferred to the AQ2/3 biocontainment zone of the CDRF, acclimated for 1 week, and intraperitoneally (ip) injected with a very high dose (1 × 10 8 CFU/dose) of each isolate. Fish ip injected with PBS were used as a negative control, and fish ip injected with M. viscosa J311 were used as a positive control. As expected, M. viscosa killed all the animals quickly, and all the animals ip injected with PBS survived ( Figure S2). Only the strain J383 (SP6) caused mortality and clinical signs of ulcer disease. Mortality associated with J383 infection in Atlantic salmon indicates that it causes a chronic type of infection rather than an acute infection. This is consistent with the infection event from which these samples were obtained.
The first infection trial was conducted to evaluate Koch's postulates under rising temperature conditions (10 • C for 4 weeks, 12 • C for 4 weeks, and 16 • C for 4 weeks). Approximately 22.5% mortality was recorded in the high-dose (10 8 CFU/dose) group, 5% mortality rate in the medium-dose (10 7 CFU/dose) group, and no mortality was observed in the low-dose (10 6 CFU/dose) group ( Figure 2A). Mortality started at 10 • C, peaked at 12 • C, but was not reported at 16 • C. Clinical signs and presence of Vibrio sp. J383 were observed in all moribund Atlantic salmon ( Figure 2B). Vibrio sp. J383 was not detected in internal organs and blood samples at 2 wpi, but it was detected in blood in the high-dose infection group at 4, 6, 8 and 10 wpi. Vibrio sp. J383 was detected in the blood of fish infected with the medium dose at 6 and 10 wpi. However, at 12 wpi, no bacteria were observed in the collected samples from the different doses ( Table 2). The isolation of Vibrio sp. J383 from blood samples confirmed Koch's postulates.    A second infection assay was performed at 12 • C to determine the infection kinetics of Vibrio sp. J383, with 10 8 CFU/dose. Mortality started at 5 dpi and reached 20% by 90 dpi (Figure 3). Vibrio sp. J383 caused bacteremia in about 30-100% of the infected fish, and reached a peak at 6 wpi, which is consistent with the mortality levels (Figures 3 and 4A). Bacteremia started to decrease by 8-10 wpi. No bacteria in the blood were detected after 12 wpi ( Figure 4A), but Vibrio sp. J383 was detected in the spleen, liver, and head kidney at this sampling point ( Figure 4B). Higher bacterial loads were observed in the spleen samples compared to the liver and head kidney ( Figure 4B).   Figure 5A-C). The coverage assembly for chromosome 1, chromosome 2, and the parge plasmid was 306, 27, and 14 times, respectively. The plasmid profile agrees with the genomic analysis, supporting the theory that this Vibrio sp. possesses one large plasmid and no small plasmids. Vibrio sp. J383 s genome was submitted to NCBI under the BioProject (PRJNA836625) and BioSample (SAMN28165975). Vibrio sp. J383 genome has a total estimated length of 5.9 Mb and a G + C content of 44.3 and 44.1% for chromosomes 1 and 2, respectively. RAST pipeline annotation predicted a total of 309 subsystems and 3235 coding sequences (CDS) for chromosome 1, a total of 101 subsystems and 1866 CDSs for chromosome 2, and a total of 8 subsystems and 237 CDSs for the large plasmid p.J383 (Table 3). The NCBI Prokaryote Genome Annotation pipeline (PGAP) presented a total of 5288 genes predicted, a total of 16 (5S), 15 (16S), and 15 (23S) rRNAs, 138 tRNAs, and 5 ncRNAs for the whole genome (Table 4).     J383 genome has a total estimated length of 5.9 Mb and a G + C content of 44.3 and 44.1% for chromosomes 1 and 2, respectively. RAST pipeline annotation predicted a total of 309 subsystems and 3235 coding sequences (CDS) for chromosome 1, a total of 101 subsystems and 1866 CDSs for chromosome 2, and a total of 8 subsystems and 237 CDSs for the large plasmid p.J383 (Table 3). The NCBI Prokaryote Genome Annotation pipeline (PGAP) presented a total of 5288 genes predicted, a total of 16 (5S), 15 (16S), and 15 (23S) rRNAs, 138 tRNAs, and 5 ncRNAs for the whole genome (Table 4).

Genomic Islands (GIS)
Twenty-four putative GIs were identified within the chromosomes and plasmid: sixteen GIs in chromosome 1, seven GIs in chromosome 2, and one GI in the plasmid ( Figure 6A-C). The GIs' size ranged from 6 kb to 50 kb, with a total of 929 genes (Supplementary Files S1-S3). Genes encoding for integrases, transposases, phage integrase, and multidrug-resistance transporters were found in GIs 6, 7, 9 and 12, respectively. Also, UDP-3-O- [3-hydroxymyristoyl] glucosamine N-acyltransferase, virulence factor VirK and alpha-galactosidase were in GIs 3, 7 and 15, respectively. Zonula occludes toxin (Zot)-like phage protein was found in the chromosome 1 of Vibrio sp. J383. Glutaredoxin encoding gene was detected in genomic island 20 of chromosome 2. The type VI secretion system (T6SS) is an important virulence factor detected in Vibrio sp. J383. The SecReT6 v3 web server identified and annotated the T6SS genes that share sequence homology with characterized T6SSs and indicated that the Vibrio sp. J383 genome encodes two distinct T6SSs in GI 12 of chromosome 1 and GI 21 of chromosome 2 (Table 5, Figure 7, and Supplementary Files S4 and S5). The Vibrio sp. J383 secretion systems belong to the T6SS i family and are most closely related to the T6SS in Vibrio coralliilyticus OCN008. Some genes encoded within these two loci do not share sequence homology with known T6SS genes; however, predicting the structures of the encoded proteins with AlphaFold and comparison to solved structures in the Protein Data Bank (http://rcsb.org)( last accessed in June 2023) allowed us to identify additional T6SS genes [41]. These analyses indicate that both T6SS loci in the Vibrio sp. J383′s genome encode for the full complement of components required for T6SS assembly and activity [42]. . Type VI secretion system (T6SS) gene cluster in Vibrio sp. J383. The T6SS in chromosomes 1 and 2 are made from 13 Tss (Type Six Subunits) proteins that are known as the "core components". TssC gene duplicated in chromosome 1. Several uncharacterized genes are present in two T6SS loci and these may encode toxins secreted by the apparatus; however, further characterization is required to elucidate the role of these genes. The type VI secretion system (T6SS) is an important virulence factor detected in Vibrio sp. J383. The SecReT6 v3 web server identified and annotated the T6SS genes that share sequence homology with characterized T6SSs and indicated that the Vibrio sp. J383 genome encodes two distinct T6SSs in GI 12 of chromosome 1 and GI 21 of chromosome 2 (Table 5, Figure 7, and Supplementary Files S4 and S5). The Vibrio sp. J383 secretion systems belong to the T6SS i family and are most closely related to the T6SS in Vibrio coralliilyticus OCN008. Some genes encoded within these two loci do not share sequence homology with known T6SS genes; however, predicting the structures of the encoded proteins with AlphaFold and comparison to solved structures in the Protein Data Bank (http://rcsb.org) (last accessed on 20 June 2023) allowed us to identify additional T6SS genes [41]. These analyses indicate that both T6SS loci in the Vibrio sp. J383 s genome encode for the full complement of components required for T6SS assembly and activity [42]. The type VI secretion system (T6SS) is an important virulence factor detected in Vibrio sp. J383. The SecReT6 v3 web server identified and annotated the T6SS genes that share sequence homology with characterized T6SSs and indicated that the Vibrio sp. J383 genome encodes two distinct T6SSs in GI 12 of chromosome 1 and GI 21 of chromosome 2 (Table 5, Figure 7, and Supplementary Files S4 and S5). The Vibrio sp. J383 secretion systems belong to the T6SS i family and are most closely related to the T6SS in Vibrio coralliilyticus OCN008. Some genes encoded within these two loci do not share sequence homology with known T6SS genes; however, predicting the structures of the encoded proteins with AlphaFold and comparison to solved structures in the Protein Data Bank (http://rcsb.org)( last accessed in June 2023) allowed us to identify additional T6SS genes [41]. These analyses indicate that both T6SS loci in the Vibrio sp. J383′s genome encode for the full complement of components required for T6SS assembly and activity [42]. . Type VI secretion system (T6SS) gene cluster in Vibrio sp. J383. The T6SS in chromosomes 1 and 2 are made from 13 Tss (Type Six Subunits) proteins that are known as the "core components". TssC gene duplicated in chromosome 1. Several uncharacterized genes are present in two T6SS loci and these may encode toxins secreted by the apparatus; however, further characterization is required to elucidate the role of these genes. . Type VI secretion system (T6SS) gene cluster in Vibrio sp. J383. The T6SS in chromosomes 1 and 2 are made from 13 Tss (Type Six Subunits) proteins that are known as the "core components". TssC gene duplicated in chromosome 1. Several uncharacterized genes are present in two T6SS loci and these may encode toxins secreted by the apparatus; however, further characterization is required to elucidate the role of these genes.

Comparative Genomic Analysis
The phylogenetic analysis of Vibrio sp. J383 chromosomes indicated that it was closely related to V. splendidus ( Figure 8A-D). The average nucleotide identity (ANI) analysis between Vibrio sp. J383 and V. splendidus showed 95.75% identity for chromosome 1 ( Figure S3A) and 93.31% identity for chromosome 2 ( Figure S3B), which suggests that these two strains share a common ancestor.
The phylogenetic analysis of Vibrio sp. J383 chromosomes indicated that it was closely related to V. splendidus ( Figure 8A-D). The average nucleotide identity (ANI) analysis between Vibrio sp. J383 and V. splendidus showed 95.75% identity for chromosome 1 ( Figure  S3A) and 93.31% identity for chromosome 2 ( Figure S3B), which suggests that these two strains share a common ancestor. However, the dot plot showed significant differences between Vibrio sp. J383 and V. splendidus in both chromosomes ( Figure 9A,C), including genome gaps and one inversion event in chromosome 1 ( Figure 9A,B). However, the dot plot showed significant differences between Vibrio sp. J383 and V. splendidus in both chromosomes ( Figure 9A,C), including genome gaps and one inversion event in chromosome 1 ( Figure 9A,B). The whole genome alignment identified 18 locally collinear blocks (LCBs) in Vibrio sp. J383, which are conserved segments with no genomic rearrangements [43]. The comparative alignment analysis of each chromosome showed 9 LCBs in chromosome 1 ( Figure  9B) and 9 LCBs in chromosome 2 ( Figure 9D).
Also, comprehensive genome analysis in PATRIC indicated that there were some pathogenesis-associated genes. Chromosomes 1 and 2 contain several transporter-related genes as well as genes linked to virulence or antibiotic resistance. RAST and PATRIC comprehensive genome analyses identified the presence or absence of specific genes in the chromosomes and the plasmid (Table S2).

Discussion
The causes of ulcerative skin disease in Atlantic salmon are not fully understood. Several causative agents are being described in the North Atlantic rim, including M. viscosa, Tenecebacillum spp, and A. wondandis [4,5]. However, differences in disease etiology indicate that several undescribed pathogens might also cause skin ulcers in Atlantic salmon. In the present study, we isolated Vibrio sp. strain J383 from the spleen of Atlantic salmon exhibiting skin ulcers (Figure 1). The phenotypic characterization indicates that this strain is marine and requires the presence of at least 1% of NaCl for survival (Table 1 and Figure S1A). Also, Vibrio sp. J383 showed substantial growth between 4 and 15 °C, but no growth at temperatures over 28 °C, suggesting that this strain is adapted to cold temperatures. Vibrio sp. J383 also possesses several virulence factors, including hemolysins, siderophores, and LPS (Table 1, Figure S1B,D,G,I), indicating that it has pathogenic properties. Finally, Vibrio sp. J383 constitutively synthesizes siderophores, indicating that this The whole genome alignment identified 18 locally collinear blocks (LCBs) in Vibrio sp. J383, which are conserved segments with no genomic rearrangements [43]. The comparative alignment analysis of each chromosome showed 9 LCBs in chromosome 1 ( Figure 9B) and 9 LCBs in chromosome 2 ( Figure 9D).
Also, comprehensive genome analysis in PATRIC indicated that there were some pathogenesis-associated genes. Chromosomes 1 and 2 contain several transporter-related genes as well as genes linked to virulence or antibiotic resistance. RAST and PATRIC comprehensive genome analyses identified the presence or absence of specific genes in the chromosomes and the plasmid (Table S2).

Discussion
The causes of ulcerative skin disease in Atlantic salmon are not fully understood. Several causative agents are being described in the North Atlantic rim, including M. viscosa, Tenecebacillum spp, and A. wondandis [4,5]. However, differences in disease etiology indicate that several undescribed pathogens might also cause skin ulcers in Atlantic salmon. In the present study, we isolated Vibrio sp. strain J383 from the spleen of Atlantic salmon exhibiting skin ulcers (Figure 1). The phenotypic characterization indicates that this strain is marine and requires the presence of at least 1% of NaCl for survival (Table 1 and Figure S1A). Also, Vibrio sp. J383 showed substantial growth between 4 and 15 • C, but no growth at temperatures over 28 • C, suggesting that this strain is adapted to cold temperatures. Vibrio sp. J383 also possesses several virulence factors, including hemolysins, siderophores, and LPS (Table 1, Figure S1B,D,G,I), indicating that it has pathogenic properties. Finally, Vibrio sp. J383 constitutively synthesizes siderophores, indicating that this strain can scavenge essential iron within and outside the host. The synthesis of siderophores is usually regulated [44], but Vibrio sp. J383 possesses a natural constitutive expression that requires further study.
Infection assays indicated that Vibrio sp. J383 is a non-acute, chronic pathogen that can produce skin ulcers and kill fish, especially at 1 2 • C (Figures 2 and 3). Vibrio sp. J383 triggered clinical signs of ulcer disease in vaccinated farmed Atlantic salmon, indicating that the generic vaccine utilized does not confer protection against this novel chronic pathogen ( Figure 2B). Skin ulcer severity ranged from mild to severe in some of the fish that were infected with medium and high doses (10 7 and 10 8 CFU/dose, respectively) ( Figure 2B). Although clinical signs of ulcer disease were evident in the infected fish, the low mortality rates indicated that this strain is not an acute pathogen. Around 5% to 22.5% mortality was recorded in the infected fish given the medium and high doses of Vibrio sp. J383 (Figure 2A). Our result is consistent with the literature, which indicates that the mortality during outbreaks in sea-cages with skin ulcers ranged from 0.01 to 23.32% [13]. Vibrio sp. J383 caused bacteremia, and it was detected until 10 wpi in Atlantic salmon at 12 • C. Bacteremia showed the same patterns in both infection assays ( Figure 4A, Table 2). This indicates that Vibrio sp. J383 can cause a systemic infection. Vibrio sp. J383 was detected at 6 wpi in most of the blood samples in the high-dose infection group (Table 2), and isolated in all the blood samples at 12 • C ( Figure 4A). Bacteremia was not recorded at 10 • C at 2 wpi but was reported at 12 • C in some blood samples (Table 2, Figure 4A). No bacteremia and mortality were recorded when the temperature increased to 16 • C ( Table 2). These results are consistent with previous field observations, which indicated that the water temperature during skin ulcerative disease outbreaks ranged from 10 • C to 13 • C [13]. Our findings indicated that Vibrio sp. J383 becomes more invasive at 12 • C compared to 10 • C. Regardless of water temperature, Vibrio sp. J383 was not detected in the blood at 12 wpi (Table 2, Figure 4A). However, Vibrio sp. J383 was detected in the spleen, head kidney, and liver at 12 and 14 wpi in Atlantic salmon infected with the high dose at 12 • C ( Figure 4B). Bacterial loads were significantly higher in the spleen compared to the liver and head kidney at 12 and 14 wpi. Vibrio sp. J383 loads decreased substantially from 12 wpi to 14 wpi ( Figure 4B). Collectively, our findings suggest that Vibrio sp. J383 is a chronic pathogen.
The biochemical characterization and identification of environmental Vibrio species have been complicated because of their notable diversity [45,46]. The biochemical profile attained using API 20NE showed that Vibrio sp. J383 was unable to reduce urea but reduces nitrates and produces indole, suggesting 64.8% and 34.6% similarity to V. vulnificus and V. alginolyticus, respectively (Table S1). Phylogenetic and comparative analyses showed that Vibrio sp. J383 is closely related to V. splendidus, with 93% identity, and this suggests that these two strains share a common ancestor. However, phenotypical tests revealed significant differences between Vibrio sp. J383 and other Vibrio strains, indicating that Vibrio sp. J383 could be a novel species.
V. anguillarum J360, a virulent strain isolated from the North Atlantic, displayed thermo-inducible α-hemolysin activity at 28 • C, but no hemolytic activity at 15 • C [15]. In contrast, Vibrio sp. J383 showed hemolysin activity at 15 • C but not at 28 • C. Vibrio spp. utilize hemolysins to lyse host erythrocytes to acquire nutrients, such as iron [47]. Hemolysins are crucial virulence factors for V. anguillarum and play a key role in boosting its pathogenicity [48,49]. The presence of hemolytic activity in Vibrio sp. J383 indicates its potential pathogenicity. Vibrio sp. J383 showed growth at 4 • C, optimal growth around 15 • C, weak growth at 28 • C and no growth at 37 • C. Also, Vibrio sp. J383 produces catalase and oxidase. These results are consistent with most pathogenic marine Vibrio spp. [50], and the ability of Vibrio sp. J383 to grow at 4 • C indicates that this novel pathogen is well adapted to cold environments.
The genotypic characterization of Vibrio sp. J383 indicated the presence of two chromosomes. The existence of two chromosomes is a basic characteristic of Vibrio spp. that developed as a survival means, and allows for the rapid adaptation of the pathogen to different environments and hosts [51]. Vibrio sp. J383 has a genome size of 5,902,734 bp, very similar to V. splendidus BST 398, (5,508,387 bp) ( Table 3). The genome of V. splendidus BST 398 included 4700 predicted open reading frames, a G + C content of 44.12%, 137 tRNA genes and 46 rRNA genes [52]. In contrast, the whole genome of the novel strain Vibrio sp. J383 has a total of 16 (5S), 15 (16S), and 15 (23S) rRNAs, 138 tRNAs, and 5 ncRNAs (Table 4). These results suggest that although V. splendidus and Vibrio sp. J383 share a common ancestor, they are different strains with distinct genomic characteristics.
Some specific genes associated with virulence and antibiotic resistance were detected in the PATRIC comprehensive genome analysis. The tetracycline resistance subsystem was found in Vibrio sp. J383, which makes this strain resistant to antibiotics and toxic compounds. Hydroxyacylglutathione hydrolase was also found in Vibrio sp. J383, which is a virulence-related gene and contributes to the stress response in bacteria.
Cold shock proteins of the CSP family were found in chromosome 2 of Vibrio sp. J383 (Table S2). Cold shock proteins help cells adapt by reducing some of the negative effects of temperature changes [53]. They play a crucial role in the cold shock response, and recent data suggest that CSPs may have a greater role in bacterial stress tolerance [54][55][56]. Following the initial cold shock response, the production of CSPs declines while the production of other proteins increases. This helps the cells to grow at a low temperature, but at a slower rate [57]. This finding can explain why Vibrio sp. J383 has fast growth at 15 • C and slightly slower growth at 4 • C (Table 1).
Flagellum was detected in chromosome 1 of Vibrio sp. J383 (Table S2), which plays a key role in bacteria motility, and can contribute to biofilm formation, protein export and adhesion [58]. Chemotaxis present in chromosome 1 of Vibrio sp. J383 plays several roles, including biofilm formation, auto aggregation and swarming, as well as in bacterial interactions with their hosts [59]. Toxin-antitoxin systems are common in bacterial genomes and found in chromosome 2 of Vibrio sp. J383 (Table S2). They are normally made of two elements: a toxin that inhibits a vital cellular process and an antitoxin that hinders its cognate toxin [60].
A total of 24 genomic islands were detected in Vibrio sp. J383. Genomic Islands have also been identified in the species V. anguillarum. For example, V. anguillarum J360 has 21 GIs [15]. Genes encoding for integrase and transposase were found in the mentioned Vibrio strains. Glutaredoxin was detected in chromosome 2 of Vibrio sp. J383. and performs a critical role in the protection against oxidative stress in bacteria [61]. VirK is a virulence factor present in several bacterial pathogens that has been shown to contribute to Salmonella enterica serovar Typhimurium and Escherichia coli virulence [62][63][64]. It may contribute to Vibrio sp.'s pathogenicity as well. Also, Zot was detected in the genomes of V. parahaemolyticus [65], indicating a correlation between this gene and the cytotoxicity of bacteria [65,66]. In V. cholerae, Zot is an important toxin after the classical cholera toxin (CT), and it is encoded by the CTX prophage [67]. It has been shown that Zot has enterotoxic activity and it is hypothesized that it plays a role in the classic diarrhea symptom of V. cholerae infections [68,69]. The presence of Zot in the genome of Vibrio sp. J383 perhaps contributes to its virulence in fish hosts.
The T6SS is a contractile nanomachine that plays a role in interbacterial competition and bacterial pathogenesis by secreting toxic effector proteins into adjacent cells [70][71][72][73][74][75]. The Vibrio sp. J383 genome includes two T6SSs on genomic islands 12 and 21 (Table 5, Figure 7, and Supplementary File S4). T6SS apparatuses are composed of 13 core components that form two parts, the membrane complex and phage-like tail [76,77]. Three membraneassociated proteins comprise the membrane complex, TssJ, TssL, and TssM [75,76,78,79]. Consistent with this, the Vibrio sp. J383 homologs are predicted to be lipidated or to contain transmembrane domains. The T6SS tail is evolutionarily related to the contractile tail of the T4 bacteriophage and is composed of several subassemblies. The baseplate is composed of tssE, tssF, tssG, and tssK [80,81], and the Vibrio sp. J383 homologs share 60-80% identity with other Vibrio T6SS counterparts. tssB, tssC, and tssA comprise the sheath and cap of the T6SS and are encoded in both loci [82][83][84]. Of note, tssB and tssC share greater than 95% identity with other Vibrio T6SSs. Hcp and VgrG play a role in substrate recognition and are co-secreted with associated toxins, and both are encoded in the T6SS loci in Vibrio sp. J383 [76,78,85].
The genes that encode these T6SS apparatus subunits share high levels of homology between genomic islands 12 and 21. Based on homology with well-characterized T6SSs, both apparatuses encoded by Vibrio sp. J383 belong to the T6SS i family. It is impossible to identify T6SS-exported toxins based on sequence alone; however, some of the hypothetical genes in these two loci encode hallmarks that suggest that they may be effector proteins. The presence of toxic effectors genes in close proximity to T6SS apparatus genes does not preclude other effectors encoded in distant genomic loci, and further research can identify these toxins in the future.
Gene duplications are essential prerequisites for gene innovation, which may assist in adaptation to changing environmental conditions [86]. In chromosome 1 of Vibrio sp. J383, tssC is duplicated (Figure 7). T6SSs have been strongly linked to a variety of biological processes, including biofilm formation, bacterial survival in the environment, virulence, and host adaptation; therefore, the duplication of tssC in Vibrio sp. J383 may be an important step toward increasing the fitness of this strain in the environment, amongst the microbiota, or as a pathogen [87][88][89].
In summary, Vibrio sp. J383 has several virulence factors and genes that are associated with pathogens.
Vibrio sp. J383 has one large plasmid, and an analysis based on PATRIC annotation found one subsystem, the MazEF toxin-antitoxin (program cell death) system (Table S2). Toxin-antitoxin (TA) systems, initially discovered in plasmids, were recognized as extrachromosomal genes responsible for post-segregationally killing, which protects plasmid integrity [90]. Toxin-antitoxin (TA) systems have been reported in many bacterial genomes and mediate program cell death (PCD), and are therefore attractive targets for new antimicrobial drugs since they are recognized to "kill from within" [91,92]. The antitoxins neutralize the toxin using different mechanisms and play vital roles, including the maintenance of genomic stability, and assist in biofilm formation in some bacteria [90]. The presence of the TA system in the plasmid of Vibrio sp. J383 indicates that it may play may the same role and enable Vibrio sp. J383 to survive in different temperatures and maintain its virulence factors.

Conclusions
The biochemical profile showed that Vibrio sp. J383 is similar to V. vulnificus, with 64.8% identity. However, phylogenetic and comparative analyses showed that Vibrio sp. J383 is closely related to V. splendidus, with 93% identity. The isolation of Vibrio sp. J383 from blood samples confirmed Koch's postulates. Mortality was approximately 20% in the vaccinated fish infected with a 10 8 CFU/dose, but no mortality was observed in fish infected with a low dose (10 6 CFU/dose). Vibrio sp. J383 was detected for 10 wpi in the blood and at up to 14 wpi in the internal organs (spleen and kidney). The pathogenicity of this new strain is supported by Koch's postulates and the presence of pathogenic genomic islands (GIs 12 and 21) containing virulence factors such as type VI secretion system (T6SS) genes and multidrug-resistance transporter/family protein. Vibrio sp. J383 displays unique characteristics, and has notable differences, compared to other Vibrio strains. The results of this study revealed that Vibrio sp. J383 is potentially a new species that can trigger clinical signs of ulcer disease and cause chronic infections in vaccinated farmed Atlantic salmon. The impact of the co-infection of Vibrio sp. J383 with other etiological agents, like Moritella viscosa, on Atlantic salmon remains to be investigated.