Abstract Cunner (Tautogolabrus adspersus) is a cleaner fish being considered for utilized in the North Atlantic salmon (Salmo salar) aquaculture industry to biocontrol sea lice infestations. However, bacterial diseases due to natural infections in wild cunners have yet to be described. This study reports the isolation of Pseudomonas sp. J380 from infected wild cunners and its phenotypic, genomic, and transcriptomic characterization. This Gram-negative motile rod-shaped bacterium showed a mesophilic (4–28 °C) and halotolerant growth. Under iron-limited conditions, Pseudomonas sp. J380 produced pyoverdine-type fluorescent siderophore. Koch’s postulates were verified in wild cunners by intraperitoneally (i.p.) injecting Pseudomonas sp. J380 at 4 × 10^3, 4 × 10^5, and 4 × 10^7 colony forming units (CFU)/dose. Host-range and comparative virulence were also investigated in lumpfish and Atlantic salmon i.p. injected with ~10^6 CFU/dose. Lumpfish were more susceptible compared to cunners, and Atlantic salmon was resistant to Pseudomonas sp. J380 infection. Cunner tissues were heavily colonized by Pseudomonas sp. J380 compared to lumpfish and Atlantic salmon suggesting that it might be an opportunistic pathogen in cunners. The genome of Pseudomonas sp. J380 was 6.26 megabases (Mb) with a guanine-cytosine (GC) content of 59.7%. Biochemical profiles, as well as comparative and phylogenomic analyses, suggested that Pseudomonas sp. J380 belongs to the P. fluorescens species complex. Transcriptome profiling under iron-limited vs. iron-enriched conditions identified 1159 differentially expressed genes (DEGs). Cellular metabolic processes, such as ribosomal and energy production, and protein synthesis, were impeded by iron limitation. In contrast, genes involved in environmental adaptation mechanisms including two-component systems, histidine catabolism, and redox balance were transcriptionally up-regulated. Furthermore, iron limitation triggered the differential expression of genes encoding proteins associated with iron homeostasis. As the first report on a bacterial infection in cunners, the current study provides an overview of a new marine pathogen, Pseudomonas sp. J380. Keywords: cleaner fish, Salmo salar, bacterial infection, comparative genomics, transcriptomics, iron homeostasis 1. Introduction The ectoparasitic infestation by sea louse (e.g., Lepeophtheirus salmonis) is the most serious threat to both wild and cage-cultured salmonids (e.g., Salmo salar) in the Northern hemisphere [[42]1]. During the infestation, sea lice feed on salmonid tissues leading to skin erosion, osmoregulatory failure, immunosuppression, increased disease susceptibility and, ultimately, death [[43]1,[44]2]. Sea lice escalate the economic burden on salmon farmers both by causing mortalities and demanding expensive control regimes [[45]3]. Over the years, the Atlantic salmon industry has used several methods to combat sea lice infestation. Chemotherapeutic treatment was the dominant sea lice control method for decades until the recent emergence of resistance in sea lice to several active components of drugs [[46]4] and increasing negative public opinion about the impacts of anti-lice drugs on ecological equilibrium. Mechanical abrasion and thermal treatment are routinely used as physical delousing methods, but have adverse effects on salmon health and welfare [[47]5]. Integrated pest management (IPM) was then introduced to the salmon industry, in which multiple non-medicinal methods are being utilized in different combinations (e.g., physical barriers—curtains and skirts, anti-lice diets, laser, water jets, and ultrasound) [[48]6]. Biological pest control by using cleaner fishes is another commonly used lice-controlling strategy in North Atlantic salmon farms [[49]7]. Cleaner fish are considered as ‘green alternatives’ and the potential solution to lice infestation from both economic and ecological points of view. The mutually beneficial cleaner fish—salmonid association reduces the parasite burden on salmonids while providing a food source to cleaner fish. In salmonid aquaculture, different wrasse (e.g., rock cook (Centrolabrus exoletus), goldsinny (Ctenolabrus rupestris), cuckoo (Labrus mixtus), ballan (Labrus bergylta), and corkwing (Symphodus melops) and lumpfish (Cyclopterus lumpus) species are used [[50]8,[51]9,[52]10]. Wrasses and lumpfish possess contrasting season-dependent feeding behaviors. Wrasses tend to consume more parasites than lumpfish, but they reduce their activity during winter and eventually enter into hypometabolic winter dormancy (torpor) state at water temperatures below 5 °C [[53]11]. However, as a cold-water cleaner fish, lumpfish can effectively delouse at or even below, 5 °C. Therefore, the usage of multiple cleaner fish species in combination (e.g., cohabiting wrasse and lumpfish) might be advantageous, since they are efficient during the spring–summer period and autumn–winter period, respectively [[54]10,[55]12]. The potential of cunner (Tautogolabrus adspersus), a native Atlantic Canadian species [[56]13], as a delousing fish was first studied by MacKinnon in the 1990s [[57]14]. A more recent study also described the cleaning behavior and delousing efficiency of cunners using a S. salar—L. salmonis model [[58]15]. The increasing interest in the commercialization of cleaner fish in Atlantic Canada may lead to the consideration of cultured cunners for hatchery production in the future [[59]16]. However, infectious diseases in cunners are not well described [[60]17]. Pseudomonas spp. exist in a diverse range of environments including marine habitats. Some pathogenic species, such as P. fluorescens [[61]18], P. putida [[62]19], and P. plecoglossicida [[63]20], have been reported to cause diseases in rainbow trout (Oncorhynchus mykiss), ayu (Plecoglossus altivelis), and tilapia (Oreochromis niloticus) [[64]21]. Iron is indispensable for several bacterial processes including DNA synthesis, enzyme and redox catalysis, electron transport, and respiration, and is considered a central determinant for growth, survival, and pathogenesis. Bacterial habitats are commonly limited in iron availability and iron withholding is a principal defense strategy imposed by the host to prevent microbial outgrowth. In order to successfully establish an infection, pathogens have evolved many sophisticated evading mechanisms (e.g., siderophore iron-sequesters) for iron piracy from their hosts [[65]22,[66]23]. On the other hand, due to the extreme toxicity resulting from iron overload, bacteria employ a variety of mechanisms to regulate intracellular iron concentrations by coordinating complex transcriptional regulatory schemes to control multiple aspects of iron homeostasis. The uptake, transport, storage, and mobilization of iron are controlled in an iron-dependent manner [[67]22,[68]23]. In this study, we isolated and studied a bacterial pathogen (named Pseudomonas sp. strain J380) that caused skin lesions and ulcers in wild-caught cunners from the coast of Eastern Newfoundland (Canada). Koch’s postulates and infection kinetics were studied in cunners under controlled conditions. Then, host-range was evaluated in lumpfish and Atlantic salmon. Pseudomonas sp. J380 was determined to infect both cleaner fish species, but not Atlantic salmon. This new species was then characterized at phenotypic, genomic, and transcriptomic levels. Comparative genomics and phylogenetic inference indicated that Pseudomonas sp. J380 is a new strain and closely related to the P. fluorescens species complex. Finally, RNA-Seq transcriptome profiling under iron-enriched vs. iron-limited conditions revealed a modulated expression of canonical and unique genes potentially linked to metabolism, environmental adaptation mechanisms, and iron homeostasis. Collectively, our findings provide insights into the biology of a novel marine pathogen, Pseudomonas sp. J380. 2. Materials and Methods 2.1. Pseudomonas sp. Strain J380 Isolation Pseudomonas sp. strain J380 was isolated from the head kidney, liver, and spleen of infected wild cunners captured at Conception Bay, Newfoundland and Labrador (NL), Canada ([69]Figure S1A,B). Fish with skin lesions ([70]Figure S1C,D) were netted and euthanized with an overdose of MS222 (400 mg/L) (Syndel Laboratories, BC, Canada). Samples were dissected and placed into sterile homogenizer bags (Nasco Whirl-Pak^®, Madison, WI, USA). The infected tissues were weighed and homogenized in 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)) up to a final volume of 1 mL. Homogenized and serially diluted tissue suspension (100 µL) was plated onto Trypticase Soy Agar (TSA) plates supplemented with 2% NaCl and incubated at 15 °C for 48 h. Isolated colonies were selected and purified for further analysis. Bacterial stocks were preserved at −80 °C in 10% glycerol and 1% peptone solution. Three selected colonies were phenotypically characterized and subjected to 16S sequencing using universal primers, 27F (5’-AGAGTTTGATYMTGGCTCAG-3’) and 1492R (5′-TACGGYTACCTTGTTACGACTT-3′) [[71]24] at Core Science Facility, Memorial University of Newfoundland (MUN). 2.2. Bacterial Growth Conditions Single colonies of Pseudomonas sp. J380 were grown routinely in 3 mL of Trypticase Soy Broth (TSB, Difco, Franklin Lakes, NJ, USA) at 15 °C in a 16 mm diameter glass tube and placed in a roller for 24 h. To conduct the assays under iron-enriched and -limited conditions, TSB was supplemented with 100 µM FeCl[3] or 100 µM 2,2’-dipyridyl (DIP), respectively. Siderophore synthesis was detected on Chrome Azurol S (CAS) agar plates [[72]25] using bacterial cells harvested at the mid-log phase at an optical density (OD; at 600 nm) of ~0.7 (~1 × 10^8 CFU/mL). 2.3. Biochemical, Enzymatic, and Physiological Characterization The biochemical and enzymatic profiles of Pseudomonas sp. J380 were characterized using standard strips systems, including API20E, API20NE, and APY ZYM (BioMerieux, Marcy-l’Etoile, France) according to the manufacturer’s instructions. Following the incubation of strips with Pseudomonas sp. J380 at 15 °C for 48 h, the results were analyzed using APIWEB (BioMerieux). The growth of Pseudomonas sp. J380 was evaluated at different temperatures (4 °C, 15 °C, 28 °C, and 37 °C) and NaCl concentrations (0%, 1%, and 2%). Motility, siderophore synthesis, hemolysis activity, catalase activity, and oxidase activity were evaluated using standard methods [[73]26]. The antibiogram of Pseudomonas sp. J380 was determined for tetracycline (TET; 30 µg), oxytetracycline (OTC; 30 µg), ampicillin (AMP; 10 µg), chloramphenicol (CHL; 30 µg), trimethoprim-sulfamethoxazole (SXT; 25 µg), cefotaxime (CXT; 30 µg), and oxolinic acid (OXA; 2 µg) using standard methods [[74]27]. 2.4. Siderophores Synthesis The secretion of siderophores was tested using CAS plates by previously described assays [[75]28,[76]29]. Pseudomonas sp. J380 was grown under previously described conditions. Briefly, bacterial cells grown in TSB were harvested at the mid-log phase. Thirty microliters of bacterial inoculum were added to 3 mL TSB supplemented with 100 µM of FeCl[3] or 100 µM of DIP and grown at 15 °C for 24 h with aeration. As a control, Vibrio anguillarum J360 was grown under the same conditions but in TSB supplemented with 2% NaCl [[77]30]. Following the incubation period, the cells were harvested (6000 rpm, 10 min), washed twice with PBS, and resuspended in 50 µL of PBS. Ten microliters of the concentrated bacterial pellet were inoculated onto CAS agar plates and incubated at 15 °C for 48–72 h. Following the incubation, the secreted siderophores were visualized as a yellow-orange halo around the bacterial colony. The plates were also observed using a UV trans-illuminator. To photo-document the fluorescent siderophores, bacteria grown under previously described conditions were washed and resuspended in 1 mL of PBS and stained with 4’,6-diamidino-2-phenylindole (DAPI; Thermo Fisher, USA) for 30 min in darkness. After washing cells with PBS three times (6000 rpm, 10 min), cells were visualized using DAPI (461 nm) and EtBr (358 nm) filters through a Nikon AR1 laser scanning confocal microscope. 2.5. Fish Capture and Holding Fish were captured or produced and maintained in tanks within the Dr. Joe Brown Aquatic Research Building (JBARB) at the Department of Ocean Sciences (DOS), MUN, under animal protocols #18-01-JS, #18-03-JS (21 Jan 2018), and biohazard L-01. All the protocols were reviewed and approved by the Institutional Animal Care Committee ([78]https://www.mun.ca/research/about/acs/acc/) following the Canadian Council of Animal Care guidelines ([79]https://www.ccac.ca/). Cunner fish (Tautogolabrus adpersus) captured in Conception Bay, NL, Canada ([80]Figure S1), were transferred to JBARB for 4 weeks of quarantine. Upon arrival, fish were acclimated to ~8–10 °C in 500 L tanks supplied with oxygen saturation of 95–110%, UV-treated, and filtered flow-through seawater, with 12–12 photoperiod and illuminance of 10–15 lux. The animals were fed with chopped capelin (Mallotus villosus) two times per week ad libitum. Lumpfish were acclimated to ~8–10 °C in 500 L tanks supplied with oxygen saturation of 95–110%, UV-treated, and filtered flow-through seawater, and 12–12 photoperiod with an illuminance of 30 lux. Biomass was maintained at 6.6 kg/m^3. The fish were fed a commercial diet (Skretting—Europa, BC, Canada) daily at a rate of 0.5% of body weight per day using automated feeders. Farmed Atlantic salmon were held under optimal conditions in 3800 L tanks as described previously [[81]31], with some modifications (at 10 °C, and a 12/12 photoperiod with an illuminance of 40–60 lux). The fish were fed three days per week at a level of 1% body weight per day using a commercial dry pellet (Skretting–Europa). 2.6. Infection Assays Wild cunners (~150 ± 50 g), cultured lumpfish (~140 ± 20 g), and cultured Atlantic salmon (~800 ± 100 g) were transferred from JBARB to AQ3 biocontainment at Cold-Ocean Deep-Sea Research Facility (CDRF) at DOS, MUN, for infection assays and acclimated for 2 weeks under the above-described conditions. Cunners and lumpfish were separated into 500 L tanks containing 60 fish per tank and Atlantic salmon were separated into six tanks containing 15 fish each (two tanks/dose; 30 fish/dose). The infection procedures were conducted according to established protocols [[82]30,[83]32,[84]33]. Briefly, fish were anesthetized with 50 mg/L MS222 per liter of seawater and intraperitoneally (i.p.) injected with 100 µL of inoculum. Cunners were injected with three doses of 4 × 10^3 (low), 4 × 10^5 (medium), and 4 × 10^7 (high) CFU/dose, and a control group (n = 60) was mock injected with PBS. In an independent experiment, lumpfish and Atlantic salmon were injected with 1 × 10^6 and 2.2 × 10^6 CFU/dose of Pseudomonas sp. J380, respectively. Mortality was monitored daily until ~30 days post-injection (dpi). Samples of liver, spleen, and head kidney were taken from moribund fish to re-isolate the pathogen. 2.7. Tissue Sampling and Analysis After infection, samples of internal organs (liver, spleen, head kidney, brain, and blood), were taken from a minimum of 5 fish euthanized with an overdose of MS222 (400 mg/L) in every sampling. Tissue samples were aseptically removed at 7, 14, 21, and 35 dpi for cunners; 7, 14, 21, and 28 dpi for lumpfish; and 7 and 14 for Atlantic salmon, based on the progression of the disease. The samples were placed into sterile homogenizer bags. Subsequently, they were weighed and homogenized in PBS in a final volume of 1 mL (weight: volume), serially diluted in PBS (1:10), and plated on TSA plates before incubating at 15 °C for 3–4 days. The following formula was used to determine the CFU of Pseudomonas sp. J380 per g of tissue [[85]34]: [MATH: CFU × gram−1 = CFU< mo> × (1/Dilution Factor) ×10 Tissue w eight (grams) :MATH] (1) 2.8. DNA Extraction and Sequencing Pseudomonas sp. J380 genomic DNA was extracted from cultures grown to mid-logarithmic phase as described in [86]Section 2.2. The Wizard DNA extraction high molecular weight kit (Promega, Madison, WI, USA) was used to extract and purify the genomic DNA. Integrity and purity of the DNA were evaluated by gel electrophoresis (0.8% agarose) and spectrophotometry (Genova-Nano spectrophotometer, Jenway, UK). Library preparations and sequencing were conducted by Genome Quebec (ON, Canada) using PacBio RS II and Miseq Illumina sequencers. 2.9. Genome Assembly, Annotation, and Mapping Celera Assembler (August 2013; version v7.0) at Genome Quebec was used to assemble the PacBio reads. Rapid Annotation Subsystem Technology pipeline (RAST) was used for the annotation [[87]35]. The Pseudomonas sp. J380 chromosome was submitted to the National Center for Biotechnology Information (NCBI) and re-annotated using the NCBI Prokaryotic Genome Annotation Pipeline (PGAP; 4.10). To detect small plasmids, the raw Illumina reads were trimmed using CLC Genomics Workbench v20.0 (CLCGWB; Qiagen, Hilden, Germany) and examined for quality using FastQC version 0.11.9 (Babraham Institute, Cambridge, UK) [[88]36]. High-quality Illumina reads were assembled using de novo tool (CLCGWB) and aligned to the reference Pseudomonas sp. J380 chromosome using the genome finishing module tools with default parameters. Illumina sequences that did not align with the chromosome were analyzed and annotated. Pseudomonas sp. J380 whole-genome illustration was developed by using CGview Server [[89]37]. 2.10. Multi-Locus Sequence Analysis (MLSA) Using Housekeeping Genes To infer the phylogenetic history of Pseudomonas sp. J380, several reference genes from different pseudomonads with complete genomes were considered. These genes included 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) and were used in MLSA. The gene loci and accession numbers for each species are listed in [90]Table S1. Sequences were aligned using CLCGWB. Concatenation of locus sequence was made using Sequence Matrix software v1.7.8 [[91]38]. The phylogenetic history was estimated using the neighbour-joining (NJ) method [[92]39] with a bootstrap consensus of 1000 replicates using MEGA X. 2.11. Whole Genome Comparison and Evolutionary Analysis The genomes used in this study are listed in [93]Table S2, and the analyses were conducted using the whole-genome analysis tool (CLCGWB) as described earlier [[94]30]. To determine the average nucleotide identity (ANI), genomes were aligned with default parameters (min. initial seed length = 15; allow mismatches = yes; min. alignment block = 100). A correlation matrix was generated with default parameters (Euclidean distance method and complete cluster linkages). The phylogenetic history was determined using NJ method [[95]39] with a bootstrap consensus of 1000 replicates using CLCGWB. Aeromonas salmonicida strain J223 ([96]NZ_CP048223) was used as an outgroup. The analysis was repeated using MEGA X using the same parameters [[97]40]. Whole-genome dot plots between closely related strains were also constructed to visualize and further analyze genomic differences. 2.12. Bacterial Growth under Iron-Enriched and -Limited Conditions, and Total RNA Extraction Triplicate cultures of 50 mL of Pseudomonas sp. J380 grown under iron-enriched and iron-limited conditions at 15 °C with aeration (180 rpm) to mid-log phase were subjected for RNA extraction according to established protocols [[98]31,[99]41]. Briefly, cells were harvested (6000 rpm, 10 min, 4 °C) and washed twice with PBS. The cell pellet was used for total RNA extraction using MirVana following the manufacturer’s instructions (Invitrogen, Waltham, MA, USA). To degrade any residual genomic DNA, the RNA samples were treated with 2 U of DNase (TURBO DNA-free™ Kit, Invitrogen) and incubated at 37 °C for 30 min. Then, 2.5 μL of DNase Inactivation Reagent was added, and further incubated for 5 min at room temperature. The supernatant containing clean RNA was harvested by centrifuging samples at 10,000× g for 1.5 min. RNA samples were quantified and evaluated for purity (A260/280 and A260/230 ratios) using a GenovaNano-spectrophotometer (Jenway, UK). RNA integrity was evaluated by agarose gel electrophoresis. The A260/280 and A260/230 ratios of purified RNA samples were 1.9–2.1 and 1.9–2.2, respectively. 2.13. Library Preparation and RNA-Sequencing For each condition, there were three biological replicates (total = 6 samples). Library preparation and sequencing were done commercially at Genome Quebec. Briefly, RNA quality was evaluated using a NanoDrop spectrophotometer (Thermo Scientific) and a Bioanalyzer 2100 (Agilent; [100]Figure S2). Libraries were generated using the NEBNext^® Multiplex Oligos for Illumina^® (Dual Index Primers Set 1; Adapter 1: 3’-AGATCGGAAGAGCACACGTCTGAACTCCAGTCAC-5’; Adapter 2: 3’-AGATCGGAAGAGCGTCGTGTAGGGAAAGAGTGT-5’) and RNA depleted (1 ng; 5S, 16S, 23S) using NEBNext^® rRNA Depletion Kit (Bacteria) according to the manufacturer’s instructions. Sequencing runs were performed on NovaSeq 6000 Sequencer (Illumina) using a NovaSeq 6000 S4 100 bp PE flow cell. 2.14. RNA-Seq Data Analysis The RNA-Seq data is available at the NCBI database under the accession number PRJNA717273. Obtained paired-end raw reads were mate-paired and filtered to remove low-quality reads using CLCGWB with default parameters (mate-paired read information, minimum distance = 1; maximum distance = 1000) ([101]Table S3). Adapter trimming was realized by CLCGWB using the trim reads tool with default parameters (quality trimming, trim using quality scores, limit: 0.05, and trim ambiguous nucleotides, maximum number of ambiguities = 2). The number of reads and nucleotides removed are indicated in [102]Table S3. The quality of the reads was checked using FastQC [[103]36] before and after trimming. Trimmed high-quality reads were, then, mapped by CLCGWB against the Pseudomonas sp. J380 genome ([104]NZ_CP043060.1) using the RNA-Seq analysis tool. Reads mapping and transcript counts were performed using the following settings: mismatch cost = 2, insertion and deletion costs = 3, minimum length fraction and minimum similarity fraction = 0.8, maximum number of hits for a read = 10, and strand-specific = both. Gene expression quantification and normalization of the mapped reads were performed by alignment-dependent expectation-maximization (EM) algorithm [[105]42] based on the RESM and eXpress methods [[106]43]. The transcript per million reads (TPM) values were, then, computed from the counts assigned to each transcript, after normalization by the trimmed mean of M-values (TMM) [[107]44]. A global correlation analysis was performed using log[2]-transformed TPM values (x + 1) of each gene under iron-enriched and iron-limited conditions. The correlation was estimated by Pearson method. Abundance data were subsequently subjected to differential expression analyses using the CLCGWB and the differential expression tool based on a negative binomial general linear model (GLM) [[108]45]. A standard selection of biologically significant differentially expressed genes (DEGs) was performed with cut-off values of log[2] fold-change (FC) ≥ |1| and false discovery rate (FDR), p ≤ 0.05. 2.15. Analyses of Enriched Gene Ontology (GO) Terms and KEGG Kyoto Encyclopedia of Genes and Genomes Pathways DEGs identified in this study under standard selection criteria were subjected to Gene Ontology (GO) and KEGG enrichment analyses by ShinyGO version 0.61 [[109]46]. GO terms and KEGG pathways with FDR, p ≤ 0.05 were considered as significantly enriched, and the top 20 elements were selected to construct graphical illustration using ggplot2 version 3.3.1 [[110]47]. A GO term-gene network analysis was performed using ClueGO plug-in [[111]48] in Cytoscape (v3.8.2). Three GO category resources (updated on 23.03.2021) for biological process (BP), molecular function (MF), and cellular component (CC) were used. 2.16. In Silico Tools Pseudomonas Genome DB (PGDB; [112]https://pseudomonas.com/) and homology search by DIAMOND Blast against other pseudomonads were used to determine the identity of unknown genes. The genes encoding the virulence factors were surveyed by homology search at the virulence factor database (VFDB) [[113]49] using other P. fluorescens strains (i.e., pf-5, pf0-1, and SBW25) as references. In addition, the