https://orcid.org/0009-0000-4535-8119
Figures
Abstract
Vibrio species are an emerging public and animal health risk in marine environments and the opportunistic bacterial pathogen Vibrio harveyi is a major disease risk for tropical aquaculture. Current understanding of virulence in V. harveyi is limited by strain-specific variability and complex host-pathogen dynamics. This study sought to integrate genomic investigation, phenotypic characterisation and in vivo challenge trials in barramundi (Lates calcarifer) to increase our understanding of V. harveyi virulence. We identified two hypervirulent isolates, Vh-14 and Vh-15 that caused 100% mortality in fish within 48 hours, and that were phenotypically and genotypically distinct from other V. harveyi isolates. Virulent isolates contained multiple plasmids, including a 105,412 bp conjugative plasmid with type III secretion system genes originally identified in Yersinia pestis. The emergence of this hypervirulent plasmid-mediated patho-variant poses a potential threat to the sustainable production of marine finfish in Southeast Asia, the Mediterranean and Australia. In addition, we observed an effect of temperature on phenotypic indicators of virulence with an increase in activity at 28°C and 34°C compared to 22°C. This suggests that temperature fluctuations associated with climate change may act as a stressor on bacteria, increasing virulence gene secretion and host adaptation. Our results utilising a myriad of technologies and tools, highlights the importance of a holistic view to virulence characterisation.
Citation: Sullivan R, Becker JA, Zadoks RN, Venturini C, Esteves AIS, Benedict S, et al. (2025) Vibrio harveyi plasmids as drivers of virulence in barramundi (Lates calcarifer). PLoS One 20(5): e0319450. https://doi.org/10.1371/journal.pone.0319450
Editor: Inbakandan Dhinakarasamy, Sathyabama Institute of Science and Technology, India
Received: February 5, 2025; Accepted: April 21, 2025; Published: May 19, 2025
Copyright: © 2025 Sullivan et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: Raw reads for both the Illumina short-read (paired) sequencing and Oxford nanopore long-read (single) sequencing have been deposited into the NCBI Sequence Read Archive under the Bioproject PRJNA1210070. Other data is within the manuscript and its Supporting information files.
Funding: This study was funded by the B Richards Fund Vet Path Bequest grant (N2151 B1038 1181185) from the Sydney School of Veterinary Science, Sydney Infectious Disease Institute seed grant Genomics in a Backpack (IRMA ID 222916) and Sydney Vietnam Institute seed grant (IRMA ID 218910) awarded to F.S. R.S is supported by an Australian Government Research Training Program (RTP) Scholarship and the Sydney Institute of Agriculture Christian Rowe Thornett, Sydney School of Veterinary Science James Ramage Wright and School of Life and Environmental Sciences James Vincent supplementary scholarships. J.A.B received funding from the School of Life and Environmental Sciences. BVL received funding from the Australian Government Department of Agriculture, Fisheries and Forestry (DAFF) as part a Northern Australia biosecurity sequencing (NABSeq) project (Grant C08586). The funders of this work had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
1. Introduction
The genus Vibrio encompasses a broad group of opportunistic and highly pathogenic marine bacteria which are anticipated to increase in prevalence and distribution as climate change continues [1]. There are more than 70 described Vibrio spp. which are considered ubiquitous in marine and brackish water worldwide and can cause disease in a wide array of species including humans, finfish, crustaceans and molluscs [2]. Whilst Vibrio spp. can exist in a wide-range of thermal conditions, they thrive in warmer environments and as average sea surface temperatures are increasing, so too has the extent and prevalence of Vibrio spp. globally [3]. In the past decade, there has been a sustained emergence and detection of Vibrio spp. in previously unaffected regions and outside expected detection windows directly in relation to increasing sea surface temperatures and changing climatic conditions [4]. Notable examples include the 2014 incursion of non-O1 V. cholera into subarctic regions of Sweden and Finland [5], the first detection of non-O1 V. cholera in British Columbia, Canada, following marine heat-waves in 2018 [3], and the increase in V. vulnificus infections associated with Hurricane Ian’s landfall in Florida in 2022 [6].
Whilst there is a clear and direct public health risk for humans with increased incursions and detections of pathogenic Vibrio spp., they also pose a major barrier to the expansion of aquaculture and the health of farmed and wild aquatic organisms [7]. Similar to the incursion of human pathogenic non-O1 V. cholera into previously unaffected regions, the extent and occurrence of Vibrio spp. that are pathogenic for aquaculture species is also increasing. Alongside new incursions and the prolonged detection of V. rotiferianus and V. jasicida in shellfish in Wales and England [8], increasing reports of V. harveyi are occurring in regions of the Mediterranean, with V. harveyi now considered one of the major disease threats to European seabass (Dicentrarchus labrax) and gilthead seabream (Sparus aurata) production [9,10]. This trend is echoed in Southeast Asia, with production losses due to vibriosis accounting for more than 7% of production costs in sea-cage barramundi in Malaysia [11]. In China, which accounts for over 30% of global aquatic animal production [12], V. harveyi was the dominant bacterial pathogen identified in diseased marine finfish farmed in the southern Guangdong and Hainan regions in 2020 [13]. This echoed earlier findings which identified V. harveyi as the dominant pathogen in diseased finfish sampled from 2015 to 2018 in China, Malaysia and Vietnam [14].
Whilst V. harveyi is typically more opportunistic than other Vibrio spp., such as V. vulnificus or V. alginolyticus, it has increasingly been detected as the causative agent of disease in a wide range of host species including seahorses (Hippocampus kuda) [15], European abalone (Haliotis tuberculate) [16], mud crabs (Scylla spp.) [17], prawns [18] and numerous finfish species [19]. Clinical signs of disease can be highly varied and differ depending on the life-stage and species affected, a comprehensive review of clinical vibriosis is provided by Mohamad and colleagues [19]. In finfish, vibriosis associated with V. harveyi commonly manifests as gastroenteritis, muscle necrosis, blindness and eye lesions, fin erosion and skin lesions [20]. This heterogeneity in clinical manifestations of infection may in part be due to the fact that V. harveyi may be reliant on co-stressors or host compromise for disease to occur [21]. To date, the study of virulence in V. harveyi has mainly focused on PCR-based analysis of known and atypical virulence genes from other Vibrio spp. such as flaC from V. anguillarum and ctxA from V. cholerae [13,22]. Although several whole genome investigations have been performed for V. harveyi [10,23,24], a holistic understanding of virulence in V. harveyi is still lacking – particularly in comparison to V. cholerae [25,26], and V. parahaemolyticus [27].
This study aims to advance our understanding of virulence in V. harveyi by combining in silico genomic characterisation with traditional in vitro phenotypic analysis and in vivo challenges using juvenile barramundi (Lates calcarifer) as a model tropical fish.
2. Methods
2.1. Sample collection and bacterial isolation
Twelve isolates from diseased fish were obtained from Australia in 2022 and Vietnam in 2023 (Table 1). Australian Vibrio isolates originated from clinically moribund farmed barramundi (Lates calcarifer) that were submitted to Berrimah Veterinary Laboratory, Northern Territory (BVL) for culture and identification of bacterial pathogens. Upon arrival at BVL, an eye and three skin swabs collected post handling and/or grading (n = 4) of moribund barramundi were inoculated onto marine agar (Difco 2216), tryptone soy agar with 5% sheep’s blood (ThermoFisher Scientific, Australia PP2166) and thiosulfate-citrate-bile salts-sucrose (TCBS) agar (ThermoFisher PP2013) and incubated at 26°C until suspected V. harveyi colony growth was observed. Vietnamese isolates were collected in June 2023, during a syndromic surveillance study in marine polyculture sea cage farms in the Quang Ning Province of Ha Long Bay (Vietnam). Using selective media (TCBS agar plates), Vibrio spp. were isolated from moribund fish showing gross external pathology consistent with bacterial infections (scale loss and skin ulcers and reddening at the base of the fins and ventral areas of the body). Fish were humanely euthanised with a rapid blow to the head and sampled by aseptically swabbing the head-kidney with a 1 µL inoculating loop or for one fish, the caudal margins of a large skin lesion onto TCBS agar for selective isolation of Vibrio spp. Eight Vibrio spp. displaying characteristic yellow colonies on TCBS plates, were identified as Vibrio harveyi by matrix-assisted laser desorption-ionization time of flight mass spectrometry (MALDI-ToF MS; ThermoFisher Scientific). These isolates came from six out of thirty fish sampled during this surveillance study and included spotted scat (Scatophagus argus) (n = 1), rabbitfish (Siganus gutatus) (n = 2) and grouper (Epinephelus fuscoguttatus) (n = 3). The eight isolates were passaged onto marine agar (Difco 2216) twice and incubated at 28°C to obtain pure colonies. All twelve suspected V. harveyi isolates (n = 4 collected from Australia and n = 8 collected from Vietnam) were identified to species level by MALDI-Tof MS [28] before sequencing as described in Section 2.2.
2.2. Virulence typing in silico
Bacterial DNA was extracted from the isolates using the MagMAX CORE Nucleic Acid Purification Kit (327-000) on the KingFisher Flex System (ThermoFisher Scientific) or with the Quick-DNA Miniprep Plus Kit (ZymoResearch D4068, USA) according to the respective manufacturer’s protocol. DNA quality was checked with a QubitTM fluorometer (ThermoFisher Scientific). Following extraction, short-read libraries were prepared with the Illumina DNA Prep kit and sequenced as 300-bp paired-end reads on an Illumina MiSeqTM (Illumina, USA) at BVL to a depth of at least 60x coverage. Following sequencing, genomes were de novo assembled using SPAdes v3.15.5 [29] with default parameters and annotated with Prokka v1.14.1 [30]. After assembly, the average nucleotide identity (ANI) of the genomes was assessed using FastANI v1.32 [31] against the NCBI reference genomes for V. harveyi (SB1, NCBI: PRJNA972608) and V. campbelli (BoB-53, NCBI: PRJNA429202) because of high phenotypic and genomic similarities between the species [32,33]. Antimicrobial resistance (AMR) genes were investigated in the short-read assemblies of V. harveyi isolates using AMRFinder v3.12.8 [34] with default parameters.
Long-read sequencing was performed for the Vietnamese isolates with a R10.4.1 flow cell (FLO-MIN114) on a MinIONTM Mk1C (Oxford Nanopore Technologies, UK) and libraries were prepared with the Rapid Barcoding Kit V14 (SQK-RBK114). Following sequencing, POD5 files were basecalled in MinKNOW v.24.02.8 using Dorado super accuracy basecaller model (SUP; dna_r10.4.1_e8.2_400bps_5khz_sup) enabling barcode and adapter trimming at basecalling. Raw fastq files were then assembled using the Genomics in a Backpack pipeline [35] developed by the Sydney Informatics Hub in collaboration with the Sydney School of Veterinary Science (https://github.com/Sydney-Informatics-Hub/ONT-bacpac-nf). Briefly, the quality of raw reads was assessed using NanoPlot v.1.42.0 [36] and PycoQC v2.5.2 [37]. Reads were then screened with Kraken2 v2.1.3 [38] to check for contamination prior to assembly. Where possible a hybrid assembly was performed by generating a consensus assembly using Flye v2.9.3 [39,40] and Unicycler v0.4.8 [41]. If the two long-read assemblies could not be hybridised, only the Flye assembly was kept. For isolates which contained plasmids, the plasmids were assembled using Plassembler v1.6.2 [42]. Following consensus assembly, all contigs were annotated with Bakta v1.9.2 [43]. The presence of virulence genes was assessed using Abricate v1.0.1 (https://github.com/tseemann/abricate) and the virulence factor database (VFDB) housed within abricate (https://doi.org/10.1093/nar/gkv1239). Prophage content of the genomes was predicted in the long-read assemblies using PHASTEST v3.0 [44].
To investigate virulence in silico, the protein-coding sequences of virulence genes commonly described in V. harveyi, or identified in other Vibrio species (atypical to V. harveyi) were curated into a localised BLAST database for BLASTx analysis of the Prokka-annotated short-read assemblies. In this analysis, publicly available V. harveyi genomes were also included (see below under phylogenetic investigation). Forty-one virulence genes were included as identified in prior studies [22,45–49] and an additional eight genes were included following manual identification in silico and via Abricate VFDB output resulting in 49 protein-coding sequences for investigation (S1 Table). All confirmed V. harveyi genomes included for phylogenetic analysis, including the NCBI reference genome (SB1), were analysed using blast + /2.12.0 with flag -task blastx and -evalue 0.00001. Furthermore, for the plasmids identified via long-read sequencing using Plassembler in Vh-14 and Vh-15, an additional BLASTx was performed as described above against the protein-coding sequences for the reference genomes of Yersinia enterocolitica (Y11, NCBI: SAMEA2271938), Y. pestis (A1122, NCBI: SAMN02603531) and Y. pseudotuberculosis (NCTC10275, NCBI: SAMEA4442458) following the identification of ypkA in plasmids within these isolates. Visualisation of the plasmids was performed using Proksee [50] and the gggenes package in R [51] for further visualisation of the locus of interest.
Phylogenetic investigation of the isolates was conducted using Snippy v4.1.0 [52] by identifying single nucleotide polymorphisms (SNPs) in each isolate against the V. harveyi reference genome (SB1) and generating a core genome SNP alignment with command snippy-core. For comparative analysis, publicly available genomes were identified and screened in the National Centre for Biotechnology Information (NCBI) Sequence Read Archive (S2 Table). Genomes were included if isolates originated from vertebrates or environmental samples, included geographical and isolation source data, were sequenced on an Illumina platform and fastq files were readily available. Genomes (n = 28) were downloaded and annotated using Prokka as above. Following SNP identification for all genomes, a recombination free-alignment was created using Gubbins v3.3.5 [53] and a maximum-likelihood phylogeny constructed in IQ-TREE v1.6.12 [54] using a GTR model and 1000 bootstrap replicates [55]. The tree was then annotated in FigTree v1.4.4 [56] and Microreact [57].
2.3. Virulence typing in vitro
For the V. harveyi isolates collected from moribund fish in Vietnam and V. harveyi isolate TCFB-0558 (NCBI: PRJNA940306) [21], eight phenotypic assays were performed as well as infrared biotyping using Fourier Transform infrared spectroscopy (Fig 1). All isolates were grown aerobically from cryo-preserved frozen aliquots on marine-agar (Difco 2216) at 28°C overnight, and passaged onto a new marine agar plate and incubated aerobically overnight at 28°C. This subculture was then used for testing of lipase, phospholipase, caseinase, haemolysis and urease activity. For each assay, each isolate was stabbed into three different plates to ensure three replicates, and this was repeated for three temperatures: 22°C, 28°C and 34°C. These three temperatures were selected for incubation based on the typical thermal range experienced by barramundi in pond production systems [58,59] and to subsequently explore realistic temperatures in which V. harveyi may persist in aquaculture settings [60]. For each plate-based assay, E. coli (ATCC 11775) was used as a negative control. For urease, Yersinia enterocolitica (ATCC 23715) was used as a positive control and for the other tests, P. aeruginosa (NCTC 10662) was used as the positive control. For the other phenotypic assays, one colony of each subculture was inoculated into marine broth (Difco 2216) and incubated overnight at 28°C. Density of the overnight broth cultures was standardized by diluting the suspensions in sterile PBS to an OD600 of 0.5 before testing swarming, gelatinase and biofilm formation activity. For gelatinase and biofilm testing, each isolate was inoculated into three separate tubes or plates to ensure three replicates, and this was repeated for each temperature. To evaluate swarming, five replicates were performed for each isolate at each temperature. For gelatinase, E. coli (ATCC 11775) was used as a negative control and for biofilm formation, uninoculated marine broth was used as the negative control. P. aeruginosa (NCTC 10662) was used as the positive control for each tube-based test. A detailed description of each test, including composition of the media, is provided in S1 Appendix. Plates were observed every 24 hours and activity zones were measured using a ruler at 24, 48 or 72 hours as specified in S1 Appendix. This was true for all assays with the exception of urease (read at 48 hours only), biofilm formation (peg-based assay, see S1 Appendix) and infrared biotyping. For infrared biotyping, isolates were grown on marine agar from frozen aliquots and incubated for 24 hours at 28°C, isolates were then sub-cultured on marine agar and incubated again for 24 hours at 28°C. From this second passage, colonies were collected for biotyping and plated in triplicate for each isolate. To ensure enough biological replicates for each isolate alongside sufficient technical replicates (n = 3), sub-culture from the original passage was repeated and incubated for 24 hours at 28°C (n = 3).
Graphical summary of phenotypic tests (lipase, phospholipase, caseinase, haemolysis, urease, swarming, gelatinase, biofilm formation and infrared (IR) biotyping performed for Vibrio harveyi isolates Vh-14, Vh-15, Vh-26, Vh-28, Vh-31, Vh-33, Vh-47 and TCFB-0558 and positive and negative assay controls. Swarming was tested using five replicates, all other assays were performed in triplicate, and conducted at 22°C, 28°C or 34°C. Created in BioRender. Sullivan, R. (2025) https://BioRender.com/k48o607.
2.4. Pathogenicity in vivo
2.4.1. Animal ethics statement.
All animal procedures were approved by The University of Sydney Animal Ethics Committee (AEC) under AEC project number 2023/2397 according to the Australian Code of Practice for the Care and Use of Animals for Scientific Purposes (8th edition, 2013). All researchers and staff involved in the handling and monitoring of fish received animal ethics training provided by the University of Sydney prior to the trials and had experience in fish husbandry. Fish were monitored twice a day for seven days post infection. All anaesthesia and euthanasia procedures were performed by hand-netting fish into a small tank which had been dosed with the appropriate amount of benzocaine required for juvenile barramundi: anaesthesia (60 mg L-1), and euthanasia (>150 mg L-1). Any fish showing severe clinical disease (e.g., lethargy, weak or erratic swimming, keeling on the side and minimal response to stimulus) were euthanised as quickly as possible with an overdose of benzocaine (>150 mg L-1), and counted as a mortality for analysis in accordance with animal welfare and ethics requirements of the project. Any fish remaining in holding tanks following the experimental period were hand-netted and humanely euthanised using an overdose of benzocaine (>150 mg L-1).
2.4.2. Fish care and maintenance.
Experimental challenge studies were performed in juvenile barramundi to complement in silico and in vitro virulence characterisation. Juvenile fish were selected due to the reported high susceptibility of juvenile finfish to severe V. harveyi infections [9,61] and the feasibility of keeping smaller fish in our challenge facility. Fish were purchased from a commercial hatchery (Mainstream Aquaculture, Werribee, Victoria) and acclimated for at least two weeks prior to experimental challenges. Three hundred and fifty fish (ca. 0.5 g) were divided into three 500 L circular tanks containing 350 L of seawater (30 ppt) (Red Sea Salt, Red Sea Aquatics) and fed a commercial diet (Otohime C2, BioKyowa, USA) at 8% body weight twice daily for two weeks during the acclimation period. Grading was performed weekly, and fish were grouped based on size to reduce cannibalism. Fish were kept under a 12 hour light:dark schedule and water quality was monitored daily with temperature maintained at 28 ± 1°C, pH at 8.0 ± 0.2 and dissolved oxygen above 6 mg L−1. Water quality parameters were monitored using colorimetric test kits (API Saltwater Master Test Kit) and partial water changes were performed as required to maintain total ammonia (<1.0 mg L−1), nitrite (<0.5 mg L−1) and nitrate (<80 mg L−1) at specified levels.
An experimental trial (Experiment 1) was conducted to assess virulence of the V. harveyi isolates in vivo with two biological replicates from the same cohort of fish. In Replicate 1, fish had a mean weight of 4.5 g ± 1.7 g (mean + SD) and a total length of 74.4 mm ± 8.5 mm, and in Replicate 2, fish had a mean weight of 11.9 g ± 3.4 g and a total length of 103.8 ± 10.0 mm. All fish used in the experimental challenge trials were apparently healthy based on visual assessment of external and internal gross pathology and the absence of systemic bacterial infection in the population was confirmed by random sampling of three fish from each of the holding tanks and plating of a head-kidney swab onto TCBS and tryptone soy agar with 5% sheep blood (ThermoFisher PP2166). Plates were incubated for up to 72 hours at 28°C to confirm the absence of bacterial colonies.
Following confirmation of the health status of fish in the holding tanks, fish were hand netted and transferred to the experimental aquaria for five days of acclimation prior to challenge. Experimental aquaria (100 L) contained 90 L of seawater and water quality was monitored daily to maintain conditions. Each experimental tank was set-up with a biofilter and filtration unit (Aquaclear Power 285 Filter 110), an airstone and air pump (Aqua One stellar 200D Air Pump) for aeration, a heater (Aqua One 200W quartz heater) to maintain water temperature, and a UV steriliser (AA Aquarium, 12V GKM24W). Water quality was monitored daily and maintained as above, the room was temperature-controlled with an air-conditioning system (Hyper Inverter, Mitsubishi Heavy Industries, Australia) to assist in maintaining water temperature. Each tank contained 10 fish to minimise inter-fish aggression.
2.4.3. Challenge trials.
Experiment 1 (Replicates 1 and 2) involved randomly assigning nine tanks to one of seven V. harveyi isolates collected by the University of Sydney, a known virulent V. harveyi isolate (TCFB-0558; [21]), or a negative control (sham-inoculated with phosphate-buffered saline [PBS]) (n = 180 fish in total). Replication (n = 2) for each isolate was achieved by replicating the experiment over time rather than using spatial replication. Using a challenge concentration of approximately 3.5 × 108 CFU mL-1, based on previous findings for isolate TCFB-0558 [21], we assessed whether these isolates could cause mortality in in juvenile barramundi without additional stressors. For the isolate with the highest and most rapid mortality across the two replicates, a dose-titration experiment was subsequently conducted (Experiment 2).
Inoculum preparation was performed according to the protocol of Samsing and colleagues [21]. Briefly, isolates were grown as a lawn on marine agar at 28°C. The suspensions were harvested, resuspended, homogenized and diluted to an OD600 of ca. 0.250 that was estimated to be 3.5 × 108 CFU mL-1. The CFU count was retrospectively determined using the drop count method [62] on 3% marine agar to minimise swarming. Dilutions were only counted if they contained between 5–80 colonies per drop to minimize variability.
For Experiment 1, fish were hand-netted and anaesthetised with benzocaine (60 mg L-1) prior to intra-muscular (IM) injection in the left flank directly under the dorsal fin with 50 µL of adjusted V. harveyi inoculum in PBS giving a final inoculum per fish ranging from 1.4–2.2 × 107 CFU fish-1, or 50 µL of sterile PBS (S3 Table). Fish were monitored twice a day for seven days post infection. Any fish showing severe clinical disease (e.g., lethargy, weak or erratic swimming, keeling on the side and minimal response to stimulus) were euthanised as quickly as possible with an overdose of benzocaine (>150 mg L-1), and counted as a mortality for analysis in accordance with animal welfare and ethics requirements of the project. During the experimental period, all moribund and dead fish were sampled to confirm V. harveyi infection and the head-kidney swabbed with a 1 µL inoculating loop onto TCBS plates. For fish with visible skin lesions, the caudal margins of the lesion were also swabbed and streaked onto plates. On day 8, the remaining fish in each tank were euthanised, and up to 4 fish per tank (depending on the number of fish left in each tank) were sampled at random to assess infection by culturing head-kidney swabs on TCBS. All TCBS plates were incubated for up to 48 hours at 28°C and inspected for growth. To confirm V. harveyi infection, colony PCR was performed on a pool of 4–6 colonies from each plate. Colonies were collected using a 10 µL loop and resuspended in 300 µL of sterile PBS and flick mixed before being heated at 56°C for 10 minutes to lyse the cells. This was then centrifuged at 17,000 g for 10 minutes and the supernatant transferred to a new tube and stored at -20°C until required. This supernatant was used as the template for a conventional PCR using the toxR gene primers (382 bp product) from the protocol of Pang and colleagues [63]. Briefly, a 12.5 µL reaction was prepared containing 1 µL of template DNA, 2.5 µL of 5x MyTaqTM Red Reaction buffer, 0.25 µL of MyTaq Red DNA Polymerase, 0.5 µM of forward and reverse primer and nuclease-free water. Cycling conditions included an initial denaturation at 95°C for 1 minute followed by 32 cycles of 95°C for 15 seconds, 59.5°C for 15 seconds and 72°C for 10 seconds and a final extension at 72°C for one minute in a thermocycler (Bio-Rad 382 T100 Thermal Cycler, USA). A non-template control (no DNA template) was included in each run and a 2% agarose gel was run to confirm the presence of a band of the correct size (382 bp).
Based on results from Experiment 1, a dose titration trial was performed (Experiment 2) using 10-fold dilutions of isolate Vh-14 such that the final doses were 4.0 × 108, 3.0 × 107, 3.0 × 106 and 3.2 × 105 CFU mL-1. Fish in Experiment 2 had a mean weight of 22.8 g ± 5.7 g (mean + SD) and a total length of 127.8 mm ± 11.6 mm. Fish were anaesthetised with benzocaine (60 mg L-1) prior to IM injection and were injected with 50 µL of Vh-14 adjusted for a final inoculum per fish between 104–107 CFU fish-1. The experimental design consisted of two replicate tanks for each challenge dose and one tank for the negative control (sham inoculation with PBS) and each tank contained 10 fish (n = 90 fish total). After monitoring fish for eight days, on day 9, remaining fish were euthanised, and up to 4 fish per tank were randomly sampled to assess infection by culturing head-kidney swabs on TCBS plates followed by colony PCR as described for Experiment 1.
2.5. Statistical analysis
For infrared biotyping and the measurements from the IR Biotyper system, principal coordinate analysis (PCoA) was performed in R v4.2.1 [64] using cmdscale in stats 3.6.2 [65] and ggplot2 v3.5.1 [66] for visualisation. For experimental challenges, survival analysis was conducted using the Kaplan-Meier method and comparisons between survival curves made using Cox proportional hazard models followed by log-rank test and statistical significance was defined as p-value < 0.05. Moribund or dead fish were defined as experiencing the event of interest if they had clinical signs consistent with vibriosis and growth from a head-kidney swab on TCBS that was confirmed as V. harveyi by colony PCR. For Experiment 1, which tested different V. harveyi isolates at one dose (ca. 3.2 x 108 CFU mL-1), isolate ID was fitted as a fixed effect. For the second experiment, which was the dose titration study of Vh-14, the Cox proportional hazard model was fitted with dose as a fixed effect and tank as a random clustering factor to incorporate tank variation into the model. All analysis was performed in R v4.2. and survival analysis was conducted with packages survminer v0.5.0 [67], survival v3.6.4 [68], coxme v2.2.20 [69].
3. Results
3.1. In silico predictors of virulence
MALDI-ToF MS identified all newly obtained isolates sequenced in this study as V. harveyi (S4 Table) however average nucleotide identity (ANI) analysis showed that one isolate (Vh-24) had a 97.3% ANI to the V. campbellii reference strain (BoB-53) compared to 89.9% similarity to the V. harveyi reference strain (SB1). The remaining isolates sequenced in this study were confirmed as V. harveyi with > 97% ANI to SB1. Due to discrepancies in identification, we re-analysed the 28 genomes sourced from the SRA using fastANI leading to the re-classification of three as V. campbellii: MWU1063 (SRR25169360), CUB2 (SRR2931635) and RT-6 (SRR5410471). Inclusion of these isolates for phylogenetic investigation resulted in a distinct outgroup from the other strains (S1 Fig) and recombination could not be calculated for these isolates due to the significant distances between SNPs when compared to the other isolates. Vh-24 was therefore excluded from phenotypic and in vivo study and all V. campbellii genomes were removed from phylogenetic and virulence gene investigation resulting in 37 genomes remaining, including the V. harveyi reference genome SB1.
For short-read sequencing of the V. harveyi isolates collected from Vietnam, the mean number of bases per isolate was 5,863,706 (± 250,299) with a mean N50 of 378,834 bp and high coverage (mean 93x; S4 Table). For long-read sequencing, the total number of bases sequenced was 7.23 Gb across 1,945,996 reads with an average genome size of 5,870,692 bp (± 154,712) across the isolates. The N50 read length was 6,850 bp with a median read length of 2,145 bp and a median Phred score of 19.34. Both sequencing methods revealed an average GC content of 45%. Vh-14 and Vh-15 had larger genomes than those of the other isolates from Vietnam for both short (average 6,277,849 ± 4302 bp for Vh-14 and Vh-15 versus 5,713,333 ± 32,459 bp for the rest, respectively) and long-read sequencing (average 6,095,344 ± 569 bp versus 5,778,011 ± 22,549 bp, respectively), due to the presence of additional plasmids (S4 Table).
Plasmids > 5000 bp in length were detected via long-read sequencing in three of the isolates: Vh-14, Vh-15 and Vh-31 (S5 Table). In Vh-14 and Vh-15, eight and nine plasmids larger than 5,000 bp were assembled with the long-read data, respectively, and confirmed with Plassembler. Whilst four of the plasmids in Vh-14 were circular according to Plassembler, only one small 5,097 bp sequence was circular in Vh-15. In addition, Plassembler identified two circular plasmids of 61,899 and 24,313 bp in Vh-31. In the absence of circularisation, a match between each plasmid and the plasmid database (PLSDB) [70,71] with a mash distance of 0.1 or less was considered a key criterion for inclusion as a true plasmid [42]. Vh-14 had six hits to five distinct plasmids in PLSDB whilst Vh-15 had five hits to three plasmids (S5 Table). In Vh-14, the largest non-circular plasmid with a match in PLSDB was 105,412 bp in length and contained 131 coding sequences and two non-coding RNAs encoding for the sok antitoxin regulator, which is involved in plasmid maintenance and stability in conjunction with the hok gene [72].
Across all isolates sequenced from Australia and Vietnam and the V. harveyi genomes downloaded from the NCBI, the same three antimicrobial resistance genes (S6 Table): bla(VHH-1), a type A beta-lactamase that facilitates hydrolysis of beta-lactams [73]; Tet(34) which is proposed to inhibit tetracycline activity via activation of Mg2+-dependent purine nucleotide synthesis [74]; and Tet(35), an efflux pump involved in transporting tetracycline out of the cells [75] were present. Aside from these three universal antimicrobial resistance genes, twelve other genes were detected across the short-read assemblies, although the majority of genomes (30/37) contained only blaVHH-1, Tet(34) and Tet(35). The proportion of prophage content detected in the long-read sequences was variable across the isolates yet all isolates except Vh-33 contained at least 1 putative prophage region (S7 Table). Vh-14 and Vh-15 had the highest number of chromosomally dispersed prophage regions with three intact phage regions each.
Of the 49 protein-encoding virulence genes that were investigated, 38 were detected in all V. harveyi genomes analysed, indicating a large proportion of virulence genes are conserved and form part of the essential virulome of V. harveyi (Fig 2), while three atypical virulence genes originally identified in V. cholerae (ctxA, tcpA) and V. parahaemolyticus (trh) were not identified in any of the genomes. Four virulence genes were detected in one or two genomes only, including vhml (SRR11841529 and SRR17393361), hylA (SRR11841535), vvh (SRR11841535) and ypkA (yopO) (Vh-14 and Vh-15). There were five genes with more variable presence across the genomes including zot (present in 24 isolates), ureB and ureG, which encode for urease production (present in 32 isolates), and virD4 (present in 22 isolates). The average number of virulence genes present in each genome was 41 (± 1) and all genomes in this study encoded a type III secretion system on Chromosome I (vscN, vscF, vscB, vcrD and vcrH). In addition to this chromosomal type III secretion system, Vh-14 and Vh-15 were the only two to encode ypkA – a serine/threonine kinase and effector protein within an additional type III secretion system (Lee et al., 2017). The 2,142 bp ypkA gene was located on the large 105,412 bp conjugative plasmid of Vh-14 and was also in a fragmented plasmid in Vh-15 (Fig 3A). YpkA is a crucial virulence gene in Yersinia species [76]. Several additional genes involved in the type III secretion system, yopH, lcrG and lcrQ and a type III chaperone were identified through manual screening against the Yersinia spp. reference genomes (Fig 3B). Surrounding the type III secretion system were numerous transposases and an integrase which can assist in the uptake and stabilisation of genes within bacterial plasmids [77].
Presence-absence matrix for the occurrence of 49 protein-encoding virulence genes across 37 Vibrio harveyi genomes including the V. harveyi reference genome (SB1, NCBI: PRJNA972608; bolded). BLASTx was performed to determine whether a gene was present or absent within the genomes with a cut-off e-value of < 0.00001. Figure generated using ggplot in R.
A). Map of 105,412 bp plasmid identified in Vh-14. Major genes of interest are labelled, and colours correspond to the legend on the right of the circular map. An additional blastn was performed against Vh-15 and is the outermost ring on the map with identity to Vh-14 shaded in pink. B) Annotated map of genes of interest carried on 105,412 bp plasmid in Vh-14 and Vh-15 which contains components of a second type III secretion system. Genes annotated with Bakta or via manual BLASTx investigation of the plasmid against the Yersinia species reference genomes. Protein-coding genes of interest (yopH, lcrG, lcrQ, ypkA) are labelled and putative function of these genes and the surrounding ones are colour-coded. Figures generated in Proksee with genes annotated via Bakta or manual BLASTx investigation and with gggenes package in R.
Whilst isolates collected from the same disease investigations and hence typically the same site clustered together on the phylogeny, there was no other clear geographical pattern to phylogenetic relations (Fig 4). Two of the isolates collected in Vietnam (Vh-14 and Vh-15) clustered together with three isolates sequenced from moribund barramundi in Australia (1W, 3W and 5W) and formed a sub-clade with the reference (SB1) genome and SRR25067778. When compared to the reference genome SB1, the Vietnamese isolates Vh-14, Vh-15 and the three Australia genomes 1W, 3W and 5W had large levels of genomic recombination.
Maximum-likelihood phylogenetic tree of 37 V. harveyi genomes including the NCBI reference genome (SB1, NCBI: PRJNA972608; bolded). Tip colour indicates source of isolate (fish or environmental), and external blocks indicate country of origin, presence of ureB/ureG (Urease), virD4, ypkA and zot genes (indicated by darker shading) and homologous recombination. Scale bar indicates average number of base substitutions per 1000 nucleotides. Figure generated using Microreact (https://microreact.org/).
3.2. In vitro virulence characteristics
For most phenotypic tests, including caseinase, gelatinase, lipase, phospholipase and swarming, there was an evident effect of temperature on observed activity, which was typically higher at 28°C or 34°C than at 22°C (Fig 5A). Interestingly, biofilm production was the inverse with the highest levels of biofilm formation observed at 22°C for five of the eight V. harveyi isolates (S8 Table). Although activity generally increased at higher temperatures, a large degree of individual variation was observed between isolates. This was most evident with haemolysis with only P. aeruginosa (the positive control) and Vh-31 producing complete β-haemolysis on sheep blood agar after 72 hours. For negative controls, there was no growth or activity with the exception of E. coli at 28°C on casein-specific media. At 28°C, small amounts of E. coli growth were observed below the agar following stabbing but without the formation of true clearing zones and hence was not considered true caseinase activity. Urease activity was observed in all isolates and at all temperatures with the exception of Vh-14 and Vh-15 (S8 Table). These isolates were urease negative at all three temperatures and lacked urease genes: ureB, ureG_1 and ureG_2 identified in the other isolates in silico.
A) Bar plots depicting the variation in phenotypic activity of Vibrio harveyi isolates at three temperatures (22°C, 28°C and 34°C). Bar colours indicate different isolates on the x-axis; measurements of interest on the y-axis are diameter in mm of clearing or opalescent zones for caseinase, gelatinase, lipase, phospholipase and swarming assays and OD590nm for biofilm formation). Bar height is representative of the mean measurement of three replicates (five for swarming) with standard error bars. Where no negative control bar (light grey) can be seen, results were negative. B) Principal Coordinate Analysis plot of V. harveyi isolates investigated with the IR Biotyper system (Bruker Daltonics). Each isolate was typed in triplicate, by performing the run on three separate days (biological) with three replicates (technical) on each day.
Fourier Transform Infrared Spectroscopy was performed to assess the extent of each isolates’ distinct ‘fingerprints’. Similar to the genomic and other phenotypic findings, Vh-14 and Vh-15 were highly distinctive (Fig 5B). Based on PCoA, they were highly similar with an average Euclidean distance of 6.30 to one another, below the IR Biotyper determined clustering cut-off of 6.93, but with an average distance of 109.87 to the other isolates (S9 Table). Vh-33 was also distinct from the other isolates with an average distance of 85.49. The remaining isolates clustered together in one cluster with an average distance of 35.98 from one another. Nested within this was a subcluster formed by Vh-28 and Vh-47 with an average distance of 3.38, below the IR biotyper clustering cut-off.
3.3. In vivo experimental challenges
In Experiment 1 (n = 2 replicates per isolate), there was significant (p-value < 0.001) and rapid mortality in juvenile barramundi injected with Vh-14 and Vh-15 with 100% cumulative mortality in less than 48 hours (Fig 6A) regardless of fish size. Significant cumulative mortality (p-value < 0.001) was also observed in fish infected with TCFB-0558 (positive control) with 70 ± 10% standard error (mean ± SE), although mortality with this isolate was lower in large fish than small fish (p-value = 0.006, pairwise log-rank test). Several isolates caused low level mortality in vivo, e.g., a cumulative mortality of 9.5 ± 6.4% was observed for Vh-33, but this was not significantly different from the negative control (p-value = 0.24, pairwise log-rank test). Lower mortalities (one fish recorded for Vh-26 and Vh-47) were also not significantly different (p-value = 0.40) from the negative control group. No mortalities were recorded for Vh-28 or Vh-31 across the replicates. In Replicate 2, one fish from the negative control tank was euthanised for animal welfare reasons after presenting skin lesions around the head consistent with inter-fish aggression and cannibalism in barramundi. The head-kidney of the euthanised control fish was swabbed onto TCBS and sheep blood TSA to confirm the absence of infection. No colonies grew from the head-kidney swab or from subsequent swabs from any of the control fish across all experiments.
(A) Kaplan-Meier survival curves for intra-muscular (IM) challenge trials with eight Vibrio harveyi isolates in juvenile barramundi (Lates calcarifer). Control fish (n = 20) were IM injected with sterile PBS, whilst V. harveyi challenged fish (n = 10/tank per replicate) were IM injected with ca. 3.2 × 108 CFU mL-1. (B) Fish experiencing scale loss (red arrow) and gastroenteritis, visible as white casts (blue arrow), 3 days post-challenge in Experiment 1 (replicate 1), scale bar = 2 cm. (C) Kaplan-Meier survival curves for Experiment 2, with IM challenge doses of 105 to 108 CFU mL-1 of isolate Vh-14. Censored events such as survival or a lack of V. harveyi confirmed infection are indicated by a ‘+’.
Across all the isolates tested in vivo, clinical signs were observed as early as six hours post injection and continued until the conclusion of the observation period on day 8. Clinical signs were highly varied between individuals and included scale loss, skin discolouration, lethargy, inappetence and erratic swimming. As infection progressed, increased severity of the above clinical signs and additional ones such as gastroenteritis (evidenced by white casts protruding from the vent, Fig 6B), loss of equilibrium and skin and fin ulceration were observed (Fig 6B). Ulceration and scale loss continued and resulted in the sloughing off of skin and muscle from 3 days post-injection. Despite the skin sloughing and exposure of the subcutaneous tissue, most fish continued to eat and swim normally and during the observation period were actively responding to stimuli. At the completion of the trial, many fish had experienced substantial wound-healing of their lesions. There was an apparent effect of fish length and weight on mortality kinetics with smaller fish experiencing more rapid mortality compared to larger fish. Across the two replicates, the nature and onset of clinical signs was consistent across fish sizes as was mortality for each isolate.
In Experiment 1, head-kidney swabs from 53 of the 56 (95%) moribund and dead fish yielded growth on TCBS plates, with characteristic yellow, smooth and round colonies that were all confirmed as V. harveyi by conventional PCR. The three fish (two injected with Vh-14 and one injected with Vh-15) that did not develop positive head-kidney swabs died within the first 12 hours following challenge, displaying behavioural and clinical signs consistent with vibriosis (keeling on their side, unresponsive to stimuli, darker coloration of the skin). Of the head-kidney swabs collected at the completion of the trial, 93% of fish (41/44) presented growth on TCBS plates, with 87% (36/41) confirmed as V. harveyi by conventional PCR.
Experiment 2 investigated the dose kinetics of Vh-14 based on the rapid and high cumulative mortality observed for Vh-14 in Experiment 1. In comparison to our prior challenge model study with TCFB-0558 [21], Vh-14 caused mortality at lower doses with significantly lower survival for Doses 1, 2 and 3 (p-value < 0.001) in comparison to the negative control treatment (Fig 6C). For Dose 1 (4.0 × 108 CFU mL-1), total cumulative mortality was 79 ± 9.4% across the trial and mortality occurred later than in the smaller fish used in Experiment 1. Mortalities were lower (26 ± 10% and 5 ± 4.6%, respectively) and occurred later (4- and 5-days post injection) for Dose 2 (3.0 × 107 CFU mL-1) and Dose 3 (3.0 × 106 CFU mL-1) compared to Dose 1. Clinical signs were consistent with those observed in Experiment 1 and additional gross internal pathology was observed at the conclusion of the trial including splenomegaly and discoloration of the liver in fish receiving Dose 1 or 2. Fish injected with Dose 4 (3.2 × 105 CFU mL-1) experienced some lethargy following injection but exhibited no other clinical signs and were responsive and feeding within 24 hours with no detectable V. harveyi in the head-kidney or skin at the completion of the trial. All moribund fish and dead fish recorded in Experiment 2 were positive for V. harveyi by culture and conventional PCR, with the exception of one fish that died overnight and had started decomposing, resulting in release of gastrointestinal bacteria. At the end of the trial, all fish sampled in Dose 1 and Dose 2 had growth on TCBS plates that was confirmed by PCR as V. harveyi, except for one head-kidney swab taken from a fish exposed to Dose 2. For Doses 3 and 4, 37.5% (6/16) of swabs taken from fish yielded growth on TCBS plates with characteristic colonies, but only one from Dose 3 was confirmed as V. harveyi.
4. Discussion
In the face of climate change and rising sea-surface temperatures, an increase in the prevalence of Vibrio spp. worldwide is already occurring [78,79]. Increasingly attention to the threat of Vibrio spp. that are relevant to public health, such as V. vulnificus [80], non-O1 V. cholerae [3] and V. parahaemolyticus [81] is occurring as human case numbers rise. In contrast, the impacts of Vibrio spp. on sustainable aquaculture and food security have received limited attention, despite the recognition of the genus as a dominant bacterial agent associated with disease and mortality in finfish and crustacean aquaculture globally [10,13,46]. Although some virulence characteristics of V. harveyi have been described [82], it remains unclear why certain strains cause severe disease while others remain avirulent unless the host experiences considerable stress [21,83]. Our study sought to integrate genomic analyses, conventional phenotypic assays and in vivo trials in barramundi fingerlings to expand understanding of V. harveyi pathogenesis.
Across all isolates, phenotypic activity indicative of virulence was increased at higher temperatures with the exception of urease activity and biofilm formation. Whether this effect is mediated simply by preferred growth conditions or virulence gene expression remains unclear. Evidence for increased growth within the first 48 hours at higher temperatures was observed for V. harveyi cultured in seawater at 25°C and 30°C compared to 20°C [84] and at 30°C compared to 26°C [85], although long-term this was associated with a decrease in seawater culturability after two weeks. Transcriptomic analysis is suggestive of higher virulence gene expression for V. harveyi at 30°C with increased expression of chemotaxis and several type III and IV secretion systems compared to 26°C [85]. This change in gene expression was also observed for V. vulnificus at higher temperatures (28°C vs 20°C) with the activation of metabolic components such as chemotaxis and proteases to assist in host colonisation within European eels (Anguilla anguilla) [86]. Increases in temperature could therefore act as a stressor and in conjunction with other stressors such as high salinity or pH contribute to the increased secretion of virulence components and host adaptation as previously shown for the food-borne Enterococcus spp. [87] and the fish pathogen Aeromonas hydrophila [88]. The in vivo effect of temperature on virulence has been seen with an increase in mortality of red tilapia (Oreochromis sp.) when challenged with Streptococcus agalactiae [89]. Similar effects have been observed in Pacific oysters (Crassostrea gigas) with increases in mortality and the presence of Vibrio spp. in the microbiome during simulated marine heatwaves [90] and summer mortality outbreaks [91]. The effect of thermal stressors on increasing virulence is supported by predictive modelling showing that an increase of just 1°C environmentally could lead to a 3.5% increase in mortality of aquatic organisms [92]. Despite the expression of virulence characteristics in vitro, most isolates did not cause mortality in our challenge study, although they may have contributed to the disease syndromes in farmed marine fish in Vietnam observed during the isolation process. In addition to exerting pressure on the pathogen, stressors play a major role in host response and susceptibility to V. harveyi. Typically co-stressors such as skin lesions [93] or thermal stressors [21] are necessary for disease to occur in otherwise healthy animals. Without testing the effect of these stressors alone or in combination, experimental infection models cannot always replicate the mortality kinetics observed on farm. Given that the Quang Ninh province in the North of Vietnam is a key tourism area and experiences heavy anthropogenic impacts within the surrounding regions reporting increased eutrophication [94] and pollution. Finfish in this region likely experience significant environmental stress facilitating infection and disease progression in the field. Interestingly, fish challenged experimentally with Vh-14 and Vh-15 did experience 100% mortality at a high dose (ca. 3.5 × 108 CFU mL-1) without any additional stressors. Fish challenged with Vh-14 also exhibited dose-dependent mortality kinetics, not seen in our previous work with TCFB-0558 [21], suggesting a potentially highly virulent isolate capable of causing disease without the need for co-participating factors when injected in a sufficient dose.
Isolates Vh-14 and Vh-15 were also distinctive in vitro and in silico, in that they were urease negative and lacked the genes for ureB and ureG1/2. Urease activity is typical of the V. harveyi clade with reports upwards of 50% of isolates are urease positive [95–97]. In the highly specialised fish pathogen Photobacterium damselae subsp. piscicida, the loss of urease catalysation resulted in an increase in host adaption and virulence from the generalist P. damselae subsp. damselae [98]. This shedding of non-essential genes including urease has also been observed in the evolution of Y. pestis from Y. pseudotuberculosis, facilitating greater transmission via the flea vector [99].
Across all the genomes analysed in this study, there was conservation of three antimicrobial resistance genes (ARGs) conferring resistance to beta-lactams and tetracycline (S6 Table) regardless of where the genomes originated. There were twelve additional ARGs detected in seven (18%) of the analysed genomes which all originated from East Asia. The region has recorded relatively high levels of multidrug resistance within the high production regions of China and parts of Southeast Asia, with 64% of the strains analysed by Deng and colleagues [13] recording resistance to three or more antibiotics. Whilst this study provided in silico identification of ARGs across the genomes and identified three highly conserved resistance genes, utilising functional antimicrobial resistance susceptibility testing and minimum inhibition concentrations (MICs) would be beneficial to expand these findings. Although this was an unavoidable constraint when performing in silico investigation of publicly available genomes. Mirroring this conservation of ARGs was the high maintenance of virulence genes detected across the genomes with 38 of the investigated genes (77.5%) present in all genomes. The high level of conservation suggests that these 38 genes may be core to the virulome of V. harveyi and have practical applications in the development of new typing schemes for species identification due to the high level of conservation [100]. Whilst all of the analysed V. harveyi genomes carried a type III secretion system 1 (T3SS1: vopD, vscB/F/N/P, vcrD/H) on chromosome I [101], hypervirulent isolates Vh-14 and Vh-15 maintained a secondary plasmid-associated type III secretion system. Highly similar to the type III secretion system first identified in Yersinia spp. [102], this type III secretion system is also the major virulence mechanism of the economically important aquaculture pathogen Aeromonas salmonicida [103,104]. A homologue of this additional type III secretion system has been identified in V. cholerae and V. parahaemolyticus [105] as type III secretion system 2 (T3SS2), although a universal nomenclature for the genes involved in the T3SS2 in Vibrio spp. is yet to be developed [106]. Upon first contact with host cells [107], the six effector proteins of the T3SS2 are responsible for the prevention of phagocytosis by the host [108]. Each of the effector proteins performs this is in a distinct manner with YpkA responsible for actin-filament disruption, impairing the cytoskeletal function [109] whilst YopH is antiphagocytic via inactivating neutrophils and by dephosphorylating focal adhesion complexes [110]. The other two protein-encoding genes lcrQ and lcrG identified in Vh-14 and Vh-15 are negative regulators of the effector proteins and needle apparatus contributing to cytoplasm stability in the bacteria [111].
Whilst plasmids are associated with some V. harveyi isolates [48,49,112,113], there is currently no confirmed virulence plasmid of V. harveyi. Our study used long-read sequencing, to identify a conjugative plasmid in two isolates carrying components of a secondary type III secretion system suggesting the emergence of a hypervirulent plasmid-mediated patho-variant of V. harveyi with implications for marine aquaculture. The assembly of these large plasmids carrying multiple repeated elements (e.g., transposons, insertion sequences) is generally not possible with short-read sequencing alone [114], highlighting the importance of long-read sequencing. In public health investigations, long-read sequencing is increasingly used for the monitoring of hypervirulent plasmids in multi-host pathogens including Klebsiella pneumoniae [115,116] and group B Streptococcus (Streptococcus agalactiae), which affects aquatic and terrestrial species [117]. In aquatic systems, a key example of plasmid-mediated virulence has been documented for acute hepatopancreatic necrosis disease (APHND) in whiteleg shrimp (Penaeus vannamei) caused by the hypervirulent pVA1-type V. parahaemolyticus plasmid, which subsequently spread to other Vibrio spp. including V. campbellii [118]. Marine environments are a hotspot for horizontal gene transfer, with higher gene transfer rates observed for bacteria co-existing with archaea in high temperature environments [119] and within ocean environments [120]. As such, the increased use of long-read sequencing for V. harveyi and other aquatic bacteria of interest in aquaculture will aid the characterisation of previously unknown virulence components, improving our understanding of virulence evolution in key marine pathogens [121].
Although this study identified two new hypervirulent patho-variants of V. harveyi and the subsequent first detection of a T3SS2 in V. harveyi, our methodology was primarily in silico based and opens an avenue for further functional plasmid analysis and antimicrobial resistance study. Functional characterisation of the plasmids in Vh-14 and Vh-15 could provide insights into the potential for horizontal gene transfer, confirm the role of the plasmid-mediated T3SS2 in hypervirulence and the risk for spread and transmission to other V. harveyi strains or Vibrio spp. in production settings. Whilst this study provided the identification of this T3SS2 akin to those found in Yersinia spp. and Aeromonas salmonicida, it also highlights the power of long-read sequencing for plasmid identification and provides support that utilising an integrated approach combining next-generation sequencing, phenotypic investigation and experimental trials is pivotal for expanding our understanding of virulence in opportunistic bacteria.
In conclusion, this study identified two potentially hypervirulent plasmid-mediated patho-variants of V. harveyi (Vh-14 and Vh-15) encoding components of an additional type III secretion system on a conjugative plasmid. Further investigation into the role of this plasmid, its prevalence in the global aquaculture industry, and the likelihood of conjugation and gene transfer to other V. harveyi strains under a range of environmental conditions is crucial given that such patho-variants could significantly affect fish mortality in farmed settings if they were to become widespread.
Supporting information
S1 Appendix. Supplemental methods for Section 2.3.
https://doi.org/10.1371/journal.pone.0319450.s001
(DOCX)
S2 Appendix. Questionnaire-Inclusivity in global research.
https://doi.org/10.1371/journal.pone.0319450.s002
(DOCX)
S1 Fig. Phylogeny including mis-identified Vibrio campbellii.
SNP-based phylogeny for the 12 isolates sequenced with short-read sequencing in this study (labelled Vh- or -W), the twenty-eight sequences downloaded from the SRA labelled with their unique SRR number and the Vibrio harveyi reference genome (SB1, NCBI: PRJNA972608). A distinct outgroup of four isolates (shaded in blue) formed from the other isolates. Vh-24 had an 97.3% ANI to the Vibrio campbellii reference genome (BoB-53, NCBI: PRJNA429202).
https://doi.org/10.1371/journal.pone.0319450.s003
(TIF)
S1 Table. 49 protein-encoding virulence genes identified in Vibrio harveyi or other species of relevance.
https://doi.org/10.1371/journal.pone.0319450.s004
(XLSX)
S2 Table. Geographical and host origin of publicly available Vibrio harveyi genomes included for phylogenetic analyses.
https://doi.org/10.1371/journal.pone.0319450.s005
(XLSX)
S3 Table. Doses of Vibrio harveyi determined using colony forming unit (CFU) counts performed retrospectively.
https://doi.org/10.1371/journal.pone.0319450.s006
(XLSX)
S4 Table. Summary of short and long-read sequencing for the Vibrio isolates sequenced in the study.
Additional identification to a species level using Matrix-assisted laser-desorption ionisation time-of-flight mass spectrometry (MALDI-ToF MS) and average nucleotide identity (ANI, as determined with FastANI) is included.
https://doi.org/10.1371/journal.pone.0319450.s007
(XLSX)
S5 Table. Plassembler results for Vibrio harveyi isolates containing plasmids detected during long-read sequencing with coding sequences and non-coding RNA annotated using Bakta.
https://doi.org/10.1371/journal.pone.0319450.s008
(XLSX)
S6 Table. Antimicrobial resistance (AMR) genes detected using AMRFinder in short-read assemblies of V. harveyi.‘+’ indicates gene found in nucleotide sequence of strain whilst ‘-’ is when the gene was not identified.
https://doi.org/10.1371/journal.pone.0319450.s009
(XLSX)
S7 Table. Prophage regions predicted using PHASTEST in Vibrio harveyi isolates sequenced in this study.
https://doi.org/10.1371/journal.pone.0319450.s010
(XLSX)
S8 Table. Phenotypic characteristics of Vibrio harveyi at three different temperatures: 22°C, 28°C (optimal growth temperature) and 34°C.
For biofilm, mean OD + SE at 600nm is reported. For urease, a positive or negative result is reported and for haemolysis, a positive or negative result is reported and the mean + SE in mm is reported for those exhibiting haemolysis. All other tests are reported as the mean + SE in mm for all replicates for each isolate under each condition.
https://doi.org/10.1371/journal.pone.0319450.s011
(XLSX)
S9 Table. Distance matrix generated from IR Biotyper software for Fourier Transform Infrared Spectroscopy for the eight Vibrio harveyi isolates phenotypically investigated in this study.
https://doi.org/10.1371/journal.pone.0319450.s012
(XLSX)
Acknowledgments
The authors acknowledge the technical assistance provided by Stuart Glover, Jade Sutton, Cara Jeffrey, Anna Waldron and Slavicka Patten. We acknowledge the efforts of Aakriti Regmi for assistance with sample preparation and experimental trial data entry. We thank Andrew Eliot at the Elizabeth Macarthur Agricultural Institute for providing access to incubators and Elise Beggs for her assistance in the lab. We acknowledge the Sydney Informatics Hub at the University of Sydney for providing access to high-performance computing and technical assistance, with special thanks to Georgie Samaha, Fred Jaya and Nandan Deshpande for their assistance and pipeline development.
References
- 1. Baker-Austin C, Trinanes J, Gonzalez-Escalona N, Martinez-Urtaza J. Non-Cholera Vibrios: The Microbial Barometer of Climate Change. Trends Microbiol. 2017;25(1):76–84. pmid:27843109
- 2. Baker-Austin C, Oliver JD, Alam M, Ali A, Waldor MK, Qadri F, et al. Vibrio spp. infections. Nat Rev Dis Primers. 2018;4(1):8. pmid:30002421
- 3. Vezzulli L, Baker-Austin C, Kirschner A, Pruzzo C, Martinez-Urtaza J. Global emergence of environmental non-O1/O139 Vibrio cholerae infections linked with climate change: a neglected research field?. Environ Microbiol. 2020;22(10):4342–55. pmid:32337781
- 4. Roberts SD, Van Ruth PD, Wilkinson C, Bastianello SS, Bansemer MS. Marine Heatwave, Harmful Algae Blooms and an Extensive Fish Kill Event During 2013 in South Australia. Front Mar Sci. 2019;6.
- 5. Baker-Austin C, Trinanes JA, Salmenlinna S, Löfdahl M, Siitonen A, Taylor NGH, et al. Heat Wave-Associated Vibriosis, Sweden and Finland, 2014. Emerg Infect Dis. 2016;22(7):1216–20. pmid:27314874
- 6.
Sodders N, Stockdale K, Baker K, Ghanem A, Vieth B, Harder T. Notes from the field: vibriosis cases associated with flood waters during and after Hurricane Ian — Florida, September–October 2022. In: Services USDoHaH, editor. 2022.
- 7. Manchanayake T, Salleh A, Amal MNA, Yasin ISM, Zamri-Saad M. Pathology and pathogenesis of Vibrio infection in fish: A review. Aquaculture Reports. 2023;28:101459.
- 8. Harrison J, Nelson K, Morcrette H, Morcrette C, Preston J, Helmer L, et al. The increased prevalence of Vibrio species and the first reporting of Vibrio jasicida and Vibrio rotiferianus at UK shellfish sites. Water Res. 2022;211:117942. pmid:35042073
- 9. Triga A, Smyrli M, Katharios P. Pathogenic and Opportunistic Vibrio spp. Associated with Vibriosis Incidences in the Greek Aquaculture: The Role of Vibrio harveyi as the Principal Cause of Vibriosis. Microorganisms. 2023;11(5):1197. pmid:37317171
- 10. Zupičić IG, Oraić D, Križanović K, Zrnčić S. Whole genome sequencing of Vibrio harveyi from different sites in the Mediterranean Sea providing data on virulence and antimicrobial resistance genes. Aquaculture. 2024;581:740439.
- 11. Mohd Yazid SH, Mohd Daud H, Azmai MNA, Mohamad N, Mohd Nor N. Estimating the Economic Loss Due to Vibriosis in Net-Cage Cultured Asian Seabass (Lates calcarifer): Evidence From the East Coast of Peninsular Malaysia. Front Vet Sci. 2021;8:644009. pmid:34692800
- 12.
FAO. The State of World Fisheries and Aquaculture 2024 - Blue Transformation in action. Rome, Italy: FAO; 2024. p. 264.
- 13. Deng Y, Xu L, Chen H, Liu S, Guo Z, Cheng C, et al. Prevalence, virulence genes, and antimicrobial resistance of Vibrio species isolated from diseased marine fish in South China. Sci Rep. 2020;10(1):14329. pmid:32868874
- 14. Zhu Z, Duan C, Dong C, Weng S, He J. Epidemiological situation and phylogenetic relationship of Vibrio harveyi in marine-cultured fishes in China and Southeast Asia. Aquaculture. 2020;529:735652.
- 15. Xie J, Bu L, Jin S, Wang X, Zhao Q, Zhou S, et al. Outbreak of vibriosis caused by Vibrio harveyi and Vibrio alginolyticus in farmed seahorse Hippocampus kuda in China. Aquaculture. 2020;523:735168.
- 16. Cardinaud M, Barbou A, Capitaine C, Bidault A, Dujon AM, Moraga D, et al. Vibrio harveyi adheres to and penetrates tissues of the European abalone Haliotis tuberculata within the first hours of contact. Appl Environ Microbiol. 2014;80(20):6328–33. pmid:25107972
- 17.
Chong RS-M. Chapter 20 - Crab larval luminescent vibriosis. In: Kibenge FSB, Baldisserotto B, Chong RS-M, editors. Aquaculture Pathophysiology: Academic Press; 2022. p. 163–5.
- 18. de Souza Valente C, Wan AHL. Vibrio and major commercially important vibriosis diseases in decapod crustaceans. J Invertebr Pathol. 2021;181:107527. pmid:33406397
- 19. Mohamad N, Amal MNA, Yasin ISM, Zamri Saad M, Nasruddin NS, Al-saari N, et al. Vibriosis in cultured marine fishes: a review. Aquaculture. 2019;512:734289.
- 20. Zhang X-H, He X, Austin B. Vibrio harveyi: a serious pathogen of fish and invertebrates in mariculture. Mar Life Sci Technol. 2020;2(3):231–45. pmid:32419972
- 21. Samsing F, Zhang W, Zadoks RN, Whittington R, Venturini C, Giles C, et al. Cold temperature stress and damaged skin induced high mortality in barramundi (Lates calcarifer) challenged with Vibrio harveyi. J Fish Dis. 2023.
- 22. Ruwandeepika HAD, Defoirdt T, Bhowmick PP, Shekar M, Bossier P, Karunasagar I. Presence of typical and atypical virulence genes in vibrio isolates belonging to the Harveyi clade. J Appl Microbiol. 2010;109(3):888–99. pmid:20345385
- 23. Tu Z, Li H, Zhang X, Sun Y, Zhou Y. Complete genome sequence and comparative genomics of the golden pompano (Trachinotus ovatus) pathogen, Vibrio harveyi strain QT520. PeerJ. 2017;5:e4127. pmid:29230364
- 24. Triga A, Issa ZA, Smyrli M, Fenske L, Katharios P. Virulence and pangenome analysis of Vibrio harveyi strains from Greek and Red Sea marine aquaculture. Aquaculture. 2024;587:740839.
- 25. Herzog R, Peschek N, Fröhlich KS, Schumacher K, Papenfort K. Three autoinducer molecules act in concert to control virulence gene expression in Vibrio cholerae. Nucleic Acids Res. 2019;47(6):3171–83. pmid:30649554
- 26. Gallego-Hernandez AL, DePas WH, Park JH, Teschler JK, Hartmann R, Jeckel H, et al. Upregulation of virulence genes promotes Vibrio cholerae biofilm hyperinfectivity. Proc Natl Acad Sci U S A. 2020;117(20):11010–7. pmid:32355001
- 27. Li L, Meng H, Gu D, Li Y, Jia M. Molecular mechanisms of Vibrio parahaemolyticus pathogenesis. Microbiol Res. 2019;222:43–51. pmid:30928029
- 28. Singhal N, Kumar M, Kanaujia PK, Virdi JS. MALDI-TOF mass spectrometry: an emerging technology for microbial identification and diagnosis. Front Microbiol. 2015;6:791. pmid:26300860
- 29. Prjibelski A, Antipov D, Meleshko D, Lapidus A, Korobeynikov A. Using SPAdes De Novo Assembler. Curr Protoc Bioinformatics. 2020;70(1):e102. pmid:32559359
- 30. Seemann T. Prokka: rapid prokaryotic genome annotation. Bioinformatics. 2014;30(14):2068–9. pmid:24642063
- 31. Jain C, Rodriguez-R LM, Phillippy AM, Konstantinidis KT, Aluru S. High throughput ANI analysis of 90K prokaryotic genomes reveals clear species boundaries. Nat Commun. 2018;9(1):5114. pmid:30504855
- 32. Ke H-M, Prachumwat A, Yu C-P, Yang Y-T, Promsri S, Liu K-F, et al. Comparative genomics of Vibrio campbellii strains and core species of the Vibrio Harveyi clade. Sci Rep. 2017;7:41394. pmid:28145490
- 33. Gomez-Gil B, Soto-Rodríguez S, García-Gasca A, Roque A, Vazquez-Juarez R, Thompson FL, et al. Molecular identification of Vibrio harveyi-related isolates associated with diseased aquatic organisms. Microbiology (Reading). 2004;150(Pt 6):1769–77. pmid:15184563
- 34. Feldgarden M, Brover V, Gonzalez-Escalona N, Frye JG, Haendiges J, Haft DH, et al. AMRFinderPlus and the Reference Gene Catalog facilitate examination of the genomic links among antimicrobial resistance, stress response, and virulence. Sci Rep. 2021;11(1):12728. pmid:34135355
- 35.
Samaha G, Samsing F, O’Brien M, Jaya F. ONT-bacpac-nf. 1.0.0 ed. 2025.
- 36. De Coster W, Rademakers R. NanoPack2: population-scale evaluation of long-read sequencing data. Bioinformatics. 2023;39(5):btad311. pmid:37171891
- 37. Leger A, Leonardi T. pycoQC, interactive quality control for Oxford Nanopore Sequencing. JOSS. 2019;4(34):1236.
- 38. Wood DE, Lu J, Langmead B. Improved metagenomic analysis with Kraken 2. Genome Biol. 2019;20(1):257. pmid:31779668
- 39. Kolmogorov M, Yuan J, Lin Y, Pevzner PA. Assembly of long, error-prone reads using repeat graphs. Nat Biotechnol. 2019;37(5):540–6. pmid:30936562
- 40. Lin Y, Yuan J, Kolmogorov M, Shen MW, Chaisson M, Pevzner PA. Assembly of long error-prone reads using de Bruijn graphs. Proc Natl Acad Sci U S A. 2016;113(52):E8396–405. pmid:27956617
- 41. Wick RR, Judd LM, Gorrie CL, Holt KE. Unicycler: Resolving bacterial genome assemblies from short and long sequencing reads. PLoS Comput Biol. 2017;13(6):e1005595. pmid:28594827
- 42. Bouras G, Sheppard AE, Mallawaarachchi V, Vreugde S. Plassembler: an automated bacterial plasmid assembly tool. Bioinformatics. 2023;39(7):btad409. pmid:37369026
- 43. Schwengers O, Jelonek L, Dieckmann MA, Beyvers S, Blom J, Goesmann A. Bakta: rapid and standardized annotation of bacterial genomes via alignment-free sequence identification. Microb Genom. 2021;7(11):000685. pmid:34739369
- 44. Wishart DS, Han S, Saha S, Oler E, Peters H, Grant JR, et al. PHASTEST: faster than PHASTER, better than PHAST. Nucleic Acids Res. 2023;51(W1):W443–50. pmid:37194694
- 45. Ruwandeepika HAD, Karunasagar I, Bossier P, Defoirdt T. Expression and Quorum Sensing Regulation of Type III Secretion System Genes of Vibrio harveyi during Infection of Gnotobiotic Brine Shrimp. PLoS One. 2015;10(12):e0143935. pmid:26636765
- 46. Huang Z, Anokyewaa MA, Wang J, Jian J, Lu Y. Pathogenicity and antibiotic resistance analysis of Vibrio species found in coastal water at mainly beach of Shenzhen, China. Front Mar Sci. 2022;9.
- 47. Fu S, Ni P, Yang Q, Hu H, Wang Q, Ye S, et al. Delineating the key virulence factors and intraspecies divergence of Vibrio harveyi via whole-genome sequencing. Can J Microbiol. 2021;67(3):231–48. pmid:32941745
- 48. Zhang Y, Deng Y, Feng J, Guo Z, Chen H, Wang B, et al. Functional characterization of VscCD, an important component of the type Ⅲ secretion system of Vibrio harveyi. Microb Pathog. 2021;157:104965. pmid:34015493
- 49. Worden PJ, Bogema DR, Micallef ML, Go J, Deutscher AT, Labbate M, et al. Phylogenomic diversity of Vibrio species and other Gammaproteobacteria isolated from Pacific oysters (Crassostrea gigas) during a summer mortality outbreak. Microb Genom. 2022;8(12):mgen000883. pmid:36748707
- 50. Grant JR, Enns E, Marinier E, Mandal A, Herman EK, Chen C-Y, et al. Proksee: in-depth characterization and visualization of bacterial genomes. Nucleic Acids Res. 2023;51(W1):W484–92. pmid:37140037
- 51.
Wilkins D. gggenes: Draw Gene Arrow Maps in ‘ggplot2’. R package version 0.5.1. 2023.
- 52.
Seemann T. Snippy: fast bacterial variant calling from NGS reads. 2015.
- 53. Croucher NJ, Page AJ, Connor TR, Delaney AJ, Keane JA, Bentley SD, et al. Rapid phylogenetic analysis of large samples of recombinant bacterial whole genome sequences using Gubbins. Nucleic Acids Res. 2015;43(3):e15. pmid:25414349
- 54. Nguyen L-T, Schmidt HA, von Haeseler A, Minh BQ. IQ-TREE: a fast and effective stochastic algorithm for estimating maximum-likelihood phylogenies. Mol Biol Evol. 2015;32(1):268–74. pmid:25371430
- 55. Hoang DT, Chernomor O, von Haeseler A, Minh BQ, Vinh LS. UFBoot2: Improving the Ultrafast Bootstrap Approximation. Mol Biol Evol. 2018;35(2):518–22. pmid:29077904
- 56.
Rambaut A. FigTree v 1.4.4. Edinburgh: Institute of Evolutionary Biology, University of Edinburgh; 2018.
- 57. Argimón S, Abudahab K, Goater RJE, Fedosejev A, Bhai J, Glasner C, et al. Microreact: visualizing and sharing data for genomic epidemiology and phylogeography. Microb Genom. 2016;2(11):e000093. pmid:28348833
- 58. Katersky RS, Carter CG. Growth efficiency of juvenile barramundi, Lates calcarifer, at high temperatures. Aquaculture. 2005;250(3–4):775–80.
- 59. Bermudes M, Glencross B, Austen K, Hawkins W. The effects of temperature and size on the growth, energy budget and waste outputs of barramundi (Lates calcarifer). Aquaculture. 2010;306(1–4):160–6.
- 60. Montánchez I, Kaberdin VR. Vibrio harveyi: A brief survey of general characteristics and recent epidemiological traits associated with climate change. Mar Environ Res. 2020;154:104850. pmid:32056705
- 61. Shen GM, Shi CY, Fan C, Jia D, Wang SQ, Xie GS, et al. Isolation, identification and pathogenicity of Vibrio harveyi, the causal agent of skin ulcer disease in juvenile hybrid groupers Epinephelus fuscoguttatus × Epinephelus lanceolatus. J Fish Dis. 2017;40(10):1351–62.
- 62. Miles AA, Misra SS, Irwin JO. The estimation of the bactericidal power of the blood. J Hyg (Lond). 1938;38(6):732–49. pmid:20475467
- 63. Pang L, Zhang X-H, Zhong Y, Chen J, Li Y, Austin B. Identification of Vibrio harveyi using PCR amplification of the toxR gene. Lett Appl Microbiol. 2006;43(3):249–55. pmid:16910927
- 64.
R Core Team. R: A language and environment for statistical computing. 4.2.1 ed. Vienna, Austria: R Foundation for Statistical Computing; 2022.
- 65.
Team RC. Stats: R stats package v 3.6.2. 2018.
- 66.
Wickham H. ggplot2: Elegant Graphics for Data Analysis. New York: Springer-Verlag; 2016.
- 67.
Kassambara A, Kosinski M, Biecek P. survminer: Drawing Survival Curves using ‘ggplot2’. R package version 0.5.0.999. 2024.
- 68.
Therneau TM. A package for survival analysis in R. R package version 3.6.4. 2024.
- 69.
Therneau TM. coxme: Mixed Effects Cox Models. R package version 2.2.20. 2024.
- 70. Molano L-AG, Hirsch P, Hannig M, Müller R, Keller A. The PLSDB 2025 update: enhanced annotations and improved functionality for comprehensive plasmid research. Nucleic Acids Res. 2025;53(D1):D189–96. pmid:39565221
- 71. Galata V, Fehlmann T, Backes C, Keller A. PLSDB: a resource of complete bacterial plasmids. Nucleic Acids Res. 2019;47(D1):D195–202. pmid:30380090
- 72. Faridani OR, Nikravesh A, Pandey DP, Gerdes K, Good L. Competitive inhibition of natural antisense Sok-RNA interactions activates Hok-mediated cell killing in Escherichia coli. Nucleic Acids Res. 2006;34(20):5915–22. pmid:17065468
- 73. Teo JW, Suwanto A, Poh CL. Novel beta-lactamase genes from two environmental isolates of Vibrio harveyi. Antimicrob Agents Chemother. 2000;44(5):1309–14. pmid:10770767
- 74. Nonaka L, Suzuki S. New Mg2+-dependent oxytetracycline resistance determinant tet 34 in Vibrio isolates from marine fish intestinal contents. Antimicrob Agents Chemother. 2002;46(5):1550–2. pmid:11959596
- 75. Teo JWP, Tan TMC, Poh CL. Genetic determinants of tetracycline resistance in Vibrio harveyi. Antimicrob Agents Chemother. 2002;46(4):1038–45. pmid:11897587
- 76. Pha K, Navarro L. Yersinia type III effectors perturb host innate immune responses. World J Biol Chem. 2016;7(1):1–13. pmid:26981193
- 77. Aziz RK, Breitbart M, Edwards RA. Transposases are the most abundant, most ubiquitous genes in nature. Nucleic Acids Res. 2010;38(13):4207–17. pmid:20215432
- 78. Vezzulli L, Pezzati E, Brettar I, Höfle M, Pruzzo C. Effects of Global Warming on Vibrio Ecology. Microbiol Spectr. 2015;3(3):10.1128/microbiolspec.VE-0004–2014. pmid:26185070
- 79. Froelich BA, Daines DA. In hot water: effects of climate change on Vibrio-human interactions. Environ Microbiol. 2020;22(10):4101–11. pmid:32114705
- 80. Archer EJ, Baker-Austin C, Osborn TJ, Jones NR, Martínez-Urtaza J, Trinanes J, et al. Climate warming and increasing Vibrio vulnificus infections in North America. Sci Rep. 2023;13(1):3893. pmid:36959189
- 81. Fleischmann S, Herrig I, Wesp J, Stiedl J, Reifferscheid G, Strauch E, et al. Prevalence and distribution of potentially human pathogenic Vibrio spp. on German north and Baltic sea coasts. Front Cell Infect Microbiol. 2022;12:846819.
- 82. Darshanee Ruwandeepika HA, Sanjeewa Prasad Jayaweera T, Paban Bhowmick P, Karunasagar I, Bossier P, Defoirdt T. Pathogenesis, virulence factors and virulence regulation of vibrios belonging to the Harveyi clade. Reviews in Aquaculture. 2012;4(2):59–74.
- 83. Samsing F, Barnes AC. The rise of the opportunists: What are the drivers of the increase in infectious diseases caused by environmental and commensal bacteria? Reviews in Aquaculture. 2024;16(4):1787–97.
- 84. Gundogdu K, Orus Iturriza A, Orruño M, Montánchez I, Eguiraun H, Martinez I, et al. Addressing the Joint Impact of Temperature and pH on Vibrio harveyi Adaptation in the Time of Climate Change. Microorganisms. 2023;11(4):1075. pmid:37110498
- 85. Montánchez I, Ogayar E, Plágaro AH, Esteve-Codina A, Gómez-Garrido J, Orruño M, et al. Analysis of Vibrio harveyi adaptation in sea water microcosms at elevated temperature provides insights into the putative mechanisms of its persistence and spread in the time of global warming. Sci Rep. 2019;9(1):289. pmid:30670759
- 86. Hernández-Cabanyero C, Sanjuán E, Fouz B, Pajuelo D, Vallejos-Vidal E, Reyes-López FE, et al. The Effect of the Environmental Temperature on the Adaptation to Host in the Zoonotic Pathogen Vibrio vulnificus. Front Microbiol. 2020;11:489. pmid:32296402
- 87. Zarzecka U, Zadernowska A, Chajęcka-Wierzchowska W. Effects of osmotic and high pressure stress on expression of virulence factors among Enterococcus spp. isolated from food of animal origin. Food Microbiol. 2022;102:103900. pmid:34809932
- 88. Nagar V, Bandekar JR, Shashidhar R. Expression of virulence and stress response genes in Aeromonas hydrophila under various stress conditions. J Basic Microbiol. 2016;56(10):1132–7. pmid:27163835
- 89. Phuoc NN, Linh NTH, Crestani C, Zadoks RN. Effect of strain and enviromental conditions on the virulence of Streptococcus agalactiae (Group B Streptococcus; GBS) in red tilapia (Oreochromis sp.). Aquaculture. 2021;534:736256.
- 90. Green TJ, Siboni N, King WL, Labbate M, Seymour JR, Raftos D. Simulated Marine Heat Wave Alters Abundance and Structure of Vibrio Populations Associated with the Pacific Oyster Resulting in a Mass Mortality Event. Microb Ecol. 2019;77(3):736–47. pmid:30097682
- 91. King WL, Jenkins C, Go J, Siboni N, Seymour JR, Labbate M. Characterisation of the Pacific Oyster Microbiome During a Summer Mortality Event. Microb Ecol. 2019;77(2):502–12. pmid:29987529
- 92. Reverter M, Sarter S, Caruso D, Avarre J-C, Combe M, Pepey E, et al. Aquaculture at the crossroads of global warming and antimicrobial resistance. Nat Commun. 2020;11(1):1870. pmid:32312964
- 93. Chin YK, Ina-Salwany MY, Zamri-Saad M, Amal MNA, Mohamad A, Lee JY, et al. Effects of skin abrasion in immersion challenge with Vibrio harveyi in Asian seabass Lates calcarifer fingerlings. Dis Aquat Organ. 2020;137(3):167–73. pmid:31942862
- 94. Le ND, Nguyen TH, Duong TT, Rochelle-Newall E, Hoang TTH, Vu TH, et al. Risk of eutrophication in the seawater of the coastal Red River aquaculture zone (Thai Binh province, Vietnam). Regional Studies in Marine Science. 2022;55:102587.
- 95. Defoirdt T, Vlaeminck SE, Sun X, Boon N, Clauwaert P. Ureolytic Activity and Its Regulation in Vibrio campbellii and Vibrio harveyi in Relation to Nitrogen Recovery from Human Urine. Environ Sci Technol. 2017;51(22):13335–43. pmid:29083891
- 96. Pavlinec Ž, Zupičić IG, Oraić D, Lojkić I, Fouz B, Zrnčić S. Biochemical and molecular characterization of three serologically different Vibrio harveyi strains isolated from farmed Dicentrarchus labrax from the Adriatic Sea. Sci Rep. 2022;12(1):7309. pmid:35508554
- 97.
Carson J, Douglas M, Wilson T, Gudkovs N. Identification of Vibrionaceae from Australian and New Zealand aquatic animals using phenotypic and molecular procedures. Australian and New Zealand Standard Diagnostic Procedures (ANZSDP) Department of Agriculture, Fisheries and Forestry; 2020.
- 98. Baseggio L, Rudenko O, Engelstädter J, Barnes AC. The Evolution of a Specialized, Highly Virulent Fish Pathogen through Gene Loss and Acquisition of Host-Specific Survival Mechanisms. Appl Environ Microbiol. 2022;88(14):e0022222. pmid:35862683
- 99. Chouikha I, Hinnebusch BJ. Silencing urease: a key evolutionary step that facilitated the adaptation of Yersinia pestis to the flea-borne transmission route. Proc Natl Acad Sci U S A. 2014;111(52):18709–14. pmid:25453069
- 100. Cano-Gomez A, Høj L, Owens L, Andreakis N. Multilocus sequence analysis provides basis for fast and reliable identification of Vibrio harveyi-related species and reveals previous misidentification of important marine pathogens. Syst Appl Microbiol. 2011;34(8):561–5. pmid:22055753
- 101. Wu C, Zhao Z, Liu Y, Zhu X, Liu M, Luo P, et al. Type III Secretion 1 Effector Gene Diversity Among Vibrio Isolates From Coastal Areas in China. Front Cell Infect Microbiol. 2020;10:301. pmid:32637366
- 102. Cornelis GR, Boland A, Boyd AP, Geuijen C, Iriarte M, Neyt C, et al. The virulence plasmid of Yersinia, an antihost genome. Microbiol Mol Biol Rev. 1998;62(4):1315–52. pmid:9841674
- 103. Frey J, Origgi FC. Type III Secretion System of Aeromonas salmonicida Undermining the Host’s Immune Response. Front Mar Sci. 2016;3.
- 104. Burr SE, Wahli T, Segner H, Pugovkin D, Frey J. Association of Type III secretion genes with virulence of Aeromonas salmonicida subsp. salmonicida. Dis Aquat Organ. 2003;57(1–2):167–71. pmid:14735937
- 105. Park K-S, Ono T, Rokuda M, Jang M-H, Okada K, Iida T, et al. Functional characterization of two type III secretion systems of Vibrio parahaemolyticus. Infect Immun. 2004;72(11):6659–65. pmid:15501799
- 106. Miller KA, Tomberlin KF, Dziejman M. Vibrio variations on a type three theme. Curr Opin Microbiol. 2019;47:66–73. pmid:30711745
- 107. Lee WL, Singaravelu P, Wee S, Xue B, Ang KC, Gunaratne J, et al. Mechanisms of Yersinia YopO kinase substrate specificity. Sci Rep. 2017;7:39998. pmid:28051168
- 108. Plano GV, Schesser K. The Yersinia pestis type III secretion system: expression, assembly and role in the evasion of host defenses. Immunol Res. 2013;57(1–3):237–45. pmid:24198067
- 109. Peter MF, Tuukkanen AT, Heubach CA, Selsam A, Duthie FG, Svergun DI, et al. Studying Conformational Changes of the Yersinia Type-III-Secretion Effector YopO in Solution by Integrative Structural Biology. Structure. 2019;27(9):1416-1426.e3. pmid:31303480
- 110. Rolán HG, Durand EA, Mecsas J. Identifying Yersinia YopH-targeted signal transduction pathways that impair neutrophil responses during in vivo murine infection. Cell Host Microbe. 2013;14(3):306–17. pmid:24034616
- 111. Fei K, Yan H, Zeng X, Huang S, Tang W, Francis MS, et al. LcrQ Coordinates with the YopD-LcrH Complex To Repress lcrF Expression and Control Type III Secretion by Yersinia pseudotuberculosis. mBio. 2021;12(3):e0145721. pmid:34154409
- 112. Schikorski D, Renault T, Paillard C, Bidault-Toffin A, Tourbiez D, Saulnier D. Development of TaqMan real-time PCR assays for monitoring Vibrio harveyi infection and a plasmid harbored by virulent strains in European abalone Haliotis tuberculata aquaculture. Aquaculture. 2013;392–395:106–12.
- 113. Deng Y, Xu H, Su Y, Liu S, Xu L, Guo Z, et al. Horizontal gene transfer contributes to virulence and antibiotic resistance of Vibrio harveyi 345 based on complete genome sequence analysis. BMC Genomics. 2019;20(1):761. pmid:31640552
- 114. Wick RR, Judd LM, Gorrie CL, Holt KE. Completing bacterial genome assemblies with multiplex MinION sequencing. Microb Genom. 2017;3(10):e000132. pmid:29177090
- 115. Russo TA, Marr CM. Hypervirulent Klebsiella pneumoniae. Clin Microbiol Rev. 2019;32(3):e00001-19. pmid:31092506
- 116. Lam MMC, Wyres KL, Duchêne S, Wick RR, Judd LM, Gan Y-H, et al. Population genomics of hypervirulent Klebsiella pneumoniae clonal-group 23 reveals early emergence and rapid global dissemination. Nat Commun. 2018;9(1):2703. pmid:30006589
- 117. Crestani C, Forde TL, Lycett SJ, Holmes MA, Fasth C, Persson-Waller K, et al. The fall and rise of group B Streptococcus in dairy cattle: reintroduction due to human-to-cattle host jumps?. Microb Genom. 2021;7(9):000648. pmid:34486971
- 118. Dong X, Chen J, Song J, Wang H, Wang W, Ren Y, et al. Evidence of the horizontal transfer of pVA1-type plasmid from AHPND-causing V. campbellii to non-AHPND V. owensii. Aquaculture. 2019;503:396–402.
- 119. Fuchsman CA, Collins RE, Rocap G, Brazelton WJ. Effect of the environment on horizontal gene transfer between bacteria and archaea. PeerJ. 2017;5:e3865. pmid:28975058
- 120. McDaniel LD, Young E, Delaney J, Ruhnau F, Ritchie KB, Paul JH. High frequency of horizontal gene transfer in the oceans. Science. 2010;330(6000):50. pmid:20929803
- 121. Baseggio L, Rudenko O, Buller N, Landos M, Englestädter J, Barnes AC. Complete, closed and curated genome sequences of Photobacterium damselae subsp. piscicida isolates from Australia indicate mobilome-driven localized evolution and novel pathogenicity determinants. Microb Genom. 2021;7(4):000562. pmid:33885359