Reproducible challenge model to investigate the virulence of Flavobacterium columnare genomovars in rainbow trout Oncorhynchus mykiss

Flavobacterium covae is the predominant species of columnaris-causing bacteria impacting the Channel Catfish industry in the southeastern United States

OBJECTIVE

Columnaris disease is a leading cause of disease-related losses in the catfish industry of the southeastern United States. The term "columnaris-causing bacteria" (CCB) has been coined in reference to the four described species that cause columnaris disease: Flavobacterium columnare, F. covae, F. davisii, and F. oreochromis. Historically, F. columnare, F. covae, and F. davisii have been isolated from columnaris disease cases in the catfish industry; however, there is a lack of knowledge of which CCB species are most prevalent in farm-raised catfish. The current research objectives were to (1) sample columnaris disease cases from the U.S. catfish industry and identify the species of CCB involved and (2) determine the virulence of the four CCB species in Channel Catfish Ictalurus punctatus in controlled laboratory challenges.

METHODS

Bacterial isolates or swabs of external lesions from catfish were collected from 259 columnaris disease cases in Mississippi and Alabama during 2015-2019. The DNA extracted from the samples was analyzed using a CCB-specific multiplex polymerase chain reaction to identify the CCB present in each diagnostic case. Channel Catfish were challenged by immersion with isolates belonging to each CCB species to determine virulence at ~28°C and 20°C.

RESULT

Flavobacterium covae was identified as the predominant CCB species impacting the U.S. catfish industry, as it was present in 94.2% (n = 244) of diagnostic case submissions. Challenge experiments demonstrated that F. covae and F. oreochromis were highly virulent to Channel Catfish, with most isolates resulting in near 100% mortality. In contrast, F. columnare and F. davisii were less virulent, with most isolates resulting in less than 40% mortality.

CONCLUSION

Collectively, these results demonstrate that F. covae is the predominant CCB in the U.S. catfish industry, and research aimed at developing new control and prevention strategies should target this bacterial species. The methods described herein can be used to continue monitoring the prevalence of CCB in the catfish industry and can be easily applied to other industries to identify which Flavobacterium species have the greatest impact.

Transcriptome analysis of immune-related gene expression in Yellow River carp (Cyprinus carpio var.) after challenge with Flavobacterium columnare

Yellow River carp (Cyprinus carpio) is an economically-important freshwater fish. It is the common host of the epizootic bacterium Flavobacterium columnare, a common fish pathogen that causes columnaris disease resulting in aquacultural losses. However, information on the functions and mechanisms of the immune system of Yellow River carp infected with F. columnare is limited. Therefore, the aim of this study is to evaluate the genetic and histopathological effects of an experimentally-induced F. columnare infection in Yellow River carp. Sixty fish were divided into control (CT group) and challenged groups. The gills were collected for histological and transcriptome analysis to understand the host immune response after challenge with F. columnare. The infected fish of the IF group presented typical columnaris disease symptoms and higher mortality, as well as histological changes. However, some challenged fish showed asymptomatic infection (IC group). Additionally, there were 1776 significant differentially-expressed genes (DEGs) between the IC and CT groups, 1853 DEGs between the IF and CT groups, and 1836 DEGs between the IF and IC groups, All the DEGs were classified into three gene ontology categories, which were allocated to 158 KEGG pathways. Moreover, immune-related genes were confirmed by qRT-PCR. we quantified the level of IL-1, IL-6, TNF-α and IL-8 by ELISA. The results showed the highest expression levels of inflammatory cytokines as well as stress proteins and the adhesion molecules in the lF group, which may contribute to severe infection, and a higher case fatality rate, while the high expression of chemokines, costimulatory molecules and the up regulation of antigen presentation function could help the carp resist F. columnare infection.

Growth of Flavobacterium columnare Genomovars in the Presence or Lack of Supplemental Cations

Flavobacterium columnare is a problematic pathogen for the aquaculture industry where isolates are classified by genomovars. Suspended growth in a low nutrient media, like tryptone yeast extract salts, is a common method used for laboratory study. The presence of calcium and magnesium is the factor contributing to growth, virulence, and biofilm formation for F. columnare. Exponential growth occurs within 24 h for F. columnare when grown in complete tryptone yeast extract salts medium at 30 ºC. Withholding CaCl₂ and MgSO₄ components from a complete TYES formulation reduced or completely inhibited growth of genomovar I isolates but not the growth of genomovar II, IIB, or III isolates. Only 3 of 20 genomovar I isolates, MS-FC-4, FC-CSF-53, and 023-08-3, could achieve O.D. 540 readings ≥ 0.3 but only after 48-h incubation in cation-restricted TYES. Independently adding CaCl₂ or MgSO₄ to tryptone and yeast extracts did not result in a genomovar-specific growth phenotype, but generally demonstrated increased clumping with individual isolates presenting abnormal growth. Clumping formed filamentous strings that migrated to the top of the culture tube when isolates were grown in TYE+CaCl₂. Several of the F. columnare isolates from all the genomovars exhibited delayed growth when a single cation source was provided. This study demonstrates phenotypic differences between and within genomovars of a single bacterial species when grown under different TYES media conditions.

Genetic characterization of Flavobacterium columnare isolates from the Pacific Northwest, USA

Flavobacterium columnare is the causative agent of columnaris disease. Previous work has demonstrated a high degree of genetic variability among F. columnare isolates, identifying 4 genetic groups (GGs) with some host associations. Herein, a total of 49 F. columnare isolates were characterized, the majority of which were collected from 15 different locations throughout the US Pacific Northwest. Most isolates were collected from 2015-2018 and originated from disease outbreaks in salmonid hatcheries and rearing ponds, sturgeon hatcheries and ornamental fish. Other isolates were part of collections recovered from 1980-2018. Initial identification was confirmed by F. columnare species-specific qPCR. Study isolates were further characterized using a multiplex PCR that differentiates between the 4 currently recognized F. columnare GGs. Multiplex PCR results were supported by repetitive sequence-mediated PCR fingerprinting and gyrB sequence analysis. F. columnare GG1 was the most prevalent (83.7%, n = 41/49), represented by isolates from salmonids (n = 32), white sturgeon (n = 2), channel catfish (n = 1), ornamental goldfish (n = 1), koi (n = 3), wild sunfish (n = 1) and 1 unknown host. Six isolates (12.2%, n = 6/49) were identified as GG3, which were cultured from rainbow trout (n = 3) and steelhead trout (n = 3). Two isolates were identified as GG2 (4.1%, n = 2/49) and were from ornamental fish. No GG4 isolates were cultured in this study. The biological significance of this genetic variability remains unclear, but this variation could have significant implications for fish health management. The results from this study provide baseline data for future work developing strategies to ameliorate columnaris-related losses in the US Pacific Northwest.

Virulence variations of Flavobacterium columnare in rainbow trout (Oncorhynchus mykiss) eyed eggs and alevin

Flavobacterium columnare (Fc) is the causative agent for columnaris disease (CD) in several fish species and an emerging problem for rainbow trout aquaculture. We characterize the virulence phenotype of two Fc isolates, CSF-298-10 and MS-FC-4, against trout from two sources, NCCCWA and a production stock (PS), at the eyed egg and alevin life stages. Immersion challenges demonstrated that NCCCWA eyed eggs were susceptible to the Fc isolate MS-FC-4 (>97% mortality) but no mortality was observed against PS eyed eggs. The CSF-298-10 had little effect on any eyed eggs tested and was not highly virulent to any alevin till day six post-hatch, up to 38% for NCCCWA and ~80% PS alevin. The MS-FC-4 strain produced ≥80% mortality any day an immersion challenge occurred post-hatch. Significant difference in CFU counts was recorded between the Fc strains on 2 days post-hatch immersion challenges. Counts for the NCCCWA alevin were 4.4 × 10³  CFU/ml^-1 and 1.8 × 10⁶  CFU/ml^-1 for the CSF-298-10 strain and MS-FC-4 strain, respectively, and for the PS alevin CSF-298-10 measured 9.9 × 10¹  CFU/ml^-1 and 3.8 × 10⁵  CFU/ml^-1 for MS-FC-4. These two Fc isolates present stark differences in virulence phenotypes to both eyed eggs and alevin and present an interesting model system for virulence kinetics and potentially alternative pathogenic pathways.

Isolation and experimental challenge of cultured burbot (Lota lota maculosa) with Flavobacterium columnare and Aeromonas sp. isolates

Burbot (Lota lota maculosa) are a potential new species for commercial aquaculture. As burbot culture expands, there is a need to further define pathogen susceptibility and characterize aspects of the burbot immune response in an effort to assess fish health. A recent clinical diagnostic case from juvenile burbot reared at a commercial production facility resulted in the isolation and identification of Flavobacterium columnare along with several Aeromonas spp. The F. columnare isolate was assigned to genetic group 1 via multiplex PCR, a genetic group commonly associated with columnaris disease cases in rainbow trout (Oncorhynchus mykiss). Virulence of the F. columnare isolate was assessed in vivo in both juvenile burbot and rainbow trout. Additionally, several of the Aeromonas sp. case isolates were identified via sequencing (16S rRNA, gyrB and rpoD) and a putative A. sobria isolate (BI-3) was used to challenge burbot, along with a known virulent Aeromonas sp. (A141), but BI-3 was not found to be virulent. Burbot were refractory to F. columnare when challenged by immersion, and it is likely that this is a secondary pathogen for burbot. Although refractory in burbot, the identified F. columnare isolate (BI-1) was found to be virulent in rainbow trout.

Recirculation versus flow-through rainbow trout laboratory Flavobacterium columnare challenge

Flavobacterium columnare immersion challenges are affected by water-related environmental parameters and thus are difficult to reproduce. Whereas these challenges are typically conducted using flow-through systems, use of a recirculating challenge system to control environmental parameters may improve reproducibility. We compared mortality, bacterial concentration, and environmental parameters between flow-through and recirculating immersion challenge systems under laboratory conditions using 20 rainbow trout families. Despite identical dose concentration (1:75 dilution), duration of challenge, lot of fish, and temperature, average mortality in the recirculating system (42%) was lower (p < 0.01) compared to the flow-through system (77%), and there was low correlation (r = 0.24) of family mortality. Mean days to death (3.25 vs. 2.99 d) and aquaria-to-aquaria variation (9.6 vs. 10.4%) in the recirculating and flow-through systems, respectively, did not differ (p ≥ 0.30). Despite 10-fold lower water replacement rate in the recirculating (0.4 exchanges h-1) compared to flow-through system (4 exchanges h-1), differences in bacterial concentration between the 2 systems were modest (≤0.6 orders of magnitude) and inconsistent throughout the 21 d challenge. Compared to the flow-through system, dissolved oxygen during the 1 h exposure and pH were greater (p ≤ 0.02), and calcium and hardness were lower (p ≤ 0.03), in the recirculating system. Although this study was not designed to test effects of specific environmental parameters on mortality, it demonstrates that the cumulative effects of these parameters result in poor reproducibility. A recirculating immersion challenge model may be warranted to empirically identify and control environmental parameters affecting mortality and thus may serve as a more repeatable laboratory challenge model.

Mechanisms and the role of probiotic Bacillus in mitigating fish pathogens in aquaculture

Diseases are natural components of the environment, and many have economic implications for aquaculture and fisheries. Aquaculture is a fast-growing industry with the aim to meet the high protein demand of the ever-increasing global population; however, the emergence of diseases is a major setback to the industry. Probiotics emerged as a better solution to curb the disease problem in aquaculture among many alternatives. Probiotic Bacillus has been proven to better combat a wide range of fish pathogens relative to other probiotics in aquaculture; therefore, understanding the various mechanisms used by Bacillus in combating diseases will help improve their mode of action hence yielding better results in their combat against pathogens in the aquaculture industry. Thus, an overview of the mechanisms (production of bacteriocins, suppression of virulence gene expression, competition for adhesion sites, production of lytic enzymes, production of antibiotics, immunostimulation, competition for nutrients and energy, and production of organic acids) used by Bacillus probiotics in mitigating fish pathogens ranging from Aeromonas, Vibrio, Streptococcus, Yersinia, Pseudomonas, Clostridium, Acinetobacter, Edwardsiella, Flavobacterium, white spot syndrome virus, and infectious hypodermal and hematopoietic necrosis virus proven to be mitigated by Bacillus have been provided.

Multiplex PCR for genotyping Flavobacterium columnare

Recent research has identified four distinct genetic groups among isolates of Flavobacterium columnare through multilocus phylogenetic analyses; however, there are no quick methods to determine the genotype of an isolate. The objective of this research was to develop a multiplex PCR to rapidly genotype F. columnare to genetic group. Comparative bacterial genomics was used to identify regions in the genomes unique to each genetic group, and primers were designed to specifically amplify different sized amplicons for each genetic group. The optimized assay was demonstrated to be specific for each genetic group and F. columnare, and no specific amplicons were generated using gDNA from a panel of other Flavobacterium spp. and bacterial fish pathogens. The analytical sensitivity of the assay ranged from 209 to 883 genome equivalents depending on the genetic group. The multiplex PCR was evaluated by genotyping a panel of 22 unknown F. columnare isolates and performing DNA sequencing of the dnaK gene in parallel. The results demonstrated 100% accordance between multiplex PCR results and assignment to genetic group via phylogenetic analysis. The multiplex PCR provides a useful tool for assigning an unknown isolate to genetic group and may be used to determine which genetic groups of F. columnare are circulating and most predominant in different aquaculture industries.

Influence of native catfish mucus on Flavobacterium columnare growth and proteolytic activity

Flavobacterium columnare causes columnaris disease of farmed and wild freshwater fish. Skin mucus is an important factor in early stages of columnaris pathogenesis, albeit little studied. Our objectives were to (a) characterize the terminal glycosylation pattern (TGP) of catfish mucus, (b) determine the growth of F. columnare in formulated water (FW)-containing channel catfish (Ictalurus punctatus) or hybrid catfish (Ictalurus punctatus X Ictalurus furcatus) mucus and (c) examine extracellular protease activity of two F. columnare isolates differing in virulence. The TGP of catfish mucus by lectin binding was as follows: alpha-D-mannose/alpha-D-glucose >N-acetyl-beta-D-glucosamine >N-acetyl-beta-D-glucosamine/N-acetylneuraminic acid >N-acetyl-D-galactosamine >alpha-D-galactose/N-acetyl-alpha-D-galactosamine >beta-D-galactose = alpha-L-fucose. Virulence studies demonstrated isolate AL-02-36 was highly virulent in channel catfish fry (0.1 g) with cumulative mortality of 90%-100% versus 60% for isolate ALG-00-530 at equivalent doses (~3 × 10⁶  CFU/ml); a similar result was observed in larger (0.7 g) catfish. In multiple experiments, F. columnare replicated (2-3 logs) and survived (28 days) in formulated water-containing catfish mucus. Highly virulent isolate AL-02-36 possessed at least 2.5- to fivefold higher protease activity following growth in mucus than the less virulent ALG-00-530. Flavobacterium columnare utilized catfish mucus as a nutrient source and mucus presence modulated extracellular protease production.

Hepcidin protects grass carp (Ctenopharyngodon idellus) against Flavobacterium columnare infection via regulating iron distribution and immune gene expression

Columnaris disease (CD) caused by Flavobacterium columnare (F. columnare) is lack of knowledge on effective treatment measures. Bacterial pathogens require iron as an essential nutrient to infect the host. While hepcidin acts as a master regulator in iron metabolism, its contribution to host defense is emerging as complex and multifaceted. In vitro, recombinant Ctenopharyngodon idellus (C. idellus) hepcidin (CiHep) and synthetic CiHep both showed the ability to increase the expression of hepcidin and ferritin in C. idellus kidney cells, especially the recombinant CiHep. In vivo, recombinant CiHep improved the survival rate of C. idellus challenged with F. columnare. In addition, the fish fed diet containing recombinant CiHep (group H-1) had a higher survival rate than other pretreatment groups. The study showed that recombinant CiHep regulated iron metabolism causing iron redistribution, decreasing serum iron levels and increasing iron accumulation in the hepatopancreas. Moreover, the expression of iron-related genes was upregulated in various degrees at a different time except for group H-1. Immune-related genes were also evaluated, showing higher expression in the groups pretreated with CiHep at an early stage of infection. Of note, a clear upregulation of more immune genes occurred in the groups pretreated with recombinant CiHep than that pretreated with synthetic CiHep in the late stage of infection. In conclusion, the recombinant CiHep has a protective effect on the host response to bacterial pathogens. We speculate that hepcidin protects C. idellus against F. columnare infection via regulating the iron distribution and immune gene expression.

Identification of Four Distinct Phylogenetic Groups in Flavobacterium columnare With Fish Host Associations

Columnaris disease, caused by the Gram-negative bacterium Flavobacterium columnare, is one of the most prevalent fish diseases worldwide. An exceptionally high level of genetic diversity among isolates of F. columnare has long been recognized, whereby six established genomovars have been described to date. However, little has been done to quantify or characterize this diversity further in a systematic fashion. The objective of this research was to perform phylogenetic analyses of 16S rRNA and housekeeping gene sequences to decipher the genetic diversity of F. columnare. Fifty isolates and/or genomes of F. columnare, originating from diverse years, geographic locations, fish hosts, and representative of the six genomovars were analyzed in this study. A multilocus phylogenetic analysis (MLPA) of the 16S rRNA and six housekeeping genes supported four distinct F. columnare genetic groups. There were associations between genomovar and genetic group, but these relationships were imperfect indicating that genomovar assignment does not accurately reflect F. columnare genetic diversity. To expand the dataset, an additional 90 16S rRNA gene sequences were retrieved from GenBank and a phylogenetic analysis of this larger dataset also supported the establishment of four genetic groups. Examination of isolate historical data indicated biological relevance to the identified genetic diversity, with some genetic groups isolated preferentially from specific fish species or families. It is proposed that F. columnare isolates be assigned to the four genetic groups defined in this study rather than genomovar in order to facilitate a standard nomenclature across the scientific community. An increased understanding of which genetic groups are most prevalent in different regions and/or aquaculture industries may allow for the development of improved targeted control and treatment measures for columnaris disease.

The Type IX Secretion System Is Required for Virulence of the Fish Pathogen Flavobacterium columnare

Flavobacterium columnare, a member of the phylum Bacteroidetes, causes columnaris disease in wild and aquaculture-reared freshwater fish. The mechanisms responsible for columnaris disease are not known. Many members of the phylum Bacteroidetes use type IX secretion systems (T9SSs) to secrete enzymes, adhesins, and proteins involved in gliding motility. The F. columnare genome has all of the genes needed to encode a T9SS. gldN, which encodes a core component of the T9SS, was deleted in wild-type strains of F. columnare The F. columnare ΔgldN mutants were deficient in the secretion of several extracellular proteins and lacked gliding motility. The ΔgldN mutants exhibited reduced virulence in zebrafish, channel catfish, and rainbow trout, and complementation restored virulence. PorV is required for the secretion of a subset of proteins targeted to the T9SS. An F. columnare ΔporV mutant retained gliding motility but exhibited reduced virulence. Cell-free spent media from exponentially growing cultures of wild-type and complemented strains caused rapid mortality, but spent media from ΔgldN and ΔporV mutants did not, suggesting that soluble toxins are secreted by the T9SS.IMPORTANCE Columnaris disease, caused by F. columnare, is a major problem for freshwater aquaculture. Little is known regarding the virulence factors produced by F. columnare, and control measures are limited. Analysis of targeted gene deletion mutants revealed the importance of the type IX protein secretion system (T9SS) and of secreted toxins in F. columnare virulence. T9SSs are common in members of the phylum Bacteroidetes and likely contribute to the virulence of other animal and human pathogens.

Comparative genome analysis of fish pathogen Flavobacterium columnare reveals extensive sequence diversity within the species

Flavobacterium columnare is one of the deadliest fish pathogens causing devastating mortality in various freshwater fish species globally. To gain an insight into bacterial genomic contents and structures, comparative genome analyses were performed using the reference and newly sequenced genomes of F. columnare including genomovar I, II and I/II strains isolated from Thailand, Europe and the USA. Bacterial genomes varied in size from 3.09 to 3.39Mb (2714 to 3101 CDSs). The pan-genome analysis revealed open pan-genome nature of F. columnare strains, which possessed at least 4953 genes and tended to increase progressively with the addition of a new genome. Genomic islands (GIs) present in bacterial genomes were diverse, in which 65% (39 out of 60) of possible GIs were strain-specific. A CRISPR/cas investigation indicated at least two different CRISPR systems with varied spacer profiles. On the other hand, putative virulence genes, including those related to gliding motility, type IX secretion system (T9SS), outer membrane proteins (Omp), were equally distributed among F. columnare strains. The MLSA scheme categorized bacterial strains into nine different sequence types (ST 9-17). Phylogenetic analyses based on either 16S rRNA, MLSA and concatenated SNPs of core genome revealed the diversity of F. columnare strains. DNA homology analysis indicated that the estimated digital DNA-DNA hybridization (dDDH) between strains of genomovar I and II can be as low as 42.6%, while the three uniquely tilapia-originated strains from Thailand (1214, NK01 and 1215) were clearly dissimilar to other F. columnare strains as the dDDH values were only 27.7-30.4%. Collectively, this extensive diversity among bacterial strains suggested that species designation of F. columnare would potentially require re-emendation.

Isolation and characterization of Flavobacterium columnare strains infecting fishes inhabiting the Laurentian Great Lakes basin

Flavobacterium columnare, the aetiological agent of columnaris disease, causes significant losses in fish worldwide. In this study, the prevalence of F. columnare infection was assessed in representative Great Lakes fish species. Over 2000 wild, feral and hatchery-propagated salmonids, percids, centrarchids, esocids and cyprinids were examined for systemic F. columnare infections. Logistic regression analyses showed that the prevalence of F. columnare infection varied temporally and by the sex of the fish, whereby females had significantly higher prevalence of infection. A total of 305 isolates of F. columnare were recovered. Amplification of the near complete 16S rRNA gene from 34 representative isolates and subsequent restriction fragment length polymorphism analyses demonstrated that all belonged to F. columnare genomovar I. Phylogenetic analysis of near complete 16S rRNA gene sequences also placed the isolates in genomovar I, but revealed some intragenomovar heterogeneity. Together, these results suggest that F. columnare genomovar I is widespread in the Great Lakes Basin, where its presence may lead to mortality.

Comparative Genomics and Transcriptional Analysis of Flavobacterium columnare Strain ATCC 49512

Flavobacterium columnare is a Gram-negative fish pathogen causing columnaris disease in wild and cultured fish species. Although the pathogen is widespread in aquatic environments and fish worldwide, little is known about biology of F. columnare and mechanisms of columnaris disease pathogenesis. Previously we presented the complete genome sequence of F. columnare strain ATCC 49512. Here we present a comparison of the strain ATCC 49512 genome to four other Flavobacterium genomes. In this analysis, we identified predicted proteins whose functions indicate F. columnare is capable of denitrification, which would enable anaerobic growth in aquatic pond sediments. Anaerobic growth of F. columnare ATCC 49512 with nitrate supplementation was detected experimentally. F. columnare ATCC 49512 had a relatively high number of insertion sequences and genomic islands compared to the other Flavobacterium species, suggesting a larger degree of horizontal gene exchange and genome plasticity. A type VI subtype III secretion system was encoded in F. columnare along with F. johnsoniae and F. branchiophilum. RNA sequencing proved to be a valuable technique to improve annotation quality; 41 novel protein coding regions were identified, 16 of which had a non-traditional start site (TTG, GTG, and CTT). Candidate small noncoding RNAs were also identified. Our results improve our understanding of F. columnare ATCC 49512 biology, and our results support the use of RNA sequencing to improve annotation of bacterial genomes, particularly for type strains.

Draft Genome Sequence of the Fish Pathogen Flavobacterium columnare Strain CSF-298-10

We announce here the draft genome assembly of Flavobacterium columnare CSF-298-10, a strain isolated from an outbreak of columnaris disease at a commercial trout farm in Hagerman Valley, Idaho, USA. The complete genome consists of 13 contigs totaling 3,284,579 bp, with an average G+C content of 31.5% and 2,933 predicted coding genes.

Virulence of Flavobacterium columnare genomovars in rainbow trout Oncorhynchus mykiss

Flavobacterium columnare is the causative agent of columnaris disease and is responsible for significant economic losses in aquaculture. F. columnare is a Gram-negative bacterium, and 5 genetic types or genomovars have been described based on restriction fragment length polymorphism of the 16S rRNA gene. Previous research has suggested that genomovar II isolates are more virulent than genomovar I isolates to multiple species of fish, including rainbow trout Oncorhynchus mykiss. In addition, improved genotyping methods have shown that some isolates previously classified as genomovar I, and used in challenge experiments, were in fact genomovar III. Our objective was to confirm previous results with respect to genomovar II virulence, and to determine the susceptibility of rainbow trout to other genomovars. The virulence of 8 genomovar I, 4 genomovar II, 3 genomovar II-B, and 5 genomovar III isolates originating from various sources was determined through 3 independent challenges in rainbow trout using an immersion challenge model. Mean cumulative percent mortality (CPM) of ~49% for genomovar I isolates, ~1% for genomovar II, ~5% for the II-B isolates, and ~7% for the III isolates was observed. The inability of genomovar II isolates to produce mortalities in rainbow trout was unanticipated based on previous studies, but may be due to a number of factors including rainbow trout source and water chemistry. The source of fish and/or the presence of sub-optimal environment may influence the susceptibility of rainbow trout to different F. columnare genomovars.

Virulence assay of rhizoid and non-rhizoid morphotypes of Flavobacterium columnare in red tilapia, Oreochromis sp., fry

Numerous isolates of Flavobacterium columnare were previously recovered from red tilapia, Oreochromis sp., exhibiting columnaris-like disease in Thai farms, and the phenotypic and genetic characteristics were described. The objective of this study was to determine the virulence of two morphotypes (rhizoid and non-rhizoid colonies) of F. columnare and to determine their ability to adhere to and persist in red tilapia fry. The results showed that the typical rhizoid isolate (CUVET1214) was a highly virulent isolate and caused 100% mortality within 24 h following bath challenge of red tilapia with three different doses. The non-rhizoid isolate (CUVET1201) was avirulent to red tilapia fry. Both morphotypes adhered to and persisted in tilapia similarly at 0.5 and 6 h post-challenge as determined by whole fish bacterial loads. At 24 and 48 h post-challenge, fry challenged with the rhizoid morphotype exhibited significantly higher bacterial loads than the non-rhizoid morphotype. The results suggested that an inability of the non-rhizoid morphotype to persist in tilapia fry may explain lack of virulence.

Rainbow trout (Oncorhynchus mykiss) resistance to columnaris disease is heritable and favorably correlated with bacterial cold water disease resistance

Columnaris disease (CD), caused by Flavobacterium columnare, is an emerging disease affecting rainbow trout aquaculture. Objectives of this study were to 1) estimate heritability of CD resistance in a rainbow trout line (ARS-Fp-R) previously selected 4 generations for improved bacterial cold water disease (BCWD) resistance; 2) estimate genetic correlations among CD resistance, BCWD resistance, and growth to market BW; and 3) compare CD resistance among the ARS-Fp-R, ARS-Fp-S (selected 1 generation for increased BCWD susceptibility), and ARS-Fp-C (selection control) lines. Heritability of CD resistance was estimated using data from a waterborne challenge of 44 full-sib ARS-Fp-R families produced using a paternal half-sib mating design, and genetic correlations were estimated using these data and 5 generations of BCWD resistance, 9-mo BW (approximately 0.5 kg), and 12-mo BW (approximately 1.0 kg) data from 405 ARS-Fp-R full-sib families. The CD and BCWD challenges were initiated at approximately 52 and 84 d posthatch, or approximately 650 and 1,050 degree days (°C × d), respectively. Survival of ARS-Fp-R families ranged from 0 to 48% following CD challenge and heritability estimates were similar between CD (0.17 ± 0.09) and BCWD (0.18 ± 0.03) resistance, and the genetic correlation between these 2 traits was favorable (0.35 ± 0.25). Genetic correlations were small and antagonistic (-0.15 ± 0.08 to -0.19 ± 0.24) between the 2 resistance traits and 9- and 12-mo BW. Two challenges were conducted in consecutive years to compare CD resistance among ARS-Fp-R, ARS-Fp-C, and ARS-Fp-S families. In the first challenge, ARS-Fp-R families (83% survival) had greater CD resistance than ARS-Fp-C (73.5%; P = 0.02) and ARS-Fp-S (68%; P < 0.001) families, which did not differ (P = 0.16). In the second challenge, using an approximately 2.5-fold greater challenge dose, ARS-Fp-R families exhibited greater CD resistance (56% survival) than ARS-Fp-S (38% survival; P = 0.02) families. The favorable genetic correlation between CD and BCWD resistance is supported by greater CD resistance of the ARS-Fp-R line compared to the ARS-Fp-C and ARS-Fp-S lines and suggests that both traits will be improved simultaneously when selection is practiced on only 1 trait. In summary, these data indicate the feasibility of further selective breeding of the BCWD-resistant ARS-Fp-R line for increased CD resistance to produce a double pathogen-resistant line of rainbow trout.

Effect of Nanophyetus salmincola and Bacterial Co-Infection on Mortality of Juvenile Chinook Salmon

The freshwater trematode Nanophyetus salmincola has been demonstrated to impair salmonid immune function and resistance to the marine pathogen Vibrio anguillarum, potentially resulting in ocean mortality. We examined whether infection by the parasite N. salmincola similarly increases mortality of juvenile Chinook Salmon Oncorhynchus tshawytscha when they are exposed to the freshwater pathogens Flavobacterium columnare or Aeromonas salmonicida, two bacteria that juvenile salmonids might encounter during their migration to the marine environment. We used a two-part experimental design where juvenile Chinook Salmon were first infected with N. salmincola through cohabitation with infected freshwater snails, Juga spp., and then challenged with either F. columnare or A. salmonicida. Cumulative percent mortality from F. columnare infection was higher in N. salmincola-parasitized fish than in nonparasitized fish. In contrast, cumulative percent mortality from A. salmonicida infection did not differ between N. salmincola-parasitized and nonparasitized groups. No mortalities were observed in the N. salmincola-parasitized-only and control groups from either challenge. Our study demonstrates that a relatively high mean intensity (>200 metacercariae per posterior kidney) of encysted N. salmincola metacercariae can alter the outcomes of bacterial infection in juvenile Chinook Salmon, which might have implications for disease in wild fish populations.

New hosts and genetic diversity of Flavobacterium columnare isolated from Brazilian native species and Nile tilapia

Flavobacterium columnare is responsible for disease outbreaks in freshwater fish farms. Several Brazilian native fish have been commercially exploited or studied for aquaculture purposes, including Amazon catfish Leiarius marmoratus × Pseudoplatystoma fasciatum and pacamã Lophiosilurus alexandri. This study aimed to identify the aetiology of disease outbreaks in Amazon catfish and pacamã hatcheries and to address the genetic diversity of F. columnare isolates obtained from diseased fish. Two outbreaks in Amazon catfish and pacamã hatcheries took place in 2010 and 2011. Four F. columnare strains were isolated from these fish and identified by PCR. The disease was successfully reproduced under experimental conditions for both fish species, fulfilling Koch's postulates. The genomovar of these 4 isolates and of an additional 11 isolates from Nile tilapia Oreochromis niloticus was determined by 16S rRNA restriction fragment length polymorphism PCR. The genetic diversity was evaluated by phylogenetic analysis of the 16S rRNA gene and repetitive extragenic palindromic PCR (REP-PCR). Most isolates (n = 13) belonged to genomovar II; the remaining 2 isolates (both from Nile tilapia) were assigned to genomovar I. Phylogenetic analysis and REP-PCR were able to demonstrate intragenomovar diversity. This is the first report of columnaris in Brazilian native Amazon catfish and pacamã. The Brazilian F. columnare isolates showed moderate diversity, and REP-PCR was demonstrated to be a feasible method to evaluate genetic variability in this bacterium.

The Influence of Infective Dose on the Virulence of a Generalist Pathogen in Rainbow Trout (Oncorhynchus mykiss) and Zebra Fish (Danio rerio)

Pathogen density and genetic diversity fluctuate in the outside-host environment during and between epidemics, affecting disease emergence and the severity and probability of infections. Although the importance of these factors for pathogen virulence and infection probability has been acknowledged, their interactive effects are not well understood. We studied how an infective dose in an environmentally transmitted opportunistic fish pathogen, Flavobacterium columnare, affects its virulence both in rainbow trout, which are frequently infected at fish farms, and in zebra fish, a host that is not naturally infected by F. columnare. We used previously isolated strains of confirmed high and low virulence in a single infection and in a co-infection. Infection success (measured as host morbidity) correlated positively with dose when the hosts were exposed to the high-virulence strain, but no response for the dose increase was found when the hosts were exposed to the low-virulence strain. Interestingly, the co-infection resulted in poorer infection success than the single infection with the high-virulence strain. The rainbow trout were more susceptible to the infection than the zebra fish but, in both species, the effects of the doses and the strains were qualitatively similar. We suggest that as an increase in dose can lead to increased host morbidity, both the interstrain interactions and differences in infectivity in different hosts may influence the severity and consequently the evolution of disease. Our results also confirm that the zebra fish is a good laboratory model to study F. columnare infection.

Phenotypic characterization and genetic diversity of Flavobacterium columnare isolated from red tilapia, Oreochromis sp., in Thailand

Flavobacterium columnare is the aetiological agent of columnaris disease and severely affects various freshwater aquaculture fish species worldwide. The objectives of this study were to determine the phenotypic characteristics and genetic variability among F. columnare isolates isolated from red tilapia in Thailand. Forty-four F. columnare isolates were recovered from diseased fish in different geographical locations. The isolates exhibited homologous phenotypic characteristics but exhibited genetic diversity. One isolate was assigned to genomovar I, and the remainder were assigned to genomovar II, indicating the coexistence of these genomovars but predominance of genomovar II. Phylogenetic analysis of the 16S-23S ISR sequences revealed that a subset of the Thai isolates (n = 25) contained a smaller intergenic spacer region (ISR) (523-537 bp) and formed a unique ISR phylogenetic group. Phylogenetic analysis of the 16S rRNA gene supported the unique cluster of Thai isolates. This is the first description of the phenotypic and molecular characteristics of F. columnare isolated from red tilapia in Thailand as well as five isolates of F. columnare derived from other fish species including Nile tilapia, koi carp and striped catfish.

Lack of association between Flavobacterium columnare genomovar and virulence in hybrid tilapia Oreochromis niloticus (L.)×Oreochromis aureus (Steindachner)

Columnaris disease can be problematic in tilapia (Oreochromis spp.) production. An understanding of the pathogenesis and virulence of Flavobacterium columnare is needed to develop prevention strategies. The objective of this study was to determine the virulence of genetically defined isolates of F. columnare in sex-reversed hybrid tilapia, Oreochromis niloticus (L.)×O. aureus (Steindachner). A series of immersion challenge trials were performed using isolates of the five established genomovars of F. columnare: I, II, II-B, III and I/II. The mean per cent mortality of fish challenged with genomovar I, II and III isolates ranged from 0 to 100, 3.3-78 and 3.3-75%, respectively. The mean per cent mortality of fish challenged with genomovar II-B ranged from 35 to 96.7%, and the only genomovar I/II isolate tested caused no mortality. Contrary to previous work in other fish species, there did not appear to be an association between F. columnare genomovar and virulence in tilapia. The challenge model used resulted in acute mortality. An alternative challenge model was tested by cohabitating healthy fish with dead fish infected with F. columnare. This method resulted in rapid appearance of clinical signs and mortality, suggesting the potential for F. columnare to increase in virulence upon growth on/in a fish host.

Interactions of highly and low virulent Flavobacterium columnare isolates with gill tissue in carp and rainbow trout

The interactions of Flavobacterium columnare isolates of different virulence with the gills of carp (Cyprinus carpio L.) and rainbow trout (Oncorhynchus mykiss Walbaum) were investigated. Both fish species were exposed to different high (HV) or low virulence (LV) isolates and sacrificed at seven predetermined times post-challenge. Histopathological and ultrastructural examination of carp and rainbow trout inoculated with the HV-isolate disclosed bacterial invasion and concomitant destruction of the gill tissue, gradually spreading from the filament tips towards the base, with outer membrane vesicles surrounding most bacterial cells. In carp, 5-10% of the fish inoculated with the LV-isolate became moribund and their gill tissue displayed the same features as described for the HV-isolate, albeit to a lesser degree. The bacterial numbers retrieved from the gill tissue were significantly higher for HV- compared to LV-isolate challenged carp and rainbow trout. TUNEL-stained and caspase-3-immunostained gill sections demonstrated significantly higher apoptotic cell counts in carp and rainbow trout challenged with the HV-isolate compared to control animals. Periodic acid-Schiff/alcian blue staining demonstrated a significantly higher total gill goblet cell count for HV- and LV-isolate challenged compared to control carp. Moreover, bacterial clusters were embedded in a neutral matrix while being encased by acid mucins, resembling biofilm formation. Eosinophilic granular cell counts were significantly higher in the HV-isolate compared to LV-isolate inoculated and control carp. The present data indicate a high colonization capacity, and the destructive and apoptotic-promoting features of the HV-isolate, and point towards important dynamic host mucin–F. columnare interactions warranting further research.

Gill infection model for columnaris disease in common carp and rainbow trout

Challenge models generating gill lesions typical for columnaris disease were developed for the fry of both Common Carp Cyprinus carpio and Rainbow Trout Oncorhynchus mykiss by means of an immersion challenge and Flavobacterium columnare field isolates were characterized regarding virulence. Carp inoculated with highly virulent isolates revealed diffuse, whitish discoloration of the gills affecting all arches, while in trout mostly unilateral focal lesions, which were restricted to the first two gill arches, occurred. Light microscopic examination of the gills of carp exposed to highly virulent isolates revealed a diffuse loss of branchial structures and desquamation and necrosis of gill epithelium with fusion of filaments and lamellae. In severe cases, large parts of the filaments were replaced with necrotic debris entangled with massive clusters of F. columnare bacterial cells enwrapped in an eosinophilic matrix. In trout, histopathologic lesions were similar but less extensive and much more focal, and well delineated from apparently healthy tissue. Scanning and transmission electron microscopic observations of the affected gills showed long, slender bacterial cells contained in an extracellular matrix and in close contact with the destructed gill tissue. This is the first study to reveal gill lesions typical for columnaris disease at macroscopic, light microscopic, and ultrastructural levels in both Common Carp and Rainbow Trout following a challenge with F. columnare. The results provide a basis for research opportunities to examine pathogen-gill interactions.

Epidemiology of columnaris disease affecting fishes within the same watershed

In the southeastern USA, columnaris disease (caused by Flavobacterium columnare) typically affects catfish raised in earthen ponds from early spring until late summer. Recently, unusually severe outbreaks of columnaris disease occurred at the E. W. Shell Fisheries Center located in Auburn, AL, USA. During these outbreaks, catfish and other aquaculture and sport fish species that were in ponds located within the same watershed were affected. Our objective was to investigate the genetic diversity among F. columnare isolates recovered from different sites, sources, and dates to clarify the origin of these outbreaks and, ultimately, to better understand the epidemiology of columnaris disease. A total of 102 F. columnare isolates were recovered from catfishes (channel catfish Ictalurus puntactus, blue catfish I. furcatus, and their hybrid), bluegill Lepomis microchirus, Nile tilapia Oreochromis niloticus, largemouth bass Micropterus salmoides, egg masses, and water during columnaris outbreaks (from spring 2010 to summer 2012). Putative F. columnare colonies were identified following standard protocols. All isolates were ascribed to Genomovar II following restriction fragment length polymorphism analysis of the 16S rRNA gene. Genetic variability among the isolates was revealed by amplified fragment length polymorphism. Date of isolation explained most of the variability among our isolates, while host was the least influential parameter, denoting a lack of host specificity within Genomovar II isolates. The susceptibility of each of the isolates against commonly used antibiotics was tested by antibiogram. Our data showed that 19.6 and 12.7% of the isolates were resistant to oxytetracycline and kanamycin, respectively.

Intragenomic heterogeneity in the 16S rRNA genes of Flavobacterium columnare and standard protocol for genomovar assignment

Genetic variability in 16S rRNA gene sequences has been demonstrated among isolates of Flavobacterium columnare, and a restriction fragment length polymorphism (RFLP) assay is available for genetic typing of this important fish pathogen. Interpretation of restriction patterns can be difficult due to the lack of a formal description of the expected number and sizes of DNA fragments generated for each of the described genomovars. In this study, partial 16S rRNA gene sequences (ca. 1250-bp fragment) from isolates representing each described genomovar and isolates generating unique restriction patterns were cloned and sequenced. The results demonstrated that some isolates contained up to three different 16S rRNA genes whose sequences generate different RFLP patterns due to intragenomic heterogeneity within HaeIII restriction sites. The occurrence of HaeIII restriction sites within the portion of the 16S rRNA gene used for typing the F. columnare isolates and intragenomic heterogeneity within these sites explained the restriction patterns observed following RFLP analyses. This research provides a standard protocol for typing isolates of F. columnare by RFLP and a formal description of the expected restriction patterns for the previously described genomovars I, II, II-B and III. Additionally, we describe a new genomovar, I/II.

Columnaris disease in fish: a review with emphasis on bacterium-host interactions

Flavobacterium columnare (F. columnare) is the causative agent of columnaris disease. This bacterium affects both cultured and wild freshwater fish including many susceptible commercially important fish species. F. columnare infections may result in skin lesions, fin erosion and gill necrosis, with a high degree of mortality, leading to severe economic losses. Especially in the last decade, various research groups have performed studies aimed at elucidating the pathogenesis of columnaris disease, leading to significant progress in defining the complex interactions between the organism and its host. Despite these efforts, the pathogenesis of columnaris disease hitherto largely remains unclear, compromising the further development of efficient curative and preventive measures to combat this disease. Besides elaborating on the agent and the disease it causes, this review aims to summarize these pathogenesis data emphasizing the areas meriting further investigation.