We characterized six filamentous, gram-negative bacterial isolates, each recovered from a naturally occurring salmonid bacterial gill disease (BGD) outbreak in Ontario. The morphological, physiological, biochemical, and antigenic traits of the Ontario isolates were similar to those of Flavobacterium branchiophilum (ATCC 35035, ATCC 35036) recovered from salmonid BGD in Japan and Oregon. We conclude that the Ontario BCD isolates are strains of F. branchiophilum.

No abstract

Three trials were conducted to determine if feeding influenced the development of bacterial gill disease (BGD) in fingerling rainbow trout Oncorhynchus mykiss following bath exposure to Flavobacterium branchiophilum (LAB 4a). In each trial (Trial 1. n = 70 X 3; Trial 2, n = 140 X 3 ; Trial 3, n = 250 X 3), the comparison groups (fed, unfed, negative controls) were held in similar conditions and (where applicable) fed prior to exposure to the pathogen. Severe clinical signs and high rates of mortality developed within 24 h in fish which were provided feed following the bath exposure. Cumulative mortalities ranged from 43 to 63% by 12 d post-exposure Unfed fish demonstrated only mild and transient clinical signs of BGD, and cumulative mortalities of 0 to 2 % 12 d post-exposure. In one trial, resumption of feeding 6 d post-exposure resulted in 26% cumulative mortality within 7 d. Estimation of the quantity of F branchiophilum antigen by enzyme-linked immunosorbent assay (ELISA) indicated that feeding facilitated colonization of the gill. The evaluation of water quality and direct observations suggest that the influence of feeding on BGD is linked to the consumption of feed by the fish, rather than to environmental changes arising from feeding. Alterations in the unstirred layer on the gill, secondary to feed consumption and waste excretion, may aid bacterial colonization.

No abstract

The ability of selected strains of Flavobacterium branchiophilum to attach to and colonize the gills of five species of teleosts (four salmonid and one cyprinid) and cause mortality was investigated. In virulence studies with F. branchiophilum strain LAB4a, cumulative mortality was dose-dependent in rainbow trout, Oncorhynchus mykiss (Walbaum), and ranged from 0 to 75%. However, regardless of dose, the relative amount of gill-associated F. branchiophilum antigen inereased lh after challenge. The gill-associated F. branchiophilum antigen in fish which survived (moribund fish) increased by four to six times compared to that detected on the gills lh after challenge. The gill-associated antigen on moribund fish was not significantly different from that on the gills of dead fish. Flavobacterium branchiophilum strain LAB4a also attached to the gills of brook trout, Salvelinus fontinalis (Mitchell), rainbow trout, chinook salmon, Oncorhynchus tshawytscha (Walbaum), Arctic charr, Salvehnus alpinus (L.), and common shiners, Luxilus cornutus (Mitchell), following a 1-h bath exposure, and caused mortality in all species. The virulence of eight strains of F. branchiophilum for rainbow trout was examined. Both virulent and avirulent strains adhered to the gills following bath exposure (fimbriae were observed on all strains), but only virulent strains had the capacity to further colonize the gills and cause mortality.

Rainbow trout were experimentally infected with the causative agent of bacterial gill disease (BGD) (Flavobacterium branchiophilum) via bath challenge. All fish were cannulated with dorsal aortic catheters, had nasogastric tubes sutured in place for feeding, and were maintained individually, in plexiglass boxes with a flow-through water system. Fish were either fed, or unfed during the trial. Acute changes in blood gas, serum biochemistry and clinical parameters were monitored. By 24h post-challenge, BGD-infected trout that had been fed had significant hypoxemia, hypercapnia, increased blood ammonia, hypoosmolality, hyponatremia, hypochloremia, and increased cough and respiratory rates when compared to control levels. Unfed BGD-infected trout had similar, but less severe blood gas and clinical changes, and no electrolyte disturbances. The BGD-induced hypoxemia is likely exacerbated by increased oxygen demands brought on by feeding. It is not known what association feeding has with the development of low serum ion levels in BGD-infected trout. This is the first study to report the use of fed fish, as opposed to unfed or starved trout, in obtaining blood chemistry values from indisturbed and cannulated animals.

An unusual form of bacterial gill disease (BGD) was identified which affected five species of cultured salmonids from Canada (i.e. rainbow trout, chinook salmon and Atlantic salmon), Norway (i.e. brown trout) and Chile (i.e. coho salmon). All outbreaks occurred at low water temperatures (< 10 °C) and with clinical presentations distinct from classical BGD, which is caused by Flavobacterium branchiophilum. In contrast to classical BGD, fish did not show marked respiratory distress with flaring of the opercula, the animals did not orientate at the surface of the water column near inflow water or at the margins of the tanks, and the feed response of the fish was varied. While mortality was increased, it was not precipitous as in classical BGD. Eight outbreaks were examined in greater detail using histopathology, scanning electron microscopy, bacteriology and immunohistochemistry. Large numbers of small bacterial rods were seen adhering to the lamellar epithelium of affected gills from all outbreaks. Histologically, the lamellar epithelium appeared swollen, often with evidence of single cell degeneration and exfoliation. In more severe instances, the formation of lamellar synechiae was seen, usually associated with sequestration of bacteria between fused lamellae. By contrast with typical BGD, overt epithelial hyperplasia, lamellar fusion and filamental clubbing were not common sequelae to infection; instead, the end result was shortened and somewhat stubby lamellae covered with swollen epithelial cells. The predominant bacterium recovered from affected gills was a small, Gram‐negative, motile, fluorescent pigment‐producing rod that shared phenotypic characteristics with Pseudomonas fluorescens. Polyclonal antisera prepared against three representative isolates indicated a weak antigenic similarity among them. Immunohistochemistry corroborated this finding, in that the antisera reacted strongly with gill sections containing the homologous bacteria, but not against morphologically similar bacteria in heterologous sections. A Gram‐negative, yellow pigmented bacterium (YPB), identified as Flavobacterium psychrophilum, was also recovered, but only from the gills in the Ontario outbreaks. Antiserum prepared against this YPB indicated an antigenic similarity among isolates recovered from the Ontario outbreaks, but immunohistochemistry failed to recognize antigenically related bacteria on the gills of fish from the other outbreaks. Based on the unusual clinical presentation and the histopathological appearance of the gills, in conjunction with the absence of filamentous bacteria associated with and recovered from affected gills, the present authors have called this condition ‘atypical bacterial gill disease’ or ABGD.

Preliminary experiments are described which aimed to identify compounds that could inhibit the attachment of Flavobacterium branchiophilum strains LAB4a and ATCC 35035 to the gills of rainbow trout. Total inhibition was never achieved, regardless of the compound tested. Formalin‐killed or acetone‐killed F. branchiophilum cells retained at least some of their adherent nature, relative to untreated (live) cells. Adherence was reduced by 22–33% following immersion of fish in one litre of water containing 0.21 mg of a homologous crude fimbrial extract. When fish were immersed in water containing hyperimmune rainbow trout antiLAB4a serum, a dose‐dependent decrease in attachment (a reduction of 15% to 63%) of LAB4a to the gills was observed. Rainbow trout anti‐LAB4a serum also reduced the attachment of ATCC 35035 to the gills, but this reduction was not significant. Adherence of LAB4a was not inhibited following exposure of fish to group 1 carbohydrates (arabinose, mannose and xylose), group 2 carbohydrates (dextrose, galactose and lactose), group 3 carbohydrates (galactosamine, glucosamine and fucose) or group 4 carbohydrates (N‐acetyl‐D‐glucosamine, N‐acetyl‐D‐galactosamine, N‐acetyl‐neuraminic acid and the globoceramide glycolipid from human erythrocytes). In contrast, when rainbow trout erythrocytes were incubated with the bacteria prior to bath challenge, this resulted in an 87% and 53% reduction in gill‐associated LAB4a and ATCC 35035 antigen, respectively, following immersion of rainbow trout in this suspension.