FEMS Microbiol Lett. 2014 Sep;358(1):14-20. doi: 10.1111/1574-6968.12543.

Attempts at validating a recombinant Flavobacterium psychrophilum gliding motility protein N as a vaccine candidate in rainbow trout, Oncorhynchus mykiss (Walbaum) against bacterial cold-water disease.

Plant KP, LaPatra SE, Call DR, Cain KD.

Hagerman Fish Culture Experiment Station, University of Idaho, Hagerman, ID, USA.



The Flavobacterium psychrophilum gliding motility N (GldN) protein was investigated to determine its ability to elicit antibody responses and provide protective immunity in rainbow trout Oncorhynchus mykiss (Walbaum). GldN was PCR-amplified, cloned into pET102/D-TOPO, and expressed in Escherichia coli. Bacteria expressing recombinant GldN (rGldN) were formalin-inactivated and injected intraperitoneally (i.p.) into rainbow trout with Freund’s complete adjuvant (FCA) in four separate studies that used two different immunization protocols followed by challenge evaluations. Fish injected with E. coli only in FCA served as the control. Antibody responses to F. psychrophilum whole-cell lysates measured by ELISA were low in all four studies. Protection against F. psychrophilum challenge was observed in the first study, but not in the three following studies. The discrepancies in results obtained in the later studies are unclear but may relate to formalin treatment of the antigen preparations. Overall, it appeared that rGldN delivered i.p. as a crude formalin-killed preparation is not a consistent vaccine candidate, and more work is required. Additionally, this study illustrates the importance of conducting multiple in vivo evaluations on potential vaccine(s) before any conclusions are drawn.

© 2014 Federation of European Microbiological Societies. Published by John Wiley & Sons Ltd.

KEYWORDS: Flavobacterium psychrophilum; formalin; gliding motility; vaccination

PMID: 25053267



Bacterial cold water disease (BCWD) caused by Flavobacterium psychrophilum is a major contributor to losses in rainbow trout, Oncorhynchus mykiss (Walbaum) and salmon farming around the world (Figure 1). Outbreaks of BCWD can result in 50-60% mortality but further losses due to deformities and reduced growth in survivors contribute to additional economic impacts for the farmer (Figure 2). Despite a number of publications addressing vaccination against BCWD there is still no commercially available vaccine. Proteins of F. psychrophilum have been identified including those in the 70-100 and 41-49 kDa range that can provide protection against BCWD in laboratory evaluations (1). A number of proteins within these fractions were subsequently identified as being immunogenic (2) along with other proteins of the bacteria. We have previously evaluated five of these  proteins for their efficacy as recombinant protein vaccines (3,4). The present study continues our previous work that failed to identify any potential recombinant protein vaccine candidates and examines the potential of recombinant gliding motility protein N (rGldN) as a vaccine against BCWD.


1 Figure 1. Freshwater rainbow trout (Oncorhynchus mykiss) aquaculture facilities located along the Snake River in southern Idaho, USA.


Figure 2. Vertebral compression deformity in a freshwater rainbow trout (Oncorhynchus mykiss) that survived a bacterial cold water disease epizootic.

Gliding motility is a common characteristic of members of the Bacteroidetes phylum of which F. psychrophilum is a member. Gliding occurs despite a lack of flagella or pili and the mechanism responsible in F. psychrophilum is unknown (Figure 3). However, in Flavobacterium johnsoniae gliding motility is thought to involve cell surface complexes consisting of the Gld proteins which facilitate the movement of adhesive proteins from one to another. A substantial amount of research has been carried out on gliding motility in F. johnsoniae that has identified multiple genes involved in gliding. As a result of the sequencing of the complete F. psychrophilum genome (5) comparison of the gliding motility genes between F. psychrophilum and F. johnsoniae revealed extensive similarities. In F. johnsoniae GldN is essential for motility and it is thought to be involved in the secretion of motility proteins to the cell surface in concert with a number of other Gld proteins. Recently, a link was found between protein translocation and the motility apparatus in members of the Bacteroidetes phylum. A protein secretion system was identified and termed the Por secretion system (PorSS) of which GldN is thought to be one component. The PorSS is thought to have a similar role to the type III secretion system present in other bacterial species and therefore could be exploited for disease control.



Figure 3. Spleen imprint from a freshwater rainbow trout (Oncorhynchus mykiss) infected with Flavobacterium psychrophilum. Note the long filamentous, rod shaped bacteria extracellularly and intracellularly.

Because our previous work on potential candidate antigen(s) was unsuccessful at identifying one for a recombinant vaccine, the F. psychrophilum GldN protein was examined. This choice was supported by: (1) identification of GldN as an immunogenic protein within the protective high- and mid-molecular mass fractions; (2) GldN is upregulated in vivo and in vitro under iron-restricted conditions; (3) GldN is probably exposed on the cell membrane; and (4) GldN has a potential role in the Por secretion system in F. johnsoniae.

The first study used a statistically reliable number of fish and the results suggested that rGldN expressed and delivered using whole cell E.coli was a potential vaccine candidate.  The overall results of the entire study, however, illustrate the problems with demonstrating such results consistently. Expression of rGldN in E. coli was relatively simple and over expression of rGldN was clearly observed by staining. Nevertheless, we were unable to purify the protein despite attempting many different methods.

The disappearance of rGldN at its predicted molecular weight after formalin treatment suggested the protein was cross-linked by the formalin and the positive results of the first study led us to believe it was a suitable inactivation method. Formaldehyde is commonly used to inactivate pathogens and toxins for vaccine use and it is quick to penetrate cells, inactivate enzymes and cross link proteins to proteins, DNA and RNA. Cross linking is dependent on the lysine residues on the surface of proteins and also the physical proximity of the proteins. The conditions used for formalin inactivation in the current study (0.5% formalin overnight at 4 °C) are not dissimilar from those used by Shimmoto et al. [6 (0.3% formalin for 24 h at 37 °C)] who demonstrated significant protection in red sea bream, Pagrus major   (Temminck & Schlegel), vaccinated with a major capsid protein of red sea bream iridovirus delivered as a crude preparation of recombinant protein expressed in formalin inactivated E. coli.

However, in the four in-vivo evaluations that were conducted the formalin killed E. coli expressing rGldN conferred good protection in the first study but did not confer significant protection in the subsequent studies (Figure 4). Although GldN is considered to be a membrane protein, it may not be completely surface exposed and it is possible this may account for the lack of protection observed in the latter studies. Another possibility is that a single protein may not be able to adequately stimulate the immune response to confer protection against BCWD. There are many other proteins involved in gliding motility and the putative PorSS in F. psychrophilum. It is possible that a combination of these or other proteins may be required to elicit a significant, reproducible protective response and should be examined in future studies.



Figure 4. Laboratory facility used for the in-vivo evaluation of potential vaccines for the control of bacterial cold water disease in freshwater farmed rainbow trout (Oncorhynchus mykiss).


In summary, F. psychrophilum rGldN expressed and delivered using whole cell E.coli delivered by intraperitoneal injection as a vaccine gave variable results. However, potentially alternative formalin concentrations and/or other inactivation methods could improve the consistency of this antigen along with including other proteins involved with gliding motility. This work also illustrates the importance of validating the results of potential vaccine candidates. If only the first challenge evaluation had been carried out the conclusion would have been that rGldN expressed and delivered via whole E. coli cellswas a viable vaccine candidate. Variation is inherent with in vivo evaluations and demonstration of reproducible results can only be accomplished through multiple challenge evaluations.



  1. LaFrentz BR, LaPatra SE, Jones GR, Cain KD 2004 Protective immunity in rainbow trout Oncorhynchus mykiss following immunization with distinct molecular mass fractions isolated from Flavobacterium psychrophilum. Diseases of Aquatic Organisms 59: 17-26.
  2. LaFrentz BR, LaPatra SE, Call DR, Wiens GD, Cain KD 2011 Identification of immunogenic proteins within distinct molecular mass fractions of Flavobacteriumpsychrophilum. Journal of Fish Diseases 34: 823-830.
  3. Plant KP, LaPatra SE, Cain KD 2009 Vaccination of rainbow trout, Oncorhynchus mykiss (Walbaum), with recombinant and DNA vaccines produced to Flavobacterium psychrophilum heat shock proteins 60 and 70. Journal of Fish Diseases 32: 521-534.
  4. Plant KP, LaPatra SE, Call DR, Cain KD 2011 Immunization of rainbow trout, Oncorhynchus mykiss (Walbaum) with Flavobacterium psychrophilum proteins elongation factor-Tu, SufB Fe-S assembly protein and ATP synthaseb. Journal of Fish Diseases34: 247-250.
  5. Duchaud E, Boussaha M, Loux V, Bernardet J, Michel C, Kerouault B, Mondot S, Nicolas P, Bossy R, Caron C, Bessières P, Gibrat J, Claverol S, Dumetz F, Le Hénaff M, Benmansour A 2007 Complete genome sequence of the fish pathogen Flavobacterium psychrophilum. Nature Biotechnology 25: 763-769.
  6. Shimmoto H, Kawai K, Ikawa T, Oshima S 2010 Protection of red sea bream Pagrus major against red sea bream iridovirus infection by vaccination with a recombinant viral protein. Microbiology immunology 54: 135-142.
Multiselect Ultimate Query Plugin by InoPlugs Web Design Vienna | Webdesign Wien and Juwelier SchönmannMultiselect Ultimate Query Plugin by InoPlugs Web Design Vienna | Webdesign Wien and Juwelier Schönmann