Biofilm formation, virulence attenuation and comparative genomics of the fish pathogen Flavobacterium columnare

Abstract

Flavobacterium columnare is an important bacterial fish pathogen that causes great economic losses in aquaculture. In this dissertation I aimed at characterizing part of its life cycle including the transition between planktonic and biofilm stages and understanding its pathogenicity mechanisms using an avirulent mutant as model. In the first part, I evaluated biofilm formation on common aquaculture substrates. I found that F. columnare efficiently formed biofilm on pond liner, flexible PVC, and nets, while plant material prevented F. columnare attachment and inhibited cell growth. Biofilm formation on specific substrates was confirmed using Scanning Electron Microscope (SEM). In addition, I evaluated the role of calcium in biofilm formation using a microtiter plate assay, and the results showed that calcium supplementation greatly enhanced the biofilm formation. To understand the molecular mechanism involve in the transition between planktonic and biofilm stages as well as to characterize the role of calcium in biofilm formation, I used a transcriptome approach to identify Differential Expressed Genes (DEGs) among different life stages (i.e. planktonic and biofilm cells) and different calcium concentrations. Results showed that oxidative stress and nutrient starvation are predominant drivers in biofilm metabolic pathways, and that aerobic respiration is greatly limited during biofilm development. The DEGs under calcium simulation were also evaluated. I identified 175 DEGs (6.30% of genomic protein-coding sequences), which fall into functional categories including iron acquisition, biofilm signaling, T9SS system, and calcium homeostasis. Together, our data suggested that that biofilm is significantly affected by calcium, which seems to serve as a critical signal in controlling bacterial surface adhesion and biofilm formation in F. columnare. Our group had previously patented a modified-live vaccine (a rifampicin-resistant mutant) against columnaris disease. To understand the molecular basis for attenuation, the mutant and its parent strains were sequenced, and comparative genomic analysis was conducted to identify specific point mutations. Sequence-based genome comparison identified 16 single nucleotide polymorphisms (SNP) unique to the mutant. Genes that contain mutations were involved in rifampicin resistance, gliding motility, DNA transcription, toxin secretion, and protease synthesis. I also compared biofilm production between the mutant and the parent strain, and the results showed that the vaccine strain formed biofilm at a significantly lower level than the parent strain. F. columnare is a genetically heterogeneous species, which is comprised of several genetic groups (i.e. genomovar). In the last chapter, comparative genomic were conducted to further elucidate the genetic diversity behind this species. Three strains representing 3 different lineages within the species were sequenced and compared with 5 additional strains whose complete genome sequences were available. Results showed that all average nucleotide identity (ANI) values between the genomovars were lower than the recommended cut-off point of 95% for species delineation. The pan- and core-genomes were evaluated, and genes unique to each genomovar were retrieved. Our results revealed an extensively diversity within F. columnare species, whose genomic relatedness between the genomovars was below the cut-off threshold for species. Therefore, I propose to consider the species F. columnare as a species complex. In summary, my results identified critical genes/pathways related to surface colonization, biofilm formation, and pathogenicity of F. columnare. In addition, I confirmed the need for describing at least two cryptic species within the F. columnare species complex.