Discovering novel functions of carbohydrate-active enzymes to unravel novel mechanisms of bacterial virulence

The main aim of the project has been to unravel the putative virulence properties of a novel family of carbohydrate active enzymes called the lytic polysaccharide monooxygenases (LPMOs). Since we originally had hypothesized that the enzymes were important for the colonization and invasion of mucosal our initial studies involved analyzing the bacterial community present in the Atlantic salmon skin mucus and the proteins they secreted when allowed to proliferate in the mucus. Our results showed that there were not a rich bacterial flora on the Atlantic salmon skin, but they were able to grow in the mucus. No LPMOs were identified, but rather an abundance of proteases thought to help the microbes digest the mucin glycoprotein present in the skin mucus. We next brought our attention to the fish pathogen Aliivibrio salmonicida which has two LPMOs enoded in its genome. Firstly, we were able to show that the LPMOs were able to cleave chitin chains, indicating that they may be involved in acquisition of nutrients for the bacterium from the abundant marine polysaccharide chitin. Indeed, this was shown to be the case as variants of A. salmonicida not having these LPMO genes were not able to grow on chitin as efficiently as the non-modified bacterium. Secondly we investigated the ability of the A. salmonicida to infect and cause disease in Atlantic salmon smolts. Interestingly, the A. salmonicida variants lacking the LPMO genes showed a reduced ability to enter the invasive phase of the infection compared to the normal A. salmonicida strain. This indicates that the LPMOs are virulence factors that help the bacterium in certain phases of the infection process. It also shows that the LPMOs are multifunctional since they can be used for both nutrient acquisition and infection processes. To follow up our A. salmonicida results, we embarked on a similar study using the human pathogen Pseudomonas aeruginosa. This bacterium has one LPMO called CbpD. In a comprehensive study published in Nature Communications we were able to show that CbpD is of critical importance for the ability of the bacterium to maintain systemic or lung infection. These results were obtained by investigation of a P. aeruginosa variant lacking the CbpD gene in mouse infection models. We also discovered that the mechanism underlying the CbpD function was related to its ability to protect the bacterium against the parts of the complement system which is an important part of the immune system that for example kills bacteria. An important aspect of the project was obtaining a deeper understanding of LPMO reaction mechanism in general since such knowledge could give us important clues to how the LPMOs find and react with their target molecules in the host. In this research we studied LPMOs with known carbohydrate substrates and found, among others, that LPMOs are quickly inactivated if they are in a reactive environment in the absence of their substrate. In a study that was associated with the project, it was also found that hydrogen peroxide was an important co-substrate for LPMOs. This was crucial knowledge, especially for virulence-related LPMOs as hydrogen peroxide is encountered by bacterial pathogens in many phases of infection (e.g. on mucosal surfaces, in blood and in confrontation with immune cells). This finding was followed up by a study where we determined the proteome response of the fish pathogen Vibrio anguillarum when exposed to various concentrations of hydrogen peroxide. The main finding was that the bacterium changes its metabolism and also secretes enzymes that may help reduce the oxidative stress caused by the hydrogen peroxide. The LPMOs of the bacterium were not identified indicating that they are not involved in managing oxidative stress. In conclusion, the project has been a success as we have pushed the state of the art of LPMO research showing that these enzymes not only are essential for biomass degradation, but also highly important virulence factors for pathogenic bacteria.

Den faglige nytteverdien av prosjektet forventes å være stor da prosjektet har vist at LPMOer har en viktig rolle som virulensfaktorer for bakterier. Dette er helt ny viten og vi forventer at mange forskningsgrupper vil starte forskning på disse proteinene. Nå som LPMOer har blitt knyttet til bakterielle infeksjoner vil viten om disse enzymene også kunne ha betydning og nytteverdi innen medisinsk forskning og anvendelse. Siden disse enzymene er så viktige for bakterien for overlevelse i verten, kan det tenkes at medisiner som hemmer eller slår ut aktiviteten/funksjonen til LPMOene kan bidra til bekjempelse av bakterielle infeksjoner. Prosjektet har gitt stort utbytte når det gjelder utvikling av kompetanse for gjennomføring av multidisiplinære studier, både for forskningsgruppen og samarbeidspartnerne. Prosjektet har også gitt opphav til flere nye samarbeid og prosjekter som allerede er i gang.

Lytic polysaccharide monooxygenases (LPMOs) and chitinases are enzymes know to cleave the glycosidic bonds of chitin chains (GlcNAc-GlcNAc bonds). An accumulating body of evidence is now suggesting prominent roles for these enzymes in infection, but most reports are only indicative and lack multidisciplinary validation. The importance of these proteins as virulence factors may have been overlooked by a one-sided focus on elucidating their function in chitin metabolism. The overarching goal of the proposed project is to explore LPMOs and chitinases as virulence factors. We will specifically investigate Atlantic salmon and its bacterial pathogens, comprising a well established, experimentally accessible and highly relevant biological system. We will employ state-of-the-art -omics and enzymological tools to identify the regulation, properties and functions of chitinases and LPMOs during infection, and of unknown proteins co-expression with these enzymes. Enzyme activity of LPMOs and chitinases from the major pathogens of Atlantic salmon will initially be evaluated towards chitinous substrates, followed by a comprehensive search for activity towards host glycans. The latter approach will be based on using highly complex salmon derived substrates (e.g. cell cultures, salmon skin mucus, scales, tissue) for activity reactions with the pure enzymes. Product detection will be enabled by the use of a set of complementary (advanced) methods involving glycan array screens, "partner fishing" and stable isotope labelling methods. In parallel to the enzymology approach, we will specifically investigate one of the major salmon pathogens, Aliivibrio salmonicida, by transcriptomic and proteomics methods on bacteria growing on salmon derived substrates. LPMO/chitinase knockout mutants will be evaluated for infectiousness in vivo and time-resolved transcriptomics will be used to determine the spatiotemporal regulation (time and location in the fish) of these novel virulence factors.