Characterization and engineering of a polysaccharide mono-oxygenase involved in the degradation of (ligno) cellulosic biomass (178BW0913)
Lignocellulosic biomass is believed to be the most promising renewable feedstock for sustainable production of biofuels and commodity chemicals in biorefineries. While the need for sustainable systems is continuously increasing, the depolymerization of lignocellulose is impeded by its high recalcitrance. This makes the development of a cost-effective process extremely challenging. In enzymatic degradation, the classical view of synergistic endo- and exocellulases working together has recently been revolutionized by the discovery of the lytic polysaccharide monooxygenases (LPMOs). These metalloproteins oxidize the glycosidic linkage, thereby rendering the crystalline substrate more amorphous and thus more amenable for traditional cellulases. In that way, LPMOs boost the efficiency of the degradation process. It is clear that LPMOs have a huge potential for cost reduction in biomass-converting industry and the knowledge on these enzymes is quickly expanding. Nonetheless, many aspects of this interesting group of enzymes still have to be elucidated. One of the concerns with LPMOs is protein stability, which is an industrially important parameter. Processes at higher temperatures indeed offer advantages such as a reduced risk of contamination and an increased reaction rate. Therefore, this PhD thesis aims to study and increase the stability of 2 LPMOs via various protein engineering techniques. Since only two classes of LPMOs show activity on (hemi)cellulose, only these are important for the biomass industry and they are therefore the focus of this PhD thesis. More specifically, one member of auxiliary activity family 9 (AA9) and one from AA10 were selected to study. For family AA9, on the one hand, an LPMO of the well known cellulase-producing fungus Trichoderma reesei was examined, namely TrCel61A. For family AA10 on the other hand, the first described cellulose-active bacterial LPMO was studied, i.e. LPMO10C (formerly CelS2) from the soil bacterium Streptomyces coelicolor. The first enzyme, TrCel61A, was heterologously expressed in the eukaryotic host Pichia pastoris to enable post-translational modifications. A peculiar concern in expressing an LPMO is the special characteristic that LPMOs require a histidine residue at their N-terminus (His-1) in order to remain active. To that end, different secretion signals were compared. The protein’ native secretion signal outperformed the widely used -mating factor of S. cerevisiae among others. The highest LPMO yield in P. pastoris ( > 400 mg/ L during fermentation) was described and a 100 % correct processing was guaranteed. Next, the activity and stability were measured of the previously expressed protein. Phosphoric acid swollen cellulose (PASC) served as substrate in the activity test, while the analysis of the reaction products was performed by high performance anion-exchange chromatography (HPAEC, specialized HPLC). For the first time LPMO activity of TrCel61A on PASC was demonstrated and furthermore, the protein was found to be a type-3 LPMO, generating neutral, C1- and C4-oxidized products. Furthermore, the enzyme’ thermostability was measured via differential scanning fluorimetry (DSF), which yields a thermal denaturation curve. The resulting parameter that can be compared between enzymes is the apparent melting temperature (Tm), which matches the inflection point. The method was found to be sensitive and reproducible with a low technical and biological variability. Moreover, for TrCel61A, an apparent melting temperature of 62 1 °C was measured. Finally, some protein engineering strategies were applied to TrCel61A. First, the catalytic domain (CD) only of the protein was expressed, but did not show a higher stability. Therefore, the full-length protein (including linker and carbohydrate binding module) was taken as starting point for engineering. Second, even though N-glycosylation and native disulfide bridges were found to contributed to the wild-type protein stability, a reverse strategy of adding extra positions for such post-translational modifications did not yield more stable variants. Finally, the principle of consensus engineering was applied after building a phylogenetic tree. However, these de novo designed enzymes showed a significantly decreased thermodynamic stability. The second enzyme, ScLPMO10C, was heterologously expressed in the prokaryotic host Escherichia coli. Taking the His-1 requirement into account, the protein was expressed in the periplasmic space by preceding the coding sequence by its native secretion signal. ScLPMO10C was also active on PASC as a type-1 LPMO, producing only neutral and C1- oxidized products and had an apparent melting temperature of 51 1 °C. Despite this rather moderate melting temperature, an unusual 34 % residual activity was measured after 2 hours incubation at 80 °C, which could be attributed to the native disulfide bridges. Finally, ScLPMO was also subjected to protein engineering strategies in view of increasing its stability. Even though some rational mutations at the protein’ surface did not have the desired effect, disulfide engineering yielded different variants with improved apparent melting temperature, ranging from +2 to +9 °C. By combining the positive disulfide introductions, the best variant obtained, displayed a +12 °C increase in Tm and was able to retain no less than 60 % of its activity after heat treatment (compared to only 34 % for the wild-type). This improvement brings the enzyme’ apparent melting point to an industrially relevant temperature. In conclusion, even though not all engineering strategies yielded more stable LPMO variants, some interesting stability characteristics were discovered for LPMOs. Furthermore, disulfide bridges are thought to be essential elements in LPMO stability.