Characterization of fungal endo-β(1→4)-mannanases for biomass conversion

Abstract

Biomass from wood is an abundant renewable resource and is rapidly gaining interest as a raw material that can be converted into fuels and other value-added chemicals. Historically, research has mainly focused on cellulose, but interest in utilizing all components of the wood, including the lignin and the hemicellulose, has increased in recent years. In softwoods the dominant hemicellulose is O-acetylated galactoglucomannans, which constitutes up to 25 % by weight of the wood dry matter. Endo-β(1→4)-mannanases (endomannanases) catalyze the hydrolysis of βmannans including galactoglucomannans. However, galactose substitutions on the mannan backbone have been shown to compromise enzymatic degradation of mannans. In addition, the recalcitrant nature of the wood biomass also seems to challenge the enzyme catalyzed degradation. To overcome these challenges and enhance the industrial softwood saccharification, it is important to understand the molecular background of the natural specificity and substrate interactions of endomannanases. This thesis aimed to improve knowledge on the natural specificity of fungal GH5 and GH26 endomannanases. Another aim was to evaluate performance of fungal endomannanases on softwood saccharification and to examine whether any differences in performance might correlate with specific enzyme characteristics. The influence of galactose substitutions in hydrolysis of mannans catalyzed by enzymes was examined using five fungal endomannanases from GH family 5 and 26. The initial hydrolysis rate and the degree of conversion of two galactomannans, locust bean gum and the more heavily substituted guar gum, were determined for all five enzymes. Product profiles were analyzed using the DNA sequencer-Assisted Saccharide analysis in High throughput (DASH) method. In addition, the accommodation of galactopyranosyl residues in the enzymes active site clefts was analyzed by docking analyses using previously determined crystal structures and new homology models. Based on these investigations, it was shown that the fungal GH26 endomannanases, including the novel AnidMan26 from Aspergillus nidulans, had a very open active site cleft and accommodated galactopyranosyl residues in at least the -2, -1 and +1 subsites. This novel structural feature enabled full conversion of guar gum galactomannan and resulted in the production of multiple galactomanno-oligosaccharides from guar gum hydrolysis. In contrast, the fungal GH5 endomannanases were more restricted by the galactose substitutions, as indicated by a lower conversion of guar gum and in general from a more narrow active site cleft. To investigate if the capability to accommodate multiple galactopyranosyl moieties in the active site cleft is beneficial in softwood saccharification, ten fungal endomannanases of GH5 and GH26 were assessed on a variety of pure mannans as well as in enzymatic cellulose saccharification of pretreated softwood. The results obtained emphasize that the saccharification performance of endomannanases varies significantly, but cannot be predicted by initial rates on pure mannans or the ability to accommodate galactose substitutions. Rather, the enzymes ability to act on the mannan in complex with the lignocellulosic matrix that determined its performance. The best performing endomannanase in softwood saccharification was TresMan5A from Trichoderma reesei. The catalytic efficiency of the core module and the presence of the CBM1 both played important roles in the superior performance. Substrate binding amino acids in the fungal GH26 endomannanses were identified by solving and analyzing the crystal structure of a novel GH26 endomannanase from Yunnania penicillata in complex with α-62-61-di-galactosyl-mannotriose (MGG). The capability of accommodating multiple galactopyranosyl side-groups in the binding cleft was found to be conserved among the eight fungal GH26 endomannanases examined, as seen from the identified substrate binding amino acids which were highly conserved among these enzymes. Furthermore, all eight GH26 endomannanases reached full conversion of guar gum. YpenMan26A mutated variants were designed based on the few variations found in the substrate binding amino acids among the studied GH26 endomannanases. A novel mass spectrometry-based method was used to determine kcat/KM of the YpenMan26A wild type and the D37T mutant on α-64-63-di-galactosylmannopentaose (MGGMM). The results showed that this single amino acid substitution in the 2 subsite of YpenMan26A compromised interactions with the galactose side group in this subsite. A more pronounced structural difference between the GH26 endomannanases was found in the area of the core module approaching the CBM35. The enzymes carrying a CBM35 all seem to have an α-helix, which allows ordered interactions with the binding domain, whereas the enzymes without a CBM had a flexible surface loop. The results emphasize that great diversity exists in the specificity of fungal endomannanases and that this specificity affects endomannanase performance in mannan biomass conversion. The data also supported the assumption that structural features play significant roles in substrate recognition, i.e. galactose-accommodation, of endomannanases. The results have potential to affect future selection of endomannanases for industrial applications as well as the design of future screening campaigns.