Removal of N-linked glycans in cellobiohydrolase Cel7A from Trichoderma reesei reveals higher activity and binding affinity on crystalline cellulose

Cellobiohydrolase from glycoside hydrolase family 7 is a major component of commercial enzymatic mixtures for lignocellulosic biomass degradation. For many years, Trichoderma reesei Cel7A (TrCel7A) has served as a model to understand structure-function relationships of processive cellobiohydrolases. The architecture of TrCel7A includes an N-glycosylated catalytic domain, which is connected to a carbohydrate-binding module through a flexible, O-glycosylated linker. Depending on the fungal expression host, glycosylation can vary not only in glycoforms, but also in site occupancy, leading to a complex pattern of glycans, which can affect the enzyme's stability and kinetics.

The kinetic changes of TrCel7A imposed by removal of N-linked glycans reflected modifications of substrate accessibility. The presence of N-glycans with extended structures increased K M and decreased attack site density of TrCel7A likely due to steric hindrance effect and distance between the enzyme and the cellulose surface, preventing the enzyme from achieving optimal conformation. This knowledge could be applied to modify enzyme glycosylation to engineer enzyme with higher activity on the insoluble substrates.

Two expression hosts, Aspergillus oryzae and Trichoderma reesei, were utilized to successfully express wild-types TrCel7A (WT Ao and WT Tr ) and the triple N-glycosylation site deficient mutants TrCel7A N45Q, N270Q, N384Q (ΔN-glyc Ao and ΔN-glyc Tr ). Also, we expressed single N-glycosylation site deficient mutants TrCel7A (N45Q Ao , N270Q Ao , N384Q Ao ). The TrCel7A enzymes were studied by steady-state kinetics under both substrate- and enzyme-saturating conditions using different cellulosic substrates. The Michaelis constant (K M ) was consistently found to be lowered for the variants with reduced N-glycosylation content, and for the triple deficient mutants, it was less than half of the WTs' value on some substrates. The ability of the enzyme to combine productively with sites on the cellulose surface followed a similar pattern on all tested substrates. Thus, site density (number of sites per gram cellulose) was 30-60% higher for the single deficient variants compared to the WT, and about twofold larger for the triple deficient enzyme. Molecular dynamic simulation of the N-glycan mutants TrCel7A revealed higher number of contacts between CD and cellulose crystal upon removal of glycans at position N45 and N384. Conclusions: The kinetic changes of TrCel7A imposed by removal of N-linked glycans reflected modifications of substrate accessibility. The presence of N-glycans with extended structures increased K M and decreased attack site density of TrCel7A likely due to steric hindrance effect and distance between the enzyme and the cellulose surface, preventing the enzyme from achieving optimal conformation. This knowledge could be applied to modify enzyme glycosylation to engineer enzyme with higher activity on the insoluble substrates.