Hierarchical model screening on enzymatic hydrolysis of microcrystalline cellulose

Though a large number of simple to complex models are theorized in the kinetic description of cellulose hydrolysis, the reason for their selection in specific studies remains obscure. Considering different combinations of substrate biodegradability-accessibility, hydrolysis step, and the fate of enzyme (adsorption/deactivation) in a hierarchical approach, 36 dynamic model structures of different complexity were tested in the present study for microcrystalline cellulose hydrolysis at different substrate/enzyme loadings. The quality of the candidate models was assessed based on R², adjusted R², Aike information criteria (AIC), and statistical analysis. The results suggest that consideration of substrate accessibility can influence model fitness only when the substrate is biodegradable. However, the selection of hydrolysis kinetics and enzyme fate is remained as significant. A simple 4-parameter model assuming complete biodegradable-accessible substrate, homogeneous Michaelis–Menten (MM), and enzyme deactivation is selected (R² = 0.946, adj. R² = 0.920), even though a few complex models offered marginally better R². Further, dynamic relative sensitivity analysis revealed that reducing sugar profile was sensitive to the entire set of parameters. Though the complex models reveal the cellulose hydrolysis dynamics on a greater extent, such an attempt should be performed with caution unless experimental data on other process variables (apart from reducing sugar) are made available in regression. Estimation of parameters relevant to enzyme inactivation or biomass adsorption from independent batch experiments can be invaluable in deciphering complex hydrolysis kinetics.