Date Available


Year of Publication


Degree Name

Doctor of Philosophy (PhD)

Document Type

Doctoral Dissertation




Chemical and Materials Engineering

First Advisor

Dr. Christina M. Payne


In nature, organisms secrete synergistic enzyme cocktails to deconstruct crystalline polysaccharides, such as cellulose and chitin, to soluble sugars. The cocktails consist of multiple classes of processive and non-processive glycoside hydrolases (GH) that aid in substrate accessibility and reduce product inhibition. Processive GHs attach to chain ends and hydrolyze many glycosidic linkages in sequence to produce disaccharide units before dissociation, and as such, are responsible for the majority of hydrolytic bond cleavages. Accordingly, processive GHs are targets for activity improvements towards efficient and economical biomass conversion. However, the mechanism and factors responsible for processivity are still not understood completely at the molecular level. Specifically, the relationship between processive GH function and the enzyme active site topology and chemical composition has yet to be elucidated. Using molecular simulation and free energy calculations, this work presents a molecular-level understanding of the protein-carbohydrate interactions governing processive GHs, which will facilitate rational design of GHs for enhanced biomass conversion. We hypothesize that processive GHs, having long tunnels or deep active site clefts, will allow more amino acids to interact with the ligand and exhibit strong ligand binding and low substrate dissociation rate constants; whereas non-processive enzymes, having more open tunnels or clefts, will exhibit comparatively weak binding and high dissociation rate constants. Moreover, the ligand binding free energy of a processive enzyme must also be more thermodynamically favorable than the work required to decrystallize a polymer from the substrate matrix. We selected the Serratia marcescens Family 18 chitinase model system, including processive chitinases, ChiA and ChiB, and a non-processive chitinase, ChiC, to test our hypotheses. We find that processive ChiA and ChiB exhibit ligand binding free energies that are more thermodynamically favorable than the work to decrystallize a chito-oligosaccharide from the crystalline chitin surface, which is essential for forward processive movement. The non-processive ChiC binds chito-oligosaccharides with a free energy that is significantly less favorable than the work of decrystallization. In general, our findings suggest that processive GH function necessitates tight binding within the enzyme active site. We also observed that aromatic and polar residues close to the catalytic center of ChiA and ChiB have a greater effect on ligand binding and processivity than the residues at the entrance or exit of the cleft. Mutation of active site aromatic and polar residues generally resulted in reduction in processivity and substantial reduction in substrate binding. Overall, our work demonstrates the existence of a fundamental relationship between ligand binding free energy and processive GH active site characteristics.

Digital Object Identifier (DOI)