C-Glycosides belong to a class of bioactive compounds biosynthesized by C-glycosyltransferases, also known as C-GTs. Despite their practical significance, C-GTs have scarcely been studied due to the limited availability of their crystal structures. In this study, we applied molecular dynamics (MD) simulations and density functional theory (DFT) calculations to investigate Glycyrrhiza glabra C-GT (GgCGT), focusing on the impact of protonation states of two histidine residues and specific mutations on enzyme-substrate configurations. We explored nine native ternary models, considering all possible combinations of protonation states for the His351/His373 pair, which we proposed as fundamental for catalysis. We also included four different mutants designed to assess the role of residues found to be essential for catalytic activity through mutagenesis experiments: His12Ala, His12Lys, His12Lysn and Asp375Ala. MD simulations revealed that only two models (M1 and M3) satisfied the criteria for catalytic competence, where the protonation states of His351 and His373 significantly influenced the relative position between donor and acceptor substrates, as well as the acceptor substrate conformation, adopting extended and packed states. DFT calculations confirmed that these conformations impact the electron density distribution, influencing substrate reactivity. Mutant simulations further supported the experimental data: His12Ala, His12Lys, and Asp375Ala mutants failed to meet the catalytic distance criteria, leading us to infer that these mutations prevented the formation of a reactive enzyme-substrate complex. Conversely, the His12Lysn mutant partially meets the criteria, which could help to explain the catalytic activity of this mutant. These findings provide the first molecular interpretation of the role of key residues in substrate binding and catalysis, which are essential for understanding catalytic activity. Furthermore, they offer new structural insights into residues such as His351/His373, which are often overlooked in GT modeling despite their potential to modulate the Michaelis complex. We hope that these findings will contribute to the rational engineering of more efficient C-GTs for biotechnological applications.