(680e) A Three-Enzyme Cascade During N-Glycan Branching: the Minimal Module for Ultrasensitivity

Introduction: Human cell-surface and secreted proteins often bear carbohydrate structures or glycans. Among the glycoforms, protein asparagine linked- (N-) glycosylation is an important type of post-translational modification (PTM). This PTM regulates the pharmacokinetics of biotherapeutic proteins, and also cellular adhesion and signaling properties. Here, N-glycan branching is an important step since it regulates the distribution of hybrid and complex N-glycan structures. Wet-lab experiments and computational models developed by Lau et al. [1] reveal that the production of tri- and tetra- antennary N-glycans on cell-surface receptors is ultrasensitive to hexosamine flux. However, the precise components of the reaction network that contribute to ultrasensitivity are yet unidentified.

Materials and Methods: Lau et al. [1] describe an initial value problem for N-linked glycosylation with 144 species and 166 reactions. In the current paper, we have systematically reduced this model to yield a three enzyme minimal module that exhibits ultrasensitivity and other functionality that is very similar to the more complex LD2007 model (Figure 1). Such analysis quantifies the concentrations of tri-antennary N-glycans and also other species as a function of hexosamine flux. The source for sensitivity regulation and the role of different model components were quantified using three parameters that fit the Hill equation: i) the Hill coefficient, h; ii) the half maximal effective concentration, D₅₀ and iii) the maximal response R_m.

Results and Discussion: The minimal module for ultrasensitivity in the synthesis of tri-antennary glycans from high mannose structure is presented. It includes a linear three-enzyme cascade (E1-E2-E3), the inter-Golgi transports of glycans, and product recycling steps providing feedback control. Sensitivity analysis was performed on this reduced model by perturbing individual enzymatic rate constants or transport rates by 5% of their nominal values and determining the effect on h, D₅₀ and R_m with respect to the final tri-antennary product. Such computational analysis reveals that the enzyme E1 (GnTI) regulates ultrasensitivity steepness (h) and branching activation threshold (D₅₀). E3 (GnTIV) regulates maximum response (R_m). Further, a step-wise decrease of binding affinity for substrate and enzymes in the cascade favors higher ultrasensitivity and lower activation threshold. Finally, the minimal model predicts that fine tuning of the characteristics for ultrasensitivity can be performed by varying enzyme concentration or enzyme binding affinity at various enzyme cascade levels. Such perturbations may regulate N-glycan multiplicity on cell-surface glycoproteins, cell growth and differentiation characteristics.

Conclusions:  We identify the minimal components of the N-glycosylation model that impart ultrasensitivity. The fundamental mechanism leading to ultrasensitivity has been studied extensively in systems like the MAPK pathway. These studies demonstrate a significant role for cooperativity, zero-order ultrasensitivity (arising from enzyme cycles operating under saturation regimes) and multisite phosphorylation during MAPK activation [2]. In contrast, our study shows that different mechanisms contribute to ultrasensitivity during N-glycosylation. 

References:

1. Lau, K.S., et al. Cell, 2007. 129(1): p. 123-34.

2. Huang, C.Y., et al. PNAS, 1996. 93(19): p. 10078-10083.