|
Public viva-voce Notification Date: Thursday, 4 September 2025 Time: 11.30 AM Venue: Seminar Hall(Hybrid mode) Pareto-optimal Golgi design and decoding cell-surface glycans Aashish Satyajith Chennai Mathematical Institute. 04-09-25 Abstract A forest of tree-shaped sugar chains called glycans coat the surfaces of eukaryotic cells, embedded to proteins in their plasma membrane. They are manufactured step by step in enzyme-catalyzed (stochastic) reactions as they traverse the multiple compartments of the Golgi apparatus. One function of glycans is that they mediate communication between living cells. How information gets encoded by one cell and decoded by other cells remains unclear. This question becomes especially relevant considering that there is significant variability across cells in their surface glycan distribution due to the stochastic nature of glycan manufacture. We answer this question in the specific context of Golgi dysfunction. We perturb the Golgi to various extents and measure the corresponding cell-surface glycans. Then we ask if we can infer the amount of perturbation that was made by measuring cell-surface glycans alone. To do this, we first model the dynamics of cell-surface glycans and fit the model to experimental data. We use the model to construct an optimal Bayesian decoder that predicts Golgi dysfunction given single-cell glycan distribution. We show that having multiple types of glycans on the cell surface helps in faster, more accurate information dissipation. Secondly, of the billions of glycans that can theoretically be manufactured, cells must manufacture only those glycans that are required for their function. The amount of each glycan made is regulated in two ways: the distribution of enzymes across the compartments of the Golgi apparatus, and the amount of time a growing glycan spends in a compartment. We posit that the enzyme distribution and the residence times must be such that the chances of producing useful glycans is maximized. When multiple glycans are to be manufactured, competing manufacturing requirements can create trade-offs. We study the solution space reconciling this trade-off using a Pareto-optimality framework. We show that enzyme splits change discontinuously as residence times are varied, creating an interesting space of growth strategies. In summary, this thesis forms a bridge between the form and function of the Golgi apparatus. All are invited to attend the viva-voce examination
|
