Comparison of Microtubule Dynamics for A- and B-Lattice Geometries

Abstract

Accurate quantitative interpretation of experimental data and prediction of the effects of microtubule-targeted anti-mitotic drugs require a detailed model of the events that occur at microtubule ends. Before searching the large parameter space of a model with few constraints on lattice symmetry, binding site configuration, GTP-hydrolysis rate, and oligomerisation state of the associating and dissociating species, we performed an extensive, systematic investigation into the dynamics of a series of simplified models with significantly smaller parameter spaces. The models had regular A or B-lattice geometries, tightly coupled GTP-hydrolysis, and association-dissociation events involving the formation or breakage of just two lateral bonds. GTP-hydrolysis weakened the two lateral bonds to the β-tubulin subunit by 4.6 kBT, in either a balanced (+2.3 kBT each) or an unbalanced way (+4.6 kBT for one and 0 for the other bond). Association rate constants were 1 μM-1s-1, and dissociation rates were thus dependent on the lateral bond energies. We observed the following:1. Values for CC (the concentration of free tubulin-GTP at which the net growth is zero) varied from 1.2 to 80 μM2. All configurations showed discernable phases of growth (G) and shrinkage (S) around their CC3. Effective growth rates at CC (average growth rate during the G-phase divided by the maximum attainable growth rate at that CC) varied from less than 0.1 in most of the B-lattice geometries to 0.9 in the A-lattices with a balanced effect of hydrolysis4. G-phase lifetimes were relatively short (10-15 s), and growth was significantly more uniform in the balanced A-lattice geometries, compared with those in the unbalanced geometries (lifetimes > 100 s)Thus, balanced A-lattice configurations support efficient growth on relatively unstable microtubule ends, whereas most other configurations grow less efficiently on more stable ends.