Coarse-grained models of biomolecule dynamics and allostery

Recently, it has become increasingly accepted that thermal fluctuations take active part in functional tasks of biological molecules. We employ a set of coarse-grained models to investigate the mechanism of transmission of allosteric signal via spatial fluctuations. Our models are coarser than those in computational techniques established in molecular biology, but allow for both the identification of candidates for the essential physical structures and also the analytical determination of thermodynamic quantities that define ligand binding. The models are constructed for general classes of macromolecules and
are validated through parameterisation from experiments and atomistic simulations.

In the first part of this thesis we investigate the “dynamic allostery” in dimeric proteins
composed of two identical subunits. We demonstrate that cooperative effects upon binding of two identical ligands can arise purely through modification of slow global
vibrational modes of the protein. We parameterise the model on a test case, the CAP homodimer. Finally, we explain the role of local, fast vibrations in the allosteric effect
and propose a general protocol for interpreting thermodynamic parameters of dynamically allosteric homodimers.

The second part of this thesis considers allosteric effects in DNA, an example of nearly uniform elastic medium. The DNA is modeled as an elastic rod and substrate binding
as local increase of its bending and twisting rigidity. This results in altered structure of normal modes and leads to qualitatively different type of dynamic allostery compared to that of the discrete models previously employed to study allosteric effects in proteins. Dynamic allostery in DNA is found always to be negative, due to an anti-correlated amplitude of thermal fluctuations at the binding site and around it. This allows us to draw conclusions about general design rules of allosteric molecules and highlight the controlling feature that biological molecules evolved to optimize their dynamics for their function.