Solvent-quality dependent contact formation dynamics in proteins

Abstract

The mean time of contact formation between two ends of a protein chain shows power law dependence with respect to the number of residues, $\tau_{CF} \sim N^{\alpha}$. Fluorescence quenching measurements based on triplet-triplet energy transfer show variation in the value of scaling exponent $\alpha$ for different protein-solvent systems. Here, starting from a non-Markovian diffusion equation supplemented with an exponential sink term that accounts for the energy transfer reaction between donor and acceptor groups, we calculate the mean time of contact formation using the Wilemski-Fixman closure approximation. The non-Markovian diffusion-reaction equation includes the effects of solvent quality and hydrodynamic interaction in a mean-field fashion. It shows that the contact formation dynamics is mainly governed by two time scales, the reciprocal of the intrinsic rate of quenching $(k_0^{ET})^{-1}$, and the relaxation time $\tau_0 = \eta b^3/k_B T$ of the coarse-grained residue of an effective size $b$ with solvent viscosity $\eta$. In the limit of $k_0^{ET} \tau_0 \ll 1$, the dominating effect of the reaction-controlled kinetics yields the scaling exponents as $0.89$, $1.47$ and $1.79$ in poor, theta and good solvents respectively. In the opposite limit $k_0^{ET} \tau_0 \gg 1$, the dominating influence of the diffusion-controlled kinetics results in $\alpha$ as $1.90$, $2.17$, $2.36$ for a freely-draining and $1.31$, $1.77$, $2.06$ for a non-freely-draining chain in poor, theta and good solvents respectively. In the intermediate limit, $k_0^{ET} \tau_0 \approx 1$, the increase in the number of residues switches the kinetics from reaction-controlled at low $N$ to diffusion-controlled at large $N$. These general results suggest that experimental estimates of the scaling exponents reflect solvent-quality dependence of the mean contact formation time in the reaction-controlled limit.