Active diffusion of particles in a dynamic network

Diffusion is a phenomenon well understood for microscopic particles as arising from random molecular collisions. However, these interactions are typically non-specific and cannot be tuned. In contrast, macromolecular diffusion through networks can be controlled by binding and unbinding events between passaging molecules and flexible chains. Some biophysical examples include the central channel of the nuclear pore complex and liquid drops formed from multivalent interactions. The precise effect of properties such as binding and unbinding rates, number of binding sites and chain elasticity, on diffusion is still poorly understood. Following a statistical mechanics approach, we have developed a diffusion model which shows that the maximum diffusion occurs for few occupied binding sites independent of other parameter choices. We show the validity of our findings by comparing model predictions with a macroscopic diffusion experiment designed to contain similar driving mechanisms and to allow tuning of key parameters including the number binding sites, activity and kinetic parameters. These findings will drive future research work and understanding of controlled active diffusion in dynamic networks.