A Putative Network of Interactions Controls the Opening and Closing Dynamics of the ATPase Domain of E. Coli Hsp70, DNAk

Abstract

Hsp70 chaperones play important roles in cells including protein folding, trafficking, degradation and enabling survival under stress conditions. DnaK is an E. coli Hsp70 homolog comprising an ATPase domain and a substrate-binding domain. DnaK has two substrate-affinity states: ATP binding lowers the affinity of the substrate, whereas its hydrolysis leads to a higher affinity of substrate for binding. ATP-dependent communication between the two domains is essential for chaperone function and mediated via a conserved hydrophobic linker (384GDVKDVLLL392). Previous studies showed that when the linker interacts with the ATPase domain, which was studied by the construct containing the entire linker, DnaK(1-392), an enhanced ATPase rate is observed compared to the construct lacking the conserved 389VLLL392 linker region, DnaK(1-388). This observation suggests that structural rearrangements caused by linker docking adopt the ATPase domain in a closed conformation, leading to an enhanced, pH-dependent ATPase activity. Here, our aim is to delineate the residues that are responsible for the linker induced conformational rearrangements. In that line, using molecular dynamic simulations we identified a putative network of interactions through Arg71-Glu81-Asp85-Thr225-His226 at the lobe interface of the ATPase domain that might be critical in stabilization of the domain in the so called “open” and “closed” conformational equilibrium. We made point mutations for these sites on the two ATPase domain constructs, and studied the structural and functional effects of these residues on the ATPase domain as a function of pH using various biophysical and biochemical methods. Mutations' effects studied by equilibrium thermodynamic measurements showed variations for the constructs, but dramatic changes were observed in the dynamics of the constructs. Our results suggest the linker as a controller for ATPase domain dynamics changes partly through the identified network.