Molecular motors perform diverse functions in cells, ranging from muscle contraction, cell division, DNA/RNA replication, protein degradation, and vesicle transport. The majority of molecular motors use energy from the ATP hydrolysis cycle, converting chemical energy into mechanical work in cells. All ATP-dependent molecular motors have a similar ATP binding site, although the functions can be drastically different. Myosins comprise a large group of ATP-dependent molecule motors. The structure-function relationship governing different functions for different myosin families remains elusive. Hypothesizing that members of each family possess conserved residues for their consensus functions and residues distinctive from those of other families to differentiate their functions from functions of other myosin families, we developed an algorithm for comparative sequence analysis in a phylogenic hierarchy to identify family-specific residues for 38 myosin families/subfamilies that comprise human myosin members.

We found a number of family-specific residues that have been reported, such as residues in β-cardiac myosin associated with hypertrophic cardiomyopathy and residues in myosin 7A associated with hereditary deafness. We also identified distinct features among myosin families that have never been reported, including a unique signature of the SH1 domain in each of the myosin families, residues differentiating α- and β-cardiac myosins, and a unique converter domain of myosin VI. We further examined myosin VI to understand why it moves toward the (-)-end of actin filaments, opposite to the direction of all other myosins and to shed light on their links to prostate cancer and ovarian cancer, where myosin VI is over-expressed. We found that many of myosin VI specific residues locate in or adjacent to the converter domain, including a cluster of unique residues at the interface between the motor domain and the converter. Using molecular dynamics (MD) simulation, we found mutations of M701 on the SH1 helix and F763 on a helix of the converter caused the separation of the motor domain and the converter, indicating their important roles in linking the converter and the motor domain in the pre-power stroke state structure, potentially critical for positioning of lever arm.

Using the location of the unique residues at the interface of the motor domain and the converter as the site of drug docking, we identified a set of candidate small molecules binding to this unique binding site selectively, potentially blocking the converter rotation of myosin VI. A benzoic acid (C15H17N3O3) was found to have the best score in docking, binding to both the converter and motor domain stably in a 200 ns MD simulation run. This molecule can be a good lead to be optimized to inhibit myosin VI functions in cancer patients. We have also applied our algorithm to other ATP-dependent molecular motors, including hepatitis C virus NS3 helicase and DEAD box helicase Mss116. We found an important residue, T324, in NS3 helicase connecting domains 1 and 2 acting as a flexible hinge for opening of the ATP-binding cleft and an atomic interaction cascade from T324 to residues in domains 1 and 2 controls the flexibility of the ATP-binding cleft in NS3 helicase. We also found a conserved flexible linker for Mss116, and the tight interactions between the Mss116-specific flexible linker and the two RecA-like domains are mechanically required to crimp RNA for the unique RNA processes of yeast Mss116.