Probing the structural effects on directionality and processivity of myosin VI using artificial lever arms
Jung-Chi Liao, Zev Bryant, Mary W Elting, Scott L Delp, James A Spudich
Myosin uses ATP chemical energy to perform diverse functions such as muscle contractions, cell division, and vesicle transports. My interests lie in integrating experimental and computational techniques together to understand its mechanics and energy transduction. I have designed myosin VI with artificial lever arms to identify key structural elements of myosin VI dictating its reverse directionality. I implemented computational modeling and molecular dynamics simulation to guide the protein design. I then used in vitro motility experiments, optical tweezers, and total internal reflection fluorescence microscopy to test these designs. Our results demonstrated that the calmodulin-bound portion is not an integral mechanical component of the reverse stroke. Furthermore, we were able to alter the moving direction of myosin on actin filaments with the change of only 18 amino acids. The following movies from in vitro motility experiments show the actin filaments moved by designed myosins toward the opposite directions. Field size: 12.8 μm x 6.4 μm
Two designed myosins moving toward the opposite directions |
||
(-) end directed motor (moving toward the Cy5 red label) |
(+) end directed motor (moving away from the Cy5 red label) |
|
 |
 |
|
 |
 |
|
Computer-aided protein engineering
Model structures of designed myosins were first predicted computationally. We have examined constructs fused at 5 different locations, i.e. residue numbers 789, 790, 791, 792, and 793 of myosin VI. Model structures predicted 789 and 793 to have structural collisions, so we were unable to run simulation for these constructs. I ran molecular dynamics simulation for constructs 790, 791, and 792. The construct 792 showed the artificial lever arm denatures a key alpha-helix in the myosin converter domain, while 790 and 791 were structurally stable for at least 2 ns. Construct 790 appeared to have the least structure hindrance if myosin undergoes conformational changes. In vitro motility experiments were conducted to test all these constructs. Experiments showed indeed that constructs 789, 792, and 793 could not move the actin filaments, while 790 and 791 were active (see movies below).
Experimental validations of computational predictions |
||||
Construct |
In vitro motility |
Predicted structure / Molecular Dynamics |
Predicted Velocity |
Measured Velocity |
789 |
 |
Severe structural collision in the model |
0 |
0 |
790 |
 |
 |
37 nm/s |
40+-8 nm/s |
791 |
 |
 |
37 nm/s |
17+-4 nm/s |
792 |
 |
 |
0 |
0 |
793 |
 |
Severe structural collision in the model |
0 |
0 |