Clifford Bryant SB 228 Feb 14th 2001
Here is Assignment #4.
Proteins DO6 and DQK were downloaded from PDB after morphing for further study.
(from the assignment...)
5.Make a webpage that showcases this protein motion (you can use images provided by the server, but it would be even better if you create your own since the provided ones are rather bland)
REFERENCE: Biochemistry Vol. 39, No. 10, 2000 2499-2508 This is a unique morph because it involves a pronounced movement of one portion, a loop near the C-Terminal Most of the secondary structure is preserved. There is slight perturbation at the ends of beta sheets at another site adjacent in space. The loop (center of upper picture) containing the glutamate (Glu 14) in the middle picture (green) is seen to shift positions in a representative single subunit of this multisubunit protein (homo tetramer, reduced and homo dimer, oxidized). The portions of the chain involved in subunit interaction do not move much. Although the differences are hard to discern on the ribbon model, ball and stick shows difference at glutamate (final picture).
6.Add a small description of why this motion is BIOLOGICALLY interesting (one paragraph) and any information that you think the morph, or any data generated on the morph webpage may have revealed.
The Morph is biologically interesting because it relates directly to the catalytic mechanism of this nonheme iron enzyme. In the reduced state, (1DQK) we have Fe(II) which by the rule of 18 d-electrons is tetrahedral, chelated by His 16, 41, 47, 114 and Cys 111. In the oxidized state we have Fe(III) and the ligands are now joined by Glu 14 in an octahedral scheme. The movement shown in the morph presumably allows this. Torsion angle analysis included with the morph shows marked shifts in the dCA parameter from residues 9-16 only, indicating that sidechains moved around specifically in the catalytic region.
7.Add answers to the following questions to your webpage:
1.Do you think the trajectory suggested by the morph server is accurate/actually occurs in reality? Why or why not?
It has been said some models are useful, but no model is 100.00% accurate. Our morph model nicely allows for catalysis, but it stops far from explaining the mechanism. Indeed, the predicted movement may not happen. What we have is a theoretical continuous path between conformations. There is no way to confirm that it stays on this path until we develop technology to visualize things on the nanosecond time scale. In this case, there doesn't seem to be anything that might perturb the motion such as remote interference from other regions/subunits of the protein.
2.Give one way you think we could speed up molecular dynamics simulations. Explain. ("faster computers" or "parallel code" is not good enough. Hardware based answers need to be heavily justified!)
(Adapted from a conversation with Swami Swaminathan of Gilead Sciences 2/14/01). The thing we want to do here is boost up the step time in our simulation (50 pS/step sounds right?), such that we can move forward with larger jumps and ultimately run for longer times with confidence in our results. To do this, energy balance must be maintained by stopping accumulation/ compounding of residual terms in our mathematical model. Ways to do this: simplifications: i) Base your energy function on torsion angles. Develop a set of rules based on energies of known conformations of molecules (say, peptides) that produce certain relationships between substituents. When one of these patterns is spotted, apply the rules and iterate. ii) Neglect things: Hydrogens are small with respect to the other atoms. Subtrating them off reduces complexity and calculation error.