Membrane-bound cytochrome P450 1 A2 system. The cytochrome is shown in transparent surface, with the heme and bound ligand in the active site. The membrane is shown in red.


Tiffany (Yu-Ting) Sun, PRIME 2012 (Malaysia)

This research project focused on computational simulation, i.e., molecular dynamics, of the enzyme cytochrome P450 1A2 in two environments: membrane-bound, and water. One of the main functions of cytochrome P450 1A2 is drug oxidation, i.e. drug digestion. Its large binding (active) site can accommodate various sizes of molecules (substrates) to oxidize different kinds of drugs. The purpose of this project is to determine if, and how, the presence of the membrane has any impact on the stability of the enzyme, including any structural changes in one particular component of the enzyme (i.e., helix structures B, C, F, G) and within binding (active) site.

The two simulations–one, membrane-bound system and another, water system (without the membrane)–were set up using the CHARMM force field, which models the force relationship among the atoms in the systems using two topology files: CHARMM 27 for the protein and CHARMM 36 for the lipid ( in this case a structure component of the membrane). The lipid is set up with a ratio of 2:1 of 1-Palmitoyl-2-oleoylsn-glycero-3-phosphocholine (POPC) and 1-Palmitoyl-2-oleoyl-sn-glycero-3-phosphoethanolamine (POPE) to mimic the environment of the endoplasmic reticulum (ER), a cellular subunit, in human liver, where the 1A2s are mostly found. The massively parallelizable NAMD2 program was used for the molecular dynamics simulations, and 250 nanoseconds (ns) of sampling were gathered for each system.

A number of analyses were performed through the extension tools in VMD (Visual Molecular Dynamics) and other programs, including: root mean square deviation (RMSD), root mean square fluctuation (RMSF), GROMOS clustering, active site analysis, hydration analysis, and a heme angle analysis. The result of the baseline RMSD indicates that the membrane-bound protein with a value of 2.29 angstroms is more stable compared to the water system with a value of 2.41 angstroms. In addition, the RMSF graph helps identify specific residues with significant RMSF differences, which corresponds to regions in the protein that undergo major structural rearrangements in the presence of the membrane.

The selected residues or the corresponding helices are then highlighted as areas of interest for further structural analysis, including the interaction with the substrate, the active site, and the membrane. Structural clustering helps identify the different protein structures sampled in each environment, which can be used subsequently as the input for the active site analysis and hydration analysis. In a hydration analysis we identify and track pathways of water molecules that residue in the active site for more than 2 ns to determine potential water channels in both systems. A major result of the project shows that the size of the active site in the membrane-bound system is significantly smaller with the value of 840 cubic angstroms, whereas the volume in only water, the active site is roughly 2.5 times larger, 2168 cubic angstroms. In addition, in the membrane bound system, a major portion of a helix structure (14) unfolds back to its primary structure. Together, these results will help us understand how the P450 enzymes are able to recognize their many molecular and atomic-level interactions governing the recognition process.

PARTICIPATING RESEARCHERS: USM: Sy Bing Choi, Habibah Wahab; UCSD: Rommie Amaro, Luke Czapla.