News & Events

David Space

Posted on October 26, 2016

When

Date - October 26, 2016
1:00 pm


What

Ph.D. Thesis Defense – 752

Title: “Construction of a Homology Model of the Orexin-1 Receptor in its Active and Inactive States and Subsequent Molecular Dynamics Experiments in a Simulated Lipid Bilayer”

Abstract:

The purpose of Phase I of this research was to construct two homology models, one each inactive (R state) and active (R* state), of the orexin-1 receptor (ox1r).  The ox1r is a Class A (rhodopsin-like) G protein-coupled receptor (GPCR) that couples to the Gq protein.  Homology models were built using an existing crystal structure of a Class A GPCR.  The ligands 1-(5-(2-Fluoro-phenyl)-2-methylthiazol-4-yl)-1-((S)-2-(5-phenyl-(1,3,4)oxadiazol-2-ylmethyl)-pyrrolidin-1-yl)-methanone (SB-674042), and the orexin-A peptide [pyroglutamate]PLPDCCRQKTCSCRLYELLHGAGNHAAGILTL-NH2 with residues C6-C12 and C7-C14 disulfide-bonded, were docked in the ox1r R and R* states, respectively.

Another orexin receptor (ox2r) subtype’s x-ray crystal structure was published (Yin et al., 2014).  The high level of identity between the ox1r and ox2r, and the release of peptide-bound neurotensin crystal structures (Egloff et al., 2014, Krumm et al., 2015), led to the latter being used as a guide to dock orexin-A’s C-terminus into the binding pocket.  The ligands were docked in ox1r models created by directly mutating ox2r’s crystal structure, only using Conformational Memories (CM) (Whitnell et al., 2008) to select an R* TMH6 that kicks out.  The new ox1r R* model has orexin-A hold the ox1r’s toggle switch residue, Y6.48, in a trans 1 and has part of orexin-A accept a hydrogen bond from the ox1r R* state’s H7.39 as per Darker et al., 2001 and Malherbe et al., 2010.  These interactions promote the ox1r R* state.  The new ox1r R model has SB-674042 hold the ox1r’s Y6.48 in a g+ 1 and accepts a hydrogen bond from the R state’s Q3.32, as per Malherbe et al., 2010.  These interactions promote the ox1r R state.  New IC3 loops and termini were created with Modeller (Sali et al., 1993, University of California San Francisco, San Francisco, CA), and the receptor models were palmitoylated.

The purpose of Phase II was to examine the homology models’ behavior in a simulated lipid bilayer environment.  To do so, CHARMM force field parameters were created based on SB-674042 and on pyroglutamate using the global minimum structure of each, all calculated at the HF 6-31G* level of theory using Spartan (Wavefunction Inc., Irvine, CA).  The result was a list of new parameters that were used in the Molecular Dynamics (MD) simulation, and in turn, the results of the lipid bilayer simulation of each ligand/receptor complex indicated that each complex was stable and retained important GPCR features.  Furthermore, each complex was compared with recently released ox1r crystal structures (Yin et al., 2016), and the transmembrane helices match almost perfectly, with an RMSD of 0.5383 for TMH1, 0.2374 for TMH2, 0.3416 for TMH3, 0.3731 for TMH4, 0.3839 for TMH5, 0.2851 for TMH6, 0.3074 for TMH7, and 0.5060 for Helix 8 between the Cαs of the ox1r R model and the ox1r crystal structure 4zjc.  The residues in the ox1r R* model’s EC2 loop that bind to orexin-A’s residues L16, L19, and L20, after running MD on it, match extremely well with the ox1r crystal structures’ corresponding EC2 residues, and face the same way as well.  SB-674042 in one of those crystal structures accepts no hydrogen bonds, as opposed to the complex.  In fact, orexin-B was docked in the ox1r, and the binding interaction energy thereof was calculated as it was for the ox1r ligands, and the results were consistent with experimental data.