Skip to main content »

Trinity College Dublin

MOLDY: Biomolecular Component

Abstract

The aim of the current study is to explore the conformational flexibility of α1-adrenoceptors in various membrane environments. As membrane bound proteins, adrenoceptors are a key component of signal transduction and have been linked closely to vasoconstriction and the control of blood pressure. Recent studies have suggested that, in cases of membrane-bound proteins, function is affected by the local membrane environment. Such affects are induced not simply by direct lipid-protein interactions but also by collective bilayer properties such as lateral pressure profiles, surface tensions and bilayer curvature. Variability in lateral pressure profiles has been linked to the degree of lipid saturation[1] and it is this correlation we hope to explore. Through molecular dynamic (MD) simulation we aim to characterize the dynamics of α1-adrenoceptors embedded in different membrane environments. This information in conjunction with more rigorous ligand binding free energy calculations should allow for a correlation to be drawn between protein activity and membrane composition.

Experimental Outline

  • Choose an accurate membrane force field model, based on the simulations of pure membranes.
  • Obtain a reasonable model of α1-adrenoceptors.
  • Embed the models in the validated membrane environments and proceed with MD simulation on the order of 10ns. Analyse the resulting protein dynamics.

The Membrane Environment

Three membrane environments were chosen based on variability of their published lateral pressure profiles.[1] The phospholipid components are illustrated in Figure 1.

Lipids

Figure 1: Structures of the three lipid bilayers to be studied. 1,2-dipalmitoyl-sn-glycero-3-phosphatidylcholine (DPPC), 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphatidylcholine (POPC) and 1-palmitoyl-2-docosahexaenoyl-sn-glycero-3-phosphatidylcholine (PDPC). The sn-1 chain remains constant in all three cases.

Pre-equilibrated membrane coordinates were acquired, with thanks, from the Vattulainen group at the Tampere University of Technology, Finland. As our interest lay in all-atom simulation we chose to use the CHARMM27 force field to model the resulting lipid systems. However, recent studies have suggested that the CHARMM27 charge set does not accurately reproduce experimental membrane behavior. Thus, a modified charge set has been proposed by Sonne et al. and validated using DPPC[3]. We are currently simulating pure membrane systems with both the original and the modified charge set using NAMD2.6[2]. Snapshots from the first 5ns of simulation for PDPC with both charge sets are shown in Figure 2.

PDPC Snapshots

Figure 2: Snapshots from the first 5ns of simulation for a PDPC bilayer with either the CHARMM27 charge set (top) or the modified charge set (bottom) proposed by Sonne et al. Nitrogen and phosphorus atoms from the lipid head group are shown in blue and brown respectively.

During the first 5ns of simulation we see that the PDPC bilayer modeled with the modified charges appears to compress to a lesser extent. This observation is supported by measuring the bilayer thickness as shown in Figure 3. At this point in the simulation, the modified charges result in a bilayer thickness closer to the experimental observed value and thus more accurately represent the membrane environment.

PDPC P-to-P Width

Figure 3: PDPC bilayer thickness measured as the average distance between the phosphorus atoms in the upper and lower leaflet.

Once the simulations have completed (run to 10 ns) the membranes will be fully analyzed with a view to the following characteristics:

  • Area per lipid
  • Bilayer thickness
  • Number densities
  • Lateral pressure profile
  • Deuterium order parameters

A set of analysis tools will be written in C++ to facilitate analysis with the intent of making such tools available to the greater scientific community. The above metrics when compared to experimental data should allow for the quality of the force fields to be assessed and an informed choice of charge set to be made.

Adrenoceptor Models and Future Work

Lacking an experimental xray crystal structure, work previously completed in the Watson research group has focused on the construction and validation of homology models for all three α1-adrenoceptors subtypes. Several models were built for each subtype and validated though MD simulations in a chloroform membrane mimic as show in Figure 4 (left). The starting structures for this work have been taken from these models. Using the results of the aforementioned membrane validation work, the homology models will be inserted into the membrane environments, as shown in Figure 4 (right).

A1A in ChCl3 A1A in ChCl3 (2)

Figure 4: α1-adrenoceptor. (left) Embedded in a chloroform/water membrane mimic, snaps approximately 3.1 ns. (right) Embedded in a PDPC bilayer, snapshot at approximately 0.5 ns.

Approximately 10 ns of equilibrated simulation will be performed in each membrane environment and the resulting protein dynamics characterized using the following methods:

  • RMSD analysis
  • Trans-membrane bundle metrics
  • Helix compression and rotation metrics
  • Principle component analysis
  • Secondary structure analysis
  • Residue contacts

This preliminary work should lend insight as to the impact of variable membrane lateral pressure profiles on alpha1-adrenoceptor dynamics.

References

[1] Ollila, S.; Hyvönen, M.T.; Vattulainen, I. J. Phys.Chem. B 2007, 111, 3139-3150

[2] NAMD website http://www.ks.uiuc.edu/Research/namd/

[3] Sonne, J.; Jensen, M.Ø; Hansen, F.Y.; Hemmingsen, L.; Peters, G.H. Biophys. J. 2007, 92, 4157-4167.


Last updated 01 Jul 2010Contact TCHPC: info | support.