Modeling the Structure of SARS 3A Transmembrane Protein Using a Minimum Unfavorable Contact

SUPPLEMENTARY INFORMATION

Modeling the structure of SARS 3a transmembrane protein using a minimum unfavorable contact approach

S RAMAKRISHNA, SILADITYA PADHI and U DEVA PRIYAKUMAR*

Center for Computational Natural Sciences and Bioinformatics, International Institute of Information Technology, Hyderabad 500 032, India

e-mail:

*For correspondence

Table of Contents

Supplementary text: Computational details

Implicit membrane simulations

The three transmembrane (TM) α-helices of SARS 3a, TM1 (LQAS40 LPFGWLVIGV50 AFLAVFQSA), TM2 (GFQFI80 CNLLLLFVTI90 YSHLLLVA), and TM3 (EAQFLYLYA110 LIYFLQCINA120 CRII) were modeled as ideal α-helix structures using the Sybyl 7.2 program.1 To these three TM domains, three additional residues in the C- and N-terminal ends were included for modeling the structures. Hence the number of residues in each of the TM domains is 23 + 3 + 3 = 29. These ideal helix geometries were used as the starting configurations for the implicit membrane molecular dynamics (MD) simulations. The MD simulations in the implicit membrane environment were carried out using the CHARMM program applying the all-atom CHARMM22 protein force field along with the CMAP corrections.2-4 The generalized Born model with a switching function (GBSW) proposed by Brooks and coworkers was used for the MD simulations.5,6 A surface tension coefficient of 0.03 kcal/mol/Å2 was used. The membrane smoothing length, which defines the gradual change of hydrophobic region to the solvent region, was taken as 0.4 Å. Four different thicknesses (33, 35, 36 and 37 Å) were chosen to model the hydrophobic region of the membrane. The normal of the membrane was aligned along the Z-axis and the hydrophobic core was centered at the origin. Each of the three TM domains was placed at the center of the membrane and the helical axis was aligned along the membrane normal for all the four different thicknesses used, which resulted in 12 independent systems. These systems were then subjected to a 500-step minimization using adopted basis Newton-Raphson (ABNR) method followed by production simulations at 300 K for 50 ns with an integration time step of 1 fs. The coordinates during the simulation were saved every 500 steps, and were used for analysis.

Explicit membrane simulations

The NAMD program was used to perform the MD simulations of the TM domains in explicit lipid bilayer environment.7 The CHARMM22 all-atom protein force field including the CMAP corrections, the CHARMM36 all-atom lipid force field, and the TIP3P water model were used for all these simulations.3,4,8,9 The CHARMM GUI membrane builder was used for the initial setup of the lipid bilayer simulations.10,11 For this purpose, four explicit membranes were considered: DLPE (2,3dilauroyl-D-glycero-1-phosphatidylethanolamine), DMPC (2,3dimyristoyl-D-glycero-1-phosphatidylcholine), POPC(3-palmitoyl-2-oleoyl-D-glycero-1-phosphatidylcholine), and DPPC (2,3 dipalmitoyl-D-glycero-1-phosphatidylcholine). The final configurations of the three TM domains obtained from the implicit membrane simulations with four varying thicknesses were taken as the initial structures for the explicit membrane simulations. The three TM domains in four different explicit membranes with varying thicknesses resulted in 12 more systems which were equilibrated in six steps as given in the Table S1. In short, during these different stages of the equilibrium simulations, the positional constraints on the backbone atoms, side chain atoms, and restraints on the dihedral angles of the lipid tails were relaxed, starting from 10, 5 and 500 kcal/mol/Å2, respectively, and were gradually reduced to 0. The planar and the dihedral harmonic restraints mentioned were defined using the COLVAR module in the NAMD program. Prior to the MD simulations, the systems were subjected to 10000-step conjugate gradient minimization with the constraints mentioned above. During all equilibration simulations, a nonbonded cutoff of 12 Å was employed, and a switching function starting at 10 Å was used. Nonbonded pair lists were generated based on a distance cutoff of 16 Å, with the pair list for calculation of nonbonded interactions being updated after every 10 steps. All the covalent bonds involving hydrogen atoms were restrained using the SHAKE algorithm.12 Periodic boundary conditions were used, and a flexible cell with a constant x:y ratio for the x-y plane was used. Langevin dynamics with a damping constant of 1.0 ps-1 was used. Langevin piston target was set to 1.0 atm and Langevin piston period was set to 50 fs, with the Langevin piston decay set to 25 fs. Particle mesh Ewald summation method was employed for electrostatic calculations with an interpolation order of 6.13 After equilibration, the resultant configurations were subjected to production simulations for 10 ns with 2 fs time step, with the parameters used being the same as those used during the equilibration, but without any restraints.

Simulations of the bundle

The bundle with the three TM domains was modeled in an implicit membrane environment using the same parameters as those used for the individual TM domains in implicit membrane environment. The bundle was equilibrated for 5 ns, after which loops were added to the TM domains using the CHARMM program.2 The structure with loops was equilibrated in an implicit membrane environment for 10 ns with a hydrophobic core of thickness 33 Å. This was followed by 30 ns of equilibration in an explicit lipid bilayer of POPC molecules. POPC was preferred over other phospholipids since it is the most abundant phospholipid in cell membranes,14 and can therefore closely resemble the features of a typical membrane. The simulation parameters used for this simulation were the same as those used in the explicit membrane simulations described above.

References

1.  The Sybyl Software, version 7.2 Tripos Inc, St Louis, MO, USA 2006.

2.  Brooks B R, Brooks III C L, MacKerell A D Jr, Nilsson L, Petrella R J, Roux B, Won Y, Archontis G, Bartels C, Boresch S, Caflisch A, Caves L, Cui Q, Dinner AR, Feig M, Fischer S, Gao J, Hodoscek M, Im W, Kuczera K, Lazaridis T, Ma J, Ovchinnikov V, Paci E, Pastor R W, Post C B, Pu J Z, Schaefer M, Tidor B, Venable R M, Woodcock H L, Wu X, Yang W, York D M and Karplus M 2009 J. Comput. Chem. 30 1545

3.  MacKerell A D Jr, Bashford D, Bellott M, Dunbrack R L, Evanseck J D, Field M J, Fischer S, Gao J, Guo H, Ha S, Joseph-McCarthy D, Kuchnir L, Kuczera K, Lau F T, Mattos C, Michnick S, Ngo T, Nguyen D T, Prodhom B, Reiher W E, Roux B, Schlenkrich M, Smith J C, Stote R, Straub J, Watanabe M, Wiórkiewicz-Kuczera J, Yin D and Karplus M 1998 J. Phys. Chem. B 102 3586

4.  MacKerell A D Jr, Feig M and Brooks III CL 2004 J. Comput. Chem. 25 1400

5.  Im W, Lee M S and Brooks III C L 2003 J. Comput. Chem. 24 1691

6.  Im W M, Feig M and Brooks III C L 2003 Biophys. J. 85 2900

7.  Phillips J C, Braun R, Wang W, Gumbart J, Tajkhorshid E, Villa E, Chipot C, Skeel R D, Kale L and Schulten K 2005J. Comput. Chem.26 1781

8.  Klauda J B, Venable R M, Freites J A, O’Connor J W, Tobias D J, Mondragon-Ramirez C, Vorobyov I, MacKerell A D Jr and Pastor R W 2010 J. Phys. Chem. B 114 7830

9.  Jorgensen W L, Chandrasekhar J, Madura J D, Impey R W and Klein M L 1983 J. Chem. Phys. 79 926

10.  Jo S, Kim T and Im W 2007 PLoS ONE 2 880

11.  Jo S, Lim J B, Klauda J B and Im W 2009 Biophys. J. 97 50

12.  Ryckaert J P, Ciccotti G and Berendsen H J C 1977 J. Comput. Chem. 23 327

13.  Essmann U, Perera L, Berkowitz M L, Darden T A, Lee H and Pedersen L G 1995 J. Chem. Phys. 103 8577

14.  Seelig A and Seelig J 1977Biochemistry16 45

Table S1. Equilibration steps with the corresponding parameter values used in each step

Figure S1. (a) Schematic representation of the calculation of kink angle. (b) Kink angle vs simulation time for TM1 in implicit membrane simulations. (c) Kink angle vs simulation time for TM1 in explicit membrane simulations.

Figure S2. (a) Schematic representation of the tilt angle of a TM domain with respect to the membrane normal (vertical line in the figure). (b) Tilt angle of the various TM domains in implicit membrane models. (c) Tilt angle of the various TM domains in explicit membrane models.

Figure S3. RMSD of the protein backbone during equilibration of the bundle in an implicit membrane environment.

1