For this exercise, a small peptide will be constructed and then it's conformational energy minimized. To make it more interesting, a cash prize will be awarded to the person that achieves the lowest conformational energy. Review the concept of a force field and the problems and approaches to finding a minimum in a potential energy surface in the class text. You may also want to review criteria that are generally used to determine whether or not a minimum has been found.
For all of the exercises indicated below the octapeptide asp-arg-val-tyr-ile-his-pro-phe (DRVYIHPF) will be studied. This peptide is the hormone angiotensin II that is involved in the regulation of blood pressure in mammals, although the hormone is found in many other living species. The following steps will be carried out to the extent feasible.
(a) Build the peptide by linking together the component amino acids,
(b) Consider what should be the protonation state on each of the residues
with acidic or basic groups attached,
(c) Take into account somehow the effects of solvent molecules,
(d) Determine a (the) low energy conformation for the peptide by conformational
energy minimization,
(e) Examine differences in conclusions that result from the use of
different empirical force fields.
You may wish to draw out the covalent structure of the peptide in order to review the nature of the sidechains of the component amino acids.
A. Start the modeling package SYBYL and work through the tutorial on peptide building that is attached. This tutorial is also available in color at http://www.kncv.nl/tutorials/studs/sybyl/ tut.pep001.html. and from the Tripos Bookshelf.
The SYBYL tutorial uses a Tripos force field for peptide structures. An alternative and generally more popular force field is that provided by Kollman's group in the package AMBER (http://www.amber.ucsf.edu/amber/amber.html). This representation includes an expanded definition of amino acids that takes into account charge on side chains. The AMBER amino acids are listed below. (Generally, an amino acid name that ends in Z has the ionization state that would not be present near neutral pH.)
Name Property
ASP Aspartic acid, side chain ionized
ASZ Aspartic acid, side chain not ionized
GLU Glutamic acid, side chain ionized
GLZ Glutamic acid, side chain not ionized
LYS Lysine, side chain ionized (protonated)
LYZ Lysine, side chain neutral (not protonated)
HID Histidine, with the d-proton present
HIE Histidine, with the e-proton present
HIP Histidine, with the imidazole ring protonated
CYZ Cysteine that will eventually become part of a disulfide
bond
Consider what will be the ionization state of each amino acid of DRVYIHPF at neutral pH so that the correct amino acid type in the AMBER representation can be chosen.
B. Now build the peptide using the beta_sheet conformation. Go to Biopolymer >>> Monomer Dictionary… >>> Open >>> and select the Amber95protein data base. This will cause the peptide building process to be done with the most recent AMBER force field parameters published by the Kollman group. Build DRVYIHPF in SYBYL setting the initial conformation to beta_sheet. Add hydrogen to the carbon skeleton (Biopolymer >>> Add Hydrogens… >>> , select All, OK, and then add ALL hydrogen. The "essential" ones are those along the peptide backbone. The Add Hydrogens function will not add hydrogen to the side chains that have been defined to have a certain ionziation state.) Color the atoms (By Atom Type) and locate each of the acidic or basic groups of the molecule. Check that these have the correct number of hydrogens attached. The C- and N-terminal amino acids can be fixed by Biopolymer >>> Modify >>> Fix End Groups… Now load the partial charges on each atom that are needed by the force field, choosing the AMBER95_ALL set of charges. Store a copy of this structure in your directory exer3.
You may want to label the various residues of the peptide to make it easier to examine what is going on in the structure. Click View >>> Label >>> Substructure… Choose ALL, then OK. The residue names and position in the sequence should appear. These labels will rotate with the structure.
To turn off labels repeat this sequence, selecting Unlabel instead of Label.
C. Minimize the conformational energy of the peptide by selecting the Compute menu, then Minimize. The SYBYL minimization procedure includes a rapid Simplex optimization at the start of the run to remove bad contacts. This can be turned off by clicking on the Simplex button and changing it to None. Note that various minimization algorithms can be selected. For this exercise select the Conjugate Gradient option. Click on Modify… and set the Force Field to Amber4.1. Set the Charges: to Use Current (The Amber95 charges were installed earlier.). Note that details of how the dielectric constant and non-bonded terms are treated can be altered. Click OK and then OK in the main menu. An energy minimization calculation will be run (in foreground) for the number of iterations specified. Details of the minimization calculation are reported in the text window. Find and record the best (lowest) energy you can obtain when the dielectric constant is set to 4 (distance independent) and the distance cut-off is 30 A. Record the value of all energy terms, as well as the total energy.
D. Rebuild the peptide in the alpha helix conformation and store a copy of this structure. Minimize its conformational energy by the same processes used above. Find and report the best (lowest) energy you can when the dielectric constant is set to 4 (distance independent) and the distance cut-off is 30 A. Include all of the energy terms that contribute to the total energy.
Recall that any minimization procedure can only find a local minimum, not the global minimum. How can you be sure that the energy and attending conformation that has been produced corresponds to the global minimum? One approach is to start the minimization calculation from several different initial conformations and determine whether or not the same (minimum energy) structure is obtained in all cases. You have already calculated the minimum energy structure when the canonical helix and sheet structures are the starting points. Other starting points could be a "random" structure or perhaps a part of a protein structure in the PDB that is reasonably homologous to the sequence of angiotensin II. Or maybe you can look at the structures that have been produced and define some conformational change (probably a rotation) that would make the structure "better".
(1) Find the lowest conformational energy you can for angiotensin II using the AMBER95 force field, a distance-independent dielectric constant of 4 and a cut-off of 30 A. Report your best energy term-by-term. Create a structure file (.mol2 format) in your exer3 directory that corresponds to your best minimum-energy structure obtained under these assumptions. Call it yourname-min1.mol2.
(2) Find the lowest conformational energy you can for angiotensin II using the AMBER95 force field, a distance-dependent dielectric constant of 4 and a cut-off of 30 A. Report your best energy term-by-term. Create a structure file (.mol2 format) in your exer3 directory that corresponds to your best minimum-energy structure obtained under these assumptions. Call it yourname-min2.mol2.
Your structure files will be checked to see if you have done the calculation correctly and that your structure corresponds to the energy you report.
Hints: Forgetting to use the correct charges and/or force field are
the most common errors in doing this exercise. There can only be a fair
comparison of results if everyone uses exactly the same set of assumptions.
Be aware that some minimization calculations can unexpectedly run for a
long time.
By February 8, report the energies and have available the structure files described above.