The main advantages of the Darling models for representing proteins are the rigidity of peptide planes and the ease with which hydrogen bonds are represented.
I have found that the easiest way to build protein structures rapidly with the Darling models is to start by building a stock of pre-built, interchangeable components. For proteins, the most important component to have is a peptide plane, containing four atoms: the carbonyl carbon and oxygen from one residue, the nitrogen of the next residue, and the hydrogen bonded to the nitrogen. To make these pieces maximally useful, we want to be able to form a hydrogen bond from any backbone NH to any backbone O, and we would like the rod/tube connections to the C-alpha atoms at the two ends of the peptide plane to be of opposite types (consistently for all peptide planes).
Steve Darling provides instructions for one way to build peptide planes, which I will explain first, then I'll give a somewhat simpler way to build the planes that is faster and cheaper, but does not represent the pi-bonds as explicitly. The two are interchangeable, so you can build as many of each as you choose (limited by the available parts).
Here are the pieces needed for the complex peptide plane (without the hydrogen): 1 red linear, 1 red half-pi, 1 grey trigonal 2t/1r, 2 grey SP-2, 1 blue trigonal 2r/1t, 1 blue half-pi. The red is for the carbonyl oxygen, which is double-bonded to the grey carbonyl carbon, which is double-bonded to the blue nitrogen. Because the pi bonds are shared by the oxygen and nitrogen, the carbonyl carbon still needs to make one covalent bond.
Start by snapping the half-pi pieces onto the oxygen and nitrogen, so that the tube of the oxygen is covered and the central rod of the nitrogen is covered. It does not matter which way round the half-pi snaps into the oxygen, but once that is set, there is only one good way to put the nitrogen together.
Now let's go through the same construction for a simpler way to get the same geometry. This time, I'll include the hydrogen. The pieces needed are 1 red linear, 1 grey half-pi, 1 grey trigonal 2t/1r (2 tubes and 1 rod), 1 blue trigonal 2r/1t (2 rods and 1 tube), 1 blue half-pi, and 1 white linear. The white linear pieces is for forming a hydrogen bond---you can use a white ball if you don't want to make an H-bond.
The first thing to do is again to snap on the half-pi pieces, making sure that the rods of the carbon and nitrogen will end up on the same side of the peptide plane:
The next step, after assembling some peptide planes, is to add some alpha carbons. I find it easiest to add one black SP-3 piece for each peptide plane and build the peptide backbone before adding the other SP-3 piece to finish the tetrahedral alpha carbons.
The last step in building the protein backbone is to finish the alpha carbons. For each alpha carbon, we'll add one more black SP-3 piece and a white ball to represent the hydrogen. To make sidechains be interchangeable, we'll want the hydrogen to be always on the tube, with the free rod left for attaching the beta carbon of a sidechain. There are two ways to put in this SP-3 piece, corresponding to L-amino and D-amino acids. This is the only place where there is chirality in the backbone---everything we have done up to now has mirror symmetry.
The mnemonic I use for remembering the orientation for the normal (L-amino) acids is "clockwise CORN". If you look down the hydrogen-C-alpha bond, you should see the CO group, the Residue, and the N arranged clockwise, as in the picture below which has a single methyl group as the sidechain (making an alanine). We are looking down the H-CA bond of the amino acid that is on the left of the picture, and the CO, R (methyl group), and N residues are clockwise around the CA atom.
Also note that in the "extended" conformation shown here, the alanine sidechain on the left points downward while the one on the right points upward. This alternation of sides is a standard arrangement for the sidechains in a beta sheet.
Here is a picture of a full 3-alanine peptide, include an initial NH3+ and a final carboxyl group. Note that if you begin and end with a peptide plane, then you have not included the first and last residues of the chain.
|BME 205 home page||Karplus's lab page||UCSC Bioinformatics research|
Questions about page content should be directed to
University of California, Santa Cruz
Santa Cruz, CA 95064