Optimal potentials for predicting inter-helical packing in transmembrane proteins.

A set of pairwise contact potentials between amino acid residues in transmembrane helices was determined from the known native structure of the transmembrane protein (TMP) bacteriorhodopsin by the method of perceptron learning, using Monte Carlo dynamics to generate suitable "decoy" structures. The procedure of finding these decoys is simpler than for globular proteins, since it is reasonable to assume that helices behave as independent, stable objects and, therefore, the search in the conformational space is greatly reduced. With the learnt potentials, the association of the helices in bacteriorhodopsin was successfully simulated. The folding of a second TMP (the helix-dimer glycophorin A) was then accomplished with only a refinement of the potentials from a small number of decoys.

Simple model to study insertion of a protein into a membrane.

A simple coarse grained model on a two-dimensional lattice is presented to elucidate the main effects ruling the insertion of a protein into a polar environment such as a lipidic membrane. The amino acids are divided into two classes (hydrophobic or polar), and they behave differently according to their surroundings. In aqueous solution the hydrophobic amino acids are forced to minimize contacts with water, whereas in the apolar environment all the amino acids try to aggregate regardless to their specificity. The lattice is employed in order to perform exact calculations and to generate a fictitious protein data bank. Despite the simplicity of the model, some morphological features of the protein-like lattice structures obtained by our model are compatible with the observed phenomenology of transmembrane proteins. These results seem to corroborate the hypothesis that the number of classes into which the amino acids can be divided that correctly describe the phenomena may be extremely low.