DFT studies on helix formation in N-acetyl-(L-alanyl)(n)-N '-methylamide for n=1-20

We compare the geometries and relative energies of important secondary stru ctural elements, the 3.6(13) helix, 3(10) helix and C-5(ext) structures, fo r a set of blocked peptide models, N-acetyl-(L-alanyl)(n)-N'-methylamide, f or n = 1-20. We use full density-functional theory (DFT) calculations at th e B3LYP/6-31G* level (for peptides up toll residues), the self-consistent-c harge density-functional tight binding (SCC-DFTB) and the semiempirical AM1 method. The 3.6(13) and 3(10) structures are found to be not inherently st able in general. Their stability is dependent on peptide length, other stru ctural motifs and aqueous or membrane environments. For short peptides with less than eight residues, the 3.6(13) helix relaxes into the 3(10) structu re. For longer peptides, the 3.6(13) is Stable in the middle of the chain, while the ends assume 3(10) conformations, at the C-terminus additionally a beta II type turn is formed. The relative energies and structures calculat ed with the recently developed SCC-DFTB method are in very good agreement w ith the results from the B3LYP density-functional calculations. Therefore, we use the SCC-DFTB method to look at helix formation in N-acetyl-(L-alanyl )(n)-N'-methylamide for n = 11, 14, 17 and 20. On the SCC-DFTB potential en ergy surface, we find the 3(10) helix to be more stable than the 3.6(13) he lix for all peptide sizes. However, the effects of solution might change th is picture and favor the 3.6(13) motif. (C) 2000 Elsevier Science B.V. All rights reserved.