SOYBEAN AND HEALTH - University of Macau Library

44Soybean and Healthrespectively, were selected by the design approach for unconstrained peptides (Pak et al.,2006). GLPTGG and GFPTGG, acting as a bisubstrate-mimic, were chosen among designedpeptides, having the most constrained structures (Fig. 8) (Pak et al., 2008a). Active peptideswere selected by the design parameter ‘V’ or ‘Pr’, which reflects the probability of activepeptide conformations for constrained and unconstrained peptides, respectively. (Table 7)(Pak et al., 2006, 2007). According to the ‘Pr’ value, FVAE and F(4-Fluoro)VAE haverelatively higher values compared with WVAE and halogen-containing peptides,HOO5O4 OHHN OO NHFOO5HOHO OH4OHOHHOO5O4OHOHNHOONH 21 HN23OO1 N 2 NNSO3NH 21OHNHNO23HYVAE3C OHONN 4CH 2 NH 1C HN 2 OOC HNOH5C y0zHNOHOH6 OHC Simvastatin and Rosuvastatina-TxOHOO5'N4NH ONHOON HOHNHONH 2-H-TIVAEOGFPTGG (H:-T)GFPTGG (-H:-T)bFig. 8. Models of HMG-CoA competitive inhibitory peptides (a) and peptide acting as abisubstrate (b). (a) Model of an ‘active space’ (tetrahedron 1-2-3-4) based on a superpositionof the bioactive conformations of simvastatin and rosuvastatin extracted from the crystalstructure of HMGR–statin complex (PDB codes: 1HW9 (simvastatin), 1DQA (rosuvastatin)),and the average structures determined for IVAE and YVAE peptides. (b) Model of the typeII β-turn in peptide conformer (-H:-T) with an arrangement of peptide head (H-bold) andtail (T-short-dashed) relative to the plane passing through the three α-carbon of 2, 3, and 5atoms of the peptide backbone, and the peptide conformation (-H:-T) used in this study. The4–5 and 4–50 fragments in the peptide models were fixed as an HMG-mimic duringmodeling (Pak et al., 2008a).