Peptide fragmentation as an approach in modeling of an active peptide and designing a competitive inhibitory peptide for HMG-CoA reductase

This study presents a simple method to design an active peptide based on a description of the structural preferences of peptide via its peptide fragments. In a previous design, while searching for lead peptide candidates, the efficacy of a design approach that was based on the use of a cyclic peptide as a model of linear analog was demonstrated. Analysis of the conformational behavior of the peptide models showed that an analogical approach could be applied in order to assess the conformational space that was occupied by a peptide by using peptide fragments. In order to assess the proposed method, a design of a competitive inhibitor for HMG-CoA reductase (HMGR) was performed. Two starting points were used in the design: (1) determined recognized residues and (2) the structural preference of a peptide, such as a beta-turn conformation in the present design. Two sets of peptides were developed based on the different location of a beta-turn structure relative to a recognized residue. Set 1 contains peptides in which a recognized residue is included in turn conformation. In Set 2, the turn structure is located distantly from the recognized residues. The design parameter 'V' that was applied in previous studies was slightly modified for the purpose of the current research. The 17 previously and 8 newly designed peptides were estimated by this parameter. In each set, one sequence was selected as a lead peptide candidate for each set: GF(4-fluoro)PEGG for Set 1 and DFGYVAE for Set 2. The inhibitory activities improved in each set. The IC(50) for the GF(4-fluoro)PEGG peptide was found to be 0.75 microM, while the linear DFGYVAE peptide (IC(50)=0.16 microM) showed a 3000-fold increase in inhibitory activity compared to the first isolated LPYP peptide (IC(50)=484 microM) from soybeans. The comparison of the structure-activity relationship (SAR) data between Set 1 and 2 provided an opportunity to design the peptides in terms of peptide selectivity. A structural analysis of the modeled peptides confirmed the appropriateness of the proposed method for the design of active peptides.