Integrating Computational Modeling and Experimental Assay to Discover New Potent ACE‐Inhibitory Peptides

Abstract

Human angiotensin-I-converting enzyme (ACE) is an important target of antihypertensive therapy, which possesses a bulky, hydrophobic pocket that is physicochemically compatible with a wide variety of peptide substrates and small-molecule ligands. Rational design of potent ACE inhibitors has long been an attractive topic in the chemical, biological and medical communities. In the present study, an integrative protocol is described to optimize and modify peptides bound with ACE based on their complex three-dimensional structures. The protocol combines a number of sophisticated computational methods including molecular dynamics (MD) simulation, quantitative structure-activity relationship (QSAR), molecular mechanics/Poisson-Boltzmann surface area (MM-PB/SA) and quantum mechanics/molecular mechanics (QM/MM) to discover new potent ACE-inhibitory peptides. With this strategy hundreds of potential peptides are generated virtually, from which several promising candidates are synthesized and assayed in vitro using a standard Fmoc -protected amino acid solid phase synthesis and spectrophotometric method, respectively. Six peptides are found to have potency against ACE, four of which (LVY, VLKP, MLPVY and LKIPLY) show satisfactory inhibitory capability (pIC50 =5.84, 5.27, 5.40 and 5.57, respectively). Subsequently, the complex structures of tripeptide LVY and hexapeptide LKIPLY with ACE are computationally modeled and their binding free energies are estimated as - 34.2 and - 57.8 kcal/mol, respectively, by using a rigorous QM/MM scheme. Intensive steric collisions are observed around the C-terminus of peptide ligands in the bound state. Based upon the findings the C-terminal residues of the two peptides are modified to small, hydrophobic amino acids Val and Ala, resulting in mutated peptides LVA and LKIPVA, respectively. Consequently, a significant improvement in ACE-inhibitory activity is observed for the LKIPVA mutant (pIC50 increases from 5.57 to 6.07), whereas only a modest activity change was associated with the LVA mutant (pIC50 decreases from 5.84 to 5.80).

Evolution of Optimal Descriptors

“…The negative logarithm of peptide concentration at which the enzymatic activity of ACE is reduced by 50% (pIC50, n=234) has been examined. In the literature (Ren et al 2014) data for 289 peptides are available, but for the method used in this work, peptides which contain two amino acids cannot be involved in building up model. The above datasets were randomly split into the visible training (≈80%), calibration (≈10%), and validation set (≈10%).…”

Activity and bioavailability of food protein‐derived angiotensin‐I‐converting enzyme–inhibitory peptides

“…Molecular simulation is increasingly accepted and used by researchers, especially for drug development, but has not been widely applied in ACE-resistant peptides research. With the development of molecular simulation technique, artificial intelligence, and big data accumulated from previous studies, researchers could optimize simulation models to better predict ACE-inhibitory activity of new or untested peptides, or to design new peptides (Ren, Wang, Chen, & Cao, 2014). This is undoubtedly a promising area of future research.…”