Predictive modelling of angiotensin converting enzyme inhibitory dipeptides

The ability of docking to predict angiotensin converting enzyme (ACE) inhibitory dipeptide sequences was assessed using AutoDock Vina. All potential dipeptides and phospho-dipeptides were docked and scored. Peptide intestinal stability was assessed using a prediction amino acid clustering model. Selected dipeptides, having AutoDock Vina scores ≤ -8.1 and predicted to be 'stable' intestinally, were characterised, using LIGPLOT and for ACE-inhibitory potency. Two newly identified ACE-inhibitory dipeptides, Asp-Trp and Trp-Pro, having Vina scores of -8.3 and -8.6 gave IC ₅₀ values of 258 ± 4.23 and 217 ± 15.7 μM, respectively. LIGPLOT analysis indicated no zinc interaction for these dipeptides. Phospho-dipeptides were predicted to have a good affinity for ACE. However, the experimentally determined IC ₅₀ results did not correlate since, for example, Trp-pThr and Pro-pTyr, having Vina scores of -8.5 and -8.1, respectively, displayed IC ₅₀ values of >500 μM. While docking allowed identification of new ACE inhibitory dipeptides, it may not be a fully reliable predictive tool in all cases.