A heuristic, computer-driven and top-down approach to identify novel bioactive peptides: A proof-of-principle on angiotensin I converting enzyme inhibitory peptides

Bioactive peptides are short peptides (3–20 amino acid residues in length) endowed of specific biological activities. The identification and characterization of bioactive peptides of food origin are crucial to better understand the physiological consequences of food, as well as to design novel foods, ingredients, supplements, and diets to counteract mild metabolic disorders. For this reason, the identification of bioactive peptides is also relevant from a pharmaceutical standpoint. Nevertheless, the systematic identification of bioactive sequences of food origin is still challenging and relies mainly on the so defined “bottom-up” approaches, which rarely results in the total identification of most active sequences. Conversely, “top-down” approaches aim at identifying bioactive sequences with certain features and may be more suitable for the precise identification of very potent bioactive peptides. In this context, this work presents a top-down, computer-assisted and hypothesis-driven identification of potent angiotensin I converting enzyme inhibitory tripeptides, as a proof of principle. A virtual library of 6840 tripeptides was screened in silico to identify potential highly potent inhibitory peptides. Then, computational results were confirmed experimentally and a very potent novel sequence, LMP was identified. LMP showed an IC50 of 15.8 and 6.8 µM in cell-free and cell-based assays, respectively. In addition, a bioinformatics approach was used to search potential food sources of LMP. Yolk proteins were identified as a possible relevant source to analyze in further experiments. Overall, the method presented may represent a powerful and versatile framework for a systematic, high-throughput and top-down identification of bioactive peptides.