Integration of QSAR modelling and QM/MM analysis to investigate functional food peptides with antihypertensive activity

Antihypertensive peptides derived from dietary proteins have long been recognised as an important source of developing functional foods with blood pressure-lowering effect. However, most of such peptides exhibit diverse tastes, such as sweet, bitter, sour and salty, which is a non-negligible aspect considered in the food development process. In the present study, several predictive quantitative structure–activity relationship (QSAR) models that correlate peptide's structural features with their multi-bioactivities and bitter taste are established at both sequence and structure levels, and the models are then used to conduct extrapolation on thousands of randomly generated, structurally diverse peptides with chain lengths ranging from two to six amino acid residues. Based on the statistical results gained from QSAR modelling, the relationship between the antihypertensive activity and bitter taste of peptides at different sequence lengths is investigated in detail. Moreover, the structural basis, energetic property and biological implication underlying peptide interactions with angiotensin-converting enzyme (ACE), a key target of antihypertensive therapy, are analysed at a complex three-dimensional structure level by using a high-level hybrid quantum mechanics/molecular mechanics scheme. It is found that (a) bitter taste is highly dependent on peptide length, whereas ACE inhibitory potency has only a modest correlation with the length, (b) dipeptides and tripeptides perform a moderate relationship between their ACE inhibition and bitterness, but the relationship could not be observed for those peptides of more than three amino acid residues and (c) the increase in sequence length does not cause peptides to exhibit substantial enhancement of antihypertensive activity; this is particularly significant for longer peptides such as pentapeptides and hexapeptides.