Wu Jianping, Aluko Rotimi E, Nakai Shuryo
Department of Human Nutritional Sciences, University of Manitoba, Winnipeg, Manitoba, R3T 2N2, Canada.
J Agric Food Chem. 2006 Feb 8;54(3):732-8. doi: 10.1021/jf051263l.
A database consisting of 168 dipeptides and 140 tripeptides was constructed from published literature to study the quantitative structure--activity relationships of angiotensin I-converting enzyme (ACE) inhibitory peptides. Two models were computed using partial least squares regression based on the three z-scores of 20 coded amino acids and further validated by cross-validation and permutation tests. The two-component model could explain 73.2% of the Y-variance (inhibitor concentration that reduced enzyme activity by 50%, IC50) with the predictive ability of 71.1% for dipeptides, while the single-component model could explain 47.1% of the Y-variance with the predictive ability of 43.3% for tripeptides. Amino acid residues with bulky side chains as well as hydrophobic side chains were preferred for dipeptides. For tripeptides, the most favorable residues for the carboxyl terminus were aromatic amino acids, while positively charged amino acids were preferred for the middle position, and hydrophobic amino acids were preferred for the amino terminus. According to the models, the IC50 values of seven new peptides with matchable primary sequences within pea protein, bovine milk protein, and soybean were predicted. The predicted peptides were synthesized, and their IC50 values were validated through laboratory determination of inhibition of ACE activity.
从已发表的文献中构建了一个由168种二肽和140种三肽组成的数据库，用于研究血管紧张素I转换酶（ACE）抑制肽的定量构效关系。基于20种编码氨基酸的三个z分数，使用偏最小二乘回归计算了两个模型，并通过交叉验证和排列检验进一步验证。双组分模型可以解释73.2%的Y方差（使酶活性降低50%的抑制剂浓度，IC50），对二肽的预测能力为71.1%，而单组分模型可以解释47.1%的Y方差，对三肽的预测能力为43.3%。二肽优选具有大侧链以及疏水侧链的氨基酸残基。对于三肽，羧基末端最有利的残基是芳香族氨基酸，而中间位置优选带正电荷的氨基酸，氨基末端优选疏水氨基酸。根据模型，预测了豌豆蛋白、牛乳蛋白和大豆中具有匹配一级序列的七种新肽的IC50值。合成了预测的肽，并通过实验室测定ACE活性抑制来验证其IC50值。
