Ring C S, Cohen F E
Department of Pharmaceutical Chemistry, University of California, San Francisco 94143-0446.
FASEB J. 1993 Jun;7(9):783-90. doi: 10.1096/fasebj.7.9.8330685.
Although no general solution to the protein folding problem exists, the three-dimensional structures of proteins are being successfully predicted when experimentally derived constraints are used in conjunction with heuristic methods. In the case of interleukin-4, mutagenesis data and CD spectroscopy were instrumental in the accurate assignment of secondary structure. In addition, the tertiary structure was highly constrained by six cysteines separated by many residues that formed three disulfide bridges. Although the correct structure was a member of a short list of plausible structures, the "best" structure was the topological enantiomer of the experimentally determined conformation. For many proteases, other experimentally derived structures can be used as templates to identify the secondary structure elements. In a procedure called modeling by homology, the structure of a known protein is used as a scaffold to predict the structure of another related protein. This method has been used to model a serine and a cysteine protease that are important in the schistosome and malarial life cycles, respectively. The model structures were then used to identify putative small molecule enzyme inhibitors computationally. Experiments confirm that some of these nonpeptidic compounds are active at concentrations of less than 10 microM.
尽管蛋白质折叠问题尚无通用解决方案，但当将实验得出的限制条件与启发式方法结合使用时，蛋白质的三维结构正在被成功预测。就白细胞介素-4而言，诱变数据和圆二色光谱对于二级结构的准确归属起到了重要作用。此外，三级结构受到六个被多个残基分隔的半胱氨酸的高度限制，这些半胱氨酸形成了三个二硫键。尽管正确结构是一小串合理结构中的一员，但“最佳”结构却是实验确定构象的拓扑对映体。对于许多蛋白酶而言，其他实验得出的结构可作为模板来识别二级结构元件。在一种称为同源建模的过程中，已知蛋白质的结构被用作支架来预测另一种相关蛋白质的结构。该方法已被用于分别对在血吸虫和疟疾生命周期中起重要作用的一种丝氨酸蛋白酶和一种半胱氨酸蛋白酶进行建模。然后，模型结构被用于通过计算识别假定的小分子酶抑制剂。实验证实，其中一些非肽类化合物在浓度低于10微摩尔时具有活性。
