Polgár Tímea, Menyhárd Dóra K, Keseru György M
Gedeon Richter plc, P. O. Box 27, 1475, Budapest, Hungary.
J Comput Aided Mol Des. 2007 Sep;21(9):539-48. doi: 10.1007/s10822-007-9137-8. Epub 2007 Oct 25.
An effective virtual screening protocol was developed against an extended active site of CYP2C9, which was derived from X-ray structures complexed with flubiprofen and S-warfarin. Virtual screening has been effectively supported by our structure-based pharmacophore model. Importance of hot residues identified by mutation data and structural analysis was first estimated in an enrichment study. Key role of Arg108 and Phe114 in ligand binding was also underlined. Our screening protocol successfully identified 76% of known CYP2C9 ligands in the top 1% of the ranked database resulting 76-fold enrichment relative to random situation. Relevance of the protocol was further confirmed in selectivity studies, when 89% of CYP2C9 ligands were retrieved from a mixture of CYP2C9 and CYP2C8 ligands, while only 22% of CYP2C8 ligands were found applying the structure-based pharmacophore constraints. Moderate discrimination of CYP2C9 ligands from CYP2C18 and CYP2C19 ligands could also be achieved extending the application domain of our virtual screening protocol for the entire CYP2C family. Our findings further demonstrate the existence of an active site comprising of at least two binding pockets and strengthens the need of involvement of protein flexibility in virtual screening.
针对细胞色素P450 2C9(CYP2C9)扩展的活性位点开发了一种有效的虚拟筛选方案,该活性位点源自与氟比洛芬和S-华法林复合的X射线结构。基于结构的药效团模型有效地支持了虚拟筛选。在富集研究中首先评估了通过突变数据和结构分析确定的热点残基的重要性。还强调了精氨酸108(Arg108)和苯丙氨酸114(Phe114)在配体结合中的关键作用。我们的筛选方案成功地在排名数据库的前1%中识别出76%的已知CYP2C9配体,相对于随机情况富集了76倍。在选择性研究中进一步证实了该方案的相关性,当从CYP2C9和CYP2C8配体的混合物中检索到89%的CYP2C9配体时,而应用基于结构的药效团限制仅发现22%的CYP2C8配体。通过扩展我们的虚拟筛选方案在整个CYP2C家族中的应用领域,也可以实现对CYP2C9配体与CYP2C18和CYP2C19配体的适度区分。我们的研究结果进一步证明了存在一个至少由两个结合口袋组成的活性位点,并强化了在虚拟筛选中考虑蛋白质柔性的必要性。