Kaserer Teresa, Höferl Martina, Müller Klara, Elmer Sebastian, Ganzera Markus, Jäger Walter, Schuster Daniela
Computer-Aided Molecular Design Group, Institute of Pharmacy/Pharmaceutical Chemistry and Center for Molecular Biosciences Innsbruck (CMBI), University of Innsbruck, Innrain 80-82, 6020 Innsbruck, Austria phone/fax:+43 512 507 58253/+43 512 507 58299.
Department of Pharmaceutical Chemistry, Division of Clinical Pharmacy and Diagnostics, University of Vienna, Althanstr. 14, 1090 Vienna, Austria.
Mol Inform. 2015 Jun;34(6-7):431-57. doi: 10.1002/minf.201400192. Epub 2015 Jun 23.
The cytochrome P450 (CYP) superfamily represents the major enzyme class responsible for the metabolism of exogenous compounds. Investigation of clearance pathways is therefore an integral part in early drug development, as any alteration of metabolic enzymes may markedly influence the toxicological profile and efficacy of novel compounds. In silico methods are widely applied in drug development to complement experimental approaches. Several different tools are available for that purpose, however, for CYP enzymes they have only been applied retrospectively so far. Within this study, pharmacophore- and shape-based models and a docking protocol were generated for the prediction of CYP1A2, 2C9, and 3A4 inhibition. All theoretically validated models, the validated docking workflow, and additional external bioactivity profiling tools were applied independently and in parallel to predict the CYP inhibition of 29 compounds from synthetic and natural origin. After subsequent experimental assessment of the in silico predictions, we analyzed and compared the prospective performance of all methods, thereby defining the suitability of the applied techniques for CYP enzymes. We observed quite substantial differences in the performances of the applied tools, suggesting that the rational selection of that virtual screening method that proved to perform best can largely improve the success rates when it comes to CYP inhibition prediction.
细胞色素P450（CYP）超家族是负责外源性化合物代谢的主要酶类。因此，清除途径的研究是早期药物开发中不可或缺的一部分，因为代谢酶的任何改变都可能显著影响新化合物的毒理学特征和疗效。计算机模拟方法在药物开发中被广泛应用以补充实验方法。为此有几种不同的工具可用，然而，对于CYP酶，到目前为止它们仅被用于回顾性研究。在本研究中，生成了基于药效团和形状的模型以及对接方案，用于预测CYP1A2、2C9和3A4的抑制作用。所有经过理论验证的模型、经过验证的对接工作流程以及额外的外部生物活性分析工具被独立且并行地应用，以预测29种合成和天然来源化合物对CYP的抑制作用。在对计算机模拟预测进行后续实验评估后，我们分析并比较了所有方法的前瞻性性能，从而确定所应用技术对CYP酶的适用性。我们观察到所应用工具的性能存在相当大的差异，这表明合理选择被证明表现最佳的虚拟筛选方法在预测CYP抑制作用时可以很大程度上提高成功率。
