PDF(Pubmed)
Abstract：
OBJECTIVE: Nowadays, cyclin-dependent kinase 4/6 (CDK4/6) inhibitors have been approved for treating metastatic breast cancer and have achieved inspiring curative effects. But some discoveries have indicated that CDK 4/6 are not the requisite factors in some cell types because CDK2 partly compensates for the inhibition of CDK4/6. Thus, it is urgent to design CDK2/4/6 inhibitors for significantly enhancing their potency. This study aims to explore the mechanism of the binding of CDK2/4/6 kinases and their inhibitors to design novel CDK2/4/6 inhibitors for significantly enhancing their potency in different kinds of cancers.
METHODS: A series of 72 disparately functionalized 4-substituted N-phenylpyrimidin-2-amine derivatives exhibiting potent inhibitor activities against CDK2, CDK4 and CDK6 were collected to apply to this research. The total set of these derivatives was divided into a training set (54 compounds) and a test set (18 compounds). The derivatives were constructed through the sketch molecule module in SYBYL 6.9 software. A Powell gradient algorithm and Tripos force field were used to calculate the minimal structural energy and the minimized structure was used as the initial conformation for molecular docking. By the means of 3D-QSAR models, partial least squares (PLS) analysis, molecular dynamics (MD) simulations and binding free energy calculations, we can find the relationship between structure and biological activity.
RESULTS: In this study, we used molecular docking, 3D-QSAR and molecular dynamics simulation methods to comprehensively analyze the interaction and structure-activity relationships of 72 new CDK2/4/6 inhibitors. We used detailed statistical data to reasonably verify the constructed 3D-QSAR models for three receptors (q2 of CDK2 = 0.714, R2pred = 0.764, q2 = 0.815; R2pred of CDK4 = 0.681, q2 = 0.757; R2pred of CDK6 = 0.674). MD simulations and decomposition energy analysis validated the reasonability of the docking results and identified polar interactions as crucial factors that influence the different bioactivities of the studied inhibitors of CDK2/4/6 receptors, especially the electrostatic interactions of Lys33/35/43 and Asp145/158/163. The nonpolar interaction with Ile10/12/19 was also critical for the differing potencies of the CDK2/4/6 inhibitors. We concluded that the following probably enhanced the bioactivity against CDK2/4/6 kinases: (1) electronegative groups at the N1-position and electropositive and moderate-sized groups at ring E; (2) electrogroups featured at R2; (3) carbon atoms at the X-position or ring C replaced by a benzene ring; and (4) an electrogroup as R4.
CONCLUSIONS: Previous studies, to our knowledge, only utilized a single approach of 3D-QSAR and did not integrate this method with other sophisticated techniques such as molecular dynamics simulations to discover new potential inhibitors of CDK2, CDK4, or CDK6. So we applied the intergenerational technology, such as 3D-QSAR technology, molecular docking simulation techniques, molecular dynamics simulations and MMPBSA19/MMGBSA20-binding free energy calculations to statistically explore the correlations between the structure with biological activities. The constructed 3D-QSAR models of the three receptors were reasonable and confirmed by the excellent statistical data. We hope the results obtained from this work will provide some useful references for the development of novel CDK2/4/6 inhibitors.
METHODS: A series of 72 disparately functionalized 4-substituted N-phenylpyrimidin-2-amine derivatives exhibiting potent inhibitor activities against CDK2, CDK4 and CDK6 were collected to apply to this research. The total set of these derivatives was divided into a training set (54 compounds) and a test set (18 compounds). The derivatives were constructed through the sketch molecule module in SYBYL 6.9 software. A Powell gradient algorithm and Tripos force field were used to calculate the minimal structural energy and the minimized structure was used as the initial conformation for molecular docking. By the means of 3D-QSAR models, partial least squares (PLS) analysis, molecular dynamics (MD) simulations and binding free energy calculations, we can find the relationship between structure and biological activity.
RESULTS: In this study, we used molecular docking, 3D-QSAR and molecular dynamics simulation methods to comprehensively analyze the interaction and structure-activity relationships of 72 new CDK2/4/6 inhibitors. We used detailed statistical data to reasonably verify the constructed 3D-QSAR models for three receptors (q2 of CDK2 = 0.714, R2pred = 0.764, q2 = 0.815; R2pred of CDK4 = 0.681, q2 = 0.757; R2pred of CDK6 = 0.674). MD simulations and decomposition energy analysis validated the reasonability of the docking results and identified polar interactions as crucial factors that influence the different bioactivities of the studied inhibitors of CDK2/4/6 receptors, especially the electrostatic interactions of Lys33/35/43 and Asp145/158/163. The nonpolar interaction with Ile10/12/19 was also critical for the differing potencies of the CDK2/4/6 inhibitors. We concluded that the following probably enhanced the bioactivity against CDK2/4/6 kinases: (1) electronegative groups at the N1-position and electropositive and moderate-sized groups at ring E; (2) electrogroups featured at R2; (3) carbon atoms at the X-position or ring C replaced by a benzene ring; and (4) an electrogroup as R4.
CONCLUSIONS: Previous studies, to our knowledge, only utilized a single approach of 3D-QSAR and did not integrate this method with other sophisticated techniques such as molecular dynamics simulations to discover new potential inhibitors of CDK2, CDK4, or CDK6. So we applied the intergenerational technology, such as 3D-QSAR technology, molecular docking simulation techniques, molecular dynamics simulations and MMPBSA19/MMGBSA20-binding free energy calculations to statistically explore the correlations between the structure with biological activities. The constructed 3D-QSAR models of the three receptors were reasonable and confirmed by the excellent statistical data. We hope the results obtained from this work will provide some useful references for the development of novel CDK2/4/6 inhibitors.
摘要：
目标：如今，细胞周期蛋白依赖性激酶4/6（CDK4/6）抑制剂已被批准用于治疗转移性乳腺癌，并取得了令人鼓舞的疗效。但是一些发现表明CDK4/6不是某些细胞类型的必需因子，因为CDK2部分补偿了CDK4/6的抑制。因此，迫切需要设计CDK2/4/6抑制剂以显着增强其效力。本研究旨在探讨CDK2/4/6激酶及其抑制剂的结合机制，以设计新型CDK2/4/6抑制剂，显着增强其在不同类型癌症中的效力。
方法:收集了一系列72种不同官能化的4-取代的N-苯基嘧啶-2-胺衍生物,其表现出对CDK2、CDK4和CDK6的有效抑制剂活性,以应用于该研究。将这些衍生物的总集分成训练集(54个化合物)和测试集(18个化合物)。通过SYBYL6.9软件中的草图分子模块构建衍生物。使用Powell梯度算法和Tripos力场来计算最小结构能量，并将最小结构用作分子对接的初始构象。通过3D-QSAR模型,偏最小二乘(PLS)分析，分子动力学（MD）模拟和结合自由能计算，我们可以找到结构和生物活性之间的关系。
结果：在这项研究中，我们用分子对接,3D-QSAR和分子动力学模拟方法综合分析了72种新型CDK2/4/6抑制剂的相互作用和构效关系。我们使用详细的统计数据来合理地验证三个受体的3D-QSAR模型（CDK2的q2=0.714，R2pred=0.764，q2=0.815；CDK4的R2pred=0.681，q2=0.757；CDK6的R2pred=0.674）。MD模拟和分解能量分析验证了对接结果的合理性，并确定了极性相互作用是影响CDK2/4/6受体抑制剂不同生物活性的关键因素。特别是Lys33/35/43和Asp145/158/163的静电相互作用。与Ile10/12/19的非极性相互作用对于CDK2/4/6抑制剂的不同效力也是关键的。我们得出的结论是，以下内容可能增强了对CDK2/4/6激酶的生物活性：（1）N1位置的负电基团以及E环的正电性和中等大小的基团；（2）R2上的电基团；（3）X位或C环的碳原子被苯环取代；（4）作为R4的电基团。
结论：以前的研究，根据我们的知识，仅使用了单一的3D-QSAR方法,并且没有将该方法与其他复杂的技术如分子动力学模拟相结合,以发现新的CDK2,CDK4或CDK6的潜在抑制剂.所以我们应用了代际技术，例如3D-QSAR技术，分子对接模拟技术，分子动力学模拟和MMPBSA19/MMGBSA20结合自由能计算，以统计方式探索结构与生物活性之间的相关性。所构建的三种受体的3D-QSAR模型是合理的，并得到了良好的统计数据的证实。希望本文的研究结果能够为新型CDK2/4/6抑制剂的开发提供参考。
方法:收集了一系列72种不同官能化的4-取代的N-苯基嘧啶-2-胺衍生物,其表现出对CDK2、CDK4和CDK6的有效抑制剂活性,以应用于该研究。将这些衍生物的总集分成训练集(54个化合物)和测试集(18个化合物)。通过SYBYL6.9软件中的草图分子模块构建衍生物。使用Powell梯度算法和Tripos力场来计算最小结构能量，并将最小结构用作分子对接的初始构象。通过3D-QSAR模型,偏最小二乘(PLS)分析，分子动力学（MD）模拟和结合自由能计算，我们可以找到结构和生物活性之间的关系。
结果：在这项研究中，我们用分子对接,3D-QSAR和分子动力学模拟方法综合分析了72种新型CDK2/4/6抑制剂的相互作用和构效关系。我们使用详细的统计数据来合理地验证三个受体的3D-QSAR模型（CDK2的q2=0.714，R2pred=0.764，q2=0.815；CDK4的R2pred=0.681，q2=0.757；CDK6的R2pred=0.674）。MD模拟和分解能量分析验证了对接结果的合理性，并确定了极性相互作用是影响CDK2/4/6受体抑制剂不同生物活性的关键因素。特别是Lys33/35/43和Asp145/158/163的静电相互作用。与Ile10/12/19的非极性相互作用对于CDK2/4/6抑制剂的不同效力也是关键的。我们得出的结论是，以下内容可能增强了对CDK2/4/6激酶的生物活性：（1）N1位置的负电基团以及E环的正电性和中等大小的基团；（2）R2上的电基团；（3）X位或C环的碳原子被苯环取代；（4）作为R4的电基团。
结论：以前的研究，根据我们的知识，仅使用了单一的3D-QSAR方法,并且没有将该方法与其他复杂的技术如分子动力学模拟相结合,以发现新的CDK2,CDK4或CDK6的潜在抑制剂.所以我们应用了代际技术，例如3D-QSAR技术，分子对接模拟技术，分子动力学模拟和MMPBSA19/MMGBSA20结合自由能计算，以统计方式探索结构与生物活性之间的相关性。所构建的三种受体的3D-QSAR模型是合理的，并得到了良好的统计数据的证实。希望本文的研究结果能够为新型CDK2/4/6抑制剂的开发提供参考。
