Computer-Aided Drug Design And Partial Synthesis Of Novel VEGF Receptor Inhibitors

Tumors are malignant diseases that threaten human health. It is well known that thetraditional anti-cancer drugs always behave high toxicity and drug-resistance becauseof their inherent cytotoxic properties. Therefore, anti-cancer research turns tosearching new targets with high-specificity and well tolerability. Anti-angiogenesis isthe imminent breakthrough in cancer therapy, following the recognition thatangiogenesis is an important mechanism governing tumor growth and metastasis.VEGF, a potent specific mitogen for vascular endothelial cells, has become the mostimportant target for anti-angiogenesis since discoverd. The angiogenic function ofVEGF is mediated through binding to its specific receptor KDR. Thus, inhibitingKDR can serve as an efficient approach to inhibit growth and metastasis of tumors. The lead compounds of reported VEGF receptor inhibitors are discovered byrandom screening, whose derivatives showing improved activity and selectivity.However, there are still many limitations and deficiencies on further structuremodification and optimization because of lacking information of the mechanism ofinteracting between inhibitors and the target enzyme. In 1999, McTigue et al. reportedthe crystal structure of tyrosine kinase domain of KDR, which offers importantinformation for structure-based drug design. Based on their research, we designednovel lead compounds by using computer-aided drug design methods in order todevelop VEGF receptor inhibitors of independent intellectual proprietary rights. Firstly, by docking two classes of VEGFR inhibitors into the activity cavity ofKDR using AutoDock3.05, we identified the binding conformation of the inhibitorsand the functionally important residues of the active cavity of KDR. Docking resultsshow that the inhibitors occupy the site of adenine of ATP and form two conservedhydrogen bonds with the hinge region backbone. The substituent on the skeleton ofthe inhibitors extend into the hydrophobic pocket of KDR and make a hydrophobiccontact with KDR. Some residues of the activity cavity of KDR can form anadditional hydrogen bond with the inhibitors. According to these results, we propose - iii -第二军医大学·硕士学位论文·摘要that the general features of these inhibitors are: (1) a hydrophobic planar skeleton; (2)a hydrogen bond acceptor and a hydrogen bond donor which form the two conservedhydrogen bonds; (3) a hydrophobic substituent which can make a hydrophobic contactwith the hydrophobic pocket of KDR; (4) a additional hydrogen bond acceptor whichcan form a additional hydrogen bond. Then we studied the 3D-QSAR of the two classes of these inhibitors. The inhibitormolecules were aligned by rms Fit and a best grid was obtained by using the programwritten by us. A maximum of Q2 and ONC was obtained by Leave-One-Out PLS. TheCoMFA and CoMSIA model was derived through none-cross-validation PLS. All ofthe values of Q2 are greater than 0.5, indicating that the models have a satisfactorypredictive ability. Based on the models, we elucidated the distribution of the stericfield, the hydrophobic field, the electrostatic field, the hydrogen bond acceptor andhydrogen bond donor, which contribute to the activity. Finally, we designed a series of novel 7-benzyl- 4-pyridinylmethyl-1-isoquinolinamine compounds and their synthesis routes. Partially of the syntheticexperiments have been already done.