Anti-cancer Drug Paclitaxel Analogs Qsar And Quantum Chemical Study

Taxol isolated from bark of Taxus brevifolia in the late 1960s has become the drugs of choice for the treatment of ovarian and breast cancer. In attempts to understand the action of Taxol at its microtubule target, extensive SAR studies have been conducted. The main conclusions of these studies are that C-13 side chain, the ester groups at C-2 and C-4, the oxetane ring are all essential for biological activity. In this thesis Taxol analogs have been calculated by using of molecular mechanics, quantum chemistry methods and statistic methods for obtaining the models of Quantitative structure-activity relationships. With the QSAR models and geometry conformation discussion we search the good geometry of taxol, discuss the necessary of the four-member D-ring, explore influence of the configuration at C-2' and C-3' for biological activity. All of above should be the theory foundations for obtaining some Taxol analogs with high biological activity, good water-solubility and low toxicity.The main conclusions are as follows:1. For QSAR research, the taxotere's geometry optimization conformations that were calculated by MM3, MM+, PM3, MNDO respectively have been discussed. The geometry optimization conformation of MM3 is the best one, and that means MM3 is the best method for geometry optimization in them; An ab initio GIAO method at HF/6-31G* level has been used to theoretically predict 1H-NMR and 13C-NMR chemical shifts for the conformations that has been optimized by MM3, PM3 and ab initio methods at RHF/STO-3G respectively. The Results show that the calculated 6 1H-NMR and 6 13C-NMR for the geometry conformation optimized by HF/STO-3G are most accurate in comparison with experimentally observed data. It is inferred that, therefore, the geometry conformation optimized by ab initio method at RHF/STO-3G must probably be approximate to the real one of taxol. HF/STO-3G is the best method on accurate theoretical study for taxol.2. After the MM3 geometry optimizating, the electronic structure of taxol analogues both a group containing 28 ones and a group containing 43 ones have been calculated by MNDO quantum chemical method. To analyze the QSAR of taxol analogue anti-tumor drugs, the pattern recognition of BP artificial neural network have been performed. It can be concluded as below: (1) The relationship between logP values of C-13 side chain in taxol analogues and activity is parabola model. (2) The more of the negative charge of Bz on 2-OBz, the higher activities they will be have. (3) The pattern recognition of BP artificial neural network has a 98% accuracy, which may predict the activities of taxol analogues.3. Four-member D-ring is not necessary to maintain activity of Taxol analogue. If only the substituents that come from opening of D-ring don't transfigure the concave cavity of the molecule, the biological activity of Taxol analogues will not decrease greatly. Furthermore the results of geometry conformation show that for the configuration (2'R,3'S) both 3'-NH and 2'-OH extent outwards from the molecule, so both the 3'-NH and 2'-OH can probably interact with receptors at the same time, which could explain that Taxol analogues achieve the hightest biological activity when C13 side chain adopts natural conformation (2'R,3'S).