Structure-activity Relationship Studies On Polyhydroxylated Xanthones

Quantitative structure-activity relationship (QSAR) is an important means of environmental and ecological risk assessment on organic compounds. It can be used to predict the migration, transformation and distribution behavior of organic compounds in the environment. In QSAR studies on organic compounds, density functional theory (DFT) is a major method to obtain thermodynamic parameters and structural parameters of molecules.Polyhydroxylated xanthones (PHOXTHs) is a kind of xanthone derivatives, which have extensive applications in pharmaceutical and medical industries due to their biological activities and pharmacological properties. So far, there are no complete and comprehensive reports concerning the thermodynamic and physicochemical parameters of PHOXTHs. Understanding the thermodynamic, physicochemical and spectroscopic properties of PHOXTHs is of great significance for the design of new drugs and the researches on their environmental behavior. In the current paper, PHOXTHs were studied, and the appropriate models were constructed. The main contents and the corresponding conclusions are as follows:1. The spatial orientation of hydroxyl groups in PHOXTHs and the type and strength of the intramolecular hydrogen bonds, were analyzed. There are three types of intramolecular hydrogen bonds in PHOXTHs. The bond energy is about52kJ·mol-1,12kJ·mol-1,20kJ·mol-1, respectively. The influences of the intramolecular hydrogen bonds on molecular stability were discussed. The standard enthalpies of formation (△fHθ) and the standard Gibbs energies of formation (△fGθ) of XTH and135PHOXTHs were calculated with the design of isodesmic reactions. The QSAR models between thermodynamic parameters (Sθ,△fHθ and△fGθ) and the number and position of HO-substitution (NPHOS) were established. All the squared correlation coefficients R2are larger, with values being0.986,1.000and1.000, and all the standard error SE are smaller, with values being3.910,5.350and5.304, indicating these thermodynamic parameters all have close relationships with NPHOS· Consequently, these thermodynamic parameters can be predicted from the NPHOS·2. Two new reaction routes were devised and optimized to synthesize xanthone,3methoxyl-substituted xanthones and3hydroxyl-substituted xanthones. The starting materials of the synthetic route1are o-dibromobenzene and salicylaldehyde or methoxyl-substituted salicylaldehydes, and the substrate of the synthetic route2is salicylic acid. Infrared spectra and1H NMR spectra were adopted for structural identification of the samples synthesized, high performance liquid chromatography was employed to determine the purities of the samples, and the melting points were measured. The advantages and disadvantages of the two synthetic routes were compared with each other. For the route1, the reaction steps are less and the post-processing is relatively simple, but the substrate o-dibromobenzene and the catalyst triphenylphosphine palladium are expensive. Although the route2has less toxic and cheaper raw materials, and mild reaction conditions, it is time-consuming with more reaction steps.3. The infrared spectra of the several xanthones synthesized were analyzed and compared. The infrared spectrum of2-hydroxyxanthone was comprehensively studied, the vibrational modes of various chemical bonds in the compound were determined, and the correlation equation between the experimental frequencies and the scaled theoretical frequencies calculated at the two computational levels was obtained. The squared correlation coefficients R2are larger than0.98, indicating that good linearity relationships exist between the scaled and experimental frequencies. However, the larger set does not show evident improvement in the accuracy of the vibration frequencies. Consequently, in order to shorten the calculation time, one can adopt the smaller basis set without reducing the computational accuracy.4. The n-octanol-water partition coefficient (logKow) values of136polyhydroxylated xanthones were calculated using the Molinspiration method, based on group contributions. The correlation equation between logKow and structural parameters or NPHOS was obtained respectively, and the stability analysis of the two equations was carried out and comparisons were made. The results show that the model with NPHOS as the theoretical descriptors has a higer fitting goodness due to the larger squared correlation coefficient R2(0.981) and the smaller standard error SE (0.069). Meanwhile, the logKow values of xanthone and3hydroxyl-substitued xanthones were determined with the slow-stirring method, the logKow correlation equation was then corrected using the experimental values, and the best model for the prediction of logKow was finally determined.