Institute of Pharmaceutical Biology and Phytochemistry (IPBP), University of Muenster, PharmaCampus Corrensstrasse 48, D-48149 Muenster, Germany.
Department of Chemistry, Egerton University, P.O. Box 536, Egerton 20115, Kenya.
Int J Mol Sci. 2018 Nov 22;19(12):3721. doi: 10.3390/ijms19123721.
Three complementary quantitative structure⁻activity relationship (QSAR) methodologies, namely, regression modeling based on (i) "classical" molecular descriptors, (ii) 3D pharmacophore features, and (iii) 2D molecular holograms (HQSAR) were employed on the antitrypanosomal activity of sesquiterpene lactones (STLs) toward , the causative agent of the East African form of human African trypanosomiasis. In this study, an extension of a previous QSAR study on 69 STLs, models for a much larger and more diverse set of such natural products, now comprising 130 STLs of various structural subclasses, were established. The extended data set comprises a variety of STLs isolated and tested for antitrypanosomal activity within our group and is furthermore enhanced by 12 compounds obtained from literature, which have been tested in the same laboratory under identical conditions. Detailed QSAR analyses yielded models with comparable and good internal and external predictive ability. For a set of compounds as chemically diverse as the one under study, the models exhibited good coefficients of determination (R²) ranging from 0.71 to 0.85, as well as internal (leave-one-out Q values ranging from 0.62 to 0.72) and external validation coefficients (² values ranging from 0.54 to 0.73). The contributions of the various tested descriptors to the generated models are in good agreement with the results of previous QSAR studies and corroborate the fact that the antitrypanosomal activity of STLs is very much dependent on the presence and relative position of reactive enone groups within the molecular structure but is influenced by their hydrophilic/hydrophobic properties and molecular shape.
三种互补的定量构效关系（QSAR）方法，即基于（i）“经典”分子描述符、（ii）3D 药效团特征和（iii）2D 分子全息图（HQSAR）的回归建模，被用于研究倍半萜内酯（STLs）对非洲锥虫病东非型病原体的抗锥虫活性。在本研究中，对先前关于 69 种 STLs 的 QSAR 研究进行了扩展，建立了更大、更多样化的此类天然产物的模型，现在包括 130 种具有各种结构子类的 STLs。扩展数据集包括我们小组分离并测试抗锥虫活性的各种 STLs，并且通过 12 种从文献中获得的化合物进行了增强，这些化合物在相同的实验室中在相同的条件下进行了测试。详细的 QSAR 分析产生了具有可比和良好内部和外部预测能力的模型。对于一组像研究中那样化学多样性的化合物，模型表现出良好的决定系数（R²），范围从 0.71 到 0.85，以及内部（留一法 Q 值），范围从 0.62 到 0.72 和外部验证系数（²值），范围从 0.54 到 0.73。各种测试描述符对生成模型的贡献与先前 QSAR 研究的结果非常一致，并证实了 STLs 的抗锥虫活性非常依赖于分子结构中反应性烯酮基团的存在和相对位置，但受其亲水性/疏水性性质和分子形状的影响。
