Szatkowska-Wandas Paulina, Koba Marcin, Smolinski Grzegorz, Wandas Jacek
Department of Toxicology, Faculty of Pharmacy, Collegium Medicum of Bydgoszcz, Nicolaus Copernicus University, Torun, Poland.
Med Chem. 2016;12(6):592-600. doi: 10.2174/1573406411666151002130028.
The QSRRs and QSARs are relatively new approaches to relate internal chemical structure and particular biological activity. This methodology is based on theory that mechanisms which took place into chromatography column are similar to those that occur in a living organism at the molecular level, for example when compounds penetrate into cells.
In this paper, we aim to describe different cytostatic activities of selected anticancer drugs as QSRR and QSAR models, and prove usefulness of connected QSRR and QSAR methodology in different types of studies.
Chromatographic experiments using gradient RP-HPLC method and different C18 stationary phases were performed. As a result we obtained retention parameter log kw. Moreover, to calculate descriptors, which characterize lipophilicity of analyzed antitumor drugs, DryLab program was utilized. Molecular modeling studies were performed by using HyperChem program. Dragon software was used to calculate structural descriptors, and then selected descriptors were used to build QSRR and QSAR models. Obtained data were analyzed by multiple regression analysis (MLR).
Experimental log kw and predicted log kw from QSRR models developed, were further used in QSAR analysis. The goodness of fit was in the range of R2= 0.75 - 0.95, and the predictive performance of the models was Q2 = 0.6 - 0.81.
Both QSRR and QSAR strategies presented in this paper, allowed predicting HPLC retention parameters and cytotoxic activities of anticancer medicines without the necessity to carry out time-consuming and expensive experimental tests.
定量结构保留关系（QSRRs）和定量构效关系（QSARs）是将内部化学结构与特定生物活性相关联的相对较新的方法。该方法基于这样一种理论，即发生在色谱柱中的机制与在分子水平上活生物体中发生的机制相似，例如当化合物渗透到细胞中时。
在本文中，我们旨在将选定抗癌药物的不同细胞抑制活性描述为QSRR和QSAR模型，并证明相关的QSRR和QSAR方法在不同类型研究中的有用性。
使用梯度反相高效液相色谱法（RP-HPLC）和不同的C18固定相进行色谱实验。结果我们获得了保留参数log kw。此外，为了计算表征所分析抗肿瘤药物亲脂性的描述符，使用了DryLab程序。通过HyperChem程序进行分子建模研究。使用Dragon软件计算结构描述符，然后使用选定的描述符构建QSRR和QSAR模型。通过多元回归分析（MLR）对获得的数据进行分析。
从所开发的QSRR模型获得的实验log kw和预测log kw进一步用于QSAR分析。拟合优度在R2 = 0.75 - 0.95范围内，模型的预测性能为Q2 = 0.6 - 0.81。
本文提出的QSRR和QSAR策略都能够预测抗癌药物的高效液相色谱保留参数和细胞毒性活性，而无需进行耗时且昂贵的实验测试。
