Marinković Marija, Nikolić Nemanja, Nikolić Tamara, Božanić Borislav, Topalović Marija, Rančić Andrija, Šajnović Stefan, Veselinović Aleksandar M
Pulmonology Clinic, University Clinical Center Niš, Niš, Serbia.
Policlinic "Dr Nikolić", Niš, Serbia.
J Mol Graph Model. 2025 Dec;141:109155. doi: 10.1016/j.jmgm.2025.109155. Epub 2025 Aug 20.
Endothelin-1 (ET-1), a potent vasoconstrictor peptide, plays a critical role in cardiovascular pathologies and remains a key target for therapeutic intervention. Despite its clinical significance, the development of selective and potent ET-1 antagonists continues to present major challenges. Computational methods, particularly Quantitative Structure-Activity Relationship (QSAR) modeling, offer a rational and efficient framework for designing such compounds. In this study, conformation-independent QSAR models were developed using molecular descriptors derived from SMILES notation and local molecular graph invariants. The Monte Carlo method was employed for descriptor selection and weight optimization, resulting in statistically robust models. Critical molecular fragments associated with antagonist activity were identified and applied in the computer-aided design (CAD) of new ET-1 inhibitors. The optimal QSAR model exhibited strong predictive performance, with high correlation coefficients for both the training set (r = 0.9362, q = 0.9314) and the test set (r = 0.9006, q = 0.8655). To further validate the structural plausibility of the designed molecules, molecular docking simulations were conducted against the ETA receptor. The docking results were in agreement with QSAR-predicted activity, revealing favorable binding poses, strong interaction energies, and consistent structure-activity trends across all six designed compounds. This methodological convergence strengthens the credibility of the in silico predictions. Additionally, computational analysis of physicochemical and pharmacokinetic parameters indicated favorable ADME profiles, high drug-likeness, and efficient gastrointestinal absorption, suggesting suitability for medicinal chemistry development. This study introduces a reliable and mechanistically interpretable computational pipeline for the discovery of novel ET-1 antagonists. The proposed compounds demonstrate promising pharmacological characteristics and represent viable candidates for future experimental validation. These findings underscore the value of integrated in silico strategies in accelerating cardiovascular drug discovery.
内皮素 -1（ET -1）是一种强效血管收缩肽，在心血管疾病中起关键作用，仍是治疗干预的关键靶点。尽管其具有临床意义，但开发选择性强效ET -1拮抗剂仍面临重大挑战。计算方法，特别是定量构效关系（QSAR）建模，为设计此类化合物提供了合理且高效的框架。在本研究中，使用源自SMILES符号和局部分子图不变量的分子描述符开发了构象无关的QSAR模型。采用蒙特卡罗方法进行描述符选择和权重优化，得到了统计稳健的模型。确定了与拮抗剂活性相关的关键分子片段，并将其应用于新型ET -1抑制剂的计算机辅助设计（CAD）。最佳QSAR模型表现出强大的预测性能，训练集（r = 0.9362，q = 0.9314）和测试集（r = 0.9006，q = 0.8655）的相关系数都很高。为了进一步验证所设计分子的结构合理性，针对ETA受体进行了分子对接模拟。对接结果与QSAR预测的活性一致，揭示了所有六种设计化合物的良好结合姿势、强相互作用能和一致的构效趋势。这种方法上的融合增强了计算机模拟预测的可信度。此外，对物理化学和药代动力学参数的计算分析表明其具有良好的ADME特性、高药物相似性和高效的胃肠道吸收，表明适合药物化学开发。本研究引入了一种可靠且可从机制上解释的计算流程，用于发现新型ET -1拮抗剂。所提出的化合物表现出有前景的药理学特性，是未来实验验证的可行候选物。这些发现强调了综合计算机模拟策略在加速心血管药物发现中的价值。
