Mechanism-specific identification of antihypertensive phytochemicals targeting CACNA1C, ADRB1, and AGTR1 through...

Natural products are a valuable source of antihypertensive agents; however, most studies have focused on multitarget effects or extract-level activity, limiting mechanism-specific interpretation and drug development. Here, a structure-driven screening strategy was applied to identify phytochemicals targeting CACNA1C, ADRB1, and AGTR1, key pharmacological targets of antihypertensive therapy. Quantitative structure-activity relationship models were developed using ChEMBL datasets (11,452 for CACNA1C, 612 for ADRB1, and 291 for AGTR1) and demonstrated robust predictive performance (AUC = 0.973), which was maintained under scaffold split and cross-validation. The SHAP-based interpretation revealed target-specific structural features, followed by pharmacophore profiling, chemical space analysis, and network-based prioritization. Overall, 10,586 natural compounds were screened and the top 5% per target (530 compound-target edges) were selected, revealing that 83.5% of the compounds were target-specific. Chemical space analysis revealed the predicted compounds occupied diverse and partially orthogonal regions compared with approved drugs, indicating scaffold diversification. Representative compounds selected based on docking, ADMET, and structural criteria predicted to form stable binding interactions in 100 ns molecular dynamics simulations, with consistent RMSD stabilization and favorable MM-PBSA binding free energies (-29.17 to -36.59 kJ/mol). Although source plants have reported antihypertensive effects, direct evidence in the literature linking individual compounds to specific pharmacological targets remains limited. Overall, this study presents a multilayer validation framework for the computational prioritization of antihypertensive phytochemicals, facilitating their potential integration of natural products into established pharmacological paradigms. This approach provides a strategy for prioritizing structurally diverse and computationally target-aligned candidates and may facilitate the translation of traditional herbal knowledge into modern target-based drug discovery.