ISSN print edition: 0366-6352 ISSN electronic edition: 1336-9075 Registr. No.: MK SR 9/7 Published monthly

Bioisosteric coumarin–pyrimidine hybrids for multi-target breast-cancer inhibition: an integrated in Silico study

Veda B. Hacholli, M. R. Shubha, Łukasz Szeleszczuk, B. H. Prabhanjan, Abhishek Kumar, and Marcin Gackowski

Department of Pharmaceutical Chemistry, Nitte College of Pharmaceutical Sciences (Nitte Deemed to be University), Bengaluru, India

E-mail: abhi12bunty@nitte.edu.in

Abstract:

Breast cancer heterogeneity and resistance motivate the development of polypharmacological agents that engage multiple oncogenic nodes within a single chemotype. We designed and evaluated 64 coumarin–pyrimidine hybrids—organized into hydroxy- and sulfhydryl-pyrimidine series—against PARP-1, EGFR, HER2, and BCL-2 using an integrated in silico workflow. Molecular docking provided primary ranking and top docking affinities reached − 8.7 kcal/mol for PARP-1, − 10.6 kcal/mol for EGFR, − 10.1 kcal/mol for HER2, and − 8.1 kcal/mol for BCL-2, exceeding the reference ligands (olaparib − 6.2; erlotinib − 7.3; lapatinib − 6.9; doxorubicin − 7.0 kcal/mol). Molecular dynamics (MD) simulations (200 ns) were used to assess pose persistence and protein compactness. MM-GBSA rescoring on MD snapshots refined binding energetics; MM-GBSA ΔG_bind values were as favorable as − 51.10 kcal/mol (5 m–EGFR) and − 50.94 kcal/mol (9 s–HER2). Time-resolved MM-PBSA profiles monitored the stability of binding free energy along trajectories, and pharmacophore modeling rationalized key interaction features. Hydroxy-pyrimidines generally outperformed sulfhydryl analogues at PARP-1 and in kinase pockets, with para electron-withdrawing and heteroaryl substituents strengthening hinge-directed hydrogen bonding, π-stacking, and lipophilic packing. Four complexes—5 h–4R6E (PARP-1), 6f–1M17 (EGFR), 9 s–3RCD (HER2), 10d–4IEH (BCL-2)—showed consistent performance across methods, characterized by low MD fluctuations, compact radius-of-gyration ranges, persistent hydrogen bonds, and persistently favorable MM-PBSA energy profiles. Time-averaged MM-PBSA binding free energies were approximately − 140 to − 160 kJ/mol across the four systems. In silico ADMET supported their developability with molecule-specific considerations: 5 h (potent, but low predicted oral bioavailability), 6f (drug-like, predicted gastrointestinal absorption and potential blood–brain barrier penetration), 9 s (balanced profile with predicted BCRP (breast cancer resistance protein) interaction), and 10d (strong binding with hepatotoxicity alerts to de-risk). These findings nominate 5 h, 6f, 9 s, and 10d as prioritized computational leads for synthesis and experimental validation in relevant breast cancer models.

Full paper is available at www.springerlink.com.

DOI: 10.1007/s11696-026-04721-9

Chemical Papers 80 (5) 5415–5444 (2026)