| Sumario: | Fritillaria thunbergii Miq. (Zhebeimu, ZBM) is traditionally recognized in Chinese medicine for its effects of clearing heat, resolving phlegm, suppressing cough, detoxifying, dissipating nodules, and resolving abscesses—properties that align closely with the pathogenesis of breast cancer. Classical texts including Shennong Bencao Jing and Bencao Zheng document its historical use in breast cancer treatment. This study aims to evaluate the potential therapeutic mechanisms and safety profile of the traditional Chinese medicine ZBM against triple-negative breast cancer (TNBC) using in silico methodologies. This computational study integrated network pharmacology, machine learning, and single-cell RNA sequencing to systematically identify ZBM's bioactive components and potential targets against TNBC. Molecular docking and dynamics simulations were employed to characterize interactions between key compounds and targets, while network toxicology assessed potential toxicity risks. Our multi-omics approach identified 42 bioactive ZBM components and 148 potential TNBC targets. Among these, machine learning algorithms prioritized five key autophagy-related genes, with CXCR4 selected as the core autophagy-associated target for computational validation. Molecular simulations predicted strong binding between hapepunine and CXCR4. Meanwhile, subtype analysis at the single-cell level revealed that BL TNBC subtypes may be particularly sensitive to ZBM compounds. Network toxicology revealed potential hepatotoxicity/nephrotoxicity risks. These risks were computationally mitigated through structural optimization of hapepunine derivatives. This study not only provides a novel mechanistic framework for ZBM's anti-TNBC activity but also demonstrates the utility of network toxicology coupled with structural optimization in proactively identifying and mitigating potential toxicity liabilities of natural product derivatives. [Display omitted] • Proposes potential anti-TNBC mechanism of Fritillaria thunbergii Miq., filling a key theoretical gap. • Integrated multi-omics deciphers pharmacology of Fritillaria thunbergii Miq. against TNBC. • Machine learning and scRNA-seq identify CXCR4 as Fritillaria thunbergii Miq's key target activating autophagy in TNBC. • Network toxicology and database mining reveal Fritillaria thunbergii Miq's multi-organ toxicity risks. • Computer-aided drug design-guided structural optimization reduces hapepunine's toxicity while enhancing its affinity for CXCR4.
|
|---|