2. Academic Validation
  3. Amentoflavone from Ginkgo biloba inhibits EMT-driven lung cancer metastasis by targeting TGFBR2: Integrative network pharmacology, machine learning, and experimental validation

Amentoflavone from Ginkgo biloba inhibits EMT-driven lung cancer metastasis by targeting TGFBR2: Integrative network pharmacology, machine learning, and experimental validation

  • J Ethnopharmacol. 2026 Mar 25:359:121061. doi: 10.1016/j.jep.2025.121061.
Kaile Liu 1,  Kaiting Chen 1,  Xiaoxue Zhao 1,  Yahui Zhang 2,  Yuejiao Cai 1,  Xiaojie Fu 1,  Zhongqi Wang 3,  Juying Jiao 4,  Haibin Deng 5
Affiliations

Affiliations

  • 1 Department of Oncology, Longhua Hospital, Shanghai University of Traditional Chinese Medicine, Shanghai, 200032, China.
  • 2 Shanghai Seventh People's Hospital, Shanghai University of Traditional Chinese Medicine, Shanghai, 200137, China.
  • 3 Department of Oncology, Longhua Hospital, Shanghai University of Traditional Chinese Medicine, Shanghai, 200032, China. Electronic address: [email protected].
  • 4 Department of Oncology, Longhua Hospital, Shanghai University of Traditional Chinese Medicine, Shanghai, 200032, China. Electronic address: [email protected].
  • 5 Department of Oncology, Longhua Hospital, Shanghai University of Traditional Chinese Medicine, Shanghai, 200032, China. Electronic address: [email protected].
PMID: 41412234 DOI: 10.1016/j.jep.2025.121061
Abstract

Ethnopharmacological relevance: In non-small cell lung Cancer (NSCLC), fatal outcomes predominantly result from metastatic dissemination, highlighting an urgent need for metastasis-directed therapies. Amentoflavone (AMF), a natural biflavonoid found in Ginkgo biloba and Other medicinal Plants, exhibits known antitumor activity. However, its role in NSCLC metastasis and the underlying mechanisms remain poorly understood.

Aim of the study: This study aimed to investigate the anti-metastatic effects of AMF from Ginkgo biloba in NSCLC and to elucidate the underlying molecular mechanisms.

Materials and methods: A network pharmacology approach combined with machine learning was used to predict AMF targets and derive a core diagnostic gene signature. Molecular docking and 100-ns molecular dynamics simulations were performed to characterize AMF-target binding and stability. In vitro, NSCLC cell migration was assessed using wound-healing (scratch) assays and Transwell migration chambers, with mechanistic insights obtained by Western blot analysis of EMT-related proteins. In vivo efficacy was evaluated in a murine pulmonary metastasis model established via tail-vein tumor cell injection. The machine learning model achieved an AUC of 0.951 in the internal testing set and 0.814 in the external validation set (GSE31210), indicating high robustness.

Results: Convergent computational analyses identified TGFBR2 as a key candidate target of AMF. Docking and 100-ns MD simulations predicted a strong and dynamically stable binding mode of AMF to TGFBR2. In vitro, AMF significantly reduced NSCLC cell migration, accompanied by downregulation of TGFBR2 and inhibition of SMAD2/3 phosphorylation, which collectively reversed epithelial-mesenchymal transition (EMT). Consistently, AMF suppressed NSCLC metastasis in vivo by inhibiting the TGF-β/Smad signaling pathway.

Conclusion: Our data indicate that AMF functions as an anti-metastatic agent in NSCLC at least in part by downregulating TGFBR2, thereby dampening TGF-β/Smad signaling and preventing EMT. These findings highlight AMF as a promising natural candidate for developing anti-metastatic therapies that appear well tolerated under the conditions tested.

Keywords

AMF; EMT; Metastasis; NSCLC; TGFBR2.

Figures
Products