Tumor Fibroblasts Drive Chemo Resistance in Gastric Cancer via GAS6-AXL-ABCG1 Pathway
Cancer-associated fibroblasts secrete GAS6, activating AXL/STAT3/ABCG1 to shield gastric cancer cells from chemotherapy — and an AXL inhibitor reverses it.
Summary
Researchers at Ajou University discovered that cancer-associated fibroblasts (CAFs) — the connective tissue cells surrounding tumors — secrete a protein called GAS6 that makes gastric cancer cells resistant to chemotherapy. GAS6 activates a receptor called AXL on cancer cells, which triggers a signaling chain through STAT3 that ramps up a drug-efflux pump called ABCG1, physically expelling chemotherapy drugs before they can kill the cancer. Blocking this pathway with a selective AXL inhibitor called 9im restored chemotherapy sensitivity in lab and animal models. In patient tissue samples, high ABCG1 expression combined with high CAF marker levels strongly predicted poor survival outcomes. The findings identify a targetable mechanism for overcoming one of gastric cancer's most stubborn treatment problems.
Detailed Summary
Gastric cancer remains one of the deadliest malignancies worldwide, and a major reason patients fail chemotherapy is the tumor microenvironment — specifically the dense network of cancer-associated fibroblasts (CAFs) that infiltrate gastric tumors and actively communicate with cancer cells. This study from Ajou University School of Medicine set out to decode one key signaling axis mediating this communication: GAS6 (Growth Arrest-Specific 6), a secreted protein, and its receptor AXL, a receptor tyrosine kinase.
Using RNA-in situ hybridization (RNA-ISH) of clinical gastric cancer tissue samples, the researchers confirmed that GAS6 expression was markedly elevated in CAFs compared to cancer epithelial cells. They then used co-culture transwell systems to model CAF-GC cell interactions, demonstrating that conditioned medium from CAFs significantly promoted gastric cancer cell migration and increased resistance to standard chemotherapy agents (5-fluorouracil and oxaliplatin) compared to cancer cells cultured alone. These effects were abrogated when cells were treated with 9im, a highly selective small-molecule AXL inhibitor, confirming the GAS6/AXL axis as the responsible mechanism.
Transcriptome analysis of gastric cancer cells exposed to CAF-conditioned medium identified ABCG1 (ATP-binding cassette transporter G1) as a top upregulated gene associated with chemoresistance. Luciferase reporter assays demonstrated that STAT3 — a transcription factor activated downstream of AXL — directly binds to the ABCG1 promoter and drives its expression. Pharmacological inhibition of STAT3 reduced ABCG1 protein levels, and drug efflux assays confirmed that ABCG1 upregulation enabled gastric cancer cells to actively pump out chemotherapy agents. Treatment with 9im suppressed STAT3 phosphorylation, reduced ABCG1 expression, and restored intracellular drug accumulation, thereby re-sensitizing cancer cells to chemotherapy.
In vivo mouse xenograft experiments showed that combination treatment with 9im and chemotherapy produced significantly greater tumor growth inhibition than either agent alone, validating the translational potential of AXL inhibition as a chemosensitizing strategy. Tumor volumes in the combination group were substantially smaller than monotherapy groups, with acceptable tolerability.
Clinical relevance was assessed using public transcriptomic databases and patient cohorts. High ABCG1 expression correlated significantly with poor overall survival in gastric cancer patients. Crucially, co-expression of ABCG1 with established CAF markers (such as FAP and αSMA) identified a subgroup with particularly dismal prognosis, suggesting that the GAS6-AXL-STAT3-ABCG1 axis is clinically active and prognostically significant. These findings position ABCG1 as a potential stratification biomarker and AXL inhibitors as rational combination partners for chemotherapy in CAF-rich gastric cancers. The authors acknowledge that further clinical trials are needed to translate these findings into patient benefit.
Key Findings
- GAS6 was significantly overexpressed in cancer-associated fibroblasts (CAFs) compared to gastric cancer epithelial cells, confirmed by RNA-ISH in patient tissue samples.
- CAF-conditioned medium increased gastric cancer cell resistance to 5-fluorouracil and oxaliplatin; this resistance was reversed by the selective AXL inhibitor 9im.
- Transcriptome analysis identified ABCG1 as the top upregulated chemoresistance gene in gastric cancer cells exposed to CAF-derived GAS6/AXL signaling.
- Luciferase assays confirmed direct STAT3 binding to the ABCG1 promoter; STAT3 inhibition reduced ABCG1 protein expression and intracellular drug efflux.
- In vivo xenograft experiments showed combination 9im + chemotherapy produced significantly greater tumor suppression than either monotherapy alone.
- High ABCG1 expression in patient datasets correlated with significantly worse overall survival in gastric cancer.
- Co-expression of ABCG1 with CAF markers (FAP, αSMA) identified a patient subgroup with the poorest prognosis, validating the clinical relevance of the GAS6-AXL-STAT3-ABCG1 axis.
Methodology
The study used in vitro transwell co-culture systems, western blotting, cell viability and drug efflux assays, flow cytometry, and transcriptome (RNA-seq) analysis to characterize CAF-induced chemoresistance in gastric cancer cell lines. Mechanistic validation of STAT3-ABCG1 regulation used luciferase promoter reporter assays. Clinical relevance was assessed via RNA-ISH of patient-derived gastric cancer tissue samples and analysis of public transcriptomic survival databases. In vivo validation used mouse xenograft tumor models treated with 9im alone, chemotherapy alone, or the combination; statistical comparisons used standard parametric and nonparametric tests appropriate to each assay.
Study Limitations
The study is primarily preclinical, relying on cell lines and mouse xenograft models that may not fully recapitulate the complexity of human gastric tumors or patient immune contexts. Clinical survival correlations are based on retrospective public database analyses rather than a prospectively validated cohort with matched treatment data. The authors do not report specific conflicts of interest beyond institutional funding from the Korean National Research Foundation and a Korea-US Collaborative Cancer R&D Program.
Enjoyed this summary?
Get the latest longevity research delivered to your inbox every week.
Enter your email to subscribe: