TGF-β2 Rewires Colon Cancer Cells to Survive Doxorubicin Through Autophagy and HMGB1

Chemotherapy’s most stubborn adversary is not the tumor it fails to shrink on the first pass, but the population of cancer cells that quietly learns to survive the drug. A new study published in Medical Oncology by Sruthi Sritharan, Tamilselvan Jayavelu, and Nageswaran Sivalingam offers a detailed look at one such survival program in metastatic colon cancer. Working with the Colo205 cell line, a metastatic colon cancer model that carries a mutation in the SMAD4 gene, the team found that the signaling molecule transforming growth factor beta 2, or TGF-β2, appears to determine which molecular route colorectal cancer cells take toward resistance against doxorubicin, one of the most widely used and time-tested anticancer drugs in clinical oncology.

The significance of the finding lies in the tumor microenvironment. Colorectal cancer progression has long been associated with elevated levels of TGF-β in the blood of patients, and previous work by the same group had suggested that colon cancer cells themselves can secrete interleukin-6 and TGF-β2 in ways that may undermine chemotherapy. Earlier studies have also shown that TGF-β2 serves as a prognostic biomarker correlated with immune cell infiltration in colorectal cancer, and that higher frequencies of SMAD4 mutations occur in colorectal cancers with distant metastasis. What remained unclear was precisely how the presence of this ligand changes the cellular response to a cytotoxic drug, and whether it steers cells toward one resistance mechanism over another.

To answer that question, the researchers treated Colo205 cells with doxorubicin alone or in combination with TGF-β2 for seven days, then tracked a battery of cellular and molecular readouts over time. Cell proliferation was inhibited after the first day of treatment, while measurable loss of viability and cytotoxicity emerged after day three. Crucially, by the end of the seven-day window, no comparable loss of viability persisted, indicating that a fraction of the treated cells had adapted to the drug rather than succumbed to it. That temporal pattern, initial damage followed by recovery, is the classic fingerprint of acquired chemoresistance developing in a cell population under selective pressure.

One of the earliest and most striking differences appeared in the co-treated group. The multidrug resistance protein 1, known as MDR1, an efflux pump that expels chemotherapeutic agents from cells, was significantly elevated at both the gene and protein levels after just one day, but only in the cells that received both doxorubicin and TGF-β2. By days three and seven, all treatment groups showed elevation of MDR1, but the early, ligand-dependent surge suggests that TGF-β2 accelerates the deployment of the pump. The team confirmed this functionally using rhodamine 123 efflux assays, which measure how efficiently cells pump out a fluorescent dye handled by the same transporter, and assessed MDR1 at the gene, cell-surface, and total protein levels to build a consistent picture of enhanced drug export.

Beyond pumping drugs out, the surviving cells appeared to change their internal chemistry. The researchers detected induction of autophagy, the cellular recycling process in which cells digest their own components to survive stress, using three complementary approaches: acridine orange staining to visualize acidic organelles, monodansylcadaverine staining to label autophagic vacuoles, and measurement of the LC3II/I protein ratio, a standard molecular marker of autophagic flux. Autophagy has previously been implicated in doxorubicin resistance in breast cancer and hepatocellular carcinoma models, and HMGB1, the high mobility group box 1 protein, is known to drive autophagy-mediated resistance to doxorubicin through the AMPK/mTOR pathway in liver cancer cells. The new study places TGF-β2 upstream of this machinery in colon cancer.

The phenotypic shift extended to cell identity. After seven days of treatment, the co-treated cells showed increased expression of mesenchymal and stem cell markers, alongside downregulation of the Dab2 protein, a regulator linked to TGF-β signaling whose loss has been documented in pancreatic cancer progression and whose cleavage is known to regulate TGF-β-induced autophagy. The gain of mesenchymal traits, a process called epithelial-mesenchymal transition, is a well-established route by which epithelial tumors acquire both invasive capacity and drug tolerance. Notably, prior research has shown that SMAD4 mutations do not preclude epithelial-mesenchymal transition in colorectal cancer, which helps explain how a SMAD4-mutant line like Colo205 can still execute these TGF-β2-driven programs.

Perhaps the most unexpected result concerned HMGB1 and the transcription factor NF-κB. In the TGF-β2-containing groups, HMGB1 levels were significantly decreased, and the researchers observed arrest of NF-κB p65 nuclear translocation, meaning the inflammatory transcription factor failed to move into the nucleus where it would normally activate resistance genes. This is counterintuitive on the surface, since HMGB1 and NF-κB are often associated with pro-survival signaling and MDR1 regulation, and NF-κB inhibitors have been shown to revert multidrug resistance in breast cancer cells. The finding suggests that in this SMAD4-mutant, metastatic context, the TGF-β2 ligand reroutes the resistance program away from the canonical HMGB1-NF-κB axis and toward an autophagy- and stemness-dominated pathway, even though the study’s title frames HMGB1 as a participant in the process.

The authors interpret this constellation of results as evidence that the presence of the TGF-β2 ligand might determine which resistance pathway doxorubicin-treated metastatic colon cancer cells adopt. In practical terms, this matters because resistance is not a single switch but a branching decision tree, and different branches respond to different therapeutic countermeasures. If a patient’s tumor microenvironment is rich in TGF-β2, the cancer cells may be primed to survive doxorubicin through autophagy and stem-like phenotypes rather than through the routes a clinician might otherwise expect. Prior clinical research has explored TGF-β blockade as a strategy, including the anti-TGF-β monoclonal antibody fresolimumab in advanced melanoma and renal cell carcinoma, and animal work has shown that TGF-β blockade can normalize tumor stroma and improve drug distribution in breast carcinoma models.

The study also connects to a broader literature on colorectal cancer drug resistance. Doxorubicin, though more classically associated with breast cancer, leukemia, and sarcoma treatment, has served as a valuable experimental tool in colon cancer models, and previous work from the same laboratory demonstrated that TGF-β2 aggravates doxorubicin-induced chemoresistance in a Duke’s type B colon adenocarcinoma cell line through non-apoptotic cell death. Other groups have shown that chemoresistance to doxorubicin induces epithelial-mesenchymal transition via upregulation of TGF-β signaling in HCT116 colon cancer cells, and that targeting the stem cell marker CD133 reverses drug resistance via the AKT/NF-κB/MDR1 pathway in colorectal cancer. The Colo205 findings add a ligand-dependent dimension to this picture, implying that the same drug can push cells down different resistance routes depending on the signaling molecules floating in their surroundings.

As with all cell-line studies, the caveats are real. Colo205 is a single metastatic line with a defined SMAD4 mutation, and the seven-day treatment window captures early adaptation rather than the long evolution of resistance in a patient over months of therapy. The data were generated with established assays, including the MTT-based colorimetric method for proliferation and cytotoxicity introduced decades ago, and the autophagy measurements were cross-validated with multiple stains and the LC3II/I ratio, which strengthens confidence in the core observations. The authors report no conflicts of interest, and the datasets are available from the corresponding author upon reasonable request. The work was supported by a DST-INSPIRE fellowship from the Department of Science and Technology, Government of India, and conducted at SRM Institute of Science and Technology in Tamil Nadu.

The immediate value of the study is conceptual: it argues that chemoresistance in metastatic colon cancer is context-dependent, shaped by paracrine and autocrine TGF-β2 signaling, and funneled through autophagy, efflux pump upregulation, and stem-like reprogramming in ways that a SMAD4 mutation does not prevent. If follow-up work in additional cell lines, organoids, and ultimately patient tumors confirms that TGF-β2 levels predict which resistance pathway dominates, the finding could inform combination strategies that pair cytotoxic drugs with autophagy inhibitors or TGF-β pathway blockers, timed to the biology of the individual tumor. For now, the Colo205 experiments provide a mechanistic map of how a single signaling molecule in the tumor microenvironment can tip the balance between cell death and cell survival under chemotherapy, and they underscore why the microenvironment deserves equal billing with the cancer cell itself in the ongoing effort to make treatment stick.

Subject of Research: TGF-β2-driven autophagy-mediated doxorubicin resistance in metastatic colon cancer cells

Article Title: TGF-β2 promotes autophagy-mediated resistance to doxorubicin in metastatic colon cancer cells via HMGB1

Article References: Sritharan, S., Jayavelu, T., & Sivalingam, N. (2026). TGF-β2 promotes autophagy-mediated resistance to doxorubicin in metastatic colon cancer cells via HMGB1. Medical Oncology, 43(11), Article 298. https://doi.org/10.1007/s12032-026-03404-5

Image Credits: AI Generated

DOI: 10.1007/s12032-026-03404-5

Keywords: TGF-β2, doxorubicin, chemoresistance, autophagy, HMGB1, MDR1, colorectal cancer, NF-κB, SMAD4, epithelial-mesenchymal transition, cancer stem cells, tumor microenvironment