FDA-Approved Antibiotics Show Promise Against Wnt-Driven Colorectal Cancer

Colorectal cancer remains one of the deadliest malignancies worldwide, and at the heart of most of these tumors lies a single misbehading molecular circuit: the Wnt/β-catenin signaling pathway. When this pathway runs unchecked, cells proliferate relentlessly, resist death, and acquire the invasive traits that make colorectal cancer so difficult to treat. Now, a team of Turkish researchers has taken aim at the critical molecular handshake that keeps this circuit switched on, and their results suggest that two familiar antibiotics, already sitting on pharmacy shelves, may be repurposed to break it.

In a study published in the journal Molecular Diversity, Kader Sahin of Bahçeşehir University and Bandırma Onyedi Eylül University, Ajda Coker of Marmara University, and Serdar Durdağı of Bahçeşehir University describe a structure-based virtual screening campaign that combed through FDA-approved drugs in search of molecules capable of disrupting the interaction between β-catenin and T-cell factor 4, or Tcf4. This protein–protein interaction is the final, decisive step in Wnt signaling: when β-catenin accumulates in the cell nucleus and binds Tcf4, the complex switches on a battery of genes that drive tumor growth, survival, metastasis, and drug resistance. Block that binding, the logic goes, and the entire oncogenic program collapses.

The catch has always been the shape of the target. The β-catenin/Tcf4 interface sprawls across roughly 4,800 square angstroms, a vast and shallow surface with few of the deep, well-defined pockets that medicinal chemists traditionally exploit. For years, interfaces like this one were written off as undruggable. Yet structural biology has gradually revealed that even sprawling protein–protein interfaces concentrate their binding energy in a handful of so-called hot spots, short stretches of residues that contribute disproportionately to the affinity between the two partners. If a small molecule can be steered into those hot spots, the reasoning goes, it might achieve with a compact drug what would otherwise require a much larger molecule.

That is precisely the strategy the researchers pursued. Rather than screening synthetic chemical libraries filled with untested compounds, they turned to drug repurposing, focusing on medicines that have already passed safety scrutiny and are in clinical use. Repurposing offers a practical shortcut: known pharmacology, established safety profiles, and in many cases existing manufacturing and distribution infrastructure. The team first mapped the critical hot spot residues of β-catenin that engage Tcf4, then used molecular docking to position each candidate drug within the interface, asking which approved molecules could physically occupy and destabilize the contact points that hold the complex together.

The computational pipeline did not stop at docking. To separate plausible binders from artifacts, the researchers subjected the top candidates to physics-based molecular dynamics simulations, watching in atomic detail how each ligand behaved over time when bound to β-catenin. They then calculated binding free energies using the MM/GBSA method, a hybrid approach that combines molecular mechanics energies with implicit solvent models to estimate how strongly a ligand clings to its target. This layered filtering, docking followed by simulation followed by energetic scoring, is designed to weed out compounds that look good in a single static snapshot but fall apart under the thermal motion of a real cellular environment.

Three candidates emerged from the screen with stable binding to the critical hot spot residues of β-catenin, specifically asparagine 430, lysine 435, histidine 470, arginine 474, and lysine 508. They were Leucovorin, also known as folinic acid, a vitamin derivative routinely used to enhance chemotherapy; Carbenicillin, a penicillin-class antibiotic; and Ceforanide, a second-generation cephalosporin. All three are approved, well-characterized medicines, which means that if any of them proves effective against Wnt-driven cancers, the path from bench to bedside could be considerably shorter than for a brand-new chemical entity.

Computational predictions, however convincing, mean little until they are tested in living cells. The team moved their three hits into the laboratory, using HT-29 colorectal cancer cells, a widely studied cell line in which Wnt signaling is aberrantly active. The cells were treated with each compound and assessed for viability and for their ability to form colonies, a measure of long-term proliferative capacity that often correlates with tumorigenic potential. The results were striking: Carbenicillin and Ceforanide reduced cell viability and impaired colony formation, exhibiting superior anti-proliferative activity compared with the other candidates, including Leucovorin.

The convergence between the computational and experimental arms of the study is what gives the findings their weight. The two antibiotics that performed best in cell assays were the same molecules whose simulated binding poses anchored them most securely to the β-catenin hot spots. This consistency suggests that the anti-proliferative effects observed in HT-29 cells may indeed arise from disruption of the β-catenin/Tcf4 interaction, although the researchers are careful to frame the work as a proof of concept rather than a finished therapy. Further studies will be needed to confirm the mechanism directly, to characterize selectivity for cancer cells over healthy tissue, and to establish whether the compounds can achieve therapeutic concentrations in tumors.

Even so, the implications are considerable. Wnt/β-catenin signaling is implicated not only in colorectal cancer but in a broad family of malignancies, including hepatocellular carcinoma and other Wnt-driven tumors, and it also maintains cancer stem cell populations that fuel relapse and metastasis. A validated small-molecule inhibitor of the β-catenin/Tcf4 interaction would be a genuinely new class of anticancer agent, one that attacks the transcriptional engine of the tumor rather than a single downstream symptom. Prior efforts, including peptide mimetics and dedicated small molecules developed in academic labs, have struggled to reach the clinic, which makes the repurposing angle particularly attractive: these are drugs whose toxicity, metabolism, and drug interactions are already documented.

The study also offers a template for tackling other supposedly undruggable targets. By combining hot spot analysis, docking against a library of approved drugs, rigorous molecular dynamics validation, and energetic scoring, and only then moving to cell-based experiments, the researchers demonstrated a workflow that is both computationally disciplined and experimentally grounded. As the authors conclude, their findings provide compelling evidence that targeting the β-catenin/Tcf4 interaction with small molecules is feasible, and that Carbenicillin and Ceforanide may serve as promising repurposed candidates for therapeutic intervention in Wnt-driven malignancies. For patients with colorectal cancer, whose treatment options too often narrow as the disease progresses, the idea that answers might be hiding in plain sight among existing medicines is a hopeful one, and this study shows exactly how such hidden answers can be systematically found.

Subject of Research: Virtual drug repurposing to discover β-catenin/Tcf4 interaction inhibitors for colorectal cancer treatment

Article Title: Hot spot-driven discovery of β-catenin/Tcf4 interaction inhibitors via virtual drug repurposing study and experimental validation in colorectal cancer cells

Article References: Sahin, K., Coker, A., & Durdağı, S. (2026). Hot spot-driven discovery of β-catenin/Tcf4 interaction inhibitors via virtual drug repurposing study and experimental validation in colorectal cancer cells. Molecular Diversity. https://doi.org/10.1007/s11030-026-11720-1

Image Credits: AI Generated

DOI: 10.1007/s11030-026-11720-1

Keywords: β-catenin, Tcf4, Wnt signaling, colorectal cancer, drug repurposing, virtual screening, molecular docking, molecular dynamics, MM/GBSA, Carbenicillin, Ceforanide, protein-protein interaction inhibitors