A team led by Min Deng of the State Key Laboratory of Molecular Oncology at the Cancer Hospital of the Chinese Academy of Medical Sciences, working with Zhenkun Lou of the Mayo Clinic and Bin Chen of Anhui University, has identified a tumor-intrinsic checkpoint that operates specifically in human cancer cells and is essentially absent from the murine models on which most radioimmunotherapy research is built. The circuit centers on three proteins: the farnesoid X receptor, or FXR, a nuclear receptor best known for its role in bile acid metabolism; TRIM22, an E3 ubiquitin ligase of the tripartite motif family; and STAT1, the transcription factor that sits at the heart of interferon signaling. In human tumor cells, the researchers found, radiation inadvertently activates this axis, and the result is a molecular strangulation of the very interferon program that radiation is supposed to unleash.

The mechanistic story begins with protein stability. The team showed that ionizing radiation stabilizes FXR in human cancer cells through the action of ubiquitin-specific peptidase 7, or USP7, a deubiquitinating enzyme that removes the ubiquitin tags that would otherwise mark FXR for proteasomal destruction. Once stabilized, FXR translocates its transcriptional activity toward a specific target: the gene encoding TRIM22. TRIM22, in turn, selectively binds STAT1 and catalyzes the attachment of K48-linked polyubiquitin chains, the canonical signal for degradation by the 26S proteasome. With STAT1 levels falling, the downstream interferon-stimulated gene program, the battery of ISGs that coordinates antigen presentation, T-cell recruitment, and inflammatory amplification, is extinguished precisely when radiation has done its work and the immune system should be mobilizing.

What makes this discovery particularly striking is its species specificity. When the researchers compared human and murine cancer cells, they found that mouse tumor cells simply lack a TRIM22 family member capable of destabilizing STAT1 in this way. Radiation therefore produces robust, sustained STAT1-dependent interferon signaling in mouse tumors, while the same insult in human cells is blunted by the FXR-TRIM22 circuit. This single proteostatic difference offers a mechanistic explanation for one of the most persistent frustrations in translational radiation oncology: the repeated failure of dramatic preclinical abscopal effects to translate into consistent clinical benefit. Decades of combination trials pairing radiation with immune checkpoint inhibitors have been designed around mouse data that, it now appears, were structurally incapable of revealing this human brake.

The consequences of dismantling the brake were demonstrated through genetic and pharmacologic means. When the team disrupted FXR or TRIM22 in human cancer cells, STAT1 was preserved, type I interferon signaling was amplified, and the irradiated tumor cells became markedly better at activating CD8-positive T cells. These effects were tested in peripheral blood mononuclear cell-humanized mouse models, which carry functional human immune components and human tumor cells, providing a far more faithful platform than conventional murine systems. In these humanized settings, loss of the FXR-TRIM22 axis enhanced abscopal tumor control, meaning that untreated tumors at distant sites regressed as the systemic immune response gathered strength, a phenomenon that has long been the holy grail of radiation-immunology research.

The pharmacologic arm of the study is where the work acquires its most immediate clinical resonance. Ursodeoxycholic acid, or UDCA, is a bile acid that has been used clinically for decades, primarily to treat cholestatic liver diseases and primary biliary cholangitis, with a long and well-characterized safety record. Prior work, including studies in morbidly obese patients, had suggested that UDCA exerts antagonistic effects on FXR signaling. The team reasoned that a safe, approved drug capable of dampening FXR might suppress the radiation-induced TRIM22 response and thereby preserve STAT1. That is exactly what they observed: UDCA treatment suppressed radiation-induced TRIM22 expression in human tumor cells, restored STAT1 protein levels and interferon-stimulated gene signaling, and potentiated the antitumor effects of radiotherapy in humanized models.

Crucially, the benefits extended to combination therapy. When UDCA was paired with radiotherapy and an anti-PD-1 immune checkpoint antibody in humanized mice, the triple combination produced superior tumor control without overt toxicity. This is a significant point, because the tolerability profile of UDCA is exceptionally well understood from its long hepatology track record, and any strategy that could enhance radioimmunotherapy using an already-approved, orally available drug would face a far shorter path to clinical testing than a purpose-built inhibitor of a novel target. The authors also note that the ubiquitin-proteasome machinery involved, including the E3 ligase TRIM22, belongs to target classes increasingly considered druggable in modern oncology, offering additional avenues if dedicated FXR or TRIM22 inhibitors are developed.

The study also contributes to a broader rethinking of how radiation and immunity interact in human tumors. It has long been appreciated that type I interferon signaling is a double-edged sword in cancer, with chronic or dysregulated interferon sometimes promoting immune escape, but the acute interferon burst following DNA damage is widely regarded as essential for effective antitumor immunity. By showing that a human-specific proteostatic checkpoint throttles that burst at the level of STAT1 stability, the work reframes the problem of radiation resistance: the issue is not that human tumors fail to sense DNA damage through cGAS-STING, but that a downstream transcriptional regulator is actively degraded before it can broadcast the alarm. Intervening at the level of protein stability, rather than at pathway activation, may therefore be the more effective strategy in patients.

Caveats remain, as they always do at this stage of translation. The evidence for the FXR-TRIM22-STAT1 axis and the efficacy of UDCA comes from cell lines and humanized mouse models, not yet from clinical trials, and the precise dosing, timing, and patient selection needed to deploy UDCA alongside radiation and checkpoint blockade in humans remain to be established. The authors declare no competing interests, and the RNA-sequencing data underpinning the transcriptional analyses have been deposited in public repositories for independent scrutiny. Still, the conceptual payoff is immediate: a human-biased mechanism that has been hiding in plain sight, masked by the very mouse models used to study it, now has a name, a circuit diagram, and a candidate drug. If the findings hold in patients, a decades-old bile acid could become an unexpected partner for radiation in the fight against cancer.

Subject of Research: A human-biased FXR-TRIM22-STAT1 immune checkpoint that limits radiation-induced antitumor immunity and its pharmacologic modulation by ursodeoxycholic acid

Article Title: Ursodeoxycholic acid enhances cancer radioimmunotherapy by releasing a human-biased FXR-TRIM22-STAT1 immune brake

Article References: Li, X., Jiang, P., Yin, H., Huo, Z., Xu, X., Nowsheen, S., Aziz, K., Zheng, S., Sun, N., Chen, B., Lou, Z., & Deng, M. (2026). Ursodeoxycholic acid enhances cancer radioimmunotherapy by releasing a human-biased FXR-TRIM22-STAT1 immune brake. Cell Research. https://doi.org/10.1038/s41422-026-01293-x

Image Credits: AI Generated

DOI: 10.1038/s41422-026-01293-x

Keywords: radiotherapy, radioimmunotherapy, FXR, TRIM22, STAT1, ursodeoxycholic acid, type I interferon, cGAS-STING, abscopal effect, CD8 T cells, ubiquitination, cancer immunotherapy