KRAS-mutant tumors flood themselves with adenosine to blind immune cells

One of the most stubborn problems in modern oncology may have just given up a crucial secret. Cancers driven by mutations in the KRAS gene—among the most common genetic lesions in pancreatic, colorectal, and lung cancers—are notoriously resistant to immunotherapy drugs that work brilliantly in other tumor types. Now, a team of researchers reporting in Molecular Cancer has traced that resistance to a surprising source: a metabolic sabotage scheme in which mutant KRAS floods the tumor microenvironment with adenosine, an immunosuppressive molecule that effectively puts nearby T cells to sleep. The finding, published as an open-access study by a team led by Jing Yuan, Wenwen Ding, Yuhong Zhang, and Rui Liu at Sun Yat-sen University Cancer Center, together with collaborators including Peng Huang, Panpan Liu, Shijun Wen, Christophe Nicot, Paul J. Chiao, and Christophe Glorieux, not only explains a long-standing clinical puzzle but also points to concrete pharmacological strategies for dismantling it.

The logic of the discovery is elegant in its biochemical economy. Adenosine is a well-known signaling molecule in healthy tissue, where it accumulates during periods of stress, inflammation, or hypoxia and helps dampen excessive immune responses to prevent collateral tissue damage. Tumors have long been suspected of hijacking this safety valve, but the question of how a specific oncogene might actively orchestrate adenosine accumulation had remained poorly defined. The new study answers that question with unusual precision, showing that oncogenic KRAS attacks adenosine metabolism from two directions at once—boosting the machinery that produces the molecule while simultaneously disabling the machinery that destroys it.

On the production side, the researchers found that mutant KRAS acts through the BRAF signaling pathway to upregulate ecto-5′-nucleotidase, an enzyme encoded by the NT5E gene and better known by its alternative name, CD73. This enzyme sits on the outer surface of cancer cells and performs the final step in a cascade that begins with extracellular ATP, the energy currency released by dying and stressed cells. Enzymes such as ENTPD1 strip phosphate groups from extracellular ATP to yield AMP, and NT5E then cleaves the remaining phosphate to liberate free adenosine directly into the extracellular space. By cranking up NT5E expression, KRAS-mutant cells essentially convert the debris of tumor turnover into a potent immunosuppressant, manufacturing adenosine right at the front line where immune cells patrol.

Production alone, however, would not be enough if the accumulating adenosine were quickly degraded. This is where the second arm of the scheme comes in. The team demonstrated that KRAS signaling, acting through the AKT pathway, suppresses adenosine deaminase, or ADA—the enzyme responsible for breaking adenosine down into inosine. The consequence of this coordinated manipulation is pathological: adenosine is produced faster and cleared more slowly, allowing it to build up to concentrations that would never be tolerated in healthy tissue. The researchers quantified this accumulation using liquid chromatography–tandem mass spectrometry, confirming that adenosine levels rise dramatically in KRAS-driven tumors compared with their non-malignant or KRAS-wildtype counterparts.

The functional consequences for the immune system are exactly what one would predict from adenosine’s known biology. When T cells encounter high concentrations of extracellular adenosine, they engage A2A adenosine receptors on their surface, which triggers intracellular signaling that blunts the production of interferon-gamma and interleukin-2, the cytokines that drive effective antitumor attack. Proliferation slows, cytotoxic activity wanes, and the T cells that should be tearing into the tumor instead drift toward a dysfunctional, exhausted state. Using immune–tumor co-culture assays, the team showed that adenosine-rich conditioned media from KRAS-mutant cells markedly impaired T-cell function, and that this impairment could be reversed when adenosine was removed from the equation.

Crucially, the study did not stop at cell culture. The investigators employed inducible KRAS cell models, human tumor specimens, and syngeneic mouse models to establish that the adenosine axis operates in living tumors. Flow cytometry and immunohistochemistry revealed that adenosine-laden KRAS tumors were infiltrated by fewer functional T cells and harbored a more immunosuppressive microenvironment overall. Analysis of human tumor samples corroborated the mouse data, showing the expected relationship between KRAS-driven signaling, NT5E and ADA expression, and the immune landscape of the tumors. The consistency across model systems and human tissue lends considerable weight to the conclusion that this is not a laboratory artifact but a genuine feature of KRAS-mutant cancers in patients.

The therapeutic implications are where the study becomes truly exciting. The researchers tested three complementary ways of attacking the adenosine axis. First, genetic silencing of NT5E in tumor cells restored antitumor immunity, demonstrating that the ectonucleotidase is a viable point of intervention. Second, and perhaps most creatively, they deployed nanoparticles loaded with adenosine deaminase itself—packaging the very enzyme that KRAS-mutant tumors suppress into poly(lactic-co-glycolic acid) delivery vehicles that could enzymatically deplete adenosine in the tumor microenvironment. The strategy is conceptually striking: rather than blocking a receptor or inhibiting a kinase, it simply replenishes the missing metabolic cleanup crew. When ADA nanoparticles converted adenosine into inosine, they performed a kind of metabolic alchemy, transforming an immunosuppressive metabolite into an immune-stimulatory one, since inosine has been shown in prior work to fuel T cell activity rather than dampen it.

Third, the team turned to pharmacological blockade of the adenosine A2A receptor, the molecular dock where adenosine delivers its calming signal to immune cells. A2AR inhibition suppressed tumor growth in mouse models through a dual mechanism that the authors highlight as particularly attractive: the drug simultaneously enhanced immune activation by preventing adenosine from silencing T cells and directly impaired the viability of KRAS-driven cancer cells themselves. That second, tumor-intrinsic effect suggests that KRAS-mutant cells may become dependent on adenosine signaling in ways that extend beyond immune manipulation, raising the possibility that A2AR antagonists could deliver a one-two punch against this tumor type.

Perhaps the most clinically resonant result came from combining these metabolic interventions with PD-1 checkpoint blockade, the immunotherapy class that has transformed treatment for many cancers but frequently fails in KRAS-mutant disease. In the syngeneic mouse models, both genetic NT5E silencing and ADA nanoparticle treatment potentiated the efficacy of anti-PD-1 antibodies, converting a weak or absent response into a robust one. The interpretation is straightforward: checkpoint inhibitors release the brakes on T cells, but if those T cells are metabolically sedated by adenosine, releasing the brakes accomplishes little. By clearing adenosine from the microenvironment first, the metabolic interventions allow the checkpoint blockade to function as intended. This framing positions adenosine targeting as a rational partner for existing immunotherapy rather than a replacement for it.

The study, published on 13 October 2026 in Molecular Cancer with the DOI 10.1186/s12943-026-02810-7, was supported by funding from the National Natural Science Foundation of China and several institutional programs, and the authors declare no competing interests. As with any preclinical work, the road from mouse models to patients is long, and the safety and pharmacology of ADA nanoparticles in humans remain to be established. But the conceptual contribution is substantial: a previously unrecognized KRAS-controlled adenosine metabolic axis now stands revealed as a driver of immune evasion, and three distinct points along that axis—NT5E, ADA, and A2AR—offer druggable targets. For the large population of patients whose KRAS-mutant tumors have shrugged off immunotherapy, the prospect of restoring T-cell function simply by correcting a metabolic imbalance is a compelling one, and it ensures that adenosine metabolism will be watched closely as a next frontier in cancer immunotherapy research.

Subject of Research: KRAS-driven adenosine metabolic reprogramming and immune evasion in cancer

Article Title: Oncogenic KRAS drives immune evasion through adenosine metabolic reprogramming mediated by NT5E and ADA

Article References: Oncogenic KRAS drives immune evasion through adenosine metabolic reprogramming mediated by NT5E and ADA. (n.d.). https://doi.org/10.1186/s12943-026-02810-7

Image Credits: AI Generated

DOI: 10.1186/s12943-026-02810-7

Keywords: KRAS, adenosine, NT5E, ADA, A2AR, immune evasion, tumor microenvironment, PD-1 blockade, cancer metabolism, T cells, immunotherapy resistance, inosine