Immune-cell switch flips the metabolic war between tumors and T cells in immunotherapy-resistant cancers

Tumors are not simply masses of renegade cells multiplying without opposition. They are ecosystems, miniature economies in which cancer cells and immune cells compete for the same scarce metabolic resources inside a hostile, immunosuppressive landscape. In work published in Nature Immunology, researchers led by Semir Beyaz of Cold Spring Harbor Laboratory report that the lipid-sensing transcription factor PPARδ acts as a decisive arbiter of this competition, and that the outcome depends entirely on which side of the battlefield activates it. When PPARδ is switched on in cancer cells, tumors grow faster and immunity weakens. When it is switched on in immune cells, tumors shrink, T cells flood into previously impenetrable cancers, and therapies that normally fail against immunotherapy-resistant tumors begin to work.

The finding addresses one of the most stubborn problems in oncology: colorectal cancer that is microsatellite stable (MSS). Most colorectal tumors fall into this category, in contrast to the smaller subset with microsatellite instability that responds well to immune checkpoint blockade. MSS metastatic colorectal cancer resists both checkpoint inhibitors and chimeric antigen receptor (CAR) T cell therapy, and its incidence in people under fifty has been rising to the point that colorectal cancer has become a leading cause of cancer death in young adults in the United States. The therapeutic challenge, the authors argue, is not merely to remove inhibitory signals but to equip immune cells with the metabolic and epigenetic fitness required to function inside immune-evasive tumors.

To dissect the compartment question, the team built genetically engineered mouse models in which a constitutively active VP16/VP64–PPARδ fusion protein was overexpressed either in cancer cells or across the immune compartment using Vav1-cre. When PPARδ activity was forced in MC38 colorectal adenocarcinoma cells, tumors grew more aggressively, consistent with prior reports that cancer-intrinsic PPARδ promotes tumorigenesis. The striking result came when the same pathway was activated in immune cells: growth of both control tumors and PPARδ-activated tumors was suppressed. Single-cell RNA sequencing revealed that immune-intrinsic activation drove expression of apoptosis, anti-proliferative and immune recognition genes within cancer cells, and reversed the expression and chromatin accessibility of exhaustion markers such as Ctla4, Pdcd1, Havcr2 and Lag3 in tumor-infiltrating CD8⁺ T cells. Crucially, when PPARδ was active in both compartments simultaneously, the immune cells won; the T cell state resembled that seen with immune activation alone.

The researchers then moved to a far more demanding model. They injected genetically engineered Apc-null, Kras^G12D, Tp53-null (AKP) organoids, and more aggressive AKPS organoids lacking Smad4, directly into the colon submucosa of mice. These orthotopic tumors recapitulate the histopathology and immune evasion of human colorectal cancer, and AKPS tumors spontaneously metastasize to the liver at high frequency, mimicking metastatic disease. Immune-intrinsic PPARδ activation restricted the growth of these tumors, and adoptive transfer of splenocytes from PPARδ-activated mice curtailed both primary growth and metastatic spread. The effect was not limited to colorectal cancer: B16-F10 melanoma and EO771 breast tumors were similarly restrained, indicating that reinforcing the immune compartment can generalize across cancer types.

Mechanistically, PPARδ-activated immune cells remodeled the tumor microenvironment from an immunosuppressive to an immunostimulatory state. Single-cell profiling showed upregulation of genes involved in antigen presentation, survival, chemokine and interferon signaling across immune cell populations. Immunofluorescence imaging revealed increased infiltration of CD45⁺ immune cells and the formation of ‘immune hubs’, multicellular aggregates previously linked to robust antitumor responses in human colorectal cancer, even in the poorly infiltrated MSS tumors that normally exclude lymphocytes. Treated tumors were also less fibrotic, suggesting resolution of tumor-related inflammation. Antigen receptor sequencing showed that T cells and B cells within normally immune-evasive AKPS tumors underwent clonal expansion under PPARδ activation, a hallmark of a genuine, targeted adaptive immune response rather than nonspecific inflammation.

The CD8⁺ T cell emerged as the central effector. Depleting these cells erased the tumor suppression conferred by immune PPARδ activation, while selectively activating PPARδ only in CD8⁺ T cells was sufficient to slow tumor growth and enrich for polyfunctional T cells producing interferon gamma, tumor necrosis factor and granzyme B. Single-cell ATAC sequencing showed that PPARδ opened chromatin at effector-associated loci including Prf1, Gzmb, Ccl4 and Klrg1, with enrichment of AP-1 transcription factor motifs, a chromatin state associated with T cell potency and resistance to exhaustion. The chemokine receptor CXCR3 was upregulated, and depleting CXCR3⁺ cells abolished much of the antitumor benefit, identifying the CXCL9/CXCL10–CXCR3 axis as a required route for effector T cell recruitment into tumors.

On the metabolic side, PPARδ reprogrammed T cell lipid handling through carnitine palmitoyltransferase 1a (Cpt1a), the rate-limiting enzyme that shuttles fatty acids into mitochondria for oxidation. Mass spectrometry revealed reduced carnitine and enhanced fatty acid oxidation in activated T cells, while MALDI mass spectrometry imaging of tumors showed depletion of fatty acid species such as palmitic, palmitoleic and oleic acid, consistent with metabolically empowered immune cells outcompeting cancer cells for nutrients. Genetic ablation of Cpt1a in immune cells impaired PPARδ-driven fat oxidation, reduced CD8⁺ T cell infiltration and weakened antitumor immunity, whereas a malonyl-CoA-insensitive Cpt1a gain-of-function mutant boosted oxidation and promoted memory-associated transcriptional programs but did not fully reproduce the effector phenotype. The authors conclude that Cpt1a-mediated fatty acid oxidation is necessary but not sufficient, and that PPARδ integrates metabolic support with additional chromatin remodeling to deliver the full fitness program.

Safety data from the mouse models were reassuring within the limits of the study. PPARδ activation in immune cells did not provoke systemic inflammation, did not alter splenic immune composition, did not worsen dextran sodium sulfate-induced colitis, and produced no histological evidence of autoimmunity or oncogenic transformation in surveyed tissues. Lifespan and body weight were comparable to controls. The authors caution, however, that longer-term studies across additional genetic backgrounds will be required before clinical translation, particularly because PPARδ has context-dependent effects in regulatory T cells, natural killer cells, myeloid cells and endothelial cells that were not the focus of this work.

The therapeutic implications are the most striking part of the study. In mice bearing established AKPS tumors, adoptive transfer of PPARδ-activated CD8⁺ T cells combined with anti-PD-1 checkpoint blockade dramatically reduced tumor growth and metastasis, and half of the recipient mice completely rejected their tumors, an outcome essentially unheard of in this immune-evasive model. Combining the transferred cells with dual anti-PD-1 and anti-CTLA4 blockade conferred a significant survival advantage. Human translation was explored in two ways. In an autologous co-culture system pairing patient-derived colorectal cancer organoids with the same patients’ peripheral blood mononuclear cells, pretreating the immune cells with the PPARδ agonist GW501516 increased tumor cell killing regardless of mismatch repair status. In CAR-T engineering, both pharmacological activation and genetic expression of the VP64–PPARδ fusion improved the cytotoxicity of EGFR-targeting and HER-2-targeting CAR-T cells against solid tumor targets, with the engineered cells showing elevated fatty acid oxidation, open chromatin at activation genes, reduced dysfunction markers such as BTLA and TOX2, and better tumor control in NSG mice bearing SK-OV-3 xenografts.

Beyond the immediate target, the study articulates a conceptual advance: pathways that confer fitness benefits in both cancer and immune cells can be redirected as therapies if they are selectively engaged in the immune compartment. PPARδ serves as proof of principle, resolving a literature in which the factor has been described as pro-tumorigenic, anti-tumorigenic, immunostimulatory and immunosuppressive depending on context. The lesson is that compartment, activation state, kinetics and ligand dependence determine outcome, and that a shared fitness program is not inherently good or bad but a resource to be seized by whichever cell uses it best. If selective, immune-restricted activation of such programs can be achieved safely in patients, the competitive balance inside immunologically cold tumors could be tipped toward the immune system, offering a route to sensitize MSS colorectal cancer and other resistant malignancies to immunotherapies that currently leave them largely untouched.

Subject of Research: The role of compartment-specific PPARδ activity in cancer–immune cell metabolic competition and immunotherapy efficacy

Article Title: Compartment-specific PPARδ activity controls the outcome of cancer–immune cell competition and immunotherapy efficacy

Article References: Compartment-specific PPARδ activity controls the outcome of cancer–immune cell competition and immunotherapy efficacy. (n.d.). https://doi.org/10.1038/s41590-026-02685-2

Image Credits: AI Generated

DOI: 10.1038/s41590-026-02685-2

Keywords: PPARδ, cancer immunotherapy, CD8 T cells, colorectal cancer, tumor microenvironment, fatty acid oxidation, Cpt1a, checkpoint blockade, CAR-T cells, metabolic competition, immune evasion, chromatin remodeling