A metabolic enzyme found in the support cells surrounding tumors may hold the key to predicting which cancer patients will benefit from immunotherapy, according to new peer-reviewed research. The enzyme, nicotinamide N-methyltransferase (NNMT), appears to drive immune suppression in the tumor microenvironment by reprogramming nearby fibroblasts, and blocking it restored the ability of the immune system to attack cancer in mouse models. The findings open a potential path toward both a predictive biomarker and a new drug target, though any combination therapy for human patients would still need regulatory approval.
How a Support Cell Enzyme Shuts Down Immune Attacks
Most immunotherapy research focuses on cancer cells and the T cells meant to destroy them. But the new study, published in Nature, shifts the focus to cancer-associated fibroblasts (CAFs), which are structural cells that form the scaffolding of a tumor. Researchers found that NNMT acts as a metabolic-epigenetic regulator inside these fibroblasts, altering their gene expression through chemical modifications to DNA-packaging proteins. Those epigenetic changes cause CAFs to secrete complement proteins, which in turn recruit myeloid-derived suppressor cells (MDSCs) into the tumor. MDSCs are a class of immune cells that actively block T cell function, creating a shield around the cancer.
The practical effect is that even when checkpoint inhibitors like PD-1 blockers are administered, T cells cannot penetrate the immunosuppressive barrier that NNMT-high fibroblasts build. This helps explain a persistent clinical puzzle: why a large share of patients receiving checkpoint immunotherapy see little or no tumor shrinkage. The problem, this research suggests, is not always that the patient’s T cells are weak or that the tumor has evolved to hide. Sometimes the surrounding stroma simply smothers the immune response before it begins.
Spatial Mapping of Human Tumors Confirms the Pattern
The Nature paper drew on spatial transcriptomics data from human tissue samples cataloged in the Gene Expression Omnibus. The tissue microarray design spanned non-tumorous adjacent stroma, ovarian tumors, and omental metastases, with annotated areas of interest including cancer epithelium, macrophage and monocyte populations, immune cells, and stroma. By mapping where NNMT expression was highest and correlating those zones with immune cell infiltration, the researchers could show that NNMT-rich fibroblast neighborhoods were consistently depleted of active T cells and enriched for MDSCs.
Mouse experiments confirmed the findings. When NNMT was inhibited in fibroblasts, the immunosuppressive complement cascade weakened, MDSC recruitment dropped, and T cells regained access to the tumor. Combined with checkpoint blockade, NNMT inhibition produced stronger antitumor responses than either approach alone, according to an institutional summary of the research. Any combination therapy in humans would require U.S. Food and Drug Administration approval, so clinical translation remains years away.
Where NNMT Fits in the Metabolic Immune Puzzle
NNMT does not operate in isolation. It belongs to a broader family of metabolic enzymes that regulate one-carbon (1C) metabolism, a network of biochemical reactions that shuffles single-carbon units among amino acids, nucleotides, and methyl donors. A Cell Research commentary on the Nature study placed NNMT within this wider framework, noting that its role as a metabolic-epigenetic regulator in fibroblasts clarifies how stromal metabolism can dictate whether checkpoint blockade succeeds or fails. The commentary described complement-driven MDSC recruitment as the central link in the mechanistic chain and flagged the proposed combination with checkpoint inhibitors as a logical next step.
Other enzymes in the 1C pathway have also drawn attention as potential cancer targets. Methylenetetrahydrofolate dehydrogenase 2 (MTHFD2) drives the folate cycle, sustaining uridine diphosphate N-acetylglucosamine (UDP-GlcNAc) levels that are necessary for T cell proliferation and cytokine production. MTHFD2 is also upregulated in tumor cells and associated with S-adenosylmethionine (SAM) production, a key methyl donor. Targeting serine-glycine metabolism, which feeds into the same 1C network, has attracted growing research interest because related metabolic enzymes show potential as therapeutic targets in cancer.
The connection between 1C metabolism and immune function extends beyond individual enzymes. A review in Clinical and Translational Medicine reported that one-carbon metabolism helps predict prognosis for patients receiving immunotherapy and informs drug discovery efforts. What makes NNMT distinctive in this group is that its most consequential activity appears to occur not in the cancer cells themselves but in the surrounding fibroblasts that shape the tumor microenvironment. Rather than directly feeding tumor growth, NNMT remodels stromal cells into active enforcers of immune suppression.
From Mechanism to Biomarker
The mechanistic data suggest that NNMT levels in CAFs could serve as a biomarker for immunotherapy response. In principle, patients whose tumors harbor NNMT-high fibroblasts might be less likely to benefit from checkpoint inhibitors alone, because their T cells would remain trapped outside an MDSC-fortified stromal barrier. Conversely, tumors with low NNMT activity in the stroma might be more permissive to T cell infiltration once PD-1 or CTLA-4 brakes are released.
Translating that concept into a usable clinical test would require several steps. Pathologists would need standardized assays to measure NNMT expression in tumor stroma, likely through immunohistochemistry or RNA-based profiling of biopsy samples. Clinicians would then have to validate whether NNMT levels correlate with response rates, progression-free survival, or overall survival in patients receiving checkpoint inhibitors across different cancer types. Only after such prospective validation could NNMT be considered for routine use in treatment selection.
Even if NNMT proves predictive, it is unlikely to act alone. The tumor microenvironment reflects a complex interplay among metabolism, cytokine signaling, vascular structure, and immune cell composition. One-carbon metabolism intersects with many of these layers, and a recent overview of metabolic and immune crosstalk emphasized that multiple pathways often converge to shape immunotherapy outcomes. NNMT might therefore become part of a broader panel of stromal and metabolic markers rather than a single deciding factor.
Prospects for Targeted NNMT Inhibitors
The therapeutic implications of the new work are equally significant. If NNMT-driven epigenetic reprogramming of fibroblasts is essential for building an MDSC-rich shield, then pharmacologically blocking NNMT could weaken that barrier and allow T cells to reach their targets. In the mouse models described in the Nature paper, genetic and pharmacologic inhibition of NNMT both reduced complement secretion and restored T cell infiltration, particularly when combined with checkpoint blockade.
Designing NNMT inhibitors suitable for human use will pose familiar challenges. The enzyme participates in normal cellular metabolism, so systemic inhibition could carry safety risks, especially in tissues that rely heavily on methylation reactions. Medicinal chemists would need to develop molecules with sufficient selectivity and pharmacokinetic properties to modulate NNMT activity in the tumor microenvironment without causing unacceptable toxicity elsewhere. Early-stage compounds might first be tested in advanced cancers that have exhausted standard options, with careful monitoring for immune-related and metabolic side effects.
Another open question is whether NNMT inhibition would synergize equally well with all forms of immunotherapy. The current data focus on checkpoint inhibitors, but other modalities, such as adoptive T cell transfer or cancer vaccines, also depend on robust T cell access to tumor tissue. If NNMT shapes the stromal gatekeeping function more generally, then blocking it could potentially enhance a wider range of immune-based treatments. Answering that will require additional preclinical studies that systematically vary both the immunotherapy platform and the timing of NNMT blockade.
Balancing Promise and Uncertainty
For now, NNMT stands out as a compelling example of how noncancerous cells in the tumor microenvironment can determine whether immunotherapy works. By illuminating a metabolic-epigenetic circuit that links fibroblasts, complement proteins, MDSCs, and T cell exclusion, the new research adds a crucial piece to the puzzle of variable patient responses. It also underscores the importance of looking beyond tumor-intrinsic mutations and considering stromal metabolism as both a biomarker and a therapeutic target.
Yet substantial uncertainty remains. The mouse models and spatial analyses provide strong mechanistic support, but human tumors are more heterogeneous, and clinical behavior can diverge from preclinical expectations. It is not yet clear how consistently NNMT levels in CAFs will predict outcomes across cancer types, or whether other stromal pathways might compensate when NNMT is blocked. Regulatory hurdles, safety evaluations, and the logistics of integrating new biomarkers into oncology workflows will all shape how quickly these insights reach patients.
Even with those caveats, the work on NNMT offers a clear conceptual shift. Rather than viewing the tumor microenvironment as a passive backdrop, it frames stromal cells as active participants whose metabolic choices can silence or unleash the immune system. If future studies confirm that inhibiting NNMT in fibroblasts safely dismantles the immunosuppressive shield in people, oncologists could gain both a new way to forecast immunotherapy benefit and a fresh target to help more patients respond.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
