The research team, led by investigators at Fudan University, set out to determine whether immunosenescence, traditionally considered a hallmark of aging, might also be operating in the inflamed intestines of people with ulcerative colitis. Their starting point was a large-scale analysis of single-cell RNA sequencing data from the colonic mucosa of patients and healthy controls. When the researchers scored every immune cell against established senescence gene signatures, a striking pattern emerged: T cells from patients carried significantly higher senescence scores than T cells from healthy donors, while B cells and myeloid cells showed no comparable increase. The finding was replicated in an independent patient cohort, strengthening the case that senescent T cells genuinely accumulate in the ulcerative colitis microenvironment rather than appearing as a statistical fluke.

Digging deeper into the T cell compartment, the team stratified the cells into a high-senescence group, comprising the top ten percent by senescence score, and a low-senescence group containing the rest. Computational trajectory analysis placed the high-senescence cells at the terminal end of differentiation, suggesting they had reached an irreversible state. Gene set enrichment analysis revealed that these cells were loaded with inflammatory response programs, and cell-to-cell communication modeling showed that they interacted far more intensively with epithelial cells than their low-senescence counterparts did. In other words, the senescent T cells were not inert bystanders; they appeared to be active participants in the inflammatory dialogue that sustains mucosal damage.

To validate these observations in living systems, the researchers turned to a chronic colitis mouse model induced by repeated cycles of dextran sulfate sodium, a chemical that damages the colonic epithelium and triggers persistent inflammation. Flow cytometry of immune cells isolated from the colon revealed a stark increase in the frequency of T cells positive for senescence-associated beta-galactosidase, a classic enzymatic marker of senescence. Immunofluorescence staining confirmed the accumulation of T cells bearing gamma-H2A.X, a marker of DNA damage, and p21, a protein enforcing cell-cycle arrest. The colonic T cells also displayed shortened telomeres and elevated intrinsic beta-galactosidase activity. Notably, B cells and myeloid cells did not show these changes to the same degree, reinforcing the specificity of the T cell response.

The pivotal question was whether these senescent cells actually drive disease or merely accompany it. The team administered a senolytic regimen, the combination of dasatinib and quercetin, which selectively eliminates senescent cells, to mice with chronic colitis. The treatment significantly reduced weight loss, improved disease activity scores, and lowered the expression of pro-inflammatory cytokines including tumor necrosis factor, interleukin-6, and interleukin-1 beta. When the researchers examined which immune populations had been depleted, the answer was unambiguous: the frequency of senescent CD8-positive T cells dropped markedly, while senescent CD4-positive T cells and regulatory T cells were largely untouched. Moreover, the abundance of senescent CD8-positive T cells correlated strongly with disease severity metrics. A depletion experiment using an anti-CD8 antibody, alone or combined with senolytics, showed no additive benefit, indicating that the therapeutic effect of the senolytic cocktail was primarily mediated through clearing senescent CD8-positive T cells.

With the pathogenic cell type identified, the researchers turned to the metabolic underpinnings of its senescent state. Untargeted metabolomic profiling of the intestinal mucosa revealed a profound disturbance of nicotinate and nicotinamide metabolism, the biochemical pathway that generates and recycles nicotinamide adenine dinucleotide, or NAD+, a coenzyme essential for hundreds of cellular reactions. In colitic mice, the precursors of NAD+ synthesis, tryptophan and nicotinic acid, were depleted, while the catabolic byproducts nicotinamide and ADP-ribose accumulated. Direct measurement confirmed reduced NAD+ levels in inflamed colons, accompanied by upregulation of NAD+-consuming enzymes such as CD38, SIRT1, and PARP1, alongside the salvage enzyme NAMPT. Human transcriptomic datasets mirrored these findings, and spatial transcriptomics of patient tissue revealed that areas rich in senescent CD8-positive T cells co-localized with regions of high NAD+ metabolic activity, expanding from a rare niche in healthy colons to a prominent one in diseased tissue.

To test whether NAD+ depletion is sufficient to induce senescence, the researchers treated primary murine CD8-positive T cells with FK866, an inhibitor of NAMPT that restricts intracellular NAD+ availability. Within seventy-two hours, the treated cells exhibited elevated beta-galactosidase activity, upregulated p16 and p21, impaired proliferation, and accelerated telomere shortening. Supplementing the cells with the NAD+ precursor nicotinamide mononucleotide largely reversed these changes, as did genetic knockdown experiments that confirmed the specificity of the effect. Transcriptomic profiling of the NAD+-depleted cells showed a characteristic senescence program: cell-cycle inhibitors and the terminal differentiation marker KLRG1 went up, while co-stimulatory molecules and cytotoxic effectors such as granzyme B and perforin went down. Intriguingly, the cells did not upregulate canonical exhaustion markers like PD-1, CTLA-4, or LAG-3, underscoring that senescence and exhaustion are biologically distinct dysfunctional states. The senescent cells also secreted abundant interferon-gamma and tumor necrosis factor alpha, an antigen-independent inflammatory program that persisted even as their antigen-specific responsiveness collapsed.

The causal role of these metabolically induced senescent cells was demonstrated through adoptive transfer. When FK866-treated CD8-positive T cells were transferred into mice with dextran sulfate sodium colitis, the recipients experienced accelerated weight loss and higher disease activity scores than controls receiving untreated cells. The transferred cells engrafted in the inflamed colon, expanded the CD8-positive compartment, and increased the frequency of interferon-gamma-producing CD8-positive T cells in the mucosa. This experiment established that NAD+-depleted senescent CD8-positive T cells are not merely a correlate of inflammation but active agents capable of worsening disease on their own.

Mechanistically, the study traced the pathway from metabolic starvation to inflammatory output through the mitochondria. NAD+ depletion suppressed oxidative phosphorylation, reduced ATP production, and shrank mitochondrial mass, as confirmed by transmission electron microscopy showing aberrant, diminished organelles. Mitochondrial NAD+ levels fell sharply, mitochondrial reactive oxygen species accumulated, and, critically, mitochondrial DNA leaked into the cytosol. This mislocalized DNA acted as an endogenous danger signal, activating the cGAS-STING innate immune sensing pathway and its downstream effectors TBK1, IRF3, and IRF7. Pharmacological blockade of STING attenuated the senescent phenotype, while a STING agonist reproduced it, and CD8-positive T cells lacking STING were completely resistant to FK866-induced senescence. Blocking VDAC1 oligomerization, the pore through which mitochondrial DNA escapes, similarly prevented both DNA leakage and senescence, cementing the mitochondria-to-cytosol DNA pathway as the indispensable link between metabolic stress and the senescence program.

The clinical implications of these findings are considerable. In two independent patient cohorts, individuals with high scores for both NAD+ metabolic dysregulation and senescent CD8-positive T cell infiltration showed strong enrichment of pro-inflammatory pathways and, more importantly, significantly lower response rates to frontline biologics, including the anti-TNF agent infliximab and the gut-selective anti-integrin vedolizumab. Responders had lower baseline scores and showed greater reductions after treatment, and the combined metric correlated positively with endoscopic disease activity. These results position the NAD-plus-senescence axis not only as a mechanistic driver of ulcerative colitis but as a potential prognostic biomarker capable of identifying patients destined to fail standard therapies before treatment begins. They also suggest a therapeutic strategy that departs from conventional immunosuppression: rather than broadly suppressing T cell activity, clinicians might one day selectively eliminate the senescent CD8-positive population or protect T cells from mucosal NAD+ exhaustion, thereby interrupting the self-perpetuating cycle of metabolic damage, mitochondrial leakage, and inflammatory secretions that keeps the colon in a state of chronic war. The authors caution that the dextran sulfate sodium model does not fully recapitulate human disease and that the senolytic regimen used is not T-cell specific, so lineage-specific studies will be needed. Even so, the work opens a genuinely new front in inflammatory bowel disease research, one where the tools of geroscience, the biology of aging, may prove as relevant as immunology itself.

Subject of Research: NAD+ metabolic reprogramming driving CD8+ T cell senescence in ulcerative colitis

Article Title: NAD+ Metabolic Reprogramming Drives CD8+ T Cells Senescence and Exacerbates Ulcerative Colitis

Article References: Ye, M., Zhou, Q., Kong, M., Zhao, S., Lin, L., Chen, L., Yu, J., Luo, F., Liu, J., & Zhang, J. (2026). NAD + Metabolic Reprogramming Drives CD8 + T Cells Senescence and Exacerbates Ulcerative Colitis. Aging Cell, 25(9), Article e70706. https://doi.org/10.1111/acel.70706

Image Credits: AI Generated

DOI: 10.1111/acel.70706

Keywords: ulcerative colitis, CD8+ T cells, NAD+ metabolism, cellular senescence, cGAS-STING, mitochondrial dysfunction, senolytics, inflammatory bowel disease, immunosenescence, SASP, biologic therapy resistance, spatial transcriptomics