Fat-Laden T Cells Burst in Inflamed Joints and Fuel Rheumatoid Arthritis

Roughly 10% of people develop a clinically significant autoimmune disease, and no therapy restores immune tolerance. Chronic tissue inflammation is a major obstacle, and researchers are increasingly looking at antigen-independent drivers such as the metabolic state of effector cells. In rheumatoid arthritis, CD4+ T cells have defective DNA repair and impaired mitochondria. How these metabolically weakened cells survive and behave inside the glucose-poor synovium has been unclear.

The Mayo Clinic and Stanford team combined human samples, patient-derived T cells and a human synovium–NSG mouse chimera. Mice were engrafted with human synovial tissue and reconstituted with patient PBMCs. They quantified free fatty acids (FFAs) in synovial biopsies and tested oleic acid (10 mg/kg intraperitoneally for one week) in the chimeras. They also ran lipidomics, electron microscopy, organelle imaging, live-cell death assays and lipid droplet (LD) proteomics on CD4+ T cells from healthy donors, RA patients and psoriatic arthritis controls.

Inflamed synovium contained FFA concentrations about 100-fold higher than plasma, and these tracked with tissue inflammation scores. Oleic acid worsened synovitis in the chimeras, with more immune infiltration, dead cells and lipid deposits, plus higher cytokine production by macrophages and T cells. IFN-γ+ T-cell frequency correlated with LD formation and cell death. RA CD4+ T cells held about 40% more intracellular FFAs than controls, and their LDs were enriched in oleic and myristic acid. The cells showed signs of failed lipid catabolism, including acetyl-CoA and citrate accumulation, ER expansion with ER stress markers, and upregulated CD36. In a 17-patient cohort, the ratio of reduced to oxidized FFA tracked with disease activity (CDAI).

LDs in RA cells sat at the uropod plasma membrane, with LD–membrane distances shrinking from 300–400 nm in healthy cells to under 20 nm. Oleic acid pushed LDs into the membrane and incorporated the LD protein PLIN2 there. Healthy T cells survived oleic acid, but circulating RA cells died faster, and RA tissue T cells were mostly dead within an hour. LD proteomics showed enrichment of gasdermin D, caspase-1, mitochondrial-contact proteins and the depalmitoylase LYPLA2. According to the abstract, LDs sequester gasdermin D and the acyltransferase zDHHC5 and carry them to the membrane, triggering pyroptotic rupture. PLIN2 knockdown or blocking protein acylation then suppressed synovitis in the model.

The work suggests that mitochondrially exhausted T cells are primed to die in lipid-rich tissue, releasing pro-inflammatory cargo that sustains inflammation. This opens possible targets in LD biogenesis and gasdermin acylation. Caveats are the preclinical, humanized-mouse evidence, small patient cohorts, and that the full text available stopped partway through the results. The zDHHC5, PLIN2 knockdown and acylation-inhibition findings are therefore summarized from the abstract. A correction notice for this paper was also published in October 2026.