Primary human CD4+ CD25+ CD127lo/− regulatory T (T reg) cells were isolated and subjected to repeated cycles of T cell receptor stimulation (anti-CD3/CD28) with or without proinflammatory cytokines. The investigators compared individual and combined cytokine conditions and found that the Th17-promoting combination of IL-6, IL-1β and IL-23 most consistently induced progressive loss of the lineage-specifying transcription factors FOXP3 and HELIOS.
FOXP3 and HELIOS expression were assessed at day 7 of each stimulation cycle to avoid transient activation-associated FOXP3 expression in conventional CD4+ T cells. Destabilization was both concentration- and duration-dependent: higher doses and prolonged exposure to the cytokine cocktail produced a larger fraction of FOXP3− HELIOS− cells. Transient early exposure still increased the frequency of destabilized cells in later cycles, indicating the cytokine effect can induce a durable change in identity rather than only transient repression.
Cytokine-exposed T reg cells progressively downregulated canonical T reg surface and functional markers including CD25, CTLA-4, ICOS and TIGIT, while upregulating markers linked to dysfunction such as CD226 and CD40LG. Functionally, these cells showed reduced in vitro suppressive capacity in co-culture assays with peripheral blood mononuclear cells.
Destabilized T reg cells also displayed increased proliferation relative to stable T reg cells and acquired the ability to produce proinflammatory cytokines, notably IFN-γ, TNF and IL-17A. Together, these phenotypic and functional shifts indicate conversion toward a less suppressive, more proinflammatory state.
To evaluate epigenetic changes, the authors examined DNA methylation at the FOXP3 conserved noncoding sequence 2 (CNS2, also called the T reg–specific demethylated region, TSDR). HELIOS+ cells retained predominant TSDR demethylation consistent with committed T reg identity. By contrast, FOXP3− HELIOS− cells showed partial remethylation of the TSDR, with variable remethylation at individual CpG sites, supporting the interpretation that these cells are undergoing a transitional state rather than arising from expansion of TSDR-methylated conventional T cell contaminants.
The study used single-cell paired chromatin accessibility and transcriptome profiling from nuclei of primary human cells at 16 and 25 days of culture. Ten clusters emerged on a weighted nearest-neighbor UMAP integrating both modalities. Clusters annotated as resting T reg (Trrest), effector T reg (Treff) and TNFR2-hi T reg (TrTNFR2hi) retained high expression of FOXP3, IKZF2 (HELIOS), CD27, CD25, CTLA4, ICOS and TIGIT, consistent with stable suppressive phenotype.
Other clusters—cycling T reg (Trcyc), cytotoxic T reg (Trcytox) and inflammatory T reg (Trinfl)—displayed downregulation of FOXP3 and HELIOS together with upregulation of proinflammatory or cytotoxic genes such as IFNG, TNF, IL17A, NKG7, GZMB and GZMA. The authors collectively term these clusters destabilized or putative (ex)T reg cells.
Changes in gene expression were accompanied by coordinated alterations in chromatin accessibility. For example, decreased FOXP3 transcription in destabilized clusters coincided with reduced accessibility at the FOXP3 CNS2 (TSDR). Conversely, increased accessibility at IFNG and GZMB loci aligned with transcriptional upregulation in destabilized cells. Global analysis identified loss of accessibility at OCRs associated with stable T reg identity and acquisition of OCRs not present in stable T reg cells; some acquired OCRs overlapped with conventional T cell accessibility patterns, while others were unique to destabilized T reg cells, indicating a distinct epigenetic state.
Differential motif enrichment analysis between stable and destabilized T reg clusters revealed relative depletion of FOXP3 and bHLH E-box motifs in destabilized cells and enrichment for AP-1, RUNX, ETS and IRF4 motifs. Because AP-1 and ETS family members are known to cooperate with IRF4 to shape transcriptional programs, these motif shifts suggested a cooperative AP-1–IRF4 axis as a driver of cell-state transition.
At the IRF4 locus, the authors observed reciprocal chromatin accessibility changes: a proximal OCR (OCRprox) located within intron 6 lost accessibility in destabilized cells, while a distal OCR (OCRdist) became more accessible during destabilization. Increased accessibility at this distal element correlated with increased IRF4 expression in destabilized cells. Bulk ATAC-seq across cytokine conditions showed changes in accessibility at these OCRs, and correlations across samples supported a relationship between distal element accessibility and IRF4 regulation.
The source document reports that excision of the distal IRF4 element conferred resistance to inflammatory cytokine–induced reprogramming, indicating the element functions as a regulatory node enabling heightened AP-1–IRF4 cooperative activity that promotes T reg destabilization. Conversely, forced co-expression of IRF4 with BATF promoted T reg destabilization. These experimental manipulations implicate the distal IRF4 regulatory element and AP-1–IRF4 activity as critical contributors to cytokine-driven loss of T reg identity.
The study reports two orthogonal perturbations: targeted excision of the distal IRF4 regulatory element reduced susceptibility of human T reg cells to cytokine-driven reprogramming, and enforced expression of IRF4 together with BATF enhanced destabilization. These results support a model in which increased accessibility of a distal IRF4 element amplifies IRF4 expression and cooperativity with AP-1 factors to drive transcriptional and epigenetic reprogramming away from the T reg program.
Collectively, the data identify a distal regulatory element at the IRF4 locus as a critical node that enables AP-1–IRF4 cooperative activity to drive inflammation-induced T reg cell destabilization. This insight suggests potential engineering strategies to preserve T reg lineage stability—for example, by targeting regulatory elements or modulating IRF4/AP-1 pathways—to improve the safety and efficacy of T reg cell–based immunotherapies. The source article did not provide extended clinical trial data or detailed therapeutic protocols; such translational steps would require further experimental validation and safety assessment.