Paper Details

ABSTRACT

Introduction: Nodal T-follicular helper lymphomas (n-TFHL; AITL, follicular, NOS) are aggressive mature TFH-cell neoplasms with limited therapeutic advances. Recurrent TET2 and DNMT3A mutations are common in n-TFHL, unlike myeloid malignancies or CHIP conditions, where mutations are typically monoallelic and mutually exclusive. We hypothesize that cooperative Tet2 and Dnmt3a deficiency establishes a distinct epigenetic state reprogramming the CD4? T-cell transcriptome, driving aberrant T_FH differentiation and malignant transformation.

Methods: We generated conditional CD4? T-cell murine models with single (Tet2?^/?) or combined Tet2/Dnmt3a deficiency (DKO) and established DKO tumor cell allograft models in NOD/Scid γ-(NSG) mice for preclinical therapeutic evaluation. We used functional genomics approaches (RNA-seq, DNA 5hmC and 5mC profiling) for preneoplastic and neoplastic CD4⁺ T-cells. Cell proliferation, apoptosis  and polarization assays were performed.

Results: Genomic analysis among PTCL entities demonstrated a significant enrichment of co-occurring TET2 and DNMT3A mutations in the n-TFHL subtype (51%). Clinically, double mutant cases trended toward inferior survival vs WT (p=0.14). A large subset of cases exhibited higher variant allele frequencies than expected for the tumor population, supporting a hematopoietic stem cell origin. Most of these mutations were either biallelic or involved multiple mutations within the same tumor. Double mutant cases were enriched for proliferation-promoting and stress-response-associated gene signatures.

Conditional biallelic Tet2/Dnmt3a deletion in murine CD4? T cells yielded a more aggressive T_FH lymphoma phenotype than Tet2 loss alone, with significantly accelerated disease onset and reduced overall survival (median OS, 13.4 vs 20.4 months; P<0.0001). These mice exhibited splenomegaly, lymphadenopathy, hepatomegaly, and disrupted splenic architecture. Histology showed T-cell lymphoma in 15/23 DKO cases (65%): CD4? (9/15), CD8? (4/15), B-cell neoplasm (2/15); others were non-lymphoid. Clonal T-cell expansion was associated with increased CD3?CD4? T cells and expanded CXCR5?PD-1? TFH cells vs WT (≈ 5-fold; p<0.01).

Neoplastic DKO CD4? T cells showed global 5mC/5hmC depletion, increased expression of proliferation/activation markers (Ki67, CD44, ICOS, PD-1, pFOXO1, LAG-3), and reduced TCF-7. They displayed active TCR signaling with elevated phosphorylation of PLCγ1, AKT, ERK, LCK, and STAT3 without external stimulation. Pre-neoplastic DKO CD4? T cells (48 weeks) showed increased exhaustion markers (LAG-3/TIM-3), an effector-memory phenotype (CD44?CD62L?), and activation of the TCR–PI3K–AKT axis.

Functionally, DKO CD4? T cells from 12-week mice exhibited enhanced proliferative capacity, increased survival, elevated activation marker expression (CD69), accelerated S-phase entry, and augmented proximal TCR signaling upon CD3 CD28 stimulation, characterized by increased phosphorylation of LCK and SLP-76. These cells exhibited hyper plasticity in vitro, with increased differentiation toward TH1, TH2, and TFH lineages, while retaining epigenetic dysregulation (decreased 5hmC and increased 5mC).

Integrated multi-omics analysis (RNA-seq, 5hmC, and 5mC) of CD4? T cell at pre-neoplastic (12 and 48 weeks) and at neoplastic stages revealed progressive transcriptional reprogramming with activation of cell-cycle, DNA replication, and DNA damage response pathways increasing over time.

Serial transplantation of DKO tumors in NOD/Scid γ-(NSG) mice showed transplantable, aggressive, fully penetrant, clonally related CD4? T-cell lymphomas with shortened latency and preserved T_FH phenotype. Multi-omics analysis of DKO tumors identified actionable therapeutic pathways, and combination treatment of allograft-bearing NSG mice with 5-azacytidine plus copanlisib (PI3K) or ruxolitinib (JAK1/2) inhibitors significantly extended survival, with dual epigenetic and pathway-targeted therapy outperforming either agent alone.

Conclusion: This study reveals a mechanistic basis for n-TFHL pathogenesis driven by concurrent TET2 and DNMT3A loss. Using genetically engineered mouse models, we show that combined TET2/DNMT3A deficiency recapitulates defining molecular and phenotypic features of TFH cell transformation. These findings uncover PI3K and JAK–STAT signaling as actionable therapeutic vulnerabilities and provide a rationale for precision treatment strategies in n-TFHL patients harboring TET2/DNMT3A co-mutations.