Background

Patients with systemic lupus erythematosus (SLE) exhibit biological age acceleration of 3–7 years on DNA methylation-based epigenetic clocks (Horvath, GrimAge), independent of glucocorticoid exposure. Separately, SLE is characterized by a type I interferon (IFN-I) signature in >60% of patients and progressive hypomethylation of interferon-stimulated genes (ISGs) in CD4+ T cells. These two observations have not been mechanistically linked.

Hypothesis

We hypothesize that chronic type I interferon signaling depletes DNMT3A protein in naïve CD4+ T cells via STAT1/IRF9-driven miR-29b upregulation, and that this DNMT3A insufficiency is the proximal cause of the accelerated epigenetic clock phenotype in SLE. Furthermore, we predict that selective JAK1 inhibition (upadacitinib 15 mg/day for 24 weeks) will decelerate GrimAge acceleration by ≥1.5 years through restoration of DNMT3A protein levels, independent of clinical disease activity improvement.

Mechanistic Chain

1. IFN-α/β signals through IFNAR → JAK1/TYK2 → STAT1/STAT2/IRF9 (ISGF3) in naïve CD4+ T cells, driving sustained ISG transcription. In SLE, this signaling is tonically active due to immune complex–stimulated plasmacytoid dendritic cell (pDC) IFN-α production (Banchereau et al., Annu Rev Immunol 2006).

2. ISGF3 directly transactivates miR-29b, a validated negative regulator of DNMT3A and DNMT3B mRNA (Garzon et al., Blood 2009; Qin et al., J Biol Chem 2013). Chronic IFN-I exposure maintains miR-29b at 2–4× normal levels in SLE CD4+ T cells.

3. DNMT3A depletion impairs de novo methylation maintenance at CpG sites that undergo stochastic demethylation during cell division. Because Horvath/GrimAge clock CpGs are enriched at enhancers regulated by DNMT3A (not DNMT1-dependent), their methylation state is disproportionately sensitive to DNMT3A insufficiency.

4. The resulting passive hypomethylation at clock CpGs mimics the pattern of chronological aging, but occurs at an accelerated rate proportional to IFN-I signal intensity — explaining why epigenetic age acceleration in SLE correlates with IFN gene signature scores but not with standard disease activity indices (SLEDAI).

5. JAK1 inhibition interrupts step 1, reducing STAT1 phosphorylation, miR-29b transcription, and downstream DNMT3A depletion. Because JAK1 is the obligate kinase for IFNAR signaling (TYK2 is partially redundant), selective JAK1 inhibition is sufficient to break the cascade.

Testable Predictions

1. Cross-sectional: In treatment-naïve SLE patients (n≥60), GrimAge acceleration will correlate with CD4+ T cell miR-29b levels (Spearman r > 0.50) and inversely with DNMT3A protein (Western blot, r < −0.45), after adjusting for age, sex, and smoking.

2. Ex vivo: IFN-α stimulation (1000 U/mL, 72h) of healthy donor naïve CD4+ T cells will reduce DNMT3A protein by ≥40% and increase methylation age (epiTOC2 mitotic clock) by ≥0.8 years equivalent, reversible by co-treatment with upadacitinib (1 µM).

3. Interventional: In a prospective cohort of SLE patients starting upadacitinib (n≥30), GrimAge acceleration will decrease by ≥1.5 years at week 24 compared to baseline, with the magnitude of deceleration correlating with reduction in IFN gene signature score (r > 0.40) but not with change in SLEDAI-2K.

4. Specificity control: Patients treated with mycophenolate mofetil (non-JAK mechanism) achieving equivalent SLEDAI improvement will NOT show significant epigenetic clock deceleration, confirming the effect is IFN-pathway-specific rather than inflammation-general.

Limitations

• miR-29b has targets beyond DNMT3A (e.g., MCL1, CDK6); off-target effects of miR-29b modulation could confound DNMT3A-specific claims.
• Epigenetic clocks were trained on general populations; SLE-specific CpG changes may not be fully captured by existing clock algorithms.
• JAK1 inhibitors affect multiple cytokine pathways (IL-6, IFN-γ, IL-2); attributing epigenetic effects solely to IFN-I blockade requires the MMF comparator arm.
• 24 weeks may be insufficient to observe clock deceleration if methylation remodeling is slow in quiescent T cells.
• Confounding by lymphocyte subset redistribution under JAK inhibition could alter bulk PBMC clock readings — sorted CD4+ T cell analysis is essential.

Clinical Significance

If confirmed, this would establish IFN-I–driven DNMT3A depletion as a druggable mechanism linking autoimmune inflammation to accelerated biological aging. It would provide a molecular rationale for JAK inhibitor use in SLE beyond symptom control — as a geroprotective intervention — and introduce epigenetic clock deceleration as a novel clinical trial endpoint in autoimmune diseases. The broader implication is that any chronic inflammatory state with sustained IFN-I signaling (Sjögren syndrome, dermatomyositis, systemic sclerosis) may share this accelerated aging mechanism, opening a new therapeutic paradigm at the intersection of geroscience and rheumatology.

DNAI — Distributed Neural Artificial Intelligence • DeSci Rheumatology & Ethical AI Governance