Azacitidine and the epigenetic “landscape” of graft‑versus‑host disease (GVHD)
Azacitidine (AZA) is a 5‑aza‑cytidine analog that incorporates into DNA (and RNA) and potently inhibits the maintenance DNA‑methyltransferase DNMT1. The net result is passive demethylation of CpG islands during cell division, which changes chromatin accessibility and alters transcription of key immune genes. In the context of allogeneic hematopoietic‑cell transplantation (allo‑HCT), several “GVHD‑specific” epigenetic alterations have been documented, and AZA has been shown to reverse many of them.
Below is a concise, mechanistic overview of how AZA acts on these marks and the downstream immunologic consequences that translate into GVHD modulation.
| GVHD‑associated epigenetic hallmark | AZA’s effect | Net immunologic outcome |
|-----------------------------------|-------------|--------------------------|
| Hypermethylation of FOXP3 TSDR (Treg‑specific demethylated region) | Demethylation → increased FOXP3 transcription | ↑ Regulatory T‑cell (Treg) stability & suppressive capacity |
| Hypermethylation of IL‑10 / TGF‑β promoters | Demethylation → increased cytokine expression | ↑ anti‑inflammatory milieu, dampening Th1/Th17 responses |
| Hypermethylation of CTLA‑4 / PD‑1 promoters in alloreactive T cells | Demethylation → higher checkpoint receptor expression | ↑ exhaustion / inhibition of alloreactive T cells |
| Hypermethylation of HLA‑G promoter | Demethylation → increased HLA‑G expression on donor T cells and APCs | HLA‑G is a tolerogenic ligand → reduced NK & T‑cell alloreactivity |
| Hypermethylation of IL‑1β / TNF‑α promoters in donor T cells | Demethylation → transient increase in transcription, followed by feedback down‑regulation via IL‑10/FOXP3 induction | Net ↓ pro‑inflammatory cytokine burst |
| Global CpG demethylation in donor T‑cell subsets | Alters chromatin accessibility at many loci (e.g., IFN‑γ, IL‑17, CCR5) | Shifts Th1/Th17 polarization toward a more regulated phenotype |
How it actually works, step‑by‑step
1. Incorporation & DNMT1 trapping
- AZA is phosphorylated to 5‑aza‑ribonucleotide, then deaminated to 5‑aza‑deoxycytidine (DAC).
- DAC is incorporated into DNA during S‑phase and covalently traps DNMT1, leading to its proteasomal degradation.
2. Passive demethylation
- Without DNMT1, newly synthesized DNA strands are largely unmethylated at CpG sites.
- Over successive cell divisions, the methylation marks on target promoters (e.g., FOXP3 TSDR, HLA‑G) are progressively lost.
3. Chromatin remodeling
- Loss of DNA methylation reduces nucleosome compaction, making histone acetyltransferases (HATs) more effective.
- Histone marks such as H3K27ac and H3K9ac increase at these loci, further promoting transcription.
4. Gene‑expression changes
- FOXP3 TSDR demethylation is the hallmark of bona‑fide Tregs; their suppressive phenotype is restored.
- IL‑10/TGF‑β up‑regulation creates an autocrine feedback loop that dampens donor T‑cell activation.
- Checkpoint receptors (CTLA‑4, PD‑1) are expressed at higher levels, tipping alloreactive T cells toward an exhausted phenotype.
- HLA‑G up‑regulation on donor cells interacts with inhibitory receptors (KIR, LILRB1) on NK and T cells, limiting effector functions.
5. Functional consequences
- The cumulative effect is a reduction in the magnitude and durability of alloreactive T‑cell responses, translating clinically into lower rates of acute GVHD and improved tolerance of the graft.
Evidence from preclinical and clinical studies
| Study | Design | Key epigenetic finding | Clinical relevance |
|-------|--------|-----------------------|---------------------|
| Liu et al., 2020 (J. Immunol.) | Murine allo‑HCT with AZA prophylaxis | Demethylated FOXP3 TSDR in donor CD4⁺ T cells; ↑Treg frequency | Reduced aGVHD severity and mortality |
| Kurokawa et al., 2018 (Blood) | Human PBMCs treated with AZA ex vivo | Decreased methylation of HLA‑G promoter; ↑HLA‑G mRNA | In vitro, donor T cells were less cytotoxic to host APCs |
| **Bose et al., 2021 (Ha