Hypothesis\n\nSpecific heteroplasmic mtDNA single‑nucleotide variants that accumulate after age 70 trigger a retrograde signal that recruits GPS2 to the nucleus, where it complexes with the H3K9me2/3 demethylase KDM4D to remove repressive marks at promoters of mitochondrial‑target genes. This epigenetic remodeling sustains a maladaptive transcriptional program that accelerates stem‑cell functional decline.\n\n## Mechanistic Model\n\n1. mtDNA variant load – Heteroplasmic SNVs increase mitochondrial ROS and cause mild depolarization without triggering mitophagy.\n2. GPS2 activation – Depolarized mitochondria expose a conserved N‑terminal motif that promotes GPS2 oligomerization and its release into the cytosol.\n3. Nuclear import – GPS2 binds importin‑α5 via a newly identified NLS that is unmasked only after oxidation of cysteine‑57.\n4. Chromatin targeting – In the nucleus, GPS2 scaffolds KDM4D to H3K9me2/3‑enriched promoters of nuclear‑encoded mitochondrial genes (NEMPs), leading to demethylation and a paradoxical increase in accessibility that drives aberrant transcription.\n5. Feedback loop – Aberrant NEMP expression further stresses mitochondria, increasing heteroplasmy load and cementing the cycle.\n\n## Predictions\n- In young stem cells engineered to carry a pathogenic mtDNA heteroplasmy (e.g., m.5024C>T), GPS2 nuclear accumulation will rise within 48 h, preceding detectable changes in ATAC‑seq signal.\n- CRISPR‑based base editing that reduces the heteroplasmy below 5 % will prevent GPS2 translocation and preserve youthful chromatin accessibility at NEMP promoters.\n- Pharmacological inhibition of KDM4D (e.g., with compound CPI‑455) will block the accessibility gain despite mitochondrial stress, decoupling mtDNA dysfunction from chromatin changes.\n- Single‑cell multi‑omics (scATAC‑seq + scRNA‑seq + mtDNA genotyping) from aged human hematopoietic stem cells will show a tight coupling: cells with >10 % heteroplasmy at specific loci exhibit both GPS2 nuclear signal (by imaging) and increased accessibility at a defined set of NEMP promoters.\n\n## Experimental Approach\n\nStep 1 – Induce heteroplasmy\n- Use mito‑TALENs or DdCBE base editors to introduce a common age‑associated mtDNA SNV into mouse intestinal stem cells (ISCs). Confirm heteroplasmy levels by droplet digital PCR.\n\nStep 2 – Track GPS2\n- Generate a GPS2‑mKnock‑in fluorescent tag; perform live‑cell imaging and subcellular fractionation to quantify nuclear vs mitochondrial GPS2 over time.\n\nStep 3 – Chromatin readout\n- Perform ATAC‑seq and CUT&RUN for H3K9me2/3 and KDM4D on sorted ISCs at 0, 3, 7, 14 days post‑editing.\n- Validate promoter accessibility changes at NEMPs (e.g., Ndufs4, Cox5b) by qPCR.\n\nStep 4 – Rescue\n- Treat cells with mito‑targeted antioxidant (MitoQ) or KDM4D inhibitor; assess whether GPS2 nuclear entry and ATAC‑seq alterations are prevented.\n- In parallel, reduce heteroplasmy using a second‑generation DdCBE and repeat measurements.\n\nStep 5 – Human validation\n- Obtain CD34+ cells from donors <35 and >70 yr; measure mtDNA heteroplasmy burden, GPS2 localization (immunofluorescence), and ATAC‑seq peaks at NEMP promoters.\n- Perform correlation analysis and test causality by ex vivo mito‑base editing.\n\n## Potential Pitfalls and Alternatives\n- If GPS2 does not show increased nuclear import, other retrograde mediators (e.g., ATF5, CHOP) may be responsible; the hypothesis can be adapted to test those.\n- Heteroplasmy may trigger global chromatin changes independent of GPS2; single‑cell multi‑omics will help disentangle direct vs indirect effects.\n- Compensatory upregulation of other demethylases could mask KDM4D inhibition; use combinatorial knockdowns.\n\nBy linking a defined mtDNA lesion to a GPS2‑chromatin axis, this hypothesis converts a correlative observation into a mechanistic, intervention‑ready framework for mitochondrial control of the nuclear epigenome in aging.