Epigenetic scar hypothesis: aged neurons retain proteostatic stress marks that resist iPSC reprogramming and predispose to complement-mediated pruning

We hypothesize that neurons surviving the age‑dependent eviction process encode a reprogramming‑resistant epigenetic signature at loci governing autophagy and ubiquitin‑proteasome function. This signature persists after iPSC re‑programming and redifferentiation, rendering the resulting neurons more susceptible to microglia‑mediated complement pruning and less capable of clearing aggregated synaptic proteins. Specifically, prolonged proteostatic stress in aged neurons elevates H3K9me3 and DNA methylation at promoters of key autophagy genes (e.g., ATG5, BECN1) and proteasome subunits (e.g., PSMB5), while simultaneously reducing H3K27ac at enhancer regions of CD47, a “don’t eat me” signal. These chromatin changes are not erased by the pluripotent state, leading to a neuronal phenotype that mirrors the original aged state: higher basal C1q deposition on synapses, lower phagocytic resistance, and accelerated loss of synapses when exposed to microglia in vitro.

Testable predictions

1. iPSC‑derived neurons from donors >70 years will show significantly higher H3K9me3 and CpG methylation at ATG5 and PSMB5 promoters compared to iPSC‑neurons from donors <30 years, even after identical reprogramming protocols.
2. Restoring autophagy flux (e.g., via rapamycin treatment) or demethylating agents (e.g., 5‑azacytidine) in aged‑donor iPSC‑neurons will reduce C1q‑synapse colocalization and increase CD47 surface expression to levels comparable to young‑donor neurons.
3. Co‑culture of aged‑donor iPSC‑neurons with microglia will produce greater synapse loss (quantified by PSD‑95/VGLUT1 puncta) than young‑donor counterparts, an effect attenuated by blocking C1q or overexpressing CD47.
4. ATAC‑seq on sorted neuronal nuclei will reveal reduced accessibility at autophagy gene loci specifically in aged‑donor iPSC‑neurons, correlating with proteomic decreases in LC3‑II and ubiquitin‑conjugated species.

Falsifiability If aged‑donor iPSC‑neurons fail to display any of the predicted epigenetic or functional differences, or if pharmacological rescue of autophagy does not normalize pruning susceptibility, the hypothesis would be refuted. This would suggest that age‑related neuronal eviction does not leave a durable epigenetic scar, prompting alternative explanations for the lag in neural rejuvenation relative to other tissues.