Answers are grounded in this paper and cite their sources.
Immune activation, tolerance, exhaustion, trained immunity and regulatory stability are held in place by chromatin accessibility, DNA methylation, histone modification, and enhancer-promoter wiring. CRISPR-based epigenome editors rewrite these layers without cutting DNA, but most reports address editing efficiency and leave time unanswered. Target duration is a design input to be specified before an editor is chosen; the duration achieved remains an experimental outcome. Memory-window engineering defines a candidate immune edit by three parameters: duration after editor clearance, retention through cell division, and amplitude at single-cell resolution, with target-compartment division rate as an explicit modifier. For passively diluted marks, division rate constrains post-clearance duration; transcriptional and phenotypic persistence may follow different trajectories. Published results from methylation editing at IL1RN, FOXP3 enhancer demethylation, PD-1 enhancer tuning, exhaustion circuits and CRISPRoff programming of primary human T cells are interpreted using their readouts and observation windows; some persistence endpoints remain censored by study duration. Three testable predictions concern CpG density, transcript-versus-mark recovery and division-dependent loss, with competing mechanisms and measurement artefacts considered. Specificity, the nuclease-knockout comparison, and the ethics of durable edits are treated as parts of the same specification. Two design consequences follow. TET-dependent erasure requires oxygen and 2-oxoglutarate; reduced erasure in hypoxic or metabolically inhibitory tumour regions is therefore a hypothesis for edited immune cells. Sustained programming of circulating monocyte output requires an upstream progenitor strategy, whereas long-lived resident myeloid cells are a separate case. A worked CAR T example and five-item reporting minimum distinguish development-stage characterisation from lot-release testing.
No takes yet. Share an insight, caveat, or question.
Jack Riegler (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: