44. SETD1A LOSS OF FUNCTION DRIVES STAGE-SPECIFIC H3K4ME3 REDUCTION IN HUMAN NEURAL LINEAGES AND ASTROCYTE VULNERABILITY IN SCHIZOPHRENIA

Background Rare variants now define schizophrenia's highest-confidence risk genes, but how they reshape cellular biology during neurodevelopment remains opaque. SETD1A, SCHEMA's most significant rare-variant hit and the only chromatin regulator in that set, offers a tractable entry point. As a catalytic subunit of the COMPASS complex, SETD1A deposits H3K4me3 at active promoters. When SETD1A is lost, does it disrupt many neurodevelopmental promoters at once, including those of other risk genes? If so, when does it happen, and do its effects persist into mature cells? Methods From 4 SETD1A loss-of-function carriers in the SUPER-Finland cohort, 4 matched SETD1A+/+ controls, 2 HDR-corrected isogenic lines from those carriers, and 3 CRISPRi knockdowns in WTC-11 dCas9-KRAB iPSCs, we profiled iPSCs, neural progenitor cells (NPCs), mature astrocytes (day 28), and mature neurons (day 28) with and without LSD1 inhibition in biological triplicate (> 200 samples), all by paired H3K4me3 CUT and Tag and RNA-seq. To test whether early chromatin changes leave a lasting trace, day-14 astrocytes underwent confocal EdU imaging (proliferation) and hydrogen peroxide challenge (oxidative-stress tolerance) in triplicate across matched pairs. Results To ask whether SETD1A loss reshapes promoter chromatin, we profiled H3K4me3 by CUT and Tag (> 80% of peaks at promoters; reproducible TSS enrichment). In early NPCs (72 h post-differentiation) from carriers, we observed a broad leftward shift in H3K4me3 relative to controls, including at SCHEMA-associated loci. Affected promoters were enriched for neurodevelopmental and synaptic programs (GO biological process terms including nervous system development, axonogenesis, synapse organization, trans-synaptic signaling, etc), confirming SETD1A coordinates many neurodevelopmental gene programs at once. Both CRISPRi and HDR-corrected lines recapitulated this NPC pattern, establishing causality across two independent isogenic systems. By day 28, mature astrocytes had largely rebalanced their H3K4me3 signal. Yet at day 14, SETD1A loss-of-function astrocytes still showed reduced EdU incorporation across three matched donor pairs (p = 0.005 and 0.002 in two pairs; p = 0.062 in the third) and significantly diminished viability after acute hydrogen peroxide exposure at 350 and 500 µM in all three pairs (all p < 0.05). Something set in motion during the NPC chromatin window outlasted the chromatin mark itself. LSD1 inhibition partially rescued H3K4me3 in SETD1A-deficient NPCs, indicating the early defect is at least partly reversible. Discussion SETD1A behaves as a developmental-window regulator: it coordinates H3K4me3 across many promoter programs that build the brain, and its loss leaves a transient but consequential mark in early neural progenitor cells. The chromatin signature normalizes by mature astrocytes, yet at intermediate stages the cells already carry forward what happened. This temporal uncoupling, defect at one stage and phenotype at another, reframes chromatin-based psychiatric risk: early windows of vulnerability can program cellular liabilities that outlive the defect that caused them. LSD1 rescue is encouraging, but restoring chromatin alone may be insufficient once the developmental window has closed. Used as a molecular probe, SETD1A converts a SCHEMA association into a defined developmental mechanism and a candidate intervention window, charting a path from genetic discovery toward translatable intervention in psychiatry.