ATF3-dependent formation of inclusion bodies in polyQ-expressing human iPSC-derived neurons confers cellular protection [RNAseq]

Huntington’s disease (HD) is an incurable, neurodegenerative disorder. While the causative mutation – CAG expansions within the coding region of the Huntingtin (HTT) gene – has been identified over 30 years ago, the pathological mechanisms underlying HD are still not clear. The abnormal CAG track encodes a polyglutamine (polyQ) expanded protein, which leads to HTT protein misfolding. These polyQ aggregates can form insoluble inclusion bodies (IBs); however, whether IBs are protective or detrimental remains debatable. To ask whether IBs are protective or harmful in human neuronal cells, and to elucidate the pathological pathways involved in polyQ IB formation, we developed an iPSC-based human neuronal models for polyQ-related disorders. We ectopically expressed the first 21 exons of the HTT gene, containing either 18Q or 105Q fused with GFP in HD patient-derived iPSCs containing 180Q or their isogenic corrected cells, respectively, differentiated them into neuronal progenitor cells (NPCs), and enriched for a pure population of polyQ IBs-containing cells (IB+). Comparing gene expression profiles between IB+ and IB- NPCs, growing side-by-side, in both systems, we identified ATF3 as responsible for inducing neuroinflammatory genes specifically in IB+ cells. ChIP-seq for ATF3 confirmed direct binding at their promoters, exclusively in IB+ NPCs. Remarkably, knocking out (KO) ATF3 completely prevented polyQ-IB formation in NPCs and rendered the cells more vulnerable to induced stress. Finally, using time-lapse imaging comparing cell death in IB+ and IB- NPCs, we demonstrate that polyQ IBs have a significant protective effect. Taken together, our results highlight ATF3’s role in polyQ IB formation in human NPCs, acting as an activator of inflammatory genes, and demonstrate that polyQ IBs protect cells from stress-induced death.