<EM>IN VITRO</EM> AND <EM>IN VIVO</EM> CHARACTERIZATION OF HOW<EM> </EM>ALS MUTATION ON THE RNA-BINDING PROTEIN FUS CHANGES GENE EXPRESSION AND SYNAPSES IN HUMAN STEM CELL-DERIVED NEURONS

Fused in Sarcoma (FUS) gene mutation is one of the leading genetic causes of amyotrophic lateral sclerosis (ALS). This gene encodes a multi-functional DNA/RNA-binding protein. In post-mortem brain samples and human induced pluripotent stem cell (hiPSCs)-derived neurons, the ALS-related FUS mutant proteins form cytoplasmic aggregates or inclusions which are believed to cause neuronal damage. Most human FUS-ALS studies involve differentiating hiPSCs into spinal motoneurons. Given the non-motor deficits of transgenic mouse models of FUS-ALS, it is imperative to determine how ALS-FUS mutations affect hiPSC-derived neurons in the brain. Here we employ CRISPR/Cas9 to knock-in missense mutation (F525L) into the nuclear localization signal of the FUS protein. Compared to wild-type FUS which is mostly in the nucleus, the P525L-FUS mutant exhibits diffuse cytoplasmic localization in the iPSC-derived cortical neurons. Despite lacking FUS aggregate, the mutant neurons possess significantly fewer dendrite arborizations and PSD-95 puncta. RNAseq identified massive differences in expression of mRNAs that encode synaptic proteins between neurons of the two FUS genotypes. To study the effects of the FUS mutations on human neurons in vivo, GFP-expressing neural stem cells were transplanted into the prefrontal cortex and dorsal hippocampus of NOD/SCID mice. Dendritic spines are observed in the grafted hiPSC-derived cortical neurons 4 months after transplantation, and the P525L-FUS mutant neurons possess significantly fewer dendritic spines than the control neurons. Our study provides new insights into human ALS-FUS pathology by showing that the P525L-FUS missense mutation can lead to synaptic phenotypes independently without cytoplasmic FUS aggregation.