Single-nucleus transcriptome atlas of orbitofrontal cortex in ALS with a deep learning-based decoding of alternative polyadenylation mechanisms

Amyotrophic lateral sclerosis (ALS) and frontotemporal lobar degeneration (FTLD) are fatal neurodegenerative diseases sharing clinical and pathological features. Both involve complex neuron-glia interactions, but cell-type-specific alterations remain poorly defined. We performed single-nucleus RNA sequencing of the frontal cortex from C9orf72-related ALS (with and without FTLD) and sporadic ALS (sALS). Neurons showed prominent changes in mitochondrial function, protein homeostasis, and chromatin remodeling. Comparison with independent datasets from other cortical regions revealed consistent pathway alterations, including upregulation of STMN2 and NEFL across brain regions and subtypes. We further examined dysregulation of alternative polyadenylation (APA), an understudied post-transcriptional mechanism, uncovering cell-type-specific APA patterns. To investigate its regulation, we developed the alternative polyadenylation network (APA-Net), a multi-modal deep learning model integrating transcript sequences and RNA-binding protein (RBP) expression profiles to predict APA. This atlas advances our understanding of ALS/FTLD molecular pathology and provides a valuable resource for future mechanistic studies.