Whether neurodevelopmental defects contribute to selective neuronal vulnerability in neurodegenerative diseases is an emerging and intriguing question. We addressed this in the context of Spinal Muscular Atrophy (SMA), an early-onset ND affecting spinal motor neurons (MNs). SMA is caused by the deficiency of Survival of Motor Neuron (SMN) protein, which is critical for RNA splicing and highly expressed during development. We thus hypothesized that SMN deficiency alters development and renders MNs vulnerable to stress.
To test this, we established a guided ventral neuromuscular organoid (vNMO) protocol from neuromesodermal progenitors (NMPs) and applied it to a cohort of isogenic hiPSC lines, including two healthy controls, two SMA type 1 patient-derived lines, and their isogenic controls in which SMN levels were restored via CRISPR-Cas9 editing. This 3D system enabled modeling of spinal cord and neuromuscular development with lineage fidelity.
To uncover potential alterations throughout development we performed longitudinal scRNA-seq of vNMOs at an early timepoint corresponding to neuromesodermal lineage acquisition, and a late timepoint in which MNs and myocytes are present. Cell clustering revealed a mesodermal fate bias in early SMA vNMOs compared to the healthy vNMOs and rescued in isogenic controls. This phenotype was mirrored in E10.5 SMA mouse embryos, which showed reduced neural tube area and increased mesodermal tissue compared to healthy littermates. In later-stage vNMOs, this skewed lineage allocation resulted in reduced MN output, potentially impairing neuromuscular junction formation. Mechanistically, SMA NMPs showed decreased SOX2 and upregulated canonical WNT signaling, two key regulators of NMP fate. WNT inhibition partially reversed mesodermal bias in SMA vNMOs. We also identified an SMN-dependent isoform switch in a known SOX2 regulator. Knockdown of this regulator restored SOX2 levels and prevented mesodermal lineage bias in SMA vNMOs. Notably, continuous knockdown of this regulator was sufficient to prevent MN death in late SMA vNMOs, which indicates the necessity to tackle early alterations in development to fully preserve MNs postnatally.
Taken together, we found a novel developmental role for SMN in NMP lineage specification and demonstrate how organoid models can uncover early cellular mis-specifications underlying neuromuscular diseases. This approach provides novel therapeutic entry points for developmentally rooted disorders.