Incomplete cerebellar circuit restoration limits functional recovery following SMN therapy in severe spinal muscular atrophy

<jats:p>Spinal muscular atrophy (SMA) is caused by a deficiency in the survival motor neuron (SMN) protein, resulting in degeneration of spinal motor neurons (MNs). However, persistent neurological deficits despite postnatal SMN-restoring therapies suggest that recovery of sensorimotor and supraspinal circuits may be incomplete. The cerebellum has recently emerged as a supraspinal contributor to motor deficits in the severe SMNΔ7 mouse model, yet it remains unclear whether cerebellar pathology is a conserved and therapeutically reversible feature across severe SMA mouse models and clinical subtypes. Here, we identify cerebellar pathology in Taiwanese SMA mice, characterized by hypoplasia, disrupted organization and loss of Purkinje cells (PCs), altered synaptic circuitry, and impaired cerebellar cortical output. Unlike the previously described p53-dependent PC degeneration in SMNΔ7 mice, cerebellar pathology in Taiwanese SMA mice was associated with developmental disorganization and external granule layer (EGL)-restricted p53 activation. Human cerebellar tissue mirrored this distinction, with p53 activation found in PCs from SMA Type I and in the EGL from SMA Type 0 individuals, indicating that cerebellar pathology arises through distinct mechanisms across severe forms of SMA. Importantly, two SMN-restoring strategies produced divergent therapeutic outcomes. In SMNΔ7 mice, AAV9-SMN prevented PC degeneration yet incompletely restored cerebellar circuitry. AAV9-SMN-treated Taiwanese mice developed severe ataxia-like deficits, retained profound cerebellar pathology, and survived to approximately one month of age. In contrast, systemic risdiplam rescued cerebellar pathology, motor behavior, and survival in both models. Together, these findings identify cerebellar pathology as a conserved yet distinct feature across severe forms of SMA and reveal cell type-specific tropism as a critical determinant of therapeutic outcome. More broadly, these findings suggest that successful recovery requires restoration of distributed supraspinal circuit integrity in addition to rescue of spinal motor pathways.</jats:p>