Motor Neuron-Specific Restoration of SMN in Two SMA Mouse Models: Insights into the Role of Motor Neurons in Spinal Muscular Atrophy

Spinal muscular atrophy (SMA) results from α-motor neuron loss in the spinal cord due to low levels of the survival of motor neuron (SMN) protein, required for proper spliceosome assembly. The reduced levels of SMN cause muscle atrophy and ultimately death in the most severe cases. Although mouse models of SMA recapitulate many features of the human disease, it is still unclear whether their phenotypes are primarily due to motor neuron deficits. If so, motor neuron-selective restoration of normal SMN levels should have a great positive impact on SMA mice. To test this, we first exogenously raised normal human SMN in severe SMA mice that die perinatally, by driving its expression selectively in motor neurons with an Hb9 promoter. We found no extension of survival. We detected motor neuronal-SMN protein expression in Hb9-SMN transgenic mice from mid embryogenesis to postnatal day 6. However, mRNA for transgenic SMN was undetected by late embryogenesis. These results suggest that expression of Hb9-SMN declines before SMN levels are most needed perinatally. Second, we increased endogenous motor neuronal-SMN expression following embryonic Hb9-dependent Cre recombination of a conditional hybrid mutant allele (Smn^(res)) in another severe SMA mouse model (SMA∆7-like). Cre recombination irreversibly transforms the Smn^(res) allele to WT. We confirmed that recombination of Smn^(res) occurred exclusively in the spinal cord. Yet, unlike a previous study that used choline acetyltransferase (ChAT) as a driver on the same mice, we found no improvement in survival, weight, motor behavior or presynaptic neurofilament accumulation. However, like in ChAT^(Cre+) SMA mice, we detected rescue of endplate size and mitigation of neuromuscular junction (NMJ) denervation status. Real time-PCR showed that the expression of spinal cord SMN transcript was sharply reduced in Hb9^(Cre+) SMA mice relative to ChAT^(Cre+) SMA mice. This suggests that our lack of overall phenotypic improvement was most likely due to an unexpectedly poor recombination efficiency driven by Hb9^(Cre). Nonetheless, the low levels of SMN were sufficient to rescue two NMJ structural parameters indicating that these motor neuron cell-autonomous phenotypes are very sensitive to changes in motor neuronal-SMN levels