Motor unit recovery following Smn restoration in mouse models of spinal muscular atrophy

Abstract

Spinal muscular atrophy (SMA) is a childhood motor neuron disease caused by anomalies in the SMN1 gene. Although therapeutics have been approved for the treatment of SMA, there is a therapeutic time window, after which efficacy is reduced. Hallmarks of motor unit pathology in SMA include loss of motor-neurons and neuromuscular junction (NMJs). Following an increase in Smn levels, it is unclear how much damage can be repaired and the degree to which normal connections are re-established. Here, we perform a detailed analysis of motor unit pathology before and after restoration of Smn levels. Using a Smn-inducible mouse model of SMA, we show that genetic restoration of Smn results in a dramatic reduction in NMJ pathology, with restoration of innervation patterns, preservation of axon and endplate number and normalized expression of P53-associated transcripts. Notably, presynaptic swelling and elevated Pmaip levels remained. We analysed the effect of either early or delayed treated of an antisense oligonucleotide (ASO) targeting SMN2 on a range of differentially vulnerable muscles. Following ASO administration, the majority of endplates appeared fully occupied. However, there was an underlying loss of axons and endplates, which was more prevalent following a delay in treatment. There was an increase in average motor unit size following both early and delayed treatment. Together this work demonstrates the remarkably regenerative capacity of the motor neuron following Smn restoration, but highlights that recovery is incomplete. This work suggests that there is an opportunity to enhance neuromuscular junction recovery following administration of Smn-enhancing therapeutics.

Graphical abstract

Figure 1

There is significant NMJ pathology at P4 in Smn^Res/Res mice before Smn restoration. (A, B) Bar chart (meanSEM) showing significant decrease in body weight (A) and significant increase in time taken to self-right (B) in SMA (Smn^Res/Res;SMN2;SMNΔ7) mice compared with control (Smn^+/Res;SMN2;SMNΔ7 and Smn^+/+;SMN2;SMNΔ7) at P4. (C) Images showing neuromuscular junctions (NMJs) in the TVA muscle labelled with antibodies against NF and SV2(green) and BTX (magenta) in control and SMA mice at P4. Note presence of denervated endplates (BTX lacking presynaptic terminal) in SMA which does not occur in control. Scale Bar = 16 μm (D) Bar chart (meanSEM) showing percentage of fully occupied, partially occupied and vacant endplates in AS, TVA and Levator Auris Longus Caudal (LALc) muscles in control and SMA mice at P4. Note increase in vacant and partially occupied endplates in SMA and corresponding decrease in fully occupied endplates. (E) Bar chart (meanSEM) showing percentage of NMJs with no, mild, moderate or severe swelling in AS, TVA and LALc muscles in control and SMA mice at P4. Note increase in presynaptic swelling in SMA and corresponding decrease NMJs with no swelling. (F) Bar chart (meanSEM) showing increase in relative transcript levels of P53, Cdkn1a, Fas and Pmaip in the spinal cord of P4 SMA mice compared with control. ^*P < 0.05; ^**P < 0.01, ^***P < 0.001, ^****P < 0.0001 by Student’s t test (A, B) or Mann–Whitney U test (D, E, F). N = 3/4 mice, 6 muscles per genotype.

Figure 2

Restoration of Smn levels at P4 results in a significant rescue of NMJ pathology by P10. (A, B) Line chart (meanSEM) showing body weight (A) and time taken to self right (B) in Rescue (Smn^Res/Res;SMN2;SMNΔ7;Cre^+/−), SMA (Smn^Res/Res;SMN2;SMNΔ7;Cre^−/−) and control mice at P4, 6, 8 and 10 days of age following treatment with Tamoxifen at P4. (C) Images showing neuromuscular junctions (NMJs) in the AS muscle labelled with antibodies against NF and SV2 (green) and BTX (magenta) in Rescue, Control and SMA mice at P10. Note presence of denervated endplates in SMA which was significantly reduced in Rescue mice. Scale bar = 20 μm. (D) Bar chart (mean ± SEM) showing percentage of fully occupied, partially occupied and vacant endplates in AS, TVA and Levator Auris Longus Caudal (LALc) muscles in Control, SMA and Rescue mice at P10. Stats show significant increase in percentage of fully occupied endplates in Rescue mice compared with SMA. (E) Bar chart (mean ± SEM) showing percentage of NMJs with no, mild, moderate or severe swelling in AS, TVA and LALc muscles in Control, SMA and Rescue mice at P10. Stats show comparison of percentage of NMJs with no swelling, comparing Rescue to SMA. (F) Bar chart (mean ± SEM) showing average endplate area in AS, TVA and LALc muscles in Control, SMA and Rescue mice at P10. (G) Bar chart (mean ± SEM) relative transcript levels of P53, Cdkn1a, Fas and Pmaip in the spinal cord of P10 Control, SMA and Rescue mice. Stats show SMA versus control and Rescue versus SMA. Note significant reduction in Cdkn1a and Fas in Rescue Mice compared with SMA compared with control. ^*P < 0.05; ^**P < 0.01, ^***P < 0.001, ^****P < 0.0001 by two-way ANOVA (A,B), one-way ANOVA with Tukey’s post hoc test (F) or Kruskal–Wallis with Dunn’s Post Hoc (D, E, G). N = 12/3/3 mice per Control/Rescue/SMA (A, B); N = 6/3/6 muscles per Control/SMA/Rescue (D–F) and N = 3 mice per genotype (G).

Figure 3

Recovery from denervation is due to reinnervation of denervated endplates. (A–C) Montages fluorescent micrographs showing C2 region of LALc muscle labelled with antibodies against NF and SV2 (green) and BTX (magenta) in Control (A), SMA (B) and Rescue (C) mice at P10. Scale Bar = 200 μm. (D) Bar chart (meanSEM) showing significant decrease in total endplate number in SMA compared with both Control and SMA, but no difference in total endplate number in Rescue versus Control. ^*P < 0.05; ns, non-significant by Student’s t test. N = 6/5/6 muscles per Control/SMA/Rescue.

Figure 4

Smn restoration at P4 preserves axon number and motor unit size. (A, C) Images showing axon bundles (A) and neuromuscular junctions (NMJs) (C) labelled with antibodies against NF and SV2 (green) and BTX (magenta) in control, SMA and Rescue mice at P10. Scale bar = 20 μm (A) and 50 μm (C). (B, D) Bar chart (meanSEM) showing number of axons (B) or average motor unit size (D) in AS muscle in Control, Rescue and SMA mice at P10. Note decrease in axon number and increase in motor unit size in SMA which was not seen in Rescue mice ^*P < 0.05; ^**P < 0.01, ^***P < 0.001, ns non-significant by Student’s t test (B) or Mann–Whitney U test (C). N = 5/6/5 muscles per Control/SMA/Rescue.

Figure 5

Delayed treatment with a Smn-inducing ASO leads to more widespread loss of axons and endplates. (A) Confocal images showing NMJs labelled with antibodies against NF and SV2 (green) and BTX (magenta) in SmnΔ7 mice treated with ASO at P1 (ASO1) or P6 (ASO6) with their respective controls. Muscles used are the LALr (least vulnerable), AAL (moderately vulnerable) and AS (most vulnerable). Vacant and partially occupied endplates in the AS after early treatment (ASO1) are marked by an asterisk and arrow heads, respectively. Scale bar = 20 μm. (B–D) Bar charts (mean ± SEM) showing the percentage of fully occupied (black), partially occupied (grey) or vacant (white) endplates in P12 Smn^−/−;SMN2;SMNΔ7 mice treated with ASO at P1 (ASO1) or P6 (ASO6) compared with similarly treated controls. Muscles used are the LALr (least vulnerable), AAL (moderately vulnerable) and AS (most vulnerable). (E–J) Bar charts (mean ± SEM) showing the number of axons (E–G) and innervated NMJs (H–J) in the three muscle groups in P12/P17 Smn^−/−;SMN2;SMNΔ7 mice following early (ASO1) or delayed (ASO6) treatment compared with controls. Note the significant loss of axons in the moderately vulnerable AAL muscle following delayed P6 treatment which was not seen in the early P1-treated cohort. Significant loss of both axons and endplates was seen in the severely vulnerable AS muscle following both P1 and P6 treatment. ^*P < 0.05; ^**P < 0.01, ^***P < 0.001, ^****P < 0.0001, ns non-significant by Mann–Whitney U test. N = 3–10 muscles per group.

Figure 6

Axonal sprouting leads to increased motor unit size in both vulnerable and moderately resistant muscles following ASO treatment. (A) Confocal images showing NMJs labelled with antibodies against NF and SV2 (green) and BTX (magenta) in the AS muscle of a P12 Smn^−/−;SMN2;SMNΔ7 mouse treated with ASO at P1. Note presence of large motor units (Ai) and evidence of significant presynaptic remodelling and terminal sprouting (Aii and (iii) respectively). Scale bar = 150 μm (Ai), 20 μm (Aii and Aiii). (B, D) Bar charts (mean ± SEM) showing average motor unit size in LALr, AAL and AS muscles of mice treated with ASO at P1 (ASO1) or P6 (ASO6) compared with similarly treated controls. Note that the significant loss of axons in the moderately vulnerable AAL muscle is masked by compensatory sprouting leading to enlarged motor units (B). ^*P < 0.05; ^**P < 0.01, ^***P < 0.001, ^****P < 0.0001, ns non-significant by Mann–Whitney U test. N = 3–10 muscles per group.