NCALD Antisense Oligonucleotide Therapy in Addition to Nusinersen further Ameliorates Spinal Muscular Atrophy in Mice

Abstract

Spinal muscular atrophy (SMA) is a neuromuscular disease causing the most frequent genetic childhood lethality. Recently, nusinersen, an antisense oligonucleotide (ASO) that corrects SMN2 splicing and thereby increases full-length SMN protein, has been approved by the FDA and EMA for SMA therapy. However, the administration of nusinersen in severe and/or post-symptomatic SMA-affected individuals is insufficient to counteract the disease. Therefore, additional SMN-independent therapies are needed to support the function of motoneurons and neuromuscular junctions. We recently identified asymptomatic SMN1-deleted individuals who were protected against SMA by reduced expression of neurocalcin delta (NCALD). NCALD reduction is proven to be a protective modifier of SMA across species, including worm, zebrafish, and mice. Here, we identified Ncald-ASO3-out of 450 developed Ncald ASOs-as the most efficient and non-toxic ASO for the CNS, by applying a stepwise screening strategy in cortical neurons and adult and neonatal mice. In a randomized-blinded preclinical study, a single subcutaneous low-dose SMN-ASO and a single intracerebroventricular Ncald-ASO3 or control-ASO injection were presymptomatically administered in a severe SMA mouse model. NCALD reduction of >70% persisted for about 1 month. While low-dose SMN-ASO rescues multiorgan impairment, additional NCALD reduction significantly ameliorated SMA pathology including electrophysiological and histological properties of neuromuscular junctions and muscle at P21 and motoric deficits at 3 months. The present study shows the additional benefit of a combinatorial SMN-dependent and SMN-independent ASO-based therapy for SMA. This work illustrates how a modifying gene, identified in some asymptomatic individuals, helps to develop a therapy for all SMA-affected individuals.

Keywords: ASO therapy; NCALD; SMA; SMN1; SMN2; modifier gene; motor neuron disorder; neuromuscular disorder; neuromuscular junction; spinal muscular atrophy.

Figure 1

Target Regions of Tested Ncald-ASOs and Knockdown Efficiency in Spinal Cord and Brain of Adult Wild-Type Mice (A) Numbers of generated Ncald-ASOs and their respective target site in the mouse Ncald gene are shown. Used reference sequence is the Mus musculus strain C57BL/6J chromosome 15, GRCm38.p4 C57BL/6J (GenBank: NC_000081.6). Exons are labeled in yellow, introns in pink. (B) The 22 most efficient Ncald-ASOs were applied in adult wild-type mice by intracerebroventricular (i.c.v.) bolus injection and knockdown efficiency was determined by qRT-PCR of Ncald (primer probe set flanks exon 5–6 junction) relative to Gapdh (control) expression in spinal cord and brain. Ncald-ASOs marked by black triangle were tested in neonatal mice. (C) NCALD protein levels in the spinal cord and the brain of P10 animals (n = 4 animals per genotype) treated i.c.v. at P2 with 100 μg Ncald-ASO3 were more than 70% reduced compared to mice injected with 100 μg of Ctrl-ASO. Numbers on the left indicate respective band size in kDa. ACTB, loading control. Unpaired, two-tailed Student’s t test; ^∗p < 0.05, ^∗∗∗p < 0.001.

Figure 2

Experimental Design and Time Points of Individual Analyses A graph showing the breeding scheme to obtain severe SMA and HET mice on mixed₅₀ background (upper left), ASO treatments (blue boxes), and individual analyses at the indicated time points P21 and 3 and 6 months (MO). s.c., subcutaneous; i.c.v., intracerebroventicular; CMAP, compound muscle action potential; MUNE, motor unit number estimation; NMJ, neuromuscular junction.

Figure 3

Temporal Progression of NCALD Protein Levels in Mice Injected with SMN-ASO at P1 and Ctrl-ASO or Ncald-ASO3 at P2 Western blot analysis of NCALD and SMN levels in the spinal cord and the brain at (A) P21 and (B) 3 months after injection with respective ASOs (n = 3 per genotype and age). NCALD protein levels normalized to ACTB (loading control) are shown on the right side. NCALD levels are significantly decreased in both spinal cord and brain of HET and SMA animals co-injected with SMN-ASO and Ncald-ASO3 at P21 (A) but not at 3 months (B). Color legend for graphs is displayed at the bottom. For statistics, ordinary one-way ANOVA was applied with Tukey posthoc test for multiple comparisons; n.s., not significant, ^∗∗p < 0.01, ^∗∗∗p < 0.001.

Figure 4

Electrophysiological, Structural, and Motoric Analysis of Mice Injected with SMN-ASO at P1 and Ctrl-ASO or Ncald-ASO3 at P2 (A) Dot plots of CMAP and MUNE values at P21 in HET and SMA mice with mean ± SD. Animals used for CMAP: HET Ctrl-ASO n = 12; HET Ncald-ASO3 n = 13; SMA Ctrl-ASO n = 12; SMA Ncald-ASO3 n = 13. MUNE: HET Ctrl-ASO n = 11; HET Ncald-ASO3 n = 11; SMA Ctrl-ASO n = 9; SMA Ncald-ASO3 n = 9. (B) Representative NMJ images in HET and SMA animals at P21 showing postsynaptic NMJ region stained with bungarotoxin (BTX, magenta) and presynaptic nerve with neurofilament (NF, green). Scale bar: 10 μm. Graphs below the images show single dot plot values of NMJ areas of all grouped animals with mean ± SD and quantification of maturity of NMJs with mean values of mice per group ± SD. Statistics was performed with mean values of animals per group. n = 4 animals per group, 30–55 NMJs per animal. (C) Representative pictures and quantifications of hematoxylin and eosin-stained gastrocnemius (GC) muscle fibers from HET and SMA animals at P21. Scale bar: 50 μm. Graphs on the right show single dot plot values of GC areas of all grouped animals with mean ± SD (n = 4 animals per genotype, n = 50 fibers/mouse). For visualization, muscle fibers were grouped according to the area intervals of 100 μm². (D) Dot plot of grip strength in HET and SMA animals at 3 months. Single values of each animal are shown as mean ± SD. Number of animals used are: HET Ctrl-ASO n = 14, HET Ncald-ASO3 n = 9, SMA Ctrl-ASO n = 10, SMA Ncald-ASO3 n = 10. Color legend for all graphs is displayed in (A). For all analyses, ordinary one-way ANOVA with Tukey posthoc test for multiple comparisons was applied. n.s., not significant, ^∗p < 0.05, ^∗∗p < 0.01, ^∗∗∗p < 0.001.