Gene therapy ameliorates neuromuscular pathology in CLN3 disease

Abstract

CLN3 disease is a neuronopathic lysosomal storage disorder that severely impacts the central nervous system (CNS) while also inducing notable peripheral neuromuscular symptoms. Although considerable attention has been directed towards the neurodegenerative consequences within the CNS, the involvement of peripheral tissues, including skeletal muscles and their innervation, has been largely neglected. We hypothesized that, CLN3 deficiency could directly influence peripheral nerves and investigated the neuromuscular system in Cln3^Δex7/8 mice. Our study found no overt loss of sciatic nerve axons or demyelination in 18-month-old Cln3^Δex7/8 mice at disease endstage, but a marked reduction of terminal Schwann cells (tSCs) at lower limb neuromuscular junctions (NMJs), culminating in progressive denervation of these NMJs which appeared abnormal. This led us to investigate skeletal muscle where we found significant myofiber atrophy and decreased and misplaced myofibril nuclei. Similar myopathic alterations were present in a muscle biopsy from an 8-year-old human CLN3 patient shortly after diagnosis. To assess a potential therapeutic intervention, we administered intravenous gene therapy using AAV9.hCLN3 to neonatal Cln3^Δex7/8 mice, which at disease endstage, entirely prevented tSC loss and NMJ abnormalities, while also averting skeletal muscle atrophy. These findings underscore the underappreciated, yet substantial effects of CLN3 disease beyond the CNS, highlighting peripheral neuromuscular pathologies as novel features of this disorder. Our findings also indicate that these manifestations could be amenable to treatment via gene therapy, opening new therapeutic strategies in the management of CLN3 disease.

Keywords: AAV9 gene therapy; CLN3 disease; Muscle atrophy; Neuromuscular junction; Peripheral nervous system; Terminal Schwann cells.

Fig. 1

Sciatic nerve in Cln3^Δex7/8 mice. Semi-thin (1 μm) sections of the sciatic nerve stained with toluidine blue show no significant (ns) differences in the number of axon, axon area or its distribution, or myelin area in 18 months (mo) Cln3^Δex7/8 mice vs. age matched wildtype (WT) controls. Arrow indicates an area of vacuolation, which affected less than 1% of fibers. Scale bar = 100 μm for lower power view; 1 μm for higher power view. Histograms show measurements of axon number, axon area (and its distribution) and myelin area made from sciatic nerve sections. Data ± SEM. The number of mice (n=) is shown below the X-axis. Unpaired t-test.

Fig. 2

Neuromuscular junction (NMJ) pathology in Cln3^Δex7/8 mice. (A) Representative images of NMJ morphology in the extensor digitorum longus (EDL) muscle. Immunofluorescence staining for acetylcholine receptors (BTX = α-bungarotoxin for AChRs; red), terminal Schwann cells (tSCs, S100B; green), and neurofilament (NF200; green) shows altered NMJ innervation patterns in Cln3^Δex7/8 mice, with fragmented Schwann cell coverage. Scale bar 20 = µm. (B) Quantification reveals a significant reduction in the number of tSCs, an increase in NMJs with fragmented tSCs and atypical AChRs in Cln3^Δex7/8 mice vs. age wildtype (WT) at 12 and 18 months (mo). (C)Cln3^Δex7/8 mice show a progressive loss of fully innervated NMJs compered to age matched WT mice. NMJs were defined as fully innervated if nerve terminals occupied ≥ 75% of endplate area. An NMJ was determined as partially innervated if nerve terminals cover < 75% of the α-BTX stained endplate. Denervation was defined as no neurofilaments overlapping α-BTX staining. NMJ percentages were normalized to the total NMJ number in the microscopic field of view. Data ± SEM. The number of mice (n=) is shown below the X-axis. Unpaired t-test (B, C), one-way ANOVA with a post-hoc Bonferroni correction, * p ≤ 0.05, ** p ≤ 0.01, **** p ≤ 0.0001. In (C) * p ≤ 0.05 (% fully innervated NMJs, 12-month Cln3^Δex7/8 mice vs. age-matched WT mice); # p ≤ 0.05 (% denervated NMJs, 18-month Cln3^Δex7/8 mice vs. age-matched WT mice); § p ≤ 0.05 (% fully innervated MJs, 18-month Cln3^Δex7/8 mice vs. age-matched WT mice)

Fig. 3

Evidence for myopathy in skeletal muscles in Cln3^Δex7/8 mice. (A) Hematoxylin and eosin staining reveals the histopathological changes in Cln3^Δex7/8 quadriceps muscle compared to age matched wildtype (WT) controls at 18 months. Arrow indicates displaced myofibril nucleus. Changes were also pronounced in extracellular deposition of ß-Laminin (green) between myofibers in Cln3^Δex7/8 mice vs. WT at 18 months of age. Sections were counterstained with the nuclear stain DAPI (blue). Scale bar = 100 μm, 50 μm in inserts. (B) The muscle fiber diameter and total number (No.) of nuclei per myofiber cross-section was measured in the quadriceps muscles at disease endstage. A size distribution plot of individual myofiber diameters reveals the atrophy of myofibers in Cln3^Δex7/8 mice, with a shift to smaller size vs. the distribution of myofiber diameters in WT mice. Unpaired t-test, * p ≤ 0.05, **** p ≤ 0.0001. Data ± SEM. The number of mice (n=) is shown below the X-axis

Fig. 4

Evidence for myopathy of bowel smooth muscle in Cln3^Δex7/8 mice. (A) Hematoxylin and eosin staining reveals the histopathological changes in Cln3^Δex7/8 smooth muscle layers in jejunum compared to age matched wildtype (WT) controls at 18 months. Scale bar = 200 μm (full thickness sections), 50 μm in inserts (closer view of smooth muscle layers). (B) Analysis of bowel demonstrated atrophy of smooth muscle layers and decreased smooth muscle area in jejunum Cln3^Δex7/8 mice vs. WT at 18 months. Unpaired t-test, *** p ≤ 0.001, **** p ≤ 0.0001. Data ± SEM. The number of mice (n=) is shown below the X-axis

Fig. 5

Evidence for myopathy in an early-stage human CLN3 skeletal muscle biopsy. Hematoxylin and eosin staining reveals evidence for myopathy in a human CLN3 gastrocnemius muscle biopsy collected approximately one year after diagnosis. This biopsy shows evidence of reduced muscle fiber cross-sectional area and accumulation of both subunit c of mitochondrial ATP synthetase (SCMAS, red) immunoreactivity and acid phosphatase positive inclusions (brown). These phenotypes are absent in a similarly aged control biopsy sample. Scale bar = 50 μm. Dotted lines show the boundaries of individual myofibrils

Fig. 6

Treatment effect of gene therapy upon neuromuscular junction (NMJ) pathology in Cln3^Δex7/8 mice at 18 months (mo). (A) Representative images of NMJ morphology in the extensor digitorum longus (EDL) muscle in untreated Cln3^Δex7/8 mice, Cln3^Δex7/8 mice treated with AAV9.hCLN3 and age matched wildtype (WT). Immunofluorescence staining for acetylcholine receptors (BTX = α-bungarotoxin for AChRs; red), terminal Schwann cells (tSCs, S100B; green), and neurofilament (NF200; green), with DAPI for nuclear staining (blue). Scale bar 20 = µm. (B) Quantification of NMJs shows the protective effect of gene therapy upon the number of tSCs per NMJ, a decreased proportion of NMJs with fragmented tSCs and atypical AChRs. Data ± SEM. * p  ≤ 0.05; **** p  ≤ 0.0001; ns = Not significant. The number of mice (n=) is shown below the X-axis. (C) Innervation analysis showed an increase in fully innervated NMJs in 18-month-old Cln3^Δex7/8 mice treated as neonates with AAV9.hCLN3 compared to untreated Cln3^Δex7/8 mice and WT mice at 18 months. NMJs were defined as fully innervated if nerve occupied ≥ 75% of endplate area. NMJ percentages were normalized to the total NMJ number in the microscopic field of view. Data ± SEM. The number (n=) of mice is shown below the X-axis. Unpaired t-tests, * p  ≤ 0.05 (% fully innervated, 12-month Cln3^Δex7/8 mice vs. age-matched WT mice); #, p  ≤ 0.05 (% denervated, 18-month Cln3^Δex7/8 mice vs. age-matched WT mice); §, p  ≤ 0.05 (% fully innervated, 18-month Cln3^Δex7/8 mice vs. age-matched WT mice); ø, p  ≤ 0.05 (% fully innervated, 18-month AAV9.hCLN3 treated Cln3^Δex7/8 vs. 18-month untreated Cln3^Δex7/8 mice); øø, p  ≤ 0.001 (% partially innervated, 18-month AAV9.hCLN3 treated Cln3^Δex7/8 vs. 18-month untreated Cln3^Δex7/8 mice); $, p  ≤ 0.05 (% partially innervated, 18-month AAV9.hCLN3 treated Cln3^Δex7/8 vs. age-matched WT); Ω, p  ≤ 0.05 (% fully innervated, 18-month AAV9.hCLN3 treated Cln3^Δex7/8 vs. age-matched WT). Scale bar 20 μm. Figure  2 A data are reprinted here to simplify comparison to AAV9.hCLN3 treated mice

Fig. 7

Treatment effect of gene therapy upon myopathy in Cln3^Δex7/8 mice at 18 months (mo). (A) Analysis of quadriceps muscle shows less deposition of ß-Laminin (green) between myofibrils in AAV9.hCLN3 treated Cln3^Δex7/8 mice than in untreated Cln3^Δex7/8 mice. This treatment also prevented the reduction in muscle fiber diameter (histogram of average diameter and size-distribution line graph), while the average number of nuclei per myofiber remained significantly less than in wildtype (WT) mice. Sections counterstained with the nuclear stain DAPI (blue). Scale bar = 50 μm. (B) AAV9.hCLN3 gene therapy prevented thinning of bowel smooth muscle in the jejunum of Cln3^Δex7/8 mice at 18 months of age. Data ± SEM. The number of mice (n =) is shown below the X-axis. One-way ANOVA with a post-hoc Bonferroni correction (A, B), * p ≤ 0.05, ** p ≤ 0.01, *** p ≤ 0.001, **** p ≤ 0.0001. ns = Not significant. Figure 3A data are reprinted here to simplify comparison to AAV9-hCLN3 treated mice