Symptomatic treatment by a BBB-permeable AAV engineered to restore TDP-43 function slows motor neuron disease and prevents paralysis

Abstract

TAR DNA-binding protein 43kDa (TDP-43) dysfunction is an early pathogenic mechanism that underlies amyotrophic lateral sclerosis (ALS), a devastating neurodegenerative disorder that lacks disease modifying therapies. We previously developed a mouse model in which TDP-43 is selectively deleted from motor neurons (ChAT-Cre;Tardbp ^f/f ) that mimics the early stages of ALS. Here, we demonstrate that intravenous delivery of a blood-brain-barrier (BBB) permeable AAV capsid expressing our rationally designed splicing repressor CTR (AAV-PHP.eB-CTR) in symptomatic ChAT-Cre;Tardbp ^f/f mice markedly slowed disease progression and prevented paralysis. Systemic delivery of AAV-PHP.eB-CTR led to transduction of ~80% of spinal motor neurons, repression of TDP-43-associated cryptic exons within motor neurons expressing CTR, and attenuation of motor neuron loss. Notably, the addition of the TARDBP 3'UTR autoregulatory element to CTR maintained its expression within a physiological range. In control littermates that received AAV-PHP.eB-CTR and were monitored for >20 months, grip strength and body weight remained normal, and no histopathological abnormalities were observed, underscoring a favorable safety profile for this gene therapy. These results provide preclinical proof-of-concept that BBB-crossing AAV delivery of CTR can rescue motor neuron disease through the restoration of TDP-43 function, offering a promising mechanism-based therapeutic strategy for ALS.

Keywords: Amyotrophic Lateral Sclerosis; Blood-brain-barrier permeable AAV; Cryptic exons; Gene therapy; Motor neuron; Splicing repressor; Symptomatic treatment; TDP-43 autoregulatory element; TDP-43 dysfunction.

Fig. 1:. Intravenous delivery of AAV-PHP.eB-CTR confers high efficiency of transduction in adult motor neurons.

(A) Schematic diagram showing the study design. (B) Representative images showing the accumulation of CTR in ventral motor neurons of ChAT-Cre;Tardbp^f/f mice. CTR protein staining was carried out using the anti-human TDP-43 antibody (indicated by arrows in Panel 2). The administration of AAV-PHP.eB-CTR via tail vein injection results in a mean infectivity rate of 79±6.2% (n=3). Immunostaining for mouse TDP-43 indicates ~60% depletion of endogenous TDP-43 (indicated by arrowheads in Panels 2 and 3) and CTR as detected by anti-human TDP-43 in lumbar spinal cord sections of three- and five-month-old ChAT-Cre;Tardbp^f/f mice. Panel 1. AAV-PHP.eB-CTR treated ChAT-Cre;Tardbp^f/+ mice, Panel 2. AAV-PHP.eB-CTR treated ChAT-Cre;Tardbp^f/f mice and Panel 3. AAV-PHP.eB-GFP treated ChAT-Cre;Tardbp^f/f mice. Scale bar=50 μm. (C) Single cell quantification of AAV genome copy number (i) and CTR mRNA expression (ii) using BaseScope-ISH probes and immunohistochemical staining of CTR (iii) show that the autoregulatory domain expressed along with CTR prevents the overexpression of CTR in cells with normal expression of TDP-43. (D) Representative images showing accumulation of CTR in the facial motor nuclei of ChAT-Cre;Tardbp^f/f mice. The administration of AAV-PHP.eB-CTR via tail vein injection results in a mean infectivity rate of 72±11.7% (n=3). Arrows indicate anti-human TDP-43 immunostaining for CTR, and arrowheads indicate endogenous TDP-43 in Panels 2 and 3. Panel 1. AAV-PHP.eB-CTR treated ChAT-Cre;Tardbp^f/+ mice, Panel 2. AAV-PHP.eB-CTR treated ChAT-Cre;Tardbp^f/f mice and Panel 3. AAV-PHP.eB-GFP treated ChAT-Cre;Tardbp^f/f mice. For all groups, n=3. Scale bar=50 μm.

Fig. 2:. Efficacy of intravenous delivery of AAV-PHP.eB-CTR to ChAT-Cre;Tardbp^f/f mice after symptom onset.

(A) Grip strength analysis using performance in hanging wire experiments revealed marked amelioration in motor deficits (characteristic of the TDP-43 knockout mice) in CTR treated adult ChAT-Cre;Tardbp^f/f mice (*p<0.0001, two-way ANOVA test). (B) Intravenous administration of CTR resulted in the mitigation of behavioral deficits. Significant improvement was observed in the age-dependent reduction of body weight associated with the ChAT-Cre;Tardbp^f/f mice (*p<0.0001, two way ANOVA using Turkey’s multiple comparison test). (C) Intravenous administration of CTR resulted in the extension of survival. Kaplan-Meier survival curve of ChAT-Cre;Tardbp^f/+ and ChAT-Cre;Tardbp^f/f mice administered AAV-PHP.eB expressing either CTR or GFP as control. Data from male and female cohorts aged between 23–24 months are shown together. For all groups, n=15.

Fig. 3:. AAV-PHP.eB-CTR prevents inclusion of TDP-43 cryptic exons.

(A) Representative images of simultaneous detection of the cryptic exon in Synj2bp using BaseScope-ISH probe and antibodies against ChAT or mouse TDP-43 reveals an elevated presence of transcripts containing Synj2bp cryptic exon in some ChAT+ neurons or in neurons lacking mouse TDP-43 in ChAT-Cre;Tardbp^f/f mice. White arrows indicate Synj2bp cryptic exon in ChAT+ neurons. Black arrows indicate Synj2bp cryptic exon in TDP-43 deficient cells. Scale bar=100 μm. (B) Representative images of co-detection of cryptic exon in Synj2bp with CTR (using anti-human TDP-43) shows the rescue of cryptic exon expression in CTR expressing cells in ChAT-Cre;Tardbp^f/f mice. Arrowheads indicate CTR expressing neurons, and black arrows show CTR or mouse TDP-43 deficient cells expressing transcripts containing Synj2bp cryptic exon (n=5). Scale bar=100 μm. (C) Representative images from BaseScope-ISH or Co-detection assay to detect Synj2bp cryptic exon in facial motor nuclei. Arrows depict TDP-43 deficient cells containing transcripts with Synj2bp cryptic exons in ChAT-Cre;Tardbp^f/f mice (n=5). Scale bar=50 μm. (D) Quantitative analysis of ChAT positive motor neurons expressing Synj2bp cryptic exon in ChAT-Cre;Tardbp^f/f mice treated with AAV-PHP.eB-GFP and AAV-PHP.eB-CTR (n=5 per group).

Fig. 4:. AAV-PHP.eB-CTR attenuates motor neuron loss, axon degeneration, and skeletal muscle atrophy.

(A) Representative images from pathological analysis showing that CTR protects against motor neuron loss in adult ChAT-Cre;Tardbp^f/f mice treated with AAV-PHP.eB-CTR at symptomatic stage (treated at 1.5 months of age and analyzed at 5 months of age, n=8). (B) Quantification of number of ChAT positive motor neurons in the lumbar spinal cord (n=8). Scale bar=50 μm. (C) Representative toluidine blue-stained light microscopic images of transversely sectioned motor roots (n=3). (D) A notable reduction in the number of large axonal fibers is observed in the knockout mice treated with AAV-PHP.eB-GFP compared to those treated with AAV-PHP.eB-CTR. Scale bar=50 μm. (E) Representative images after H&E staining of gastrocnemius muscle sections (10 μm) from ChAT-Cre;Tardbp^f/f treated with AAV-PHP.eB-CTR and ChAT-Cre;Tardbp^f/f mice treated with AAV-PHP.eB-CTR or AAV-PHP.eB-GFP at 5 months of age. Arrows represent the muscle fibers undergoing degeneration in ChAT-Cre;Tardbp^f/f mice treated with AAV-PHP.eB-GFP, which is rescued in ChAT-Cre;Tardbp^f/f mice treated with AAV-PHP.eB-CTR. Scale bar=50 μm.