Novel AAV-Capsid-Mediated Delivery of an RNAi Targeting Atxn2 Extends Survival and Improves Strength

Defne A. Amado1,2, Alejandro Mas Monteys2, Katherine

Whiteman2, Alicia R. Smith2, Guillem Chillon Bosch2,

Aleksandar Izda2, Beverly L. Davidson2

1Neurology, University of Pennsylvania, Philadelphia, PA,2Children’s Hospital

of Philadelphia, Philadelphia, PA

Amyotrophic lateral sclerosis (ALS) is a fatal disease characterized by

death of motor neurons. A key pathologic feature is the cytoplasmic

mislocalization of a nuclear transcription and splice regulator, Tar-

DNA binding protein of 43kDa (TDP-43). TDP-43 aggregates in

cytoplasmic stress granules (SGs) and leads to toxicity through both

cytoplasmic gain- and nuclear loss-of-function. Inhibiting formation

of SGs is therefore a promising strategy, and downregulating the SGassociated

protein Ataxin-2 (Atxn2) using antisense oligonucleotides

(ASOs) prolongs survival by 35% in a mouse model of sporadic ALS

(Becker et al. Nature 2017), a strategy that is now in human clinical

trials. However, this strategy requires frequent CNS readministration

of the Atxn2-targeting ASO and may not effectively reach the brain

after intrathecal injections, limiting efficacy and safety. An alternative

approach is to provide lasting knockdown throughout the brain and

spinal cord after one treatment using AAV-mediated RNAi delivery. If

successful, this strategy could be used to treat the vast majority of ALS.

We designed miRNAs targeting Atxn2 and tested their efficacy in N2A

cells, packaging the most effective candidate into a novel AAV9 capsid

variant, AAV1999, engineered in our lab for superior CNS targeting

in both mice (Figure, A) and nonhuman primates. A dosing study

demonstrated 55% knockdown of Atxn2 in the frontal cortex and

25% knockdown throughout the brainstem and cervical and lumbar

spinal cord after a one-time intracerebroventricular injection, with

GFP-tagging demonstrating selective localization within the spinal

cord to anterior horn cells. We then conducted an efficacy study in

the same ALS mouse model used in the prior ASO study, in which

wildtype human TDP-43 is overexpressed in neurons starting at P7

and mice exhibit a rapid decline in strength, succumbing around

P22. After treatment, mean survival was increased by 54% and

median survival by 45% (p<0.002; Figure, B). Mice showed marked

improvement across several strength-related measures, including

rotarod (2X duration, p<0.02); gait (30% improvement, p<0.001);

kyphosis (66% improvement, p<0.001); tremor (34% improvement,

p<0.005); and foot angling (48% improvement, p<0.01), with a trend

towards improvement in abdominal droop and limping. Interestingly,

mice showed an increase in vertical activity above that seen in wildtype

mice, perhaps suggesting an unmasking of an FTD phenotype in the

setting of improved strength. Histologically, treated mice showed

normalization of the astrogliosis seen in mutant mice to wildtype

levels (p<0.05). AAV-mediated RNAi targeting Atxn2 is therefore a

promising strategy for treatment of the 97% of ALS characterized by

TDP-43 pathology.