Alzheimer's disease (AD) is characterized by both amyloid and Tau pathologies. The amyloid component and altered cholesterol metabolism are closely linked, but the relationship between Tau pathology and cholesterol is currently unclear. Brain cholesterol is synthesized in situ and cannot cross the blood–brain barrier: to be exported from the central nervous system into the blood circuit, excess cholesterol must be converted to 24S-hydroxycholesterol by the cholesterol 24-hydroxylase encoded by the CYP46A1 gene. In AD patients, the concentration of 24S-hydroxycholesterol in the plasma and the cerebrospinal fluid are lower than in healthy controls. The THY-Tau22 mouse is a model of AD-like Tau pathology without amyloid pathology. We used this model to investigate the potential association between Tau pathology and CYP46A1 modulation. The amounts of CYP46A1 and 24S-hydroxycholesterol in the hippocampus were lower in THY-Tau22 than control mice. We used an adeno-associated virus (AAV) gene transfer strategy to increase CYP46A1 expression in order to investigate the consequences on THY-Tau22 mouse phenotype. Injection of the AAV-CYP46A1 vector into the hippocampus of THY-Tau22 mice led to CYP46A1 and 24S-hydroxycholesterol content normalization. The cognitive deficits, impaired long-term depression and spine defects that characterize the THY-Tau22 model were completely rescued, whereas Tau hyperphosphorylation and associated gliosis were unaffected. These results argue for a causal link between CYP46A1 protein content and memory impairments that result from Tau pathology. Therefore, CYP46A1 may be a relevant therapeutic target for Tauopathies and especially for AD.

No treatment is available for early-onset forms of metachromatic leukodystrophy (MLD), a lysosomal storage disease caused by autosomal recessive defect in arylsulfatase A ( ARSA ) gene causing severe demyelination in central and peripheral nervous systems. We have developed a gene therapy approach, based on intracerebral administration of AAVrh.10-hARSA vector, coding for human ARSA enzyme. We have previously demonstrated potency of this approach in MLD mice lacking ARSA expression. We describe herein the preclinical efficacy, safety, and biodistribution profile of intracerebral administration of AAVrh.10-hARSA to nonhuman primates (NHPs). NHPs received either the dose planned for patients adjusted to the brain volume ratio between child and NHP (1×dose, 1.1×10 ¹¹ vg/hemisphere, unilateral or bilateral injection) or 5-fold this dose (5×dose, 5.5×10 ¹¹ vg/hemisphere, bilateral injection). NHPs were subjected to clinical, biological, and brain imaging observations and were euthanized 7 or 90 days after injection. There was no toxicity based on clinical and biological parameters, nor treatment-related histological findings in peripheral organs. A neuroinflammatory process correlating with brain MRI T2 hypersignals was observed in the brain 90 days after administration of the 5×dose, but was absent or minimal after administration of the 1×dose. Antibody response to AAVrh.10 and hARSA was detected, without correlation with brain lesions. After injection of the 1×dose, AAVrh.10-hARSA vector was detected in a large part of the injected hemisphere, while ARSA activity exceeded the normal endogenous activity level by 14–31%. Consistently with other reports, vector genome was detected in off-target organs such as liver, spleen, lymph nodes, or blood, but not in gonads. Importantly, AAVrh.10-hARSA vector was no longer detectable in urine at day 7. Our data demonstrate requisite safe and effective profile for intracerebral AAVrh.10-hARSA delivery in NHPs, supporting its clinical use in children affected with MLD.

Charcot-Marie-Tooth disease 1 A (CMT1A) results from a duplication of the PMP22 gene in Schwann cells and a deficit of myelination in peripheral nerves. Patients with CMT1A have reduced nerve conduction velocity, muscle wasting, hand and foot deformations and foot drop walking. Here, we evaluate the safety and efficacy of recombinant adeno-associated viral vector serotype 9 (AAV2/9) expressing GFP and shRNAs targeting Pmp22 mRNA in animal models of Charcot-Marie-Tooth disease 1 A. Intra-nerve delivery of AAV2/9 in the sciatic nerve allowed widespread transgene expression in resident myelinating Schwann cells in mice, rats and non-human primates. A bilateral treatment restore expression levels of PMP22 comparable to wild-type conditions, resulting in increased myelination and prevention of motor and sensory impairments over a twelve-months period in a rat model of CMT1A. We observed limited off-target transduction and immune response using the intra-nerve delivery route. A combination of previously characterized human skin biomarkers is able to discriminate between treated and untreated animals, indicating their potential use as part of outcome measures.