Direct Intracranial Injection of AAVrh8 Encoding Monkey β-N-Acetylhexosaminidase Causes Neurotoxicity in the Primate Brain

Abstract

GM2 gangliosidoses, including Tay-Sachs disease and Sandhoff disease, are lysosomal storage disorders caused by deficiencies in β-N-acetylhexosaminidase (Hex). Patients are afflicted primarily with progressive central nervous system (CNS) dysfunction. Studies in mice, cats, and sheep have indicated safety and widespread distribution of Hex in the CNS after intracranial vector infusion of AAVrh8 vectors encoding species-specific Hex α- or β-subunits at a 1:1 ratio. Here, a safety study was conducted in cynomolgus macaques (cm), modeling previous animal studies, with bilateral infusion in the thalamus as well as in left lateral ventricle of AAVrh8 vectors encoding cm Hex α- and β-subunits. Three doses (3.2 × 10¹² vg [n = 3]; 3.2 × 10¹¹ vg [n = 2]; or 1.1 × 10¹¹ vg [n = 2]) were tested, with controls infused with vehicle (n = 1) or transgene empty AAVrh8 vector at the highest dose (n = 2). Most monkeys receiving AAVrh8-cmHexα/β developed dyskinesias, ataxia, and loss of dexterity, with higher dose animals eventually becoming apathetic. Time to onset of symptoms was dose dependent, with the highest-dose cohort producing symptoms within a month of infusion. One monkey in the lowest-dose cohort was behaviorally asymptomatic but had magnetic resonance imaging abnormalities in the thalami. Histopathology was similar in all monkeys injected with AAVrh8-cmHexα/β, showing severe white and gray matter necrosis along the injection track, reactive vasculature, and the presence of neurons with granular eosinophilic material. Lesions were minimal to absent in both control cohorts. Despite cellular loss, a dramatic increase in Hex activity was measured in the thalamus, and none of the animals presented with antibody titers against Hex. The high overexpression of Hex protein is likely to blame for this negative outcome, and this study demonstrates the variations in safety profiles of AAVrh8-Hexα/β intracranial injection among different species, despite encoding for self-proteins.

Keywords: AAV; Tay-Sachs disease; adeno-associated virus; gene therapy; hexosaminidase; intracranial delivery.

Figure 1.

Corresponding coronal sections of brain T2-weighted magnetic resonance imaging before surgery and before necropsy indicating signal alterations at the study endpoint in the treated animal. Hyper-intensities (arrows) detected in the thalami of an animal injected with 1.1 × 10¹¹ vg (1/30th dose) of AAVrh8 vector formulation encoding cmHexα or cmHexβ subunits. No hyper-intensities were noted at any time point in the thalami of animals treated with a transgene empty vector.

Figure 2.

Histological findings in the thalamus. (A) Neurons with granular eosinophilic inclusions were present in the thalamus of AAVrh8-Hex injected animals. Box indicates location of magnified region. 40 × . (B) Necrotic area in the thalamus with cell loss and vacuolation. 10 × . (C) Example of perivascular cuffing in the thalamus. 20 × . (D) Luxol fast blue staining revealed a large area (delineated by arrows) of necrosis, vascular proliferation, and white matter loss (note pallor compared to adjacent dark blue color). 10 × . (E) Neuropathological findings were identical in the lowest dosing cohort. 10 × . (F) Normal histology of the thalamus in phosphate-buffered saline (PBS)-injected animal; 10 × (1 × cohort, A–D; 1/30th cohort, E; PBS-injected, F). Tissue sections for histological assessment shown in (A), (B), (C), (E), and (F) were stained with hematoxylin and eosin.

Figure 3.

Immunological profile of perivascular cuffs in AAVrh8-cmHexα/β injected monkey thalamus. (A) CD3 immunostaining for T cells. (B) CD20 immunostaining for B cells. (C) CD68 immunostaining for macrophages/microglia. (A–C; 1/10th cohort animal). 20 × .

Figure 4.

Increased Hex expression and activity in the central nervous system. (A) Western blot analysis of Hex α- or β-subunit protein levels in thalamus lysates from the PBS-injected control and two animals from the 1/10th cohort. (B) Hex activity was measured in the thalamus and thoracic spinal cord on all cohorts. 4-Methylumbelliferyl N-acetyl-β-D-glucosaminide substrate is cleaved by HexA, HexB, and HexS isozymes. 4-Methylumbelliferyl 6-sulfo-2-acetamido-2-deoxy-β-D-glucopyranoside potassium salt substrate is preferentially cleaved by HexA and HexS isozymes. Note: Due to an emergency euthanasia and necropsy, the identity of the thalamus (left or right) in animal is 909589 is unknown.

Figure 5.

Hexα- or Hexβ-subunit localization in monkey thalamus. (A) Immunofluorescence staining for Hexα (green) and Hexβ (magenta) subunits and their co-localization (white). Box indicates location of magnified region. 40 × . (B) Immunostaining for Hexα-subunit (green) and NeuN (magenta). 40 × . (C) RNA in situ hybridization with probe for WPRE (red) expressed by AAV transduced cells. DAPI stain (cyan). 20 × .