AAV9/MFSD8 gene therapy is effective in preclinical models of neuronal ceroid lipofuscinosis type 7 disease

Neuronal ceroid lipofuscinosis type 7 (CLN7) disease is a lysosomal storage disease caused by mutations in the facilitator superfamily domain containing 8 (MFSD8) gene, which encodes a membrane-bound lysosomal protein, MFSD8. To test the effectiveness and safety of adeno-associated viral (AAV) gene therapy, an in vitro study demonstrated that AAV2/MFSD8 dose dependently rescued lysosomal function in fibroblasts from a CLN7 patient. An in vivo efficacy study using intrathecal administration of AAV9/MFSD8 to Mfsd8– /– mice at P7–P10 or P120 with high or low dose led to clear age- and dose-dependent effects. A high dose of AAV9/MFSD8 at P7–P10 resulted in widespread MFSD8 mRNA expression, tendency of amelioration of subunit c of mitochondrial ATP synthase accumulation and glial fibrillary acidic protein immunoreactivity, normalization of impaired behaviors, doubled median life span, and extended normal body weight gain. In vivo safety studies in rodents concluded that intrathecal administration of AAV9/MFSD8 was safe and well tolerated. In summary, these results demonstrated that the AAV9/MFSD8 vector is both effective and safe in preclinical models.


INTRODUCTION
The objective of this study was to characterize the toxicity and gene expression of scAAV9/Jet-hCLN7opt-SV40pA (AAV9/CLN7) following a single intrathecal (IT) injection in wild type (WT) CH57BL/6J mice. The AAV9/CLN7 is being developed for treatment of the neurodegenerative Batten disease (CLN7 subtype).

Study Design
WT CH57BL/6J mice from Jackson Laboratories were assigned to the study as indicated in Table 1. The non-GLP studies presented in Table 1 were designed to identify any long-term safety issues of the experimental therapy. The mice were randomized to different groups and injected IT with 5 µL of vehicle or different doses of AAV9/CLN7 vectors. Two lots of AAV9/CLN7 vectors were made by UNC-VC (University of North Carolina -Vector Core) or Vigene (Vigene Biosciences, Inc.). Both vectors were titered in parallel at UNC. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and silver stain analysis showed no visible contaminating protein in the UNC-VC lot, but the Vigene lot had host cell protein contamination (not shown). The equivalence of the two AAV9/CLN7 vector lots made at UNC-VC and Vigene was assessed by injecting the same doses of each vector in WT mice and measuring the biodistribution of the vectors to the heart, liver, and brain after 1 week. As shown in Figure 1, both vectors showed equivalent in vivo biodistribution patterns, indicating similar biopotency, and therefore were employed in this toxicity study.

Figure 1. In vivo equivalence of UNC and Vigene preclinical lots of AAV9/CLN7 vectors.
Mice (n=3) in each group were injected with the vector via tail vein in a 200 µL volume. The dose administered was 2×10 11 vg/mouse and tissues were collected a week later for biodistribution analysis.
Mice were monitored for changes in body weight, clinical signs, adverse events, and mortality following the treatment. All mice were weighed twice weekly for the first 2 months and then biweekly thereafter. Any clinical signs or adverse events including neurological symptoms were investigated, evaluated, and recorded. Appropriate supportive or therapeutic interventions were offered per Institutional Animal Care and Use Committee (IACUC) and veterinary guidance. Blood and tissue samples were collected from mice that are euthanized for humane reasons. Where possible, a detailed necropsy was performed to investigate or identify the reason for the ailment by a trained technician or veterinary staff. Terminal tissue samples at 12 months following the treatment were collected for histopathological assessment. The final histopathological evaluation on collected tissue samples was performed and reported by Dr. Mary Wight-Carter, DVM, DACVP, Veterinary Pathologist at Animal Research Center, University of Texas Southwestern Medical Center.

Body Weight
IT AAV9/CLN7 doses up to 9.50×10 11 vg/mouse were administered in WT C57BL/6J mice to assess the safety of IT dosing and hCLN7 overexpression (Table 1). The vector manufactured by UNC-VC (dosed 21 December 2017) and Vigene (dosed 19 or 23 October 2017) were tested in separate cohorts. Body weight difference was monitored to assess the overall health of the animals. There was no significant difference in body weight between groups within male or female mice at any point of assessment ( Figure  2 and Table 3), demonstrating that doses up to 9.50×10 11 vg/mouse are well tolerated in the WT C57BL/6J mice up to 12 months following the treatment.

Figure 2. No significant difference of body weight between groups within male or female mice.
WT CH57BL/6J mice (n=5 per group/sex) were dosed with vehicle or AAV9/CLN7 via IT injection. Left panel: UNC manufactured AAV9/CLN7 was dosed at 9.50×10 11 vg/mouse at 8 weeks of age. Right panel: Vigene manufactured AAV9/CLN7 was dosed at 4.47×10 10 , 1.48×10 11 , or 4.47×10 11 vg/mouse at 6 weeks of age. Mice were weighed twice weekly for the first two months following the treatment and then biweekly thereafter.

Clinical Signs or Adverse Events
No outward signs of toxicity were noted over the duration of the study.

Survival Rate
There were no obvious clinical signs of morbidity in the adult WT mice dosed with AAV9/CLN7 at doses up to 9.50×10 11 vg/mouse (Table 1). There were four unexpected death in this study: one found death in the control group injected with vehicle with no obvious reason, one found death in the treated group injected with 4.47×10 11 vg/mouse with no obvious reason, one found death in the treated group injected with 9.50×10 11 vg/mouse which was most likely caused by overgrooming-induced severe back injury, and one euthanized death for animal welfare in the treated group injected with 9.50×10 11 vg/mouse because of overgrooming-induced severe back and leg injury. There was no significant difference of survival rates between groups within male or female mice (Figure 3), further demonstrating that doses up to 9.50×10 11 vg/mouse are well tolerated in WT C57BL/6J mice up to 12 months following the treatment.

Histopathology
At the end of the experiment, 46 survival mice were perfused with phosphate-buffered saline containing 1 U/mL heparin and tissues were fixed in 10% formalin for 3 days. The tissues were then transferred to 70% ethanol and sent out for histopathology evaluation by Dr. Mary Wight-Carter, DVM, DACVP. Dr. Wight-Carter concluded that the tumors, increased number of inflammatory cell infiltrates, and degenerative lesions within multiple organs that were seen are considered to be common background lesions in aged mice. None of the microscopic findings are suggestive of adverse effects related to vector administration in these mice (see below the detailed report from Dr. Wight-Carter).
The seminiferous tubules of animal ID WT27 and 109 had multiple variably sized vacuoles that replaced various levels of the seminiferous epithelium in a few tubules. There was no evidence of accompanying germ cell degeneration. Because very few tubules were affected and there was no accompanying degeneration, it suggests that this was an incidental finding. Animal ID 101 had multiple variably sized vacuoles that replaced various levels of the seminiferous epithelium in multiple tubules. Sertoli cells were present; however, other levels of germ cells were not. No other lesions were present in the mouse testicles.
All of the ovaries that were present were normal and the structures within the ovaries were consistent with various points in the estrus cycle.
The hearts were normal. There was no evidence of heart failure in the other organs.
The kidneys of 107 showed multifocal mild to moderate thickening of the glomerular tufts, multifocal tubular regeneration, mild multifocal interstitial fibrosis, and mild to moderate multifocal interstitial and perivascular infiltrates with mononuclear cells (glomerulonephropathy). The multiple glomerular tufts were thickened with eosinophilic proteinaceous material in kidneys of animal ID 105 (mild glomerulopathy).
The tubules of kidneys of WT8 had a few small areas of mineralization. The renal pelvis had mild to moderate dilation of a few tubules of kidneys of WT12, WT14, WT15, and WT23.
The above described lesions are not uncommon in adult or aged mice and typically are more frequent in male mice.
The livers of the following mice had mild to moderate peribiliary infiltrates with mononuclear cells with no corresponding fibrosis or hepatocellular necrosis: 107, WT1, WT5, WT15, WT17, and WT26. Minimal to moderate peribiliary and perivascular infiltrates are a common finding in mice and increase in incidence as the mice age.
The livers of the following mice contained multifocal infiltrates with small numbers of mixed inflammatory cell infiltrates with hepatocellular necrosis (micro-abscess): 102,103,104,105,106,111,112,113,114, WT1, WT4, WT12, WT14, WT15, WT17, and WT25. Areas with 1 to 2 cell hepatocyte necrosis accompanied by inflammatory cells can occur spontaneously in the mouse liver with increased incidence as the mice age.
Animal numbers 101 and 102 had multifocal areas of extramedullary hematopoiesis present in the liver parenchyma. This is less common in rodents as they age and typically occurs in response to increased hematopoietic demand.
Animal 105 had a liver nodule grossly evident that microscopically was morphologically consistent with a hepatocellular adenoma which expanded the parenchyma and compressed the adjacent normal tissue. Animal 106 had a hepatocellular adenoma which expanded the parenchyma and compressed the adjacent normal tissue. This tumor was not grossly evident. Adenomas are common findings in adult B6 mice.
Mouse 102 had a focal area of mineralization surrounded by a small amount of fibrosis.
The spleen of all mice had variable amounts of extramedullary hematopoiesis and hemosiderin within the macrophages of the red pulp. This is considered normal in older mice. Mouse WT2 had enlarged spleen grossly and the white pulp was expanded with neoplastic small lymphocytes (lymphoma).

CONCLUSIONS
IT AAV9/CLN7 doses up to 9.50×10 11 vg/mouse are safe and well tolerated in WT mice. There were no toxicities observed in either the in-life portion of the study or after microscopic examination of major tissues. If scaled to humans by cerebrospinal fluid (CSF) volume (assuming 35 uL CSF volume in mice and 140 mL CSF volume in humans), the mouse dose of 9.5×10 11 vg in 5 uL would translate to a human dose of 3.8×10 15 vg in a volume of 20 mL. This highest dose injected in the mice is a 3.8-fold higher titer than the highest dose proposed in humans, in twice the volume proposed in humans. Thus, the maximum tolerated dose in mice up to one year post-injection provides a further safety margin above what is proposed in humans.