Synergistic effects of central nervous system-directed gene therapy and bone marrow transplantation in the murine model of infantile neuronal ceroid lipofuscinosis

Abstract

Objective: Infantile neuronal ceroid lipofuscinosis (INCL) is an inherited childhood neurodegenerative disorder caused by the loss of palmitoyl protein thioesterase-1 (PPT1) activity. Affected children suffer from blindness, epilepsy, motor dysfunction, cognitive decline, and premature death. The Ppt1(-/-) mouse shares the histological and clinical features of INCL. Previous single-therapy approaches using small molecule drugs, gene therapy, or neuronal stem cells resulted in partial histological correction, with minimal improvements in motor function or lifespan. Here, we combined central nervous system (CNS)-directed adeno-associated virus (AAV)2/5-mediated gene therapy with bone marrow transplantation (BMT) in the INCL mouse.

Methods: At birth, Ppt1(-/-) and wild-type mice were given either intracranial injections of AAV2/5-PPT1 or bone marrow transplantation, separately as well as in combination. To assess function, we measured rotorod performance monthly as well as lifespan. At terminal time points, we evaluated the therapeutic effects on several INCL-specific parameters, such as cortical thickness, autofluorescent accumulation, and glial activation. Finally, we determined levels of PPT1 enzyme activity and bone marrow engraftment in treated mice.

Results: AAV2/5-mediated gene therapy alone resulted in significant histological correction, improved motor function, and increased lifespan. Interestingly, the addition of BMT further increased the lifespan of treated mice and led to dramatic, sustained improvements in motor function. These data are truly striking, given that BMT alone is ineffective, yet it synergizes with CNS-directed gene therapy to dramatically increase efficacy and lifespan.

Interpretation: AAV2/5-mediated gene therapy in combination with BMT provides an unprecedented increase in lifespan as well as dramatic improvement on functional and histological parameters.

Figure 1

(A) Lifespan of treated Ppt1^−/− mice. The median lifespan of the AAV2/5+BMT group (n=10) was significantly increased to more than double the median lifespan of the BMT only (n=5) and untreated Ppt1^−/− mice (n=10). AAV2/5 treatment also significantly increased the median lifespan (n=11). (B) Motor function using the constant speed rotarod paradigm. At 6 months, Ppt1^−/− mice (n=10) performed poorly on the rotarod. The BMT-only performed worse than untreated controls (n=5; ω = p<0.001). By 10 months, AAV2/5 performance declined (n=15) and was significantly decreased compared to AAV2/5+BMT (n=10) or WT mice (n=15; γ = p<0.01). The AAV2/5+BMT group was indistinguishable from WT until 13 months (ψ = p<0.001). Error bars reflect standard error of the mean (±SEM).

Figure 2

Brain pathology. (A) Images of cortical thinning in primary motor (M1) cortex at terminal time points. (B) There was a significant decrease in M1 thickness in both BMT (n=2) and untreated Ppt1^−/− mice (n=4), while AAV2/5 (n=4) or AAV2/5+BMT (n=3) displayed significantly less atrophy compared to these groups. (C) Images of autofluorescent accumulation in the S1BF cortex. (D) There was a significant increase in storage burden in both the untreated Ppt1^−/−and BMT brains. Conversely, treatment with either AAV2/5 or AAV2/5+BMT decreased storage material such that these treated groups were indistinguishable from WT (n=4). (E) Changes in CD68+ immunostaining at terminal time points. (F) There was a significant increase in CD68+ staining in the S1BF cortex of untreated Ppt1^−/− and BMT-treated mice. Following treatment with AAV2/5, either alone or in combination with BMT, there was a significant decrease in CD68 staining with levels comparable to WT brains. The morphology of individual CD68+ cells (insets) in the AAV2/5 or AAV2/5+BMT mice more closely resembled WT cells. (G) Alterations in GFAP immunoreactivity in the S1BF cortex. (H) There was a significant increase in GFAP staining in untreated Ppt1^−/− and BMT brains. There appeared to be less GFAP staining in the S1BF cortex in mice treated with either AAV2/5 or AAV2/5-BMT; however, this did not reach statistical significance. Error bars reflect standard error of the mean (±SEM). (* = p<0.05; ** = p<0.01; *** = p<0.001; white scale bar = 50μm; black scale bar = 20μm)

Figure 3

(A) Comparable levels of bone marrow engraftment were achieved in the BMT (n=5) and AAV2/5+BMT (n=7) groups. (B) PPT1 assays performed on the brains from treated and untreated mice demonstrated a 1.5-fold increase in PPT1 activity in the CNS of AAV2/5-treated (n=5) mice when compared to WT (n=4) mice. The AAV2/5+BMT (n=7) group had a 6-fold increase compared to WT. No PPT1 activity was detected in BMT (n=2) or Ppt1^−/− (n=4) brains. (C) Although AAV2/5-treated mice had an increase in liver PPT1 activity compared to Ppt1^−/− mice, it was only 25% of WT levels. Conversely, AAV2/5+BMT mice had liver activity 1.5-fold higher than WT mice. (D) qPCR analysis investigating the relative copy number of AAV-PPT1 genomes in AAV2/5− (n=4) and AAV2/5+BMT− (n=8) treated mice. Four out of five AAV2/5-only treated mice had undetectable levels of AAV-PPT1 genome within their livers (level of detection = 0.00005). In comparison, all eight of the AAV2/5-BMT-treated livers had detectable levels of viral genomes. There was a ~3.4-fold increase in the ratio of vector genomes/diploid genomes within the AAV2/5-BMT mice (range=0.001–0.04) compared to AAV2/5-only mice (range=0.0–0.019). Error bars reflect standard error of the mean (±SEM). (* = p<0.05; ** = p<0.01)