Disease characterization and pre-clinical testing in canine models of the neuronal ceroid lipofuscinoses

Abstract

The neuronal ceroid lipofuscinoses (NCLs) are inherited lysosomal storage disorders characterized by intracellular accumulation of autofluorescent storage material and by progressive degeneration of neurons in the brain and retina. Clinically, the NCLs are characterized by progressive cognitive and motor impairment, seizures and progressive loss of vision culminating in premature death. Variants in at least 14 genes are known to cause forms of NCL. The CLN2 form of NCL is caused by variants in the TPP1 gene, which encodes the soluble lysosomal hydrolase tripeptidyl peptidase 1 (TPP1), while the CLN5 form is caused by variants in CLN5, which encodes a lysosomal protein the function of which has not been definitively determined. CLN2 disease has a late-infantile onset and typically culminates in death in the early teenage years. A canine disease model of affected Dachshunds with CLN2 disease recapitulates the features of the human disease and has proven to be an excellent model for preclinical therapy testing. CLN2-affected Dachshunds are homozygous for a single base pair deletion in the canine ortholog of TPP1 which results in a premature termination codon and lack of functional TPP1 protein. Affected dogs present with progressive motor, cognitive and behavioral changes, myoclonic seizures and brain atrophy which progress to end-stage disease requiring humane euthanasia around 10 months of age. Affected dogs also exhibit vison loss associated with progressive retinal degeneration, electroretinogram (ERG) deficits and degeneration of visual centers in the brain. Previous work in the CLN2 Dachshund model demonstrated that delivering functional TPP1 protein to the central nervous system via enzyme replacement therapy (ERT) or gene therapy extends lifespan and delays neurological disease progression. However, these treatments did not preserve retinal structure or function. We therefore conducted studies to assess three different methods of delivering functional TPP1 to the retina: (1) intravitreal enzyme replacement therapy, (2) ex vivo cell-mediated gene therapy and (3) in vivo intravitreal gene therapy. Additionally, we tested an ex vivo cell-mediated gene therapy approach to deliver TPP1 to the central nervous system as an alternative to the ERT and gene therapy methods previously assessed in the canine model. Intravitreal ERT was tested both before and after onset of declines in ERG amplitudes. ERT was able to fully preserve ERG b-wave amplitudes when administered pre-symptomatically. Post-symptomatic treatment halted progression of disease-related declines in amplitudes through end stage neurological disease at approximately 10 months of age. The major side effect of treatment was transient intraocular inflammation characterized by anterior and posterior uveitis and by the production of anti-TPP1 antibodies. Treatment also preserved retinal structure, prevented disease-related loss cells in the inner and outer nuclear layers and reduced the accumulation of intracellular storage material in retinal ganglion cells. These results were incorporated into an application that is being submitted to the Food and Drug Administration and European regulatory agencies for approval to conduct a clinical trial of this treatment to prevent progressive vision loss in children with CLN2 disease. A single intravitreal administration of TPP1-overexpressing autologous mesenchymal stem cells (TPP1-MSCs) was able to partially preserve retinal function. ERG rod b-wave amplitudes remained larger in treated eyes than in control eyes through 10 months of age, although amplitudes continued to decline in treated eyes. TPP1-MSCs were safe and well-tolerated and did not cause inflammation. While these results are encouraging, further studies are necessary to optimize this therapy. One-time intravitreal administration of a gene therapy vector (AAV2.CAG.TPP1) prior to the onset of retinal function impairment partially preserved retinal function through end-stage neurological disease at approximately 10 months of age. The vector transduced primarily cells of the inner retina and resulted in transgene expression for at least 6 months after treatment. There was also transgene expression in the retinal pigment epithelium. As with intravitreal ERT, a side effect was intraocular inflammation characterized by anterior and posterior uveitis. Inflammation was only partially alleviated by treatment with various anti-inflammatory medications. These results support further investigation to develop this approach to preserve vision in CLN2 disease. To date, infusion of TPP1-expressing mesenchymal stem cells into the cerebrospinal fluid has not been effective in treating central nervous system pathology in CLN2 disease. So far, with two minor exceptions, this method has not extended lifespan, slowed neurological disease progression or impeded brain atrophy. These exceptions, a slightly increased lifespan of one dog and temporarily improved performance on a cognitive test by another, may be indicative of some therapeutic benefit in those dogs. Work is ongoing to determine why this approach has not been successful and to modify the approach to enhance the potential for efficacy. This current work suggests that reducing the in vitro expansion of MSCs and/or removing immunosuppressive medications from the protocol may have improved efficacy, since these modifications were implemented for the dogs that demonstrated slight increases in lifespan and cognitive function. However, even in these two dogs any improvements were minor. Lack of success could be due to failure of the MSCs to survive long-term in the CNS or to secrete therapeutic amounts of hTPP1. Future studies could include alternative delivery methods and repeated injections over the disease course. Going forward, efforts should focus on determining if MSCs survive in the CNS, and if so for how long, as well as whether therapeutic concentrations of hTPP1 are obtainable with this method. The CLN5 form of NCL is characterized by progressive neurological decline and vision loss. As with CLN2 disease, a CLN5 canine model disease would be a valuable resource for testing potential therapeutic interventions. We characterized visual system pathology in five Golden Retrievers homozygous for a two base pair deletion and frameshift in the canine ortholog of CLN5. The dogs exhibited visual impairment by 22 months of age, progressive decline in ERG amplitudes primarily affecting cones and altered visual evoked potential recordings indicating impaired visual signal processing in the brain. While there were no gross retinal abnormalities by 23 months of age, there was extensive storage material accumulation in the retina and occipital cortex by 20 months of age. The visual system pathology in these dogs was similar to that seen in human patients. The baseline data obtained from the untreated affected dogs with CLN5 disease will serve as a foundation for designing therapeutic intervention studies.