Modeling CLN3 and ATP13A2 deficiency in yeast and zebrafish and use of the ATP13A2 models for drug repurposing

Neuronal ceroid lipofuscinoses (NCL) are a heterogeneous group of inherited
recessive neurodegenerative disorders that appear during childhood and result in
premature death. Nowadays, mutations in 14 genes are known to cause NCL and
this project focused on CLN3 and ATP13A2 (CLN12), two genes linked to a juvenile
form of NCL (JNCL). Mutations in CLN12 are known to cause two additional rare
neurodegenerative disorders called Kufor-Rakeb syndrome and spastic paraplegia-
78. Since the number of people affected with a rare disease is relatively small and
the cost of the drug development process is high, the chance for a patient to get
therapeutic treatment is very low. Therefore, the aim of this PhD project was to
develop a new drug screening pipeline for the identification of drug candidates that
could be used for the treatment of some of these rare diseases.
In this work, we successfully developed a phenotypic high-throughput assay based
on a decreased zinc resistance phenotype in an ATP13A2-deficient yeast model and
we screened more than 2500 compounds, resulting in the identification of 11 hits.
Subsequently, we created a stable ATP13A2 knockout line in zebrafish and
developed a validation platform based on decreased manganese resistance in this
line. Using this approach, N-acetylcysteine and furaltadone emerged as promising
compounds for follow-up studies. A similar strategy could not be implemented for
CLN3, due to failure, despite extensive efforts, to find a suitable phenotype in yeast
for a drug screening. Nevertheless, we successfully created two stable cln3 mutant
lines in zebrafish. No overt phenotype was initially observed, but behavioral tests
suggested that cln3 mutants display subtle neurological dysfunction, making them
more susceptible to treatment with picrotoxin, a pro-convulsive drug. Further
investigation is needed, but our preliminary data indicate that cln3 mutant larvae may
recapitulate certain aspects of JNCL pathology.
On the whole, this work provides a time- and cost-efficient pipeline for the discovery
of drugs against ATP13A2 deficiencies, which can be applied for the screening of
larger compound libraries in the future. In addition, we generated a new CLN3
disease model in zebrafish that will be instrumental for the development of drug
screens and also may help to elucidate the molecular disease mechanism of JNCL.