Project Portugal 2030

Summary

In MJDPPITT we will characterize the full ataxin-3 protein network in tissues relevant to MJ disease, to identify in-silico and in-vitro the proteins that cause this disease. To get a complete understanding of the role of these proteins we will perform in-vivo analyses, using Drosophila ataxin-3 mutant lines, with the final goal of using them as novel therapeutic targets. For that we will: 1- Develop an automated pipeline, using in-silico methodologies, that take in consideration the role of different 3D protein structure predictions as well as different docking algorithms. This tool is crucial along the project for the identification of interacting regions, as well as determining the interaction strength of ataxin-3 interactors with the two ataxin-3 forms. 2- Infer the complete ataxin-3 network using machine learning approaches: For the ataxin-3 interactors described in main databases (from humans, mouse and zebrafish) as well as in mutant cell lines and mutant model species (available in EvoPPI3 [30]) we will identify the proteins that are true ataxin-3 interactors. These proteins will be used to identify the interacting regions with ataxin-3 (as performed for ataxin-3, [32]). Using these interacting regions, machine learning algorithms, will be used to derive rules to be used in the identification of novel ataxin-3 interactors. For the new interactors, in-silico analyses will be performed using the automated pipeline to confirm that these proteins interact with ataxin-3 using the ataxin-3 interacting regions. 3- Predicting the ataxin-3 interactors that matter to MJ disease: Using the automated pipeline to be obtained in 1, we will infer from the ataxin-3 network, obtained in 2, those proteins that show a strong interaction strength with the expanded ataxin-3 form when compared with the wt ataxin-3 form. 4- In-vitro validation of the ataxin-3 interactors that show a 20 % increase in the interaction strength with exp ataxin-3 compared with the wt form, using commercial SCA3 cell lines and protein interaction coupling (PICO) technology combined with quantitative PCR. 5- For the proteins validated in 4, we will address their role as SCA3 suppressers/enhancers using Drosophila SCA3 mutants and RNAi stocks for those proteins. Proteomic/ transcriptomic analyses will be performed for the most severe phenotypes. 6- Address the effect of the amino acid polymorphism at the proteins identified in 4, in patients with SCA3, from public exome data. The MJDPPITT team includes experts in bioinformatics and machine learning, molecular biologists, Drosophila experts and clinicians, that bring together all the needed expertise to successfully tackle the ambitious objective of identifying novel SCA3 therapeutic targets.