Etude de la deregulation des microRNA dans les mitochondries des patients atteints de FXTAS. – MITO-FXTAS

Thanks to this ANR funding, we developed novel cell and animal model of FXTAS.

1 - Novel cell model of FXTAS:
First, we developed neuron from iPS cells generated from fibroblasts of three control and three FXTAS patients. These neuronal cultures present cell “markers” typical of FXTAS such as the disruption of the nuclear lamina and the presence of enlarged swelling dendrites. Importantly, FXTAS neurons show accumulation of the FMRpolyGlycine protein translated from the expanded CGG repeats, but no RNA aggregates of expanded CGG repeats or microRNA alterations. These results suggest that the accumulation of RNA aggregates is a late event in FXTAS patient or a bias of autopsied samples. In contrast, translation of the CGG repeats into a polyGlycine protein appears as a robust and early marker of the disease.

2 - Novel mouse model of FXTAS:
In parallel, we also developed two novel transgenic mouse models. The first one contains the expanded CGG repeats embedded in the full 5’UTR of FMR1, allowing translation of the polyGlycine protein. The second mouse model presents a deletion in the 5’UTR of FMR1 impairing translation of the CGG repeats into the polyGlycine protein such that mouse only expresses the CGG RNA. We controlled that both mouse models express identical levels of CGG RNA and that only the mouse with full 5’UTR sequence express, as expected, the polyGlycine protein. Importantly, only the mice expressing the polyGlycine protein show locomotors alterations. In contrast, the mice with a deletion in the 5’UTR of FMR1 that express the CGG RNA but not the FMRpolyGlycine protein are normal. This is an important result confirming our observation in patient iPS cells and suggesting that the prime determinant of neuronal toxicity in FXTAS is the FMRpolyGlycine protein and not the accumulation of CGG RNA.

Overall, our novel cell and animal models demonstrate that the key pathogenic event in FXTAS is the expression of a toxic polyglycine-containing protein. An article presenting these results is in preparation (Sellier et al. Translation of expanded CGG repeats is pathogenic in Fragile X Tremor Ataxia Syndrome. Neuron. In review).

In parallel, experiments based on cell transfection in Dr. Singh laboratory demonstrated that FMRpolyGlycine was inducing mitochondrial dysfunctions with increased generation of Radical Oxygen Species and expression of inflammatory cytokines ultimately resulting in cell death. Both Dr Singh and our groups are now investigating by which mechanisms this polyGlycine protein is toxic to mitochondria and promote neuronal cell dysfunction and death. We postulate that FMRpolyGlycine toxicity may originate from faulty association with other proteins. We thus propose to identify by proteomic analysis candidate proteins interacting with FMRpolyGlycine. Then, we will test whether functions of these candidate interacting proteins are altered in our iPS cell and mouse models of FXTAS, and whether expression of any of these candidates would correct FXTAS symptoms. Finally, having identified the key pathogenic event in FXTAS, we hope to screen and identify small pharmacological compounds and / or antisense oligonucleotide impairing synthesis of this toxic FMRpolyGlycine protein.

If successful, this proposal will clarify the molecular mechanisms underlying FXTAS as well as open route for identification of biomarkers and treatment for this devastating neurodegenerative disease.

Through developing novel cell and animal models of FXTAS, we identified that the key pathogenic event in that disease is the expression of a toxic polyGlycine-containing protein. We will continue this work and we will notably investigate by which mechanisms this polyGlycine protein is toxic and promote neuron dysfunction. FMRpolyGlycine protein is composed of a short N-terminus, a central polyglycine stretch which length corresponds to the number of expanded CGG repeats and a C-terminus of 42 amino acids with no predicted structure or homology. To identify how this polyglycin-containing protein promotes neuronal cell death, we constructed various mutants and found that expression of the polyglycine stretch in isolation was driving protein aggregation but was not toxic, while expression of the C-terminal part induced neuronal cell death. We propose to investigate how this C-terminal part promotes cell toxicity. Work in Dr Singh group indicates that this toxicity may occur through dysfunction of the mitochondria. We postulate that it may be through faulty association with specific proteins, and thus propose to express double-tagged FLAG-HA-FMRpolyGlycine in neuronal cells to perform tandem tag protein purification and identify by nano-LC-MS/MS interacting candidates. Then, we will validate whether these candidates interact with the polyglycine protein and if so, whether the functions of these candidates are altered in the iPS cell and mouse models of FXTAS that we developed. The key experiment will be then to test whether expression of any of these candidate proteins would correct FXTAS symptoms in our iPS or mouse models.

Finally, having identified the key pathogenic event in FXTAS, we hope to screen and identify small pharmacological compounds and / or antisense oligonucleotide impairing synthesis of this toxic FMRpolyGlycine protein.

An article presenting our results is in preparation (Sellier et al. Translation of expanded CGG repeats is pathogenic in Fragile X Tremor Ataxia Syndrome. Neuron. In review).

Le Syndrome d’Ataxie et de Tremblement Associé à l’X Fragile (FXTAS) est une maladie neurodégénérative relativement fréquente (1/4000), caractérisée par des tremblements, une ataxie cérébelleuse et un déclin cognitif. Cette maladie génétique a pour origine la présence de répétitions anormalement longues des trinucléotides CGG dans la région 5’UTR non traduite du gène FMR1. Ces répétitions CGG sont transcrites mais non-traduites, et s’accumulent sous forme d’agrégats nucléaires toxiques qui séquestrent des protéines spécifiques.

Nous avons récemment montré que ces répétitions CGG séquestrent le complexe enzymatique DROSHA-DGCR8 (Sellier et al., Cell reports, 2013). Ce complexe est déterminant pour la synthèse de petits ARN essentiels à la survie et au fonctionnement cellulaire, les microARNs. Nous avons alors montré que la séquestration du complexe DROSHA-DGCR8 chez les patients atteints de FXTAS conduit à une diminution de l’expression des microARNS neuronaux, responsable in fine de la mort de ces neurones (Sellier et al., Cell reports, 2013). Nous souhaitons continuer ce travail et étudier précisément comment cette diminution de microARN conduit à la mort neuronale. Une piste prometteuse est l’altération connue des mitochondries dans des modèles cellulaire de FXTAS. Le Prof. Rajesh Singh ayant démontré récemment que les mitochondries contiennent des microARNS spécifiques (Sripada et al., 2012), nous souhaitons établir une collaboration avec son groupe de recherche. Nos objectifs sont de :

(1) déterminer si ces microARNS mitochondriaux sont altérés chez les patients FXTAS (modèle cellulaire, murins et tissus cérébraux) ;
(2) déterminer les conséquences cellulaires de la diminution d’expression de ces microARNS, notamment sur le fonctionnement mitochondrial et la survie neuronale ;
(3) déterminer si ré-exprimer certains de ces microARNS mitochondriaux pourrait augmenter la survie neuronale de modèle FXTAS.

En conclusion, notre projet a pour objectif de mieux comprendre les causes moléculaires et cellulaires à l’origine du syndrome FXTAS, afin d’établir des approches thérapeutiques.