Abstract:
Fragile X syndrome (FXS) is the leading cause of inherited intellectual disability (ID) and autism. This neurodevelopmental disorder is caused by silencing of the FMR1 gene and lack of its product, Fragile X Mental Retardation Protein (FMRP). FMRP, an RNA-binding protein, is highly expressed in the brain and plays an important role in the transport and translation of many different mRNA targets. Lack of FMRP causes disruption in synapse morphology and function, as well as disruption in synaptic plasticity. These molecular and synaptic abnormalities cause FXS symptoms like ID, autism, hyperactivity, epilepsy and anxiety. Therefore, understanding of the molecular pathogenesis of FXS is also important for other neurodevelopmental and psychiatric disorders. In last two decades, extensive research in basic neurobiology and pathophysiology led to significant advances in FXS field. However, although caused by only one gene, the molecular mechanisms of FXS are not yet well understood. Since FMRP is characterized as a translational regulator with multiple mRNA targets, it is crucial to understand the changes of the proteome in the absence of FMRP. As multiple FMRP-regulated proteins are involved in conserved neuronal signal transduction pathways, it is also necessary to study FMRP-dependent changes in the phosphoproteome dynamics. In my thesis I applied mass spectrometry-based proteomics to address several aspects of the disorder in different FXS models, such as Fmr1-KO cell lines and mice. To analyze the influence of FMRP on major signal transduction networks, I first performed a global analysis of the proteome and phosphoproteome of Fmr1- and Fmr1+ mouse embryonic fibroblast (MEF) cell lines using stable isotope labeling by amino acids in cell culture (SILAC) and high resolution mass spectrometry. In this study I detected 6,703 proteins and 9,181 phosphorylation events, and mapped 266 significantly changing proteins and 142 phosphorylation sites onto major signal transduction pathways. My results confirmed a downregulation of the MEK/ERK pathway in absence of FMRP, with decreased phosphorylation on ERK1/2. Several proteins involved in mTOR, Wnt, p53 and MAPK signaling cascades were also significantly regulated; they were previously shown to be associated with autism, but not with FXS. Additionally, I detected a significant increase of p53 and proteins linked to p53 signaling, as well as a decreased level of the major prion protein (Prp) in Fmr1- cells. Decreased p53 signaling is the likely cause for previously observed dysregulation of cell cycle control in FXS, whereas reduced levels of Prp could contribute to the cognitive deficits observed in the FXS patients. These proteins and signal transduction pathways may represent novel targets for treatment of FXS symptoms. In the second part of the thesis, I used the same MEF cell line in order to analyze potential substrates of glycogen synthase kinase 3β (GSK-3β), an emerging therapeutic target for treatment of FXS. I used two inhibitors of GSK3-β kinase, lithium and TDZD-8, to identify potential substrates of the kinase in Fmr1- and Fmr1+ cells. Lithium treatment was poorly reproducible on the proteome and phosphoproteome level, reflecting its reported low specificity for GSK3- β, whereas TDZD-8 treatment showed good reproducibility between biological replicates. Of a total of 7,285 detected phosphorylation events, 91 were significantly decreased upon TDZD-8 treatment in Fmr1+ and 146 in Fmr1- MEF cells – these phosphorylation events were likely targets of GSK-3β. Overlap of these potential substrates was poor in Fmr1+ and Fmr1- cells, pointing to different substrates of the GSK-3β kinase in Fmr1+ and Fmr1- MEF cells. Importantly, downregulation of multiple phosphorylation events on MAP1B, a well characterized GSK-3β substrate whose mRNA is a known FMRP target, was observed only in Fmr1+ MEF cell line. In healthy neurons, MAP1B is coordinating microtubule dynamics. Since abnormal axon branching is one of the leading symptoms of FXS, I postulate that the lack of regulation of MAP1B by GSK-3β is the likely cause for aberrant morphology of neurons in FXS. Functional enrichment analysis revealed an implication of the potential GSK-3β substrates in different processes in MEF cell lines. For example, in Fmr1- MEFs potential substrates seem to be more involved in cell cycle, DNA replication and RNA processing. Since downregulation of DNA damage/repair pathway in FXS patients was recently reported, this data will shed new light on differential activity of GSK-3β in Fmr1+ and Fmr1- MEF cells and deepen our understanding of molecular mechanisms regulated by this kinase. In the third part of my thesis, I postulated that increased protein synthesis in FXS is accompanied by increased protein degradation in order to maintain cellular homeostasis. Since increased protein synthesis and degradation lead to increased protein turnover, I used stable isotope labeling (“dynamic SILAC” approach) to measure protein turnover rate in mouse primary cortical neurons derived from the WT and Fmr1-KO model. Analysis showed that most of the proteins have a similar half-life in WT and Fmr1-KO, although calculated median protein half-life was higher in Fmr1-KO neurons than in WT. Functional enrichment analysis showed that proteins with the lowest turnover rates are involved in oxidative phosphorylation, Alzheimer’s, Parkinson’s and Huntington’s diseases, whereas proteins with the highest turnover rates are involved in phagosome, Wnt signaling pathway and gap junction, for both, WT and Fmr1-KO neurons. The experimental design employed did not allow for detection of low SILAC incorporation rates and proteins with very short half-lifes and further optimization of experimental conditions is needed to fully address differences in protein turnover between Fmr1-KO and WT cells. In the fourth and final part of the thesis, I investigated molecular mechanisms of the genetic “rescue” of the FXS, recently reported to occur in Fmr1-KO mice after reduction of activity of the mGluR5 receptor. To that end, I performed a proteome-wide quantitative comparison of protein levels in the hippocampi between WT, Fmr1-KO mice and Fmr1-KO/mGluR5-het cross mice (the “rescued” FXS model), obtained from the laboratory of Mark Bear (MIT). Pearson correlation of protein intensities showed good technical reproducibility between biological replicates and also showed that different genotypes are very similar to each other – of 5,238 detected proteins, only about 198 were significantly changing between genotypes. Yet, pairwise comparison of Fmr1-KO and WT revealed proteins that are known to be involved in memory, learning and long term potentiation. Moreover, the data suggests disturbed mitochondrial transcription regulation in FXS model. One of the significantly changing proteins, the major prion protein, had a lower expression level in Fmr1-KO in comparison with WT. Since I detected the same expression pattern in MEF cell lines, I postulate that Prp plays a significant role in FXS pathogenesis. In the FXS rescue model, most of the significantly changing proteins are involved in metabolic processes, with the exception of Citron Rho-interacting kinase which is functionally linked to Fmr1 gene and it is interacting with mGluR5 receptor, and therefore may be involved in the rescue mechanism. I also addressed absolute protein levels in analyzed mouse models. This approach showed no difference in the total proteome abundance between the three genotypes; however, it did show that proteins with higher upregulation levels in Fmr1-KO are more abundant that those that were found to be downregulated. This imbalance may be the cause of the overall increased protein levels previously detected in brains of FXS mice. However, a portion of changing proteins is rather small and therefore not significant and further experiments need to be performed to address this. Overall, this thesis presents an extensive proteomics analysis of the different cellular and animal models of FXS. Combined, these approaches compile a substantial source of information for the FXS research community. In addition, this data is contributing to the better understanding of FXS pathophysiology and development of a potential new treatments.
Print-on-Demand