Commonalities between SMA and ALS: investigation of ribosome heterogeneity and translational defects

According to the “ribosome heterogeneity hypothesis”, ribosomes are not all identical, and specialized ribosomes, capable of finely tuning translation of selected mRNAs, do exist. Multiple types of ribosomes are defined by several elements, such as their composition, and their association with different proteins, known as ribosome-associated proteins (RAPs). In particular, these proteins can exert a direct role on mRNA selection and translation efficiency. At present, some RNA-binding proteins (RBPs), such as Fragile X Messenger Ribonucleoprotein 1 (FMRP) and Fused in sarcoma (FUS) are considered putative RAPs and likely to regulate translation of specific mRNA transcripts via their binding to ribosomes. Even if dysregulation of translation has been associated with several neurodegenerative diseases, it remains unclear whether ribosome heterogeneity perturbations may be central to human pathologies. In this thesis I explored the hypothesis that disruption of ribosome heterogeneity in spinal muscular atrophy (SMA) and amyotrophic lateral sclerosis (ALS) is a common hallmark of these devastating neurodegenerative disorders. Despite different genetic causes, they share some notable commonalities, both at clinical and molecular level. Both pathologies, characterized primarily by lower motor neuron degeneration, are caused by mutations in genes encoding for ubiquitously expressed RBPs with common functions in mRNA metabolism. Recent studies have revealed that the survival motor neuron protein (SMN), which is involved in SMA, is a RAP and that the RBPs TAR DNA binding protein 43 (TDP43) and Matrin3 (MATR3), involved in ALS, are associated with the translation machinery. In addition, translatome changes are commonly observed in both diseases and loss of SMN and dysregulation of TDP43 and MATR3 are associated with translation defects of specific mRNAs. Therefore, exploration of a direct involvement of these proteins in translation may reveal a still uncharacterized level of heterogeneity which may account for most of translational defects and may reveal common dysregulated pathways. This thesis has three specific aims: i) characterize the RNA features of mRNAs bound by SMN-primed ribosomes and of their defects in in cellulo and in vivo models of SMA; ii) investigate ribosome heterogeneity and translational defects in TDP43-dependent ALS models and iii) explore the mutual role of SMN and MATR3 in ribosome heterogeneity and the impact of their loss in ALS and SMA models of disease. The first specific aim stemmed from previous findings obtained at the Institute of Biophysics (CNR, Italy) on the role played by SMN in translation and the consequent translational defects in SMA. SMN co-sediments with ribosomes and polysomes and acts as a RAP, influencing the translation efficiency of specific mRNAs. Computational analyses identified specific features in the sequences of transcripts bound by SMN-primed ribosomes: translational enhancer sequences in the 5’ untranslated region (UTR) and rare codons at the beginning of the coding sequence (CDS). SMN loss causes translational downregulation of transcripts bound by SMN-primed ribosomes (named SMN-specific mRNAs) due to ribosome-associated defects in SMA. Here, I validated these features analyzing the translation efficiency of reporter constructs bearing c-Myc translational enhancer at the 5’UTR or rare codons at the beginning of the CDS in an in cellulo model of SMA. Results of luciferase assays indicated that the two features are required for SMN-specific mRNAs to be translationally controlled by SMN-primed ribosomes. Co-sedimentation analysis of selected SMN-specific mRNAs along sucrose gradients of control and early symptomatic SMA brain from the Taiwanese model of the disease confirmed that these transcripts are depleted from the polysomal fractions, thus leading to possible changes at the protein level. I found that SMN-primed ribosomes play a critical role in regulating translation of AChE mRNA, a neuron-specific gene encoding for Acetylcholinesterase (AChE), molecular marker of the neuromuscular junction (NMJ) impairment at early stages of SMA. The second specific aim of this thesis is focused on the role played by TDP43 in translation in the context of ALS. ALS is associated with the dysregulation of multiple RBPs that play important roles in mRNA localization, maturation and translation, similarly to SMN. Overexpression or mutations of TDP43 are major contributors to ALS. Recently, MATR3 has also been associated with ALS cases. Here, I investigated the association of TDP43 with ribosomes and polysomes in neuronal compartments by taking advantage of a tag-free polysome profiling approach applied in axonal and cell body lysates obtained from neurons cultured in microfluidic chambers. Thanks to this technique, I explored the hypothesis that the overexpression of TDP43 WT and its mutated form TDP43 A315T cause axonal specific impairment in the balance between granule- and polysome-associated mRNAs, leading to disruption of local protein synthesis in ALS. By sequencing the mRNAs associated with subcellular regions (cell body or axon) and sub-compartments (RNA granules or polysomes), I found a loss of balance between polysome-engaged transcripts and free RNA in the axonal compartment upon overexpression of TDP43 WT and TDP43 A315T. These observations, together with results of axonal puromycilation assay, suggested that the imbalance between mRNAs associated with granules and polysomes derives from robust degradation of free RNA targets of TDP43, which causes a release of ribosomes free to bind to and hyper-translate remaining mRNAs, in particular TDP43 non-target mRNAs. To test this hypothesis, I used a set of constructs in cell line models of ALS that reveled the existence of a translational burden effect caused by decreasing levels of TDP43 mRNA targets. This revealed possibly completely new mechanisms leading to axonal loss in a cell model of ALS. Finally, the third specific aim of this thesis derived from the observation that SMN-bound ribosomes are associated with specific mRNAs that are involved not only in SMA but also in ALS. Among the mRNAs targeted by SMN-primed ribosomes involved in ALS, I observed the MATR3 mRNA, suggesting the existence of a common molecular trait connecting SMN and MATR3. Interactomics analysis of SMN-primed ribosomes revealed that MATR3 protein is also a top interactor. In addition, MATR3 has recently been shown to associate with the translational machinery and control the translation efficiency of a specific subset of mRNAs involved in the maintenance of human cell pluripotency. Given these findings, I hypothesized that MATR3 acts as a RAP and that both MATR3 and SMN converge on the same subset of ribosomes. Using biochemical methods, I confirmed the association of MATR3 with ribosomes in hiPSCs and found that this association is RNA-independent. Furthermore, I found that downregulation of MATR3 decreases the association of MATR3 and SMN with the translation machinery. Conversely, downregulation of SMN causes a decrease in the co-sedimentation of MATR3 with ribosomes. These results strongly support the SMN and MATR3 mutual-dependency in converging on the translation machinery and open a new translation-centered scenario to understand common mechanisms leading to ALS and SMA. The discovery of common molecular mechanisms between SMA and ALS could prove to be pivotal to better understanding these diseases and enable the development of novel therapies.