Spinal muscular atrophy (SMA), the leading genetic cause of infant death, is a motor neuron disease (MND) caused by mutations or deletion of the survival motor neuron 1 (SMN1) gene, which codes for SMN protein. While SMN protein is ubiquitously expressed and crucial for the survival of all types of cells, motor neuron (MN) degeneration is the primary pathological result of SMN protein reduction. The origin of this selective vulnerability in SMA remains unsolved. In agreement with the large number of identified SMN binding partners, SMN has been linked to a vast number of cellular functions (e.g. splicing, transport and local translation of messenger ribonucleic acid (mRNA), endocytosis or autophagy), many of which impact protein homeostasis. The correct functionality of the mentioned housekeeping processes is critical for all cellular types, and thus it is puzzling why MNs are especially vulnerable to the reduction of SMN protein. The role that SMN plays in the regulation of the autophagy-lysosomal pathway (ALP), a major cellular degradative system, is not well studied. Recent studies have shown that SMN deficient cells display defects in the catabolic endosomal-autophagy pathway, leading to accumulation of autophagosomes (APs) and their undegraded cargo. The fact that APs form properly in SMN deficient cells, but are not correctly cleared from the cell, suggests a failure in the final step of the ALP, the AP degradation mediated by lysosomes. The main goal of this thesis was therefore to investigate the molecular mechanisms underlying the regulation of ALP by SMN, and whether and how alterations in this axis can result in the selective degeneration of MNs in SMA. To this end, MNs, derived from human induced pluripotent stem cell (hiPSC) lines, generated from patients affected by SMA and healthy individuals, have been used to uncover specific alterations in the ALP upon SMN reduction and the underlying molecular factors. Utilizing image-based experiments, I was able to discover that SMA MNs display a reduced number of lysosomes, compared to healthy MNs and isogenic controls, which leads to a defective AP-lysosome fusion. Interestingly, the remaining reduced pool of lysosomes in these SMA MNs exhibits an increased acidity, protease activity and axonal transport, none of which, seems to be sufficient to prevent MN loss. These findings demonstrate that SMN loss leads to a dysregulation of several key elements of the ALP, ultimately resulting in a reduced capacity of SMA MNs to degrade superfluous and potentially harmful material and to obtain essential building blocks from its recycling. To assess if the observed alterations in the ALP are specific to MNs or shared among other neuronal types that are typically not affected upon SMN reduction, I generated cortical neurons (CxNs) from the same hiPSCs and performed similar studies. These SMA CxNs did not show a reduction in the number of lysosomes or a change in their acidification status. Therefore, these findings indicate that the defective ALP upon SMN protein deficiency seems to be specific to spinal MNs and does not occur in all neuronal types. To explore the potential origin of the observed ALP abnormalities in SMA MNs, I focused on the transcription factor EB (TFEB), a master regulator of lysosomal biogenesis and autophagy functionality. The main inhibitor of TFEB activity, the mammalian target of rapamycin (mTOR), has been previously shown to be over-activated in SMA, but whether SMN protein exerts any regulation over TFEB had not been explored before. During this thesis, I found that TFEB, and several of its downstream targets essential for autophagy function, are indeed dysregulated in SMA. The decline in the expression levels of several TFEB target genes that I have discovered in SMA MNs confirmed the reduced activity of TFEB in these cells. In addition, overexpression of TFEB in SMA MNs and in an in vivo SMA zebrafish model did ameliorate the reduced survival of MNs and axonal dystrophy characteristic of these models, respectively, further confirming TFEB as a potential key protein in the loss of protein homeostasis of SMA MNs. These results also strengthened the observed over-activation of mTOR as potential key link between SMN reduction, defective ALP and MN vulnerability, but the mechanistic origin of this abnormally active mTOR in SMA MNs is unknown. I was able to find a potential candidate for this link in a previously published RNA-sequencing dataset, namely the mTOR activating tumor protein, translationally-controlled 1 (TPT1). A previous study showed that TPT1 acts as a negative regulator of basal autophagy, through activation of mTOR. Intriguingly, TPT1 was over-represented in the mentioned dataset. Upon TPT1 knock down in SMN deficient cells, autophagy flux and MN survival was ameliorated, which suggests TPT1 as a promising candidate downstream of SMN loss to revert the lysosomal and autophagic defects identified in SMA MNs. The dysregulation of the ALP, including alterations in TFEB levels, has been linked to the appearance of toxic protein aggregates in many neurodegenerative diseases (NDs). I therefore wondered if the observed reduction in ALP functionality in SMA MNs could result in an overlooked aberrant protein aggregation phenotype, similar to other NDs. Indeed, I showed in this study that three commonly used markers for protein aggregation, p62 protein - an autophagy cargo that accumulates when autophagy does not function properly -, vimentin - an integral component of the aggresome structure - and Proteostat - a fluorescent dye that binds protein aggregates - were increased in SMA hMNs compared to healthy controls. Together, these findings show that SMA MNs selectively display an accumulation of undegraded material, including APs, likely due to a dysregulation of TFEB, which additionally leads to a reduction in the number of lysosomes per MN and therefore to a decreased proteostasis capacity. Additionally, clear signs of intracellular protein aggregation were observed in SMA MNs, which could further increase the vulnerability of these neurons. These phenomena seem to be specific to MNs as no similar decrease in survival or lysosomal defects were observed in SMA CxNs and could at least partially explain the observed selective vulnerability of spinal MNs in SMA patients. Collectively, the presented Ph.D. thesis demonstrates that SMN protein regulates the correct activity of the ALP, and that low SMN levels result in the dysfunction of this critical pathway, specifically in MNs. This study highlights the importance of this axis in the survival of MNs, and places it in the spotlight for further research aiming to improve MN health, not only in SMA but potentially as well for other MNDs.

Les β2-agonistes sont couramment utilisés pour prévenir et réduire les symptômes de l'asthme et de la bronchoconstriction induite par l'exercice. Mais, pris en quantités supérieures aux doses thérapeutiques, les β2-agonistes ont un effet anabolisant qui a été clairement démontré in vivo. Un certain nombre d’acteurs sont mis en jeu dans la réponse biologique du tissu musculaire aux β2-agonistes. L’un de ces acteurs est la voie de signalisation PI3K/Akt/mTOR, voie d’initiation de la traduction, ayant un rôle majeur dans la synthèse protéique. Dans ce contexte, notre première étude avait pour objectif de déterminer la cinétique des événements moléculaires responsables de l’hypertrophie du muscle squelettique de rat après administration de formotérol pendant 1 jour (J1), 3 jours (J3) et 10 jours (J10). Nous avons montré que l’administration de formotérol induisait une hypertrophie musculaire à J3 et J10 associée à l’activation transitoire de la voie de signalisation PI3K/Akt/mTOR (J1 et J3), et à une diminution de l’expression de l’E3 ubiquitine ligase MAFbx/Atrogin-1 (J3). La voie autophagie lysosome ne semblait pas être affectée. Ainsi, l’ensemble de ces résultats suggère que l’activation de la voie PI3K/Akt/mTOR est associée à la voie ubiquitine-protéasome mais pas à la voie autophagie-lysosome. La régulation transitoire de la voie PI3K/Akt/mTOR suggère que d’autres voies de signalisation sont impliquées dans l’hypertrophie musculaire induite par le formotérol. Le 007-AM, analogue de l’AMPc, a été décrit comme pouvant stimuler la voie de signalisation PI3K/Akt/mTOR via l’activation de la protéine Epac, suggérant que le 007-AM puisse constituer une molécule de substitution à l’utilisation des β2-agonistes. Notre seconde étude avait pour but de déterminer si le 007-AM avait une action anabolisante sur le tissu musculaire, mais également de déterminer si la 007-AM était une molécule stable permettant d’envisager son usage dans un cadre pharmacologique. L’administration de 007-AM pendant 7 jours chez des souris n’engendrait pas d’hypertrophie musculaire. En revanche, in vitro sur cellules C2C12, le 007-AM activait la voie de signalisation PI3K/Akt/mTOR comme en témoignait l’augmentation de la phosphorylation des protéines rpS6 et 4E-BP1. Nos résultats montraient également que le 007-AM était instable dans le plasma alors que son produit de dégradation, le 007 était plus stable. Pris ensembles, ces résultats suggèrent qu’un traitement de 7 jours au 007-AM n’est pas suffisant pour induire une hypertrophie musculaire et que l’absence d’hypertrophie musculaire pourrait provenir de l’instabilité du 007-AM dans le plasma. Toutefois, des études supplémentaires seront nécessaires pour confirmer ces résultats / Β2-agonists are traditionally used to prevent and reduce asthma symptoms and bronchoconstriction induced by exercise. Nevertheless, when administrated in vivo, at relatively high, far away from therapeutic doses, β2-agonists induce anabolic effects. Numerous actors are involved in biological response of the skeletal muscle, induced by β2-agonists. PI3K/Akt/mTOR signaling pathway, which initiates translation, is one of these actors. In this context, our first study aimed at determined the kinetic of molecular events responsible for skeletal muscle hypertrophy after 1 day (D1), 3 days (D3) and 10 days (D10) of formoterol administration. We have shown that formoterol administration induced skeletal muscle hypertrophy at D3 and D10 associated with a transient activation of PI3K/Akt/mTOR signaling pathway (D1 and D3), and, with a decrease in E3 ubiquitin ligase MAFbx/atrogin-1 expression (D3). The autophagy-lysosome pathway seems not to be regulated by formoterol administration. Taken together, these results suggest that PI3K/Akt/mTOR activation is temporally associated with the regulation of ubiquitin-proteasome but not the autophagy-lysosome pathway. The transient nature of the regulation of PI3K/Akt/mTOR signaling pathway also indicates that other unidentified pathways are probably activated to sustain the increase in skeletal muscle mass. Recently, 007-AM synthetic molecule has been described to stimulate PI3K/Akt/mTOR signaling pathway through Epac protein activation, suggesting that 007-AM could be an alternative to the use of β2-agonists. The purpose of our second study was to determine whether 007-AM had an anabolic action on skeletal muscle and if 007-AM was stable allowing considering its use in pharmacology. 007-AM administration for 7 days to mice does not lead to muscle hypertrophy. Nonetheless, in vitro on C2C12 cells, 007-AM activated PI3K/Akt/mTOR signaling pathway by increasing phosphorylation of rpS6 and 4E-BP1. Our results showed that contrary to 007, 007-AM was instable in plasma. Altogether, these results suggest that a 7-day 007-AM treatment is not sufficient to induce skeletal muscle hypertrophy. This lack of hypertrophy could be due to 007-AM instability in plasma. However, supplemental studies are needed to confirm these results