La reprogrammation du métabolisme est maintenant considérée comme des caractéristiques des cellules cancéreuses et une conséquence de leur adaptation à un microenvironnement hostile se traduisant par une baisse de la concentration d’oxygène et de la disponibilité des nutriments. Donc, les cellules cancéreuses sont capables d’adapter leur métabolisme pour survivre et proliférer. Des avancées récentes dans la connaissance de ces modifications permettent l’émergence de nouvelles approches thérapeutiques ciblant spécifiquement ces changements métaboliques. Un des principaux régulateurs du métabolisme cellulaire est le coactivateur transcriptionnel PGC-1α (PPARgamma coactivator1-alpha). PGC-1α contrôle, entre autres, la biogénèse mitochondriale, la phosphorylation oxydative et l’oxydation des acides gras. Récemment, il a été montré que PGC-1α facilite la biogénèse mitochondriale dans les cellules cancéreuses du sein et augmentent significativement leurs potentiels métastatiques. Au contraire, il a été montré que la surexpression de PGC-1α diminue la formation de métastases dans le mélanome et l’adénocarcinome prostatique. Cependant, les modifications métaboliques et moléculaires conduisant à l’agressivité du cancer de la prostate sont, à l’heure actuelle, peu connues. Dans ce contexte, le but de ma thèse était d’étudier le rôle de PGC-1α sur le métabolisme et l’agressivité des cellules cancéreuses de prostate. Au cours de ma thèse, nous avons démontré que la diminution de l’expression de PGC-1α augmente les trois caractéristiques fondamentales de l’agressivité tumorale : la prolifération, la migration et l’invasion. Afin de déterminer les modifications métaboliques impliquées dans ce phénotype, nous avons réalisé des expériences de métabolomiques en comparant les cellules contrôles aux cellules dont l’expression de PGC-1α est diminuée (shPGC-1α). Nous avons montré que la baisse de PGC-1α augmente significativement la biosynthèse des polyamines. Les polyamines sont impliquées dans de nombreuses fonctions cellulaires, en particulier la prolifération et la migration cellulaire. Ainsi, nous avons inhibé la synthèse des polyamines avec le DFMO, l’inhibiteur de l’enzyme limitante de la voie : ODC, ou bien des siRNA dirigés contre ODC. Nous avons montré que les effets pro-migratoires et pro-invasifs dus à l’invalidation de PGC-1α sont bloqués par le DFMO et les siRNA ODC. De façon intéressante, l’ajout de polyamines exogènes restaure partiellement l’agressivité des cellules. En accord avec ces résultats, nous montrons que ODC est surexprimée quand PGC-1α est diminué et que l’expression de ODC est régulée positivement par l’oncogène c-MYC. En s’intéressant plus en détail à cet oncogène, nous observons que son niveau d’expression augmente dans les cellules invalidées pour PGC-1α et que l’inhibition de c-MYC bloque les effets pro-migratoires et pro-invasifs dus à l’invalidation de PGC-1α. Donc c-MYC participe au phénotype agressif lié à l’augmentation de la voie de biosynthèse des polyamines. Ces résultats in vitro ont été confirmés in vivo par l’analyse des micro-métastases, ils démontrent que les cellules shPGC-1α forment plus de métastases et le traitement par le DFMO inhibe la formation de micro-métastases. Finalement, les données cliniques démontrent que l’expression de PGC-1α est diminuée chez des patients atteints de cancer de la prostate, et cette diminution est corrélée avec une augmentation de c-MYC et ODC. En conclusion, nous avons démontré que PGC-1α est le régulateur majeur d’une voie onco-métabolique par c-MYC et qui promeut l’agressivité du cancer de la prostate par l’intermédiaire de la voie de biosynthèse des polyamines. Ce nouveau circuit métabolique représente une cible thérapeutique intéressante pouvant aider à freiner les formes avancées du cancer de la prostate. / Metabolism reprogramming are now considered to be characteristic of cancer cells and a consequence of their adaptations to a hostile microenvironment resulting in a decrease in oxygen concentration (hypoxia) and the availability of nutrients, particularly glucose and glutamine. Thus, cancer cells can adapt their metabolism to survive and proliferate. Recent advances in the knowledge of these modifications allow the emergence of new therapeutic approaches targeting these metabolic changes. One of the main regulators of cellular metabolism is the transcriptional coactivator PGC-1α (peroxisome proliferator-activated receptor gamma coactivator 1-alpha). PGC-1α controls mitochondrial biogenesis, oxidative phosphorylation and fatty acid oxidation. Recently, it has been shown that PGC-1α promotes mitochondrial biogenesis in cancer cells and dramatically increases their metastatic potential. On the contrary, it appears that overexpression of PGC-1α decreases the formation of metastases in melanoma and prostatic adenocarcinoma. However, the metabolic and molecular changes leading to the aggressiveness of prostate cancer are unclear. Oncogenes and tumor suppressor genes are known to be able to regulate the metabolic adaptations of cancer cells. Several studies show that the number of copies of the gene is increased in 30% of prostate cancers. Transgenic mice overexpressing c-MYC in the prostate develop prostatic intraepithelial neoplasia followed by prostatic adenocarcinoma. At the cellular level, c-MYC expression has been shown to stimulate glycolysis and glutaminolysis in tumor cells, by controlling the expression of genes involved in these metabolic pathways. In addition, c-MYC is also able to increase the polyamine synthesis pathway by inducing the expression of the limiting enzyme of this pathway, ornithine decarboxylase (ODC). In this context, the purpose of my thesis was to study the role of PGC-1α on the metabolism and aggressiveness of prostate cancer cells. During my thesis, we have shown that the decrease of PGC-1α expression increases the three fundamental characteristics of tumor aggressiveness: proliferation, migration and invasion. To determine the metabolic changes involved in this phenotype, we performed metabolic experiments and compared control cells to cells where PGC-1α expression is decreased. We show that the decrease of PGC-1α significantly increases the biosynthesis of polyamines. Polyamines are involved in many cellular functions, particularly in proliferation and cell migration. Thus, we inhibit the synthesis of polyamines with DFMO, an inhibitor ODC which is the rate limiting enzyme of this pathway. We have shown that pro-migratory and pro-invasive effects due to PGC-1α knockout are blocked by DFMO and ODC siRNA. Interestingly, the addition of exogenous polyamine partially restores the aggressiveness of the cells. Consistent with these results, we show that ODC is overexpressed when PGC-1α is decreased and that ODC expression is upregulated by the c-MYC oncogene. In addition, we observe that c-MYC expression increases in cells invalidated for PGC-1α and that the inhibition of c-MYC blocks the pro-migratory and pro-invasive effects due to the invalidation of PGC-1α. Therefore, c-MYC participates in the aggressive phenotype related to the increase of the polyamine biosynthesis pathway. These in vitro results were confirmed in vivo by micro-metastasis analysis, we demonstrate that shPGC-1α cells form more metastases and treatment with DFMO inhibits the formation of micro-metastases. Clinical data indicate that PGC-1α expression is decreased in patients with prostate cancer, and this decrease correlates with an increase in c-MYC and ODC. In conclusion, we show that PGC-1α is the major regulator of an onco-metabolic which promotes prostate cancer aggressiveness via the polyamine biosynthesis pathway.

Arginine decarboxylase is a key enzyme involved in the polyamine pathway of organisms. Pyruvoyl-dependent arginine decarboxylases are expressed in the form of proenzymes that self-cleave to form N-terminal [beta] and C-terminal [alpha] subunits generating an active pyruvoyl group at the [alpha] terminus. We have identified an archaeal homolog of a pyruvoyl-dependent arginine decarboxylase in Chlamydophila pneumoniae that could play a role in the persistence of the organism in the host. The recombinant enzyme showed highest activity at pH 3.4, which is the lowest optimum pH ever reported for a pyruvoyl dependent arginine decarboxylase. The proton-consuming decarboxylation raises intracellular pH, and thereby plays a role in acid-resistance. It could inhibit the pro-inflammatory nitric oxide synthase resulting in asymptomatic infection. A variant protein Thr⁵²Ser at the predicted cleavage site showed less pro-enzyme cleavage and activity compared to the wild-type. The homologs of arginine decarboxylase and flanking arginine-agmatine antiporter were also found in different biovariants of Chlamydia trachomatis. In the invasive L2 strain of C. trachomatis, the presence of a nonsense codon in the gene encoding arginine decarboxylase enzyme prevented the expression of an active enzyme. The variant protein with tryptophan replacing nonsense codon restored arginine decarboxylase activity. The non-invasive D strain of C. trachomatis had an intact arginine decarboxylase gene, but it was recombinantly expressed as a proenzyme that was uncleaved. The arginine-agmatine antiporters from both the strains were active and transported tritiated arginine into their cells. The polyamine pathway of the crenarchaeon Sulfolobus solfataricus uses arginine to make putrescine, but the organism lacks homologs of arginine decarboxylase. However, it has two paralogs of pyruvoyl dependent S-adenosylmethionine decarboxylase − SSO0536 and SSO0585. These enzymes were recombinantly expressed as pro-enzymes that self-cleaved into [beta] and [alpha] subunits. Even with a 47% amino acid sequence identity, the SSO0536 protein exhibited significant arginine decarboxylase activity whereas SSO0585 protein had significant S-adenosylmethionine decarboxylase activity. This is the first report of an S-adenosylmethionine decarboxylase enzyme showing alternative decarboxylase activity. The chimeric protein with the [alpha]-subunit of SSO0585 and [beta]-subunit of SSO0536 had arginine decarboxylase activity, suggesting that the residues responsible for substrate recognition are located in the amino terminus. / text