The multidrug resistance plasmid TP228 replicates at low copy number in Escherichia coli. Stable partitioning of this plasmid is mediated by three essential components: a ParA homologue, ParF; a centromere binding protein, ParG; and a centromere site, parH. ParF and ParG jointly assemble on the parH centromere forming the segrosome complex, and thereby direct intracellular plasmid transport. ParG belongs to the ribbon-helix-helix (RHH) class of dimeric DNA binding proteins. ParG specifically binds the parH site and also is a transcriptional repressor of the parFG genes. Previous studies demonstrated that unstructured N-terminal tails in ParG are not important for dimerization. Instead the tails are implicated in assembly of higher order nucleoprotein complexes essential for transcriptional repression and segrosome assembly, and also influence ParF nucleotide hydrolysis and polymerization. In this study we defined the role of residues in the RHH folded domain that are crucial for ParG dimerization and function. To achieve our goal the two α-helices, the intervening loop, and two C-terminal residues were analyzed fully by alanine scanning mutagenesis. Initially, ParG mutants were constructed and assessed for effects on normal plasmid partition activity and on dimerization. In vivo segregation assays and bacterial two-hybrid studies revealed mutation of residues F49 in α-helix 1 and W71 and L72 in α-helix 2 of ParG each resulted in defective plasmid partition activity and impaired dimerization. In vitro chemical cross-linking of purified proteins ParG-F49A, ParG-W71A and ParG-L72A demonstrated predominant monomeric species whereas wild-type ParG formed dimeric species as noted previously. Multiangle light scattering and sedimentation equilibrium analysis of the mutant proteins showed shifts in molar mass towards monomeric species with increased Kd values for dimerization. Protein-DNA interactions studied by gel retardation assays showed impaired interactions of ParG-F49A, ParG-W71A and ParG-L72A with parH. Results of conserved substitutions at position 71 showed that aromatic substitutions of W71 to Y71 or F71 are tolerated and have no apparent effects on ParG mediated plasmid segregation, but the non-aromatic W71L mutation blocked the segregation. However, a ParG double mutant bearing the ‘reversed’ amino acid pair (W71L-A52Y) retained plasmid segregation activity and behaved like wild-type ParG in dimerization assays in vitro and in vivo. Thus, substitution of W71 by tyrosine or phenylalanine does not disturb the monomer-monomer interface interactions that pack α-helix 2 from one monomer against residues of α-helix 1 and α-helix 2 of the partner monomer. Moreover, the permissible amino acid combinations at interacting positions 52 and 71 in ParG show significant flexibility and reveal key roles for these residues in function and dimerization of ParG. Overall, our in vivo and in vitro interaction studies provide novel information about the role of hydrophobic residues F49, W71 and L72 in ParG dimerization and activity. In the longer term, interference with dimerization by ParG and other centromere binding proteins using artificial ligands may provide a novel strategy for destabilization of antibiotic multiresistance plasmids.

Because of the severe toxicity and limiting side effects of traditional chemotherapy, there exists a critical need to develop better-tolerated, safer drugs to treat cancer. Recent advances in our understanding of the molecular mechanisms governing carcinogenesis have ushered in a new age in drug discovery and have enabled the design of much more sophisticated agents to treat cancer. This work describes two approaches to the development of novel, specifically targeted cancer therapeutics.The first approach involves the synthesis of a class of a new class of small molecules called epidithiodiketopiperazines (ETPs) designed to inhibit hypoxia-induced transcription. Specifically, these agents block the interaction of the transcription factor HIF-1α (hypoxia inducible factor-1α) and its required coactivator p300/CBP by inducing a structural change in p300 that renders it incapable of binding to HIF-1α. Preventing hypoxia-mediated transcription has the potential to stop the process of angiogenesis that is critical for sustained tumor growth and metastasis. Moreover, because HIF-1α also controls genes for energy production and matrix remodeling, ETPs may also halt metabolic adaptation and tumor progression. Our results show that ETPs prevent the association of HIF-1α and p300 and abrogate hypoxia signaling on both the transcriptional and translational levels in endogenous systems. In addition, they do not exhibit broad-spectrum cytotoxicity or global inhibition of the transcriptional response.The second approach addresses the validation of poly(ADP-ribose) glycohydrolase (PARG) as a new therapeutic target. This project describes studies aimed to further our understanding of the interaction between poly(ADP-ribose) polymerases (PARPs) and PARG with the ultimate goal of using this knowledge to design novel therapeutics. This portion of the dissertation involves a series of studies in mouse embryonic fibroblasts (MEFs) with genetic mutations in their PARP and PARG function. MEF cell lines containing a truncated form of PARG lacking the regulatory domain demonstrate over-activation of PARP-1, but not PARP-2. Additionally, deletion of the PARG regulatory domain impairs the DNA damage response to SSBs and DSBs and significantly increases cell death resulting from genotoxic stress. Taken together, these studies suggest a specific interaction between PARP-1 and the regulatory domain of PARG that is critical for proper PARP-1 function.

Poly(ADP)ribose (PAR) metabolism is essential to many cellular functions, including the maintenance of genomic integrity, the regulation of cell death mechanisms, as well as the regulation of gene expression. Recent work has uncovered many new players in the expanding effort to understand PAR metabolism and its cellular impact. PARP-1, the prototypical poly(ADP)ribose polymerase, was the first to be discovered, and has since been shown to be vital in the cellular response to DNA damage. Indeed, one report demonstrating that PARP-1 activation is required for apoptosis-inducing factor (AIF) release from mitochondria uncovered a novel link between DNA damage and signaling for cell death. The events following PARP activation, leading to signaling for AIF release, however, are still poorly understood. Based on our observations, we have developed a model to explain the nuclear/mitochondrial crosstalk that occurs following PARP activation. The work presented here answers several important questions regarding the relationship between ADP-ribose metabolism and mitochondria, including the role of PAR in signaling for the release of AIF, the presence of ADP-ribose metabolism protein members in mitochondria, and mitochondrial transcriptional effects following PARP activation. This work presents several novel findings, including the first report of a mitochondrial matrix isoform of poly(ADP-ribose) glycohydrolase (PARG) as well as direct evidence of mitochondria-associated PARP activity. Furthermore, it provides evidence for a novel effect of PARP-1 activation, in the specific transcriptional upregulation of the mitochondrial gene, NADH dehydrogenase, subunit 1 (ND1). Our data is consistent with the hypothesis that uncontrolled PARP activity results in energy metabolism dysfunction and cell death. Furthermore, it supports a model in which PARP activity is required for normal transcriptional responses in mitochondria following DNA damage. In total, this report adds to the body of work outlining the roles of PARP following DNA damage recognition and activation, demonstrating that ADP-ribose metabolism plays an important role in cell death regulation by both direct and indirect means.

Cancer cachexia is a wasting syndrome causing involuntary weight loss and muscle atrophy. PARP1 is a nicotinamide dinucleotide-dependent enzyme that modifies target proteins by PARylation. The reversal process, dePARylation, is mediated by the PARG enzyme. PARP1 inhibitors are potent cancer agents, while PARG inhibitors are in clinical trials for similar cancers. Here we examine the role of PARylation on muscle homeostasis in cancer cachexia. We employed mouse models with inducible muscle specific knockouts of Parp1 (Parp1-IMKO) or Parg (Parg-IMKO) to investigate their implications on skeletal muscle in a cancer cachexia model. We assessed muscle loss, grip strength, and gene expression. Results show that Parp1- IMKO mice had increased muscle wasting, while Parg-IMKO had degradation rates similar to wild-type mice during cancer cachexia. This suggests reduced PARylation might worsen cancer cachexia, while an increase does not. This supports PARG inhibitor development as anticancer alternatives. Our study highlights challenges with PARP1 inhibitors and the need to study PARylation and dePARylation in muscle health during cancer cachexia, impacting clinical strategies using PARP1 or PARG inhibitors.

L’ADN cellulaire qu’il soit nucléaire ou mitochondrial est constamment soumis à l’action de stress d’origine exogène ou même endogène à la base d’altérations plus ou moins profondes de sa structure. Ces modifications chimiques sont très variées et peuvent aller de l’oxydation d’une base aux cassures double-brins de la molécule d’ADN. Ces dernières sont considérées comme les dommages les plus agressifs pour la cellule car peuvent conduire à la perte d’information et donc à la mort cellulaire. Parmi les systèmes de surveillance de la stabilité du génome figure la Poly(ADP-ribosyl)ation (PARylation). Cette modification post-traductionnelle est assurée essentiellement par les protéines PARP et PARG et est caractérisée par l’incorporation des polymères d’ADP ribose (pADPr) sur des protéines cibles. La PARylation constitue un élément clé dans plusieurs voies de maintien de l’intégrité génomique (BER, NHEJ, HR). La PARylation est aussi décrite au niveau de la mitochondrie mais son rôle dans la gestion des DSBs de l’ADNmt n’est pas connu. Le travail, objet de cette thèse, consiste à étudier le rôle de la PARylation dans le cas des DSB au niveau général chez la drosophile et ensuite de comprendre les mécanismes de gestion des DSB mitochondriales et évaluer l’implication de la PARylation dans ce processus. Nos résultats montrent que : (1) le comportement de la PARylation ne varie pas au cours du processus de cassures et de réparation de l’ADN nucléaire, alors que l’expression des ARNm de PARP-I et PARP-II augmente durant la phase de réparation ; (2) les cassures de l’ADN mitochondrial, induites par la bléomycine, entraînent une augmentation du nombre de copies de l’ADNmt. Cette augmentation transitoire de la quantité de l’ADNmt est observée durant la phase des dommages et retourne à la valeur initiale durant la phase de la réparation. Ce comportement semble être régulé par PARP. L’ensemble de ces résultats suggère que la réparation des DSBs est indépendante de la PARylation au niveau nucléaire mais que la présence de PARP est importante. De plus, PARP semble avoir un rôle dans la régulation de la réplication de l’ADNmt en réponse à un stress génotoxique. / Both nuclear and mitochondrial DNA alterationsarise following exposure to environmental and endogenous stresses. These genomic alterations are various, ranging from base oxidation to DNA strand breaks, single- and double-strand breaks. These damages are highly detrimental to the cell because they can lead to loss of genetic information and thus to cell death. However, cells have developed various mechanisms to counteract this biological issue and to lead up to a complex DNA damage response (DDR). The Poly (ADP- ribosyl) ation (PARylation) is among these DDR systems. This post-translational modification is mainly carried out by PARP and PARG proteins and is characterized by the incorporation of polymers of ADP-ribose on target proteins. The majority of the PARylationfunctions are related to cellular stress response, particulary in response to genomic damages where it is implicated in many DNA integrity pathways such as Base Excision Repair, Non Homologous End Joining and Homologous Recombination. In contrast to the nucleus, PARylation is also described in the mitochondria but its role in mtDNA integrityis still a heavily debate issue, particularly in case of mtDNA DSBs.To understand it, we used Drosophila model wherePARP-B isoform (human PARP-1 ortholog) is the only enzymatically active form in Drosophila PARP family. The aim of this thesis is to study the role of PARylation in response to DSBs induction in nucleus and mitochondrial DNAand then to understand the mechanisms involved in mtDNA integrity and to evaluate the role of PARylation in this process. Our results show that PARylation level remains stable during DSBs induction and also during repair process,contrary to what is shown in Human cells.However, PARP-I and PARP-II mRNA expression increase during repair period. In mitochondria compartment,our data show an increase of mtDNA copy number in presence of mtDNA DSBs. This increased level returns to normal during repair period and seems to be dependent on PARP. All these results suggest that DSBs repair is PARylation independent at the nuclear level but that the presence of PARP is important. In addition, PARP appears to have a role in the regulation of mtDNA replication in response to genotoxic stress.

Les mitochondries sont des organites essentiels pour la production d'énergie cellulaire grâce à la synthèse d'ATP au cours des étapes de phosphorylations oxydatives (OXPHOS). Les complexes de la chaine respiratoire sont en partie codés par le génome mitochondrial (ADNmt), dont la structure est très sensible aux facteurs exogènes ou endogènes. De nombreuses mutations de l'ADNmt sont associées à des dysfonctionnements de la chaine respiratoire conduisant à des pathologies. La production d’Espèces Oxygénées Réactives (EOR) mitochondriale est la principale source de dommages à l’ADNmt. Une voie de réparation particulière, le système de réparation par excision de bases (BER) est mis en oeuvre dans ce cas. Nous avons, au cours de notre étude, analysé le système BER mitochondrial chez la drosophile. Dans une première approche, nous avons caractérisé de manière globale par une technologie de puces à ADN un ensemble de glycosylases et endonucléases impliquées dans la voie BER mitochondriale et comparé leur variation au cours du vieillissement. Cette étude a été complétée par une analyse transcriptionnelle sur des modèles de drosophiles mutantes pour des enzymes spécifiques de la voie BER, ceci afin de déterminer les éventuelles interactions transcriptionnelles entre les acteurs de cette voie. L’ARNm de Parp présentait de fortes variations dans les différents contextes mutants testés. C’est une molécule essentielle de la réparation BER. Elle a fait l’objet dans un deuxième temps, d’une étude plus approfondie. Dans le modèle des cellules S2, PARP bien que majoritairement nucléaire est également présent dans la mitochondrie. Le comportement différentiel des deux variants ARNm de Parp a pu être mis en évidence lors de stress cellulaires. Les isoformes protéiques de PARP observées dans nos études apparaissent différentes de celles habituellement décrites dans la littérature. Cet aspect a été discuté. / Mitochondria are key organelles mainly devoted to energy production through ATP synthesis. Such a function is permitted by oxidative phosphorylation (OXPHOS) within mitochondria inner membrane. Key components of the OXPHOS processes are encoded by mitochondrial DNA (mtDNA) that is particularly sensitive to exogenous or endogenous insults. As a result, mtDNA mutations are often correlated with OXPHOS dysfunction leading to diseases. ROS production in mitochondria is the main source of mtDNA damage. Such DNA damages are mainly taken over by BER systems within mitochondria. In this study, we focused on this peculiar mitochondrial DNA repair system in Drosophila. In a first step, we analysed in a comprehensive manner through microarray, most glycosylases and endonucleases involved in mitochondrial BER and compared their evolution during aging. Using mutant flies for specific BER enzymes, we started to decipher some of the transcriptional interactions between key BER actors. In a second step, Parp molecule was further studied due its changes in all mutant contexts and for its importance in several cellular processes. We described its nuclear but also its mitochondrial location in S2 cells. Interestingly, two Parp mRNA variants were observed showing distinct regulations following stress induction. However, PARP protein isoforms observed in this study were different compared to what was described in literature. This discrepancy is discussed.

The metabolism of ADP-ribose polymers (PAR) is involved in several cellular processes with a primary focus on maintaining genomic integrity. PAR metabolism following genotoxic stress is transient due to a close coordination between poly(ADP-ribose) polymerases (PARPs) which synthesize PAR and poly(ADP-ribose) glycohydrolase (PARG) which degrades PAR. PARP-1 inhibitors have emerged as promising anticancer therapeutics by increasing chemotherapy sensitivity and selectively target tumors harboring DNA repair defects. Several pharmaceutical companies have PARP-1 inhibitors in clinical trials for treatment of cancer. PARP-1 inhibitors are generally well tolerated, although they typically have poor selectivity among PARPs, and potentially other NAD binding enzymes. The promise of PARP-1 inhibitors as cancer therapeutics has led this dissertation research towards developing alternative tools and approaches to target PAR metabolism.One approach described is an evaluation of high-throughput PARP-1 screening assays as potential tools to discover new classes of PARP-1 inhibitors. These assays were compared to a widely used radiolabeling PARP-1 assay. They were found to offer several advantages that include simplicity, sensitivity, reproducibility, accuracy and eliminating the need for radioactive materials.The primary focus of this dissertation research was to develop PARG inhibitors as an alternative way of targeting PAR metabolism. Lack of viable genetically engineered animals, effective siRNA, and useful pharmacological 20 inhibitors has prevented PARG from being evaluated as a therapeutic target. This dissertation describes the first systematic approach, using Target related Affinity Profiling (TRAP) technology, for the discovery of PARG inhibitors. Identification of several hits led to the first detailed structural activity relationship (SAR) studies defining a pharmacophore for PARG inhibition. Interestingly, these molecules show varying degrees of PARP-1 inhibition, providing the first direct evidence for homology in the active sites of PARP-1 and PARG. Evaluation of a lead inhibitor has provided the first evidence for PARG inhibition in intact cells. Further optimization resulted in a cell permeable inhibitor with reduced toxicity and poor selectivity, providing evidence for a new class of inhibitors that disrupt PAR metabolism by inhibiting both enzymes. The use of dual PARG/PARP-1 inhibitors represents a new approach for therapeutic development of anticancer agents. Finally, directions aimed to overcome remaining challenges are discussed.

Adenosine diphosphate (ADP) ribosylation is a post-translational modification dependent on the transfer of ADPr units from nicotinamide adenine dinucleotide (NAD+) on to a plethora of biomolecules (i.e., proteins, DNA, RNA, etc.) in response to physiological stressors (i.e., nutrient deprivation, oxidative stress, DNA strand breaks). Poly-ADP-ribosylation (PARylation) is primarily mediated by the family of poly(ADP-ribose) polymerases (PARPs) and enzymatically degraded (dePARylation) by hydrolases such as poly(ADP-ribose) glycohydrolase (PARG). This thesis characterizes the role of poly(ADP-ribose) polymerase 1 (PARP1) and PARG in the skeletal muscle of healthy mature mice under normal physiological conditions. Specifically, we validate the deletion of Parp1 and Parg in inducible skeletal muscle-specific KO mouse models followed by performing general phenotyping of both male and female mice. The thesis concludes that under normal physiological conditions the activity of Parp1 or Parg in (de)PARylation is dispensable for maintaining skeletal muscle mass, function, and homeostasis in healthy mature mice.

La Poly(ADP-ribosyl)ation est une modification post-traductionnelle de protéines, impliquée dans un grand nombre de processus biologiques, dont la réparation de l’ADN. Alors que la fonction et le mode d’action de la Poly(ADP-ribose) (PAR) Polymérase 1 (PARP1), activée en réponse aux dommages de l’ADN sont bien compris, on en sait beaucoup moins sur la fonction et la régulation de l’enzyme de dégradation du PAR, la Poly(ADP-ribose) glycohydrolase (PARG). Dans le contexte de ce projet de thèse, nous décrivons de nouvelles lignées U2OS stables, déficientes pour toutes les isoformes de PARG, permettant la complémentation inductible avec chacun des isoformes de PARG. Ces modèles nous ont permis d’évaluer les contributions relatives des isoformes à la réparation de dommages à l’ADN. Nous avons identifié un nouveau partenaire cellulaire de PARG : la protéine-kinase dépendante des dommages à l’ADN (DNA-PK). Nous explorons l’interaction fonctionnelle de ces deux protéines dans le contexte de la réponse cellulaire à la camptothécine (CPT), un agent anticancéreux inhibant la topoisomérase I et provoquant l’activation simultanée de PARP1 et DNA-PK. / Poly (ADP-ribosyl) ation is a post-translational modification of proteins involved in a large number of biological processes, including DNA repair. While the function and mode of action of Poly (ADP-ribose) (PAR) Polymerase 1 (PARP1), activated in response to DNA damage, is well understood, much less is known about the function and regulation the PAR degrading enzyme, Poly (ADP-ribose) glycohydrolase (PARG). In the context of this thesis project, we describe new stable U2OS lines, deficient for all PARG isoforms, allowing the inducible complementation with each of the PARG isoforms. These models allowed us to evaluate the relative contributions of the isoforms to DNA damage repair. We have identified a new cellular partner of PARG: the DNA-dependent protein kinase-dependent kinase (DNA-PK). We explore the functional interaction between these two proteins in the context of the cellular response to camptothecin (CPT), an anticancer drug that inhibits topoisomerase I and induces the simultaneous activation of PARP1 and DNA-PK.

Transforming growth factor-β (TGF-β) signaling is initiated when the ligand binds to type II and type I serine/threonine kinase receptors at the cell surface. Activated TGF-β type I receptors phosphorylate R-Smads which relocate, together with co-Smads, to the cell nucleus and regulate transcription. Enhancement or repression of Smad-specific gene targets leads to intracellular protein compositions which organize functional complexes and thus govern cellular processes such as proliferation, migration and differentiation. TGF-β/Smad signaling relays are regulated by various post-translational modifications. From receptors to gene promoters, intricate interplays between phosphorylation, acetylation, ubiquitination and numerous other modifications, control Smad signaling initiation and duration. However, many steps in the cascade, including receptor internalization, Smad nuclear shuttling and transcriptional termination, still remain elusive. The open gaps in our understanding of these mechanisms most likely involve additional post-translational regulations. Thus, the aim of the present investigation was to identify novel modulators of TGF-β/Smad signaling. In the first part of this thesis, we show the importance of ADP-ribosylation in Smad-mediated transcription. We identified poly(ADP-ribose) polymerase 1 (PARP-1) as a Smad interacting protein. Our work revealed that PARP-1 forms direct interactions with Smad3/4, and PARylates residues in their MH1 domains. This modification restricts Smads from binding to DNA and attenuates Smad-activated transcription. PARylation is reversed by the glycohydrolase PARG. We provide evidence that PARG can de-ADP-ribosylate Smads, which enhances Smad-promoted gene regulation. In the second part, we examine a Smad-dependent gene target of TGF-β signaling, salt inducible kinase 1 (SIK). After induction, SIK cooperates with Smad7 and Smurf2 to downregulate the TGF-β type I receptor. The mechanism relies on both the kinase and UBA domain of SIK as well as the E3-ligase activity of Smurf2. In summary, we have unveiled two enzyme-dependent TGF-β/Smad modulatory mechanisms; SIK promoted receptor turnover and PARP-1/PARG-regulated Smad signaling.