Nous cherchons à comprendre le mode de fonctionnement d’HACE1, un suppresseur de tumeur majeur, dont l’expression est altérée dans 50% des tumeurs humaines. Notre équipe a démontré que cette ubiquitine-ligase cible et provoque la dégradation de Rac1 au niveau du protéasome. Rac1 est une protéine oncogénique promouvant la croissance tumorale. En ciblant Rac1, HACE1 permet de restreindre le stress oxydatif des cellules, diminuant ainsi leurs dommages à l’ADN. Pour comprendre les voies de signalisation cellulaires ciblées par HACE1, nous avons ensuite recherché des protéines interagissant avec elle. Mes travaux révèlent que l'optineurine (OPTN), une protéine jusqu’alors connue dans des désordres neuro-dégénératifs, forme un complexe avec HACE1. Nous avons montré que ce complexe régule transcriptionnellement et traductionnellement la Cycline-D1. Nous avons également montré que l’OPTN se localise dans les points d'ancrage des cellules à la matrice extracellulaire (MEC) et régule leur formation. En réponse à l’élasticité de la MEC, nous avons montré que le complexe HACE1-OPTN réprime, au travers du métabolisme, la division cellulaire. Une étude que nous avons effectuée sur des données cliniques indiquerait l'importance d’une perte d’expression d’OPTN dans le cancer du sein, associée à des dérégulations des tensions de la MEC tumorale / We have established a novel regulatory mechanism that restricts Rac1 activity through ubiquitylation and targeting to the proteasome of the active form of the GTPase for degradation and signal termination. This regulation is dominant over the classical GEF/GAP cycle of regulation. We identified the E3 ubiquitin ligase (E3L) HACE1 as the main enzyme that catalyzes Rac1 ubiquitylation. HACE1 is a major tumor suppressor that limits Rac-dependent NADPH oxidase complex activity, S phase entry, cell migration, mammary cell transformation and tumor growth in animal models. In order to determine how HACE1 activity is controlled, we conducted a whole genome two-hybrid screen and identified Optineurin (OPTN) as a primary partner of the E3L HACE1. We report that OPTN is a new Extracellular matrix (ECM) stiffness sensor that activates HACE1 E3L activity. OPTN controls adhesion-mediated ubiquitin-proteasome degradation of Rac1 to tune Rac1 signaling with tissue stiffness. Loss of OPTN is associated with a gain of cell-ECM adhesive properties and enhanced integrin-mediated proliferative signaling and metabolic activity. Interestingly, cells that loose OPTN display atypical mechanical properties and apparent uncoupling of Focal Adhesions growth from acto-myosin contractile activity. Together, our findings establish the first link between the Rac1 ubiquitylation pathway and ECM compliance sensing and define OPTN as a previously unknown mechanical sensor. OPTN and HACE1 would then act as a tumor-suppressor complex that adapts cell proliferative response to ECM mechanical properties in order to insure both Tensional integrity and Redox homeostasis of cells

Thesis advisor: Marc A. T. Muskavitch / Notch signaling is an evolutionarily conserved developmental pathway regulated by two classes of transmembrane proteins: the Notch receptors and the Delta/Serrate/LAG-2 (DSL) ligands. Notch and DSL ligands mediate cell-cell communication that results in a downstream signaling cascade that affects many aspects of metazoan development. Additional regulatory mechanisms that affect Notch signaling are being discovered continuously, and recent findings highlight the importance of endocytosis, ubiquitylation and subcellular trafficking as essential requirements for proper signaling. In order to obtain further insights into the regulation of Notch signaling, I took a two-fold approach, combining genetic and molecular techniques in <italic>Drosophila</italic>. First, I took part in a large-scale transposon-based screen in the developing <italic>Drosophila</italic> eye to identify additional genes involved in the pathway. We screened 10,447 transposon lines from the Exelixis collection for modifiers of cell fate alterations caused by overexpression of Delta, and we identified 170 distinct modifier lines that may affect up to 274 genes. I further analyzed a previously uncharacterized gene, which we have named <italic>Amun</italic>, and showed that it localizes to the nucleus and contains a putative DNA glycosylase domain. Further analyses of Amun reveal that altered levels of Amun function interfere with cell fate specification during eye and sensory organ development. Second, to investigate structural requirements for ubiquitylation of Delta, I analyzed four individual lysine residues in the Delta intracellular domain, and assessed their necessity for Delta signaling activity. I find that a conserved residue, DeltaK742, is essential for Notch signaling in the <italic>Drosophila</italic> imaginal wing disc and is apparently required for ubiquitylation of Delta by the E3 ubiquitin ligase, Mind bomb1 (one of two E3 ubiquitin ligases required for Delta signaling activity). Taken together, the findings from this thesis research contribute to the advancement of our understanding of different aspects of Notch signaling and Notch-mediated development. / Thesis (PhD) — Boston College, 2009. / Submitted to: Boston College. Graduate School of Arts and Sciences. / Discipline: Biology.

Macrophages are sentinels present in most tissues of the body, where they recognise and respond to biological dangers. Recognition and uptake of particles is mediated through phagocytic receptors which upon activation induce appropriate responses. These responses need to be tightly regulated in order to destroy pathogens but prevent uncontrolled inflammation. Phagocytosis is an evolutionarily conserved process required for host defence and homeostasis. During phagocytosis, particles are recognised by cell surface receptors that trigger rearrangement of the actin cytoskeleton and internalization of the bound particle into a de novo, membranous organelle known as the phagosome. Regulation of phagocytosis and phagosome maturation can be achieved through changes in transcription/translation and differential recruitment of proteins but also through their non-translational modifications. Here I explored the role of ubiquitylation in the phagosome biogenesis of Interferon-gamma (IFN-ɣ) activated macrophages. Ubiquitylation is a diverse, reversible post-translational modification which is not only involved in protein degradation but also in vesicle trafficking and immune signalling. My data shows that phagosomes are enriched in polyubiquitylation, which is further enhanced by IFN-ɣ. I applied a targeted AQUA peptide approach by which we quantified ubiquitin chain linkage peptides from phagosome samples by PRM. This data shows that all chain linkages apart from M1/linear chains are present on phagosomes. Furthermore, IFN-ɣ activation enhanced K11, K48 and K63 chains significantly. In order to identify the molecular function of this polyubiquitylation, I characterized the ubiquitinome of phagosomes of IFN-γ activated macrophages and can demonstrate that ubiquitylation is preferentially attached to proteins involved in vesicle trafficking, thereby delaying fusion with late endosomes and lysosomes. I demonstrated that most ubiquitin chains are on the cytoplasmic site of the phagosome enabling an interaction of ubiquitin chains with cytosolic proteins such as Rab7. Rab7 a major regulator of vesicle trafficking could be shown to be ubiquitylated on phagosomes. I further showed that phagosomal recruitment of the E3 ligase RNF115 is enhanced upon IFN-γ stimulation and RNF115 is responsible for most of the increase of K63 polyubiquitylation of phagosomal proteins. Knock-down of RNF115 promotes phagosome maturation and induces an increased pro-inflammatory response to Toll-like receptor (TLR) agonists, indicating that RNF115 is a negative regulator of vesicular trafficking to the lysosome and disruption of this pathway induces pro-inflammatory responses in macrophages. In conclusion, this is the first study showing unbiasedly that ubiquitylation plays an important role in vesicle trafficking to the lysosome.

Yes / Rpn13 is an intrinsic ubiquitin receptor of the 26S proteasome regulatory subunit that facilitates substrate capture prior to degradation. Here we show that the C-terminal region of Rpn13 binds to the tetratricopeptide repeat (TPR) domain of SGTA, a cytosolic factor implicated in the quality control of mislocalised membrane proteins (MLPs). The overexpression of SGTA results in a substantial increase in steady-state MLP levels, consistent with an effect on proteasomal degradation. However, this effect is strongly dependent upon the interaction of SGTA with the proteasomal component Rpn13. Hence, overexpression of the SGTA-binding region of Rpn13 or point mutations within the SGTA TPR domain both inhibit SGTA binding to the proteasome and substantially reduce MLP levels. These findings suggest that SGTA can regulate the access of MLPs to the proteolytic core of the proteasome, implying that a protein quality control cycle that involves SGTA and the BAG6 complex can operate at the 19S regulatory particle. We speculate that the binding of SGTA to Rpn13 enables specific polypeptides to escape proteasomal degradation and/or selectively modulates substrate degradation. / BBSRC [grant number: BB/L006510/1] and the Wellcome Trust [grant number: 092107/Z/10/Z]. K.K. was supported by the UPStream network [EU, FP7, ITN project 290257]

UHRF1 is an E3 ubiquitin ligase and a key epigenetic regulator establishing a crosstalk between DNA methylation and histone modification. Despite the important biochemical role of UHRF1 in cells, its overexpression has been found in almost all primary cancer types including breast cancer, lung cancer and so on. Numerous evidence indicates a strong link between tumorigenesis and UHRF1 overexpression, supporting its potential as a universal biomarker for cancer. However, UHRF1 is “yet-to-be drugged” and no highly potent chemical probes have been developed to target UHRF1 to date. In this study, we proposed two drug design approaches for UHRF1. The first approach is to construct multivalent DNA-peptide nucleosome mimetics that can target UHRF1 directly. For UHRF1 to promote DNA methylation, the interaction with nucleosomes, both through a DNA-binding (SRA) and histone-binding domain (TTD-PHD), and ubiquitylation of histone H3 are necessary to recruit DNA methyltransferase. We utilized the natural binding activity between UHRF1 and nucleosome in cells to develop a DNA-peptide hybrid that mimics UHRF1’s interaction with nucleosomes, thereby inhibiting UHRF1-dependent histone ubiquitylation and impairing its function in controlling DNA methylation. Here, we described the synthesis of the DNA-peptide hybrids using different lengths of PEG linkers including PEG2, 6, 8, 16 and 24. We purified and characterized the molecules with RP-HPLC and ESI-MS. Biophysical assays such as ITC and METRIS were conducted to study about the binding affinities of these DNA-peptide hybrids. In vitro UHRF1 ubiquitylation assays were performed to investigate the inhibition efficacy of these inhibitors, and pull-down assays were conducted to study their selectivity. In addition, mass photometry assays were used to study the stoichiometry of the binding between UHRF1 and the DNA-peptide hybrids. We demonstrated that multivalent DNA-peptide hybrids possess high affinity for UHRF1 and can inhibit histone ubiquitylation. Among them, In16 can form a 1:1 binding complex with UHRF1, substantiating its ability to be used as a molecular tool for structural analysis of UHRF1. In the second approach, we designed and constructed three generations of multi-domain protein inhibitors of E2 enzyme Ube2D, including RING-UBL (RU), UBOX-UBL (UU) and UBOX-UbvD1.1 short/long (UD1 and UD2). Through targeting both the RING- and backside-binding sites on Ube2D, UHRF1 enzymatic function can be indirectly inhibited as Ube2D is the only cognate E2 enzyme that cooperates with UHRF1 for histone H3 ubiquitylation. In this study, ITC was used to measure the binding affinities of these inhibitors, showing an increasing affinity from the first inhibitor RU to the last one UD2, ranging from 10-6 M to

Mémoire numérisé par la Division de la gestion de documents et des archives de l'Université de Montréal.

L’ubiquitylation est une modification post-transcriptionelle qui est nécessaire pour la dégradation des protéines mais aussi pour la régulation et la localisation de nombreux facteurs. Un grand nombre de protéines impliquées dans la réplication de l’ADN et dans la réponse aux lésions de l’ADN sont ubiquitylées. L’ubiquitylation durant la réponse aux dommages de l’ADN dépend de l’enzyme d’activation de l’ubiquitine UBA1 qui est située au sommet de la cascade d’ubiquitination. Durant ma thèse, j’ai mis à jour le mécanisme de recrutement d’UBA1 au niveau de l’ADN endommagé ainsi qu’un rôle majeur de l’ubiquitylation dans la voie de signalisation ATR.A l’aide d’une approche in vitro qui mime la voie de signalisation ATR, j’ai montré qu’UBA1 est recrutée au niveau de molécules d’ADN linéaire et qu’elle est nécessaire à l’ubiquitylation des protéines recrutées sur ce substrat. J’ai également découvert que l’ubiquitylation et le recrutement d’UBA1 in vitro sont dépendants de la kinase DNA-PKcs et de la poly(ADP-ribose) polymérase PARP1, deux senseurs majeurs des lésions de l’ADN. Il apparait que PARP1 régule le recrutement d’UBA1 via la formation de chaines de poly(ADP)-ribose (PAR). De plus, j’ai montré qu’UBA1 est capable de se lier aux chaines PAR. J’ai identifié la région d’UBA1 capable d’interagir avec les chaines PAR : il apparait que cette région est désorganisée et riche en acide aminés hydrophobes. La comparaison de la protéine UBA1 de levure et la protéine UBA6 humaine qui ne lient pas PAR nous a permis d’identifier les acides aminées hydrophobes nécessaires pour la lésion à PAR.J’ai aussi démontré qu’UBA1 est nécessaire pour la réponse aux lésions de l’ADN. En effet, l’inhibition ou la déplétion d’UBA1 conduit à une perte de la phosphorylation de Chk1 dans notre système in vitro. De même, le traitement avec des molécules induisant des lésions de l’ADN telles que le CPT, le MMS et la bléocine conduit à une phosphorylation moindre de Chk1 lorsqu’UBA1 est inhibée. Afin de démontrer que la liaison d’UBA1 aux chaines PAR est cruciale pour la réponse aux dommages, j’ai mis en place un système in vivo permettant d’exprimer des mutants d’UBA1 incapable de lier les chaines PAR.Globalement mes résultats indiquent qu’UBA1 est recrutée au niveau de l’ADN endommagé à l’aide de PARP1 et DNA-PKcs. Plus précisément, il apparait que la liaison avec les chaines PAR et l’ubiquitylation de protéines spécifiques est nécessaire pour la mise en place de la voie de signalisation ATR. L’importance d’UBA1 dans le processus est soulignée par le fait que son inhibition ou son inactivation conduit à une phosphorylation moindre de Chk1. Il est raisonnable de penser que des inhibiteurs d’UBA1 puissent être utilisés pour cibler la voie ATR dans les cellules cancéreuses. Finalement, cette étude devrait permettre de mieux comprendre comment les interactions entre les processus d’ubiquitylation et de PARylation permettent de réguler la réponse aux dommages. / Ubiquitylation is an important posttranslational modification that is necessary for protein degradation as well as for the regulation and the localization of many cellular factors. A number of proteins implicated in DNA replication and DNA damage response are ubiquitylated. Ubiquitylation during the DNA damage response is selectively dependent on the ubiquitin-activating enzyme UBA1, which functions at the apex of the ubiquitylation cascade. In this thesis, I describe the mechanism of UBA1 recruited at damaged sites and uncover the role of ubiquitylation in ATR signaling.Using a cell free system developed in the lab that recapitulates ATR kinase-signaling pathway, I present evidence that, UBA1 is recruited to linear DNA substrates and mediate ubiquitylation of DNA-bound proteins. I found that protein ubiquitylation and the recruitment of UBA1 to DNA in cell-free extracts was dependent on the kinase DNA-PKcs and on the poly ADP-ribose polymerase PARP1, two sensors of DNA lesions. PARP1 regulates UBA1 recruitment in large part through poly (ADP)-ribose (PAR) chain formation. UBA1 exhibited affinity for PARP1 and for PAR chains. Furthermore, we have identified minimal region on UBA1 which is prominently hydrophobic and disordered region of UBA1 which are required for its PAR binding activity. Through comparison with yeast UBA1 and human UBA6 which failed to bind with PAR chains, we identified hydrophobic amino acid residues which are critical for PAR binding.I also show evidence that UBA1 is required for efficient DNA damage signaling. In a cell free system, chemical inhibition or siRNA depletion of UBA1 led to the loss of ChK1 phosphorylation, suggesting that UBA1 activity is required for efficient DNA damage response. Consistent with these observations, when cells were treated with DNA lesion inducing drugs like CPT, MMS and Bleocin, we observed less efficient Chk1 phosphorylation. I have developed UBA1 replacement system to demonstrate functional significance of mutation in PAR binding residues of UBA in DNA damage response.Collectively, these results indicate that UBA1 is recruited to DNA damaged sites in a DNA-PKcs and PARP1 dependent-manner, in a larger part through its interaction with PAR chains and that protein ubiquitylation on DNA damages is necessary for the assembly of a productive ATR signaling complex. The importance of role of UBA1 in DNA damage response is underscored by the finding that UBA1 inhibition leads to inefficient Chk1 phosphorylation which is required for efficient DNA damage response. Thus, UBA1 inhibitors could be used to target ATR signaling in cancer cells. This study should eventually lead us to provide more insights on how cell maintains genome integrity through crosstalk between posttranslational modifications including ubiquitylation and PARylation.

Mémoire numérisé par la Division de la gestion de documents et des archives de l'Université de Montréal

L'ubiquitylation des protéines est une modification post-traductionnelle qui joue un rôle capital dans la régulation des nombreuses fonctions cellulaires, y compris la croissance cellulaire et la prolifération. Les dysfonctionnements de ce mécanisme sont à l'origine de diverses maladies telles que le cancer par exemple. Le processus d'ubiquitylation implique une série des réactions enzymatiques en cascade, catalysées par une famille des enzymes, structuralement très proches. Cette famille est composée des enzymes activateurs d'ubiquitine (E1s), des enzymes de conjugaison d'ubiquitine (E2s) et des ligases d'ubiquitine (E3s). Les interactions entre E2s et E3s sont dans le centre de la cascade d'ubiquitylation. Une combinaison particulière des pairs E2/E3 va déterminer le type de chaînes d'ubiquitine qui seront attachées à la protéine d'intérêt pour ensuite déterminer la fonction régulatrice de la voie d'ubiquitylation. A ce jour, seulement une petite fraction de paires possibles entre E2 et E3 a été investiguée par des approches biochimiques et in vitro. Cependant ces approches ne reflètent pas forcément des conditions qu'on trouve dans une cellule vivante. Prenant ceci en considération, les principales objectives de ma thèse seront comme suit : identifier et optimiser une méthode de détection et de quantification des interactions E2/E3 dans une cellule vivante de la levure de boulanger (Saccharomyces cerevisiae) ; construire une bibliothèque de souches de la levure qui permettrait d'établir des interactions entre E2 et E3 ; chercher de nouvelles potentielles paires E2/E3 ; caractériser fonctionnellement une potentielle paire E2/E2. Il est difficile de trouver une méthodologie appropriée afin d'étudier les interactions entre E2 et E3 parce qu'ils sont relativement faibles et transitoires. Leurs études nécessitent donc des techniques de détection avec une grande sensibilité. Parmi différentes techniques nous avons testé et choisi la complémentation bimoléculaire de la fluorescence, BiFC. Kurtosis, une mesure permettant localiser et quantifier la fluorescence BiFC-spécifique. Nos résultats nous nous avons permis à identifier 117 putatives paires E2/E3 parmi quels, 23 paires ont été déjà décrit dans la littérature. Parmi 94 nouvelles paires, certains E3s interagissent avec seulement une seule E2 ou d'autres donnent un signal BiFC avec plusieurs E2s. Ubc13, Ubc1 et Ubc4 sont les E2s qui interagissent le plus souvent. Nous avons identifié aussi une interaction entre les protéines Asi1 et Asi3 et les enzymes de conjugaison d'ubiquitine Ubc6 et Ubc7. Asi1 et 3 sont connus de former un complexe Asi1/3 sur la membrane intérieure du noyau impliqué dans la réponse de la cellule aux acides aminés extracellulaires. Ces protéines contiennent un domaine RING caractéristique pour les ligases d'ubiquitine mais cette activité n'était pas démontrée auparavant. / Protein ubiquitylation is a post-translational modification that plays a crucial role in regulating many cellular functions, including cell growth and proliferation. Defects in this control mechanism cause cancer and other diseases. The ubiquitylation process involves a cascade of enzymatic reactions catalyzed by a family of structurally-related enzymes, namely ubiquitin activating enzymes (E1s), ubiquitin conjugating enzymes (E2s) and ubiquitin ligases (E3s). Interactions between E2s and E3s are in the centre of ubiquitylation cascade and it is a combination of particular E2/E3 pairs that determine what types of ubiquitin chains are made, thus determining the regulatory functions of the ubiquitin pathway. To date, only a small fraction of all possible E2/E3 pairs have been investigated, mainly using biochemical and in vitro approaches that may not accurately reflect the conditions that occur in living cells. We aimed to develop a method capable of detecting specific E2-E3 interactions under physiological conditions. Using budding yeast as a model organism, we found that the Bimolecular Fluorescence Complementation (BiFC) enables sensitive detection of the well described Ubc4-Ufd4 pair under endogenous conditions. The assay is specific since the interaction signal is lost in yeasts expressing Ubc4 mutants truncated in its E3 interaction domain. We then used this system to further analyze the physiological network of E2 and E3 enzymes in living yeast. We performed a microscopy screen to assay all interactions between eleven E2s and 56 E3s. Our results show that approximately 20% of all E2/E3 combinations give a detectable BiFC signal. Few E3s interacted only with a single E2, whereas most E3s produced a BiFC signal with multiple E2s. Ubc13, Ubc1 and Ubc4 were found to be the most frequently interacting E2s. Our results match many examples from current literature but we also detected 94 new E2/E3 interactions, in particular we identified an interaction between the proteins Asi1 and Asi3 and E2s Ubc6 and Ubc7. Asi1 and Asi3 are known to form a complex (the Asi1/3 complex) at the inner nuclear membrane and are involved in the regulation of the response to extracellular amino acids. The Asi1/3 complex was suspected to function as a ubiquitin ligases because they contain a RING domain, but this has previously not been demonstrated. We therefore further characterized them functionally.

Résumé : Le “Suppressor of cytokine signaling -1” (SOCS1) est une protéine de 24 kD qui fonctionne principalement comme un régulateur négatif des voies de signalisation intracellulaires. SOCS1 inhibe l’axe JAK-STAT et induit l’ubiquitylation et la dégradation de certaines protéines cibles. L’expression de SOCS1 est diminuée par l’hyperméthylation de son promoteur dans plus de 65% des cas de carcinome hépatocellulaire. Les souris déficientes en SOCS1 ne survivent que trois semaines après la naissance à cause d’une hyperinflammation induite par IFN-[gamma]. Afin d’étudier le rôle anti tumoral de SOCS1 dans le foie, nous avons utilisé les souris Socs1[indice supérieur -/-]Ifng[indice supérieur -/-], Ifng[indice supérieur -/-] et les souris de type sauvage. Nous avons démontré que le taux de la régénération du foie après une hépatectomie partielle était augmenté chez les souris déficientes en SOCS1. De plus, les souris Socs1[indice supérieur -/]-Ifng[indice supérieur -/-] étaient plus susceptibles pour le développement des nodules hépatiques suite à un traitement avec diethylnitrosamine (DEN). Par contre, les souris déficientes en IFN-[gamma] ont démontré une résistance contre le cancer du foie. Néanmoins, au contraire de nos attentes préliminaires, nous n’avons pas observé une augmentation des taux sériques d’IL-6. Pourtant, la prolifération compensatoire et la synthèse de l’ADN étaient élevées chez les souris SOCS1 KO. Afin d’expliquer cette observation, nous avons étudié l’activation de p53. Nous n’avons pas trouvé une réponse différente de stabilisation ni de phosphorylation de p53 (Ser15) après traitement au cisplatin (in vitro) ou DEN (in vivo). Par contre, nous avons observé que l’expression du gène Cdkn1a était élevée chez les hépatocytes déficients en SOCS1. De plus, l’expression ectopique de SOCS1 pouvait supprimer l’expression de p21 chez les cellules HepG2 traitées au cisplatin. Nous avons aussi constaté que la stabilité de p21 était augmentée chez les hépatocytes primaires déficients en SOCS1. En effet, SOCS1 induisait l’ubiquitylation et la dégradation de p21. SOCS1 pouvait interagir avec p21 par son domaine SH2. De plus, SOCS1 pouvait contrôler la localisation cytoplasmique de p21 en régulant l’activité d’AKT. Bien que p21 soit connu comme un inhibiteur du cycle cellulaire, il peut également participer à l’assemblage des complexes CDK4-Cyclin D. Nous avons démontré que l’expression de p21 et des cyclines de type D était augmentée chez les souris déficientes en SOCS1 après l’hépatectomie partielle. En diminuant l’expression de p21 par shRNA, nous avons empêché la réponse proliférative des hépatocytes SOCS1 KO. Finalement, nous avons trouvé que l’expression élevée de p21 chez les hépatocytes déficients en SOCS1 rendait les cellules plus résistantes contre l’apoptose. En conclusion, nos résultats suggèrent que SOCS1 protège contre le cancer du foie par la régulation des activités oncogéniques de p21. // Abstract : Suppressor of cytokine signaling - 1 (SOCS1) is an inducible 24 kD protein that principally acts as a negative regulator of different intracellular signaling pathways. SOCS1 exerts its regulatory feedback by blocking the JAK - STAT axis and inducing ubiquitylation and subsequent proteasomal degradation of target proteins. The gene coding for Socs1 has a CpG - rich promoter and can be methylated by methyltransferases. SOCS1 is silenced due to hypermethylation of its promoter in almost 65% of hepatocellular carcinoma cases. SOCS1 deficient mice cannot survive more than three weeks of age because of enhanced IFN - [gamma] induced inflammation. To better understand the tumor suppressor role of SOCS1 in the liver we used Socs1[superscript - / -]Ifng[superscript - / -], while Ifng[superscript - / -] and wild type mice served as controls. We found that SOCS1 deficient mice showed accelerated liver regeneration following standard partial hepatectomy (PH). Moreover, Socs1 null mice were susceptible to development of hepatic nodules after treatment with diethylnitrosamine (DEN). Interestingly, the IFN - [gamma] deficient mice showed reduced number of liver tumors. In contrast to our preliminary expectations, we did not observe elevated IL - 6 serum levels in SOCS1 deficient mice compared to the controls. Nevertheless, loss of SOCS1 was associated with increased compensator y proliferation and DNA synthesis after PH and DEN treatment. To find an explanation for the increased tumorigeneis in the SOCS1 deficient liver, we examined the activation of p53 and its target genes. Although we observed neither a variable phosphorylation (Ser15), nor an impaired stabilization of p53 after cisplatin ( in vitro ) or DEN treatment ( in vivo ), Cdkn1a expression was increased in the absence of SOCS1. We also found that ectopic expression of SOCS1 could suppress the mRNA levels of p21 in HepG2 cells treated with cisplatin. In addition, we found that loss of SOCS1 increased p21 stability in hepatocytes and that SOCS1 could induce p21 ubiquitylation and subsequent proteasomal degradation. We showed that SOCS1 could bind directly to p21 via its SH2 domain. Furthermore, in SOCS1 deficient hepatocytes, p21 was retained in the cytosol in an AKT dependent fashion. While classically known as a cell cycle inhibitor, p21 can promote the assembly and kinase activity of CDK4 - cyclin D complexes. We showed that D - type cyclins and p21 levels were increased in the liver of SOCS1 deficient mice following PH. Suppression of p21 by transient shRNA transfection in SOCS1 deficient primary hepatocytes could reverse their increased proliferative response to mitogens. Finally, we found that increased p21 expression in SOCS1 deficient hepatocytes renders them resistant to apoptosis. In conclusion, our findings suggest that SOCS1 protects against liver cancer via inhibiting the oncogenic potential of p21.