A la suite d’un criblage in silico nous avons identifié un nouveau gène codant pour une protéine à domaine BTB-POZ, XBTBD6. Nous avons déterminé que la protéine XBTBD6 est une protéine cytoplasmique. Dans les cellules Hela, CHO, U2OS et COS7 la protéine XBTBD6 est localisée dans des corpuscules cytoplasmiques, localisation similaire à celle des protéines XBTBD3, HBTBD1 et HBTBD2. Nous avons observé que la partie N-terminale de la protéine, contenant le domaine BTB-POZ, est localisée dans la cellule comme la protéine entière ; par contre la partie C-terminale est exclusivement nucléaire. De plus, nous avons observé que XBTBD6 est localisée de façon diffuse dans le cytoplasme des cellules Neuro2A, 9L et 518A2e. Nous avons montré que la protéine XBTBD6 homodimérise et hétérodimérise avec XBTBD3 et XBTBD2 et qu’elle interagit avec l’ubiquitine ligase E3 XCullin 3. L’ensemble de ces interactions nécessite la présence du domaine BTB-POZ. Ces données montrent que les protéines BTBD6, BTBD3, BTBD1 et BTBD2 possèdent des propriétés communes indiquant qu’elles appartiennent à un sous groupe de la famille des protéines à domaine BTB-POZ. Le profil d’expression a été analysé par la technique de protection à la RNAse et par hybridation in situ. Les résultats montrent que ce gène est fortement exprimé dans le système nerveux adulte et embryonnaire. Des expériences de surexpression par micro-injection d’ARNm ont permis de placer le gène XBTBD6 dans la cascade d’activation des gènes proneuraux en aval de XNgnr-1, XNeuroD, Xath3 et Xebf3. Ces résultats montrent que XBTBD6 est un marqueur neuronal chez le xénope. Au cours de l’étude de la fonction du gène XBTBD6, nous avons montré que la surexpression et la perte de fonction de ce gène dans l’embryon de xénope n’induit pas de variation du nombre de neurones dans la plaque neurale. Par contre nous avons observé que la surexpression du gène XBTBD6 dans des cellules Neuro2A en différentiation régule négativement la croissance des neurites. Nous avons élaboré un modèle de fonctionnement biochimique hypothétique où la protéine XBTBD6 fonctionnerait comme protéine adaptatrice dans un complexe d’ubiquitination permettant l’ubiquitination d’une protéine cible. Nous avons recherché les partenaires potentiels de XBTBD6 en utilisant la technique du double hybride en levure mais sans y parvenir.

Selective protein degradation by the ubiquitin-proteasome system is essential for cellular homeostasis and the regulation of diverse biological processes. The selectivity of this system is imparted by hundreds of ubiquitin ligases that specifically recognise substrates and catalyse their ubiquitination, thereby targeting them for degradation. Among ubiquitin ligases, multisubunit cullin-RING ubiquitin ligases constitute the largest group. However, despite significant advances in understanding their assembly, regulation, and molecular architecture, the substrates and functions of most of them remain unknown. This thesis focuses on two ubiquitin ligases from the cullin-RING ubiquitin ligase 4 (CRL4) subfamily: CRL4DCAF4 and CRL4DCAF12 . To identify their candidate substrates and to address their biological roles, several different approaches have been employed. First, proteomic screening revealed a wide range of candidate substrates. Next, detailed characterisation of the identified interactions and exploration of the condition under which candidate substrates undergo degradation was performed. Finally, knockout human cell lines and mice with a targeted disruption of genes encoding DCAF4 and DCAF12 were generated to explore the physiological roles of CRL4DCAF4 and CRL4DCAF12 . In summary, the herein...

Marking proteins for degradation by the proteasome is a classical function of ubiquitination. This process of covalent attachment of a chain of ubiquitin molecules to target proteins is governed by the ubiquitin-activating enzyme (E1), the ubiquitin-conjugating enzyme (E2) and the ubiquitin ligase (E3). The von Hippel-Lindau (VHL) tumour suppressor protein forms an E3 ubiquitin ligase, ECV (Elongins BC/Cul2/VHL), which targets the alpha subunit of hypoxia-inducible factor (HIF) for ubiquitin-mediated destruction under normal oxygen tension. Tumour hypoxia promotes accumulation of HIFalpha, whose expression is associated with cancer progression, poor prognosis and resistance to conventional therapies, thus establishing HIF as a therapeutic target. Notably, VHL is functionally inactivated in VHL disease, a hereditary cancer syndrome characterized by the formation of tumours in multiple organs, as well as in the majority of sporadic clear-cell renal cell carcinomas (CCRCC) and haemangioblastomas. Recently, certain VHL mutations have been shown to cause the congenital disorder Chuvash polycythemia. Work contained in this thesis describes the temporally coordinated activation of the ECV, whereby oxygen-dependent recognition of HIFalpha by VHL triggers Cul2 modification by the ubiquitin-like molecule NEDD8, which enhances ECV ubiquitin ligase activity by recruiting the E2. In addition, the feasibility of ‘bio-tailored’ enzymes in the treatment of cancer is introduced by creating a bioengineered VHL capable of targeting HIFalpha for degradation irrespective of oxygen tension, which leads to the dramatic inhibition of CCRCC tumour growth and angiogenesis in a xenograft model. Furthermore, a ubiquitin ligase composed of two F-box proteins, VHL and suppressor of cytokine signalling 1 (SOCS1), was identified and shown to be paramount for the negative regulation of erythropoiesis by targeting phosphorylated Janus kinase 2 (JAK2) for ubiquitin-mediated destruction. The malfunction of this ubiquitin ligase explains the excessive erythrocytosis observed in Chuvash polycythemia patients and reveals a novel genetic link between the seemingly distinct genes VHL and JAK2 in the development of polycythemia.

Marking proteins for degradation by the proteasome is a classical function of ubiquitination. This process of covalent attachment of a chain of ubiquitin molecules to target proteins is governed by the ubiquitin-activating enzyme (E1), the ubiquitin-conjugating enzyme (E2) and the ubiquitin ligase (E3). The von Hippel-Lindau (VHL) tumour suppressor protein forms an E3 ubiquitin ligase, ECV (Elongins BC/Cul2/VHL), which targets the alpha subunit of hypoxia-inducible factor (HIF) for ubiquitin-mediated destruction under normal oxygen tension. Tumour hypoxia promotes accumulation of HIFalpha, whose expression is associated with cancer progression, poor prognosis and resistance to conventional therapies, thus establishing HIF as a therapeutic target. Notably, VHL is functionally inactivated in VHL disease, a hereditary cancer syndrome characterized by the formation of tumours in multiple organs, as well as in the majority of sporadic clear-cell renal cell carcinomas (CCRCC) and haemangioblastomas. Recently, certain VHL mutations have been shown to cause the congenital disorder Chuvash polycythemia. Work contained in this thesis describes the temporally coordinated activation of the ECV, whereby oxygen-dependent recognition of HIFalpha by VHL triggers Cul2 modification by the ubiquitin-like molecule NEDD8, which enhances ECV ubiquitin ligase activity by recruiting the E2. In addition, the feasibility of ‘bio-tailored’ enzymes in the treatment of cancer is introduced by creating a bioengineered VHL capable of targeting HIFalpha for degradation irrespective of oxygen tension, which leads to the dramatic inhibition of CCRCC tumour growth and angiogenesis in a xenograft model. Furthermore, a ubiquitin ligase composed of two F-box proteins, VHL and suppressor of cytokine signalling 1 (SOCS1), was identified and shown to be paramount for the negative regulation of erythropoiesis by targeting phosphorylated Janus kinase 2 (JAK2) for ubiquitin-mediated destruction. The malfunction of this ubiquitin ligase explains the excessive erythrocytosis observed in Chuvash polycythemia patients and reveals a novel genetic link between the seemingly distinct genes VHL and JAK2 in the development of polycythemia.

The covalent modification of proteins with a suite of molecular tags, a process termed post-translational modification, is a powerful means to enhance the proteomic complexity of an organism far beyond that which is directly encoded by its genome. A particularly widespread form of modification involves the conjugation of the protein ubiquitin to specified substrates, which serves to regulate numerous cellular processes. The mechanism of ubiquitin conjugation, known as ubiquitylation, requires E3 ubiquitin ligases that specify and recruit substrate proteins for ubiquitin conjugation. Recent insights into the mechanisms of ubiquitylation demonstrate that E3 ligases can possess active regulatory properties beyond those of a simple assembly scaffold. This thesis describes the dimeric structure of the E3 ligase adaptor protein SPOP in complex with the N-terminal domain of Cul3 at 2.4 Å resolution. Here, it is demonstrated that SPOP forms large oligomers that can form heteromeric species with the closely related paralog SPOPL. In combination, SPOP and SPOPL form a molecular rheostat that can fine-tune E3 ubiquitin ligase activity by affecting the oligomeric state of the E3 complex. These results reveal a mechanism through which adaptor protein self-assembly may provide a graded level of regulation of the SPOP/Cul3 E3 ligase toward its multiple protein substrates.

The covalent modification of proteins with a suite of molecular tags, a process termed post-translational modification, is a powerful means to enhance the proteomic complexity of an organism far beyond that which is directly encoded by its genome. A particularly widespread form of modification involves the conjugation of the protein ubiquitin to specified substrates, which serves to regulate numerous cellular processes. The mechanism of ubiquitin conjugation, known as ubiquitylation, requires E3 ubiquitin ligases that specify and recruit substrate proteins for ubiquitin conjugation. Recent insights into the mechanisms of ubiquitylation demonstrate that E3 ligases can possess active regulatory properties beyond those of a simple assembly scaffold. This thesis describes the dimeric structure of the E3 ligase adaptor protein SPOP in complex with the N-terminal domain of Cul3 at 2.4 Å resolution. Here, it is demonstrated that SPOP forms large oligomers that can form heteromeric species with the closely related paralog SPOPL. In combination, SPOP and SPOPL form a molecular rheostat that can fine-tune E3 ubiquitin ligase activity by affecting the oligomeric state of the E3 complex. These results reveal a mechanism through which adaptor protein self-assembly may provide a graded level of regulation of the SPOP/Cul3 E3 ligase toward its multiple protein substrates.

The need for anabolic therapies to increase bone formation in difficult orthopaedic circumstances and to treat osteoporosis is an area of intense research focus. There is a current interest in the Wnt signalling pathway as a target for such treatment, with accumulating evidence for a role of this pathway in bone formation. Ectoderm Neural Cortex 1 (ENC1) is a Wnt target gene, not previously studied in bone, which was observed in our laboratory to be up-regulated in an anabolic surgical model of bone formation. The involvement of ENC1 in the differentiation of neuronal and adipocytic cells has previously been reported; therefore, this thesis investigates the expression of ENC1 in cells of the bone and the role of ENC1 during osteoblast differentiation. ENC1 transcript expression was localised to osteoblastic, chondrocytic and osteocytic cells in sections of healing fracture callus and normal mouse bone by in situ hybridisation. The expression of ENC1 was confirmed in differentiating primary osteoblasts and in osteoblastic and osteosarcoma cell lines by quantitative real time PCR and western blotting. ENC1 exists as two protein isoforms of 67 and 57kD in size, which are translated from alternatively spliced ENC1 transcripts. Both isoforms of the protein were detected in differentiating cultures of the pre-osteoblast cell line MC3T3-E1. To address the function of ENC1 in osteoblast differentiation, shRNA knockdown of the endogenous transcript was undertaken in MG63 osteosarcoma cells and in the MC3T3-E1 pre-osteoblastic differentiation model. Stable expression of shRNA targeted to both ENC1 spliceforms resulted in reduced accumulation of alkaline phosphatase positive nodules and alkaline phosphatase transcripts in MG63 cell culture. This reduction was not seen with targeted knockdown of 67kD ENC1 alone. Stable tetracycline-inducible shRNA knockdown targeted to both 57 and 67kD ENC1 isoforms in MC3T3-E1 cells resulted in a significant reduction of Alizarin Red S stained mineralised nodules. When expression of 67kD ENC1 alone was reduced, however, a significant increase in MC3T3-E1 nodule formation was observed. This knockdown had no effect on the expression of early genes involved in osteoblast differentiation Runx2 and osterix, but changes in expression of alkaline phosphatase and osteocalcin mRNA mirrored nodule formation. ENC1 is a member of the BTB-Kelch family of proteins. Some members of this family have recently been found to act as substrate adaptors for the E3 ubiquitin ligase, binding to the cullin 3 component of the complex. These adaptor proteins function to bring a substrate protein within the vicinity of the E2 ubiquitin-conjugating enzyme, thus targeting it for ubiquitination and subsequent proteasomal degradation. The ability of ENC1 to interact with cullin 3 was investigated as a possible mechanism by which it may affect a role in osteoblast differentiation. Full length ENC1 showed robust binding to cullin 3 and weak binding was seen between the N-terminally truncated 57kD isoform and cullin 3. ENC1, therefore, may act as a substrate adaptor protein for the cullin 3 based E3 ubiquitin ligase. These data present ENC1 as a novel candidate protein involved in osteoblast differentiation, and suggest the possible involvement of this protein in proteasomal degradation of a substrate involved in osteoblast differentiation. The ENC1 isoforms and the associated functional pathways thus are possible future therapeutic targets to treat bone loss and enhance or accelerate fracture healing.

Thesis (Ph. D.)--University of Washington, 2006. / Vita. Includes bibliographical references (leaves 92-104).

Nierenzellkarzinome (RCC) gehören zu den häufigsten malignen Tumoren weltweit. Aufgrund der alarmierend hohen Inzidenz- und Sterberate besteht ein dringender Bedarf an neuen therapeutischen Targets zur Behandlung von RCCs. Punktmutationen in der Codesequenz von Proteinen führen zu einer Anhäufung von fehlgefalteten Proteinen in Tumorzellen und erfordern eine stärkere Kontrolle der Proteinqualität. Das Ubiquitin-Proteasome-System (UPS) bietet daher ein ideales therapeutisches Target für die RCC Therapie. Aktuelle Veröffentlichungen deuten auf eine Deregulation des COP9 Signalosome (CSN)-Cullin-RING-Ubiquitin-Ligase-(CRL)-Signalweges hin, einem Bestandteil des UPS. In der vorliegenden Arbeit wurden ausgewählte Komponenten des CSN-CRL Signalweges im RCC Gewebe und in vier RCC Zelllinien untersucht. In immunohistochemischen Studien am klarzelligen RCC-Gewebe konnte keine Hochregulierung einer einzelnen CSN-Untereinheit gezeigt werden. Höchstwahrscheinlich ist der gesamte CSN-Komplex im klarzelligen Nierenkarzinom im Vergleich zu nicht-malignem Nierengewebe stärker exprimiert. Die Untersuchung von vier RCC-Zelllinien zeigte eine interessante Deregulierung der CAND1-Skp2-p27 Achse in einer der Zelllinien. 786-O Zellen wiesen zwei p27Kip1 (p27) Varianten (p27V109G und p27I119T), eine Erhöhung des Skp2 und eine Reduktion des CAND1 Levels auf. Die Expression und Lokalisation von CAND1 wurde weiter in einer größeren RCC-Kohorte untersucht. Dabei zeigte sich eine negative Korrelation zwischen einer hohen zytosolischen CAND1 Expression und dem Gesamtüberleben von Patienten mit klarzelligen renalen Tumoren. Beide p27 Varianten werden durch das UPS abgebaut und binden an das CSN, Skp2, Cdks sowie an Cyclin E. Interessanterweise zeigte die p27 Mutanten beinhaltende Zelllinie 786-O eine höhere Proliferationsrate als die p27-Wildtyp-Zelllinie A498. In einem im Rahmen dieser Arbeit entwickelten Genotypisierungs-Assay konnte eine große RCC-Kohorte nach den beiden p27-Mutanten untersucht werden. In 42,5% der RCC Patienten konnte die Mutante p27V109G heterozygot nachgewiesen werden. Die Präsenz der beiden Mutanten p27V109 und p27I119T im RCC-Gewebe sowie die veränderte Expression von Skp2 und CAND1 machen den CSN-CRL Signalweg zu einem attraktiven therapeutischen Target für die Behandlung von Patienten mit Nierenzellkarzinom. / Renal cell carcinomas (RCC) belong to the most common malignant tumors worldwide. Alarming high incidence and mortality rates elucidate the urgent need for new therapeutic targets in RCCs. Point mutations in protein coding sequences lead to numerous unfolded proteins in cancer cells, requiring effective protein quality control. Therefore, components of the ubiquitin proteasome system (UPS) might be a promising new approach for RCC therapy. Recent publications in renal cancers point to a deregulated COP9 signalosome (CSN)-Cullin-RING Ubiquitin Ligase (CRL) pathway, a segment of the UPS. In the present thesis, selected components of the CSN-CRL pathway were studied in RCC tissues and four RCC cell lines. Immunohistochemistry results did not show an overexpression of a single CSN subunit in clear cell RCC tissues (ccRCC). However, it seems that the CSN holo complex is upregulated in analyzed ccRCCs. Examination of four RCC cell lines revealed a deregulation of the CAND1-Skp2-p27 axis in 786-O cells. These cells harbor two p27Kip1 (p27) mutants (p27V109G and p27I119T), high Skp2 and decreased CAND1 levels. Expression and localization of CAND1 was studied in a larger cohort of RCC tissues and revealed high cytosolic levels of CAND1 to be negatively correlated with overall survival in ccRCC patients. Both p27 variants were found to be degraded by the UPS and bound to the CSN, Skp2, Cdks and cyclin E. Interestingly, 786-O cells appear to grow 3-fold faster than A496 cells expressing p27wt. Further, a large cohort of RCC was screened for both p27 variants using a genotyping assay, specifically designed within the present thesis. 42.5% of the RCC patients harbor p27V109G heterozygously. The occurrence of p27V109G and p27I119T in RCC tissues as well as changed expression of Skp2 and CAND1 make the CSN-CRL pathway an attractive therapeutic target for the treatment of patients with RCC.

L’ubiquitination est une modification post-traductionnelle impliquée dans de nombreux mécanismes cellulaires. L’E3-ubiquitine ligase CULLIN 3 (CUL3) est un régulateur essentiel de la progression mitotique, ubiquitinant d’importants régulateurs mitotiques et contrôlant leur localisation subcellulaire. Plus particulièrement, notre travail décrit le rôle de la nouvelle E31 ligase CUL3/KLHL22 dans la régulation de l’activité localisée de Polo-like kinase 1 (PLK1) et de ce fait dans l’établissement d’une progression mitotique précise. Néanmoins, les mécanismes moléculaires qui régissent la reconnaissance de son substrat par CUL3 demeurent inconnus. L’activité catalytique de PLK1 ne semble pas être nécessaire à son interaction avec KLHL22, mais aussi bien son domaine kinase que Polo-box (PBD) suffisent à co-purifier KLHL22. Des mutations au niveau du motif DFG, situé en amont du domaine kinase,et du tryptophane 414 au sein du PBD semblent influer sur la reconnaissance de KLHL22. Les résultats obtenus montrent les premières indications biochimiques du mode d’interaction du complexe CUL3/KLHL22/PLK1. / Ubiquitination is a post-translational modification involved in many cellular processes. The E3 ubiquitin-ligase based on CULLIN 3 protein (CUL3) is an essential regulator of mitotic division in human cells by ubiquitinating several important mitotic regulators and controlling their subcellular localization. In particular, our work described the role of novel CUL3/KLHL22 E3-ligase in regulation of localized activity of Polo-like kinase 1 (PLK1) and there by faithful mitotic progression. However, the molecular mechanisms of substrate recognition by CUL3 remain unknown. The catalytic activity of PLK1 may not be required for binding KLHL22 but both the kinase and the Polo-box domains are sufficient to co-purify KLHL22. Mutating the DFG motif within the kinase domain and the tryptophan 414 within the PBD influence the binding to KLHL22. These results provide first insights into molecular mechanisms of CUL3/KLHL22/PLK1complex.