Dissertations / Theses on the topic 'UBE2J1'

The ubiquitin proteasome system (UPS) is responsible for the degradation and turnover of proteins in eukaryotes. As such it is a key process that is involved in normal and in some cases, abnormal cellular functions. Ubiquitin conjugating enzymes (UBCs) are key components of the UPS and may serve as therapeutic targets. The aims of this project are the structural and functional analyses of UBCs in eukaryotic organisms whose genomes have been fully sequenced, and also the review of the nomenclature of yeast and human UBCs. The main findings were: 1) The successful construction of Phylogenetic trees containing all 14 yeast UBCs and UBC-like proteins and their homologues in selected species whose genomes have been fully sequenced. The phylogenetic tree consists of 15 different branches. Thirteen of the branches contain a member of the yeast UBC or UBC-like family (MMS2) and their homologues. The remaining two branches contain firstly the human UBE2L proteins and secondly the TSG101 UBC like proteins, both of which branches do not appear to have any yeast orthologues. 2) The UBC phylogenetic study was also used to identify the previously unknown UBC active site of the <i>Drosophila </i>transcription factor, TAF_II250. Using multiple sequence manual alignments of known TAF_II250 and UBC protein sequences a putative UBC active site in <i>Drosophila melanosaster</i> and <i>Apis mellifera</i> TAF_II250 was successfully identified, and was shown to share approximately 70% homology to the known UBC PROSITE signature. 3) Using multiple sequence alignments the hitherto unknown PROSITE signature of the clinically important UBE2J family was identified. This PROSITE signature is very different from all other UBCs suggesting that this family of enzymes has significant structural changes at their active sites. Homology modelling proved to be a successful approach to obtain structural information of the UBE2J1 active site. Superimposition studies using the previously solved structures for human UBE2J2 and human UBC9 were carried out. Significant differences were observed near the active sites of human UBE2J proteins compared to the active site for human UBC9.

Made available in DSpace on 2016-06-02T20:21:38Z (GMT). No. of bitstreams: 1 DissLFR.pdf: 1075822 bytes, checksum: f9dc443682a3269a2f32e2a5579089a0 (MD5) Previous issue date: 2005-06-10<br>Financiadora de Estudos e Projetos<br>The ubiquitin system represents a selective mechanism for intracellular proteolysis in eukaryotic cells that involves the sequential activity of three enzymes, E1 (Ubiquitin activating enzyme), E2 (Ubiquitin-conjugating enzyme), and E3 (Ubiquitin-protein ligase). The identification of these proteins and their targets as well as structural data is essential to understand their function in the eukaryotic cell. In the present study the open reading frame of human Ubiquitin- conjugating enzyme UBE2G2 was isolated from a human brain cDNA panel, cloned into pET28 vector and expressed in Escherichia coli. His-tagged protein was then purified by nickel-affinity chromatography and subjected to structural and functional studies using circular dichroism (CD) and an in vitro ubiquitin-binding assay, respectively. The affinity chromatography assay rendered approximately the 27 mg of the soluble recombinant HisUBE2G2 after expressed in bacteria at low amounts of IPTG (0,1mM) in 3 hours of induction. The CD spectra of recombinant pure protein showed a secondary structure content according with the expected for a member of the E2 family (Ubiquitin-conjugating enzyme), with 35 % of &#945;-Helix, 21 % of &#946; sheets and 23 % of turns. Moreover, purified protein was able to bind ubiquitin molecules when mixed with a HeLa cell extract during the pull-down assay. Taken together the results presented in this work allow inferring that HisUBE2G2 was expressed in their active form.<br>O Sistema de Ubiquitinação representa o mecanismo de degradação protéica intracelular mais importantes em todos nas células eucarióticas e envolve a atividade seqüencial de três enzimas conhecidas como E1, enzima ativadora de ubiquitina, a E2 enzima conjugante de ubiquitina e a E3 enzima ligante de ubiquitina. A identificação destas proteínas e seus alvos protéicos, assim como as obtenções de dados estruturais são essenciais para entender a função deste sistema dentro da célula eucariótica. No presente trabalho, a fase de leitura aberta (ORF) do gene humano ube2g2 foi isolado de um painel de cDNA de cérebro humano, foi clonado no vetor pET28a e expressado em bactérias Escherichia coli. A proteína de 18,5 kDa em fusão com uma cauda de histidinas foi posteriormente purificada por cromatografia de afinidade e submetido a ensaios estruturais e funcionais a traves do medição do CD e a traves do ensaio de pull-down, respectivamente. A cromatografia de afinidade rendeu 27 mg de proteína solúvel, logo após de tê-la expressado heterologamente a baixas concentrações de indutor IPTG 0,1 mM em três horas de indução. O espectro de CD da proteína purificada mostrou o conteúdo de estrutura secundaria de acordo ao esperado para um membro típico da família de enzimas E2, apresentando um 35 % de hélices &#945;, 21 % de folhas &#946; e 23 % de giros. Alem do mais, a proteína purificada foi capaz de ligar molécula de ubiquitina quando misturada com um extrato de células HeLa, durante o ensaio de pull-down. Desta maneira e com esses resultados apresentados aqui pode se inferir que a proteína humana UBE2G2 foi expressa na sua forma ativa.

In most eukaryotic organisms, the ubiquitination pathway is one of the most important and versatile signaling systems in use. It is integral to processes such as protein degradation and homeostasis, DNA repair cell cycle regulation, signaling and regulation, epigenetics, and many more. Ubiquitin (Ub) is a short polypeptide of 8.6 kDa, 76 residues that functions as a reversible post-translation modification (PTM). It furthermore contains 7 different lysine residues (K6, K11, K27, K29, K33, K48, K63), all of which can form isopeptide linkages with one another to link individual Ub moieties to form unique polyUb chains onto substrates. The type of polyUb chain a substrate gets labeled with can determine the subsequent activity of that substrate. Substrate ubiquitination is achieved through an enzymatic cascade. First, an E1-activating enzyme activates a free Ub moiety. Then Ub is transferred onto an E2-conjugating enzyme, and finally an E3 ligase interacts with both substrate and E2~Ub complex to facilitate Ub transfer onto a substrate. Within this scheme, the E2-enzyme acts as a master manipulator in that, it controls when, where and how a ubiquitin chain is transferred onto a substrate. Irregular activity of E2-conjugating enzyme has been implicated in a wide variety of diseases such as cancer, neurodegenerative diseases, muscular dystrophy, genetic azoospermia and more. While attempts have been made to inhibit other ubiquitination cascade enzymes such as E3 ligases and E1-activating enzymes, there is a strikingly small number of inhibitors specifically targeting E2 enzymes mainly due to the high degree of structural conservation that exists among members of the E2 enzyme family. In this work, we introduce 3 novel linked-domain protein inhibitors of the E2-conjugating enzyme Ube2D. We covalently attached either UHRF1 RING domain or an affinity optimized U-box domain, with UHRF1 UBL domain or UbvD1.1 (A ubiquitin variant specific for Ube2D), through a glycine-serine linker, producing 3 unique inhibitors: Ring-UBL (RU), U-box-UBL (UU), and U-box-UbvD1.1 (UUD1.1). In this way, we attempt to specifically inhibit Ube2D for two purposes : 1) While Ube2D can interact with the largest number of E3 ligases and facilitate the largest number of polyUb chains, very little is known about cellular phenotypes specifically associated with Ube2D; 2) We want to establish whether targeting the E2 enzyme in general can be utilized as a viable therapeutic treatment for cancer. We show that all three inhibitors are able to inhibit ubiquitin assays using Ube2D and using ITC we measured binding affinities of UUD1.1 (5 nM) > UUWT (300 nM) > RU WT (3 µM). Furthermore, we found that all inhibitors could prevent E1, E3 and backside binding domain interactions simultaneously, which single domain UBL could not. UU and RU showed specificity towards Ube2D when tested against APC/C and Cullin1 E3 ligases and their cognate E2 enzymes. We propose that linking domains in this way, by targeting the backside binding domains of E2 enzymes, could be a strategy that can be standardized and applied to the rest of the E2 enzyme family as well. In vivo testing must now elucidate whether these inhibitors can provide more information about the cellular role of Ube2D and whether it is a viable therapeutic target to treat cancer.

Les ciliopathies sont des maladies multisystémiques dues à des mutations dans des gènes codant des protéines localisées au cil. Les cils sont des organites constitués de microtubules présents à la surface de quasi toutes les cellules de vertébrés. IFT52 est un composant majeur du complexe du transport intraflagellaire IFT-B, indispensable à la formation et fonction du cil primaire. Des mutations faux-sens, non-sens ou de délétion en phase ont été identifiées dans IFT52 chez des individus présentant des phénotypes distincts de ciliopathies : une dysplasie rénale multikystique (DRK) ou des dysplasies squelettiques de type côtes courtes-polydactylie (CCP) ou Sensenbrenner (SB). Afin de mieux comprendre la variabilité phénotypique, nous avons étudié la pathogénicité de chacune des mutations grâce à des modèles cellulaires et animaux. D'abord, j'ai montré que les mutations de délétion en phase (CCP) et non-sens (SB), mènent en fait respectivement à un décalage de phase de lecture et un saut d'exon en phase, ce qui permettait d'expliquer en partie les relations génotype/phénotype. Par ailleurs, l'analyse des mutations faux-sens CCP et DRK montrait que la mutation CCP avait un effet plus délétère sur la fonction de IFT52 que celle de DRK, confirmant leur pathogénicité et expliquant le phénotype squelettique du cas CCP. Toutefois, cela n'expliquait pas la variabilité du phénotype rénal observée parmi les individus. De fait, nous avons considéré une deuxième mutation homozygote faux-sens présente dans le cas DRK dans le gène UBE2C, codant une enzyme E2 de conjugaison d'ubiquitine impliquée dans la sortie de mitose. La modélisation de la structure 3D de la protéine a montré que la mutation affectait la liaison à l'ubiquitine et de fait, probablement la fonction de la protéine. Des études in vitro ont confirmé que la mutation ralentissait la sortie de mitose des cellules. Par ailleurs, afin d'étudier l'impact de la mutation sur le développement rénal, une souris portant la mutation DRK (knock-in) a été générée par CRISPR/Cas9. Bien que les souris Ube2cKI/KI présentent un défaut de croissance et une létalité précoce similaire au modèle knock-out, validant ainsi la pathogénicité de la mutation, les reins sont normaux chez ces animaux. Nous avons donc émis l'hypothèse que le phénotype rénal résultait de la synergie entre les mutations IFT52 et UBE2C. Afin de tester cette hypothèse, des études de synergie ont été réalisées chez le poisson zèbre, par injection de morpholino ift52 dans des embryons issus du croisement d'animaux ube2c+/-. Cette approche n'a malheureusement pas permis d'induire un défaut rénal chez les embryons ube2c-/-. Ainsi soit le cas DRK présente une autre mutation pouvant expliquer le phénotype rénal, soit les modèles animaux (souris, poisson zèbre) ne sont pas appropriés à l'étude de l'implication de UBE2C dans le développement rénal humain. De manière inattendue, l'étude de IFT52 a permis de mettre en évidence un espacement anormal des centrioles composant le centrosome des cellules Ift52-/-. La cohésion du centrosome est maintenue par deux mécanismes : le lien protéique entre les parties proximales des centrioles, et la force des microtubules (MTs) qui s'exerce sur le centrosome. Aucune altération des principaux composants du lien protéique, notamment c-Nap1 et rootletine, n'a pu être observée dans les cellules Ift52-/-. En revanche, nous avons montré que Ift52 interagissait et co-localisait partiellement avec centrine au niveau de la partie distale du centrosome, suggérant un rôle de IFT52 dans le mécanisme des MTs. En effet, nous avons montré que les MTs présentaient un défaut d'ancrage au centrosome et qu'ils étaient moins dynamiques dans les cellules Ift52-/-. Nos résultats démontrent un rôle extra-ciliaire de IFT52 dans l'ancrage des MTs et la cohésion du centrosome, à la manière de son partenaire IFT88 au niveau des MTs mitotiques, ajoutant un autre mécanisme physiopathologique aux ciliopathies associées à IFT52<br>Ciliopathies are multisystemic disorders due to mutations in genes encoding proteins localizing at the cilium. Cilia are microtubule-based organelles present at the surface of almost all vertebrate cells. IFT52 is a key component of the intraflagellar transport IFT-B complex that guarantees cilium formation and functions. Missense, nonsense mutations and in-frame deletions have been identified in IFT52 in three families presenting distinct phenotypes of ciliopathies: multicystic kidney dysplasia (MCKD) or skeletal dysplasia such as short ribs-polydactyly (SRP) or Sensenbrenner (SB). To understand the phenotype variability, we sought to characterize the pathogenicity of the mutations by using patient fibroblasts and CRISPR/Cas9-induced cellular and animal models. First, we showed that the nature of the in-frame (SRP) and nonsense mutations (SB) actually led to a frameshift and an in- frame exon skipping, respectively that partially explained the genotype/phenotype correlation. On the contrary, analyses of missense mutations from SRP and MCDK cases indicated that SRP mutation had a more severe impact on IFT52 function than the DRK mutation, confirming the pathogenicity of the mutations and explaining the skeletal defects of SRP case. However, it did not explain the renal phenotype variability observed amongst the individuals. Thus, we considered a second homozygous missense mutation in the MCDK case in UBE2C, a gene encoding an ubiquitin conjugating enzyme involved in metaphase/anaphase transition. Tridimensionnal modelisation of the protein structure showed that the mutation affected the binding with ubiquitin and likely the function of the protein. In vitro studies confirmed that the mutation delayed mitosis exit. In order to study the role of Ube2c in kidney development, a knock-in (KI) mouse line of the patient mutation was generated by CRISPR/Cas9 technique. Although Ube2cKI/KI failed to thrive and died early, similarly to the knock-out line, kidneys appeared normal in these animals. So we hypothesized thet the renal phenotype of the MCDK case could come from a synergy between the mutations in IFT52 and UBE2C. To test this hypothesis, we performed synergy experiments by injecting ift52 morpholino in zebrafish embryos issued from ube2c+/- incrosses, but we did not observe kideny defects in ube2c-/- embryos. Thus, we conclude that either the MCDK individual present another mutation responsible for the renal phenotype or the animal models we used (mouse, zebrafish) were not appropriate to study the involvement of UBE2C in human kidney development. Surprisingly, the study on IFT52 mutations highlighted an abnormal splitting of the centrioles of the same centrosome in Ift52-/- cells. Centrosome cohesion is guaranteed by two mechanisms: a protein linker between proximal parts of the centrioles and the microtubule (MTs) forces exerted on the centrosome. The proteins from the protein linker, such as c-Nap1 and Rootletin, did not seem to be affected by the loss of Ift52. However, we showed that Ift52 interacted and partially co-localized with centrin at the distal part of the centriole, suggesting a role for Ift52 in the MTs mechanism. Indeed, we showed that in Ift52-/- cells, MTs presented failed to anchor at the centrosome and their dynamic was reduced. These defects could be the cause of the centrosome splitting and suggest an extra-ciliary role for IFT52 in MTs anchoring and centrosome cohesion, as its partner IFT88 in mitosis, adding another physiopathological mechanism to the IFT52-associated ciliopathies

Em trabalho anterior, identificamos a mutação c.382C8594;T no gene UBE2A, localizado em Xq24 e codificador de enzima conjugadora de ubiquitina, como causa de nova síndrome de deficiência mental (DM) de herança ligada ao cromossomo X. Foi a primeira descrição de mutação nesse gene e a primeira associação de mutação em gene que codifica conjugase de ubiquitina com patologia humana. Neste trabalho, focalizamos o gene UBE2A quanto a expressão dos transcritos alternativos, função das isoformas por eles codificadas e o efeito da mutação c.382C 8594; T como causa de deficiência mental (DM). No Capítulo I, revisamos os aspectos genéticos da DM, dando ênfase à herança ligada ao cromossomo X, principal causa de DM herdada e resumimos o estudo que levou à identificação da mutação em UBE2A como causa de quadro sindrômico de DM. Revisamos o papel da via de ubiquitinação de proteínas e das enzimas que participam do processo, em especial as conjugases de ubiquitina. Levantamos evidências na literatura que não deixam dúvida sobre a importância da via de ubiquitinação no sistema nervoso, tanto em processos de neurodesenvolvimento como neurodegeneração. No Capítulo II, avaliamos a contribuição de mutações em UBE2A como causa de DM. Apresentamos os resultados do sequenciamento direto da região codificadora do gene UBE2A em afetados de 23 famílias em que a DM segrega ligada à segmento que inclui Xq24, onde está localizado o gene UBE2A. Uma dessas famílias foi averiguada no Serviço de Aconselhamento Genético do Laboratório de Genética Humana do Departamento de Genética e Biologia Evolutiva, Instituto de Biociência, USP (LGH-IB/USP), coordenado pelo Dr. Paulo A. Otto e pela Dra. Angela Vianna Morgante. As demais 22 famílias pertencem ao banco de amostras do Consórcio Europeu de Deficiência Mental (European Mental Retardation Consortium EURO-MRX). A triagem foi também realizada em um indivíduo afetado por DM sindrômica que compartilha características clínicas com nossos pacientes. Como acontece com a maioria dos genes do cromossomo X, o gene UBE2A não parece ser responsável por parcela significativa dos casos de DM, já que novas mutações em UBE2A não foram detectadas nessa triagem. Avaliamos o efeito da mutação c.382C 8594; T nos níveis da transcrição e da tradução em homem afetado e em mulher portadora. A presença da mutação que leva a um códon de parada prematura não resultou na degradação do RNA, que detectamos nas células do afetado. Já na mulher portadora, apenas o transcrito normal foi detectado, de acordo com nossos dados anteriores que mostraram desvio completo no padrão de inativação do cromossomo X nas portadoras da mutação, tendo um mesmo cromossomo X ativo nas células do sangue. A vantagem proliferativa das células em que o cromossomo X com alelo mutado estava inativo deve ter levado a esse padrão de inativação desviado do casual, evidenciando o efeito deletério da mutação. Entretanto, o mecanismo pelo qual a mutação afeta a via de UBE2A permanece interrogado. A proteína UBE2A alterada foi encontrada em baixa quantidade nas células do paciente, o que pode ser o resultado de síntese prejudicada ou de degradação pós-traducão. Independente do mecanismo responsável, o fato de apenas uma pequena quantidade da proteína mutada ter sido encontrada, nos permite afirmar que, nas células desses indivíduos, há perda de função de UBE2A. Devemos, contudo, considerar que a proteína mutada é sintetizada e que, no caso de a menor quantidade dever-se à degradação pós-tradução, esse processo pode prejudicar a homeostase celular e contribuir para o quadro clínico. O capítulo II focaliza os transcritos alternativos de UBE2A. Diversos bancos de dados apontam para a existência de três transcritos alternativos do gene UBE2A humano, mas não há trabalho científico que caracterize os tecidos em que os transcritos são expressos ou a função das proteínas por eles codificadas. A mutação c.382C 8594; T localiza-se no éxon 6 do gene, comum a todos os transcritos, de forma que, no caso de eles codificarem proteínas funcionais, a mutação comprometeria três proteínas, e não apenas uma. Demonstramos que os três transcritos de UBE2A são xpressos em leucócitos, pré-adipócitos, placenta, córtex cerebral e hipocampo humanos. Detectamos também os três transcritos nas células de sangue e pré-adipócitos de um de nossos pacientes portador da mutação c.382C 8594; T. Embora os bancos de dados apontem para a existência de apenas um transcrito de UBE2A em camundongos, identificamos um transcrito alternativo correspondente ao transcrito alternativo 3 humano. Este foi detectado inclusive em camundongos nocaute quanto ao gene UBE2A o processo de geração do animal nocaute foi realizado por recombinação homóloga em que o cassete de neomicina foi inserido no éxon 1 do gene, de maneira que não eliminou a existência do transcrito correspondente ao transcrito 3 humano, que utiliza uma 5 UTR alternativa localizada no íntron 3. Entretanto, as proteínas codificadas pelos transcritos alternativos não foram detectadas nos extratos protéicos analisados humanos e de camundongo. Esse resultado poderia ser explicado pela falta de especificidade do anticorpo utilizado ou por essas isoformas representarem pequena parcela do pool de proteínas da célula. Os anticorpos comerciais anti-RAD6 e anti-HR6A/HR6B foram produzidos após imunização de coelhos com a porção N-terminal da isoforma 1. Seria, portanto, possível que não fossem capazes de detectar as isoformas 2 e 3, em que o segmento utilizado para a produção dos anticorpos está total ou parcialmente ausente. No caso de as proteínas estarem pouco representadas na célula, experimentos de co-imunoprecipitação auxiliariam na identificação dessas isoformas nos extratos protéicos. Entretanto, para a detecção de todas as isoformas de UBE2A seria necessário anticorpo que reconhecesse a porção C-terminal de UBE2A. Em 2009, duas novas mutações em UBE2A foram descritas em estudo colaborativo realizado no Welcome Trust Sanger Institute, Hinxton, Cambridge, Reino Unido, após sequenciamento em larga escala de aproximadamente 700 genes do cromossomo X de cerca de 200 indivíduos com DM de herança ligada ao X. Ambas as mutações c.215C 8594; T e c.328C 8594; G eram do tipo missense. Não foram fornecidas informações quanto ao quadro clínico dos portadores dessas mutações e também não foi esclarecido porque apenas a alteração c.215C 8594; T que resulta na troca do resíduo de fenilalanina da posição 72 por um resíduo de serina (F72S) foi considerada pelos autores como possivelmente patogênica. Nossos estudos in vitro, apresentados no Capítulo III, sugerem que ambas as alterações afetam a função de UBE2A. Em 2010, foram publicados dois trabalhos associando novas alterações em UBE2A a quadro de DM. Honda e col. (2010) descreveram uma microdeleção em Xq24 que inclui UBE2A e outros oito genes, em um menino com DM e características também presentes em nossos pacientes. Budny e col. (2010) descreveram mutações missense em UBE2A em duas famílias em que segregava quadro de DM sindrômica semelhante ao de nossos pacientes. Por comunicação pessoal de Arjan de Brouwer (Departamento de Genética Humana da Universidade Radboud, Nijmegen, Holanda), soubemos da existência de três outras microdeleções de segmentos do cromossomo X que incluem UBE2A, em pacientes do sexo masculino, não aparentados. Comparamos as características clínicas de nossos pacientes com as dos portadores das microdeleções em Xq24 e com aquelas dos portadores de mutações missense em UBE2A. A DM grave e o comprometimento significativo ou ausência de fala são comuns a todos. Outras características como baixa estatura, sinófris, boca grande e lábios finos com comissuras voltadas para baixo, pescoço curto e largo, implantação baixa de cabelos na nuca, mamilos espaçados, pênis pequeno, hirsutismo generalizado e a ocorrência de convulsões parecem predominar. Entretanto, enquanto a microcefalia aparece em dois dos três portadores de microdeleções avaliados, a macrocefalia parece predominar no grupo em que ocorrem as mutações de ponto. No Capítulo III, abordamos estudos funcionais in vivo e in vitro para avaliar se as isoformas alternativas de UBE2A compartilham suas funções de conjugase de ubiquitina e compreender o efeito da mutação c.382C8594;T na função de UBE2A. Buscamos estabelecer modelo celular para avaliar o efeito da mutação na formação de neuritos. Trabalho previamente publicado havia demonstrado que a diferenciação neuronal de células PC12 concomitantemente com a inibição parcial do mRNA de UBE2B (parálogo de UBE2A) resultava na redução de 20-30% do comprimento de neuritos. Entretanto, nossos ensaios de diferenciação de pré-adipócitos não responderam nossas questões sobre o efeito da mutação na formação de neuritos, pois não conseguimos obter, nas células do controle ou nas do paciente, a densidade de neuritos descrita anteriormente na diferenciação de pré-adipócitos. Diferentemente das células PC12, de origem ectodérmica, os pré-adipócitos tem origem mesodérmica, o que dificulta sua diferenciação em linhagem derivada de outro folheto germinativo. A elevada conservação entre as proteínas ortólogas UBE2A e UBE2B humanas e RAD6 de levedura e a observação de que ambas as parálogas humanas são capazes de complementar os fenótipos apresentados pela linhagem 916;rad6 de Saccharomyces cerevisiae nos levou a considerar a linhagem de levedura 916;rad6 como modelo para nossos estudos funcionais. Também, avaliamos a capacidade das isoformas 2 e 3 e da isoforma Q128X de UBE2A para ubiquitinar histonas H2A in vitro, conforme previamente descrito para a isoforma UBE2A/1. Os resultados dos ensaios in vivo indicam que apenas a expressão do transcito 1 de UBE2A é capaz de complementar os fenótipos apresentados pela linhagem 916;rad6 de S. cerevisiae. A expressão dos transcritos 2 ou 3 não resulta na restituição do fenótipo de sensibilidade à UV - a expressão gera certa toxicidade, agravada quando as células são cultivadas a 37o C. Entretanto, as isoformas por eles codificadas não parecem ser estáveis na levedura: assim como nos tecidos humanos testados, não conseguimos detectá-las nos extratos protéicos das leveduras que expressavam esses transcritos. A expressão do transcrito 1 contendo a mutação c.382CT revelou que a isoforma UBE2A/Q128X, por sua vez, é estável na linhagem 916;rad6, porém, além de não restituir o fenótipo de sensibilidade à UV, foi, dentre as isoformas de UBE2A, a mais tóxica. Os fenótipos de toxicidade não foram observados após expressão em linhagem selvagem de S. cerevisiae. Esses resultados indicam que as isoformas 2 e 3 de UBE2A não apresentam atividade de conjugase de ubiquitina e que são, aparentemente, degradadas imediatamente após sua expressão em levedura. O fato de o fenótipo de toxicidade ser agravado, em condições de choque térmico, apóia a hipótese de degradação dessas isoformas, em levedura. A degradação pode ser resultado da ausência de parceiro que permita sua estabilidade, mas a ausência das isoformas também em extratos protéicos de tecidos humanos sugere que o mesmo processo de degradação ocorra em mamíferos. Segundo a classificação das E2, as diversas conjugases de ubiquitina têm em comum o domínio UBC altamente conservado e as variações observadas consistem em inserções ou extensões C-terminais, mas nunca deleções, como ocorre nas isoformas 2 e 3 de UBE2A. Os transcritos alternativos teriam, assim, função regulatória. Os ensaios in vitro confirmaram a capacidade de UBE2A/1 ubiquitinar histonas H2A. Os ensaios com UBE2A/2 e UBE2A/3 não foram conclusivos, uma vez que a incapacidade de ubiquitinação de histonas que observamos pode ter consequência da renaturação in vitro, que pode ter ocorrido prejudicando sua função. Entretanto a obtenção das isoformas puras nos permitiu verificar que, caso a isoforma 2 estivesse presente nos extratos de levedura, ela seria reconhecida pelo anticorpo anti-RAD6. Verificamos que a proteína mutada UBE2A/Q128X é capaz de interagir com a E1, da qual recebe a molécula de ubiquitina, mas não é capaz de transferi-la para a histona. O segmento C-terminal ausente nessa isoforma é, portanto, importante nesse processo. Os ensaios in vitro despertaram nossa atenção para o fenômeno de autoubiquitinação de UBE2A, possível mecanismo de autorregulação previamente considerado na literatura. O fato de alguns trabalhos sugerirem que as E2 atuem também como dímeros in vivo e in vitro e a elevada conservação entre as parálogas humanas UBE2A e UBE2B nos levaram a considerar a possibilidade de mecanismo de regulação recíproca. Dessa maneira, a degradação de UBE2A/Q128X nas células do paciente poderia ser dependente de UBE2B. A reduzida capacidade de autoubiquitinação da isoforma mutada dificultaria sua degradação e tornaria necessária a atividade da paráloga. Isso explicaria porque ela é estável quando expressa na linhagem de levedura 916;rad6, mas não nas células do paciente. A presença de RAD6 estaria diretamente relacionada à ausência de toxicidade após a expressão de UBE2A/Q128X em linhagem selvagem a degradação da isoforma mutada está ocorrendo nessas células. A não viabilidade de camundongo duplo-nocaute quanto as parálogas UBE2A e UBE2B não permite testar a estabilidade da isoforma mutada em células de mamíferos. Observamos, de fato, que a inibição do proteassoma nas células do paciente leva ao acúmulo dessa proteína. A presença da mutação c.382C8594;T nas células do paciente parece resultar no fenótipo de DM devido à perda de função de UBE2A: a isoforma mutada não restitui o fenótipo de sensibilidade à UV de S. cerevisiae e não foi capaz de ubiquitinar histonas H2A in vitro. Além disso, indivíduos com microdeleções de UBE2A apresentam fenótipo semelhante ao de nossos pacientes. Por outro lado, a presença da proteína mutada que necessitaria de UBE2B para ser degradada pode caracterizar um ganho tóxico de função - comprometeria a função de ambas as parálogas. É possível que os dois mecanismos contribuam para o quadro clínico. Os dados dos ensaios in vivo e in vitro abrem caminhos de investigação do processo de regulação de UBE2A e UBE2B no nível da proteína, sugerindo a autoubiquitinação e a ubiquitinaão recíproca como possíveis mecanismos reguladores, que podem explicar a conservação das duas parálogas de RAD6 em mamíferos.<br>We have previously described a nonsense mutation (c.382C8594;T) in the UBE2A gene, at Xq24, which encodes a ubiquitin conjugating enzyme (E2), as the cause of a new X-linked mental retardation syndrome. The predicted protein lacks the 25 C-terminal amino acid residues conserved in vertebrates and in Drosophila. This was the first description of a mutation in a ubiquitin conjugating enzyme gene causative of a human disease. In the present work, we focused on the UBE2A gene, its alternative transcripts and isoforms, and the effect of the c.382C8594;T mutation. We screened for UBE2A mutations 23 males presenting X-linked mental retardation (XLMR), previously mapped to the interval encompassing this gene, and one isolated case, who shared clinical features with our previously described patients. No mutations were detected in this selected series of patients suggesting that mutations in UBE2A is not a common cause of XLMR, similarly to the majority of the XLMR genes hereto described. Very recently four Xq24 microdeletions encompassing UBE2A and three missence mutations were found by other groups in mentally retarded males that shared several clinical features with our patients. Comparing these and our patients, a clinical picture emerges of mental retardation associated with severe speech impairment, present in all of them. Short stature, large mouth with downturned corners and thin lips, short and broad neck, low posterior hairline, widely spaced nipples, marked generalized hirsutism and seizures are common features. However, microcephaly was observed only in patients carrying UBE2A deletions, while carriers of missense or nonsense mutations showed macrocephaly. We evaluated the effect of the UBE2A c.382C8594;T mutation on transcription and translation. This mutation affects the last UBE2A exon and, as expected, does not lead to nonsense mediated RNA decay, demonstrated by the presence of UBE2A mRNA in leucocytes of an affected male. However, only a small amount of the mutated protein was detected in the patients cells, suggesting the loss of UBE2A function as the cause of the syndrome. The posttranslational degradation of the mutated protein could also disturb the cellular homeostasis, a gain of function that remained a possibility. The detrimental effect of the c.382C8594;T mutation was further supported by the presence of only the normal transcript in leucocytes of a heterozygous woman, who had completely skewed X inactivation, thus pointing to the selective advantage of lymphocytes carrying the normal allele on the active X chromosome. Our search in DNA and protein sequence databases suggested that the UBE2A gene produces three alternative transcripts all classified as protein coding. These three tanscripts contain the mutation site (c.382C8594;T). We showed that all three UBE2A transcripts are expressed in human leucocytes, adipocytes, placenta, cerebral cortex and hippocampus. We also detected an alternative transcript in murine, which corresponds to the human transcript 3. This alternative transcript was present in all murine tissues analyzed, including samples from a UBE2A knockout mouse. However, we failed to detect the proteins encoded by the alternative transcripts. This could result from low affinity of the used commercial antibody to the isoforms. Alternatively, a small amount of these proteins in the pool of cellular proteins, might have not been detected by Western blotting. We performed in vivo and in vitro assays to address the role of the alternative UBE2A isoforms, and to evaluate the effect of c.382C-T mutation on UBE2A function. Taking into account the high amino acid conservation between the human UBE2A and the Saccharomyces cerevisiae ortholog RAD6, we used a 916;rad6 yeast strain to verify whether UBE2A alternative and mutated isoforms were able to complement its UV-sensitivity phenotype, as previously demosntrated for UBE2A isoform 1. We also performed in vitro assays to evaluate their ubiquitination activity towards histone H2A, a known in vitro substrate of RAD6 and UBE2A. Only UBE2A isoform 1 could rescue the UV sensitivity phenotype of the knockout yeast strain. The expression of the alternative isoforms 2 and 3 was partially toxic to this yeast strain, and toxicity increased under heat shock conditions. However, these two isoforms do not seem to be stable in yeast cells: as in human tissues, we failed to detect UBE2A isoforms 2 and 3 in yeast cells expressing the corresponding transcripts. The mutant isoform was stable in yeast, but was unable to rescue the UV-sensitivity phenotype, its expression resulting in severe toxicity to the 916;rad6 strain. On the other hand, toxicity was not observed when the mutant UBE2A isoform was expressed in wild type yeast. These findings suggest that isoforms 2 and 3 do not have ubiquitin conjugating activity and, apparently, are degraded immediately after translation. The fact that toxicity is enhanced when these isoforms are expressed under heat shock conditions supports Degradation hypothesis. The degradation could also be due to the absence of a functional partner, in yeast, that could contribute to their stability. Since the alternative isoforms were not detected in the human tissues analyzed, the degradation might occur in human cells as well. E2 enzymes share a catalytic domain and variations among them consist of insertions or terminal extensions, never deletions. Both isoforms 2 and 3 would have deletions of the catalytic domain, suggesting that they are not functional. A regulatory role for these transcripts is a possibility. Our in vitro assays confirmed that UBE2A isoform 1 is capable of histone H2A ubiquitination. The assays for isoforms 2 and 3 were inconclusive, since their lack of ubiquitin conjugating activity could be caused by incorrect in vitro refolding, required because the proteins were obtained from bacterial inclusion bodies after heterologous expression. The mutated protein, however, was able to interact with the ubiquitin molecule, but failed to transfer it to histones, thus pointing to the importance of the C-terminal segment in this process. Our in vitro assays stongly suggested that UBE2A autoubiquitination occur, an activity previously considered a possible E2 regulatory mechanism. Since there is evidence that some E2s form functional dimers, we hypothesized that, due to their high amino acid conservation, UBE2A and its paralog UBE2B might form heterodimers in vivo, as a mutual regulating mechanism. Under this hypothesis, the degradation of the mutated protein could be UBE2B dependent. The reduced autoubiquitination capacity of the mutated isoform could impair its degradation, and require the participation or the paralog. This would explain why the mutated protein was stable in the 916;rad6 yeast strain, but not in the patient´s cells with a functional UBE2B. Following the same reasoning, in wild type yeast, the presence of RAD6 would explain the absence of the mutated protein and toxicity. The non-viability of the double (UBE2A and UBE2B) knockout cells prevented testing whether the mutated protein was stable in the absence of its paralog. However, proteasome inhibition in cultured cells from one of our patients resulted in accumulation of the mutated protein, confirming its degradion via the ubiquitin-proteasome pathway. In conclusion, the UBE2A c.382C8594;T mutation seems to lead to mental retardation in our patients due to loss of UBE2A function: the mutated isoform is unable to rescue the UV-sensitivity phenotype of 916;rad6 yeast or to ubiquitinate histones in vitro. In addition, patients carrying UBE2A deletions share clinical manifestations with our patients. On the other hand, the possibility remains of a clinical effect of the requirement of UBE2B for degrading the mutated UBE2A. Our data suggest reciprocal ubiquitination in addition to autoubiquitination as UBE2A and UBE2B regulatory mechanism that would explain the conservation of the two paralog genes in mammals.

La Leucemia Mieloide Cronica(CML) è caratterizzata dalla presenza del gene di fusione BCRABL, prodotto dalla traslocazione tra i cromosomi 9 e 22. Se non trattata questa patologia progredisce entro 3 anni da una forma cronica(CP) a una forma acuta, la crisi blastica(BC). I meccanismi molecolari alla base della progressione della malattia non sono ancora completamente chiariti. Per raggiungere questo scopo ho quindi utilizzato un duplice approccio: 1)Grazie alla disponibilità di campioni CP/BC derivanti dagli stessi pazienti progrediti in BC dopo terapia standard, abbiamo sequenziato l’intero esoma(Whole Exome Sequencing-WES) e analizzato i dati utilizzando il campione CP come controllo. In questo modo l’analisi ha permesso di evidenziare solamente le alterazioni genetiche acquisite dopo progressione in BC. Ho quindi individuato la presenza di 2 mutazioni ricorrenti a carico dei geni RUNX1 e UBE2A, con quest’ultimo associato per la prima volta all’evoluzione in leucemia acuta e trovato mutato in circa l’11per cento dei campioni BC. Analisi successive invitro e invivo permetteranno di chiarire in particolare il ruolo delle mutazioni a carico di UBE2A nella progressione di CML. 2)Il secondo approccio si basa sullo studio del promotore del gene BCR. Dopo la traslocazione il gene di fusione BCRABL è sotto il controllo del promotore di BCR. Attualmente i meccanismi di regolazione di questo promotore sono poco chiari. Nel nostro laboratorio è stata precedentemente identificata una deregolazione trascrizonale di entrambi i geni BCR e BCRABL durante la progressione in BC. Ho quindi messo a punto un’analisi insilico per identificare i fattori di trascrizione che possano avere un ruolo nella regolazione del promotore BCR. Utilizzando tecniche di immunoprecipitazione di cromatina ho confermato in-vitro il legame dei fattori di trascrizione MYC e MAX a siti di legame specifici sul promotore di BCR(PBS1-4). Ho quindi osservato come l’overespressione di MYC e MAX sia in grado di indurre un aumento dell’espressione di BCR e di BCRABL, e che questo aumento risulta più evidente quando entrambi i fattori di trascrizione sono overespressi. Il silenziamento specifico di MYC nelle stesse linee cellulari ha dimostrato ulteriormente l’effetto regolatorio su entrambi i promotori. Per confermare questi risultati ho messo a punto un saggio di luciferasi inserendo il promotore di BCR nel plasmide pGL3. Cellule silenziate o meno per il gene MYC sono state poi trasfettate con i plasmidi di interesse. I risultati di questi esperimenti dimostrano l’importanza di MYC nella regolazione del promotore BCR. Ho inoltre osservato come il silenziamento di MYC in cellule BCRABL+ induca anche una alterazione del potenziale proliferativo e un aumento del tasso apoptotico. Questi dati descrivono quindi per la prima volta un meccanismo di regolazione del promotore BCR basato sul legame specifico di MYC e MAX e indicano una diretta associazione tra i livelli di espressione di MYC e quelli di BCRABL, entrambi upregolati durante l’evoluzione della malattia. E’ quindi suggerito un meccanismo molecolare alla base dell’aumento dell’espressione di BCRABL e dell’evoluzione in BC.<br>Chronic Myeloid Leukemia (CML) is caused by the BCR/ABL fusion gene. If untreated CML progresses within 3 years from a mild and easy to control form, called chronic phase (CP), into an aggressive and deadly acute leukemia called blast crisis (BC). Despite the aggressiveness of BC and the poor overall survival of BC patients, little is known about the molecular mechanisms responsible for the progression of the disease. Therefore to gain insight into the molecular lesions responsible for BC, I used a two prong approach. First, I performed whole-exome SEQ analysis of paired CP/BC CML samples from patients that underwent progression to BC after standard therapy. By comparing exome-sequences of 11 paired CP (used as a control) and BC samples we found a total of 38 single nucleotide mutations occurring in BC and that were absent in the corresponding CP sample. By using this approach we found recurrent somatic single nucleotide mutations in RUNX1 and UBE2A in 2 out of 11 BC samples. UBE2A is here associated with CML progression for the first time. In addition, Copy Number Alteration analysis of 9 matched BC/CP exomes reveals the presence of large chromosomal alterations acquired during BC transformation, among which the bulky alteration of chromosome 7 in 3/9 BC samples. In conclusion, despite some heterogeneity in the genetic alterations identified in BC samples, we were able to find 2 recurrently mutated genes associated with blastic transformation RUNX1 and UBE2A, with the last one never been detected in CML samples. Ongoing analysis on additional BC/CP samples and in vitro experiments will help to clarify the role of UBE2A mutations in CML progression. The second approach involves the study of the BCR promoter. After the oncogenic translocation, the BCRABL gene is transcriptionally controlled by the BCR promoter. In spite of all the research performed in the field, little is known in the regulation of BCR promoter. We thus need to understand what are the transcription factors or the mechanisms that regulate BCRABL expression. By in-silico analysis and in-vitro Chromatin Immunoprecipitation experiments we found that COUP-TF1, CTCF, MYC and MAX proteins bind at BCR promoter at specific sites. As many studies have indicated the involvement of MYC in CML progression, we here focused the study on MYC and its co-factor MAX for our further analysis. In the present study we demonstrate that MYC_MAX heterocomplex binds to the BCR promoter at four different loci, leading to upregulation of BCR and BCRABL at both transcriptional and protein level. In contrast, silencing of MYC expression in various BCRABL positive cell lines causes significant downregulation of BCR and BCRABL, which consequently leads to decreased proliferation and induction of apoptosis. Taken together all these results demonstrate that MYC_MAX heterocomplex regulates BCR promoter basal activity and can contribute to BCRABL overexpression and blast aggressiveness of CML advanced phase.

Maintaining genome integrity is indispensible for cells to prevent and limit accruement of deleterious mutations and to promote viable cell growth and proliferation. Cells possess a myriad of mechanisms to detect, prevent and repair incurred cellular damage. Here we discuss various proteins and their accompanying cellular pathways that promote genome stability. We first investigate the NEDD8 protein and its role in promoting homologous recombination repair via multiple Cullin E3 ubiquitin ligases. We provide specific mechanisms through which, UBE2M, an E2 conjugating enzyme, neddylates various Cullin ligases to render them catalytically active to degrade their substrates by the proteasome. We show that CUL1, CUL2 and CUL4 are important in regulating various steps in the DNA damage response. Our data indicates that UBE2M and the neddylation pathway are important for genome stability. Our second topic discusses the role of the USP1- UAF1 deubiquitinating enzyme in promoting homologous recombination. We show that USP1-UAF1 interact with and stabilize RAD51AP1 (RAD51- Associated Protein 1). RAD51AP1 has previously been reported to promote homologous recombination by facilitating recombinase activity of RAD51, an essential protein involved in homologous recombination repair. We show that USP1, UAF1 and RAD51AP1 depletion leads to genome instability. Our data demonstrates the importance of these proteins in promoting genome integrity via homologous recombination.

Chaque année, plus de 150 000 décès sont associés aux cancers épithéliaux de l’ovaire dans le monde, notamment en raison du développement d’une résistance à la chimiothérapie. Environ la moitié de ces cancers présentent des altérations moléculaires provoquant une déficience de la réparation de l’ADN par recombinaison homologue (HRD) qui les sensibilise à l’action des inhibiteurs de la protéine PARP (PARPi). A ce jour, il n’existe pas de test capable d’appréhender le phénotype HRD dans sa globalité, limitant ainsi l’accès à ces traitements. Dans ce contexte, nous avons entrepris de mettre au point des tests fonctionnels basés sur l’utilisation d’explants tumoraux tranchés puis sur l’utilisation d’organoïdes tumoraux dérivés de tumeurs ovariennes de patientes chimio-naïves ou antérieurement traitées. La culture d’explants s’est révélée inappropriée pour la réalisation de ces tests et nous avons alors focalisé nos travaux sur les organoïdes tumoraux. Ces derniers ont été exposés au carboplatine (traitement de 1e ligne) et à deux inhibiteurs de PARP (l’olaparib et le niraparib) utilisés en traitement d’entretien. En parallèle, nous avons collecté les données cliniques des patientes (survie, intervalle sans platine, RECIST, traitements) afin d’évaluer le potentiel prédictif de ces modèles. Les organoïdes tumoraux établis ont répondu de façon hétérogène aux différents médicaments testés, et nos résultats montrent que les tests réalisés sur les organoïdes sont capables d’identifier des patientes présentant un niveau de résistance élevé au carboplatine, suggérant que ce test fonctionnel pourrait présenter un intérêt prédictif vis-à-vis de ce médicament. Concernant le potentiel prédictif des organoïdes vis-à-vis des PARPi, des profils de sensibilité variés ont été identifiés, mais la corrélation avec la réponse clinique reste à établir par des études menées sur des échantillons de tumeurs issus de patientes traitées par ces médicaments<br>Worldwide each year, more than 150 000 women die from epithelial ovarian cancer largely due to emergence of resistance to chemotherapy. Approximately half of these cancers display molecular alterations that cause deficiency of DNA repair via homologous recombination (HRD), which confer sensitivity to PARP protein inhibitors (PARPi). To date, there is no test capable of fully identifying the HRD phenotype, thus limiting access to these treatments. In this context, we are developing functional assays based on the use of tumor explant slices and then, on the use of tumor organoids derived from ovarian tumors of chemotherapy-naive or previously treated patients. The culture of explants was unsuitable for this application and we then focused our work on tumor organoids. Tumor organoids were exposed to carboplatin (first-line treatment) and two PARP inhibitors (olaparib and niraparib) used for maintenance therapy. In parallel, we collected clinical data from patients (survival, platinum-free interval, RECIST, treatments) to evaluate the predictive potential of these models. The established tumor organoids responded heterogeneously to different drugs, and our results show that the organoid-based assay is capable of identifying patients highly resistant to carboplatin, suggesting that this functional assay could have a predictive value for patients treated with carboplatin. Regarding the potential of organoids in predicting PARPi response, multiple sensitivity profiles have been identified, but the correlation with clinical response has yet to be determined by studies conducted on tumor samples from patients treated with these drugs

The ubiquitination of proteins serves as molecular signal to control an enormous number of physiological processes and its dysregulation is connected to human diseases like cancer. The versatility of this signal stems from the diverse ways by which ubiquitin can be attached to its targets. Thus, specificity and tight regulation of the ubiquitination are pivotal requirements of ubiquitin signaling. Ubiquitin-conjugating enzymes (E2s) act at the heart of the ubiquitination cascade, transferring ubiquitin from a ubiquitin-activating enzyme (E1) to a ubiquitin ligase (E3) or substrate. When cooperating with a RING-type E3, ubiquitin-conjugating enzymes can determine linkage specificity in ubiquitin chain formation. Our understanding of the regulation of E2 activities is still limited at a structural level. The work described here identifies two regulation mechanisms in UBE2S, a cognate E2 of the human RING-type E3 anaphase-promoting complex/cyclosome (APC/C). UBE2S elongates ubiquitin chains on APC/C substrates in a Lys11 linkage-specific manner, thereby targeting these substrates for degradation and driving mitotic progression. In addition, UBE2S was found to have a role in DNA repair by enhancing non-homologous end-joining (NHEJ) and causing transcriptional arrest at DNA damage sites in homologous recombination (HR). Furthermore, UBE2S overexpression is a characteristic feature of many cancer types and is connected to poor prognosis and diminished response to therapy. The first regulatory mechanism uncovered in this thesis involves the intramolecular auto-ubiquitination of a particular lysine residue (Lys+5) close to the active site cysteine, presumably through conformational flexibility of the active site region. The Lys+5-linked ubiquitin molecule adopts a donor-like, ‘closed’ orientation towards UBE2S, thereby conferring auto-inhibition. Notably, Lys+5 is a major physiological ubiquitination site in ~25% of the human E2 enzymes, thus providing regulatory opportunities beyond UBE2S. Besides the active, monomeric state and the auto-inhibited state caused by auto-ubiquitination, I discovered that UBE2S can adopt a dimeric state. The latter also provides an auto-inhibited state, in which ubiquitin transfer is blocked via the obstruction of donor binding. UBE2S dimerization is promoted by its unique C-terminal extension, suppresses auto-ubiquitination and thereby the proteasomal degradation of UBE2S. Taken together, the data provided in this thesis illustrate the intricate ways by which UBE2S activity is fine-tuned and the notion that structurally diverse mechanisms have evolved to restrict the first step in the catalytic cycle of E2 enzymes<br>Die Ubiquitinierung von Proteinen fungiert als molekulares Signal zur Kontrolle einer Vielzahl physiologischer Prozesse, wobei eine gestörte Regulation der Ubiquitinierung eng mit zahlreichen Erkrankungen, wie beispielsweise Krebs, verbunden ist. Aufgrund der verschiedenen Verknüpfungsmöglichkeiten von Ubiquitin, die das zelluläre Schicksal des Zielproteins bestimmen, sind Spezifität und stringente Regulation unabkömmliche Voraussetzungen im Ubiquitinierungsprozess. Ubiquitin-konjugierende Enzyme (E2s) fungieren in der Mitte der Ubiquitinierungskaskade. Sie übernehmen ein Ubiquitinmolekül vom Ubiquitin-aktivierenden Enzym (E1) und übertragen es auf eine Ubiquitin-Ligase (E3) oder direkt auf das Zielprotein. Arbeiten Ubiquitin-konjugierende Enzyme mit E3s des RING-Typus zusammen, so bestimmen E2s die Art der Verknüpfung. Die Regulation der Aktivität Ubiquitin-konjugierender Enzyme auf struktureller Ebene ist jedoch bisher nur bedingt verstanden. Die hier dargelegte Arbeit umfasst die Identifizierung zweier Regulationsmechanismen des Ubiquitin-konjugierenden Enzyms UBE2S. UBE2S arbeitet mit einem humanen E3 des RING-Typus‚ dem ‚Anaphase Promoting Complex/Cyclosome‘ (APC/C) zusammen und bildet Lys11-spezifische Ubiquitinketten auf Substraten des APC/Cs. Hierdurch werden die Substrate für den Abbau durch das Proteasom markiert, was das Fortschreiten der Mitose bedingt. Zusätzlich wird UBE2S eine Rolle in der DNS-Reparatur zugeschrieben. Hierbei verstärkt UBE2S die nicht-homologe Rekombination (NHEJ) und verhindert außerdem die Transkription an DNS-Bruchstellen, die durch Homologe Rekombination (HR) repariert werden. Die Überexpression von UBE2S ist ein Charakteristikum verschiedenster Krebsarten, vermindert den Erfolg herkömmlicher Krebstherapien, und führt somit zu schlechten Prognosen für betroffenen Patienten. Der erste hier beschriebene Regulationsmechanismus beinhaltet die intramolekulare Ubiquitinierung eines Lysins (Lys+5) nahe des katalytischen Cysteins, mutmaßlich durch strukturelle Flexibilität der Region des aktiven Zentrums. Das Lys+5-verknüpfte Ubiquitin nimmt eine Donorubiquitin-ähnliche Position auf UBE2S ein, wodurch UBE2S gehemmt wird. Da ein Lysin an der Position +5 in ~25% der humanen E2-Enzyme vorhanden und eine physiologische Ubiquitinierungsstelle ist, birgt dieser Mechanismus Regulationsmöglichkeiten über UBE2S hinaus. Zusätzlich zum aktiven monomeren Zustand und dem durch Autoubiquitinierung ausgelösten inhibierten Zustand, kann UBE2S auch als Dimer vorliegen. In diesem Zustand ist es ebenfalls inaktiv, da die Donorubiquitin-Bindestelle auf UBE2S durch ein zweites Molekül des E2s blockiert wird. Begünstigt wird die Dimerisierung durch die C-terminale Verlängerung von UBE2S und verhindert so deren Autoubiquitinierung, und folglich den proteasomalen Abbau von UBE2S. Es handelt sich hierbei somit um einen zweiten Regulationsmechanismus von UBE2S. Zusammenfassend veranschaulichen die in dieser Arbeit dargelegten Daten die komplexen Möglichkeiten, durch die die Aktivität von UBE2S reguliert werden kann, sowie die Erkenntnis, dass strukturell unterschiedliche Mechanismen existieren, um den ersten Schritt der von Ubiquitin-konjugierenden Enzymen katalysierten Reaktion zu hemmen

Due to the action of endogenous and exogenous agents, DNA is subject up to 70,000 lesions per day, thus the existence of repair mechanisms and enzymes is more than necessary. We know basic mechanisms of several specific DNA repair pathways, of which the Fanconi anaemia (FA) repair pathway is one of the least explored. FA is a rare, autosomal recessive disorder characterized by early onset bone marrow failure, developmental defects, genomic instability and predisposition to acute myeloid leukaemia and solid tumours. The primary diagnosis of FA is a hypersensitivity to cross-linking agents of DNA due to inactivation of one of the 21 genes from the FA repair pathway, the so-called FANC genes (FA complementation group). The molecular defect in FA is an impaired repair of DNA interstrand cross-links (ICLs). The ICLs are cytotoxic lesions that inhibit the process of DNA replication and transcription. A crucial step in the FA pathway that initiates ICL repair is a monoubiquitination of FANCD2. FANCD2 monoubiquitination is a base for the recruitment of additional proteins that coordinate DNA repair. Ubiquitin is recruited via activating enzyme E1 (UBA1), ubiquitin-conjugating enzyme E2T (UBE2T) and transferred onto FANCD2 by multisubunit E3 ligase (FA core complex). There are up to 11 different proteins...

碩士<br>國立中山大學<br>生物科學系研究所<br>101<br>Breast microcalcification is an important feature of early breast carcinoma. Mammography and histology diagnosis reveal that 15% of breast microcalcification could be further defined as breast cancer. UBE2C, an ubiquitin conjugation E2C, is a newly identified biomarker candidate in various types of cancers. It contributes to the ubiquitin-mediated proteasome degradation in cell cycle. To evaluate the clinical implication of UBE2C in early breast cancer diagnosis, we measured the mRNA expression level of UBE2C from 56 sets of breast biopsy samples; each set contains non-microcalcification and microcalcification tissues from individual. Approximately 70% microcalcification samples with 2 fold increased UBE2C mRNA expression were detected by quantitative RT-PCR; in addiction, the immunohistochemistry also revealed that 68% microcalcification tissues were UBE2C positive. The UBE2C expression level and pathology diagnosis data of the microcalcification biopsies were future analysed. The results indicated a correlation between highly expression of UBE2C and early breast cancer, which suggested that UBE2C might be a comportable biomarker of early breast cancer. To further investigate the role of UBE2C in breast cancer oncogenesis, we determined the UBE2C expression level in 3 types of breast cancer cells. We found that the invasive cells, MDA-MB-231 and MDA-MB-361, highly express UBE2C, and the relative lower expression of UBE2C were observed in non-invasive cells, MCF-7. By using sh-RNA gene silencing vector of UBE2C, the cell growth reduction and cell death increasing were found in UBE2C knockdown MCF7 cells. In contrast, the MCF-7 growth rate was enhanced by over-expression of UBE2C. Our data suggest that the UBE2C plays the important role in promoting cell proliferation and breast oncogenesis.

碩士<br>國立臺灣大學<br>生化科技學系<br>102<br>Ub and ubiquitin-like protein modification involve in lots of important physiological pathways. Of all known ubiquitin-like proteins, NEDD8 (Neuronal precursor cell Expressed, Developmentally Down-regulated 8) is the closest relative to ub in amino acid sequence identity. In spite of their highly identity, ub and NEDD8 participate in distinct pathways, indicating that they should be distinguished and activated by individual specific enzyme systems. Previous studies showed that ub-specific peptidase USP2 and NEDD8-specific peptidase SENP8 perform different substrate specificity for recognizing ub and NEDD8. By investigating the crystal structure of ub-USP2 binding complex and comparing the amino acid sequence of ub and NEDD8, residues 4, 12 and 14 of ub were proposed as important molecular determinants. Mutations at F4K, T12E and T14E of ub partially inhibited its hydrolysis by USP2. In this study, the experimental results further revealed that residues 4, 12, 14 and 72 of ub can also be the molecular determinants for specific recognition by ubiquitin E1 activating enzyme Ube1. The present data showed that Ube1 cannot catalyze the ubiquitin mutant with the simultaneous mutations at residues 4, 12, 14 and 72, suggesting that these four molecular determinants are conserved among different enzymes in the ubiquitin processing and activation pathways.

Le régulateur transcriptionnel BAP1 est une déubiquitinase nucléaire (DUB) dont le substrat est l’histone H2A modifiée par monoubiquitination au niveau des residus lysines 118 et 119 (K118/K119). Depuis les dernières années, BAP1 emerge comme un gene suppresseur de tumeur majeur. En effet, BAP1 est inactivé dans un plethore de maladies humaines héréditaires et sporadiques. Cependant, malgré l’accumulation significative des connaissances concernant l’occurrence, la pénétrance et l’impact des défauts de BAP1 sur le développement de cancers, ses mécanismes d’action et de régulation restent très peu compris. Cette étude est dédiée à la caractérisation moléculaire et fonctionnelle du complexe multi-protéique de BAP1 et se présente parmi les premiers travaux décrivant sa régulation par des modifications post-traductionnelles. D’abord, nous avons défini la composition du corps du complexe BAP1 ainsi que ses principaux partenaires d’interaction. Ensuite, nous nous sommes spécifiquement intéressés a investiguer d’avantage deux principaux aspects de la régulation de BAP1. Nous avons d’abord décrit l’inter-régulation entre deux composantes majeures du complexe BAP1, soit HCF-1 et OGT. D’une manière très intéressante, nous avons trouvé que le cofacteur HCF-1 est un important régulateur des niveaux protéiques d’OGT. En retour, OGT est requise pour la maturation protéolytique de HCF-1 en promouvant sa protéolyse par O-GlcNAcylation, un processus de régulation très important pour le bon fonctionnement de HCF-1. D’autre part, nous avons découvert un mécanisme unique de régulation de BAP1 médiée par l’ubiquitine ligase atypique UBE2O. en effet, UBE2O se caractérise par le fait qu’il s’agit aussi bien d’une ubiquitine conjuratrice et d’une ubiquitine ligase. UBE2O, multi-monoubiquitine BAP1 au niveau de son domaine NLS et promeut son exclusion du noyau, le séquestrant ainsi dans le cytoplasme. De façon importante, nos travaux ont permis de mettre de l’emphase sur le rôle de l’activité auto-catalytique de chacune de ces enzymes, soit l’activité d’auto-déubiquitination de BAP1 qui est requise pour la maintenance de sa localisation nucléaire ainsi que l’activité d’auto-ubiquitination d’UBE2O impliquée dans son transport nucléo-cytoplasmique. De manière significative, nous avons trouvé que des défauts au niveau de l’auto-déubiquitination de BAP1 due à des mutations associées à certains cancers indiquent l’importance d’une propre regulation de cette déubiquitinase pour les processus associés à la suppression de tumeurs.<br>BAP1 is a nuclear deubiquitinating enzyme (DUB) that acts as a transcription regulator and a DUB of nucleosomal histone H2AK119. In the recent years, it has become clear that BAP1 is a major tumor suppressor, inactivated in a plethora of hereditary and sporadic human malignancies. Although, we now accumulated a significant body of knowledge in respect to the occurrence, penetrance and impact of BAP1 disruption in cancer, its mechanism of action and regulation remained poorly defined. This work is dedicated to the biochemical and functional characterization of the BAP1 multiprotein complex and presents one of the first cases regarding its regulation by post-translational modifications. First, we defined the initial composition of the BAP1 complex and its main interacting components. Second, we specifically focused on two aspects of BAP1 regulation. We described the cross regulation between the two major components of the complex namely HCF-1 and OGT. We found that HCF-1 is important for the maintenance of the cellular levels of OGT. OGT, in turn, is required for the proper maturation of HCF-1 by promoting O-GlcNAcylation-mediated limited proteolysis of its precursor. Third, we discovered an intricate regulatory mechanism of BAP1 mediated by the atypical ubiquitin ligase UBE2O. UBE2O multi-monoubiquitinates BAP1 on its NLS and promotes its exclusion from the nucleus. Importantly, our work emphasises the role of the autocatalytic activity of both enzymes namely the auto-deubiquitination activity of BAP1, required for the maintenance of nuclear BAP1 and the auto-ubiquitination of UBE2O implicated in its nucleocytoplasmic transport. Significantly, we found that auto-deubiquitination of BAP1 is disrupted by cancer-associated mutations, indicating the involvement of this process in tumor suppression.