As metaloproteases AAA/FtsH são componentes chave do controle da qualidade das proteínas inseridas nas membranas de mitocôndrias e cloroplastos. Em Arabidopsis thaliana, as proteases FtsH presentes nas membranas dos tilacóides formam um complexo heterohexamérico composto pelas subunidades FtsH1/FtsH5 (tipo A) e FtsH2/FtsH8 (tipo B). Este complexo está envolvido na reciclagem de proteínas foto-danificadas, especialmente da proteína D1, centro de reação do PSII. Algumas linhas de evidências indicam ainda que existe um limiar de concentração das proteases FtsH, necessário para a correta formação e desenvolvimento dos cloroplastos. Apesar da extensiva caracterização genética e molecular das proteases FtsH, o mecanismo regulatório do complexo FtsH dos cloroplastos não foi totalmente elucidado até o momento, contudo existem evidências de que a sua ativação pode estar relacionada a alta incidência luminosa e a outras condições de estresse. A presença de fatores proteicos auxiliares, foi testada como hipótese alternativa por nosso grupo, através do uso da protease FtsH5 como isca em um ensaio de duplo híbrido de leveduras. Este ensaio identificou uma proteína interagente putativa, nomeada FIP (FtsH5 Interacting Protein), a qual comprovadamente interage com FtsH5 e está localizada nas membranas dos tilacóides. De modo a investigar o papel regulatório putativo de FIP sobre a atividade do complexo FtsH, nós analisamos os padrões de expressão em uma ampla gama de condições de estresse a partir de dados públicos de microarranjos de DNA. Os perfis de expressão indicam que FIP pode ser um regulador negativo da atividade do complexo. Os resultados também sugerem que o complexo pode estar envolvido na resposta do cloroplasto a diferentes tipos de condições de estresse. O estudo da história evolutiva das proteínas interagentes FtsH5 e FIP evidenciou que as sequências homólogas a FIP são encontradas exclusivamente em musgos e plantas superiores, sugerindo assim que a origem de FIP pode estar relacionada a colonização terrestre. Todos os genes codificantes das proteases FtsH do complexo foram usados como \"query\" na busca por sequências homólogas, permitindo a classificação das proteases FtsH nos tipos A e B por inferência filogenética Bayesiana. Análises filogenéticas Bayesianas também foram feitas para FIP e as proteases FtsH tipos A e B, independentemente. A análise Mirrortree suportou a existência de coevolução entre FIP e as proteases FtsH tipo A. Por outro lado, nenhuma correlação foi encontrada entre FIP e as proteases FtsH tipo B, o que corrobora nossas observações experimentais anteriores. Além disso, o agrupamento portador de homólogos FIP pôde ser recuperado em uma filogenia mais completa das proteases FtsH do tipo A. Análises subsequentes mostraram que ambas as proteínas interagentes estão extensivamente sobre seleção negativa e que proteases FtsH tipo A são bastante conservadas, principalmente nos seus domínios internos. / Eukaryotic AAA/FtsH metalloproteases display a key role in the protein quality control of membrane-inserted proteins in mitochondria and chloroplasts. In Arabidopsis thaliana, chloroplast thylakoidal membranes FtsH proteases form a heterohexameric complex made by FtsH1/FtsH5 (type A) and FtsH2/FtsH8 (type B) subunits. This complex is involved in protein turnover of photo-damaged proteins, in particular the D1 protein at the PSII reaction center. Several lines of evidence also indicate that a FtsH threshold level is necessary for the proper formation and development of chloroplasts. Despite extensive genetic and molecular characterization of the FtsH proteases, the regulatory mechanism of the FtsH complex in chloroplasts has not yet been fully elucidated, however, there are evidences that its activation might be related to high light incidence and other stress conditions. The presence of auxiliary protein factors, as an alternative hypothesis, was tested by our group, through the use of the protease FtsH5 as bait in a yeast two-hybrid assay. This essay identified a putative interacting protein named FIP (FtsH5 Interacting Protein), which has been proved to interact with FtsH5 and be located at the thylakoid membranes. In order to investigate a putative regulatory role of FIP on FtsH complex activity, we analyzed gene expression patterns in a wide range of stress conditions from public DNA microarray data. The expression profiles indicate that FIP could be a negative regulator of the FtsH complex activity. The results also suggest that the complex may be involved in the chloroplast response to different types of stress conditions. In order to shed some light on the evolutionary history of FtsH5 and FIP interacting proteins, we have shown that FIP\'s homologous sequences were exclusively found in mosses and higher plants, suggesting that FIP origin might be related to the plant terrestrial colonization. All Arabidopsis FtsH complex-encoding genes were used as \"query\" sequences in search for homologous sequences, allowing us to classify the FtsH proteases in type A and B by Bayesian phylogenetic inference. Bayesian phylogenetic analyses were also run for FIP and FtsH types A and B proteases, independently. Mirrortree analysis supported coevolution between FIP and type A FtsH proteases. On the other hand, no correlation was found between FIP and type B FtsH homologues, which support our previous experimental observations. In addition, the FIP bearing cluster could be recovered in a more complete type A FtsH phylogeny. Subsequent analyzes have shown that both interacting proteins are extensively under negative selection and that type A FtsH are very conserved, mainly in its inner domains.

Orientador: Orientador : Kleber Gomes Franchini / Tese (doutorado) - Universidade Estadual de Campinas, Faculdade de Ciências Médicas / Made available in DSpace on 2018-08-21T12:58:21Z (GMT). No. of bitstreams: 1 Cardoso_AlissonCampos_D.pdf: 4487510 bytes, checksum: eec48cf16c52d89f7324a6a56476ca59 (MD5) Previous issue date: 2012 / Resumo: Durante os primeiros dias do desenvolvimento pós-natal, os miócitos cardíacos perdem a capacidade de proliferação, sendo o crescimento adicional do coração decorrente de hipertrofia e não hiperplasia dos miócitos cardíacos. No entanto, em situações de estresse os miócitos cardíacos diferenciados podem apresentar desdiferenciação e reestabelecimento do ciclo celular. Os mecanismos envolvidos nesse fenômeno são ainda pouco compreendidos. No presente estudo, demonstramos que a ativação do fator de transcrição MEF2C (Myocyte Enhancer Factor 2-C) tem papel crítico no processo de desdiferenciação de miócitos cardíacos. Essa conclusão foi obtida por meio de experimentos de ganho de função pela superexpressão de MEF2C em miócitos ventriculares de ratos neonatos em cultura (MVRNs). Demonstramos que a superexpressão de MEF2C em MVRNs induziu a desdiferenciação e a ativação de mecanismos envolvidos na progressão do ciclo celular. Esses resultados foram obtidos por meio de experimentos de microarranjo de DNA, PCR em tempo real, western blotting e análise do fenótipo celular por microscopias de luz, confocal e eletrônica de transmissão. Esses fenômenos foram atenuados pela superexpressão da quinase de adesão focal (FAK), uma proteína que reconhecidamente exerce efeitos pró-hipertróficos em miócitos cardíacos adultos. Experimentos in vivo e in vitro demonstraram a interação direta entre o fator de transcrição MEF2C e a FAK. Estudos com base em ensaios de reação cruzada associada à espectrometria de massas, dinâmica molecular, espalhamento de raios-X a baixos ângulos e mutação sítio dirigida, demonstraram que as hélices 1 e 4 do domínio FAT da FAK interagem diretamente com a domínio de ligação ao DNA do dímero de MEF2C. Estudos de afinidade e de gel shift demonstraram que a porção FAT da FAK desloca a interação MEF2C/DNA in vitro. Ensaios de gene repórter demonstraram que a FAK, mediada pela região C-terminal, diminui a atividade transcricional de MEF2C em células C2C12. O conjunto de dados demonstra que a ativação do fator de transcrição MEF2C em MVRNs induz a desdiferenciação e ativação de mecanismos de progressão do ciclo celular e que a FAK impede esses efeitos através da interação inibitória no domínio de ligação de MEF2C ao DNA / Abstract: During the first days of postnatal development, cardiac myocytes lose their ability to proliferate, and the further growth of the heart is due to hypertrophy and not hyperplasia of cardiac myocytes. However, in response to stress, cardiac myocytes may have dedifferentiation and re-establishment of the cell cycle. The mechanisms involved in this phenomenon are still poorly understood. In the present study, we demonstrated that activation of the transcription factor MEF2C (myocyte enhancer factor 2-C) plays a critical role in the process of dedifferentiation of cardiac myocytes. This conclusion was obtained by gain-of-function experiments through overexpressing MEF2C in neonatal rat ventricular myocytes in culture (NRVMs). We also showed that overexpression of MEF2C in NRVMs induced the dedifferentiation and activation of mechanisms involved on cell cycle progression. These results were obtained by DNA microarray experiments, real time PCR, western blotting and cell phenotype analysis by light microscopy, confocal and electronic transmission. These effects were attenuated by overexpression of focal adhesion kinase (FAK) protein known to exert pro-hypertrophic effects on adult cardiac myocytes. In vivo and in vitro experiments demonstrated the direct interaction between the transcription factor MEF2C and FAK. A model based on crosslinking technology coupled with mass spectrometry, small angle X-ray scattering and the site directed mutation analyses indicated that alpha-helices 1 and 4 of FAK FAT domain interacts directly with the DNA binding domain of MEF2C dimer. Affinity studies and gel shift assay demonstrated that the FAK FAT domain displaces the MEF2C/DNA interaction in vitro. Reporter gene assays demonstrated that FAK, mediated by the C-terminal region, decreases the transcriptional activity of MEF2C in C2C12 cells. The data set shows that the activation of the transcription factor MEF2C in MVRNs induces dedifferentiation and activation of cell cycle progression and that FAK prevents these effects by inhibitory interaction with DNA binding domain of MEF2C / Doutorado / Biologia Estrutural, Celular, Molecular e do Desenvolvimento / Doutor em Ciências

Doctor of Philosophy / Department of Physics / Jeremy D. Schmit / Diseases such as Alzheimer’s, Parkinson’s, eye lens cataracts, and Type 2 diabetes are the results of protein aggregation. Protein aggregation is also a problem in pharmaceutical industry for designing protein based drugs for long term stability. Disordered states such as precipitates and gels and ordered states such as crystals, micro tubules and capsids are both possible outcomes of protein–protein interaction. To understand the outcomes of protein–protein interaction and to find the ways to control forces, it is required to study both kinetic and equilibrium factors in protein–protein interactions. Salting in/salting out and Hofmeister effects are familiar terminologies used in protein science field from more than a century to represent the effects of salt on protein solubility, but they are yet to be understood theoretically. Here, we build a theory accounting both attractive and repulsive electrostatic interactions via the Poisson Boltzmann equation, ion–protein binding via grand cannonical partition function and implicit ion–water interaction using hydrated ion size, for describing salting in/salting out phenomena and Hofmeister and/or salt specific effect. Our model free energy includes Coulomb energy, salt entropy and ion–protein binding free energy. We find that the salting in behavior seen at low salt concentration near the isoelectric point of the protein is the output of Coulomb energy such that the addition of salt not only screens dipole attraction but also it enhances the monopole repulsion due to anion binding. The salting out behavior appearing after salting in at high salt concentration is due to a salt mediated depletion interaction. We also find that the salting out seen far from the isoelectric point of the protein is dominated by the salt entropy term. At low salt, the dominant effect comes from the entropic cost of confining ions within the aggregates and at high salt, the dominant effect comes from the entropy gain by ions in solution by enhancing the depletion attraction. The ion size has significant effects on the entropic term which leads to the salt specificity in the protein solubility. Crystal growth of anisotropic and fragile molecules such as proteins is a challenging task because kinetics search for a molecule having the correct binding state from a large ensemble of molecules. In the search process, crystal growth might suffer from a kinetic trap called self–poisoning. Here, we use Monte Carlo simulation to show why protein crystallization is vulnerable to the poisoning and how one can avoid such trap or recover crystal growth from such trap during crystallization. We show that self–poisoning requires only three minimal ingredients and these are related to the binding affinity of a protein molecule and its probability of occurrence. If a molecule attaches to the crystal in the crystallographic state then its binding energy will be high but in protein system this happens with very low probability (≈ 10−5). On the other hand, non–crystallographic binding is energetically weak, but it is highly probable to happen. If these things are realized, then it will not be surprising to encounter with self–poisoning during protein crystallization. The only way to recover or avoid poisoning is to alter the solution condition slightly such as by changing temperature or salt concentration or protein concentration etc.

Les IDPs sont des protéines dépourvues de structure 3D unique en solution et en l'absence de leur(s) partenaire(s). Ces protéines possèdent des propriétés d'interactions avec leurs partenaires uniques.L'extrêmité C-terminale de la nucléoprotéine du virus de la rougeole, NTAIL, est une IDP. NTAIL interagit avec XD, le domaine C-terminal globulaire de la phosphoprotéine virale, via la box2 qui est un a-MoRE. Cette interaction permet le recrutement de la protéine L afin de former la polymérase virale.J'ai pu montrer que la contribution des différents résidus au sein de l'a-MoRE dépend de l'orientation de leur chaîne latérale, et que la substitution d'un seul acide aminé crucial a des effets dramatiques sur la réplication virale.Les IDPs conservent un désordre résiduel non négligeable au sein du complexe. Cela se manifeste par la présence des régions « fuzzy » de part et d'autres du MoRE. Nous avons montré que la région « fuzzy » en amont de la box2 inhibe l'établissement de l'interaction entre NTAIL et ses partenaires.Lors de l'interaction avec leurs partenaires les IDPs subissent en général un gain de structure. Le repliement associé à l'interaction peut avoir lieu avant ou après interaction. Des études précédentes ont montré l'existence d'une pré-structuration partielle de l'alpha-MoRE de NTAIL. L'interaction entre NTAIL et XD a été étudiée et a permis de conclure que NTAIL se replie selon un mécanisme de repliement induit.Dans leur ensemble, ces études contribuent à éclaircir les mécanismes moléculaires qui gouvernent la reconnaissance de partenaires par les IDPs. / IDPs are proteins devoided of a unique and stable 3D structure in solution and in the absence of their partners. Those proteins possess unique properties, as well as mechanisms of interaction with their partners.The C-terminal region of the nucleoprotein of measles virus, NTAIL, is an IDP and interacts with XD, the globular C-terminal domain of the viral phosphoprotein, via the box2 region that is an (alpha-MoRE). This interaction allows to recruit the L protein in order to form the viral polymerase.The aim of my work was to characterize the molecular basis of NTAIL-XD interaction. I was able to show that the contribution of the amino acids among box2 depends on the orientation of their lateral chain, and that the substitution of one single amino acid has drastic effect upon the viral replication.IDPs keep a non-negligible amount of residual disorder among the complex. This fuzziness can have multiple forms, like the presence of fuzzy regions from either side of the MoRE. The impact fuzzy regions have within the complex is not well known. We demonstrated that the fuzzy region upstream box2 inhibits the settlement of the interaction between NTAIL and its partners.When interacting with their partners, IDPs generally undergo a folding associated with binding that can take place either before or after the interaction. The interaction between NTAIL and XD was investigated and monitored by fluorescence kinetic measurements, using variants bearing a tryptophan substitution. We concluded without any doubt that NTAIL folds under an induced folding mechanism.Those studies together contribute to enlighten the molecular mechanisms that govern partner recognition by the IDPs.

No description available.

La croissance et le développement des plantes sont fortement régulés par plusieurs hormones végétales, dont l’auxine qui joue un rôle prépondérant. La modification de l’expression de certains gènes en réponse à l’auxine est contrôlée par des interactions spécifiques entre les facteurs de transcription ARF (Auxin Response Factors) et les protéines Aux/IAA. Des études sur Arabidopsis thaliana ont aussi montré l’implication de corépresseurs de la famille TOPLESS pour réprimer les gènes cibles des ARF. Toutefois, cette régulation transcriptionnelle a surtout été caractérisée chez la plante modèle Arabidopsis et la validité de ce modèle n’a pas encore été confortée par l’étude d’autres modèles. La tomate (Solanum lycopersicon), espèce modèle tant pour les Solanacées que pour les plantes à fruits constitue une bonne alternative pour élucider les caractères généraux liés à la signalisation auxinique. Dans notre travail, nous avons d’abord mis en place des protocoles expérimentaux – double-hybride, pull-down, complémentation de fluorescence (BiFC, Bifluorescence Complementation) – permettant d’étudier les interactions protéines-protéines. Ces méthodes ont d’abord été validées sur des couples Aux/IAA – ARF étant connus chez la tomate pour leur implication dans le développement et la maturation des fruits (SlIAA9, SlARF8, SlIAA3, SlARF4, SlIAA27). L’utilisation du double hybride a également permis de construire une carte d’interactions entre les Aux/IAA et les ARF de tomate. Dans un deuxième temps, la disponibilité de la séquence du génome de la tomate a permis d’entreprendre une étude globale de la famille des corépresseurs TOPLESS. Cette étude a inclus : la caractérisation et le clonage des gènes, l’analyse de la séquence protéique, une analyse phylogénétique de la famille sur un ensemble de génome séquencés, la caractérisation du profil d’expression des différentes isoformes ainsi qu’une analyse comparative de leur capacité d’interaction avec les protéines Aux/IAA. Enfin, dans un dernier temps, nous avons souhaité construire des premiers outils permettant d’entreprendre une recherche non-ciblée de nouveaux partenaires interagissant avec les protéines ARF ou Aux/IAA en partant de protoplastes de cellules BY-2 de tabac exprimant de façon transitoire des gènes codant des protéines chimères (tagged proteins). Même si ce travail reste préliminaire, il a pu notamment illustrer l’importance de l’intégrité des noyaux pour la stabilité des Aux/IAA, même en l’absence d’auxine. / The plant hormone auxin plays a central role in plant growth and development. The specific Aux/IAAs and Auxin Response Factors (ARFs) interactions are involved in auxin signaling pathway to regulate the auxin-responsive gene expression. Studies in Arabidopsis showed that TOPLESS family (TPLs) also was recruited by some Aux/IAAs to repress the function of ARFs. The whole machinery of the auxin signaling pathway is not clear yet, and most of this knowledge comes from the research on Arabidopsis. As a reference for Solanaceae and fleshy fruit plant, tomato (Solanum lycopersicon) is a good alternative model to better understand general traits of the auxin regulation process. In our work, we first established in our labs three experimental protocols – Yeast two-Hybrid, Pull-down and Bifluorescence complementation to unravel protein-protein interactions. These methods were first challenged on specific Aux/IAA and ARF proteins that were already characterized as major actors in fruit tomato development or ripening (SlIAA9, SlARF8, SlIAA3, SlARF4, SlIAA27). This also enabled us to build an ARF-Aux/IAA interaction map. In a second part, taking advantage of the tomato genome sequence, we carried a whole-genome study on tomato TOPLESS family. This investigation included gene cloning and characterization, protein sequence analysis, phylogenetic analyses, expression pattern and construction of protein-protein interaction maps. In a last part, we developed tools to start a non-targeted approach aiming at identifying new potential partners or protein complex involved in auxin signaling pathway using BY-2 tobacco cell protoplasts transiently expressing tagged-proteins. Although this study is still preliminary, it demonstrated the importance of nucleus integrity for Aux/IAA stability even in absence of auxin.

Domínio proteico é uma sequência de aminoácidos evolutivamente conservada e funcionalmente independente. Um dos aspectos mais importantes do estudo de uma proteína que contem múltiplos domínios é o entendimento da comunicação, entre os diferentes domínios, e seu papel biológico. Essa comunicação em maior parte é feita pela interação direta entre domínios. A interação poderia ser tratada como uma clássica interação proteína-proteína. Entretanto, proteínas multidomínio possuem restrições determinadas por suas regiões conectoras. Os conectores interdomínio impõem restrições e limitam espaço conformacional dos domínios. Apresentamos aqui o MAD, uma rotina capaz de obter modelos tridimensionais de alta resolução para proteínas, contendo qualquer número de domínios, a partir de sua sequencia primária. Os domínios conservados são identificados utilizando a base de domínios conservados (CDD) e seus limites são utilizados para definir as regiões conectoras. É criado um ensamble de possíveis dobramentos dos conectores e sua distribuição de distâncias C/N-terminais são utilizadas como restrição espacial na busca pela interação entre os domínios.Os modelos dos domínios são obtidos por uma modelagem comparativa. Foi implementada uma heurística, capaz de lidar com a natureza combinatorial dos múltiplos domínios e com a necessidade imposta pela limitação computacional de realizar o docking dos domínios em forma de pares. Todas combinações de domínios são submetidas as rotinas de docking. Aplica-se filtro de distância e energético, excluindo as conformações que apresentam distância C/N-terminal entre domínios maior do que o valor máximo observado no ensamble de conectores e seleciona as conformações energeticamente mais favoráveis. As conformações são submetidas a uma rotina de agrupamento hierárquico baseada em sua similaridade estrutural. Para a segunda fase as conformações selecionadas são pareadas com seu domínio complementar e ressubmetidas a rotina de docking até que todas as fases tenham sido completadas. Foi criado um conjunto de testes a partir do Protein Data Bank contendo 54 proteínas multidomínio para que a rotina de docking do MAD fosse comparada com outros softwares utilizados pela comunidade cientifica, mostrou-se superior ou equivalente aos métodos testados. A capacidade de utilizar dados experimentais foi demostrada através da proposição de um modelo da forma ativa da enzima tirosina fosfatase 2, nunca observado experimentalmente. A rotina de docking foi expandida paralelamente em uma aplicação standalone e utilizada na resolução de diversos problemas biológicos. Concluímos que a inovação metodológica proposta pelo MAD é de grande valia para a modelagem molecular e tem potencial de gerar uma nova perspectiva a respeito da interação de proteína multidomínio, visto que é possível analisar essas proteínas em sua plenitude e não como domínios separados. / Protein domain is an evolutionary conserved and functionally independent amino acid sequence. One of the most important aspects of the study of a protein that contains multiple domains is the understanding of communication between the different areas, and their biological role. This communication is made mostly by direct interaction between domains. The interaction could be treated as a classical protein-protein interaction. However, multidomain proteins have certain restrictions for its connector regions. The intra connectors impose restrictions and limit conformational space of the domains. We present the MAD, a routine able to get three-dimensional models of high-resolution protein, containing any number of domains, from its primary sequence. The conserved domains are identified using the basic conserved domains database (CDD) and its boundaries are used to define the connector regions. This creates a ensemble of possible folding of the connectors and distribution of distances C/N-terminals are used as spatial restriction in the search for interaction between domains.Os models of the domains are obtained by comparative modelling. A heuristic able to handle the combinatorial nature of the multiple areas and the need imposed by the computer to perform the limitation of the docking areas as pairs was implemented. All combinations of domains are referred to the docking routines. Distance and energy filters are applied, excluding conformations that have C/N-terminal domains distances larger than the maximum value observed in the connectors ensemble and selects the most favourable energy conformations. Conformations are subjected to hierarchical clustering routine based on their structural similarity. For the second phase, the selected conformations are paired with its complementary domain and resubmitted to the docking routine until all phases have been completed. A test set has been created from the Protein Data Bank containing 54 multidomain proteins so that the docking routine of MAD could be compared with other software used by the scientific community, it has been shown to be superior or equivalent to the tested methods. The ability to use experimental data was demonstrated by proposing a model of the active form of tyrosine phosphatase enzyme 2, never observed experimentally. The docking routine was expanded in a standalone application and used in solving various biological problems. We conclude that the methodological innovation proposed by the MAD is very useful for molecular modelling and has the potential to generate a new perspective on multidomain protein interaction as you can analyse these proteins in its entirety and not as separate domains.

京都大学 / 新制・課程博士 / 博士(理学) / 甲第23460号 / 理博第4754号 / 新制||理||1681(附属図書館) / 京都大学大学院理学研究科生物科学専攻 / (主査)教授 杤尾 豪人, 教授 森 和俊, 教授 望月 敦史 / 学位規則第4条第1項該当 / Doctor of Science / Kyoto University / DGAM

Tato práce se zabývá zbůsoby, jimiž je možné získávat data z bioinformatických databází obsahujících data týkajících se interakcí mezi proteiny. Od souvislostí okolo vzniku bioinformatiky sloučením informatiky a biologie tato práce uvede čtenáře do problematiky přístupu k datům týkajících se interakcí mezi proteiny. Tato práce vysvětlí důvody vzniku IMEx konsorcia, jeho cíle a prostředky, kterými svých cílů dosahuje. IMEx konsorcium dalo vzniknout mnoha standardům, které usnadňují přístup k datům členů konsorcia a výměnu těchto dat mezi nimi. Jedním z výtvorů IMEx konsorcia je i webová služba PSICQUIC, která byla navržena s využitím architektonického stylu REST, a která je přístupná i pomocí protokolu SOAP. Obě tyto kategorie přístupů k webových službám jsou v rámci této práce studovány a na základě výsledků výzkumu je implementována aplikace pro získávání interakcí mezi proteiny z databází, jenž jsou členy IMEx konsorcia.

Les interactions protéine-protéine (IPP) jouent un rôle essentiel dans le vivant. Mon travail de thèse s’est concentré sur développement de méthodes bio-informatiques pour la prédiction et la modélisation structurale des IPP. Mon objectif était d'améliorer le pouvoir prédictif des méthodes permettant de prédire les structures d’assemblages macromoléculaires (docking) et d'aborder les problèmes rencontrés par les biologistes sur des cas réels d’interactions.Pour obtenir des modèles de protéines isolées de meilleure qualité, j’ai tout d’abord développé le serveur HHalign-Kbest basé sur des algorithmes d’alignements sous-optimaux. Ensuite, dans le domaine du « docking », j’ai élaboré le serveur InterEvDock qui prend en compte les informations de coévolution entre protéines. Les validations en aveugle montrent que ce serveur atteint de meilleures performances que d’autres serveurs de référence lorsque l’information évolutive est disponible.Afin de tester plus à fond nos méthodes, nous avons participé au concours CAPRI - un concours international pour la prédiction des interactions protéiques. Sur les sessions couvrant la période 2013-2016, notre groupe s’est classé 1er. Enfin, j'ai développé un jeu de données d’apprentissage et de test, PPI4DOCK. Il contient un très grand nombre de cibles de complexes (plus de 1000) et permettra d'améliorer les méthodes de docking à partir des structures expérimentales ou de modèles.En termes d'applications, je me investis dans différents projets collaboratifs, qui touchent des domaines aussi variés que, la recherche de partenaires pour le chaperon d’histone Asf1; la prédiction des modes d’interaction entre CENP-F et Nup133 dans le contexte de la mitose et de Exo70 et Abi dans celui de la régulation de la mobilité cellulaire; la simulation des modes de liaison entre le complexe Ku et ses partenaires peptidiques, dans les voies de réparation de l'ADN. / Protein-protein interactions (PPIs) play essential roles in life. My PhD work aimed at developing advanced bioinformatics methods in the field of PPI prediction at the structural scale. My goal was to improve the predictive power of methods which model the structures of macromolecular assemblies (docking) and to tackle real-life problems faced by biologists.First, I developed HHalign-Kbest server using algorithms for the search of suboptimal solutions to gain better-quality models. Second, in the field of protein docking, I built InterEvDock server which can take co-evolutionary information into account. It yields better performance than other state-of-the-art servers. In order to further test our methods, we participated in CAPRI – an international challenge for prediction of protein interactions. Over years 2013-2016, our group ranked 1st at the 6th CAPRI evaluation meeting. At last, I developed a realistic benchmark dataset PPI4DOCK, largest dataset so far, in order to improve docking methods for the scientific community.In terms of applications, I was involved in a variety of collaborative projects with different labs. As representative examples, I searched for binding partners of the histone chaperone Asf1; I studied the CENP-F/Nup133 interaction in the context of mitosis and the Exo70/Abi interaction related to cell mobility regulation; I also simulated the binding modes of multiple peptides, partners of Ku complex involved in DNA repair pathway.