La biologie des systèmes actuelle s’appuie sur des techniques d’analyse biologique à haut débit comme la transcriptomique ou la métabolomique. Cependant, ces techniques haut débit ont leurs limites et peuvent générer des erreurs. En croisant les résultats de différentes techniques d’analyse biologique, nous espérons pallier une partie de leurs limites. À cet effet, nous avons commencé à développer une plateforme de modélisation, MPSA (Metabolic Pathways Software Analyzer), permettant d’intégrer les données générées à des réseaux métaboliques. MPSA permet de représenter les graphes de voies métaboliques, d’effectuer des simulations basées sur la résolution de systèmes d’équations différentielles et d’étudier la structure des réseaux métaboliques par le calcul et la représentation des modes élémentaires. Nous avons développé différentes applications web permettant, d’une part, l’interprétation des résultats biologiques en utilisant des bases de données et, d’autre part, leur export vers MPSA. La base de données centrale de ce développement est myKegg, incluant l’ensemble des voies métaboliques humaines de la base de données publique KEGG ainsi qu’une base de synonymes construite elle aussi à partir de KEGG. Cette base permet d’identifier des voies métaboliques et de les importer dans MPSA. Une base de données de métabolomique, BioNMR, a aussi été construite spécifiquement pour organiser les résultats générés à partir de spectres de RMN. Une autre application web, GeneProm, a été développée pour l’analyse de promoteurs de gènes ou promotologie. Un protocole d’étude a été mis au point et testé sur un groupe de 4 gènes codant pour les isoformes 1 à 4 de la protéine ANT, transporteur mitochondrial d’ATP, chacune ayant un rôle et un profil d’expression spécifique dans la bioénergétique cellulaire. L’étude par promotologie de ces 4 gènes a permis d’identifier des éléments de régulation spécifiques dans leurs séquences promotrices et d’identifier des gènes potentiellement co-régulés. Ces gènes peuvent ensuite être exportés vers notre plateforme MPSA. L’ensemble de ce développement sera inclus au projet de plateforme intégrative de l’Unité de Nutrition Humaine de l’INRA. / Current systems biology relies on high-throughput biological analysis techniques such as transcriptomics or metabolomics. However, these techniques may generate errors. By crossing results from different analysis techniques, we hope to avoid at least part of these limits. For that purpose, we started to develop a modeling platform, MPSA (Metabolic Pathways Software Analyzer). MPSA allows integrating biological data on metabolic pathways. MPSA also ensures the display of metabolic pathways graphs, the simulation of models based on ordinary differential equations systems solving and the study of network structures using elementary flux modes. We have developed several web applications allowing on the one hand to interpret biological results by using databases, and on the other hand to export these data to MPSA. The main database of this work is myKegg. It includes all human KEGG metabolic pathways and a list of synonyms for human KEGG entries. This base allows to identify metabolic pathways from a list of biological compounds and to import them in MPSA. Another database, BioNMR, has been developed to organize the data extracted from NMR spectra. Another web application named GeneProm has been developed to analyze gene promoters. A promotology protocol was developed and tested on a set of four genes coding for the four ANT (adenine nucleotide translocator) protein isoforms. Each ANT isoform has a specific expression profile and role in cell bioenergetics. The promotology study of these four genes led us to construct specific regulatory models from identified regulatory elements in their promoter sequence. Potentially co-regulated genes were deduced from these models. Then they can be exported to our MPSA platform. This whole development will be included in the project of Integrative Biology platform in the INRA Human Nutrition Unit.

As células produtoras de anticorpos desempenham um papel chave na resposta efetora a microrganismos, sendo o foco principal da maioria das vacinas existentes. Entretanto, elas podem ter efeitos deletérios em doenças autoimunes e na rejeição a transplantes. Apesar de grandes avanços no controle da resposta humoral, alguns desafios permanecem e neste contexto, novos alvos terapêuticos têm sido explorados. Sabe-se que alterações metabólicas decorrentes da ativação de células B estão intimamente relacionadas com a função efetora destas células, o que anteriormente se imaginava ser apenas um reflexo de crescimento e proliferação celular. Estas alterações são controladas por sensores metabólicos ativados logo após a ativação de células B, como o mTOR, o qual é componente central de dois complexos: mTORC1 e mTORC2. Estudos anteriores já reportaram o papel positivo exercido por mTOR na via glicolítica em células T, bem como na função efetora destas células. Aqui, nós formulamos a hipótese de que a via do mTOR favorece a via glicolítica em detrimento de OXPHOS em células B, e que estas alterações metabólicas impactam as funções efetoras das mesmas. Desta maneira, para investigarmos alterações nestas vias em decorrência de mTOR, células B foram isoladas de animais controle (CT), ou de animais com células B deficientes de mTORC1 (RaptorΔB) ou mTORC2 (RictorΔB) e então estimuladas com LPS (lipopolissacarídeo) in vitro. Nossos dados indicam que a deficiência de mTORC2 beneficia OXPHOS em detrimento da via gicolítica, bem como a ativação de células B e a formação de plasmablastos. Na sequência, confirmamos que a redução nas taxas de glicólise, assim como a elevação da oxidação lipídica e de OXPHOS são cruciais para manter a elevada ativação de células B e formação de plasmablastos a partir de células B RaptorΔB e RictorΔB. Constatou-se ainda que a produção total de IgM é elevada em células RictorΔB após estímulo com LPS. Entretanto, identificamos que isso é decorrente do aumento de plasmablastos formados e não da capacidade individual de secreção dos mesmos. Diferentemente das células B deficientes, observamos que plasmablastos RaptorΔB e RictorΔB reduzem a atividade mitocondrial. Na sequência, confirmamos que a atividade mitocondrial via oxidação lipídica é fundamental para a produção de anticorpos. Além disso, demonstramos que a deficiência de mTORC2 eleva a troca de isotipo, enquanto a de mTORC1 a diminui após estimulo com LPS e IL4. Posteriormente, o impacto da deficiência de mTORC2 em células B foi avaliado in vivo em modelo de transplante de pele. Neste caso, não observamos diferenças significativas na sobrevida do enxerto entre CT e RictorΔB, mas foi constatado que apesar de ambos apresentarem formação de plasmócitos similares, animais deficientes apresentaram um número significativamente menor de células B na periferia. Assim, concluímos que a deficiência de mTORC1 ou mTORC2 em células B implica em uma maior diferenciação de plasmablastos e em uma maior produção total de anticorpos em RictorΔB in vitro, enquanto um papel funcional para essas moléculas no contexto das células B in vivo ainda precise ser determinado. / Antibodies are produced by Antibody Secreting Cells (ASCs), which are essential to fight infections. They are also the basis of most successful vaccines available, however they can present deleterious effects in autoimmune diseases and in graft rejection. Even though there have been great improvements in controlling the humoral response, its proper manipulation still remains a challenge, thus new targets need to be explored. It is known that metabolic shifts that occur upon B cell activation are not only essential for cell growth and proliferation, but are also interconnected with these cells effector function. Metabolic shifts are controlled by metabolic sensors, as the mTOR, which is a core component of two complexes, mTORC1 and mTORC2. Previous studies with T cells have already reported that mTOR exerts a positive role on glycolysis, which in turn impacts the effector function of T cells. We then hypothesized that mTOR favors glycolysis over Oxidative Phosphorylation (OXPHOS) in B cells, and that these metabolic changes impact the effector function of B cells. Thus, to investigate the impact of the mTOR pathway on B cells, we isolated B cells from mice with mTORC1 deficient B cells (RaptorΔB) or mTORC2 deficient B cells (RictorΔB) or Control mice (CT), and stimulated them in vitro with lipopolysaccharide (LPS). Our results indicate that the deficiency of mTORC2 favors OXPHOS over glycolysis, as well as B cell activation markers expression and plasmablast formation. Next, we confirmed that the reduced glycolysis levels, improved lipid oxidation and OXPHOS are in fact crucial for the enhanced activation and plasmablast formation observed in RaptorΔB and RictorΔB B cells. We also described that IgM secretion was elevated in B cells from RictorΔB after stimuli with LPS, however we found that this increase was due to the overall increase in plasmablasts in this group and not to their individual antibody secretion capacity. Interestingly, RaptorΔB and RictorΔB plasmablasts differently from B cells, reduce their mitochondrial activity. Subsequently, we confirmed that the mitochondria via lipid oxidation is actually essential for antibody secretion. In addition, we showed that mTORC2 deficiency increases isotype switching, while mTORC1 deficiency diminishes it when IL4 was added to LPS. We then sought to determine if mTORC2 deficiency in B cells would present an impact in vivo in a skin graft model. However, we did not observe a significant difference in graft survival between the CT and RictorΔB mice and in plasmacyte numbers, but we did observe a significant reduction in B cells in the periphery. Thus, we conclude that mTORC1 and mTORC2 deficiency leads to an improved plasmablast differentiation and an overall increase in antibody secretion in the last one in vitro, whereas the role of these sensors remains to be determined in vivo.

The MYC onocogene is frequently overexpressed in human cancer due to its capacity to promote cell growth and cell proliferation. MYC overexpression activates the p53 tumour suppressor pathway, which resists the pro-tumourigeneic program elicited by MYC. How MYC overexpression engages p53 is yet to be elucidated, and in this study I carried out a large metabolic siRNA screen to determine whether p53 responds to a specific MYC-driven metabolic pathway. Two clear lipid metabolic pathways emerged from the siRNA screen: PPARγ/arachidonate metabolism and de novo sphingolipid synthesis. Knockdown or inhibition of PPARγ increased p53 levels, and PPARγ ligands decreased following MYC overexpression. Knockdown of ceramide synthesis depleted p53 levels, and MYC overexpression increased de novo ceramide synthesis. This demonstrated that MYC-driven ceramide synthesis positively regulates p53, and highlights the role of cell metabolism in the tumour suppressor response to MYC deregulation.

Background: The unicellular green alga Chlamydomonas reinhardtii is an important eukaryotic model organism for the study of photosynthesis and plant growth. In the era of modern highthroughput technologies there is an imperative need to integrate large-scale data sets from highthroughput experimental techniques using computational methods and database resources to provide comprehensive information about the molecular and cellular organization of a single organism. Results: In the framework of the German Systems Biology initiative GoFORSYS, a pathway database and web-portal for Chlamydomonas (ChlamyCyc) was established, which currently features about 250 metabolic pathways with associated genes, enzymes, and compound information. ChlamyCyc was assembled using an integrative approach combining the recently published genome sequence, bioinformatics methods, and experimental data from metabolomics and proteomics experiments. We analyzed and integrated a combination of primary and secondary database resources, such as existing genome annotations from JGI, EST collections, orthology information, and MapMan classification. Conclusion: ChlamyCyc provides a curated and integrated systems biology repository that will enable and assist in systematic studies of fundamental cellular processes in Chlamydomonas. The ChlamyCyc database and web-portal is freely available under http://chlamycyc.mpimp-golm.mpg.de.

Metabolic networks are the collections of all cellular activities taking place in a living cell and all the relationships among biological elements of the cell including genes, proteins, enzymes, metabolites, and reactions. They provide a better understanding of cellular mechanisms and phenotypic characteristics of the studied organism. In order to reconstruct a metabolic network, interactions among genes and their molecular attributes along with their functions must be known. Using this information, proteins are distributed among pathways as sub-networks of a greater metabolic network. Proteins which carry out various steps of a biological process operate in same pathway.<p>The metabolic network of <i>Caenorhabditis elegans</i> was reconstructed based on current genomic information obtained from the KEGG database, and commonly found in SWISS-PROT and WormBase. Assuming proteins operating in a pathway are interacting proteins, currently available protein-protein interaction map of the studied organism was assembled. This map contains all known protein-protein interactions collected from various sources up to the time. Topology of the reconstructed network was briefly studied and the role of key enzymes in the interconnectivity of the network was analysed. The analysis showed that the shortest metabolic paths represent the most probable routes taken by the organism where endogenous sources of nutrient are available to the organism. Nonetheless, there are alternate paths to allow the organism to survive under extraneous variations. <p>Signature content information of proteins was utilized to reveal protein interactions upon a notion that when two proteins share signature(s) in their primary structures, the two proteins are more likely to interact. The signature content of proteins was used to measure the extent of similarity between pairs of proteins based on binary similarity score. Pairs of proteins with a binary similarity score greater than a threshold corresponding to confidence level 95% were predicted as interacting proteins. The reliability of predicted pairs was statistically analyzed. The sensitivity and specificity analysis showed that the proposed approach outperformed maximum likelihood estimation (MLE) approach with a 22% increase in area under curve of receiving operator characteristic (ROC) when they were applied to the same datasets. When proteins containing one and two known signatures were removed from the protein dataset, the area under curve (AUC) increased from 0.549 to 0.584 and 0.655, respectively. Increase in the AUC indicates that proteins with one or two known signatures do not provide sufficient information to predict robust protein-protein interactions. Moreover, it demonstrates that when proteins with more known signatures are used in signature profiling methods the overlap with experimental findings will increase resulting in higher true positive rate and eventually greater AUC. <p>Despite the accuracy of protein-protein interaction methods proposed here and elsewhere, they often predict true positive interactions along with numerous false positive interactions. A global algorithm was also proposed to reduce the number of false positive predicted protein interacting pairs. This algorithm relies on gene ontology (GO) annotations of proteins involved in predicted interactions. A dataset of experimentally confirmed protein pair interactions and their GO annotations was used as a training set to train keywords which were able to recover both their source interactions (training set) and predicted interactions in other datasets (test sets). These keywords along with the cellular component annotation of proteins were employed to set a pair of rules that were to be satisfied by any predicted pair of interacting proteins. When this algorithm was applied to four predicted datasets obtained using phylogenetic profiles, gene expression patterns, chance co-occurrence distribution coefficient, and maximum likelihood estimation for S. cerevisiae and <i>C. elegans</i>, the improvement in true positive fractions of the datasets was observed in a magnitude of 2-fold to 10-fold depending on the computational method used to create the dataset and the available information on the organism of interest. <p>The predicted protein-protein interactions were incorporated into the prior reconstructed metabolic network of <i>C. elegans</i>, resulting in 1024 new interactions among 94 metabolic pathways. In each of 1024 new interactions one unknown protein was interacting with a known partner found in the reconstructed metabolic network. Unknown proteins were characterized based on the involvement of their known partners. Based on the binary similarity scores, the function of an uncharacterized protein in an interacting pair was defined according to its known counterpart whose function was already specified. With the incorporation of new predicted interactions to the metabolic network, an expanded version of that network was resulted with 27% increase in the number of known proteins involved in metabolism. Connectivity of proteins in protein-protein interaction map changed from 42 to 34 due to the increase in the number of characterized proteins in the network.

Metabolic networks are the collections of all cellular activities taking place in a living cell and all the relationships among biological elements of the cell including genes, proteins, enzymes, metabolites, and reactions. They provide a better understanding of cellular mechanisms and phenotypic characteristics of the studied organism. In order to reconstruct a metabolic network, interactions among genes and their molecular attributes along with their functions must be known. Using this information, proteins are distributed among pathways as sub-networks of a greater metabolic network. Proteins which carry out various steps of a biological process operate in same pathway.<p>The metabolic network of <i>Caenorhabditis elegans</i> was reconstructed based on current genomic information obtained from the KEGG database, and commonly found in SWISS-PROT and WormBase. Assuming proteins operating in a pathway are interacting proteins, currently available protein-protein interaction map of the studied organism was assembled. This map contains all known protein-protein interactions collected from various sources up to the time. Topology of the reconstructed network was briefly studied and the role of key enzymes in the interconnectivity of the network was analysed. The analysis showed that the shortest metabolic paths represent the most probable routes taken by the organism where endogenous sources of nutrient are available to the organism. Nonetheless, there are alternate paths to allow the organism to survive under extraneous variations. <p>Signature content information of proteins was utilized to reveal protein interactions upon a notion that when two proteins share signature(s) in their primary structures, the two proteins are more likely to interact. The signature content of proteins was used to measure the extent of similarity between pairs of proteins based on binary similarity score. Pairs of proteins with a binary similarity score greater than a threshold corresponding to confidence level 95% were predicted as interacting proteins. The reliability of predicted pairs was statistically analyzed. The sensitivity and specificity analysis showed that the proposed approach outperformed maximum likelihood estimation (MLE) approach with a 22% increase in area under curve of receiving operator characteristic (ROC) when they were applied to the same datasets. When proteins containing one and two known signatures were removed from the protein dataset, the area under curve (AUC) increased from 0.549 to 0.584 and 0.655, respectively. Increase in the AUC indicates that proteins with one or two known signatures do not provide sufficient information to predict robust protein-protein interactions. Moreover, it demonstrates that when proteins with more known signatures are used in signature profiling methods the overlap with experimental findings will increase resulting in higher true positive rate and eventually greater AUC. <p>Despite the accuracy of protein-protein interaction methods proposed here and elsewhere, they often predict true positive interactions along with numerous false positive interactions. A global algorithm was also proposed to reduce the number of false positive predicted protein interacting pairs. This algorithm relies on gene ontology (GO) annotations of proteins involved in predicted interactions. A dataset of experimentally confirmed protein pair interactions and their GO annotations was used as a training set to train keywords which were able to recover both their source interactions (training set) and predicted interactions in other datasets (test sets). These keywords along with the cellular component annotation of proteins were employed to set a pair of rules that were to be satisfied by any predicted pair of interacting proteins. When this algorithm was applied to four predicted datasets obtained using phylogenetic profiles, gene expression patterns, chance co-occurrence distribution coefficient, and maximum likelihood estimation for S. cerevisiae and <i>C. elegans</i>, the improvement in true positive fractions of the datasets was observed in a magnitude of 2-fold to 10-fold depending on the computational method used to create the dataset and the available information on the organism of interest. <p>The predicted protein-protein interactions were incorporated into the prior reconstructed metabolic network of <i>C. elegans</i>, resulting in 1024 new interactions among 94 metabolic pathways. In each of 1024 new interactions one unknown protein was interacting with a known partner found in the reconstructed metabolic network. Unknown proteins were characterized based on the involvement of their known partners. Based on the binary similarity scores, the function of an uncharacterized protein in an interacting pair was defined according to its known counterpart whose function was already specified. With the incorporation of new predicted interactions to the metabolic network, an expanded version of that network was resulted with 27% increase in the number of known proteins involved in metabolism. Connectivity of proteins in protein-protein interaction map changed from 42 to 34 due to the increase in the number of characterized proteins in the network.

No description available.

No description available.

Orientador: Guilherme Targino Valente / Resumo: Enzimas são uma classe de proteínas responsáveis por catalisar diversos tipos de reações químicas presentes em diferentes rotas metabólicas, sendo assim o principal foco de estudo nas áreas de engenharia metabólica e biologia sintética. Contudo, a anotação de enzimas e a identificação da rota metabólica em que atuam, são frequentemente baseados na similaridade de sequências previamente descritas. A falta e dificuldade de anotação das enzimas se devem pela diversidade funcional em sequências similares de famílias proteicas, sequências espécie-específicas e a dificuldade na definição de homologia em larga escala. De modo a auxiliar a superar tais problemas, o presente trabalho objetivou criar um classificador de rotas metabólicas de enzimas baseado inteiramente nas características da estrutura primária de enzimas e utilizando aprendizado de máquina. A ferramenta computacional criada (mAppLe - Metabolic Pathway Prediction of Enzymes) é composta por 11 preditores de rotas metabólicas de fungos, podendo assim auxiliar nas anotações dos bancos de dados e em trabalhos nas diferentes áreas de pesquisa, como biologia sintética e engenharia metabólica. As performances médias de predição foram de 94% de acurácia, 44% de taxa de falsa descoberta, 67% de F-​ score , ​ 98% de sensitividade, 93% de especificidade e 0,69 para coeficiente de correlação de Matthews​ . Com base no desempenho dos preditores criados, constata-se que a ferramenta computacional criada pode ser aplicada com grande s... (Resumo completo, clicar acesso eletrônico abaixo) / Abstract: Enzymes are a class of proteins that are responsible for catalyzing chemical reactions in numerous metabolic pathways and are often "main targets" in metabolic engineering and synthetic biology. However, enzyme annotation and metabolic pathway identifications are often based on sequence similarities to previously well-described enzymes. Functional diversity in similar sequences of protein families, species-specificity, and difficult-to-define large-scale homologies results in difficulties and a lack of annotation. Here, we present the mAppLe (Metabolic Pathway Prediction of Enzymes), the first metabolic pathway classifier for enzymes based only on primary structure features and a machine learning approach, surpassing limitations imposed by sequence similarities. This tool is composed of 11 pathways predictors for fungi, that can help databank annotations and several type of researches like synthetic biology and metabolic engineering. Results show an average performance of 94% to accuracy, 44% false discovery rate, 67% F-score, 98% sensitivity, 93% specificity and 0.69 to Matthews coefficient correlation. Based on the performance of this predictors, the computational tool created (mAppLe) can be applied successfully to predict pathways of enzymes of the fungi, independent of sequence similarity. / Doutor

Senescence is a cell stress program characterized by a stable cell cycle arrest and thus aims to protect against replication of potentially harmful cells. In oncogene-induced senescence (OIS) the cell cycle arrest is brought about by activation of an oncogene. This in turn initiates a DNA damage response and subsequently, the DDR induces p53-p21 and RB tumour suppressor pathways. The metabolism of senescent cells is highly altered, notably there is increased secretion of proteins and increased functional activity of certain metabolic enzymes. There have been many recent studies investigating the role of specific metabolic pathways in OIS and how they may be targeted for therapeutic benefit. This thesis aims to identify novel metabolic regulators of OIS, by combining high throughput RNAi screening and LC-MS based methods. This thesis has identified and validated 17 essential OIS metabolic genes; in this list, there was enrichment for genes involved in lipid biosynthetic processes. Lipid metabolism was an attractive focus for this thesis as it has not been extensively studied in current literature. Next, ceramide synthase 4 (CERS4) was extensively validated as a key enzyme for both OIS and replicative senescence. Using LC-MS based lipidomics, CERS4-driven rewiring of lipid metabolism in OIS was revealed and this corresponded with an accumulation of ceramides due to increased de novo ceramide synthesis. It was then confirmed OIS-related ceramide is mechanistically linked to cell cycle via the PP1-RB-E2F axis. Ceramide activates PP1, which physically binds to RB in a CERS4-dependent manner. PP1 is then able to dephosphorylate and activate RB, which inhibits transcription of E2F targets (cell cycle genes). Overall, this thesis identifies a metabolic checkpoint that links altered lipid metabolism with OIS.