<p>Transcriptional regulation has been studied intensively in recent decades. One important aspect of this regulation is the interaction between regulatory proteins, such as transcription factors (TF) and nucleosomes, and the genome. Different high-throughput techniques have been invented to map these interactions genome-wide, including ChIP-based methods (ChIP-chip, ChIP-seq, etc.), nuclease digestion methods (DNase-seq, MNase-seq, etc.), and others. However, a single experimental technique often only provides partial and noisy information about the whole picture of protein-DNA interactions. Therefore, the overarching goal of this dissertation is to provide computational developments for jointly modeling different experimental datasets to achieve a holistic inference on the protein-DNA interaction landscape. </p><p>We first present a computational framework that can incorporate the protein binding information in MNase-seq data into a thermodynamic model of protein-DNA interaction. We use a correlation-based objective function to model the MNase-seq data and a Markov chain Monte Carlo method to maximize the function. Our results show that the inferred protein-DNA interaction landscape is concordant with the MNase-seq data and provides a mechanistic explanation for the experimentally collected MNase-seq fragments. Our framework is flexible and can easily incorporate other data sources. To demonstrate this flexibility, we use prior distributions to integrate experimentally measured protein concentrations. </p><p>We also study the ability of DNase-seq data to position nucleosomes. Traditionally, DNase-seq has only been widely used to identify DNase hypersensitive sites, which tend to be open chromatin regulatory regions devoid of nucleosomes. We reveal for the first time that DNase-seq datasets also contain substantial information about nucleosome translational positioning, and that existing DNase-seq data can be used to infer nucleosome positions with high accuracy. We develop a Bayes-factor-based nucleosome scoring method to position nucleosomes using DNase-seq data. Our approach utilizes several effective strategies to extract nucleosome positioning signals from the noisy DNase-seq data, including jointly modeling data points across the nucleosome body and explicitly modeling the quadratic and oscillatory DNase I digestion pattern on nucleosomes. We show that our DNase-seq-based nucleosome map is highly consistent with previous high-resolution maps. We also show that the oscillatory DNase I digestion pattern is useful in revealing the nucleosome rotational context around TF binding sites. </p><p>Finally, we present a state-space model (SSM) for jointly modeling different kinds of genomic data to provide an accurate view of the protein-DNA interaction landscape. We also provide an efficient expectation-maximization algorithm to learn model parameters from data. We first show in simulation studies that the SSM can effectively recover underlying true protein binding configurations. We then apply the SSM to model real genomic data (both DNase-seq and MNase-seq data). Through incrementally increasing the types of genomic data in the SSM, we show that different data types can contribute complementary information for the inference of protein binding landscape and that the most accurate inference comes from modeling all available datasets. </p><p>This dissertation provides a foundation for future research by taking a step toward the genome-wide inference of protein-DNA interaction landscape through data integration.</p> / Dissertation

O desenvolvimento do padrão de segmentação de Drosophila é estabelecido por uma cascata de genes de segmentação zigóticos. Estes genes são divididos em três classes (gap, pair-rule e segment-polarity) e codificam para fatores de transcrição (FT) que se ligam a módulos cis-reguladores (CRMs), reprimindo ou ativando genes alvo. A faixa 2 do gene pair-rule even-skipped (eve 2) é ativada pelos fatores maternos Bicoid e Hunchback e reprimida pelas proteínas gap Giant (Gt) e Krüppel. Estudos posteriores mostraram que o FT Sloppy-paired 1 (Slp 1) e provavelmente um outro FT forkhead também atuam na repressão de eve 2. O gene anotado CG9571 foi isolado em uma varredura como proteína forkhead candidata a repressão de eve 2. Estudos genéticos confirmaram essa possibilidade e revelaram que eve 1 também pode ser alvo deste FT. Este trabalho teve como objetivo verificar a interação de CG9571 com os CRMs eve 1 e eve 2. Para tanto, planejamos obter o domínio de ligação da proteína (CG9571 BD) e da proteína completa (CG9571 FL) e testar suas interações in vitro com fragmentos dos CRMs por meio da técnica de retardo da mobilidade eletroforética (EMSA). Obtivemos a quantidade necessária de DNA para os experimentos através de PCR e preparações plasmidiais de versões clonadas destes CRMs que já dispúnhamos em laboratório. Realizamos tentativas de obtenção de CG9571 BD por transcrição e tradução in vitro, mas esta estratégia não foi bem-sucedida e adotamos a estratégia de clonagem em vetor para expressão em células competentes bacterianas. O fragmento de CG9571 BD foi clonado com sucesso, mas não conseguimos verificar a expressão do polipeptídeo em duas linhagens de E. coli. Alteramos novamente nossa estratégia e clonamos o fragmento correspondente a CG9571 FL em vetor de expressão e conseguimos induzir sua expressão em bactéria, embora não tenha sido obter a proteína recombinante em forma solúvel. Prosseguimos para tentativas de recuperação da proteína a partir de corpos de inclusão. Foram empregados diferentes métodos para solubilização, renovelamento e purificação da proteína. Extratos da fração insolúvel solubilizada em diferentes concentrações de ureia foram submetidos a tentativas de purificação e renaturação por cromatografia de afinidade, mas não houve adsorção significativa de CG9571 FL em colunas com Ni2+ imobilizado. Preparações não puras contendo CG9571 FL foram obtidas através de procedimentos de renaturação destes extratos e foram utilizadas em ensaios de interação com os CRMs. Não houve detecção de retardo da mobilidade eletroforética dos fragmentos em gel. Foram observados efeitos de redução da quantidade de DNA detectado com brometo de etídio nas interações, mas este efeito foi considerado produto da ação de possíveis nucleases contaminantes nas preparações após investigação. Preparações de CG9571 FL puras foram obtidas por purificações a partir de SDS-PAGE, mas a maioria das interações da proteína solúvel com eve 1 e eve 2 não indicou formação de complexo. Obtivemos um único resultado positivo para a interação entre CG9571 FL e eve 2. Por não ter sido reproduzido, consideramos o resultado inconclusivo e novos experimentos serão conduzidos para dar continuidade à investigação da hipótese do trabalho / The development of Drosophila segmentation pattern is established by a cascade of zygotic segmentation genes. The zygotic genes are grouped in three classes (gap, pair-rule and segment-polarity) and code for transcription factors (TF) that bind to cis-regulatory modules (CRMs) with activation or repression roles. The stripe 2 of the pair-rule gene even-skipped (eve 2) is activated by the maternal factors Bicoid and Hunchback and repressed by the gap proteins Giant and Krüppel. Later studies showed that Sloppy-paired 1 (Slp 1) and probably another forkhead transcription factor also act for eve 2 repression. The annotated gene CG9571 was isolated in a search for putative forkhead protein repressors of eve 2. Genetic studies confirmed this possibility and reveled that eve 1 could also be a target for this TF. The aim of this work was to verify the interaction of CG9571 with the CRMs eve 1 and eve 2. To reach this aim, we planned to obtain the binding domain of the protein (CG9571 BD) or of the full-length protein (CG9571 FL) and to test their in vitro interactions with the eve 1 and eve 2 fragments by the electrophoretic mobility shift assay (EMSA). We obtained the necessary amount of DNA for the tests by PCR and plasmidial preparations of cloned versions of these CRMs that we already had in our laboratory. We made attempts to obtain CG9571 BD by in vitro transcription and translation system, but this strategy did not work and we adopted the cloning strategy to obtain the protein expressed by bacterial competent cells. CG9571 BD was cloned successfully, but we were not able to detect the polypeptide expression in two E. coli strains. We then turned to the CG9571 FL protein that we cloned and succeed to express it in bacteria, although we were not able to obtain the soluble recombinant form. We proceed for attempts of protein recovering from inclusion bodies. Different methods for solubilization, refolding and purification of the proteins were used. Extracts of the insoluble fraction solubilized in solutions with different urea concentrations were used in attempts of refolding and purification by affinity chromatography, but there was not significant CG9571 FL adsorption on columns with Ni2+ immobilized. We obtained impure preparations with CG9571 FL through procedures of refolding of these extracts and employed them on binding assays with the CRMs, but there was no gel shift detection. We observed reduction of the amount of DNA present in the binding reaction samples detected by ethidium bromide, but after further investigations this effect was considered the product of contaminant nuclease reaction from bacteria. Pure preparations of CG9571 FL were obtained by purification from SDS-PAGE, but there was no indication of complex formation on the most binding reaction assays with eve 1 and eve 2. We obtained only one positive result for the interaction between CG9571 FL and eve 2. However, the result was considered inconclusive because we were not able to reproduce it and new experiments will be conducted to investigate the hypothesis of this work

Diss. (sammanfattning) Stockholm : Karol. inst., 2003. / Härtill 8 uppsatser.

No description available.

Helicobacter pylori est une bactérie à Gram-négatif qui colonise la muqueuse de l’estomac humain. Elle se distingue des autres bactéries par un nombre de gènes très limité et de nombreuses particularités physiologiques et biochimiques. Elle provoque des infections associées à différentes maladies gastro-duodénales (ulcères et cancers). Depuis quelques années, une recrudescence de multi-résistances aux antibiotiques est observée. La transformation naturelle est l’un des processus clés qui les propage. Il s’agit d’un mécanisme de transfert horizontal de gènes qui permet aux bactéries de s’adapter à leur environnement, en internalisant des fragments d’ADN exogène à travers leur membrane, puis en les intégrant dans le chromosome par recombinaison homologue. Mes travaux ont visé à étudier de façon structurale et fonctionnelle trois protéines d’H. pylori décrites comme étant essentielles dans le processus de transformation naturelle: NucT, DprA et ComFc. La première partie de ce travail s’est concentrée sur la nucléase périplasmique NucT, supposée être impliquée dans la transformation chez H. pylori. Cependant, la délétion de son gène a permis de démontrer qu’elle ne joue en fait qu’un rôle mineur dans ce processus. La résolution de sa structure 3D a permis de mieux comprendre sa spécificité pour les acides nucléiques simple brin. Dans la seconde partie, la protéine DprA, responsable du chargement de la recombinase RecA sur l’ADN internalisé, a été étudiée. DprA d’H. pylori n’est composée que de 2 des 3 domaines qui constituent habituellement DprA, et fixe aussi bien l’ADN double brin que l’ADN simple brin mais uniquement via son domaine RF. Malgré son homologie structurale avec le domaine WH de liaison à l’ADN, le domaine C-terminal de HpDprA n’a pas d’affinité pour l’ADN. Nous avons mis en évidence des acides aminés conservés dans ce domaine dont l’étude pourrait permettre de comprendre son rôle. Enfin, une étude structurale de la protéine ComFc dont la délétion du gène entraîne la disparition totale de la capacité de transformation d’H. pylori a été réalisée. L’obtention de sa structure 3D a permis de mettre en évidence la présence d’un domaine catalytique phosphoribosyl-transférase ainsi que d’un domaine en doigt en zinc. Ce dernier pourrait être responsable de la capacité de ComFc à fixer l’ADN. Le substrat naturel de cette enzyme reste à découvrir.L’ensemble de ce travail a permis de contribuer à une meilleure compréhension à l’échelle moléculaire du mécanisme de transformation génétique naturelle d’H. pylori. L’avancement sur ces connaissances pourrait à long terme aider à réduire la propagation des multi-résistances par l’élaboration de nouvelles thérapies.Mots-clés : H. pylori, transformation naturelle, NucT, DprA, ComFc, interaction protéine-ADN / Helicobacter pylori is a Gram-negative bacterium that colonizes the mucus of the human stomach. It is distinguished from other bacteria by a limited number of genes and many physiological and biochemical characteristics. It causes infections associated with various gastro-duodenal diseases (ulcers and gastric cancers). In recent years, an increase in multi-resistance to antibiotics has been observed. Natural transformation is one of the key processes that spreads these multi-resistances. It is a horizontal gene transfer mechanism that allows bacteria to adapt to their environment by internalizing exogenous DNA fragments through their membrane and then integrating them into the chromosome by homologous recombination. My work aimed to study in a structural and functional approach three proteins of H. pylori described as essential in the natural transformation process: NucT, DprA and ComFc. The first part of this work focused on periplasmic nuclease, NucT, which is supposed to be involved in transformation in H. pylori. However, the deletion of its gene has shown that it actually plays only a minor role in this process. The resolution of its 3D structure has led to a better understanding of its specificity for single-stranded nucleic acids. In the second part, the protein DprA, responsible for loading RecA recombinase onto internalized DNA, was studied. HpDprA is composed of only 2 of the 3 domains that usually constitute DprA, and binds both double-stranded and single-stranded DNA but only via its RF domain. Despite its structural homology with the WH DNA binding domain, the C-terminal domain of HpDprA has no affinity for DNA. We have identified conserved amino acids in this domain that could be studied to understand its role. Finally, a structural study of ComFc, whose deletion of the gene leads to the total disruption of the transformation capacity of H. pylori, has been carried out. The acquisition of its 3D structure has highlighted the presence of a phosphoribosyl transferase catalytic domain as well as a zinc finger domain. The latter could be responsible for capacity of ComFc to bind DNA. The natural substrate of this enzyme remains to be discovered.All this work has contributed to a better knowledge at the molecular level of the natural genetic transformation mechanism of H. pylori. Advancing this knowledge could in the long term help to reduce the spread of multiresistance through the development of new therapies.Keywords: Helicobacter pylori, natural transformation, NucT, DprA, ComFc, protein-DNA interaction

No description available.

No description available.

No description available.

This dissertation focuses on rationalised DNA design as a tool for the discovery and development of new therapeutic entities, as well as understanding the biological function of DNA beyond the storage of genetic information. The study is comprised of two main areas of study: (i) the use of DNA as a coding unit to illustrate the relationship between code-diversity and dynamics of self-assembly; and (ii) the use of DNA as an active unit that interacts and regulates a target protein. In the study of DNA as a coding unit in code-diversity and dynamics of self-assembly, we developed the DNA-Based Diversity Modelling and Analysis (DDMA) method. Using Polymerase Chain Reaction (PCR) and Real Time Polymerase Chain Reaction (RT-PCR), we studied the diversity and evolution of synthetic oligonucleotide populations. The manipulation of critical conditions, with monitoring and interpretation of their effects, lead to understanding how PCR amplification unfolding could reshape a population. This new take on an old technology has great value for the study of: (a) code-diversity, convenient in a DNA-based selection method, so semi-quantitation can evaluate a selection development and the population\'s behaviour can indicate the quality; (b) self-assembly dynamics, for the simulation of a real evolution, emulating a society where selective pressures direct the population's adaptation; and (c) development of high-entropy DNA structures, in order to understand how similar unspecific DNA structures are formed in certain pathologies, such as in auto-immune diseases. To explore DNA as an active unit in Tumour Necrosis Factor α (TNF-α) interaction and activity modulation, we investigate DNA's influence on its spatial conformation by physical environment regulation. Active TNF-α is a trimer and the protein-protein interactions between its monomers are a promising target for drug development. It has been hypothesised that TNF-α forms a very intricate network after its activation between its subunits and receptors, but the mechanism is still not completely clear. During our research, we estimate the non-specific DNA binding to TNF-α in the low micro-molar range. Cell toxicity assays confirm this interaction, where DNA consistently enhances TNF-α's cytotoxic effect. Further binding and structural studies lead to the same conclusion that DNA binds and interferes with TNF-α structure. From this protein-DNA interaction study, a new set of tools to regulate TNF-α's biological activity can be developed and its own biology can be unveiled.

Os projetos genoma têm fornecido uma grande quantidade de informação sobre a arquitetura gênica e sobre a configuração física de suas respectivas regiões flanqueadoras (RF). Estas RF contêm informações com o potencial de auxiliar na elucidação de vários processos biológicos, como os mecanismos de expressão gênica e de sua regulação. Estes mecanismos são de extrema importância para a compreensão do correto funcionamento dos organismos e das patologias que os afetam. Uma parte significativa dos mecanismos de controle de expressão gênica atuam na fase transcricional. Na base destes mecanismos está o recrutamento de proteínas que se ligam às regiões promotoras da transcrição, as quais são segmentos específicos de DNA que podem estar localizados tanto próximos à região de início da transcrição (TSS) quanto a centenas ou até a milhares de pares de bases dela. Essas proteínas compõem a maquinaria transcricional e podem ativar ou inibir o processo de transcrição. Experimentalmente, os segmentos regulatórios podem ser identificadas utilizando métodos complexos de biologia molecular, tais como SELEX, ChiP-ChiP, ChIP-Seq, dentre outros. Uma estratégia alternativa aos métodos experimentais é a utilização de metodologias computacionais. Análises computacionais tendem a ser mais rápidas, baratas e flexíveis do que protocolos experimentais, além de poderem ser utilizadas em larga escala. Atualmente, os métodos computacionais disponíveis necessitam de informações experimentais para a definição de padrões globais de preferências de sequências de DNA para a ligação de fatores de transcrição (TFBS, em inglês transcription factor binding sites). Entretanto, esses métodos apresentam uma elevada taxa de falso positivos e, por vezes, apresentam também taxas significativas de falso negativos, além de serem limitados ao estudo de fatores de transcrição de espécies bem conhecidas, o que diminui a área de aplicação dos mesmos. Diante deste cenário, o uso de métodos computacionais que não necessitem da informação referente aos sítios de ligação, bem como os que utilizem parâmetros mais robustos de detecção dos resultados, em detrimento dos escores de pontuação provindos de alinhamentos, podem acrescentar uma sensível melhoria ao processos de predição de regiões regulatórias. Neste projeto, foi desenvolvido um novo modelo computacional (TFBSAnalyzer) para análise e identificação de TFBS em elementos regulatórios, que utiliza técnicas de modelagem molecular para a construção de complexos entre um fator de transcrição ancorado a estruturas de DNA com sequências variáveis de bases e, através de cálculos termodinâmicos de entalpia de ligação, determina uma função de pontuação baseada na energia de ligação e realiza a predição de sítios de ligação ao DNA para o fator de transcrição em análise. Esta abordagem foi testada com três fatores de transcrição como sistemas-modelo, pertencentes à família dos receptores nucleares, a saber: o receptor de estrógeno ER-alfa (Estrogen Receptor Alpha), o receptor de ácido retinoico RAR-beta (Retinoid Acid Receptor Beta) e o receptor X retinóico RXR (Retinoid X Receptor). Os modelos previstos computacionalmente foram comparados aos dados experimentais disponíveis para estes receptores nucleares, os quais apresentaram as seguintes taxas de FP/FN: 10%/0 para RAR-beta e RXR, 21%/6% para ER-alfa. Também simulamos um experimento de ChIP-seq do ER-alfa no genoma humano, cujos genes selecionados foram submetidos a uma análise de enriquecimento de fatores de transcrição curados experimentalmente, que fazem sua regulação, revelando que o receptor de estrógeno está realmente envolvido no processo. Para mostrar a aplicabilidade geral de nosso método, nós modelamos a distribuição de energia de ligação para o receptor NHR-28 isoforma a de Caenorhabditis elegans com DNA . Obtivemos distribuições de energia semelhantes àquelas encontradas para os NRs modelos, portanto seria possível aplicar o método para buscar possíveis TFBSs para este receptor no genoma de C. elegans. Os dados gerados e as metodologias desenvolvidas neste projeto devem acrescentar uma sensível melhoria aos processos de predição de regiões regulatórias e consequentemente auxiliar no entendimento dos mecanismos envolvidos no processo de expressão gênica e de sua regulação. / The genome projects have provided a lot of information about the genetic architecture, as well as on the physical configuration of their flanking regions (FR). These FR have the potential to aid in the elucidation of many biological processes, such as the mechanisms involved in gene expression and its regulation. These mechanisms are extremely important for undeerstanfind the correct functioning of organisms as well as the pathologies that affect them. A significant part of the control mechanisms of gene expression act during transcription. On the basis of this mechanisms is the recruitment of proteins that bind to promoter regions of transcription, which are specific segments of DNA that can be located either near the transcription start site or at hundreds or even thousands of base pairs away. These proteins form the transcription machinery, which can activate or inhibit the transcription process. The regulatory segments can be identified experimentally using complex methods of molecular biology, such as SELEX, ChIP-chip, ChIP-seq, among others. An alternative strategy to these experimental methods is the use of computational methodologies for predicting regulatory regions. Computational analysis tend to be faster, cheaper and more flexible than the experimental protocols, and can be used on a larger scale. Currently, the available computational methods require information previously obtained from experiments in order to define global standards of preference of DNA-Binding sequences for transcription factors (TFBS - Transcription Factor Binding Sites). However, these methods have a high rate of false positives and sometimes also have significant rates of false negatives, besides being limited to the study of transcription factors of well-known species, which decreases their application area. In this scenario, the use of computational methods that do not require previous information concerning the binding sites and use more robust parameters of results detection, instead of alignment scores, may add significant improvement to the processes of predicting regulatory regions. In this project, we developed a new computational model TFBSAnalyzer) for analysis and identification of regulatory elements using molecular modeling techniques for the construction of complexes between a transcription factor bound to specific DNA structures with variable sequences of bases and, by means of thermodynamic calculations of bond enthalpy, provides a scoring function based on the binding energy and predicts the DNA binding sites for the transcription factor in analysis. This approach was tested initially with three transcription factors as models, belonging to the nuclear receptor family, namely estrogen receptor ER-alpha (Estrogen Receptor Alpha), the retinoic acid receptor RAR-beta (Retinoid Acid Receptor Beta) and the retinoic X receptor RXR (Retinoid X Receptor). The computationally predicted models were compared to experimental data available for these nuclear receptors, and presented the following rates of FP/FN: 10%/0 for RAR-beta and RXR, 21%/6% for ER-alpha. We also simulated an experiment of ChIP-seq with ER-alpha with the human genome, where the selected genes were subjected to a transcription factor enrichment analysis, with curated information, revealing that the estrogen receptor is indeed involved in their regulation. To show that our method has a general applicability, we modeled the binding energy distribution for the NHR-28 receptor, isoform a, from Caenorhabditis elegans. The energy distributions obtained were similar to the ones obtained for the model NR, so it would be possible to use the method and search for possible TFBS in the C. elegans genome. The data generated and the methodologies developed in this project should add a significant improvement to the prediction processes of regulatory regions and, consequently, help to understand the mechanisms involved in the gene expression process and its regulation.