Cation-exchange fast protein liquid chromatography separated whole casein into β-casein A2, A1, and B, K-casein, αs1-casein, and αs2-casein fractions as well as γ-caseins and several unidentified peaks using a urea-acetate buffer at pH 5 and a NaCl gradient. The whole casein fractions eluted in the following order: breakdown products of β-casein and unidentified peaks; β-casein A2, Al, and B; additional breakdown products of β-casein and unidentified peaks; K-casein; αs1-casein; and αs2-casein. The calculated composition of the four major caseins correlated well with values obtained using anion-exchange fast protein liquid chromatography at pH 7. An acid-PAGE gel confirmed that the three β-casein peaks were variants of β-casein. Incubating herd bulk whole casein with neuraminidase (EC 3.2.1.18) removed carbohydrate from K-casein. Anion-exchange fast protein liquid chromatography separated whole casein into β-casein breakdown products, K-casein A and B, β-casein, αs2-casein, and αs1-casein peaks as well as three unidentified fractions using bis-Tris-propane-urea buffer at pH 7 and a NaCl gradient. Fractions of whole casein eluted in the following order: breakdown products of β-casein and unidentified fractions A and B; K-casein fraction; unidentified C fraction; β-casein; αs2-casein; and αs1-casein. Following treatment with neuraminidase, K-casein eluted as K-casein B and A rather than a series of peaks. Casein samples from individual cows containing known combinations of K-casein A and B confirmed that the peaks were K-casein variants. Isoelectric focusing on a PhastSystem™ separated K-casein A and B; β-casein A1, A2, A3, and B; αs1-casein Band C; β-lactoglobulin A and B; αs2-casein A; and α-lactalbumin B. Minimal preparation and a short separation time enabled many whole milk and whole casein samples to be phenotyped daily. Stepwise regression equations derived to predict samples as homozygous or heterozygous for K-casein A and B and β-casein A1, A2, and B had coefficient of determination values of .18, .58, .82, and .72 for K-casein A and B, β-casein A1, β-casein A2, and β-casein B. Although amino acid analysis can identify whether β-casein A1, A2, or B variants are present, it cannot identify whether K-casein A and B variants are present. Percentages of K-casein, β-casein, αs1-casein, and αs2-casein obtained with isoelectric focusing, cation-exchange fast protein liquid chromatography, and anion-exchange fast protein liquid chromatography compare well with published results. Isoelectric focusing and anion-exchange fast protein liquid chromatography methods separated K-casein into its A and B variants. Isoelectric focusing and cation-exchange fast protein liquid chromatography separated β-casein into its A1, A2, and B variants. Individual cows homozygous for K-casein A or B expressed the same amount of K-casein. When results from individual cows heterozygous for K-casein are combined, the two alleles are expressed equally; on an individual cow basis, however, some cows expressed more K-casein B than K-casein A. Individual cows homozygous for β-casein A1, A2 or B expressed the same amount of β-casein. When the results for individual cows heterozygous for β-casein are combined, the two β-casein alleles are expressed equally. In milk from individual cows typed β-casein A2B, slightly more B than A2 was expressed with cation-exchange fast protein liquid chromatography.

Nous étudions dans cette thèse le problème de stabilisation de systèmes physiques par l'intermédiaire d'un réseau de communication induisant des retards de dynamique connue. Ce problème apparaît lorsque la loi de commande est mise en ouvre à distance et conduit au problème de la stabilisation d'un système instable en boucle ouverte avec un retard variant dans le temps. Nous utilisons un prédicteur d'état à horizon variable afin d'établir une loi de commande stabilisante qui place les pôles du système en boucle fermée. Le calcul de l'horizon de prédiction est détaillé afin d'établir un contrôleur qui utilise de manière explicite les dynamiques du retard et est robuste vis-à-vis d'erreurs d'estimation induites par le modèle de réseau utilisé.<br />Cette analyse de robustesse permet de prendre en compte les caractéristiques non déterministes du réseau. Nous proposons finalement une loi de commande basée sur un observateur pour le cas où seule la sortie est mesurable et considérons aussi le<br />cas de poursuite de trajectoire. Des simulations et résultats expérimentaux sont proposés.

Les microcystines (MCs), hépatotoxines produites par des cyanobactéries d'eau douce, sont incriminées dans des intoxications humaines et animales. L'ingestion est leur principale voie d'exposition chez l'homme. Parmi les quelques 90 variants décrits, la MC-LR constitue le variant le plus toxique et le mieux connu sur lequel est fondé l'évaluation du risque associé aux MCs. Afin de mieux caractériser le danger pour l'homme par ingestion, nous avons choisi d'étudier et de comparer l'absorption et la toxicité de deux variants, la MC-LR et la MC-RR, sur un modèle cellulaire intestinal humain, les Caco-2. Ces deux variants, de structure quasi identique, sont connus pour leur différence de toxicité aiguë. Nous avons montré que, contrairement aux cellules hépatiques, les Caco-2 absorbent les deux variants de manière similaire. De plus, nos travaux suggèrent l'existence d'un efflux actif de la MC-LR et de la MC-RR par les Caco-2. Enfin, une étude à l'échelle pangénomique a permis de mettre en évidence des différences de réponses cellulaires, avec des atteintes plus marquées de la MC-LR sur la réponse au stress oxydant, la régulation du cycle cellulaire, le stress du réticulum endoplasmique et la dégradation des protéines. Ainsi, les mécanismes de toxicité mis en jeu par les deux variants semblent diverger. Nos travaux soulignent donc la nécessité d'obtenir de données complémentaires sur d'autres variants que la MC-LR pour affiner l'évaluation du risque associé aux MCs.

The fundamental building blocks of chromatin are the nucleosomes. Each such unit is composed of about 200 bp of DNA, the well-conserved core histones (H2A, H2B, H3 and H4) and a linker histone (H1). The DNA is wound around two dimers of H2A–H2B and a tetramer comprising two molecules each of H3 and H4, and there is approximately one linker histone molecule positioned on the exterior of the DNA–protein octamer complex. The nucleosome directs the various structural transitions in chromatin that are needed for proper transcriptional regulation during differentiation and development of the organism in question. The gene activity can be regulated by different histone variants, DNA–protein interactions, and protein–protein interactions, all of which are influenced by the enormous amounts of post-translational modifications that occur in the histone tails. The research underlying this thesis focused on different aspects of post-translational modifications during aging, differentiation, and progression of the cell cycle, and also on expression of linker histone variants and linker histone-chromatin interactions in a variety of cells and tissues. The present results are the first to show that H4 can be trimethylated at lysine 20 in mammalian cells. The trimethylated H4K20 was found in rat kidney and liver at levels that rose with increasing age of the nimals, and it was also detected in trace amounts in human cell lines. Furthermore, in differentiating MEL cells, trimethylated H4K20 was localized to heterochromatin, and levels of trimethylated H4K20 increased during the course of cell differentiation and were correlated with the increasing compaction of the chromatin. The chromatin of terminally differentiated chicken and frog erythrocytes is highly condensed, and the linker histone variants it contains vary between the two species. Cytofluorometric analyses revealed that the linker histones in the chicken erythrocytes exhibited higher affinity for chromatin than did those in the frog erythrocytes. Characterization of the H1° in frog erythrocytes proved it to be the H1°-2 subvariant. Other experiments demonstrated that normal human B lymphocytes expressed the linker histone variants H1.2, H1.3, H1.4, and H1.5, and that B cells from patients with B-CLL expressed the same variants although in different amounts. The most striking dissimilarity was that amounts of H1.3 in the cells were decreased or undetectable in some samples. Sequencing did not discern any defects in the H1.3 gene, and thus the absence of H1.3 is probably regulated at the post-translational level. It was also observed that the levels of linker histone phosphorylation in EBV-transformed B lymphocytes were already increased in the G1 phase of the cell cycle, which is earlier than previously thought. This increase in phosphorylation is probably responsible for the lower affinity of linker histones for chromatin in EBV-transformed cells in the G1 phase of the cell cycle.

The rapid emergence of bacterial resistance to antibiotics has necessitated the development of alternative treatment strategies. Antimicrobial peptides (AMPs) are important immune system components that kill microbes rapidly and have broad activity-spectra, making them promising leads for new pharmaceuticals. Although the need for novel antimicrobials is great, we also need a better understanding of the mechanisms underlying resistance development to enable design of more efficient drugs and reduce the rate of resistance development. The focus of this thesis has been to examine development of bacterial resistance to AMPs and the resulting effects on bacterial physiology. The major model organism used was Salmonella enterica variant Typhimurium LT2. In Paper I, we observed that bacteria resistant to PR-39 appeared at a high rate, and that the underlying sbmA resistance mutations were low cost or even cost-free. Such mutants are more likely to rapidly appear in a population and, most importantly, will not disappear easily once the selective pressure is removed. In paper II, we isolated protamine-resistant hem- and cydC-mutants that had reduced growth rates and were cross-resistant to several other antimicrobials. These mutants were small colony variants (SCVs), a phenotype often associated with persistent infections. One SCV with a hemC-mutation reverted to faster growth when evolved in the absence of protamine. In paper III, the mechanism behind this fitness compensation was determined, and was found to occur through hemC gene amplification and subsequent point mutations. The study provides a novel mechanism for reversion of the SCV-phenotype and further evidence that gene amplification is a common adaptive mechanism in bacteria. In Paper IV, the antibacterial properties of cyclotides, cyclic mini-proteins from plants, were evaluated. Cycloviolacin O2 from violets was found to be bactericidal against Gram-negative bacteria. Cyclotides are very stable molecules and may be potential starting points for development of peptide antibiotics.

There is solid evidence that complex human diseases can be caused by rare variants. Next generation sequencing technology has revolutionized the study of complex human diseases, and made possible detecting associations with rare variants. Traditional statistical methods can be inefficient for analyzing sequence data and underpowered. In addition, due to high cost of sequencing, it is also necessary to explore novel cost effective studies in order to maximize power and reduce sequencing cost. In this thesis, three important problems for analyzing sequence data and detecting associations with rare variants are presented. In the first chapter, we presented a new method for detecting rare variants/binary trait associations in the presence of gene interactions. In the second chapter, we explored cost effective study designs for replicating sequence based association studies, combining both sequencing and customized genotyping. In the third chapter, we present a method for analyzing multiple phenotypes in selected samples, such that phenotypes that are commonly measured in different studies can be jointly analyzed to improve power. The methods and study designs presented are important for dissecting complex trait etiologies using sequence data.

H2A.Z is an essential histone variant that has multiple chromosomal functions. One such role is transcriptional regulation. However, its role in this process is complex since it has been reported to function both as a repressor and activator. Earlier work in our lab showed that H2A.Z can be post-translationally modified with monoubiquitin (H2A.Zub1) and this form of H2A.Z is linked to transcriptional silencing. We further predicted that changes in the H2A.Z ubiquitylation status directly modulated its function in transcription. Furthermore, H2A.Z-containing nucleosomes possess a unique set of post-translational modifications (PTMs), compared to H2A nucleosomes, many of which are linked to transcriptional activation. The central aim of this thesis was to characterize the role of PTMs on H2A.Z nucleosomes in transcriptional regulation. To this end, I have provided the first evidence linking H2A.Z deubiquitylation to transcriptional activation. I demonstrated that ubiquitin specific protease 10 (USP10) is a deubiquitylase that targets H2A.Z in vitro and in vivo. Moreover, I found that both H2A.Z and USP10 are required for activation of androgen-receptor (AR)-regulated genes, and that USP10 regulates the levels of H2A.Zub1 at these genes. To understand how H2A.Z engages downstream effector proteins, in the nucleosome context, we used a mass spectrometry approach to identify H2A.Z-nucleosome-interacting proteins. Many of the identified proteins contained conserved structural motifs that bind post-translationally modified histones. For example, we found that Brd2 contains tandem bromodomains that engage H2A.Z nucleosomes through acetylated H4 residues. To investigate the biological relevance of this interaction, I present evidence that Brd2 is recruited to AR-regulated genes in a manner dependent on H2A.Z and the bromodomains of Brd2. Consistent with this observation, chemical inhibition of Brd2 recruitment greatly inhibited AR-regulated gene expression. Collectively, these studies have defined how H2A.Z mediates transcriptional regulation through multiple mechanisms and pathways.

Copy number variations (CNVs) are a major source of inter-individual genetic difference, accounting for a greater proportion of the human genome than other forms of variation. Recently, the identification of benign and pathogenic CNVs has improved due to arrays with increased coverage. Nevertheless, most CNVs have not been studied with great precision and questions persist regarding their exact breakpoint, gene content, frequency and functional impact. This is especially true in cancer, in which a role for CNVs as risk factors is under-explored. Li-Fraumeni syndrome (LFS) is a dominantly inherited disorder with an increased risk of early-onset breast cancer, sarcomas, brain tumors and other neoplasms in individuals harboring germline TP53 mutations. Known genetic determinants of LFS do not fully explain its clinical phenotype. In this thesis we describe the association between CNVs and LFS. First, by examining DNA from a healthy population and an LFS cohort using oligonucleotide arrays, we show that the number of CNVs per genome is well conserved in the healthy population, but remarkably enriched in these cancer-prone individuals. We found a significant increase in CNVs among carriers of germline TP53 mutations with a familial cancer history. Second, we find that ii specific CNVs at 17p13.1 are associated with LFS or developmental delay, depending on the exact breakpoint with respect to TP53. Using a purpose built array with 93.75% accuracy, we fine-mapped these microdeletions and find that they arise by Alu-mediated non-allelic homologous recombination, and contain common genes, whose under-expression distinguishes the two phenotypes. Third, we explore somatic CNVs in choroid plexus carcinoma tumor genomes. We show that this tumor is over-represented in LFS, and the number of somatic CNVs is associated with TP53 mutations and disease progression. These studies represent the first genomic analyses of LFS, and suggest a more generalized association between CNVs and cancer.

Copy number variations (CNVs) are a major source of inter-individual genetic difference, accounting for a greater proportion of the human genome than other forms of variation. Recently, the identification of benign and pathogenic CNVs has improved due to arrays with increased coverage. Nevertheless, most CNVs have not been studied with great precision and questions persist regarding their exact breakpoint, gene content, frequency and functional impact. This is especially true in cancer, in which a role for CNVs as risk factors is under-explored. Li-Fraumeni syndrome (LFS) is a dominantly inherited disorder with an increased risk of early-onset breast cancer, sarcomas, brain tumors and other neoplasms in individuals harboring germline TP53 mutations. Known genetic determinants of LFS do not fully explain its clinical phenotype. In this thesis we describe the association between CNVs and LFS. First, by examining DNA from a healthy population and an LFS cohort using oligonucleotide arrays, we show that the number of CNVs per genome is well conserved in the healthy population, but remarkably enriched in these cancer-prone individuals. We found a significant increase in CNVs among carriers of germline TP53 mutations with a familial cancer history. Second, we find that ii specific CNVs at 17p13.1 are associated with LFS or developmental delay, depending on the exact breakpoint with respect to TP53. Using a purpose built array with 93.75% accuracy, we fine-mapped these microdeletions and find that they arise by Alu-mediated non-allelic homologous recombination, and contain common genes, whose under-expression distinguishes the two phenotypes. Third, we explore somatic CNVs in choroid plexus carcinoma tumor genomes. We show that this tumor is over-represented in LFS, and the number of somatic CNVs is associated with TP53 mutations and disease progression. These studies represent the first genomic analyses of LFS, and suggest a more generalized association between CNVs and cancer.

<p>Transcription initiation is a key component in the regulation of gene expression. Recent high-throughput sequencing techniques have enhanced our understanding of mammalian transcription by revealing narrow and broad patterns of transcription start sites (TSSs). Transcription initiation is central to the determination of condition specificity, as distinct repertoires of transcription factors (TFs) that assist in the recruitment of the RNA polymerase II to the DNA are present under different conditions. However, our understanding of the presence and spatiotemporal architecture of the promoter patterns in the fruit fly remains in its infancy. Nucleosome organization and transcription initiation have been considered hallmarks of gene expression, but their cooperative regulation is also not yet understood.</p> <p>In this work, we applied a hierarchical clustering strategy on available 5' expressed sequence tags (ESTs), and developed an improved paired-end sequencing strategy to explore the transcription initiation landscape of the D.melanogaster genome. We distinguished three initiation patterns: 'Peaked or Narrow Peak TSSs&#8219;, 'Broad Peak TSSs&#8219;, and 'Broad TSS cluster groups or Weak Peak TSSs&#8219;. The promoters of peaked TSSs contained the location specific sequence elements, and were bound by TATA Binding Protein (TBP), while the promoters of broad TSS cluster groups were associated with non-location-specific elements, and were bound by the TATA-box related Factor 2 (TRF2).</p> <p>Available ESTs and a tiling array time series enabled us to show that TSSs had distinct associations to conditions, and temporal patterns of embryonic activity differed across the majority of alternative promoters. Peaked promoters had an association to maternally inherited transcripts, and broad TSS cluster group promoters were more highly associated to zygotic utilization. The paired-end sequencing strategy identified a large number of 5' capped transcripts originating from coding exons that were unlikely the result of alternative TSSs, but rather the product of post-transcriptional modifications.</p> <p>We applied an innovative search program called FREE to embryo, head, and testes specific core promoter sequences and identified 123 motifs: 16 novel and 107 supported by other motif sources. Motifs in the embryo specific core promoters were found at location hotspots from the TSS. A family of oligos was discovered that matched the Pause Button motif that is associated with RNA pol II stalling.</p> <p>Lastly, we analyzed nucleosome organization, chromatin structure, and insulators across the three promoter patterns in the fruit fly and human genomes. The WP promoters showed higher associations with H2A.Z, DNase Hypersensitivity Sites (DHS), H3K4 methylations, and Class I insulators CTCF/BEAF32/CP190. Conversely, NP promoters had higher associations with polII and GAF binding. BP promoters exhibited a combination of features from both promoter patterns. Our study provides a comprehensive map of initiation sites and the conditions under which they are utilized in D. melanogaster. The presence of promoter specific histone replacements, chromatin modifications, and insulator elements support the existence of two divergent strategies of transcriptional regulation in higher eukaryotes. Together, these data illustrate the complex regulatory code of transcription initiation.</p> / Dissertation