Cells within multicellular organisms share the same genetic information, yet their shape and function can differ dramatically. This diversity of form and function is established by differential use of the genetic information. Early embryonic development describes the processes that lead to a fully differentiated embryo starting from a single fertilized cell - the zygote. Interestingly, in metazoan species this early development is governed by maternally provided factors (nutrients, RNA, protein), while the zygotic genome is transcriptionally inactive. Only at a specific developmental stage, the zygotic genome becomes transcriptionally active, and zygotic transcripts drive further embryonic development. This major change is called zygotic genome activation (ZGA). While major regulators of activation of early zygotic genes could be identified recently, the molecular mechanisms that contribute to robust global genome activation during embryonic development is not fully understood. In this study, I investigated whether the establishment of histone H3 lysine 4 trimethylation (H3K4me3) is involved in zebrafish zygotic transcription activation and early embryonic development. H3K4me3 is a chromatin modification that is implicated in transcription regulation. H3K4me3 has been shown to be enriched at Transcription Start Sites (TSS) of genes prior to their activation, and is postulated facilitate transcription activation of developmentally important genes. To interfere with H3K4me3 establishment, I generated histone methyltransferase mutants. I further inhibited H3K4me3 establishment by introduction of histones with lysine 4-to-methionine (K4-to-M) substitution, which act as dominant-negative inhibitors of H3K4me3 establishment. Upon H3K4me3 reduction, I studied the resulting effect on early transcription activation. I found that H3K4me3 is not involved in transcription activation during early zebrafish embryogenesis. Finally I analyzed possible cues in DNA sequence and chromatin environment that might favor early H3K4me3 establishment. These studies show that H3K4me3 is established during ZGA, yet it is not involved in transcription activation during early zebrafish development. Establishment of H3K4me3 might be a consequence of histone methyltransferase recruitment during a permissive chromatin state, and might be targeted to CpG-rich promoter elements that are enriched for the histone variant H2A.z. / Jede Zelle eines multizellulären Organismus enthält dieselbe Erbinformation, und doch können Form und Funktion von Zellen untereinander sehr unterschiedlich sein. Diese Diversität wird durch unterschiedliches Auslesen - Transkribieren - der Erbinformation erreicht. Embryogenese beschreibt den Prozess, der aus einer einzelnen Zelle - der Zygote - einen multizellulären Embryo entstehen lässt. Interessanterweise laufen frühe Stadien der Embryogenese ohne Transkription der embryonalen Erbinformation ab, sondern werden durch maternal bereitgestellte Faktoren ermöglicht. Erst nach einer spezies-spezifischen Entwicklungsphase wird das Erbgut der Zygote aktiv transkribiert und ermöglicht die weitere Embryonalentwicklung. Obwohl bereits wichtige Regulatoren dieser globalen Genomaktivierung identifiziert werden konnten, sind viele molekulare Mechanismen, die zur Aktivierung des zygotischen Genoms beitragen, bisher unbekannt. In der hier vorliegenden Doktorarbeit habe ich die Rolle von Histon H3 Lysin 4 Trimethylierung (H3K4me3) während der frühen Embryogenese des Zebrafischs untersucht. H3K4me3 ist eine Chromatinmodifikation, die mit aktiver Transkription in Verbindung gebracht wird. H3K4me3 ist an Transkriptions-Start-Stellen von aktiv ausgelesenen Genen angereichert und es wird vermutet, dass diese Modifikation das Binden von Transkriptionsfaktoren und der Transkriptionsmaschinerie erleichtert. Während meiner Arbeit habe ich durch Mutation verschiedener Histon-Methyltransferasen beziehungsweise die Überexpression eines dominant-negativen Histonsubstrats versucht, die Etablierung von H3K4me3 in frühen Entwicklungsstadien des Zebrafischs zu verhindern. Anschliessend habe untersucht, welchen Effekt H3K4me3-Reduktion auf Tranksriptionsaktivität entsprechender Gene hat. Allerdings konnte ich keinen Zusammenhang zwischen H3K4me3-Reduktion und Transkriptionsaktivität beobachten. Um herauszufinden, weshalb H3K4me3 dennoch während früher Embryonalstadien etabliert wird, habe ich nachfolgend untersucht, ob möglicherweise bestimmte DNASequenzen oder Chromatin-Modifikationen zur Etablierung von H3K4me3 wahrend der Embryogenese des Zebrafischs beitragen. Aus der hier vorliegenden Arbeit lässt sich schlussfolgern, dass H3K4me3 in Tranksriptionsaktivierung während früher Embryonalstadien des Zebrafischs nicht involviert ist. Möglicherweise wird H3K4me3 in diesen Stadien in einer permissiven Chromatinumgebung etabliert, bevorzugt an Promotoren mit starker H2A.z-Anreicherung und CpG-reichen DNA-Elementen.

N"alpha"-Acetylierung, eine der häufigsten eukaryontischen Proteinmodifikationen, wird von N-terminalen Acetyltransferasen (NATs) katalysiert. NatA, die bedeutendste NAT in Saccharomyces cerevisiae, besteht aus den Untereinheiten Nat1, Ard1 und Nat5, und ist am silencing, d.h. am Aufbau repressiver Chromatinstrukturen an Telomeren und den Paarungstyp-Loci HML und HMR beteiligt. Die vorliegende Arbeit demonstriert eine Rolle von NatA auch beim rDNA-silencing, und zeigt erstmals, dass die silencing-Faktoren Orc1 und Sir3 funktionell von der N"alpha"-Acetylierung durch NatA abhängen. Orc1, die größte Untereinheit des origin recognition complex (ORC), wurde in vivo durch NatA N"alpha"-acetyliert. Mutationen, die dies verhinderten, bewirkten eine starke telomerische Derepression. NatA wirkte genetisch über die ORC Bindungsstelle des HMR-E-silencers. Die artifizielle Bindung von Orc1 an HMR-E machte HMR-silencing NatA-unabhängig. Auch die synthetische Letalität von nat1"DELTA" orc2-1 Doppelmutanten wies auf eine funktionelle Verbindung zwischen NatA und ORC hin. Als weiteres NatA-Substrat wurde Sir3 identifiziert, dessen zelluläre Lokalisierung von NAT1 abhing. Die schwächeren silencing-Defekte der unacetylierten orc1 sir3 Doppelmutante im Vergleich zu nat1"DELTA" implizierten allerdings, dass noch weitere silencing-Proteine die N"alpha"-Acetylierung für ihre Funktion bedürfen. Weitere Ergebnisse dieser Arbeit belegen eine Funktion N-terminalen 100 Aminosäuren von Orc1 im silencing. Deletionen innerhalb dieses Bereichs erzeugten silencing-Defekte. Das Fehlen von 51 Aminosäuren vom N-Terminus von Orc1 unterbrach die Interaktion mit Sir1, verstärkte aber auch den silencing-Defekt von sir1"DELTA". Dies ergibt ein Model, in dem Orc1 neben Sir1 ein weiteres silencing-Protein rekrutiert, das zu seiner Bindung einen intakten, acetylierten N-Terminus von Orc1 benötigt. Zusammenfassend sprechen die Ergebnisse für eine Rolle der N"alpha"-Acetylierung durch NatA bei der Modellierung der Chromatinstruktur. / N"alpha"-acetylation, one of the most abundant eukaryotic protein modifications, is catalyzed by N-terminal acetyltransferases (NATs). NatA, the major NAT in Saccharomyces cerevisiae, consists of the subunits Nat1, Ard1 and Nat5 and is necessary for the assembly of repressive chromatin structures at the silent mating type loci and telomeres. This thesis shows that NatA also acts in rDNA repression and it provides the first direct evidence for the functional regulation of the silencing factors Orc1 and Sir3 by NatA-dependent N"alpha"-acetylation.Orc1, the large subunit of the origin recognition complex (ORC), was N"alpha"-acetylated in vivo by NatA. Mutations that abrogated this acetylation caused strong telomeric derepression. NatA functioned genetically through the ORC binding site of the HMR-E silencer. Direct tethering of Orc1 to HMR-E circumvented the requirement for NatA in silencing. The synthetic lethality of nat1"DELTA" orc2-1 double mutants further supported a functional link between NatA and ORC.Sir3 was also indentified as a NatA substrate. Its localization to perinuclear foci was NAT1 dependent. Unacetylated sir3 orc1 double mutants did not resemble the nat1"DELTA" silencing phenotype. Thus, we suggest that further silencing components require NatA-dependent N"alpha"-acetylation for their function. We further identified the N-terminal 100 amino acids of Orc1 to be important for silencing, since truncations within this region impaired silencing. The deletion of 51 amino acids from the Orc1 N-terminus interrupted the interaction with Sir1 and also reduced silencing in sir1"DELTA" strains. We thus propose that the silencing function of Orc1 is not restricted to Sir1 recruitment, but also comprises the interaction with another protein. The silencing function of this hypothesized interaction partner may depend on the N"alpha"-acetylation and integrity of the N-terminus of Orc1.In summary, we propose that N"alpha"-acetylation by NatA represents a protein modification that modulates chromatin structure in yeast.

ABSTRACT Chromatin insulators are required for proper temporal and spatial expression of genes in metazoans. Insulators are thought to play an important role in the regulation of gene expression through the formation of higher-order chromatin structures. One of the best characterized insulators is the Drosophila gypsy insulator, which is located in the gypsy retrovirus. Several proteins are required for gypsy insulator function, including Su(Hw), Mod(mdg4), and CP190. In addition to the gypsy insulator, these proteins are located throughout the genome at sites which are thought to correspond to endogenous insulators. Analysis of the distribution of insulator proteins across a region of chromosome 2R in Drosophila polytene chromosomes shows that Su(Hw) is found in three structures differentially associated with insulator proteins: bands, interbands and domains of coexpressed genes. Bands are formed by condensation of chromatin within genes containing one or more Su(Hw) binding sites, while Su(Hw) sites in interbands appear to form structures normally associated with open chromatin. Bands characterized by the lack of CP190 and BEAF-32 insulator proteins are formed by clusters of coexpressed genes, and these bands correlate with the distribution of specific chromatin marks. Conservation of the band interband pattern, as well as the distribution of insulator proteins in nurse cells, suggests that this organization may represent the basic organization of interphasic chromosomes. We also show that, in addition to the gypsy insulator, sequence analysis predicts the presence of Su(Hw) binding sites within a number of transposable elements. Su(Hw) binds to predicted sites within gtwin and jockey, which possesses enhancer-blocking activity. Su(Hw) affects the tissue-specific expression of transposable elements, although this effect is unrelated to the presence of Su(Hw) binding sites within the element or control of the elements via the piRNA pathway. Additionally, the effect of Su(Hw) on transposable element expression often differs from that of Mod(mdg4). Taken together, these results suggest that insulator proteins associate specifically with, and may help to define, various levels of chromatin organization on polytene chromosomes. Also, gypsy insulator proteins may influence the expression of transposable elements in a way that does not depend on Su(Hw) binding sites within the elements themselves.

Chromatin is a highly dynamic, regulatory component in the process of transcription, repair, recombination and replication. The BRG1 and SNF2H proteins are ATP-dependent chromatin remodeling proteins that modulate chromatin structure to regulate DNA accessibility for DNA-binding proteins involved in these processes. The BRG1 protein is a central ATPase of the SWI/SNF complexes involved in chromatin remodeling associated with regulation of transcription. SWI/SNF complexes are biochemically hetero-geneous but little is known about the unique functional characteristics of the various forms. We have shown that SWI/SNF activity in SW13 cells affects actin filament organization dependent on the RhoA signaling pathway. We have further shown that the biochemical composition of SWI/SNF complexes qualitatively affects the remodeling activity and that the composition of biochemically purified SWI/SNF complexes does not reflect the patterns of chromatin binding of individual subunits. Chromatin binding assays (ChIP) reveal variations among subunits believed to be constitutive, suggesting that the plasticity in SWI/SNF complex composition is greater than suspected. We have also discovered an interaction between BRG1 and the splicing factor Prp8, linking SWI/SNF activity to mRNA processing. We propose a model whereby parts of the biochemical heterogeneity is a result of function and that the local chromatin environment to which the complex is recruited affect SWI/SNF composition. We have also isolated the novel B-WICH complex that contains WSTF, SNF2H, the splicing factor SAP155, the RNA helicase II/Guα, the transcription factor Myb-binding protein 1a, the transcription factor/DNA repair protein CSB and the RNA processing factor DEK. The formation of this complex is dependent on active transcription and links chromatin remodeling by SNF2H to RNA processing. By linking chromatin remodeling complexes with RNA processing proteins our work has begun to build a bridge between chromatin and RNA, suggesting that factors in chromatin associated assemblies translocate onto the growing nascent RNA.

Nuestro grupo ha estudiado la estructura de la cromatina de núcleos de eritrocitos de pollo (Bartolomé et al., 1994; Bartolomé et al., 1995; Bermúdez et al., 1998). La consecuencia de estos estudios ha sido la elaboración de un modelo para el plegamiento de la fibra de cromatina con una elevada concentración local del DNA (Daban y Bermúdez, 1998; Daban, 2000). Sin embargo, el nivel máximo de condensación en la cromatina, se encuentra en el interior de los cromosomas metafásicos. Aunque la bibliografía ha planteado diferentes modelos para el plegamiento de la cromatina en el interior de éstos, existe un conocimiento muy escaso acerca de la estructura molecular de la cromatina en los cromosomas condensados.Se ha realizado un estudio exhaustivo de microscopía electrónica de transmisión sobre la estructura de los cromosomas metafásicos de células HeLa. Se han estudiado un total de 4410 micrografías de cromosomas metafásicos, que en su mayor parte han sido tratados con diversos medios parcialmente desnaturalizantes, para poder analizar su estructura interna.Morfológicamente, los cromosomas estudiados en este trabajo pueden agruparse en tres tipos diferentes: compactos, granulados y fibrilados. La morfología más abundante es la compacta y se observa en presencia de cationes monovalentes y divalentes a concentración similar a la presente en la cromatina metafásica (Mg2+ 1.7-40 mM). Estos cromosomas tienen las cromátidas muy densas y en sus bordes se aprecian una serie de estructuras planas superpuestas. En condiciones de menor concentración de cationes (Mg2+£ 1.7 mM), la morfología dominante es la granular. Estos cromosomas están compuestos principalmente por gran cantidad de cuerpos circulares de 30-40 nm de diámetro. Únicamente en condiciones de fuerza iónica extremadamente baja podemos encontrar la morfología fibrilar, la cual se caracteriza por la abundancia de fibras de 30-40 nm.Los resultados obtenidos con cromosomas parcialmente desnaturalizados nos permiten concluir que existen tres elementos estructurales en el interior de los cromosomas metafásicos: la fibra, el gránulo y la placa.Las fibras gruesas con diámetros que oscilan entre los 100 y los 500 nm son el resultado de la deformación plástica de las cromátidas durante los diferentes procesos de preparación de las muestras. En función de las condiciones iónicas del medio las fibras gruesas muestran gránulos o placas en su interior. Las fibras delgadas están formadas por una sucesión de cuerpos de 30-40 nm de diámetro unidos irregularmente mediante interacciones cabeza-cola. Las fibras delgadas se observan dominantemente en condiciones de concentración salina extremadamente baja.Los gránulos son unos cuerpos circulares compactos de unos 30-40 nm de diámetro. Estos cuerpos compactos descritos previamente por nuestro grupo y se interpretaron como una forma de plegamiento solenoidal de la fibra de 30 nm (Daban y Bermúdez, 1998). Se encuentran presentes en todas las condiciones estudiadas en este trabajo, siendo especialmente abundantes en presencia de iones divalentes a concentración baja y en muestras tratadas con nucleasa micrococal. La placa es un elemento estructural característico de los cromosomas cuando éstos se encuentran en su forma más compacta, en presencia de concentraciones elevadas de cationes divalentes. Esta estructura no había sido descrita previamente por otros laboratorios. Es una estructura cromatínica de gran regularidad y con una superficie muy lisa. Hemos estimado la altura de estas placas a través de muestras sombreadas unidireccionalemente con platino. El promedio de los valores obtenidos es de 6.7 ± 1.4 nm.En conjunto los resultados obtenidos en esta tesis permiten sugerir que el componente principal de la cromatina en los cromosomas metafásicos es el gránulo de 30-40 nm. Dependiendo de las condiciones iónicas, este elemento estructural fundamental se agrega a través de uniones cabeza-cola para formar fibras (fuerza iónica muy baja), o bien se agrega mediante interacciones laterales para formar placas (condiciones salinas próximas a las de la cromatina metafásica). / Our group has studied the chromatin structure in the chicken erythrocyte nuclei (Bartolome et al., 1994; Bartolomé et al., 1995; Bermúdez et al., 1998). The consequences of this studies has been the elaboration of a folding model of the chromatin fiber with a high local concentration of DNA. However, the maximum level of chromatin condensation, is found in the metaphase chromosomes. Although the bibliography has proposed different models to explain the chromatin folding inside the chromosomes, there is a low knowledge about the molecular structure of chromatin in the condensed chromosomes. In this thesis, we have carried out an exhaustive electron microscopy study about the HeLa cells metaphase chromosomes. We have studied a large number of chromosome electron micrographs (4410). Chromosomes were partially denaturated under a wide variety of conditions in order to observe some chromatin structural element inside them.Our studies indicate that chromosomes can adopt three global structural forms in function of the ionic conditions: compact, granular and fibrillar.The compact form is the most frequent and we can observe it in the presence of monovalent and divalent cations in similar concentrations than the ones found in metaphase chromatin (Mg2+ 1.7-40 mM). These chromosomes have highly condensed chromatids and we can appreciate overlapped chromatin plates around the chromosomes edges. When the chromosomes are incubated with solutions containing lower cations concentration (Mg 2+£ 1.7 mM) they become granular. The granular structures seen inside these chromosomes show a diameter of about 35 nm. Fibrillar chromosomes are observed only at very low ionic strength. The fibers seen emanating from the chromatids have a diameter of 30-40 nm.Our results obtained from partially denaturated chromosomes show that there are three structural elements inside the metaphase chromosomes: the fiber, the 30-40 nm chromatin granule and the plate.The largest fibers with a diameter of 100-400 nm, presumably are produced by mechanical deformation of chromosomes during the preparation processes. Depending of the ionic conditions these fibrillar structures are composed by plates or granules. The thinnest fibers are formed by face to face association of the 30-40 nm chromatin granules. These kind of fibers are usually found only at very low ionic strength.The chromatin granules are compact bodies with 35 nm of diameter. These compact bodies were previously described in our laboratory and were modeled as compact solenoids of nucleosomes forming (Daban and Bermúdez, 1998). They are usually seen at low divalent cation concentrations and in chromosome samples treated with micrococal nuclease.The plate is the most frequent structural element when the chromosomes are in their compact form (high ionic strength, similar to physiological conditions). This element has not been described by any group. It is a chromatin element with a regular structure and very smooth surface. We have estimated the height of the steps between layers in unidirectional shadowing experiments. The value obtained is 6.7 ± 1.4 nm.Our results suggest that the fundamental component inside the metaphase is the 30-40 nm chromatin granules. Depending of the ionic conditions, this basic structural element forms fibers through face to face interactions (very low ionic strength) or form plates through side to side interactions (high ionic strength similar to metaphase chromatin).

<p>Heterochromatin, or condensed chromatin, is a transcriptionally repressive form of chromatin that occurs in many eukaryotic organisms. At its natural locations, heterochromatin is thought to play important roles in genome organization as well as gene expression. Just as important is the restriction of this repressive form of chromatin to appropriate regions of the genome. In the budding yeast <italic>Saccaromyces cerevisiae</italic>, domains of condensed, transcriptionally silenced chromatin are found at telomeres and at the silent-mating type cassettes, <italic>HML<italic/> and <italic>HMR</italic>. At these locations, a complex of Silent Information Regulator (SIR) proteins gets recruited to DNA through discrete silencer elements. Once recruited, the Sir protein complex then spreads along chromosomes in a step-wise manner. This process results in the silencing of gene expression. It is unclear whether silenced chromatin is established in the same manner at different genomic locations. Understanding how silenced chromatin is formed is important for determining how these chromatin structures are regulated.</p><p>To better understand how silenced chromatin is established in different genomic contexts, I used chromatin immuoprecipitation to follow the rate of silenced chromatin formation at different locations. The rates of Sir protein assembly were compared at two locations, telomere VI-R and <italic>HMR</italic>. I discovered that the silencers at these two locations were equally proficient at recruiting Sir proteins. However, the rate of Sir protein assembly onto nucleosomes was far more rapid at <italic>HMR</italic> than at the telomere VI-R. Furthermore, the rate of Sir protein assembly was more rapid on one side of the <italic>HMR-E</italic> silencer at <italic>HMR</italic> than the other. Moreover, insertion of the <italic>HMR-E</italic> silencer adjacent to the telomere VI-R significantly improved the rate of Sir protein assembly onto nucleosomes. Additionally, observations that the association of Sir protein occurs simultaneously across several kilobases at <italic>HMR</italic> and that silencing at <italic>HMR</italic> is insensitive to co-expression of wild-type and catalytically inactive Sir2 proteins suggest that <italic>HMR-E</italic> enables the assembly of silenced chromatin in a non-linear fashion. These results suggest that <italic>HMR-E</italic> functions to both recruit Sir proteins and promote their assembly across several kilobases.</p><p>In addition to the <italic>HMR-E</italic> silencer, <italic>HMR</italic> is also characterized by the presence of a second auxiliary <italic>HMR-I</italic> silencer and a tRNA gene that functions as a boundary element to restrict the spread of silenced chromatin. I used chromatin immunoprecipitation to determine how each of these regulatory elements contribute to the steady-state levels of Sir protein association with chromatin. Consistent with a role for <italic>HMR-E</italic> beyond recruitment, I discovered that the <italic>HMR-E</italic> silencer alone promoted higher levels of Sir proteins on nucleosomes compared to the telomere VI-R. The levels of Sir protein association with <italic>HMR</italic> were further elevated by the <italic>HMR-I</italic> silencer, even though this silencer does not recruit Sir proteins on its own and does not contribute to any of the known functions of silenced chromatin at <italic>HMR</italic>. Additionally, although the tRNA gene did block the spread Sir proteins, I discovered that the capacity for Sir proteins to spread beyond a few kilobases was severely limited even in the absence of the boundary.</p><p>The results of this thesis work provide new insights into the mechanisms of silenced chromatin establishment and regulation in budding yeast. I show here that the capacity of Sir proteins to assemble onto nucleosomes is inherently limited. Additionally, silencers vary in their ability to promote this assembly. I conclude that the silencer is a key factor in determining the relative size, efficiency, and location of silenced chromatin domains in the cell.</p> / Dissertation

Covalent histone modifications such as methylation, acetylation as well as differential incorporation of histone variants are shown to coincide with different chromatin compartments and mark active or repressed genes. Msc1 is one of the seven JmjC Domain Proteins (JDPs) in Fission Yeast. JDPs are known to function in chromatin and some act as histone demethylases. We found that Msc1 is a member of Swr1 Complex which is known to exchange histone H2A variant H2A.z in nucleosomes. We purified H2A.z as a member of Swr1 Complex and its interaction with Swr1 Complex depends on Swr1. We’ve shown that histone H4 Lysine 20 trimethylation (H4 K20 Me3) is lost in h2A.z and msc1 deletion strains and these strains are sensitive to UV. Deletion strain of h2A.z is sensitive to Camptothecin. Histones H3 and H4 are obtained in Msc1 and H2A.z purifications and we’ve shown that histone H4 from these purifications has low level of Lysine 16 acetylation (H4 K16 Ac). Deletion strains of h2A.z, swr1 and msc1 are shown to be sensitive to TSA, a histone deacetylase (HDAC) inhibitor suggesting that H2A.z cooperates with HDACs. TSA treatment of wild type cells cause an increase in H4 K16 Ac and a decrease in H4 K20 Me3. Gene expression profiles of h2A.z, swr1 and msc1 are significantly similar and upregulated genes in deletion strains localize at chromosome ends (a region of 160 kb for each end). The number of stress or meiotic inducible genes is increased in deletion strains suggesting that H2A.z has a role in regulation of inducible genes. We suggest that H2A.z, in cooperation with HDACs, functions in regulation of chromatin accessibility of inducible promoters.

Das Tumormikromilieu beeinflusst maßgeblich Tumorwachstum und -progression sowie das Ansprechen von Tumorzellen auf Strahlen- und Chemotherapie. Weiterhin ist bekannt, dass Wachstumsfaktoren, Sauerstoffgehalt und extrazelluläre Matrix (EZM) als Resistenzfaktoren das Zellüberleben nach Exposition mit ionisierender Strahlung oder zytotoxischen Substanzen bestimmen. Weitere zelluläre Parameter, wie Zellmorphologie, Zytoskelettarchitektur und Chromatinkondensation, werden ebenfalls in Abhängigkeit der Wachstumsbedingungen moduliert, wie vergleichende Untersuchungen an physiologischeren drei- (3D) mit herkömmlichen zwei-dimensionalen (2D) Zellkulturen zeigen. Veränderungen der Chromatindichte beeinflussen zudem die Genexpression, wodurch wichtige zelluläre Prozesse, wie Überleben, Proliferation und Differenzierung der Zellen, reguliert werden. Außerdem ist die Chromatinkondensation für eine effektive Reparatur strahleninduzierter DNA-Schäden, wie DNA-Doppelstrangbrüche (DSB), von großer Bedeutung. Das Ziel der vorliegenden Arbeit bestand darin, die zelluläre Strahlenempfindlichkeit unter Berücksichtigung der Chromatinkondensation in humanen Bronchial- (A549) und Plattenepithelkarzinomzellen (UTSCC-15) in Abhängigkeit der Wachstumsbedingungen zu analysieren. Da die molekularen Mechanismen der Wechselwirkung zwischen Chromatindichte und Reparatur strahleninduzierter DSB bis heute unklar sind, war die Untersuchung dieser Zusammenhänge unter 2D, 3D und in vivo Wachstumsbedingungen von besonderem Interesse. Die Ergebnisse dieser Arbeit zeigen, dass das Zellwachstum in einer physiologischen 3D Matrix im Vergleich zur herkömmlichen 2D Zellkultur zu einer geringeren Anzahl an strahleninduzierten residuellen DSB (rDSB) und letalen Chromosomenaberrationen führen kann, was wiederum für ein verbessertes Zellüberleben nach Bestrahlung verantwortlich sein könnte. Des Weiteren konnte in 3D im Zusammenhang mit einer höheren Chromatinkondensierung eine Erhöhung der zellulären Strahlenresistenz gezeigt werden. Auf molekularer Ebene zeigen die Ergebnisse dieser Arbeit außerdem, dass eine siRNA-vermittelte Hemmung chromatinmodifizierender Histondeacetylasen (HDAC 1, 2 und 4) zu keiner Strahlensensibilisierung führt, während durch die Behandlung mit dem pharmakologischen HDAC-Inhibitor Panobinostat (LBH589) neben der Chromatindekondensierung auch eine erhöhte Strahlenempfindlichkeit der Zellen erreicht werden konnte. In Abhängigkeit der untersuchten Wachstumsbedingungen konnten Unterschiede in der Verteilung strahleninduzierter DSB zwischen hetero- und euchromatischen DNA-Bereichen nachgewiesen werden. Interessanterweise nimmt in 2D dosisabhängig der prozentuale Anteil der Heterochromatin (HC)-assoziierten Foci ab, wohingegen in 3D und im Xenografttumormodell dosisunabhängig etwa die Hälfte der Foci mit heterochromatischen DNA-Bereichen assoziiert sind. Diese Daten zeigen, dass Tumorzellen in 3D und in vivo in Abhängigkeit der veränderten Zellmorphologie und Chromatinkondensierung deutlich mehr HC-assoziierte rDSB besitzen als in 2D, was die Hypothese einer beeinträchtigten Reparatur im HC unterstützt. Dennoch zeigt die Korrelation zwischen der deutlich geringeren rDSB Gesamtanzahl und dem erhöhtem Zellüberleben in 3D, dass neben dem Anteil an kondensiertem Chromatin auch die Gesamtanzahl rDSB ein wichtiger Einflussfaktor der zellulären Strahlenempfindlichkeit zu sein scheint. Die Ergebnisse dieser Arbeit liefern somit wichtige Erklärungsansätze für einen direkten Zusammenhang zwischen Zellmorphologie, Chromatinkondensation und zellulärer Strahlenempfindlichkeit. Des Weiteren unterstreichen diese Untersuchungen das verwendete 3D Zellkulturmodell als Annäherung an die in vivo Situation. Damit sind diese Daten von großer Relevanz für ein besseres Verständnis der zellulären Strahlenempfindlichkeit auf molekularer Ebene und können entscheidend dazu beitragen die Behandlung von Tumorerkrankungen sowie die Heilungschancen der Patienten zu verbessern.

Thesis (Ph. D.)--University of Washington, 2001. / Vita. Includes bibliographical references (leaves 72-93).

Epigenetics is an important layer of transcriptional control necessary for cell-type specific gene regulation. We developed computational methods to analyze the combinatorial effect and large-scale organizations of genome-wide distributions of epigenetic marks. Throughout this dissertation, we show that regions containing multiple genes with similar epigenetic patterns are found throughout the genome, suggesting the presence of several chromatin domains. In Chapter 1, we develop a hidden Markov model (HMM) for detecting the types and locations of epigenetic domains from multiple histone modifications. We use this method to analyze a published ChIP-seq dataset of five histone modification marks in mouse embryonic stem cells. We successfully detect domains of consistent epigenetic patterns from ChIP-seq data, providing new insights into the role of epigenetics in longrange gene regulation. In Chapter 2, we expand our model to investigate the genome-wide patterns of histone modifications in multiple human cell lines. We find that chromatin states can be used to accurately classify cell differentiation stage, and that three cancer cell lines can be classified as differentiated cells. We also found that genes whose chromatin states change dynamically in accordance with differentiation stage are not randomly distributed across the genome, but tend to be embedded in multi-gene chromatin domains. Moreover, many specialized gene clusters are associated with stably occupied domains. In the last chapter, we develop a more sophisticated, tiered HMM to include a domain structure in our chromatin annotation. We find that a model with three domains and five sub-states per domain best fits our data. Each state has a unique epigenetic pattern, while still staying true to its domain’s specific functional aspects and expression profiles. The majority of the genome (including most introns and intergenic regions) has low epigenetic signals and is assigned to the same domain. Our model outperforms current chromatin state models due to its increased domain coherency and interpretation.