The ‘epigenome’ refers to a difference in the transcriptional behaviour of a given allele that cannot be explained by differences in the genetic code, or ‘mutation’ at that locus. The epigenome is associated with certain biochemical marks, and generally exerts a silencing effect upon the transcription of genes under its influence. In Angiosperms such as Arabidopsis thaliana, fertilisation of both the egg and central cell– referred to as ‘double fertilisation’ – giving rise to the embryo and endosperm respectively, the latter of which is one of the most important tissues in the human food chain. Upon double fertilisation the gametes of each parent are known to contribute differing epigenetic ‘imprints’, where one gamete contributes a copy of a given allele in a transcriptionally unavailable state, while the other parent’s copy is in an available state. When a gene resides in such a region, the result is that only one parent’s copy is hence transcribed; such a gene is said to be ‘imprinted.’ Imprinting is known to affect the development of the placenta as well as some of the adult tissues in mammalian models, and in plants is most extensively found in the endosperm, where without the imprint of both parents the resulting seed exhibits reduced viability and defective endosperm development. A relative dearth in the number of known imprinted loci in the model angiosperm, A. thaliana makes is difficult to make reliable assessments of its role and regulation. This thesis initially aimed to extend the count of known imprinted genes using a model that proved insufficient to identify novel imprinted genes, and presents a meta-analysis showing that the reliable attribution of imprinted status to a gene is difficult using high-throughput methods as well. In addition, the further characterisation of a novel imprinted gene identified previously by this lab, MPC, with a view to acquiring a more detailed understanding of its role using mutants carrying point mutations in the MPC protein showed only a subtle phenotype to discern them from wild-type plants. There has also been recent speculation of a role for repeat elements in imprinting. This thesis presents findings suggesting that the apparent association of repeat elements with imprinted genes is an artefact rather of an association of endosperm-expressed genes with transposable elements, rather than genes that are specifically imprinted.

Eine intrauterine Gestationsdiabetes (GDM) Exposition induziert in den betroffenen Nachkommen eine lebenslang erhöhte Prädisposition für metabolische und komplexe Erkrankungen. Die Krankheitssuszeptibilität wird dabei durch epigenetische Veränderungen vermittelt, die sich über die Regulation der Genaktivität auch auf das Expressionsniveau und den Phänotypen auswirken. Um neue Gene zu finden, die eine Rolle in der fetalen Programmierung spielen, wurden in dieser Arbeit genomweite Methylierungsmuster von Nabelschnurbluten (FCBs) aus GDM-Schwangerschaften und Kontrollen miteinander verglichen. Mit Illumina Infinium HumanMethylation 450K Arrays konnten signifikante Gruppenunterschiede für insgesamt 65 CpG-Stellen (52 davon genassoziiert) festgestellt werden, die multiplem Testen standhielten. Mittels Pyrosequenzierung wurden vier dieser Kandidaten-Loci (ATP5A1, MFAP4, PRKCH, SLC17A4), sowie ein Gen aus der Literatur (HIF3A) genauer untersucht und die Effekte erfolgreich validiert. Für das zugrundeliegende multivariate Regressionsmodell wurden die potenziellen Störfaktoren Gestationsalter, kindliches Geschlecht und mütterlicher BMI berücksichtigt. Der GDM-Effekt zeigte sich stärker in der insulinbehandelten Subgruppe (I-GDM) als in der diätisch behandelten (D GDM) und scheint insgesamt multifaktoriell bedingt zu sein, da viele Gene betroffen waren, jedoch alle mit einer vergleichsweise niedrigen Effekt-Größe. Zusätzlich konnten für den MEG3 Promotor, MEST und PEG3, drei von vier geprägten Genen, die mittels Deep Bisulfite Sequencings (DBS) analysiert wurden, ebenfalls signifikante Methylierungs-unterschiede zwischen der GDM- und Kontroll-Gruppe detektiert werden. Die identifizierten Gene stellen labile Zielregionen für die GDM-induzierte intrauterine Programmierung dar und können zukünftig nützliche Biomarker für Krankheitsdiagnosen und Prognosen sein. Mittels DBS können darüber hinaus Einzelmolekül-Analysen durchgeführt werden, für die in differentiell methylierten Regionen (DMRs) anhand eines informativen SNPs die parentale Allel-Herkunft bestimmt und bei der Berechnung von Epimutationsraten einbezogen werden kann. Epimutationen wurde als solche gewertet, wenn sie ein > 50 % abnormal (de)methyliertes Methylierungsprofil aufwiesen. Die DBS-Daten wurden mit zwei verschiedenen Sequenzierplattformen generiert (Roche GS Junior und Illumina MiSeq). Für Zweitere wurde ein eigenes, unabhängiges Library-Präparations-Protokoll entwickelt. In Nabelschnurblut, adultem Blut und Viszeralfett wurden für die paternal exprimierte MEST Promotor DMR und die maternal exprimierte MEG3 intergenic (IG) DMR hohe Epimutationsraten für das jeweils unmethylierte Allel detektiert. Die geprägten (methylierten) Allele wiesen dagegen nur niedrige Epimutationsraten auf. Da MEST und MEG3 invers geprägte Gene sind, war die Hypermethylierung des nicht geprägten Allels (HNA) demnach unabhängig von der parentalen Allel-Herkunft. Die HNA scheint außerdem erst nach der Fertilisation aufzutreten, da in Spermien nur sehr wenige Epimutationen gefunden wurden. Für die sekundäre MEG3 Promotor DMR (deren Prägung von der primären MEG3 IG-DMR reguliert wird) wurde ein deutlich schwächerer, wenngleich signifikanter HNA-Effekt im FCB gemessen, für die paternal exprimierte PEG3 Promotor DMR konnte dagegen kein signifikanter Unterschied zwischen den beiden parentalen Epimutationsraten festgestellt werden. Der HNA-Effekt für die MEST DMR, MEG3 IG-DMR und MEG3 Promotor DMR war weder mit GDM noch mit Adipositas assoziiert und zeigte allgemein eine große interindividuelle Varianz. Die Aufrechterhaltung differenzieller Methylierungsmuster in Imprinting Kontrollregionen (ICRs) scheint in manchen Entwicklungs-Zeitspannen von großer Bedeutung und damit streng kontrolliert zu sein, später jedoch redundant zu werden, was sich in der Anreicherung von stochastischen sowie umweltinduzierten Fehlern auf dem nicht geprägten Allel äußern kann. HNA-suszeptible geprägte Gene ähneln in mancherlei Hinsicht metastabilen Epiallelen. Diese Studie zeigt, dass sowohl stochastische Faktoren als auch Umweltstimuli während der frühen embryonalen Entwicklung u.a. über HNA-Effekte geprägte Gen-Netzwerke programmieren, die in Wachstumsprozesse involviert sind. Um tiefere Einblicke in allelspezifische Prägungsprofile zu erhalten, wären umfangreiche DBS HNA-Längsschnittstudien aller 50-100 human geprägten Gene in unterschiedlichen Gewebetypen und Differenzierungsstadien wünschenswert. / Intrauterine exposure to gestational diabetes mellitus (GDM) induces lifelong increased predisposition for metabolic and complex diseases in the exposed progeny. The elevated disease susceptibility is transmitted via epigenetic alterations that influence gene expression levels and phenotypes through regulation of gene activity. Genome-wide methylation profiles of fetal cord bloods (FCBs) were investigated in GDM and control pregnancies in order to identify new genes susceptible to fetal programming. After multiple testing correction, we found 65 significantly differentially methylated CpG sites between GDM and control groups (52 of which were gene associated) within the Illumina Infinium HumanMethylation 450K array data. Using pyrosequencing, we successfully confirmed the observed results in four of these candidate loci (ATP5A1, MFAP4, PRKCH, SLC17A4) and one gene from the literature (HIF3A). A multivariate regression model was adjusted for the confounding factors gestational age, fetal sex and maternal BMI. The GDM effect was stronger within the insulin treated subgroup (I-GDM) compared to the dietary subgroup (D GDM), suggesting that GDM is a multifactorial disease evidenced by changes of small effect size in multiple genes. Significant mean methylation differences were detected between the GDM group and controls in three out of four imprinted genes (MEG3 promoter, MEST and PEG3) that were analyzed with Deep Bisulfite Sequencing (DBS). The identified genes represent labile target regions for GDM-induced intrauterine programming and could represent future biomarkers for disease diagnosis and prognosis. Furthermore, DBS enables sequencing at a single allele resolution and the calculation of allele specific epimutation rates by differentiating the parental origin of alleles via an informative SNP within differentially methylated regions (DMRs). Epimutations were characterized as alleles showing > 50 % aberrantly (de)methylated CpG sites. DBS data were generated using two different sequencing platforms (Roche GS Junior and Illumina MiSeq). An independent library preparation protocol was established for Illumina MiSeq sequencing. The paternally expressed MEST promoter DMR and the maternally expressed MEG3 intergenic (IG) DMR showed high epimutation rates for the unmethylated alleles in FCB, as well as adult blood and visceral adipose tissue. On the contrary, only minor epimutation rates were displayed by the imprinted (methylated) alleles. Thus, hypermethylation of the non-imprinted allele (HNA) was independent of parental origin, as MEST and MEG3 are opposingly imprinted genes. Very low epimutation rates in sperm indicate that the HNA effect arises after fertilization. A weak but significant HNA was also found for the secondary MEG3 promoter DMR (which is known to be regulated by the MEG3 IG-DMR). The paternally expressed PEG3 promoter DMR showed no HNA and no difference in parental epimutation rates. The observed HNA effect (for the MEST DMR, the MEG3 IG-DMR and the MEG3 promoter DMR) was neither associated with GDM nor obesity and exhibited a large interindividual variance. Maintenance of differential methylation profiles in imprinting control regions (ICRs) seems to be of great importance during some developmental periods and is therefore strictly controlled in germ cells. Later on, it might become redundant manifested in the accumulation of stochastic as well as environmentally-induced errors on the non-imprinted allele. There is evidence that HNA-susceptible imprinted genes resemble metastable epialleles in many aspects. Therefore, we suggest that stochastic as well as environmental stimuli program imprinted gene networks that are important for growth related processes during early development using HNA. Further longitudinal studies of all 50 – 100 imprinted genes would benefit in a deeper insight in allele-specific imprinting patterns of various human tissues.

Currently, very little information is available for an in-depth understanding of the molecular binding interactions with molecularly imprinted polymers (MIPs). To address this issue MIPs that have high binding affinities for their template compounds were made so that the nature of these interactions could be elucidated using spectroscopic techniques. 12 functional MIPs were prepared using a series of azobenzene and anthracenyl derivatives as the templates. Affinities of these MIPs for the corresponding templates and analogues were determined by performing batch and competitive binding tests. It was found that extensively conjugated compounds that contain at least two OH groups, an electron-withdrawing substituent and have limited conformational freedom were effective templates. The most efficient MIP, M34, was prepared with 4-[(4-nitrophenyl)azo]-1,2-benzenediol (12). M34 exhibited high affinities for azobenzene derivatives of catechol, and bound those that did not contain non electron-withdrawing substituents more specifically. M34 did not lose affinity for 12 in the presence of analogues, and vice versa, in competitive binding tests. These observations suggested a distribution of different binding sites on M34. M34 bound substrates rapidly, which was attributed to its highly porous polymer matrix giving ready access to binding sites. Formation of the porous matrix was facilitated by the use of DMF as the porogen in the preparation of M34. DMF is not a conventional choice of porogen because use of such highly polar H-bonding solvents is thought to disrupt complexation between template and polymer precursors, which is required for the formation of binding sites. Significant changes in the wavenumbers and the intensities of absorption bands assigned to the catechol substructure of 12 were observed in the FT-Raman spectra of 12 bound to M34. These findings suggested that the catechol substructure was responsible for interactions of 12 with M34 that are critical to rebinding and imprinting. In-situ analyses of dithranol (8) being removed from and bound to its MIP, M23, were performed using ATR-IR spectroscopy. Only one band, assigned to the aromatic substructure of 8, was not obstructed by solvent bands in the spectra of unwashed M23 and washed M23 that was treated with the rebinding solution. The wavenumbers of the corresponding bands in the two spectra were significantly different. This observation suggested that there were differences in the vibrational characteristics of 8 bound to M23 under the two conditions. Evidence was found for H-bonding between OH groups of 8 and C=O group of methacrylic acid using transmission FT-IR spectroscopy. However, no evidence was found that showed significant interactions between 12 and 2-vinylpyridine. Methacrylic acid and 2-vinylpyridine were used as the functional monomers in the preparations of M23 and M34. The FT-IR spectra of mixtures of 12 and 4-vinylpyridine showed three new bands assigned to H-bonded OH stretches. These observations indicated that 4-vinylpyridine H-bonds with 12, and would be a more effective functional monomer than 2-vinylpyridine in the preparation of the MIPs for 12. Titration of 12 with 2-vinylpyridine was analysed by �H NMR spectroscopy. Only small changes to the signals of the corresponding compounds were observed. This lack of change was attributed to the use of d₇DMF, which would compete against 2-vinylpyridine for H-bonding interactions. The findings made using ATR-IR spectroscopy and FT-Raman were novel because previously reported data on bound templates obtained using the corresponding techniques did not show changes in the vibrational characteristics of templates as they bind to MIPs. This investigation has shown that rebinding and spectroscopic studies can provide information about the nature of the binding interactions in MIPs.

The outcome of an infection is often determined by the qualitative nature of the immune response generated against the infectious agent. Various intracellular pathogens, including those that cause leprosy, tuberculosis, leishmaniasis, and most probably malaria and AIDS appear to require a predominant cell-mediated, Th1, response for effective containment, whereas the generation of a mixed Th1/Th2 or predominantly Th2 response is associated with progressive disease. Therefore, any attempt to develop universally efficacious vaccination against these pathogens must generate an immunological imprint that ensures a strong and stable cell-mediated response upon natural infection with the relevant pathogen. We report here critical tests of a strategy designed to achieve such an imprint using Bacille-Calmette-Guérin (BCG) vaccine. BCG vaccine is an attenuated form of M. bovis, the causative agent of tuberculosis in cattle, and is the most widely used vaccine in humans. However, considerable uncertainty still surrounds its efficacy against tuberculosis both in man and animals. As the protective dose is not known, BCG has been given at the maximum tolerable dose. However, recent studies in mice and other animals have shown that the dose of an antigen can be a critical factor in determining the type of immune response generated. I tested the general hypothesis that low dose vaccination would preferentially induce cell-mediated immune response and generate a Th1 imprint that can protect the host against intracellular pathogens in the particular case of mycobacteria. To this end, both adult and newborn mice were vaccinated with different doses of BCG or saline and cell-mediated and humoral immune responses were assessed at different times post-vaccination. Several weeks after vaccination, mice from each group were challenged with a dose of BCG that induces a mixed Th1/Th2 response in naïve mice, and the T-cell and antibody responses were assessed using ELISPOT and ELISA assays, respectively. The splenic bacterial burden was also determined using colony formation on agar plates. <p>Our results show that the class of immunity induced by BCG depends on the dose employed for vaccination, independent of the route of administration and the age and strain of mice used. In all cases, lower doses induce an exclusive cell-mediated, Th1, response with no antibody production, while higher doses induce either a mixed Th1/Th2 response or a predominantly Th2, humoral response, with higher titers of both IgG1 and IgG2a antibodies. Following intravenous high dose BCG challenge, all mice in the vaccinated groups developed a Th1 response associated with a more efficient clearance of BCG from the spleen. The greatest clearance of mycobacteria was generated following vaccination with lower doses, as low as 33 cfu of BCG. In addition, our findings demonstrate that newborn mice are not inherently biased towards generating Th2 responses, but they can generate Th1 responses and Th1 imprints if appropriate vaccination protocols (dose, route and time) are employed. Furthermore, subcutaneous exposure of young mice to environmental mycobacteria can induce a mixed Th1/Th2 response that can abrogate the potential to generate Th1 responses and Th1 imprints upon vaccination with low doses of BCG vaccine. Low dose neonatal BCG vaccination can circumvent the interference caused by impingement of environmental mycobacteria on the immune system. Therefore, our observations strongly support a neonatal low dose BCG vaccination strategy to provide universally efficacious protection against infections by pathogenic mycobacteria.

The outcome of an infection is often determined by the qualitative nature of the immune response generated against the infectious agent. Various intracellular pathogens, including those that cause leprosy, tuberculosis, leishmaniasis, and most probably malaria and AIDS appear to require a predominant cell-mediated, Th1, response for effective containment, whereas the generation of a mixed Th1/Th2 or predominantly Th2 response is associated with progressive disease. Therefore, any attempt to develop universally efficacious vaccination against these pathogens must generate an immunological imprint that ensures a strong and stable cell-mediated response upon natural infection with the relevant pathogen. We report here critical tests of a strategy designed to achieve such an imprint using Bacille-Calmette-Guérin (BCG) vaccine. BCG vaccine is an attenuated form of M. bovis, the causative agent of tuberculosis in cattle, and is the most widely used vaccine in humans. However, considerable uncertainty still surrounds its efficacy against tuberculosis both in man and animals. As the protective dose is not known, BCG has been given at the maximum tolerable dose. However, recent studies in mice and other animals have shown that the dose of an antigen can be a critical factor in determining the type of immune response generated. I tested the general hypothesis that low dose vaccination would preferentially induce cell-mediated immune response and generate a Th1 imprint that can protect the host against intracellular pathogens in the particular case of mycobacteria. To this end, both adult and newborn mice were vaccinated with different doses of BCG or saline and cell-mediated and humoral immune responses were assessed at different times post-vaccination. Several weeks after vaccination, mice from each group were challenged with a dose of BCG that induces a mixed Th1/Th2 response in naïve mice, and the T-cell and antibody responses were assessed using ELISPOT and ELISA assays, respectively. The splenic bacterial burden was also determined using colony formation on agar plates. <p>Our results show that the class of immunity induced by BCG depends on the dose employed for vaccination, independent of the route of administration and the age and strain of mice used. In all cases, lower doses induce an exclusive cell-mediated, Th1, response with no antibody production, while higher doses induce either a mixed Th1/Th2 response or a predominantly Th2, humoral response, with higher titers of both IgG1 and IgG2a antibodies. Following intravenous high dose BCG challenge, all mice in the vaccinated groups developed a Th1 response associated with a more efficient clearance of BCG from the spleen. The greatest clearance of mycobacteria was generated following vaccination with lower doses, as low as 33 cfu of BCG. In addition, our findings demonstrate that newborn mice are not inherently biased towards generating Th2 responses, but they can generate Th1 responses and Th1 imprints if appropriate vaccination protocols (dose, route and time) are employed. Furthermore, subcutaneous exposure of young mice to environmental mycobacteria can induce a mixed Th1/Th2 response that can abrogate the potential to generate Th1 responses and Th1 imprints upon vaccination with low doses of BCG vaccine. Low dose neonatal BCG vaccination can circumvent the interference caused by impingement of environmental mycobacteria on the immune system. Therefore, our observations strongly support a neonatal low dose BCG vaccination strategy to provide universally efficacious protection against infections by pathogenic mycobacteria.

<p>Genomic imprinting is the parent-of-origin dependent monoallelic expression of select developmentally important genes that are regulated by epigenetic mechanisms. It is believed to have co-evolved with placentation in the Therian lineage, but it is unclear whether this phenomenon arose in a convergent or divergent manner in the Metatherians (those with a rudimentary placenta) and Eutherians (true placental mammals). Moreover, the precise epigenetic mechanisms involved in establishing genomic imprinting (DNA methylation or histone modifications) are still poorly defined. Thus, I studied Metatherian orthologues of Eutherian imprinted loci using Monodelphis domestica as a model organism. L3MBTL and HTR2A were monoallelically expressed; PEG1/MEST had one imprinted and one non-imprinted transcript, while IMPACT, COPG2 and PLAGL1 were not imprinted, thus revealing that this phenomenon is conserved at some, but not all loci between the two groups of Therians. Moreover, differential methylation patterns and the associated regulatory non-coding RNA are also not conserved amongst them, exemplified by the novel DMR identified within IGF2R which had no associated anti-sense transcript. However, histone modifications, specifically the activating H3 Lysine 4 dimethylation mark at the active allele's promoter seems to be important in both lineages and probably serves as the primordial imprint mark. Although the evidence does not resolve the issue of convergence or divergence, it raises the intriguing possibility that both forms of evolution occurred during establishment of imprinting in these mammals.</p> <p>The imposition of functional haploidy in the genome by such epigenetic mechanisms necessarily makes imprinted genes more susceptible to deleterious mutations and regulatory perturbations. Thus, imprinting is implicated in a number of developmental disorders, but its role in the etiology of complex human diseases and neurological disorders, like autism and schizophrenia, remains to be determined. I chose to investigate the imprint status of the duplicated locus DGCR6/DGCR6L lying within the 22q11.2 microdeletion causative of DiGeorge Syndrome (DGS), because our lab previously predicted genes at this genomic location to be imprinted. My studies revealed that both genes DGCR6 and DGCR6L are monoallelically expressed in the primate lineage, but not in a parent-of-origin dependent manner. Interestingly, DGCR6L is not present in the mouse, and Dgcr6 is expressed from both parental alleles.</p> <p>Although DGS primarily manifests as facial, limb and heart abnormalities in children, a number of these patients also ultimately present with variable neurocognitive defects. Thus, I focused my studies on determining the effect of the microdeletion at this chromosomal region on DGCR6 and DGCR6L expression because of their potential role in neural crest cell migration. This revealed that DGS subjects have a highly dysregulated pattern of DGCR6 and DGCR6L expression as compared to that in controls. Moreover, increased expression of these genes correlated significantly with decreased performance in sustained-attention tests. This provides the first evidence that disruption of the normal monoallelic expression pattern of DGCR6 and DGCR6L by hemizygous deletion is involved in the variability in neurocognitive symptoms associated with DiGeorge Syndrome. The results of my studies highlight the importance of searching for novel imprinted domains to better understand not only their evolution, but also the potential role of such epigenetically labile regions in modulating complex human diseases and neurological disorders.</p> / Dissertation

The advent of somatic cell nuclear transfer in cattle has provided the opportunity for researchers to generate genetically identical animals as well as animals that possess precise genetic modifications for agriculture and biomedical purposes. However, in spite of the revolutionary impact this technology presents, problems remain which hinder the production of healthy animals on a consistent basis. Research on cloned mice implicates improper reprogramming of epigenetic modifications and genomic imprinting for the low pregnancy rates and high incidence of abnormalities that are manifested in cloned animals; however, a systematic and comprehensive analysis of nuclear reprogramming in cloned cattle remains undone. The purpose of this research is to assess and characterize the patterns of genomic imprinting in normal and nuclear transfer derived bovine fetuses. To facilitate the identification of imprinted genes in the bovine, a Bos gaurus/Bos taurus interspecies model has been incorporated to maximize the genetic heterozygosity that exists between the alleles of putative imprinted genes for allelic discrimination and parental inheritance. The sequence of twenty-six genes, previously reported as imprinted in mice and humans, was analyzed in Bos gaurus (Gaur) and Bos taurus (Angus) cattle for the presence of single nucleotide polymorphisms (SNP). SNPs were detected in the Gene trap locus 2 (GTL2), Insulin like growth factor 2 (IGF2), Wilms tumor 1 (WT1) and the X chromosome inactivation specific transcript (XIST). Allelic expression analysis in interspecies hybrids indicated maternal genomic imprinting at the IGF2 and XIST loci, paternal genomic imprinting at the GTL2 locus and no imprinting at the WT1 locus. Analysis in cloned hybrids indicated fidelity of allelic expression at the IGF2 and GTL2 loci, however disruption of imprinting was observed at the XIST locus in the placenta of clones. These results are the largest identification of imprinted genes in the bovine and the first identification of the disruption of an imprinted gene in an animal derived from somatic cell nuclear transfer.

Thesis (Ph. D.)--West Virginia University, 2003. / Title from document title page. Document formatted into pages; contains xi, 204 p. : ill. (some col.) Includes abstract. Includes bibliographical references.

published_or_final_version / Medical Sciences / Master / Master of Medical Sciences

Not available / text