Die Entdeckung adulter dentaler Stammzellen eröffnete ein neues Forschungsfeld im Hinblick auf die Regeneration dentaler Gewebe. Bisher liegen nur wenige Studien vor, in denen das Genexpressionsprofil dentaler Stammzellen im Vergleich zu den Knochenmarkstammzellen analysiert wurde. Diese Untersuchungen wurden vorwiegend an Mischkulturen vorgenommen. Im Gegensatz dazu war es daher das Ziel der vorliegenden Arbeit, das Genexpressionsprofil einer bestimmten Stammzell-Population, nämlich der Stro1-positiven pulpalen mesenchymalen Zahnkeimstammzellen (Stro1+ZK) im Vergleich zu Stro1-positiven mesenchymalen Knochenmarkstammzellen (Stro1+BK), zu untersuchen. Die Genexpression beider Zelltypen wurde anhand von Microarrays ermittelt. Insgesamt gingen 22.454 Gene in die Auswertung ein, wovon bei einem konservativ festgesetzten Schwellenwert einer FDR≤1% 2730 Gene eine hochsignifikant differentielle Expression zeigten. Die Analyse dieser differentiell exprimierten Gene mithilfe der Programme „DAVID“ und „Ingenuity“ ergab, dass in den Stro1+ZK vermehrt Gene heraufreguliert sind, die mit Zellfunktionen wie beispielsweise Proliferationsregulation, der Zell-zu-Zell-Signalleitung und der Organisation des Zytoskeletts verknüpft sind. Die Stro1+BK hingegen exprimieren verstärkt Gene, die mit der Organisation der extrazellulären Knochenmatrix und Zell-Adhäsion assoziiert sind. Des Weiteren findet sich in diesen Zellen eine verstärkte Expression von Genen, die mit der Struktur- und Formgebung des Skeletts in Verbindung stehen. Trotz identischem Stammzellmarker-Typus (Stro1) weisen die untersuchten mesenchymalen Stammzelltypen stark unterschiedliche Hox-Gen-Signaturen auf. Dabei zeigt sich sowohl eine Variation in Anzahl und Art der Hox-Gene als auch in deren Expressionsmuster. Stro1+BK exprimieren verstärkter Hox-Gene der Cluster A bis D (HOXA-D), die Segment- und Positionsinformationen codieren. Hingegen sind in den Stro1+ZK die Hox-Gene BARX1, MSX1, MSX2, DLX1, DLX2, PAX9 und LEF1 hochreguliert, welche eine tragende Rolle in der Zahnentwicklung spielen. So ist z.B. bereits bekannt, dass Mutationen dieser Gene zu Fehlbildungen von Zähnen wie Aplasien oder Hypoplasien führen können. Die vorliegende Arbeit zeigt, dass insbesondere hinsichtlich der Hox-Gene signifikante Unterschiede zwischen den Stro1+ZK und Stro1+BK bestehen. Weiterführende Experimente zur Aufklärung der Funktionsweise von Genen, die in den Stro1+ZK von Bedeutung sein könnten, einschließlich deren Erforschung auf proteinbiochemischer und zellbiologischer Ebene, wären wünschenswert.

Tuberculosis, caused by infection with Mycobacterium tuberculosis (M.tb), remains a global health problem. Drug resistance and high rates of HIV infection have fuelled the pandemic and, although a vaccine exists, its ability to protect from pulmonary tuberculosis varies between 0 and 80%. Bacille Calmette Guerin (BCG) has been administered to billions worldwide yet its protective mechanisms remain unknown, as do the reasons for its failure to protect in many parts of the world. Modified Vaccinia virus Ankara expressing antigen 85A (MVA85A) is a novel candidate vaccine designed to boost immune responses to BCG and improve protection. An aim of this thesis has been to characterise the innate immune response to an MVA85A boosting vaccination in both UK adults and South African infants. In the former, volunteers develop a strong innate response following vaccination however this does not always translate into a robust adaptive response to antigen 85A (Ag85A), which is determined in part by Treg expansion and the nuclear protein HMGB1 signaling through the TLR1-2-6 axis. By contrast, not all South Africa infants mount a strong innate immune response to MVA85A yet this response is correlated with the magnitude of the adaptive response. The immune response to BCG in both populations is also characterised and an association found between increased production of IL-17, IL-22 and IFN-γ in response to BCG stimulation and control of mycobacterial growth. The results presented here further the knowledge on the links between innate and adaptive responses to vaccination with BCG and MVA85A and the variation in mechanisms involved in different populations.

Tuberculosis (TB) is still a global epidemic disease despite its discovery over 100 years ago. It is caused by Mycobacterium tuberculosis, which invades and replicates within macrophages, key cells of the innate immune system. The hallmark of tuberculosis is the granuloma which is an accumulation of Mycobacterium-infected cells surrounded by immune cells, and the containment of the bacteria is assured as long as the host immune response remains intact. Despite a well-developed immune response in the infected host, reactivation of latent tuberculosis infection (LTBI) may occur through the introduction of other bacterial pathogens, re-infection with M. tuberculosis or due to other immunosuppression, e.g. AIDS or cancer. The zebrafish–M. marinum model provides an ideal system for examining the pathogenesis of tuberculosis and the associated immune response of the host due to its vertebrate-like immune system, and the close phylogenetic relationship of M. marinum to M. tuberculosis. Granuloma formation and immune response to M. marinum have been investigated mainly in zebrafish embryos or larvae, which lack an adaptive immune response, and little work has been performed in adult fish. This complicates the transfer of findings in these models to chronic, latent or re-activated disease stages in humans, where adaptive immunity plays an important part. The aim of the research presented here was to investigate the immune response of the adult zebrafish to M. marinum infection, with the focus on the kidney as one of the major immune organs in fish. The results obtained support further use of the adult zebrafish-M. marinum model for human tuberculosis infections in the future. In the present study, adult zebrafish were infected with low doses of M. marinum (NCIMB 1297 or NCIMB 1298) and the kidney was investigated for histopathological changes in the form of granulomas over a period of two months(Chapter 3). No granulomas were detected in the fish infected with M. marinum NCIMB 1298 while in zebrafish infected with NCIMB 1297, macrophage aggregation and granuloma formation were detected as early as day 11 post-infection. Occurrence and severity of granulomas and the presence of replicating bacteria increased over time, resulting in a high density of non-caseating and caseating granulomas in the head and posterior kidney after two months of infection. Interleukin 1 beta (IL-1β), Interleukin-12 (IL-12), Tumor necrosis factor alpha (TNFα) and Interferon gamma (IFNγ) have been shown to be important cytokines functioning in defence against tuberculosis, especially IFNγ which is considered to play an important part in acute, chronic and latent tuberculosis. Changes in gene expression of these immune genes in adult zebrafish were investigated over the first two weeks of infection with M. marinum NCIMB 1298 and NCIMB 1297. The results obtained in the first week after infection were inconclusive for both strains investigated. In agreement with the results presented in Chapter 3, no specific immune response was detectable in fish infected with M. marinum NCIMB 1298. However, after 14 days, a high-fold change in IL-12 and TNFα expression were detected in fish infected with M. marinum NCIMB 1297, while IL-1β showed no changes compared to the control fish. Furthermore, no IFNγ expression was detectable over the first two weeks of infection. The delay in the expression of IL-12 and the lack of IFNγ expression can be explained by the ability of M. marinum to manipulate the host immune response, as described for M. tuberculosis and other intracellular bacteria. Besides in vivo investigations of the host-pathogen interactions, in vitro primary macrophage cultures from individual zebrafish kidneys were developed to investigate macrophage-specific gene expression to M. marinum infection (Chapter 4). Although the results looked promising, further optimization is required before the results of the in vitro assays can be fully compared to the in vivo results. Our understanding of reactivation in latent tuberculosis infection (LTBI) both in healthy and immune compromised individuals is insufficient and is delaying the development of treatments for the disease. Therefore, the transcriptome profile of long-term infections (26 weeks) with M. marinum NCIMB 1297 in adult zebrafish was investigated to determine whether the gene expression in this model is comparable to LTBI in humans or other vertebrate model organisms (Chapter 5). In addition, transcriptome profiling was investigated in a group of long-term infected zebrafish exposed to stress to induce re-activation of the disease. Expression profiles in the long-term infected fish and the infected plus stressed fish differed from each other and displayed similar gene profiles to those found in the latent or re-activated disease states, respectively, in human and other vertebrate models. Infected fish displayed a profile highlighted by IFNγ, TNFα, NOS2b and IL-8 expression alongside activating and regulatory T cell responses, including involvement of cytotoxic T cells (CTLs). The transcriptome profile of the group of fish that had been infected and then stressed was distinguished by the lack of IFNγ expression and reduction in TNFα and NOS2b expression, as well as a lack of T cell response compared to the infected fish. In conclusion, the results obtained from Chapters 3 and 4 showed that M. marinum NCIMB 1298 is non-pathogenic to zebrafish. Infection with M. marinum NCIMB 1297, on the other hand, resulted in a similar immune response to that described for human and other mammalian vertebrate models (Chapters 3-5). These results support the use of the adult zebrafish-M. marinum model to investigate LTBI and disease reactivation, and will aid our understanding host-pathogen interactions for tuberculosis in the future.

Receiver operating characteristic (ROC) curves are widely used in medical decision making. It was recognized in the last decade that only a specific region of the ROC curve is of clinical interest, which can be summarized by the partial area under the ROC curve (partial AUC). Early statistical methods for evaluating partial AUC assume that the data are from a specified underlying distribution. Nonparametric estimators of the partial AUC emerged recently, but there are theoretical issues to be addressed. In this dissertation, we propose two new nonparametric statistics, partially integrated ROC and partially integrated weighted ROC, for estimating partial AUC. We show that our partially integrated ROC statistic is a consistent estimator of the partial AUC, and derive its asymptotic distribution which is distribution free under the null hypothesis. In the partially integrated ROC statistic, when the ROC curve crosses the Uniform distribution function (CDF) and if the partial area evaluated contains the crossing point, or when there are multiple crossing, the partially integrated ROC statistic might not perform well. To address this issue, we propose the partially integrated weighted ROC statistic. This statistic evaluates the partially weighted AUC, where larger weight is given when the ROC curve is above the Uniform CDF and smaller weight is given when the ROC curve is below the Uniform CDF. We show that our partially integrated weighted ROC statistic is a consistent estimator of the partially weighted AUC. We derive its asymptotic distribution which is distribution free under the null hypothesis. We propose to apply our two nonparametric statistics to functional category analysis in microarray experiments. We define the functional category analysis to be the statistical identification of over-represented functional gene categories in a microarray experiment based on differential gene expression. We compare our statistics with existing methods for the functional category analysis both via simulation study and application to a real microarray data, and demonstrate that our two statistics are effective for identifying over-represented functional gene categories. We also emphasize the essential role of the empirical distribution function plots and the ROC curves in the functional category analysis.

This dissertation includes three topics. First topic: Regularized estimation in the AFT model with high dimensional covariates. Second topic: A novel application of quantile regression for identification of biomarkers exemplified by equine cartilage microarray data. Third topic: Normalization and analysis of cDNA microarray using linear contrasts.

The endoplasmic reticulum (ER) is the organelle responsible for synthesis and folding of secreted and membranous protein and lipid biosynthesis. It also functions as one of the main cellular calcium stores. Pancreatic beta-cells evolved to produce and secrete insulin upon demand in order to regulate blood glucose homeostasis. In response to increases in serum glucose, insulin synthesis represents nearly 50% of the total protein biosynthesis by beta-cells. This poses an enormous burden on the ER, rendering beta-cells vulnerable to agents that perturb ER function. Alterations of ER homeostasis lead to accumulation of misfolded proteins and activation of an adaptive response named the unfolded protein response (UPR). The UPR is transduced via 3 ER transmembrane proteins, namely PERK, IRE-1 and ATF6. The signaling cascades activated downstream of these proteins: a) induce expression of ER resident chaperones and protein foldases. Increasing the protein folding capacity of the ER; b) attenuate general protein translations which avoids overloading the stressed ER with new proteins; c) upregulate ER-associated degradation (ERAD) genes, which decreases the unfolded protein load of the ER. In severe cases, failure by the UPR to solve the ER stress leads to apoptosis. The mechanisms linking ER stress to apoptosis are still poorly understood, but potential mediators include the transcription factors Chop and ATF3, pro-apoptotic members of the Bcl-2 familly, the caspase 12 and the kinase JNK. Accumulating evidence suggest that ER stress contributes to beta-cell apoptosis in both type 1 and type 2 diabetes. Type 1 diabetes is characterized by a severe insulin deficiency resulting from chronic and progressive destruction of pancreatic beta-cells by the immune system. During this autoimmune assault, beta-cells are exposed to cytokines secreted by the immune cells infiltrating the pancreatic islets. Our group has previously shown that the pro-inflamatory cytokines interleukin-1beta (IL1-beta and interferon-gamma (IFN-gamma), via nitric oxide (NO) formation, downregulate expression and function of the ER Ca2+ pump SERCA2. This depletes beta-cell ER Ca2+ stores, leading to ER stress and apoptosis. Of note, IL1-beta alone triggers ER stress but does not induce beta-cell death, while IFN-gamma neither causes ER stress nor induces beta-cell death. Together, these cytokines cause beta-cell apoptosis but the mechanisms behind this synergistic effect were unknown. Type 2 diabetes is characterized by both peripheral resistance to insulin, usually as a result of obesity, and deficient insulin secretion secondary to beta cell failure. Obese patients have high levels of circulating free fatty acids (FFA) and several studies have shown that the FFA palmitate induces ER stress and beta-cell apoptosis. In the present work we initially established an experimental model to specifically activate the ER stress response in pancreatic beta-cells. For this purpose, insulinoma cells (INS-1E) or primary rat beta-cells were exposed to the reversible chemical SERCA pump blocker cyclopiazonic acid (CPA). Dose-response and time course experiments determined the best conditions to induce a marked ER stress without excessive cell death (<25%). The first goal of the work was to understand the synergistic effects of IL1-beta and IFN-gamma leading to pancreatic beta-cell apoptosis. Our group previously observed, by microarray analysis of primary beta-cells, that IFN-gamma down-regulates mRNAs encoding for some ER chaperones. Against this background, our hypothesis was that IFN-gamma aggravates beta-cell ER stress by decreasing the ability of these cells to mount an adequate UPR. To test this hypothesis, we investigated whether IFN-gamma pre-treatment augments CPA-induced ER stress and beta cell death. The results obtained indicated that IFN-gamma pre-treatment potentiates CPA-induced apoptosis in INS-1E and primary beta-cells. This effect was specific for IFN-gamma since neither IL1-beta nor a low dose CPA pre-treatment potentiated CPA-induced apoptosis in INS-1E cells. These effects of IFN-gamma were mediated via the down regulation of genes involved in beta cell defense against ER stress, including the ER chaperones BiP, Orp150 and Grp94 as well as Sec61, a component of the ERAD pathway. This had functional consequences as evidenced by a decreased basal and CPA-induced activity of a reporter construct for the unfolded protein response element (UPRE) and augmented expression of the pro-apoptotic transcription factor Chop. We next investigated the molecular regulation of the Chop gene in INS-1E cells in response to several pro-apoptotic and ER stress inducing agents, namely cytokines (IL1-beta and IFN-gamma), palmitate, or CPA. Detailed mutagenesis studies of the Chop promoter showed differential regulation of Chop transcription by these compounds. While cytokines (via NO production)- and palmitate-induced Chop expression was mediated via a C/EBP-ATF composite and AP-1 binding sites, CPA induction required the C/EBP-ATF site and the ER stress response element (ERSE). Cytokines, palmitate and CPA induced ATF4 protein expression and further binding to the C/EBP-ATF composite site, as shown by Western blot and EMSA experiments. There was also formation of distinct AP-1 dimers and binding to the AP-1 site after exposure to cytokines or palmitate. The third objective of this work was to obtain a broad picture of the pancreatic beta-cell molecular responses during and after (recovery period) a severe ER stress. For this purpose, we utilized an “in home” spotted microarray, the APOCHIP, containing nearly 600 probes selected for the study of beta-cell apoptosis. Time-dependent gene expression profiles were measured in INS-1E cells exposed to CPA. CPA-induced ER-stress modified expression of 183 genes in at least one of the time points studied. Most of theses genes returned to control levels 3h after CPA removal from the culture medium. We observed full beta-cell recovery and survival, indicating that these cells trigger efficient defenses against ER stress. Beta-cell recovery is associated with a sustained increase in the expression of ER chaperones and a rapid decrease of pro-apoptotic mRNAs following CPA removal. Two groups of genes were particularly affected by CPA, namely those related to the cellular responses to ER stress, which were mostly up-regulated, and those related to differentiated beta-cell functions, which were down-regulated. Among this last group, we observed a 40-90% decrease of the mRNAs for insulin-1 and -2. These findings were confirmed in INS-1E cells exposed to cytokines or thapsigargin (another SERCA blocker), and in primary beta-cells exposed to the same treatments. This decrease in insulin mRNA expression is due to transcript degradation, most probably caused by IRE-1 activation and triggering of its endoribonuclease activity, as recently described in Drosophila cells. In conclusion, our work enabled a better understanding of the pancreatic beta-cell responses to ER stress: 1.)We identified a sensitizing effect of IFN-gamma to ER stress in beta-cells via downregulation of key ER chaperones. 2.)We observed a differential regulation of Chop transcription by different treatments suggesting distinct responses of pancreatic beta-cells to diverse ER stress inducers. 3.)We provided the first global analysis of gene expression modifications in pancreatic beta-cells following ER stress. 4.)We demonstrated a high capacity of beta-cells to cope and recover from a severe ER stress. 5.)We identified a new protective mechanism against ER stress, namely the degradation of insulin mRNA which limits the load posed on the ER by insulin synthesis. This, coupled to a marked increase in ER chaperones and a fast degradation of pro-apoptotic mRNAs, enables beta cells to recover from ER stress after the causes of this stress are removed.

<p>In this thesis, methods to study gene and protein expression in cells and tissues were developed and utilized in combination with protein-specific antibodies, with the overall objective to attain greater understanding of protein function.</p><p>To analyze protein expression in <i>in vitro</i> cultured cell lines, a cell microarray (CMA) was developed, facilitating antibody-based protein profiling of cell lines using immunohistochemistry (IHC). Staining patterns in cell lines were analyzed using image analysis, developed to automatically identify cells and immunohistochemical staining, providing qualitative and quantitative measurements of protein expression. Quantitative IHC data from CMAs stained with nearly 3000 antibodies was used to evaluate the adequacy of using cell lines as models for cancer tissue. We found that cell lines are homogenous with respect to protein expression profiles, and generally more alike each other, than corresponding cancer cells <i>in vivo</i>. However, we found variability between cell lines in regards to the level of retained tumor phenotypic traits, and identified cell lines with a preserved link to corresponding cancer, suggesting that some cell lines are appropriate model systems for specific tumor types. </p><p>Specific gene expression patterns were analyzed in vitiligo vulgaris and malignant melanoma. Transcriptional profiling of vitiligo melanocytes revealed dysregulation of genes involved in melanin biosynthesis and melanosome function, thus highlighting some mechanisms possibly involved in the pathogenesis of vitiligo. Two new potential markers for infiltrating malignant melanoma, Syntaxin-7 and Discs large homolog 5, were identified using antibody-based protein profiling of melanoma in a tissue microarray format. Both proteins were expressed with high specificity in melanocytic lesions, and loss of Syntaxin-7 expression was associated with more high-grade malignant melanomas.</p><p>In conclusion, the combination of antibody-based proteomics and microarray technology provided valuable information of expression patterns in cells and tissues, which can be used to better understand associations between protein signatures and disease.</p>

Prognostic models have clinical appeal to aid therapeutic decision making. In the UK, the Nottingham Prognostic Index (NPI) has been used, for over two decades, to inform patient management. However, it has been commented that NPI is not capable of identifying a subgroup of patients with a prognosis so good that adjuvant therapy with potential harmful side effects can be withheld safely. Tissue Microarray Analysis (TMA) now makes possible measurement of biological tissue microarray features of frozen biopsies from breast cancer tumours. These give an insight to the biology of tumour and hence could have the potential to enhance prognostic modelling. I therefore wished to investigate whether biomarkers can add value to clinical predictors to provide improved prognostic stratification in terms of Recurrence Free Survival (RFS). However, there are very many biomarkers that could be measured, they usually exhibit skewed distribution and missing values are common. The statistical issues raised are thus number of variables being tested, form of the association, imputation of missing data, and assessment of the stability and internal validity of the model. Therefore the specific aim of this study was to develop and to demonstrate performance of statistical modelling techniques that will be useful in circumstances where there is a surfeit of explanatory variables and missing data; in particular to achieve useful and parsimonious models while guarding against instability and overfitting. I also sought to identify a subgroup of patients with a prognosis so good that a decision can be made to avoid adjuvant therapy. I aimed to provide statistically robust answers to a set of clinical question and develop strategies to be used in such data sets that would be useful and acceptable to clinicians. A unique data set of 401 Estrogen Receptor positive (ER+) tamoxifen treated breast cancer patients with measurement for a large panel of biomarkers (72 in total) was available. Taking a statistical approach, I applied a multi-faceted screening process to select a limited set of potentially informative variables and to detect the appropriate form of the association, followed by multiple imputations of missing data and bootstrapping. In comparison with the NPI, the final joint model derived assigned patients into more appropriate risk groups (14% of recurred and 4% of non-recurred cases). The actuarial 7-year RFS rate for patients in the lowest risk quartile was 95% (95% C.I.: 89%, 100%). To evaluate an alternative approach, biological knowledge was incorporated into the process of model development. Model building began with the use of biological expertise to divide the variables into substantive biomarker sets on the basis of presumed role in the pathway to cancer progression. For each biomarker family, an informative and parsimonious index was generated by combining family variables, to be offered to the final model as intermediate predictor. In comparison with NPI, patients into more appropriate risk groups (21% of recurred and 11% of non-recurred patients). This model identified a low-risk group with 7-year RFS rate at 98% (95% C.I.: 96%, 100%).

Microarrays emerged in the 1990s as a consequence of the efforts to speed up the process of drug discovery. They revolutionized molecular biological research by enabling monitoring of thousands of genes together. Typical microarray experiments measure the expression levels of a large numberof genes on very few tissue samples. The resulting sparsity of data presents major challenges to statistical methods used to perform any kind of analysis on this data. This research posits that phenotypic classification and prediction serve as good objective functions for both optimization and evaluation of microarray data analysis methods. This is because classification measures whatis needed for diagnostics and provides quantitative performance measures such as leave-one-out (LOO) or held-out prediction accuracy and confidence. Under the classification framework, various microarray data normalization procedures are evaluated using a class label hypothesis testing framework and also employing Support Vector Machines (SVM) and linear discriminant based classifiers. A novel normalization technique based on minimizing the squared correlation coefficients between expression levels of gene pairs is proposed and evaluated along with the other methods. Our results suggest that most normalization methods helped classification on the datasets considered except the rank method, most likely due to its quantization effects.Another contribution of this research is in developing machine learning methods for incorporating an independent source of information, in the form of gene annotations, to analyze microarray data. Recently, genes of many organisms have been annotated with terms from a limited vocabulary called Gene Ontologies (GO), describing the genes' roles in various biological processes, molecular functions and their locations within the cell. Novel probabilistic generative models are proposed for clustering genes using both their expression levels and GO tags. These models are similar in essence to the ones used for multimodal data, such as images and words, with learning and inference done in a Bayesian framework. The multimodal generative models are used for phenotypic class prediction. More specifically, the problems of phenotype prediction for static gene expression data and state prediction for time-course data are emphasized. Using GO tags for organisms whose genes have been studied more comprehensively leads to an improvement in prediction. Our methods also have the potential to provide a way to assess the quality of available GO tags for the genes of various model organisms.

A controlled and efficient coordination of gene expression is the key for normal development of an organism. In mammals, a subset of autosomal genes is expressed monoallelically depending on the sex of the transmitting parent, a phenomenon known as genomic imprinting. The imprinted state of the H19 and Igf2 genes is controlled by a short stretch of sequences upstream of H19 known as the imprinting control region (ICR). This region is differentially methylated and is responsible for the repression of the maternally inherited Igf2 allele. It harbors hypersensitive sites on the unmethylated maternal allele and functions as an insulator that binds a chromatin insulator protein CTCF. Hence the H19 ICR, which plays an important role in maintaining the imprinting status of H19 and Igf2, was shown to lose the insulator property upon CpG methylation. Another ICR in the Kcnq1 locus regulates long-range repression of p57Kip2 and Kcnq1 on the paternal allele, and is located on the neighboring subdomain of the imprinted gene cluster containing H19 and Igf2, on the distal end of mouse chromosome 7. Similarly to the H19 ICR, the Kcnq1 ICR appears to possess a unidirectional and methylation-sensitive chromatin insulator property in two different somatic cell types. Hence, methylation dependent insulator activity emerges as a common feature of imprinting control regions. The protein CTCF is required for the interpretation and propagation of the differentially methylated status of the H19 ICR. Work in this thesis shows that this feature applies genomewide. The mapping of CTCF target sites demonstrated not only a strong link between CTCF, formation of insulator complexes and maintaining methylation-free domains, but also a network of target sites that are involved in pivotal functions. The pattern of CTCF in vivo occupancy varies in a lineage-specific manner, although a small group of target sites show constitutive binding. In conclusion, the work of this thesis shows that epigenetic marks play an important role in regulating the insulator property. The studies also confirm the importance of CTCF in maintaining methylation-free domains and its role in insulator function. Our study unravels a new range of target sites for CTCF involved in divergent functions and their developmental control.