Preview

PDF, English - main document Download (11MB) | Terms of use

Abstract

Malignant transformation of cells from the lymphoid lineage can occur at various developmental stages. These stages carry distinct patterns of epigenetic modifications as a result of dynamic epigenetic programming during physiological cell maturation that enables stage specific gene regulation. Transformation frequently involves the disruption of these epigenetic patterns, which can potentially be modulated by drug treatment. Thus, epigenetic profiles of cancer cells comprise both disease specific modifications as well as signatures that reflect the original cell type and developmental stage of the cell of origin from which they arise. Deregulated epigenetic signaling via activity changes of promoters and enhancers is likely to be a major factor for the establishment and maintenance of cancer specific gene expression patterns. The activity of regulatory promoter and enhancer elements and the binding of transcription factors in the genome can be mapped by the assay for transposase-accessible chromatin using sequencing (ATAC seq). Thus, ATAC seq is a highly informative readout about the local epigenetic state of chromatin. However, the epigenetic profiles of the non-malignant tumor cell of origin needs to be considered to reveal by contrast disease specific aberrant gene regulation mechanisms and analyze the effect of drug treatment on deregulated chromatin patterns. Here, I dissected deregulated chromatin states in blood cancers by mapping the chromatin accessibility of gene regulatory elements and their relation to gene expression during physiological B-cell development, malignant transformation and drug treatment. First, I established and adjusted the experimental and bioinformatic procedures with quality controls to map chromatin accessible regions by ATAC-seq in primary human cells. By applying this method, I could demonstrate that the chromatin accessibility pattern of the genome changes during B cell development and can be used to reconstruct the developmental phylogeny of sorted healthy B cell subsets. Next, I characterized how the chromatin accessibility landscape was disrupted in chronic lymphocytic leukemia (CLL) compared to CD19+ B cells of healthy donors. The epigenetic deregulation affected in addition to many enhancer and promoter regions also loci with repressive or bivalent histone modifications. While enhancers with gained accessibility in CLL were enriched for transcription factor binding sites of NFAT and E2A, enhancers with lost accessibility were enriched for binding sites of EBF1. To further refine the definition of the physiological state, the accessibility-based phylogenetic tree of B cell development was used to infer the maturation stage and the chromatin accessibility pattern of the cell-of-origin for each individual CLL patient. The results of the cell of origin in silico modeling for individual patients were in agreement with the IGHV mutational status, an established marker for the maturation stage, and enabled a more detailed classification. The comparison of the CLL state with the direct cell-of-origin allowed to discriminate between developmental changes and disease specific aberrant accessibility. With this refined analysis I was able to follow chromatin accessibility changes in CLL patients that underwent standard treatment with ibrutinib, an inhibitor of disease enhanced B cell signaling. Regions with CLL specific aberrant chromatin accessibility approached physiological accessibility levels with the course of treatment. However, a single-cell analysis of the DNA mutations revealed changes in the clonal heterogeneity in two patients during treatment. Furthermore, the presence of an ibrutinib-resistant subclone was detected already during remission. In order to further analyze the druggability of disrupted epigenetic patterns, the effect of treatment with the histone deacetylase inhibitor panobinostat was evaluated in multiple myeloma. Panobinostat induced a global increase in histone acetylation, mainly attributable to a broadening of acetylation within the gene body of expressed genes. No significant changes of accessibility were detected. However, the degree of transcriptional drug response between different patients and multiple myeloma cell lines varied considerably. Only the transcriptionally responsive cell lines showed a G1 cell cycle arrest and an enrichment of early senescence pathway genes in the panobinostat responding genes. In summary, this thesis contributes to explore novel aspects of the deregulation of gene regulatory elements in blood cancers, which holds great potential for clinical application. The approach of identifying disease specific changes in the chromatin accessibility and transcription factor occupancy while accounting for the tumor cell of origin profiles might also help to better understand epigenetic deregulation and the effect of drugs in other tumor entities.