| 總結: | <p>Chromatin organization is a critical regulator of gene expression and cell phenotype, and is frequently dysregulated in cancer. Endometrial cancer (EC) is the most common gynecological malignancy, and casues significant morbidity and mortality. EC is notable for recurrent alterations in chromatin including the <em>ARID1A</em> gene – a key component of the SWI/SNF remodeling complex – has emerged as a prevalent driver in EC, along with other remodelers such as <em>CHD4</em> and <em>BCOR</em>. However, a systematic analysis of chromatin modifier alterations and their functional consequences in EC has not been done.</p>
<p>This thesis presents a comprehensive investigation of genomic alterations in chromatin modifiers using whole genome sequencing (WGS) data from the Genomics England(GEL) 100,000 Genomes Project, the largest EC cohort to date. I demonstrate that while mutational processes vary across molecular subtypes, numerous chromatin modifiers are consistently altered across all subtypes. These genomic alterations frequently occur in different subunits of the same complex, such as alterations in <em>CHD3</em>, <em>CHD4</em> and <em>MBD3</em>, subunits of the ATP-dependent chromatin remodeling complex NuRD. Additionally, I examine the correlation between driver mutations and patient survival, revealing that mutations in <em>PBRM1</em> and <em>CHD4</em> are associated with an increased risk of death after accounting for age, molecular subtype, and tumor mutation and copy number alteration burden.</p>
<p>To complement the correlative analysis, I employ CRISPR-Cas9 gene editing to study the functional consequences of perturbations in selected chromatin modifiers (<em>ARID1A, ARID1B, ARID5B, EP300, KMT2C,</em> and <em>SETD1B</em>) in normal and malignant endometrial cells using transcriptomic and chromatin accessibility data. Furthermore, I explore the implications of the N1459S <em>BCOR</em> mutation, a hotspot mutation near-unique to EC.</p>
<p>Considering the frequent occurrence of <em>ARID1A</em> mutations in malignant tissues and their absence in normal endometrium, I investigate tumor heterogeneity in endometrial cancer. I discuss the limitations of current methodologies and propose a deep learning approach to uncover the hidden evolutionary trajectories. With additional research, this approach could potentially facilitate understanding the sequence in which driver alterations occur.</p>
<p>In summary, this work presents a resource for investigating chromatin organization in EC. The functional analyses using gene editing techniques confirm that EC-associated drivers disrupt essential cellular processes involved in oncogenesis. By providing the first systematic correlative and functional analyses of chromatin modifiers in EC, this thesis offers novel insights into EC biology.</p>
|
|---|