| Abstract |
The purpose of this study is the development of a computational framework for studying the dynamic changes of active transcription, and its interaction with chromatin remodeling and chromatin alterations during cellular responses to genotoxic stress. For this purpose, ultraviolet light C (UVC) was used as a genotoxic stress factor, damaging skin cells, specifically skin fibroblasts (VH10, CSB and 1BR.3), while the activity of Nucletiode Excision Repair (NER) pathway and the repair products of Global Genome NER (GG-NER) and Transcription Coupled NER (TC-NER) sub-pathways were used to evaluate the examined mechanisms. Various types of Next Generation Sequencing (NGS) experiments have been used to study the stages of the transcription cycle in normal conditions, and in response to Ultraviolet C irradiation (UVC) induced stress. Specifically, for studying the kinetics of RNA Polymerase 2 (RNAPII) molecules from the transcription initiation state, to promoter proximal pausing (PPP), and the transition to productive elongation, Chromatin immunoprecipitation sequencing (ChIP-seq) data of the hypophosphorylated RNAPII (RNAPII-hypo), the elongating isoform of RNAPII (RNAPII-ser2P), and the RNAPII-ser5P isoform (transcription initiation) was generated and analyzed. To study the productivity of RNAPII molecules during the above stages, Capped Analysis of Gene expression sequencing (CAGE-seq) data and nascent RNA synthesis sequencing (nRNA-seq) data was used. To study the interactions of chromatin with active transcription and its alteration during the states of active transcription, Assay for Transposase-Accessible Chromatin (ATAC-seq) data was generated and analyzed, and ChIP-seq data of H3K27ac and H3K27me3 histone modifications. To study the effectiveness and genomic landscape of NER repair-synthesis events, for both GG-NER and TC-NER sub-pathways, a novel assay called aniFOUND-seq was developed and analyzed, coupled with data of excised DNA during NER activity (XR-seq) and NER damage sequencing data (damage-seq). The functional assessment of TC-NER at active genes was carried out through the study of mutations in melanoma and lung adenocarcinoma cancer genomes, and XR-seq data meta-analysis respectively.
The results of these essays are divided into four sections:
(1) Development and application of algorithms for the analysis of NGS data related to human disease. (a) Implementation of stand-alone analysis pipelines for the analysis of ChIP-seq, nRNA-seq, and ATAC-seq datasets that include: Quality control (QC) assessment of sequenced short-reads, short-read preprocessing, short-read mapping against the under study reference genome/transcriptome, alignment processing, alignment summarization in genomic features and visualization via heatmaps and average profiles, generation of genomic tracks viewable in genome browsers (IGV, UCSC), NGS signal clustering upon functional genomic regions, correlation of biological and technical replicates, dimensionality reduction methods to identify technical/biological similarities/differences between samples, differential expression analysis, peak calling analysis, differential binding analysis, differential accessibility analysis and other statistical comparisons between biological conditions.
(b) Implementation of a “de novo” elongation wave identification algorithm using Hidden Markov Models (HMMs), and DRB-nRNA-seq datasets.
(2) Cellular responses under genotoxic stress conditions. (a) Development of a computational pipeline for the study of the reorganization of transcription and the chromatin rearrangements upon UV-induced stress that include: genome annotation reconstruction, and characterization of transcribed units’ activity (promoters, genes, enhancers, PROMoter uPstream Transcripts - asPROMPs) along the human genome, the quantification of the RNAPII release from PPP sites, and the evaluation of the RNAPII elongation wave kinetics.
(b) A proposed model describing the ‘safe’ mode mechanism of transcription elongation; upon UVC-induced stress, steady-state transcription levels of virtually all actively transcribed genes are re-adjusted to fast and uniform release of RNAPII elongation waves from PPP sites that scan the transcribed genome for DNA lesions.
This mechanism maximizes the speed of lesion sensing, the probability that a damage will be identified by an elongating RNAPII molecule and removed by TC-NER along the actively transcribed elements. As a result, environmentally exposed genomes are characterized by a modest and homogeneous mutation prevalence across the actively transcribed genome in both strands, as opposed to the non-transcribed elements where higher mutation rates are observed. In case NER is unsuccessful or is not recruited efficiently during the stress recovery process, unrepaired DNA lesions can provoke error-prone DNA synthesis and result in mutagenesis during replication.
(3) Extending the previously described ‘safe’ mode mechanism of transcription elongation, the results of the particular dissertation also support a model of continuous transcription initiation that can fuel the widespread UV-triggered escape of RNAPII into transcription elongation, that safeguards the integrity of the actively transcribed genome. The particular mechanism is supported by a global increase of chromatin accessibility at all actively transcribed promoters serving as a platform that favors unrestrained transcription initiation, coupled by preservation of the active mark H3K27ac and repressive mark H3K27me3 mark during early response to genotoxic stress.
(4) A genome-wide analysis pipeline for the evaluation of aniFOUND-seq methodology.
aniFOUND, is the first methodology (at the time of writing this thesis) that can exclusively label, capture and map the post-damage newly synthesized repaired chromatin in its native form (see materials and methods). Coupling of aniFOUND to NGS, allows the mapping and characterization of the NER efficacy of different chromosomal regions of the human genome. aniFOUND-seq was successfully applied to map the repair-synthesis activity along damaged skin fibroblasts (1BR.3 cells) with particular attention to promoter and enhancer sequences. Furthermore, aniFOUND-seq was applied for the assessment of NER-UDS activity in several chromosomal regions, including the fraction of repetitive DNA. Specifically, the repair efficacy during the first 4 hours after damage induction was clarified for rDNA and telomeres, for which contradictory explanatory models have been suggested. This is the first time that NGS-based approaches are adopted for shedding light in the above-mentioned inquiries regarding repair of telomeric DNA. Evidently, the cumulative nature of aniFOUND-seq (in terms of both damage types and repair assessment period) renders it applicable for the cases that require capturing of
the whole repair process, or the repair activity during moderately-to-considerably long-time windows.
|
Ανάπτυξη υπολογιστικών μεθόδων ανάλυσης της απόκρισης του κυττάρου στο γενοτοξικό στρες
