Super Resolution Microscopy to investigate epigenetic mechanisms driving platinum resistance in ovarian High Grade Serous Carcinoma
File(s)
Author(s)
Lim, Mi Qi
Type
Thesis
Abstract
The high mortality rate of ovarian High Grade Serous Carcinoma (HGSC) partly results from frequent relapses with platinum-resistant disease, and accumulating evidence indicate the importance of epigenetic modifications, which influence higher order chromatin structure to regulate gene expression, driving this resistance. Currently, this structural information is typically inferred through bulk biochemical assays. Conversely, single molecule localisation microscopy (SMLM) methods like stochastic optical reconstruction microscopy (STORM) can directly probe these structures with single cell resolution, but most implementations are expensive and low-throughput. This thesis presents work towards developing a high-throughput SMLM assay to investigate epigenetic mechanisms driving carboplatin resistance in HGSC using easySTORM implemented on a novel low-cost and open-source microscopy platform (‘openFrame’).
Sample preparation protocols and imaging parameters were optimised, during which nuclear pore complexes in U2OS-Nup96-SNAP cells were imaged to serve as reference structures for calibration. Subsequently, different SMLM cluster analysis pipelines were compared, as most pipelines are computationally intensive and require optimising user-selected parameters. A novel multi-component pair-correlation function fitting algorithm yielded similar results to established methods (e.g., DBSCAN) without requiring user-set parameters, and could analyse SMLM data from 100 nuclei in 40 minutes with parallel processing on 12-core high-performance computing (HPC) 128 GB RAM nodes. Physiologically relevant carboplatin-treatment models of HGSC were also studied to provide the biological basis for subsequent assays. Preliminary results indicate diverse early transcriptomic changes and increased accessibility of AP-1 binding sites across primary and established cell lines following treatment, suggesting an epigenetic role in resistance.
Carboplatin-induced changes in chromatin structure of HGSC cells were investigated with automated multiwell plate imaging on the openFrame-based easySTORM platform. Preliminary results suggest a relative decompaction of chromatin in resistant cells, potentially correlating with a more stem-like phenotype. Overall, this work demonstrates the potential of this platform as a practical assay for investigating changes in chromatin structure.
Sample preparation protocols and imaging parameters were optimised, during which nuclear pore complexes in U2OS-Nup96-SNAP cells were imaged to serve as reference structures for calibration. Subsequently, different SMLM cluster analysis pipelines were compared, as most pipelines are computationally intensive and require optimising user-selected parameters. A novel multi-component pair-correlation function fitting algorithm yielded similar results to established methods (e.g., DBSCAN) without requiring user-set parameters, and could analyse SMLM data from 100 nuclei in 40 minutes with parallel processing on 12-core high-performance computing (HPC) 128 GB RAM nodes. Physiologically relevant carboplatin-treatment models of HGSC were also studied to provide the biological basis for subsequent assays. Preliminary results indicate diverse early transcriptomic changes and increased accessibility of AP-1 binding sites across primary and established cell lines following treatment, suggesting an epigenetic role in resistance.
Carboplatin-induced changes in chromatin structure of HGSC cells were investigated with automated multiwell plate imaging on the openFrame-based easySTORM platform. Preliminary results suggest a relative decompaction of chromatin in resistant cells, potentially correlating with a more stem-like phenotype. Overall, this work demonstrates the potential of this platform as a practical assay for investigating changes in chromatin structure.
Version
Open Access
Date Issued
2023-12
Date Awarded
2024-03
Copyright Statement
Creative Commons Attribution NonCommercial Licence
License URL
Advisor
McNeish, Iain
French, Paul
Dunsby, Christopher
Sponsor
Cancer Research UK
Grant Number
PS3700
Publisher Department
Faculty of Medicine
Publisher Institution
Imperial College London
Qualification Level
Doctoral
Qualification Name
Doctor of Philosophy (PhD)