Approaches to improve in vivo RNA structure prediction through chemical structure probing
Restricted (Penn State Only)
- Author:
- Douds, Catherine
- Graduate Program:
- Biochemistry, Microbiology, and Molecular Biology
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- March 14, 2024
- Committee Members:
- Paul Babitzke, Major Field Member
Sarah Assmann, Outside Field Member
Shaun Mahony, Outside Unit Member
Lu Bai, Major Field Member
Philip Bevilacqua, Chair & Dissertation Advisor
Santhosh Girirajan, Program Head/Chair - Keywords:
- RNA
RNA structure
structure probing
E. coli ribosome - Abstract:
- At the center of the central dogma, RNA plays numerous roles in gene regulation in a cell, many of which continue to be discovered. RNA folds into complex and dynamic structures with which it regulates gene expression through interaction with factors such as temperature, metabolites, and protein binding. Thus, to understand and discover new functions, accurately determining RNA structure as it exists in a cell is key. Chemical structure probing is at the forefront of the RNA field to better understand RNA structure in vivo. By this method, cells are treated with membrane-permeable reagents that react with the RNA in a structurally specific manner, leaving a covalent modification that can be detected by reverse transcription (RT). Although these techniques are ever improving and the chemical suite with which to probe RNA is ever growing, there are persistent limitations in the field. The work in this thesis aims to take steps towards some of these limitations that include partial nucleotide information, misinterpretation of low reactivities, and poor structure prediction accuracy. Chapter 1 begins by outlining the utility of chemical structure probing methods towards understanding RNA structure in biology, as well as practical applications for using reactivities. Additionally, it outlines shortcomings of these techniques facing the field of in vivo RNA structure probing and recent advancements towards addressing them. Chapter 2 aims to address the problem of partial nucleotide specificity for chemicals probing at the Watson-Crick-Franklin (WCF) face of RNAs. We present a novel in vivo chemical probe for Gs and Us with reduced RNA degradation that is detectable by MaP, 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide methiodide (ETC). This chemical, when combined with dimethylsulfate (DMS) that probes the WCF face of As and Cs, provides full nucleotide information at the WCF face. And, upon combining these reactivities, we show an improvement in both accuracy and reliability of empirically driven structure predictions. In Chapter 3, the reactivities calculated in Chapter 2 are explored in more depth by using cheminformatics measurements from the crystal structure of the Escherichia coli ribosome to address the misinterpretation of low reactivities. In this work, we provide physical evidence that non-base pairing interactions like hydrogen bonds to the reactive atom or occlusion of solvent accessibility influence reactivity values. We also identify sequence- and chemical-dependent trends in reactivity that can be applied to new pseudo-free energy equations to further improve structure prediction accuracy. The last research chapter, Chapter 4, presents data characterizing DMS-stimulated RNA degradation by phosphate methylation in vivo and in vitro emphasizing the importance of pH in this reaction and suggesting best practices to avoid DMS-stimulated degradation. Although the results are inconclusive and preliminary, we also provide initial trouble shooting steps for assays to directly measure DMS phosphate methylation and indirectly measure its ability to probe RNA structure at the phosphate. Finally, in Chapter 5, I conclude this work and describe future directions towards the continued improvement of in vivo RNA structure determination. Overall, this work aims to address issues within this methodology by avoiding degradation, expanding the chemical toolbox for probing structures, and deeply studying the resulting data in the context of tertiary structure, thereby laying the groundwork for future steps towards improvement of empirically driven RNA structure determination.