Computational modeling of mRNA degradation in Mycolicibacterium smegmatis

Sun, Huaming

Etd

Computational modeling of mRNA degradation in Mycolicibacterium smegmatis

Public Deposited

The increasingly severe and frequent drug resistance highlights the need to better understand the stress response strategies that the causative agent of tuberculosis, Mycobacterium tuberculosis, employs to successfully adapt and persist within the host. One of the stress response strategies is regulation of mRNA degradation, which can contribute to mycobacterial survival in energy-limited environments by reprogramming gene expression, altering mRNA abundance, and modulating energy usage. However, the regulatory mechanisms that control mRNA degradation are not well understood. In this work, I investigated mRNA degradation mechanisms in the nonpathogenic model Mycolicibacterium smegmatis from two perspectives: a targeted study of the impact of an important RNase, and an agnostic study of the impact of a diverse compendium of mRNA properties on degradation rates. In Chapter 2, we characterized the role and cleavage site preferences of an essential endoribonuclease, RNase E, in mycobacteria. By repressing transcription of rne, the gene encoding RNase E, we showed that RNase E has a major impact on mRNA degradation rates transcriptome-wide in M. smegmatis. Through the comparison of RNAseq coverage between rne knockdown and control strains, we showed that RNase E cleavage regions are enriched for cytidines in both M. smegmatis and M. tuberculosis, allowing us to attribute to RNase E a number of cleavage sites previously mapped with high resolution in vivo. These preferences for cytidines at RNase E cleavage sites were further confirmed in vitro for M. smegmatis. Together, these findings defined the dominant role of RNase E in transcriptome-wide mRNA degradation along with its cleavage targets at high resolutions in mycobacteria. In Chapter 3, we developed an experimental and computational framework to identify the intrinsic transcript properties that are associated with transcript stability in M. smegmatis. We quantified transcriptome-wide mRNA half-life in log phase growth and hypoxia-induced growth arrest using RNAseq. Through machine learning, we showed that transcript stability is influenced by the collective effect of diverse transcript features. Our results highlighted the impact of 5’ UTRs on the stability of leadered transcripts. We also identified transcript properties whose associations with transcript stability differ between leadered and leaderless transcripts as well as between different growth conditions. In sum, these results provided a comprehensive and enhanced understanding of the impacts of intrinsic transcript features on mRNA degradation rates in M. smegmatis.

Creator