In silico methods for co-transcriptional RNA secondary structure prediction and for investigating alternative RNA structure expression

Abstract

RNA transcripts are the primary products of active genes in any living organism, including many viruses. Their cellular destiny not only depends on primary sequence signals, but can also be determined by RNA structure. Recent experimental evidence shows that many transcripts can be assigned more than a single functional RNA structure throughout their cellular life and that structure formation happens co-transcriptionally, i.e. as the transcript is synthesised in the cell. Moreover, functional RNA structures are not limited to non-coding transcripts, but can also feature in coding transcripts. The picture that now emerges is that RNA structures constitute an additional layer of information that can be encoded in any RNA transcript (and on top of other layers of information such as protein-context) in order to exert a wide range of functional roles. Moreover, different encoded RNA structures can be expressed at different stages of a transcript's life in order to alter the transcript's behaviour depending on its actual cellular context. Similar to the concept of alternative splicing for protein-coding genes, where a single transcript can yield different proteins depending on cellular context, it is thus appropriate to propose the notion of alternative RNA structure expression for any given transcript. This review introduces several computational strategies that my group developed to detect different aspects of RNA structure expression in vivo. Two aspects are of particular interest to us: (1) RNA secondary structure features that emerge during co-transcriptional folding and (2) functional RNA structure features that are expressed at different times of a transcript's life and potentially mutually exclusive.

A genetic selection reveals functional metastable structures embedded in a toxin-encoding mRNA

“…Thus, the metastable RNA structures identified in the present study are essential to uncouple transcription and translation processes and allow the presence of type I TA systems on bacterial chromosomes. Although transient RNA structures can be predicted in silico (Meyer, 2017a), their in vivo characterization remains challenging. Several high-resolution methods have been recently devised for analyzing the co-transcriptional folding of regulatory RNAs, both in vitro (Uhm et al, 2018; Watters et al, 2016) and in vivo (Incarnato et al, 2017).…”

Know Your Enemy: Successful Bioinformatic Approaches to Predict Functional RNA Structures in Viral RNAs

“…The elements considered above form mainly local structures. These local structures may form in nascent RNA (Meyer, 2017 ) or be stabilized by protein or RNA binding. Local structures can be predicted using appropriate windows of sequence (e.g., 80–200 bases) and it is also practical to analyse local alignments of similar lengths (Lange et al, 2012 ).…”

Investigating the concept of accessibility for predicting novel RNA-RNA interactions

“…Right now, none of the computational methods conceptually aim to disentangle the conserved cis and trans features into self-consistent configurations of mutually compatible cis and trans features that could be expressed at different points in time. As we discover more complexities in vivo, we may thus be required to expand the notion of alternative RNA structure expression (59) to the notion of alternative trans RNA-RNA interactions. 1.…”