Computationally reconstructing cotranscriptional RNA folding from experimental data reveals rearrangement of non-native folding intermediates

Conformational Ensembles of Noncoding Elements in the SARS-CoV-2 Genome from Molecular Dynamics Simulations

“…The modeling protocol used in the present study implicitly considers helix formation and internal loop dynamics as partially decoupled events. While this is a relatively common assumption in computational studies of RNA, 18 , 27 , 31 we can envisage situations in which the rearrangement of a loop requires one or more base pairs to break transiently. In our partial tempering MD simulations the secondary structure is not fixed: these rare events can thus be observed, albeit with a low frequency due to the absence of auxiliary replicas enhancing the breakage and formation of stem regions.…”

“… 11 While several fragment-based tools exist to perform such a task, 23 , 24 we here adapt an MD-based procedure. 25 Similar secondary structure restraining approaches have been employed in all-atom RNA folding simulations 26 , 27 and to study internal loop dynamics. 28 These initial structures are subject to extensive MD sampling, amounting to more than 0.5 ms of aggregate simulation time.…”

Cotranscriptional RNA strand exchange underlies the gene regulation mechanism in a purine-sensing transcriptional riboswitch

“…Preliminary secondary structure predictions for informing SHAPE-Seq experiments were generated using RNAstructure version 6.3 from the Mathews Lab (https://rna.urmc.rochester.edu/RNAstructure.html) (49). SHAPE-Seq-informed secondary structure predictions were generated using Reconstructing RNA Dynamics from Data (R2D2) with the protocol and code detailed in methods of Yu, et al (42). Additional options used during R2D2 analysis are provided in Supplementary Information.…”

“…To investigate the structural intermediates that mediate the yxjA riboswitch regulatory mechanism, we performed cotranscriptional in vitro Selective 2'-Hydroxyl Acylation Analyzed by Primer Extension Sequencing (cotranscriptional SHAPE-Seq) experiments using benzoyl cyanide (BzCN) to chemically modify the RNA at more flexible residues during stalled transcription at various lengths (27,64). The resulting chemical probing reactivity data was then used as inputs into the Reconstructing RNA Dynamics from Data (R2D2) algorithm to generate experimentally informed models of cotranscriptionally folded structural intermediates (27,42). Our goal with these experiments was to observe evidence of a central helix -a structure formed by the complementary regions of P1 and the terminator (Figure 1C) -that could form during the folding pathway, or an alternative structural model that we could use to further investigate the riboswitch mechanism.…”

“…This observation raises an important question for understanding the mechanisms of these riboswitches: how can ligand binding in the AD "communicate" downstream structural changes and determine folding outcomes of the EP (37,38)? Given the role of strand displacement mechanisms among diverse riboswitches (29,40,41), as well as within the cotranscriptional folding pathways of other important non-coding RNAs such as the signal recognition particle (SRP) RNA (42), we hypothesize that riboswitches without direct AD and EP overlap still rely on a ligand-dependent strand exchange process to differentiate their folded states and enact regulatory function, however with an unknown mechanism to inhibit strand exchange in the presence of ligand.…”