Figure 2 - available via license: Creative Commons Attribution 4.0 International
Content may be subject to copyright.
SMS-seq reveals structural dependencies in RNA. (A) dAMI dependencies for pairs of bases in the synthetic hairpin construct. (B) Z-score normalized co-openness and co-agreement scores for the synthetic hairpin construct. (C) The predicted secondary structure obtained from a previous study(23) for the TPP riboswitch in unbound (left) and bound (right) configurations. Domains are indicated in colors: first stem (yellow), aptamer domain (blue), tertiary interaction of loop-stem (purple), and ribosome binding site (green). Only adenosine residues are colored. (D) Log 2 fold change in accessibility between bound and unbound states of TPP. Only adenosine residues are depicted. Red bars indicate significant changes (FDR adjusted Cochran-MantelHaenszel test q < 0.1) in the ligand-binding domain. (E) dAMI dependencies between pairs of bases for the bound (upper) and unbound (lower) state of the TPP riboswitch. Domains are colored according to subfigure C. Three black rectangle labels as I, II and III show some of the differences between bound and unbound TPP.
Source publication
RNA molecules can form secondary and tertiary structures that can regulate their localization and function. Using enzymatic or chemical probing together with high-throughput sequencing, secondary structure can be mapped across the entire transcriptome. However, a limiting factor is that only population averages can be obtained since each read is an...
Contexts in source publication
Context 1
... evaluate it further we checked that we can still separate modified adenosine from the unmodified at the consensus level when using our threshold on individual reads (ROC curve Figure 1D), resulting in AUC values ∼0.95 for different modification rates. Furthermore, we calculated FDR for the unmodified ligated RNA part of the (+)DEPC hairpin ( Figure 1E) to be 0.16 for the cutoff of 0.2 on the Tombo statistic (Supplementary Figure S2A, B). This indicates that the variance between different replicates is not influencing the FDR greatly and the Tombo method is stable. ...
Context 2
... we focused on evaluating the quality of calling open-closed regions using modified-unmodified states. To achieve that we calculated ROC curves for the highly stable hairpin construct, with a well-defined loop region (Supplementary Figure S2). For determining the accuracy of modification calling on the hairpin (ROC curves) we excluded reads with < 4 and > 18 modifications, which we deemed to be under-and over-modified, respectively (Supplementary Figure S2C). ...
Context 3
... achieve that we calculated ROC curves for the highly stable hairpin construct, with a well-defined loop region (Supplementary Figure S2). For determining the accuracy of modification calling on the hairpin (ROC curves) we excluded reads with < 4 and > 18 modifications, which we deemed to be under-and over-modified, respectively (Supplementary Figure S2C). Additionally, Tombo does not output any statistics for the 10 last bases of the read, and due to thefishers-method-context set to 3 we lose an additional three bases of the read. ...
Context 4
... with a known, stable structure ( Supplementary Fig- ure S2A). We first determined DEPC modification through footprinting analysis where cleavage patterns indicate modifications in a single-stranded region of the hairpin ( Figure 1E). ...
Context 5
... then quantified the modification level of the hairpin RNA treated with DEPC using proton NMR. This showed a decrease of adenosine integral by 24.7 ± 4.27% ( Figure 1F Figure S2B) confirming that modifications are readily detectable using nanopore sequencing. ...
Context 6
... FDR was calculated by predicting modifications in the untreated sequences ligated to the hairpins and reads from the unmodified 5-mer construct (see Materials and Methods). We selected a threshold of 0.2, corresponding to a false discovery rate of ∼15%, which when applied to the hairpin ( Figure 1G), resulted in an average of 7.43 modified bases over full-length reads (Supplementary Figure S2C). This corresponds to a modification rate of 26% and is consistent with the modification rate obtained from NMR. ...
Context 7
... z-score normalization (see Methods), this provided us with a distance-normalized AMI (dAMI). Applying this metric to the hairpin revealed that bases that are part of the same structural unit share a high dAMI (Figure 2A). In particular, we observed that the two base-paired regions forming the stem of the hairpin had a modification state that depends on each other ( Figure 2A, black rectangle) and to a lesser extent the bases in the loop (Figure 2A, black triangle). ...
Context 8
... this metric to the hairpin revealed that bases that are part of the same structural unit share a high dAMI (Figure 2A). In particular, we observed that the two base-paired regions forming the stem of the hairpin had a modification state that depends on each other ( Figure 2A, black rectangle) and to a lesser extent the bases in the loop (Figure 2A, black triangle). While the dAMI can infer dependencies it does not distinguish between open and closed states, but only estimates the extent of information captured by one position in relation to the other. ...
Context 9
... this metric to the hairpin revealed that bases that are part of the same structural unit share a high dAMI (Figure 2A). In particular, we observed that the two base-paired regions forming the stem of the hairpin had a modification state that depends on each other ( Figure 2A, black rectangle) and to a lesser extent the bases in the loop (Figure 2A, black triangle). While the dAMI can infer dependencies it does not distinguish between open and closed states, but only estimates the extent of information captured by one position in relation to the other. ...
Context 10
... therefore, also calculated to what extent bases are co-observed in the same state. Specifically, we calculated co-openness, the probability that one base is modified given that the other is modified; and co-agreement, the fraction of reads where the pair share the same state ( Figure 2B). These metrics highlighted the open and stem regions, respectively. ...
Context 11
... next asked whether SMS-seq could capture the dynamics and dependencies of structured riboswitches. We probed two riboswitches in their ligand-bound and unbound states: thiamine pyrophosphate (TPP (23), Fig- ure 2C) and F. nucleatum Flavin mononucleotide (FMN (11), Supplementary Figure S6A). As expected, both riboswitches became overall less accessible upon ligandbinding ( Figure 2D, Supplementary Figure S6B). ...
Context 12
... probed two riboswitches in their ligand-bound and unbound states: thiamine pyrophosphate (TPP (23), Fig- ure 2C) and F. nucleatum Flavin mononucleotide (FMN (11), Supplementary Figure S6A). As expected, both riboswitches became overall less accessible upon ligandbinding ( Figure 2D, Supplementary Figure S6B). Comparing between bound and unbound states of TPP revealed that, in particular, the aptamer domain (positions 11-88) exhibited significant differences in accessibility ( Figure 2D). ...
Context 13
... expected, both riboswitches became overall less accessible upon ligandbinding ( Figure 2D, Supplementary Figure S6B). Comparing between bound and unbound states of TPP revealed that, in particular, the aptamer domain (positions 11-88) exhibited significant differences in accessibility ( Figure 2D). The structural dependencies predicted by dAMI further showed that the aptamer domain and those that encompass the ribosome binding site (positions 125-146) are interdependent. ...
Context 14
... observation is consistent with the reduced access for ribosomes and translational inhibition upon ligand binding as previously observed (26). The high accessibility of the P1 helix, consisting of the pairing of base 11-14 with 85-88 (represented by position 12, Figure 2C, yellow dot, and Figure 2D) in the unbound state showed a strong decrease in accessibility in the bound state indicating its role in conformational change upon ligand recognition. Our dAMI analysis further supported this by showing the role of the first stem (represented by position 12) on the overall architecture of TPP ( Figure 2E, a diagonal line extending from 12) including the ribosome binding site (Fig- ure 2E, box I). ...
Context 15
... observation is consistent with the reduced access for ribosomes and translational inhibition upon ligand binding as previously observed (26). The high accessibility of the P1 helix, consisting of the pairing of base 11-14 with 85-88 (represented by position 12, Figure 2C, yellow dot, and Figure 2D) in the unbound state showed a strong decrease in accessibility in the bound state indicating its role in conformational change upon ligand recognition. Our dAMI analysis further supported this by showing the role of the first stem (represented by position 12) on the overall architecture of TPP ( Figure 2E, a diagonal line extending from 12) including the ribosome binding site (Fig- ure 2E, box I). ...
Context 16
... high accessibility of the P1 helix, consisting of the pairing of base 11-14 with 85-88 (represented by position 12, Figure 2C, yellow dot, and Figure 2D) in the unbound state showed a strong decrease in accessibility in the bound state indicating its role in conformational change upon ligand recognition. Our dAMI analysis further supported this by showing the role of the first stem (represented by position 12) on the overall architecture of TPP ( Figure 2E, a diagonal line extending from 12) including the ribosome binding site (Fig- ure 2E, box I). The dependency of the expression domain on the conformational change of the TPP binding pocket in the presence of TPP is also evident ( Figure 2E, box II). ...
Context 17
... high accessibility of the P1 helix, consisting of the pairing of base 11-14 with 85-88 (represented by position 12, Figure 2C, yellow dot, and Figure 2D) in the unbound state showed a strong decrease in accessibility in the bound state indicating its role in conformational change upon ligand recognition. Our dAMI analysis further supported this by showing the role of the first stem (represented by position 12) on the overall architecture of TPP ( Figure 2E, a diagonal line extending from 12) including the ribosome binding site (Fig- ure 2E, box I). The dependency of the expression domain on the conformational change of the TPP binding pocket in the presence of TPP is also evident ( Figure 2E, box II). ...
Context 18
... dAMI analysis further supported this by showing the role of the first stem (represented by position 12) on the overall architecture of TPP ( Figure 2E, a diagonal line extending from 12) including the ribosome binding site (Fig- ure 2E, box I). The dependency of the expression domain on the conformational change of the TPP binding pocket in the presence of TPP is also evident ( Figure 2E, box II). Interestingly, dAMI also revealed that SMS-seq is capable of capturing the tertiary interactions between the two helices ( Figure 2E, box III), which has been demonstrated in mutational profiling (RING-MaP) and high-resolution TPP structure mapping (3,23), where the interaction between nucleotides in the L5 loop (bases from 67 to 72) and the P3 helix (a stem structure formed by base pairing of nucleotide 21-27 and 33-38) enables the formation of the ligand-binding pocket. ...
Context 19
... dependency of the expression domain on the conformational change of the TPP binding pocket in the presence of TPP is also evident ( Figure 2E, box II). Interestingly, dAMI also revealed that SMS-seq is capable of capturing the tertiary interactions between the two helices ( Figure 2E, box III), which has been demonstrated in mutational profiling (RING-MaP) and high-resolution TPP structure mapping (3,23), where the interaction between nucleotides in the L5 loop (bases from 67 to 72) and the P3 helix (a stem structure formed by base pairing of nucleotide 21-27 and 33-38) enables the formation of the ligand-binding pocket. Although RING-Map and SMSseq agree on loop-helix interaction, they are not completely consistent in their predictions (Supplementary Figure S7). ...
Similar publications
RNA flexibility is reflected in its heterogeneous conformation. Through direct visualization using atomic force microscopy (AFM) and the adenosylcobalamin riboswitch aptamer domain as an example, we show that a single RNA sequence folds into conformationally and architecturally heterogeneous structures under near-physiological solution conditions....
Citations
... RNA conformational dynamics have essential roles in splicing 1 , packaging 2 , cellular transactivation 3 and responding to environmental stimulation 4,5 . Indeed, ample experimental evidence 1,6-18 , including from single molecule-based approaches 5,[19][20][21] , indicates that RNA molecules exist in a much broader conformational space than is depicted by a small number of static snapshots 22,23 . Various techniques have been applied to study RNA conformational diversity and dynamics, which probe either ensemble behaviours 10,12,[24][25][26][27] or characterize single molecules using sparse distances 21,28,29 . ...
... In this procedure, we start with the fitting of each of 158 conformers, independently, mapping the χ 2 values, and then correlating them with the class (C1,C2 and C3) to which that particle belongs. Then, we add in permutations of a given number (1,2,3,5,10,20,50) of random conformers from each class, map the χ 2 for 10,000 rounds, in which (1) assumes one structure from each class, (2) assumes two structures from each class, (3) assumes three structures from each class, and so on. ...
RNA conformational diversity has fundamental biological roles1, 2, 3, 4–5, but direct visualization of its full conformational space in solution has not been possible using traditional biophysical techniques. Using solution atomic force microscopy, a deep neural network and statistical analyses, we show that the ribonuclease P (RNase P) RNA adopts heterogeneous conformations consisting of a conformationally invariant core and highly flexible peripheral structural elements that sample a broad conformational space, with amplitudes as large as 20–60 Å in a multitude of directions, with very low net energy cost. Increasing Mg²⁺ drives compaction and enhances enzymatic activity, probably by narrowing the conformational space. Moreover, analyses of the correlations and anticorrelations between spatial flexibility and sequence conservation suggest that the functional roles of both the structure and dynamics of key regions are embedded in the primary sequence. These findings reveal the structure–dynamics basis for the embodiment of both enzymatic precision and substrate promiscuity in the RNA component of the RNase P. Mapping the conformational space of the RNase P RNA demonstrates a new general approach to studying RNA structure and dynamics.
... We experimentally validated that the evolved sequences behaved as modeled by in vitro transcribing WT CLN3 and the 6 structure mutants and using RNA structure probing with singlemolecule, long read, direct RNA sequencing using Nanopore technology (21). This approach enabled us to measure the pairwise distance between the structures of individual mRNA molecules in the population. ...
Most amino acids are encoded by multiple codons, making the genetic code degenerate. Synonymous mutations affect protein translation and folding, but their impact on RNA itself is often neglected. We developed a genetic algorithm that introduces synonymous mutations to control the diversity of structures sampled by an mRNA. The behavior of the designed mRNAs reveals a physical code layered in the genetic code. We find that mRNA conformational heterogeneity directs physical properties and functional outputs of RNA-protein complexes and biomolecular condensates. The role of structure and disorder of proteins in biomolecular condensates is well appreciated, but we find that RNA conformational heterogeneity is equally important. This feature of RNA enables both evolution and engineers to build cellular structures with specific material and responsive properties.
... Single-molecule methods are gaining ground in bioanalytical applications like nucleic acid sequencing and sensitive analyte detection. [1][2][3][4] For Point-of-Care (POC) detection in lowtechnology environments, however, single-molecule approaches, are still considered prohibitively expensive due to their reliance on sophisticated setups. For instance, single-molecule detection using fluorescence often requires either molecular multiplication to detect the target's signal against background noise or expensive instrumentation to detect single fluorescent molecules directly. ...
Creating increasingly sensitive and cost-effective nucleic acid detection methods is critical for enhancing point-of-care (POC) applications. This involves capturing all desired biomarkers in a sample with high specificity and transducing the capture events to a detector. However, the signal from biomarkers present at extremely low amounts often falls below the detection limit of typical fluorescence-based methods, making molecular amplification a necessary step. Here, we present a nucleic acid detection assay of a 151-nucleotide sequence specific to antibiotics-resistant Klebsiella pneumoniae, based on single-molecule fluorescence detection of non-amplified DNA down to the attomolar level, using Trident NanoAntennas with Cleared HOtSpots (NACHOS). Our NACHOS-diagnostics assay leverages a compact microscope with a large field-of-view and cost-efficient components, including microfluidic flow to enhance capturing efficiency. Fluorescence enhancement is provided by DNA origami NanoAntennas, arranged in a dense array using a combination of nanosphere lithography and site-specific DNA origami placement. Our method can detect 200 ± 50 out of 600 molecules in a 100 μL sample volume within an hour. This represents typical number of pathogens in clinical samples commonly detected by Polymerase Chain Reaction but without the need for molecular amplification. We achieve similar sensitivity in untreated human blood plasma, enhancing the practical applicability of the system. Our platform can be adapted to detect shorter nucleic acid fragments that are not compatible with traditional amplification-based technologies. This broadens its potential for diverse diagnostic and healthcare applications, providing a robust and scalable solution for sensitive nucleic acid detection in various clinical settings.
... DRS can, to some extent, offer an opportunity to explore the combinatorial interplay between RNA modifications and many other RNA features such as APA, secondary structures, and even other types of co-occurring modification ( Figure 4B), as well as the possible roles of modifications within fusion transcripts and circRNAs . Recent advances in the structural interrogation of single RNA molecules through chemical probing based on DRS and the coordinated regulation of gene expression by both m 6 A and m 5 C demonstrate the feasibility of at least some of these ideas (Bizuayehu et al., 2022;Alan et al., 2023;Yu et al., 2023a). Integrating data that detect specific RNA-binding proteins can facilitate the determination of some poorly understood mechanisms by which modifications influence RNA metabolism in plants. ...
The transcriptome serves as a bridge that links genomic variation and phenotype diversity. A vast number of studies using next-generation RNA sequencing (RNA-seq) in the last two decades emphasize the essential roles of plant transcriptome in response to developmental and environmental conditions, leading to numerous insights into the dynamic change, evolutionary trace and elaborate regulation of plant transcriptome. With substantial improvement in accuracy and throughput, direct RNA sequencing (DRS) has emerged as a new and powerful sequencing platform for the precise detection of native and full-length transcripts, which overcomes many limitations such as read length and PCR bias that are inherent to short-read RNA-seq. Here, we reviewed recent advances in dissecting the complexity and diversity of plant transcriptome utilizing DRS as a main technological mean from many aspects of RNA metabolism, including novel isoforms, poly(A) tail and RNA modification, and proposed a comprehensive workflow for the data process of plants DRS. Many challenges concerning the application of DRS in plants, such as machine learning tools tailored to plant transcriptome, remain to be solved, and together we prospect the future biological questions that can be potentially answered by DRS such as allele-specific RNA modification. This technology provides convenient support on which the connection of distinct RNA features is tightly built, sustainably refining our understanding of the biological functions of plant transcriptome.
... Regulatory RNAs exhibit highly dynamic properties due to their multidimensional structures and their ability to interact with RNA, DNA, and proteins. These RNAs are involved in regulated events, such as transcription, translation, molecular localization and stability, and enzymatic degradation cascades, on many levels [1][2][3][4] . A broad spectrum of RNAs that have regulatory functions are collectively recognized as regulatory RNAs. ...
It is apparent that various functional units within the cellular machinery are derived from RNAs. The evolution of sequencing techniques has resulted in significant insights into approaches for transcriptome studies. Organisms utilize RNA to govern cellular systems, and a heterogeneous class of RNAs is involved in regulatory functions. In particular, regulatory RNAs are increasingly recognized to participate in intricately functioning machinery across almost all levels of biological systems. These systems include those mediating chromatin arrangement, transcription, suborganelle stabilization, and posttranscriptional modifications. Any class of RNA exhibiting regulatory activity can be termed a class of regulatory RNA and is typically represented by noncoding RNAs, which constitute a substantial portion of the genome. These RNAs function based on the principle of structural changes through cis and/or trans regulation to facilitate mutual RNA‒RNA, RNA‒DNA, and RNA‒protein interactions. It has not been clearly elucidated whether regulatory RNAs identified through deep sequencing actually function in the anticipated mechanisms. This review addresses the dominant properties of regulatory RNAs at various layers of the cellular machinery and covers regulatory activities, structural dynamics, modifications, associated molecules, and further challenges related to therapeutics and deep learning.
... Instead, this signal could be used to study noncanonical RNA structures involving Hoogsteen interactions, such as G-quadruplexes 48,49 . Our method is related to recent advances in long-read RNA structural probing by nanopore direct RNA sequencing (dRNA-seq) of chemically modified transcripts [50][51][52] . However, dRNA-seq structural probing requires specialized modification detection algorithms which may need continual updates as nanopore sequencing chemistry changes. ...
Genome-wide measurements of RNA structure can be obtained using reagents that react with unpaired bases, leading to adducts that can be identified by mutational profiling on next-generation sequencing machines. One drawback of these experiments is that short sequencing reads can rarely be mapped to specific transcript isoforms. Consequently, information is acquired as a population average in regions that are shared between transcripts, thus blurring the underlying structural landscape. Here, we present nanopore dimethylsulfate mutational profiling (Nano-DMS-MaP)—a method that exploits long-read sequencing to provide isoform-resolved structural information of highly similar RNA molecules. We demonstrate the value of Nano-DMS-MaP by resolving the complex structural landscape of human immunodeficiency virus-1 transcripts in infected cells. We show that unspliced and spliced transcripts have distinct structures at the packaging site within the common 5′ untranslated region, likely explaining why spliced viral RNAs are excluded from viral particles. Thus, Nano-DMS-MaP is a straightforward method to resolve biologically important transcript-specific RNA structures that were previously hidden in short-read ensemble analyses.
... The measurements in this study provide insight into the salient features of the misfolded structures adopted by ctRSD gates ( Figure S1, Supporting Note 5), but more detailed measurements are required to fully understand the failure mechanisms that lead to poor performance. For example, RNA sequencing techniques could be used to understand the secondary structures of misfolded gates [40][41][42] and leak mechanisms 43,44 . These measurements would guide further improvements to ctRSD gate design and could enable predictive models of the ctRSD gate sequence-structure-function relationship that optimize gate sequences in silico for a given circuit. ...
Cotranscriptionally encoded RNA strand displacement (ctRSD) circuits are an emerging tool for programmable molecular computation with potential applications spanning in vitro diagnostics to continuous computation inside living cells. In ctRSD circuits, RNA strand displacement components are continuously produced together via transcription. These RNA components can be rationally programmed through base pairing interactions to execute logic and signaling cascades. However, the small number of ctRSD components characterized to date limits circuit size and capabilities. Here, we characterize 220 ctRSD gate sequences, exploring different input, output, and toehold sequences and changes to other design parameters, including domain lengths, ribozyme sequences, and the order in which gate strands are transcribed. This characterization provides a library of sequence domains for engineering ctRSD components, i.e., a toolkit, enabling circuits with up to four-fold more inputs than previously possible. We also identify specific failure modes and systematically develop design approaches that reduce the likelihood of failure across different gate sequences. Lastly, we show ctRSD gate design is robust to changes in transcriptional encoding, opening a broad design space for applications in more complex environments. Together, these results deliver an expanded toolkit and design approaches for building ctRSD circuits that will dramatically extend capabilities and potential applications.
... Similar to proteins, some non-coding RNAs are highly structured and undergo thermodynamic and kinetically controlled conformational changes in the folding free energy landscape [99][100][101]. RNA structures and their dynamics have been studied by various groups using the combination of two or more single-molecule techniques such as AFM, TIRF microscopy, confocal microscopy, optical and magnetic tweezers, and most recently using nanopore technology [100,[102][103][104][105][106][107][108][109][110]. The combination of optical tweezers, microfluidics, and fluorescence microscopy has emerged as a great tool to understand the dynamics and functions of RNA structures. ...
Understanding complex biological events at the molecular level paves the path to determine mechanistic processes across the timescale necessary for breakthrough discoveries. While various conventional biophysical methods provide some information for understanding biological systems, they often lack a complete picture of the molecular-level details of such dynamic processes. Studies at the single-molecule level have emerged to provide crucial missing links to understanding complex and dynamic pathways in biological systems, which are often superseded by bulk biophysical and biochemical studies. Latest developments in techniques combining single-molecule manipulation tools such as optical tweezers and visualization tools such as fluorescence or label-free microscopy have enabled the investigation of complex and dynamic biomolecular interactions at the single-molecule level. In this review, we present recent advances using correlated single-molecule manipulation and visualization-based approaches to obtain a more advanced understanding of the pathways for fundamental biological processes, and how this combination technique is facilitating research in the dynamic single-molecule (DSM), cell biology, and nanomaterials fields.
Cells must limit RNA–RNA interactions to avoid irreversible RNA entanglement. Cells may prevent deleterious RNA-RNA interactions by genome organization to avoid complementarity however, RNA viruses generate long, perfectly complementary antisense RNA during replication. How do viral RNAs avoid irreversible entanglement? One possibility is RNA sequestration into biomolecular condensates. To test this, we reconstituted critical SARS-CoV-2 RNA–RNA interactions in Nucleocapsid condensates. We observed that RNAs with low propensity RNA–RNA interactions resulted in more round, liquid-like condensates while those with high sequence complementarity resulted in more heterogeneous networked morphology independent of RNA structure stability. Residue-resolution molecular simulations and direct sequencing-based detection of RNA–RNA interactions support that these properties arise from degree of trans RNA contacts. We propose that extensive RNA–RNA interactions in cell and viral replication are controlled via a combination of genome organization, timing, RNA sequence content, RNA production ratios, and emergent biomolecular condensate material properties.
Probabilistic modeling of RNA-RNA interactions
and complex formation using nanopre dricet sequencing
