RNA Contacts Research Articles

Leveraging data-efficient RNA contact prediction toward reliable RNA structure prediction

Leveraging data-efficient RNA contact prediction toward reliable RNA structure prediction

Long-range RNA structures in the human transcriptome beyond evolutionarily conserved regions.

RNA structure has been increasingly recognized as a critical player in the biogenesis and turnover of many transcripts classes. In eukaryotes, the prediction of RNA structure by thermodynamic modeling meets fundamental limitations due to the large sizes and complex, discontinuous organization of eukaryotic genes. Signatures of functional RNA structures can be found by detecting compensatory substitutions in homologous sequences, but a comparative approach is applicable only within conserved sequence blocks. Here, we developed a computational pipeline called PHRIC, which is not limited to conserved regions and relies on RNA contacts derived from RNA in situ conformation sequencing (RIC-seq) experiments. It extracts pairs of short RNA fragments surrounded by nested clusters of RNA contacts and predicts long, nearly perfect complementary base pairings formed between these fragments. In application to a panel of RIC-seq experiments in seven human cell lines, PHRIC predicted ~12,000 stable long-range RNA structures with equilibrium free energy below -15 kcal/mol, the vast majority of which fall outside of regions annotated as conserved among vertebrates. These structures, nevertheless, show some level of sequence conservation and remarkable compensatory substitution patterns in other clades. Furthermore, we found that introns have a higher propensity to form stable long-range RNA structures between each other, and moreover that RNA structures tend to concentrate within the same intron rather than connect adjacent introns. These results for the first time extend the application of proximity ligation assays to RNA structure prediction beyond conserved regions.

RNA contact prediction by data efficient deep learning

On the path to full understanding of the structure-function relationship or even design of RNA, structure prediction would offer an intriguing complement to experimental efforts. Any deep learning on RNA structure, however, is hampered by the sparsity of labeled training data. Utilizing the limited data available, we here focus on predicting spatial adjacencies ("contact maps”) as a proxy for 3D structure. Our model, BARNACLE, combines the utilization of unlabeled data through self-supervised pre-training and efficient use of the sparse labeled data through an XGBoost classifier. BARNACLE shows a considerable improvement over both the established classical baseline and a deep neural network. In order to demonstrate that our approach can be applied to tasks with similar data constraints, we show that our findings generalize to the related setting of accessible surface area prediction.

RNAcontacts: A Pipeline for Predicting Contacts from RNA Proximity Ligation Assays.

High-throughput RNA proximity ligation assays are molecular methods that are used to simultaneously analyze the spatial proximity of many RNAs in living cells. Their principle is based on cross-linking, fragmentation, and subsequent religation of RNAs, followed by high-throughput sequencing. The generated fragments have two different types of splits, one resulting from pre-mRNA splicing and the other formed by the ligation of spatially close RNA strands. Here, we present RNAcontacts, a universal pipeline for detecting RNA-RNA contacts in high-throughput RNA proximity ligation assays. RNAcontacts circumvents the inherent problem of mapping sequences with two distinct types of splits using a two-pass alignment, in which splice junctions are inferred from a control RNA-seq experiment on the first pass and then provided to the aligner as bona fide introns on the second pass. Compared to previously developed methods, our approach allows for a more sensitive detection of RNA contacts and has a higher specificity with respect to splice junctions that are present in the biological sample. RNAcontacts automatically extracts contacts, clusters their ligation points, computes the read support, and generates tracks for visualizing through the UCSC Genome Browser. The pipeline is implemented in Snakemake, a reproducible and scalable workflow management system for rapid and uniform processing of multiple datasets. RNAcontacts is a generic pipeline for the detection of RNA contacts that can be used with any proximity ligation method as long as one of the interacting partners is RNA. RNAcontacts is available via the GitHub repository https://github.com/smargasyuk/ RNAcontacts/.

RNA in the Loop: Probing T-loop/T-loop Receptor Interactions as Mediators of Long-range RNA Contacts that Influence Gene Regulation

RNA in the Loop: Probing T-loop/T-loop Receptor Interactions as Mediators of Long-range RNA Contacts that Influence Gene Regulation

Structural basis for recognition of transcriptional terminator structures by ProQ/FinO domain RNA chaperones

The ProQ/FinO family of RNA binding proteins mediate sRNA-directed gene regulation throughout gram-negative bacteria. Here, we investigate the structural basis for RNA recognition by ProQ/FinO proteins, through the crystal structure of the ProQ/FinO domain of the Legionella pneumophila DNA uptake regulator, RocC, bound to the transcriptional terminator of its primary partner, the sRNA RocR. The structure reveals specific recognition of the 3’ nucleotide of the terminator by a conserved pocket involving a β-turn-α-helix motif, while the hairpin portion of the terminator is recognized by a conserved α-helical N-cap motif. Structure-guided mutagenesis reveals key RNA contact residues that are critical for RocC/RocR to repress the uptake of environmental DNA in L. pneumophila. Structural analysis and RNA binding studies reveal that other ProQ/FinO domains also recognize related transcriptional terminators with different specificities for the length of the 3’ ssRNA tail.

Intrinsically disordered interaction network in an RNA chaperone revealed by native mass spectrometry

RNA-binding proteins contain intrinsically disordered regions whose functions in RNA recognition are poorly understood. The RNA chaperone Hfq is a homohexamer that contains six flexible C-terminal domains (CTDs). The effect of the CTDs on Hfq's integrity and RNA binding has been challenging to study because of their sequence identity and inherent disorder. We used native mass spectrometry coupled with surface-induced dissociation and molecular dynamics simulations to disentangle the arrangement of the CTDs and their impact on the stability of Escherichia coli Hfq with and without RNA. The results show that the CTDs stabilize the Hfq hexamer through multiple interactions with the core and between CTDs. RNA binding perturbs this network of CTD interactions, destabilizing the Hfq ring. This destabilization is partially compensated by binding of RNAs that contact multiple surfaces of Hfq. By contrast, binding of short RNAs that only contact one or two subunits results in net destabilization of the complex. Together, the results show that a network of intrinsically disordered interactions integrate RNA contacts with the six subunits of Hfq. We propose that this CTD network raises the selectivity of RNA binding.

The influenza A virus genome packaging network - complex, flexible and yet unsolved.

The genome of influenza A virus (IAV) consists of eight unique viral RNA segments. This genome organization allows genetic reassortment between co-infecting IAV strains, whereby new IAVs with altered genome segment compositions emerge. While it is known that reassortment events can create pandemic IAVs, it remains impossible to anticipate reassortment outcomes with pandemic prospects. Recent research indicates that reassortment is promoted by a viral genome packaging mechanism that delivers the eight genome segments as a supramolecular complex into the virus particle. This finding holds promise of predicting pandemic IAVs by understanding the intermolecular interactions governing this genome packaging mechanism. Here, we critically review the prevailing mechanistic model postulating that IAV genome packaging is orchestrated by a network of intersegmental RNA–RNA interactions. Although we find supporting evidence, including segment-specific packaging signals and experimentally proposed RNA–RNA interaction networks, this mechanistic model remains debatable due to a current shortage of functionally validated intersegmental RNA–RNA interactions. We speculate that identifying such functional intersegmental RNA–RNA contacts might be hampered by limitations of the utilized probing techniques and the inherent complexity of the genome packaging mechanism. Nevertheless, we anticipate that improved probing strategies combined with a mutagenesis-based validation could facilitate their discovery.

SRPK1 regulates RNA binding in a pre-spliceosomal complex using a catalytic bypass mechanism.

Serine-arginine protein kinase 1 (SRPK1) phosphorylates serine-arginine (SR) proteins in the cytoplasm, directing them to the nucleus for splicing function. SRPK1 has also been detected in the nucleus but its function here is still not fully understood. We now demonstrate that nuclear SRPK1 can regulate U1-70K, a protein component of the uridine-rich 1 small nuclear ribonucleoprotein (U1 snRNP) that binds SR proteins and facilitates 5' splice-site selection in precursor mRNA. We found that SRPK1 uses a large, disordered domain to bind U1-70K, regulating the interaction of an exonic splicing enhancer (ESE) to the associated SR protein. Surprisingly, the catalytic activity of SRPK1 is not required for this phenomenon. Instead, SRPK1 associates directly with the N-terminus of U1-70K and alters the regulatory function of the distal C-terminus, modifying interactions between the U1-70K:SR protein complex and the ESE. Disruption of SRPK1 binding to this complex affects the alternative splicing of genes modulated by the C-terminus of U1-70K. Such findings suggest that, in addition to operating as a traditional serine-modifying catalyst, SRPK1 can also bypass this intrinsic activity to regulate RNA contacts in an early pre-spliceosomal complex.

A novel, essential trans-splicing protein connects the nematode SL1 snRNP to the CBC-ARS2 complex.

Spliced leader trans-splicing is essential for gene expression in many eukaryotes. To elucidate the molecular mechanism of this process, we characterise the molecules associated with the Caenorhabditis elegans major spliced leader snRNP (SL1 snRNP), which donates the spliced leader that replaces the 5′ untranslated region of most pre-mRNAs. Using a GFP-tagged version of the SL1 snRNP protein SNA-1 created by CRISPR-mediated genome engineering, we immunoprecipitate and identify RNAs and protein components by RIP-Seq and mass spectrometry. This reveals the composition of the SL1 snRNP and identifies associations with spliceosome components PRP-8 and PRP-19. Significantly, we identify a novel, nematode-specific protein required for SL1 trans-splicing, which we designate SNA-3. SNA-3 is an essential, nuclear protein with three NADAR domains whose function is unknown. Mutation of key residues in NADAR domains inactivates the protein, indicating that domain function is required for activity. SNA-3 interacts with the CBC-ARS2 complex and other factors involved in RNA metabolism, including SUT-1 protein, through RNA or protein-mediated contacts revealed by yeast two-hybrid assays, localisation studies and immunoprecipitations. Our data are compatible with a role for SNA-3 in coordinating trans-splicing with target pre-mRNA transcription or in the processing of the Y-branch product of the trans-splicing reaction.

The low-resolution structural models of hepatitis C virus RNA subdomain 5BSL3.2 and its distal complex with domain 3'X point to conserved regulatory mechanisms within the Flaviviridae family.

Subdomain 5BSL3.2 of hepatitis C virus RNA lies at the core of a network of distal RNA–RNA contacts that connect the 5′ and 3′ regions of the viral genome and regulate the translation and replication stages of the viral cycle. Using small-angle X-ray scattering and NMR spectroscopy experiments, we have determined at low resolution the structural models of this subdomain and its distal complex with domain 3′X, located at the 3′-terminus of the viral RNA chain. 5BSL3.2 adopts a characteristic ‘L’ shape in solution, whereas the 5BSL3.2–3′X distal complex forms a highly unusual ‘Y’-shaped kissing junction that blocks the dimer linkage sequence of domain 3′X and promotes translation. The structure of this complex may impede an effective association of the viral polymerase with 5BSL3.2 and 3′X to start negative-strand RNA synthesis, contributing to explain the likely mechanism used by these sequences to regulate viral replication and translation. In addition, sequence and shape features of 5BSL3.2 are present in functional RNA motifs of flaviviruses, suggesting conserved regulatory processes within the Flaviviridae family.

How non-coding RNAs shape nuclear structure to regulate gene expression

The nucleus is a highly organized arrangement of RNA, DNA, and protein molecules that are compartmentalized within three-dimensional (3D) structures involved in shared functional and regulatory processes. Although RNA has long been proposed to play a global role in organizing nuclear structure, exploring this role has remained a challenge because no existing methods can simultaneously measure RNA and DNA contacts within 3D structures. To address this, we developed RNA & DNA SPRITE (RD-SPRITE) to comprehensively map the spatial organization of all RNAs relative to DNA.

RNA promotes the formation of spatial compartments in the nucleus

RNA, DNA, and protein molecules are highly organized within three-dimensional (3D) structures in the nucleus. Although RNA has been proposed to play a role in nuclear organization, exploring this has been challenging because existing methods cannot measure higher-order RNA and DNA contacts within 3D structures. To address this, we developed RNA & DNA SPRITE (RD-SPRITE) to comprehensively map the spatial organization of RNA and DNA. These maps reveal higher-order RNA-chromatin structures associated with three major classes of nuclear function: RNA processing, heterochromatin assembly, and gene regulation. These data demonstrate that hundreds of ncRNAs form high-concentration territories throughout the nucleus, that specific RNAs are required to recruit various regulators into these territories, and that these RNAs can shape long-range DNA contacts, heterochromatin assembly, and gene expression. These results demonstrate a mechanism where RNAs form high-concentration territories, bind to diffusible regulators, and guide them into compartments to regulate essential nuclear functions.

Expanding the binding specificity for RNA recognition by a PUF domain

The ability to design a protein to bind specifically to a target RNA enables numerous applications, with the modular architecture of the PUF domain lending itself to new RNA-binding specificities. For each repeat of the Pumilio-1 PUF domain, we generate a library that contains the 8,000 possible combinations of amino acid substitutions at residues critical for RNA contact. We carry out yeast three-hybrid selections with each library against the RNA recognition sequence for Pumilio-1, with any possible base present at the position recognized by the randomized repeat. We use sequencing to score the binding of each variant, identifying many variants with highly repeat-specific interactions. From these data, we generate an RNA binding code specific to each repeat and base. We use this code to design PUF domains against 16 RNAs, and find that some of these domains recognize RNAs with two, three or four changes from the wild type sequence.

Structure and dynamics of the quaternary hunchback mRNA translation repression complex.

A key regulatory process during Drosophila development is the localized suppression of the hunchback mRNA translation at the posterior, which gives rise to a hunchback gradient governing the formation of the anterior-posterior body axis. This suppression is achieved by a concerted action of Brain Tumour (Brat), Pumilio (Pum) and Nanos. Each protein is necessary for proper Drosophila development. The RNA contacts have been elucidated for the proteins individually in several atomic-resolution structures. However, the interplay of all three proteins during RNA suppression remains a long-standing open question. Here, we characterize the quaternary complex of the RNA-binding domains of Brat, Pum and Nanos with hunchback mRNA by combining NMR spectroscopy, SANS/SAXS, XL/MS with MD simulations and ITC assays. The quaternary hunchback mRNA suppression complex comprising the RNA binding domains is flexible with unoccupied nucleotides functioning as a flexible linker between the Brat and Pum-Nanos moieties of the complex. Moreover, the presence of the Pum-HD/Nanos-ZnF complex has no effect on the equilibrium RNA binding affinity of the Brat RNA binding domain. This is in accordance with previous studies, which showed that Brat can suppress mRNA independently and is distributed uniformly throughout the embryo.

Structure and mechanism of Mycobacterium smegmatis polynucleotide phosphorylase.

Polynucleotide phosphorylase (PNPase) catalyzes stepwise phosphorolysis of the 3′-terminal phosphodiesters of RNA chains to yield nucleoside diphosphate products. In the reverse reaction, PNPase acts as a polymerase, using NDPs as substrates to add NMPs to the 3′-OH terminus of RNA chains while expelling inorganic phosphate. The apparent essentiality of PNPase for growth of M. tuberculosis militates for mycobacterial PNPase as a potential drug target. A cryo-EM structure of Mycobacterium smegmatis PNPase (MsmPNPase) reveals a characteristic ring-shaped homotrimer in which each protomer consists of two RNase PH-like domains and an intervening α-helical module on the inferior surface of the ring. The carboxy-terminal KH and S1 domains, which impart RNA specificity to MsmPNPase, are on the opposite face of the core ring and are conformationally mobile. Single particle reconstructions of MsmPNPase in the act of poly(A) synthesis highlight a 3′-terminal (rA)4 oligonucleotide and two magnesium ions in the active site and an adenine nucleobase in the central tunnel. We identify amino acids that engage the 3′ segment of the RNA chain (Phe68, Arg105, Arg112, Arg430, Arg431) and the two metal ions (Asp526, Asp532, Gln546, Asp548), and we infer those that bind inorganic phosphate (Thr470, Ser471, His435, Lys534). Alanine mutagenesis pinpointed RNA and phosphate contacts as essential (Arg105, Arg431, Lys534, Thr470 + Ser471), important (Arg112, Arg430), or unimportant (Phe68) for PNPase activity. Severe phosphorolysis and polymerase defects accompanying alanine mutations of the enzymic metal ligands suggest a two-metal mechanism of catalysis by MsmPNPase.

The SARS-CoV-2 RNA\u2013protein interactome in infected human cells

Characterizing the interactions that SARS-CoV-2 viral RNAs make with host cell proteins during infection can improve our understanding of viral RNA functions and the host innate immune response. Using RNA antisense purification and mass spectrometry, we identified up to 104 human proteins that directly and specifically bind to SARS-CoV-2 RNAs in infected human cells. We integrated the SARS-CoV-2 RNA interactome with changes in proteome abundance induced by viral infection and linked interactome proteins to cellular pathways relevant to SARS-CoV-2 infections. We demonstrated by genetic perturbation that cellular nucleic acid-binding protein (CNBP) and La-related protein 1 (LARP1), two of the most strongly enriched viral RNA binders, restrict SARS-CoV-2 replication in infected cells and provide a global map of their direct RNA contact sites. Pharmacological inhibition of three other RNA interactome members, PPIA, ATP1A1, and the ARP2/3 complex, reduced viral replication in two human cell lines. The identification of host dependency factors and defence strategies as presented in this work will improve the design of targeted therapeutics against SARS-CoV-2.

Evaluating DCA-based method performances for RNA contact prediction by a well-curated data set.

RNA molecules play many pivotal roles in a cell that are still not fully understood. Any detailed understanding of RNA function requires knowledge of its three-dimensional structure, yet experimental RNA structure resolution remains demanding. Recent advances in sequencing provide unprecedented amounts of sequence data that can be statistically analyzed by methods such as direct coupling analysis (DCA) to determine spatial proximity or contacts of specific nucleic acid pairs, which improve the quality of structure prediction. To quantify this structure prediction improvement, we here present a well curated data set of about 70 RNA structures of high resolution and compare different nucleotide–nucleotide contact prediction methods available in the literature. We observe only minor differences between the performances of the different methods. Moreover, we discuss how robust these predictions are for different contact definitions and how strongly they depend on procedures used to curate and align the families of homologous RNA sequences.

Assessing the accuracy of direct-coupling analysis for RNA contact prediction.

Many noncoding RNAs are known to play a role in the cell directly linked to their structure. Structure prediction based on the sole sequence is, however, a challenging task. On the other hand, thanks to the low cost of sequencing technologies, a very large number of homologous sequences are becoming available for many RNA families. In the protein community, the idea of exploiting the covariance of mutations within a family to predict the protein structure using the direct-coupling-analysis (DCA) method has emerged in the last decade. The application of DCA to RNA systems has been limited so far. We here perform an assessment of the DCA method on 17 riboswitch families, comparing it with the commonly used mutual information analysis and with state-of-the-art R-scape covariance method. We also compare different flavors of DCA, including mean-field, pseudolikelihood, and a proposed stochastic procedure (Boltzmann learning) for solving exactly the DCA inverse problem. Boltzmann learning outperforms the other methods in predicting contacts observed in high-resolution crystal structures.

Mapping RNA-chromatin interactions by sequencing with iMARGI.

RNA-chromatin interactions represent an important aspect of the transcriptional regulation of genes and transposable elements. However, analyses of chromatin-associated RNAs (caRNAs) are often limited to one caRNA at a time. Here, we describe the iMARGI (in situ mapping of RNA-genome interactome) technique, which is used to discover caRNAs and reveal their respective genomic interaction loci. iMARGI starts with in situ crosslinking and genome fragmentation, followed by converting each proximal RNA-DNA pair into an RNA-linker-DNA chimeric sequence. These chimeric sequences are subsequently converted into a sequencing library suitable for paired-end sequencing. A standardized bioinformatic software package, iMARGI-Docker, is provided to decode the paired-end sequencing data into caRNA-DNA interactions. Compared to its predecessor MARGI (mapping RNA-genome interactions), the number of input cells for iMARGI is 3-5 million (a 100-fold reduction), experimental time is reduced, and clear checkpoints have been established. It takes a few hours a day and a total of 8 d to complete the construction of an iMARGI sequencing library and 1 d to carry out data processing with iMARGI-Docker.