Principles for targeting RNA with drug-like small molecules

19F NMR in RNA structural biology: exploring structures, dynamics, and small molecule interactions

RNA molecules play essential roles in numerous biological pathways, making them attractive targets for drug discovery. Despite the challenges in developing small molecules targeting RNA, the success in developing compounds that modulate RNA function underscores its therapeutic potential. 19F NMR spectroscopy has emerged as a powerful tool in structural biology and drug discovery, particularly for studying macromolecular structures and ligand interactions. As RNA continues to gain prominence as a drug target, 19F NMR is expected to play a pivotal role in advancing RNA-focused drug discovery. This review describes the diverse applications of 19F NMR in RNA biology, including its use in characterizing RNA structures, probing molecular dynamics, identifying small-molecule binders, and investigating interaction mechanisms of small-molecule ligands. By providing detailed structural and ligand binding insights, 19F NMR will facilitate the discovery of RNA-targeting therapeutics and deepen our understanding of RNA modulatory mechanisms.

Engineering covalent small molecule-RNA complexes in living cells

Covalent labeling of RNA in living cells poses many challenges. Here we describe a structure-guided approach to engineer covalent RNA aptamer-ligand complexes. The key is to modify the cognate ligand with an electrophilic handle that allows it to react with a guanine at the RNA binding site. We illustrate this for the preQ1-I riboswitch, in vitro and in vivo. Further, we demonstrate the versatility of the approach with a covalent fluorescent light-up aptamer. The coPepper system maintains strong fluorescence in live-cell imaging even after washing, can be used for super-resolution microscopy and, most notably, is uniquely suited for fluorescence recovery after photobleaching to monitor intracellular RNA dynamics. In addition, we have generated a Pepper ligand with a second handle for bioorthogonal chemistry to allow easily traceable pull-down of the covalently linked target RNA. Finally, we provide evidence for the suitability of this tethering strategy for drug targeting.

Specific Interaction between a Fluoroquinolone Derivative, KG022, and RNAs with a Single Bulge

Small compounds targeting RNAs are recognized as a promising modality in drug discovery. We have found that a fluoroquinolone derivative, KG022, binds to RNAs with single-bulged residues. It has been demonstrated by 1H NMR that KG022 binds to RNAs with a bulged G or C and a GC or AU base pair at the 3' adjacent to the bulged residues. In the present study, the effects of the base pairs at the 5' adjacent to the bulged residues on the interaction of KG022 were analyzed mainly by 1H NMR. It was found that KG022 prefers UA and CG base pairs at the 5' adjacent to the bulged residues, indicating that a stable complex is formed by the stacking interaction among the fluoroquinolone ring and the purine bases of the 5' and 3' sides. In addition, this was confirmed by analysis of the 19F-NMR spectra. Analysis of temperature dependences of NMR spectra revealed that KG022 forms a more stable complex with RNAs having CG base pairs at the 5' adjacent position than those with UA base pairs. This work presented useful information for the development of small molecules having higher affinity to target RNAs.

Molecular insights into de novo small-molecule recognition by an intron RNA structure

Despite the promise of vastly expanding the druggable genome, rational design of RNA-targeting ligands remains challenging as it requires the rapid identification of hits and visualization of the resulting cocomplexes for guiding optimization. Here, we leveraged high-throughput screening, medicinal chemistry, and structural biology to identify a de novo splicing inhibitor against a large and highly folded fungal group I intron. High-resolution cryoEM structures of the intron in different liganded states not only reveal molecular interactions that rationalize experimental structure-activity relationship but also shed light on a unique strategy whereby RNA-associated metal ions and RNA conformation exhibit exceptional plasticity in response to small-molecule binding. This study reveals general principles that govern RNA-ligand recognition, the interplay between chemical bonding specificity, and dynamic responses within an RNA target.

RNA sample optimization for cryo-EM analysis

RNAs play critical roles in most biological processes. Although the three-dimensional (3D) structures of RNAs primarily determine their functions, it remains challenging to experimentally determine these 3D structures due to their conformational heterogeneity and intrinsic dynamics. Cryogenic electron microscopy (cryo-EM) has recently played an emerging role in resolving dynamic conformational changes and understanding structure-function relationships of RNAs including ribozymes, riboswitches and bacterial and viral noncoding RNAs. A variety of methods and pipelines have been developed to facilitate cryo-EM structure determination of challenging RNA targets with small molecular weights at subnanometer to near-atomic resolutions. While a wide range of conditions have been used to prepare RNAs for cryo-EM analysis, correlations between the variables in these conditions and cryo-EM visualizations and reconstructions remain underexplored, which continue to hinder optimizations of RNA samples for high-resolution cryo-EM structure determination. Here we present a protocol that describes rigorous screenings and iterative optimizations of RNA preparation conditions that facilitate cryo-EM structure determination, supplemented by cryo-EM data processing pipelines that resolve RNA dynamics and conformational changes and RNA modeling algorithms that generate atomic coordinates based on moderate- to high-resolution cryo-EM density maps. The current protocol is designed for users with basic skills and experience in RNA biochemistry, cryo-EM and RNA modeling. The expected time to carry out this protocol may range from 3 days to more than 3 weeks, depending on the many variables described in the protocol. For particularly challenging RNA targets, this protocol could also serve as a starting point for further optimizations.

Ligand-binding pockets in RNA and where to find them

RNAs are critical regulators of gene expression, and their functions are often mediated by complex secondary and tertiary structures. Structured regions in RNA can selectively interact with small molecules-via well-defined ligand-binding pockets-to modulate the regulatory repertoire of an RNA. The broad potential to modulate biological function intentionally via RNA-ligand interactions remains unrealized, however, due to challenges in identifying compact RNA motifs with the ability to bind ligands with good physicochemical properties (often termed drug-like). Here, we devise fpocketR, a computational strategy that accurately detects pockets capable of binding drug-like ligands in RNA structures. Remarkably few, roughly 50, of such pockets have ever been visualized. We experimentally confirmed the ligandability of novel pockets detected with fpocketR using a fragment-based approach introduced here, Frag-MaP, that detects ligand-binding sites in cells. Analysis of pockets detected by fpocketR and validated by Frag-MaP reveals dozens of sites able to bind drug-like ligands, supports a model for RNA secondary structural motifs able to bind quality ligands, and creates a broad framework for understanding the RNA ligand-ome.

Electrostatic Anchoring in RNA-Ligand Design─Dissecting the Effects of Positive Charges on Affinity, Selectivity, Binding Kinetics, and Thermodynamics

Targeting RNA with small molecules is an emerging field in medicinal chemistry. However, highly potent ligands are often challenging to achieve. One intuitive strategy to enhance ligand's potency is the implementation of positively charged moieties to interact with the negatively charged RNA phosphate backbone. We investigated the effect of such "electrostatic anchors" on binding affinity, kinetics, thermodynamics, and selectivity by MST, SPR, and ITC experiments, respectively, with the Ba SAM-VI riboswitch and the Tte preQ1 riboswitch aptamer model systems. RNA-ligand interactions were dominated by enthalpy, and electrostatic anchors had moderate effects on binding affinity driven by faster association rates for higher charged ligands. Despite the observations of loose binding interactions in SPR experiments with multibasic ligands, selectivity over structurally unrelated RNA off-targets was maintained. Therefore, the addition of positively charged moieties is no universal RNA-ligand design principle, but a purposefully implemented ionic RNA-ligand interaction can enhance potency without impairing selectivity.

Genome-Wide Analysis of Stable RNA Secondary Structures across Multiple Organisms Using Chemical Probing Data: Insights into Short Structural Motifs and RNA-Targeting Therapeutics

Small molecules targeting specific RNA-binding sites, including stable and transient RNA structures, are emerging as effective pharmacological approaches for modulating gene expression. However, little is understood about how stable RNA secondary structures are shared across organisms, which is an important factor in controlling drug selectivity. In this study, I provide an analytical pipeline named RNA secondary structure finder (R2S-Finder) to discover short, stable RNA structural motifs in humans, Escherichia coli (E. coli), SARS-CoV-2, and Zika virus by leveraging existing in vivo and in vitro genome-wide chemical RNA-probing datasets. I found several common features across the organisms. For example, apart from the well-documented tetraloops, AU-rich tetraloops are widely present in different organisms. I also validated that the 5' untranslated region (UTR) contains a higher proportion of stable structures than the coding sequences in humans and Zika virus. In general, stable structures predicted from in vitro (protein-free) and in vivo datasets are consistent across different organisms, indicating that stable structure formation is mostly driven by RNA folding, while a larger variation was found between in vitro and in vivo data for certain RNA types, such as human long intergenic noncoding RNAs (lincRNAs). Finally, I predicted stable three- and four-way RNA junctions that exist under both in vivo and in vitro conditions and can potentially serve as drug targets. All results of stable structures, stem-loops, internal loops, bulges, and n-way junctions have been collated in the R2S-Finder database (https://github.com/JingxinWangLab/R2S-Finder), which is coded in hyperlinked HTML pages and tabulated in CSV files.

Identifying RNA-small Molecule Binding Sites Using Geometric Deep Learning with Language Models

RNAs are emerging as promising therapeutic targets, yet identifying small molecules that bind to them remains a significant challenge in drug discovery. This underscores the crucial role of computational modeling in predicting RNA-small molecule binding sites. However, accurate and efficient computational methods for identifying these interactions are still lacking. Recently, advances in large language models (LLMs), previously successful in DNA and protein research, have spurred the development of RNA-specific LLMs. These models leverage vast unlabeled RNA sequences to autonomously learn semantic representations with the goal of enhancing downstream tasks, particularly those constrained by limited annotated data. Here, we develop RNABind, an embedding-informed geometric deep learning framework to detect RNA-small molecule binding sites from RNA structures. RNABind integrates RNA LLMs into advanced geometric deep learning networks, which encodes both RNA sequence and structure information. To evaluate RNABind, we first compile the largest RNA-small molecule interaction dataset from the entire multi-chain complex structure instead of single-chain RNAs. Extensive experiments demonstrate that RNABind outperforms existing state-of-the-art methods. Besides, we conduct an extensive experimental evaluation of eight pre-trained RNA LLMs, assessing their performance on the binding site prediction task within a unified experimental protocol. In summary, RNABind provides a powerful tool on exploring RNA-small molecule binding site prediction, which paves the way for future innovations in the RNA-targeted drug discovery.

Stacking Interactions of Druglike Heterocycles with Nucleobases

Stacking interactions contribute significantly to the interaction of small molecules with RNA, and harnessing the power of these interactions will likely prove important in the development of RNA-targeting inhibitors. To this end, we present a comprehensive computational analysis of stacking interactions between a set of 54 druglike heterocycles and the natural nucleobases. We first show that heterocycle choice can tune the strength of stacking interactions with nucleobases over a large range and that heterocycles favor stacked geometries that cluster around a discrete set of stacking loci characteristic of each nucleobase. Symmetry-adapted perturbation theory results indicate that the strengths of these interactions are modulated primarily by electrostatic and dispersion effects. Based on this, we present a multivariate predictive model of the maximum strength of stacking interactions between a given heterocycle and nucleobase that depends on molecular descriptors derived from the electrostatic potential. These descriptors can be readily computed using density functional theory or predicted directly from atom connectivity (e.g., SMILES). This model is used to predict the maximum possible stacking interactions of a set of 1854 druglike heterocycles with the natural nucleobases. Finally, we show that trivial modifications of standard (fixed-charge) molecular mechanics force fields reduce errors in predicted stacking interaction energies from around 2 kcal/mol to below 1 kcal/mol, providing a pragmatic means of predicting more reliable stacking interaction energies using existing computational workflows. We also analyze the stacking interactions between ribocil and a bacterial riboswitch, showing that two of the three aromatic heterocyclic components engage in near-optimal stacking interactions with binding site nucleobases.

Advances in Isotope Labeling for Solution Nucleic Acid Nuclear Magnetic Resonance Spectroscopy

The availability of structural biology methods for nucleic acid still lags behind that of proteins, as evidenced by the smaller number of structures (DNA: 2513, RNA: 1899, nucleic acid-protein complexes: 13 842, protein: 196 887) deposited in the protein database. The skewed ratio of nucleic acid structures, relative to proteins (≈1:50), is inverted with respect to the cellular output of RNA and proteins in higher organisms (≈50:1). While nuclear magnetic resonance (NMR) is an attractive biophysical tool capable of bridging this gap at the molecular level, the conformational flexibility, line broadening, and low chemical shift dispersion of nucleic acids have made the NMR method challenging, especially for structures larger than 35 nucleotides. The incorporation of NMR-active isotopes is a f strategy to combat these problems. Significant strides made to push the size limits of nucleic acid structures solved by NMR using chemoenzymatic 13C- methyl and aromatic 15N- and 19F-13C-labeling are reviewed and challenges and opportunities are evaluated. Combining these isotopic labeling patterns with superior NMR spectroscopic properties, and new DNA/RNA synthesis methods (palindrome-nicking-dependent amplification and segmental labeling and site-specific modifications by template-directed tension), may stimulate advances in NMR studies of large DNA/RNA and their complexes with important biological functions.

The prediction of RNA-small-molecule ligand binding affinity based on geometric deep learning

Small molecule-targeted RNA is an emerging technology that plays a pivotal role in drug discovery and inhibitor design, with widespread applications in disease treatment. Consequently, predicting RNA-small-molecule ligand interactions is crucial. With advancements in computer science and the availability of extensive biological data, deep learning methods have shown great promise in this area, particularly in efficiently predicting RNA-small molecule binding sites. However, few computational methods have been developed to predict RNA-small molecule binding affinities. Meanwhile, most of these approaches rely primarily on sequence or structural representations. Molecular surface information, vital for RNA and small molecule interactions, has been largely overlooked. To address these gaps, we propose a geometric deep learning method for predicting RNA-small molecule binding affinity, named RNA-ligand Surface Interaction Fingerprinting (RLASIF). In this study, we create RNA-ligand interaction fingerprints from the geometrical and chemical features present on molecular surface to characterize binding affinity. RLASIF outperformed other computational methods across ten different test sets from PDBbind NL2020. Compared to the second-best method, our approach improves performance by 10.01 %, 6.67 %, 2.01 % and 1.70 % on four evaluation metrics, indicating its effectiveness in capturing key features influencing RNA-ligand binding strength. Additionally, RLASIF holds potential for virtual screening of potential ligands for RNA and predicting small molecule binding nucleotides within RNA structures.

LncRNA in gastric cancer drug resistance: deciphering the therapeutic strategies

Gastric cancer (GC) is an exceedingly aggressive disease and ranks as the third leading cause of cancer-related deaths, which poses a huge health burden globally. Chemotherapy is commonly employed during the middle to advanced stages of cancer, although it faces frequent treatment failures attributed to drug resistance. Thus, it is imperative for researchers to identify potential targets for overcoming therapeutic resistance, thereby facilitating the development of novel anti-cancer agents for GC patients with advanced stages. Long noncoding RNAs (lncRNAs) are a diverse group of transcripts with limited protein-coding capacity, which have been recognized for functional molecules for regulating cancer progression including cell proliferation, metastasis, and drug resistance in GC. In this review, we examine the intricate molecular networks on the role of lncRNAs in drug resistance of GC. LncRNAs conferred cancer cell resistance to anti-cancer drug through various molecular mechanisms, therefore functioning as promising therapeutic targets for GC patients. Additionally, we discuss current advancements of strategies targeting lncRNAs in cancer therapy, which may pave the way for lncRNA-mediated precision medicine for this malignant disease.

fpocketR: A platform for identification and analysis of ligand-binding pockets in RNA

Small molecules that bind specific sites in RNAs hold promise for altering RNA function, manipulating gene expression, and expanding the scope of druggable targets beyond proteins. Identifying binding sites in RNA that can engage ligands with good physicochemical properties remains a significant challenge. fpocketR is a software package for identifying, characterizing, and visualizing ligand-binding sites in RNA. fpocketR was optimized, through comprehensive analysis of currently available RNA-ligand complexes, to identify pockets in RNAs able to bind small molecules possessing favorable properties, generally termed drug-like. Here, we demonstrate use of fpocketR to analyze RNA-ligand interactions and novel pockets in small and large RNAs, to assess ensembles of RNA structure models, and to identify pockets in dynamic RNA systems. fpocketR performs best with RNA structures visualized at high (≤3.5 Å) resolution, but also provides useful information with lower resolution structures and computational models. fpocketR is a powerful, freely available tool for discovery and analysis of ligand-binding pockets in RNA molecules.

rCGMM: A Coarse-Grained Force Field Embedding Elastic Network for Studying Small Noncoding RNA Dynamics

Short noncoding RNA molecules play significant roles in catalysis, biological regulation, and disease pathways. Their assessment through sequence-based approaches has been a challenge, compounded by the significant structural flexibility accrued from six free backbone torsions per nucleotide. To efficiently study the structure and dynamics of an extensive repertoire of these molecules in a high throughput mode, we have built a coarse-grained force field using one, two, three, and four pseudoatoms to represent the phosphate, sugar, pyrimidines, and purines, respectively. The Boltzmann inversion method was applied to structures of 5 piRNA, 8 miRNA, and 13 siRNA from the Nucleic Acid Database (NDB) to estimate the initial force field parameters and iteratively optimized through 1 μs molecular dynamics run by comparing against an equivalent all-atom simulation using the CHARMM36 force field. We applied an elastic net to model the hydrogen bond network stabilizing the local structure for double-stranded cases. A spine using pseudoatoms was calculated for the same from the coarse-grain beads, and all beads within a threshold radial distance were constrained using soft distance potentials. Lennard-Jones and Coulomb's potential function modeled the nonbonded interaction. Benchmarks on 26 molecules compared through root-mean-square deviation graphs against all-atom simulation show close concurrence for single- and double-stranded small noncoding RNA molecules. The rCGMM force field is available for download at https://github.com/majumderS/rCGMM.

Eurycomalactone switched hepatocellular carcinoma cells into quiescence through 5'tRFAla/DVL/β-catenin pathway inhibition

Although tsRNA has been demonstrated to modulate various physiological processes analogous to miRNA, the potential regulatory functions and mechanisms of tsRNAs related to the pharmacological effects of small molecule drugs remain unclear. Herein, it is shown that eurycomalactone (ELT), a natural product, can reversibly switch hepatocellular carcinoma (HCC) PLC/PRF/5 and HUH7 cells into a quiescent state. This quiescence is characterized by cell proliferation inhibition without cytotoxicity, cell cycle arrest at the G0/G1 phase, and cell reactivation following the removal of ELT. Given the established role of β-catenin activity in mediating cancer cellular quiescence or proliferation, a notable reduction in total, cytoplasmic, and nuclear β-catenin expression, along with its downstream targets Survivin, c-myc, and Cyclin D1, was observed in ELT-treated cells. Subsequently, two new tsRNAs, namely 5'tRFAla and 5'tiRNAAla, which match well with the mRNAs of two pivotal upstream regulators (DVL2 and DVL3) of β-catenin based on bioinformatics analyses, were detected to be significantly decreased in ELT-treated PLC/PRF/5 cells using Arraystar small RNA microarray analyses. Consistently, the concentrations of the DVL2 and DVL3 proteins were also found to be reduced by ELT. The mimic of 5'tRFAla could increase the relative expression of DVL2 and DVL3 mRNA and rescue their decrease induced by ELT, while the mimic of 5'tiRNAAla could not. It therefore seems that ELT could down-regulate the expression of 5'tRFAla, leading to the suppression of DVL2 and DVL3 mRNA translation, consequently inhibiting the β-catenin signaling pathway and reversibly switching HCC cells into a quiescent state. Conclusively, our findings imply that tsRNAs, like miRNAs, might activate the translation of their matched mRNAs in non-dividing cells and provide a possible potential for repressing tumor cell growth, although further evidence is still needed.

Functional expression mechanisms of food-derived components based on target proteins

Food-derived polyphenols and some alkaloids have reported bioactivities related to the prevention of systemic metabolic disorders such as obesity, glucose intolerance, and dyslipidemia. For food-derived components to exert their functions in vivo, it is essential to interact with biological factors such as proteins, lipids, and nucleic acids. However, it is still unclear whether bioactive components in foods express functions related to their target factors. In this review, I introduce the target proteins in which food-derived components express functions in cells.

RNAmigos2: accelerated structure-based RNA virtual screening with deep graph learning

RNAs are a vast reservoir of untapped drug targets. Structure-based virtual screening (VS) identifies candidate molecules by leveraging binding site information, traditionally using molecular docking simulations. However, docking struggles to scale with large compound libraries and RNA targets. Machine learning offers a solution but remains underdeveloped for RNA due to limited data and practical evaluations. We introduce a data-driven VS pipeline tailored for RNA, utilizing coarse-grained 3D modeling, synthetic data augmentation, and RNA-specific self-supervision. Our model achieves a 10,000x speedup over docking while ranking active compounds in the top 2.8% on structurally distinct test sets. It is robust to binding site variations and successfully screens unseen RNA riboswitches in a 20,000-compound in-vitro microarray, with a mean enrichment factor of 2.93 at 1%. This marks the first experimentally validated success of structure-based deep learning for RNA VS.

DRLiPS: a novel method for prediction of druggable RNA-small molecule binding pockets using machine learning

Ribonucleic Acid (RNA) is the central conduit for information transfer in the cell. Identifying potential RNA targets in disease conditions is a challenging task, given the vast repertoire of functional non-coding RNAs in a human cell. A potential druggable target must satisfy several criteria, including disease association, cellular accessibility, binding pockets for drug-like molecules, and minimal cross-reactivity. While several methods exist for prediction of druggable proteins, they cannot be repurposed for RNAs due to fundamental differences in their binding modality. Taking all these constraints into account, a new structure-based model, Druggable RNA-Ligand binding Pocket Selector (DRLiPS), is developed here to predict binding site-level druggability of any given RNA target. A novel strategy for sampling negative binding sites in RNA structures using three parallel approaches is demonstrated here to improve model specificity: backbone motif search, exhaustive pocket prediction, and blind docking. An external blind test dataset has also been curated to showcase the model's generalizability to both experimental and modelled apo state RNA structures. DRLiPS has achieved an F1-score of 0.70, precision of 0.61, specificity of 0.89, and recall of 0.73 on this external test dataset, outperforming two existing methods, DrugPred_RNA and RNACavityMiner. Further analysis indicates that the features selected for model-building generalize well to both apo and holo states with a backbone RMSD tolerance of 3 Å. It can also predict the effect of binding site single point mutations on druggability, which can aid in optimizing synthetic RNA aptamers for small molecule recognition. The DRLiPS model is freely accessible at https://web.iitm.ac.in/bioinfo2/DRLiPS/.

Ligand-binding pockets in RNA, and where to find them

RNAs are critical regulators of gene expression, and their functions are often mediated by complex secondary and tertiary structures. Structured regions in RNA can selectively interact with small molecules - via well-defined ligand binding pockets - to modulate the regulatory repertoire of an RNA. The broad potential to modulate biological function intentionally via RNA-ligand interactions remains unrealized, however, due to challenges in identifying compact RNA motifs with the ability to bind ligands with good physicochemical properties (often termed drug-like). Here, we devise fpocketR, a computational strategy that accurately detects pockets capable of binding drug-like ligands in RNA structures. Remarkably few, roughly 50, of such pockets have ever been visualized. We experimentally confirmed the ligandability of novel pockets detected with fpocketR using a fragment-based approach introduced here, Frag-MaP, that detects ligand-binding sites in cells. Analysis of pockets detected by fpocketR and validated by Frag-MaP reveals dozens of newly identified sites able to bind drug-like ligands, supports a model for RNA secondary structural motifs able to bind quality ligands, and creates a broad framework for understanding the RNA ligand-ome.

Chemotranscriptomic profiling with a thiamine monophosphate photoaffinity probe

RNA is a multifaceted biomolecule with numerous biological functions and can interact with small molecule metabolites as exemplified by riboswitches. Here, we profile the Escherichia coli transcriptome on interactions with the metabolite Thiamine Monophosphate (TMP). We designed and synthesized a photoaffinity probe based on the scaffold of TMP and applied it to chemotranscriptomic profiling. Using next-generation RNA sequencing, several potential interactions between bacterial transcripts and the probe were identified. A remarkable interaction between the TMP probe and the well-characterized Flavin Mononucleotide (FMN) riboswitch was validated by RT-qPCR, and further verified with competition assays. Localization of the photocrosslinked nucleotides using reverse transcription and docking predictions of the probe suggested binding to the riboswitch aptamer. After examining binding of unmodified TMP to the riboswitch using SHAPE, we found selective yet moderate binding interactions, potentially mediated by the phosphate group of TMP. Lastly, TMP appeared to enhance gene expression of a reporter gene that is under riboswitch control, while the natural ligand FMN displayed an inhibitory effect, hinting at a potential biological role of TMP. This work showcases the possibility of chemotranscriptomic profiling to identify new RNA-small molecule interactions.

Harnessing RNA-Protein Interactions for Therapeutic Interventions

Interactions between RNAs and proteins play a crucial role in various diseases, including viral infections and cancer. Hence, understanding and inhibiting these interactions are important for the development of novel therapeutics. However, the identification of drugs targeting RNA-protein interactions with high specificity and affinity is challenged by our limited molecular understanding of these interactions. Recent focus on structural and biochemical characterization, coupled with high-throughput screening technologies and computational modeling, have accelerated the identification of new RBPs and optimization of potential inhibitors. This review discusses key examples of inhibitors developed over the past decade that effectively disrupt pathogenic RNA-protein interactions. We focus on small molecule and peptide-based inhibitors that have shown promise in disrupting crucial RNA-protein interactions in eukaryotes, prokaryotes, and viruses. We also present the challenges and future directions in this field, emphasizing the need to achieve improved specificity and reduce the off-target effects of the inhibitors. This review aims to contribute to ongoing efforts towards the development of novel therapeutic agents targeting RNA-protein interactions by providing an in-depth analysis of significant developments and emerging trends in this rapidly growing field.

Development of a DNA-encoded library screening method "DEL Zipper" to empower the study of RNA-targeted chemical matter

To date, RNA-targeted chemical matter is under explored due to a lack of robust screening assays. In this study, we present a novel RNA-targeted small molecule screening approach using a specialized DNA-encoded library (DEL). Our findings reveal that the specialized DEL library, called "DEL Zipper", can significantly reduce single-stranded DNA-RNA region interaction signals during various kinds of RNA selection. By performing the selection against both G-quadruplex, we have identified novel hits that interact with RNA targets and the results are validated through binding. This study demonstrates that the "DEL Zipper" method is a robust screening assay that has potential for discovering small molecule ligands for diverse RNA targets.

Identification of chemical inhibitors targeting long noncoding RNA through gene signature-based high throughput screening

Scalable methods for functionally high-throughput screening of RNA-targeting small molecules are currently limited. Here, an RNA knockdown gene signature and high-throughput sequencing-based high-throughput screening (HTS2) were integrated to identify RNA-targeting compounds. We first generated a gene signature characterizing the knockdown of the long non-coding RNA LINC00973. Then, screening of 8199 compounds by HTS2 assay identified that treatments of Hesperadin and GSK1070916 significantly mimic the expression pattern of the LINC00973 knockdown gene signature. Functionally, cell phenotype changes after treatments of these two compounds also mimic the losing function of LINC00973 in multiple types of cancer cells. Mechanistically, the inhibitory action of these two compounds on LINC00973 primarily operates via the AURKB-mediated MAPK signaling pathway, resulting in reduced expression of the transcription factor c-Jun. Consequently, this leads to the suppression of LINC00973 transcription. Moreover, these two compounds significantly inhibit xenograft tumor growth in vivo. Clinically, we further found that breast tumors with high expression of LINC00973 also show relatively high expression of AURKB or JUN, and vice versa. In summary, we established a novel high-throughput screening strategy to identify small molecules capable of targeting RNA, provided two promising compounds targeting LINC00973 and further shed light on the underlying transcriptional upregulation mechanism of LINC00973 within cancer cells.

Refining Ligand Poses in RNA/Ligand Complexes of Pharmaceutical Relevance: A Perspective by QM/MM Simulations and NMR Measurements

Predicting the binding poses of ligands targeting RNAs is challenging. Here, we propose that using first-principles quantum mechanics/molecular mechanics (QM/MM) simulations, which incorporate automatically polarization effects, can help refine the structural determinants of ligand/RNA complexes in aqueous solution. In fact, recent advances in massively parallel computer architectures (such as exascale machines), combined with the power of machine learning, are greatly expanding the domain of applicability of these types of notoriously expensive simulations. We corroborate this proposal by carrying out a QM/MM-based study on a ligand targeting CAG repeat-RNA, involved in Huntington's disease. The calculations indeed show a clear improvement in the ligand binding properties, and they are consistent with the NMR measurements, also performed here. Thus, this type of approach may be useful for practical applications in the design of ligands targeting RNA in the near future.

Stimulus-activated ribonuclease targeting chimeras for tumor microenvironment activated cancer therapy

RNA degradation using ribonuclease targeting chimeras (RiboTACs) is a promising approach for cancer therapy. However, potential off-target degradation is a serious issue. Here, a RiboTAC is designed for tumor microenvironment triggered activation. The tumor microenvironment activated RiboTAC (TaRiboTAC) incorporates two pre-miR-21 binders, a near-infrared fluorophore IR780, an RGD targeting peptide and a phenylboronic acid caged ribonuclease recruiter. The caged ribonuclease recruiter is embedded in the molecule and exposed in acidic pH, the phenylboronic acid cage is removed by H2O2 making the TaRiboTAC responsive to the acidic and high H2O2 levels in the tumor microenvironment. It is shown the TaRiboTAC targets tumor tissue and degrades pre-miR-21. The degradation of pre-miR-21 by TaRiboTACs significantly increases the radiotherapeutic susceptibility of cancer cells achieving efficient suppression of human lung adenocarcinoma A549 tumors in living mice.

Polysaccharide-based biomaterials for regenerative therapy in intervertebral disc degeneration

Intervertebral disc (IVD) degeneration represents a significant cause of chronic back pain and disability, with a substantial impact on the quality of life. Conventional therapeutic modalities frequently address the symptoms rather than the underlying etiology, underscoring the necessity for regenerative therapies that restore disc function. Polysaccharide-based materials, such as hyaluronic acid, alginate, chitosan, and chondroitin sulfate, have emerged as promising candidates for intervertebral disc degeneration (IVDD) therapy due to their biocompatibility, biodegradability, and ability to mimic the native extracellular matrix (ECM) of the nucleus pulposus (NP). These materials have demonstrated the capacity to support cell viability, facilitate matrix production, and alleviate inflammation in vitro and in vivo, thus supporting tissue regeneration and restoring disc function in comparison to conventional treatment. Furthermore, polysaccharide-based hydrogels have demonstrated the potential to deliver bioactive molecules, including growth factors, cytokines and anti-inflammatory drugs, directly to the degenerated disc environment, thereby enhancing therapeutic outcomes. Therefore, polysaccharide-based materials provide structural support and facilitate the regeneration of native tissue, representing a versatile and effective approach for the treatment of IVDD. Despite their promise, challenges such as limited long-term stability, potential immunogenicity, and the difficulty in scaling up production for clinical use remain. This review delineates the potential of various polysaccharides during the fabrication of hydrogels and scaffolds for disc regeneration, guiding and inspiring future research to focus on optimizing these materials for clinical translation for IVDD repair and regeneration.

A journey into siRNA therapeutics development: A focus on Pharmacokinetics and Pharmacodynamics

siRNA therapeutics are emerging novel modalities targeting highly specific mRNA via RNA interference mechanism. Its unique pharmacokinetics (PKs) and pharmacodynamics (PDs) are significant challenges for clinical use. Furthermore, naked siRNA is a highly soluble macromolecule with a negative charge, making plasma membrane penetration a significant hurdle. It is also vulnerable to nuclease degradation. Therefore, advanced formulation technologies, such as lipid nanoparticles and N-acetylgalactosamine conjugation, have been developed and are now used in clinical practice to enhance target organ delivery and stability. The innate complex biological mechanisms of siRNA, along with its formulation, are major determinants of the PK/PD characteristics of siRNA products. To systematically and quantitatively understand these characteristics, it is essential to develop and utilize quantitative PK/PD models for siRNA therapeutics. In this review, the effects of formulation on the PKs and PK/PD models of approved siRNA products were presented, highlighting the importance of selecting appropriate biomarkers and understanding formulation, PKs, and PDs for quantitative interpreting the relationship between plasma concentration, organ concentration, biomarkers, and efficacy.

Designing small molecules that target a cryptic RNA binding site via base displacement

Most RNA-binding small molecules have limited solubility, weak affinity, and/or lack of specificity, restricting the medicinal chemistry often required for lead compound discovery. We reasoned that conjugation of these unfavorable ligands to a suitable "host" molecule can solubilize the "guest" and deliver it site-specifically to an RNA of interest to resolve these issues. Using this framework, we designed a small molecule library that was hosted by cobalamin (Cbl) to interact with the Cbl riboswitch through a common base displacement mechanism. Combining in vitro binding, cell-based assays, chemoinformatic modeling, and structure-based design, we unmasked a cryptic binding site within the riboswitch that was exploited to discover compounds that have affinity exceeding the native ligand, antagonize riboswitch function, or bear no resemblance to Cbl. These data demonstrate how a privileged biphenyl-like scaffold effectively targets RNA by optimizing π-stacking interactions within the binding pocket.

RNAbpFlow: Base pair-augmented SE(3)-flow matching for conditional RNA 3D structure generation

Motivation

Despite the groundbreaking advances in deep learning-enabled methods for bimolecular modeling, predicting accurate three-dimensional (3D) structures of RNA remains challenging due to the highly flexible nature of RNA molecules combined with the limited availability of evolutionary sequences or structural homology.

Results

We introduce RNAbpFlow, a novel sequence- and base-pair-conditioned SE(3)-equivariant flow matching model for generating RNA 3D structural ensemble. Leveraging a nucleobase center representation, RNAbpFlow enables end-to-end generation of all-atom RNA structures without the explicit or implicit use of evolutionary information or homologous structural templates. Experimental results show that base pairing conditioning leads to broadly generalizable performance improvements over current approaches for RNA topology sampling and predictive modeling in large-scale benchmarking.

Availability

RNAbpFlow is freely available at https://github.com/Bhattacharya-Lab/RNAbpFlow.

A prodrug nanodevice co-delivering docetaxel and ROR1 siRNA for enhanced triple negative breast cancer therapy

Triple-negative breast cancer (TNBC) has been a clinical challenge due to its high recurrence and metastasis rates. Chemotherapy remains the primary treatment for TNBC after surgery ablation, but it lacks targeted specificity and causes side effects in normal tissues. Receptor tyrosine kinase-like orphan receptor 1 (ROR1) is significantly expressed in TNBC cells, and small interference RNA (siRNA) targeting ROR1 can effectively suppress ROR1 gene expression, thereby inhibiting proliferation and metastasis. However, clinical application of ROR1 siRNA is limited by susceptibility to clearance and difficulty in endosomal escape. In this study, the docetaxel (DTX) prodrug nanoparticle BBRM delivering ROR1 siRNA was constructed. The BBRM could be effectively internalized by tumor cells and endosomal escape to release DTX and ROR1 siRNA. In 4T1 tumor-bearing mice, BBRM could be targeting delivered to tumor and lung tissues, with good biosafety, achieving a tumor inhibition rate of 74.1 % and inhibiting lung metastasis. By integrating chemotherapy and RNA interference therapy, BBRM successfully co-delivered chemotherapeutic agents and siRNA to improve the therapeutic efficacy of triple-negative breast cancer and provided a promising strategy for clinical transformation. STATEMENT OF SIGNIFICANCE: Chemotherapy is still the primary treatment for triple-negative breast cancer (TNBC) after surgery ablation, but it causes side effects without targeting capacity. ROR1 is significantly expressed in TNBC cells, and RNA interference for ROR1 can suppress ROR1 gene expression to inhibit tumor proliferation. However, as oligonucleotides, effect of ROR1 siRNA is limited by susceptibility to clearance and difficulty in endosomal escape. In this work, we designed a nanodevice based on a docetaxel (DTX) prodrug that targets ROR1 for the synergistic therapy of TNBC. We constructed a nanoparticle (BBRM) for co-delivery of the DTX and ROR1 siRNA. The BBRM could be effectively internalized by tumor cells and endosomal escape. The ROR1 siRNA downregulated ROR1 protein expression and improved the anti-proliferative and anti-metastatic effects. In addition, BBRM reversed the immunosuppressive tumor microenvironment, thus improving breast cancer therapeutic efficacy. It was a pioneering investigation in synergistic chemo-gene therapy by co-delivering DTX and ROR1 siRNA for TNBC treatment.

A general strategy for engineering GU base pairs to facilitate RNA crystallization

X-ray crystallography is a fundamental technique that provides atomic-level insights into RNA structures. However, obtaining crystals of RNA structures diffracting to high resolution is challenging. We introduce a simple strategy to enhance the resolution limit of RNA crystals by the selective substitution of Watson-Crick pairs by GU pairs within RNA sequences. Our approach has successfully yielded high-resolution structures for eight unique RNA crystals. Notably, six instances showed marked resolution enhancement upon GC/AU to GU base pair substitution, with two cases achieving high-resolution structures from initially poor data. In one case, reverting GU to GC base pairs also improved resolution. Our method facilitated the first structural determinations of the Long Interspersed Nuclear Element-1 and Olfactory Receptor family 4 subfamily K member 15 ribozymes, the 2'-deoxyguanosine-III riboswitch and the Broccoli RNA aptamer. The placement of GU base pairs within the first 5' helical stem of any given RNA species, or in one peripheral stem, is shown to be sufficient. These results offer a simple and effective approach for designing sequences or selecting sequences from homologous sequences, for high-resolution RNA structure determination.

Design and Evaluation of Azaspirocycles as RNA binders

This study presents efficient synthetic pathways for preparing novel azaspirocycles. These methodologies involve functionalizing key bicyclic hydrazines with a substituent on one of their bridgehead carbon atoms. The desired spirocyclic cores were successfully obtained through double reductive amination reactions, intramolecular cyclizations, and cleavages of the N-N bond. The isolated molecules possess unique three-dimensional structures, suggesting potential applications in medicinal chemistry and drug discovery. With the growing interest in targeting nucleic acids as a complementary approach to protein-targeting strategies for developing novel active compounds, we investigated the potential of the synthesized azaspirocycles as RNA binders. As a proof of concept, we highlight the promising activity of some compounds as strong binders of HIV-1 TAR RNA and inhibitors of Tat/TAR interactions.

Scaffold-enabled high-resolution cryo-EM structure determination of RNA

Cryo-EM structure determination of protein-free RNAs has remained difficult with most attempts yielding low to moderate resolution and lacking nucleotide-level detail. These difficulties are compounded for small RNAs as cryo-EM is inherently more difficult for lower molecular weight macromolecules. Here we present a strategy for fusing small RNAs to a group II intron that yields high resolution structures of the appended RNA. We demonstrate this technology by determining the structures of the 86-nucleotide (nt) thiamine pyrophosphate (TPP) riboswitch aptamer domain and the recently described 210-nt raiA bacterial non-coding RNA involved in sporulation and biofilm formation. In the case of the TPP riboswitch aptamer domain, the scaffolding approach allowed visualization of the riboswitch ligand binding pocket at 2.5 Å resolution. We also determined the structure of the ligand-free apo state and observe that the aptamer domain of the riboswitch adopts an open Y-shaped conformation in the absence of ligand. Using this scaffold approach, we determined the structure of raiA at 2.5 Å in the core. Our versatile scaffolding strategy enables efficient RNA structure determination for a broad range of small to moderate-sized RNAs, which were previously intractable for high-resolution cryo-EM studies.

Recognition of Noncanonical RNA Base Pairs Using Triplex-Forming Peptide Nucleic Acids

Noncanonical base pairs play an important role in enabling the structural and functional complexity of RNA. Molecular recognition of such motifs is challenging because of their diversity, significant deviation from the Watson-Crick structures, and dynamic behavior, resulting in alternative conformations of similar stability. Triplex-forming peptide nucleic acids (PNAs) have emerged as excellent ligands for the recognition of Watson-Crick base-paired double helical RNA. The present study extends the recognition potential of PNA to RNA helices having noncanonical GoU, AoC, and tandem GoA/AoG base pairs. The purines of the noncanonical base pairs formed M+·GoU, T·AoC, M+·GoA, and T·AoG Hoogsteen triples of similar or slightly reduced stability compared to the canonical M+·G-C and T·A-U triples. Recognition of pyrimidines was more challenging. While the P·CoA triple was only slightly less stable than P·C-G, the E nucleobase did not form a stable triple with U of the UoG wobble pair. Molecular dynamics simulations suggested the formation of expected Hoogsteen hydrogen bonds for all of the stable triples. Collectively, these results expand the scope of triple helical recognition to noncanonical structures and sequence motifs common in biologically relevant RNAs.

LncRNA 3222401L13Rik Is Upregulated in Aging Astrocytes and Regulates Neuronal Support Function Through Interaction with Npas3

Aging leads to cognitive decline and increased risk of neurodegenerative diseases. While molecular changes in central nervous system (CNS) cells contribute to this decline, the mechanisms are not fully understood. Long non-coding RNAs (lncRNAs) are key regulators of cellular functions. Background/Objectives: The roles of lncRNAs in aging, especially in glial cells, are not well characterized. Methods: We investigated lncRNA expression in non-neuronal cells from aged mice and identified 3222401L13Rik, a previously unstudied lncRNA, as upregulated in astrocytes during aging. Results: Knockdown of 3222401L13Rik in primary astrocytes revealed its critical role in regulating genes for neuronal support and synapse organization, a function conserved in human iPSC-derived astrocytes. A 3222401L13Rik interacts with the transcription factor Neuronal PAS Domain Protein 3 (Npas3), and overexpression of Npas3 rescues deficits in astrocytes lacking 3222401L13Rik. Conclusions: These data suggest that 3222401L13Rik upregulation may help delay age-related cognitive decline.

Detailed mechanism of a DNA/RNA nucleobase substituting bridging ligand in diruthenium(II,III) and dirhodium(II,II) tetraacetato paddlewheel complexes: protonation of the leaving acetate is crucial

Paddlewheel complexes of bimetallic scaffolds are emerging metallic agents in the bioinorganic chemistry landscape. In the most commonly employed construct, these complexes are decorated by the carboxylate moiety, prompting their possible deployment to target either protein or nucleic acid targets. In this study, density functional investigation was performed to assess viable mechanistic routes for the substitution of one acetate ligand with one chelating purine, i.e. adenine or guanine, in diruthenium and dirhodium tetraacetate paddlewheel complexes. This study evidenced the relevant stages of the process at an atomistic scale of resolution and provided for the encompassed rate-determining chemical events. Therefore, calculations indicated that acetate decomplexation as well as the concomitant nucleobase bridging proceeded gradually via a multistep process that included protonation of the leaving acetate. The present picture of the mechanism is envisioned to be relevant to the design and interpretation of experiments focused on the reaction of diruthenium and/or dirhodium tetracarboxylate complexes with nucleobases and eventuating in the formation of either nucleobase bridged-complexes or in the dismantling of the bimetallic construct.

Targeting noncoding RNAs to treat atherosclerosis

Noncoding RNAs (ncRNAs) are pivotal for various pathological processes, impacting disease progression. The potential for leveraging ncRNAs to prevent or treat atherosclerosis and associated cardiovascular diseases is of great significance, especially given the increasing prevalence of atherosclerosis in an ageing and sedentary population. Together, these diseases impose a substantial socio-economic burden, demanding innovative therapeutic solutions. This review explores the potential of ncRNAs in atherosclerosis treatment. We commence by examining approaches for identifying and characterizing atherosclerosis-associated ncRNAs. We then delve into the functional aspects of ncRNAs in atherosclerosis development and progression. Additionally, we review current RNA and RNA-targeting molecules in development or under approval for clinical use, offering insights into their pharmacological potential. The importance of improved ncRNA delivery strategies is highlighted. Finally, we suggest avenues for advanced research to accelerate the use of ncRNAs in treating atherosclerosis and mitigating its societal impact. LINKED ARTICLES: This article is part of a themed issue Non-coding RNA Therapeutics. To view the other articles in this section visit http://onlinelibrary.wiley.com/doi/10.1111/bph.v182.2/issuetoc.

Artificial dynamic structure ensemble-guided rational design of a universal RNA aptamer-based sensing tag

Artificially functional RNAs, such as fluorogenic RNA aptamer (FRApt)-based biosensing tag, represent significant advancements in various biological applications but are limited by the lack of insight into dynamic structure ensembles and universal design concepts. Through the development of an artificial RNA structure ensemble, we rationally established an RNA reconstitution model, "SSPepper-Apt," to generate a universal fluorogenic RNA biosensing tag. By utilizing various target-recognizing RNA motifs, SSPepper-Apt enables the modular generation of sensing tags for low-background, highly selective imaging of metabolites, peptides, and proteins in living cells. Additionally, by employing single guide RNA (sgRNA) as the recognition RNA motif, SSPepper-Apt generates fluorescence in both CRISPR-mediated imaging and gene editing only when the Cas9-sgRNA complex is successfully assembled; therefore, it can be an effective sgRNA screening tool for gene editing. Our fluorogenic RNA-sensing tag provides a universal approach for constructing functional RNA systems, avoiding the laborious and time-consuming process of sequence combination, and expanding the application of synthetic biological tools.

A guide to RNA structure analysis and RNA-targeting methods

RNAs are increasingly recognized as promising therapeutic targets, susceptible to modulation by strategies that include targeting with small molecules, antisense oligonucleotides, deoxyribozymes (DNAzymes), or CRISPR/Cas13. However, while drug development for proteins follows well-established paths for rational design based on the accurate knowledge of their three-dimensional structure, RNA-targeting strategies are challenging since comprehensive RNA structures are yet scarce and challenging to acquire. Numerous methods have been developed to elucidate the secondary and three-dimensional structure of RNAs, including X-ray crystallography, cryo-electron microscopy, nuclear magnetic resonance, SHAPE, DMS, and bioinformatic methods, yet they have often revealed flexible transcripts and co-existing populations rather than single-defined structures. Thus, researchers aiming to target RNAs face a critical decision: whether to acquire the detailed structure of transcripts in advance or to adopt phenotypic screens or sequence-based approaches that are independent of the structure. Still, even in strategies that seem to rely only on the nucleotide sequence (like the design of antisense oligonucleotides), researchers may need information about the accessibility of the compounds to the folded RNA molecule. In this concise guide, we provide an overview for researchers interested in targeting RNAs: We start by revisiting current methodologies for defining secondary or three-dimensional RNA structure and then we explore RNA-targeting strategies that may or may not require an in-depth knowledge of RNA structure. We envision that complementary approaches may expedite the development of RNA-targeting molecules to combat disease.

Development of an In Silico Platform (TRIPinRNA) for the Identification of Novel RNA Intramolecular Triple Helices and Their Validation Using Biophysical Techniques

There are surprisingly few RNA intramolecular triple helices known in the human transcriptome. The structure has been most well-studied as a stability-element at the 3' end of lncRNAs such as MALAT1 and NEAT1, but the intrigue remains whether it is indeed as rare as it is understood to be or just waiting for a closer look from a new vantage point. TRIPinRNA, our Python-based in silico platform, allows for a comprehensive sequence-pattern search for potential triplex formation in the human transcriptome─noncoding as well as coding. Using this tool, we report the putative occurrence of homopyrimidine type (canonical) triple helices as well as heteropurine-pyrimidine strand type (noncanonical) triple helices in the human transcriptome and validate the formation of both types of triplexes using biophysical approaches. We find that the occurrence of triplex structures has a strong correlation with local GC content, which might be influencing their formation. By employing a search that encompasses both canonical and noncanonical triplex structures across the human transcriptome, this study enriches the understanding of RNA biology. Lastly, TRIPinRNA can be utilized in finding triplex structures for any organism with an annotated transcriptome.

Technologies for Targeted RNA Degradation and Induced RNA Decay

The vast majority of the human genome codes for RNA, but RNA-targeting therapeutics account for a small fraction of approved drugs. As such, there is great incentive to improve old and develop new approaches to RNA targeting. For many RNA targeting modalities, just binding is not sufficient to exert a therapeutic effect; thus, targeted RNA degradation and induced decay emerged as powerful approaches with a pronounced biological effect. This review covers the origins and advanced use cases of targeted RNA degrader technologies grouped by the nature of the targeting modality as well as by the mode of degradation. It covers both well-established methods and clinically successful platforms such as RNA interference, as well as emerging approaches such as recruitment of RNA quality control machinery, CRISPR, and direct targeted RNA degradation. We also share our thoughts on the biggest hurdles in this field, as well as possible ways to overcome them.

Perspectives on Applications of 19F-NMR in Fragment-Based Drug Discovery

Fragment-based drug discovery is a powerful approach in drug discovery, applicable to a wide range of targets. This method enables the discovery of potent compounds that can modulate target functions, starting from fragment compounds that bind weakly to the targets. While biochemical, biophysical, and cell-based assays are commonly used to identify fragments, 19F-NMR spectroscopy has emerged as a powerful tool for exploring interactions between biomolecules and ligands. Because fluorine atoms are not naturally present in biological systems, 19F-NMR serves as a sensitive method for fragment screening against diverse targets. Herein, we reviewed the applications of 19F-NMR in fragment screening, highlighting its effectiveness in identifying fragments that bind weakly to various targets such as proteins and RNA. The accumulated evidence suggests that 19F-NMR will continue to be a crucial tool in drug discovery.

Harnessing Computational Approaches for RNA-Targeted Drug Discovery

RNA molecules have emerged as promising therapeutic targets due to their diverse functional and regulatory roles within cells. Computational modeling in RNA-targeted drug discovery presents a significant opportunity to expedite the discovery of novel small molecule compounds. However, this field encounters unique challenges compared to protein-targeted drug design, primarily due to limited experimental data availability and current models' inability to adequately address RNA's conformational flexibility during ligand recognition. Despite these challenges, several studies have successfully identified active RNA-targeting compounds using structure-based approaches or quantitative structure-activity relationship (QSAR) models. This review offers an overview of recent advancements in modeling RNA-small molecule interactions, emphasizing practical applications of computational methods in RNA-targeted drug discovery. Additionally, we survey existing databases that catalog nucleic acid-small molecule interactions. As interest in RNA-small molecule interactions grows and curated databases expand, the field anticipates rapid development. Novel computational models are poised to enhance the identification of potent and selective small-molecule modulators for therapeutic needs.

Accurate RNA 3D structure prediction using a language model-based deep learning approach

Accurate prediction of RNA three-dimensional (3D) structures remains an unsolved challenge. Determining RNA 3D structures is crucial for understanding their functions and informing RNA-targeting drug development and synthetic biology design. The structural flexibility of RNA, which leads to the scarcity of experimentally determined data, complicates computational prediction efforts. Here we present RhoFold+, an RNA language model-based deep learning method that accurately predicts 3D structures of single-chain RNAs from sequences. By integrating an RNA language model pretrained on ~23.7 million RNA sequences and leveraging techniques to address data scarcity, RhoFold+ offers a fully automated end-to-end pipeline for RNA 3D structure prediction. Retrospective evaluations on RNA-Puzzles and CASP15 natural RNA targets demonstrate the superiority of RhoFold+ over existing methods, including human expert groups. Its efficacy and generalizability are further validated through cross-family and cross-type assessments, as well as time-censored benchmarks. Additionally, RhoFold+ predicts RNA secondary structures and interhelical angles, providing empirically verifiable features that broaden its applicability to RNA structure and function studies.

Structural comparison of homologous protein-RNA interfaces reveals widespread overall conservation contrasted with versatility in polar contacts

Protein-RNA interactions play a critical role in many cellular processes and pathologies. However, experimental determination of protein-RNA structures is still challenging, therefore computational tools are needed for the prediction of protein-RNA interfaces. Although evolutionary pressures can be exploited for structural prediction of protein-protein interfaces, and recent deep learning methods using protein multiple sequence alignments have radically improved the performance of protein-protein interface structural prediction, protein-RNA structural prediction is lagging behind, due to the scarcity of structural data and the flexibility involved in these complexes. To study the evolution of protein-RNA interface structures, we first identified a large and diverse dataset of 2,022 pairs of structurally homologous interfaces (termed structural interologs). We leveraged this unique dataset to analyze the conservation of interface contacts among structural interologs based on the properties of involved amino acids and nucleotides. We uncovered that 73% of distance-based contacts and 68% of apolar contacts are conserved on average, and the strong conservation of these contacts occurs even in distant homologs with sequence identity below 20%. Distance-based contacts are also much more conserved compared to what we had found in a previous study of homologous protein-protein interfaces. In contrast, hydrogen bonds, salt bridges, and π-stacking interactions are very versatile in pairs of protein-RNA interologs, even for close homologs with high interface sequence identity. We found that almost half of the non-conserved distance-based contacts are linked to a small proportion of interface residues that no longer make interface contacts in the interolog, a phenomenon we term "interface switching out". We also examined possible recovery mechanisms for non-conserved hydrogen bonds and salt bridges, uncovering diverse scenarios of switching out, change in amino acid chemical nature, intermolecular and intramolecular compensations. Our findings provide insights for integrating evolutionary signals into predictive protein-RNA structural modeling methods.

New insights into the function and therapeutic potential of RNA-binding protein TRBP in viral infection, chronic metabolic diseases, brain disorders and cancer

RNA-binding proteins (RBPs) and non-coding RNAs are crucial trans-acting factors that bind to specific cis-acting elements in mRNAs, thereby regulating their stability and translation. The trans-activation response (TAR) RNA-binding protein (TRBP) recognizes precursor microRNAs (pre-miRNAs), modulates miRNA maturation, and influences miRNA interference (mi-RNAi) mediated by the RNA-induced silencing complex (RISC). TRBP also directly binds and mediates the degradation of certain mRNAs. Thus, TRBP acts as a hub for regulating gene expression and influences a variety of biological processes, including immune evasion, metabolic abnormalities, stress response, angiogenesis, hypoxia, and metastasis. Aberrant TRBP expression has been proven to be closely related to the initiation and progression of diseases, such as viral infection, chronic metabolic diseases, brain disorders, and cancer. This review summarizes the roles of TRBP in cancer and other diseases, the therapeutic potential of TRBP inhibition, and the current status of drug discovery on TRBP.

DeepRSMA: a cross-fusion-based deep learning method for RNA-small molecule binding affinity prediction

MOTIVATION

RNA is implicated in numerous aberrant cellular functions and disease progressions, highlighting the crucial importance of RNA-targeted drugs. To accelerate the discovery of such drugs, it is essential to develop an effective computational method for predicting RNA-small molecule affinity (RSMA). Recently, deep learning-based computational methods have been promising due to their powerful nonlinear modeling ability. However, the leveraging of advanced deep learning methods to mine the diverse information of RNAs, small molecules, and their interaction still remains a great challenge.

RESULTS

In this study, we present DeepRSMA, an innovative cross-attention-based deep learning method for RSMA prediction. To effectively capture fine-grained features from RNA and small molecules, we developed nucleotide-level and atomic-level feature extraction modules for RNA and small molecules, respectively. Additionally, we incorporated both sequence and graph views into these modules to capture features from multiple perspectives. Moreover, a transformer-based cross-fusion module is introduced to learn the general patterns of interactions between RNAs and small molecules. To achieve effective RSMA prediction, we integrated the RNA and small molecule representations from the feature extraction and cross-fusion modules. Our results show that DeepRSMA outperforms baseline methods in multiple test settings. The interpretability analysis and the case study on spinal muscular atrophy demonstrate that DeepRSMA has the potential to guide RNA-targeted drug design.

AVAILABILITY AND IMPLEMENTATION

The codes and data are publicly available at https://github.com/Hhhzj-7/DeepRSMA.

Cyclic peptides targeting the SARS-CoV-2 programmed ribosomal frameshifting RNA from a multiplexed phage display library

RNA provides the genetic blueprint for many pathogenic viruses, including SARS-CoV-2. The propensity of RNA to fold into specific tertiary structures enables the biomolecular recognition of cavities and crevices suited for the binding of drug-like molecules. Despite increasing interest in RNA as a target for chemical biology and therapeutic applications, the development of molecules that recognize RNA with high affinity and specificity represents a significant challenge. Here, we report a strategy for the discovery and selection of RNA-targeted macrocyclic peptides derived from combinatorial libraries of peptide macrocycles displayed by bacteriophages. Specifically, a platform for phage display of macrocyclic organo-peptide hybrids (MOrPH-PhD) was combined with a diverse set of non-canonical amino acid-based cyclization modules to produce large libraries of 107 structurally diverse, genetically encoded peptide macrocycles. These libraries were panned against the -1 programmed ribosomal frameshifting stimulatory sequence (FSS) RNA pseudoknot of SARS-CoV-2, which revealed specific macrocyclic peptide sequences that bind this essential motif with high affinity and selectivity. Peptide binding localizes to the FSS dimerization loop based on chemical modification analysis and binding assays and the cyclic peptides show specificity toward the target RNA over unrelated RNA pseudoknots. This work introduces a novel system for the generation and high-throughput screening of topologically diverse cyclopeptide scaffolds (multiplexed MOrPH-PhD), and it provides a blueprint for the exploration and evolution of genetically encoded macrocyclic peptides that target specific RNAs.

Small Molecule RNA Degraders

RNA is a central molecule in life, involved in a plethora of biological processes and playing a key role in many diseases. Targeting RNA emerges as a significant endeavor in drug discovery, diverging from conventional protein-centric approaches to tackle various pathologies. Whilst identifying small molecules that bind to specific RNA regions is the first step, the abundance of non-functional RNA segments renders many interactions biologically inert. Consequently, small molecule binding does not necessarily meet stringent criteria for clinical translation, calling for solutions to push the field forward. Converting RNA-binders into RNA-degraders presents a promising avenue to enhance RNA-targeting. This mini-review outlines strategies and exemplars wherein simple small molecule RNA binders are reprogrammed into active degraders through the linkage of functional groups. These approaches encompass mechanisms that induce degradation via endogenous enzymes, termed RIBOTACs, as well as those with functional moieties acting autonomously to degrade RNA. Through this exploration, we aim to offer insights into advancing RNA-targeted therapeutic strategies.