RNA Binding Affinity Research Articles

IGF2BP1 phosphorylation in the disordered linkers regulates ribonucleoprotein condensate formation and RNA metabolism.

The insulin-like growth factor 2 mRNA binding protein 1 (IGF2BP1) is a conserved RNA-binding protein that regulates RNA stability, localization and translation. IGF2BP1 is part of various ribonucleoprotein (RNP) condensates. However, the mechanism that regulates its assembly into condensates remains unknown. By using proteomics, we demonstrate that phosphorylation of IGF2BP1 at S181 in a disordered linker is regulated in a stress-dependent manner. Phosphomimetic mutations in two disordered linkers, S181E and Y396E, modulate RNP condensate formation by IGF2BP1 without impacting its binding affinity for RNA. Intriguingly, the S181E mutant, which lies in linker 1, impairs IGF2BP1 condensate formation in vitro and in cells, whereas a Y396E mutant in the second linker increases condensate size and dynamics. Structural approaches show that the first linker binds RNAs nonspecifically through its RGG/RG motif, an interaction weakened in the S181E mutant. Notably, linker 2 interacts with IGF2BP1's folded domains and these interactions are partially impaired in the Y396E mutant. Importantly, the phosphomimetic mutants impact IGF2BP1's interaction with RNAs and remodel the transcriptome in cells. Our data reveal how phosphorylation modulates low-affinity interaction networks in disordered linkers to regulate RNP condensate formation and RNA metabolism.

Recurrent Neurodevelopmentally Associated Variants of the Pre-mRNA Splicing Factor U2AF2 Alter RNA Binding Affinities and Interactions.

De novo mutations affecting the pre-mRNA splicing factor U2AF2 are associated with developmental delays and intellectual disabilities, yet the molecular basis is unknown. Here, we demonstrated by fluorescence anisotropy RNA binding assays that recurrent missense mutants (Arg149Trp, Arg150His, or Arg150Cys) decreased the binding affinity of U2AF2 for a consensus splice site RNA. Crystal structures at 1.4 Å resolutions showed that Arg149Trp or Arg150His disrupted hydrogen bonds between U2AF2 and the terminal nucleotides of the RNA site. Reanalysis of publicly available RNaseq data confirmed that U2AF2 depletion altered splicing of transcripts encoding RNA binding proteins (RBPs). These results confirmed that the impaired RNA interactions of Arg149Trp and Arg150His U2AF2 variants could contribute to dysregulating an RBP-governed neurodevelopmental program of alternative splicing.

Sequencing CURLY LEAF-associated RNAs in Arabidopsis revealed prevalent intergenic RNAs from the nuclear mitochondrial sequence

Polycomb Group (PcG) proteins play key roles in development by repressing thousands of targets through histone modifications. However, how PcG is recruited to specific targets is poorly understood. In Arabidopsis, certain noncoding RNAs are necessary for recruiting the PcG protein CURLY LEAF (CLF) to its target sites. However, RNAs associated with CLF have not been analyzed on a genomic scale, thus it is unknown whether long noncoding RNA (lncRNA)-mediated PcG recruitment is a widespread mechanism. Here, we systematically searched for CLF-associated RNAs by RNA immunoprecipitation followed by deep sequencing. We identified 1299 genic and 138 intergenic regions that produced CLF-associated mRNAs and putative lncRNAs, respectively. The genes producing CLF-associated RNAs are depleted in PcG targets, carry active chromatin marks, and are highly expressed, suggesting that CLF may have a non-specific or promiscuous RNA-binding affinity, similar to animal PcG proteins. Notably, a significant portion of the CLF-associated lncRNAs is derived from the nuclear mitochondrial sequence, which is extensively marked by H3K27me3. These findings indicate that while CLF-RNA interactions are widespread, they may not always correlate with PcG target sites, highlighting the complexity of PcG recruitment mechanisms in Arabidopsis.

Towards combining backbone and sugar constraint in 3'-3' bis-phosphonate tethered 2'-4' bridged LNA oligonucleotide trimers.

Therapeutic oligonucleotides are chemically modified to enhance their drug-like properties - including binding affinity for target RNA. Many nucleic acid analogs that enhance RNA binding affinity constrain the furanose sugar in an RNA-like sugar pucker. The improvements in binding affinity result primarily from increased off-rates with minimal effects on on-rates for hybridization. To identify alternate chemical modification strategies that can modulate on- and off-rates for oligonucleotide hybridization, we hypothesized that extending conformational restraint across multiple nucleotides could modulate hybridization kinetics by restricting rotational freedom of the sugar-phosphate backbone. As part of that effort, we recently reported that using hydrocarbon tethers to bridge adjacent phosphodiester linkages as phosphonate tethered bridges can pre-organize nucleic acids in conformations conducive for Watson-Crick base-pairing and modulate hybridization kinetics. In this report, we describe the synthesis of locked nucleic acid (LNA) trimers linked through alkylphosphonate tethers which restrict conformation of the furanose sugar in addition to restricting conformational mobility of the sugar-phosphate backbone across three nucleotide units.

Mitochondrial sequencing identifies long noncoding RNA features that promote binding to PNPase.

Extranuclear localization of long noncoding RNAs (lncRNAs) is poorly understood. Based on machine learning evaluations, we propose a lncRNA-mitochondrial interaction pathway where polynucleotide phosphorylase (PNPase), through domains that provide specificity for primary sequence and secondary structure, binds nuclear-encoded lncRNAs to facilitate mitochondrial import. Using FVB/NJ mouse and human cardiac tissues, RNA from isolated subcellular compartments (cytoplasmic and mitochondrial) and cross-linked immunoprecipitate (CLIP) with PNPase within the mitochondrion were sequenced on the Illumina HiSeq and MiSeq, respectively. lncRNA sequence and structure were evaluated through supervised [classification and regression trees (CART) and support vector machines (SVM)] machine learning algorithms. In HL-1 cells, quantitative PCR of PNPase CLIP knockout mutants (KH and S1) was performed. In vitro fluorescence assays assessed PNPase RNA binding capacity and verified with PNPase CLIP. One hundred twelve (mouse) and 1,548 (human) lncRNAs were identified in the mitochondrion with Malat1 being the most abundant. Most noncoding RNAs binding PNPase were lncRNAs, including Malat1. lncRNA fragments bound to PNPase compared against randomly generated sequences of similar length showed stratification with SVM and CART algorithms. The lncRNAs bound to PNPase were used to create a criterion for binding, with experimental validation revealing increased binding affinity of RNA designed to bind PNPase compared to control RNA. The binding of lncRNAs to PNPase was decreased through the knockout of RNA binding domains KH and S1. In conclusion, sequence and secondary structural features identified by machine learning enhance the likelihood of nuclear-encoded lncRNAs binding to PNPase and undergoing import into the mitochondrion.NEW & NOTEWORTHY Long noncoding RNAs (lncRNAs) are relatively novel RNAs with increasingly prominent roles in regulating genetic expression, mainly in the nucleus but more recently in regions such as the mitochondrion. This study explores how lncRNAs interact with polynucleotide phosphorylase (PNPase), a protein that regulates RNA import into the mitochondrion. Machine learning identified several RNA structural features that improved lncRNA binding to PNPase, which may be useful in targeting RNA therapeutics to the mitochondrion.

Natural evidence of coronaviral 2′-O-methyltransferase activity affecting viral pathogenesis via improved substrate RNA binding

Previous studies through targeted mutagenesis of K-D-K-E motif have demonstrated that 2′-O-MTase activity is essential for efficient viral replication and immune evasion. However, the K-D-K-E catalytic motif of 2′-O-MTase is highly conserved across numerous viruses, including flaviviruses, vaccinia viruses, coronaviruses, and extends even to mammals. Here, we observed a stronger 2′-O-MTase activity in SARS-CoV-2 compared to SARS-CoV, despite the presence of a consistently active catalytic center. We further identified critical residues (Leu-36, Asn-138 and Ile-153) which served as determinants of discrepancy in 2′-O-MTase activity between SARS-CoV-2 and SARS-CoV. These residues significantly enhanced the RNA binding affinity of 2′-O-MTase and boosted its versatility toward RNA substrates. Of interest, a triple substitution (Leu36 → Ile36, Asn138 → His138, Ile153 → Leu153, from SARS-CoV-2 to SARS-CoV) within nsp16 resulted in a proportional reduction in viral 2′-O-methylation and impaired viral replication. Furthermore, it led to a significant upregulation of type I interferon (IFN-I) and proinflammatory cytokines both in vitro and vivo, relying on the cooperative sensing of melanoma differentiation-associated protein 5 (MDA5) and laboratory of genetics and physiology 2 (LGP2). In conclusion, our findings demonstrated that alterations in residues other than K-D-K-E of 2′-O-MTase may affect viral replication and subsequently influence pathogenesis. Monitoring changes in nsp16 residues is crucial as it may aid in identifying and assessing future alteration in viral pathogenicity resulting from natural mutations occurring in nsp16.

Hfq-Antisense RNA I Binding Regulates RNase E-Dependent RNA Stability and ColE1 Plasmid Copy Number.

The mechanisms and consequences of gene regulation by Hfq on trans-encoded small RNAs (sRNAs) have been well studied and documented. Recent employment of Genomic SELEX to search for Hfq-binding motifs has indicated that Hfq might frequently regulate gene expression controlled by cis-antisense RNAs. Here, we use the classic ColE1 plasmid antisense RNA-based regulation model (i.e., RNA I) to study the role of Hfq in controlling antisense regulatory functions. We show that Hfq exhibits a high binding affinity for RNA I and that binding limits RNase E cleavage, thereby stabilizing RNA I and reducing the plasmid copy number. Full-length RNA I displays a binding affinity for Hfq in the sub-micromolar range. In vivo overexpression of Hfq prolongs RNA I stability and reduces the ColE1 plasmid copy number, whereas deletion of hfq reduces RNA I stability and increases the plasmid copy number. RNA I predominantly binds to the proximal face of Hfq and exhibits competitive ability against a chromosome-borne proximal face-bound sRNA (DsrA) for Hfq binding. Through its strong promoter and high gene dosage features, plasmid-encoded antisense RNA I results in high RNA I expression, so it may antagonize the effects of trans-encoded RNAs in controlling target gene expression.

Contrastive pre-training and 3D convolution neural network for RNA and small molecule binding affinity prediction.

The diverse structures and functions inherent in RNAs present a wealth of potential drug targets. Some small molecules are anticipated to serve as leading compounds, providing guidance for the development of novel RNA-targeted therapeutics. Consequently, the determination of RNA-small molecule binding affinity is a critical undertaking in the landscape of RNA-targeted drug discovery and development. Nevertheless, to date, only one computational method for RNA-small molecule binding affinity prediction has been proposed. The prediction of RNA-small molecule binding affinity remains a significant challenge. The development of a computational model is deemed essential to effectively extract relevant features and predict RNA-small molecule binding affinity accurately. In this study, we introduced RLaffinity, a novel deep learning model designed for the prediction of RNA-small molecule binding affinity based on 3D structures. RLaffinity integrated information from RNA pockets and small molecules, utilizing a 3D convolutional neural network (3D-CNN) coupled with a contrastive learning-based self-supervised pre-training model. To the best of our knowledge, RLaffinity was the first deep learning based method for the prediction of RNA-small molecule binding affinity. Our experimental results exhibited RLaffinity's superior performance compared to baseline methods, revealed by all metrics. The efficacy of RLaffinity underscores the capability of 3D-CNN to accurately extract both global pocket information and local neighbor nucleotide information within RNAs. Notably, the integration of a self-supervised pre-training model significantly enhanced predictive performance. Ultimately, RLaffinity was also proved as a potential tool for RNA-targeted drugs virtual screening. https://github.com/SaisaiSun/RLaffinity.

Distinct and overlapping RNA determinants for binding and target-primed reverse transcription by Bombyx mori R2 retrotransposon protein.

Eukaryotic retrotransposons encode a reverse transcriptase that binds RNA to template DNA synthesis. The ancestral non-long terminal repeat (non-LTR) retrotransposons encode a protein that performs target-primed reverse transcription (TPRT), in which the nicked genomic target site initiates complementary DNA (cDNA)synthesis directly into the genome. The best understood model system for biochemical studies of TPRT is the R2 protein from the silk moth Bombyx mori. The R2 protein selectively binds the 3' untranslated region of its encoding RNA as template for DNA insertion to its target site in 28S ribosomal DNA. Here, binding and TPRT assays define RNA contributions to RNA-protein interaction, template use for TPRT and the fidelity of template positioning for TPRT cDNA synthesis. We quantify both sequence and structure contributions to protein-RNA interaction. RNA determinants of binding affinity overlap but are not equivalent to RNA features required for TPRT and its fidelity of template positioning for full-length TPRT cDNA synthesis. Additionally, we show that a previously implicated RNA-binding protein surface of R2 protein makes RNA binding affinity dependent on the presence of two stem-loops. Our findings inform evolutionary relationships across R2 retrotransposon RNAs and are a step toward understanding the mechanism and template specificity of non-LTR retrotransposon mobility.

Understanding Antisense Oligonucleotide Efficiency in Inhibiting Prokaryotic Gene Expression.

Oligonucleotides offer a unique opportunity for sequence specific regulation of gene expression in bacteria. A fundamental question to address is the choice of oligonucleotide, given the large number of options available. Different modifications varying in RNA binding affinities and cellular uptake are available but no comprehensive comparisons have been performed. Herein, the efficiency of blocking expression of β-galactosidase (β-Gal) in E. coli was evaluated utilizing different antisense oligomers (ASOs). Fluorescein (FAM)-labeled oligomers were used to understand their differences in bacterial uptake. Flow cytometry analysis revealed significant differences in uptake, with high fluorescence seen in cells treated with FAM-labeled peptidic nucleic acid (PNA), phosphorodiamidate morpholino oligonucleotide (PMO) and phosphorothioate (PS) oligomers, and low fluorescence observed in cells treated with phosphodiester (PO) oligomers. Thermal denaturation (Tm) of oligomer:RNA duplexes and isothermal titration calorimetry (ITC) studies reveal that ASO binding to target RNA demonstrates a good correlation between Tm and Kd values. There was no correlation between Kd values and reduction of β-Gal activity in bacterial cells. However, cell-free translation assays demonstrated a direct relationship between Kd values and inhibition of gene expression by antisense oligomers, with tight binding oligomers such as LNA being the most efficient. Membrane active compounds such as polymyxin B and A22 further improved the cellular uptake of FAM-PNA and FAM-PS oligomers in wild-type E. coli cells. PNA and PMO were most effective in cellular uptake and reducing β-Gal activity as compared to oligomers with PS or those with PO linkages. Overall, cell uptake of the oligomers is shown as the key determinant in predicting their differences in bacterial antisense inhibition, and the RNA affinity is the key determinant in inhibition of gene expression in cell free systems.

Guanidine modification improves functions of natural RNA-targeting alkaloids

Guanidine modification on berberine improved RNA binding affinity and enhanced suppression of protein expression.

Highly sequence-specific, timing-controllable m6A demethylation by modulating RNA-binding affinity of m6A erasers.

Recent advancements in tools using programmable RNA binding proteins and m6A-erasers enable sequence-selective and timing-controllable m6A demethylation. However, off-target effects are still a concern. This study addresses the problem by reducing the RNA-binding ability of m6A-erasers. The modulated m6A-erasers achieved sequence-specific and timing-controllable m6A demethylation with minimal off-target activity.

Unraveling blunt-end RNA binding and ATPase-driven translocation activities of the RIG-I family helicase LGP2.

The RIG-I family helicases, comprising RIG-I, MDA5and LGP2, are cytoplasmic RNA sensors that trigger an antiviral immune response by specifically recognizing foreign RNAs. While LGP2 lacks the signaling domain necessary for immune activation, it plays a vital role in regulating the RIG-I/MDA5 signaling pathway. In this study, we investigate the mechanisms underlying this regulation by examining the oligomeric state, RNA binding specificity, and translocation activity of human LGP2 and the impact of ATPase activity. We show that LGP2, like RIG-I, prefers binding blunt-ended double-stranded (ds) RNAs over internal dsRNA regions or RNA overhangs and associates with blunt-ends faster than with overhangs. Unlike RIG-I, a 5'-triphosphate (5'ppp), Cap0, or Cap1 RNA-end does not influence LGP2's RNA binding affinity. LGP2 hydrolyzes ATP in the presence of RNA but at a 5-10 fold slower rate than RIG-I. Nevertheless, LGP2 uses its ATPase activity to translocate and displace biotin-streptavidin interactions. This activity is significantly hindered by a methylated RNA patch, particularly on the 3'-strand, suggesting a 3'-strand tracking mechanism like RIG-I. The preference of LGP2 for blunt-end RNA binding, its insensitivity to Cap0/Cap1 modification, and its translocation/protein displacement ability have substantial implications for how LGP2 regulates the RNA sensing process by MDA5/RIG-I.

Intra- and inter-molecular regulation by intrinsically-disordered regions governs PUF protein RNA binding

PUF proteins are characterized by globular RNA-binding domains. They also interact with partner proteins that modulate their RNA-binding activities. Caenorhabditis elegans PUF protein fem-3 binding factor-2 (FBF-2) partners with intrinsically disordered Lateral Signaling Target-1 (LST-1) to regulate target mRNAs in germline stem cells. Here, we report that an intrinsically disordered region (IDR) at the C-terminus of FBF-2 autoinhibits its RNA-binding affinity by increasing the off rate for RNA binding. Moreover, the FBF-2 C-terminal region interacts with its globular RNA-binding domain at the same site where LST-1 binds. This intramolecular interaction restrains an electronegative cluster of amino acid residues near the 5′ end of the bound RNA to inhibit RNA binding. LST-1 binding in place of the FBF-2 C-terminus therefore releases autoinhibition and increases RNA-binding affinity. This regulatory mechanism, driven by IDRs, provides a biochemical and biophysical explanation for the interdependence of FBF-2 and LST-1 in germline stem cell self-renewal.

Gadd45β is critical for regulation of type I interferon signaling by facilitating G3BP-mediated stress granule formation

Stress granules (SGs) constitute a signaling hub that plays a critical role in type I interferon responses. Here, we report that growth arrest and DNA damage-inducible beta (Gadd45β) act as a positive regulator of SG-mediated interferon signaling by targeting G3BP upon RNA virus infection. Gadd45β deficiency markedly impairs SG formation and SG-mediated activation of interferon signaling invitro. Gadd45β knockout mice are highly susceptible to RNA virus infection, and their ability to produce interferon and cytokines is severely impaired. Specifically, Gadd45β interacts with the RNA-binding domain of G3BP, leading to conformational expansion of G3BP1 via dissolution of its autoinhibitory electrostatic intramolecular interaction. The acidic loop 1- and RNA-binding properties of Gadd45β markedly increase the conformational expansion and RNA-binding affinity of the G3BP1-Gadd45β complex, thereby promoting assembly of SGs. These findings suggest a role for Gadd45β as a component and critical regulator of G3BP1-mediated SG formation, which facilitates RLR-mediated interferon signaling.

RNA binding by the glucocorticoid receptor attenuates dexamethasone-induced gene activation

The glucocorticoid receptor (GR) is a ligand-activated transcription factor that regulates a suite of genes through direct binding of GR to specific DNA promoter elements. GR also interacts with RNA, but the function of this RNA-binding activity remains elusive. Current models speculate that RNA could repress the transcriptional activity of GR. To investigate the function of the GR-RNA interaction on GR’s transcriptional activity, we generated cells that stably express a mutant of GR with reduced RNA binding affinity and treated the cells with the GR agonist dexamethasone. Changes in the dexamethasone-driven transcriptome were quantified using 4-thiouridine labeling of RNAs followed by high-throughput sequencing. We find that while many genes are unaffected, GR-RNA binding is repressive for specific subsets of genes in both dexamethasone-dependent and independent contexts. Genes that are dexamethasone-dependent are activated directly by chromatin-bound GR, suggesting a competition-based repression mechanism in which increasing local concentrations of RNA may compete with DNA for binding to GR at sites of transcription. Unexpectedly, genes that are dexamethasone-independent instead display a localization to specific chromosomal regions, which points to changes in chromatin accessibility or architecture. These results show that RNA binding plays a fundamental role in regulating GR function and highlights potential functions for transcription factor-RNA interactions.

Electronegative clusters modulate folding status and RNA binding of unstructured RNA-binding proteins.

Electronegative clusters (ENCs) made up of acidic residues and/or phosphorylation sites are the most abundant repetitive sequences in RNA-binding proteins. Previous studies have indicated that ENCs inhibit RNA binding for structured RNA-binding domains (RBDs). However, this is not the case for the unstructured RBD in histone pre-mRNA stem-loop binding protein (SLBP). The SLBP RBD contains 70 amino acids and is followed by a phosphorylatable ENC. ENC phosphorylation increases RNA-binding affinity of SLBP to the sub-picomolar range. In this study, we use NMR and molecular dynamics simulations to elucidate the mechanism for this tight binding. Our NMR data demonstrate that the ENC transiently folds apo SLBP into an RNA-bound resembling state. We find that in the RNA-bound state, the phosphorylated ENC interacts with the loop region opposite to the RNA-binding site. This allosteric interaction stabilizes the complex and therefore enhances RNA binding. To evaluate the generality of our findings, we graft an ENC onto endoribonuclease homolog 1's first double-stranded RNA-binding motif (DRBM1), an unstructured RBD that shares no homology with SLBP. We find that the engineered ENC increases the folded species of DRBM1 and inhibits RNA binding. On the contrary, introducing basic residues to DRBM1 makes the domain more unfolded, enhances RNA binding, and mitigates the inhibitory effect of the engineered ENC. In summary, our study suggests that ENCs promote folding of unstructured RNA-binding domains, and their effects on RNA binding depend on the electropositive charges on the RBD surface.

Systematic Investigation of Tether Length and Phosphorus Configuration in Backbone Constrained Macrocyclic Nucleic Acids to Modulate Binding Kinetics for RNA.

We recently described a chemical strategy to pre-organize a trinucleotide subunit in a conformation suitable for Watson-Crick base pairing for modulating the binding kinetics of single-stranded oligonucleotides (ONs) using bis-phosphonate esters bridging hydrocarbon tethers to provide 11- and 15-membered macrocyclic analogues. In this manuscript, we describe the synthesis of all eight P-stereoisomers of macrocyclic 12-, 13-, 14-, and 16-membered hydrocarbon-bridged nucleotide trimers, their incorporation into ONs, and biophysical characterization of the modified ONs. The size of the macrocyclic tether and configuration at phosphorus had profound effects on hybridization kinetics. ONs containing 12- and 13-membered rings exhibited faster on-rates (up to 5-fold) and off-rates (up to 161-fold). In contrast, ONs using the larger ring size macrocycles generally exhibited smaller changes in binding kinetics relative to unmodified DNA. Interestingly, several of the analogues retained significant binding affinity for RNA based on their dissociation constants, despite being modestly destabilizing in the thermal denaturation experiments, highlighting the potential utility of measuring dissociation constants versus duplex thermal stability when evaluating novel nucleic acid analogues. Overall, our results provide additional insights into the ability of backbone-constrained macrocyclic nucleic acid analogues to modulate hybridization kinetics of modified ONs with RNA.

Expression and functional characterization of odorant-binding protein2 in the predatory mite Neoseiulus barkeri.

Olfactionplays a crucial role for arthropods in foraging, mating, and oviposition. The odorant-binding protein (OBP) gene is considered one of the most important olfactory genes. However, little is known about its functions in predatory mites. Here, we used Neoseiulus barkeri, an important commercialized natural pest control, to explore the chemosensory characteristics of OBP. In this study, N.barkeri was attracted by methyl salicylate (MeSA) and showed higher crawling speeds under MeSA treatment. Then, we identified and cloned an OBP gene named Nbarobp2 and analyzed its expression profiles in the predatory mite. Nbarobp2 was 663bp, was highly expressed in larval and nymphal stages, and was significantly upregulated in N.barkeri under MeSA treatment. Nbarobp2 encoded 202 amino acid residues with a molecular weight of 23kDa (after removing the signal peptide). Sequence comparisons revealed that the OBPs in Arachnida shared 6 conserved cysteine sites, but were distinguishable from the OBPs of Insecta on the phylogenetic tree. RNA interference, Western blotting, and binding affinity assays further proved that Nbarobp2 was involved in volatile perception in predatory mites. This study shed light on the functional characteristics of OBPs in predatory mites, providing a new insight for better biological control.

A structural dynamics model for how CPEB3 binding to SUMO2 can regulate translational control in dendritic spines.

A prion-like RNA-binding protein, CPEB3, can regulate local translation in dendritic spines. CPEB3 monomers repress translation, whereas CPEB3 aggregates activate translation of its target mRNAs. However, the CPEB3 aggregates, as long-lasting prions, may raise the problem of unregulated translational activation. Here, we propose a computational model of the complex structure between CPEB3 RNA-binding domain (CPEB3-RBD) and small ubiquitin-like modifier protein 2 (SUMO2). Free energy calculations suggest that the allosteric effect of CPEB3-RBD/SUMO2 interaction can amplify the RNA-binding affinity of CPEB3. Combining with previous experimental observations on the SUMOylation mode of CPEB3, this model suggests an equilibrium shift of mRNA from binding to deSUMOylated CPEB3 aggregates to binding to SUMOylated CPEB3 monomers in basal synapses. This work shows how a burst of local translation in synapses can be silenced following a stimulation pulse, and explores the CPEB3/SUMO2 interplay underlying the structural change of synapses and the formation of long-term memories.