Integrative Approach to Probe Alternative Redox Mechanisms in RNA Modifications.

ConspectusRNA modifications found in most RNAs, particularly in tRNAs and rRNAs, reveal an abundance of chemical alterations of nucleotides. Over 150 distinct RNA modifications are known, emphasizing a remarkable diversity of chemical moieties in RNA molecules. These modifications play pivotal roles in RNA maturation, structural integrity, and the fidelity and efficiency of translation processes. The catalysts responsible for these modifications are RNA-modifying enzymes that use a striking array of chemistries to directly influence the chemical landscape of RNA. This diversity is further underscored by instances where the same modification is introduced by distinct enzymes that use unique catalytic mechanisms and cofactors across different domains of life. This phenomenon of convergent evolution highlights the biological importance of RNA modification and the vast potential within the chemical repertoire for nucleotide alteration. While shared RNA modifications can hint at conserved enzymatic pathways, a major bottleneck is to identify alternative routes within species that possess a modified RNA but are devoid of known RNA-modifying enzymes. To address this challenge, a combination of bioinformatic and experimental strategies proves invaluable in pinpointing new genes responsible for RNA modifications. This integrative approach not only unveils new chemical insights but also serves as a wellspring of inspiration for biocatalytic applications and drug design. In this Account, we present how comparative genomics and genome mining, combined with biomimetic synthetic chemistry, biochemistry, and anaerobic crystallography, can be judiciously implemented to address unprecedented and alternative chemical mechanisms in the world of RNA modification. We illustrate these integrative methodologies through the study of tRNA and rRNA modifications, dihydrouridine, 5-methyluridine, queuosine, 8-methyladenosine, 5-carboxymethylamino-methyluridine, or 5-taurinomethyluridine, each dependent on a diverse array of redox chemistries, often involving organic compounds, organometallic complexes, and metal coenzymes. We explore how vast genome and tRNA databases empower comparative genomic analyses and enable the identification of novel genes that govern RNA modification. Subsequently, we describe how the isolation of a stable reaction intermediate can guide the synthesis of a biomimetic to unveil new enzymatic pathways. We then discuss the usefulness of a biochemical "shunt" strategy to study catalytic mechanisms and to directly visualize reactive intermediates bound within active sites. While we primarily focus on various RNA-modifying enzymes studied in our laboratory, with a particular emphasis on the discovery of a SAM-independent methylation mechanism, the strategies and rationale presented herein are broadly applicable for the identification of new enzymes and the elucidation of their intricate chemistries. This Account offers a comprehensive glimpse into the evolving landscape of RNA modification research and highlights the pivotal role of integrated approaches to identify novel enzymatic pathways.

S9.6 Antibody-Enzyme Conjugates for the Detection of DNA-RNA Hybrids.

Diagnosis of infectious agents is increasingly done by the detection of unique nucleic acid sequences, typically using methods such as PCR that specifically amplify these sequences. A largely neglected alternative approach is to use antibodies that recognize nucleic acids. The unique monoclonal antibody S9.6 recognizes DNA-RNA hybrids in a largely sequence-independent manner. S9.6 has been used in several cases for the analysis of nucleic acids. Extending our recent determination of the structure of S9.6 Fab bound to a DNA-RNA hybrid, we have developed reagents and methods for the sensitive detection of specific DNA and RNA sequences. To facilitate the use in diagnostics, we conjugated the S9.6 Fab to the highly active and well-characterized reporter enzyme human-secreted embryonic alkaline phosphatase (SEAP). Two approaches were utilized for conjugation. The first used sortase A (SrtA), which generates a covalent peptide bond between short amino acid sequences added to recombinantly produced S9.6 Fab and SEAP. The second approach was to genetically fuse the S9.6 Fab and SEAP so that the two are produced as a single molecule. Using these two antibody-SEAP proteins, we developed a simplified ELISA format for the identification of synthetic DNA-RNA hybrids, which can be optimized for detecting nucleic acids of pathogens, as well as for other applications. We successfully used this immunosorbent assay, HC-S, to identify DNA-RNA hybrids in solution with high specificity and sensitivity.

The search for a PKR code-differential regulation of protein kinase R activity by diverse RNA and protein regulators.

The interferon-inducible protein kinase R (PKR) is a key component of host innate immunity that restricts viral replication and propagation. As one of the four eIF2α kinases that sense diverse stresses and direct the integrated stress response (ISR) crucial for cell survival and proliferation, PKR's versatile roles extend well beyond antiviral defense. Targeted by numerous host and viral regulators made of RNA and proteins, PKR is subject to multiple layers of endogenous control and external manipulation, driving its rapid evolution. These versatile regulators include not only the canonical double-stranded RNA (dsRNA) that activates the kinase activity of PKR, but also highly structured viral, host, and artificial RNAs that exert a full spectrum of effects. In this review, we discuss our deepening understanding of the allosteric mechanism that connects the regulatory and effector domains of PKR, with an emphasis on diverse structured RNA regulators in comparison to their protein counterparts. Through this analysis, we conclude that much of the mechanistic details that underlie this RNA-regulated kinase await structural and functional elucidation, upon which we can then describe a "PKR code," a set of structural and chemical features of RNA that are both descriptive and predictive for their effects on PKR.

Molecular basis for transfer RNA recognition by the double-stranded RNA-binding domain of human dihydrouridine synthase 2.

Double stranded RNA-binding domain (dsRBD) is a ubiquitous domain specialized in the recognition of double-stranded RNAs (dsRNAs). Present in many proteins and enzymes involved in various functional roles of RNA metabolism, including RNA splicing, editing, and transport, dsRBD generally binds to RNAs that lack complex structures. However, this belief has recently been challenged by the discovery of a dsRBD serving as a major tRNA binding module for human dihydrouridine synthase 2 (hDus2), a flavoenzyme that catalyzes synthesis of dihydrouridine within the complex elbow structure of tRNA. We here unveil the molecular mechanism by which hDus2 dsRBD recognizes a tRNA ligand. By solving the crystal structure of this dsRBD in complex with a dsRNA together with extensive characterizations of its interaction with tRNA using mutagenesis, NMR and SAXS, we establish that while hDus2 dsRBD retains a conventional dsRNA recognition capability, the presence of an N-terminal extension appended to the canonical domain provides additional residues for binding tRNA in a structure-specific mode of action. Our results support that this extension represents a feature by which the dsRBD specializes in tRNA biology and more broadly highlight the importance of structural appendages to canonical domains in promoting the emergence of functional diversity.

Unveiling structural and functional divergences of bacterial tRNA dihydrouridine synthases: perspectives on the evolution scenario.

Post-transcriptional base modifications are important to the maturation process of transfer RNAs (tRNAs). Certain modifications are abundant and present at several positions in tRNA as for example the dihydrouridine, a modified base found in the three domains of life. Even though the function of dihydrourine is not well understood, its high content in tRNAs from psychrophilic bacteria or cancer cells obviously emphasizes a central role in cell adaptation. The reduction of uridine to dihydrouridine is catalyzed by a large family of flavoenzymes named dihydrouridine synthases (Dus). Prokaryotes have three Dus (A, B and C) wherein DusB is considered as an ancestral protein from which the two others derived via gene duplications. Here, we unequivocally established the complete substrate specificities of the three Escherichia coli Dus and solved the crystal structure of DusB, enabling for the first time an exhaustive structural comparison between these bacterial flavoenzymes. Based on our results, we propose an evolutionary scenario explaining how substrate specificities has been diversified from a single structural fold.

Structural Insights into RNA Dimerization: Motifs, Interfaces and Functions.

In comparison with the pervasive use of protein dimers and multimers in all domains of life, functional RNA oligomers have so far rarely been observed in nature. Their diminished occurrence contrasts starkly with the robust intrinsic potential of RNA to multimerize through long-range base-pairing ("kissing") interactions, self-annealing of palindromic or complementary sequences, and stable tertiary contact motifs, such as the GNRA tetraloop-receptors. To explore the general mechanics of RNA dimerization, we performed a meta-analysis of a collection of exemplary RNA homodimer structures consisting of viral genomic elements, ribozymes, riboswitches, etc., encompassing both functional and fortuitous dimers. Globally, we found that domain-swapped dimers and antiparallel, head-to-tail arrangements are predominant architectural themes. Locally, we observed that the same structural motifs, interfaces and forces that enable tertiary RNA folding also drive their higher-order assemblies. These feature prominently long-range kissing loops, pseudoknots, reciprocal base intercalations and A-minor interactions. We postulate that the scarcity of functional RNA multimers and limited diversity in multimerization motifs may reflect evolutionary constraints imposed by host antiviral immune surveillance and stress sensing. A deepening mechanistic understanding of RNA multimerization is expected to facilitate investigations into RNA and RNP assemblies, condensates, and granules and enable their potential therapeutical targeting.

Conformational Stability Adaptation of a Double-Stranded RNA-Binding Domain to Transfer RNA Ligand.

The double-stranded RNA-binding domain (dsRBD) is a broadly distributed domain among RNA-maturing enzymes. Although this domain recognizes dsRNA's structures via a conserved canonical structure adopting an α1-ß1ß2ß3-α2 topology, several dsRBDs can accommodate discrete structural extensions expanding further their functional repertoire. How these structural elements engage cooperative communications with the canonical structure and how they contribute to the dsRBD's overall folding are poorly understood. Here, we addressed these issues using the dsRBD of human dihydrouridine synthase-2 (hDus2) (hDus2-dsRBD) as a model. This dsRBD harbors N- and C-terminal extensions, the former being directly involved in the recognition of tRNA substrate of hDus2. These extensions engage residues that form a long-range hydrophobic network (LHN) outside the RNA-binding interface. We show by coarse-grain Brownian dynamics that the Nt-extension and its residues F359 and Y364 rigidify the major folding nucleus of the canonical structure via an indirect effect. hDus2-dsRBD unfolds following a two-state cooperative model, whereas both F359A and Y364A mutants, designed to destabilize this LHN, unfold irreversibly. Structural and computational analyses show that these mutants are unstable due to an increase in the dynamics of the two extensions favoring solvent exposure of α2-helix and weakening the main folding nucleus rigidity. This LHN appears essential for maintaining a thermodynamic stability of the overall system and eventually a functional conformation for tRNA recognition. Altogether, our findings suggest that functional adaptability of extended dsRBDs is promoted by a cooperative hydrophobic coupling between the extensions acting as effectors and the folding nucleus of the canonical structure.

Ultrafast photoinduced flavin dynamics in the unusual active site of the tRNA methyltransferase TrmFO.

Flavoproteins often stabilize their flavin coenzyme by stacking interactions involving the isoalloxazine moiety of the flavin and an aromatic residue from the apoprotein. The bacterial FAD and folate-dependent tRNA methyltransferase TrmFO has the unique property of stabilizing its FAD coenzyme by an unusual H-bond-assisted π-π stacking interaction, involving a conserved tyrosine (Y346 in Bacillus subtilis TrmFO, BsTrmFO), the isoalloxazine of FAD and the backbone of a catalytic cysteine (C53). Here, the interaction between FAD and Y346 has been investigated by measuring the photoinduced flavin dynamics of BsTrmFO in the wild-type (WT) protein, C53A and several Y346 mutants by ultrafast transient absorption spectroscopy. In C53A, the excited FAD very rapidly (0.43 ps) abstracts an electron from Y346, yielding the FAD˙-/Y346OH˙+ radical pair, while relaxation of the local environment (1.3 ps) of the excited flavin produces a slight Stokes shift of its stimulated emission band. The radical pair then decays via charge recombination, mostly in 3-4 ps, without any deprotonation of the Y346OH˙+ radical. Presumably, the H-bond between Y346 and the amide group of C53 increases the pKa of Y346OH˙+ and slows down its deprotonation. The dynamics of WT BsTrmFO shows additional slow decay components (43 and 700 ps), absent in the C53A mutant, assigned to excited FADox populations not undergoing fast photoreduction. Their presence is likely due to a more flexible structure of the WT protein, favored by the presence of C53. Interestingly, mutations of Y346 canceling its electron donating character lead to multiple slower quenching channels in the ps-ns regime. These channels are proposed to be due to electron abstraction either (i) from the adenine moiety of FAD, a distribution of the isoalloxazine-adenine distance in the absence of Y346 explaining the multiexponential decay, or (ii) from the W286 residue, possibly accounting for one of the decays. This work supports the idea that H-bond-assisted π-π stacking controls TrmFO's active site dynamics, required for competent orientation of the reactive centers during catalysis.

Dissociation of the Dimer of the Intrinsically Disordered Domain of RNase Y upon Antibody Binding.

Although RNase Y acts as the key enzyme initiating messenger RNA decay in Bacillus subtilis and likely in many other Gram-positive bacteria, its three-dimensional structure remains unknown. An antibody belonging to the rare immunoglobulin G (IgG) 2b λx isotype was raised against a 12-residue conserved peptide from the N-terminal noncatalytic domain of B. subtilis RNase Y (BsRNaseY) that is predicted to be intrinsically disordered. Here, we show that this domain can be produced as a stand-alone protein called Nter-BsRNaseY that undergoes conformational changes between monomeric and dimeric forms. Circular dichroism and size exclusion chromatography coupled with multiangle light scattering or with small angle x-ray scattering indicate that the Nter-BsRNaseY dimer displays an elongated form and a high content of α-helices, in agreement with the existence of a central coiled-coil structure appended with flexible ends, and that the monomeric state of Nter-BsRNaseY is favored upon binding the fragment antigen binding (Fab) of the antibody. The dissociation constants of the IgG/BsRNaseY, IgG/Nter-BsRNaseY, and IgG/peptide complexes indicate that the affinity of the IgG for Nter-BsRNaseY is in the nM range and suggest that the peptide is less accessible in BsRNaseY than in Nter-BsRNaseY. The crystal structure of the Fab in complex with the peptide antigen shows that the peptide adopts an elongated U-shaped conformation in which the unique hydrophobic residue of the peptide, Leu6, is completely buried. The peptide/Fab complex may mimic the interaction of a microdomain of the N-terminal domain of BsRNaseY with one of its cellular partners within the degradosome complex. Altogether, our results suggest that BsRNaseY may become accessible for protein interaction upon dissociation of its N-terminal domain into the monomeric form.

Electrostatic Potential in the tRNA Binding Evolution of Dihydrouridine Synthases.

Dihydrouridine (D) is an abundant modified base of tRNA found in the majority of living organisms. This base is synthesized via an NADPH-dependent reduction of specific uridines by the dihydrouridine synthases (Dus), a large family of flavoenzymes comprising eight subfamilies. Almost all of these enzymes function with only two conserved domains, an N-terminal catalytic domain (TBD) adopting a TIM barrel fold and a unique C-terminal helical domain (HD) devoted to tRNA recognition, except for the animal U20-specific Dus2 enzyme. Curiously, this enzyme is distinguished from paralogues and its fungi orthologues by the acquisition of an additional domain, a double stranded RNA binding domain (dsRBD), which serves as the main tRNA binding module. On the basis of a homology model of yeast Dus2 and the crystallographic structure of a human Dus2 variant (TBD + HD) lacking dsRBD, we herein show that the HD surface of the human enzyme is less electropositive than that of its yeast orthologue. This is partly due to two positively charged residues, K304 and K315, present in yeast and more broadly in fungi Dus2 that are replaced by E294 and Q305 in human and conserved among animals Dus2. By artificially reintroducing these positive charges in human Dus2 lacking dsRBD, we restored a functional tRNA binding in this enzyme variant. Altogether, these results suggest that the electrostatic potential changes of HD have likely played a key role in the emergence of a new tRNA binding mode among Dus2 enzymes.

Power of protein/tRNA functional assembly against aberrant aggregation.

Understanding the mechanisms of protein oligomerization and aggregation is a major concern for biotechnology and medical purposes. However, significant challenges remain in determining the mechanism of formation of these superstructures and the environmental factors that can precisely modulate them. Notably the role that a functional ligand plays in the process of protein aggregation is largely unexplored. We herein address these issues with an original flavin-dependent RNA methyltransferase (TrmFO) used as a protein model since this protein employs a complex set of cofactors and ligands for catalysis. Here, we show that TrmFO carries an unstable protein structure that can partially mis-unfold leading to either formation of irregular and nonfunctional soluble oligomers endowed with hyper-thermal stability or large amorphous aggregates in the presence of salts. Mutagenesis confirmed that this peculiarity is an intrinsic property of a polypeptide and it is independent of the flavin coenzyme. Structural characterization and kinetic studies identified several regions of the protein that enjoy conformational changes and more particularly pinpointed the N-terminal subdomain as being a key element in the mechanisms of oligomerization and aggregation. Only stabilization of this region via tRNA suppresses these aberrant protein states. Although protein chaperones emerged as major actors against aggregation, our study emphasizes that other powerful mechanisms exist such as the stabilizing effect of functional assemblies that provide an additional layer of protection against the instability of the proteome.

An extended dsRBD is required for post-transcriptional modification in human tRNAs.

In tRNA, dihydrouridine is a conserved modified base generated by the post-transcriptional reduction of uridine. Formation of dihydrouridine 20, located in the D-loop, is catalyzed by dihydrouridine synthase 2 (Dus2). Human Dus2 (HsDus2) expression is upregulated in lung cancers, offering a growth advantage throughout its ability to interact with components of the translation apparatus and inhibit apoptosis. Here, we report the crystal structure of the individual domains of HsDus2 and their functional characterization. HsDus2 is organized into three major modules. The N-terminal catalytic domain contains the flavin cofactor involved in the reduction of uridine. The second module is the conserved α-helical domain known as the tRNA binding domain in HsDus2 homologues. It is connected via a flexible linker to an unusual extended version of a dsRNA binding domain (dsRBD). Enzymatic assays and yeast complementation showed that the catalytic domain binds selectively NADPH but cannot reduce uridine in the absence of the dsRBD. While in Dus enzymes from bacteria, plants and fungi, tRNA binding is essentially achieved by the α-helical domain, we showed that in HsDus2 this function is carried out by the dsRBD. This is the first reported case of a tRNA-modifying enzyme carrying a dsRBD used to bind tRNAs.

Enzyme Activation with a Synthetic Catalytic Co-enzyme Intermediate: Nucleotide Methylation by Flavoenzymes.

To facilitate production of functional enzymes and to study their mechanisms, especially in the complex cases of coenzyme-dependent systems, activation of an inactive apoenzyme preparation with a catalytically competent coenzyme intermediate is an attractive strategy. This is illustrated with the simple chemical synthesis of a flavin-methylene iminium compound previously proposed as a key intermediate in the catalytic cycle of several important flavoenzymes involved in nucleic acid metabolism. Reconstitution of both flavin-dependent RNA methyltransferase and thymidylate synthase apoproteins with this synthetic compound led to active enzymes for the C5-uracil methylation within their respective transfer RNA and dUMP substrate. This strategy is expected to be of general application in enzymology.

Flavin-Protein Complexes: Aromatic Stacking Assisted by a Hydrogen Bond.

Enzyme-catalyzed reactions often rely on a noncovalently bound cofactor whose reactivity is tuned by its immediate environment. Flavin cofactors, the most versatile catalyst encountered in biology, are often maintained within the protein throughout numbers of complex ionic and aromatic interactions. Here, we have investigated the role of π-π stacking and hydrogen bond interactions between a tyrosine and the isoalloxazine moiety of the flavin adenine dinucleotide (FAD) in an FAD-dependent RNA methyltransferase. Combining several static and time-resolved spectroscopies as well as biochemical approaches, we showed that aromatic stacking is assisted by a hydrogen bond between the phenol group and the amide of an adjacent active site loop. A mechanism of recognition and binding of the redox cofactor is proposed.

Structural basis of MALAT1 RNA maturation and mascRNA biogenesis.

The metastasis-associated lung adenocarcinoma transcript 1 (MALAT1) long noncoding RNA (lncRNA) has key roles in regulating transcription, splicing, tumorigenesis, etc. Its maturation and stabilization require precise processing by RNase P, which simultaneously initiates the biogenesis of a 3' cytoplasmic MALAT1-associated small cytoplasmic RNA (mascRNA). mascRNA was proposed to fold into a transfer RNA (tRNA)-like secondary structure but lacks eight conserved linking residues required by the canonical tRNA fold. Here we report crystal structures of human mascRNA before and after processing, which reveal an ultracompact, quasi-tRNA-like structure. Despite lacking all linker residues, mascRNA faithfully recreates the characteristic 'elbow' feature of tRNAs to recruit RNase P and ElaC homolog protein 2 (ELAC2) for processing, which exhibit distinct substrate specificities. Rotation and repositioning of the D-stem and anticodon regions preclude mascRNA from aminoacylation, avoiding interference with translation. Therefore, a class of metazoan lncRNA loci uses a previously unrecognized, unusually streamlined quasi-tRNA architecture to recruit select tRNA-processing enzymes while excluding others to drive bespoke RNA biogenesis, processing and maturation.

Shigella IpaA mediates actin bundling through diffusible vinculin oligomers with activation imprint.

Upon activation, vinculin reinforces cytoskeletal anchorage during cell adhesion. Activating ligands classically disrupt intramolecular interactions between the vinculin head and tail domains that bind to actin filaments. Here, we show that Shigella IpaA triggers major allosteric changes in the head domain, leading to vinculin homo-oligomerization. Through the cooperative binding of its three vinculin-binding sites (VBSs), IpaA induces a striking reorientation of the D1 and D2 head subdomains associated with vinculin oligomerization. IpaA thus acts as a catalyst producing vinculin clusters that bundle actin at a distance from the activation site and trigger the formation of highly stable adhesions resisting the action of actin relaxing drugs. Unlike canonical activation, vinculin homo-oligomers induced by IpaA appear to keep a persistent imprint of the activated state in addition to their bundling activity, accounting for stable cell adhesion independent of force transduction and relevant to bacterial invasion.

Rational engineering enables co-crystallization and structural determination of the HIV-1 matrix-tRNA complex.

Host tRNAs specifically interact with the matrix domain (MA) of HIV-1 major structural polyprotein, Gag, to control its membrane localization and virion assembly. In this protocol, we describe the purification and engineering of HIV-1 MA and tRNA, and the co-crystallization and structure determination of the complex using X-ray crystallography. Rational engineering of the tRNA surface created tRNA-tRNA packing contacts that drove the formation of diffraction-quality co-crystals. This protocol can be adapted to solve other ribonucleoprotein complex structures containing structured RNAs. For complete details on the use and execution of this protocol, please refer to Bou-Nader et al. (2021).

Structural basis of R-loop recognition by the S9.6 monoclonal antibody.

R-loops are ubiquitous, dynamic nucleic-acid structures that play fundamental roles in DNA replication and repair, chromatin and transcription regulation, as well as telomere maintenance. The DNA-RNA hybrid-specific S9.6 monoclonal antibody is widely used to map R-loops. Here, we report crystal structures of a S9.6 antigen-binding fragment (Fab) free and bound to a 13-bp hybrid duplex. We demonstrate that S9.6 exhibits robust selectivity in binding hybrids over double-stranded (ds) RNA and in categorically rejecting dsDNA. S9.6 asymmetrically recognizes a compact epitope of two consecutive RNA nucleotides via their 2'-hydroxyl groups and six consecutive DNA nucleotides via their backbone phosphate and deoxyribose groups. Recognition is mediated principally by aromatic and basic residues of the S9.6 heavy chain, which closely track the curvature of the hybrid minor groove. These findings reveal the molecular basis for S9.6 recognition of R-loops, detail its binding specificity, identify a new hybrid-recognition strategy, and provide a framework for S9.6 protein engineering.

An enzymatic activation of formaldehyde for nucleotide methylation.

Folate enzyme cofactors and their derivatives have the unique ability to provide a single carbon unit at different oxidation levels for the de novo synthesis of amino-acids, purines, or thymidylate, an essential DNA nucleotide. How these cofactors mediate methylene transfer is not fully settled yet, particularly with regard to how the methylene is transferred to the methylene acceptor. Here, we uncovered that the bacterial thymidylate synthase ThyX, which relies on both folate and flavin for activity, can also use a formaldehyde-shunt to directly synthesize thymidylate. Combining biochemical, spectroscopic and anaerobic crystallographic analyses, we showed that formaldehyde reacts with the reduced flavin coenzyme to form a carbinolamine intermediate used by ThyX for dUMP methylation. The crystallographic structure of this intermediate reveals how ThyX activates formaldehyde and uses it, with the assistance of active site residues, to methylate dUMP. Our results reveal that carbinolamine species promote methylene transfer and suggest that the use of a CH2O-shunt may be relevant in several other important folate-dependent reactions.

HIV-1 matrix-tRNA complex structure reveals basis for host control of Gag localization.

The HIV-1 virion structural polyprotein, Gag, is directed to particle assembly sites at the plasma membrane by its N-terminal matrix (MA) domain. MA also binds to host tRNAs. To understand the molecular basis of MA-tRNA interaction and its potential function, we present a co-crystal structure of HIV-1 MA-tRNALys3 complex. The structure reveals a specialized group of MA basic and aromatic residues preconfigured to recognize the distinctive structure of the tRNA elbow. Mutational, cross-linking, fluorescence, and NMR analyses show that the crystallographically defined interface drives MA-tRNA binding in solution and living cells. The structure indicates that MA is unlikely to bind tRNA and membrane simultaneously. Accordingly, single-amino-acid substitutions that abolish MA-tRNA binding caused striking redistribution of Gag to the plasma membrane and reduced HIV-1 replication. Thus, HIV-1 exploits host tRNAs to occlude a membrane localization signal and control the subcellular distribution of its major structural protein.