An AI-informed NMR structure reveals an extraordinary LETM1 F-EF-hand domain that functions as a two-way regulator of mitochondrial calcium.

AlphaFold can accurately predict static protein structures but does not account for solvent conditions. Human leucine zipper EF-hand transmembrane protein-1 (LETM1) has one sequence-identifiable EF-hand but how calcium (Ca2+) affects structure and function remains enigmatic. Here, we used highly confident AlphaFold Cα predictions to guide nuclear Overhauser effect (NOE) assignments and structure calculation of the LETM1 EF-hand in the presence of Ca2+. The resultant NMR structure exposes pairing between a partial loop-helix and full helix-loop-helix, forming an unprecedented F-EF-hand with non-canonical Ca2+ coordination but enhanced hydrophobicity for protein interactions compared to calmodulin. The structure also reveals the basis for pH sensing at the link between canonical and partial EF-hands. Functionally, mutations that augmented or weakened Ca2+ binding increased or decreased matrix Ca2+, respectively, establishing F-EF as a two-way mitochondrial Ca2+ regulator. Thus, we show how to synergize AI prediction with NMR data, elucidating a solution-specific and extraordinary LETM1 F-EF-hand.

The 1H, 15N, and 13C resonance assignments of a single-domain antibody against immunoglobulin G.

Research on camelid-derived single-domain antibodies (sdAbs) has demonstrated their significant utility in diverse biotechnological applications, including therapy and diagnostic. This is largely due to their relative simplicity as monomeric proteins, ranging from 12 to 15 kDa, in contrast to immunoglobulin G (IgG) antibodies, which are glycosylated heterotetramers of 150-160 kDa. Single-domain antibodies exhibit high conformational stability and adopt the typical immunoglobulin domain fold, consisting of a two-layer sandwich of 7-9 antiparallel beta-strands. They contain three loops, known as complementary-determining regions (CDRs), which are assembled on the sdAb surface and are responsible for antigen recognition. The single-domain antibody examined in this study, sdAb-mrh-IgG, was engineered to recognize IgG from rats, mice, but it also weakly recognizes IgG from humans (Pleiner et al. 2018). A search of the Protein Data Bank revealed only one NMR structure of a single-domain antibody, which is unrelated to sdAb-mrh-IgG. The NMR chemical shift assignments of sdAb-mrh-IgG will be utilized to study its molecular dynamics and interactions with antigens in solution, which is fundamental for the rational design of novel single-domain antibodies.

Hidden Structural States of Proteins Revealed by Conformer Selection with AlphaFold-NMR.

Recent advances in molecular modeling using deep learning can revolutionize our understanding of dynamic protein structures. NMR is particularly well-suited for determining dynamic features of biomolecular structures. The conventional process for determining biomolecular structures from experimental NMR data involves its representation as conformation-dependent restraints, followed by generation of structural models guided by these spatial restraints. Here we describe an alternative approach: generating a distribution of realistic protein conformational models using artificial intelligence-(AI-) based methods and then selecting the sets of conformers that best explain the experimental data. We applied this conformational selection approach to redetermine the solution NMR structure of the enzyme Gaussia luciferase. First, we generated a diverse set of conformer models using AlphaFold2 (AF2) with an enhanced sampling protocol. The models that best-fit NOESY and chemical shift data were then selected with a Bayesian scoring metric. The resulting models include features of both the published NMR structure and the standard AF2 model generated without enhanced sampling. This "AlphaFold-NMR" protocol also generated an alternative "open" conformational state that fits nearly as well to the overall NMR data but accounts for some NOESY data that is not consistent with first "closed" conformational state; while other NOESY data consistent with this second state are not consistent with the first conformational state. The structure of this "open" structural state differs from that of the "closed" state primarily by the position of a thumb-shaped loop between α-helices H5 and H6, revealing a cryptic surface pocket. These alternative conformational states of Gluc are supported by "double recall" analysis of NOESY data and AF2 models. Additional structural states are also indicated by backbone chemical shift data indicating partially-disordered conformations for the C-terminal segment. Considered as a multistate ensemble, these multiple states of Gluc together fit the NOESY and chemical shift data better than the "restraint-based" NMR structure and provide novel insights into its structure-dynamic-function relationships. This study demonstrates the potential of AI-based modeling with enhanced sampling to generate conformational ensembles followed by conformer selection with experimental data as an alternative to conventional restraint satisfaction protocols for protein NMR structure determination.

Structure-based investigation of a DNA aptamer targeting PTK7 reveals an intricate 3D fold guiding functional optimization.

DNA aptamers have emerged as novel molecular tools in disease theranostics owing to their high binding affinity and specificity for protein targets, which rely on their ability to fold into distinctive three-dimensional (3D) structures. However, delicate atomic interactions that shape the 3D structures are often ignored when designing and modeling aptamers, leading to inefficient functional optimization. Challenges persist in determining high-resolution aptamer-protein complex structures. Moreover, the experimentally determined 3D structures of DNA molecules with exquisite functions remain scarce. These factors impede our comprehension and optimization of some important DNA aptamers. Here, we performed a streamlined solution NMR-based structural investigation on the 41-nt sgc8c, a prominent DNA aptamer used to target membrane protein tyrosine kinase 7, for cancer theranostics. We show that sgc8c prefolds into an intricate three-way junction (3WJ) structure stabilized by long-range tertiary interactions and extensive base-base stackings. Delineated by NMR chemical shift perturbations, site-directed mutagenesis, and 3D structural information, we identified essential nucleotides constituting the key functional elements of sgc8c that are centralized at the core of 3WJ. Leveraging the well-established structure-function relationship, we efficiently engineered two sgc8c variants by modifying the apical loop and introducing L-DNA base pairs to simultaneously enhance thermostability, biostability, and binding affinity for both protein and cell targets, a feat not previously attained despite extensive efforts. This work showcases a simplified NMR-based approach to comprehend and optimize sgc8c without acquiring the complex structure, and offers principles for the sophisticated structure-function organization of DNA molecules.

One N-glycan regulates natural killer cell antibody-dependent cell-mediated cytotoxicity and modulates Fc γ receptor IIIa / CD16a structure.

Both endogenous antibodies and a subset of antibody therapeutics engage Fc gamma receptor (FcγR)IIIa / CD16a to stimulate a protective immune response. Increasing the FcγRIIIa/IgG1 interaction improves the immune response and thus represents a strategy to improve therapeutic efficacy. FcγRIIIa is a heavily glycosylated receptor and glycan composition affects antibody-binding affinity. Though our laboratory previously demonstrated that natural killer (NK) cell N-glycan composition affected the potency of one key protective mechanism, antibody-dependent cell-mediated cytotoxicity (ADCC), it was unclear if this effect was due to FcγRIIIa glycosylation. Furthermore, the structural mechanism linking glycan composition to affinity and cellular activation remained undescribed. To define the role of individual amino acid and N-glycan residues we measured affinity using multiple FcγRIIIa glycoforms. We observed stepwise affinity increases with each glycan truncation step with the most severely truncated glycoform displaying the highest affinity. Removing the N162 glycan demonstrated its predominant role in regulating antibody-binding affinity, in contrast to four other FcγRIIIa N-glycans. We next evaluated the impact of the N162 glycan on NK cell ADCC. NK cells expressing the FcγRIIIa V158 allotype exhibited increased ADCC following kifunensine treatment to limit N-glycan processing. Notably, an increase was not observed with cells expressing the FcγRIIIa V158 S164A variant that lacks N162 glycosylation, indicating the N162 glycan is required for increased NK cell ADCC. To gain structural insight into the mechanisms of N162 regulation, we applied a novel protein isotope labeling approach in combination with solution NMR spectroscopy. FG loop residues proximal to the N162 glycosylation site showed large chemical shift perturbations following glycan truncation. These data support a model for the regulation of FcγRIIIa affinity and NK cell ADCC whereby composition of the N162 glycan stabilizes the FG loop and thus the antibody-binding site.

Effects of interrupting residues on DNA dumbbell structures formed by CCTG tetranucleotide repeats associated with myotonic dystrophy type 2.

Myotonic dystrophy type 2 (DM2) is a neurogenerative disease caused by caprylic/capric triglyceride (CCTG) tetranucleotide repeat expansions in intron 1 of the cellular nucleic acid-binding protein (CNBP) gene. Non-B DNA structures formed by CCTG repeats can promote genetic instability, whereas interrupting motifs of NCTG (N = A/T/G) within CCTG repeats help to maintain genomic stability. However, whether the interrupting motifs can affect DNA structures of CCTG repeats remains unclear. Here, we report that four CCTG repeats with an interrupting 3'-A/T/G residue formed dumbbell structures, whereas a non-interrupting 3'-C residue resulted in a multi-loop structure exhibiting conformational dynamics that may contribute to a higher tendency of escaping from DNA mismatch repair and causing repeat expansions. The results provide new structural insights into the genetic instability of CCTG repeats in DM2.

Validating the 15N-1H HSQC-ROESY experiment for detecting 1HN exchange broadening in proteated proteins.

It is advantageous to investigate milli-to-micro-second conformational exchange data contained in the solution NMR protein relaxation data other than 15N nuclei. Not only does one search under another lamp post, one also looks at dynamics at other time scales. The HSQC-ROESY 1HN relaxation dispersion experiment for amide protons as introduced by Ishima, et al (1998). J. Am. Soc. 120, 10534-10542, is such an experiment, but has by the authors been advised to only be used for perdeuterated proteins to avoid complication with the 1H-1H multiple-spin effects. This is regretful, since not all proteins can be perdeuterated. Here we analyze in detail the 1HN relaxation terms for this experiment for a fully proteated protein. Indeed, the 1HN relaxation theory is in this case complex and includes dipolar-dipolar relaxation interference and TOCSY transfers. With simulate both of these effects and show that the interference can be exploited for detecting exchange broadening. The TOCSY effect is shown to minor, and when it is not, a solution is provided. We apply the HSQC-ROESY experiment, with a small modification to suppress ROESY crosspeaks, to a 7 kDa GB1 protein that is just 15N and 13C labeled. At 10 °C we cannot detect any conformational exchange broadening: the 1HN R2 relaxation rates with 1.357 kHz spinlock field not larger than those recorded with a 12.136 kHz spinlock field. This means that there is no exchange broadening that can be differentially suppressed with the applied fields. Either there is no broadening, or the broadening is effectively suppressed by all fields, or the broadening cannot be suppressed by either of the fields. While initially this seems to be a disappointing result, we feel that this work establishes that the HSQC-ROESY experiment is very robust. It can indeed be utilized for proteated proteins upto about 30 kDa. This could be opening the study the milli-microsecond conformational dynamics as reported by 1HN exchange broadening for many more proteins.

Improved analysis of NMR chemical shift perturbations through an error estimation method.

In solution NMR, chemical shift perturbation (CSP) experiments are widely employed to study intermolecular interactions. However, excluding the nonsignificant peak shift is difficult because little is known about errors in CSP. Here, to address this issue, we introduce a method for estimating errors in CSP based on the noise level. First, we developed a technique that involves line shape fitting to estimate errors in peak position via Monte Carlo simulations. Second, this technique was applied to estimate errors in CSP. In intermolecular interaction analysis of VAP-A with SNX2, error estimation of CSP enabled the evaluation of small but significant changes in peak position and yielded detailed insights that are unattainable with conventional CSP analysis. Third, this technique was successfully applied to estimate errors in residual dipolar couplings. In conclusion, our error estimation method improves CSP analysis by excluding the nonsignificant peak shift.

Affinity-directed substrate/H+-antiport by a MATE transporter.

Multidrug and toxin extrusion (MATE) family transporters excrete toxic compounds coupled to Na+/H+ influx. Although structures of MATE transporters are available, the mechanism by which substrate export is coupled to ion influx remains unknown. To address this issue, we conducted a structural analysis of Pyrococcus furiosus MATE (PfMATE) using solution nuclear magnetic resonance (NMR). The NMR analysis, along with thorough substitutions of all non-exposed acidic residues, confirmed that PfMATE is under an equilibrium between inward-facing (IF) and outward-facing (OF) conformations, dictated by the Glu163 protonation. Importantly, we found that only the IF conformation exhibits a mid-µM affinity for substrate recognition. In contrast, the OF conformation exhibited only weak mM substrate affinity, suitable for releasing substrate to the extracellular side. These results indicate that PfMATE is an affinity-directed H+ antiporter where substrates selectively bind to the protonated IF conformation in the equilibrium, and subsequent proton release mechanistically ensures H+-coupled substrate excretion by the transporter.

NMR methods to detect fluoride binding and transport by membrane proteins.

Solid-state nuclear magnetic resonance (NMR) methods can probe the motions of membrane proteins in liposomes at the atomic level, and propel the understanding of biomolecular processes for which static structures cannot provide a satisfactory description. High-resolution crystallography snapshots have provided a structural basis for fluoride channels. NMR is a powerful tool to build upon these snapshots and depict a dynamic picture of fluoride channels in native-like lipid bilayers. In this contribution, we discuss solid-state and solution NMR experiments to detect fluoride binding and transport by fluoride channels. Ongoing developments in membrane protein sample preparation and ssNMR methodology, particularly in using 1H, 19F and 13C-detection schemes, offer additional opportunities to study structure and functional aspects of fluoride channels.

High-Resolution NMR Structures of Intrastrand Hairpins Formed by CTG Trinucleotide Repeats.

The CAG and CTG trinucleotide repeat expansions cause more than 10 human neurodegenerative diseases. Intrastrand hairpins formed by trinucleotide repeats contribute to repeat expansions, establishing them as potential drug targets. High-resolution structural determination of CAG and CTG hairpins poses as a long-standing goal to aid drug development, yet it has not been realized due to the intrinsic conformational flexibility of repetitive sequences. We herein investigate the solution structures of CTG hairpins using nuclear magnetic resonance (NMR) spectroscopy and found that four CTG repeats with a clamping G-C base pair was able to form a stable hairpin structure. We determine the first solution NMR structure of dG(CTG)4C hairpin and decipher a type I folding geometry of the TGCT tetraloop, wherein the two thymine residues form a T·T loop-closing base pair and the first three loop residues continuously stack. We further reveal that the CTG hairpin can be bound and stabilized by a small-molecule ligand, and the binding interferes with replication of a DNA template containing CTG repeats. Our determined high-resolution structures lay an important foundation for studying molecular interactions between native CTG hairpins and ligands, and benefit drug development for trinucleotide repeat expansion diseases.

Probing amino acid side chains of the integral membrane protein PagP by solution NMR: Side chain immobilization facilitates association of secondary structures.

Solution NMR spectroscopy of large protein systems is hampered by rapid signal decay, so most multidimensional studies focus on long-lived 1H-13C magnetization in methyl groups and/or backbone amide 1H-15N magnetization in an otherwise perdeuterated environment. Herein we demonstrate that it is possible to biosynthetically incorporate additional 1H-12C groups that possess long-lived magnetization using cost-effective partially deuterated or unlabeled amino acid precursors added to Escherichia coli growth media. This approach is applied to the outer membrane enzyme PagP in membrane-mimetic dodecylphosphocholine micelles. We were able to obtain chemical shift assignments for a majority of side chain 1H positions in PagP using nuclear Overhauser enhancements (NOEs) to connect them to previously assigned backbone 1H-15N groups and newly assigned 1H-13C methyl groups. Side chain methyl-to-aromatic NOEs were particularly important for confirming that the amphipathic α-helix of PagP packs against its eight-stranded ß-barrel, as indicated by previous X-ray crystal structures. Interestingly, aromatic NOEs suggest that some aromatic residues in PagP that are buried in the membrane bilayer are highly mobile in the micellar environment, like Phe138 and Phe159. In contrast, Tyr87 in the middle of the bilayer is quite rigid, held in place by a hydrogen bonded network extending to the surface that resembles a classic catalytic triad: Tyr87-His67-Asp61. This hydrogen bonded arrangement of residues is not known to have any catalytic activity, but we postulate that its role is to immobilize Tyr87 to facilitate packing of the amphipathic α-helix against the ß-barrel.

Structural Heterogeneity in a Phototransformable Fluorescent Protein Impacts its Photochemical Properties.

Photoconvertible fluorescent proteins (PCFP) are important cellular markers in advanced imaging modalities such as photoactivatable localization microscopy (PALM). However, their complex photophysical and photochemical behavior hampers applications such as quantitative and single-particle-tracking PALM. This work employs multidimensional NMR combined with ensemble fluorescence measurements to show that the popular mEos4b in its Green state populates two conformations (A and B), differing in side-chain protonation of the conserved residues E212 and H62,  altering the hydrogen-bond network in the chromophore pocket. The interconversion (protonation/deprotonation) between these two states, which occurs on the minutes time scale in the dark, becomes strongly accelerated in the presence of UV light, leading to a population shift. This work shows that the reversible photoswitching and Green-to-Red photoconversion properties differ between the A and B states. The chromophore in the A-state photoswitches more efficiently and is proposed to be more prone to photoconversion, while the B-state shows a higher level of photobleaching. Altogether, this data highlights the central role of conformational heterogeneity in fluorescent protein photochemistry.

The protease associated (PA) domain in ScpA from Streptococcus pyogenes plays a role in substrate recruitment.

Annually, over 18 million disease cases and half a million deaths worldwide are estimated to be caused by Group A Streptococcus. ScpA (or C5a peptidase) is a well characterised member of the cell enveleope protease family, which possess a S8 subtilisin-like catalytic domain and a shared multi-domain architecture. ScpA cleaves complement factors C5a and C3a, impairing the function of these critical anaphylatoxins and disrupts complement-mediated innate immunity. Although the high resolution structure of ScpA is known, the details of how it recognises its substrate are only just emerging. Previous studies have identified a distant exosite on the 2nd fibronectin domain that plays an important role in recruitment via an interaction with the substrate core. Here, using a combination of solution NMR spectroscopy, mutagenesis with functional assays and computational approaches we identify a second exosite within the protease-associated (PA) domain. We propose a model in which the PA domain assists optimal delivery of the substrate's C terminus to the active site for cleavage.

Lipid- and substrate-induced conformational and dynamic changes in a glycosyltransferase involved in E. coli LPS synthesis revealed by 19F and 31P NMR.

WaaG is a glycosyltransferase (GT) involved in the synthesis of the bacterial cell wall, and in Escherichia coli it catalyzes the transfer of a glucose moiety from the donor substrate UDP-glucose onto the nascent lipopolysaccharide (LPS) molecule which when completed constitutes the major component of the bacterium's outermost defenses. Similar to other GTs of the GT-B fold, having two Rossman-like domains connected by a short linker, WaaG is believed to undergo complex inter-domain motions as part of its function to accommodate the nascent LPS and UDP-glucose in the catalytic site located in the cleft between the two domains. As the nascent LPS is bulky and membrane-bound, WaaG is a peripheral membrane protein, adding to the complexity of studying the enzyme in a biologically relevant environment. Using specific 5-fluoro-Trp labelling of native and inserted tryptophans and 19F NMR we herein studied the dynamic interactions of WaaG with lipids using bicelles, and with the donor substrate. Line-shape changes when bicelles are added to WaaG show that the dynamic behavior is altered when binding to the model membrane, while a chemical shift change indicates an altered environment around a tryptophan located in the C-terminal domain of WaaG upon interaction with UDP-glucose or UDP. A lipid-bound paramagnetic probe was used to confirm that the membrane interaction is mediated by a loop region located in the N-terminal domain. Furthermore, the hydrolysis of the donor substrate by WaaG was quantified by 31P NMR.

Visualizing a two-state conformational ensemble in stem-loop 3 of the transcriptional regulator 7SK RNA.

Structural plasticity is integral to RNA function; however, there are currently few methods to quantitatively resolve RNAs that have multiple structural states. NMR spectroscopy is a powerful approach for resolving conformational ensembles but is size-limited. Chemical probing is well-suited for large RNAs but provides limited structural and no kinetics information. Here, we integrate the two approaches to visualize a two-state conformational ensemble for the central stem-loop 3 (SL3) of 7SK RNA, a critical element for 7SK RNA function in transcription regulation. We find that the SL3 distal end exchanges between two equally populated yet structurally distinct states in both isolated SL3 constructs and full-length 7SK RNA. We rationally designed constructs that lock SL3 into a single state and demonstrate that both chemical probing and NMR data fit to a linear combination of the two states. Comparison of vertebrate 7SK RNA sequences shows conservation of both states, suggesting functional importance. These results provide new insights into 7SK RNA structural dynamics and demonstrate the utility of integrating chemical probing with NMR spectroscopy to gain quantitative insights into RNA conformational ensembles.

Prostaglandins as Candidate Ligands for a Per-ARNT-Sim (PAS) Domain of Steroid Receptor Coactivator 1 (SRC1).

Steroid receptor coactivators (SRCs) comprise a family of three paralogous proteins commonly recruited by eukaryotic transcription factors. Each SRC harbors two tandem Per-ARNT-Sim (PAS) domains that are broadly distributed that bind small molecules and regulate interactions. Using computational docking, solution NMR, mass spectrometry, and molecular dynamics simulations, we show that the SRC1 PAS-B domain can bind to certain prostaglandins (PGs) either non-covalently to a surface that overlaps with the site used to engage transcription factors or covalently to a single, specific, conserved cysteine residue next to a solvent accessible hydrophobic pocket. This pocket is in proximity to the canonical transcription factor binding site, but on the opposite side of the domain, suggesting a potential mode of regulating transcriptional activator-coactivator interactions.

Two-Dimensional NMR Spectroscopy of the G Protein-Coupled Receptor A2AAR in Lipid Nanodiscs.

Eight hundred and twenty-six human G protein-coupled receptors (GPCRs) mediate the actions of two-thirds of the human hormones and neurotransmitters and over one-third of clinically used drugs. Studying the structure and dynamics of human GPCRs in lipid bilayer environments resembling the native cell membrane milieu is of great interest as a basis for understanding structure-function relationships and thus benefits continued drug development. Here, we incorporate the human A2A adenosine receptor (A2AAR) into lipid nanodiscs, which represent a detergent-free environment for structural studies using nuclear magnetic resonance (NMR) in solution. The [15N,1H]-TROSY correlation spectra confirmed that the complex of [u-15N, ~70% 2H]-A2AAR with an inverse agonist adopts its global fold in lipid nanodiscs in solution at physiological temperature. The global assessment led to two observations of practical interest. First, A2AAR in nanodiscs can be stored for at least one month at 4 °C in an aqueous solvent. Second, LMNG/CHS micelles are a very close mimic of the environment of A2AAR in nanodiscs. The NMR signal of five individually assigned tryptophan indole 15N-1H moieties located in different regions of the receptor structure further enabled a detailed assessment of the impact of nanodiscs and LMNG/CHS micelles on the local structure and dynamics of A2AAR. As expected, the largest effects were observed near the lipid-water interface along the intra- and extracellular surfaces, indicating possible roles of tryptophan side chains in stabilizing GPCRs in lipid bilayer membranes.

Cost-effective selective deuteration of aromatic amino acid residues produces long-lived solution 1H NMR magnetization in proteins.

Solution NMR studies of large proteins are hampered by rapid signal decay due to short-range dipolar 1H-1H and 1H-13C interactions. These are attenuated by rapid rotation in methyl groups and by deuteration (2H), so selective 1H,13C-isotope labelling of methyl groups in otherwise perdeuterated proteins, combined with methyl transverse relaxation optimized spectroscopy (methyl-TROSY), is now standard for solution NMR of large protein systems > 25 kDa. For non-methyl positions, long-lived magnetization can be introduced as isolated 1H-12C groups. We have developed a cost-effective chemical synthesis for producing selectively deuterated phenylpyruvate and hydroxyphenylpyruvate. Feeding these amino acid precursors to E. coli in D2O, along with selectively deuterated anthranilate and unlabeled histidine, results in isolated and long-lived 1H magnetization in the aromatic rings of Phe (HD, HZ), Tyr (HD), Trp (HH2, HE3) and His (HD2 and HE1). We are additionally able to obtain stereoselective deuteration of Asp, Asn, and Lys amino acid residues using unlabeled glucose and fumarate as carbon sources and oxalate and malonate as metabolic inhibitors. Combining these approaches produces isolated 1H-12C groups in Phe, Tyr, Trp, His, Asp, Asn, and Lys in a perdeuterated background, which is compatible with standard 1H-13C labeling of methyl groups in Ala, Ile, Leu, Val, Thr, Met. We show that isotope labeling of Ala is improved using the transaminase inhibitor L-cycloserine, and labeling of Thr is improved through addition of Cys and Met, which are known inhibitors of homoserine dehydrogenase. We demonstrate the creation of long-lived 1H NMR signals in most amino acid residues using our model system, the WW domain of human Pin1, as well as the bacterial outer membrane protein PagP.

Chaperone Recycling in Late-Stage Flagellar Assembly.

The flagellum is a sophisticated nanomachine responsible for motility in Gram-negative bacteria. Flagellar assembly is a strictly choreographed process, in which the motor and export gate are formed first, followed by the extracellular propeller structure. Extracellular flagellar components are escorted to the export gate by dedicated molecular chaperones for secretion and self-assembly at the apex of the emerging structure. The detailed mechanisms of chaperone-substrate trafficking at the export gate remain poorly understood. Here, we structurally characterized the interaction of Salmonella enterica late-stage flagellar chaperones FliT and FlgN with the export controller protein FliJ. Previous studies showed that FliJ is absolutely required for flagellar assembly since its interaction with chaperone-client complexes controls substrate delivery to the export gate. Our biophysical and cell-based data show that FliT and FlgN bind FliJ cooperatively, with high affinity and on specific sites. Chaperone binding completely disrupts the FliJ coiled-coil structure and alters its interactions with the export gate. We propose that FliJ aids the release of substrates from the chaperone and forms the basis of chaperone recycling during late-stage flagellar assembly.