Structure–function analysis of an ancient TsaD–TsaC–SUA5–TcdA modular enzyme reveals a prototype of tRNA t6A and ct6A synthetases

Mengqi Jin; Zelin Zhang; Zhijiang Yu; Wei Chen; Xiaolei Wang; Dongsheng Lei; Wenhua Zhang

doi:10.1093/nar/gkad587

. 2023 Jul 10;51(16):8711–8729. doi: 10.1093/nar/gkad587

Structure–function analysis of an ancient TsaD–TsaC–SUA5–TcdA modular enzyme reveals a prototype of tRNA t⁶A and ct⁶A synthetases

Mengqi Jin ¹, Zelin Zhang ², Zhijiang Yu ³, Wei Chen ⁴, Xiaolei Wang ⁵, Dongsheng Lei ^6,^✉, Wenhua Zhang ^7,^✉

PMCID: PMC10484737 PMID: 37427786

Abstract

N ⁶-threonylcarbamoyladenosine (t⁶A) is a post-transcriptional modification found uniquely at position 37 of tRNAs that decipher ANN-codons in the three domains of life. tRNA t⁶A plays a pivotal role in promoting translational fidelity and maintaining protein homeostasis. The biosynthesis of tRNA t⁶A requires members from two evolutionarily conserved protein families TsaC/Sua5 and TsaD/Kae1/Qri7, and a varying number of auxiliary proteins. Furthermore, tRNA t⁶A is modified into a cyclic hydantoin form of t⁶A (ct⁶A) by TcdA in bacteria. In this work, we have identified a TsaD–TsaC–SUA5–TcdA modular protein (TsaN) from Pandoraviruses and determined a 3.2 Å resolution cryo-EM structure of P. salinus TsaN. The four domains of TsaN share strong structural similarities with TsaD/Kae1/Qri7 proteins, TsaC/Sua5 proteins, and Escherichia coli TcdA. TsaN catalyzes the formation of threonylcarbamoyladenylate (TC-AMP) using L-threonine, HCO₃⁻ and ATP, but does not participate further in tRNA t⁶A biosynthesis. We report for the first time that TsaN catalyzes a tRNA-independent threonylcarbamoyl modification of adenosine phosphates, leading to t⁶ADP and t⁶ATP. Moreover, TsaN is also active in catalyzing tRNA-independent conversion of t⁶A nucleoside to ct⁶A. Our results imply that TsaN from Pandoraviruses might be a prototype of the tRNA t⁶A- and ct⁶A-modifying enzymes in some cellular organisms.

Graphical Abstract

INTRODUCTION

So far, more than 150 non-canonical nucleobases have been identified in various RNAs in nature (1). Transfer RNAs (tRNAs) contain the largest variety of these chemically distinct nucleosides (1–4). N⁶-threonylcarbamoyladenosine (t⁶A) is a threonine-modified carbamoyl adenosine universally found at position 37 of tRNAs that decipher ANN-codons (N being A, U, C or G) (5,6). t⁶A belongs to a core set of 18 ‘universal’ non-canonical nucleosides that are found in tRNAs throughout the three kingdoms of life (3,4). tRNAs play a pivotal role in protein translation, which is a central commonality of all cellular life. Structural studies have revealed that tRNA t⁶A37 interacts with its 5′-adjacent base U36, preventing the canonical Watson–Crick pairing between U33 and A37, and enhancing tRNA–mRNA–rRNA interactions in the ribosomal A-site (6,7). Biochemical analysis demonstrated that tRNA t⁶A is required for maintaining translational fidelity and proteostasis (7–9). The deficiency in tRNA t⁶A biosynthesis leads to cell death in unicellular organisms (10–12), and is also implicated in a number of human disorders (13), including the Galloway–Mowat syndrome (14,15).

t⁶A is deemed as one of molecular fossils of the life's origin and evolution (16,17). Indeed, Schneider et al. has reported a spontaneous chemical synthesis of t⁶A nucleoside using CH₄, HNCO, L-threonine and adenosine in a controlled prebiotic condition (17). However, there are currently no reports of enzyme-assisted threonylcarbamoyl modification occurring at the nucleotide or nucleoside level. In cellular organisms, the biosynthesis of tRNA t⁶A is catalyzed by members of two evolutionarily conserved protein families TsaC/Sua5 (COG0009) and TsaD/Kae1/Qri7 (COG0533) (10,11,18). They carry out tRNA t⁶A biosynthesis in two consecutive steps (Figure 1) (19): in the first step, TsaC/Sua5 protein utilizes L-threonine, CO₂/HCO₃⁻ and ATP to generate an intermediate threonylcarbamoyladenylate (TC-AMP) (20,21); in the second step, TC-moiety of TC-AMP is transferred onto N6 atom of tRNA A37 by TsaD/Kae1/Qri7 proteins on their own in eukaryotic mitochondria (21,22) or with support of a varying number of auxiliary proteins in bacteria, archaea and eukaryotic cytosols (19,23,24). In bacteria, TsaD is catalytically activated by TsaB and TsaE (20,25–29). TsaD–TsaB heterodimer interacts with TsaE in the presence of ATP, leading to an ATP-mediated formation of the ternary complex TsaD–TsaB–TsaE (26,27,29,30). The t⁶A-catalysis cycle of TsaD involves the dynamic assembly of TsaD–TsaB–TsaE complex that is regulated by TsaE-catalyzed hydrolysis of ATP (26,27,29). Small angle X-ray scattering (SAXS) model of TsaD−TsaB−tRNA complex and competitive binding assay demonstrated that tRNA and TsaE compete for the same binding site at the interacting interface between TsaD and TsaB (27,29). In archaea and eukaryotes (e.g. yeast and human), Kae1/OSGEP functions in a KEOPS complex that comprises other subunits Bud32/TP53RK, Cgi121/TPRKB, Pcc1/LAGE3 and Pcc2 or Gon7/GON7 (31–39). The docking model KEOPS−tRNA complex exhibits that Cgi121, Bud32 and Pcc1 are needed to help dock tRNA A37 into the active site of Kae1 (23,33), whose catalytic activity is directly regulated by the ATPase Bud32 (31,33,35). It appears that life has evolved diversified tRNA t⁶A biosynthetic systems and mechanisms in coping with the increasing biological complexity (19,23). However, how the prototypical tRNA t⁶A biosynthetic systems originated and evolved remains elusive.

Figure 1. — Diagram of the biosynthetic pathways of tRNA t⁶A and ct⁶A. In the first step, TsaC/Sua5 family protein utilizes L-threonine, HCO₃⁻ and ATP to generate an intermediate L-threonylcarbamoyladenylate (TC-AMP) and one molecule of pyrophosphate (PPi); in the second step, TsaD/Kae1/Qri7 family protein binds with TC-AMP and tRNA, and catalyzes the transfer of TC-moiety onto the N6 atom of tRNA A37 with support of auxiliary proteins, leading to tRNA t⁶A and releasing one molecule of AMP; in a subsequent step, TcdA catalyzes the conversion of tRNA t⁶A to ct⁶A at the cost of ATP hydrolysis into AMP and PPi.

Phylogenetic analysis shows that the single-domain TsaC/YRDC proteins are mostly distributed in bacteria, multicellular eukaryotes and a few archaea species, whereas Sua5 proteins that comprise a TsaC domain and a SUA5 domain are confined to archaea, unicellular eukaryotes and a few bacteria (24). The phylogenetic distribution pattern of TsaC/Sua5 members remains enigmatic. Crystal structures of E. coli TsaC (40), S. tokodaii Sua5 (41,42) and P. abyssi Sua5 (43) reveal that TsaC folds into a compact α/β twisted open-sheet ‘baseball glove’-like structure with a large positively charged concave cavity and the SUA5 domain adopts a reduced Rossmann fold consisting of only five adjacent β-sheets sandwiched by three α-helices. Biochemical analysis demonstrates that the SUA5 domain functions to promote the solubility and stability of the TsaC domain (43). TC-AMP catalytic site is formed principally in the TsaC domain (40–43) and also constitutes the C-terminal tail of TsaC/YRDC proteins (40,44) or the inter-domain linker (∼40 amino acids) connecting the TsaC domain and the SUA5 domain of Sua5 (43). Structure–function analysis reveals that a conserved tetrad motif KxR–SxN in the TsaC domain is essential for TC-AMP biosynthesis (41–45). The interdomain linker acts as a ‘gating’ loop to regulate the TC-AMP catalysis cycle of Sua5 (43). The mutation in the conserved motifs PGM or HY on the interdomain linker diminishes the catalytic activity of Sua5 (43). Biochemical and structural studies proposed a mechanistic model of TC-AMP biosynthesis (20,41–43,45): L-threonine first reacts with HCO₃⁻ to generate an intermediate N-carboxy-L-threonine. Then, the linker closes upon binding of ATP and creates a catalytically active conformation, leading to nucleophilic attack on the carboxyl group of N-carboxy-L-threonine. The activated N-carboxy-L-threonine subsequently condensates with ATP via an adenylation reaction (R–COOH + ATP→R–C = O–AMP + PPi) (20,45), leading to the generation of TC-AMP and pyrophosphate (PPi) (Figure 1).

Members of TsaD/Kae1/Qri7 family are encoded in genomes of almost all the cellular organisms (16,18,24). Crystal structures of E. coli TsaD (26), P. abyssi Kae1 (46), S. cerevisiae Qri7 (21) and H. sapiens OSGEP (15) reveal that TsaD/Kae1/Qri7 proteins consist of two similar ancestral oligonucleotide-binding sub-domains composed of five β-sheets flanked by three α-helices (15,19,21,26,29,46). TsaD/Kae1/Qri7 proteins adopt a conserved bi-lobal fold and differ in local variations incurred by insertion or deletion of connecting loops (15,19,21,26,46). In particular, variation in α1/α2 and α2/α3 of TsaD/Kae1/Qri7 proteins renders different types of interfacial interactions that are mediated by a four-helical bundle (19). ATP, ADP and AMP are bound in coordination with Fe²⁺ or Zn²⁺ ion in the nucleotide-binding site in the cleft between the two sub-domains of P. abyssi Kae1 (46), E. coli TsaD (26) and S. cerevisiae Qri7 (47), respectively. E. coli TsaD and P. abyssi Kae1 possess weak activities in hydrolyzing ATP into ADP (26,28). S. cerevisiae Kae1 is also active in hydrolyzing ADP and GDP into adenosine and guanosine, and PPi, respectively (48). However, ATP, ADP or AMP bound in the nucleotide-binding site of TsaD/Kae1/Qri7 proteins are not chemically required for tRNA t⁶A biosynthesis (Figure 1). Crystal structures of E. coli TsaD in complex with BK951 (a TC-AMP mimic) (49) and T. maritima TsaD in complex with carboxy-AMP (29) revealed that the ATP-binding site in TsaD/Kae1/Qri7 proteins is a bona fide TC-transfer catalytic site. A highly conserved HxxxH−D motif in TsaD/Kae1/Qri7 proteins functions to coordinate a divalent metal ion (Zn²⁺ or Mg²⁺) and the phosphonate oxygen of TC-AMP (29,49). Mutation of the HxxxH−D motif abolishes the t⁶A-catalytic activity of TsaD/Kae1/Qri7 proteins (11,21,26,31,35). So far, it is unknown how A37 of tRNA is docked into the catalytic site to accept the TC-moiety of TC-AMP.

tRNA t⁶A can be enzymatically converted to a cyclic hydantoin form of t⁶A (ct⁶A), which is found only in a few species of bacteria, fungi and plants (50,51). In E. coli, the formation of tRNA ct⁶A is catalyzed by TcdA/CsdL (50), which probably requires additional support of CsdE and CsdA (50,52,53). In vitro assay demonstrated that TcdA alone is active in catalyzing the conversion of tRNA t⁶A to tRNA ct⁶A in the presence of ATP (Figure 1) (50). Knockout of tcdA gene compromises the decoding efficiency and causes observable growth defects (50). Deletion of either TCD1/YHR003C or TCD2/YKL027W (homologs of E. coli tcdA) led to loss of tRNA ct⁶A and compromised the respiratory growth of S. cerevisiae under non-fermenting conditions (50). Crystal structures revealed that E. coli TcdA folds into a two-domain globular structure consisting of seven parallel β-strands in a continuous sheet surrounded by eight α-helices (53,54). Structure of the N-terminal domain of TcdA resembles the ATP-binding Rossmann fold of ThiF/MoeB/E1 superfamily, whose catalytic functions depend on a classical adenylation reaction (53,55,56). Binding of ATP/AMP in the catalytic site of TcdA is coordinated by a number of conserved residues that include Arg⁷² and Lys⁸⁵ (53,54). Arg⁷² stabilizes the β- and γ-phosphates of ATP; Lys⁸⁵ directly interacts with β-phosphate and stabilizes α-phosphate via hydrogen-bonding with one water molecule. The C-terminal domain of TcdA adopts a novel fold comprising three β-strands flanked by three α-helices, which mediates the homodimerization of TcdA (53,54). TcdA forms a symmetrical homodimer with six α-helices from both domains. The binding assay demonstrated that the C-terminal domain of TcdA is involved in the interaction between TcdA and tRNA (54). A SAXS model of TcdA−tRNA complex shows that two molecules of tRNA are independently bound to the positively-charged outer rims of the TcdA dimer (53).

Comparative genetic analysis revealed that a TsaC-like domain and a Kae1-like domain are fused in a few bacteria (57). In S. tenebrarius TobZ (45) and C. subterraneu HypF (58), a TsaC-like domain and a Kae1-like domain cooperate to catalyze an ancient carbamoylation reaction that is analogous to the threonylcarbamoylation reaction during tRNA t⁶A biosynthesis (19,45). Crystal structures of TobZ and HypF reveal that the inter-domain interaction creates a channel for efficient delivery of the intermediate carbamoyladenylate (CA) from the TsaC-like domain to the Kae1-like domain (45,58). How the chemically unstable TC-AMP generated by TsaC/Sua5 proteins is efficiently delivered to the catalytic sites of TsaD/Kae1/Qri7 proteins remains unknown. It was suggested that TsaC/Sua5 proteins function in spatial proximity to TsaD/Kae1/Qri7 proteins via transient interactions to overcome the limitation of diffusion-based delivery of TC-AMP (20,25). In this study, we searched against NCBI nr database for encoded proteins that include at least a TsaC/Sua5-like domain and a TsaD/Kae1/Qri7-like domain. We identified genes encoding a consecutive fusion of a TsaD-like domain, a TsaC-like domain, a SUA5-like domain and a TcdA-like domain in a number of Pandoraviruses genomes. We characterized the structure and function of P. salinus YP_009430141.1 (59) and propose to rename it as TsaN. Our 3.2 Å resolution cryo-EM structure of P. salinus TsaN manifests that the four domains of TsaN share strong structural similarity with TsaD/Kae1/Qri7 proteins, TsaC/Sua5 proteins and E. coli TcdA. We demonstrate that TsaN catalyzes TC-AMP formation but does not participate further in tRNA t⁶A biosynthesis. We discovered that TsaN catalyzes the threonylcarbamoyl modification of ATP and ADP, but neither AMP nor A, using TC-AMP as the TC-moiety donor. Furthermore, we show that TsaN is highly active in catalyzing tRNA-independent conversion of synthetic t⁶A to ct⁶A. Our work lays foundation for future in-depth analysis on the evolutionary origin and catalytic mechanisms of tRNA t⁶A- and ct⁶A-modifying enzymes.

MATERIALS AND METHODS

Bioinformatics

We performed sequence alignment of the four domains of P. salinus TsaN (YP_009430141.1) using Clustal Omega (60). The TsaD domain (amino acids 30−392) was aligned to representative sequences of E. coli TsaD, T. maritima TsaD, P. abyssi Kae1, S. cerevisiae Kae1, S. cerevisiae Qri7, H. sapiens OSGEP and H. sapiens OSGEPL1; the TsaC domain (amino acids 405−613) and SUA5 domain (amino acids 614−751) were aligned to representative sequences of E. coli TsaC, T. maritima Sua5, S. tokodaii Sua5, P. abyssi Sua5, S. cerevisiae Sua5, and H. sapiens YRDC; the TcdA domain (amino acids 752–1008) was aligned to the sequences of E. coli TcdA, S. cerevisiae Tcd1 and Tcd2, and A. thaliana AT5G37530. The NCBI Standard Protein BLAST was performed against the NCBI nr database to identify orthologs of the four domains of TsaN in virus, bacteria, archaea and eukarya. The sequence identities were calculated with Clustal Omega.

Cloning, mutagenesis and expression procedures

P. salinus TsaN (YP_009430141.1) gene was designed with codons optimized for E. coli expression and chemically synthesized (Tsingke, Beijing, China). The DNA was cloned into a pET24a vector using restriction sites of NdeI and XhoI, yielding a plasmid named pET24a−TsaN that expresses a recombinant full-length TsaN with a six-histidine (6His) tag added at the C-terminus. Sub-cloning and site-directed mutagenesis were performed with PCR using primers listed in Supplementary Table S2, following the manufacturer's protocols (Takara, Beijing, China). The expression plasmids of the TsaD domain (TsaN¹⁻³⁹²), TsaD−TsaC−SUA5 domain (TsaN¹⁻⁷⁴⁸) and TcdA domain (TsaN^749−1008) were named as pET24a−TsaN^1–392, pET24a−TsaN¹⁻⁷⁴⁸ and pET24a−TsaN^749−1008, respectively, to which 6His tag was added at the C-terminus for purification. The gene of E. coli TcdA was amplified from E. coli K12 cells by PCR using primers (Supplementary Table S2) and cloned into pET21a vector using restriction sites of NdeI and XhoI, generating an expression plasmid named pET21a−EcTcdA with a 6His tag added at its C-terminus.

For expression of TsaN, E. coli BL21 (DE3) AI cells were transformed with plasmids of pET24a–TsaN and cultured at 37°C in 2YT medium supplemented with 50 μg/ml kanamycin. The overexpression, upon the OD₆₀₀ of cell culture reached 0.6–0.8, was induced with addition of 0.5 mM IPTG and 2 g/l arabinose at 16°C for 18 h. Expression of TsaN^749−1008 was carried out in E. coli Rosetta (DE3) at 37°C for 3 h, following induction with addition of 0.5 mM IPTG at OD₆₀₀ of 0.6−0.8. The expression of mutants of each protein was carried out using procedures as for the wild type protein. Expression of yeast KEOPS complex and yeast Sua5 was carried out with E. coli BL21 (DE3) AI cells transformed with pJ241−ScKEOPS (61) and pET21a−ScSua5 (28), respectively. Cells were incubated with 0.5 mM IPTG and 2 g/l arabinose and were grown in 2YT medium at 16°C for 16 h. E. coli TcdA was expressed using E. coli BL21 (DE3) cells grown in LB medium at 22°C for 18 h. Cells were harvested by centrifugation and suspended in Buffer A (20 mM Tris−HCl pH 7.5, 300 mM NaCl and 5 mM β-mercaptoethanol). Cells were first lysed by sonication and centrifuged at 15 000 rpm for 30 min. The supernatant was applied to immobilized metal affinity chromatography using Ni-NTA resins, followed by protein elution using Buffer A supplemented with increasing concentrations of imidazole (10−400 mM). The protein-containing fractions were analysed by SDS−PAGE and applied to purification by size-exclusion chromatography (SEC) using column of HiLoad 16/600 Superdex 200 prep grade or HiLoad 16/600 Superdex 75 prep grade (GE Healthcare) equilibrated with Buffer A. Protein-containing fractions were analysed by SDS−PAGE, followed by centrifugal concentration using centrifugal filters (Amicon). Protein concentration was determined by BioPhotometer D30 (Eppendorf) or NanoDrop 2000 (ThermoFisher Scientific), followed by immediate use or flash-freezing in liquid nitrogen for storage at −80°C.

Cryo-EM sample preparation and data collection

3.5 μl freshly purified TsaN (0.5 mg/ml) was added to a glow-discharged Quantifoil Au R2/2 grid (Quantifoil Micro Tools). The grid was blotted for 5 s and then plunged into pre-cooled liquid ethane using a Vitrobot mark IV (ThermoFisher Scientific). The chamber of Vitrobot was set to 100% humidity at 4°C. All the cryo-EM data were collected on a Thermo Fisher Scientific Titan Krios G3i microscope equipped with a Gatan Bioquantum K3 direct electron camera. EPU (ThermoFisher Scientific) was used to collect super-resolution movies (30 frames, 3 seconds exposure) with a total dose of ∼60 e⁻/Å², a pixel size of 0.41 Å, and a defocus of −2.0 μm to −0.6 μm.

Cryo-EM data processing

The frames of each collected raw movie were subjected to motion correction by using MotionCor2 (62), and then CTF estimation by using CTFFIND4 (63). The CTF fitting of each micrograph was examined and screened based on Thon ring. All subsequent image processing was performed by using CryoSPARC (64,65). About 100 particles were manually picked for 2D classification, and the best classifications were used as templates for automatic particle picking. A total of 700 000 particles were picked and subjected to multiple rounds of 2D classification and 3D classification. Finally, particles from two best 3D classes were selected and subjected to 3D auto-refinement separately. One of the refined reconstructions (115 363 particles) showed a symmetric dimer structure. However, it was significantly blurred at two ends with obvious missing densities, thus was not suitable for model building. The other reconstruction (152 435 particles) was quite similar to half of the first one, but showed more complete density at one end. This reconstruction was used for subsequent model building. The resolution for all reconstructions were evaluated by using the gold standard Fourier Shell Correlation (FSC) of 0.143.

Model building, refinement and validation

The 3D reconstruction showed presence of two classes of particles, one for symmetrical TsaN dimer and the other for incomplete TsaN dimer with unsymmetrical shape. The obtained electron density map of the complete TsaN dimer show significant density absence at two far ends, and only the density map of the incomplete TsaN dimer allowed us to build a model. It should be noted that, this model comprises not just a TsaN monomer, but also a complete dimerizing TcdA protomer. Besides, it fits well with the density map of the complete TsaN dimer, with an overall correlation of over 0.73 and a local correlation nearly 0.8 within the TcdA protomer (measured via Chimera). Therefore, although this model was obtained from reconstruction of incomplete TsaN dimer particles, it could represent the conformation and structural features of TsaN upon dimerization. This model was finally used to construct the complete TsaN dimer model based on symmetry of the TcdA protomer. Crystal structures of E. coli TsaD (PDB: 4YDU), the TsaC domain and the SUA5 domain of P. abyssi Sua5 (PDB: 6F89), and E. coli TcdA (PDB: 4RDI) were rigid-body fitted into the cryo-EM map using Phenix (66). The complete model was improved with real-space refinement using Phenix, followed by manual building with COOT (67). The final model was validated by MolProbity (68), and the refinement statistics are presented in Table 1. The structure alignment was performed with COOT and PyMOL (Schrödinger, LLC). The visualization and graphic representation of the structures were generated using PYMOL and UCSF Chimera (69). The electrostatic potential molecular surface was generated with Adaptive Poisson–Boltzmann Solver (70). The interface, buried area and presence of salt bridges between interacting domains of TsaN were assessed and calculated by PISA (71).

Table 1.

Cryo-EM data collection, processing, refinement and validation statistics

P. salinus TsaN (EMD-35365, PDB: 8IDE)
Data collection
Electron microscope	Thermo Fisher Titan Krios G3i
Camera	Gatan Bioquantum K3
Magnification	×105 000
Pixel size (Å/pix)	0.41
Defocus range (μm)	−2.0 to −0.6
Total electron dose (e-/Å²)	60
Total micrographs (no.)	7263
Reconstruction
Software	cryoSPARC
Particles for 2D classification	700 000
Particles for 3D classification	490 000
Particles in the final map (no.)	152 435
Symmetry	C1
Final resolution (Å)	3.21
FSC threshold	0.143
Map sharpening B factor (Å²)	−82.8
Model Building and validation
MolProbity score	2.31
Clash score	14.03
Sidechain outliers (%)	1.1
R.M.S.D. of bond lengths (Å)	0.003
R.M.S.D. of angles (^∘)	0.726
Ramachandran plot
Favored (%)	85.38
Allowed (%)	14.10
Outliers (%)	0.52

Open in a new tab

SEC coupled multi-angle light scattering (SEC-MALS)

100 μl protein (1.25 mg/ml) was chromatographed by Superdex 200 10/300 GL column (GE Healthcare) using a HPLC system (GE Healthcare, USA) connected in-line with a DAWN HELEOS II (Wyatt Technology, Santa Barbara, CA, USA) eight-angle light-scattering detector, followed by a refractive-index detector (Wyatt Technology, Santa Barbara, CA, USA). SEC-MALS system was equilibrated with Buffer A and was run at a flow rate of 0.75 ml/min. Astra software (Wyatt Technology, version 7.1.3) was used to collect and determine the absolute molar mass of each protein sample.

Preparation of tRNA

For in vitro tRNA transcription, the template DNAs for P. salinus tRNA-Met (gene ID: 16607007), P. salinus tRNA-Pro (gene ID: 16605227) and P. salinus tRNA-Trp (gene ID: 16606877) were constructed by PCR using synthetic DNAs (Tsingke, Beijing, China, Supplementary Table S2). The transcription was carried out at 30°C for 6 h in a reaction mixture consisting 40 mM Tris−HCl pH 8.0, 4 mM NTP mix, 15 mM GMP, 5 mM DTT, 2 mM spermidine, 0.5% Triton X-100, 33 mM MgCl₂, 5 μM T7 RNA polymerase and 5 μM pyrophosphatase. In vitro transcripts were separated by 10% denaturing Urea−PAGE, followed by visualization with UV light. The shadowed bands were cut out and eluted from the gel slices by crush and soak in elution buffer (0.5 M NaAc pH 5.2, 2 mM EDTA, 0.5% SDS) at 37°C overnight, followed by precipitation with addition of 3 volumes of 100% ethanol and cooling at −80°C for 2 h. The tRNA transcript was pelleted by centrifugation at 12 000 rpm for 15 min at 8°C and resuspended in buffer containing 20 mM Tris−HCl pH 8.0, 100 mM KCl and 5 mM MgCl₂. tRNA was unfolded by heating up to 95°C for 5 min and gradually annealed by cooling down in water bath to room temperature. The correct folding of tRNA was confirmed by Circular Dichroism following the manufacture's protocol (Jasco Corporation, Japan). The concentration of tRNA was determined by measuring the absorbance at 258 nm using a NanoDrop 2000.

Yeast bulk tRNAs devoid of t⁶A were extracted from S. cerevisiae sua5Δ strain cells (72). Bacterial bulk tRNAs were extracted from E. coli K12 cells. S. cerevisiae sua5Δ strain cells were grown to supersaturation in YPD medium at 30°C for 72 h, followed by harvest with centrifugation and lysis by grinding with liquid nitrogen. E. coli K12 cells were grown in LB medium at 37°C for 16 h. Total RNAs were extracted with TRIzol (ThermoFisher Scientific) and precipitated using absolute ethanol. Bulk tRNAs were separated and purified by 10% denaturing Urea−PAGE, followed by precipitation, recovery and refolding procedures as that for in vitro transcription.

Chemical synthesis of t⁶A

t⁶A compound was chemically synthesized following the protocol by Schneider, et al. (17).

Enzymatic assays for biosynthesis of TC-AMP, t⁶A and ct⁶A

TC-AMP assay was performed at 30°C for 1 h in a 50 μl reaction mixture containing 5 μM enzyme (P. salinus TsaN or S. cerevisiae Sua5), 1 mM ATP, 4 mM L-threonine and 10 mM NaHCO₃ in assay buffer B (20 mM Tris−HCl pH 8.0, 300 mM NaCl, 1 mM MgCl₂, 1 mM MnCl₂ and 1 mM DTT). Reconstitution of tRNA t⁶A was performed with TC-AMP assay supplemented with 100 μM tRNA and 5 μM protein (P. salinus TsaN or S. cerevisiae KEOPS). In order to remove the reactants, tRNA in reaction mixture was further purified using 10% Urea−PAGE, precipitated and recovered following the protocols used for preparation of tRNAs. The digestion of tRNA into nucleotides was carried out with 0.1 U/μl nuclease P1 (Sigma) for 1 hour at 37°C and the dephosphorylation of nucleotides was carried out with 0.05 U/μl alkaline phosphatase (Sigma) overnight at 37°C. The enzymatic assay of ct⁶A was performed by incubating 5 μM P. salinus TsaN (or E. coli TcdA) with 0.5 mM synthetic t⁶A, 0.5 mM ATP, 1 mM MgCl₂ and 1 mM MnCl₂ for 1 h at 30°C.

Preparation of TC-AMP compound and nucleotide threonylcarbamoylation assay

For preparation of TC-AMP compound, multiple injections of 100 μl TC-AMP reaction mixtures were chromatographed by HPLC using a Hypersil GOLD column (5 μm; 250 × 4.6 mm, ThermoFisher Scientific). The collected samples were immediately frozen in liquid nitrogen and subsequently subjected to freeze-drying in a lyophilizer (Beijing Sihuan Technology). The white-coloured compound was stored at –20°C. TC-AMP compound was dissolved to a final concentration of 25 μM in assay buffer B containing 2 μM TsaN¹⁻³⁹² and 1 mM ATP, ADP, AMP or adenosine (A) for 1 h at 25°C. The reaction mixture was treated with 0.05 U/μl alkaline phosphatase (Sigma).

Analysis of the formation of t⁶ADP and t⁶ATP in bacterial cells

E. coli BL21 (DE3) AI cells transformed with pET24a−TsaN or pET24a−TsaN^H155A/H159A were induced by IPTG at an OD₆₀₀ of 0.6 and grown for up to 3 h. The expression of TsaN and TsaN^H155A/H159A was determined by western blot against 6His tag using primary anti-6xHis Tag mouse monoclonal antibody (Sangon Biotech, Shanghai) and secondary HRP-conjugated Goat Anti-Mouse IgG (Sangon Biotech, Shanghai). Cells from 5 ml cell culture were centrifuged and resuspended in 1 ml deionized water, followed by sonification lysis. The lysate was centrifuged at 15 000 rpm for 10 min at room temperature. The supernatant was filtered by a centrifugal filter (3 kDa cut-off, Amicon) to remove soluble macromolecules. The filtrate was enriched by freeze-drying in a lyophilizer (Beijing Sihuan Technology), followed by addition of 20 μl deionized water. The suspension solution was directly subjected to LC–MS analysis or treated with 0.05 U/μl alkaline phosphatase (Sigma) before LC–MS analysis.

Liquid chromatography−mass spectrometry analysis

The mixture of digested nucleosides was subjected to liquid chromatography–mass spectrometry (LC–MS) using an LCQ-Advantage ion trap mass spectrometer (ThermoFisher Scientific), equipped with an electrospray ionization source and an HP1100 LC system (Agilent Technologies). For LC, 10 μl sample of the digested nucleosides was chromatographed with a C18 column (5 μm; 250 × 4.6 mm, Ecosil) and eluted with 1‰ trifluoroacetic acid (solvent A) and acetonitrile containing 1‰ trifluoroacetic acid (solvent B) at a flow rate of 0.8 ml/min with a multistep gradient: 2% solvent B for 5 min, 2–20% solvent B for 15 min, 20−98% solvent B for 7 min and then initialized with 98% solvent B for 2 min. The presence of nucleosides was monitored by recording UV absorbance at 254 nm. The retention time and mass-to-charge ratio of chemicals were collected using OpenLab CDS ChemStation Edition Online (Agilent). The UV peak areas at 254 nm corresponding to specific nucleotides or nucleosides were integrated, summed and normalized to standard areas of internal references. To determine the chemical identities of t⁶ADP and t⁶ATP, a reaction mixture containing 10 mM L-threonine, 15 mM NaHCO₃, 2 mM ATP and 10 μM TsaN was incubated for 6 h at 25°C. The mixture was directly chromatographed in reverse phase with a C18 column (1.9 μm, 150 × 2.1 mm; ThermoFisher Scientific) and analysed by Q Exactive Hybrid Quadrupole-Orbitrap mass spectrometer (ThermoFisher Scientific) with a linear gradient of 100:0 to 2:98 of water containing 1‰ formic acid and methanol for 50 minutes. The acquired UPLC–MS data were analysed using Xcalibur 4.5 (ThermoFisher Scientific). The UPLC–MS chromatographic profile of chemical species was generated by searching the theoretic molecular weight (m/z).

Isothermal titration calorimetry

Isothermal titration calorimetry (ITC) measurements were performed with MicroCal PEAQ-ITC. Routinely, 30 μM proteins and 360 μM ligands were equilibrated in Buffer A. For each measurement, the protein was titrated with a total of 19 injections of 2 μl ligand at intervals of 180 seconds under continuous stirring at 25°C or 20°C. A single-site binding model was fit to the data by a nonlinear regression analysis. The stoichiometry (N), binding affinity (K_D), enthalpy (ΔH) and change in free energy (ΔG) of the protein and ligand were analysed using MicroCal PEAQ-ITC Analysis Software.

ATPase assay

The ATPase activity quantification was carried out using a nicotinamide adenine dinucleotide dehydrogenase (NADH)-coupled assay, in which the hydrolysis of ATP to ADP was coupled to the oxidation of NADH to NAD⁺, following the protocols previously reported (26). 5 μM TsaN or variant was added in 200 μl reaction mixture buffer containing 100 mM Tris−HCl pH 7.5, 100 mM KCl, 10 mM MgCl₂, 0.5 mM NADH (Sigma), 4 mM phospho(enol)pyruvic acid (PEP) and 1 mM pyruvate kinase/lactate dehydrogenase (PK/LD) and varying concentrations of ATP. Assays were performed in a 96-well plate at 25°C. The consumption of NADH was monitored by measuring the absorbance at 340 nm at an interval of 30 s by a spectrophotometer (Multiskan GO). Standard curves were plotted for ADP and the measurements were performed independently in triplicates.

Electrophoretic mobility shift assays

Gel retardation assay on the interaction between tRNAs and proteins was performed with 2% agarose native gel in buffer C composed of 20 mM Tris−base pH 9.6 and 50 mM glycine. 20 μg bulk tRNAs and equimolar proteins (170 μg TsaN, 65 μg TsaN¹⁻⁷⁴⁸, 96 μg ScKEOPS, 24 μg TsaN^749−1008, or 24 μg EcTcdA) were preincubated in buffer supplemented with 25% glycerol for 10 min before loaded on the gel. Migration was run at constant current of 100 V for 60 min at 4°C in prechilled buffer C. Protein bands were stained with Coomassie Brilliant Blue, and tRNA bands were stained with Ethidium Bromide and visualized under UV at 254 nm.

Pyrophosphate assay

The amount of inorganic pyrophosphate was determined using the Pyrophosphate assay kit (MAK168, Sigma), following the manufacturer's protocol. 50 μl reaction mixture was incubated with 50 μl of sensor reagent in a 96 well plate for 30 min at room temperature. The fluorescence intensity was measured (λ_ex = 316 nm and λ_em = 456 nm) with Varioskan Flash (ThermoFisher Scientific). All measurements were performed in triplicates and with appropriate control.

RESULTS

TsaD–TsaC–SUA5–TcdA fusion protein encoded in Pandoravirus genomes

Prompted by the modular fusion and functional association of a TsaC-like domain and a Kae1-like domain in TobZ (45) and HypF (58), we searched through the genomes for encoding proteins with fusion of a TsaC/Sua5-like domain and a TsaD/Kae1/Qri7-like domain. The search hit a Pandoravirus salinus YP_009430141.1 (59), which was previously annotated as a threonylcarbamoyl AMP synthase and renamed as TsaN in this study. TsaN contains four consecutive domains that share significant sequence similarities with TsaD/Kae1/Qri7 proteins, TsaC/Sua5 proteins and TcdA/Tcd1/Tcd2 proteins (Supplementary Figure S1A–D and Supplementary Table S1). Sequence alignment analysis reveals that the four domains of TsaN are widely distributed in the three kingdoms of life and a number of DNA viruses (Supplementary Table S1). However, a consecutive fusion of TsaD–TsaC–SUA5–TcdA is only encoded in 7 out of 20 sequenced genomes of Pandoraviruses (Figure 2 and Supplementary Table S1), which include P. salinus, P. belohorizontensis, P. celtis, P. dulcis, P. japonicus, P. macleodensis and P. quercus. Interestingly, UMO80163.1 (a putative TsaD-like protein) and UMO80162.1 (a putative TsaC–SUA5–TcdA-like protein) are separately encoded in the genome of P. aubagnensis (Figure 2 and Supplementary Table S1). The genomes of the remaining 12 species of Pandoraviruses do not encode any orthologs of the four domains of TsaN (Supplementary Table S1).

Figure 2. — Domain organization and distribution of TsaN orthologs in *Pandoraviruses* (*P. salinus*: YP_009430141.1; *P. belohorizontensis*: UMO79024.1; *P. celtis*: QBZ81682.1; *P. dulcis*: YP_008318818.1; *P. japonicus*: BCU03410.1; *P. macleodensis*: YP_009481386.1; *P. quercus*: YP_009483771.1; *P. aubagnensis*: UMO80163.1 and UMO80162.1). The TsaD domain, TsaC domain, SUA5 domain and TcdA domain of TsaN are shown in red, blue, yellow and purple, respectively. The conserved functional motifs HxxxH–D, KxR–SxN, HYxP and D–R–K are indicated in the TsaD domain, TsaC domain, SUA5 domain and TcdA domain, respectively.

Purification and characterization of recombinant TsaN and truncated domains

Based on the protein sequence alignment identities, TsaN is defined to be composed of an N-terminal TsaD-like domain (amino acids 1–392), a TsaC-like domain (amino acids 393–613), a SUA5-like domain (amino acids 614–748), and a C-terminal TcdA-like domain (amino acids 749–1008) (Figure 3A). We constructed expression plasmids and purified well-behaved recombinant proteins of full-length TsaN¹⁻¹⁰⁰⁸, TsaN¹⁻⁷⁴⁸ (TsaD−TsaC−SUA5 domain), TsaN¹⁻³⁹² (TsaD domain) and TsaN^749−1008 (TcdA domain) (Figure 3B and Supplementary Figure S2A−F). Meanwhile, we also constructed expression plasmids of TsaN^393−748 (TsaC−SUA5 domain) and TsaN^393−1008 (TsaC−SUA5−TcdA domain). Both the TsaC−SUA5 domain and the TsaC−SUA5−TcdA domain were well expressed but not soluble (data not shown), suggesting a role of the TsaD domain in promoting the solubility of the TsaC−SUA5 domain.

Figure 3. — Cryo-EM structure of TsaN. (A) Domain organization of TsaN. (B) SDS−PAGE analysis of recombinant proteins TsaN¹⁻¹⁰⁰⁸, TsaN¹⁻⁷⁴⁸, TsaN¹⁻³⁹² and TsaN^749−1008. (C) SEC-MALS analysis of TsaN, TsaN¹⁻⁷⁴⁸ and TsaN^749−1008. (D) Experimental cryo-EM density map coloured according to domain identity shown in (A). (E) Cartoon representation of the atomic model built into the electron density map with same colour and view as in (D). (F) Zoom-in stereo view of the interface between the TsaD domain and the TsaC−SUA5 domain. (G) The structural model of TsaN. Three catalytic sites in the TsaD domain, TsaC domain and TcdA domain are indicated in cyan. (H) The structure model of the TsaN dimer. The model was generated using the dimer of the TcdA domain as a superposing template.

Size-exclusion chromatography (SEC) profiles over protein purification suggest that TsaN exists as a dimer (Supplementary Figure S2A). To confirm the dimeric state of TsaN, we used SEC coupled to multi-angle light scattering (SEC-MALS) to determine the actual molar masses of TsaN, TsaN¹⁻⁷⁴⁸ and TsaN^749−1008 (Figure 3C). Our SEC-MALS analysis estimated a molar mass of 209.5, 82.8 and 57.7 kDa for full-length TsaN, TsaN^1–748 and TsaN^749−1008, whose theoretic molecular weight is 106.7, 78.9 and 28.7 kDa, respectively. Our SEC-MALS results demonstrated that both the full-length TsaN and the TcdA domain exist as dimers in solution while the TsaD−TsaC−SUA5 domain alone exits as a monomer. In addition, SEC profile of TsaN^1–392 shows the TsaD domain exists as a monomer (Supplementary Figure S2C).

The cryo-EM structure of TsaN

In order to perform the structure−function relationship analysis, we used single-particle cryo-electron microscopy (Cryo-EM) to determine the structure of TsaN. Fresh TsaN were initially screened by negative-staining EM, followed by optimization that yielded homogenous particles suitable for analysis by cryo-EM (Supplementary Figure S3). Over 7000 micrographs were collected from grids prepared with Quantifoil Au R2/2 grids. The micrograph showed presence of a dimer with a symmetric structure, mixed with monomer or incomplete dimeric structures (Supplementary Figure S3B). A total of 700 000 particles were picked and subjected to multiple rounds of 2D classification. 152 435 selected particles permitted a 3D reconstruction (Supplementary Figure S3C−F) and yielded a 3.21 Å resolution electron density (Figure 3D). We initially performed rigid fit of the map using structures of E. coli (Ec) TsaD (26), TsaC (40) and TcdA (54), and the SUA5 domain of P. abyssi (Pa) Sua5 (43), followed by manual building and refinement. The final built model contains one molecule of TsaN (32−1008) and one molecule of the TcdA domain (755−1005) from the dimerizing TsaN protomer (Figure 3E). The density for the part of TsaD−TsaC−SUA5 in protomer 2 was extremely weak probably due to structural damage at air-water interface during sample preparation. In addition, regions of 1−31, 77−91, 617−637, 696−700 and 952−958 were not modelled due to the absence of clear electron densities. The modelling, refinement and validation statistics are summarized in Table 1.

In the resolved structure of TsaN (Figure 3E), the TsaD domain interacts simultaneously with the TsaC domain and the SUA5 domain (Supplementary Figure S4A). Specifically, Asp³⁶⁸ of the TsaD domain interacts with Arg⁴¹⁸ and Arg⁴²² of the TsaC domain via electrostatic interactions; Val³⁷⁷Cys³⁷⁸Leu³⁷⁹ of the TsaD domain forms an anti-parallel β-strands with Lys⁶⁸⁶Val⁶⁸⁷Trp⁶⁸⁸Tyr⁶⁸⁹ of the SUA5 domain (Figure 3F). Interestingly, these residues are not conserved in their corresponding orthologs (Supplementary Figure S1). The TcdA domain is connected to the TsaD−TsaC−SUA5 domain via a short linker (Figure 3G), and its homodimerizing protomer makes few direct contacts with the SUA5 domain (Supplementary Figure S4A). The buried surfaces of TsaD−TsaC, TsaD−SUA5, TsaC−SUA5, TcdA₁−TcdA₂, and SUA5−TcdA₂ are 113, 628, 723, 3162 and 219 Å², which account for 1.1%, 6.0%, 9.3%, 25.0% and 2.3% of the total surface areas of the two interacting subunits, respectively.

The final model of TsaN monomer is presented in Figure 3G. The structural comparison suggests that the four domains of TsaN share strong structural similarities with TsaD/Kae1/Qri7 proteins (Supplementary Figure S4B), TsaC/Sua5 proteins (Supplementary Figure S4C and D) and EcTcdA (Supplementary Figure S4E). The four domains of TsaN can be structurally aligned to EcTsaD, EcTsaC, the SUA5 domain of StSua5, and EcTcdA, giving RMSD (root-mean-square deviation) values of 1.5, 2.1, 2.7 and 1.0 Å, respectively (Table 2). Upon structural juxtaposition, the catalytic sites of TsaD/Kae1/Qri7 proteins, TsaC/Sua5 proteins and EcTcdA are projected onto TsaN, showing that three putative catalytic sites of TsaN are distantly located (Figure 3G). Moreover, the complete structure of the TcdA domain dimer is perfectly matched to the EcTcdA dimer (Supplementary Figure S4F), with an RMSD value of 1.34 Å (348 C_α). We built a model of TsaN dimer using the dimer of the TcdA domain (Figure 3H). The model and symmetry were in agreement with the 3D electron density map (Supplementary Figure S3C). The model of TsaN dimer shows that except the dimerizing interface of the TcdA domain, only the TcdA domain from one protomer has limited interaction with the SUA5 domain from the other protomer, and vice versa.

Table 2.

Summary of the alignment statistics of the four domains of TsaN. Sequence identities and RMSD values of C_α atoms are given. The aligned structures are presented in Supplementary Figure S4

TsaN domain	Comparing protein (PDB ID)	Sequence identity (%)	RMSD (Å)
TsaD (32–392)	E. coli TsaD (4YDU)	32.6%	1.500 (205 C_α)
	P. abyssi Kae1 (2IVN)	24.7%	1.696 (194 C_α)
	S.cerevisiae Qri7 (4K25)	30.9%	1.584 (205 C_α)
	H. sapiens OSGEP (6GWJ)	22.1%	2.294 (204 C_α)
TsaC (412–616)	E. coli TsaC (1HRU)	26.4%	2.101 (149 C_α)
	TsaC domain of P. abyssi Sua5 (6F87)	49.7%	1.657 (158 C_α)
	TsaC domain of S. tokodaii Sua5 (3AJE)	42.5%	1.438 (152 C_α)
SUA5 (638–751)	SUA5 domain of P. abyssi Sua5 (6F87)	37.5%	2.733 (82 C_α)
	SUA5 domain of S. tokodaii Sua5 (3AJE)	29.6%	2.679 (70 C_α)
TcdA (752–1008)	E. coli TcdA (4D79)	32.5%	0.998 (157 C_α)

Open in a new tab

Structural superposition of the TsaD domain of TsaN and the Kae1-like domain of TobZ shows that the TsaC domain of TsaN is located at an opposite position (Supplementary Figure S4G). Likewise, the relative location of the TsaC domain to the TsaD domain in TsaN does not conform to that of the TsaC-like domain to the Kae1-like domain in HypF (Supplementary Figure S4H). In both TobZ (45) and HypF (58), the TsaC-like domain is located near the cleft between the two subdomains of the Kae1-like domain, allowing formation of a 15 Å-long channel between two catalytic sites. In contrast, the TsaC domain is located at the extremity of the N-terminal subdomain of the TsaD domain in TsaN, manifesting a completely different spatial organization of the TsaD domain and the TsaC domain. The cryo-EM structure of TsaN revealed no presence of a channel among the three projected catalytic sites, which are located more than 60 Å away from each other (Figure 3G).

Enzymatic synthesis of TC-AMP by TsaN

Sequence alignment and structural comparison suggest that a putative TC-AMP biosynthetic center resides in the TsaC domain of TsaN (Figure 4A), including the catalytic motif K⁴⁵³xR⁴⁵⁵−S⁵⁴⁶xN⁵⁴⁸ that is conserved in all TsaC/Sua5 proteins (Supplementary Figure S1B) (10). However, structural juxtaposition of the TsaC domains of TsaN and Sua5 proteins shows that their SUA5 domains are distinctly positioned in relation to the corresponding TsaC domains (Figure 4B). Whereas the SUA5 domain is located in proximity to the catalytic site on the ‘palm’ side of the TsaC domain in Sua5 proteins, it is positioned far away from the active site on the other side of the TsaC domain in TsaN (Figure 4B). The interdomain linker (amino acids 617−637) between the TsaC domain and the SUA5 domain of TsaN was not modelled due to the lack of clear electron density. Sequence alignment shows that the interdomain linkers in TsaN and Sua5 proteins are conserved (Supplementary Figure S1C), including two functional motifs PGM and HYxP (43). The structure of the TsaC−SUA5 domain suggests that the interdomain linker between the TsaC domain and the SUA5 domain is proximal to the TC-AMP binding site and might participate in TC-AMP catalytic function of TsaN (Figure 4B).

We determined that TsaN actively catalyzes TC-AMP formation using L-threonine, NaHCO₃ and ATP (Figure 4C). The catalytic activity of TsaN and TsaN¹⁻⁷⁴⁸ is comparable to that of ScSua5 (Figure 4C), suggesting that the absence of the TcdA domain does not affect the catalytic activity of the TsaC−SUA5 domain. The structural juxtaposition suggests that Arg⁴⁵⁵ likely plays a role in the adenylation reaction (Figure 4A). To confirm the catalytic function of Arg⁴⁵⁵ in the KxR−SxN motif (Figure 4A), we chose to substitute Arg⁴⁵⁵ and generated its variant TsaN^R455A, which was purified in good quality (Supplementary Figure S2A and B). The LC–MS analysis on the TC-AMP formation demonstrated that TsaN^R455A was inactive in catalyzing TC-AMP biosynthesis (Figure 4D). Moreover, to determine the role of Tyr⁶⁴⁰ in the HYxP motif that belongs to the interdomain loop (Figure 4B), we substituted Tyr⁶⁴⁰ and generated its variant TsaN^Y640A (Supplementary Figure S2A and B). The variant TsaN^Y640A exhibited no measurable activity in generating TC-AMP (Figure 4D). Besides, a large quantity of AMP was produced during TC-AMP generation (Figure 4C). Notably, a great amount of AMP is still produced in the assays using catalytically dead variants TsaN^R455A and TsaN^Y640A (Figure 4D), suggesting that Arg⁴⁵⁵ and Tyr⁶⁴⁰ are not essentially involved in ATP hydrolysis.

Characterization of the nucleotide binding property of the TsaD domain

The TsaD domain of TsaN strikes great similarities with TsaD/Kae1/Qri7/OSGEP proteins, such as the bi-lobal architecture and a nucleotide-binding Rossmann fold. Clearly defined electron density allowed us to model the H¹⁵⁵xxxH¹⁵⁹−D³⁵⁰ triad motif in the cryo-EM structure of TsaN (Figure 5A). In structures of TsaD/Kae1/Qri7/OSGEP proteins (19,21,26,29,46), HxxxH−D motif is involved in coordinating the divalent metal ions (Fe²⁺, Zn²⁺ or Mg²⁺ ion) and nucleotides (ATP, ADP or AMP). Based on the essentiality of MnCl₂ for the activity of the TsaD domain, we modelled Mn²⁺ in the place of metal ions (Figure 5A). Juxtaposition of the TsaD domain of TsaN, PaKae1 (46) and EcTsaD (49) revealed a conserved nucleotide binding site in the TsaD domain of TsaN, which could accommodate one molecule of ATP (Figure 5B) or TC-AMP (Figure 5C). We used isothermal titration calorimetry (ITC) to investigate the nucleotide-binding properties of the TsaD domain (TsaN¹⁻³⁹²). The titration profiles are presented in Supplementary Figure S5, and the thermodynamic parameters are summarized in Table 3. The ITC data demonstrate the binding of ATPγS (a non-hydrolyzable analogue of ATP), ADP and AMP with TsaN¹⁻³⁹², with a K_D value of 10.90, 5.55 and 14.50 μM, respectively. In contrast, the titration of adenosine (A) to the TsaD domain did not produce any binding curves. We then tested the ATP hydrolysis activity of TsaN¹⁻³⁹² using an NADH-coupled ATPase assay. Using standard curves titrated with ADP, we determined that TsaN and TsaN¹⁻³⁹² exhibit a comparable ATPase activity with an apparent K_m and V_Max of about 40 μM and 0.015 μM/s, respectively (Figure 5D). We simultaneously mutated His¹⁵⁵ and His¹⁵⁹, and generated a variant TsaN^H155A/H159A, which was well purified (Supplementary Figure S2A and B). As expected, the production of ADP was not detected in the ATPase assay using TsaN^H155A/H159A (Figure 5D). We then generated a variant of the TsaD domain–TsaN^{(1−392)H155A/H159A} (Supplementary Figure S2C and D). ITC assay demonstrated that TsaN^{(1−392)H155A/H159A} does not bind with ATP (Supplementary Figure S5E), suggesting that the mutation of HxxxH–D motif disrupts the binding site of Mn²⁺ ion and also interferes with the binding of ATP, and TC-AMP as well.

Figure 5. — Structure−function analysis of the TsaD domain of TsaN. (A) Electron density is shown as transparent surface for His¹⁵⁵xxxHis¹⁵⁹−Asp³⁵⁰ motif and Mn²⁺. (B) Structural superposition of the TsaD domain and PaKae1 in complex with AMPPNP (PDB: 2IVN). (C) Superposition of the TsaD domain and EcTsaD in complex with BK951 (PDB: 6Z81). (D) Kinetic measurements of the ATPase activity of 5 μM TsaN, TsaN¹⁻³⁹² and TsaN^H155A/H159A using an NADH-coupled assay. The initial velocity of the ATP hydrolysis was measured at the indicated concentrations of ATP. Experiments were performed in triplicates. (E) LC–MS analysis of the t⁶A formation in assay containing bulk SctRNAs in presence of TsaN or TsaN and ScKEOPS. (F) Native gel analysis of the interaction between bulk SctRNAs and TsaN, TsaN¹⁻⁷⁴⁸ or ScKEOPS. The upper panel shows the migration of tRNAs and the lower panel shows the migration of proteins. (G) LC–MS analysis of the t⁶A formation from alkaline phosphatase-treated assays containing 4 mM L-threonine, 10 mM NaHCO₃, 1 mM ATP and 5 μM proteins of TsaN, TsaN¹⁻³⁹² and ScSua5, or TsaN^H155A/H159A. (H) LC–MS analysis of the t⁶ADP formation in *E. coli* BL21 AI cells expressing TsaN or TsaN^H155A/H159A. The mass spectrum detected at 15.476 min is shown in the insert. (I) LC–MS analysis of the t⁶A formation from alkaline phosphatase-treated assay containing 2 μM TsaN¹⁻³⁹², 25 μM TC-AMP and 1 mM ATP, ADP, AMP or adenosine (A). The assay that contained only TsaN¹⁻³⁹² and TC-AMP was used as a control. (J) Effect of different divalent metal ions on the catalytic efficiency of TsaN. The pre-bound metal ions were removed by dialysis with 2 mM EDTA. The activity was expressed as a ratio of t⁶A/A. The assay that contained no metal ions was used as a control. Experiments were performed in triplicates. (K) Structural model of TsaN–tRNA. The model was generated by structurally aligning the TsaD domain of TsaN to Kae1 subunit in structure of *M. jannaschii* KEOPS–tRNA. The TsaN dimer is shown and KEOPS is omitted for clarity. (L) A docking model of TsaN–TC-AMP–ADP. BK951 (a TC-AMP mimic) was modelled in TsaN by superposing EcTsaD in complex with BK951 onto the TsaD domain of TsaN; based on the overlaid structures of the TsaD domain of TsaN and the Kae1-like domain of TobZ in complex with carbamoyladenylate (CA) and tobramycin (PDB: 3VET and 3VER), ADP was manually docked in the TsaD domain structure using tobramycin as a superposing template. The O6 atom of tobramycin accepts the carbamoyl-moiety from CA, which is analogous to the transfer of TC-moiety from TC-AMP onto the N6 atom of ADP.

Table 3.

The summary of ITC thermodynamic parameters. The binding curves and plots are presented in Supplementary Figure S5 and Figure 6G and H. * the binding was performed in the presence of 100 μM synthetic t⁶A

Titrant (μM)	Injectant (μM)	N (sites)	K _D (μM)	ΔH (kcal/mol)	ΔG (kcal/mol)
TsaN¹⁻³⁹² (30)	ATPγS (360)	1.18 ± 0.14	10.9 ± 8.91	−6.02 ± 2.38	−6.77
	ADP (360)	1.10 ± 0.10	5.55 ± 3.53	−2.74 ± 0.62	−7.05
	AMP (360)	1.15 ± 0.14	14.5 ± 7.41	−4.74 ± 1.46	−6.49
	A (360)	No binding
TsaN^{(1−392)H155A/H159A} (30)	ATP (360)	No binding
TsaN^749−1008 (30)	ATP (360)	No binding
	ADP (360)	No binding
	AMP (360)	No binding
	A (360)	No binding
	t⁶A (360)	0.80 ± 0.03	9.14 ± 2.06	−6.02 ± 0.62	−6.76
	AMPCPP (360)*	No binding
TsaN^{(749−1008)R814A} (30)	t⁶A (360)	1.02 ± 0.06	10.6 ± 4.13	−8.98 ± 1.49	−6.67
TsaN^{(749−1008)K827A} (30)	t⁶A (360)	0.94 ± 0.04	11.2 ± 3.26	−9.38 ± 1.25	−6.65
EcTcdA (30)	t⁶A (360)	No binding

Open in a new tab

TsaN does not catalyze tRNA t⁶A biosynthesis

The genome of P. salinus encodes three tRNAs: tRNA-Met (CAU), tRNA-Trp (CCA) and tRNA-Pro (CGG) (59). We prepared P. salinus tRNA-Met, tRNA-Pro and tRNA-Trp (Supplementary Figure S2G) by in vitro transcription and confirmed its folding by CD spectra (Supplementary Figure S2H). We supplemented the TC-AMP assay with P. salinus tRNA substrates (tRNA-Met, tRNA-Pro and tRNA-Trp) and observed no formation of tRNA t⁶A by TsaN (Supplementary Figure S2I). We noticed that tRNA-Met contains C33, which was reported to be an anti-determinant of tRNA as a t⁶A-modifying substrate (37). We mutated C33 into U33 and prepared a variant tRNA-Met^C33U (Supplementary Figure S2H). However, no formation of tRNA t⁶A was observed in assay using the variant tRNA-Met^C33U (Supplementary Figure S2I). We also determined that the four P. salinus tRNA substrates could not be t⁶A-modified by ScSua5 and ScKEOPS (Supplementary Figure S2J). This suggests that the three P. salinus tRNAs are not substrates of TsaN and yeast tRNA t⁶A-modifying enzymes. To further confirm the inactivity of TsaN and to exclude false negative incurred by using P. salinus tRNAs, we examined the enzymatic activity of TsaN with bulk yeast tRNAs that were isolated from S. cerevisiae sua5Δ cells (72). Bulk SctRNAs were t⁶A-modified by ScKEOPS when supplied with TC-AMP that was generated by TsaN (Figure 5E). In contrast, TsaN was inactive in catalyzing the t⁶A modification of bulk SctRNAs (Figure 5E). We performed native gel retardation assays to compare the tRNA-binding properties of TsaN and ScKEOPS. We first showed that TsaN does not bind to tRNA-Met, tRNA-Pro, tRNA-Trp and tRNA-Met^C33U (Supplementary Figure S2K), none of which binds to ScKEOPS (Supplementary Figure S2L). Then, we showed that ScKEOPS binds to bulk SctRNAs when mixed at molar ratio of 1:1 (Figure 5F). In contrast, TsaN or TsaN¹⁻⁷⁴⁸ does not bind to bulk SctRNAs (Figure 5F).

Biosynthesis of t⁶ATP and t⁶ADP by the TsaD domain of TsaN

Unexpectedly, we found that t⁶A was formed when the TC-AMP assay mixture was dephosphorylated with alkaline phosphatase, irrespective of tRNAs (Figure 5G). Using similar analysis we confirmed t⁶A formation in the presence of ScSua5 and TsaN¹⁻³⁹², but not TsaN^H155A/H159A (Figure 5G), suggesting that the TsaD domain of TsaN performs the t⁶A-catalytic function. As the formation of t⁶A nucleoside in the assay relies on the alkaline phosphatase treatment, we reasoned that t⁶A might be derived from N⁶-threonylcarbamoyladenosine triphosphates (t⁶ATP), N⁶-threonylcarbamoyladenosine diphosphates (t⁶ADP) and/or N⁶-threonylcarbamoyladenosine monophosphate (t⁶AMP). To determine the presence of t⁶ATP, t⁶ADP and t⁶AMP, we incubated 2 mM ATP, 10 mM L-threonine, 15 mM NaHCO₃ and 10 μM TsaN for 6 h and subjected the reaction mixture to UPLC–MS high-resolution mass analysis. We first confirmed the formation of TC-AMP, ADP and AMP in the reaction mixture (Supplementary Figure S6). The relative intensity of the mass chromatograms demonstrated an abundance of ADP and AMP, and fractional portion of ATP, whose nominal level was 9.76e7, 8.68e7 and 3.71e6, respectively (Supplementary Figure S6). We then determined the presence of t⁶ADP and t⁶ATP, whose nominal level was 5.30e7 and 2.36e6, respectively (Supplementary Figure S6). As t⁶AMP and TC-AMP share the same molar mass with a nominal level of 5.99e6 in the reaction mixture (Supplementary Figure S6), we could not determine the formation of t⁶AMP in the reaction assay by mass spectrometry.

As the enzymatic assay identified t⁶ATP and t⁶ADP as products of TsaN in vitro, we set out to test for TsaN-dependent formation of these modified nucleotides in bacterial cells. We extracted the soluble small molecules from E. coli BL21 AI cells that expressed TsaN or TsaN^H155A/H159A (Supplementary Figure S2M) and analyzed the formation of t⁶ATP, t⁶ADP and t⁶AMP by LC–MS. We first confirmed that a large quantity of t⁶ADP was generated in cells that expressed TsaN, but not in cells that expressed the catalytically dead variant TsaN^H155A/H159A (Figure 5H). In contrast, we have not detected the presence of t⁶ATP or t⁶AMP in cells that expressed TsaN. The alkaline phosphatase treatment of cell extract that expressed TsaN leads to t⁶A (Supplementary Figure S2N). We further explored the acceptor(s) of TC-moiety with a TC-transfer assay using isolated TC-AMP, which was purified by HPLC from TC-AMP assay and prepared in compound to prevent decomposition. Equal amount of TC-AMP compound in a final concentration of about 25 μM (Figure 4D) was dissolved in assay containing 2 μM TsaN¹⁻³⁹² and 1 mM nucleotide (ATP, ADP, AMP or A) and incubated for 1 h at 25°C, followed by alkaline phosphatase treatment. LC–MS analysis confirmed t⁶A formation in assays containing either ATP or ADP, but neither AMP nor A (Figure 5I). The results demonstrated that the TC-moiety was transferred from TC-AMP onto N6 atom of ATP and ADP, leading to t⁶ATP and t⁶ADP. In addition, metal ion is an integral part of the catalytic site of TsaD/Kae1/Qri7 protein and functions to coordinate the carboxylate group of the threonine moiety of TC-AMP, forming an oxyanion hole that is probably involved in catalysis (49). Structural analysis and enzymatic assays showed that TsaD/Kae1/Qri7 proteins utilize Fe²⁺/Fe³⁺, Zn²⁺ or Mg²⁺ ions to perform their catalytic function (19,22,29). We removed the pre-bound metals in TsaN through dialysis with EDTA, and tested the essentiality and stimulating activity of divalent metal ions (Mn²⁺, Mg²⁺, Ca²⁺, Zn²⁺ or Fe²⁺) in biosynthesis of t⁶ADP and t⁶ATP by TsaN (Figure 5J). The results demonstrated only MnCl₂ stimulated the catalytic activity of TsaN. Based on this finding, we modelled Mn²⁺ in the metal binding site in the structure of the TsaD domain (Figure 5A).

The KEOPS–tRNA model shows that all the subunits (Kae1, Cgi121, Bud32 and Pcc1) interact with the bound tRNA and contribute to docking of anticodon stem loop of tRNA into the catalytic site of Kae1 (23,33). We generated a TsaN–tRNA model by juxtaposing the TsaD domain of TsaN and the Kae1 domain of KEOPS–tRNA (Figure 5K). The model suggests that a tRNA molecule could be accommodated between two sub-domains of the TsaD domain without creating steric clashes with the other three domains of TsaN. However, our binding assays demonstrate that TsaN does not interact with tRNA (Figure 5F). It implies that TsaN lacks key structural determinants in support of tRNA binding.

The Kae1-like domain of TobZ catalyzes the transfer of carbamoyl-moiety from carbamoyladenylate (CA) onto O6 atom of tobramycin (45), an ancient carbamoylation reaction analogous to the catalytic transfer of TC-moiety from TC-AMP onto N6 atom of tRNA A37 (23,45). To gain structural insights into the catalytic specificity of the TsaD domain, we juxtaposed the structures of the TsaD domain, EcTsaD in complex with BK951 (a TC-AMP mimic) (Figure 5C) (49) and the Kae1 domain of TobZ in complex with CA and tobramycin (Supplementary Figure S4G) (45). The overlaid structures revealed the positions and orientations of BK951, CA and tobramycin in the catalytic center of TsaD/Kae1-like domains (Figure 5L). Promoted by the docking models of adenine in the active site of EcTsaD bound to BK951 (49) and the docking model of tobramycin and BK951 in the active site of Kae1 (23), we manually modelled the binding modes of BK951 and ADP in the active site of the TsaD domain of TsaN by constraining that the N6 atom of ADP is in a nucleophilic attack range to the carbonyl group of BK951 (Figure 5L). In the overlaid structures, it shows that docked ADP in the TsaD domain of TsaN clashes with α1 (amino acids 32−46) of EcTsaD, wherein the equivalent segment (amino acids 77−91) was disordered in the cryo-EM structure of TsaN. Sequence alignment reveals that the equivalent segments that contain a GxxP motif are conserved in all TsaD/Kae1/Qri7 proteins but not in the TsaD domain of TsaN (Supplementary Figure S1A). The models suggest that the specific segment connecting β3 and α1 of the TsaN is likely involved in recognition and binding of the TC-acceptors ADP and ATP (Figure 5L).

Conversion of t⁶A to ct⁶A by the TcdA domain of TsaN

Overlaid structures of the TcdA domain of TsaN and EcTcdA in complex with ATP or AMP project the ATP/AMP binding site in the TcdA domain of TsaN and show the positions of Asp⁸⁰³, Arg⁸¹⁴ and Lys⁸²⁷ in relation to the ATP/AMP bound to EcTcdA (Figure 6A). Strong conservation in structures also implies that the TcdA domain of TsaN possesses an enzymatic activity related to tRNA ct⁶A biosynthesis. We confirmed that bulk EctRNAs contained t⁶A, but have not managed to detect ct⁶A in assays that contained bulk EctRNAs and EcTcdA or TsaN (data not shown) in the presence of ATP. We reasoned that a small amount of tRNA ct⁶A was possibly hydrolyzed or reacted with amines in Tris−HCl buffer (50). We then opted to determine the enzymatic activity of TsaN in catalyzing the conversion of synthetic t⁶A to ct⁶A. The LC–MS analysis on ct⁶A formation showed that either TsaN or the TcdA domain (TsaN^749−1008) catalyzed an efficient conversion of synthetic t⁶A to ct⁶A in the presence of ATP (Figure 6B). To confirm the essentiality of Arg⁸¹⁴ and Lys⁸²⁷, whose counterparts–Arg⁷² and Lys⁸⁵–in E. coli TcdA are directly involved in ATP binding (Figure 6A), we generated and purified two variants of the TcdA domain–TsaN^{(749−1008)R814A} and TsaN^{(749−1008)K827A} (Supplementary Figure S2E and F). TsaN^{(749−1008)R814A} and TsaN^{(749−1008)K827A} were inactive in catalyzing the conversion of synthetic t⁶A to ct⁶A. In comparison, only trace ct⁶A was formed in assays that contained same concentrations of synthetic t⁶A and EcTcdA (Figure 6B). Our quantification shows that the catalytic efficiency of EcTcdA is <5% of TsaN^749−1008 (Figure 6C).

We observed that the conversion of synthetic t⁶A to ct⁶A by TsaN^749−1008 is coupled with the hydrolysis of ATP into AMP and PPi (Figure 6B). Then, we explored the role of ATP in the conversion of synthetic t⁶A to ct⁶A by TsaN^749−1008. Through ITC measurement, we were not able to determine the binding of nucleotide (ATP, ADP, AMP or A) with TsaN^749−1008 in the absence of synthetic t⁶A (Supplementary Figure S5F–I, Table 3). Alternatively, our quantification of the enzymatic activity showed efficient conversion of t⁶A to ct⁶A in the presence of ATP. In contrast, we have not observed the formation of ct⁶A in assays that contained AMP, AMPCPP (a nonhydrolyzable ATP analogue) or AMP plus PPi (Figure 6D), implying that ATP hydrolysis is required to stimulate the ct⁶A-catalytic activity of TsaN. Moreover, we showed that the catalytically dead variants TsaN^{(749−1008)R814A} and TsaN^{(749−1008)K827A} are still capable of binding with synthetic t⁶A (Supplementary Figure S5J and K, Table 3), suggesting that Arg⁸¹⁴ and Lys⁸²⁷ are not essentially involved in the binding of t⁶A and may play a role in binding and/or hydrolysis of ATP. To test whether the binding of ATP to TsaN^749−1008 depends on the pre-binding of t⁶A, we performed the ITC analysis on the binding of AMPCPP to t⁶A-bound TsaN^749−1008. However, we have not observed a binding signal in the ITC assay (Supplementary Figure S5L and Table 3). Furthermore, we determined that ATP is not hydrolyzed by TsaN^749−1008 in the absence of synthetic t⁶A (data not shown), suggesting that the conversion of t⁶A to ct⁶A is coupled to ATP hydrolysis.

Comparison of the surface electrostatic potentials of the TcdA domain of TsaN and EcTcdA reveals a large positive electrostatic patch surrounding the ATP-binding site in EcTcdA and a greatly reduced positive electrostatic surface on the TcdA domain of TsaN (Figure 6E). Using native gel retardation assay we first determined that EcTcdA binds to bulk EctRNAs while TsaN^749−1008 does not bind EctRNAs (Figure 6F). We then determined by ITC that TsaN^749−1008 binds to synthetic t⁶A with a K_D value of 9.14 μM (Figure 6G and Table 3) while titration of EcTcdA with synthetic t⁶A did not produce a binding signal (Figure 6H). The binding data suggests that TsaN^749−1008 specifically binds with t⁶A nucleoside and EcTcdA preferentially interacts with tRNAs.

Notably, we have not detected ct⁶ATP or ct⁶ADP in the enzymatic assay that contained t⁶ATP, t⁶ADP and TsaN (Supplementary Figure S6), nor have we detected ct⁶ADP in the cells that contained a great amount of t⁶ADP and TsaN (Figure 5H and Supplementary Figure S2M and N), implying that the phosphates of t⁶ADP and t⁶ATP might be anti-determinants for the conversion of t⁶ADP to ct⁶ADP by TsaN.

DISCUSSION

As one of a few non-canonical RNA nucleosides that are deemed as molecular fossils of life's origin, t⁶A might have existed in pre-RNAs and possibly functions, as part of a universal chemical code, to increase the chemical diversity of RNA to broaden its folding and catalytic capabilities before the genetic code arose in living cells (16,17). Chemical assay demonstrated non-enzymatic formation of t⁶A at the nucleoside level under simulated prebiotic and anaerobic conditions (17). However, an enzymatic synthesis of t⁶A or t⁶A derivatives at the nucleoside or nucleotide level has never been reported. In this work, we identified a TsaD–TsaC–SUA5–TcdA modular enzyme TsaN from Pandoravirus genomes (Figure 2). For this study, we have chosen to characterize the structure and function of P. salinus TsaN. Our cryo-EM structure of TsaN reveals that the four domains share strong structural similarities with TsaD/Kae1/Qri7 proteins, TsaC/Sua5 proteins and EcTcdA. We demonstrate that TsaN is active in catalyzing the biosynthesis of TC-AMP but inactive in t⁶A modification of tRNAs. We discovered that TsaN is an ancient enzyme that catalyzes the tRNA-independent biosynthesis of t⁶ADP and t⁶ATP, and report for the first time an enzymatic threonylcarbamoyl modification at the nucleotide level. Moreover, we report that the TcdA domain of TsaN is active in catalyzing the conversion of t⁶A to ct⁶A at the nucleoside level.

Though the four domains of TsaN are structurally similar and functionally related to those of TsaD/Kae1/Qri7 proteins, TsaC/Sua5 proteins and EcTcdA (Supplementary Figure S4), their spatial organization is distinct from those of related proteins, e.g. TobZ, HypF and Sua5. The fusion of a Kae1-like domain and a TsaC-like domain in TobZ (45) and HypF (58) creates a channel that delivers the intermediate carbamoyladenylate (CA) from the catalytic site in the TsaC-like domain onto the catalytic site in the Kae1-like domain. In TsaN, however, no such analogous channel was formed between the TsaC domain and TsaD domain (Figure 3G). Our enzymatic assays demonstrate that the TsaD domain functions independently while function of the TsaC–SUA5 domain depends on the TsaD domain. In TsaN, the TsaC domain and the SUA5 domain are connected by a long linker, of which 20 amino acids were not observed in our structure. In Sua5 proteins, the SUA5 domain is positioned in proximity to the catalytic site in the TsaC domain (41–43); in TsaN, the SUA5 domain is largely rotated by 75° with respect to the TsaC domain (Figure 4B). Consequently, conformation of the interdomain linker and two conserved motifs PGM and HYxP in TsaN are substantially different from those in Sua5 proteins. Tyr⁶⁴⁰ of the HYxP motif in TsaN is positioned around 15 Å far away from the TC-AMP catalytic site than its counterpart Tyr²³⁵ in PaSua5 (43). Interestingly, mutation of Tyr⁶⁴⁰ abolishes the catalytic activity of TsaN in generating TC-AMP but does not affect its catalytic activity in hydrolyzing ATP into AMP and PPi (Figure 4D). However, the negative regulation of inter-domain linker by Tyr⁶⁴⁰ mutation is presently unexplainable. Our biochemical characterization suggests that the inter-domain interactions between the TsaD domain and the TsaC–SUA5 domain might be required to overcome the insolubility issue of the TsaC–SUA5 domain. Our structural and functional characterization of TsaN suggests that the formation of the TsaN dimer is probably a mere result of dimerization of the TcdA domain and is not essentially needed for the enzymatic functions, but might promote the protein solubility and enhance conformational stability. Similar dimerizing phenomena occur with a number of tRNA t⁶A-modifying enzymes (19). For instance, TsaB forms a homodimer that mediates the dimerization of TsaD–TsaB dimer, leading to the formation of a TsaD₂–TsaB₂ tetramer (27,29,30); Pcc1 also exists as a homodimer and mediates the formation of KEOPS dimer (32,34,36,39,61).

The bi-lobal structure and catalytic center configuration of the TsaD domain of TsaN are highly similar to those of TsaD/Kae1/Qri7 family proteins (Supplementary Figure S4B and Table 2), which bind TC-AMP and catalyze the TC-transfer onto A37 of tRNAs (19,23). However, the TsaD domain does not participate in tRNA t⁶A modification but catalyzes the threonylcarbamoyl modification of ADP and ATP (Figure 5I), suggesting that TC-acceptor ADP or ATP is bound in proximity to TC-AMP (Figure 5L). We generated a model of TsaN in complex with TC-AMP and ADP, in which TC-AMP and ADP mimic the position and orientation of CA and tobramycin (45) bound in TobZ. Our docking model of TsaN–TC-AMP–ADP model suggests that the TC-moiety acceptor ADP could be accommodated in a manner that conforms with the binding of tRNA A37 (Figure 5L). The docking model suggests that a segment (amino acids 77−91) connecting β3 and α1 of TsaN might participate in the binding of TC-acceptor ADP (Figure 5L). Unfortunately, the segment was not observed in the cryo-EM structure of TsaN. Sequence alignment shows that the disordered segment of TsaN lacks a GxxP motif that is extremely conserved in TsaD/Kae1/Qri7 proteins (Supplementary Figure S1A). The equivalent position in TsaN is a DxxK motif (Supplementary Figure S1A) that is conserved in all TsaN orthologs from Pandoraviruses. We determined that Mn²⁺ is the third type of divalent metal ion coordinated by the HxxxH–D motif in the TsaD domain (Figure 5A and J). We were not able to test the interaction between the TsaD domain and TC-AMP due to its chemical instability. The TsaD domain binds with ATP (Table 3) and its variant TsaN^{(1–392)H155A/H159A} that bears two mutations in the conserved HxxxH–D motif does not bind with ATP (Table 3) and is inactivated in catalyzing the formation of t⁶ADP (Figure 5G). Our ITC analysis of the binding between the TsaD domain and ATP, ADP or AMP only determines a single-binding site model (Supplementary Figure S5 and Table 3), which suggests that one molecule of ATP is bound in coordination with the HxxxH–D motif in the TC-transfer site. The TsaD domain binds with AMP in the TC-AMP site (Table 3) but not in the TC-acceptor site, as it does not catalyze the threonylcarbamoyl modification of AMP (Figure 5I). We tentatively titrated the 30 μM AMP-bound TsaN^1-392 with 360 μM ADP in ITC measurements and observed no binding curves (data not shown), suggesting the AMP occupied the TC-AMP binding site and ADP cannot bind to the TC-acceptor site in the absence of TC-AMP. It may suggest that the binding of TC-acceptor ADP or ATP depends on the pre-binding of TC-AMP to the TsaD domain. Moreover, the catalytic inactivity of TsaN on AMP and A suggests that the β- and/or γ-phosphate plays a key role in the molecular recognition between TsaN and the TC-acceptor ADP or ATP.

Miyauchi et al. reported that EcTcdA catalyzed the biosynthesis of ct⁶A using a t⁶A-containing RNA fragment (50). Kim et al. mentioned that EcTcdA was not active in catalyzing the conversion of synthetic t⁶A to ct⁶A in the presence of ATP (54). We discovered that TsaN catalyzes an efficient conversion of synthetic t⁶A to ct⁶A at the nucleoside level (Figure 6B). In contrast, EcTcdA exhibits only a weak activity that is less than 5% of the TcdA domain (Figure 6B and C). Our binding assays demonstrate that the synthetic t⁶A is bound to the TcdA domain but not EcTcdA (Table 3) and explains the functional difference between the TcdA domain and EcTcdA. Interestingly, we have not detected ct⁶ADP or ct⁶ATP in the enzymatic products that contained t⁶ADP, t⁶ATP and TsaN (Supplementary Figure S6), nor in cells that produced a large quantity of t⁶ADP and TsaN (Figure 5H), implying that phosphates of t⁶ADP/t⁶ATP are anti-determinant. Conversion of t⁶A nucleoside to ct⁶A by TsaN is coupled with the hydrolysis of ATP into AMP and PPi (Figure 6B). Neither nonhydrolyzable AMPCPP nor AMP plus PPi could stimulate the ct⁶A-catalytic activity of the TcdA domain (Figure 6D), suggesting that the conversion of t⁶A to ct⁶A is driven by the ATP-hydrolyzing process per se and/or the energy released is required. Through ITC measurement, we show that the TcdA domain binds with synthetic t⁶A (Table 3). Mutations in the conserved D–R–K motif (Arg⁸¹⁴ and Lys⁸²⁷) of the TcdA domain lead to loss of the ct⁶A-catalytic activity (Figure 6B) but do not interfere with the binding of synthetic t⁶A (Table 3), implying that the D–R–K motif is involved in ATP binding and adenylation reaction. However, we did not observe binding between the TcdA domain and ATP or AMP in the absence or presence of synthetic t⁶A (Table 3) and could not examine the effect of Arg⁸¹⁴ and Lys⁸²⁷ mutations on either the binding or the hydrolysis of ATP. It's presently unknown how t⁶A and ATP are simultaneously bound to the catalytic site of the TcdA domain of TsaN or EcTcdA.

We explored the structural basis for the inactivity of TsaN on tRNAs. Our docking model of TsaN–tRNA shows that the anticodon stem loop (ASL) of tRNA could be accommodated between the two sub-domains of the TsaD domain (Figure 5K). However, our binding analysis demonstrate that TsaN and the TsaD–TsaC–SUA5 domain do not interact with tRNAs (Figure 5F and Supplementary Figure S2K). In KEOPS–tRNA model, tRNA ASL is bound between two sub-domains of Kae1 with support of Bud32, Cgi121 and Pcc1 (23,33). In TsaD₂–TsaB₂–tRNA model, tRNA ASL is docked into the catalytic site of TsaD with support of TsaB (29,30). We searched against the Pandoravirus genomes for orthologs of Kae1 auxiliary proteins (Bud32, Cgi121, Pcc1, Pcc2 and Gon7) and TsaD auxiliary proteins (TsaB and TsaE) (19,23), and identified no protein-coding sequence (CDS) with significant similarity to these TsaD/Kae1 auxiliary proteins. In the case of the TcdA domain of TsaN, surface electrostatic potential distribution reveals that the positively charged area surrounding the catalytic site is greatly smaller than that of EcTcdA (Figure 6E). Our binding assays demonstrate that EcTcdA binds strongly to EctRNAs while the TcdA domain of TsaN does not bind to tRNAs (Figure 6F). We assume that bases adjacent to tRNA A37 contribute to the interaction between EcTcdA and tRNA t⁶A37 (50).

The evolutionary relationship between TsaN and TsaC/Sua5 family (COG0009), TsaD/Kae1/Qri7 family (COG0533), and TcdA/Tcd1/Tcd2 family (COG1179) remains unexplored. We found that TsaN orthologs are encoded in 7 genomes of Pandoraviruses (Figure 2), which are giant double-stranded DNA viruses infecting species of Acanthamoeba (59,73–75). Comparative genomics analysis found that horizontal gene transfer (HGT) occurs frequently between viruses and hosts and is hypothesized to be one main course of genetic evolution (76–78). The vast majority of HGT occurred in double-stranded DNA viruses and most HGT underwent unidirectional transfer, i.e. eukaryote-to-virus or virus-to-eukaryote transfer (73,77). Analysis on the relationship between gene function and HGT direction uncovered that a great number of genes involved in RNA modification and tRNA processing underwent a virus-to-eukaryote transfer (77). Acanthamoeba genomes seem to encode their own tRNA t⁶A-modifying system composed of Sua5 and KEOPS complex–Kae1, Bud32 and Pcc1, and a TcdA-like protein (79). Therefore, TsaN of Pandoraviruses does not seem to be a result of HGT from Sua5 and Kae1 of Acanthamoeba, vice versa. It has been hypothesized that a number of hallmark genes of cellular life, such as genes of tRNAs and tRNA-processing enzymes, may have originated from ancestral giant viruses that predate last universal common ancestor (76–78). A distance-based phylogenomic analysis on the evolutionary relationships between cells and viruses suggests an early origin of a large variety of DNA/RNA viruses predating the origin of cells (78). In this scenario, TsaN seems to be an ancient fusion protein that has not evolved to catalyze the tRNA t⁶A modification yet, but has acquired prototypic activity in catalyzing the threonylcarbamoyl modification at the nucleotide level.

Viral proteins can function to facilitate the intracellular replication by a number of means that include signalling modulation and metabolic reprogramming (76). For example, P. massiliensis genome encodes 8 orthologs of cellular enzymes involved in the universal tricarboxylic acid cycle and these genes are transcribed during the infectious cycles of P. massiliensis in Acanthamoeba (75). Acanthamoeba seems to have evolved a typical eukaryotic tRNA t⁶A-modifying system. Therefore, TsaN might not participate in t⁶A modification of tRNAs of Acanthamoeba when infected with Pandoraviruses. However, TC-AMP produced by TsaN could be supplied to KEOPS to t⁶A-modify tRNAs of Acanthamoeba. In our experiment, the expression of TsaN and production of t⁶ADP conferred no observable effects on the bacterial growth. The physiological roles of t⁶ADP or t⁶ATP remain to be explored in the future. ct⁶ADP, ct⁶ATP and ct⁶A are not produced by TsaN in the cellular context.

CONCLUSION

In this study, we have identified a TsaD–TsaC–SUA5–TcdA modular protein–TsaN in Pandoraviruses. The four domains of TsaN are structurally and functionally related to TsaD/Kae1/Qri7 proteins, TsaC/Sua5 proteins, and TcdA/Tcd1/Tcd2 proteins, all of which participate in the biosynthesis of tRNA t⁶A and ct⁶A. TsaN catalyzes the biosynthesis of TC-AMP, an intermediate substrate of TsaD/Kae1/Qri7 proteins, but does not participate further in tRNA t⁶A biosynthesis. However, we discovered that TsaN catalyzes tRNA-independent threonylcarbamoyl modification of ADP and ATP, leading to t⁶ADP and t⁶ATP, respectively. Moreover, TsaN is highly active in catalyzing the conversion of t⁶A nucleoside to ct⁶A, which is coupled with the hydrolysis of ATP into AMP and PPi. The role of TsaN during the infection cycles of Pandoraviruses in Acanthamoeba remains to be explored in the future. The structure and functions of P. salinus TsaN imply that TsaN from Pandoraviruses might represent a prototype of the tRNA t⁶A- and ct⁶A-modifying enzymes of cellular organisms. This work contributes to understanding the origins of the universal and ancient tRNA t⁶A biosynthetic pathway.

Supplementary Material

gkad587_Supplemental_File

Click here for additional data file.^{(32.5MB, pdf)}

ACKNOWLEDGEMENTS

We are grateful to Prof. Herman van Tilbeurgh for providing expression plasmids of S. cerevisiae KEOPS and S. cerevisiae Sua5, to Prof. Jinqiu Zhou for providing S. cerevisiae sua5Δ strain, to Dr. Jinsai Shang for his help with SEC-MALS experiment, to Supercomputing Center of Lanzhou University for providing computational resource, and to Prof. Frank Sicheri for providing coordinate of the MjKEOPS–tRNA model. We appreciate the support of Core Facility of School of Life Sciences, Lanzhou University.

Author contributions: W.Z. conceived and supervised the study. M.J., Z.Y. and W.C. produced recombinant proteins and performed the biochemical assays, Z.Z. and D.L. prepared the cryo-EM grids, collected data, determined cryo-EM structures. D.L. and W.Z. analysed the structural data. X.W. synthesized the t⁶A compound and helped with LC–MS analysis. W.Z. wrote the manuscript with inputs from M.J. and all other co-authors.

Contributor Information

Mengqi Jin, School of Life Sciences, Key Laboratory of Cell Activities and Stress Adaptation of the Ministry of Education, Lanzhou University, Lanzhou 730000, China.

Zelin Zhang, Key Laboratory for Magnetism and Magnetic Materials of the Ministry of Education, Electron Microscopy Centre of Lanzhou University, Lanzhou University, Lanzhou 730000, China.

Zhijiang Yu, School of Life Sciences, Key Laboratory of Cell Activities and Stress Adaptation of the Ministry of Education, Lanzhou University, Lanzhou 730000, China.

Wei Chen, School of Life Sciences, Key Laboratory of Cell Activities and Stress Adaptation of the Ministry of Education, Lanzhou University, Lanzhou 730000, China.

Xiaolei Wang, State Key Laboratory of Applied Organic Chemistry, College of Chemistry and Chemical Engineering, Lanzhou University, Lanzhou 730000, China.

Dongsheng Lei, Key Laboratory for Magnetism and Magnetic Materials of the Ministry of Education, Electron Microscopy Centre of Lanzhou University, Lanzhou University, Lanzhou 730000, China.

Wenhua Zhang, School of Life Sciences, Key Laboratory of Cell Activities and Stress Adaptation of the Ministry of Education, Lanzhou University, Lanzhou 730000, China.

DATA AVAILABILITY

The cryo-EM electron density map and the atomic model of the P. salinus TsaN have been deposited in the Protein Data Bank (https://www.wwpdb.org) under the accession codes: EMD-35365 and PDB ID 8IDE.

SUPPLEMENTARY DATA

Supplementary Data are available at NAR Online.

FUNDING

Natural Science Foundation of China [32000847, 32171300]; Natural Science Foundation of Gansu Province [20JR10RA618]; Fundamental Research Funds for the Central Universities [lzujbky-2021-ct05, 2022021zr46]. Funding for open access charge: Natural Science Foundation of China.

Conflict of interest statement. None declared.

REFERENCES

1. Boccaletto P., Stefaniak F., Ray A., Cappannini A., Mukherjee S., Purta E., Kurkowska M., Shirvanizadeh N., Destefanis E., Groza P.et al.. MODOMICS: a database of RNA modification pathways. 2021 update. Nucleic Acids Res. 2022; 50:D231–D235. [DOI] [PMC free article] [PubMed] [Google Scholar]
2. Machnicka M.A., Olchowik A., Grosjean H., Bujnicki J.M.. Distribution and frequencies of post-transcriptional modifications in tRNAs. RNA Biol. 2014; 11:1619–1629. [DOI] [PMC free article] [PubMed] [Google Scholar]
3. Jackman J.E., Alfonzo J.D.. Transfer RNA modifications: nature's combinatorial chemistry playground. Wiley Interdiscip. Rev. RNA. 2013; 4:35–48. [DOI] [PMC free article] [PubMed] [Google Scholar]
4. Grosjean H., Sprinzl M., Steinberg S.. Posttranscriptionally modified nucleosides in transfer RNA: their locations and frequencies. Biochimie. 1995; 77:139–141. [DOI] [PubMed] [Google Scholar]
5. Parthasarathy R., Ohrt J.M., Chheda G.B.. Conformation and possible role of hypermodified nucleosides adjacent to 3'-end of anticodon in tRNA: n-(purin-6-ylcarbamoyl)-L-threonine riboside. Biochem. Biophys. Res. Commun. 1974; 60:211–218. [DOI] [PubMed] [Google Scholar]
6. Stuart J.W., Gdaniec Z., Guenther R., Marszalek M., Sochacka E., Malkiewicz A., Agris P.F.. Functional anticodon architecture of human tRNALys3 includes disruption of intraloop hydrogen bonding by the naturally occurring amino acid modification, t6A. Biochemistry. 2000; 39:13396–13404. [DOI] [PubMed] [Google Scholar]
7. Murphy F.V.t., Ramakrishnan V., Malkiewicz A., Agris P.F.. The role of modifications in codon discrimination by tRNA(Lys)UUU. Nat. Struct. Mol. Biol. 2004; 11:1186–1191. [DOI] [PubMed] [Google Scholar]
8. Yarian C., Townsend H., Czestkowski W., Sochacka E., Malkiewicz A.J., Guenther R., Miskiewicz A., Agris P.F.. Accurate translation of the genetic code depends on tRNA modified nucleosides. J. Biol. Chem. 2002; 277:16391–16395. [DOI] [PubMed] [Google Scholar]
9. Thiaville P.C., Legendre R., Rojas-Benitez D., Baudin-Baillieu A., Hatin I., Chalancon G., Glavic A., Namy O., de Crecy-Lagard V.. Global translational impacts of the loss of the tRNA modification t(6)A in yeast. Microb. Cell. 2016; 3:29–45. [DOI] [PMC free article] [PubMed] [Google Scholar]
10. El Yacoubi B., Lyons B., Cruz Y., Reddy R., Nordin B., Agnelli F., Williamson J.R., Schimmel P., Swairjo M.A., de Crecy-Lagard V.. The universal YrdC/Sua5 family is required for the formation of threonylcarbamoyladenosine in tRNA. Nucleic Acids Res. 2009; 37:2894–2909. [DOI] [PMC free article] [PubMed] [Google Scholar]
11. El Yacoubi B., Hatin I., Deutsch C., Kahveci T., Rousset J.P., Iwata-Reuyl D., Murzin A.G., de Crecy-Lagard V.. A role for the universal Kae1/Qri7/YgjD (COG0533) family in tRNA modification. EMBO J. 2011; 30:882–893. [DOI] [PMC free article] [PubMed] [Google Scholar]
12. Thiaville P.C., El Yacoubi B., Kohrer C., Thiaville J.J., Deutsch C., Iwata-Reuyl D., Bacusmo J.M., Armengaud J., Bessho Y., Wetzel C.et al.. Essentiality of threonylcarbamoyladenosine (t(6)A), a universal tRNA modification, in bacteria. Mol. Microbiol. 2015; 98:1199–1221. [DOI] [PMC free article] [PubMed] [Google Scholar]
13. Suzuki T. The expanding world of tRNA modifications and their disease relevance. Nat. Rev. Mol. Cell Biol. 2021; 22:375–392. [DOI] [PubMed] [Google Scholar]
14. Braun D.A., Rao J., Mollet G., Schapiro D., Daugeron M.C., Tan W., Gribouval O., Boyer O., Revy P., Jobst-Schwan T.et al.. Mutations in KEOPS-complex genes cause nephrotic syndrome with primary microcephaly. Nat. Genet. 2017; 49:1529–1538. [DOI] [PMC free article] [PubMed] [Google Scholar]
15. Arrondel C., Missoury S., Snoek R., Patat J., Menara G., Collinet B., Liger D., Durand D., Gribouval O., Boyer O.et al.. Defects in t(6)A tRNA modification due to GON7 and YRDC mutations lead to Galloway-Mowat syndrome. Nat. Commun. 2019; 10:3967. [DOI] [PMC free article] [PubMed] [Google Scholar]
16. Weiss M.C., Sousa F.L., Mrnjavac N., Neukirchen S., Roettger M., Nelson-Sathi S., Martin W.F.. The physiology and habitat of the last universal common ancestor. Nat. Microbiol. 2016; 1:16116. [DOI] [PubMed] [Google Scholar]
17. Schneider C., Becker S., Okamura H., Crisp A., Amatov T., Stadlmeier M., Carell T.. Noncanonical RNA nucleosides as molecular fossils of an early earth-generation by prebiotic methylations and carbamoylations. Angew Chem. Int. Ed. Engl. 2018; 57:5943–5946. [DOI] [PubMed] [Google Scholar]
18. Galperin M.Y., Koonin E.V.. Conserved hypothetical’ proteins: prioritization of targets for experimental study. Nucleic Acids Res. 2004; 32:5452–5463. [DOI] [PMC free article] [PubMed] [Google Scholar]
19. Su C., Jin M., Zhang W.. Conservation and diversification of tRNA t6A-modifying enzymes across the three domains of life. Int. J. Mol. Sci. 2022; 23:13600. [DOI] [PMC free article] [PubMed] [Google Scholar]
20. Lauhon C.T. Mechanism of N6-threonylcarbamoyladenonsine (t(6)A) biosynthesis: isolation and characterization of the intermediate threonylcarbamoyl-AMP. Biochemistry. 2012; 51:8950–8963. [DOI] [PubMed] [Google Scholar]
21. Wan L.C., Mao D.Y., Neculai D., Strecker J., Chiovitti D., Kurinov I., Poda G., Thevakumaran N., Yuan F., Szilard R.K.et al.. Reconstitution and characterization of eukaryotic N6-threonylcarbamoylation of tRNA using a minimal enzyme system. Nucleic Acids Res. 2013; 41:6332–6346. [DOI] [PMC free article] [PubMed] [Google Scholar]
22. Zhou J.B., Wang Y., Zeng Q.Y., Meng S.X., Wang E.D., Zhou X.L.. Molecular basis for t6A modification in human mitochondria. Nucleic Acids Res. 2020; 48:3181–3194. [DOI] [PMC free article] [PubMed] [Google Scholar]
23. Beenstock J., Sicheri F.. The structural and functional workings of KEOPS. Nucleic Acids Res. 2021; 49:10818–10834. [DOI] [PMC free article] [PubMed] [Google Scholar]
24. Thiaville P.C., Iwata-Reuyl D., de Crecy-Lagard V.. Diversity of the biosynthesis pathway for threonylcarbamoyladenosine (t(6)A), a universal modification of tRNA. RNA Biol. 2014; 11:1529–1539. [DOI] [PMC free article] [PubMed] [Google Scholar]
25. Deutsch C., Yacoubi E., B. d.C.-L., Iwata-Reuyl D. Biosynthesis of threonylcarbamoyl adenosine (t6A), a universal tRNA nucleoside. J. Biol. Chem. 2012; 287:13666–13673. [DOI] [PMC free article] [PubMed] [Google Scholar]
26. Zhang W., Collinet B., Perrochia L., Durand D., van Tilbeurgh H.. The ATP-mediated formation of the YgjD-YeaZ-YjeE complex is required for the biosynthesis of tRNA t6A in Escherichia coli. Nucleic Acids Res. 2015; 43:1804–1817. [DOI] [PMC free article] [PubMed] [Google Scholar]
27. Luthra A., Swinehart W., Bayooz S., Phan P., Stec B., Iwata-Reuyl D., Swairjo M.A.. Structure and mechanism of a bacterial t6A biosynthesis system. Nucleic Acids Res. 2018; 46:1395–1411. [DOI] [PMC free article] [PubMed] [Google Scholar]
28. Perrochia L., Crozat E., Hecker A., Zhang W., Bareille J., Collinet B., van Tilbeurgh H., Forterre P., Basta T.. In vitro biosynthesis of a universal t6A tRNA modification in Archaea and Eukarya. Nucleic Acids Res. 2013; 41:1953–1964. [DOI] [PMC free article] [PubMed] [Google Scholar]
29. Luthra A., Paranagama N., Swinehart W., Bayooz S., Phan P., Quach V., Schiffer J.M., Stec B., Iwata-Reuyl D., Swairjo M.A.. Conformational communication mediates the reset step in t6A biosynthesis. Nucleic Acids Res. 2019; 47:6551–6567. [DOI] [PMC free article] [PubMed] [Google Scholar]
30. Missoury S., Plancqueel S., Li de la Sierra-Gallay I., Zhang W., Liger D., Durand D., Dammak R., Collinet B., van Tilbeurgh H.. The structure of the TsaB/TsaD/TsaE complex reveals an unexpected mechanism for the bacterial t6A tRNA-modification. Nucleic Acids Res. 2018; 46:5850–5860. [DOI] [PMC free article] [PubMed] [Google Scholar]
31. Perrochia L., Guetta D., Hecker A., Forterre P., Basta T.. Functional assignment of KEOPS/EKC complex subunits in the biosynthesis of the universal t6A tRNA modification. Nucleic Acids Res. 2013; 41:9484–9499. [DOI] [PMC free article] [PubMed] [Google Scholar]
32. Wan L.C., Pillon M.C., Thevakumaran N., Sun Y., Chakrabartty A., Guarne A., Kurinov I., Durocher D., Sicheri F.. Structural and functional characterization of KEOPS dimerization by Pcc1 and its role in t6A biosynthesis. Nucleic Acids Res. 2016; 44:6971–6980. [DOI] [PMC free article] [PubMed] [Google Scholar]
33. Beenstock J., Ona S.M., Porat J., Orlicky S., Wan L.C.K., Ceccarelli D.F., Maisonneuve P., Szilard R.K., Yin Z., Setiaputra D.et al.. A substrate binding model for the KEOPS tRNA modifying complex. Nat. Commun. 2020; 11:6233. [DOI] [PMC free article] [PubMed] [Google Scholar]
34. Daugeron M.C., Missoury S., Da Cunha V., Lazar N., Collinet B., van Tilbeurgh H., Basta T.. A paralog of Pcc1 is the fifth core subunit of the KEOPS tRNA-modifying complex in Archaea. Nat. Commun. 2023; 14:526. [DOI] [PMC free article] [PubMed] [Google Scholar]
35. Srinivasan M., Mehta P., Yu Y., Prugar E., Koonin E.V., Karzai A.W., Sternglanz R.. The highly conserved KEOPS/EKC complex is essential for a universal tRNA modification, t6A. EMBO J. 2011; 30:873–881. [DOI] [PMC free article] [PubMed] [Google Scholar]
36. Wan L.C., Maisonneuve P., Szilard R.K., Lambert J.P., Ng T.F., Manczyk N., Huang H., Laister R., Caudy A.A., Gingras A.C.et al.. Proteomic analysis of the human KEOPS complex identifies C14ORF142 as a core subunit homologous to yeast Gon7. Nucleic Acids Res. 2017; 45:805–817. [DOI] [PMC free article] [PubMed] [Google Scholar]
37. Wang J.T., Zhou J.B., Mao X.L., Zhou L., Chen M., Zhang W., Wang E.D., Zhou X.L.. Commonality and diversity in tRNA substrate recognition in t6A biogenesis by eukaryotic keopss. Nucleic Acids Res. 2022; 50:2223–2239. [DOI] [PMC free article] [PubMed] [Google Scholar]
38. Downey M., Houlsworth R., Maringele L., Rollie A., Brehme M., Galicia S., Guillard S., Partington M., Zubko M.K., Krogan N.J.et al.. A genome-wide screen identifies the evolutionarily conserved KEOPS complex as a telomere regulator. Cell. 2006; 124:1155–1168. [DOI] [PubMed] [Google Scholar]
39. Mao D.Y., Neculai D., Downey M., Orlicky S., Haffani Y.Z., Ceccarelli D.F., Ho J.S., Szilard R.K., Zhang W., Ho C.S.et al.. Atomic structure of the KEOPS complex: an ancient protein kinase-containing molecular machine. Mol. Cell. 2008; 32:259–275. [DOI] [PMC free article] [PubMed] [Google Scholar]
40. Teplova M., Tereshko V., Sanishvili R., Joachimiak A., Bushueva T., Anderson W.F., Egli M.. The structure of the yrdC gene product from Escherichia coli reveals a new fold and suggests a role in RNA binding. Protein Sci. 2000; 9:2557–2566. [DOI] [PMC free article] [PubMed] [Google Scholar]
41. Agari Y., Sato S., Wakamatsu T., Bessho Y., Ebihara A., Yokoyama S., Kuramitsu S., Shinkai A.. X-ray crystal structure of a hypothetical Sua5 protein from Sulfolobus tokodaii strain 7. Proteins. 2008; 70:1108–1111. [DOI] [PubMed] [Google Scholar]
42. Kuratani M., Kasai T., Akasaka R., Higashijima K., Terada T., Kigawa T., Shinkai A., Bessho Y., Yokoyama S.. Crystal structure of Sulfolobus tokodaii Sua5 complexed with L-threonine and AMPPNP. Proteins. 2011; 79:2065–2075. [DOI] [PubMed] [Google Scholar]
43. Pichard-Kostuch A., Zhang W., Liger D., Daugeron M.C., Letoquart J., Li de la Sierra-Gallay I., Forterre P., Collinet B., van Tilbeurgh H., Basta T.. Structure-function analysis of Sua5 protein reveals novel functional motifs required for the biosynthesis of the universal t(6)A tRNA modification. RNA. 2018; 24:926–938. [DOI] [PMC free article] [PubMed] [Google Scholar]
44. Harris K.A., Bobay B.G., Sarachan K.L., Sims A.F., Bilbille Y., Deutsch C., Iwata-Reuyl D., Agris P.F.. NMR-based structural analysis of threonylcarbamoyl-AMP synthase and its substrate interactions. J. Biol. Chem. 2015; 290:20032–20043. [DOI] [PMC free article] [PubMed] [Google Scholar]
45. Parthier C., Gorlich S., Jaenecke F., Breithaupt C., Brauer U., Fandrich U., Clausnitzer D., Wehmeier U.F., Bottcher C., Scheel D.et al.. The O-carbamoyltransferase TobZ catalyzes an ancient enzymatic reaction. Angew. Chem. Int. Ed. 2012; 51:4046–4052. [DOI] [PubMed] [Google Scholar]
46. Hecker A., Leulliot N., Gadelle D., Graille M., Justome A., Dorlet P., Brochier C., Quevillon-Cheruel S., Le Cam E., van Tilbeurgh H.et al.. An archaeal orthologue of the universal protein Kae1 is an iron metalloprotein which exhibits atypical DNA-binding properties and apurinic-endonuclease activity in vitro. Nucleic Acids Res. 2007; 35:6042–6051. [DOI] [PMC free article] [PubMed] [Google Scholar]
47. Tominaga T., Kobayashi K., Ishii R., Ishitani R., Nureki O.. Structure of saccharomyces cerevisiae mitochondrial Qri7 in complex with AMP. Acta Crystallogr. Sect. F Struct. Biol. Commun. 2014; 70:1009–1014. [DOI] [PMC free article] [PubMed] [Google Scholar]
48. Li Q.X., Liu J.C., He M.H., Zhou J.Q.. Kae1 of Saccharomyces cerevisiae KEOPS complex possesses ADP/GDP nucleotidase activity. Biochem. J. 2022; 479:2433–2447. [DOI] [PubMed] [Google Scholar]
49. Kopina B.J., Missoury S., Collinet B., Fulton M.G., Cirio C., van Tilbeurgh H., Lauhon C.T.. Structure of a reaction intermediate mimic in t6A biosynthesis bound in the active site of the TsaBD heterodimer from Escherichia coli. Nucleic Acids Res. 2021; 49:2141–2160. [DOI] [PMC free article] [PubMed] [Google Scholar]
50. Miyauchi K., Kimura S., Suzuki T.. A cyclic form of N6-threonylcarbamoyladenosine as a widely distributed tRNA hypermodification. Nat. Chem. Biol. 2013; 9:105–111. [DOI] [PubMed] [Google Scholar]
51. Matuszewski M., Wojciechowski J., Miyauchi K., Gdaniec Z., Wolf W.M., Suzuki T., Sochacka E.. A hydantoin isoform of cyclic N6-threonylcarbamoyladenosine (ct6A) is present in tRNAs. Nucleic Acids Res. 2017; 45:2137–2149. [DOI] [PMC free article] [PubMed] [Google Scholar]
52. Bolstad H.M., Botelho D.J., Wood M.J.. Proteomic analysis of protein-protein interactions within the Cysteine Sulfinate Desulfinase Fe-S cluster biogenesis system. J. Proteome Res. 2010; 9:5358–5369. [DOI] [PMC free article] [PubMed] [Google Scholar]
53. Lopez-Estepa M., Arda A., Savko M., Round A., Shepard W.E., Bruix M., Coll M., Fernandez F.J., Jimenez-Barbero J., Vega M.C.. The crystal structure and small-angle X-ray analysis of CsdL/TcdA reveal a new tRNA binding motif in the MoeB/E1 superfamily. PLoS One. 2015; 10:e0118606. [DOI] [PMC free article] [PubMed] [Google Scholar]
54. Kim S., Lee H., Park S.. The structure of Escherichia coli TcdA (Also Known As CsdL) reveals a novel topology and provides insight into the tRNA binding surface required for N(6)-threonylcarbamoyladenosine dehydratase activity. J. Mol. Biol. 2015; 427:3074–3085. [DOI] [PubMed] [Google Scholar]
55. Duda D.M., Walden H., Sfondouris J., Schulman B.A.. Structural analysis of Escherichia coli ThiF. J. Mol. Biol. 2005; 349:774–786. [DOI] [PubMed] [Google Scholar]
56. Lake M.W., Wuebbens M.M., Rajagopalan K.V., Schindelin H.. Mechanism of ubiquitin activation revealed by the structure of a bacterial MoeB-MoaD complex. Nature. 2001; 414:325–329. [DOI] [PubMed] [Google Scholar]
57. Aravind L., Koonin E.V.. Gleaning non-trivial structural, functional and evolutionary information about proteins by iterative database searches. J. Mol. Biol. 1999; 287:1023–1040. [DOI] [PubMed] [Google Scholar]
58. Shomura Y., Higuchi Y.. Structural basis for the reaction mechanism of S-carbamoylation of HypE by HypF in the maturation of [NiFe]-hydrogenases. J. Biol. Chem. 2012; 287:28409–28419. [DOI] [PMC free article] [PubMed] [Google Scholar]
59. Philippe N., Legendre M., Doutre G., Coute Y., Poirot O., Lescot M., Arslan D., Seltzer V., Bertaux L., Bruley C.et al.. Pandoraviruses: amoeba viruses with genomes up to 2.5 mb reaching that of parasitic eukaryotes. Science. 2013; 341:281–286. [DOI] [PubMed] [Google Scholar]
60. Sievers F., Wilm A., Dineen D., Gibson T.J., Karplus K., Li W., Lopez R., McWilliam H., Remmert M., Soding J.et al.. Fast, scalable generation of high-quality protein multiple sequence alignments using Clustal Omega. Mol. Syst. Biol. 2011; 7:539. [DOI] [PMC free article] [PubMed] [Google Scholar]
61. Zhang W., Collinet B., Graille M., Daugeron M.C., Lazar N., Libri D., Durand D., van Tilbeurgh H.. Crystal structures of the Gon7/Pcc1 and Bud32/Cgi121 complexes provide a model for the complete yeast KEOPS complex. Nucleic Acids Res. 2015; 43:3358–3372. [DOI] [PMC free article] [PubMed] [Google Scholar]
62. Li X., Mooney P., Zheng S., Booth C.R., Braunfeld M.B., Gubbens S., Agard D.A., Cheng Y.. Electron counting and beam-induced motion correction enable near-atomic-resolution single-particle cryo-EM. Nat. Methods. 2013; 10:584–590. [DOI] [PMC free article] [PubMed] [Google Scholar]
63. Mindell J.A., Grigorieff N.. Accurate determination of local defocus and specimen tilt in electron microscopy. J. Struct. Biol. 2003; 142:334–347. [DOI] [PubMed] [Google Scholar]
64. Punjani A., Rubinstein J.L., Fleet D.J., Brubaker M.A.. cryoSPARC: algorithms for rapid unsupervised cryo-EM structure determination. Nat. Methods. 2017; 14:290–296. [DOI] [PubMed] [Google Scholar]
65. DiIorio M.C., Kulczyk A.W.. A robust single-particle cryo-electron microscopy (cryo-EM) processing workflow with cryoSPARC, RELION, and scipion. J. Vis. Exp. 2022; 179:e63387. [DOI] [PubMed] [Google Scholar]
66. Liebschner D., Afonine P.V., Baker M.L., Bunkoczi G., Chen V.B., Croll T.I., Hintze B., Hung L.W., Jain S., McCoy A.J.et al.. Macromolecular structure determination using X-rays, neutrons and electrons: recent developments in Phenix. Acta Crystallogr. D Struct. Biol. 2019; 75:861–877. [DOI] [PMC free article] [PubMed] [Google Scholar]
67. Emsley P., Lohkamp B., Scott W.G., Cowtan K.. Features and development of Coot. Acta Crystallogr. D Biol. Crystallogr. 2010; 66:486–501. [DOI] [PMC free article] [PubMed] [Google Scholar]
68. Williams C.J., Headd J.J., Moriarty N.W., Prisant M.G., Videau L.L., Deis L.N., Verma V., Keedy D.A., Hintze B.J., Chen V.B.et al.. MolProbity: more and better reference data for improved all-atom structure validation. Protein Sci. 2018; 27:293–315. [DOI] [PMC free article] [PubMed] [Google Scholar]
69. Pettersen E.F., Goddard T.D., Huang C.C., Couch G.S., Greenblatt D.M., Meng E.C., Ferrin T.E.. UCSF Chimera–a visualization system for exploratory research and analysis. J. Comput. Chem. 2004; 25:1605–1612. [DOI] [PubMed] [Google Scholar]
70. Konecny R., Baker N.A., McCammon J.A.. iAPBS: a programming interface to Adaptive Poisson-Boltzmann Solver (APBS). Comput. Sci. Discov. 2012; 5:015005. [DOI] [PMC free article] [PubMed] [Google Scholar]
71. Krissinel E., Henrick K.. Inference of macromolecular assemblies from crystalline state. J. Mol. Biol. 2007; 372:774–797. [DOI] [PubMed] [Google Scholar]
72. Liu Y.Y., He M.H., Liu J.C., Lu Y.S., Peng J., Zhou J.Q.. Yeast KEOPS complex regulates telomere length independently of its t(6)A modification function. J. Genet. Genomics. 2018; 45:247–257. [DOI] [PubMed] [Google Scholar]
73. Legendre M., Fabre E., Poirot O., Jeudy S., Lartigue A., Alempic J.M., Beucher L., Philippe N., Bertaux L., Christo-Foroux E.et al.. Diversity and evolution of the emerging Pandoraviridae family. Nat. Commun. 2018; 9:2285. [DOI] [PMC free article] [PubMed] [Google Scholar]
74. Alempic J.M., Lartigue A., Goncharov A.E., Grosse G., Strauss J., Tikhonov A.N., Fedorov A.N., Poirot O., Legendre M., Santini S.et al.. An update on eukaryotic viruses revived from ancient permafrost. Viruses. 2023; 15:564. [DOI] [PMC free article] [PubMed] [Google Scholar]
75. Aherfi S., Brahim Belhaouari D., Pinault L., Baudoin J.P., Decloquement P., Abrahao J., Colson P., Levasseur A., Lamb D.C., Chabriere E.et al.. Incomplete tricarboxylic acid cycle and proton gradient in Pandoravirus massiliensis: is it still a virus?. ISME J. 2022; 16:695–704. [DOI] [PMC free article] [PubMed] [Google Scholar]
76. Schulz F., Abergel C., Woyke T.. Giant virus biology and diversity in the era of genome-resolved metagenomics. Nat. Rev. Microbiol. 2022; 20:721–736. [DOI] [PubMed] [Google Scholar]
77. Irwin N.A.T., Pittis A.A., Richards T.A., Keeling P.J.. Systematic evaluation of horizontal gene transfer between eukaryotes and viruses. Nat. Microbiol. 2022; 7:327–336. [DOI] [PubMed] [Google Scholar]
78. Nasir A., Caetano-Anolles G.. A phylogenomic data-driven exploration of viral origins and evolution. Sci. Adv. 2015; 1:e1500527. [DOI] [PMC free article] [PubMed] [Google Scholar]
79. Clarke M., Lohan A.J., Liu B., Lagkouvardos I., Roy S., Zafar N., Bertelli C., Schilde C., Kianianmomeni A., Burglin T.R.et al.. Genome of Acanthamoeba castellanii highlights extensive lateral gene transfer and early evolution of tyrosine kinase signaling. Genome Biol. 2013; 14:R11. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

gkad587_Supplemental_File

Click here for additional data file.^{(32.5MB, pdf)}

Data Availability Statement

[B1] 1. Boccaletto P., Stefaniak F., Ray A., Cappannini A., Mukherjee S., Purta E., Kurkowska M., Shirvanizadeh N., Destefanis E., Groza P.et al.. MODOMICS: a database of RNA modification pathways. 2021 update. Nucleic Acids Res. 2022; 50:D231–D235. [DOI] [PMC free article] [PubMed] [Google Scholar]

[B2] 2. Machnicka M.A., Olchowik A., Grosjean H., Bujnicki J.M.. Distribution and frequencies of post-transcriptional modifications in tRNAs. RNA Biol. 2014; 11:1619–1629. [DOI] [PMC free article] [PubMed] [Google Scholar]

[B3] 3. Jackman J.E., Alfonzo J.D.. Transfer RNA modifications: nature's combinatorial chemistry playground. Wiley Interdiscip. Rev. RNA. 2013; 4:35–48. [DOI] [PMC free article] [PubMed] [Google Scholar]

[B4] 4. Grosjean H., Sprinzl M., Steinberg S.. Posttranscriptionally modified nucleosides in transfer RNA: their locations and frequencies. Biochimie. 1995; 77:139–141. [DOI] [PubMed] [Google Scholar]

[B5] 5. Parthasarathy R., Ohrt J.M., Chheda G.B.. Conformation and possible role of hypermodified nucleosides adjacent to 3'-end of anticodon in tRNA: n-(purin-6-ylcarbamoyl)-L-threonine riboside. Biochem. Biophys. Res. Commun. 1974; 60:211–218. [DOI] [PubMed] [Google Scholar]

[B6] 6. Stuart J.W., Gdaniec Z., Guenther R., Marszalek M., Sochacka E., Malkiewicz A., Agris P.F.. Functional anticodon architecture of human tRNALys3 includes disruption of intraloop hydrogen bonding by the naturally occurring amino acid modification, t6A. Biochemistry. 2000; 39:13396–13404. [DOI] [PubMed] [Google Scholar]

[B7] 7. Murphy F.V.t., Ramakrishnan V., Malkiewicz A., Agris P.F.. The role of modifications in codon discrimination by tRNA(Lys)UUU. Nat. Struct. Mol. Biol. 2004; 11:1186–1191. [DOI] [PubMed] [Google Scholar]

[B8] 8. Yarian C., Townsend H., Czestkowski W., Sochacka E., Malkiewicz A.J., Guenther R., Miskiewicz A., Agris P.F.. Accurate translation of the genetic code depends on tRNA modified nucleosides. J. Biol. Chem. 2002; 277:16391–16395. [DOI] [PubMed] [Google Scholar]

[B9] 9. Thiaville P.C., Legendre R., Rojas-Benitez D., Baudin-Baillieu A., Hatin I., Chalancon G., Glavic A., Namy O., de Crecy-Lagard V.. Global translational impacts of the loss of the tRNA modification t(6)A in yeast. Microb. Cell. 2016; 3:29–45. [DOI] [PMC free article] [PubMed] [Google Scholar]

[B10] 10. El Yacoubi B., Lyons B., Cruz Y., Reddy R., Nordin B., Agnelli F., Williamson J.R., Schimmel P., Swairjo M.A., de Crecy-Lagard V.. The universal YrdC/Sua5 family is required for the formation of threonylcarbamoyladenosine in tRNA. Nucleic Acids Res. 2009; 37:2894–2909. [DOI] [PMC free article] [PubMed] [Google Scholar]

[B11] 11. El Yacoubi B., Hatin I., Deutsch C., Kahveci T., Rousset J.P., Iwata-Reuyl D., Murzin A.G., de Crecy-Lagard V.. A role for the universal Kae1/Qri7/YgjD (COG0533) family in tRNA modification. EMBO J. 2011; 30:882–893. [DOI] [PMC free article] [PubMed] [Google Scholar]

[B12] 12. Thiaville P.C., El Yacoubi B., Kohrer C., Thiaville J.J., Deutsch C., Iwata-Reuyl D., Bacusmo J.M., Armengaud J., Bessho Y., Wetzel C.et al.. Essentiality of threonylcarbamoyladenosine (t(6)A), a universal tRNA modification, in bacteria. Mol. Microbiol. 2015; 98:1199–1221. [DOI] [PMC free article] [PubMed] [Google Scholar]

[B13] 13. Suzuki T. The expanding world of tRNA modifications and their disease relevance. Nat. Rev. Mol. Cell Biol. 2021; 22:375–392. [DOI] [PubMed] [Google Scholar]

[B14] 14. Braun D.A., Rao J., Mollet G., Schapiro D., Daugeron M.C., Tan W., Gribouval O., Boyer O., Revy P., Jobst-Schwan T.et al.. Mutations in KEOPS-complex genes cause nephrotic syndrome with primary microcephaly. Nat. Genet. 2017; 49:1529–1538. [DOI] [PMC free article] [PubMed] [Google Scholar]

[B15] 15. Arrondel C., Missoury S., Snoek R., Patat J., Menara G., Collinet B., Liger D., Durand D., Gribouval O., Boyer O.et al.. Defects in t(6)A tRNA modification due to GON7 and YRDC mutations lead to Galloway-Mowat syndrome. Nat. Commun. 2019; 10:3967. [DOI] [PMC free article] [PubMed] [Google Scholar]

[B16] 16. Weiss M.C., Sousa F.L., Mrnjavac N., Neukirchen S., Roettger M., Nelson-Sathi S., Martin W.F.. The physiology and habitat of the last universal common ancestor. Nat. Microbiol. 2016; 1:16116. [DOI] [PubMed] [Google Scholar]

[B17] 17. Schneider C., Becker S., Okamura H., Crisp A., Amatov T., Stadlmeier M., Carell T.. Noncanonical RNA nucleosides as molecular fossils of an early earth-generation by prebiotic methylations and carbamoylations. Angew Chem. Int. Ed. Engl. 2018; 57:5943–5946. [DOI] [PubMed] [Google Scholar]

[B18] 18. Galperin M.Y., Koonin E.V.. Conserved hypothetical’ proteins: prioritization of targets for experimental study. Nucleic Acids Res. 2004; 32:5452–5463. [DOI] [PMC free article] [PubMed] [Google Scholar]

[B19] 19. Su C., Jin M., Zhang W.. Conservation and diversification of tRNA t6A-modifying enzymes across the three domains of life. Int. J. Mol. Sci. 2022; 23:13600. [DOI] [PMC free article] [PubMed] [Google Scholar]

[B20] 20. Lauhon C.T. Mechanism of N6-threonylcarbamoyladenonsine (t(6)A) biosynthesis: isolation and characterization of the intermediate threonylcarbamoyl-AMP. Biochemistry. 2012; 51:8950–8963. [DOI] [PubMed] [Google Scholar]

[B21] 21. Wan L.C., Mao D.Y., Neculai D., Strecker J., Chiovitti D., Kurinov I., Poda G., Thevakumaran N., Yuan F., Szilard R.K.et al.. Reconstitution and characterization of eukaryotic N6-threonylcarbamoylation of tRNA using a minimal enzyme system. Nucleic Acids Res. 2013; 41:6332–6346. [DOI] [PMC free article] [PubMed] [Google Scholar]

[B22] 22. Zhou J.B., Wang Y., Zeng Q.Y., Meng S.X., Wang E.D., Zhou X.L.. Molecular basis for t6A modification in human mitochondria. Nucleic Acids Res. 2020; 48:3181–3194. [DOI] [PMC free article] [PubMed] [Google Scholar]

[B23] 23. Beenstock J., Sicheri F.. The structural and functional workings of KEOPS. Nucleic Acids Res. 2021; 49:10818–10834. [DOI] [PMC free article] [PubMed] [Google Scholar]

[B24] 24. Thiaville P.C., Iwata-Reuyl D., de Crecy-Lagard V.. Diversity of the biosynthesis pathway for threonylcarbamoyladenosine (t(6)A), a universal modification of tRNA. RNA Biol. 2014; 11:1529–1539. [DOI] [PMC free article] [PubMed] [Google Scholar]

[B25] 25. Deutsch C., Yacoubi E., B. d.C.-L., Iwata-Reuyl D. Biosynthesis of threonylcarbamoyl adenosine (t6A), a universal tRNA nucleoside. J. Biol. Chem. 2012; 287:13666–13673. [DOI] [PMC free article] [PubMed] [Google Scholar]

[B26] 26. Zhang W., Collinet B., Perrochia L., Durand D., van Tilbeurgh H.. The ATP-mediated formation of the YgjD-YeaZ-YjeE complex is required for the biosynthesis of tRNA t6A in Escherichia coli. Nucleic Acids Res. 2015; 43:1804–1817. [DOI] [PMC free article] [PubMed] [Google Scholar]

[B27] 27. Luthra A., Swinehart W., Bayooz S., Phan P., Stec B., Iwata-Reuyl D., Swairjo M.A.. Structure and mechanism of a bacterial t6A biosynthesis system. Nucleic Acids Res. 2018; 46:1395–1411. [DOI] [PMC free article] [PubMed] [Google Scholar]

[B28] 28. Perrochia L., Crozat E., Hecker A., Zhang W., Bareille J., Collinet B., van Tilbeurgh H., Forterre P., Basta T.. In vitro biosynthesis of a universal t6A tRNA modification in Archaea and Eukarya. Nucleic Acids Res. 2013; 41:1953–1964. [DOI] [PMC free article] [PubMed] [Google Scholar]

[B29] 29. Luthra A., Paranagama N., Swinehart W., Bayooz S., Phan P., Quach V., Schiffer J.M., Stec B., Iwata-Reuyl D., Swairjo M.A.. Conformational communication mediates the reset step in t6A biosynthesis. Nucleic Acids Res. 2019; 47:6551–6567. [DOI] [PMC free article] [PubMed] [Google Scholar]

[B30] 30. Missoury S., Plancqueel S., Li de la Sierra-Gallay I., Zhang W., Liger D., Durand D., Dammak R., Collinet B., van Tilbeurgh H.. The structure of the TsaB/TsaD/TsaE complex reveals an unexpected mechanism for the bacterial t6A tRNA-modification. Nucleic Acids Res. 2018; 46:5850–5860. [DOI] [PMC free article] [PubMed] [Google Scholar]

[B31] 31. Perrochia L., Guetta D., Hecker A., Forterre P., Basta T.. Functional assignment of KEOPS/EKC complex subunits in the biosynthesis of the universal t6A tRNA modification. Nucleic Acids Res. 2013; 41:9484–9499. [DOI] [PMC free article] [PubMed] [Google Scholar]

[B32] 32. Wan L.C., Pillon M.C., Thevakumaran N., Sun Y., Chakrabartty A., Guarne A., Kurinov I., Durocher D., Sicheri F.. Structural and functional characterization of KEOPS dimerization by Pcc1 and its role in t6A biosynthesis. Nucleic Acids Res. 2016; 44:6971–6980. [DOI] [PMC free article] [PubMed] [Google Scholar]

[B33] 33. Beenstock J., Ona S.M., Porat J., Orlicky S., Wan L.C.K., Ceccarelli D.F., Maisonneuve P., Szilard R.K., Yin Z., Setiaputra D.et al.. A substrate binding model for the KEOPS tRNA modifying complex. Nat. Commun. 2020; 11:6233. [DOI] [PMC free article] [PubMed] [Google Scholar]

[B34] 34. Daugeron M.C., Missoury S., Da Cunha V., Lazar N., Collinet B., van Tilbeurgh H., Basta T.. A paralog of Pcc1 is the fifth core subunit of the KEOPS tRNA-modifying complex in Archaea. Nat. Commun. 2023; 14:526. [DOI] [PMC free article] [PubMed] [Google Scholar]

[B35] 35. Srinivasan M., Mehta P., Yu Y., Prugar E., Koonin E.V., Karzai A.W., Sternglanz R.. The highly conserved KEOPS/EKC complex is essential for a universal tRNA modification, t6A. EMBO J. 2011; 30:873–881. [DOI] [PMC free article] [PubMed] [Google Scholar]

[B36] 36. Wan L.C., Maisonneuve P., Szilard R.K., Lambert J.P., Ng T.F., Manczyk N., Huang H., Laister R., Caudy A.A., Gingras A.C.et al.. Proteomic analysis of the human KEOPS complex identifies C14ORF142 as a core subunit homologous to yeast Gon7. Nucleic Acids Res. 2017; 45:805–817. [DOI] [PMC free article] [PubMed] [Google Scholar]

[B37] 37. Wang J.T., Zhou J.B., Mao X.L., Zhou L., Chen M., Zhang W., Wang E.D., Zhou X.L.. Commonality and diversity in tRNA substrate recognition in t6A biogenesis by eukaryotic keopss. Nucleic Acids Res. 2022; 50:2223–2239. [DOI] [PMC free article] [PubMed] [Google Scholar]

[B38] 38. Downey M., Houlsworth R., Maringele L., Rollie A., Brehme M., Galicia S., Guillard S., Partington M., Zubko M.K., Krogan N.J.et al.. A genome-wide screen identifies the evolutionarily conserved KEOPS complex as a telomere regulator. Cell. 2006; 124:1155–1168. [DOI] [PubMed] [Google Scholar]

[B39] 39. Mao D.Y., Neculai D., Downey M., Orlicky S., Haffani Y.Z., Ceccarelli D.F., Ho J.S., Szilard R.K., Zhang W., Ho C.S.et al.. Atomic structure of the KEOPS complex: an ancient protein kinase-containing molecular machine. Mol. Cell. 2008; 32:259–275. [DOI] [PMC free article] [PubMed] [Google Scholar]

[B40] 40. Teplova M., Tereshko V., Sanishvili R., Joachimiak A., Bushueva T., Anderson W.F., Egli M.. The structure of the yrdC gene product from Escherichia coli reveals a new fold and suggests a role in RNA binding. Protein Sci. 2000; 9:2557–2566. [DOI] [PMC free article] [PubMed] [Google Scholar]

[B41] 41. Agari Y., Sato S., Wakamatsu T., Bessho Y., Ebihara A., Yokoyama S., Kuramitsu S., Shinkai A.. X-ray crystal structure of a hypothetical Sua5 protein from Sulfolobus tokodaii strain 7. Proteins. 2008; 70:1108–1111. [DOI] [PubMed] [Google Scholar]

[B42] 42. Kuratani M., Kasai T., Akasaka R., Higashijima K., Terada T., Kigawa T., Shinkai A., Bessho Y., Yokoyama S.. Crystal structure of Sulfolobus tokodaii Sua5 complexed with L-threonine and AMPPNP. Proteins. 2011; 79:2065–2075. [DOI] [PubMed] [Google Scholar]

[B43] 43. Pichard-Kostuch A., Zhang W., Liger D., Daugeron M.C., Letoquart J., Li de la Sierra-Gallay I., Forterre P., Collinet B., van Tilbeurgh H., Basta T.. Structure-function analysis of Sua5 protein reveals novel functional motifs required for the biosynthesis of the universal t(6)A tRNA modification. RNA. 2018; 24:926–938. [DOI] [PMC free article] [PubMed] [Google Scholar]

[B44] 44. Harris K.A., Bobay B.G., Sarachan K.L., Sims A.F., Bilbille Y., Deutsch C., Iwata-Reuyl D., Agris P.F.. NMR-based structural analysis of threonylcarbamoyl-AMP synthase and its substrate interactions. J. Biol. Chem. 2015; 290:20032–20043. [DOI] [PMC free article] [PubMed] [Google Scholar]

[B45] 45. Parthier C., Gorlich S., Jaenecke F., Breithaupt C., Brauer U., Fandrich U., Clausnitzer D., Wehmeier U.F., Bottcher C., Scheel D.et al.. The O-carbamoyltransferase TobZ catalyzes an ancient enzymatic reaction. Angew. Chem. Int. Ed. 2012; 51:4046–4052. [DOI] [PubMed] [Google Scholar]

[B46] 46. Hecker A., Leulliot N., Gadelle D., Graille M., Justome A., Dorlet P., Brochier C., Quevillon-Cheruel S., Le Cam E., van Tilbeurgh H.et al.. An archaeal orthologue of the universal protein Kae1 is an iron metalloprotein which exhibits atypical DNA-binding properties and apurinic-endonuclease activity in vitro. Nucleic Acids Res. 2007; 35:6042–6051. [DOI] [PMC free article] [PubMed] [Google Scholar]

[B47] 47. Tominaga T., Kobayashi K., Ishii R., Ishitani R., Nureki O.. Structure of saccharomyces cerevisiae mitochondrial Qri7 in complex with AMP. Acta Crystallogr. Sect. F Struct. Biol. Commun. 2014; 70:1009–1014. [DOI] [PMC free article] [PubMed] [Google Scholar]

[B48] 48. Li Q.X., Liu J.C., He M.H., Zhou J.Q.. Kae1 of Saccharomyces cerevisiae KEOPS complex possesses ADP/GDP nucleotidase activity. Biochem. J. 2022; 479:2433–2447. [DOI] [PubMed] [Google Scholar]

[B49] 49. Kopina B.J., Missoury S., Collinet B., Fulton M.G., Cirio C., van Tilbeurgh H., Lauhon C.T.. Structure of a reaction intermediate mimic in t6A biosynthesis bound in the active site of the TsaBD heterodimer from Escherichia coli. Nucleic Acids Res. 2021; 49:2141–2160. [DOI] [PMC free article] [PubMed] [Google Scholar]

[B50] 50. Miyauchi K., Kimura S., Suzuki T.. A cyclic form of N6-threonylcarbamoyladenosine as a widely distributed tRNA hypermodification. Nat. Chem. Biol. 2013; 9:105–111. [DOI] [PubMed] [Google Scholar]

[B51] 51. Matuszewski M., Wojciechowski J., Miyauchi K., Gdaniec Z., Wolf W.M., Suzuki T., Sochacka E.. A hydantoin isoform of cyclic N6-threonylcarbamoyladenosine (ct6A) is present in tRNAs. Nucleic Acids Res. 2017; 45:2137–2149. [DOI] [PMC free article] [PubMed] [Google Scholar]

[B52] 52. Bolstad H.M., Botelho D.J., Wood M.J.. Proteomic analysis of protein-protein interactions within the Cysteine Sulfinate Desulfinase Fe-S cluster biogenesis system. J. Proteome Res. 2010; 9:5358–5369. [DOI] [PMC free article] [PubMed] [Google Scholar]

[B53] 53. Lopez-Estepa M., Arda A., Savko M., Round A., Shepard W.E., Bruix M., Coll M., Fernandez F.J., Jimenez-Barbero J., Vega M.C.. The crystal structure and small-angle X-ray analysis of CsdL/TcdA reveal a new tRNA binding motif in the MoeB/E1 superfamily. PLoS One. 2015; 10:e0118606. [DOI] [PMC free article] [PubMed] [Google Scholar]

[B54] 54. Kim S., Lee H., Park S.. The structure of Escherichia coli TcdA (Also Known As CsdL) reveals a novel topology and provides insight into the tRNA binding surface required for N(6)-threonylcarbamoyladenosine dehydratase activity. J. Mol. Biol. 2015; 427:3074–3085. [DOI] [PubMed] [Google Scholar]

[B55] 55. Duda D.M., Walden H., Sfondouris J., Schulman B.A.. Structural analysis of Escherichia coli ThiF. J. Mol. Biol. 2005; 349:774–786. [DOI] [PubMed] [Google Scholar]

[B56] 56. Lake M.W., Wuebbens M.M., Rajagopalan K.V., Schindelin H.. Mechanism of ubiquitin activation revealed by the structure of a bacterial MoeB-MoaD complex. Nature. 2001; 414:325–329. [DOI] [PubMed] [Google Scholar]

[B57] 57. Aravind L., Koonin E.V.. Gleaning non-trivial structural, functional and evolutionary information about proteins by iterative database searches. J. Mol. Biol. 1999; 287:1023–1040. [DOI] [PubMed] [Google Scholar]

[B58] 58. Shomura Y., Higuchi Y.. Structural basis for the reaction mechanism of S-carbamoylation of HypE by HypF in the maturation of [NiFe]-hydrogenases. J. Biol. Chem. 2012; 287:28409–28419. [DOI] [PMC free article] [PubMed] [Google Scholar]

[B59] 59. Philippe N., Legendre M., Doutre G., Coute Y., Poirot O., Lescot M., Arslan D., Seltzer V., Bertaux L., Bruley C.et al.. Pandoraviruses: amoeba viruses with genomes up to 2.5 mb reaching that of parasitic eukaryotes. Science. 2013; 341:281–286. [DOI] [PubMed] [Google Scholar]

[B60] 60. Sievers F., Wilm A., Dineen D., Gibson T.J., Karplus K., Li W., Lopez R., McWilliam H., Remmert M., Soding J.et al.. Fast, scalable generation of high-quality protein multiple sequence alignments using Clustal Omega. Mol. Syst. Biol. 2011; 7:539. [DOI] [PMC free article] [PubMed] [Google Scholar]

[B61] 61. Zhang W., Collinet B., Graille M., Daugeron M.C., Lazar N., Libri D., Durand D., van Tilbeurgh H.. Crystal structures of the Gon7/Pcc1 and Bud32/Cgi121 complexes provide a model for the complete yeast KEOPS complex. Nucleic Acids Res. 2015; 43:3358–3372. [DOI] [PMC free article] [PubMed] [Google Scholar]

[B62] 62. Li X., Mooney P., Zheng S., Booth C.R., Braunfeld M.B., Gubbens S., Agard D.A., Cheng Y.. Electron counting and beam-induced motion correction enable near-atomic-resolution single-particle cryo-EM. Nat. Methods. 2013; 10:584–590. [DOI] [PMC free article] [PubMed] [Google Scholar]

[B63] 63. Mindell J.A., Grigorieff N.. Accurate determination of local defocus and specimen tilt in electron microscopy. J. Struct. Biol. 2003; 142:334–347. [DOI] [PubMed] [Google Scholar]

[B64] 64. Punjani A., Rubinstein J.L., Fleet D.J., Brubaker M.A.. cryoSPARC: algorithms for rapid unsupervised cryo-EM structure determination. Nat. Methods. 2017; 14:290–296. [DOI] [PubMed] [Google Scholar]

[B65] 65. DiIorio M.C., Kulczyk A.W.. A robust single-particle cryo-electron microscopy (cryo-EM) processing workflow with cryoSPARC, RELION, and scipion. J. Vis. Exp. 2022; 179:e63387. [DOI] [PubMed] [Google Scholar]

[B66] 66. Liebschner D., Afonine P.V., Baker M.L., Bunkoczi G., Chen V.B., Croll T.I., Hintze B., Hung L.W., Jain S., McCoy A.J.et al.. Macromolecular structure determination using X-rays, neutrons and electrons: recent developments in Phenix. Acta Crystallogr. D Struct. Biol. 2019; 75:861–877. [DOI] [PMC free article] [PubMed] [Google Scholar]

[B67] 67. Emsley P., Lohkamp B., Scott W.G., Cowtan K.. Features and development of Coot. Acta Crystallogr. D Biol. Crystallogr. 2010; 66:486–501. [DOI] [PMC free article] [PubMed] [Google Scholar]

[B68] 68. Williams C.J., Headd J.J., Moriarty N.W., Prisant M.G., Videau L.L., Deis L.N., Verma V., Keedy D.A., Hintze B.J., Chen V.B.et al.. MolProbity: more and better reference data for improved all-atom structure validation. Protein Sci. 2018; 27:293–315. [DOI] [PMC free article] [PubMed] [Google Scholar]

[B69] 69. Pettersen E.F., Goddard T.D., Huang C.C., Couch G.S., Greenblatt D.M., Meng E.C., Ferrin T.E.. UCSF Chimera–a visualization system for exploratory research and analysis. J. Comput. Chem. 2004; 25:1605–1612. [DOI] [PubMed] [Google Scholar]

[B70] 70. Konecny R., Baker N.A., McCammon J.A.. iAPBS: a programming interface to Adaptive Poisson-Boltzmann Solver (APBS). Comput. Sci. Discov. 2012; 5:015005. [DOI] [PMC free article] [PubMed] [Google Scholar]

[B71] 71. Krissinel E., Henrick K.. Inference of macromolecular assemblies from crystalline state. J. Mol. Biol. 2007; 372:774–797. [DOI] [PubMed] [Google Scholar]

[B72] 72. Liu Y.Y., He M.H., Liu J.C., Lu Y.S., Peng J., Zhou J.Q.. Yeast KEOPS complex regulates telomere length independently of its t(6)A modification function. J. Genet. Genomics. 2018; 45:247–257. [DOI] [PubMed] [Google Scholar]

[B73] 73. Legendre M., Fabre E., Poirot O., Jeudy S., Lartigue A., Alempic J.M., Beucher L., Philippe N., Bertaux L., Christo-Foroux E.et al.. Diversity and evolution of the emerging Pandoraviridae family. Nat. Commun. 2018; 9:2285. [DOI] [PMC free article] [PubMed] [Google Scholar]

[B74] 74. Alempic J.M., Lartigue A., Goncharov A.E., Grosse G., Strauss J., Tikhonov A.N., Fedorov A.N., Poirot O., Legendre M., Santini S.et al.. An update on eukaryotic viruses revived from ancient permafrost. Viruses. 2023; 15:564. [DOI] [PMC free article] [PubMed] [Google Scholar]

[B75] 75. Aherfi S., Brahim Belhaouari D., Pinault L., Baudoin J.P., Decloquement P., Abrahao J., Colson P., Levasseur A., Lamb D.C., Chabriere E.et al.. Incomplete tricarboxylic acid cycle and proton gradient in Pandoravirus massiliensis: is it still a virus?. ISME J. 2022; 16:695–704. [DOI] [PMC free article] [PubMed] [Google Scholar]

[B76] 76. Schulz F., Abergel C., Woyke T.. Giant virus biology and diversity in the era of genome-resolved metagenomics. Nat. Rev. Microbiol. 2022; 20:721–736. [DOI] [PubMed] [Google Scholar]

[B77] 77. Irwin N.A.T., Pittis A.A., Richards T.A., Keeling P.J.. Systematic evaluation of horizontal gene transfer between eukaryotes and viruses. Nat. Microbiol. 2022; 7:327–336. [DOI] [PubMed] [Google Scholar]

[B78] 78. Nasir A., Caetano-Anolles G.. A phylogenomic data-driven exploration of viral origins and evolution. Sci. Adv. 2015; 1:e1500527. [DOI] [PMC free article] [PubMed] [Google Scholar]

[B79] 79. Clarke M., Lohan A.J., Liu B., Lagkouvardos I., Roy S., Zafar N., Bertelli C., Schilde C., Kianianmomeni A., Burglin T.R.et al.. Genome of Acanthamoeba castellanii highlights extensive lateral gene transfer and early evolution of tyrosine kinase signaling. Genome Biol. 2013; 14:R11. [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

Structure–function analysis of an ancient TsaD–TsaC–SUA5–TcdA modular enzyme reveals a prototype of tRNA t6A and ct6A synthetases

Mengqi Jin

Zelin Zhang

Zhijiang Yu

Wei Chen

Xiaolei Wang

Dongsheng Lei

Wenhua Zhang

Abstract

Graphical Abstract

Graphical Abstract.

INTRODUCTION

Figure 1.

MATERIALS AND METHODS

Bioinformatics

Cloning, mutagenesis and expression procedures

Cryo-EM sample preparation and data collection

Cryo-EM data processing

Model building, refinement and validation

Table 1.

SEC coupled multi-angle light scattering (SEC-MALS)

Preparation of tRNA

Chemical synthesis of t6A

Enzymatic assays for biosynthesis of TC-AMP, t6A and ct6A

Preparation of TC-AMP compound and nucleotide threonylcarbamoylation assay

Analysis of the formation of t6ADP and t6ATP in bacterial cells

Liquid chromatography−mass spectrometry analysis

Isothermal titration calorimetry

ATPase assay

Electrophoretic mobility shift assays

Pyrophosphate assay

RESULTS

TsaD–TsaC–SUA5–TcdA fusion protein encoded in Pandoravirus genomes

Figure 2.

Purification and characterization of recombinant TsaN and truncated domains

Figure 3.

The cryo-EM structure of TsaN

Table 2.

Enzymatic synthesis of TC-AMP by TsaN

Figure 4.

Characterization of the nucleotide binding property of the TsaD domain

Figure 5.

Table 3.

TsaN does not catalyze tRNA t6A biosynthesis

Biosynthesis of t6ATP and t6ADP by the TsaD domain of TsaN

Conversion of t6A to ct6A by the TcdA domain of TsaN

Figure 6.

DISCUSSION

CONCLUSION

Supplementary Material

ACKNOWLEDGEMENTS

Contributor Information

DATA AVAILABILITY

SUPPLEMENTARY DATA

FUNDING

REFERENCES

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases

Structure–function analysis of an ancient TsaD–TsaC–SUA5–TcdA modular enzyme reveals a prototype of tRNA t⁶A and ct⁶A synthetases

Chemical synthesis of t⁶A

Enzymatic assays for biosynthesis of TC-AMP, t⁶A and ct⁶A

Analysis of the formation of t⁶ADP and t⁶ATP in bacterial cells

TsaN does not catalyze tRNA t⁶A biosynthesis

Biosynthesis of t⁶ATP and t⁶ADP by the TsaD domain of TsaN

Conversion of t⁶A to ct⁶A by the TcdA domain of TsaN