Mitoribosome structure with cofactors and modifications reveals mechanism of ligand binding and interactions with L1 stalk

Abstract

The mitoribosome translates mitochondrial mRNAs and regulates energy conversion that is a signature of aerobic life forms. We present a 2.2 Å resolution structure of human mitoribosome together with validated mitoribosomal RNA (rRNA) modifications, including aminoacylated CP-tRNA^Val. The structure shows how mitoribosomal proteins stabilise binding of mRNA and tRNA helping to align it in the decoding center, whereas the GDP-bound mS29 stabilizes intersubunit communication. Comparison between different states, with respect to tRNA position, allowed us to characterize a non-canonical L1 stalk, and molecular dynamics simulations revealed how it facilitates tRNA transitions in a way that does not require interactions with rRNA. We also report functionally important polyamines that are depleted when cells are subjected to an antibiotic treatment. The structural, biochemical, and computational data illuminate the principal functional components of the translation mechanism in mitochondria and provide a description of the structure and function of the human mitoribosome.

Fig. 1. Structure of the human mitoribosome.

a Overview of the 2.2 Å resolution model with mRNA and tRNAs. Proteins are shown as cartoons colored in blue for the LSU and yellow for the SSU. The rRNA is shown as grey surface. Newly identified features are shown as spheres. b The newly identified cofactors and modifications are indicated against background of translucent rRNA and proteins with tRNAs and mRNA shown in cartoon. We observe 13 rRNA modifications (green), 6 protein modifications (pink), 3 iron-sulfur clusters (red-yellow), guanosine diphosphate (GDP), adenosine triphosphate (ATP), NAD, SPM, 4 SPDs and PUT.

Fig. 2. High resolution features.

Modified rRNA residues and disulfide-linked or reduced cysteine residues of mS37 in the SSU (yellow background). Modified rRNA and protein residues, and 2Fe-2S cluster in the LSU (cyan background). Their corresponding densities in the 2.2 Å map are shown as mesh at indicated threshold (bottom left) and local resolution (bottom right) levels.

Fig. 3. Interactions and post-transcriptional modifications in CP-tRNA of the mitoribosome.

a CP-tRNA^Val (left to right) secondary structure diagram with modified nucleotides (green) and valine (cyan) highlighted, protein-contact map, B-factor representation. b, c Interactions of the modified m¹A9 and m²G10 nucleotides of CP-tRNA^Val. d Interactions of A76 and V77 of CP-tRNA^Val with mL46. The cryo-EM density map is displayed for panels (b–d) (consensus map) as blue mesh. Panels (b, c) are related to panel (a) on the left by colored boxes.

Fig. 4. Polyamines functionally compensate for rRNA and protein alterations.

Central inset shows positions of polyamines and interacting rRNA and proteins in the LSU. For each polyamine, its environment and interacting partners are shown in zoom-in panels with E. coli counterparts (white) for comparison.

Fig. 5. Specific protein elements involved in mRNA binding.

a Electrostatic coulomb potential surface representation of the model shows a positively charged groove (blue) on mS39 that accommodates the mRNA. N-terminal loop of mS31 interacts with mS39. Y-N-C-Y motif and its interactions with mS39 residues are depicted as a schematic. Position of kink-2 is marked. b Interactions of the Y-N-C-Y motif with mS39 and density map (mesh). Protein mS31 has been removed for clarity. c Average percentage frequency of Y-N-C-Y motif (left) with +/− standard deviation for all mitochondrial mRNAs except ND6 (n = 12 biologically independent samples;; dark grey bar) and nuclear mRNAs (n = 1387 biologically independent samples; brown bar). P-value (‘****’ indicates p < 0.0001) calculated by Welch’s unpaired two-tailed t-test. Positions of the Y-N-C-Y motif in the mt-mRNAs are shown as discs (right) to depict its distribution along the mt-mRNA sequences. The total number of residues for each mt-mRNA is indicated. d Overview of uS5m showing a region of contact with mRNA. The polybasic stretch 111–124 along the mRNA is colored by electrostatic potential. Interacting residues between uS5m and mRNA are shown with the local density map. e Interactions of mS35 and uS9m with mRNA at the channel entry and the P-site. f, R46 and R48 of mS35 potentially interact with the mRNA backbone, and the density suggests flexibility of the side-chains. g uS9m forms stacking and hydrogen bonding interactions with mRNA position 15 via F54 and V55, respectively. h uS9m N-terminus (purple cartoon) adopts alternative conformations that result in mRNA channel blocked or open states regulating mRNA access to the SSU (gold surface).

Fig. 6. Analysis of mRNA binding in tRNA sites.

a Interactions of mL40 (shown with electrostatic potential of the model) with mRNA between A/A- and P/P-tRNAs. Hydrogen bonds G6 base and G7 phosphate are shown as dashed lines. b Comparison of mL40-tRNA interaction (left) with those of uS19 (middle) and uS13 (right) in human cytosolic and T. thermophilus ribosome, respectively.

Fig. 7. Post-transcriptional modifications in functional centers of the mitoribosome.

a Base-methylated m⁴C1486 and m⁵C1488 and the interactions at the P-site mRNA codon. b The dimethylated m⁶₂A1583, m⁶₂A1584 and base-methylated m⁵1076 facilitate interactions with mRNA codon and P-tRNA. c 2′-O-methyl modified Um3039 and Gm3040 lead to interactions with A-tRNA. d Gm3040 and Ψ3067 contribute to the interface with the A-tRNA. e m¹A2617 contributes to a stabilization of inter-subunit bridge and the A-loop. f Gm2815 base-pairs with P-tRNA. Cryo-EM density map (A/A P/P E/E state) is displayed as a blue mesh.

Fig. 8. Structural features of the L1 stalk and interactions with tRNA.

a The L1 stalk components uL9m (green), uL1m (purple) rRNA helix H76-77 (residues 2757-2790, light blue), as well as mL64 (magenta) interact with E/E-tRNA. The structure and interactions are different from T. thermophilus (PDB ID: 4V51) [10.2210/pdb4V51/pdb] b Components of the L1 stalk modeled into their respective densities (shown as mesh). c Comparison of uL9m conformation with that of T. thermophilus uL9 (white PDB ID 4V51). d Model of uL1m stalk with uL9m in green and uL1m in purple, highlighting mitochondria-specific protein elements. e 2D diagram of uL9m showing interactions with E/E-tRNA (brown bars), rRNA (light-blue bars), uL1m (dark-blue bars). Conserved regions are shown in light and mitochondria-specific in dark colors. f 2D diagram of uL1m color-coded as in panel d.

Fig. 9. Simulations of P/E formation reveal the influence of mitochondria-specific proteins.

a Distances R_elbow and R_CCA are used to describe the position of the tRNA during P/E formation. b Probability distribution as a function of P-site tRNA elbow position (R_elbow) and the fraction of uL1m-tRNA contacts (Q_L1-tRNA) relative to those formed in the E/E state, calculated from simulations that used the baseline model (left). When uL1m-P-tRNA contacts are attenuated (ΔuL1m model; right), Q_L1-tRNA remains low, consistent with a lack of uL1m binding. c In simulations of P/E formation, perturbations to uL1m and/or mL64 (models ΔuL1m, ΔmL64, and ΔmL64-ΔuL1m models) lead to a wider range of tRNA elbow positions and longer timescales for P/E formation. The narrower range of pathways accessed with the complete/baseline model leads to a lower average value of R_elbow as a function of the tRNA CCA position R_CCA,showing how the corridor favors a process where the elbow motion precedes CCA movement. d Probability distribution (327 simulated events) as a function of elbow and 3′-CCA position reveals distinct intermediates during P/E formation. Representative structural snapshots of I1-I3 are shown in panels (e–g). e Intermediate I1 is associated with a small displacement of the tRNA elbow and CCA tail toward the E site. f In order to reach I2, the tRNA elbow is displaced by an additional ~10 Å, where it it confined by contacts with mL64. g mL64 then guides the tRNA elbow towards the E site, allowing it to adopt I3. After reaching I3, a final ~40 Å displacement of the CCA tail results in adoption of the P/E conformation.

Fig. 10. Binding of GDP to mS29 and role in mitoribosome function.

a Superposition of GDP-bound (yellow) and GMPPNP-bound (white) states. b Close-up view illustrating side chains that coordinate nucleotide binding. c, Intersubunit contacts in the classical and hybrid states mediated by mS29 (cryo-EM density map shown as colored surface). d Immunoblot analysis comparing steady-state levels of proteins using whole cell lysates from HEK293T (WT), mS29-KO (KO), and mS29-KO cells stably expressing either WT mS29 or double mutants targeting the GDP-mS29β-hairpin binding site under the control of an attenuated CMV6 promoter (∆5pCMV6). COX1, COX2, and CTYB were used as markers of mitoribosome function, and β-Actin was used as a loading control. e Images in panel (d) were digitalized, and densitometry was conducted using the histogram function in Adobe Photoshop. The bar graphs represent the mean ± SD from three independent repetitions (n = 3). Statistical differences were estimated using one-way ANOVA with Dunnett’s multiple comparisons test (two-tailed) in GraphPad Prism. *p < 0.05, **p < 0.01; ***p < 0.001, ****p < 0.0001. f Metabolic labeling of mitochondrially translated peptides in whole cells from the indicated cell lines using ³⁵S-methionine for 15 min in the presence of emetine to inhibit cytosolic protein synthesis. Immunoblotting for β-ACTIN was used as a loading control. Newly synthesized peptides are identified on the left. g Images in panel (d) were digitalized, and densitometry was conducted using the histogram function in Adobe Photoshop. The bar graphs represent the mean ± SD from four independent repetitions (n = 4). Statistical differences were estimated using one-way ANOVA with Dunnett’s multiple comparisons test (two-tailed) in GraphPad Prism. *p < 0.05, **p < 0.01; ***p < 0.001, ****p < 0.0001. The source data file contains all original uncropped and unprocessed immunoblots as well as densitometry values used to perform statistics and generate quantification graphs in panels (d–g).