Coupling of ribosome biogenesis and translation initiation in human mitochondria

Abstract

Biogenesis of mitoribosomes requires dedicated chaperones, RNA-modifying enzymes, and GTPases, and defects in mitoribosome assembly lead to severe mitochondriopathies in humans. Here, we characterize late-step assembly states of the small mitoribosomal subunit (mtSSU) by combining genetic perturbation and mutagenesis analysis with biochemical and structural approaches. Isolation of native mtSSU biogenesis intermediates via a FLAG-tagged variant of the GTPase MTG3 reveals three distinct assembly states, which show how factors cooperate to mature the 12S rRNA. In addition, we observe four distinct primed initiation mtSSU states with an incompletely matured rRNA, suggesting that biogenesis and translation initiation are not mutually exclusive processes but can occur simultaneously. Together, these results provide insights into mtSSU biogenesis and suggest a functional coupling between ribosome biogenesis and translation initiation in human mitochondria.

Fig. 1. Loss of MTG3 leads to disturbed small subunit assembly.

a Confirmation of MTG3 knock out in two cell lines generated using CRISPR/Cas9 technology. Isolated mitochondria (10 µg) from wildtype and Mtg3^−/− cell lines (cl.1 and cl.2) were analyzed by western blotting with antibodies as indicated. Similar results were obtained in n ≥ 3 biologically independent experiments. b Schematic representation of the genomic locus of the generated Mtg3^−/− cell line (cl.1) in comparison to the wild type sequence. The guide RNA targets exon 1 of the MTG3 gene, which encodes for a 648 aa protein. A two bp deletion and a four bp insertion in the two alleles of the Mtg3^−/− cl.1 lead to premature stop codons and truncated proteins (65 aa). c Ablation of MTG3 reduces growth rate. Equal amounts of wild type and Mtg3^−/− cells were seeded in three biologically independent experiments on day 0 and counted at the indicated time points (n = 3; mean ± SEM). Significance was calculated by two-sample one-tailed Student’s t-test and defined as **p ≤ 0.01. d Translation of mtDNA-encoded proteins is disturbed upon loss of MTG3. Mitochondrial translation in wild type, Mtg3^−/− and Mtg3^−/− cells inducibly expressing MTG3^FLAG was analyzed via [³⁵S]Methionine de novo incorporation and subsequently visualized via autoradiography and with indicated antibodies. The signal in Mtg3^−/− using MTG3 antibody represents unspecific binding of the antibody in whole cell lysates as we confirmed several times the loss of MTG3 in isolated mitochondria. Similar results were obtained in n ≥ 3 biologically independent experiments. e, f MTG3 loss leads to reduced mtSSU MRP level. Steady state analysis of MRPs, assembly factors, and translation-related proteins in the Mtg3^−/− cells in comparison to wild type cells. Isolated mitochondria were analyzed via western blotting with indicated antibodies (e) and protein levels in Mtg3^−/− were quantified relative to wild type control (f). SDHA was used as a loading control. Statistical analysis was performed as two-sample one-tailed Student’s t-test with n ≥ 3 biologically independent samples shown as mean ± SEM (individual data points are shown as circles). Significance was defined as *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001. g, h Effect of MTG3 loss on rRNA and mRNA stability. g RNA isolated from Mtg3^−/− and wild type cells was subjected to northern blotting using indicated probes (MT-RNR1: 12S rRNA; MT-RNR2: 16S rRNA; MT-CO1: mRNA encoding for COX1). 18S-rRNA was used as loading control. h Quantification of RNA signals in Mtg3^−/− from (g) relative to wild type signals. Statistical analysis was performed as two-sample one-tailed Student’s t-test with n = 3 biologically independent samples shown as mean ± SEM (individual data points are shown as circles). Significance was defined as ***p ≤ 0.001. i mtSSU and monosome levels are severely reduced in Mtg3^−/− cells. Isolated mitoplasts (500 µg) were lysed and subjected to sucrose density gradient centrifugation. Fractions (1-16) were collected and analyzed via western blotting with antibodies against MRPs and assembly factors as indicated. Input = 10% of total. Similar results were obtained in n ≥ 3 biologically independent experiments.

Fig. 2. Structures of mtSSU assembly intermediates co-isolated via MTG3^FLAG.

a MTG3^FLAG co-isolates mtSSU MRPs and several mtSSU assembly factors. FLAG-immunoprecipitation was performed with lysed mitochondria from HEK293 wild type (WT) cells and a stable HEK293 cell line inducibly expressing MTG3^FLAG, and subsequently analyzed via western blotting with indicated antibodies (total = 3%, eluate = 100%). Similar results were obtained in n ≥ 3 biologically independent experiments. b Schematic depiction of the mature 12S rRNA secondary structure. Regions are depicted by their level of maturation in each state, with M corresponding to the mature mtSSU (PDB: 3J9M;). c Cryo-EM structures of the MTG3-TFB1M-mtRBFA⁽ⁱⁿ⁾-bound small mitoribosomal subunit (mtSSU) intermediate (state A), METTL15-mtRBFA⁽ⁱⁿ⁾-bound mtSSU (state B), and METTL15-mtRBFA^(out)-bound mtSSU (state C). The 12S rRNA (red) and indicated biogenesis factors are shown as cartoon and the remaining mitoribosomal proteins (MRPs) are indicated as white transparent surface. MTG3: light blue, TFB1M: lime green, mtRBFA: indigo, METTL15: turquoise, mS38: beige. Close-up views of the immature decoding center (d) and of the foot region (e) in state A. Coloring as in (c), with cryo-EM densities (from map A3) of immature rRNA helices shown as red surface. Close-up views of the immature decoding center with cryo-EM densities (from maps B-C3) (f, h) and of the foot region (g, i) in state B and C, respectively. (e, g, i) Close-up views of the foot region in each state, showing the densities for MTG3 (state A, map A3), or MTG3^NTD and h44 with altered trajectory (state B-C, 15 Å low-pass filtered maps B1 and C1). Mature h44 is depicted by red dashed lines and would clash with MTG3 and MTG3^NTD.

Fig. 3. Structural comparison of assembly intermediates.

a Table of six described assembly intermediates each compromising at least one of the assembly factors MTG3, TFB1M, or mtRBFA, as well as the mature mtSSU (PDB: 3J9M;). Maturation of 12S rRNA modules is shown for each state (matured: “+”, unmatured: “−”). b The previously described assembly state (PDB 8CSP) (left) as well as state A (this study) (right) are shown side-by-side, with additional modules highlighted (right). The 12S rRNA and indicated biogenesis factors are shown as cartoon and the remaining mitoribosomal proteins (MRPs) are indicated as white transparent surface. Depiction as follows (state A vs PDB 8CSP): 12S rRNA: red vs peach, MTG3: light blue vs dark blue, TFB1M: lime green vs dark green, mtRBFA: indigo, bS21m, uS21m, uS15m: beige, METTL17: yellow, MCAT: light blue, ERAL1: dark purple. c Zoomed-in view of h18 and TFB1M in state A d Zoomed-in view of MTG3 (light blue) and uS15m (beige) in state A vs additional ordered regions of MTG3 in a previous state (PDB 8CSP) (c, d) Densities from map A3 (c) and A2 (d) are depicted as transparent surfaces. e Side-by-side view of state B (right) and C (left) with coloring as followed: 12S rRNA: red vs peach, METTL15: dark vs light turquoise, mtRBFA: indigo vs dark violet. f Different routes of the rRNA content in state B vs state C. The model of state C is shown superimposed to the model of state B, with 12S rRNA density from state B (map B3) depicted transparent. g Close-up view of METTL15-mtRBFA⁽ⁱⁿ⁾ interaction in state B superimposed with METTL15 from state C and depicted mito-specific extension from of METTL15. The 12S rRNA and MRPs are indicated as transparent surface.

Fig. 4. Structures of four mtSSU (pre-) translation initiation states.

a Western blot analysis of mitoribosome complexes co-purified via mtIF3^FLAG. FLAG-immunoprecipitation was performed with lysed mitochondria from wild type cells and cells inducibly expressing mtIF3^FLAG. Samples (total = 3%, eluate = 100%) were subsequently analyzed via western blotting with indicated antibodies. Similar results were obtained in n ≥ 3 biologically independent experiments. b Cryo-EM structures of the mtIF3-bound small mitoribosomal subunit (mtSSU) (state D), mtIF2- and mtIF3-bound mtSSU (state E), mtIF2-bound mtSSU (state F), and mtIF2-, mRNA- and fMet-tRNA^Met-bound mtSSU (state G). The 12S rRNA and indicated initiation factors are shown as cartoon and the remaining mitoribosomal proteins (MRPs) are indicated as white transparent surface. 12S rRNA: red, mtIF3: yellow, mtIF2: blue, fMet-tRNA^Met: light yellow, mS37: beige. c Zoomed-in view of the codon-anticodon interaction in state G. Coloring as in (b) with mRNA backbone being depicted in purple and nucleotides being highlighted in red (A), blue (U), green (G), and yellow (C). The 12S rRNA and MRPs are indicated as transparent surface. The trajectory of the mRNA is shown by superimposing with a known IC state (7PO2;) and depicted as transparent cartoon. d–g Views of the foot region in each state, showing the densities for MTG3^NTD and for h44 from the 15 Å low-pass filtered maps D1-G1. Mature h44 is depicted by red dashed lines and would clash with MTG^NTD.

Fig. 5. Deletion in N-terminus affects MTG3 functionality.

a ΔN-MTG3^FLAG lacking 10 conserved aa in the N-terminus can only partially restore mitochondrial translation. Translation of mtDNA-encoded proteins in wild type, Mtg3^−/− and Mtg3^−/− cell lines expressing full length (MTG3^FLAG) or mutated MTG3 (ΔN-MTG3^FLAG), respectively, were analyzed using [³⁵S]Methionine de novo incorporation and visualized via autoradiography and western blotting. GAPDH was used as a loading control. Similar results were obtained in n ≥ 3 biologically independent experiments. b ΔN-MTG3^FLAG does not co-purify any MRPs. FLAG-immunoprecipitation was performed with wild type cells, wild type cells expressing MTG3^FLAG and Mtg3^−/− cells expressing MTG3^FLAG or ΔN-MTG3^FLAG, respectively. Samples were analyzed using western blotting and antibodies as indicated (total = 3%, eluate = 100%). Similar results were obtained in n ≥ 3 biologically independent experiments. c, d Monosome level can be restored upon ΔN-MTG3^FLAG expression. Mitoplasts (500 µg) were isolated from wild type, Mtg3^−/− and Mtg3^−/− cell lines expressing full length (MTG3^FLAG) or ΔN-MTG3^FLAG, respectively. Mitoribosomal complexes were separated via sucrose density gradient centrifugation and collected fractions (1-16) were analyzed via western blotting with indicated antibodies against MRPs and assembly factors (c) or northern blotting with probes against 12S rRNA (MT-RNR1) and 16S rRNA (MT-RNR2) (d). Similar results were obtained in n ≥ 3 biologically independent experiments. e Abundance of mt-mRNAs bound to monosomes is similar in the Mtg3^−/− + ΔN-MTG3^FLAG cell line in comparison to wild type MTG3^FLAG. Fraction 11 from sucrose gradients from (c) were used to isolate RNA and perform NanoString analysis. Level of mt-mRNAs were normalized to 16S-rRNA and RNA abundance bound to monosomes in the Mtg3^−/− + ΔN-MTG3^FLAG cell line was calculated relative to Mtg3^−/− + MTG3^FLAG cell line (n = 3 biologically independent samples shown as mean ± SEM; individual data points are shown as circles). Statistical analysis was performed as two-sample one-tailed Student’s t-test. Significance was defined as *p ≤ 0.05, **p ≤ 0.01.

Fig. 6. Model of MTG3-mediated maturation coupled to translation initiation of human mtSSU.

Intermediate states of the mtSSU are depicted as surface and factors and rRNA content is colored as in Figs. 2–4. Models for mature mtSSU IC (PDB: 7PO2), mtLSU and mature IC (PDB: 6GAW) have been reported previously.