Cytoplasmic ribosomes on mitochondria alter the local membrane environment for protein import

Abstract

Most of the mitochondria proteome is nuclear-encoded, synthesized by cytoplasmic ribosomes, and targeted to the mitochondria posttranslationally. However, a subset of mitochondrial-targeted proteins is imported co-translationally, although the molecular mechanisms governing this process remain unclear. We employ cellular cryo-electron tomography to visualize interactions between cytoplasmic ribosomes and mitochondria in Saccharomyces cerevisiae. We use surface morphometrics tools to identify a subset of ribosomes optimally oriented on mitochondrial membranes for protein import. This allows us to establish the first subtomogram average structure of a cytoplasmic ribosome at the mitochondrial surface in the native cellular context, which showed three distinct connections with the outer mitochondrial membrane surrounding the peptide exit tunnel. Further, this analysis demonstrated that cytoplasmic ribosomes primed for mitochondrial protein import cluster on the outer mitochondrial membrane at sites of local constrictions of the outer and inner mitochondrial membranes. Overall, our study reveals the architecture and the spatial organization of cytoplasmic ribosomes at the mitochondrial surface, providing a native cellular context to define the mechanisms that mediate efficient mitochondrial co-translational protein import.

Figure 1.

Cellular cryo- ET imaging and processing workflow captures cytoplasmic ribosomes positioned for protein import on mitochondrial membranes. (A) S. cerevisiae yeast expressing TIM50-GFP are grown in respiratory or fermentative conditions and treated with vehicle or CHX (100 μg/ml) prior to deposition on electron microscopy grids (black mesh circle) and vitrification via plunge freezing. (B) Vitrified yeast cells were imaged by cryo-FM to assess sample quality, cell density, and ice thickness. (C) Clumps of yeast were targeted for cryo-FIB milling to generate thin cellular sections (i.e., lamellae). (D) Cellular lamellae were imaged by standard cryo-ET acquisition procedures to generate tilt series that were further processed to generate 3D reconstructions (i.e., tomograms). Subcellular components such as mitochondria, the ER, the plasma membrane, and ribosomes are visible within the resulting tomograms. Scale bars = 250 nm. (E) Reconstructed tomograms were processed through “particle picking” software, which identified the initial positions and orientations of all visible cellular ribosomes. The positions and orientations were refined using subtomogram averaging to produce a consensus 8 Å 80S ribosome structure. (F) Mitochondrial membranes were traced, and separate three-dimensional voxel segmentations were generated for the OMM and IMM. These voxel segmentations were converted to surface mesh reconstructions using the surface morphometrics (Barad et al., 2023) pipeline such that the location of the membrane is represented by the coordinate of each triangle within the mesh. (G) The position and orientation of each ribosome relative to the OMM surface mesh reconstruction were calculated and rendered in the ArtiaX module of ChimeraX. The three-color arrows on ribosomes represent the Euler angles, with the yellow arrow representing the orientation of the ribosome peptide exit tunnel. (H) The cutoff for identifying cytoplasmic ribosomes engaged in protein import on the OMM was established by referring to the distance between the peptide exit tunnel of ER-translocon ribosome and the ER membrane. The optimal cutoff of the distance between the exit tunnel and OMM was identified as 0–95 Å in ArtiaX as we started to observe the exit tunnel pointed away from OMM in the expanded cutoff, either 0–110 or 0–120 Å. (I) Cytoplasmic ribosomes optimally positioned for protein import were identified as those with their exit tunnel closer than 95 Å from the OMM.

Figure S1.

Representative tomograms of cryo-FIB milled S. cerevisiae cell lamellae showed visible mitochondria-associated cytoplasmic ribosomes. (A) Representative X-Y slices of reconstructed tomograms collected at pixel size 2.638 Å from cryo-FIB milled S. cerevisiae yeast cells grown in different growth conditions (i.e., fermentative and respiratory) and treatment conditions (i.e., vehicle and CHX). Cytoplasmic ribosomes in close proximity to the OMM are highlighted by white arrowheads. Scale bars = 250 nm. (B) Quantification of the number of ribosomes positioned with the exit tunnel facing the OMM in CHX-treated or vehicle-treated cells grown in respiratory versus fermentative conditions. (C) Representative X-Y slices of reconstructed tomograms collected at pixel size 1.6626 Å from cryo-FIB milled S. cerevisiae grown in fermentative and respiratory conditions and treated with CHX (100 μg/ml) displaying subcellular features such as mitochondria, ribosomes, the ER, and the plasma membrane. Cytoplasmic ribosomes in close proximity to the OMM are highlighted by white arrowheads. Scale bars = 250 nm. (D) FSC plot (top) of the 80S cytoplasmic ribosome reconstruction is shown with resolution reported at 0.143 FSC and the reconstructed subtomogram average (bottom left). The 80S cytoplasmic ribosome was resolved to 8 Å from 35,784 ribosome particles with the color map (bottom right) showing the local resolution. (E) A subset of representative models of ribosomes positioned with their exit tunnels optimally positioned for protein import at the OMM surface. (F) Serial 10-fold dilutions of fluorescently-labeled mitochondrial strains show similar growth to the parent WT (BY4741) cells on a non-fermentable carbon source YPGE (Yeast extract, Peptone, 3% Glycerol, 2% Ethanol).

Figure 2.

3D subtomogram average of a cytoplasmic ribosome optimally positioned for protein import on the OMM shows multiple contact points. (A) Three views of the subtomogram average of a cytoplasmic ribosome positioned with the exit tunnel on the 60S subunit (dark blue) facing the OMM (gray). Three connecting densities (labeled 1, 2, 3 in pink, orange, and green, respectively) are visible between the 60S and the OMM surrounding the peptide exit tunnel (dashed circle). (B) The subtomogram average of the mitochondria-associated ribosome (blue transparent density) with a fitted atomic model of the S. cerevisiae 80S ribosome (PDB 4V6I). Boxed regions focus on the cryo-EM density of each of the three connections observed between the cytoplasmic ribosome and the OMM.

Figure S2.

3D subtomogram average of a cytoplasmic ribosome positioned for protein import on the OMM shows multiple contact points. (A) FSC plot of the OMM-associated 80S cytoplasmic ribosome reconstruction is shown with resolution reported at 0.143 FSC and the reconstructed subtomogram average (bottom left). The 80S cytoplasmic ribosome was resolved to 19 Å from 1,076 ribosome particles with the color map shows the local resolution (bottom right). (B) The peptide exit tunnel is visible in the subtomogram average (gray density) at higher isosurface volume thresholds, as indicated by the dashed black circle. This was used to mark its position relative to the connecting densities visible in the subtomogram average at lower isosurface volume threshold values (colored density). (C) Representative models of ribosomes positioned within 250 Å of the OMM surface with their exit tunnels facing away from the OMM. 3D refinement of these particles results in a 3D reconstruction of a ribosome that does not contain any distinguishable connecting densities between the 80S ribosome and the OMM, suggesting that these connections are specific to 80S ribosomes optimally positioned for protein import. (D) Density corresponding to the connection labeled #1 in the subtomogram average of mitochondrial-associated ribosomes correlates well with the density corresponding to the expansion segment of eS7La of the 25S rRNA in the large 60S subunit present in the ER-associated ribosome maps from S. cerevisiae (EMD-3764). (E) The density corresponding to the connection labeled #2 correlates well with the density corresponding to the region where the ribosome interacts with the import channel, Ssh1, in the ER-associated ribosome maps from S. cerevisiae (EMD-1667). (F) The density corresponding to the connection labeled #3 correlates well with the density corresponding to the rRNA expansion segment, rpL38, and the TRAP complex in the ER-associated ribosome maps from human (EMD-15884), rabbit (EMD-16232), and canine (EMD-3068). (G) Overlap between the OMM-associated 80S cytoplasmic ribosome structure from this work (transparent blue) with ER-bound ribosome (EMD-14424; green) and mitochondria-bound ribosome (EMD-14423, orange) from S. pombe and mitochondria-bound ribosome from purified S. cerevisiae mitochondria (EMD-3762, pink). Dashed colored lines show the relative orientations of the membranes for the different structures.

Figure S3.

Cytoplasmic ribosomes primed for protein import cluster on the mitochondrial membrane. (A) Representative plots of the K(r)/K_CSR(r) ratio for mitochondrial-associated cytoplasmic ribosomes oriented for protein import within a range of radius (r) values of 27–166 nm. The black line equals to 1. (B) Quantification of the maximum value of K(r)/K_CSR(r) for each tomogram at the indicated radius intervals for each ribosome class. Quantification from import-oriented ribosomes n = 87 and non-import-oriented ribosomes n = 89 tomograms are shown. P values from Mann–Whitney U test are indicated. *P < 0.05; ****P < 0.001. (C) Representative membrane surface reconstructions of mitochondria (gray) with ribosomes oriented for import relative to the OMM (blue). Insets show zoomed-in boxed regions of the ribosome models with circle overlays demarking the location of the 3′ mRNA entry (blue), the 5′ mRNA exit sites (orange), the possible pathways of interconnecting mRNA (dashed black line), and the calculated end-to-end distance from 5′ to 3′ of each interconnected mRNA (solid black line). (D) Membrane surface reconstruction of mitochondria (gray) and ER (purple) membranes with corresponding models for co-translating (blue) and non-co-translating ribosomes (pink). (E) Combined histogram of IMM–OMM distances of co-translation-associated and non-co-translation-associated patches in S. cerevisiae treated with CHX. Dashed vertical lines correspond to peak histogram values of pooled data. (F) Histograms of IMM–OMM distances of co-translation-associated and non-co-translation-associated patches in S. cerevisiae treated with vehicle (e.g., no CHX). Dashed vertical lines correspond to peak histogram values of pooled data.

Figure 3.

Cytoplasmic ribosomes primed for protein import cluster on the mitochondrial membrane. (A) Quantification of the maximum value of K(r)/K_CSR(r) for a 30–40-nm radius for each tomogram within the indicated ribosome class. Quantification from import-oriented ribosomes n = 87 and non-import-oriented ribosomes n = 89 tomograms are shown. P values from Mann–Whitney U test are indicated. ****P < 0.001. (B) Representative membrane surface reconstructions of mitochondria (gray) with ribosomes oriented for import relative to the OMM (blue). Insets show zoomed-in boxed regions of the ribosome models with circle overlays demarking the location of the 3′ mRNA entry (blue), the 5′ mRNA exit sites (orange), the possible pathways of interconnecting mRNA (dashed black line), and the calculated end-to-end distance from 5′ to 3′ of each interconnected mRNA (solid black line). (C) Representative tomogram slices showing labeled cytoplasm, ribosome, mitochondrial matrix, IMM, OMM, and CJs (upper panel) with an overlay of the surface mesh reconstructions of the IMM and OMM (red) and ribosome (blue) (lower panel). (D) Representative membrane surface reconstruction of mitochondria with the OMM surface colored by OMM–IMM membrane distance and the IMM surface shown in gray. The bottom inset labeled “1” shows inner membrane boundary (IBM) regions on OMM with more subtle OMM–IMM distance variations. In contrast, the bottom inset labeled “2” shows regions on OMM with large OMM–IMM distances corresponding to CJs. (E) Ribosome and membrane models defining the patches on the membrane surface mesh reconstruction that correspond to ribosomes oriented for import (blue, top) and ribosomes near but not oriented for import (pink, bottom). The ribosomes oriented for import are defined as those with the peptide exit tunnel (yellow arrow) pointed toward the membrane. In contrast, those not oriented for import have peptide exit tunnels facing away from the membrane. (F) Representative ribosome and membrane model with the OMM surface colored by the ribosome-associated (blue) and crista-associated OMM (orange), with areas of overlap (black). (G) Quantification of the average fraction of overlap from each tomogram between indicated ribosome class. Quantification from import-oriented ribosomes n = 71, non-import-oriented ribosomes n = 71, and random n = 71 tomograms are shown. P values from Mann–Whitney U test are indicated. *P < 0.05; ****P < 0.001.

Figure 4.

Ribosome-associated protein import alters the local architecture of the outer and inner mitochondrial membranes. (A) Ribosome and membrane model defining co-translation-associated and non-co-translation-associated patches on membrane surface mesh reconstruction for the OMM–IMM distance measurement. Co-translation-associated patches (blue) included the nearest OMM triangles (black in the middle of blue patch) to the import-oriented ribosomes and the OMM triangles within 150 Å of these nearest OMM triangles. Non-co-translation-associated patches (gray) consisted of the OMM mesh triangles that excluded co-translation-associated patches. (B) Quantification of the peak histogram values of OMM–IMM distance measurements for each tomogram within the indicated membrane patch region. The co-translation-associated and non-co-translation-associated patches are detailed in Fig. 4 A. The “all membrane” patches represent the entire OMM surface. The “randomized” patches were simulated according to the number of co-translation-associated patches, as outlined in Materials and methods. Quantification of CHX-treatment data n = 87 and of vehicle-treatment data n = 18 tomograms are shown. P values from Mann–Whitney U test are indicated. *P < 0.05; ****P < 0.001. (C) Representative membrane surface reconstruction of mitochondria colored by OMM–IMM membrane distance, with regions <10 nm shown in blue and regions >10 nm shown in gray. Ribosomes oriented for import relative to the OMM are colored blue, and the remaining ribosomes near but not oriented for import are shown in pink. Insets show zoomed-in boxed regions of the models (middle) and the local variations in OMM–IMM distance (right).