Research Article

Crystal Structure of the Eukaryotic Ribosome

See allHide authors and affiliations

Science  26 Nov 2010:
Vol. 330, Issue 6008, pp. 1203-1209
DOI: 10.1126/science.1194294

Abstract

Crystal structures of prokaryotic ribosomes have described in detail the universally conserved core of the translation mechanism. However, many facets of the translation process in eukaryotes are not shared with prokaryotes. The crystal structure of the yeast 80S ribosome determined at 4.15 angstrom resolution reveals the higher complexity of eukaryotic ribosomes, which are 40% larger than their bacterial counterparts. Our model shows how eukaryote-specific elements considerably expand the network of interactions within the ribosome and provides insights into eukaryote-specific features of protein synthesis. Our crystals capture the ribosome in the ratcheted state, which is essential for translocation of mRNA and transfer RNA (tRNA), and in which the small ribosomal subunit has rotated with respect to the large subunit. We describe the conformational changes in both ribosomal subunits that are involved in ratcheting and their implications in coordination between the two associated subunits and in mRNA and tRNA translocation.

The ribosome is responsible for translating the information encoded by mRNA into protein in all living cells. It is composed of two subunits that consist of multiple proteins and RNA. The structures of the prokaryotic ribosome and individual subunits reveal in fine detail the evolutionarily conserved catalytic core where peptide bonds are formed and correct aminoacyl–transfer RNA (tRNAs) are selected to complement mRNA codons (111). Other activities of the ribosome that also constitute integral steps in the translational process do not show such a high degree of conservation. In particular, prokaryotes and eukaryotes use fundamentally different mechanisms for initiation of protein synthesis (1214). The eukaryotic cap-dependent initiation is a highly regulated process that entails scanning of the mRNA from the 5′-cap moiety downstream to the start codon. Many viruses (e.g., hepatitis C virus) repress cap-dependant translation and utilize special sequences in their mRNA, known as internal ribosome entry sites (IRESs), that can directly recruit the ribosome to the viral start codon and thus hijack the translational machinery (14).

The eukaryotic ribosome has a sedimentation coefficient of 80S and, with a minimal mass of ~3.3 MD (yeast and plants), is roughly 40% larger than its bacterial counterpart (70S). The large 60S subunit consists of three rRNA (rRNA) molecules (25S, 5.8S, and 5S) and 46 proteins, whereas the small 40S subunit includes only one rRNA chain (18S) and harbors 33 proteins. Several components contribute to the greater weight and complexity of eukaryotic ribosomes: RNA expansion segments that are inserted into the evolutionarily conserved rRNA core, 25 additional proteins with no homologs in eubacteria, as well as extra protein segments, mainly in the N′ or C′ tails of the conserved proteins. The precise functional roles of these additional components are largely unknown. Cryo–electron microscopy (cryo-EM) provided the first models of eukaryotic ribosomes at resolutions between 6.1 and 15 Å, which revealed the locations and shapes of the RNA expansion segments and indicated the position of additional protein moieties (1517).

The higher complexity and larger size of the eukaryotic ribosome is amply reflected in the vast number of proteins that interact with it and the variety of cellular processes in which the ribosome or its subunits play key roles. Examples are nonsense-mediated decay, unfolded protein response, and various eukaryote-specific translation regulation mechanisms.

We crystallized the complete eukaryotic 80S ribosome from Saccharomyces cerevisiae and determined its structure at 4.15 Å resolution. Our crystals capture the ribosome in the so-called ratcheted state. It was postulated more than 40 years ago that the translocation of mRNA and tRNA during protein synthesis is coupled to intersubunit movements (18, 19). More recent cryo-EM studies suggested that translocation is facilitated by large-scale movements involving a rotation of the small subunit relative to the big subunit (20). As confirmed by various studies, this ratchetlike intersubunit reorganization of the ribosome is essential for translocation (21). Recently, structures of the Escherichia coli ribosome trapped in intermediate states of ratcheting were described (22); however, an x-ray structure of the ribosome in the fully ratcheted state was lacking.

Structure determination. The dynamic nature of the eukaryotic ribosome has hindered efforts to crystallize it. To overcome this obstacle, we used the fact that depletion of glucose from yeast growth medium for a short time results in inhibition of translation initiation and accumulation of a homogenous population of vacant ribosomes (23, 24). We then purified them in a way that minimized damage to ribosomes during and after cell lysis (24).

Initial crystals diffracted poorly and were improved by soaking with various dehydration agents and metal ions. These improvements yielded three crystal forms—form I with four ribosomes in the asymmetric unit and forms II and III with two—all belonging to space group P21 but differing in their cell parameters. Form III diffracts to better than 3 Å resolution, but in order to obtain a complete data set from a single crystal—to take full advantage of the anomalous signal from the bound osmium ions—we collected a data set at 4.15 Å resolution, with reflections useful for refinement to 4 Å (table S1).

Molecular replacement (MR) procedures found clear solutions in all crystal forms when a search model was used that was composed of the large subunit (50S) from Haloarcula marismortui (8) and the small subunit (30S) from Thermus thermophilus (1). This solution, after improvement by rigid-body refinement, was used to locate several hundred osmium sites in form III and thus obtain initial single-wavelength anomalous dispersion (SAD) phases. However, maps from the combined MR + SAD phases were not readily interpretable. Therefore, iterative rounds of the following scheme were performed: (i) phase improvement by using solvent flattening, and intercrystal and noncrystallographic symmetry averaging between the three forms; (ii) expansion of the model by manual building, followed by rigid-body refinement of large domains; and (iii) location of additional osmium sites by using the expanded model and recalculation of maps. When the model approached completion, the number of located osmium sites had reached 700. At that point, averaging was no longer used, and solvent-flattened SAD or model + SAD phases guided the final building steps.

The model consists of two ribosomes in the asymmetric unit of crystal form III and contains the entire rRNA moiety except for ES27 and a small part of ES7, both in 60S, as well as a few residues in ES6 of 40S. Density for the phosphates was usually clear and guided modeling of the rRNA parts. The model also contains the Cα backbone of all proteins with homologs in prokaryotic ribosome x-ray structures, including, in most cases, their eukaryote-specific additions. Three additional proteins whose location was identified by cryo-EM (L30e, S19, and RACK1) (2528) were also resolved. The model includes many additional α helices and β strands that belong to eukaryote-specific proteins. With the exception of several proteins (S25, S17, S28, L14, and L6), we refrained from assigning these secondary-structure elements to individual proteins because biochemical data that may confirm their position are lacking. The model also contains 715 osmium hexamine molecules that probably represent sites of magnesium ions.

This paper will focus on only one of the two ribosomes in the asymmetric unit, which assumes the same conformation as all the monomers in the other two crystal forms and fits well into cryo-EM maps of yeast ribosomes (16, 26). The second monomer, whose small subunit has a somewhat different conformation, may represent some intermediate state during translocation. The refinement of this monomer is ongoing, and it will not be further described here.

We use the common notation for ribosomal components, where ESx denotes expansion segment number x; hx and Hx refer to helices within the conserved core of the small and large subunit rRNA, respectively; and Lx and Sx are large subunit and small subunit proteins.

Overall view of the 80S ribosome: An elaborate network of interactions. The overall structure of the eukaryotic ribosome reveals a considerably larger assembly than its prokaryotic counterpart, but the basic architecture remains similar with common recognizable landmarks (Fig. 1, A and B). The rRNA expansion elements are located predominantly on the solvent-exposed sides at the periphery of both subunits. All intersubunit bridges described in the crystal structure of the bacterial ribosome (5) have corresponding bridges in the yeast ribosome (Fig. 1C, and table S2). In addition, there are several eukaryote-specific bridges that were first visualized by low-resolution cryo-EM studies of yeast ribosomes (15, 16) and whose components are more accurately described now by our 4.15 Å model (Fig. 1C and table S2). The quality of the electron density is shown in Fig. 1, D and E.

Fig. 1

Overall view of the x-ray structure of S. cerevisiae 80S ribosome. (A) View from the E site. Proteins and rRNA in the 40S are colored dark and light blue, respectively, and dark and pale yellow, respectively, in the 60S (in all the following figures, this color scheme will be maintained unless otherwise indicated). Expansion segments are in red. (B) View from the A site. (C) Interface views of the 60S and 40S subunits with bridges numbered essentially as in (5), and colored red. (D and E) Electron density maps, calculated with SAD + model combined phases and contoured at 1.5 σ, showing (D) H89 of 25S with an osmium hexamine molecule bound in the major grove of H89 and (E) interaction between ES39 and protein L9. This and all other figures were made with Pymol (43).

The eukaryote-specific elements substantially expand the network of interaction within the ribosome. Perhaps the most impressive example is the ~200-nucleotide-long ES6 of the small subunit (Fig. 2). This expansion segment emerges at the solvent face of the platform, just below helix 26 (h26), where it is enveloped by several eukaryote-specific protein moieties, including the 60-residue-long α-helical extension at the C terminus of protein L19e, thus forming a eukaryote-specific bridge (eB12) (Fig. 2). ES6 then sends one of its three long arms in the direction of the shoulder (Fig. 2). This arm interacts with protein S22 (S8p); its tip branches into two loops, one of which lies in close proximity to a eukaryote-specific helix-turn-helix insertion in protein S9 (S4p). The nearness of ES6 to h26 and S9, components of the exit and entry sites of the mRNA, respectively, suggests its involvement in translation initiation, perhaps as a docking surface for factors that participate in activities at both sites (29). The second long arm of ES6 might provide additional binding surface for some interacting partners of the ribosomes. The third long arm of this expansion segment runs down toward the bottom of the small subunit, where its tip is enveloped by the two internal loops of ES3. A eukaryote-specific protein that we could partially model is bound to ES3 and h9 and forms a eukaryote-specific bridge (eB11) with ES41 of 60S. Thus, at the bottom of the 40S body, several expansion segments and proteins create a eukaryote-specific network of interactions.

Fig. 2

Part of the extended network of interactions formed by eukaryote-specific elements. (A and B) ES6 (orange) forms bridge eB12 with the eukaryote-specific α-helical addition (colored in magenta) at the C terminus of protein L19e. Another eukaryote-specific bridge, eB11, is formed between ES41 of 25S rRNA (magenta) and a eukaryote-specific protein bound to ES3 and h9 of 40S. Bridges are marked with asterisks. (C) Secondary structure diagram of 18S rRNA in blue, 5S rRNA in brown, 25S rRNA in yellow, and 5.8S in dark red, showing expansion segments in red (Comparative RNA Web site of RNA: http://www.rna.ccbb.utexas.edu).

Ribosomal domain movements in the ratcheted state. Our model of the eukaryotic ribosome, when compared with the structure of the nonratcheted prokaryotic ribosome (1), shows a 4° to 5° counterclockwise rotation of 40S with respect to 60S and swiveling of the head domain of 40S by 15.5° in the direction of the tRNA exit site (Fig. 3, A and B). These are the characteristics of the ratcheted state, consistent with observations made by cryo-EM studies, that the yeast ribosome devoid of ligands assumes the same ratcheted conformation as the one stabilized by the binding of eukaryotic elongation factor 2 (16).

Fig. 3

Ratcheted state of the eukaryotic 80S ribosome. (A and B) Schematic representation of the motion from the nonratcheted to the ratcheted state. The red line indicates the outline of the 40S in the nonratcheted state, with arrows indicating the trajectory. (A) Top view of the yeast 80S ribosome. (B) View from the solvent side of 40S. (C) Bridge B1 in the ratcheted 80S. (D) Close-up view of the yeast B1a bridge. The tip of the A-site finger (ASF-H38) from 25S RNA forms interactions (colored in red) with the head of the 40S subunit, including protein S15 (magenta). (E) View of the yeast bridge B1b formed between proteins S18 and protein L11. Residues thought to interact are indicated in red. (F) Bridge B1 as found in the nonratcheted prokaryotic 70S ribosome. (G and H) The central protuberance (CP) of (G) S. cerevisiae 60S (eukaryote-specific elements marked in red) and (H) the archeal H. marismortui 50S. (I) The central protuberance of the 50S prokaryotic ribosome (light blue) superimposed on the eukaryotic ratcheted 60S (5S rRNA in brown, proteins in green, and eukaryote-specific elements in red). All parts except 5S rRNA and proteins L5 (L18p) and L11 (L5p) have been removed for clarity.

In comparison with nonratcheted prokaryotic ribosomes, the large-scale movements implicated in ratcheting result in substantial alterations in the bridges between the head domain of 40S and the large subunit (1). The first bridge between the head domain and the large subunit, bridge B1a (Fig. 1C), is formed in the nonratcheted 70S by the A-site finger (H38 of 23S) and protein S13p. Because swiveling displaces components at the periphery of the head domain by as much as 25 Å, this bridge is rearranged in our model. We find that ratcheting brings residues 1239 to 1241 at the tip of h33 (component of the beak of 40S), as well as protein S15 (S19p), into the vicinity of the tip of H38, which bends considerably in order to form interactions with these partners (Fig. 3, C and D). Conformational changes are also observed at the base of H38 where it contacts the central protuberance. In the second bridge between the head domain of 40S and the large subunit, B1b (Fig. 1C), the large shift in the position of protein S18 (S13p) places its longest helix, instead of the N-terminal loop, in contact with central protuberance protein L11 (L5p) (Fig. 3, C and E). Similar rearrangements were observed in structures of partially ratcheted E. coli 70S ribosomes (22). In addition, residues 70 to 75 from a loop in S15 (S19p) may also interact with L11 (L5p) in the ratcheted state. The prokaryotic homologs of the two proteins that are strongly shifted, S15 (S19p) and S18 (S13p), were shown in the nonratcheted state to monitor the occupancy of the A and P sites, respectively (Fig. 3F) (1, 4, 30). The direct interaction between these two proteins is probably stronger in eukaryotes because of additional residues in both.

The entire central protuberance of the large subunit, dominated by 5S rRNA and the proteins enveloping it, undergoes considerable structural rearrangement. These alterations involve tilting of 5S helices, a shift in the position of L11 (L5p) elements, and displacement by up to 7 Å of all L5 (L18p) domains except the N tail and the first helix (Fig. 3, G to I). Their plasticity in response to ratcheting and their strategic position in the large subunit suggest that the central protuberance and the A-site finger may coordinate changes in different sites of the large 60S subunit, notably the L1 protuberance, the GTPase (guanosine triphosphatase) center, and the peptidyl-transferase center, with the 40S head rotation and with mRNA translocation. This proposed requirement for plasticity of the central protuberance might underlie the need for keeping 5S as a separate rRNA chain.

The numerous interactions between the head and the large subunit may serve to limit and to monitor the extent of the head’s rotation. All these interactions are relatively weak and entail large distances but together facilitate a wide, but precise, ratcheting movement. Flexibility or plasticity of the interacting partners is likely crucial for constantly adjusting the bridges as the ratcheting movement progresses.

The central protuberance in eukaryotes is larger than its prokaryotic counterpart and forms additional interactions with other regions in the ribosome. Comparing our model with the archeal large subunit (Fig. 3, G and H) (8), we find that these additional features include ES12, which forms an RNA helix that runs parallel to 5S rRNA; the eukaryote-specific protein L6e, which binds 5S rRNA and is embedded in a region rich in other eukaryote-specific elements [protein L14e, the N-terminal domain of L7 (L30p), and ES7]; and eukaryote-specific domains in proteins L5 (L18p), L21e, and L10e (L16p).

Rearrangement of ribosome functional sites upon ratcheting. In structures of the prokaryotic ribosome at putative intermediate states along the ratcheting pathway (22), the central bridges are almost indistinguishable from those observed in the nonratcheted state. In contrast, we find that in our fully ratcheted state some of these bridges undergo considerable alterations. Bridge B2a is of special interest, as it is formed by the base of the penultimate stem (h44) of 18S and H69 of 25S. In this universally conserved region, the residues 1755 and 1756 (1492 and 1493 in E. coli) of 18S and 2256 (1913 in E. coli) of 25S RNA play a key role in the decoding of mRNA (10). We find that in the ratcheted yeast ribosome, the nucleotides 1755 and 1756 and the region of 18S rRNA encompassing these decoding elements undergo a local conformation change. The RNA strand carrying residues 1754 to 1758 (1491 to 1495) is twisted and pushed away from the mRNA pathway by up to 9 Å. This position would preclude interactions of the rRNA backbone with the mRNA-tRNA complex (Fig. 4A). In addition, the tip of H69 bends so as to maintain bridge B2a, and several residues involved in the bridge assume different orientation compared with the vacant nonratcheted ribosome. We suggest that the rotation of the small subunit results in breaking or at least considerably loosening of the interaction of the 40S body with the mRNA-tRNA complex, so that this complex would be free to follow the rotation of the head. These findings concur with the notion that ratcheting is a multistep process (22) and suggest that rotation of the body and rearrangement of the decoding region occur before rotation of the head.

Fig. 4

Rearrangement of ribosome functional sites upon ratcheting. (A) Superposition of the prokaryotic nonratcheted ribosome on the ratcheted yeast ribosome. View of the decoding region with prokaryotic 16S and 23S rRNA in gray, the A codon of mRNA in red, and eukaryotic 18S and 25S RNA in blue and yellow, respectively. (B) Top view of the superposition of the 70S ribosome containing P-tRNA (green), E-tRNA (blue), and mRNA (red) on the 80S ratcheted ribosome. The ridge consisting of nucleotides 1575 to 1578 (1338 to 1341 in prokaryotes) forming the steric block between P-tRNA and E-tRNA and the region 999 to 1002 (788 to 791 in prokaryotes) from the platform of the small subunit. (C and D) View of the intersubunit bridge B4 in yeast (C) and in prokaryotes (D). Intersubunit contacts are colored in red.

Other instances where ratcheting breaks interactions that may impede translocation include the bending of the A-site finger, which removes its interaction with the elbow of A-tRNA, and breaking of the contact between A1001 of 18S (A790) and P-tRNA through the rotation of the 40S platform. However, the swiveling of the head alone can account only for 13 to 15 Å of movement of the mRNA-tRNA complex, out of the required 20 Å (one codon) distance. To complete the translocation step and proceed to the nonratcheted posttranslocation state, further conformational changes that would result in breaking the bonds between the head and the mRNA-tRNA complex must ensue (1, 4, 5). Our model suggests that these changes would entail large structural rearrangement in the head domain. Considering, for example, the steric block between P-tRNA and the E site formed by the universally conserved ridge of residues 1575 to 1578 in 18S (1338 to 1341 in E. coli, fig. S1) (11). In our structure, this ridge has rotated with the head in a trajectory that passed through the E site, which suggests that the forward swiveling of the head maintains the position of the ridge with respect to P-tRNA (Fig. 4B and fig. S1B). Hence, the ridge creates a large physical barrier that prevents further forward movement and release of the tRNA, whereas the tRNA physically prevents back-ratcheting. A large conformational change, driven perhaps by the energy of guanosine triphosphate hydrolysis (31), is likely required to remove this barrier. The nature and timing of events that complete the translocation of the mRNA-tRNA complex are unknown.

Adjustment of intersubunit contacts upon ratcheting. Practically no alterations were observed at bridge B3, which may constitute the center of rotation for the small subunit. Bridges B2b and B2c also show small variations compared with the nonratcheted state of the prokaryotic ribosome (1). However, other bridges involving h44 undergo substantial changes and are weakened considerably because of the bending of this long rRNA helix just below bridge B3. Thus, most of the interactions that constituted bridges B5 and B6 are broken in the ratcheted state, although the bridges are still weakly maintained through a few interactions (table S2). A eukaryote-specific protein that we could not assign is bound to h44 and h8 and approaches protein L3 (L3p) to form a eukaryote-specific bridge (eB13) (Fig. 1C and table S2). The extent of h44 bending may be smaller in prokaryotes where eB13 cannot form.

Considerable adaptation of both subunits is required to maintain the bridges that are far from the center of rotation. This is particularly true in the case of bridge B4, which connects the platform of the small subunit with H34 of the large subunit. Upon ratcheting, the tip of H34 bends considerably in order to establish interaction with residues 629 to 631 in h20 of the small subunit, as well as the C-terminal helix of protein S13 (S15p), which also undergoes some changes (Fig. 4, C and D). This bridge is mediated only by RNA-protein interactions in the nonratcheted prokaryotic ribosome, whereas in the ratcheted state, RNA-RNA interactions play an important role (table S2).

The L1 protuberance. Large-scale movements of the L1 protuberance, which consists of a stalk (H76) and an upper part [protein L1 (L1p) and helices H77 and H78], are associated with translocation (32). These movements are thought to play a key role in the release of E-site tRNA. In our model, the L1 protuberance reaches deep into the intersubunit space and is stabilized by interactions with protein L42e (L33p) (Fig. 5A). A similar position of the L1 protuberance was observed in cryo-EM studies of a translocation intermediate of a prokaryotic ribosome complex with tRNA in the P/E state (33). This is in stark contrast to the position of the L1 protuberance observed in structures of nonratcheted 70S ribosomes where it extends away from the body of the large subunit (1, 11) (Fig. 5, B and C). The conformational change between the in- and outward positions likely entails tilting of the stalk around its base, as well as additional rotation and tilting of the upper part of the protuberance because of a hinge point that locally breaks the helical conformation at the middle of the stalk. At this hinge point, a eukaryote-specific bulged-out nucleotide and a series of species-dependent weak base pairs, probably render the L1 protuberance more flexible in eukaryotes (fig. S2). It is noteworthy that the largest movements of the L1 protuberance have been observed in yeast ribosomes (16). In cryo-EM studies, it was shown that the fungi-specific elongation factor (eEF3), the so called “E-site factor” (34, 35), stabilizes the outward position of the L1 protuberance and thus allows E-tRNA release. The binding site for eEF3 includes the central protuberance of 60S, where the presence of the factor prevents the L1 protuberance from adopting the fully inward position (34). However, it is not clear how eEF3 binding stabilizes the outward position that facilitates E-site tRNA release. In view of our finding regarding the central protuberance dynamics, it is possible that eEF3 could exert its effect through inducing conformational changes in this domain of the ribosome.

Fig. 5

The L1 protuberance and mRNA entry and exit sites. (A) Top view of the L1 protuberance in the yeast ratcheted ribosome. The interaction between protein L1 (magenta) and protein L42e (L33p) (orange) is marked with an asterisk. (B) Top view of the superposition of the nonratcheted vacant prokaryotic ribosome (23S shown in magenta) on the vacant eukaryotic ribosome (25S yellow, 18S blue), showing the movement of the L1 protuberance. Protein L1 has been removed for clarity. (C) Superposition showing the L1 protuberances of the yeast ribosome and the prokaryotic ribosome with bound E-tRNA (23S shown in red and E-tRNA in blue). (D and E) Solvent side view of the mRNA entry site on the small subunit in the (D) prokaryotic (mRNA in red) and in the (E) yeast ribosome. The additional prokaryotic domain of S4p is shown in magenta. (F) View of the back of the 40S, showing the connection between the mRNA exit and entry sites in the yeast ribosome (mRNA from the prokaryotic model in red). The eukaryote-specific additions to the N and C termini of S2 (S5p) are shown in magenta. (G) The eukaryotic mRNA exit site (mRNA from the prokaryotic model in red). The eukaryote-specific bridge eB8 is marked with an asterisk.

Structure of the entry and exit sites of mRNA—implication to initiation. The entry and exit sites of the mRNA (36) on the small subunit reveal features that are unique to eukaryotes and may pertain to the ribosome’s interactions with mRNA and initiation factors. At the entry site in prokaryotes, h16 assumes a closed conformation, so that its tip is in proximity to S3p, a protein that forms the mRNA entry tunnel together with proteins S4p and S5p (1, 36) (Fig. 5E). In contrast, this helix bends in eukaryotes to adopt an open orientation that extends away from the body (Fig. 5F) (15). In prokaryotes, a domain that belongs to S4p, composed of two α helices, forms strong interactions and virtually covers a large part of the RNA helix h16 (1, 4). In the eukaryotic homolog of S4p, protein S9, this two-helix domain does not exist, and the C-terminal helix that could potentially interact with h16 tilts away. Thus, in yeast, h16 is bare, with no rRNA-protein interactions, and is free to rotate around its base. These observations must be viewed within the context of the eukaryotic initiation step. Current models suggest that binding of factors eIF1 and eIF1A to 40S stimulates scanning by inducing h16 to adopt a closed conformation, which stabilizes an opening of the mRNA entry tunnel latch (13, 37). Binding of IRES to the 80S ribosome induces similar changes (38, 39).

The entry tunnel latch is formed by interactions between the beak of the small subunit and h18 in the body. The beak in eukaryotes has a considerably different structure and harbors an additional protein moiety, partially modeled here as the eukaryote-specific protein S17 (fig. S3).

The exit site of the mRNA is more intricate in eukaryotes than in prokaryotes and contains several additional components (Fig. 5G). In association with protein S5 (S7p), just above the mRNA path, a eukaryote-specific protein with a high β-strand content is poised to bind mRNA. We modeled this protein as S28e in accordance with biochemical data (40). We also located several protein secondary structure elements situated just below the proposed mRNA path, in a position similar to that occupied in prokaryotes by protein S18p, which binds the Shine-Dalgarno sequence (3, 41, 42). These helices and strands probably belong to protein S26e and one additional unidentified protein (indicated together in Fig. 5G as protein SX2). A eukaryote-specific bridge eB8 is likely formed between this protein moiety and the first of two arms of ES31 in 25S (Fig. 5G). In addition, ES7 of 18S, an extension of h26, forms part of the mRNA exit site.

Another characteristic of the eukaryotic ribosome in this region is a direct contact between the mRNA entry and exit sites, which is established by a strong interaction between S0 (S2p), a part of the mRNA exit site, and eukaryote-specific extensions to protein S2 (S5p), a component of the mRNA entry tunnel (Fig. 5F).

Conclusion. The crystal structure of the complete 80S eukaryotic ribosome presented here allows rationalization, in structural terms, of existing biochemical and genetic information and will facilitate the design of future experimental models for investigating various aspects of protein synthesis. Further high-resolution structures of the eukaryotic ribosome—from yeast (as well as other species of eukaryotes) with its plethora of substrates, factors, and protein partners—coupled with biochemical and biophysical studies will be needed to provide a molecular description of such complex phenomena as translation initiation, regulation, and ribosome assembly, as well as for the development of drugs that will target the translational machinery.

Supporting Online Material

www.sciencemag.org/cgi/content/full/330/6008/1203/DC1

Materials and Methods

Figs. S1 to S3

Tables S1 and S2

References

References and Notes

  1. Materials and methods are available as supporting material on Science Online.
  2. Single-letter abbreviations for the amino acid residues are as follows: A, Ala; C, Cys; D, Asp; E, Glu; F, Phe; G, Gly; H, His; I, Ile; K, Lys; L, Leu; M, Met; N, Asn; P, Pro; Q, Gln; R, Arg; S, Ser; T, Thr; V, Val; W, Trp; and Y, Tyr.
  3. We thank A. Meskauskas and J. Dinman for providing the strain of S. cerevisiae and for support at the early stages of efforts in development of yeast ribosome purification protocol; A. Urzhumtzev for numerous invaluable crystallographic discussions; and R. Jackson for pointing out to us the possible benefits of glucose depletion in yeast. We gratefully acknowledge help with crystallographic programs from P. Afonine, K. Cowtan, G. Kleywegt, A. Lebedev, S. Panjikar, N. Pannu, R. Read, T. Terwilliger, L. Tong, A. Vagin, C. Vonrhein, and T. Yeates. We thank S. Melnikov and N. Garreau de Loubresse for help during the final stages of this work; M. Iskakova and N. Demeshkina for their participation and encouragement in early stages of this work; the staff of the Structural Biology Department core facility at IGBMC, Strasbourg; and S. Uge for assistance with computer infrastructure; C. Schulze-Briese and the staff at the Swiss Light Source, Switzerland; and A. Thompson and P. Legrand from the Soleil Synchrotron, France, for help with data collection. This work was supported by a European Molecular Biology Organization Long-Term Fellowship (A.B.-S.), Human Frontier Science Program, French National Research Agency grants ANR BLAN07-3_190451 and ANR-07-PCVI-0015-01, and the European Commission SPINE2. Coordinates and structure factors have been deposited with the Protein Data Bank with accession codes 3O2Z, 3O30, 3O58, and 3O5H.
View Abstract

Stay Connected to Science

Navigate This Article