|Home | About | Journals | Submit | Contact Us | Français|
Ribosomal protein L3 (L3) is an essential and indispensable component for formation of the peptidyltransferase center. Atomic resolution ribosome structures reveal two extensions of L3 protruding deep into the core of the large subunit. The central extension of L3 in Saccharomyces cerevisiae was investigated using a combination of molecular genetic, biochemical, chemical probing and molecular modeling methods. A reciprocal relationship between ribosomal affinity for eEF-1A stimulated binding of aa-tRNA and for eEF2 suggests that the central extension of L3 may function as an allosteric switch in coordinating binding of the elongation factors. Opening of the aa-tRNA accommodation corridor promoted resistance to the A-site specific translational inhibitor anisomycin, suggesting a competitive model for anisomycin resistance. These changes were also found to inhibit peptidyltransferase activity, stimulating programmed -1 ribosomal frameshifting, and promoting virus propagation defects. These studies provide a basis for deeper insight for rational design of small molecule antiviral therapeutics.
Multicomponent nanoscale biomachines must be able to coordinate orderly series of functions. For example, the elongation phase of protein synthesis requires that the ribosome perform at least nine sequentially ordered steps. The emerging view is one in which each event triggers allosteric rearrangements of specific segments of ribosomal RNAs (rRNAs) and ribosomal proteins, thus enabling information transfer among, and coordination between the various functional centers of the ribosome (reviewed in Noller, 2007). Although rRNA carries out the catalytic activity and comprises the bulk of the ribosome, many ribosomal proteins are also required for proper function. Ribosomal protein L3 (L3) in particular appears to be extremely important: its sequence and structure is highly conserved across all three kingdoms (reviewed in Brodersen and Nissen, 2005); it is one of the few ribosomal proteins required for peptidyltransferase activity (Schulze and Nierhaus, 1982); and it is one of only two proteins capable of initiating in vitro assembly of E. coli large ribosomal subunits (Nowotny and Nierhaus, 1982). Like many ribosomal proteins, L3 contains a globular domain on the cytoplasmic face of the large subunit, and has two long tentacle-like structures that extend deep into the mostly rRNA core. L3’s central extension reaches all the way to the A-site side of the PTC, where a tryptophan at position 255 (in the S. cerevisiae L3 protein), which is located at its tip, makes the closest approach of any amino acid to the peptidyltransferase center active site (Klein et al., 2004; Meskauskas et al., 2005). Regions of two important rRNA helical structures are also anchored on this central extension (also called the “W-finger” after the tryptophan residue): helix 95, and the structure formed by helices 90 –92. Helix 95 contains the sarcin/ricin loop (SRL), one of the key structures recognized by the translation elongation factors, and the target of ribosome inactivating proteins such as ricin (reviewed in Irvin and Uckun, 1992). The helix 90–92 structure is important because it forms one side of the corridor along which the 3′ ends of aa-tRNAs slide during the process of accommodation (Sanbonmatsu et al, 2005). The N-terminal extension of L3 may also interact with this structure. Molecular genetic and biochemical studies have shown that L3 plays an important role in aa-tRNA binding, peptidyltransferase activity, drug resistance, translational frame maintenance, virus replication, and as a binding site for a ribosome inhibitory protein (Wickner et al., 1982; Meskauskas et al., 2003; Meskauskas et al., 2005; Peltz et al., 1999; Petrov et al., 2004; Bosling et al., 2003; Fried and Warner, 1981; Hudak et al., 1999; Jimenez et al., 1975; Pringle et al., 2004; Schindler et al., 1974). Saturation mutagenesis suggested that L3 may function to transmit information between the SRL and the PTC (Meskauskas et al., 2005).
The current study employed a combination of targeted mutagenesis, structural, biochemical, and genetic approaches to examine the function of the W-finger. Mutagenesis experiments suggest that the tip of this structure interacts with a nearby base(s) in 25S rRNA, presumably A2940 (E. coli A2572). The W-finger appears to be intrinsically flexible, but its range of movement is strictly limited in one direction. Changes in the conformation of the W-finger resulting from mutation of W255 to cysteine (W255C) induce significant conformational rearrangements of the A-site side of the PTC, the helix 90 – 92 structure, helix 95, and the SRL. These are accompanied by resistance to anisomycin, increased affinity for aa-tRNA, decreased affinity for eEF2, and decreased rates of peptidyltransfer. Other mutants of nearby histidines had the opposite effects on elongation factor and aa-tRNA binding to ribosomes. These findings suggest a model in which L3 may play an important role in synchronizing the processes of aa-tRNA accommodation and translocation by functioning as sensor of the tRNA occupancy status of the A-site region of the PTC.
Previous studies showed that substitution of W255 with cysteine (W255C) conferred resistance to anisomycin, increased rates of programmed -1 ribosomal frameshifting (-1 PRF), inability to maintain the killer virus (the Mak− phenotype), increased affinity for aa-tRNA, and decreased peptidyltransferase activity (see Meskauskas et al., 2003). This broad spectrum of phenotypes suggested a crucial role for W255 in ribosome function, raising questions regarding its nature. Molecular modeling using numerous X-ray crystal structures from H. marismortui (Ban et al., 2000), E. coli (Schuwirth et al., 2005), and T. thermophilus (Selmer et al., 2006) reveals that the structure of this core region of the large subunit is highly conserved, and that W255 is positioned at the tip of the W-finger (data not shown). In the current study, molecular modeling was based on the H. marismortui structures because 1) they provided the highest available levels of resolution, and 2) its amino acid sequence in the central extension of L3 most closely resembles that of the yeast protein. This analysis revealed a potential base stacking interaction along the A-site proximal side of the PTC between W255 of L3 and A2940 in helix 90 of 25S rRNA (A2572 in E. coli and U2607 in H. marismortui 23S rRNA Fig. 1A). The “plasmid shuffle’ method (Rose et al., 1990) was used to determine the viability of a set of amino acid substitutions for W255 as the only forms of L3. Amino acid substitutions were chosen to reflect the full range of biochemical properties of their sidechains. In agreement with the proposed aromatic stacking, substitution of other aromatic amino acids for tryptophan did not affect cell growth (Fig. 1B, Table 1). Arginine, with its long, basic side chain was also viable, but basic substitutions showed decreased viability in proportion to reductions in side chain lengths, indicating that electrostatic interactions may to some extent complement the requirement for aromatic side chains. The acidic and small amino acid substitutions, predicted to be unable to interact with nearby bases (Fig. 1A), all resulted in lethal phenotypes (Fig. 1B), as did deletion of W255 (data not shown). DNA sequence analyses of clones rescued from apparently viable W255E and W255A colonies revealed that all were revertants to tryptophan. The set of viable substitutions clearly indicates the requirement of aromatic or positively charged amino acids at position 255 for cell viability, the only exception being the original W255C mutant (see next paragraph addressing this issue). These findings also point to the importance of putative interactions between the tip of the L3 W-finger and rRNA bases, the most likely candidate being A2940.
Given that small amino acids at position 255 were lethal, how could substitution with cysteine be viable? Examination of nearby amino acids revealed a second cysteine, C251, located 4 residues in the N-terminal direction. We hypothesized that formation of a disulfide bond between C251 and C255 may promote structural rearrangements of W-finger, repositioning nearby aromatic or basic residues to establish new contacts with A2940 (E. coli A2572) or other nearby bases. If true, then disruption of the disulfide bond should prevent such repositioning, resulting in lethality. Two complementary experiments were designed to test this hypothesis. Figure 2A shows that while the W255C and C251S single mutants were viable, the W255C/C251S double mutant was lethal. Figure 2B shows that 4 mM DTT severely inhibited growth of the W255C mutant, but not of wild type cells. These results support the existence of a disulfide bond between C255-C251 in the W255C mutant.
Formation of a disulfide bond between C255 and C251 would shift the tip of central extension towards N-terminus, potentially moving basic or aromatic residues on the C-terminal side of C255 into position to interact with A2940. Examination of the proximal amino acids on the C-terminal side of position 255 revealed three candidates: H256, H259, and W262. To test this hypothesis, these residues were changed to alanine either individually or in combination with the W255C mutation (Fig. 2C). Among the single mutants, H256A and H259A did not affect cell growth, while W262A was viable, albeit it negatively impacted growth rates. Among the double mutants, W255C/H256A allele was viable but conferred reduced growth, the W255C/H259A double mutant was inviable, and growth of the W255C/W262A mutant was indistinguishable from wild type cells. These data suggest that in the W255C mutant, H259 (likely in combination with H256) is repositioned to interact with A2940, although the possibility that other bases may be involved cannot be excluded. A cartoon of this model is shown in the upper panel of Fig. 2E.
These results indicate that cell viability requirement for aromatic or positively charged residues at the tip of the L3 W-finger can be met by adjacent positive residues, provided that they are moved into correct position subsequent to formation of the putative disulfide bond between C251 and C255 in the W255C mutant. Although the data do not allow determination of the exact conformation of the W-finger in the W255C mutant, it suggests that a certain degree of flexibility of the W-finger in the direction of the N-terminus can be tolerated, and that this flexibility may be an inherent feature of L3. The next issue was to determine the extent and limitations of the W-finger’s flexibility. The amino acid sequence surrounding the tip of the W-finger is N -RKVACIGA(W255)HPAHVMW - C. On the N-terminal side, the amino acids proximal to W255 are all aliphatic, and the nearest best candidates for interactions with rRNA are 7 and 8 residues away (arginine and lysine). As discussed above, the C-terminal side contains the two nearby histidines and a tryptophan. Insertional mutagenesis of amino acids on the N-terminal side of W255 would move adjacent N-terminal residues towards the position occupied by W255, and these mutants should be inviable. Deletions of residues C-terminal to W255 should have similar effects. In contrast, deletions on the N-terminal side, and insertions on the C-terminal side of W255 were predicted to be viable because they should bend the tip of the W-finger towards the N-terminus, moving H256 into position where it could interact with A2940. To test this model, L3 mutants containing amino acid deletions or alanine insertions on either side of the W-finger were generated so as to either push or pull the W-finger in the N- or C-terminal directions as described above. The results of this experiment are shown in Fig. 2D. As predicted, changes which should bend the W-finger toward its C-terminal side (indicated by the red down arrow), e.g. insertions of alanine after V249 and A250, or deleting A258, V260, and A265, were all lethal. Conversely, three of the changes predicted to bend the W-finger toward its N-terminal side, (indicated by the green arrow), i.e. insertion of alanine after V264, and deleting A250 or A254, were viable. The lethality due to insertion of alanine after V260 or by deletion of V249 is likely to have resulted from complex and unpredictable changes, either in the local ribosome structure or in L3 itself. The effects of symmetrical double insertions and deletions mutations in the W-finger, predicted to move the tip of the W-finger further into or away from the PTC respectively, were also assayed. As shown in Table 1, the symmetrical double insertion mutants (iV249/iV264, iV260/iA250) and one of the double deletion mutants (ΔV249/ΔA265) were lethal. The ΔV260/ΔA250 double deletion mutation was viable, but conferred a very severe growth defect on cells (Table 1). These findings suggest that the W-finger cannot move in towards, or away from the PTC without affecting cell viability. In summary, three different series of targeted mutations to the W-finger of L3 are consistent with the notion that the W-finger is flexible, but that its range of motion is very tightly constrained within the densely crowded inner core of the large subunit. Specifically, the data suggest that its flexibility appears to be limited toward the N-terminal side of W255 in the direction of the extended helix formed by helices 90 – 92. A cartoon of this model is shown in the lower panel of Figure 2E.
The data presented above suggested that the W-finger mutants may affect local ribosome structure. To examine this, ribosomes purified from wild-type and mutant cells were probed for structural changes using three base-specific solvent accessible reagents: dimethylsulfate (DMS), kethoxal, and carbodiimide metho-p-toluenesulfonate (CMCT). rRNAs were extracted, and modified bases were identified by primer extension using reverse transcriptase to detect methylation at the N3 position of uridines and the N1 position of guanosines (CMCT), at the N1 and N2 positions of guanosines (kethoxal), and at the N1 position of adenosines and N3 position of cytidines (DMS) (Inoue et al., 1985). The primers used (see Experimental Procedures) were designed to probe functional regions of domain V (including the A-site and P-site loops, the PTC, and the helices adjacent to these structures, i.e. helices 73, 74, 89 – 93), and helices 94 – 96.
The results of these experiments are shown in figures 3A – 3C and are summarized in figures 3D and 3E. The W255C mutant significantly altered the structure of the PTC, especially in the A-site loop and along helices 90 – 92 (Fig. 3A). In the helix 90 – 92 structure, A2889, C2898, G2913, C2924, A2929, A2932, A2935, A2940, and C2941 were significantly deprotected from chemical modification. In helix 89, U2861 showed decreased reactivity to CMCT. Similar, albeit weaker deprotection patterns, were noted in the W255R, W255H, W255K, H256Q, and ΔA254 mutants (data not shown). The stronger effects due to the W255C mutation may reflect the large conformational changes of the W-finger due to the putative C251-C255 disulfide bond described above.
L3 also makes contacts with bases in helix 73 at the A-site proximal side of the PTC (Klein et al., 2004). Chemical protection experiments revealed increased modification of two bases, A2398 and A2400, in the W255R and W255K mutants (Fig. 3B). The region comprising the SRL, which forms part of the elongation factor binding site, was also probed for structural changes (Fig. 3C). A3020 and A3023, both located in the in SRL, and A3010 (base of helix 95), were deprotected in the W255C mutant. A few additional changes were also noted in the vicinity of the SRL in the W255R (deprotection of C3003, G3014, and G3014), H256Q (deprotection of A3023), and ΔA245 (enhanced protection of A3020) mutants. The major changes in protection patterns in the vicinity of the PTC and SRL are summarized on a 2-dimensional rendering of the yeast 25S rRNA (Fig. 3D). Molecular modeling of these changes to the three-dimensional structure of the H. marismortui large ribosomal subunit (Fig 3E) reveals that the W-finger mutants affect the conformation of regions of 25S rRNA that are involved in interactions with the aa-tRNA from where it first encounters the ribosome in complex with eEF-1A, along the pathway taken by accommodating aa-tRNA (indicated by the orange arrow) (Sanbonmatsu et al., 2005), and into the A-site region of the PTC in the core of the large subunit. Taken together, the rRNA chemical protection studies indicate that this group of W-finger mutants promoted a more accessible or “open” conformation of these regions of the large subunit.
We previously showed that the W255C mutant promoted increased affinity for aa-tRNA (Meskauskas et al., 2005). This analysis was extended in the current study where aa-tRNA binding properties were compared between wild-type and nine mutants. Preparations of non-salt washed ribosomes containing soluble factors were used to monitor aa-tRNA/ribosome interactions because binding of aa-tRNA to the ribosomal A-site requires elongation factor 1 (Schilling-Bartetzko et al., 1992). The saturation curves of [14C]-Phe-tRNA binding shown in figure 4A were analyzed using GraphPad Prism software fitted to a nonlinear regression one site binding curve [Y=Bmax*X/(Kd + X)] where Y denotes bound [14C]-Phe-tRNA values, and X is input [14C]-Phe-tRNA, and Bmax is maximal ligand binding. As summarized in Table 1, Kd values for aa-tRNA varied over an ~10-fold range, from a low of ~ 20nM (W255C, W255R) to >200 nM (H259A).
Coordination of the translation elongation cycle requires that binding of the eEF-1A·aa-tRNA·GTP ternary complex and eEF2 to the elongation factor binding region be mutually exclusive. To examine eEF2/ribosome interactions, affinity purified 6x-His tagged eEF2 was incubated with GDPNP and salt-washed ribosomes isolated from cells expressing the wild-type, and 6 mutant forms of L3. Ribosomes were harvested by centrifugation, the pellets treated with EDTA, and co-fractionating eEF2 was ADP ribosylated using diphtheria toxin and [14C]NAD+. Ribosome-associated eEF2 was then quantified by liquid scintillation. Analyses of the saturation curves (Fig. 4B) revealed notable differences in eEF2 binding among the different ribosomes. As summarized in Table 1, Kd values for eEF2 varied over a 5-fold range, from a low of ~ 8 nM (H259A) to ~ 40 nM (W255K). Inspection of these data clearly shows a reciprocal relationship between eEF-1A stimulated binding of aa-tRNA and binding of eEF2.
Ribosomes isolated from wild-type cells and from three mutants (ΔA254, W255H, iV264) were also assayed with respect to their affinities for Ac[14C]-Phe-tRNA. As summarized in Table 1 (data from saturation curves shown in Fig. 4C), only small changes in P-site binding were observed for the W255H and ΔA254 mutants. These findings are consistent with previous data showing that the W255C mutation did not affect Ac-Phe-tRNA binding to the P-site (Meskauskas et al., 2005).
As shown in Fig. 3E, anisomycin competes with the 3′ end of aa-tRNA for the same binding site in the A-site of the peptidyltransferase center (Hansen et al., 2003). Prior examination of the randomly generated series of L3 mutants demonstrated a correlation between anisomycin resistance and increased ribosomal affinity for aa-tRNA (Meskauskas et al., 2005). To further examine this relationship, ten-fold dilution spot assays were performed on rich media alone or media containing 50μg/ml of anisomycin, and growth of cells in the presence of drug relative to no drug was used to score for anisomycin resistance. The results of these experiments are summarized in Table 1. At position 255, the aromatic substitutions (W255F, W255Y) did not affect anisomycin sensitivity. The longer basic substitutions (W255R, and W255K) showed anisomycin resistance. Changing H256 to alanine (H256A) did not promote resistance to anisomycin but substitution with glutamine (H256Q) promoted drug resistance. Neither the H259L nor H259A mutants were anisomycin resistant, but these mutations in combination with W255C (H256A/W255C, W255C/H259A, and W262A/W255C) were all drug resistant. In fact, W255C/H259A, which was inviable on drug-free media, grew well in presence of anisomycin, i.e. it was anisomycin-dependent. None of the viable insertion or deletion mutants affected growth in the presence of anisomycin. Examination of Table 1 shows that decreasing dissociation constants for [14C]-Phe-tRNA correlated with the degree of anisomycin resistance. Approximately four-fold decreases in Kd values correlated with > 30-fold increases, and ~2-fold decreases correlated with 3 – 30 fold increases in anisomycin resistance respectively.
Maintenance of the yeast killer virus is exquisitely sensitive to subtle defects in the translational apparatus (reviewed in Wickner, 1996), providing a rapid and sensitive assay to monitor the effects of the W-finger mutants on ribosome function. The killer virus maintenance phenotypes of the mutants generated in this study are summarized in Table 1. At position 255, the healthy substitutions (W255F, W255Y, and W255R) were able to maintain the killer virus (Mak+ phenotype), but the less viable W255K and W255H were not (Mak−). In the C-terminal direction, changes at position 256 (H256A, H256Q) and 259 (H259L, H259A) did not affect killer virus maintenance. However, the double mutants of this series in combination with W255C revealed interesting trends: W255C/H256A and W262A/W255C were Mak−, indicating that the W255C mutation is dominant in this aspect. Of the viable insertion and deletion mutants, two promoted the Mak− phenotype (ΔA254 and iV264), and one maintained the killer virus (ΔA250). The L-A viral replicase is synthesized by a -1 PRF event; viral particle assembly and propagation of the killer virus depends on proper rates of -1 PRF (Dinman and Wickner, 1992). PRF efficiency was monitored for wild-type cells and four of the Mak− strains (W255H, W255C, ΔA254 and iV264). Consistent with previous studies, there was a strong correlation between increased levels of -1 PRF and the Mak− phenotype (see Table 1).
Prior studies suggested a correlation between decreased peptidyltransferase activity and increased -1 PRF and the Mak− phenotype (Meskauskas et al., 2003), but a later investigation suggested that other factors may influence the relationships between these parameters (Meskauskas et al., 2005). Peptidyltransferase activities were significantly decreased in all 6 mutants tested (ΔA254, W255C, W255H, W255K, W255R, and iV264, see Fig. 4D and Table 1), and in general, this corresponded with the inability of cells to maintain the killer virus (see Table 1). The W255R mutant was the exception: although peptidyltransferase activity in this mutant was inhibited to the same extent as e.g. ΔA254 (which had elevated levels of -1 PRF and was Mak−), this mutant nonetheless stably maintained the Killer+ phenotype.
The elongation cycle requires coordination of an ordered series of events. The elongation factor 1·aa-tRNA·GTP ternary complex must first interact with the ribosomes’ factor-binding site. At the decoding center in the small subunit, the presence of cognate codon:anticodon base-pairing is transduced up the body of the aa-tRNA, activating GTP hydrolysis by elongation factor 1 (Valle et al., 2002; Cochella and Green, 2005). This initiates accommodation of the aa-tRNA into the PTC. Successful accommodation has been hypothesized to be accompanied by rRNA structural changes that both close the aa-tRNA accommodation corridor (Sanbonmatsu et al., 2005), and activate the PTC (Schmeing et al., 2005). Concurrently, the structure of the factor binding site must also change to favor elongation factor 2 binding, setting the stage for translocation (Sergiev et al., 2005). The rRNA structure probing and ribosome binding data presented in the current study suggest a model describing how L3 may help to coordinate these ribosome-associated functions by acting as a “gatekeeper” to the A-site.
Computer modeling studies suggest that during accommodation, the aa-tRNA 3′ end moves through a ‘corridor’ formed by helix 89 on one side and the helix 90 – 92 structure on the other (Sanbonmatsu et al., 2005). The chemical protection data clearly shows that specific bases in the helix 90 – 92 side of this pathway are generally deprotected in the W255C mutant as compared to wild-type ribosomes. Thus, the increased accessibility of bases in this structure to small chemicals suggests that it is in a relatively ‘open’ conformation (Fig. 3A). The computational simulations of aa-tRNA accommodation also suggest that the aa-tRNA 3′ end passes through two rRNA-based structures or “gates”. The first gate encountered by aa-tRNA (gate 1) is formed by U2491 and C2556 (yeast U2860 and C2924), while the second (gate 2) is formed by C2573 (yeast C2941) (see Fig. 3D and 3E). Notably, the gate bases in the helix 90 – 92 structure (C2924 and C2941) were deprotected, consistent with an open conformation. Interestingly, U2861 was protected from chemical attack. It is possible that reorientation of the helix 89 gate 1 base U2860 had a shielding effect on U2861, consistent with the idea that the W255C mutant promotes opening of the accommodation corridor. Additionally, the chemical protection studies suggest that these conformational changes in this, and other mutants, are promulgated along the entire pathway traveled by the aa-tRNA 3′ end, from the SRL down to the PTC (Figs. 3B, 3C).
Opening of the aa-tRNA accommodation corridor should affect the biochemical properties of the ribosome. Specifically, such changes should promote increased affinity for aa-tRNA, which is supported by the observation of decreased Kd values for aa-tRNA by the W255C, R, K, and H256Q mutants. The increased eEF2 Kd values for these mutants are also consistent with the notion that ternary complex and eEF2 binding should be mutually exclusive. This is further supported by the H256A and H259A mutants, which promoted increased affinity for eEF2 at the expense of their ability to interact with aa-tRNA. These biochemical observations suggest that two classes of mutants were generated: one promoting ribosomal conformations more favorable to eEF2 binding, the other favoring binding of the eEF-1A·aa-tRNA·GTP ternary complex. These data suggest that in the L3 mutants, bases in the PTC and along the aa-tRNA accommodation pathway have been artificially shifted into conformations that are normally transitory during translation in wild type cells. Specifically, we propose that some W-finger mutants (e.g. W255C, R, K, and perhaps H256Q) favor positioning of the SRL into the eEF-1A·aa-tRNA·GTP binding state, and the accommodation corridor into the “open” conformation, while others (H256A, H269A) favor the closure of the corridor and eEF2 binding.
These findings suggest a model wherein the W-finger may normally assume different positions within the ribosome during these two factor-binding states, possibly accompanied by different patterns of putative interactions between amino acids in the W-finger and 25S rRNA bases. This is cartooned in Fig. 5. By this model, movement of the L3 W-finger may participate, either actively or passively, in an allosteric process to coordinate these events. Specifically, positioning of the W-finger toward its N-terminal side, i.e. directed away from the PTC and toward the factor binding site, may help to both open the helix 90 – 92 face of the accommodation corridor, and to position the SRL so as to favor binding of the eEF-1A·aa-tRNA·GTP ternary complex. We suggest that while in this conformation, A2940 of 25S rRNA may interact with H259 of L3. The model proposes that during accommodation, the W-finger swings toward its C-terminal side (toward the PTC) and that the A2940-W255 interaction is established upon completion of aa-tRNA accommodation into the A-site. Additionally, since the aa-tRNA and eEF2 Kd values for the H256A mutant were between those observed for the W255 mutants (W255C, R, and K) and H259A, it is also possible that interaction between A2940 and H256 could constitute a transitional intermediate between the two states. We propose that during rearrangement of the W-finger, the helix 90 – 92 structure moves to close the accommodation gates. Concurrently, the SRL atop helix 95 is repositioned so as to form the eEF2 competent binding site. This sequence of events would allow the occupancy status of the A-site to be communicated to the translation factor binding region, and prevent entry of new aa-tRNA into the A-site during translocation. It must be noted that our results have only revealed amino acid residues in the W-finger that may be involved in these conformational events, not their rRNA counterpart bases. Although the chemical properties of the substituted amino acids suggest their involvement in interactions with nucleotides, the actual rRNA base involved was not experimentally determined, but A2940 represents the most likely candidate based on the crystal structures.
Resistance of the W255C mutant to anisomycin has been known for a many years (Fried and Warner, 1981), but the underlying mechanism has been unclear. Early studies showed that the W255C mutation did not affect anisomycin binding (Jimenez and Vazquez, 1975), suggesting that structural and/or conformational changes in resistant ribosomes may reduce the ability of anisomycin to compete with aa-tRNA for entry into the A-site. The chemical protection data presented in the current study show structural changes in the large subunit consistent with a more “open” conformation in the vicinity of the A-site of the PTC and in the aa-tRNA accommodation corridor in the anisomycin resistant mutants. Opening of the accommodation corridor should facilitate aa-tRNA passage into the A-site, consistent with the decreased Kd values for aa-tRNAs. In contrast, because of its small molecular radius relative to aa-tRNA, access to the PTC by anisomycin should not be as intrinsically limited by the two gates, and hence should not be stimulated as much by gate opening. Therefore, it is possible that these mutations lower the barriers encountered by aa-tRNA relative to anisomycin during accommodation, thus altering the kinetic partitioning ratios for these two molecules, resulting in anisomycin resistance. An alternate, non-exclusive explanation may be that changes in the interactions between anisomycin and bases in the PTC (see Fig. 3E) consequent to these mutations may contribute to anisomycin resistance. Interestingly, the –OH group of the N-terminal serine of L3 (S2) is predicted to interact with C2924, the helix 90 ”Gate 1” residue (see Fig. 3E). The S2T mutant was also strongly anisomycin resistant (Meskauskas et al., 2005), and experiments are ongoing characterizing a bank of S2 mutants.
An “entropic catalysis” model of peptidyltransfer posits that precise positioning of substrates and water reorganization in the PTC allows formation of a highly ordered electrostatic network that stabilizes reaction intermediates, providing the proper environment for the ribosome to accelerate peptide bond formation (reviewed in Rodnina et al., 2007). The rRNA structure probing experiments in the current study are consistent with this model insofar as distortion of the topology of the A-site side of the PTC by the W-finger mutants could affect positioning of the aa-tRNA 3′ end, promoting the observed peptidyltransfer defects. It has also been proposed that the PTC is activated by an ‘induced fit’ mechanism, wherein accommodation of the aa-tRNA acceptor stem into the A-site reorganizes the geometry of the PTC, making the ester group of the peptidyl-tRNA accessible for nucleophilic attack (Schmeing et al., 2005). The involvement of the L3 W-finger in this process remains an intriguing possibility.
It is remarkable how atomic scale allosteric events can propagate themselves all the way to the biological level. Translating our understanding of molecular structure and biochemical properties of molecules may lead to concrete medical benefits. For example, understanding how these mutants affect the interactions between ribosomes and transacting factors, and how this leads to resistance to A-site specific antibiotics can provide the foundation for the design of a new antibacterials to counter resistance of pathogenic bacteria to drugs such as tiamulin (Pringle et al., 2004). This research may also aid in the development of new antivirals insofar as it furthers our understanding of how changes at the atomic level can affect peptidyltransfer, and hence –1 PRF and virus propagation. It is hoped that atomic resolution structures of yeast ribosomes will be solved in the near future, and that achievement of this milestone will enable the models proposed herein to be directly visualized.
E. coli DH5α was used to amplify plasmid DNA. Transformation of E. coli and yeast, and preparation of yeast growth media (YPAD, synthetic drop out medium, and 4.7 MB plates for testing the killer phenotype) were as previously reported (Dinman and Wickner, 1992). Restriction enzymes were obtained from MBI Fermentas (Vilnius, Lithuania). The QuikChange XL II site-directed specific mutagenesis kit was obtained from Stratagene (La Jolla, CA). Macrogen Inc. (Seoul, South Korea) performed DNA sequence analysis. Oligonucleotide primers were purchased from IDT (Coralville, IA). The yeast strains used in this study were all derived from the rpl3-gene disruption (rpl3Δ) strain JD1090 (MATα ura3-52 lys2-801 trp1δ leu2− his3 RPL3::HIS3 pRPL3-URA3-CEN6 [L-A HN M1]) (Meskauskas et al., 2003). Mutants of rpl3 were generated using the wild type RPL3 gene in pJD225 (Meskauskas et al., 2003), synthetic oligonucleotides, and the QuikChange XL II kit. The pYDL dual luciferase reporter series of plasmids for monitoring programmed ribosomal frameshifting (PRF) were used as previously described (Harger and Dinman, 2003). Ten-fold dilution spot assays were performed on rich media alone and media containing 50μg/ml of anisomycin, and growth of cells in the presence of drug relative to no drug was used to score for anisomycin resistance.
Yeast phenylalanyl-tRNAs were aminoacylated with [14C]Phe-tRNA to monitor A-site binding, Ac-[14C]Phe-tRNA was generated to monitor P-site binding, these were purified by HPLC, and equilibrium binding studies were performed as previously described (Meskauskas et al., 2005). Peptidyltransfer assays were performed essentially as previously described (Dresios et al., 2001; Meskauskas et al., 2005). 6xHis-tagged eEF2 was purified from TKY675 yeast cells (kindly provided by Dr. T. Kinzy) as previously described (Ortiz et al., 2006). For eEF2 binding experiments, reaction mixes containing 25 pmoles of salt washed 80S ribosomes and various concentrations of 6xHis-tagged eEF2 (5–40 pmol) in binding buffer (50 mM Tris-HCl, pH 7.5, 50 mM ammonium acetate, 10 mM magnesium acetate, 2 mM DTT, 100 μM GDPNP) were incubated for 5 min at room temperature, layered on top of the 300-μl sucrose cushion (10% sucrose in binding buffer) and centrifuged at 50,000 rpm for 20 min at 4°C in a MLS50 (Beckman) rotor. Pellets (bound fraction) were washed and resuspended in 50 μl of DT buffer (50 mM Tris-HCl, pH 7.5, 20 mM DTT, 1 mM EDTA). Ribosome bound eEF2 was ADP-ribosylated for 20 min at 37°C by adding 100 pmol of [14C]NAD+ and 0.2 μg of diphtheria toxin. Reaction mixes were precipitated with TCA, and amounts of [14C]ADP-ribosylated eEF2 were determined by liquid scintillation counting. Control values (lacking eEF2, not exceeding 10% of the lowest count) were subtracted. Kd values were determined assuming single binding sites using Graphpad Prism software.
Ribosomes were isolated as described (Meskauskas et al., 2005) and were synchronized by puromycin treatment in buffer C (50 mM Tris-HCl pH 7.5, 5 mM magnesium acetate, 50 mM ammonium chloride, 0.1 mM PMSF, 0.1 mM DTE, 25% glycerol) containing 1 mM of GTP, 1 mM of puromycin and 500 pmol of ribosomes. After incubation for 30 min at 30°C, ribosomes were sedimented through a cushion of buffer B (20 mM Tris-HCl pH 7.5, 10 mM magnesium acetate, 0.1 mM PMSF, 0.1 mM DTE, 0.5 M KCl, 25% glycerol), and suspended in buffer C at 10 pmol/μl. Chemical probing with dimethylsulphate (DMS), kethoxal, and carbodiimide metho- p-toluenesulfonate (CMCT), followed by RT primer extension analysis of modified RNAs were performed as described (Stern et al., 1988). Primers (numbered after the first transcribed base of yeast 25S rRNA) employed for these analyses were 2957 (5′-AACCTGTCTCACGACGG-3′), 3043 (5′-CCTGATCAGACAGCCGC-3′), 2877 (5′-GGTATGATAGGAAGAGC-3′), 2435 (5′-CCTCTATGTCTCTTCAC -3′), and 2675 (5′-GTTCTACTGGAGATTTCTG-3′).
The X-ray crystal structure of the H. marismortui 50S ribosomal subunit (Ban et al., 2000) and the cryo-electron microscopy (cryo-EM) reconstruction of Saccharomyces cerevisiae ribosomal proteins threaded onto the X-ray crystal structure of the H. marismortui 50S ribosomal subunit (Spahn et al., 2004) was visualized using PyMOL (DeLano Scientific LLC).
We would like to thank the members of our laboratory, with special thanks to Jennifer Baxter-Roshek, and Alexey Petrov for stimulating discussions, and to Rasa Rakauskaite and Johnathan Russ for technical assistance. This work was supported by grants from the NIH to JDD (GM58859) and from the American Heart Association to AM (AHA 0630163N).
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.