Mutations in Ribosomal Protein L10e Confer Resistance to the Fungal-specific Eukaryotic Elongation Factor 2 Inhibitor Sordarin*

The natural product sordarin, a tetracyclic diterpene glycoside, selectively inhibits fungal protein synthesis by impairing the function of eukaryotic elongation factor 2 (eEF2). Sordarin and its derivatives bind to the eEF2-ribosome-nucleotide complex in sensitive fungi, stabilizing the post-translocational GDP form. We have previously described a class of Saccharomyces cerevisiae mutants that exhibit resistance to varying levels of sordarin and have identified amino acid substitutions in yeast eEF2 that confer sordarin resistance. We now report on a second class of sordarin-resistant mutants. Biochemical and molecular genetic analysis of these mutants demonstrates that sordarin resistance is dependent on the essential large ribosomal subunit protein L10e in S. cerevisiae. Five unique L10e alleles were characterized and sequenced, and several nucleotide changes that differ from the wild-type sequence were identified. Changes that result in the resistance phenotype map to 4 amino acid substitutions and 1 amino acid deletion clustered in a conserved 10-amino acid region of L10e. Like the previously identified eEF2 mutations, the mutant ribosomes show reduced sordarin-conferred stabilization of the eEF2-nucleotide-ribosome complex. To our knowledge, this report provides the first description of ribosomal protein mutations affecting translocation. These results and our previous observations with eEF2 suggest a functional linkage between L10e and eEF2.

The natural product sordarin, a tetracyclic diterpene glycoside, selectively inhibits fungal protein synthesis by impairing the function of eukaryotic elongation factor 2 (eEF2). Sordarin and its derivatives bind to the eEF2-ribosome-nucleotide complex in sensitive fungi, stabilizing the post-translocational GDP form. We have previously described a class of Saccharomyces cerevisiae mutants that exhibit resistance to varying levels of sordarin and have identified amino acid substitutions in yeast eEF2 that confer sordarin resistance. We now report on a second class of sordarin-resistant mutants. Biochemical and molecular genetic analysis of these mutants demonstrates that sordarin resistance is dependent on the essential large ribosomal subunit protein L10e in S. cerevisiae. Five unique L10e alleles were characterized and sequenced, and several nucleotide changes that differ from the wild-type sequence were identified. Changes that result in the resistance phenotype map to 4 amino acid substitutions and 1 amino acid deletion clustered in a conserved 10-amino acid region of L10e. Like the previously identified eEF2 mutations, the mutant ribosomes show reduced sordarin-conferred stabilization of the eEF2-nucleotide-ribosome complex. To our knowledge, this report provides the first description of ribosomal protein mutations affecting translocation. These results and our previous observations with eEF2 suggest a functional linkage between L10e and eEF2.
Eukaryotic elongation factor 2 (eEF2) 1 and its prokaryotic counterpart, elongation factor G (EF-G), promote the translocation of the ribosome along messenger RNA during the elongation phase of protein synthesis. Hydrolysis of GTP to GDP drives translocation and is associated with a presumed conformational change in eEF2. Sordarin (1) and its analogs are fungal-specific translation inhibitors (2,3) that bind to the eEF2-ribosome-GDP complex in Saccharomyces cerevisiae, stabilizing the post-translocational GDP form in a manner similar to that of fusidic acid (3). However, in contrast to fusidic acid, which binds both EF-G and eEF2 and is a general translocation inhibitor, sordarin inhibits translation only in susceptible fungi, deriving its unique specificity from the source of eEF2 (3)(4)(5). The observation that eEF2 is the major determinant of sordarin specificity was confirmed by the identification of 15 unique sordarin-resistant alleles of EFT1 and EFT2 that encode eEF2 in S. cerevisiae. In our original characterization of 21 sordarin-resistant mutants, five mutations were not linked to the EFT1 or EFT2 genes. In this work, we show that these five mutations map to the essential ribosomal protein L10e.
The ribosome, although not contributing significantly to the fungal specificity of sordarin, is a critical partner in forming the stabilized post-translocational complex (3). Detection of a complex between fungal eEF2 and a labeled sordarin analog is strongly dependent upon the presence of ribosomes. L10, the prokaryotic counterpart of S. cerevisiae L10e, has been localized to the base of the stalk structure conserved in all large ribosomal subunits (6 -8). The eukaryotic ribosomal stalk proteins L10e, L12eIA, L12eIIA, L12eIB, and L12eIIB in S. cerevisiae (9, 10) comprise a pentameric structure that is similar to the L10 and L7/L12 complex in Escherichia coli ribosomes. The conservation of the stalk structure has been visualized in recent cryoimages of both 70S (11), in which the binding position of the EF-G-GDP-fusidic acid complex is observed in detail, and 80S (12) ribosomes. The prokaryotic L7/L12 ribosomal proteins have been studied extensively by many physical and biological techniques, but much less is known about the structure and function of prokaryotic or eukaryotic L10. L10 is among the proteins reported to be cross-linked to eEF2 in 80S ribosomes by bifunctional reagents (13). In yeast, mutational analysis of the L10e ribosomal protein gene has shown that only L10e is essential, and that the carboxyl-terminal 132 amino acids of the L10e protein are required for viability. However, the L12e proteins that comprise the L10e/L12e pentameric complex are not essential (14). Several findings suggest that there are associations between the stalk proteins and elongation factors. Mutations in L7/L12 perturb both EF-Tu and EF-G functions in E. coli (15). Chemical cross-links have been observed between EF-Tu and EF-G and the L7/L12 complex (reviewed in Ref. 16). Of particular relevance to the present case, cross-links are observed between the EF-G and L7/ L12 proteins in the presence of the nonhydrolyzable GTP analog GMPPCP, but not in the presence of GDP and fusidic acid (17,18). L7/L12 proteins in the EF-G-fusidic acid-ribosome complex are resistant to trypsin proteolysis (19). Our current results add to the body of information implicating L10e and eEF2 interactions to be important in translocation. These studies define a role for a small region of L10e in mediating inhibition by a new class of natural product with unprecedented selectivity for fungal protein synthesis and provide evidence for a functional interaction between L10e and eEF2.

EXPERIMENTAL PROCEDURES
Yeast Strains, Plasmids, and Compounds-Sordarin was isolated essentially as described from Sordaria arenosa (1), and preparation of * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. the sordarin analog L-793,422 has been described previously (3,20). The sordarin-resistant strains sRb1, sRb2, sRb5, sRb13, and sRb14 were generated essentially as described previously (3), except that mutant strains that demonstrated sordarin resistance unlinked to EFT1 or EFT2 were selected for characterization in this study. The wild-type yeast strain YPH54 used for genetic analysis and the plasmid YCplac111 used for subcloning the S. cerevisiae L12e and L10e genes have both been described previously (3).
Cloning of L10e and L12e Genes-The ribosomal protein genes L10e, L12eIA, L12eIB, L12eIIA, and L12eIIB were cloned by polymerase chain reaction (PCR) using synthetic oligonucleotides ( Table I) that were designed based on published sequences (10). These oligonucleotides also contain the 5Ј BglII restriction enzyme recognition sequence 5Ј-AGATCT-3Ј after the sequence 5Ј-GCGCGC-3Ј for subcloning purposes. S. cerevisiae genomic DNA from wild-type and sordarin-resistant strains was prepared by standard procedures and purified with a QIAamp Tissue Kit (Qiagen) according to the manufacturer's specifications. PCR products were amplified from genomic DNA with the synthetic oligonucleotides listed in Table I, using Klentaq polymerase (CLONTECH) according to the manufacturer's recommendations. The PCR products were purified over Qiaquick columns (Qiagen), and putative PCR-generated L10e and L12e ribosomal protein genes were identified by restriction analysis. Each PCR product was digested with BglII, followed by ligation into the BamHI site of YCplac111. Ligation mixes were transformed into the E. coli strain DH5␣ (Life Technologies Inc.). E. coli transformants were pooled, inoculated into LB broth containing 100 g/ml ampicillin, and grown overnight at 37°C. Plasmid DNA was prepared, and the constructs YCpL10e, YCpL12eIA, YCpL12eIB, YCpL12eIIA, and YCpL12eIIB were verified by restriction analysis. Plasmid DNA pools of each construct described above were used to transform the S. cerevisiae eft⌬1/eft⌬2 strain YEFD12 h (21) to leucine prototrophy. Colonies from the individual transformations were pooled, and 1 ϫ 10 6 cells were plated on SC medium containing 2% agar and 5 g/ml sordarin. Cultures were incubated at 29°C for approximately 4 days until colonies appeared. The pooled transformants were also tested in liquid culture essentially as described previously (3). In this assay, cells were inoculated in SC medium depleted of leucine, followed by incubation at 29°C in a humidified chamber for approximately 8 days. Transformants exhibiting plasmid-dependent sordarin resistance were clonally purified, followed by plasmid rescue and transformation into E. coli.
Molecular Mapping of eEF2 Mutations-Plasmid DNA from transformants with the resistance phenotype was sequenced with an ABI Prism 373 DNA Sequencer according to the manufacturer's recommendations (Perkin-Elmer Applied Biosystems). Sequences were analyzed using Sequencher DNA analysis software (Gene Codes Corp.). The products of three additional independent PCR reactions were sequenced to rule out the possibility of errors due to polymerase infidelity.
Cell Extracts-Washed cells grown to logarithmic phase (A 600 , 1-1.5) in YPAD medium at 30°C were disrupted, and post-ribosomal extracts were prepared essentially as described by Skogerson (22). Wild-type eEF2 was purified to a final specific activity of 2.5 pmol/g in the diphtheria-catalyzed ADP ribosylation assay. Ribosomes were prepared using standard procedures (23), and their concentration was estimated using the figure of 18.6 pmol/A 260 unit. Subunits were separated by sucrose gradient sedimentation (23). Ribosomal proteins were extracted using acetic acid precipitation and acetone extraction, and core particles remaining after the removal of the acidic L12e proteins were prepared by treating intact ribosomes with 50% ethanol and either 0.4 or 1.0 M KCl (24). SDS-polyacrylamide gel electrophoresis chromatography and Coomassie staining were used to monitor the release of protein of the expected size (about 13 kDa).
Binding-[ 3 H]L-793,422 was prepared at 20 mCi/mg (8000 mCi/ mmol) at a concentration of 0.004 mg/ml by the Drug Metabolism Department at Merck Research Laboratories. The gel filtration method used previously (3) was replaced by a filter binding assay. Filter choice, presoaking, and washing regimens were carefully optimized. GF/C filters (Millipore) were presoaked in 0.15% polyethyleneimine plus 0.25% Triton X-100. Binding was performed in a 1-ml volume (or greater, to check the validity of K d measurements) of 50 mM Tris-HCl, pH 7.4, 10 mM MgCl 2 , 25 M guanosine 5Ј-3-O-(thio)triphosphate, and 10 mM ␤-mercaptoethanol in glass tubes for 30 min at 22°C. Chilled 0.5% Triton X-100 (3 ml) was added, and the tube contents were immediately filtered for approximately 3 s under a vacuum maintained at a constant pressure of 300 mm Hg. Two additional 3 ml washes were performed, and the filter was removed for counting in scintillation fluid. Background binding of about 70 dpm was obtained with GTP Binding-The gel filtration assay used previously (3) was also replaced by a filter binding assay. Nitrocellulose filters (0.45 m; 25-mm diameter; BA85; Schleicher & Schuell) were presoaked in 50 mM Tris-HCl, pH 7.5, 10 mM MgCl 2 , 100 mM NH 4 Cl, and 2 mM ␤-mercaptoethanol. The binding reaction (200 l) was performed in the same buffer plus 0.92 pmol of purified eEF2, 3 pmol of salt-washed ribosomes, 0.1 Ci of [guanosine-8-3 H]GTP (12 Ci/mmol), and increasing amounts of sordarin or analogs. After 30 min at 22°C, 1 ml of ice-cold buffer was added, and the contents were filtered as described above, followed by a final 1 ml wash. Under these conditions, approximately 0.1 mol of nucleotide was trapped per mole of eEF2 in the absence of any compound.
GTPase Assay-The eEF2-dependent ribosomal GTPase activity was measured at 22°C for wild-type ribosomes and each mutant using 0.46 pmol of wild-type eEF2 and 0.8 pmol of ribosomes per 20-l reaction at 20 mM Tris, pH 7.4, 5 mM Mg(OAc) 2 , 20 mM NH 4 Cl, 2 mM dithiothreitol, and 500 M [␣-32 P]GTP (200 dpm/pmol). Duplicate 1-l samples were removed at the indicated times and quenched, and GDP release was analyzed as described previously (25).
Translation Assays-Polymerization using S30 extracts was performed at 22°C essentially as described by Hussain and Leibowitz (26) with poly(U) (Calbiochem) at 160 g/ml as a template during the linear period of incorporation at 22°C using [ 3 H]phenylalanine (4000 dpm/ pmol) as a precursor. Polymerization reactions were performed with either intact ribosomes or those reconstituted from separated subunits and contained 0.05 A 260 unit of ribosomes and 10 g of S100 protein as a source of all soluble factors. In these assays, incorporation of [ 3 H]phenylalanine was limited by the level of ribosomes.

RESULTS AND DISCUSSION
A Class of Sordarin-resistant Mutations Not Linked to eEF2-We have previously described S. cerevisiae mutants containing unique amino acid substitutions in eEF2 that confer resistance to sordarin (3). These mutants were obtained by spontaneous mutagenesis of eft1⌬/eft2⌬ deletion strains harboring an episomal copy of either EFT1 or EFT2, genes that encode eEF2. Linkage of sordarin resistance to eEF2 was confirmed by demonstrating plasmid-dependent sordarin resistance in sensitive strains. In addition to the eEF2 mutants, five additional sordarin-resistant mutants (sRb1, sRb2, sRb5, sRb13, and sRb14) were identified that harbored wild-type episomal EFT1 or EFT2. All of these strains had sordarin IC 50 s for growth inhibition in SC medium of 10 -30 g/ml (Fig. 1). No additional phenotypes were observed for these mutants in standard laboratory media, with the exception of a growth defect in strain sRb2 (data not shown).
To characterize the genetic elements that mediate sordarin resistance, diploids were generated from crosses of the sRb strains and wild-type strain YPH54. In growth inhibition assays, sRb/YPH54 diploids exhibited a codominant phenotype at low concentrations of sordarin. Tetrad analysis of all diploids showed that sordarin resistance segregated 2 s :2 r , indicating that the resistance phenotype results from mutations in a single chromosomal gene (data not shown). Complementation testing was performed on sRb/sRb diploid strains to determine the number of genes represented. Approximately the same level of resistance to sordarin was observed among the sRb/sRb diploids, suggesting that all the mutant strains belong to the same complementation group. This result was confirmed by tetrad analysis of the sRb/sRb mutant strains, indicating that the sRb mutation is defined by a single nuclear gene (data not shown).
Resistance in Vitro Is Associated with the Large Ribosomal Subunit-S30 extracts prepared from each of the sRb mutants showed a marked reduction in sordarin sensitivity in in vitro translation assays using poly(U) as a template. This is illustrated for the sRb5 mutant in Fig. 2A. The other four mutants displayed essentially identical inhibition curves (data not shown), which shifted about 6-fold from the control value at low sordarin concentrations and showed no further inhibition at levels up to 5 g/ml. This biphasic response could be due to the functional redundancy of a ribosomal protein(s) or to the limitations of the assay.
Fractionation of the S30 extracts into salt-washed ribosomes and S100, followed by reconstitution, did not alter the sordarin inhibition curve for the sRb mutants. However, experiments using S100 extracts and ribosomes from either the sordarinsensitive parental strain or the sRb mutants allowed for the identification of the resistance determinant as the ribosome in all cases (sRb5 is shown in Fig. 2B). Ribosomes prepared from strain sRb5 were further separated into large and small subunits, and, as shown in Fig. 2C, resistance was found to be associated with the large ribosomal subunit via in vitro translation experiments. Reconstitution studies between subunits of wild-type strains and those of the other sRb strains gave similar results (data not shown).
Ribosomal Protein L10e Confers Sordarin Resistance-Identification of the large ribosomal subunit as the determinant of sordarin resistance, in addition to the biochemical and genetic associations of the stalk proteins with EF-G or eEF2 (11,13,(15)(16)(17)(18)(19), focused our efforts on the ribosomal stalk proteins. Because both fusidic acid and sordarin stabilize the eEF2-GDPribosome complex, we hypothesized that proteins in the stalk were candidate biochemical targets of sordarin. In S. cerevisiae ribosomes, the stalk is composed of five proteins: L10e and a tetramer of L12eIA, L12eIIA, L12eIB, and L12eIIB. These proteins are the eukaryotic counterparts of the L10 and L7/L12 proteins in E. coli. S. cerevisiae L10e-and L12e-specific oligonucleotides were synthesized (Table I) and used in PCR amplification of genomic DNA from the sordarin-resistant strain sRb5. Putative L10e and L12e ribosomal protein genes were verified, subcloned, and transformed into sordarin-sensitive yeast. Colonies from individual transformations were pooled and plated on SC agar medium containing sordarin or tested in growth inhibition assays. After incubation for 4 days at 29°C, colonies appeared only on plates with transformed cells containing episomal L10e from sRb5 (data not shown). Representative data from a growth inhibition assay are shown in Fig. 3A. Based on these results, L10e clones from the four remaining sRb strains were generated and examined in a similar manner. As shown in Fig. 3B, L10e from all strains conferred sordarin resistance, although less for sRb2 than for the others, as was seen for the original isolated mutants (Fig. 1).
Plasmid DNA from the L10e transformed strains was recovered and sequenced. Several nucleotides that differ from the sequence published by Hunter Newton et al. (Ref. 10; Gen-Bank TM accession number M26506) were identified (Fig. 4). To determine the specific change responsible for the resistant phenotype, the L10e gene from the parental strain YEFD12h was amplified by PCR and sequenced. A comparison of the published sequence with our parental strain showed that L10e in YEFD12h contains three nucleotide changes that result in three conservative amino acid substitutions: I4V, V27I, and FIG. 1. Non-eEF2 (sRb) mutants are partially resistant to sordarin. Growth inhibition assays were performed in a microtiter format in which cells were inoculated to an A 600 of 0.02 in SC medium containing sordarin serially diluted 2-fold from 0.2 to 100 g/ml, followed by incubation at 29°C for approximately 24 h. q, sRb1; E, sRb2; ϫ, sRb5; f, sRb13; Ⅺ, sRb14; OE, wild-type (YEFD12h).
T281S. Based on this result, we were able to identify the L10e amino acid substitutions that confer sordarin resistance (Fig.  4). All of the amino acid substitutions resulted from single nucleotide changes, with the exception of sRb14 (S134⌬), which results from the deletion of three nucleotides. All of the changes conferring resistance were found to be clustered within 10 residues located approximately in the middle of the L10e polypeptide.
L-793,422 Binding Studies-Fungal specificity by the sordarin class of compounds is conferred by the source of eEF2. However, for a stable interaction to be detected with [ 3 H]L-793,422, ribosomes must be present (Table II). Titration of the ribosomes with a fixed amount of eEF2 showed that an approximately 3-fold excess of ribosomes was required to give saturation binding. Using this level, [ 3 H]L-793,422 binding reached saturation at about 0.85 pmol/pmol eEF2, with an apparent K d of 2.5 nM. No substantial differences were seen for [ 3 H]L-793,422 binding B max or the apparent K d when studies were conducted with mutant ribosomes (sRb1, sRb2, or sRb5) and wild-type eEF2. Also, the level of ribosomes required for saturation did not deviate significantly from the wild-type value for any of the mutants (data not shown).
The minimal requirements for the ribosomal contribution to binding were assessed by dissociating ribosomes into components (Table II). The L12e stalk was removed from ribosomes, i.e. core ribosomes, without significant binding loss. However, the separation of subunits caused a marked decrease in their ability to stimulate eEF2-associated binding. Residual binding in the presence of individual subunits was consistent with the approximately 10% cross-contamination of large subunits by light as estimated by Western blotting with L10e antiserum. Reconstitution of the separated subunits restored binding. Solubilized protein from complete ribosomes or large subunits did not stimulate the very low level of binding exhibited by eEF2 alone. Wild-type L10e purified by reverse phase chromatography also failed to stimulate eEF2 binding of sordarin (data not shown). Thus, the ribosomal component of binding requires both subunits and possibly the interface between them and cannot be supplied by soluble ribosomal proteins.
Sordarin as a Translocation Inhibitor-To examine translo-cation more directly, we assessed the stability of the nucleotide-eEF2-ribosome complex for each of the ribosomal mutants, using [guanosine-8-3 H]GTP and a rapid filtration assay. The amount of labeled nucleotide trapped in the presence of increasing levels of sordarin is shown in Fig. 5A for wild-type eEF2 plus wild-type or mutant sRb5 ribosomes. Without the addition of sordarin, approximately 0.1 mol of nucleotide is trapped per mole of eEF2. Sordarin increased this to a maximum of 0.28 (Ϯ 0.007) mol for wild-type ribosomes (with an IC 50 of about 25 ng/ml), 0.23 (Ϯ 0.011) mol for sRb1 ribosomes, 0.18 (Ϯ 0.007) mol for sRb2 ribosomes, 0.20 (Ϯ 0.010) mol for sRb5 ribosomes, 0.22 (Ϯ 0.009) mol for sRb13 ribosomes, and 0.16 (Ϯ 0.005) mol for sRb14 ribosomes. Thus, the eEF2 complexes containing mutant ribosomes all showed a 20 -40% reduction in sordarin-conferred stabilization when compared with wild-type ribosomal complex, consistent with their approximately equal resistance to sordarin for in vitro  3. L10e cloned from the sRb mutant strains confers sordarin resistance. Growth inhibition assays with strains harboring YCpL10e, YCpL12eIA, YCpL12eIB, YCpL12eIIA, and YCpL12eIIB. A, the L10e and L12e ribosomal protein genes from wild-type and the sordarin-resistant mutant sRb5 strain were cloned and expressed in S. cerevisiae as described under "Experimental Procedures." q, YCpL12eIA; E, YCpL12eIB; ϫ, YCpL12eIIA; f, YCpL12eIIB; Ⅺ, YCpL10e cloned from YEFD12h; OE, YCpL10e cloned from sRb5; ‚, the YCp111 yeast shuttle vector. L12e ribosomal protein genes were cloned from strain sRb5. B, L10e from all sRb mutants was cloned and expressed in S. cerevisiae. q, YCpL10e from sRb1; E, YCpL10e from sRb2; ϫ, YCpL10e from sRb5; f, YCpL10e from sRb13; Ⅺ, YCpL10e from sRb14; OE, YCp111. Growth inhibition assays were performed as described previously (3), except that SC medium and incubation to 48 h were used.

translation.
Turnover rates in translocation are reflected in the ribosomal EF2-dependent GTPase activity. Fusidic acid at low millimolar levels strongly inhibits the ribosomal-dependent GTPase activity of yeast eEF2 (27). Sordarin, although exerting its effect with much higher affinity, has less effect on GTPase activity. Sordarin concentrations of Ն100 ng/ml maximally inhibited the GTPase activity of wild-type ribosomes and eEF2 by about 60%. Similar experiments with wild-type eEF2 and each of the ribosomal mutants showed similar titration curves (data not shown), but the inhibition of GTPase activity was reduced 30 -40% relative to wild-type eEF2 and ribosomes. Fig. 5B shows a time course for sRb1 compared with wild-type ribosomes at a single saturating sordarin concentration of 1 g/ml. Approximately the same reduction was obtained whether the mutant GTPase activity was lower than the wild-type (0.8 pmol/pmol eEF2/min), as for sRb1 (0.4 pmol/pmol/min eEF2), or higher (1.1 units for sRb2 and 1.2 units for sRb5). Consistent with these elevated GTPase rates, ribosomes sRb2 and sRb5 both showed somewhat elevated translation rates with wildtype eEF2 in the absence of sordarin under the established assay conditions (data for sRb5 may be seen in the legend of Fig. 2). Overall, the magnitude of the reduction in GTPase activity exhibited by the mutants resembles that seen for the nucleotide complex. Because we have demonstrated that the ribosomal mutants are not impaired in sordarin binding, the reduced sensitivity seen in GTPase and complex stability must represent a functional alteration in translocation for the mutants. The amino acid substitutions that result in the sordarinresistant phenotype cluster to a 10-amino acid region approx- FIG. 4. Peptide sequence of L10e and the amino acid substitutions that confer sordarin resistance. The ribosomal protein gene L10e was cloned from the wild-type strain YEFD12h and the sordarin-resistant mutants as described under "Experimental Procedures." Changes in L10e that confer the resistance phenotype map to four amino acid substitutions and one amino acid deletion clustered in a conserved 10-amino acid region of L10e (sRb1-Q137P, sRb2-T143I, sRb5-Q137K, sRb13-T143A, and sRb14-S134⌬). All amino acid substitutions result from single nucleotide changes, with the exception of S134⌬, which results from the deletion of three nucleotides. The amino acid sequence of L10e from the wild-type parental strain YEFD12h contains the conservative substitutions I4V, V27I, and T281S that differ from Gen-Bank TM accession number M26506.
imately in the middle of the L10e polypeptide and suggest that this region may be important in conformational flexibility during translocation.
In our initial study, we emphasized the qualitative similarity between the action of sordarin and fusidic acid. Sordarin stabilizes the eEF2-ribosome-GDP complex in a manner similar to that of fusidic acid (Ref. 3 and this work). Quite compellingly, there is some cross-resistance between sordarin and fusidic acid resistance in eEF2 alleles; furthermore, sordarin-resistant mutations cluster in a pattern similar to that demonstrated for fusidic acid resistance alleles in S. typhimurium EF-G (28). The present data demonstrate the quantitative differences in the effects of sordarin and fusidic acid and suggest that upon closer analysis, their precise modes of action will most certainly differ.
Our results have implications for the physical and functional relationship between L10e and eEF2. Others have shown that L10 is exposed on the surface of 70S ribosomes in E. coli (29), and chemical cross-links have been observed between rat L10 and eEF2. Recently, a three-dimensional cryoelectron microscopy map of the ribosome-EF-G-GDP-fusidic acid complex has been obtained (11). In this structure, EF-G makes contacts with both the 30S and 50S subunits. Moreover, domains I and V of EF-G make contacts with the base of the L7/L12 stalk complex. Many reports have implicated the stalk complex in factor binding and translocation, and conformational changes in the stalk and EF-G are thought to drive the translocation process (6,16,29). The highly mobile nature of the stalk is well established, and a recent report proposes that conformational changes in the amino-terminal domains of L7/L12 are coupled with polypeptide synthesis (30). In the cryoelectron microscopy map of the ribosome-EF-G-GDP-fusidic acid complex, the stalk appears as a well-defined structure, presumably frozen in a stable conformation by fusidic acid. Sordarin and fusidic both inhibit translocation by stabilizing the ribosome-eEF2-GDP complex, although the precise details of this stabilization undoubtedly differ. Our results showed that whereas L10e mutant ribosomes are not detectably altered in their ability to bind [ 3 H]L-793,422, the ability of sordarin to stabilize the eEF2ribosome-nucleotide complex in these mutants is impaired. Similar results were obtained with some of the partially resistant eEF2 mutants. Many of these mutations in eEF2 overlap with regions in EF-G that give rise to fusidic acid resistance (3,28). Interestingly, these regions in EF-G contact the ribosome in the cryoelectron microscopy map, including those that contact the base of the stalk complex. Our genetic and biochemical data, taken together with previous results from L10e crosslinking experiments and the cryoelectron microscopy map of the ribosome-EF-G-GDP-fusidic acid complex, support a model in which L10e makes direct contact with eEF2 and also plays a role in conformational changes during translocation. The L10e mutants described here provide new opportunities to examine the functional relationships between L10e and eEF2 in S. cerevisiae.
FIG . 5. A, mutant ribosomes show reduced sordarin-conferred stabilization of the eEF2-nucleotide complex. Binding of [guanosine-8-3 H]GTP was performed with 0.92 pmol of wild-type eEF2 and 3 pmol of salt-washed ribosomes from wild-type (F) or sRB5 (E) on nitrocellulose filters as described under "Experimental Procedures." B, mutant ribosomes show reduced stimulation of eEF2 GTPase activity. A time course of GDP release is shown as the number of moles released per mole of eEF2 for wild-type ribosomes (circles) and sRb1 ribosomes (squares) in the absence (open symbols) and presence (filled symbols) of 1 g/ml sordarin. The value without eEF2 (reaching approximately 0.2 mol at 10 min for both wild-type and sRb1 ribosomes) was subtracted in each case. GTPase activity for wild-type ribosomes in the presence of 2 mM fusidic acid is shown (ϫ). Data shown are the averaged results of a single experiment performed in duplicate reactions with duplicate samples.