ABSTRACT
The ribosomal stalk protein P0 is involved in the susceptibility to the antifungal sordarin derivatives, as reported for a number of Saccharomyces cerevisiae resistant mutants. Mammals and some lower eukaryotes are naturally resistant to these compounds. It is shown here that expression in S. cerevisiae of the ribosomal protein P0 from Homo sapiens and from other sordarin-resistant organisms results in a decrease in the sensitivity of the cells to an agent of this class. To further characterize the P0 region responsible for inducing sordarin resistance, a series of protein chimeras containing complementary regions of the human and yeast P0 proteins were constructed and expressed in yeast. The chimeras complement the absence of the native yeast P0 except in chimeras containing the human P0 carboxyl-terminal domain. Resistance to sordarins was found to be associated with the presence of an HsP0 amino acid sequence comprising P118 to F138, which unexpectedly led to higher resistance than the presence of the complete human P0. A comparison of the corresponding region in P0 from yeast and sordarin-insensitive organisms, followed by site-directed mutagenesis, indicates that residues in positions 119, 124, and 126 have an important role in determining resistance to sordarins. Moreover, since sordarins block the eukaryotic elongation factor 2 (EF2) function, the P0 region affecting sordarin susceptibility must correspond to EF2-interacting domains of the ribosomal stalk protein, which affects the drug-binding site in the elongation factor.
Antibiotics, in addition to their primary interest as therapeutic agents, have been very useful as tools for investigating their target cellular processes. One of the most fruitful examples is probably translation. From very early on, work on the mode of action of protein synthesis inhibitors, such as streptomycin, chloramphenicol, erythromycin, etc., has been closely bound to the investigation into the bacterial translation machinery components, notably, ribosome structure and function (8, 22).
Our understanding of the eukaryotic ribosome falls considerably behind present knowledge of the prokaryotic particle and, although many findings can probably be extended to all kingdoms, there are a number of functional and structural features specific to eukaryotic systems. One of them is the ribosomal stalk, which has evolved into a much more complex structure suited to performing additional functional roles (1).
A series of new compounds derived from sordarin were described not long ago as potentially useful antifungals. The sordarin derivatives apparently block the interaction of the elongation factor in the ribosome (2, 11) in a way similar to but distinguishable from fusidic acid (4). In agreement with this mode of action, a drug-binding site has been recently located in a yeast eukaryotic elongation factor 2 (EF2)-sordarin crystal structure (10) in a previously proposed position based on the mapping of sordarin resistance mutations in the elongation factor modeled structure (2).
The ribosome, and more specifically the ribosomal stalk, also has a role in determining the drug-inhibitory action, since mutations in protein P0, one of the ribosomal stalk components, also induce resistance to sordarins in S. cerevisiae (9, 12). Moreover, the affinity of sordarin for the free EF2 is much lower than for the ribosome-bound factor (6), confirming a direct or indirect effect of the ribosome in the drug-binding site. In addition, the stalk proteins P1α and P2β affect the sordarin sensitivity of the yeast cells (9).
One of the most interesting peculiarities of these compounds is their exceptional specificity. They only inhibit translation in some lower eukaryotes, including different yeast and fungal species, but not in mammals (5); this specificity is obviously the basis for their antifungal activity. It is, indeed, surprising that the sordarin derivatives are able to differentiate the translocation step in different eukaryotes in such a specific way by interacting with highly conserved components, such as EF2 and the ribosomal stalk P0 protein. Both EF2 and the ribosomal stalk probably contribute to the specificity of the sordarin action.
It has been generally assumed that stalk P proteins interact with the elongation factor through their C-terminal peptide, EESDDDMGFGLFD, which is identical in all of them and forms the exposed stalk tip. Recently, this interaction was experimentally proved by the two-hybrid method (13). However, since the stalk protein C end is identical in yeast and humans, it is highly improbable that this interaction is the basis of the differential effects of the drug. There must be other stalk-EF2 interactions contributing to the specificity of sordarin activity. Thus, protein P0 mutations inducing resistance to sordarins in yeast have been mapped at around position 140 in the 312-amino-acid P0 (9, 12). These data suggest that this part of P0 can be involved in interactions with the elongation factors and might correspond to some of the regions in the base of the stalk that have been shown to be in contact with the bound EF2 in cryoelectron microscopy reconstructed images (20).
Determination of whether protein P0 is implicated in the sordarin resistance of human cells and, if so, identification of the responsible protein domain would provide information for a better understanding of the mechanism responsible for the specificity of sordarin antifungals and also for a more precise characterization of the stalk-EF2 interactions. This information may be useful for designing new more efficient and specific antifungals.
MATERIALS AND METHODS
Yeast strains and growth media. S. cerevisiae W303-1b (MATα leu2-3,112 ura3-1 trp1-1 his3-11,15 ade2-1 can1-100) was used as a wild-type control strain. S. cerevisiae W303dGP0 (MATα leu2-3,112 ura3-1 trp1-1 his3-11,15 ade2-1 can1-100 RPP0::URA3-GAL1-RPP0) was derived from W303-1b as previously described (18). Yeast strains were grown at 30°C in YEP medium (1% yeast extract and 2% peptone) with either 2% glucose or 2% galactose as a carbon source.
Estimation of growth inhibition.Cells were grown at 30°C in multiwell plastic plates containing 100 μl of YEP-glucose medium per well and the indicated concentration of inhibitor. Plates were placed in closed boxes over wet filter paper pads to avoid dryness. After incubation from 24 h to 72 h, depending on the strain growth rate, plates were shaken, and the A595 was estimated in a microplate automatic reader (Bio-Rad model 550). The optical density of cells growing exponentially in the absence of inhibitor, which was in the linear range of the reader, was considered as 100% growth.
Isolation and analysis of ribosomes.Cells grown to an A600 of 1.0 were resuspended in ice-cold buffer 1 (10 mM Tris-HCl [pH 7.4], 20 mM KCl, 12.5 mM MgCl2, 5 mM β-mercaptoethanol) containing a protease inhibitor cocktail (aprotinin, leupeptin, pepstatin, and phenylmethylsulfonyl fluoride at a 10-μg/ml final concentration) and broken by shaking them with glass beads in a FastPrep FP120 (Bio 101/Savant) at 4°C in the cold room. The S30 fraction was obtained by centrifuging the extract at 13,000 rpm for 20 min at 4°C in a Sorvall SS-30 rotor. Ribosomes were prepared from the S30 fraction by centrifugation at 4°C in a TL100.3 rotor for 90 min at 90,000 rpm at 4°C. The particles were washed by centrifugation under the same conditions in 20 mM Tris-HCl (pH 7.4), 500 mM ammonium acetate, 100 mM MgCl2, and 5 mM β-mercaptoethanol. Ribosomes were finally stored at −70°C in buffer 1. Ribosomal proteins were analyzed by either sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) or isoelectrofocusing in a 2.0 to 5.0 pH range as indicated in previous reports (24). Western blots were performed by using Immobilon-P membranes (21). The specific antibodies to stalk proteins used in the present study were described in previous publications (17, 23).
Enzymes and reagents.Restriction endonucleases were purchased from Roche, MBI Fermentas, New England Biolabs, and Amersham and were used as recommended by the suppliers. T4 DNA ligase, calf intestinal alkaline phosphatase, and the DNA polymerase I Klenow fragment were from Roche. DNA manipulations were performed basically as described previously (16). PCR was carried out by using Pfu DNA polymerase (Stratagene) and custom-made oligonucleotides from Isogen, according to the recommendations of Dieffenbach and Dveksler (3).
Sordarin derivative GM193663 (2) was kindly provided by GlaxoSmithKline (Tres Cantos, Madrid, Spain).
Plasmids.pFL37-HsP0. Plasmid pBSP0, containing the complete S. cerevisiae RPP0 gene was previously constructed (18). Plasmid pBSP0(Nd) derives from pBSP0 by mutagenesis of the region immediately upstream of the initiator ATG to generate an NdeI site (18). A fragment containing the coding region of human RPP0 was amplified with Pfu DNA polymerase, with plasmid pT7P0 as a template (14) and custom-made oligonucleotides (Table 1), creating an NdeI site at the initiator ATG and an NheI site at the stop codon. This fragment was cloned between the NdeI and NheI sites of plasmid pBSP0(Nd), generating plasmid pBSHsP0. The EcoRI-XhoI fragment encoding the human P0 under the control of the 5′ and 3′ regions of the yeast P0 was then cloned between the EcoRI-SalI sites of pFL37 (15), thus yielding plasmid pFL37-HsP0. A similar strategy was used to obtain pFL37-ScP0 (15).
Oligonucleotides used to prepare the different constructs
pFL37-HsP0P1 and pFL37-HsP0P2.First, human RPP1 and RPP2 coding regions were amplified from plasmids pT7P1 and pT7P2 (14) generating new NdeI sites in the ATG and stop codons with oligonucleotides in Table 1. The 5′- and 3′-flanking regions of yeast RPP1A gene were then amplified from BS-L47 (24) with custom oligonucleotides (Table 1) and the universal and reverse primers to create NdeI sites at the initiator and stop codons. Both fragments were afterward subcloned at the appropriate restriction sites in Bluescript originating pBS5.3YP1α. The NdeI site joining the two yeast P1α flanking fragments in pBS5.3YP1α was used to clone the human RPP1 and RPP2 genes. After digestion with EcoRI and BamHI and treatment with Klenow, these constructions were cloned in the EcoRI (Klenow-filled) site of pFL37-HsP0, thus generating plasmids pFL37-HsP0P1 and pFL37-HsP0P2.
Protein P0 chimeras.Different plasmids expressing human-yeast RPP0 chimeras were created (Fig. 1). The amino, central, and carboxyl regions include residues 1 to 136, 137 to 203, and 204 to 312, respectively (yeast P0 numbering). All of them are derived from pBSP0(Nd) (here named pBSP0YYY) and pBSHsP0 (here pBSP0HHH). Two strategies were followed. Plasmids pBSP0HYY, pBSP0YHH, pBSP055hYYY, pBSP098hYYY, pBSP0Yh60YY, pBSP0Yh40YY, and pBSP0Yh20YY were obtained by overlapping PCR (3) with custom made oligonucleotides (Table 1) and universal and reverse primers. Plasmids pBSP0YYH, pBSP0HHY, pBSP0HYH, and pBSP0YHY were prepared by ligation of the corresponding EcoRV fragments from human and yeast RPP0 genes, taking advantage of an EcoRV site at an equivalent position in both genes. The XhoI-EcoRI inserts of these constructions were subsequently cloned between the SalI-EcoRI sites of pFL37, yielding the respective pFLP0 plasmid series, ready to transform yeast strains. DNA sequencing confirmed all of the constructs.
Scheme of the protein P0 chimeras containing different fragments of the human and yeast proteins. Numbers (except 1) indicate the position of the carboxyl end of each fragment from S. cerevisiae P0 (□) and H. sapiens P0 (▪) in each chimera. (A) Whole P0 protein chimeras; (B) amino-terminal domain chimeras.
Mutation of C119, P124, and R126 in the P0Yh20YY chimera.Site-directed mutagenesis by overlapping PCR (3) was used to perform the changes C119→E, P124→R, and Q126→V in the chimera P0Yh20YY. In a first step, two overlapping DNA fragments containing the desired positions were amplified, with the pFL37-Yh20YY plasmid as a template. The complementary oligonucleotides P0Hs3M1-fixed and P0Hs3M2 were used as mutagenic primers at one end, and either oligonucleotides P0atgNde or P0taaXho at the other end (Table 1). These fragments were mixed and PCR amplified with the oligonucleotides P0atgNde and P0taaXho as primers to obtain the complete coding region (0.95 kb) with the desired nucleotide changes (P0Yh20***YY). This DNA fragment was digested with NdeI and EcoRV and subcloned in pBSP0(Nd) previously digested with the same enzymes, yielding pBS-P0Yh20***YY. This plasmid was then digested with EcoRI and NotI, and the 2.5-kb DNA fragment containing the coding region of the mutated chimera flanked by the S. cerevisiae RPP0 5′ and 3′ regulatory regions was cloned in pFL37-P0 digested with the same restriction enzymes. The resulting pFL37-P0Yh20***YY plasmid, carrying the three mutations simultaneously, was sequenced to confirm the presence of the desired mutations.
Genetic manipulations and recombinant DNA techniques were carried out according to standard methods as previously described (15).
RESULTS
Expression of H. sapiens P0 in S. cerevisiae W303dGP0.Plasmid pFL37-HsP0, containing the H. sapiens RPP0 gene encoding ribosomal protein HsP0, was used to transform the conditional P0 null mutant strains S. cerevisiae W303dGP0, derived from wild-type W303-1b (18), yielding the transformed strain W303dGP0-HsP0. As a control, the strain was also transformed with pFL37-ScP0, which contained the yeast ScP0 protein. In strain W303dGP0 the chromosomal RPP0 gene is under the control of the GAL1 promoter and, therefore, the cells depend on the plasmid-encoded P0 protein to grow in glucose. W303dGP0-HsP0 ribosomes were resolved by SDS-PAGE, and proteins were detected by Western blotting with monoclonal antibody 3BH5, specific for the conserved C-terminal peptide of all of the stalk components. The results showed a reduction in the amount of 12-kDa acidic proteins when the cells were grown in glucose to express HsP0 (Fig. 2A). Isoelectrofocusing of these proteins indicated that P2α and P1β were preferentially reduced (Fig. 2B).
Stalk proteins in ribosomes from cells expressing human protein P0. Ribosomes (30 μg [A] and 300 μg [B]) from yeast strain W303dGP0 transformed with either pFL37-HsP0 (HsP0) or pFL37-ScP0 (ScP0) grown in YEP-glucose were resolved by SDS-PAGE (A) and isoelectrofocusing in a 2.5 (bottom) to 5.0 (top) pH range (B). Proteins were detected by Western blotting with monoclonal 3BH5 in panel A and by silver staining in panel B. The position of the different components is marked. In panel B, the lower bands correspond to the phosphorylated form of each protein.
W303dGP0-HsP0 grows in the presence of glucose with a doubling time of 170 min, which is about twofold higher than the doubling time of W303dGP0-ScP0, the strain expressing ScP0.
The effect of HsP0 expression on the sensitivity of the cells to the sordarin derivative GM193663 was tested in liquid medium. In addition to the control, W303dGP0-ScP0, strains expressing P0 proteins from Dictyostelium discoideum, Rattus norvegicus, and Aspergillus fumigatus, obtained previously (15), were also included in the test as controls (Fig. 3). A clear increase in the resistance of all of the strains expressing heterologous proteins was found, which is higher in the case of cells expressing fungal proteins in agreement with previous reports (9, 19).
Growth inhibition by increasing concentrations of sordarin derivative GM193663 of S. cerevisiae W303dGP0 transformed with a plasmid-expressing protein P0 from H. sapiens (•), S. cerevisiae (▪), D. discoideum (□), R. norvegicus (▵), and A. fumigatus (○). Inhibition was estimated as indicated in Materials and Methods.
Effect of H. sapiens acidic P1 and P2 proteins on sordarin activity.To test whether the presence of the human 12-kDa stalk acidic proteins HsP1 and HsP2 have an effect on the HsP0-induced sordarin resistance in S. cerevisiae, strain W303dGP0 was transformed with plasmids simultaneously containing the human RPP0 gene and either human RPP1 or human RPP2. In this way, strains W303dGP0-HsP0/P1 and W303dGP0-HsP0/P2 were obtained, which showed doubling times in YEP-glucose of 170 and 175 min, respectively, indicating that coexpression of a human acidic protein and HsP0 did not have a significant effect on the growth rate in glucose of W303dGP0-HsP0.
The presence of the heterologous acidic proteins in the ribosomes from the transformed strains grown in glucose medium was confirmed by Western analysis of SDS-PAGE gels. In W303dGP0-derived strains, the total amount of acidic proteins remains lower than in the controls, even when a human acidic protein was expressed (data not shown).
The resistance to sordarin of the different strains also expressing one human acidic protein was not significantly different from the resistance found in W303dGP0-HsP0 (data not shown).
Mapping of the H. sapiens P0 domain conferring resistance to S. cerevisiae.The ScP0 and HsP0 proteins have a considerable overall sequence similarity, although they also contain domains with low homology (Fig. 4). To broadly define which region in HsP0 is involved in conferring resistance to sordarins, a series of protein chimeras were constructed in which the amino, central, and carboxyl regions were exchanged between HsP0 and ScP0 (Fig. 1A and Fig. 4). Plasmids containing the corresponding chimerical genes were used to transform S. cerevisiae W303dGP0.
Comparison of amino acid sequence of protein P0 from H. sapiens (H.s.) and S. cerevisiae (S.c.) by using the GAP program from GCG (Wisconsin University Biotechnology Center). Solid arrowheads mark the limits of the amino, central, and carboxyl fragments included in the protein chimera series described in Fig. 1A. Open arrowheads, followed by either a number or a “C” mark the amino terminus or the common carboxyl terminus, respectively, of the HsP0 fragments in the P0YhnYY series (Fig. 1B). Similarly, open arrows mark the carboxyl ends (numbers) and the common amino end (“A”) of the fragments in the P0hnYYY series (Fig. 1B).
The different transformed strains were tested for growth in glucose. The P0 chimeras were able to complement the absence of the native ScP0, except those containing the carboxyl domain of the human P0. These exceptions were unexpected, since W303dGP0 expressing the complete HsP0 was able to grow in glucose. Since the promoter and flanking regions in all of the constructs are identical, it is improbable that these noncomplementing proteins are not expressed. These proteins might be either inactive or degraded due to a wrong folding. In general, the complementing chimeras have a rather small negative effect on cell growth (Table 2).
Effect of P0 chimeras on S. cerevisiae W303dGP0 growtha
The transformed strains were tested for sensitivity to the sordarin derivative GM193663 in glucose liquid medium. As shown in Fig. 5, the strains expressing P0 chimeras that contain the amino-terminal part of the human protein, HHY and HYY, are approximately 1 order of magnitude more resistant than either the control W303dGP0-ScP0 or the strain expressing the YHY chimera.
Growth inhibition of the different S. cerevisiae W303dGP0 glucose-viable strains expressing P0 chimeras with increasing concentrations of sordarin derivative GM193663 as estimated as in Fig. 2. Cells expressing chimera are indicated as follows: HYY (□), HHY (▵), and YHY (○). The sordarin IC50s in cells expressing the wild-type proteins from H. sapiens HHH (•) and S. cerevisiae (▪) are included as a reference.
To define more accurately the region involved in sordarin susceptibility, different fragments of the HsP0 amino-terminal region were used to replace the equivalent fragments in the yeast protein (Fig. 1B and Fig. 4). Two plasmids were constructed carrying a chimerical gene encoding a ScP0 containing either the first 55 (pFL37-P0h55YYY) or the first 98 (pFL37-P0h98YYY) amino acids from HsP0. In addition, plasmids pFL37-P0Yh60YY, pFL37-P0Yh40YY, and pFL37-P0Yh20YY encoding ScP0s containing 60, 40, and 20 internal amino acids, respectively, from HsP0 were also prepared. The five constructs were used to transform W303dGP0. The transformed strains did not show a significant alteration of the growth rate in glucose medium (Table 2). An initial test on agar plates indicated that strains transformed with the ScP0 containing amino-terminal HsP0 amino acids (pFL37-P0h55YYY and pFL37-P0h98YYY) showed susceptibilities similar to that of the wild-type ScP0 (data not shown) and, therefore, they were excluded from further analysis. The strains expressing proteins P0Yh60YY, P0Yh40YY, and P0Yh20YY were tested further in liquid medium and were found to display a resistant phenotype equivalent to that displayed by cells expressing the P0HYY chimera (Fig. 6). Consequently, the resistance is associated with the HsP0 20-amino-acid region present in P0Yh20YY.
Sordarin sensitivity of S. cerevisiae W303dGP0 strains expressing P0YhnYY chimeras, i.e., P0Yh60YY (▵), P0Yh40YY (○), and P0Yh20YY (□). W303dGP0-ScP0 (▪) was used as a control strain. Growth inhibition was estimated as in Fig. 2.
Mutations in the HsP0 resistance domain increase sordarin sensitivity.An amino acid sequence comparison of the sordarin resistance domain of P0 proteins from S. cerevisiae and the four resistant organisms tested thus far, namely, D. discoideum, A. fumigatus, R. norvegicus, and H. sapiens, indicates quite a few differences (Fig. 7). However, only four positions, 119, 121, 124, and 126 (based on the H. sapiens numbering) were consistently different in the sensitive yeast protein and in the proteins from the resistant species. In position 121 an isoleucine in yeasts corresponds to valine in the resistant proteins, which probably is structurally not very relevant. In contrast, the differences in the other three positions were not conserved and might have an important role in determining the capacity of the proteins to induce resistance to sordarins. To test this possibility, Cys119, Pro124, and Gln126 were mutated to Glu, Arg, and Val, respectively, which are the amino acids present at the equivalent positions in ScP0. The P0Yh20***YY chimera carrying the three mutations simultaneously was then expressed in S. cerevisiae W303dGP0, and the strain was tested for sordarin resistance in YEP-glucose medium containing increasing concentrations of sordarin. The 50% inhibitory concentration (IC50) of sordarin for the strain expressing ScP0 was 0.05 μg/ml, while the IC50 for the strain expressing the nonmutated chimera (P0Yh20YY) was 0.45 μg/ml. P0Yh20***YY, in contrast to the nonmutated chimera, had a very small effect on the sordarin sensitivity of the yeast cells, for which the IC50 was 0.09 μg/ml.
Comparison of the amino acid sequences of the protein P0 region involved in sordarin selectivity in S. cerevisiae (S.c.), H. sapiens (H.s.), R. norvegicus (R.n.), A. fumigatus (A.f.), and D. discoideum (D.d.). Differences from yeast that are conserved among sordarin-resistant organisms are indicated in boldface.
DISCUSSION
The human ribosomal stalk HsP0 protein can be incorporated in the S. cerevisiae ribosome when expressed in a conditional null mutant strain, W303dGP0, grown under restrictive conditions. Ribosomes obtained from W303dGP0 cells expressing HsP0 contain this protein in roughly the same proportion as the ScP0 present in control yeast ribosomes. In contrast, the particles have a reduced amount of acidic 12-kDa yeast proteins. In the ribosome, mainly proteins P1α and P2β are present and bind the heterologous HsP0 with higher affinity than P1β and P2α, suggesting that they are functionally closer to the human P1 and P2. Confirming previous reports (7, 15), the results indicate that evolution has affected less the interaction of P0 with rRNA than with the other stalk proteins. The decrease in the 12-kDa protein content of the ribosome may be, at least in part, responsible for the reduction in the transformed strain growth rate. However, the additional expression of a human acidic protein, either HsP1 or HsP2, does not substantially improve the complementing capacity of HsP0.
The presence of the HsP0 in the W303dGP0 ribosome results in an appreciable reduction in the sensitivity to sordarin derivatives. A similar effect was detected by the expression of P0 from other mammals, such as rat (RnP0), whereas the proteins from two sordarin-insensitive fungi, such as D. discoideum (DdP0) and A. fumigatus (AfP0), induce greater resistance. The additional presence of human acidic proteins did not have any detectable effect on resistance levels induced by HsP0. These results confirm a direct participation of the stalk, and particularly of protein P0, in the susceptibility to sordarins previously reported for S. cerevisiae laboratory strains (9, 12).
The use of different P0 chimeras allows a sordarin resistance domain to be defined; this domain comprises residues 118 to 138 in the HsP0 amino acid sequence and is directly responsible for reducing sensitivity to the drug in S. cerevisiae. Mutations inside this region have confirmed the involvement of Cys119, Pro124, and Gln126 in the resistance mechanism. An equivalent region has also been identified as being responsible for the resistance to sordarins associated with protein P0 in A. fumigatus (19), a naturally resistant fungus causing serious nosocomial infections. This similarity indicates that the mechanism of resistance to sordarins is similar in mammals and fungi as far as the participation of P0 is concerned. Surprisingly, the yeast P0 containing this HsP0 fragment reduces the susceptibility to sordarins more than the complete human protein does. This greater reduction is probably related to conformational differences in this region when the proteins are assembled into the ribosomal stalk. However, it is difficult to definitely interpret the results without knowing the three-dimensional structure of the eukaryotic stalk.
Mutations have been mapped in S. cerevisiae P0, which also defines a region around positions 135 and 145 related to sordarin susceptibility (9, 12). At this time it is difficult to conclude whether this region and the one characterized here form a unique site or whether they are independent sites. As indicated previously, the three-dimensional structure of protein P0 has not been resolved in any eukaryotic organism. Consequently, it is difficult to draw a model for the specific role of each one of the mentioned residues in the mechanism responsible for the selectivity of these antifungal agents. However, it can be concluded that this region must be closely related, probably in direct contact, with EF2, the other sordarin sensitivity-determining component. Cryoelectron microscopic images of a S. cerevisiae 80S ribosome-EF2-sordarin complex indicate a direct interaction of the factor G domain with the stalk base, probably protein P0 (20). It is quite probable that the interaction site in P0 corresponds to the “sordarin sensitivity domain” defined previously.
Interestingly, the sordarin-binding site has been located in a pocket between domains III, IV, and V of EF2 in crystals of free factor from S. cerevisiae (10), a finding in agreement with most mapped resistance mutations (2, 11). This region is not involved in interactions of the factor with the ribosome (20). These results are, therefore, hardly compatible with a direct participation of the ribosomal stalk in sordarin binding to its target, and thus P0 must induce resistance through allosteric effects. Interestingly, the P0 resistant mutations reduce the sordarin-binding capacity of the EF2-ribosome complex from 20 to 50% (9). It is uncertain whether this reduction can account for the high-resistance phenotype of the mutant strains; however, a long-distance effect of P0 mutations on the drug-binding site is evident. In this sense, it must be taken into account that the affinity of EF2 for sordarins increases dramatically upon binding to the ribosome, implying important changes in the drug-binding site, which can be at least partially induced by interaction with protein P0. In agreement with this proposal, a couple of EF2 resistance mutations have been mapped in R180 and V187 at the G domain (11) in the stalk interaction site. All of these mutations can affect the EF2 conformational change induced by P0 interaction in the drug-binding site affecting the affinity and thus the activity of sordarins.
ACKNOWLEDGMENTS
We thank M. C. Fernández Moyano for expert technical assistance. We are grateful to J. F. García Bustos and M. Gómez Lorenzo for their cooperation and useful discussions. We are grateful to B. E. Rich for the human genes RPP0,RPP1, and RPP2.
This study was supported by grant BMC2003-03387 from the Ministerio de Ciencia y Tecnología (Spain), by grant EU QLRT-2001-00892, by a research contract with the GlaxoSmithKline Research Center (Tres Cantos, Madrid, Spain), and by an institutional grant to the Centro de Biología Molecular from the Fundación Ramón Areces (Madrid, Spain).
FOOTNOTES
- Received 23 February 2004.
- Returned for modification 1 March 2004.
- Accepted 14 March 2004.
- Copyright © 2004 American Society for Microbiology
