A role for the Ppz Ser/Thr protein phosphatases in the regulation of translation elongation factor 1Balpha.

In vivo 32P-labeled yeast proteins from wild type and ppz1 ppz2 phosphatase mutants were resolved by bidimensional electrophoresis. A prominent phosphoprotein, which in ppz mutants showed a marked shift to acidic regions, was identified by mixed peptide sequencing as the translation elongation factor 1Balpha (formerly eEF1beta). An equivalent shift was detected in cells overexpressing HAL3, a inhibitory regulatory subunit of Ppz1. Subsequent analysis identified the conserved Ser-86 as the in vivo phosphorylatable residue and showed that its phosphorylation was increased in ppz cells. Pull-down experiments using a glutathione S-transferase (GST)-EF1Balpha fusion version allowed to identify Ppz1 as an in vivo interacting protein. Cells lacking Ppz display a higher tolerance to known translation inhibitors, such as hygromycin and paromomycin, and enhanced readthrough at all three nonsense codons, suggesting that translational fidelity might be affected. Overexpression of a GST-EF1Balpha fusion counteracted the growth defect associated to high levels of Ppz1 and this effect was essentially lost when the phosphorylatable Ser-86 is replaced by Ala. Therefore, the Ppz phosphatases appear to regulate the phosphorylation state of EF1Balpha in yeast, and this may result in modification of the translational accuracy.

the ribosomal A site in polypeptide chain elongation and participates in proofreading of the codon-anticodon match (1) In the budding yeast Saccharomyces cerevisiae, EF1 consist of different subunits: EF1A (formerly EF1␣) is encoded by two different genes (TEF1 and TEF2), and it binds aminoacyl-tRNA in a GTP-dependent manner. The exchange of GDP for GTP on EF1A is stimulated by a member of the guanine nucleotide exchange factor family, EF1B␣, which is encoded by a single gene (TEF5). An additional subunit of uncertain function is encoded by genes TEF3 and TEF4. Although lack of TEF3 and TEF4 results in no observable defects in translation (2), lack of TEF5 or simultaneous deletion of TEF1 and TEF2 is lethal (3,4). Components of EF1 have been shown to be phosphorylated in vitro by diverse protein kinases in species different from yeast (5)(6)(7).
The function of EF1B␣ on EF1A has been shown to be critical for an efficient and accurate translation. For example, cells with increased expression of the EF1A subunit can bypass the lethality of cells lacking EF1B␣. However, these cells present a number of defects, including higher sensitivity to inhibitors of translation elongation and changes in translational fidelity (4). Alterations in translational fidelity have been also produced by specific mutations in EF1B␣, as it has been documented by evaluation of sensitivity to drugs such as paromomycin and analysis of translational fidelity at nonsense codons (8). The fidelity of translation may be related, at least in part, to the requirement for nucleotide exchange, as it has been tested by mutations in the GTP-binding motif of yeast EF1A (9).
The yeast Ppz phosphatases are encoded by genes PPZ1 and PPZ2 (10,11) and represent a novel type of Ser/Thr phosphatases characterized by a catalytic carboxyl-terminal half related to type 1 phosphatase. These phosphatases are involved in a variety of cell processes, including maintenance of cell integrity, in connection with the Pkc1/Mpk1 mitogen-activated protein kinase pathway (11,12), regulation of salt tolerance (13), and regulation of cell cycle at the G 1 /S transition (14). In all cases, the function of Ppz1 appears to be more important than that of Ppz2. Recently, we have identified the halotolerant determinant Hal3 as a negative regulatory subunit of Ppz1 that modulates the diverse physiological functions of the phosphatase (15).
As an attempt to better understand the physiological role of the Ppz phosphatases, we have performed a two-dimensional electrophoretic analysis of proteins from in vivo 32 P-labeled wild type and ppz strains, in search for polypeptides that might display an altered phosphorylation state in the absence of the phosphatases. This approach has led us to establish a previously unsuspected link between the Ppz phosphatases and the translation elongation factor 1B␣.
Recombinant DNA Techniques and Plasmid Construction-E. coli cells were transformed using the standard calcium chloride method (18). S. cerevisiae cells were transformed by a modification of the lithium acetate method (19), which includes treatment with dimethyl sulfoxide. Restriction digestions, DNA ligations, and other standard recombinant DNA techniques were performed essentially as described previously (18).
A vector for high copy expression of EF1B␣ as a GST fusion protein was constructed as follows. The entire EF1B␣ ORF was amplified by PCR with oligonucleotides EFb5Ј (5Ј-CGGGATCCCCATGGCATCCAC-CGATTTCTC-3Ј) and EFb3Ј (5Ј-GCGGATCCTTATAATTTTTGCATAG-CAG-3Ј). The underlined residues denote the engineered BamHI site used to clone the EF1B␣ ORF in-frame with the GST sequence of plasmid p426TEG2, to yield p426/TEF5. Plasmid p426TEG2 is based on the high copy number yeast vector pRS426 (20) and allows expression of GST fusion proteins under the control of the TEF1 promoter. For low level expression of EF1B␣, the 1.6-kbp GST-EF1B␣ fusion sequence was excised from p426/TEF5 with KpnI and SacI and cloned in the centromeric plasmid pRS416. The promoter region of TEF5 was amplified by PCR with oligonucleotides Efbprom_5Ј (5Ј-CGC GAG CTC AAT ACC GAC AGC TTT TGA C-3Ј) and Efbprom_3Ј (5Ј-CGC GAG CTC CAT TAT GTG TGT ATA TAT TCG-3Ј). The artificial SacI site (underlined) was used to clone the promoter downstream of the hybrid ORF to yield pRS/TEF5. Plasmid DNA was recovered and sequenced to identify clones with the correct orientation and lacking unwanted mutations.
Mutation of the phosphorylatable Ser-86 to Ala was made by sequential PCR. In a first step, the EF1B␣ ORF was amplified in two separate reactions by using the pair of primers EFb5Ј/EF1bSA3 (5Ј-CTTCT-TCATCGTCGGCACCGAATAAATCG-3Ј), and EFb3Ј/EF1bSA5 (5Ј-CGATTTATTCGGTGCCGACGATGAAGAAG-3Ј). The residues in boldface denote the modification introduced to change Ser-86 to Ala. In a second step, the 637-and 370-bp fragments produced in these PCR reactions were mixed, and the entire EF1B␣ ORF amplified by using oligonucleotides EFb5Ј and EFb3Ј. The amplification fragment was cloned into the BamHI site of p426TEG2, to yield p426/TEF5(S86A), and sequenced.
In Vivo 32 P-Labeling, Two-dimensional Electrophoresis, and Phosphopeptide Analysis-Yeast cells were 32 P-labeled as follows. 50-ml cultures were grown in YPD (or SD medium lacking leucine or uracil, when containing plasmids) until an OD 660 of 0.5-0.8 was reached. Cells were collected by centrifugation, washed with 10 ml of low phosphate YPD medium (22), and finally resuspended in the same medium to achieve an OD 660 of 0.2. Growth was resumed until an OD 660 0.8 was reached, and 500 Ci of carrier-free [ 32 P]orthophosphate was added. Incubation was continued for 60 min. Cells were then centrifuged, and pellets were resuspended in 10 ml of ice-cold trichloroacetic acid and incubated on ice for 10 min. After centrifugation, the mixture was vigorously resuspended in 1 ml of dimethylketone, centrifuged, and dried at room temperature. Pellets were then resuspended in 360 l of lysing buffer (0.1 M Tris-HCl, pH 8, 0.3% SDS, 2.5% ␤-mercaptoethanol) supplemented with protease and phosphatase inhibitors (150 M NaVO 4 , 1 M microcystine, 1 mM benzamidine, 1 mM PMSF), and homogenates were prepared with the aid of glass beads. Samples were boiled for 10 s then cooled on ice for 1 min, and 30 l of RNase A solution (200 units/ml de Rnase A in 0.5 M Tris-HCl pH 7, 50 mM MgCl 2 buffer) and 6 l of a 1 unit/l solution of DNase (Promega) were added. Samples were incubated for 2 min at 4°C, supplemented with 750 mg of urea and 200 l of a solution containing 4% CHAPS, 4.75 M urea, and 5% ␤-mercaptoethanol and further incubated for 10 min at room temperature. After centrifugation at 750 ϫ g for 10 min, supernatants were centrifuged at 13,000 rpm in a microcentrifuge and the second supernatants (50 -200 g of proteins) were then subjected to two-dimensional electrophoresis and transferred to membranes essentially as described earlier (23).
To identify proteins of interest, membranes were stained with Amido Black and subjected to autoradiography. Protein staining and phosphorylation patterns were compared using Melanie software (Bio-Rad). Relevant fragments were sliced, removed, digested for 90 min with 200 l of 500 mg/ml cyanogen bromide, as described (23), and subjected to amino acid sequencing in a Applied Biosystems Procise 494 apparatus. The mixed sequenced obtained were run against the FASTF data base to identify the protein (24).
To identify the nature of the phosphorylated residue(s), strain EDN75 (ppz1::KAN) was transformed with plasmid p426/TEF5, and cells were labeled with 32 P as above, recovered by centrifugation, and disrupted with glass beads in 25 mM Tris-HCl buffer (pH 7.5), containing 1 mM EDTA, 1 mM dithiothreitol, 100 mM NaCl, 0.5 mM PMSF, 0.5 mM benzamidine, 1 g/ml leupeptin, and 1 g/ml pepstatin. The homogenate was centrifuged at 100,000 ϫ g for 60 min at 4°C, and the supernatant was loaded into a 1-ml glutathione-Sepharose column equilibrated with the above mentioned buffer. The column was washed with the same buffer plus 250 mM NaCl, and the fusion protein was eluted with 10 mM glutathione. The purified GST-EF1B␣ was the only radioactive band when analyzed by SDS-PAGE. The protein (0.5 mg) was digested with endolysyl peptidase C (20 g/ml), which cuts carboxyl-terminal to lysine residues. Digests were acidified with TFA, and applied to a reverse phase column (Waters Nova-Pak C18, 3.9 ϫ 150 mm) that had been equilibrated in 0.1% TFA (buffer A). The flow rate was maintained at 1 ml/min. The column was washed for 10 min with buffer A before peptides were eluted with a linear gradient of acetonitrile (0 -80% in 80 min) in buffer A. Fractions (1 ml) were collected, and peptides containing 32 P were identified by measuring Cerenkov radiation. Those fractions were pooled and evaporated to dryness before the peptides were immobilized to Immobilon membranes (Millipore) following the manufacturer's instructions. The 32 P-labeled peptides generated from in vivo labeled EF1B␣ were identified with a vapor phase amino acid sequencer (Applied Biosystems Procise 494). Phosphorylated residues within phosphopeptides were located by determining the cycles in which 32 P was released when samples were subjected to sequential Edman degradation under conditions that optimize recovery of 32 P (25).
Immunodetection of Ppz1 Bound to Purified GST-EF1B␣ Fusion Protein-Wild type strain JA-100 and the ppz1 mutant strain EDN75 were transformed with the empty plasmid p426TEG2 and plasmid p426/TEF5. Cells were grown in synthetic medium lacking uracil up to an OD 660 of 2, recovered by centrifugation, and disrupted with the aid of glass beads in 50 mM Tris-HCl buffer (pH 8.0), containing 1 mM EDTA, 2 mM dithiothreitol, 150 mM NaCl, 0.5 mM PMSF, 0.5 mM benzamidine, 1 g/ml leupeptin and 1 g/ml pepstatin. A crude extract was prepared by centrifugation at 750 ϫ g, and 1 mg of protein was incubated with 100 l of glutathione-Sepharose beads for 90 min at 4°C. Beads were washed with the above mentioned buffer, resuspended in SDS-sample buffer, and boiled, and the supernatant was loaded in a 8% SDS-polyacrylamide gel. After transfer to Immobilon-P membranes (Millipore), the presence of Ppz1 was assessed by immunoblotting with anti-Ppz1 antibodies as described previously (15).
Nonsense Suppression Assays-Suppression of nonsense codons was estimated by using the pUKC series plasmids (26). Plasmid pUKC815 derives from YCp50 and expresses the LacZ gene from the yeast PGK promoter. Plasmids pUKC817, pUKC818, and pUKC819 are identical to pUKC815, but they carry the nonsense codons UAA, UAG, and UGA at the beginning of the ␤-galactosidase coding sequence. Yeast strains were transformed with these plasmids and grown up to an OD 660 of 0.5-1.0 in SD medium lacking uracil, and the ␤-galactosidase activity was determined as described previously (27).
Sensitivity to Drugs and Growth Assays-For testing the sensitivity to various drugs on plates, cultures were grown until an OD 660 of 2 and 100 l of the culture was spread on YPD plates. 20 l of a solution of each drug was placed on sterile disks on top of the plates, and growth resumed at 30°C for 2-3 days. Sensitivity in liquid medium was assessed as follows. Exponential cultures were diluted up to an OD 660 of 0.01. Aliquots of 200 l were supplemented with 100 l of medium containing the appropriate amount of the drug. Growth was resumed for 16 -18 h, and sensitivity to drugs was monitored by measuring the OD 660 of the culture.

Yeast EF1B␣ Is a Phosphoprotein in Vivo Whose Phosphorylation State Is Affected by Ppz
Phosphatases-To evaluate the influence of lack of Ppz phosphatases in the cellular phosphorylation pattern, wild type and ppz1 ppz2 yeast cells were 32 P-labeled and total cell extracts prepared and subjected to bidimensional electrophoresis. Global analysis of the distribution of phosphoproteins did not show a remarkable overall modification of the pattern. However, a clear shift to more acidic regions of a phosphoprotein of about 22.5 kDa and focusing at pI 4.3 was observed in the double mutant ppz1 ppz2 compared with the wild type yeast cells (Fig. 1). The acidic shift of this protein was also observed when the gel was silverstained, indicating that it was a relatively abundant component. Because shifting to more acidic regions is often associated to increased phosphate content, this result was considered as indicating that this protein could be a target for the Ppz phosphatases. To further test this possibility, a similar experiment was carried out using wild type cells that contained a high copy plasmid carrying the HAL3 gene, which codes for a negative regulator of Ppz1 (15). As shown in Fig. 1, overexpression of HAL3 resulted in a pattern identical to the one observed in ppz mutants, supporting the notion that the shift was the result of lack of Ppz activity.
The identity of the mentioned phosphoprotein was established by recovering the region of the membrane, followed by digestion with cyanogen bromide and automated Edman sequencing of the resulting peptide mixture. Analysis of yeast protein data banks indicated that this protein corresponded to the product of the single-copy, essential TEF5 gene, encoding the eukaryotic elongation factor 1B␣ (3). Therefore, yeast EF1B␣ is a phosphoprotein in vivo, and its phosphorylation state may be modified by the Ppz phosphatases.
To identify the residue(s) phosphorylated in vivo, a GST-EF1B␣ fusion protein was expressed from plasmid p426/TEF5 in 32 P-labeled yeast cells. The product was affinity purified by glutathione-Sepharose chromatography and digested with endolysyl peptidase C. Phosphopeptide mapping by HPLC analysis indicated that all in vivo phosphorylation sites are located in a single endolysyl peptidase C fragment (Fig. 2). Phosphoamino acid analysis of both the entire labeled fusion protein or the relevant chromatographic fractions revealed that all radioactivity was bound to Ser residue(s) (Fig. 2). Further analysis of the radioactive endolysyl peptidase C fragment by measuring recovery of 32 P during Edman sequencing proved that radioactivity was associated to a single residue that was determined to be Ser-86.
Affinity-purified GST-EF1B␣ from Yeast Cells Contains Bound Ppz1-The possibility that the PPZ phosphatases might interact in vivo with the elongation factor was approached by using an affinity system based in the expression of EF1B␣ in yeast as a GST fusion protein. This recombinant protein was affinity-purified, and the presence of accompanying proteins was evaluated by SDS-PAGE. When the presence of Ppz1 was tested by immunoblot in these samples (Fig. 3), it was found that affinity-purified EF1B␣ contained bound Ppz1. This finding indicates that Ppz1 and EF1B␣ can interact in vivo and leads to the possibility that Ppz1 could directly dephosphorylate EF1B␣. In fact, when the GST-EF1B␣ fusion protein was expressed in 32 P-labeled cells from the low copy plasmid pRS/ TEF5, an increase in radioactive phosphate content of about 2-fold was observed in ppz cells when compared with the wild type strain (data not shown). However, our attempts to in vitro dephosphorylate the in vivo 32 P-labeled translation factor using available bacterially expressed Ppz1 have been so far unsuccessful.
Evidence for Functional Interactions between Ppz Phosphatases and EF1B␣-The observation that the Ppz phosphatases may affect the in vivo phosphorylation state of EF1B␣ prompted us to analyze phenotypes related to changes in the function of this protein. It has been recently reported that mutations in the conserved carboxyl terminus of EF1B␣ alter the sensitivity of yeast cells to translation elongation inhibitors. We tested the sensitivity of wild type and phosphatasedeficient strains to paromomycin, hygromycin B, and cycloheximide by both liquid cultures and the halo assay. As shown in Fig. 4, deletion of both phosphatase genes clearly increased the tolerance of yeast cells to paromomycin and hygromycin, whereas tolerance was not modified in the case of cycloheximide. A rather strong effect was also observed when only the PPZ1 gene was absent (Fig. 4), whereas lack of PPZ2 resulted in minor changes (not shown).
Changes in sensitivity to paromomycin have been related to altered translational fidelity. Therefore, we sought to investigate whether the absence of Ppz phosphatase activity might affect this cellular function, by evaluating the suppressor capacity of the ppz mutant strains. To this end, we transformed  2. Phosphoamino acid analysis and phosphopeptide map  of yeast EF1B␣. A, EF1B␣ was expressed as a GST fusion protein using plasmid p426/TEF5, affinity-purified from 32 P-labeled ppz mutant cells and eluted with 10 mM glutathione. The sample was digested with endolysyl peptidase C that cuts carboxyl-terminal to Lys residues. Peptides were resolved by C18 reverse-phase HPLC and eluted with a gradient (dashed line) from 0 to 60% acetonitrile/0.1% TFA over 80 min. The radioactivity associated to the eluted peptides was monitored by Cerenkov counting. B, phosphoamino acid analysis. Lane 1, 32 P-labeled purified GST-EF1B␣; lane 2, HPLC-purified phosphopeptide F36 (fraction 36). Amino acid symbols denote the mobility of unlabeled phosphoamino acid standards. C, Amino acid sequence of the region surrounding Ser-86 in S. cerevisiae EF1B␣.
wild type and ppz1 cells with constructs that express the LacZ gene from the PGK1 promoter, carrying one of the UAA, UAG, or UGA stop codons at the beginning of the ␤-galactosidase coding region. Therefore, only when these codons are suppressed can the enzymatic activity be detected. As shown in Fig. 5, ppz1 mutants displayed, in all cases, a higher ␤-galac-tosidase activity than wild type cells (3-to 4-fold of increase), indicating a higher suppressor capacity. Experiments performed in parallel with ppz1 ppz2 mutants produced similar results (data not shown).
It has been reported that high levels of Ppz1 are detrimental for cell growth, because they lead to an expanded G 1 /S cell cycle transition. To know if this effect could be somehow related to the function of EF1B␣, we transformed wild type cells with a high copy plasmid containing the PPZ1 gene, and then we introduced either an empty p426TEG2 vector, the p426/TEF5 plasmid, or the same construct carrying a version of EF1B␣ in which the phosphorylatable Ser-86 had been changed to Ala. Although high levels of EF1B␣ did not modify the growth rate of a strain with normal levels of Ppz1 (data not shown), the presence of an excess of the factor results in a clear improvement in cell growth of cells overexpressing Ppz1 (Fig. 6). However, this effect was completely lost when the S86A version of EF1B␣ was expressed. Immunoblot experiments determined that the amount of wild type and mutated versions of EF1B␣ were virtually identical, exhausting the possibility of an artifact due to difference in expression levels. A further test was performed by transforming wild type cells with the mentioned constructs and determining paromomycin tolerance. Overexpression of the wild type version of EF1B␣ increased tolerance to the drug, and this effect was fully abolished when Ser-86 was replaced by Ala. These results provide evidence of functional changes as a result of the absence of the phosphorylation site in EF1B␣.

DISCUSSION
In this report we demonstrate that, in the yeast S. cerevisiae, translation elongation factor 1B␣ is a phosphoprotein. Phosphorylation site mapping and sequence analysis indicates that the Ser-86 is the only phosphorylatable residue in this protein, at least under standard growth conditions. It is remarkable that data base search reveals that the equivalent Ser residue (as well as its acidic environment) is also found in a large variety of organisms, including Drosophila melanogaster, Caenorhabditis elegans, mouse, and human. Phosphorylation of EF1B␣ has been reported in Artemia salina (5), wheat (6), and reticulocyte (7). In the former case, phosphorylation was ascribed to Ser-89, which is equivalent to Ser-86 in yeast EF1B␣. Interestingly, phosphorylation has been correlated to changes in its catalytic nucleotide exchange activity, although reports are somewhat contradictory (5, 7).
Our data indicate that deletion of the ppz genes and overexpression of Hal3, a negative regulatory subunit of Ppz1 (15), The fusion protein GST-EF1B␣ was expressed in wild type (PPZ1ϩ) or EDN75 cells (PPZ1Ϫ) and affinity-purified from 1 mg of total yeast extracts through a glutathione-Sepharose matrix. An additional control in which wild type cells were transformed with the empty p426 plasmid was also included. After extensive washing the samples were resuspended in sample buffer, electrophoresed in SDS-polyacrylamide gels, and transferred to membranes. The presence of interacting Ppz1 was tested by immunoblot using available anti-Ppz1 antibodies developed against recombinant GST-Ppz1 protein. Note that these antibodies also recognize the GST moiety of GST-EF1B␣. result in increased phosphorylation of the EF1B␣ protein, specifically at Ser-86. These results would be compatible with a role of Ppz1 in regulating the phosphorylation state of the translation factor and, possibly, its function. We also show here evidence that affinity-purified yeast EF1B␣ contains significant amounts of bound Ppz1, by using an approach that was pivotal in the past to identify the Hal3 protein as a subunit of Ppz1 (15). This could be taken as an indication that Ppz1 could be able to directly dephosphorylate EF1B␣. However, we have been unable to detect direct dephosphorylation of either in vivo labeled or CK-2 in vitro phosphorylated EF1B␣ in the presence of bacterially expressed Ppz1. Although at this point we cannot provide direct evidence for the translation factor being a substrate for the phosphatase, this possibility formally remains. For instance, the phosphatase might require accessory proteins (absent in our in vitro assay) to effectively use EF1B␣ as substrate. In this regard, there is a large body of evidence for the requirement of specific regulatory subunits (targeting subunits) for Ser/Thr phosphatases to localize at specific subcellular sites or to use a given phosphoprotein as an effective substrate (28,29). It must be noted that dephosphorylation events have been previously related to the control of the accuracy of protein synthesis, as it is the case of the Ppq1/Sal6 Ser/Thr protein phosphatase (30,31), the closest structural homologue of the Ppz phosphatases. However, the possible role of this phosphatase has not been worked out.
We considered that if EF1B␣ was a target (either direct or indirect) for Ppz1, it could be possible to establish some sort of functional connection between both proteins. Deletion of TEF5 is lethal for the cell, and high copy expression of TEF2 suppresses the lethal phenotype of tef5 mutants (4). However, these cells are markedly sensitive to translational inhibitors, such as paromomycin and hygromycin B. It is remarkable that lack of Ppz phosphatases also results in a change in sensitivity to these compounds, although in this case yielding more tolerant cells. Because these drugs are aminoglycosides known to enter the yeast cell driven by the membrane potential, which is mostly maintained by the function of the membrane H ϩ -ATPase (32), we considered the possibility that the increased tolerance could be an indirect effect due to altered proton efflux. However, this was ruled out by determining this parameter in wild type and ppz mutants and finding essentially identical values (data not shown).
Changes in sensitivity to paromomycin have been related to altered translational fidelity (33, 34), a phenotype also pro-duced by changes in the dosage of EF1A (35). Recent evidence (8) has been presented pointing out that mutations in the carboxyl-terminal region of EF1B␣ results in increased sensitivity to translation inhibitors and that this effect was accompanied by enhanced translational fidelity (i.e. reduced readthrough at nonsense codons). These observations are in keeping with our finding that cells lacking Ppz phosphatases, which are more tolerant to certain translation inhibitors, show an increased readthrough at nonsense codons, most likely due to a decrease in translational fidelity.
Further evidence for a functional interaction between EF1B␣ and Ppz1 comes from the observation that overexpression of the translation factor strongly attenuates the growth defect, due to a delayed G 1 /S transition, of cells containing an excess of Ppz1 activity. Although we showed in the past that this defect correlates with a delay in G 1 /S cyclin mRNA expression (14), immunoblot analysis of the protein level of different cyclins reveals that, at least in the case of Clb5, further post-transcriptional alterations (i.e. at the translation level) could exist. 2 Remarkably, a non-phosphorylatable version of EF1B␣ was unable to counteract the effect of an excess of Ppz1, suggesting that in vivo modulation of the phosphorylation state of the factor is somehow involved in the regulation of its function. It has been reported that, when expressed from the powerful GAL promoter, a carboxyl-terminal fragment of EF1B␣, lacking Ser-86, was sufficient for normal growth and did not display dramatically altered drug or temperature sensitivity (8). Furthermore, a strain containing a S86A version of EF1B␣ as the only source for the factor is viable. 3 Therefore, it must be concluded that regulation of EF1B␣ by phospho-dephosphorylation at Ser-86 (which, at least in part, would involve Ppz1) must affect the function of the translation factor in a subtle way. From our data, it can be hypothesized that changes in the phosphorylation state of EF1B␣ would result in altered nucleotide exchange on EF1A. However, alternative mechanisms cannot be excluded, because it has been postulated that EF1B␣ may have additional regulatory effects on EF1A (9). In any case, our data provides further support to the notion that phospho-dephosphorylation mechanisms are relevant for a proper regulation of protein synthesis.
FIG. 6. Overexpression of EF1B␣ attenuates the cell growth defect due to high levels of Ppz1. Wild type JA-100 cells carrying a high copy number plasmid with the PPZ1 gene were transformed with an empty p426 plasmid (q), plasmid p426/TEF5 (E), or p426/ TEF5(S86A) (). Positive clones were exponentially grown in SD medium lacking leucine and uracil and then diluted in the same medium up to an OD 660 of 0.02. Growth was resumed and monitored at different times. Data are mean Ϯ S.E. from at least eight independent clones.