Characterization of yeast translation initiation factor 1A and cloning of its essential gene.

Translation initiation factor eIF1A is required in vitro for maximal rates of protein synthesis in mammalian systems. It functions primarily by dissociating ribosomes and stabilizing 40 S preinitiation complexes. To better elucidate its precise role in promoting the translation initiation process, the yeast form of eIF1A has been identified in Saccharomyces cerevisiae and purified to homogeneity on the basis of its cross-reaction with antibodies prepared against mammalian eIF1A. The apparent mass of yeast eIF1A (22 kDa) resembles that of the mammalian homolog (20 kDa), and the yeast factor is active in stimulating methionyl-puromycin synthesis in an assay composed of mammalian components. The gene encoding yeast eIF1A, named TIF11, was cloned and shown to be single copy. TIF11 encodes a protein comprising 153 amino acids (17.4 kDa); the deduced amino acid sequence exhibits 65% identity with the sequence of human eIF1A. Both human and yeast eIF1A contain clusters of positive residues at the N terminus and negative residues at the C terminus. Deletion/disruption of TIF11 demonstrates that eIF1A is essential for cell growth. Expression of human eIF1A cDNA rescues the growth defect of TIF11-disrupted cells, indicating that the structure/function of yeast and mammalian eIF1A is highly conserved.

Translation initiation factor eIF1A is required in vitro for maximal rates of protein synthesis in mammalian systems. It functions primarily by dissociating ribosomes and stabilizing 40 S preinitiation complexes. To better elucidate its precise role in promoting the translation initiation process, the yeast form of eIF1A has been identified in Saccharomyces cerevisiae and purified to homogenity on the basis of its cross-reaction with antibodies prepared against mammalian eIF1A. The apparent mass of yeast eIF1A (22 kDa) resembles that of the mammalian homolog (20 kDa), and the yeast factor is active in stimulating methionyl-puromycin synthesis in an assay composed of mammalian components. The gene encoding yeast eIF1A, named TIF11, was cloned and shown to be single copy. TIF11 encodes a protein comprising 153 amino acids (17.4 kDa); the deduced amino acid sequence exhibits 65% identity with the sequence of human eIF1A. Both human and yeast eIF1A contain clusters of positive residues at the N terminus and negative residues at the C terminus. Deletion/disruption of TIF11 demonstrates that eIF1A is essential for cell growth. Expression of human eIF1A cDNA rescues the growth defect of TIF11-disrupted cells, indicating that the structure/function of yeast and mammalian eIF1A is highly conserved.
The initiation phase of protein synthesis in eukaryotic cells is promoted by a large number of proteins called initiation factors (eIF) 1 (for reviews, see Refs. 1 and 2). One of these, eIF1A (formerly eIF-4C), has been purified from both mammalian (3)(4)(5) and plant cells (6) and is essential for maximal in vitro protein synthesis. eIF1A is a small protein (17)(18)(19)(20)(21)(22) kDa) that appears to undergo no post-translational modification reactions (7). The initiation factor is implicated in 80 S ribosome dissociation, stabilizes initiator Met-tRNA i binding to 40 S ribosomal subunits, and facilitates mRNA binding to the 40 S preinitiation complex (2,8). Thus, eIF1A has pleiotropic effects at different steps of the initiation pathway. Purified eIF1A from wheat germ and rabbit reticulocytes functions interchangeably in vitro (6), suggesting that the functional domains are highly conserved. Nevertheless, a clear understanding of how eIF1A promotes the initiation phase of protein synthesis is lacking.
The process of translation initiation appears to be very conserved between eukaryotes as distantly related as mammals and the yeast, Saccharomyces cerevisiae. This conclusion is based on similarities of mRNA structure and the basic mechanism of protein synthesis (9). Particularly striking are the structural similarities between yeast and mammalian initiation factors, which share amino acid sequence identities ranging from 26 to 71%. This strong conservation has made it possible to recognize several yeast genes as encoding specific initiation factors based on their sequence homology to mammalian cDNAs: eIF2␣ (10), eIF2␤ (11), eIF2␥ (12), eIF2B⑀ (13), eIF4A (14), eIF4B (15,16), eIF4␣ (17), eIF4␥ (18), eIF5 (19), and eIF5A (20). In addition, there are numerous examples of yeast initiation factors functioning in mammalian assays and mammalian initiation factor cDNAs replacing the corresponding yeast genes. Therefore, studies on yeast initiation are expected to yield results that likely are applicable to the initiation process in all eukaryotic cells.
The recent cloning of the mammalian eIF1A cDNA has provided structural information about the factor (21). The failure to stimulate translation rates by overexpression of the cDNA in transiently transfected mammalian cells suggests that eIF1A is not limiting for protein synthesis (22). However, because of the intrinsic complexity of the mammalian system and the limited ability to manipulate specific gene expression, we decided to study eIF1A function in yeast. We report here the purification and biochemical characterization of yeast eIF1A, the cloning of its gene, TIF11 (for translation initiation factor 1A), and in vitro and in vivo characterization of the factor.

MATERIALS AND METHODS
Strains and Genetic Manipulations-The genotypes and sources of S. cerevisiae strains used or constructed in this work are described in Table I. The diploid strain W303D was made by mating W303-1A and W303-1B (23). Construction of the strains carrying a disrupted eIF1A gene is described below. Yeast cells were grown in YP or synthetic minimal medium (S) supplemented with the relevant amino acids and 2% glucose (D) or 2% galactose (G) as described (24). Cultures were grown at 30°C and were monitored by measuring optical density at 600 nm in a Beckman spectrophotometer. For sporulation (24), cells were grown on YPD plates for 24 h (containing 6% glucose) and then were sporulated at room temperature on Spo plates (0.3% potassium acetate, 0.02% raffinose, 10 g/ml of each amino acid) specialized for strain W303D. Tetrad dissections and DNA transformations were carried out by standard procedures (25).
Fractionation of Yeast Cell Lysates-Cells from strain W303-1A were grown in YPD medium to A 600 ϭ 0.5 Ϫ 1 and lysed in 20 mM HEPES-KOH, 100 mM potassium acetate, 2 mM magnesium acetate, 2 mM dithiothreitol, and 0.5 mM phenylmethanesulfonyl fluoride, pH 7.4, by vortexing with glass beads. The S30 lysate was centrifuged at 100,000 ϫ g for 22 min to generate an S100 supernatant and ribosomal pellet. The pellet fraction was suspended in 20 mM HEPES-KOH, 500 mM KCl, 6 mM magnesium acetate, 2 mM dithiothreitol, and 0.5 mM phenylmethanesulfonyl fluoride, pH 7.4, followed by centrifugation through a cush-ion of the same buffer containing 20% glycerol. The supernatant (called HSW) and washed ribosome pellet (called Rb) were collected separately and stored frozen at Ϫ70°C.
Purification of Yeast eIF-1A-Strain W303-1A was grown in 0.8 liters of YPD medium to an A 600 of 1.5. The cells were harvested (15.2 g, wet weight) and lysed as described above except that additional protease inhibitors were added to the lysis buffer: aprotinin (2 g/ml), leupeptin (0.5 g/ml), and pepstatin (0.7 g/ml). The crude S30 extract was adjusted to 500 mM KCl and centrifuged for 2 h at 100,000 ϫ g (4°C). The supernatant was dialyzed at 4°C against 5 liters of buffer H (20 mM HEPES-KOH, 5% glycerol, 0.2 mM EDTA, and 7 mM ␤-mercaptoethanol, pH 7.4) containing 100 mM KCl. The dialysate (106 mg of protein) was applied to a fast protein liquid chromatography Mono S 10/10 column and eluted with a 176-ml linear gradient of 150 -450 mM KCl in buffer H. The column fractions containing eIF1A were identified by Western blot analysis. Peak fractions were dialyzed against 2 liters of buffer H containing 100 mM KCl and applied to a Mono Q 5/5 column. Proteins were eluted with a 20-ml linear gradient of 100 -450 mM KCl in buffer H. Fractions containing eIF1A were concentrated and adjusted to 100 mM KCl by centrifugation in a Centricon 10 (Amicon) filtration apparatus. The column fractions were analyzed by SDS-polyacrylamide gel electrophoresis (SDS-PAGE) followed by Coomassie Blue staining.
Western Blot Analysis-Yeast protein samples were fractionated by 15% SDS-PAGE (26) and subjected to Western immunoblot analysis as described (22). Coomassie blue staining sometimes was used to confirm that the same quantity of protein was transferred in each lane.
Cloning the cDNA and Gene Encoding eIF-1A-Escherichia coli strain Y1090 was used to grow recombinant phages from a gt11 cDNA library of S. cerevisiae (Clontech). The eIF1A cDNA was isolated by immunoscreening the expression library with an affinity-purified antihuman eIF1A antibody (22). Membranes were treated with a 1:1,000 dilution of the anti-human eIF1A antibody followed by alkaline phosphatase-conjugated goat anti-rabbit IgG antibody. Immunoreactive plaques were identified with the chromogenic substrates nitro blue tetrazolium and 5-bromo-4-chloro-3-indolyl-1-phosphate, and putative positive phages were plaque purified. The 600-bp insert from a single positive phage was excised, cloned into phages M13 mp18 and mp19, grown in E. coli JM101, and sequenced as described below.
A yeast genomic library in gt11 (Clontech) was screened with the 32 P-labeled 600-bp insert as previously described (21). A 2.4-kb EcoRI DNA fragment was excised from one of the gt11 phages and subcloned into the EcoRI site of pSP72 (Promega) to generate pSP72-TIF11. Restriction enzymes HindIII, BglII, AccI, and EcoRV were used to digest the 2.4-kb fragment to obtain a restriction map. The smaller fragments generated by different combinations of restriction enzymes were then subcloned into M13 mp18 and mp19 vectors and transformed into E. coli strain JM109. Both DNA strands were sequenced with a Sequenase II kit (U. S. Biochemical Corp.) according to the manufacturer's instructions. DNA sequence comparisons with sequences in the GENEMBL data bank were carried out with the FASTA program.
Construction of Plasmids-Plasmid pW-y1A was constructed by ligating a blunt-ended SpeI-EcoRV genomic fragment from pSP72-TIF11, which carries the entire TIF11 open reading frame but not the adjacent CIFI gene (see "Results"), and pHSX3 (23) blunt ended at the BamHI site just downstream from the GAL1 promoter. For plasmid pW-h1A, two oligonucleotide primers were synthesized to amplify the coding region of the human eIF1A cDNA in pBluescript-1A (clone I) (22) (GenBank accession no. L18960 (21)) by PCR. Primer 1 (P1) is 5Ј-CCCCTGCAGCCGCCATGGCTCCCAAGAATAAAGG-3Ј. The underlined regions are PstI and NcoI sites; P1 corresponds to the region surrounding the initiation codon (Ϫ5 to ϩ20, where ϩ1 is the A of the initiator codon AUG). Primer 2 (P2) is 5Ј-CCCAAGCTT GAATTCA-GAAAAGAT GG-3Ј. The underlined regions are HindIII and EcoRI sites, respectively; P2 corresponds to the region 3Ј of the termination codon (ϩ458 to ϩ471). The PCR was carried out under standard conditions according to the manufacturer's protocol (Perkin-Elmer Corp.). The PCR product was cleaved with NcoI and EcoRI and blunt-ended with Klenow DNA polymerase; the resulting 470-bp fragment was subcloned into the blunt-ended BamHI site of pHSX3 to yield pW-h1A. Inserts with the correct orientation were identified by restriction enzyme cleavage patterns (results not shown). E. coli strain HB101 was used to propagate the plasmids.
Disruption of the Chromosomal TIF11 Gene-To remove the HpaI restriction site in pSP72-TIF11, the plasmid was altered by digestion with BglII and BspMI, which flank the HpaI site, and the ends were filled in with Klenow DNA polymerase followed by blunt-ended ligation. The resulting recombinant plasmid pSP72a-TIF11 was digested with SpeI and HpaI to remove a 414-bp fragment carrying 84% of the TIF11 coding region plus 27 nucleotides upstream of the AUG. The remaining 4.5-kb fragment was gel purified, the ends were filled in with Klenow DNA polymerase, and a blunt-ended DNA containing the HIS3 gene was inserted by ligation to generate pSP72a-tif11::HIS3. To prepare the HIS3 insert, a 1.75-kb BamHI fragment carrying the entire yeast HIS3 gene was isolated from plasmid pHYH (23), and the ends were filled in with Klenow DNA polymerase. pSP72a-tif11::HIS3 was digested with EcoRI to generate a 3.8-kb fragment carrying tif11::HIS3. The DNA fragment contains 1.4 and 0.6 kb of flanking DNA 5Ј and 3Ј to the TIF11 gene, respectively. The fragment was transformed into the diploid yeast strain W303D to create a one-step gene deletion/disruption (29). Stable His ϩ tranformants were selected, and the disruption of one of the TIF11 genes was confirmed by Southern blot analyses (results not shown).

Detection and Purification of the Yeast Homologue of eIF1A-
Given the fact that translational factors are very conserved from yeast to humans, it is reasonable to employ an immunoblot analysis using anti-human eIF1A antibodies to detect the corresponding yeast eIF1A protein. As shown in Fig. 1A, a single polypeptide migrating at approximately 22 kDa was detected by SDS-PAGE in a wild type W303-1A yeast cell lysate. After fractionating the yeast lysate into a low salt post- This work ribosomal supernatant (S100), a high salt-washed ribosomal pellet (Rb), and a high salt-washed supernatant, followed by Western blot analysis, the immunoreactive protein was found enriched in the high salt-washed fraction, although about 20% was present in the S100 fraction (Fig. 1A). This result implies that the immunoreactive protein weakly associates with ribosomes, as does mammalian eIF1A. Together, the results suggest that yeast may contain a homolog of mammalian eIF1A, which is similar in size, shares some of the same epitopes, and localizes to the same subcellular fractions.
Since the immunoreactive protein distributes in both the S100 and high salt-washed fractions, the lysate was raised to 500 mM KCl before initially pelleting the ribosomes. This high salt post-ribosomal supernatant (in effect, S100 ϩ HSW) supplemented with protease inhibitors was fractionated by fast protein liquid chromatography Mono S and Mono Q chromatography as described under "Materials and Methods." Putative eIF1A elutes from the Mono Q column at about 430 mM KCl as detected by Coomassie Blue staining (Fig. 1B). Approximately 10 g of yeast eIF1A were isolated from a 0.8-liter culture of cells (ϳ15 g, wet weight). The protein has an apparent mass of 22 kDa as determined by SDS-PAGE and represents the major protein in the preparation.
Yeast eIF1A Stimulates Methionyl-Puromycin Synthesis in Vitro-To obtain further evidence that the 22-kDa protein is indeed yeast eIF1A, the purified protein was tested in the mammalian methionyl-puromycin synthesis assay for eIF1A activity. A 3-fold stimulation of methionyl-puromycin synthesis was obtained, which compares favorably with stimulations seen with the purified recombinant human eIF1A (Fig. 2). Both proteins stimulate the assay to approximately the same extent and require the same amount of protein for maximal effect. The moles of eIF1A required to saturate the assay approximate the moles of ribosomes present, suggesting that eIF1A functions bound to ribosomes. The moles of methionyl-puromycin formed are quite low compared to the moles of initiation factors added; such results are routinely obtained in this assay system both for eIF1A and for other initiation factors (5,28) and may reflect more the activity of the ribosomes than that of the initiation factor being assayed. These results clearly demonstrate that the purified yeast protein possesses eIF1A activity when tested in vitro.
The Cloning of a cDNA and Genomic DNAs Encoding Yeast eIF1A-Having demonstrated an antibody-cross-reacting yeast protein with eIF1A activity, we set about to clone its gene. Immunoscreening of a gt11 yeast cDNA expression library with an affinity-purified rabbit anti-human eIF1A polyclonal antibody yielded one positive clone from approximately 6.5 ϫ 10 5 recombinant phage plaques as described under "Materials and Methods." The phage was plaque-purified, and PCR analysis revealed the presence of a 600-bp insert, which is sufficient in length to encode eIF1A. The insert was subcloned into M13 mp18 and mp19 for sequencing of both strands (see "Materials and Methods"), and the region encoding a 153-amino acid protein was identified (the cDNA sequence is not shown, but see Fig. 3). The 0.6-kb cDNA insert was 32 P labeled and used as a probe to screen a yeast genomic library as described under "Materials and Methods." Four positive clones from 6 ϫ 10 5 plaques were plaque purified for further study. Characterization of the four clones by restriction enzyme mapping and partial sequencing (data not shown) indicated that they carry portions of the same DNA. A 2.4-kb EcoRI DNA insert was then excised from one of the recombinant phages, and DNA encoding eIF1A was localized on a 1.1-kb HindIII-EcoRI fragment by Southern blot analysis. Sequence analysis (Fig. 3) confirmed that the genomic 1.1-kb clone contains an open reading frame encoding the same 153-residue protein as found in the cDNA clone. The first in-frame AUG codon is located at residues 92-94 and possesses the sequence 5Ј-AUCAUGG-3Ј, which is compatible with the yeast concensus context, A(A/U)AAUGU (30). We have named the gene TIF11 (translation initiation factor 1A) based on the similarity of its encoded protein to human eIF1A (see below).
Structural human proteins are very similar throughout the entire structure (Fig. 4). They share 65% sequence identity and 76% similarity, which reinforces the view that TIF11 encodes yeast eIF1A. The strong conservation of structure also is seen in the hydrophobicity profiles (results not shown). Both proteins contain basic N-terminal domains (29 mol % Lys ϩ Arg in the first 42 amino acids for human eIF-1A, 30 mol % Lys ϩ Arg in the first 40 amino acids for yeast eIF1A) and acidic C-terminal domains (54 mol % Asp ϩ Glu in the C-terminal 28 amino acids for human eIF1A, 54 mol % Asp ϩ Glu in the C-terminal 37 amino acids for yeast eIF1A). The highly charged terminal domains may be responsible for the apparent slow mobility of yeast eIF1A upon SDS-PAGE (apparent mass of 22 kDa versus a calculated mass of 17.4 kDa). The major difference between the yeast and mammalian proteins is the "insertion" of 8 amino acids in the yeast protein between residues 131 and 132 in human eIF1A. Comparison of the predicted amino acid sequence of yeast eIF1A with protein sequences in the GenBank and EMBL data banks indicate no significant amino acid sequence homology to other known proteins except human eIF1A. However, another gene is located upstream from TIF11 in the original 2.4-kb cloned genomic fragment. By searching the data bases, we determined that this gene encodes a homolog of the CIF1 gene (GenBank accession no. M88172). CIF1 encodes trehalose-6-phosphate synthetase in yeast and is required for cells to grow on glucose.
The number of TIF11 genes was investigated by Southern blot analyses of genomic DNA. The generation of a single 32 P-labeled band in every one of the nine different restriction digestions (results not shown) is consistent with a single copy gene. However, the analyses do not rule out the possibility of other genes encoding eIF1A that have diverged sufficiently to evade detection under the conditions used here.
eIF1A Is Essential for Cell Viability-Although eIF1A stimulates assays for initiation in vitro, it was not known if it is essential for protein synthesis and/or cell viability in intact eukaryotic cells. We therefore generated a null mutant strain in which TIF11 is substantially deleted and is disrupted by the HIS3 gene. The plasmid pSP72a-tif11::HIS3 was constructed and was used to disrupt TIF11 in the diploid strain W303D as described under "Materials and Methods." About 84% of the coding region of TIF11 is deleted, and the 3.8-kb tif11::HIS3 EcoRI-linearized fragment contains 1.4 and 0.6 kb of TIF11 flanking sequences separated by the HIS3 gene as shown in Fig. 5A. 10 His ϩ transformants were isolated, and their genomic DNAs were isolated. Southern blot analysis confirmed the integration of the fragment into a TIF11 gene on one of the chromosomes (results not shown). Transformants that carry a disrupted TIF11 gene were named CMD1. A number of independent isolates of CMD1 exhibit no growth defect at 30°C when measured either on plates or in liquid culture (results not shown).
To determine the phenotype of a null mutant that lacks eIF1A, independent isolates of CMD1 were sporulated and tetrads were dissected. A 2 ϩ :2 Ϫ viable to nonviable segregation pattern was seen after 2 days (Fig. 5B) and all of the viable spores showed a His Ϫ phenotype (results not shown). This suggests that the phenotype of the tif11::HIS3 allele is lethal. The nonviable spores were further examined after incubation for a week at 23°C to confirm that they failed to germinate and grow. The fact that the tetrad spores generated a 2 ϩ :2 Ϫ segregation pattern and no His ϩ segregants strongly suggests that eIF1A is required for cell viability and/or germination.
To distinguish between these two possibilities, the entire TIF11 open reading frame (but lacking the CIF gene homolog) was placed under control of the glucose-repressible GAL1 promotor as described under "Materials and Methods." The resulting plasmid pW-y1A was transformed into the diploid strain CMD1. Since pW-y1A carries TRP1 as the selectable marker, transformants were selected on a SD-trp plate and named CMD2. This strain was subsequently sporulated followed by tetrad dissection on galactose plates. Due to random plasmid segregation, the ratio of viable to nonviable spores was 2:2, 3:1, or 4:0 (Fig. 6A). Co-segregation of the His ϩ and Trp ϩ phenotypes further indicates that TIF11 expressed under galactose induction is sufficient to support the growth of the tif11:: HIS3   FIG. 3. Sequence of TIF11. The DNA sequence of the HindIII-EcoRI fragment, which contains the coding region of TIF11 and its flanking sequences, was determined as described under "Materials and Methods." The derived amino acid sequence for eIF1A is aligned below. Residue numbers for nucleotides and amino acids are shown on the right. The reported sequences as well as another 1.3 kb of DNA sequence upstream are available from GenBank under accession number U11585. FIG. 4. Comparison of yeast and human eIF1A sequences. Human (upper) and yeast (lower) eIF1A sequences were aligned using the program BESTFIT from the GCG software package on a VAX computer. Sequence identities are shown by vertical lines between the two sequences; similar residues, marked by two dots, are defined as: VIL, DE, KR, NQ, FYW, and ST. Every 10th residue has a dot above it. The sequences exhibit 65% identity and 76% similarity.
haploid strain (Fig. 6B, upper row). This fact rules out the possibility that the 2 ϩ :2 Ϫ segregation pattern seen when CMD1 was sporulated (Fig. 5B) could be due to down-regulation of the adjacent CIF homolog gene, since pW-y1A does not carry the CIF homolog gene. A haploid cell colony containing the tif11::HIS3 allele and plasmid pW-y1A was selected and named CM1.
If eIF1A is essential for cell viability, the His ϩ Trp ϩ spores will germinate and grow on plates with galactose medium, which induces eIF1A expression, but should not grow when transferred to plates with glucose medium. On the other hand, if eIF1A is required only for germination, the germinated spore colonies will continue to grow on glucose medium. When the spore colonies were streaked on glucose-containing SD-his or SD-trp plates, growth ceased (Fig. 6B, bottom row). The results show that TIF11 is required for cell growth and viability.
Human eIF1A cDNA Complements Yeast tif11::HIS3-The amino acid sequences of human and S. cerevisiae eIF-1A share 65% identity and 76% similarity. The fact that yeast eIF1A functions as well as human eIF1A in the in vitro methionylpuromycin synthesis assay reconstituted with mammalian components strongly suggests that the two proteins are functionally equivalent. Since TIF11 is essential for yeast cell growth, we are able to use the no-growth phenotype to test if the human cDNA can replace the yeast gene in vivo. cDNA encoding the entire human eIF1A open reading frame was placed under control of the yeast GAL1 promotor in plasmid pHSX3 as described under "Materials and Methods." The resulting plasmid pW-h1A was introduced into the diploid strain CMD1 to yield CMD3. The diploid strain was sporulated, and dissected spores were germinated on YPG plates, which allow the expression of the human form of eIF1A. The appearance of viable to nonviable spores in the ratio of 2:2, 3:1, and 4:0 (Fig.  7A) resembles that obtained when yeast TIF11 is expressed ( Fig. 6A) and therefore suggests that human eIF1A can function in yeast. To confirm that the human cDNA functionally replaces TIF11, spore colonies were streaked on various tester plates. All viable His ϩ cells (indicating disruption of TIF11 on the chromosome) also have a Trp ϩ phenotype (indicating the presence of the human cDNA on the plasmid). The His ϩ Trp ϩ cells grow only on galactose but not on glucose-containing plates (Fig. 7B). The growth rate of cells expressing human eIF1A appears to be somewhat slower compared to cells expressing TIF11 since their colony size is smaller. A His ϩ Trp ϩ haploid spore colony obtained from CMD3 was named CM2. The results demonstrate that human eIF1A cDNA expressed in yeast is able to suppress the lethal phenotype of the tif11::HIS3 allele and support cell growth as the sole source of eIF1A.
To show that strain CM2 actually synthesizes human eIF1A but not yeast eIF1A, Western blot analysis was carried out with polyclonal anti-human eIF1A antibody affinity purified with yeast eIF1A. The results (Fig. 8) clearly show that under galactose growth conditions, the S30 lysate prepared from strain CM2 (lane 2) contains human eIF1A (ϳ20 kDa-protein band) but no yeast eIF1A, which migrates somewhat more slowly at 22 kDa. In comparison, a lysate prepared from a strain carrying the wild type TIF11 gene and no human eIF1A cDNA contains only yeast eIF1A (lane 3, ϳ22-kDa protein band) as expected. When the two lysates are combined and analyzed, the two eIF1A bands are readily resolved (lane 1). DISCUSSION We report here the purification of a yeast initiation factor called eIF1A and the cloning of its gene, TIF11. Several lines of  6. Tetrad analysis of CMD2 spores. Panel A, spores from 9 tetrads of CMD2 were analyzed as described in Fig. 5B except that growth was on YPG plates. Panel B, following tetrad dissections of CMD2 (panel A), one tetrad giving four viable spore colonies was selected, and the four colonies were streaked onto YPG and YPD plates and onto SG and SD plates with the omission of histidine or tryptophan as indicated in the figure. The plates were incubated at 30°C for 2-4 days, photographed, and computer scanned. evidence indicate that the yeast protein corresponds to the mammalian homolog, eIF1A. 1) The apparent mass determined by SDS-PAGE is similar for the yeast (22 kDa) and mammalian (20 kDa) protein.
2) The amino acid sequence of yeast eIF1A exhibits 65% identity and 76% similarity to human eIF1A, and both proteins have basic N termini and acidic C termini. 3) Highly specific antibodies to mammalian eIF1A cross-react with the yeast protein. 4) Yeast eIF1A substitutes for mammalian eIF1A in the in vitro assay for methionyl-puromycin synthesis. 5) The human eIF1A cDNA confers growth to yeast cells lacking the TIF11 gene.
The strong conservation of primary structure between the yeast and mammalian homologs suggests that eIF1A plays an important role in these cells. A particularly striking structural feature is the highly charged termini of the factor, where the N terminus is positively charged and the C terminus is negatively charged. The functions of these domains are not known, but one may speculate that the N terminus is responsible for the RNA binding property of eIF1A (27). The yeast eIF1A sequence exhibits no significant homology to other proteins (other than human eIF1A) in the data bases. eIF1A is one of the most conserved proteins among the initiation factors. Sequence identities between the yeast and mammalian initiation factors range from 26% for eIF4B (15,16) and 33% for eIF4␣ (17), representing the least conserved proteins, to 65% for eIF4A (31) and 71% for eIF2␥ (32), the latter being the most conserved initiation factor known at this time. Thus, initiation factors are nearly as conserved as ribosomal proteins where the sequence identity between all cognate yeast and mammalian ribosomal proteins is 60%, with individual proteins falling in the range from 40 to 88% (33).
The similar primary structures of yeast, mammalian, and plant eIF1A are manifested in their biochemical activities. Either yeast or wheat germ eIF1A (6) functions in place of mammalian eIF1A in an in vitro assay for initiation based on mammalian components. Furthermore, the human cDNA encoding eIF1A complements a yeast strain lacking a functional TIF11 gene, indicating that the human protein functions in vivo with yeast components of the translational machinery. Many other mammalian cDNAs can relieve a growth defect of cells where the corresponding yeast gene has been disrupted. For example, the effects of disruption of SUI2, SUI3, TIF51A/B, and CDC33 are reversed by expressing the cDNAs for eIF2␣, eIF2␤, eIF5A, and eIF4␣, respectively (16,23). 2 Only in the case of eIF4A does the mammalian cDNA fail to relieve disruption of TIF1 and TIF2 (31).
Most initiation factor proteins in yeast are essential for cell growth and viability. The only exceptions known to date are the ␣-subunit of eIF2B (GCN3) (34) and eIF4B (TIF3) (15,16), where cells grow in the absence of the protein, albeit more slowly. The cloning of TIF11 and the demonstration that it is essential for cell growth will allow us to address the function of this protein by genetic and biochemical studies. Both approaches likely will be required to explain the pleiotropic effects of this small yet essential initiation factor. FIG. 7. Tetrad analysis of CMD3 spores. Panel A, spores from 9 tetrads of CMD3 were analyzed as described in Fig. 5B except that growth was on YPG plates. Panel B, following tetrad dissections of CMD3 (panel A), one tetrad giving four viable spore colonies was selected and analyzed as described in Fig. 6B.   FIG. 8. Western blot analysis of human and yeast eIF1A. Strains W303-1A and CM2 were lysed in 10 mM Tris-HCl, pH 7.5, 2 mM phenylmethanesulfonyl fluoride, and 10 mM dithiothreitol and centrifuged at 29,000 ϫ g for 10 min; lysate protein from each strain was fractionated by 15% SDS-PAGE and analyzed by immunoblotting with anti-human eIF1A antibodies affinity purified with yeast eIF1A as described under "Materials and Methods." The figure shows a computer scan of the blot. Lane 1, wild type strain W303-1A and strain CM2 expressing human eIF1A (10 g each); lane 2, strainW303-1A (10 g); lane 3, strain CM2 (10 g).