INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DICUSSION REFERENCES

Initiation of translation in eukaryotic cells requires the participation of GTP and a large number of specific proteins called eukaryotic translation initiation factors (eIFs). Among the initiation factors, eIF3 is the most complex initiation factor known with respect to both its function and its subunit composition. In vitro, the factor binds to the 40S ribosomal subunit and the 40S-bound eIF3 plays an essential role in the eIF1A-dependent transfer of initiator Met-tRNA_f (as Met-tRNA_f·eIF2·GTP ternary complex) to 40S subunits to form the 40S-preinitiation complex (40S·eIF3·Met-tRNA_f·eIF2·GTP) (Chaudhuri et al., 1999). The presence of eIF3 bound to the 40S preinitiation complex is also a stringent prerequisite for the 40S preinitiation complex to bind at the 5'-capped end of mRNA as well as for the subsequent scanning of the mRNA by the 40S preinitiation complex to locate the initiation AUG codon of the mRNA to form the 40S initiation complex (40S·eIF3·mRNA·Met-tRNA_f·eIF2·GTP) (Merrick and Hershey, 1996; Kozak, 1999). The precise mechanism by which eIF3 functions in each of the above steps is yet to be defined.

There is considerable uncertainty regarding the subunit composition of functional eIF3. The factor has been isolated in several laboratories from a variety of eukaryotic sources based on an assay that measured its ability to stimulate mRNA translation in a protein synthesizing system reconstituted with purified proteins (Benne and Hershey, 1976; Safer et al., 1976; Schreier et al., 1977; Merrick, 1979; Spremulli et al., 1979; Checkley et al., 1981; Brown-Luedi et al., 1982; Seal et al., 1983). Mammalian eIF3, purified in this way, was reported to consist of ten major polypeptides, p170, p116, p110, p66, p48, p47, p44, p40, p36,and p35 (Hershey et al., 1996). In contrast to mammalian eIF3, eIF3 purified from Saccharomyces cerevisiae on the basis of an assay that measures AUG-dependent methionyl-puromycin synthesis consisted of eight major polypeptides of apparent molecular masses, 135, 90, 62, 39, 33, 29, 21, and 16 kDa (Naranda et al., 1994). However, when budding yeast eIF3 was purified based on a direct assay for the presence of Prt1p, a known subunit of yeast eIF3, it was observed that Prt1p copurified with only four other polypeptides (Danaie et al., 1995; Phan et al., 1998). This five-subunit core complex, consisting of Rpg1p (also designated Tif32p), Prt1p, Nip1p, Tif34p, and Tif35p subunits, stimulated the transfer of Met-tRNA_f to 40S ribosomal subunits nearly 10-fold (Danaie et al., 1995; Phan et al., 1998).

Comparison of the predicted amino acid sequence of each of the 10 mammalian eIF3 subunits with the derived protein sequences in the yeast S. cerevisiae genomic data base revealed that all the five subunits of yeast eIF3 core complex (Rpg1p, Prt1p, Nip1p, Tif34p, and Tif35p) have corresponding homologues in mammalian eIF3, which are p170, p116, p110, p36, and p44, respectively (Asano et al., 1998). However, the other mammalian eIF3 subunits, p66, p48, p47, p40, and p35 have no structural homologues in S. cerevisiae. This was somewhat surprising in view of the well-accepted notion that the basic translation machinery including the pathway of translation initiation is highly conserved between yeast and mammals (Donahue and Huang, 1997).

Among the mammalian eIF3 subunits that have no structural homologues in budding yeast, the p48 subunit is of particular interest. Sequence analysis of the p48-cDNA showed that p48 is identical to the mouse protein INT6 (Asano et al., 1997). The genomic locus of Int6 has been a frequent integration site for mouse mammary tumor virus (MMTV) (Marchetti et al., 1995). This integration induces formation of mammary tumors in mice although the mechanism of tumor formation is not clear. The key question from the point of view of translation is what specific role p48 (INT6), as a subunit of mammalian eIF3, performs in translation initiation in mammalian cells. Interestingly, database searches revealed that p48 has a structural homologue in the fission yeast Schizosaccharomyces pombe. This observation provided a unique opportunity to study the role of p48 (Int6) in translation initiation in S. pombe.

In this paper, we describe the cloning and characterization of the S. pombe gene encoding Int6 protein. Deletion of int6⁺ (int6) is viable and exhibits only moderate inhibition in the rate of in vivo protein synthesis, with no apparent defects in translation initiation. int6 as well as expression of a C-terminally truncated version of Int6 cause phenotypes suggestive of a defect in the maintenance of cell wall/membrane integrity.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DICUSSION REFERENCES

Strains and Media

All strains used in this study were derived from wild-type strains. S. pombe was grown in standard YEA and PM media to which leucine and/or uracil were added when necessary.

Cloning of int6⁺ and Plasmid Construction

BLAST (Basic Local Alignment Search Tool) search of the web-based genome database (www.Sanger.ac.uk/BLAST) of the Sanger Centre (Cambridge, United Kingdom) database with human Int6 sequence revealed that the S. pombe homologue of human Int6 is present in cosmid No. 646. A 3.5-kb XhoI/SpeI fragment containing int6⁺ was obtained from cosmid No. 646 and cloned into pBluescript SK (-) to generate the plasmid pINT6. The coding sequence of Int6 cDNA was amplified by PCR from an S. pombe cDNA library, using pfu DNA polymerase (Stratagene. L Jolla, CA) and two primer sequences as follows: N-terminus, 5'-d GGCACTCGAGCATATGGGGTCCGAGCTTAAGAG-3' having a XhoI/NdeI overhang; C-terminus, 5'-dAGCTGAATTCGGATCCCTAAACAGTAGCATGCTT-3' having a BamHI/EcoRI overhang. The PCR product was digested with NdeI and BamHI, was cloned into the same sites of pREP41 to generate the thiamine-repressible pREP41-INT6 expression plasmid, and was sequenced to ensure error-free DNA synthesis. To express the C-terminally truncated 50 amino acid shorter version of Int6 protein, the same N-terminal primer was used in the PCR reaction, while the other primer used for C-terminus was 5'-d AGCTGAATTCGGATCCCTACTTAGCTTTAAACCCAAATTG -3'. The PCR product was digested with NdeI and BamHI and was cloned in pREP41 to generate a thiamine-repressible pREP41-C50 INT6 expression plasmid. Wild-type cells were transformed following standard procedures.

Gene Disruption

An SphI site was generated at a BamHI site on pINT6, which is 4 bp downstream of the first ATG initiation codon of the putative int6⁺ open reading frame. The modified plasmid pINT6-1 contained two SphI sites in the open reading frame of int6⁺, one generated at the BamHI site and the other ~ 1543 bp downstream of the first ATG codon. A 1.8-kb SphI fragment containing the fission yeast ura4⁺ gene was inserted between the two SphI sites resulting in plasmid pINT6, which would remove a 1539 bp segment of int6⁺ open reading frame. This construct, digested by XhoI and SpeI, was used to disrupt one copy of the int6⁺ gene in a diploid (h⁺/h leu1-32/leu1-32 ura4-d18/ura4-d18 ade6-216/210). Stable Ura⁺ transformants were selected, and the presence of ura4⁺ at the int6⁺ locus was confirmed by Southern analysis.

Construction of Myc or Green Fluorescent Protein-Tagged int6⁺

Open reading frame (ORF) of Enhanced Green Fluorescence Protein (EGFP) was amplified by PCR using pfu DNA polymerase (Stratagene) and two primer sequences as follows: N-terminus, 5'-d CCGCTCGAGGGGCATGCTAGTAAAGAGAAGAACTTTTC-3' having a XhoI/SphI overhang; C-terminus, 5'-d TCCGAATTCAGCATGCTTTTTGTATAGTTCATCCATGCC-3' having a SphI/EcoRI overhang. The PCR product was digested with SphI and was cloned into SphI linearized pINT6 plasmid, thus generating pINT6-GFP plasmid. The resulting construct was digested by XhoI and SpeI and was used to knock in Int6-GFP fusion gene in the int6 haploid in which int6⁺ locus was disrupted with ura4⁺.

The N-terminally Myc-tagged Int6-expressing strain was also generated using a similar strategy except that the BamHI site at + 4 position was utilized to insert the Myc tag at the N-terminus, maintaining the correct reading frame.

Caffeine Sensitivity Assay

The wild-type and int6 strains were streaked on fresh YEA plates and incubated at 32°C for two days. Cells from YEA plates were inoculated into YEA medium containing 10 mM caffeine and were incubated at 32°C for 12 h. Cells were harvested and washed with 1.2 M sorbitol, followed by staining with DAPI (4', 6-diamidine-2-phenylindole dihydro-chloride), then examined under a fluorescence microscope enhanced by CCD. For the experiment presented in Figure 10, fresh wild-type cells transformed with empty pREP41, pREP41-INT6, or pREP41-C50 INT6 were streaked on PM plates supplemented with 10 µg/ml thiamine and were grown at 32°C for 12 h. The cells were inoculated into PM medium containing 5 mM caffeine and were grown for another 12 h at 32°C. Cells were harvested, washed with 1.2 M sorbitol, stained with DAPI, and examined under a fluorescence microscope enhanced by CCD.

Cloning and Expression of Full-length Human Int6 Gene in S. pombe

Full-length human INT6 ORF was reverse-transcribed and PCR amplified from total HeLa RNA using Superscript preamplification system (Life Technologies-BRL Lifetech, GIBCO, Grand Island, NY). Sequences of the two primers used were as follows: N-terminus, 5'-d CCGCTCGAGCATATGGCGGAGTACGACTTGACT -3' having a XhoI/NdeI overhang; C-terminus, 5'- dCGCGGATCCGAATTCTCAGTAGAAGCCAGAATCTTGAG -3' having a BamHI/EcoRI overhang. The PCR product was digested with NdeI and BamHI, cloned into the same sites of pREP41 to generate the thiamine-repressible pREP41-hINT6 expression plasmid, and sequenced to ensure error-free DNA synthesis. The construct was transformed into int6 cells, and transformants were selected on PM plates supplemented with 10 µg/ml thiamine.

Assay for Growth Inhibition in Presence of High KCl Concentration

Wild-type and int6 strains were streaked on fresh YEA plates and incubated at 32°C for two days. Cells recovered from these plates were inoculated into YEA medium containing 1.2 M KCl and were grown in this high-salt medium for ~ 12 h at 32°C. The cells were washed and examined as described under "Caffeine Sensitivity Assay".

-Glucanase Sensitivity Assay

-glucanase sensitivity assay was performed following procedures described previously (Levin and Bishop, 1990). Briefly, 5 × 10⁷ cells of WT, pmk1, or int6 cells were harvested from midlogarithmic phase cultures in YEA medium and were washed with and resuspended in a buffer containing 10 mM Tris-HCl, pH 7.5, 1 mM EDTA, 1 mM 2-mercaptoethanol, followed by treatment with 100 µg/ml -glucanase (Zymolyase-20T, ICN) in the same buffer at 32°C with vigorous shaking. Cell lysis was monitored by measuring absorbance at 600 nm.

In Vivo Protein Synthesis

The method used for measurement of in vivo protein synthesis is an adaptation of the method of Sachs and Deardorff (Sachs and Deardorff, 1992). Exponentially growing cultures of wild-type and mutant int6 strains in YEA medium at 32°C were harvested, resuspended in fresh YEA medium to a cell density of 0.75 × 10⁶ cells/ml, and allowed to grow. When the cells reached the density of ~ 1 × 10⁶ cells/ml, ~ 2.5 ml of each culture was treated with 100 µCi of [³⁵S]methionine (1175 Ci/mmol), and cells were allowed to grow with vigorous shaking. At the indicated times, an aliquot (0.8 ml) was withdrawn, mixed with an equal volume of 20% trichloroacetic acid (TCA) containing 1.2 mg/ml unlabeled methionine, and the mixture heated at 95°C for 20 min in glass tubes. After cooling in an ice-water bath, the precipitated proteins were filtered through GF/C filters (Whatman, Maidstone, United Kingdom), which were then washed several times with 10% ice-cold TCA, followed by a final washing with 95% ethanol. The dried filters were then assayed for ³⁵S-radioactivity by counting in Econoflour (Packard Instrument, Meriden, CT) in a liquid scintillation spectrometer.

Polysome Profile Analysis

Exponentially growing cultures of the wild-type strain and the int6 strains in YEA medium at 32°C were harvested, resuspended in fresh YEA medium to a final cell density of 0.75 × 10⁶ cells/ml, and allowed to grow. When the cells reached the density of ~ 1 × 10⁶ cells/ml, a 70-ml aliquot of each culture was treated with cycloheximide (50 µg/ml), rapidly chilled in an ice water bath, and then harvested. The cells were washed twice with LHB buffer (10 mM Tris-HCl, pH 7.5, 100 mM NaCl, 30 mM MgCl₂, and 50 µg/ml cycloheximide). The washed cells were then suspended in 0.5 ml of LHB buffer, lysed by vortexing with an equal volume of glass beads, then treated with an additional 0.5 ml of LHB buffer. The lysates were clarified by centrifugation at 12,000 × g for 15 min, and equivalent amounts of A₂₅₄ absorbing material (~ 10 A₂₅₄ units) were layered on 11 ml of 7 to 47% (wt/vol) sucrose gradients in TMN buffer (10 mM Tris-Acetate, pH 7.0, 12 mM MgCl₂, 50 mM NH₄Cl) and centrifuged at 40,000 rpm for 2.5 h at 4°C in a Beckman SW41 rotor (Beckman, Fullerton, CA). The gradients were fractionated in an ISCO density gradient fractionator, and the absorbance profile at 254 nm was analyzed in an ISCO (Lincoln, NE) UA-5 absorbance monitor.

Association of Fission Yeast Int6 with 40S Ribosomal Particles

A 100-ml culture of fission yeast cells harboring N-terminally Myc tagged Int6 were grown to midlogarithmic phase in YEAU medium. Cells were harvested and lysates prepared as described under "Polysome Profile Analysis" (see above). Approximately 10 A₂₅₄ units of the lysates were layered on 11-ml of 5-30% (wt/vol) sucrose gradients in TMN buffer (10 mM Tris-Acetate, pH 7.0, 12 mM MgCl₂, 50 mM NH₄Cl) and were centrifuged at 40,000 rpm for 2.5 h at 4°C in a Beckman SW41 rotor. In a parallel tube, a preformed 43S preinitiation complex (40S·eIF3·AUG·[³⁵S]Met-tRNA_f·eIF2·GTP) formed with purified mammalian initiation components was also analyzed to define the position of sedimenting 40S particles. Fractions (500 µl) were collected from the bottom of each tube, and were treated with trichloroacetic acid (16% final concentration). The precipitated proteins in each fraction were resuspended in 50 µl of SDS-gel loading buffer, then boiled, and a 10 µl aliquot was analyzed by SDS-PAGE (10% gel) followed by immunoblotting using appropriate antibodies. In the gradient tube containing the 43S preinitiation complex, an aliquot of each fraction was counted in Aquasol (Packard Instrument, Meriden, CT) in a liquid scintillation spectrometer to determine the position of the 43S preinitiation complex.

Other Methods

eIF3 was purified from rabbit reticulocyte lysates as described by Chaudhuri et al. (1997a). The purified eIF3 preparation contained the p48 polypeptide (INT6 protein) as judged by immunoblot analysis using antihuman INT6 antibodies as probe. The 43S preinitiation complex containing bound [³⁵S]Met-tRNA_f was prepared as described by Chaudhuri et al. (1997b) as follows. A reaction mixture (225 µl), containing 20 mM Tris-HCl, pH 7.5, 100 mM KCl, 5 mM 2-mercaptoethanol (TKM buffer), 15 µg of bovine serum albumin, 40 pmol of [³⁵S]Met-tRNA_f (17,500 cpm/pmol), 18 µM GTP, and 5 µg of purified rabbit reticulocyte eIF2, was incubated for 4 min at 37°C to form the [³⁵S]Met-tRNA_f.eIF2.GTP ternary complex. The reaction mixture was then supplemented with 1.8 A₂₅₄ units of purified 40S ribosomal subunits, 0.6 µg of purified eIF1A, 5 µg of purified eIF3, 0.1 A₂₅₄ unit of AUG, and MgCl₂ (1 mM final concentration). After incubation at 37°C for 4 min, the reaction mixture was chilled in an ice-water bath, and the MgCl₂ concentration was raised to 5 mM, then analyzed by 5-30% sucrose gradients in TKM buffer containing 5 mM MgCl₂ as described (Chaudhuri et al., 1997b). Rabbit polyclonal antihuman INT6 antibodies are a kind gift from Dr. Pierre Jalinot of Laboratoire de Biologie Moléculaire et Cellulaire, Lyons, France. The antibodies were raised against the C-terminal 20 amino acids of the protein (Desbois et al., 1996).

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DICUSSION REFERENCES

Molecular Cloning of Fission Yeast int6⁺

Comparison of the predicted amino acid sequence of each of the 10 mammalian eIF3 subunits with the derived protein sequences in S. pombe and S. cerevisiae genomic databases showed that only five of the mammalian eIF3 subunits have corresponding homologues in the budding yeast, S. cerevisiae. In contrast, except p35, all other subunits of mammalian eIF3 including INT6 (p48) have structural homologues in the fission yeast, S. pombe (Table 1). The int6⁺ gene was found on a cosmid clone (No. 646) that was previously mapped on chromosome II of S. pombe genome (Mizukami et al., 1993). The Int6 protein is highly conserved among species throughout the length of the protein. Int6 from S. pombe, C. elegans, Drosophila, mouse, and human were aligned for maximum homology by Jotun Hein method (Figure 1). S. pombe Int6 is 43% identical and 59% similar to its mammalian counterparts. The fission yeast Int6, however, is ~ 15% longer than its higher eukaryotic homologues due to the presence of extra amino acids at the extreme N-terminus as well as near the C-terminus. Following appropriate enzymatic digestion of the cosmid No. 646, the entire int6⁺ gene, flanked by surrounding genomic sequence, was identified in a 3.5-kb SpeI-XhoI fragment. This fragment was inserted into the appropriate restriction sites of pBluescript SK (-) to generate pINT6. Analysis of the nucleotide sequence of the 3.5-kb SpeI-XhoI fragment indicated that the gene has a 58-bp long intron 34-bp downstream of the start ATG. The initiation ATG codon at + 1 is preceded by a translational stop codon TGA at position 15 and fairly satisfies the criteria for consensus of translational start sites ³GAUAUGG⁺⁴ (the start codon is italicized) (Kozak, 1999).

View larger version (113K): [in this window] [in a new window]   Figure 1.   Comparison of Int6 sequences. Alignment of the predicted fission yeast Int6 amino acid sequence with its structural homologues from worm, fruit fly, mouse, and human. The conserved residues are highlighted. The arrowhead marks the position where C50 ends.

Deletion of int6⁺

The int6⁺ gene was replaced with the fission yeast ura4⁺ gene as described under MATERIALS AND METHODS. The replacement (Figure 2A) would allow expression of, at most, the first one and the last two amino acids of the Int6 protein and thus would not cause any phenotypes due to the remaining part of the int6⁺ gene. After transformation of wild-type diploid cells with a plasmid (designated pINT6) that carried the ura4⁺ gene in place of the int6⁺ gene, stable Ura⁺ transformants were isolated. Genomic DNAs of several transformants were subjected to Southern blot analysis to screen for heterozygous diploids with one wild-type int6⁺ gene and one disrupted (h⁺/h leu1-32/leu1-32 ura4-d18/ura4-d18 int6⁺::ura4⁺/+ ade6-216/210). The heterozygous diploid was then sporulated for tetrad analysis. Of the 13 tetrads, eight yielded all four viable spores in which the Ura⁺ marker segregated into 2:2; indicating int6⁺ gene is not essential for viability. To further confirm the disruption of the int6⁺ locus, Southern and Northern analyses were carried out following standard protocols. The SalI restriction digest followed by hybridization of the probe (shown in Figure 2A) would identify a 1.9-kb DNA fragment in wild-type int6⁺ locus, whereas it would identify a 4.7-kb fragment from the disrupted locus. As anticipated above, a single 1.9-kb fragment was identified in the wild-type diploid (Figure 2B, lane 1), whereas a 4.7-kb fragment in addition to the 1.9-kb fragment was observed in the heterozygous diploid (lane 2). Analysis of four segregants from a tetrad of the heterozygous diploid identified the 4.7-kb DNA fragment in two of the Ura⁺ segregants (lanes 3 and 5), whereas in the other two segregants that were ura4, the 1.9-kb band was detected (lanes 4 and 6). Northern analysis of the four segregants further confirmed the disruption of the int6⁺ locus. The int6⁺ transcript was detected in the wild-type diploid, the heterozygous diploid as well as in the two ura4 segregants (Figure 2C, lanes 1, 2, 4 and 6). In contrast, no int6⁺ transcript was detected in the two Ura⁺ segregants (lanes 3 and 5), although approximately similar amounts of total yeast RNA (20 µg) were examined (Figure 2C, lower panel). These results indicated that 1) the int6⁺ gene was successfully replaced by the ura4⁺ gene, and 2) the resulting disrupted strains that expressed no int6⁺ gene were viable. The int6⁺-disrupted haploid, henceforth, will be referred to as int6.

View larger version (36K): [in this window] [in a new window]   Figure 2.   Analysis of int6+ disrupted strain. (A) Strategy of disruption of the int6+ gene and construction of C-terminally GFP tagged or N-terminally Myc tagged Int6. The entire int6+ gene was replaced with the fission yeast ura4+ gene in a diploid. The SalI restriction sites are shown as S1. The GFP or Myc tag is inserted in the SphI or BamHI site, respectively, as described in MATERIALS AND METHODS and inserted in place of the ura4+ gene. (B) Confirmation of disruption by Southern blotting. The replacement was confirmed by Southern blotting using a 32P-labeled PCR product from the 5'UTR of int6+ locus as a probe (thick bar). Genomic DNAs of a wild-type diploid (lane 1), heterozygous diploid (int6+:: ura4+/+) (lane 2), a set of four segregants from the heterozygous diploid (lanes 3 to 6) were examined. (C) Confirmation of disruption by Northern blotting. Northern blot was performed with 20 µg of total RNA from a wild-type diploid (lane 1), the heterozygous diploid (int6+ :: ura4/+) (lane 2), and a set of four segregants from the heterozygous diploid (lanes 3 to 6) using a probe generated from within the int6+ ORF (thin bar). The lower panel shows the ethidium bromide stained RNA gel.

Although the int6⁺ gene is not essential for growth and viability, int6 cells grew more slowly than a wild-type strain. At 32°C, an optimal growth temperature for fission yeast, the generation times of a wild-type and int6 cells in YEA medium were 2.0 ± 0.2 h and 2.56 ± 0.2 h, respectively (Figure 3A). The growth of int6 strain was particularly impaired in S. pombe-liquid minimal medium (PMA). In this medium, while the wild-type strain grew at a rate similar to that in YEA liquid medium, the generation time of int6 strain in PMA was greater than 3.5 h. The slow growth phenotype of int6 could be rescued by introduction of the wild-type int6⁺ gene or to a lesser extent by the human Int6 gene provided from an extrachromosomally replicating plasmid (Figure 3B).

View larger version (41K): [in this window] [in a new window]   Figure 3.   Slow growth of int6 strain and rescue by the wild-type fission yeast or human int6+ gene. (A) Growth curve. Cells were grown to ~ 5 × 106 cells/ml at 32°C and then diluted to 1.3 × 106 cells/ml at time 0. The cell density was monitored over a 6 h period. (B) int6 strain was transformed with either empty pREP41 vector or vectors bearing full length fission yeast int6+ or human Int6 gene. Transformed cells were streaked on a PM plate.

We also constructed strains that express Int6 tagged with Myc-epitope or GFP. In the parental plasmid, pINT6, appropriate restriction sites were generated into which Myc-epitope or GFP were inserted. The resulting plasmids were used to replace the disrupted int6⁺ gene with int6⁺ tagged with Myc-epitoope or GFP (Figure 2A). The ura transformants that had lost the ura4⁺ gene due to the replacement were collected and further examined for the structure of the int6⁺ gene. Using Southern and Northern blotting as well as PCR, we confirmed that the replacement was successful and that it produced strains expressing tagged Int6 protein, but were otherwise identical to the parental strain (data not shown). These strains did not show any phenotypes conferred by int6 (see below), indicating that Int6 tagged with Myc-epitope or GFP are biologically active.

Association of Int6 with 40S Ribosomal Subunits

Eukaryotic translation initiation factor 3 (eIF3), purified from mammalian cell lysates, was reported to consist of 10 major polypeptides (Hershey et al., 1996). One of these polypeptides, p48 was shown to be INT6 based on cDNA sequence analysis (Asano et al., 1997). Consistent with this observation, we also found that purified mammalian eIF3 ((Benne and Hershey, 1976; Chaudhuri et al., 1999) contained INT6 polypeptide (Figure 4A, inset). However, it should be noted that the level of INT6 in this preparation was not stoichiometric to the other subunits of eIF3 (see Discussion). Characterization of mammalian eIF3 has shown that in crude mammalian cell-free extracts, eIF3 is found exclusively bound to 40S particles (Smith and Henshaw, 1975; Thompson et al., 1977). In vitro studies using purified initiation factors have also shown that eIF3 binds to 40S subunits in the absence of all other initiation components to form the 40S.eIF3 complex that is stable to sucrose gradient centrifugation (Benne and Hershey, 1976; Chaudhuri et al., 1999). The initiator Met-tRNA_f (as Met-tRNA_f.eIF2.GTP ternary complex) then binds to the 40S.eIF3 particle to form the 40S preinitiation complex (40S.eIF3. Met-tRNA_f.eIF2.GTP), which subsequently recognizes the AUG codon of the mRNA to form the 40S initiation complex (40S.eIF3.mRNA.Met-tRNA_f.eIF2.GTP). These properties of mammalian eIF3 prompted us to examine the association of Int6 with 40S particles in fission yeast cells. We reasoned that if fission yeast Int6, like its mammalian counterpart, is a subunit of eIF3, it should remain bound to the 40S particles under conditions where mammalian eIF3 containing INT6 as one of its subunits remains bound to 40S particles.

View larger version (29K): [in this window] [in a new window]   Figure 4.   Association of Int6 with 40S ribosomal particles in yeast cells. (A) Mammalian INT6 sediments with the 43S preinitiation complex. A 43S preinitiation complex containing bound [35S]Met-tRNAf and formed in presence of purified mammalian eIF3, as described in MATERIALS AND METHODS, was analyzed by 5-30% (wt/vol) sucrose gradients. Aliquots of the gradient fractions were assayed for 35S-radioactivity by scintillation counting to determine the sedimentation position of the 43S preinitiation complex (43S IC). Each gradient fraction was also analyzed by Western blotting using antihuman INT6 antibodies as probes. Inset. Approximately 300 ng of purified rabbit reticulocyte eIF3 (Chaudhuri et al., 1997) were subjected to SDS-15% PAGE. The gel was treated with silver stain reagents (lane 1) or analyzed by Western blotting using a 1:2000 dilution of polyclonal antihuman INT6 rabbit antibodies (lane 2). The arrowhead indicates the position of INT6 polypeptide, which migrates at the expected position. A set of molecular weight standards was run in parallel lanes (not shown). (B) Fission yeast Int6 cosediments with 40S particles. An exponentially growing culture of fission yeast cells expressing N-terminally Myc tagged Int6 was harvested, the cells were washed and lysed, and the cell lysates (10 A254 units) were subjected to 5-30% sucrose gradient centrifugation to separate polysomes, free 80S, 60S, and 40S ribosomes, as described in MATERIALS AND METHODS. Fractions (0.5 ml each) from the gradient were collected and proteins from 0.25 ml of each fraction were precipitated with 16% trichloroacetic acid, analyzed by SDS-PAGE, and immunoblotted with anti-Myc antibodies as probes. Total RNA was also isolated from each gradient fraction by 1% SDS-treatment at 65°C, followed by phenol/chloroform extraction and ethanol precipitation. RNA was analyzed by 1% agarose gel electrophoresis to determine the presence of 28S and 18S rRNAs.

Cell-free extracts of a strain of S. pombe expressing N-terminally Myc-tagged Int6 were subjected to sucrose gradient centrifugation, and the gradient fractions were assayed for Int6 by Western blot analysis. A 43S preinitiation complex (40S.eIF3. Met-tRNA_f.eIF2.GTP), formed by incubating 40S ribosomal subunits with purified mammalian eIF3 and [³⁵S]Met-tRNA_f.eIF2.GTP ternary complex, was also analyzed in a parallel gradient tube (Figure 4A). Under these conditions, [³⁵S]Met-tRNA_f bound to 40S ribosomes sedimented with the 40S particles and thus defined the position of 40S ribosomal subunits. Western blot analysis of each gradient fraction showed that Myc-tagged fission yeast Int6 and mammalian Int6 sedimented at a position where the 43S preinitiation complex also sedimented (Figure 4B, upper panel). It should be noted that, in agreement with the results published previously (Chaudhuri et al., 1999), when the gradient fractions from the 43S preinitiation complex were analyzed by Western blotting using total anti-mammalian eIF3 antibodies (Chaudhuri et al., 1997a) as probes, all the mammalian eIF3 subunits cosedimented with the 43S preinitiation complex (data not shown). Western blot analysis of gradient fractions using antimammalian INT6 antibodies showed that INT6 also sedimented with the 43S preinitiation complex (Figure 4A, lower panel). Further confirmation that fission yeast Int6 sedimented with the 40S particles came from the observation that the Int6 containing fractions contained 18S rRNA, a known constituent of 40S ribosomal subunits (Figure 4B, lower panel). It should be noted that a portion of fission yeast Int6 sedimented in fractions lighter than 40S particles. The possibility exists that N-terminally tagged Int6 has a weaker association with the other subunits of eIF3. Alternatively, association of fission yeast Int6 with the other eIF3 subunits is inherently weaker than its mammalian counterpart.

Int6 Is Not Essential for Translation Initiation

The presence of INT6 in highly purified eIF3 preparations prompted us to examine the effect of deletion of the int6⁺ gene on translation of mRNA in S. pombe cells. For this purpose, exponentially growing cultures of int6 and a wild-type strain were pulsed with [³⁵S]methionine, and the rate of protein synthesis was measured over a 30-min time period. The rate of protein synthesis, measured by incorporation of [³⁵S]methionine into polypeptide chains in int6 cells, was ~ 30-40% slower than the rate of wild-type cells (Figure 5, panel A). This slower rate of methionine incorporation into cellular proteins in int6 cells is in keeping with the slower growth rate of int6 cells as compared with the wild-type strain.

View larger version (22K): [in this window] [in a new window]   Figure 5.   (A) In vivo incorporation of [35S]methionine in wild-type and int6 fission yeast cells. Total incorporation of [35S]methionine into proteins in a wild-type or int6 strain was determined. (B) Analysis of polysomes. Cell lysates were prepared from exponentially growing cultures of a wild-type strain (left) or a int6 strain (right) in YEA medium at 32°C and subjected to 7-47% (wt/vol) sucrose gradient as described in MATERIALS AND METHODS. Each gradient was fractionated in an ISCO gradient fractionator, and the absorbance profile at 254 nm was analyzed in an ISCO UA-5 absorbance monitor.

To investigate whether the 30-40% decrease in the incorporation of [³⁵S]methionine was due to slight defects in the initiation phase of protein synthesis, we examined the polyribosome profile of exponentially growing cultures of int6 and control wild-type cells. It is now well established that, when initiation of translation is slowed or blocked, ribosomes already bound to mRNAs complete translation but, after their release from mRNAs, do not reinitiate the translation process. This leads to a diminution of the size of polysomes and accumulation of 80S ribosomes. When extracts prepared from exponentially growing cultures of int6 and wild-type control cells were analyzed by sucrose gradient centrifugation, the polyribosome content of both strains showed similar profiles (Figure 5, panel B). However, a slight decrease in the polysome content and a slight increase in free 80S ribosomes were observed for int6 cells. These results may indicate a small effect of Int6 protein in initiation of translation but rule out that Int6 is essential for the initiation process in global protein synthesis.

Int6 Is a Cytoplasmic Protein

Using a strain expressing Int6 tagged with GFP, we determined the cellular localization of the protein. The majority of Int6-GFP was found in cytoplasm. This localization and the intensity of the signal seemed to be homogenous when an asynchronous culture was examined (Figure 6), suggesting that Int6 does not change its level and localization in a manner depending on progression of the cell cycle.

View larger version (47K): [in this window] [in a new window]   Figure 6.   Fission yeast Int6 is localized in the cytoplasm. Fission yeast cells expressing C-terminally GFP-tagged Int6 were grown on YEAU plates along with wild-type cells. Cells were stained with DAPI and were observed under CCD camera either in GFP channel or in DAPI channel. The inset shows the background fluorescence in GFP channel with the wild-type control.

int6 Is Hypersensitive to Caffeine

We tested the int6 strain for sensitivity to a variety of stress conditions such as treatment with UV light, caffeine, thiabendazole, and incubation at 36°C or 20°C. Among these conditions it was observed that int6 is hypersensitive to caffeine. In the YEA medium containing 10 mM caffeine, a wild-type fission yeast strain grew slowly (Figure 7B). It was also apparent that most of the cells have multiple septa (Figure 8C). The phenotype induced by the drug would suggest that the drug might have an inhibitory effect on maintenance of integrity of cell wall/membrane, or on cytokinesis. The caffeine-induced phenotype was much more pronounced in the int6 strain. The int6 cells were unable to form colonies on YEA plates containing 10 mM caffeine (Figure 7B). Microscopic observation revealed that these cells were elongated and lysed (Figures 7A, 8D and 8F). Because int6 cells lysed in the caffeine containing medium, we speculated that it could become resistant to the drug if the medium contained an osmotic stabilizer such as sorbitol. Indeed, if sorbitol was added to the medium, int6 strain could form colonies in the presence of 10 mM caffeine at a rate similar to that of the wild-type strain (Figure 7B). Although the drug could have multiple cellular targets and cause various phenotypes, the suppression by sorbitol would indicate that the hypersensitivity to the drug is most likely due to a defect in cell wall/membrane caused by lack of Int6 protein.

View larger version (40K): [in this window] [in a new window]   Figure 7.   Effect of caffeine on int6 cells. (A) A histogram of distribution of cell sizes for WT and int6 cells in YEA media lacking or containing 10 mM caffeine grown at 32°C. (B) Growth of WT or int6 cells on YEA plates containing 10 mM caffeine in presence or absence of 1.2 M sorbitol.

View larger version (118K): [in this window] [in a new window]   Figure 8.   int6 cells are sensitive to caffeine. The wild-type strain (A, C, and E) or int6 (B, D, and F) were grown under the following conditions; YEA medium (A and B), YEA + 10 mM caffeine (C to F). A to D were examined under a fluorescence microscope, while E and F were examined by conventional light microscopy.

A similar phenotype, i.e., cell elongation and lysis with multiple septation in the presence of caffeine, has previously been reported for MAPK cascade mutants and the function of the MAPK cascade in maintenance of cell wall integrity has been proposed (Loewith et al., 2000). It has also been shown that these deletion strains are sensitive to high concentration (1.2 M) of KCl in the medium. We observed that int6 cells were likewise sensitive to 1.2 M KCl concentration in the medium. Whereas a wild-type strain could form colonies on YEA plates containing 1.2 M KCl, the int6 strain failed to do so (our unpublished results).

It has been reported that the mutants of the MAP kinase cascade form a defective cell wall. Treatment with a cell wall-digesting enzyme, -glucanase, results in cell lysis more dramatically in the mutants than wild-type strains (Toda et al., 1996; Sengar et al., 1997)]. The observed similarity between int6 and the mutants of the MAP kinase cascade prompted us to test if int6 exhibits a similar defect in cell wall integrity. As shown in Figure 9A, deletion of the MAP kinase gene (pmk1) exhibited a high sensitivity to -glucanase, whereas int6 as well as a wild-type strain were resistant to the enzyme treatment. These results may suggest that the underlying mechanism of caffeine sensitivity conferred by int6⁺ deletion may not be the same as for the MAP kinase cascade mutants.

View larger version (29K): [in this window] [in a new window]   Figure 9.   (A) int6 cells are not hypersensitive to -glucanase treatment. (B) Failure of int6 cells to recover from a saturated culture. A wild-type strain (upper panel) or int6 (lower panel) cells were grown to either logarithmic phase (left) or saturated phase (right) in YEA medium at 32°C. Cells were serially diluted 5-fold each time, spotted on YEA plates, and incubated for 3 days at 32°C.

int6 Cells Fail to Recover from Saturated Growth

We also observed that the int6 cells were unable to recover from growth saturation. When they were transferred from an actively growing state to fresh YEA medium, they formed colonies at a slightly reduced rate (Figure 9B) compared with wild-type cells. In contrast, when the int6 cells were grown to a saturated state and transferred to fresh YEA medium, a major fraction of the cells were unable to form colonies (Figure 9B). Microscopic observation revealed that the int6 cells transferred from a saturated state were rounded, a phenotype similar to that of starved cells (data not shown). Due to presumed defects in maintenance of cell wall/membrane integrity, the int6 cells may be defective in their ability to uptake nutrients necessary for recovery from a saturated state.

Negative Effect of a Truncated Int6

In most cases, tumor-causing integration of MMTV genome in the mouse Int6 locus produces different forms of C-terminally truncated INT6 protein. It was therefore hypothesized that overexpression of these C-terminally truncated forms of INT6 driven by viral LTR causes tumors in mice (Marchetti et al., 1995). We examined whether overexpression of a C-terminally truncated Int6 in fission yeast causes any aberrant phenotype. For this purpose, we transformed a wild-type strain with pREP41-C50 INT6 plasmid that would allow overexpression of a truncated Int6 protein lacking the C-terminal 50 amino acids (designated C50). When the transformants were grown in the presence of 5 mM caffeine in PM medium, the cells showed a highly elongated shape (Figure 10). Transformants with a control vector or a plasmid that overexpresses the full-length int6⁺ gene did not show any apparent phenotypes under the same conditions (Figure 10). It should be noted that overexpression of C50 causes a phenotype which is slightly different from that of int6. First, the cells expressing C50 exhibited cell elongation at a lower concentration of the drug. Second, the elongated cells in the medium containing caffeine did not lyse. Overexpression of C50 appeared to affect the function of Int6 partially (see DISCUSSION). When C50 was expressed from a single copy gene integrated in the genome, it did not cause the phenotype (our unpublished results).

View larger version (125K): [in this window] [in a new window]   Figure 10.   Negative effect of C50-Int6. Wild-type cells were transformed with either empty pREP41 vector (VC) or pREP41 containing either full-length int6+ gene (FL) or C50 mutant int6+ (C50) gene. Transformants were grown in PM medium for 12 h and were inoculated in PM medium containing 5 mM caffeine. Inset: Occasionally, some cells harboring pREP41-C50 INT6 grew unusually long in caffeine containing medium.

Defect in Meiosis

In the course of strain construction, we noticed that cross between int6 and int6 produced incomplete tetrads frequently. As shown in Figure 11, 30-40% of asci of the int6 homozygous cross contained less than four spores. The defect was suppressed to a large extent when int6 was crossed to the wild-type strain.

View larger version (88K): [in this window] [in a new window]   Figure 11.   Defective meiosis in int6. Crosses between wild-type strains (WT-WT) or int6 strains (int6 -int6) were observed under a conventional light microscope.

	DICUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DICUSSION REFERENCES

Fission Yeast int6⁺ Gene

In this study we have characterized a fission yeast gene (int6⁺) that is 43% identical to the mammalian INT6 gene. The complementation of deletion of int6⁺ (int6) by the human Int6 ORF strongly suggests that they are functionally homologous.

Our interest in the Int6 protein stems from our effort to understand the process of eukaryotic translation initiation and the central role the multi-subunit initiation factor eIF3 plays in this process. Asano et al. (1997) reported INT6 as one of the 9-10 polypeptide subunits comprising purified mammalian eIF3. Although in recent years, the molecular genetic techniques in the budding yeast S. cerevisiae have been very useful to understand the function of translation initiation factors in vivo, the genome of budding yeast does not encode any genes related to INT6 (Table 1). The fission yeast model system thus offers a unique opportunity to study the functions of Int6 protein by genetic approaches. Assuming that Int6 plays a role with eIF3, fission yeast appears to share a common mechanism for translation initiation and its regulation in mammalian cells.

Association of Int6 with 40S Subunits

The question naturally arises is whether Int6 is a subunit of fission yeast eIF3. To answer this question, it is necessary to isolate and characterize eIF3 from fission yeast. This has not yet been achieved. However, an important property of mammalian eIF3 is its association with 40S subunits in cell-free extracts (Smith and Henshaw, 1975; Thompson et al., 1977). Therefore, if Int6 associates with eIF3, it would remain bound to 40S particles. This work as well as work from Norbury's laboratory has demonstrated that in cell-free extracts of fission yeast, Int6 cosediments with the 40S particles. More directly, fission yeast Int6 can be coimmunoprecipitated with a major component of eIF3, Sum1 (Dr. Chris Norbury, Imperial Cancer Research Fund, University of Oxford, Oxford, United Kingdom; personal communication). These observations suggest that the fission yeast Int6 is either associated with the core subunits of fission yeast eIF3 by protein-protein interaction or the polypeptide is a bona-fide subunit of eIF3. It should however, be noted that when we purified eIF3 from rabbit reticulocyte lysates, INT6 was found not to be stoichiometric with the other subunits of eIF3 (Figure 4A). Therefore, the possibility also exists that INT6 associates only with a subset of rabbit eIF3.

Role of int6⁺ in Translation Initiation

We show that fission yeast cells lacking Int6 (int6) support protein synthesis well exhibiting polysome-ribosome profiles similar to wild-type cells. If Int6 played an essential role in initiation of global protein synthesis, its absence in yeast cells would have caused extensive breakdown of polysomes with simultaneous increase in free 80S ribosomes as was observed in S. cerevisiae cells depleted of any essential translation initiation factor (Maiti and Maitra, 1997). These observations argue against the idea that Int6 is essential for initiation of global protein synthesis. It should, however, be emphasized that int6 strain showed ~ 30-40% slower rate of [³⁵S]methionine incorporation as compared with a wild-type strain. In view of our observation that the int6⁺ gene is required for maintenance of the integrity of cell wall/membrane, the possibility exists that the lower rate of [³⁵S]methionine incorporation into int6 yeast cells is a reflection of poorer uptake of amino acids into int6 yeast cells as compared with int6⁺ cells. Alternatively, it remains possible that Int6 is required either for translation of a small subset of mRNAs or for optimal efficiency of the rate of initiation of translation of all mRNAs. Analysis of phenotypes conferred by int6 has indicated that int6 exhibits a defect in the integrity of cell wall/membrane. The int6 phenotype would support a model in which fission yeast Int6 regulates translation initiation of a subset of proteins that are required for maintenance of the integrity of cell wall/membrane.

Caffeine Hypersensitivity and Other Phenotypes of int6

One of the apparent phenotypes conferred by int6 is the hypersensitivity to caffeine. Although the drug probably targets multiple cellular components and causes various phenotypes, it has been reported that the primary effect of caffeine on the fission yeast cell cycle is an inhibition of cytokinesis (Kumada et al., 1996). Consistent with this previous study, we have observed an inhibitory effect of the drug on cytokinesis in the wild-type strain. The caffeine-induced phenotype is much more prominent in the int6 strain. The hypersensitivity is suppressed by addition of sorbitol to the media. The suppression strongly implies that int6 loses its viability due to osmotic stress caused by the drug. Other negative effects that could be caused by the drug would not directly result in the hypersensitivity. int6 also confers a slow growth rate, sensitivity to KCl, inability to recover from a saturated state and a defect in meiosis. We speculate that int6 may form a defective cell wall/membrane which causes the phenotypes in the multiple biological processes.

It has recently been reported that the fission yeast MAP kinase cascade plays a role in the maintenance of cell wall/membrane (Loewith et al., 2000; Sugiura et al., 1999; Toda et al., 1996). Three kinases, namely Mkh1, Pek1/Skh1, and Pmk1/Spm1, form a sequential signaling cascade (Loewith et al., 2000; Sugiura et al., 1999; Toda et al., 1996). Deletion of any one of these genes causes phenotypes similar to those of int6. First, deletion causes cell elongation in presence of caffeine (Loewith et al., 2000). Second, it causes inhibition of growth in media containing KCl (Loewith et al., 2000). Finally, deletion of pmk1⁺ causes a defect in recovery from a saturated state (Toda et al., 1996). On the other hand, while the mutants in the MAP kinase cascade form cell wall that is sensitive to treatment with -glucanase, cell wall of int6 is resistant to this treatment. Likewise, the mutants in the MAP kinase cascade are not defective in meiosis. These results would suggest a partial overlap in the function of Int6 and the MAP kinase cascade. Further study would shed light on the functional link between the MAP kinase cascade and the Int6 protein.

Effect of a Truncated int6 Protein

Integration of MMTV in the mouse Int6 locus results in expression of a truncated form of the INT6 protein (Marchetti et al., 1995). At present, it is not clear how a truncated INT6 protein induces tumor formation in the mouse mammary epithelial tissue. MMTV may simply disrupt the biological activity of the affected Int6 gene causing a reduction in the dose of the normal gene product. Alternatively, a truncated form of INT6 may have a dominant negative or active effect leading to hyperplasia in the mouse mammary gland. In this study, we have tested if a truncated int6⁺ gene has a negative effect in fission yeast. For this purpose, we have expressed several versions of truncated Int6 proteins in the wild-type strain. One such construct, which overexpresses a truncated Int6 protein lacking the C-terminal 50 amino acids (designated C50), causes a phenotype. In presence of caffeine, wild-type cells expressing C50-Int6 protein exhibit a highly elongated cell shape. Interestingly, overexpression of C50 can cause the phenotype in a medium containing 5 mM caffeine, whereas int6 strain expressed the phenotype at a higher concentration of the drug,10 mM. In addition, C50 does not cause other phenotypes observed in int6. When C50 is overexpressed, it may replace the wild-type Int6 and abrogate the function.

Based on the assumption that Int6 may play a role with eIF3 in regulation of translation initiation of a subset of proteins, we would interpret the discrepancy between int6 and C50 as follows: While eIF3 without Int6 fails to initiate translation of all proteins in the subset, eIF3 with C50 can regulate translation initiation of a limited number of proteins in the subset. The difference in protein composition may lead to the discrepancy in expression of the phenotype. It should also be noted that overexpression of the human Int6 can suppress the slow growth of int6, but not the hypersensitivity to caffeine. Thus, the human Int6 in fission yeast is partially functional. A lower level of expression of C50 did not cause any phenotype, suggesting it is not dominant negative.

Implication in Tumor Biology

The Int6 gene is a common integration site for the mouse mammary tumor virus (MMTV). A human homologue of Int6 is located on chromosome region 8q22-q23. Examination of this allele revealed loss of heterozygosity (LOH) in 11 of 39 (28%) of the tumor samples. Because single-strand conformation and hybrid mismatch analysis of the remaining allele in these tumor DNAs failed to detect any mutations, it has been concluded that the target gene for LOH must be closely linked to Int6 (Miyazaki et al., 1997). Thus, it is likely that mutations in the human Int6 also cause tumors.

In this study we have shown that the fission yeast int6⁺ is required for maintenance of the integrity of cell wall/membrane. Assuming that the human INT6 functions in a similar biological process, we hypothesize that mutations in the mammalian Int6 gene cause an abnormal membrane structure, which might result in defective endocytosis. Ligand binding followed by its internalization is often used as a signal, which regulates cell growth. The Int6-induced tumor cells may not be able to correctly process such a signal, which negatively regulates cell division. In higher eukaryotes, a number of genes required for endocytosis has been characterized as tumor suppressor genes or oncogenes (Floyd and De Camilli, 1998). It should be noted that the fission yeast int6 cells fail to recover from a saturated state and exhibit a smaller, round cell shape, a characteristic phenotype of starved cells. This may imply that int6 cell is defective in endocytosis and cannot uptake components necessary for recovery from a saturated state.