Translational control of Cdc25 - CiteSeerX

3137

Journal of Cell Science 112, 3137-3146 (1999) Printed in Great Britain © The Company of Biologists Limited 1999 JCS0315

Translational control of the Cdc25 cell cycle phosphatase: a molecular mechanism coupling mitosis to cell growth Rafael R. Daga and Juan Jimenez* Unidad de Genética, Facultad de Ciencias, Universidad de Málaga, Campus Universitario de Teatinos, 29071 Málaga, Spain *Author for correspondence (e-mail: [email protected])

Accepted 26 June; published on WWW 26 August 1999

SUMMARY The eukaryotic translation initiation factor 4A (eIF4A) is an RNA helicase required for translation initiation of eukaryotic mRNAs. By engineering fission yeast mutants with diminished eIF4A activity, we have found that translation of cdc25 mRNAs (a dosage-dependent activator of mitosis in all eukaryotic cells) is particularly sensitive to limitations of protein synthesis mediated by limited eIF4A activity. Genetic and biochemical analysis indicated that a rate-limited translation initiation of cdc25 mRNAs, exerted throughout its unusual 5′′ untranslated leader, acts as a molecular sensor to ensure that a minimum cell mass

(protein synthesis) is attained before mitosis occurs. The Cdc13 cyclin B is also among the limited pool of proteins whose translation is sensitive to reduced translation initiation activity. Interestingly, the 5′′ leader sequences of cdc25 and cdc13 mRNAs have conserved features which are unusual in other yeast mRNAs, suggesting that common mechanisms operate in the expression of these two key mitotic activators at the translational level.

INTRODUCTION

a relationship between initiation of polypeptide chains and cell cycle regulation (Brenner et al., 1988; Verlhac et al., 1997). Interestingly, cells with reduced translation initiation activity can greatly exceed the critical size that normally characterises the G1/S transition (Hanic-Joyce et al., 1987), suggesting that the relationship between the initiation of protein synthesis and the G1/S control could play a role in the mechanism monitoring cell size. In budding yeasts, the G1 cyclins Cln1, Cln2 and Cln3 are rate-limiting activators of Cdc2 (Cdc28) regulating the time spent in G1 (Cross, 1998; Nash et al., 1988). Among them, Cln3 is the first step in this cascade of cyclins regulating Start (Nash et al., 1988; Dirick et al., 1995), and its synthesis is specially sensitive to rapamycine-mediated inhibition of translation (Barbet et al., 1996), making the Cln3 a good candidate to be involved in a mechanism coupling cell growth with cell division throughout a rate-limited translation of its mRNA (Polymenis and Schmidt, 1997; Hall et al., 1998). These studies in budding yeast highlight a mechanism operating at G1/S size control, but mechanisms regulating the mitotic size control remain yet unsolved. Clues for the mitotic size control arise from genetic studies carried out in the fission yeast Schizosaccharomyces pombe. In this organism entry into mitosis is triggered by the activation of the Cdc2/Cdc13 cyclin B complex and takes place when cells attain a critical size (Nurse, 1975, 1990). Certain mutations in the Cdc2 protein, or in regulators of its kinase activity, produce cells dividing at a smaller cell size, indicating that these regulators (namely wee1 and cdc25) are likely to be involved in the control that

Cells double their mass during each division cycle, maintaining a constant size as they proliferate (Mitchinson, 1971). In eukaryotes, cell division is controlled just before the G1/S transition (a point called ‘Start’ in yeast and ‘Restriction point’ in mammalian cells) and at the G2/M transition. In yeasts, both transitions are controlled by activation of the universal Cdc2 protein kinase in association with specific regulators known as cyclins. In mitosis, activation of the Cdc2/cyclin complex is achieved by the Cdc25 cell cycle phosphatase, which counteracts the inhibitory activity of the Wee1 kinase (Nurse, 1990). In higher eukaryotes, a number of other cyclin dependent kinases (Cdks) belonging to the Cdc2-family are involved in the regulation of the G1/S transition and progression through S phase (Reed et al., 1994). While mechanisms regulating cell cycle progression are well established, little is known about how cell growth (increase in cell size or cell mass) is coordinated with the cell cycle regulatory machinery, as well as the nature of the molecular system which senses ‘size or mass’ in the eukaryotic cell (Sveiczer et al., 1996). Studies carried out in the budding yeast Saccharomyces cerevisiae show that these cells move quickly through G1 in rich medium in which they grow fast, and slowly in poor medium where growth rate is slower, indicating that this organism coordinates growth and division by regulating the length of time spent in G1 (Jagadish and Carter, 1977; Johnston et al., 1977). The G1 arrest seen in S. cerevisiae mutants deficient for the translation initiation step establishes

Key words: Size control, Mitosis, Cdc25, Cdc13, Schizosaccharomyces pombe, eIF4A, Translation initiation

3138 R. R. Daga and J. Jimenez coordinates mitosis with the attainment of a particular cell size (Russell and Nurse, 1986, 1987). Here we report that expression of the Cdc25 cell cycle phosphatase, a dosage-dependent inducer of mitosis in all eukaryotic cells (Russell and Nurse, 1986), is extremely sensitive to diminished translation initiation activity mediated by reduced eIF4A function. This result establishes a novel linkage between the translation initiation machinery and the mitotic control that may cast light on the mechanism coupling cell growth and mitosis in eukaryotic cells. The level of Cdc13 (cyclin B) is also influenced by limited eIF4A activity. Implications of a common mechanism (translational control) monitoring size at the G1/S and the G2/M transitions in eukaryotic cells are discussed.

MATERIALS AND METHODS Genetic and general procedures All S. pombe strains used in this work are isogenic to the 972 h− wild-type strain. Thermosensitive mutants cdc2-33, cdc25-22 and wee1-50, strains deleted for wee1 (wee1::ura4)(∆wee1), cdc25 (cdc25::ura4, viable in a wee1-50 background at 35°C)(∆cdc25), and overproducing cdc25 with the constitutive adh promoter (adh:cdc25) have been described elsewhere (Nurse, 1975; Russell and Nurse, 1986, 1987). Double mutants were constructed by tetrad dissection. YES and synthetic EMM media, genetic procedures, northern and Southern blots, flow cytometric analysis, and microtubule staining (using anti-α-tubulin) were carried out as described (Alfa et al., 1993; Moreno et al., 1991). DNA sequencing, cloning and subcloning was carried out following described procedures (Sambrook et al., 1989). Isolation of tif1 A full length cDNA corresponding to a gene coding for the eIF4A protein (tif1) in fission yeast was fortuitously isolated by colony hybridisation. tif1 lies contiguous to etd1. The genomic DNA fragment isolated by complementation of etd1 mutants (Jimenez and Oballe, 1994) was used to probe a cDNA library, and tif1 and etd1 cDNAs were isolated. Using this tif1 cDNA fragment as a probe, tif1 was physically mapped to a single locus in chromosome II close to cdc10. These tif1 sequence data have been submitted to the GenBank database under accession number L40627. Disruption of tif1 The ura4 gene was inserted into the ClaI site of the tif1 cDNA. A DNA fragment containing the ura4 gene flanked by tif1-coding sequences was used to transform an h+/h− ade6-M210/ade6-M216 ura4-d18/ura4-d18 leu1-32/leu1-32 diploid strain, and ura+ colonies selected. The disruption (tif1:ura4) in one of the chromosomes was confirmed by Southern blot and tetrads analysis (Fig. 1A). Conditional expression of tif1 The tif1 cDNA was subcloned in plasmid pREP3X subjected to the expression of the nmt1 promoter (Maundrell, 1993) and the nmt:tif1 construction inserted into the pJK148 integrative plasmid. Diploid cells heterozygous for the tif1:ura4 disruption bearing a single integrated copy of the pJK-nmt:tif1 construction in the leu1 gene were selected; leu+ ura+ spores derived from this diploid had the endogenous tif1 gene disrupted, and expressed the tif1 cDNA under the nmt1 promoter (the tif1:ura4 nmt:tif1 strain) (Fig. 1B). Repressed cells were analysed 8-10 hours after thiamine was added to EMM liquid cultures incubated at the indicated temperature. Growth was assayed by replica-plating to EMM-thiamine and to EMM+ thiamine plates. Plates were incubated for 24 hours at the indicated temperature.

Dominant negative tif1-mutants Deletions of the tif1 cDNA were engineered by using internal restriction sites, and/or by subcloning the different deleted fragments obtained by exonuclease III treatment. The constructions were designed to generate a stop codon at the C terminus of the truncated variants. The tif1-d3 construction generated an additional glycine next to the Asn343 C-terminal residue. This truncated protein lacks 56 Cterminal amino acids, essential for ATP hydrolysis and RNA binding activity (Pause and Sonenberg, 1992). Deletions were assayed under the nmt1 promoter in plasmid pREP3X. The nmt:tif1-d3 construction from pREP3X was inserted into the pJK148 integrative plasmid, and stable leu+ transformants selected. Colonies containing four or five integrated copies (as shown by Southern blot and densitometric measurement) displayed a uniform enlarged cell morphology (cdc− phenotype). Lower tif1-d3 gene dosage had no significant effects, and more than five copies per cell resembled tif1-deficient cells. Strains with one, and with four and six tandem copies of the nmt:tif1-d3 construction (1×, 4× and 6× nmt:tif1-d3) integrated at the leu1 gene were selected for further experiments. In our experimental conditions, mRNA from tif1-d3 peaked at about 16 hours after derepression (− thiamine). In the 4× nmt:tif1-d3 strain cells were completely arrested in cultures kept from 12 to 24 hours without thiamine at the temperature range tested (from 25°C to 35°C). Samples of cells derepressed for 14 to 18 hours were routinely used. Double mutants were obtained by tetrad dissection and the construction was verified by Southern blot. Growth in these strains was monitored by replicaplating to EMM-thiamine and to EMM+thiamine plates. Plates were incubated for 48 hours at the indicated temperature. Expression analysis in tif1-deficient cells Cells G2-arrested by the dominant negative tif1-d3 were obtaining as described above and the arrest monitored by visual inspection under the microscope. Once arrested, total RNA and protein extracts were prepared. Samples of 10 µg total RNA were used for northern blot analysis. For western blot, samples of 25 µg of total protein extract (100 µg in tif1-deficient experiments) were run on SDS-PAGE minigel, transferred to nitrocellulose and probed with affinity purified antiCdc25, anti-Cdc13 and anti-Cdc2 polyclonal antibodies. Wild-type, wee1-50 cdc25::ura4 mutant and cdc2-33-arrested cells (incubated for 6 hours at 35°C) were used as controls. Incorporation of [35S]methionine For pulse-labelling experiments, 500 µCi of [35S]methionine was added to 10 ml of each culture, cells incubated for 15 minutes at 25°C, and labelled proteins measured, resolved in acrylamide gel and visualised by autoradiography (Barnes et al., 1993). Polysomal fractionation and analysis of associated mRNAs For analysis of polysomal profiles, protein synthesis was instantaneously arrested with cycloheximide (100 µg/ml) for 10 minutes. Cells were collected, washed, and kept at −80°C until analysed. Cell lysis and size fractionation of polysomes through sucrose gradients (10% to 45% sucrose) were performed as described (Sagliocco et al., 1994). Analysis of associated mRNAs was performed by semiquantitative RT-PCR (one step RT-PCR system, Promega). Oligonucleotide primers used for cdc25 were 5′GATTCTCCTACTGCCATG3′ and 5′ACTACCGCTTACAGAACATGC3′ and for cdc2 were 5′ACCAGCTAGTGA-ACGGTG3′ and 5′GGCAGGGTCATAAACAAG3′. PCR products were resolved by agarose gel electrophoresis and detected by blot hybridisation with cdc25 and cdc2 32P-labelled cDNA probes. Deletion analysis of the 5´UTR Full-length cdc25 and cdc13 cDNAs (with the expected size in accordance to the mRNA size for each gene) were isolated by colony hybridisation and the isolated cDNAs sequenced. These sequences

Translational control of Cdc25 3139 started 830 and 697 nucleotides upstream of the cdc25 and cdc13 AUG codon, respectively, indicating that transcription starts at least at this position for each gene. Oligonucleotides were designed to obtain by PCR a HindIII-XbaI fragment from the promoter region of cdc25 (5′GGTGTCATCTTTGGAACTGG3′ and 5′GCGCTCTAGATAGTTGAGGG3′) and another XbaI-XbaI from the cdc25 coding region (5′CCTCTAGAAAAATGGATTCTCCGC3′ and 5′ACTACCGCTTACAGAACATGC3′). Ligation of these two fragments regenerated a cdc25 gene with a new XbaI site in which most of this 5´UTR was removed (cdc25-d1 mutation) (see Fig. 7A). The ligated fragment was sequenced to confirm the construction, and used to transform a cdc2522 strain. Transformants were selected at 35°C, and a gene replacement occurred in about half of the isolated colonies. For over-expression analysis, the 5´ region of a full-length cDNA was replaced by the equivalent fragment corresponding to the cdc25-d1 mutant, engineering in this way a cdc25 cDNA with the short 5´UTR created in the cdc25-d1 mutant. Both full-length and cdc25-d1 cDNAs were subjected to the expression of the nmt1 promoter and the constructions were integrated in a single copy at the leu1 gene.

Therefore, this preliminary observation suggests that limited eIF4A activity could specifically delay cell cycle progression. Interestingly, this cell cycle delay was exacerbated in a cdc2522 background, yielding a synthetically lethal cdc− phenotype (Fig. 2). The Cdc25 cell cycle phosphatase links the translation initiation activity to the Cdc2-mitotic machinery The eukaryotic eIF4A functions in complexes with other translation initiation factors (Fuller-Pace, 1994; Sachs et al.,

RESULTS

tif1 codes for translation initiation factor 4A in S. pombe We isolated an S. pombe cDNA encoding a putative translation initiation factor 4A (eIF4A) (see Materials and Methods). This cDNA of S. pombe has been also identified by others (Fischli et al., 1996). The eukaryotic eIF4A belongs to the D-E-A-D family of double-stranded RNA helicases, and its activity is essential for melting secondary structures in the 5′ untranslated region of eukaryotic mRNAs, to facilitate ribosome binding and initiation of translation (Fuller-Pace, 1994; Pause and Sonenberg, 1992; Sachs et al., 1997). In budding yeast, this translation initiation factor is encoded by two genes, TIF1 and TIF2. Only the simultaneous inactivation of both genes results in the arrest of protein synthesis (Linder and Slonimski, 1989). In S. pombe, disruption analysis indicated that the identified gene (named tif1) is essential (Fig. 1A), and physical mapping showed that this gene is non-redundant in the S. pombe genome. To analyse the function of tif1 in proliferating S. pombe cells we engineered a strain in which tif1 was conditionally expressed. To this end, the tif1 cDNA was subjected to the expression of the thiamine-repressible nmt1 promoter (Maundrell, 1993) in a strain bearing the disrupted endogenous tif1 gene (the tif1:ura4 nmt:tif1 strain). As shown in Fig. 1B, when tif1 is expressed (in the absence of thiamine) cells are indistinguishable from wild type. In thiamine-containing medium expression of tif1 is repressed (mRNA was undetectable by northern blot analysis) and neither cell growth nor cell division occurred at 35°C. In these conditions, the loss of polysomes and the accumulation of 80S ribosomes (see Fig. 1B) determined by size fractionation of ribosomes indicated that tif1 plays an essential role in the initiation of protein synthesis (Zhong and Arndt, 1993), and consequently that this gene encodes the translation initiation factor 4A in S. pombe cells, as predicted from DNA sequence analysis (Fischli et al., 1996). At 25°C the leaky expression of the repressed nmt1 promoter (Maundrell, 1993) still provided sufficient eIF4A function to allow cell proliferation (Fig. 2), but these cells growing with reduced expression of tif1 were slightly elongated. In S. pombe, elongated cells are a prototype phenotype to identify cdc genes, coding for cell division control functions (Nurse et al., 1976).

Fig. 1. tif1 is an essential gene required for translation initiation. (A) Meiotic segregation (tetrads) of diploids heterozygous for a tif1 disruption (tif1:ura4). (B) Growth, cells (DAPI-stained) and polysome profile of the tif1:ura4 nmt:tif1 strain with nmt1 driven expression of tif1 (northern blot) (−Thiamine or −T: expressed, +Thiamine or +T: repressed) at 35°C; analysis of polysomes was performed by size fractionation in sucrose gradient (sedimentation was from left to right) and measurement of optical density at 260 nm. EDTA (10 mM) was used in the sucrose gradient as a control to dissociate ribosomes (80S) into ribosomal subunits (40S and 60S) (dotted line).

3140 R. R. Daga and J. Jimenez

Fig. 3. Dominant negative effects caused by the tif1-d3 allele on cell cycle progression and translation initiation. Cells (DAPI-staining) and polysome profile of strains overexpressing the dominant negative tif1-d3 mutant (nmt1-driven in the absence of thiamine) at low (1× nmt:tif1-d3) and high (6x nmt:tif1-d3) gene dosage (see northern blot). The pleiotropic phenotype and the accumulation of monosomes in cells with high gene dosage of tif1-d3 resemble that of cells lacking tif1 function. Fig. 2. Synergy between reduced activity of the Cdc25 phosphatase and reduced expression of tif1 at 25°C. Growth and cells (DAPIstained) of the nmt:tif1 tif1:ura4 strain and the nmt:tif1 tif1:ura4 cdc25-22 strain. The strain with a fully active Cdc25 protein and tif1 expression (nmt:tif1 tif1:ura4, −Thiamine) grows normally and the cells of this strain divide as wild type. Strains with either a partial reduction on cdc25 activity (nmt:tif1 tif1:ura4 cdc25-22 −Thiamine) or a reduced expression of tif1 (nmt:tif1 tif1:ura4, +Thiamine) grow, but entry into mitosis is delayed in these cells (are slightly elongated). The combination of both defects is lethal (nmt:tif1 tif1:ura4 cdc25-22 +Thiamine), yielding very elongated cells with a prototype cdc− phenotype.

1997) and eIF4A mutant proteins have been identified as dominant negative variants, which efficiently inhibit translation initiation by binding competitively with eIF4F complexes (Pause et al., 1994), composed by the cap-binding factor eIF4E, protein eIF4G, and eIF4A (Fuller-Pace, 1994; Sachs et al., 1997). Searching for similar dominant negative mutants, we studied the effects that over-expression (nmt1 driven) of a number of tif1-deletions caused in wild-type cells. In this way, we identified a non-functional tif1 mutant (tif1-d3) producing a lethal phenotype when over-expressed. In agreement with a dominant negative effect caused by tif1-d3 in the initiation step, this tif1-d3 variant worked in a dosage-dependent manner. In strains with low tif1-d3 expression (1× nmt:tif1-d3) translation initiation was not affected and normal cell growth and cell division were observed (see polysome profile and cells in Fig. 3). A high level of tif1-d3 expression (6× nmt:tif1-d3) severely inhibited the initiation of protein synthesis and neither cell growth nor cell division took place in these cells (see Fig. 3).

In strains with an intermediate tif1-d3-dosage (4× nmt:tif1-d3) the expression of tif1-d3 also arrested cell proliferation but at these conditions cell growth occurred normally, yielding elongated cells identical to cells specifically affected in cdc genes (Nurse et al., 1976). According to polysome profile analysis the initiation step was only partially affected (Fig. 4A). Therefore, the slight reduction in translation initiation activity in these cells triggered a very efficient cell cycle block, suggesting that a mechanism exists that links the translation initiation with the machinery regulating cell cycle progression. This tif1-d3-induced cell cycle arrest occurred at a wide range of incubation temperatures (25°C, 30°C and 35°C), and was more pronounced than that observed in cells with limited tif1expression at 25°C (see Fig. 2), therefore, this strain (4× nmt:tif1-d3) was further used to analyse the mechanism by which the cell cycle control is linked to the translation initiation activity of the cell. These tif1-d3-arrested cells contained a single nucleus with 2C DNA content and interphase microtubular array (Fig. 4B), indicating that expression of tif1-d3 blocked the cell cycle specifically at the G2 phase. The main regulatory mechanism controlling the G2/M transition in S. pombe is driven by the activation of the cdc2-encoded protein kinase, which is regulated by the balance of specific protein kinases and phosphatases. The wee1 and mik1-encoded kinases inhibit Cdc2 activity by phosphorylation of its tyr15 residue (Russell and Nurse, 1987; Lundgren et al., 1991), and activation occurs through the antagonistic action of the cdc25-encoded phosphatase (Russell and Nurse, 1986). To determine if any of these molecular pathways transducts a tif1-induced signal to inhibit entry into

Translational control of Cdc25 3141 mitosis through cdc2, we made double mutants to produce a tif1d3 arrest in combination with different mutations in the control of Cdc2. Cells lacking both wee1 and cdc25 (in a wee1-50 cdc25::ura4 double mutant background at 35°C) bypassed this arrest, suggesting that the effect of tif1-d3 is linked to the cdc2 cell cycle control through one of these regulators. Early genetic experiments suggested a possible relationship between protein synthesis and the cdc25 cell cycle activator (Nurse and Thuriaux, 1984). As has been shown above, a slight reduction in cdc25 activity provided by the cdc25-22 ts allele at 25°C (Wu and Russell, 1997) yielded a synthetic cdc− lethal defect in cells growing with limited expression of tif1 (see Fig. 2). The cdc25-22 allele also exacerbated the cell cycle block caused by tif1-d3 (Fig. 5). This genetic analysis strongly

suggests that a partial reduction in translation initiation activity inhibits entry into mitosis by diminishing cdc25 function. In agreement with this hypothesis, cells over-producing cdc25 from the strong adh promoter (Russell and Nurse, 1986) overcome the tif1-d3-induced arrest, and the deletion of wee1, which suppresses the loss of cdc25 function (Russell and Nurse, 1987) equally suppressed this tif1-d3-arrest (see Fig. 5), suggesting that cdc25 is the main regulator causing a G2 arrest in response to a partial inhibition of the translation initiation activity. The ectopic expression of a Drosophila RNA helicase suppresses the lethal mitosis induced by overexpression of cdc25 in cells lacking wee1 activity (Warbrick and Glover, 1994). Since this helicase is highly homologous to the S. pombe eIF4A protein, a dominant negative effect of the Drosophila protein in the S. pombe translation initiation machinery (similar to that of tif1-d3) could suppress the excess of Cdc25 by selective inhibiting translation of cdc25 mRNAs, suggesting also a discriminatory effect on the synthesis of Cdc25 polypeptides. Translation of the cdc25 messenger, a sensor of translation initiation activity As assessed from the polysome profile describe above (Fig.

Fig. 4. Conditional phenotype conferred in the strain overexpressing tif1-d3 (see northern blot) at intermediate (4× nmt:tif1-d3) gene dosage (+ Thiamine or +T: tif1-d3 is expressed, −Thiamine or −T: tif1-d3 is repressed). (A) Growth, cell phenotype (DAPI-staining) and polysome profile of this strain incubated under tif1-d3 repression and under tif1-d3 expression at 25°C (similar results are obtained at 30°C and 35°C). (B) Flow cytometry analysis and microtubule localisation of tif1-d3-arrested cells (4× nmt:tif1-d3 without thiamine).

Fig. 5. Growth of the nmt:tif1-d3 (4x) strain in cdc25-22 or adh:cdc25 genetic backgrounds incubated on medium with and without thiamine at 25°C (permissive temperature for cdc25-22). Overproduction of Cdc25 (adh:cdc25) rescues the G2-arrest induced by tif1-d3. Cells (DAPI-staining) of these strains expressing tif1-d3 (−Thiamine) in cdc25-22 (the cell cycle block is exhacerbated) or adh:cdc25 (the G2-arrest is suppressed) backgrounds are shown (microphotographs). The deletion of wee1 (∆wee1) also rescues the tif1-d3 arrest. Cells of this nmt:tif1-d3 (4x) ∆wee1 strain incubated without thiamine are also shown.

3142 R. R. Daga and J. Jimenez

Fig. 6. Disproportionate effect caused by tif1-d3 in the translation of the cdc25 mRNA. (A) Electrophoretic resolution of pulse-labelled proteins of proliferating cells (+T) and G2arrested cells by over-expression of tif1-d3 (−T). The arrow indicates a specific band whose translation could be discriminated against under tif1-d3 expression. Relative molecular mass (kDa) are indicated. (B) Distribution of cdc25 mRNA and cdc2 mRNA (control) between subpolysomal (S) and polysomal (P) pooled fractions of cells with normal (+Thiamine or +T) or reduced (−Thiamine or −T) translation initiation activity caused by tif1d3 (shown in Fig. 4A). RNA was extracted from the entire pool and analysed by RT-PCR/blot hybridisation. (C) Levels of cdc25 and cdc2 mRNAs (northern blot) and Cdc2 and Cdc25 proteins (western blot) in the above cells with normal proliferation (+T) or arrested at the G2-phase by the partial inhibition in translation initiation caused by tif1-d3 (−T). Cells deleted for cdc25 (wee1-50 ∆cdc25 at 35°C) (c1), wild-type cells (wt) and G2-arrested cells by mean of the cdc2-33 allele (c2) were used as different controls. (D) Levels of Cdc13 and tubulin (western blot) in cells with normal proliferation (+T) or arrested at the G2-phase by tif1d3 (−T). (These experiments were carried out at 35°C, the restrictive temperature for wee1-50 and cdc2-33). (E) Schematic representation of common features found in the 5′ UTR of cdc25 and cdc13 mRNAs.

T

4A), the inhibition in initiation of translation is only partial in tif1-d3-arrested cells. Accordingly, a pulse-labelling experiment indicated that translation of nearly all of the 50-100 most abundant proteins was unaffected (Fig. 6A). However, this partial inhibition had a disproportionate effect in the translation of a limited number of proteins (see arrow in Fig. 6A). A similar result has been described in S. cerevisiae. In this yeast, the inhibition of translation initiation lead to a disproportionate decrease in the synthesis of specific proteins, including the Ssa1 and Ssa2 members of the Hsp70 family (Barnes et al., 1993) and the Cln3 cyclin (Polymenis and Schmidt, 1997). To test whether reduced tif1 activity diminished cdc25 function by reducing translation of this mitotic activator, we

analysed the distribution of cdc25 mRNAs between subpolysomal and polysomal fractions in proliferating cells (+Thiamine in Fig. 4) and in tif1-d3-arrested cells (−Thiamine in Fig. 4). As shown in Fig. 6B, the bulk of cdc25 mRNA was associated with polyribosomes in tif1-d3-repressed cells, indicating that cdc25 mRNAs are engaged in active protein synthesis at these conditions, however the majority of cdc25 mRNA was found in the monoribosomal fraction in G2-arrested cells by tif1-d3 expression. The fraction of polysome-associated mRNA encoding the reference Cdc2 protein remained unaltered, indicating that translation initiation of the cdc25 mRNA is highly sensitive to perturbations caused by the eIF4A mutant in the translation initiation step. In agreement with this observation, northern and western blot analysis demonstrated

Translational control of Cdc25 3143 cdc25 gene structure (wt)

A Hin d III

Hind III

ATG

OR F

5 ´ UTR

TATA

-2 0 1 0

-1 0 8 5

3 ´ UT R +2 0 1 3

+1

Xb a I

Hin d III

F

ccd c 2 5 -ORF

Xb a I

Xba I

Hind III

cd c25-ORF -

TATA XbaI

Hin d III TATA

Xb a I 3 ´ UTR

Hind III

c d c 2 5 -ORF

Xb a I 3 ´ UTR

cdc25-d1 deletion (d1)

Fig. 7. A role for the 5´UTR in the rate-limited translation of the cdc25 mRNA. (A) Schematic representation of the cdc25 gene structure, and the strategy followed to delete a 0.8 kb fragment of the 5´UTR of cdc25 (cdc25-d1 mutant). (B) Level of Cdc25 protein (western blot) produced by the cdc25-d1 mutant strain. Deletion was verified by Southern blot (HindIII-XbaI digestion) and production of a cdc25-d1 mRNA analysed by northern blot. cdc2 mRNA and Cdc2 protein were also analysed (controls). (C) Partial suppression of the tif1-d3-arrest by the cdc25-d1 mutant. Cells (DAPI-stained) of the nmt:tif1-d3 (4×) cdc25-d1 strain incubated in media with thiamine (normal protein initiation activity) and without thiamine (reduced translation initiation activity). (D) Cells (DAPI-stained) obtained from the nmt1-driven expression of a full-length cdc25 mRNA and of a cdc25-d1 mutant.

that G2-arrested cells by over-expression of tif1-d3 had normal levels of cdc2 mRNA and protein and contained abundant cdc25 mRNA too, however, the amount of Cdc25 protein dropped down drastically (−T in Fig. 6C). Translation initiation of tif1proficient cells either during active proliferation (+T and wt in Fig. 6C) or arrested in the equivalent G2 phase by using the cdc2-33 ts mutation (control c2 in Fig. 6C) efficiently accumulated Cdc25 as expected (Kovelman and Russell, 1996). In consequence, we conclude that the cdc25 mRNA belongs to a reduced set of mRNAs whose translation is largely dependent on tif1 activity; that is, translation initiation of cdc25 mRNA is rate-limited by the eIF4A function encoded by tif1, and therefore, the accumulation of Cdc25 could be a good sensor of the rate of translational activity of the cell. Western blot analysis confirmed that levels of other proteins remained unaltered in tif1-d3-arrested cells (actin, tubulin and Wee1 were assessed). Interestingly, the level of Cdc13 was significantly reduced in these cells, indicating that translation of this B-type cyclin is also sensitive to diminished eIF4A activity (see Fig. 6D). While overexpression of cdc25 rescued the tif1-d3-induced cell cycle block (Fig. 5), high levels of cdc13 expression did not suppress this arrest. Therefore, synthesis of Cdc13 polypeptides is also inhibited but Cdc25 is the only essential protein that restricts growth under these conditions. The intrinsic translation initiation efficiency of an mRNA depends on cis-acting elements within the 5′ leader sequence (Mathews et al., 1996). Strikingly, length and secondary structures of the 5′ leader region of both cdc25 and cdc13 mRNAs were extremely unusual. In addition, a detailed sequence analysis revealed some other salient features that cdc25 and cdc13 mRNAs have in common within this region. As schematised in Fig. 6E, the 5′ leader of both type of mRNAs contains repeats of the UUUCUUUU oligopyrimidine sequence, several small upstream open reading frames (uORFs), and a similar stem-loop structure, suggesting that common mechanisms operate in the expression of these two key mitotic activators at the translational level. A role for the rate-limited translation of cdc25 mRNAs in measuring a critical size As it has been previously shown, Cdc25 is the main polypeptide involved in the G2 arrest caused by a partial inhibition of the translation initiation activity (see Fig. 5). The deletion of a large region of this 5´UTR of cdc25 (cdc25-d1) (Fig. 7A) resulted in an increased level of Cdc25 protein (Fig. 7B), indicating that the 5´UTR of the cdc25 mRNA rate-limits the accumulation of Cdc25 during cell proliferation. This cdc25-d1 mutant partially rescued the tif1-d3 cell cycle arrest (Fig. 7C), suggesting that the rate-limiting effect exerted by the uncommon 5′ domain of the cdc25 mRNA takes play at the level of translation initiation. Cells of this cdc25-d1 mutant were slightly smaller than wild-type cells (see +Thiamine in Fig. 7C), suggesting that the 5´ leader of the cdc25 mRNA is required for the cell to trigger mitosis at the normal cell size. As shown in Fig. 7D, the role that this 5′ region plays in coordinating cell growth and mitosis is more obvious in cells overexpressing the cdc25-d1 mutant variant: while the over-expression of a cdc25 full-length cDNA produced viable wee cells, over-expression of the cdc25-d1 mutant (5´ UTR-deleted) was lethal, producing cells which entered into mitosis without sufficient cell mass.

3144 R. R. Daga and J. Jimenez DISCUSSION Coupling cell growth to cell division by translational control In this study, we first showed that tif1 is required for the initiation of protein synthesis (Fig. 1B), demonstrating that this gene codes for the translation initiation factor 4A (eIF4A) in S. pombe, as predicted from sequence analysis (Fischli et al., 1996). Using the regulated expression of a truncated eIF4A protein we showed that a severe inhibition on the translation initiation was pleiotropic and affected both cell growth and cell division, whereas a moderate interference on eIF4A activity arrested cell division at the G2 phase allowing cell growth to continue (Fig. 2). This result demonstrates that mRNAs involved in both cell growth and cell division require eIF4A activity for their efficient translation, but synthesis of polypeptides controlling the G2-M transition is particularly sensitive to diminished translation initiation activity. This result agrees with the observation that cell cycle progression and cell growth are separable processes which are normally coupled, and explains why growth is dominant and rate-limiting (Johnston et al., 1977; Neufeld et al., 1998). The fact that overexpression of cdc25 (or mutants in wee1 that abolish cdc25 requirements) overcomes the G2-arrest and caused a slight reduction in eIF4A activity suggests that the arrest results from failure of Cdc25 to accumulate. This genetic result predicts that translation of cdc25 mRNAs should be extremely sensitive to the level of translation initiation activity. Biochemical analysis was further used to demonstrate this prediction (Fig. 6). Protein synthesis integrates the overall biosynthetic activity leading to cell growth, and translation initiation is the key step of the whole process (Mathews et al., 1996). Since expression of Cdc25 is strongly associated with the growth threshold required to initiate M phase (Moreno et al., 1990), the rate-limited accumulation of Cdc25 at the translation initiation step shown in this work provides a plausible mechanism for coordinating cell growth with the mitotic machinery. The degradation of Cdc25 by the ubiquitin pathway at the end of mitosis (Nefsky and Beach, 1996) warrants that the proposed mechanism is reset each cycle. Furthermore, the proteolysis of a fixed amount of Cdc25 by this pathway could also provide a homeostatic mechanism to maintain an equilibrium state in the cell when the cell size is altered. Fission yeasts coordinate growth and division by regulating the length of time spent in G2 (Sveiczer et al., 1996) whereas budding yeasts exert a size control throughout length of G1 (Jagadish and Carter, 1977; Johnston et al., 1977). It has been recently shown that synthesis of the Cln3 G1 cyclin is particularly sensitive to inhibition of translation initiation, suggesting that the G1/S size control in budding yeasts operates through the translational control of this dosage-dependent activator of the S phase (Polymenis and Schmidt, 1997; Hall et al., 1998). Cln3 is not essential for control of cell growth and division, thus, despite it significant role in coupling division to growth, it is probably one part of a more complex picture (Polymenis and Schmidt, 1999). Although biochemically different, Cln3 and Cdc25 are both dosagedependent activators of Cdc2 whose synthesis is particularly

sensitive to the capacity of the cell’s protein synthesis machinery. Therefore, translation initiation control could be a common mechanism coordinating cell growth and division in the G1 (through Cln3) as well as in the G2 phase (through Cdc25). In higher eukaryotes size is monitored both during the G1/S and the G2/M transitions. In these cells entry into mitosis is regulated by the timing of Cdc2 dephosphorylation (by the Cdc25C phosphatase; Lammer et al., 1998) while other Cdks in association with various cyclins (D, E and A) control the G1/S transition (Reed et al., 1994). Interestingly, it has been shown that coordination of growth and cell division in the Drosophila wing largely dependent on the expression of cdc25 (String) for M-phase initiation and cyclin E for S-phase initiation (Neufeld et al., 1998), suggesting that the mechanism proposed for Cdc25 in the present study, and that recently described regulating the expression of Cln3 (Polymenis and Schmidt, 1997; Hall et al., 1998) could be a mechanism coupling cell growth and cell division in higher eukaryotic cells. The activity of the translation initiation machinery is regulated by stress, cAMP, and a number of other nutritional and environmental conditions (Sonenberg and Gingras, 1998; Hall et al., 1998). Thus, key cell cycle activators whose synthesis is particularly sensitive to the translational activity could connect progress through the cell cycle with the varying growth rates imposed by different nutrients and environmental signals. Interestingly, in agreement with this suggestion, overexpression of Sum1 (homologous to the eIF3-p39 regulator of translation initiation encoded by TIF34 in S. cerevisiae) influences the Cdc2-dependent size control at the time that inhibits the cell cycle in response to osmotic stress (Humphrey and Enoch, 1998). Salient sequence features of cdc25 and cdc13 mRNAs 5′′ leaders A deletion analysis demonstrated that the 5′ UTR of the cdc25 mRNA rate-limits expression of Cdc25 at the translation initiation step (Figs 6, 7). Interestingly, the cdc25 mRNAs 5′ leader contains several salient features such as its unusual length and the presence of uORFs. A short uORF in the 5′ leader of the CLN3 mRNA has been shown to rate-limit Cln3 expression at the translation initiation step by a ‘leaky scanning’ mechanism (Polymenis and Schmidt, 1997). Thus, the existence of uORFs in the cdc25 mRNAs 5′ leader suggests that a similar mechanism could be involved in rate-limiting its expression. The cdc25 mRNAs 5′ UTR is extremely long in comparison with the average length found in yeast and other eukaryotic mRNAs (Geballe, 1996). Translation of mRNAs with long 5′ UTRs is very inefficient and highly dependent on the RNA-helicase activity provided by the eIF4A activity of the eIF4F complex (Lawson et al., 1986; Koromilas et al., 1992). Thus, in addition to uORFs, the unusually long 5′ leader region of cdc25 mRNAs could equally account for the ratelimited translation initiation observed for the expression of this cell cycle phosphatase. The ability of cdc25 to rescue the G2-arrest produced in tif1d3 mutants suggests that it is the only mitotic activator whose translation limits entry into mitosis in S. pombe cells with reduced translation initiation activity. However, western blot analysis indicated that Cdc13 is also influenced by mutations

Translational control of Cdc25 3145 in tif1 (see Fig. 6). Interestingly, cdc25 and cdc13 mRNAs have in common a long 5′ UTR with short uORFs. Furthermore, both 5′ leaders contain unusual repeats of a polypirimidine track and a stable stem-loop structure (Fig. 6E), suggesting a common mechanism regulating the expression of these mitotic activators at the translational level. Some known properties of Cdc25 and Cdc13 underlie tentative roles for these conserved motifs. For instance, it is remarkable that, albeit subjected to different patterns of transcription (Bueno et al., 1991), accumulation of these two Cdc2 activators displays identical cell cycle kinetics (Creator and Mitchinson, 1996). Thus, the above 5′ UTR features could be involved in a mechanism that coregulates translation of Cdc25 and Cdc13 to ensure their proper stoichiometry at varying cellular growth rates. In eukaryotic cells, translation initiation is reduced in mitosis concomitantly with a reduction in eIF4E phosphorylation (Sonenberg, 1996). If translation in S. pombe cells suffers an equivalent regulation (Creator and Mitchinson, 1982), the accumulation of Cdc25 and Cdc13 polypeptides during the Mphase (Creator and Mitchinson, 1996) should require a mechanism to bypass the inhibition imposed by eIF4E at the initiation step. The unusual length of their 5′ UTRs provides a simple one: the accumulation of ribosomal subunits at this long region would allow further translational activity in the absence of new initiation events. Another possibility is the presence of cap-independent internal initiation mechanisms (IRES) as an alternative pathway to ensure the efficient translation of these proteins during mitosis (Geballe, 1996). Elucidation of the biological role of these 5′ motifs conserved between cdc25 and cdc13 mRNAs promises new insights into the mechanisms that coordinate growth and mitosis in fission yeasts. The conservation among eukaryotic translation initiation factors (Fischli et al., 1996.), together with the universal role of the Cdc25 cell cycle phosphatase (Jimenez et al., 1990; Lammer et al., 1998), suggest that the regulatory mechanism presented in this study could be of widespread significance in the control of balanced growth and division of all eukaryotic cells. The translation initiation machinery plays an important role in the induction of neoplasia (Lazaris-Karatzas et al., 1990; Koromilas et al., 1992; Rosenwald et al., 1993). These results could also illuminate some mechanisms by which perturbations in translation initiation factors trigger oncogenic transformations. This work was supported by the Spanish MEC and the Junta de Andalucía. We thank Lucia Cruzado and Inmaculada Florido for technical assistance, Avelino Bueno, Sergio Moreno, Paul Russell and Paul Nurse for providing strains, plasmids, libraries and antibodies, Stuart MacNeill, Paul Nurse, Sergio Moreno and Lagar de Oliveros for useful discussions and Elmar Maier for providing membranes for physical mapping of the tif1 gene.

REFERENCES Alfa, C., Fantes, P., Hyams, P., McLeod, M. and Warbrick, E. (1993). Experiments with Fission Yeast. A Laboratory Course Manual. New York: CSHL press. Barbet, N. C., Schneider, U., Helliwell, S. B., Stansfield, I., Tuite, M. F. and Hall, M. N. (1996). TOR controls translation initiation and early G1 progression in yeast. Mol. Biol. Cell. 7, 25-42. Barnes, C. A., Singer, R. A. and Johnston, G. C. (1993). Yeast prt1 mutations alter heat-shock gene expression through transcript fragmentation. EMBO J. 12, 3323-3332.

Brenner, C., Nakayama, N., Goebl, M., Tanaka, K., Toh-e, A. and Matsumoto, K. (1988). CDC33 encodes mRNA cap-binding protein eIF4E of Saccharomyces cerevisiae. Mol. Cell. Biol. 8, 3556-3559. Bueno, A., Richardson, H., Reed, S. I. and Russell, P. (1991) A fission yeast B-type cyclin functioning early in the cell cycle. Cell 66, 149159. Creator, J. and Mitchison J. M. (1982). Patterns of protein synthesis during the cell cycle of the fission yeast Schizosaccharomyces pombe. J. Cell Sci. 58, 263-285. Creator, J. and Mitchison J. M. (1996). The kinetics of the B cyclin p56cdc13 and the phosphatase p80cdc25 during the cell cycle of the fission yeast Schizosaccharomyces pombe. J. Cell Sci. 109, 1647-1653. Cross, F. R. (1988). DAF1, a mutant gene affecting size control, pheromone arrest, and cell cycle kinetics of Saccharomyces cerevisiae. Mol. Cell. Biol. 8, 4675-4684. Dirick, L., Bohm, T. and Nasmyth, K. (1995). Roles and regulation of ClnCdc28 kinases at the start of the cell cycle of Saccharomyces cerevisiae. EMBO J. 14, 4803-4813. Fischli, A., Schmid, S. R., Coppolecchia, R. and Linder, P. (1996). The translation initiation factor eIF4A from Schizosaccharomyces pombe is closely related to its mammalian counterpart. Yeast 12, 977-981. Fuller-Pace, F. V. (1994). RNA helicases, modulators of RNA structure. Trends Cell Biol. 4, 271-274. Geballe, A. P. (1996). Translational control mediated by upstream AUG codons. In Translational control (ed. J. W. B. Hershey, M. B. Mathews and N. Sonenberg). pp 173-197. New York: CSHL press. Hall, D. D., Markwardt, D. D., Parviz, F. amd Haideman, W. (1998). Regulation of the Cln3-Cdc28 kinase by cAMP in Saccharomyces cerevisiae. EMBO J. 17, 4370-4378. Hanic-Joyce, P. J., Johnston, G. C. and Singer, R. A. (1987). Regulated arrest of cell proliferation mediated by yeast prt1 mutations. Exp. Cell Res. 172, 134-145. Humphrey, T. and Enoch, T. (1998) Sum1, a highly conserved WD-repeat protein, suppresses S-M checkpoint mutants and inhibits the osmotic stress cell cycle response in fission yeast. Genetics 148, 1731-1742. Jagadish, M. N. and Carter, B. L. (1977). Genetic control of cell division in yeast cultured at different growth rates. Nature 269, 145-147. Jimenez, J., Alphey, L. Nurse, P. and Glover, D. M. (1990). Complementation of fission yeast cdc2ts and cdc25ts mutants identifies two cell cycle genes from Drosophila, a cdc2 homologue and string. EMBO J. 9, 3565-3571. Jimenez, J. and Oballe, J. (1994). Ethanol-hypersensitive and ethanoldependent cdc− mutants in Schizosaccharomyces pombe. Mol. Gen. Genet. 245, 86-95. Johnston, G. C., Pringle, J. R. and Hartwell, L. H. (1977). Coordination of growth with cell division in the yeast Saccharomyces cerevisiae. Exp. Cell Res. 105, 79-98. Koromilas, A. E., Lazaris-Karatzas, A. and Sonenberg, N. (1992). mRNAs containing extensive secondary structure in their 5′ non-coding region translate efficiently in cells overexpressing initiation factor eIF-4E. EMBO J. 11, 4153-4158. Kovelman, R. and Russell, P. (1996). Stockpiling of Cdc25 during a DNA replication checkpoint arrest in Schizosaccharomyces pombe. Mol. Cell. Biol. 16, 86-93. Lammer, C., Wagerer, S., Saffrich, R., Mertens, D., Ansorge, W. and Hoffmann, I. (1998). The cdc25B phosphatase is essential for the G2/M phase transition in human cells. J. Cell Sci. 111, 2445-2453. Lawson, T. G., Lawson, T. G., Ray, B. K., Dodds, J. T., Grifo, J. A., Abramson, R. D., Merrick, W. C., Betsch, D. F., Weith, H. L. and Thach, R. E. (1986). Influence of 5′ proximal secondary structure on the translational efficiency of eukaryotic mRNAs and on their interaction with initiation factors. J. Biol. Chem. 261, 13979-13989. Lazaris-Karatzas, A., Montine, K. S. and Sonenberg, N. (1990). Malignant transformation by a eukaryotic initiation factor subunit that binds to mRNA 5′ cap. Nature 345, 544-547. Linder, P. and Slonimski, P. P. (1989). An essential yeast protein, encoded by duplicated genes TIF1 and TIF1 and homologous to the mammalian translation initiation factor eIF-4A, can suppress a mitochondrial missense mutation. Proc. Nat. Acad. Sci. USA 86, 2286-2290. Lundgren, K., Walworth, N., Booher, R., Dembski, M., Kirschner, M. and Beach, D. (1991). mik1 and wee1 cooperate in the inhibitory tyrosine phosphorylation of cdc2. Cell 64, 1111-1122. Mathews, M. B., Sonenberg, N. and Hershey, J. W. B. (1996). Origins and targets of translational control. In Translational control (ed. J. W. B.

3146 R. R. Daga and J. Jimenez Hershey, M. B. Mathews and N. Sonenberg), pp. 1-29. New York: CSHL press. Maundrell, K. (1993). Thiamine-repressible expresion vectors pREP and pRIP for fission yeast. Gene 123, 127-130. Mitchison, J. M. (1971). The Biology of the Cell Cycle. Cambridge University Press, London. Moreno, S., Nurse, P. and Russell, P. (1990). Regulation of mitosis by cyclic accumulation of Cdc25 mitotic inducer in fission yeast. Nature 344, 549-552. Moreno, S., Klar, A. and Nurse, P. (1991). Molecular genetic analysis of fission yeast Schizosaccharomyces pombe. Meth. Enzymol. 194, 795-823. Nash, R., Tokiwa, G., Anand, S., Erickson, K. and Futcher, A. B. (1988). The WHI1+ gene of Saccharomyces cerevisiae tethers cell division to cell size and is a cyclin homolog. EMBO J. 7, 4335-4346. Nefsky, B. and Beach, D. (1996). Pup1 acts as an E6-AP-like protein ubiquitin ligase in the degradation of cdc25. EMBO J. 15, 1301-1312. Neufeld, T. P., de la Cruz, A. F., Johnston, L. A. and Edgar, B. A. (1998). Coordination of growth and cell division in the Drosophila wing. Cell 93, 1183-1193. Nurse, P. (1975). Genetic control of cell size at cell division in yeast. Nature 256, 547-551. Nurse, P., Thuriaux, P. and Nasmyth K. (1976). Genetic control of the cell division cycle in the fission yeast Schizosaccharomyces pombe. Mol. Gen. Genet. 146, 167-178. Nurse, P. and Thuriaux, P. (1984). Temperature sensitive allosuppressor mutants of the fission yeast S. pombe influence cell cycle control over mitosis. Mol. Gen. Genet. 196, 332-338. Nurse, P. (1990). Universal control mechanism regulating onset of M-phase. Nature 344, 503-508. Pause, A. and Sonenberg, N. (1992). Mutational analysis of a DEAD box RNA helicase, the mammalian translation initiation factor eIF-4A. EMBO J. 11, 2643-2654. Pause, A., Methot, N. Svitkin, Y. Merrick, W. C. and Sonenberg N. (1994). Dominant negative mutants of mammalian translation initiation factor eIF4A define a critical role for eIF-4F in cap-dependent and cap-independent initiation of translation. EMBO J. 13, 1205-15. Polymenis, M. and Schmidt, E. V. (1997). Coupling of cell division to cell growth by translational control of the G1 cyclin CLN3 in yeast. Genes Dev. 11, 2522-2531.

Polymenis, M. and Schmidt, E. V. (1999). Coordination of cell growth with cell division. Curr. Opin. Genet. Dev. 9: 76-80. Reed, S. I., Bailly, E., Dulic, V., Hengst, L., Resnitzky, D., Slingerland, J. (1994). G1 control in mammalian cells. J. Cell Sci. Suppl. 18, 69-73. Rosenwald, I. B., Kaspar, R., Rousseau, D., Gehrke, L., Leboulch, P., Chen, J. J., Schmidt, E. V. and Sonenberg, N. (1995). Eukaryotic translation initiation factor 4E regulates expression of cyclin D1 at transcriptional and post-transcriptional levels. J. Biol. Chem. 270, 21176-21180. Russell, P. and Nurse, P. (1986). cdc25+ functions as an inducer in the mitotic control of fission yeast. Cell 45, 145-153. Russell, P. and Nurse, P. (1987). Negative regulation of mitosis by wee1+, a gene encoding a protein kinase homolog. Cell 49, 559-567. Sachs, A. B., Sarnow, P. and Hetze, M. W. (1997). Starting at the beginning, middle, and end, translation initiation in eukaryotes. Cell 89, 831-838. Sagliocco, F. A., Moore, P. A. Brown, A. J. P. (1994). Polysome analysis. In Methods in Molecular Biology, vol. 53. Yeast Protocols (ed. I. Evans), pp. 297-312. Totowa: Humana Press Inc. Sambrook, J., Fritsch, E. F. and Maniatis, T. (1989). Molecular Cloning. New York: Cold Spring Harbor Laboratory Press. Sonenberg, N. (1996) mRNA 5′ Cap-binding protein eIF4E and control of cell growth. In Translational control (ed. J. W. B. Hershey, M. B. Mathews and N. Sonenberg). pp 245-269. New York: CSHL press. Sonenberg, N. and Gingras, A. C. (1998). The mRNA 5′ cap-binding protein eIF4E and control of cell growth. Curr. Opin. Cell Biol. 10, 268-275. Sveiczer, A., Novack, B. and Mitchison, J. M. (1996). The size control of fission yeast revisited. J. Cell Sci. 109, 2947-2957. Verlhac, M. H., Chen, R. H., Hanachi, P., Hershey, J. W. and Derynck, R. (1997). Identification of partners of TIF34, a component of the yeast eIF3 complex, required for cell proliferation and translation initiation. EMBO J. 16, 6812-6822. Warbrick, E. and Glover, D. M. (1994) a Drosophila gene encoding a DEAD box RNA helicase can suppress loss of wee1/mik1 function in Schizosaccharomyces pombe. Mol. Gen. Genet. 245, 654-657. Wu, L. and Russell, P. (1997). Nif1, a novel mitotic inhibitor in Schizosaccharomyces pombe. EMBO J. 16, 1342-1350. Zhong, T. and Arndt, K. T. (1993). The yeast SIS1 protein, a DnaJ homolog, is required for the initiation of translation. Cell 73, 1175-1186.