G1 cyclins block the Ime1 pathway to make mitosis and meiosis incompatible in budding yeast
1999; Springer Nature; Volume: 18; Issue: 2 Linguagem: Inglês
10.1093/emboj/18.2.320
ISSN1460-2075
Autores Tópico(s)Plant nutrient uptake and metabolism
ResumoArticle15 January 1999free access G1 cyclins block the Ime1 pathway to make mitosis and meiosis incompatible in budding yeast Neus Colomina Neus Colomina Departament de Ciències Mèdiques Bàsiques, Universitat de Lleida, Rovira Roure 44, 25198 Lleida, Catalunya, Spain Search for more papers by this author Eloi Garí Eloi Garí Departament de Ciències Mèdiques Bàsiques, Universitat de Lleida, Rovira Roure 44, 25198 Lleida, Catalunya, Spain Search for more papers by this author Carme Gallego Carme Gallego Departament de Ciències Mèdiques Bàsiques, Universitat de Lleida, Rovira Roure 44, 25198 Lleida, Catalunya, Spain Search for more papers by this author Enrique Herrero Enrique Herrero Departament de Ciències Mèdiques Bàsiques, Universitat de Lleida, Rovira Roure 44, 25198 Lleida, Catalunya, Spain Search for more papers by this author Martí Aldea Corresponding Author Martí Aldea Departament de Ciències Mèdiques Bàsiques, Universitat de Lleida, Rovira Roure 44, 25198 Lleida, Catalunya, Spain Search for more papers by this author Neus Colomina Neus Colomina Departament de Ciències Mèdiques Bàsiques, Universitat de Lleida, Rovira Roure 44, 25198 Lleida, Catalunya, Spain Search for more papers by this author Eloi Garí Eloi Garí Departament de Ciències Mèdiques Bàsiques, Universitat de Lleida, Rovira Roure 44, 25198 Lleida, Catalunya, Spain Search for more papers by this author Carme Gallego Carme Gallego Departament de Ciències Mèdiques Bàsiques, Universitat de Lleida, Rovira Roure 44, 25198 Lleida, Catalunya, Spain Search for more papers by this author Enrique Herrero Enrique Herrero Departament de Ciències Mèdiques Bàsiques, Universitat de Lleida, Rovira Roure 44, 25198 Lleida, Catalunya, Spain Search for more papers by this author Martí Aldea Corresponding Author Martí Aldea Departament de Ciències Mèdiques Bàsiques, Universitat de Lleida, Rovira Roure 44, 25198 Lleida, Catalunya, Spain Search for more papers by this author Author Information Neus Colomina1, Eloi Garí1, Carme Gallego1, Enrique Herrero1 and Martí Aldea 1 1Departament de Ciències Mèdiques Bàsiques, Universitat de Lleida, Rovira Roure 44, 25198 Lleida, Catalunya, Spain *Corresponding author. E-mail: [email protected] The EMBO Journal (1999)18:320-329https://doi.org/10.1093/emboj/18.2.320 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info Diploid yeast cells switch from mitosis to meiosis when starved of essential nutrients. While G1 cyclins play a key role in initiating the mitotic cell cycle, entry into meiosis depends on Ime1, a transcriptional activator regulated by both nutritional and cell-type signals. We show here that G1 cyclins downregulate IME1 transcription and prevent the accumulation of the Ime1 protein within the nucleus, which results in repression of early-meiotic gene expression. As G1-cyclin deficient cells do not require nutrient starvation to undergo meiosis, G1 cyclin would exert its role by transmitting essential nutritional signals to Ime1 function. The existence of a negative cross-talk mechanism between mitosis and meiosis may help explain why these two developmental options are incompatible in budding yeast. Introduction Nutrients are among the most important trophic factors for yeast and, like most other eukaryotes, Saccharomyces cerevisiae takes different developmental options depending on environmental conditions during the G1 phase of the cell cycle. Depending on the nutrient limitation conditions, haploid cells either arrest in G1 or initiate invasive growth. Diploid cells also arrest in G1 or produce pseudo-hyphae, but they have an additional option: entry into meiosis. Ime1 is a transcriptional activator that routes both nutritional and cell-type signals to the expression of meiotic genes, and has a central role in triggering meiosis (for review see Kupiec et al., 1997). Only diploid cells are able to enter meiosis as they possess both components of the Mata1–Matα2 complex, which allows expression of the IME1 gene by two separate pathways. In addition, nutrient starvation signals regulate IME1 at both transcriptional and post-transcriptional levels. The IME1 promoter is repressed by glucose and nitrogen and is induced in the presence of acetate (Kassir et al., 1988; Sagee et al., 1998). In addition, Ime1's function as a transcriptional activator depends on its ability to interact with Ume6, a protein that binds at the promoters of early meiotic genes to inhibit or activate their expression depending on the interacting proteins (Strich et al., 1994; Bowdish et al., 1995; Kadosh and Struhl, 1997). Interaction between Ime1 and Ume6, which is elicited by the Rim11 and Rim15 kinases (Rubin-Bejerano et al., 1996; Vidan and Mitchell, 1997), has been shown to be a key target for glucose-mediated inhibition of Ime1 activity (Malathi et al., 1997). S-phase entry during meiosis is completely dependent on Ime1 (Kassir et al., 1988), partly through the Ime2 protein kinase (Foiani et al., 1996). However, the mechanisms by which this transcriptional activator is able to trigger initiation of DNA replication have not been characterized. It has been proposed that Cdc28, the central cyclin-dependent kinase that regulates the mitotic cell cycle, may not have a role since cdc28 thermosensitive cells arrest meiosis after DNA replication (Shuster and Byers, 1989), but as yet no additional evidence is available to confirm this idea. Although Cln1, Cln2 and Cln3, the three yeast G1 cyclins, show clear functional redundancy, they perform different roles during the G1–S transition in the mitotic cell cycle. Cln3 is the most potent activator regarding SBF- and MBF-dependent transcription of a set of genes including CLN1 and CLN2 (Tyers et al., 1993; Dirick et al., 1995; Stuart and Wittenberg, 1995; Levine et al., 1996). On the other hand, Cln1 and Cln2 have more specialized roles in budding initiation (Benton et al., 1993; Cvrcková and Nasmyth, 1993) and also in DNA synthesis initiation through degradation of the Clb-Cdc28 inhibitor Sic1 (Schwob et al., 1994; Schneider et al., 1996; Feldman et al., 1997; Skowyra et al., 1997). The possible role of G1 cyclins in regulating entry into pre-meiotic S phase has not been characterized. This work deals with the relationships between key molecules involved in initiating either mitosis or meiosis. Here we show that, although mitosis and meiosis share some important similarities during S-phase entry, G1 cyclins are not required to trigger pre-meiotic DNA replication. In fact, we have found that G1 cyclins block the Ime1 pathway to inhibit meiosis by two different mechanisms: (i) downregulating IME1 transcription and (ii) preventing Ime1 accumulation within the nucleus. Our results indicate that yeast cells have established a negative cross-talk mechanism between mitosis and meiosis to make these cell cycle choices incompatible. Results G1 cyclins are rapidly lost during entry into meiosis In accordance with their essential role in the G1–S transition, we have shown previously that G1-cyclin levels are downregulated very rapidly in haploid yeast cells deprived of an essential nutrient such as nitrogen (Gallego et al., 1997). Contrary to haploid cells, which arrest in G1, diploid yeast cells switch from the mitotic to the meiotic cell cycle under nitrogen starvation conditions in the presence of a non-fermentable carbon source. To understand the basis of these different cell fates, we first focused on the key molecules of the mitotic G1–S transition during entry into meiosis. To obtain synchronous entry into pre-meiotic S phase, the non-standard pre-growth regimen developed by Padmore et al. (1991) was used (see Materials and methods). Diploid wild-type 1788 cells growing exponentially in acetate-based rich medium were allowed to reach a high cell density and accumulate in G1 as their growth became limited by the carbon source. Upon transfer to sporulation medium with carbon source added but lacking the nitrogen source, these G1 cells readily initiated a pre-meiotic S phase in ∼4 h (Figure 1A) and proceeded into the meiotic nuclear divisions to produce spores in 24 h. Figure 1B shows that Cln1 and Cln2 are already absent in G1 cells limited by the carbon source, while Cln3 is lost early after transfer to sporulation conditions, suggesting that G1 cyclins might not be required at all in this specialized version of the yeast cell cycle. As Cln3 is the only detectable G1 cyclin during entry into meiosis, we used Cln3-deficient cells to analyze Ime1-dependent gene expression and pre-meiotic S-phase entry kinetics. Unexpectedly, not only did Cln3-deficient cells sporulate with high efficiency (84% compared with 75% for the wild type) but they underwent pre-meiotic S phase earlier and more efficiently. More than 80% of Cln3-deficient cells had completed DNA replication 4 h after being transferred to sporulation medium, while wild-type cells were just initiating pre-meiotic S phase at that time (Figure 1A). Induction of SPO13 (an Ime1-dependent early meiotic gene) and SPS1 (a middle meiotic gene) also occurred earlier in Cln3-deficient cells (Figure 1C). Thus, G1 cyclins do not seem to be required to initiate meiosis. The experiments shown in Figure 4 with a Δcln1 Δcln2 GAL1p–CLN3 strain (see below) provide further support to this idea. Rather, the fact that Cln3-deficient cells are able to enter pre-meiotic S phase more efficiently suggests that G1 cyclins might have a negative role in meiosis initiation. Figure 1.G1 cyclins are lost early during entry into meiosis. (A) DNA content distributions of wild-type 1788 (wt) and Cln3-deficient CML254 (Δcln3) cells during entry into meiosis. Samples were obtained during mitotic growth in YPA medium (CYC), after carbon-source limitation (G1, 0 h), and at different times under sporulation conditions (−N). (B) Protein levels for G1-cyclins Cln1, Cln2 and Cln3 as determined by Western blot of samples taken as in (A). The 12CA5 cross-reactive band (*) serves as a loading control. (C) mRNA levels for meiotic genes IME4 and IME1, as well as SPO13 (an Ime1-dependent early gene) and SPS1 (a middle gene) were determined by Northern blot analysis from samples taken as in (A). The 25S rRNA serves as a loading control. Download figure Download PowerPoint G1-cyclin overexpression inhibits meiosis and blocks Ime1 function If Cln3 exerts a negative effect on Ime1 function, entry into meiosis should be inhibited by overexpression of G1 cyclins from a constitutive promoter. We used the tetracycline-regulatable system that we had adapted to yeast (Garí et al., 1997) to drive expression of CLN1, CLN2 or CLN3 from a centromeric vector in cells growing exponentially in acetate-based rich medium without tetracycline to induce the tetO2 promoter. Figure 2A shows the percentages of asci and budded cells 24 h after nitrogen deprivation in cells overexpressing CLN1, CLN2 or CLN3. As expected from the negative role of Cln3 during entry into meiosis, G1-cyclin overexpression inhibited sporulation and forced cells to enter mitosis as deduced from the final budding indexes. Similar results were obtained in W303 diploid cells (data not shown). Constitutive overexpression of CLN3 from the tetO2 promoter did not prevent the G1 arrest produced by carbon-source limitation in acetate-based rich media as deduced from the DNA content distributions (data not shown). Wild-type cells arrested in G1 by carbon-source limitation did not increase their number significantly during the 24 h upon transfer to sporulation conditions. On the contrary, CLN3-overexpressing cells doubled their number and arrested with a high percentage of cells with a 4c DNA content, which is in agreement with the high final budding index attained by these cells. Thus, G1-cyclin overexpression not only inhibits sporulation but also drives cells into mitosis under conditions where wild-type cells enter meiosis very efficiently. Figure 2.G1-cyclin overexpression inhibits meiosis by blocking Ime1 function. (A) Sporulation and budding indexes were determined after 24 h under sporulation conditions in wild-type CML256 (wt) and Cln3-deficient CML254 (Δcln3) cells, as well as in CML256 cells overexpressing either CLN1 from pCM207 (↑CLN1), CLN2 from pCM214 (↑CLN2) or CLN3 from pCM166 (↑CLN3). (B) Wild-type CML256 cells transformed with the empty vector (wt) or pCM166 to overexpress CLN3 (↑CLN3) were used to determine mRNA levels for cyclin CLN3, the meiotic genes IME4 and IME1, as well as SPO13 (an Ime1-dependent early gene) and SPS1 (a middle gene). Cells were grown in YPA medium in the absence of tetracycline to allow for expression of the tetO2–CLN3 construct. Lanes correspond to samples obtained during mitotic growth in YPA medium (CYC), after carbon-source limitation (G1, 0 h), and at different times in sporulation medium (−N). The 25S rRNA serves as a loading control. Download figure Download PowerPoint Figure 2B shows that expression of both SPO13, which depends directly on Ime1 as an early gene, and SPS1, which is induced further downstream in the Ime1 pathway as a middle gene, is strongly repressed by CLN3 overexpression under sporulation conditions. As IME1 mRNA levels showed only a moderate decrease, these results suggest that CLN3 overexpression may block Ime1 function at a post-transcriptional level. It has been proposed that Ime1 may retro-activate its own transcription during the earliest steps of meiosis (Shefer-Vaida et al., 1995). To avoid possible effects on IME1 transcription due to Ime1 inhibition at a post-transcriptional level, Ime1-deficient cells were subject to CLN3 overexpression during entry into meiosis. IME1 expression was evaluated from a plasmid construct that lacks a functional IME1 open reading frame (ORF) (ime1-2), and from the homozygous kanMX4-disrupted chromosomal copies (ime1-1) that retain the IME1 promoter and transcription termination sequences (Figure 3). In both constructs, CLN3 overexpression was able to repress the IME1 promoter. Accordingly, CLN3-overexpression repression effects on IME1 transcription were also observed in rim11 null mutants (data not shown), where Ime1-dependent transcription is completely repressed (Bowdish et al., 1994). Thus, despite the likely existence of post-transcriptional mechanisms (see below), our results indicate that high Cln-cyclin levels have a role in repressing IME1 transcription. Figure 3.CLN3 overexpression inhibits IME1 transcription. Ime1-deficient CML342 cells carrying a homozygous ime1-1 deletion (Δime1::kanMX4) and a truncated ime1-2 allele on pCM268, were transformed with empty vector (wt) or pCM166 to overexpress CLN3 (↑CLN3). mRNA levels produced by the IME1 promoter from the chromosomal ime1-1 allele were determined with a kanMX4 probe, which also detects the kanMX4 transcript. The plasmid-borne ime1-2 mRNA was detected with an IME1 probe that does not cover any of the IME1 sequences left in ime1-1. Cells were grown in YPA medium in the absence of tetracycline to allow for expression of the tetO2–CLN3 construct. Lanes correspond to samples taken after carbon-source limitation (G1, 0 h), and at different times under sporulation conditions (−N). The 25S rRNA is shown as a loading control. The origin of the transcripts detected is outlined on the right. Download figure Download PowerPoint Figure 4.G1-cyclin-deficient cells switch from mitosis to meiosis in rich media. (A) DNA content distributions of CML353, a homozygous Δcln1 Δcln2 GAL1p–CLN3 strain that expresses IME1 constitutively from pCM284, transformed either with an empty vector (−Cln) or pCM194 (+Cln3), which contains the CLN3 gene under its own promoter sequences. Samples were obtained during mitotic growth in YPGal medium (CYC) and at different times after transfer to YPA medium (+N) to repress CLN3 expression from the GAL1 promoter. After 3 h in YPA, a portion of G1-cyclin-deficient cells (−Cln) were transferred to sporulation medium (−N) and samples were taken thereafter. Budding and sporulation indexes obtained at 24 h are indicated. (B) CML353 cells constitutively expressing IME1 (adhp–IME1) or not (IME1) from plasmid pCM284, and transformed with either an empty vector (−Cln) or pCM194 (+Cln3), were used to determine mRNA levels of IME1, SPO13 and CLN3 by Northern blot analysis from samples taken as in (A). The 25S rRNA is shown as a loading control. Download figure Download PowerPoint Cells deprived of G1 cyclins switch from mitosis to meiosis in rich media Nitrogen starvation has been the most efficient environmental condition used to induce meiosis in diploid yeast cells (Freese et al., 1982). On the other hand, nitrogen starvation causes a rapid decrease of G1-cyclin levels in haploid cells (Gallego et al., 1997). We have shown here that (i) G1 cyclins are lost early during entry into meiosis; (ii) Cln3-deficient cells undergo meiosis more efficiently than wild-type cells; and (iii) G1-cyclin overexpression inhibits meiosis and drives cells into mitosis even under nitrogen starvation conditions. Thus, nitrogen starvation could exert its essential role in inducing meiosis through downregulation of G1-cyclin levels. To test whether G1-cyclin downregulation is not only a necessary but also a sufficient condition to allow entry into pre-meiotic S phase independently of the nutritional status of the cell, we used a homozygous Δcln1 Δcln2 GAL1p–CLN3 strain that depends on the presence of galactose to execute the mitotic G1–S transition. In order to determine first the essential role of G1 cyclins on Ime1 function inhibition at a post-transcriptional level, IME1 was consitutively expressed in a centromeric vector from the Schizosaccharomyces pombe adh promoter, which attains expression levels similar to those produced by the natural IME1 promoter under sporulation conditions (data not shown). Figure 4A shows that when these cells were transferred to acetate-based rich medium to shut off CLN3 expression (Figure 4B), they arrested temporarily in G1, proceeded into an S phase with no signs of budding and finally sporulated with high efficiency, indicating that G1-cyclin downregulation is sufficient for entry into meiosis in rich media as long as IME1 constitutive expression is provided. S-phase entry and sporulation efficiencies were as high as those obtained when cells were starved for nitrogen by transfer to sporulation medium. Thus, nitrogen starvation may exert its essential role in meiosis induction by downregulating G1 cyclins. As expected, when CLN3 expression was provided from its natural promoter sequences in a centromeric vector, those cells also arrested temporarily in G1 but resumed cell proliferation by mitosis as deduced from budding indexes (Figure 4A) and cell number increase (data not shown). It has been proposed that cell size may be an important requirement for entry into meiosis (Calvert and Dawes, 1984). G1-deficient cells entered pre-meiotic S phase in rich media with a mean cell volume of 184 fl, while CLN3-containing cells entered mitotic S phase under the same conditions with a mean cell volume of 187 fl. Both mean cell volumes are very similar and, on the other hand, larger than that shown by the two strains when they entered pre-meiotic S phase in sporulation media (125 and 128 fl). Thus, the fact that cells either lacking or containing Cln3 enter pre-meiotic or mitotic S phase, respectively, cannot simply be explained as a consequence of having smaller or larger cell volumes than that required for sporulation. Although the IME1 promoter senses a variety of nutritional signals to become fully active (Sagee et al., 1998), when we performed the same cell cycle analysis in acetate-based rich medium with a Δcln1 Δcln2 GAL1p–CLN3 strain with IME1 under its own promoter, cells did indeed enter S phase with no signs of budding (data not shown) and sporulated, albeit at lower frequencies (15% compared with 40% in sporulation medium). These results suggest that IME1 expression from its own promoter in acetate-based rich medium is sufficient for entry into meiosis as long as G1 cyclins are not present in the cell, which agrees with the fact that CLN3 overexpression represses IME1 transcription (Figure 2; see above). Figure 4B shows that after an initial increase in IME1 transcription due to the transfer from galactose to acetate-based rich media, Cln-deficient cells induced IME1 transcription at levels much higher than those attained by cells containing Cln3. In addition, Ime1-dependent induction of SPO13 expression only took place in Cln-deficient cells. This transcriptional activation was not merely due to higher Ime1 levels as IME1 transcription increased. Cln3-containing cells that expressed constitutive levels of IME1 mRNA were not able to activate SPO13 transcription (Figure 4B). The same inhibitory effects on Ime1 were observed with a strain where progression through the mitotic cycle depended on Cln1 (data not shown), which agrees with the fact that constitutive overexpression of CLN1 and CLN2 inhibits meiosis under optimal nutritional conditions for sporulation (see above). These results indicate that G1 cyclins are able to downregulate Ime1 function at both transcriptional and post-transcriptional levels. Figure 4B shows that both IME1 and SPO13 expression levels were similar in Cln-deficient cells independently of the presence of the nitrogen source, supporting the idea that the essential effect of nitrogen starvation to induce meiosis is the downregulation of G1 cyclins, which will in turn allow for full activation of Ime1 function. Not all meiotic nutritional requirements could be mimicked by G1-cyclin deprivation. Upon transfer of Δcln1 Δcln2 GAL1p–CLN3 cells from galactose to glucose-based rich medium, which also represses CLN3 expression, they rapidly arrested in G1 but did not undergo meiosis as deduced from DNA content distributions, lack of SPO13 expression and absence of asci (data not shown). Transcription of IME1 was not induced under these conditions, which is in agreement with the fact that the IME1 promoter is repressed by glucose (Sagee et al., 1998). In addition, and possibly through Rim15 (Vidan and Mitchell, 1997), glucose inhibits the physical interaction between Ume6 and Ime1 proteins (Malathi et al., 1997), which is essential to activate early-gene promoters during meiosis. Accordingly, although constitutive expression of IME1 resulted also in increased Ime1 protein levels in Cln-deprived cells by glucose, it did not allow for any detectable SPO13 induction (data not shown). G1-cyclin deficient cells are able to complete meiosis in the presence of nitrogen, which agrees with the fact that glucose but not the presence of a nitrogen source inhibits late steps during meiosis (Lee and Honigberg, 1996). G1 cyclins prevent Ime1 accumulation in the nucleus To activate transcription, Ime1 must interact with Ume6, a DNA-binding protein that plays a dual role in regulating meiotic early-gene expression. While Ume6 represses early-gene promoters in mitotically active cells, its interaction with Ime1 converts the complex into a transcriptional activator under sporulation conditions (Rubin-Bejerano et al., 1996). This interaction depends on two protein kinases, Rim11 and Rim15 (Rubin-Bejerano et al., 1996; Vidan and Mitchell, 1997). Although Rim15 may transmit some nutritional signals to Ime1 function (absence of fermentable carbon sources such as glucose), no physiological role has yet been established for Rim11. Since we have shown that G1 cyclins inhibit Ime1 function at a post-transcriptional level, we wished to determine whether this effect was exerted upon the interaction between Ume6 and Ime1. To test this possibility we used the tetracycline-regulatable expression system (Garí et al., 1997) and built a two-hybrid analysis model by fusing the tetO-binding domain of the Escherichia coli Tet repressor (TetR) to the interaction domain of Ime1 (Ime1id) and, on the other hand, the VP16 transactivator to the interaction domain of Ume6 (Ume6id). Similarly to previous work where the Ime1–Ume6 interaction was first shown by two-hybrid analysis (Rubin-Bejerano et al., 1996), the presence of both TetR–Ime1id and Ume6id–VP16 constructs in diploid cells gave rise to high β-galactosidase levels (comparable to those obtained with TetR–VP16), but only under sporulation conditions, while these high expression levels were completely dependent on Rim11 (Table I). We then used wild-type and homozygous cln3 null mutant strains to determine β-galactosidase activity in cycling and G1-arrested cells in glucose-based or acetate-based media and under the nitrogen starvation conditions used to induce entry into meiosis. Table II shows that the main nutritional requirement for the Ume6–Ime1 interaction is the presence of a non-fermentable carbon source such as acetate, independently of cell cycle position (cycling versus G1-arrested cells) or the presence or absence of the nitrogen source. Cln3-deficient cells showed a very similar behavior, which indicates that the Ume6–Ime1 interaction is not modulated by the presence of Cln3. Similar negative results were obtained when using Δcln1 Δcln2 GAL1p–CLN3 cells to analyze the Ume6–Ime1 interaction by two-hybrid analysis in the experimental approach shown in Figure 4 (data not shown). Thus we concluded that G1 cyclins block Ime1 function by mechanisms that seem not to involve its interaction with Ume6. Table 1. Two-hybrid system for Ime1–Ume6 interaction analysis Protein fusions +C (glucose) 8 h −N (acetate) wt wt Δrim11 TetR–VP16 221 301 273 TetR–Ime1id 3 5 4 TetR–Ime1id + Ume6id–VP16 9 233 7 β-galactosidase activities as Miller units were determined in wild-type 1788 (wt) and Rim11-deficient CML359 (Δrim11) cells expressing different hybrid proteins from pCM293 (TetR–VP16), pCM295 (TetR–Ime1id) and pCM298 (TetR–Ime1id + Ume6id–VP16). TetR-driven expression of the lacZ gene was monitored with plasmid pCM286. Samples were taken from cells growing exponentially in glucose-based minimal media (+C), or 8 h after transfer to sporulation conditions (8 h −N). Table 2. Ume6–Ime1 interaction under different nutritional conditions Strain glucose acetate CYC G1 CYC G1 8 h −N 24 h −N wt 0.05 0.10 0.75 0.75 0.79 0.79 Δcln3 0.05 0.12 0.69 0.63 0.65 0.74 β-galactosidase activities expressed as relative values were determined in wild-type 1788 (wt) and Cln3-deficient CML254 (Δcln3) cells containing either pCM293 (TetR–VP16) or pCM298 (TetR–Ime1id + Ume6id–VP16). TetR-driven expression of the lacZ gene was monitored with plasmid pCM286. Samples were taken from cells either growing exponentially (CYC), under carbon-source limitation (G1) or nitrogen starvation conditions (−N) in the presence of a fermentable (glucose) or a nonfermentable (acetate) carbon source. β-galactosidase activities determined from pCM298 (TetR–Ime1id + Ume6id–VP16) were made relative to those obtained from pCM293 (TetR–VP16) for each condition. Ime1 is a nuclear protein under sporulation conditions (Smith et al., 1993) and its localization does not depend on Rim11 (Rubin-Bejerano et al., 1996). By using a constitutively expressed hemagglutinin (HA)-tagged IME1 gene that fully complements homozygous ime1 null mutants, we determined its cellular localization by immunofluorescence in the Δcln1 Δcln2 GAL1p–CLN3 strain after transfer to acetate-rich media (Figure 5A). We have shown that under these conditions, Cln-deficient cells induce Ime1-dependent transcription and enter into pre-meiotic S phase, while mitotically cycling cells do not. Figure 5C shows that the overall Ime1 protein levels were similar as measured in immunoblots. However, the Ime1 protein did only accumulate in the nuclei of Cln-deficient cells. Mitotically cycling cells could prevent Ime1 accumulation in the nucleus either by G1-cyclin activity or more indirectly by the action of other molecules only present in mitotically active cells. To test these two possibilities we determined the cellular localization of Ime1 in cdc28-13 cells arrested in G1 with very low Cln-Cdc28 kinase activity (Wittenberg and Reed, 1988; Wittenberg et al., 1990), and cdc34-2 cells arrested at the G1–S transition with high Cln-Cdc28 levels (Deshaies et al., 1995; Yaglom et al., 1995). As seen in Figure 5B, Ime1 was clearly detected in the nuclei of cdc28-13 cells, while cdc34-2 cells showed a non-localized signal. Equivalent results were obtained in the W303 background. Similar lower-mobility forms of Ime1 can be detected in extracts obtained from cycling and cdc34-arrested cells but to a much lesser extent in extracts from G1-cyclin deficient or cdc28-arrested cells (Figure 5C), which suggests that Ime1 is subject to similar post-translational modifications when not localized to the nucleus. SPO13 induction took only place in cdc28-13 cells, indicating that Ime1 accumulation in the nucleus may be a key target for G1-cyclin inhibition of meiotic gene expression (Figure 5D). Figure 5.G1 cyclins prevent Ime1 accumulation within the nucleus. (A) Diploid CML353 (−Cln) and CML348 (+Cln) cells were transformed with either an empty vector or pCM284 (adhp–IME1-HA). Samples were taken 9 h after transfer to YPA medium as described under Figure 4. The Ime1 protein was visualized by immunofluorescence (Ime1) and nuclei were counterstained with propidium iodide. Bar represents 10 μm. (B) Haploid cdc28-thermosensitive CML200 (cdc28-13) and cdc34-thermosensitive CML344 (cdc34-2) cells were transformed with an empty ve
Referência(s)