Abstract
The DOM34 gene of Saccharomyces cerevisiae is similar togenes found in diverse eukaryotes and archaebacteria. Analysis of dom34 strains shows that progression through the G1 phase of the cell cycle is delayed, mutant cells enter meiosis aberrantly, and their ability to form pseudohyphae is significantly diminished. RPS30A, which encodes ribosomal protein S30, was identified in a screen for high-copy suppressors of the dom34Δ growth defect. dom34Δ mutants display an altered polyribosome profile that is rescued by expression of RPS30A. Taken together, these data indicate that Dom34p functions in protein translation to promote G1 progression and differentiation. A Drosophila homolog of Dom34p, pelota, is required for the proper coordination of meiosis and spermatogenesis. Heterologous expression of pelota in dom34Δ mutants restores wild-type growth and differentiation, suggesting conservation of function between the eukaryotic members of the gene family.
THE yeast Saccharomyces cerevisiae exhibits complex responses to environmental stimuli such as nutrient availability. In the presence of abundant nutrients, S. cerevisiae grows and divides by budding. However, in response to nutrient depletion three developmental regimens are available to diploid cells (Kron and Gow 1995). Sporulation in yeast is analogous to metazoan gametogenesis in that meiosis is followed by a program of cellular differentiation. Meiosis, which consists of one round of DNA replication followed by two successive rounds of division, produces four haploid nuclei that are then encapsulated into a specialized structure called an ascus (Esposito and Klapholz 1981). Stationary phase, or G0 (reviewed in Werner-Washburneet al. 1993), is a regulated cessation of proliferation in response to nutrient starvation. As cells enter G0, physiological changes occur that allow the cells to maintain viability for extended periods of time. Pseudohyphal growth is thought to be a foraging response. Nitrogen limitation on an agar surface induces this complex program of developmental change (Gimenoet al. 1992). Growth becomes hyperpolarized, and cells invade the agar surface. Pseudohyphal growth is also characterized by a mitotic delay that results in symmetric cell division (Kronet al. 1994). These different fates share a requirement for specific alterations in cell cycle control as well as diverse physiological and morphological changes.
Proper entry into these developmental pathways requires the coordination of many cellular processes. For instance, inhibition of the Tor proteins, which have been shown to be required for initiation of translation (Barbetet al. 1996), induces premature entry into meiosis (Zheng and Schreiber 1997) or stationary phase (Barbetet al. 1996). Translation also plays a critical role in specific phases of the cell cycle. CDC33, which encodes the translation initiation factor eIF-4E, is required for progression through G1 phase of the cell cycle (Brenneret al. 1988). Similarly, cells treated with low levels of the protein synthesis inhibitor cycloheximide display an extended G1 (Hartwell and Unger 1977).
In metazoans cell cycle control is altered at different stages of development as well as in different cell types (reviewed in Edgar 1995; Edgar and Lehner 1996). Mutants in twine, a meiosis-specific homolog of the cyclin dependent kinase (CDK) activating phosphatase cdc25, have defects in both oogenesis and spermatogenesis (Alpheyet al. 1992). The primary phenotype of the Drosophila mutant pelota is a failure to properly complete spermatogenesis, resulting in male infertility (Castrillonet al. 1993). Meiosis is arrested at prophase I in males homozygous for mutations in pelota; however, postmeiotic aspects of spermatid differentiation proceed in a nearly normal fashion (Eberhart and Wasserman 1995). The details of this defect are virtually identical to those of mutants in twine (Alpheyet al. 1992). This similarity led Eberhart and Wasserman (1995) to suggest that pelota acts in the cell cycle by directly impinging on CDK activity. pelota mutants display more subtle phenotypes in eye development and oogenesis that suggest that pelota also plays a role in cellular proliferation.
Dom34p is an S. cerevisiae homolog of pelota. DOM34 was isolated in a screen for meiotic mutants that are recombination-proficient (J. Engebrecht, unpublished results). This work demonstrates that dom34 mutants not only fail to undergo sporulation properly but also exhibit a G1 delay and fail to correctly execute pseudohyphal development. Isolation of a high-copy suppressor of the dom34Δ phenotype as well as polyribosome analysis indicate that Dom34p activity enhances the efficiency of the translation machinery. Finally, heterologous expression of pelota complements the dom34 phenotype, demonstrating that these two genes are functionally homologous.
MATERIALS AND METHODS
Yeast strains, media, and growth conditions: S. cerevisiae strains Y743 (MATa/MATα ura3/ura3) and Y 739 (MATa/MATα ura3/ura3 dom34::LEU2/dom34::LEU2) are derived from the R1278b background (Gimenoet al. 1992) and are isogenic derivatives of HLY336 and HLY335 (Liuet al. 1993). Yeast strains were transformed using the lithium acetate procedure (Itoet al. 1983). The dom34::LEU2 allele was created by one-step gene replacement (Rothstein 1991) using pTK3 (described below). All integrative transformations were confirmed by Southern blot (Southern 1975). YPAD is YEPD (Roseet al. 1990) supplemented with 100 μm adenine, YPARG is YPAD except that 1% raffinose and 2% galactose replace dextrose, and SC-uracil is synthetic medium (SC) (Roseet al. 1990) lacking uracil. Growth on YPARG plates was used to induce expression from the GAL1 promoter. Sporulation was induced in 2% potassium acetate. Synthetic low ammonia histidine dextrose (SLAHD) plates (Gimenoet al. 1992) were used to induce pseudohyphal growth.
Plasmids:DOM34 was cloned into a centromere-based (CEN) vector using the polymerase chain reaction (PCR). DOM34 coding sequence was amplified using primers P58 (5′-GGAA GATCTGTTTACTGTTTATC-3′) and P59 (5′-GCATCGATG GCTTCACATCTACTTGC-3′). The 1.2-kb PCR product was digested with BglII and ClaI and then used in a three-way ligation with the 2.9-kb XbaI/BglII fragment of yeast genomic DNA from λ clone PM-3064 (Rileset al. 1993) and XbaI/ClaI-digested pBluescript SK+. This ligation resulted in pTK1, which contains full-length DOM34 as well as 2.8 kb of upstream sequence. Plasmid TK2 was constructed by moving the 4.1-kb XbaI/ClaI fragment of pTK1 containing DOM34 into the XbaI and ClaI sites of the CEN URA3 vector pUN55 (Elledge and Davis 1988). A high-copy-inducible DOM34 plasmid was created by first moving the 1.9-kb SalI/ClaI fragment of pTK1 into the XhoI and ClaI sites of pUN15 to create pLD25. pLD25 was then digested with KpnI and BamHI and the resulting fragment ligated into the KpnI and BamHI sites of pYES2 (Invitrogen, Inc., Carlsbad, CA) generating pLD78.
The dom34Δ::LEU2 deletion allele was constructed as follows. Plasmid m-Tn3(LEU2) (Seifertet al. 1986) was digested with HindIII. After the resulting overhang was filled in with the Klenow fragment of DNA polymerase I, the plasmid was digested with BamHI, releasing a 2.4-kb, LEU2-containing, BamHI/blunt fragment. This fragment was used to replace the 0.9-kb BglII/HpaI fragment of pTK1 resulting in pTK3. This allele contains a deletion of 75% of the DOM34 coding region. Integration was targeted by digesting pTK3 with NotI.
A tagged version of Dom34p was created by fusing the triple hemagglutinin epitope (HA3; Wilsonet al. 1984) to the C terminus of the protein. Primers P14 (5′-GGAGAATTCCATGGCCTC CTCACCATCGTCTT-3′) and a T3-specific primer (5′-ATTAACCCTCACTAAAG-3′) were used to amplify DOM34 and to introduce an NcoI site at the 3′ end of the coding sequence. The fusion to HA3 was created by filling in the NcoI site with a Klenow fragment of DNA polymerase I and ligating that to the similarly filled-in EcoRI site of pSK-HA3 (Neimanet al. 1997). The DOM34HA3 fusion was moved on a 1.9-kb XbaI/SalI fragment into the corresponding sites of pUN55 creating pLD116. Expression of the tagged protein from the CEN plasmid complements the dom34Δ growth defect (data not shown).
A YEp24 high-copy genomic library clone was isolated as a suppressor of dom34Δ (described below). This plasmid was digested with PvuII and religated, removing all but 1.5-kb of genomic sequence containing RPS30A, generating pLD132. The 1.4-kb PvuII/ClaI fragment containing RPS30A from pLD 132 was cloned into the SmaI/ClaI site of pUN55 to create the CEN plasmid pLD135.
pY2-K1.6 contains a 1.6-kb pelota cDNA (Eberhart and Wasserman 1995) cloned into the pYES2 vector (generously provided by C. Eberhart and S. Wasserman, University of Texas, Southwestern). pelota expression is under the control of the GAL1 promoter.
Growth assays: Growth was examined qualitatively by streaking strains to be tested on YPAD plates, followed by incubation at either 30° (2 days) or 15° (8 days). Growth of strains containing the GAL1 pelota construct was tested as above except YPARG plates were used. Plates were photographed using an Olympus (Lake Success, NY) 35 mm camera and Kodak (Rochester, NY) Pan-X film.
Growth curves were performed as follows. Overnight cultures were diluted 1:20 in YPAD and grown to mid-log phase (~1 × 107 cells/ml). These cultures were used to inoculate prewarmed or prechilled YPAD to a starting density of 1 × 105 cells/ml and then grown at 30° or 15°. Samples were taken in triplicate from two independent cultures (n = 6) of each strain, and dilutions were plated to determine the number of viable cells. The mean and standard deviation for each time point were calculated.
Cell cycle analysis: Saturated cultures were used to inoculate YPAD to a starting density of 1 × 105 cells/ml. These cultures were grown at the appropriate temperatures until they reached mid-log phase (~4 × 106 cells/ml). Cells were fixed in 3.7% formaldehyde and, after brief sonication, counted as either unbudded, small budded (diameter of buds <50% of mother), or large budded (diameter of buds >50% of mother). At least 500 cells from each culture were counted.
Flow cytometry was performed essentially as described (Sazer and Sherwood 1990; as modified by S. Forsburg http://flosun.salk.edu/fcm/protocols/ycc.html). Cells (107) from a mid-log phase culture were pelleted, washed once in dH2O, pelleted, resuspended in cold 70% ethanol, and vortexed. Cells were stored at 4° until processing. Five hundred microliters of cells were pelleted in a microcentrifuge and resuspended in 1 ml of 50 mm sodium citrate. The rehydrated cells were pelleted, resuspended in 500 μl of 50 mm sodium citrate containing 0.1 mg/ml RNase A, and incubated at 34° for 8 hr. Five hundred microliters of 50 mm sodium citrate and 20 μg/ml propidium iodide were added and incubated in the dark at 4° overnight. Just prior to sorting, the cells were sonicated to break up any clumps. Flow cytometry was performed on a Becton-Dickinson (Franklin Lakes, NJ) FACScan. At least 10,000 cells were analyzed for each histogram.
Meiosis: Cells were grown to saturation in YPAD at 30° and used to inoculate sporulation medium. Cultures were sporulated at either 30° or 15°. Samples were fixed at the indicated times and stained with 4′,6-diamidino-2-phenylindole (DAPI). Progression through meiosis was assessed by fluorescence and phase-contrast microscopy. Cells were counted as either single, bi-, tri-, or tetranucleate. Bi-, tri-, or tetranucleate cells had completed meiosis I, whereas only tri- or tetranucleate cells had completed meiosis II. The number of cells that had formed spores was also counted. At each time point at least 250 cells were examined.
Pseudohyphal assays: Induction of pseudohyphal development was performed as previously described (Gimenoet al. 1992). Pseudohyphal colonies were photographed after 3 days on SLAHD plates at 30° using an Olympus BH-2 microscope and Kodak Pan-X film. Agar invasion, cell shape, and budding pattern were assessed as previously described (Mosch and Fink 1997).
Isolation and identification of high-copy suppressor: Strain Y739 was transformed with a YEp24 based yeast genomic library (Carlson and Botstein 1982; generously provided by N. Hollingsworth, SUNY, Stony Brook NY) and plated to SC-uracil at 15°. After 12 days, 22 of the ~4,000 transformants screened had formed large colonies. These were retested and those that still formed large colonies were investigated. Cells that had lost the library plasmid were selected for on medium containing 5-fluoroorotic acid (5-FOA; Boekeet al. 1987). Three of the original suppressor strains no longer formed large colonies, indicating that suppression was plasmid dependent. Plasmids pLD127, pLD128, and pLD130 were recovered from the three corresponding library transformants. pLD130 was shown to contain DOM34 sequences by performing PCR using DOM34-specific primers P59 (described above) and P2 (5′-AATTCAAGGGTATTAGTCTGAAAAAGG-3′). Sequence analysis of pLD127 and pLD128 was performed using P57 (5′-GGTGATGTCGGCGATATAGGC-3′), a YEp24-specific primer. pLD127 and pLD128 contain an overlapping sequence from chromosome XII. The region of overlap was subcloned, and RPS30A (GenBank accession no. U48699) was shown to be responsible for suppression.
Ribosome and polysome analysis: Polyribosome profiles were performed essentially as described (Martonet al. 1997). Yeast cultures were grown in YPAD to mid-log phase (~1 × 107 cells/ml) at 30°. Cycloheximide was added to 50 μg/ml, and cultures were returned to 30° for 5 min and then transferred to centrifuge tubes and placed on ice. Cells were collected by centrifugation at 1000 × g for 4 min at 4°. The pellet was resuspended in residual media, transferred to a 15-ml tube, and centrifuged as before. Cells were washed once in ice-cold breaking buffer (20 mm Tris pH 7.5, 50 mm NaCl, 10 mm MgCl2, 2 mm DTT, 0.5 mm PMSF, 0.5 mm benzamidine, 5 μmol leupeptin, 0.15 μmol aprotinin, 5 mm NaF, 50 μg/ml cycloheximide, 200 μg/ml heparin) and centrifuged at 1000 × g for 4 min at 4°. The pellet was resuspended in 1 vol of breaking buffer followed by addition of 2 vol of glass beads. Cells were vortexed for 40 sec at 4° and then placed on ice for at least 1 min; this was repeated five times. The cell suspension was transferred to microfuge tubes, and unlysed cells were removed by centrifugation at 1000 × g for 5 min at 4°. RNase A was added to some cell extracts and incubated on ice 15 min prior to centrifugation. Supernatant (typically 15 μl) was layered on a 12-ml linear 7 to 47% sucrose gradient prepared in breaking buffer lacking heparin and centrifuged for 3 hr at 39,000 rpm in a SW41 rotor at 4°. The gradients were collected using a Buchler Densi-Flow II fractionator (BWR, Bridgeport, NJ) and A254-recorded with an Isco (Lincoln, NE) UA-5 absorbance detector.
RESULTS
DOM34 sequence: DOM34 (Duplication Of Multilocus Region) is tightly linked to the centromere of chromosome XIV (Laloet al. 1993) and encodes a protein with a predicted mass of 44-kD (Figure 1A). The Dom34 protein contains three regions that display similarity to conserved motifs: (1) A putative nuclear localization signal (NLS) is located at residues 173-177. This sequence is similar to that of simian virus 40 (SV40) large T antigen NLS (Dom34p, PKKKR; SV40, PKKKRKV), which has activity in S. cerevisiae (Nelson and Silver 1989). (2) Residues 267-323 are highly similar to a portion of eukaryotic peptide chain release factor subunit 1 (ERF1). The ERF1 family of proteins is involved in the termination step of protein synthesis (Frolovaet al. 1994). (3) A putative leucine zipper motif is located at the C terminus of Dom34p. Leucine zippers have been suggested to mediate protein-protein interactions in a diverse set of functionally unrelated proteins (Busch and Sassone-Corsi 1990).
DOM34 is a member of a gene family that is conserved between kingdoms (Figure 1B). The protein product of the Drosophila gene pelota shares 32% identity and 55% similarity with Dom34p. Alignment of Dom34p and pelota suggests that the three motifs identified in Dom34p are functionally relevant (other eukaryotic Dom34p homologs, from Caenorhabditis elegans, Arabidopsis thaliana, and humans, also share these motifs).
The recently completed genomic sequencing of Methanococcus jannaschii revealed an archaebacterial homolog of Dom34p (Bultet al. 1996). Similarly, in the course of sequencing the Sulfolobus solfataricus genome, a second archae homolog was identified (Raganet al. 1996). Both sequences contain ORFs encoding proteins that are ~20% identical and 40% similar to Dom34p. Interestingly, the similarity to ERF1 is conserved in the archaebacterial proteins. Although the C-terminal 40 amino acids of the archae proteins do not contain a leucine-zipper motif, this region is 38% identical and 70% similar to Dom34p. As archaebacteria do not contain an enclosed nucleus, it is not surprising that the archae proteins do not contain a nuclear localization signal.
The presence of Dom34p homologs is not universal. The genomic sequence of five species of eubacteria has been completed (Claytonet al. 1997), and no significant sequence similarity to DOM34 has been found.
dom34Δ strains grow slowly: DOM34 function was investigated in a strain background that is able to undergo all of the developmental programs available to S. cerevisiae. A dom34 null allele (dom34::LEU2) was introduced into the R1278b strain background (Gimenoet al. 1992) by homologous recombination (Rothstein 1991). Homozygous dom34::LEU2 diploid strains form smaller colonies than an isogenic wild-type strain; this defect is more severe at 15° than at 30° (Figure 2A; dom34::LEU2 haploid strains exhibit a similar defect; data not shown). The doubling time of dom34Δ and wild-type strains growing exponentially in liquid YPAD medium was calculated. Mutant cells grow 21% slower at 30° and 65% slower at 15° (Figure 2B; doubling times at 30°: dom34Δ, 112 min, DOM34, 92 min; 15°: dom34Δ, 11.6 hr, DOM34, 7.0 hr).
Dom34p primary structure and sequence alignment. (A) Dom34p contains a putative nuclear localization sequence (NLS) at residues 173-177 (PKKR), a region of similarity to ERF1 (Frolovaet al. 1994) between amino acids 267-323 (hatched box) and a putative leucine zipper located at the C terminus of the protein (shaded box). (B) Dom34p (GenBank accession no. X77114) is shown aligned with pelota (GenBank accession no. U27197), a Drosophila protein required for spermatogenesis (Eberhart and Wasserman 1995), as well as pelA, the predicted protein product of the Methanococcus jannaschii ORF MJ0174 (Bultet al. 1996). Identical amino acids are reverse-contrasted, and similar residues are highlighted.
dom34Δ strains exhibit a G1 delay: As a first step toward elucidating the mode of action of Dom34p, cell cycle progression of dom34Δ and isogenic wild-type strains was followed morphologically. The percentage of unbudded (G1), small-budded (S), and large-budded (G2/M) cells in logarithmically growing cultures was determined. dom34Δ cultures were found to contain a significantly higher percentage of unbudded cells (Table 1; chi-square P < 0.005). These data suggest that Dom34p is required for efficient progression through, or exit from, the G1 phase of the cell cycle.
dom34::LEU2 mutants grow slowly. (A) Y743/pUN55 (DOM34/DOM34 + vector), Y 739/pUN55 (dom34Δ/dom34Δ + vector), and Y739/pTK1 (dom34Δ/dom34Δ + pCEN-DOM34) were streaked to rich medium and grown at 30° (2 days) or 15° (8 days). (B) Saturated overnight cultures of Y 739 (dom34Δ/dom34Δ and Y743 (DOM34/DOM34) were diluted 1:20 and grown to mid-log phase (~1 × 107 cells/ml) and then used to inoculate YPAD to a starting density of 1 × 105 cells/ml. At the indicated times triplicate dilutions were made from two cultures of each strain and plated to determine the number of viable cells. Open circles (○) represent dom34Δ/dom34Δ, closed circles (●) wild type. The mean ± SE is shown.
The initiation of DNA synthesis (S phase) and bud formation are closely linked but separable events (Lew and Reed 1995). Flow cytometry was performed to determine whether the increased fraction of unbudded cells in a dom34Δ strain reflected a defect in the nuclear cycle. Consistent with the morphological analysis of the cell cycle, dom34Δ mutants displayed an increased proportion of cells, relative to wild type, with a G1 content of DNA (Figure 3). The G1 delay of dom34Δ mutants, as measured by both budding and flow cytometry, is no worse at 15° than at 30°. This suggests that the cell cycle defect is not entirely responsible for the slow growth of dom34Δ strains.
G0 in dom34Δ strains: Closer examination of all phases of growth revealed that a dom34Δ strain displays a longer lag than wild type prior to resuming growth from G0 (data not shown). Furthermore, dom34Δ mutants exit the exponential phase of growth at a lower cell density than an isogenic wild-type strain; this effect is more pronounced at 15° than at 30° (Figure 2B). The most important criterion used to establish that cells properly execute the G0 state is the ability to maintain viability for an extended period of time. After the cultures in Figure 2B were incubated at their respective temperatures for 35 days, cells were plated to assay viabilities. dom34Δ mutants maintained viability as well as wild type (i.e., >90%) at both 15° and 30° (data not shown). This indicates that the ability of a dom34Δ mutant to enter G0 is not significantly perturbed.
Meiotic progression of dom34Δ/dom34Δ strains: To determine if Dom34p plays a role in sporulation, the meiotic divisions and spore formation were monitored at both 15° and 30°. Strains Y743 and Y739 were induced to undergo meiosis by transferring saturated overnight cultures grown in YPAD at 30° to prewarmed (30°) or prechilled (15°) 2% potassium acetate. At various time points cells were fixed and stained with DAPI, a DNA-specific dye. The dom34Δ strain failed to progress through the meiotic divisions, or to sporulate, to the same extent as wild-type cells (Figure 4). At later time points the dom34Δ strain does improve relative to wild type, although sporulation never reaches wild-type levels (4 days at 30°, dom34Δ 23.2% vs. wt 41.7%; 18 days at 15°, dom34Δ 4.2% vs. wt 40.1%). Similar to the growth defect, the sporulation defect is more severe at 15° than at 30°. If the sporulation defect were simply a manifestation of the growth defect, sporulation at both 15° and 30° should be similarly impaired, because the cells were grown at 30° prior to undergoing sporulation at either 15° or 30°. The increased severity of the defect at 15°, relative to that at 30°, indicates that Dom34p functions during meiotic differentiation.
Cell cycle progression in dom34Δ mutants
DOM34 is required for proper pseudohyphal growth: A third developmental fate available to S. cerevisiae is pseudohyphal growth. The switch to pseudohyphal growth is characterized by the following: (1) a polar budding pattern, (2) an elongated morphology, (3) agar invasion, and (4) G2/M cell cycle control (Gimenoet al. 1992; Kronet al. 1994; Mosch and Fink 1997). The possibility that dom34Δ cells are defective in pseudohyphal development was investigated.
Strains Y743/pUN55 (DOM34/DOM34 + vector) and Y739/pUN55 (dom34Δ/dom34Δ + vector) were induced to undergo pseudohyphal growth by streaking cells to SLAHD medium and examining colony morphology after 72 hr. Although >85% of wild-type cells exhibited normal pseudohyphal morphology (Figure 5A), <25% of dom34Δ/dom34Δ colonies displayed any pseudohyphal extensions (Figure 5B), and those that did generally had far fewer and less intricate pseudohyphae than wild type.
Flow cytometric analysis of dom34Δ/dom34Δ cell cycle defect. Saturated overnight cultures of Y739 (dom34Δ/dom34Δ) and Y 743 (DOM34/DOM34) were diluted 1:20 and grown to mid-log phase (~1 × 107 cells/ml) and then used to inoculate YPAD to a starting density of 1 × 105 cells/ml. Cultures were grown to mid-log phase at 30° and 15°. Samples were then collected and analyzed by flow cytometry. The 2N and 4N peaks represent, respectively, cells in the G1 and G2 + M phases of the cell cycle.
Morphology, invasive growth, and budding pattern have recently been shown to be independent processes, each of which is important for proper pseudohyphal development (Mosch and Fink 1997). To elucidate the role of DOM34 in pseudohyphal development, these processes were examined in mutant and wild-type cells. After 72 hr on SLAHD medium, cells were washed from the surface of the plate, and agar invasion was assayed by light microscopy. The ability of the dom34Δ/dom34Δ strain to grow invasively was impaired (Figure 5D, Table 2) relative to wild type (Figure 5C, Table 2). Furthermore, the morphology of mutant cells that did invade the agar was significantly different from wild-type cells (Table 2). Although 60% of wild-type invasive cells had an elongated morphology (length:width >2), <30% of dom34Δ/dom34Δ cells did. Finally, the budding pattern of logarithmically growing cells was examined because it has been shown that some mutations that alter this process also impair pseudohyphal growth. dom34Δ/dom34Δ cells shift the bipolar budding pattern of wild type (>70%) such that <50% of mutant cells display the bipolar pattern. These data suggest that rather than playing a direct role in the morphology, invasion, or budding processes, DOM34 acts upstream of all three.
Meiotic progression is perturbed in dom34/dom34 mutants. Y739 (dom34Δ/dom34Δ) and Y743 (DOM34/DOM34) were grown to saturation in YPAD at 30° and used to inoculate sporulation medium. Cultures were sporulated at (A) 30° and (B) 15°. Samples were fixed at the indicated times and stained with DAPI. Progression through meiosis was assessed by fluorescence and phase-contrast microscopy. Open circles (○) represent the percentage of cells that have completed meiosis I (biplus tri- and tetranucleates), closed circles (●) the completion of meiosis II (tri- and tetranucleates), and open triangles (▿) ascus formation. At each time point at least 250 cells were examined.
RPS30A is a suppressor of dom34Δ: A high-copy suppressor screen of the dom34Δ growth defect was undertaken as a first step toward elucidating the function of Dom34p. Three plasmids, which suppressed the dom34Δ growth defect, were isolated. One of these plasmids contained full-length DOM34. The other two plasmids contained overlapping genomic sequences. RPS30A, which encodes ribosomal protein S30, was determined to be responsible for the suppression.
Strains containing either vector, pDOM34, p2μ-RPS30A (high copy) or pCEN-RPS30A (low copy) were streaked for growth to YPAD plates at 15° (Figure 6). Both high- and low-copy RPS30A suppress the dom34Δ growth defect. RPS30A also suppresses the dom34/dom34 sporulation and pseudohyphal growth defects (data not shown). These data suggest that Dom34p is involved in protein translation.
The polyribosome profile of dom34Δ mutants is altered: To directly determine the effect of deleting DOM34 on translation, polysome profiles were examined. Cell extracts from dom34Δ strains containing either pDOM34HA3, vector, or pCEN-RPS30A were fractionated on a 7 to 47% sucrose gradient, and A254 was monitored. Relative to wild type (Figure 7A), dom34Δ strains exhibit a decreased level of polyribosomes and a concomitant increase in the amount of free 40S and 60S ribosomal subunits and 80S monosomes (Figure 7C). The presence of RPS30A suppressor plasmid in the dom34Δ strain shifts the polysome profile such that the amount of polyribosomes is increased to wild-type levels at the expense of 80S monosomes and free 40S and 60S subunits (Figure 7D). Addition of RNase A to extracts prior to centrifugation has been shown to almost completely eliminate polyribosomes and to increase the 80S monosome peak (Martonet al. 1997). The RNase-treated wild-type gradient (Figure 7B) confirmed the position of the 40S and 60S subunits, the 80S monosomes, and the polysomes.
DOM34 is required for normal pseudohyphal growth. DOM34/DOM34 (A and C) and dom34Δ/dom34Δ (B and D) strains were induced to undergo pseudohyphal growth by streaking cells to SLAHD medium. After 72 hr at 30°, photographs of representative colonies were taken (A and B). Cells were washed from the agar surface, and photographs of the same colonies were taken (C and D) to illustrate invasive growth.
RPS30A is a suppressor of dom34Δ. Strains Y739/pUN55 (dom34Δ/dom34Δ + vector), Y739/pTK1 (dom34Δ/dom34Δ + pCEN-DOM34), Y739/pLD132 (dom34Δ/dom34Δ + p2μ-RPS30A), and Y739/pLD135 (dom34Δ/dom34Δ + pCEN-RPS30A) were streaked for growth to SC-uracil plates at 15°. Suppression was assessed after ~10 days.
Pseudohyphal characteristics of dom34Δ mutants
The polyribosome profile of dom34Δ mutants is altered. Strains (A and B) Y739/pLD116 (dom34Δ/dom34Δ + pCEN-DOM34HA3), (C) Y 739/pUN55 (dom34Δ/dom34Δ + vector), and (D) Y739/pLD135 (dom34Δ/dom34Δ + pCEN-RPS30A) were grown to mid-log phase (~1 × 107 cells/ml). Cell extracts (A, C, and D) and RNase-treated cell extract (B) were fractionated on a 7–47% sucrose gradient and A254-monitered. The positions of the 40S and 60S ribosomal subunits, 80S monosomes, and polysomes are indicated.
Expression of pelota complements a dom34Δ mutation: Dom34p is 32% identical and 55% similar to pelota. Mutations in either gene result in defects in several developmental processes (sporulation and pseudohyphal growth in yeast and spermatogenesis and eye development in Drosophila). Furthermore, both phenotypes are suggestive of a role in the cell cycle. To determine if Dom34p and pelota are functional homologs, pY2-K1.6 (pelota cDNA under the control of the GAL1 promoter) was transformed into Y739. The resulting transformants were streaked to YPAD medium (repressing) and YPARG medium (inducing) and grown at 15°. pelota expression complemented the growth defect of dom34Δ (Figure 8). Pseudohyphal development was perturbed under inducing conditions; therefore, we were unable to determine if pelota complements this defect. However, the sporulation defect of dom34Δ mutants is complemented by pelota (data not shown). These data strongly suggest that the function of the DOM34/pelota gene family in translation is highly conserved.
Expression of pelota rescues the dom34Δ growth defect. Y743/pUN55 (DOM34/DOM34 + vector), Y 739/pYES2 (dom34Δ/dom34Δ + vector), Y739/pY2K1.6 (dom34Δ/dom34Δ + p2μ-GAL:pelota), and Y739/pLD78 (dom34Δ/dom34Δ + p2μ-GAL:DOM34; DOM34 expression is driven by both the GAL1 and DOM34 promoters) were steaked to YPAD (A) and YPARG (B) and then grown at 15°. Complementation was assessed after ~8 days.
DISCUSSION
Dom34p, a member of a highly conserved gene family, is required for S. cerevisiae to progress efficiently through the G1 phase of the cell cycle. Furthermore, dom34Δ mutants are unable to complete sporulation or pseudohyphal development properly, although their ability to enter G0 is unperturbed as judged by long-term viability. Interestingly, even in the presence of a null allele of DOM34, both the growth and sporulation defects are more severe at 15° than at 30° (pseudohyphal development could not be assessed at 15° because wild-type cells did not form pseudohyphae). One interpretation of this effect is that Dom34p stabilizes or enhances some activity that is inherently cold-sensitive. There are several examples of null mutations whose phenotypes are manifest only at low temperatures. Among the processes affected by these mutations are (1) the resumption of cell proliferation from stationary phase (gcs1Δ; Irelandet al. 1994), (2) the actin cytoskeleton (sac1Δ; Novicket al. 1989), and (3) the assembly of the 40S ribosomal subunit (drs2Δ; Ripmasteret al. 1993).
RPS30A, which encodes ribosomal protein S30, a component of the 40S ribosomal subunit (Bakeret al. 1996), was identified as a high-copy suppressor of the dom34Δ growth defect. The fact that expression of RPS30A from a centromere-based plasmid also suppresses the dom34Δ growth defect suggests that the genetic interaction between RPS30A and DOM34 may be more complex than simple bypass suppression. Taken together with the cold sensitivity of both dom34Δ and aspects of 40S assembly, this genetic interaction suggests the possibility that Dom34p mediates its effects via the translation machinery.
A role for Dom34p in translation is also suggested by a region of sequence similarity to ERF1, which is involved in the control of the termination step of translation (Frolovaet al. 1994). This region of similarity has also been identified in ribosomal proteins L30e, L7Ae/S6e, and S12e, as well as in Escherichia coli ribosomal protein modification enzyme RimK (Kooninet al. 1994).
The strongest evidence that Dom34p affects translation is provided by the analysis of polyribosome profiles. Cells harboring a deletion of DOM34 exhibit decreased levels of polyribosomes and a concomitant increase in the amount of free 40S and 60S ribosomal subunits and 80S monosomes relative to wild type. Furthermore, RPS30A suppression of the dom34Δ phenotype shifts the polysome profile such that the amount of polyribosomes is increased to wild-type levels at the expense of 80S monosomes and free 40S and 60S subunits. These data strongly suggest that the dom34Δ phenotype is the result of an impaired translation machinery.
Although an indirect role has not been ruled out, we favor the hypothesis that Dom34p plays a direct role in translation. Although the suppression by RPS30A could be interpreted to support a role for DOM34 in ribosomal protein gene (rp-gene) expression, the following arguments make that unlikely. Ribosomal protein levels are tightly regulated, primarily at the level of transcription, in order to maintain the stoichiometry necessary to ensure proper assembly of the ribosome, which consists of ~80 proteins and four ribosomal RNAs (reviewed in Plantaet al. 1995). The coordinate nature of this regulation makes it very unlikely that overexpression of RPS30A would suppress a defect in rp-gene transcription. Similarly, the promoter region of RPS30A contains two Rap1p binding sites and a T-rich poly-pyrimidine tract (Bakeret al. 1996), both of which are found in other rp-gene promoters (reviewed in Plantaet al. 1995). Again, this suggests that RPS30A expression is regulated by the same factors as other rp-genes. Therefore, if DOM34 regulated rp-gene transcription, overexpression of RPS30A alone could not suppress the dom34Δ defect. Furthermore, preliminary experiments indicate that Dom34p is distributed throughout both the cytosol and nucleus and some of the protein sediments in the 60–80S range of the gradient (data not shown). This is consistent with a role for Dom34p, which directly affects the translation machinery, perhaps by promoting assembly of the ribosomal subunits in the nucleolus and/or translationally active 80S ribosomes in the cytosol.
The Drosophila gene pelota is 32% identical and 55% similar to DOM34 at the amino acid level. The fact that pelota expression complements a dom34Δ mutation indicates that the function of the two genes is highly conserved and, by analogy, that the other eukaryotic members of the gene family are also functional homologs. pelota was identified in a screen for recessive male-sterile mutants (Castrillonet al. 1993). Male Drosophila homozygous for mutations in pelota arrest meiosis at prophase I. However, postmeiotic aspects of spermatid differentiation proceed in a nearly normal manner, resulting in the formation of a small number of spermatid with recognizable head and tail structures (Eberhart and Wasserman 1995).
The fact that pelota has a specific developmental phenotype is by no means inconsistent with a defect in ribosome function. The Drosophila gene string of pearls encodes ribosomal protein S2 and was isolated in a screen for recessive female sterile mutants (Cramton and Laski 1994). Early stages of oogenesis in homozygous females progress normally, followed by an arrest at mid-oogenesis. Similarly, the Drosophila gene aberrant immune response 8 encodes ribosomal protein S6 and acts as a tumor-suppressor gene. Mutant flies display hypertrophy of larval hematopoietic organs because of aberrant development of hemocytes (Watsonet al. 1992; Stewart and Denell 1993). These developmental roles are probably not unique to Drosophila because ribosomal proteins have been implicated in regulatory processes that may be important for tumorigenesis in humans (Ziemieckiet al. 1990; Henryet al. 1993).
The analysis of polyribosome profiles is most consistent with the hypothesis that the DOM34/pelota gene family is involved in bulk protein translation. The G1 progression and growth defects of dom34Δ mutants are consistent with this hypothesis because CDC33, which encodes the translation initiation factor eIF-4E, is required for progression through the G1 phase of the cell cycle (Brenneret al. 1988). Sporulation and pseudohyphal development may be particularly sensitive to decreased bulk protein translation or the consequent effect on the cell cycle. However, it is possible that the translation of a subset of proteins involved in these developmental pathways is more specifically affected by the action of Dom34p. Experiments to determine the effect of the dom34Δ mutation on translation under different developmental conditions should help to differentiate between these possibilities.
Acknowledgments
We would like to thank many members of the SUNY, Stony Brook, community for invaluable assistance: N. Hollingsworth, R. Sternglanz, and A. Sutton for comments on the manuscript; C. Davis, S. Rudge, S. Strickland, and J. Trimmer for helpful discussion; J. Konopka for plasmids and reagents; members of the Williams lab for assistance with polysome profiles and T. Kessel for technical support. We are also grateful to C. Eberhart and S. Wasserman (University of Texas Southwestern) for providing the pelota plasmid. J.E. was a recipient of an American Cancer Society Junior Faculty Research Award during this study. This work was supported by National Institutes of Health Grant GM-48639 to J.E.
Footnotes
-
Communicating editor: F. Winston
- Received August 4, 1997.
- Accepted February 9, 1998.
- Copyright © 1998 by the Genetics Society of America
