Isolation and Characterization of WHI3, a Size-Control Gene of Saccharomyces cerevisiae

Corresponding author: B. Futcher, Department of Molecular Genetics and Microbiology, State University of New York, Stony Brook, NY 11794-5222., bfutcher{at}ms.cc.sunysb.edu (E-mail)

Communicating editor: M. JOHNSTON

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

WHI3 is a gene affecting size control and cell cycle in the yeast Saccharomyces cerevisiae. The whi3 mutant has small cells, while extra doses of WHI3 produce large cells, and a large excess of WHI3 produces a lethal arrest in G1 phase. WHI3 seems to be a dose-dependent inhibitor of Start. Whi3 and its partially redundant homolog Whi4 have an RNA-binding domain, and mutagenesis experiments indicate that this RNA-binding domain is essential for Whi3 function. CLN3-1 whi3 cells are extremely small, nearly sterile, and largely nonresponsive to mating factor. Fertility is restored by deletion of CLN2, suggesting that whi3 cells may have abnormally high levels of CLN2 function.

MOST eukaryotic cells become committed to cell division at a control point called the restriction point, the commitment point, or, in yeasts, Start (HARTWELL et al. 1974 ; PARDEE 1989 ; ZETTERBERG et al. 1995 ; NEUFELD and EDGAR 1998 ). In some organisms, including the yeasts Saccharomyces cerevisiae and Schizosaccharomyces pombe, growth to a critical cell volume is a prerequisite for Start (JOHNSTON et al. 1977 ; MITCHISON and NURSE 1985 ; NASH et al. 1988 ). Mutations that alter the critical volume required for Start have defined important cell cycle control genes, including the G1 cyclin gene CLN3 (originally called WHI1; CARTER and SUDBERY 1980 ; SUDBERY et al. 1980 ; CROSS 1988 ; NASH et al. 1988 ). Of particular interest are genes that affect critical volume in a dose-dependent way. For instance, two copies of CLN3 cause a smaller critical volume than one copy, and higher doses can decrease volume still further (NASH et al. 1988 ). One interpretation of dose dependence is that the amount of gene product is being titrated against some other metric (e.g., nuclear volume, amount of DNA) and the ratio is sensed by the cell.

Critical cell volume is not the only requirement for Start. For instance, cells can pass through Start only if they are growing and making protein. However, a wide range of protein synthesis rates is acceptable for Start, whereas the critical volume is relatively well defined. Thus it appears that volume is measured by the cell as one requirement for Start.

In S. cerevisiae, Start depends on the Cdc28 protein kinase and the G1 cyclins Cln1, Cln2, and Cln3 (RICHARDSON et al. 1989 ). Cdc28 becomes activated by associating with a cyclin. In small G1 cells, the only cyclin known to be present is Cln3, which is expressed relatively constitutively through the cell cycle. Cln3 forms a complex with Cdc28, and this protein kinase complex somehow, perhaps in a size-dependent way, activates the transcription factors SBF and MBF (TYERS et al. 1993 ; KOCH and NASMYTH 1994 ). These in turn cause the transcription of 200 genes in late G1 phase, including CLN1 and CLN2 (SPELLMAN et al. 1998 ). Cln1-Cdc28 and Cln2-Cdc28 then presumably phosphorylate other substrates such as Sic1, and these phosphorylation events constitute Start.

Many environmental signals regulating Start do so by influencing the synthesis or activity of the G1 cyclins. Cln1, Cln2, and Cln3 are each unstable proteins with half-lives of 10 min (SCHNEIDER et al. 1998 ), that are encoded by unstable mRNAs. This instability makes Cln levels very responsive to changing environmental conditions. Thus, a drop in the rate of protein synthesis quickly reduces Cln levels, and this is probably important for the cell's ability to arrest in G1 in response to starvation. Nutrients are known to affect cell size, and this is partly due to signaling through the RAS-cAMP pathway to Cln synthesis (TOKIWA et al. 1994 ). Mating pheromones arrest cells in G1 phase because signaling through the mating pathway shuts off the synthesis of Cln1 and Cln2, and perhaps their activity as well (GARTNER et al. 1998 ).

There is still no clear idea of how critical size is measured by the cell. That is, why do cells undertake Start at some cell sizes but not at others? What happens during the long G1 phase of a slowly growing daughter cell that eventually allows Start? We have taken a genetic approach to this problem by seeking new mutations that affect critical cell size. Mutations that allow cells to undertake Start at smaller critical cell sizes will yield strains that have small average cell sizes—a Whi^- phenotype. Such screens have been daunting because of difficulty in cloning wild-type genes whose only mutant phenotype is small cell size. We have done our screen using a marked Ty element as the mutagen, allowing direct cloning of the affected gene.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Strains and media:
S. cerevisiae strains used are listed in Table 1. Yeast were grown in YEP medium (1% yeast extract, 2% peptone) with 2% glucose (YEPD), or 2% glycerol (YEP glycerol), or 2% galactose (YEPGal), or 1% raffinose plus 1% galactose (YEPraf + gal). For scoring auxotrophic markers, or for selection of plasmid-borne markers, yeast were grown in drop-out synthetic complete (SC) media (SHERMAN 1991 ).

Molecular techniques:
Standard methods of molecular biology were used. New plasmid constructions involving WHI3 DNA typically yielded many clones with heterogeneous rearrangements at or near the WHI3 locus. However, once a plasmid had been constructed, it was reasonably stable.

Site-directed mutagenesis:
whi3RRM was constructed by PCR mutagenesis using oligos whi3RRM (AAC CCT GCT GAT CAA AAT CCA CCT TGT AAT TTT TCA AAA AAC CCA CTG GGT GTT AGG GG) and whi3RRM11-2 (CCC CTA ACA CCC AGT GGG TTT TTT GAA AAA TTA CAA GGT GGA TTT TGA TC). This resulted in a deletion of 83 amino acids including RNP1 and RNP2. whi3-Y541A was constructed using the QuickChange mutagenesis kit (Stratagene, La Jolla, CA) and primers Y-A1 and Y-A2. whi3-F586A was similarly constructed using primers F-A1 and F-A2. whi3QR-2 was constructed by PCR mutagenesis using oligos whi3Q and whi3Q-2. This resulted in a deletion of 100 amino acids from S201 to Q300. Oligo sequences are available upon request.

Isolation of mutants:
Ty transposon tagging and plasmid pJEF1105 were used for mutagenesis (GARFINKEL et al. 1988 ; GARFINKEL and STRATHERN 1991 ). Plasmid pJEF1105 carries Ty1 marked with neomycin phosphotransferase, which confers kanamycin resistance in Escherichia coli and G418 resistance in yeast. The transposase of the Ty1 in pJEF1105 is under control of the GAL1 promoter, so galactose induces transposition of the tagged element into novel positions.

Yeast strain BWG1-7a (J. Boeke; Table 1) was transformed with pJEF1105. Transformants were spread on 20 SD-ura galactose plates at 1000 cells per plate. Plates were incubated at 22° for 5 days. Colonies were heterogeneous in size and appearance, presumably due to Ty transposition. Colonies were scraped from the plates and pooled. A sample from the pooled cells was used to inoculate a culture in YEPD, a glucose medium in which further transposition was repressed.

Early exponential cells were harvested, sonicated, and 10⁹ cells were layered onto a 1.0–1.75% (w/v) sorbitol gradient in a "stay-put," which separates cells according to size (PETERSON and EVANS 1967 ). Larger cells fall through the gradient faster than smaller ones. After 3 hr of sedimentation, 3% (w/v) sorbitol was pumped into the bottom of the tank, and fractions were collected from the top. Cell size was analyzed using a Coulter (Hialeah, FL) channelyzer. Fractions containing the smallest 5% of cells were pooled and used to reinoculate a YEPD culture. The size enrichment procedure was repeated two more times.

After the third size enrichment, the smallest 5% of cells were spread on YEPD plates. Individual colonies were chosen and cell sizes in 100 cultures were assayed using a Coulter channelyzer. In 18 cultures, cells had an unusually small mode cell size. Four mutants were sterile, or would not sporulate after crossing, and in five cases, the small cell phenotype did not segregate 2:2. These nine isolates were discarded. Small cell size segregated 2:2 in crosses for the remaining nine mutants, and these were analyzed further.

Crosses to WHI1-1 (= CLN3-1) showed that none of the nine mutants was allelic to WHI1. The mutants were also crossed to each other, and in each cross small cell size segregated 4 small:0 wild type, indicating that all nine mutants were alleles of one gene that we named WHI3.

Mutants were repeatedly backcrossed to wild-type BF328-4a or -4b to see if G418 resistance cosegregated with the Whi^- phenotype. In the first cross and first backcross, nearly all spore clones were G418 resistant, indicating that each mutant contained many G418-tagged Ty1 elements. This was confirmed by Southern analysis. The original isolates had an average of about eight G418-tagged Ty elements, and also several untagged Ty1 elements at novel positions. Southern analysis showed that Ty elements were in different locations in different whi3 mutants, suggesting that the mutants were independent.

Only in one case (isolate 9, strain T404) was there a strict cosegregation of small size with G418 resistance through four backcrosses. An example of the 2:2 segregation for size is shown in Fig 1. It may be that in the other eight isolates, an unmarked Ty1 element had transposed into the WHI3 gene. All further analysis was restricted to the descendants of strain T404.

View larger version (27K): In this window In a new window Download PPT slide   Figure 1. Cosegregation of the Whi- phenotype with resistance to G418. The four spore clones of tetrad T504b-1 were grown to early exponential phase in YEPD, placed on ice, and sonicated. Cell volumes were measured using a Coulter channelyzer. Cell volume distributions are plotted as cell volume (in femtoliters) vs. cell number. Mode cell volumes for the indicated curves are shown in parentheses. The two Whi- clones were resistant to G418, while the two Whi+ clones were sensitive to G418.

Cloning of whi3::TY1::neoR and WHI3:
We cloned the whi3::TY1::neoR locus by integrating plasmid pRS305neo into the neoR gene of strain T404 by recombination (Fig 2). The DNA of this strain was digested with EcoRI, self-ligated, and used to transform E. coli to both ampicillin and kanamycin resistance. These clones contained yeast genomic DNA flanking the 3' end of the Ty1 element, and this was used as a probe to clone the wild-type WHI3 locus and the whi3::TY1::neo locus. The Ty1::neo element was inserted in open reading frame YNL197c (Fig 2A); the region of insertion was sequenced. The element was surrounded by a duplication of ATAAC (normally found at positions +285 to +289); such a 5-bp duplication is typical of Ty1 insertions.

View larger version (88K): In this window In a new window Download PPT slide   Figure 2. The WHI3 gene and gene product. (A) To clone the WHI3 locus, pRS305neo was cut with XhoI and integrated into the neo region of the marked Ty element in strain T504b-1c. Genomic DNA from the integrant was cut with EcoRI, recircularized, and used to transform E. coli to ampicillin resistance. A portion of the WHI3 gene was obtained. (B) The Whi3 protein contains an RNA recognition motif (RRM). The putative RRM of Whi3 is aligned with other representative RRMs. The octamer RNP1 and the hexamer RNP2 are shown boxed in black. Some proteins have multiple RRMs; these are numbered from the N terminus (e.g., yeast NSR1 nos. 1 and 2). The top line of the "consensus" shows the most common amino acid at each position; the second line shows a common alternative amino acid. The consensus is based on RRMs in addition to those shown here. (C) Proteins with significant similarity to Whi3. Whi3 is compared to its closest homologs. Whi3 is divided into three domains: a long N-terminal domain of 510 amino acids (shaded); a C-terminal RNA-binding domain of 125 amino acids (white); and a C-terminal tail of 26 amino acids (black). The percentage identity of homologs of Whi3 is shown, with identity to the N-terminal domain shown on the left and identity to the RNA-binding domain on the right. The homologs may have additional C-terminal sequences (not shown). The identity between these RNA-binding domains is much higher than between randomly chosen RNA-binding domains (B).

Quantitative mating:
A wild-type tester strain was mated with a control strain and various mutant strains whose mating efficiency was to be determined. Wild-type cells were mixed in a 3:1 ratio with mutant cells, resulting in a total cell number of 1 x 10⁷. Cells were collected onto 25-mm-diameter nitrocellulose filter disks with a pore size of 0.45 µm. The filters were placed on YEPD plates (cells up) and incubated at 30° for 5 hr. The mating mix was resuspended in 5 ml of sterile water, sonicated, diluted, and spread on plates. The numbers of diploid cells, experimental haploid cells, and total cells were determined using appropriate selective plates (e.g., SC-his to select for diploids). Raw mating efficiency was calculated as the number of diploid cells divided by the number of diploid plus experimental cells. The efficiency obtained for a wild-type x wild-type cross was assigned a value of 1.0, and results from mutant x wild-type crosses were normalized to this.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Isolation of the whi3 mutant
Transposon mutagenesis followed by physical enrichment for small cells was used to isolate small cell-size mutants (MATERIALS AND METHODS). Transposons were used as mutagens because they tag the mutant gene, and so allow cloning. We obtained nine apparently independent alleles of the same gene, which we called WHI3. In isolate 9, strain T404, the Whi^- phenotype and G418 resistance cosegregated perfectly (e.g., Fig 1). The Ty1::neo element was inserted inside open reading frame YNL197c (Fig 2A).

To show that YNL197c is WHI3, the wild-type YNL197c open reading frame was cloned into a CEN (centromere) and an integrating vector and shown to fully complement the whi3 small cell phenotype in both cases. In addition, a fragment internal to the open reading frame, from the SphI site at the initiator codon to a BstEII site 58 nucleotides upstream of the stop codon, was deleted and replaced with LEU2. Transplacement of this null allele into a diploid yielded a heterozygote. Sporulation gave a 2:2 segregation for small cell size that was indistinguishable from the original whi3::Ty1::neo mutant phenotype. Small cell size cosegregated with Leu⁺, and Southern blot analysis confirmed that the spore clones with small cell size contained the deletion/insertion in YNL197c. Thus WHI3 is a nonessential gene, and null alleles cause small cell size.

Predicted properties of Whi3: The protein predicted from the WHI3 sequence has a mass of 71,257 kD, an isoelectric point of 8.65, and a codon bias of 0.13 [Yeast Proteome Database (COSTANZO et al. 2000 )], suggesting it is not an abundant protein.

Computer searches found an RNA-binding motif (Fig 2B) called the RNA recognition motif (RRM; BIRNEY et al. 1993 ). An RRM is 80–90 amino acids long and is defined by conserved motifs called RNP-1 and RNP-2. The Whi3 motif has only moderate similarity to RNP-1, but strong similarity to RNP-2. Structural predictions suggest that the Whi3 motif can adopt the same structure as found in confirmed RRM proteins (E. BIRNEY and A. KRAINER, personal communication).

Whi3 has sequence similarity to many RNA-binding proteins, but the similarity is usually confined to the RRM. However, a few RRM proteins have longer stretches of similarity extending toward their N-terminal regions. These include an S. cerevisiae gene we have named WHI4 (YDL224c, see below) and two genes from S. pombe, named SPCC16C4.07 (about which nothing is known) and mde7 (induced during meiosis in a mei4-dependent fashion, ABE and SHIMODA 2000 ; PombePD, Proteome; Pombe Proteome Database). In addition, the Caenorhabditis elegans protein Ceptb-1 (a homolog of mammalian polypyrimidine tract-binding protein) and the Drosophila melanogaster protein couch potato (involved in neural development) may be weak homologs (WormPD; Worm Proteome Database; COSTANZO et al. 2000 ; Fig 2).

A second prominent feature is a glutamine-rich region between amino acids 247 and 277, where 19 of 31 residues are glutamines. One putative Whi3 homolog, the Drosophila protein couch potato, has a similarly placed glutamine-rich region.

Characterization of WHI3
WHI3 is a dose-dependent modulator of cell size: As shown in Fig 1, whi3 cells are smaller than wild-type cells. One possible reason for small cell size is low growth rate. However, whi3 cells grew at the same rate as wild type in all tested media. In glycerol medium, where the growth rate for both wild-type and whi3 cells was low, the size difference between wild type and whi3 was accentuated (Fig 3). At low growth rates, the asymmetry between mother and daughter cells increases, causing a biphasic distribution of cell size (Fig 3).

View larger version (22K): In this window In a new window Download PPT slide   Figure 3. The Whi- phenotype of whi3 is maintained at low growth rates. Strain T504b-1a (WHI3) and strain T504b-1c (whi3) were grown to early exponential phase in YEP glycerol. Cell size distributions were measured and plotted as in Fig 1. Mode cell volumes of the daughter cell peak (left) and the mother cell peak (right) of each curve are shown in parentheses.

To see if the effect of WHI3 on cell size is dose dependent, we compared the cell sizes of three diploid strains. A homozygous WHI3/WHI3 diploid had a mode cell volume of 78 fl; a heterozygous WHI3/whi3 diploid had a mode cell volume of 70 fl; and a homozygous whi3/whi3 diploid had a mode cell volume of 61 fl. Furthermore, when zero, one, two, or three copies of the WHI3 gene were present in a haploid strain, the mode cell volume increased 5 fl with each dose of WHI3 (Fig 4). Both results indicate that the effect of WHI3 is dose dependent.

View larger version (31K): In this window In a new window Download PPT slide   Figure 4. The effect of WHI3 on size is dose dependent. Haploid cells carrying zero, one, two, or three copies of the WHI3 gene were grown to early exponential phase in YEPD. Cell size distributions were measured and plotted as in Fig 1. Mode cell volumes of each curve are shown in parentheses.

The RNA recognition motif is essential for Whi3 function: We asked about the significance of the putative RRM. The RRM was altered by site-directed mutagenesis to generate three mutants: whi3RRM, in which the RRM domain was deleted; whi3-Y541A, in which a tyrosine in RNP2 was changed to alanine; and whi3-F586A, in which a phenylalanine in RNP1 was changed to alanine. The aromatic residues typically found in RNP1 and RNP2 contribute to interaction with RNA, and so these point mutants might weaken RNA binding. The three mutant alleles were transplaced into otherwise wild-type cells, and phenotypes were assayed. All three RRM mutants were phenotypically indistinguishable from a whi3 null mutation (Fig 5).

View larger version (39K): In this window In a new window Download PPT slide   Figure 5. The RRM is essential for WHI3 function. (A) Isogenic strains TV242 (WT), TV259 (whi3::URA3), TV254 (whi3RRM), TV307 (whi3-Y541A), TV306 (whi3-F586A), and TV248 (whi3QR-2) were grown in YEPD, sonicated, and cell size was measured. (B) Strain W303 was transformed with an empty GAL vector, GALWHI3, GAL-WHI3QR-2-3xHA (QR), GAL-whi3RRM-4-3xHA (RRM), or GAL-WHI3-3xHA, and streaked on YEP raffinose or YEP galactose plates. (C) The amount of protein expressed by the GAL-WHI3-3xHA, GAL-whi3RRM-4-3xHA, and GAL-WHI3QR-2-3xHA constructs was assayed by Western blotting, using anti-HA antibody 12CA5.

Each of the RRM mutants was overexpressed from a GAL promoter. Overexpression of the whi3RRM mutant (Fig 5) or the whi3-Y541A mutant (data not shown) had no effect on cells, unlike overexpression of wild-type WHI3, which was lethal (Fig 5 and see below). Overexpression of the whi3-F586A mutant caused slow growth, but was not lethal. We conclude that the RRM region, and so probably RNA binding, is essential for the function of Whi3.

We also deleted the glutamine-rich region. The WHI3QR-2 allele was wild type for the size-control function of WHI3 (Fig 5), and, like wild-type WHI3, caused lethality when overexpressed (Fig 5). Thus the glutamine-rich region was dispensable for both tested functions of WHI3.

Genetic interaction of whi3 with whi4: Open reading frame YDL224c, which we call WHI4, is related to WHI3, particularly around the RRM. We constructed a whi4 null mutant. This had little if any cell size phenotype (a mode size of 44 fl vs. 45 fl for wild type; Fig 6) and no obvious other phenotype. However, the whi3 whi4 double mutant had a more extreme Whi^- phenotype than did whi3 (a mode size of 29 fl vs. 35 fl for whi3; Fig 6). The double mutant also had a modest but noticeable slow-growth phenotype (which is not seen in CLN3-1, which is similar in size to whi3 whi4). Thus, WHI4 is partially redundant with WHI3, but WHI3 is the more important gene.

View larger version (32K): In this window In a new window Download PPT slide   Figure 6. whi4 is partially redundant with whi3. TV242 (WT), TV259 (whi3), TV206-4a (whi4), and TV207-12a (whi3 whi4) were grown in YEPD, sonicated, and cell size was measured using a Coulter channelyzer. Mode cell volumes are shown in parentheses.

Genetic interaction of whi3 with CLN3-1: The whi3 CLN3-1 double mutant has a mode cell volume of 22 fl (Fig 7), which is about one-half the volume of isogenic wild-type cells and significantly smaller than either single mutant. In fact, 22 fl is the size expected if the effects caused by whi3 and CLN3-1 are multiplied: the whi3 mutant has 0.82 the volume of a wild-type cell; the CLN3-1 mutant has 0.68 the volume of a wild-type cell; and 0.82 x 0.68 x 41 fl = 23 fl. Thus, the double mutant is the size expected if whi3 and CLN3-1 affect size independently of each other. Flow cytometry showed that only 15% of the cells in a whi3 CLN3-1 population are 1N, as compared to 40–50% in a wild-type population. This is consistent with the idea that Start occurs at a small size, and consequently cells spend most of their time in S, G2, or M.

View larger version (31K): In this window In a new window Download PPT slide   Figure 7. CLN3-1 and whi3 have multiplicative effects on cell size. The four spore clones of the tetratype tetrad T654b-4 were grown to midexponential phase in YEPD. Cell size distributions were measured and plotted as in Fig 1. Mode cell volumes of each curve are shown in parentheses.

The whi3 CLN3-1 mutants were unhealthy and slow growing, and all examined double-mutant cultures accumulated faster-growing larger cells, which proved to be diploids homozygous for mating type. These diploids were similar in size to wild-type haploids. We do not know whether these diploids arise at a high rate, or whether there is selection for rare diploids that may arise in any culture.

The whi3 single mutant is slightly (approximately three-fold) resistant to -factor. CLN3-1 mutants are highly resistant, in the sense that they eventually proliferate even in the presence of high concentrations of -factor. However, when first challenged with pheromone, they respond similarly to wild-type cells and pause for 1–2 hr before resuming proliferation. In contrast, whi3 CLN3-1 double mutants seemed to have little or no response to -factor. In -factor, whi3 CLN3-1 cell number increased exponentially with no detectable pause (Fig 8A), cell size failed to increase (a sensitive indicator of a cell cycle arrest; Fig 8B), the proportion of cells with 1N DNA failed to increase (Fig 8C), and -factor-responsive genes failed to be induced (Fig 8D). There was no shmooing or other morphological change (not shown). Finally, whi3 CLN3-1 cells were nearly sterile (Table 2), whereas single mutants were fertile.

View larger version (76K): In this window In a new window Download PPT slide   Figure 8. CLN3-1 whi3 double mutants are defective in -factor response. Isogenic WT, CLN3-1, whi3, and CLN3-1 whi3 derivatives of strain RN210-13d were studied. Each strain was grown to early exponential phase in YEPD, and then the four cultures were treated with 3 µΜ -factor. (A) Cell number per milliliter as a function of time. (B) Cell volume (assayed using a Coulter channelyzer): an increase in cell volume with time indicates a cell cycle arrest. (C) DNA content (assayed by flow cytometry): an increase in the proportion of 1N cells indicates arrest in G1. (D) Northern analysis of mRNAs extracted from cultures at various times (in minutes) after addition of -factor. Probes were for WHI3 mRNA, the -factor-inducible gene FUS1, and the loading control ADH1. FUS1 was slightly induced in the whi3 CLN3-1 culture, though this may not be apparent in the figure.

Because overexpression of CLN2 reduces cell size and also reduces or eliminates the -factor response (OEHLEN and CROSS 1994 ), we wondered whether the whi3 CLN3-1 double mutants were nonresponsive to -factor because of an increase in CLN2 expression. A whi3 CLN3-1 cln2 triple mutant regained fertility (Table 2), consistent with the idea that the whi3 CLN3-1 mutant may overexpress CLN2, or otherwise have increased Cln2 activity.

Overexpression of WHI3 causes a G1 arrest: WHI3 expressed from the GAL1 promoter was lethal. Cells ceased division within 2–3 hr of galactose induction (Fig 9). The arrested cells were unbudded (data not shown) and had a 1N DNA content (Fig 9). They had a single nucleus and a microtubule array typical of G1 phase cells (data not shown). Thus, the cells arrested in G1 phase. Despite the cell cycle arrest, cells continued to increase in volume (Fig 9), showing that only cell cycle progression was blocked and not cell metabolism in general.

View larger version (35K): In this window In a new window Download PPT slide   Figure 9. Overexpression of WHI3 causes a lethal G1 arrest. Wild-type strain BWG1-7a carrying plasmid pCGW2 (CEN GAL-WHI3) was grown to midexponential phase in SC-ura medium containing raffinose. The culture was split into two parts, and 1% (final) galactose was added to one. (A) Cell number ceases to increase 4–5 hr after addition of galactose. (B) Cell volume increases rapidly (an indicator of cell cycle arrest) 2 hr after addition of galactose. (C) DNA content was assayed in galactose-treated and control cells by flow cytometry. G1 indicates a 1N DNA content, while G2/M indicates 2N DNA content. In the galactose-treated cells, there is a peak of low-fluorescence material at the left end of the distribution caused by sonication and breakage of the large arrested cells.

We attempted to suppress the lethality of GAL-WHI3 by overexpressing G1 cyclins, which promote the G1/S transition. GAL-CLN3-1, GAL-CLN2, and ADH1-CLN2 were transformed into a GAL-WHI3 strain, but none of these constructs suppressed the lethality. However, all three constructs partially suppressed the G1 arrest phenotype to a greater or lesser extent, such that a significant proportion of the arrested cells had buds and had 2N DNA content. Deletion of FAR1 also failed to suppress the lethality of GAL-WHI3 (not shown).

whi3 mutants are sensitive to cycloheximide: Start is sensitive to cycloheximide, perhaps because a critical rate of protein synthesis is required for accumulation of G1 cyclins (MOORE 1988 ; KO and MOORE 1990 ; POLYMENIS and SCHMIDT 1997 ). Whi3 is probably an RNA-binding protein, and so could be involved in translation, the cycloheximide-sensitive process. Therefore we examined the response of whi3, cln3, and CLN3-1 to cycloheximide. The whi3 mutants had increased sensitivity to cycloheximide (Fig 10). However, doses of WHI3 beyond one had no effect (Fig 10E). Neither the CLN3-1 nor the cln3 mutation affected cycloheximide sensitivity, showing that sensitivity is not an indirect effect of cell size. Curiously, the whi3 cln3 double mutant had the same sensitivity as the cln3 mutant; i.e., for this phenotype, cln3 is epistatic to whi3.

View larger version (76K): In this window In a new window Download PPT slide   Figure 10. The whi3 mutant is sensitive to cycloheximide. (A–D) Isogenic derivatives of strain RN210-13d were grown in YEPD and sonicated. A total of 2000 cells were placed into each spot on plates containing 0 (control), 0.2, or 0.3 µΜ cycloheximide. (E) The WHI3 strain RN100-1b, the whi3 strain RN100-1d, and derivatives of RN100-1d containing one or two copies of WHI3 integrated at the URA3 locus were grown to midexponential phase and tested as above.

Regulation of WHI3 transcription: WHI3 is modestly cell cycle regulated, with a peak in G2 phase (SPELLMAN et al. 1998 ). Its expression is not strongly affected by the diauxic shift (DERISI et al. 1997 ) or by sporulation (CHU et al. 1998 ).

The level of WHI3 mRNA declines noticeably by 1 hr after treatment of cells with -factor and is down to one-third or one-quarter of its original level after 4 hr (not shown). After release from the -factor block, WHI3 levels promptly rise (not shown, but see SPELLMAN et al. 1998 ). These effects are partly due to the modest cell cycle regulation of WHI3 (i.e., the -factor arrests cells in G1, near the trough of WHI3 transcription). In addition, WHI3 transcription may be down-regulated during -factor arrest to promote recovery, since WHI3 is an inhibitor of Start.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Previous work identified two size-control genes in S. cerevisiae, CLN3 (formerly WHI1) and WHI2 (SUDBERY et al. 1980 ). CLN3 regulates cell size because it is a dose-dependent inducer of Start: extra CLN3 allows Start to occur at a relatively small cell size and so reduces average cell size. CLN3-1 was found in a genetic screen for Whi^- (small cell) mutants (CARTER and SUDBERY 1980 ). Although the CLN3-1 mutation was identified in 1980, the gene was not cloned until much later (NASH et al. 1988 ), partly because there was no selection available for the wild-type gene.

The second size-control mutation found was whi2. This was found as a naturally occurring mutation noted as causing small cells (SUDBERY et al. 1980 ). The whi2 mutation causes small cell size in cultures as they enter stationary phase, but does not affect cell size in cultures growing exponentially in glucose. Its relationship to size control at Start is unclear.

We have now found a third size-control gene, WHI3. The fact that our screen identified only a single locus is unsatisfying. Other loci (e.g., CLN1, CLN2, and CLN3) can lead to a Whi^- phenotype under appropriate conditions, but were not identified. The Ty mutagenesis may have yielded only a single locus partly because it gives mainly null mutations and partly because Ty is biased in its sites of insertion. It is likely that there are more genes that could give a Whi^- phenotype.

Whi3 appears to have an RNA-binding domain of the RRM type found in many other proteins, such as the mammalian polypyrimidine tract binding (PTB) proteins. The RNA-binding domain is essential for function, since three different mutations in the RRM create null alleles of WHI3.

A gene with a function that may be related to that of WHI3 was isolated by SUGIMOTO et al. 1995 , who screened for high-copy number suppressors of the mating defect caused by CLN3-2. They isolated the gene HMD1 (commonly known as NAB3), which has an RNA-binding domain and has been shown to associate physically with nuclear poly(A) RNA. Overexpression of HMD1 suppressed the mating defect of CLN3-2 strains by reducing the level of CLN3-2 mRNA, but not CLN2 mRNA.

Although we do not know how WHI3 affects cell size, the fact that deletion of CLN2 suppressed the sterility of a whi3 CLN3-1 mutant suggests that WHI3 may be an inhibitor of the expression or activity of CLN2 and possibly CLN1. According to this theory, extra doses of WHI3 increase cell size because the extra WHI3 inhibits CLN2 and CLN1 expression or activity. Indeed, preliminary results suggest that CLN2 mRNA levels are lowered by excess WHI3 (E. GARI, T. VOLPE, H. WANG, C. GALLEGO, B. FUTCHER and M. ALDEA, unpublished results). In contrast, the whi3 mutation would relieve inhibition of CLN1 and CLN2 and would allow greater CLN1 and CLN2 activity at smaller cell sizes, thus allowing Start at smaller cell sizes. Thus, the whi3 mutant would have small cells because of higher expression or activity of these G1 cyclins. In combination with CLN3-1, the whi3 mutation would allow such high expression or activity of CLN2 as to cause sterility and loss of -factor responsiveness.

Since Whi3 is likely to be an RNA-binding protein, it is possible that it directly inhibits CLN2 expression by binding to the CLN2 mRNA. It could then inhibit translation of Cln2 and promote turnover of the CLN2 mRNA, etc. Alternatively, Whi3 could bind to and inhibit the CLN3 mRNA, since CLN3 is important for activating the transcription factors SBF and MBF, which in turn induce expression of CLN2, CLN1, and other genes (TYERS et al. 1993 ; DIRICK et al. 1995 ; STUART and WITTENBERG 1995 ; SPELLMAN et al. 1998 ). A hint that CLN3 might be the target of WHI3 is that cln3 is epistatic to whi3, at least for cycloheximide sensitivity (Fig 10); i.e., if CLN3 is not present, then WHI3 has no effect. A further hint is that whi3 mutants are unable to undertake pseudohyphal growth (MOSCH and FINK 1997 ), while cln3 mutants are enhanced for pseudohyphal growth (LOEB et al. 1999 ), which is again consistent with the idea that WHI3 opposes the action of CLN3.

If WHI3 does act as an inhibitor of CLN3 function, this must not be its only mode of action, since GAL-WHI3 is lethal and a cln3 deletion is not. Furthermore, GAL-WHI3 is lethal even when CLN2 or CLN3 are overexpressed, arguing that excess WHI3 may interfere with other essential targets. There may be several such targets, since we were not able to identify any high-copy suppressor of WHI3 overexpression (data not shown).

	ACKNOWLEDGMENTS

We thank Laurie Littlepage for assistance with the high-copy suppression experiments, J. Boeke and P. Sudbery for plasmids and strains, A. Krainer for helpful discussions on RNA-binding proteins, and Marti Aldea for helpful discussions on WHI3 and for communicating results prior to publication. This work was supported by grant RO1 GM45410 from the National Institutes of Health to B.F.

Manuscript received September 20, 2000; Accepted for publication December 22, 2000.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

ABE, H. and C. SHIMODA, 2000  Autoregulated expression of Schizosaccharomyces pombe meiosis-specific transcription factor Mei4 and a genome-wide search for its target genes. Genetics 154:1497-1508[Abstract/Free Full Text].

BIRNEY, E., S. KUMAR, and A. R. KRAINER, 1993  Analysis of the RNA-recognition motif and RS and RGG domains: conservation in metazoan pre-mRNA splicing factors. Nucleic Acids Res. 21:5803-5816[Abstract/Free Full Text].

BOEKE, J. D., H. XU, and G. R. FINK, 1988  A general method for the chromosomal amplification of genes in yeast. Science 239:280-282[Abstract/Free Full Text].

CARTER, B. L. and P. E. SUDBERY, 1980  Small-sized mutants of Saccharomyces cerevisiae. Genetics 96:561-566[Abstract/Free Full Text].

CHU, S., J. DERISI, M. EISEN, J. MULHOLLAND, and D. BOTSTEIN et al., 1998  The transcriptional program of sporulation in budding yeast. Science 282:699-705. (Erratum: Science 282: 1421).[Abstract/Free Full Text].

COSTANZO, M. C., J. D. HOGAN, M. E. CUSICK, B. P. DAVIS, and A. M. FANCHER et al., 2000  The yeast proteome database (YPD) and Caenorhabditis elegans proteome database (WormPD): comprehensive resources for the organization and comparison of model organism protein information. Nucleic Acids Res. 28:73-76[Abstract/Free Full Text].

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[Abstract/Free Full Text].

DERISI, J. L., V. R. IYER, and P. O. BROWN, 1997  Exploring the metabolic and genetic control of gene expression on a genomic scale. Science 278:680-686[Abstract/Free Full Text].

DIRICK, L., T. BOHM, and K. NASMYTH, 1995  Roles and regulation of Cln-Cdc28 kinases at the start of the cell cycle of S. cerevisiae. EMBO J. 14:4803-4813[Medline].

GARFINKEL, D. J. and J. N. STRATHERN, 1991  Ty mutagenesis in Saccharomyces cerevisiae. Methods Enzymol. 194:342-361[Medline].

GARFINKEL, D. J., M. F. MASTRANGELO, N. J. SANDERS, B. K. SHAFER, and J. N. STRATHERN, 1988  Transposon tagging using Ty elements in yeast. Genetics 120:95-108[Abstract/Free Full Text].

GARTNER, A., A. JOVANOVIC, D. I. JEOUNG, S. BOURLAT, and F. R. CROSS et al., 1998  Pheromone-dependent G1 cell cycle arrest requires Far1 phosphorylation, but may not involve inhibition of Cdc28-Cln2 kinase, in vivo. Mol. Cell. Biol. 18:3681-3691[Abstract/Free Full Text].

HARTWELL, L. H., J. CULOTTI, J. R. PRINGLE, and B. J. REID, 1974  Genetic control of the cell division cycle in yeast. Science 183:46-51[Free Full Text].

JOHNSTON, G. C., J. R. PRINGLE, and L. H. HARTWELL, 1977  Coordination of growth with cell division in the yeast Saccharomyces cerevisiae. Exp. Cell Res. 105:79-98[Medline].

KO, H. A. and S. A. MOORE, 1990  Kinetic characterization of a prestart cell division control step in yeast. Implications for the mechanism of alpha-factor-induced division arrest. J. Biol. Chem. 265:21652-21663[Abstract/Free Full Text].

KOCH, C. and K. NASMYTH, 1994  Cell cycle regulated transcription in yeast. Curr. Opin. Cell Biol. 6:451-459[Medline].

LOEB, J. D., T. A. KERENTSEVA, T. PAN, M. SEPULVEDA-BECERRA, and H. LIU, 1999  Saccharomyces cerevisiae G1 cyclins are differentially involved in invasive and pseudohyphal growth independent of the filamentation mitogen-activated protein kinase pathway. Genetics 153:1535-1546[Abstract/Free Full Text].

MITCHISON, J. M. and P. NURSE, 1985  Growth in cell length in the fission yeast Schizosaccharomyces pombe. J. Cell Sci. 75:357-376[Abstract].

MOORE, S. A., 1988  Kinetic evidence for a critical rate of protein synthesis in the Saccharomyces cerevisiae yeast cell cycle. J. Biol. Chem. 263:9674-9681. (Erratum. J. Biol. Chem. 263: 18582).[Abstract/Free Full Text].

MOSCH, H. U. and G. R. FINK, 1997  Dissection of filamentous growth by transposon mutagenesis in Saccharomyces cerevisiae. Genetics 145:671-684[Abstract].

NASH, R., G. TOKIWA, S. ANAND, K. ERICKSON, and A. B. FUTCHER, 1988  The WHI1+ gene of Saccharomyces cerevisiae tethers cell division to cell size and is a cyclin homolog. EMBO J. 7:4335-4346[Medline].

NEUFELD, T. P. and B. A. EDGAR, 1998  Connections between growth and the cell cycle. Curr. Opin. Cell Biol. 10:784-790[Medline].

OEHLEN, L. J. and F. R. CROSS, 1994  G1 cyclins CLN1 and CLN2 repress the mating factor response pathway at Start in the yeast cell cycle. Genes Dev. 8:1058-1070[Abstract/Free Full Text].

PARDEE, A. B., 1989  G1 events and regulation of cell proliferation. Science 246:603-608[Abstract/Free Full Text].

PETERSON, E. A. and W. H. EVANS, 1967  Separation of bone marrow cells by sedimentation at unit gravity. Nature 214:824-825[Medline].

POLYMENIS, M. and E. V. SCHMIDT, 1997  Coupling of cell division to cell growth by translational control of the G1 cyclin CLN3 in yeast. Genes Dev. 11:2522-2531[Abstract/Free Full Text].

RICHARDSON, H. E., C. WITTENBERG, F. CROSS, and S. I. REED, 1989  An essential G1 function for cyclin-like proteins in yeast. Cell 59:1127-1133[Medline].

SCHNEIDER, B. L., E. E. PATTON, S. LANKER, M. D. MENDENHALL, and C. WITTENBERG et al., 1998  Yeast G1 cyclins are unstable in G1 phase. Nature 395:86-89[Medline].

SHERMAN, F., 1991 Getting started with yeast, pp. 3–21 in Guide to Yeast Genetics and Molecular Biology, edited by C. GUTHRIE and G. R. FINK. Academic Press, New York.

SPELLMAN, P. T., G. SHERLOCK, M. Q. ZHANG, V. R. IYER, and K. ANDERS et al., 1998  Comprehensive identification of cell cycle-regulated genes of the yeast Saccharomyces cerevisiae by microarray hybridization. Mol. Biol. Cell 9:3273-3297[Abstract/Free Full Text].

STUART, D. and C. WITTENBERG, 1995  CLN3, not positive feedback, determines the timing of CLN2 transcription in cycling cells. Genes Dev. 9:2780-2794[Abstract/Free Full Text].

SUDBERY, P. E., A. R. GOODEY, and B. L. CARTER, 1980  Genes which control cell proliferation in the yeast Saccharomyces cerevisiae. Nature 288:401-404[Medline].

SUGIMOTO, K., K. MATSUMOTO, R. D. KORNBERG, S. I. REED, and C. WITTENBERG, 1995  Dosage suppressors of the dominant G1 cyclin mutant CLN3–2: identification of a yeast gene encoding a putative RNA/ssDNA binding protein. Mol. Gen. Genet. 248:712-718[Medline].

THOMAS, B. J. and R. ROTHSTEIN, 1989  Elevated recombination rates in transcriptionally active DNA. Cell 56:619-630[Medline].

TOKIWA, G., M. TYERS, T. VOLPE, and B. FUTCHER, 1994  Inhibition of G1 cyclin activity by the Ras/cAMP pathway in yeast. Nature 371:342-345. [see comments][Medline].

TYERS, M., G. TOKIWA, and B. FUTCHER, 1993  Comparison of the Saccharomyces cerevisiae G1 cyclins: Cln3 may be an upstream activator of Cln1, Cln2 and other cyclins. EMBO J. 12:1955-1968[Medline].

ZETTERBERG, A., O. LARSSON, and K. G. WIMAN, 1995  What is the restriction point? Curr. Opin. Cell Biol. 7:835-842[Medline].

This article has been cited by other articles:

				N. Colomina, F. Ferrezuelo, H. Wang, M. Aldea, and E. Gari Whi3, a Developmental Regulator of Budding Yeast, Binds a Large Set of mRNAs Functionally Related to the Endoplasmic Reticulum J. Biol. Chem., October 17, 2008; 283(42): 28670 - 28679. [Abstract] [Full Text] [PDF]

				K. Flick and C. Wittenberg Multiple Pathways for Suppression of Mutants Affecting G1-Specific Transcription in Saccharomyces cerevisiae Genetics, January 1, 2005; 169(1): 37 - 49. [Abstract] [Full Text] [PDF]

				E. Queralt and J. C. Igual Functional Distinction Between Cln1p and Cln2p Cyclins in the Control of the Saccharomyces cerevisiae Mitotic Cycle Genetics, September 1, 2004; 168(1): 129 - 140. [Abstract] [Full Text] [PDF]

				Q.-W. Jin and D. McCollum Scw1p Antagonizes the Septation Initiation Network To Regulate Septum Formation and Cell Separation in the Fission Yeast Schizosaccharomyces pombe Eukaryot. Cell, June 1, 2003; 2(3): 510 - 520. [Abstract] [Full Text] [PDF]

				S. M. Honigberg and K. Purnapatre Signal pathway integration in the switch from the mitotic cell cycle to meiosis in yeast J. Cell Sci., June 1, 2003; 116(11): 2137 - 2147. [Abstract] [Full Text] [PDF]

				J. Karagiannis, R. Oulton, and P. G. Young The Scw1 RNA-Binding Domain Protein Regulates Septation and Cell-Wall Structure in Fission Yeast Genetics, September 1, 2002; 162(1): 45 - 58. [Abstract] [Full Text] [PDF]

				P. Jorgensen, J. L. Nishikawa, B.-J. Breitkreutz, and M. Tyers Systematic Identification of Pathways That Couple Cell Growth and Division in Yeast Science, July 19, 2002; 297(5580): 395 - 400. [Abstract] [Full Text] [PDF]

				E. Gari, T. Volpe, H. Wang, C. Gallego, B. Futcher, and M. Aldea Whi3 binds the mRNA of the G1 cyclin CLN3 to modulate cell fate in budding yeast Genes & Dev., November 1, 2001; 15(21): 2803 - 2808. [Abstract] [Full Text] [PDF]