Commitment to Meiosis in Saccharomyces …restored at early stages of meiosis (by transfer of cells to growth medium), the cells resume the mitotic cycle without completing meiosis.

Copyright 0 1992 by the Genetics Society of America

Commitment to Meiosis in Saccharomyces cereuisiae: Involvement of the SP014 Gene

Saul M. Honigberg,* Clara Conicella**+ and Rochelle Easton Espositio* *Department of Molecular Genetics and Cell Biology, University of Chicago, Chicago, Illinois 60637, and Istituto di Agronomia

dell’Universita di Napoli, Naples, Italy Manuscript received Februay 1, 199 1

Accepted for publication December 6, 199 1

ABSTRACT This paper describes the identification, cloning and phenotypic analysis of SP014, a new gene

required for meiosis and spore formation. Studies of strains carrying a temperature-sensitive mutation or a disruption/duplication allele indicate that sf014 mutants have the unusual property of being able to return to mitotic division, even from the late stages of meiotic development. Early meiotic events, such as DNA replication and intragenic and intergenic recombination, occur normally. In contrast, later meiotic processes are defective in sf014 mutants: the meiosis I division appears to be executed at slightly depressed levels, the meiosis I1 division is reduced more severely, and no spores are formed. Epistasis tests using mutants defective in recombination or reductional division support these findings. Based on these data, we suggest that the SP014 gene product is involved in the coordinate induction of late meiotic events and that this induction is responsible for the phenomenon of commitment.

M EIOSIS in yeast can be viewed as a process of cellular development in which a programed

series of events results in a change of cell type. The first major landmark event in meiosis is DNA repli- cation, followed by chromosome pairing and genetic recombination. Homologous chromosomes then seg- regate to opposite poles of the nucleus (the meiosis I division), and rapidly thereafter the sister chromatids segregate from each other (the meiosis I1 division). Finally, each haploid genome is packaged into a spore within the mother cell, called an ascus. The two major issues addressed below are the irreversible commit- ment of the cell to this developmental process, and the coordination of the various events with one an- other.

Diploid yeast cells initiate meiosis in response to deprivation for glucose and nitrogen; however, they become committed to the meiotic process consider- ably after this initiation. Commitment to meiosis is defined by the response of the sporulating cell to the restitution of glucose and nitrogen. If nutrients are restored at early stages of meiosis (by transfer of cells to growth medium), the cells resume the mitotic cycle without completing meiosis. In contrast, if transferred at later stages, they complete both meiotic divisions and form haploid spores before returning to mitotic growth. The point at which cells lose the capacity to return to growth and become irreversibly committed to execute the two meiotic divisions and form spores, even in the presence of glucose and nitrogen, has been designated “commitment to meiosis” (KIRSOP 1954; GANESAN, HOLTER and ROBERTS 1958). Analysis of the specific meiotic events that have occurred in cells

Genetics 130 703-716 (April, 1992)

that retain the capacity to return to growth indicates that the commitment transition occurs after the com- pletion of DNA synthesis, pairing and recombination but prior to the segregation of homologous chromo- somes in meiosis I (SHERMAN and ROMAN 1963; SIMCHEN, P I ~ O N and SALTS 1972; ESPOSITO and Es-

The molecular basis of meiotic initiation is currently much better understood than is that of meiotic com- mitment. The process of initiation is regulated by both cell type and nutritional conditions. The cell type control generally restricts meiosis to diploids and is mediated by the a1 and a2 products of the M A T locus. These proteins participate in a regulatory cascade that results in the derepression of I M E l , an inducer of meiosis (MITCHELL and HERSKOWITZ 1986; KASSIR, GRANOT and SIMCHEN 1988). The I M E l gene is in- dependently repressed by glucose and nitrogen (SMITH and MITCHELL 1989), which are sensed, in part, through the RAS2-dependent CAMP pathway (TODA et al. 1985).

While specific genes required for commitment have not yet been identified, two classes of mechanisms have been considered to account for the “immunity” of meiotic cells to glucose and nitrogen after commit- ment. In the first, commitment results from repression of mitotic functions, for example, by inactivation of permeases that transport glucose and nitrogen or pro- teins that transduce the nutritional growth signal (MITCHELL 1988). In the second, commitment is due to the induction of a meiotic event that blocks the return to mitotic division. For example, it has been suggested that spindle pole body separation during

POSIT0 1974).

704 S. M. Honigberg, C. Conicella and R. E. Esposito

meiosis I may serve this role (HORESH, SIMCHEN and FRIEDMANN 1979; ESPOSITO and KLAPHOLZ 1981; MURRAY and SZOSTAK 1985).

As described above, previous reports of meiotic commitment in yeast indicate that the meiosis I (MI) and meiosis I1 (MII) divisions and spore formation become committed at the same point. This implies that the late landmarks are coordinated, a view that is supported by the close timing of these events (MOENS and RAPPORT 197 1; PADMORE, CAO and KLECKNER 199 1). Paradoxically, however, genetic analysis of a variety of mutants indicates that these same events can occur independently of one another (reviewed in ESPOSITO and KLAPHOLZ 198 1). Thus, in the absence of either recombination (MALONE and ESPOSITO 198 1 ; KLAPHOLZ, WADDELL and ESPOSITO 1985; ENCEBRECHT and ROEDER 1989; MENEES and ROEDER 1989), meiosis I (KLAPHOLZ and ESPOSITO 1980), or meiosis I1 (BYERS 1981; THOMAS and BOT- STEIN 1986), cells can continue to proceed through the remaining stages of meiosis and package spores. The independence of these meiotic landmarks stands in contrast to the dependency relationship seen in mitotically growing cells (PRINGLE and HARTWELL 1981), even though there are many gene functions shared in common between the two processes (SIMCHEN 19'74; SCHILD and BYERS 1978; SHILO, SIMCHEN and SHILO 1978; SHUSTER and BYERS 1989). Recently, it has been shown that the mitotic depend- encies are mediated by "checkpoint" functions that prevent the execution of a given event until a prior one has been completed (WEINERT and HARTWELL 1989; HOYT, TOTIS and ROBERTS 1991; LI and MUR- RAY 1991). Since MI, MI1 and spore formation can occur independently of one another, their coordina- tion may require a different type of mechanism to control the temporal order of events.

In this paper we report that the product of the S P 0 1 4 gene is required for efficient completion of meiotic segregation and spore formation. The S P 0 1 4 gene is also required for the commitment process, since spo14 mutants are able to return to mitotic growth from late stages in meiotic development. T h e implication of this phenotype to possible models for commitment and to the coordination of late meiotic functions is discussed below.

MATERIALS AND METHODS

Strains: The primary strains used in this study are listed with their genotypes in Table 1 . The behavior of the tem- perature-sensitive spol4-1 mutant (H16) was compared at permissive (25") and restrictive (34") conditions; in addi- tion, an isogenic wild-type control was constructed by trans- formation with a CEN plasmid containing the cloned SP014 gene (see below). A disruption/duplication allele of sPol4 was also made in a related wild-type strain (see H67 and H68; Table 1) to provide additional markers for genetic studies. Mating type and heteroallele tests utilized additional

standard strains from our laboratory collection. The four testers used to determine ploidy (W303-1A, W303-1B, SH219 and SH220) are isogenic to one another. Plasmids were propagated in bacterial strains LE392 and DH5a (SAM- BROOK, FRITSCH and MANIATIS 1989).

Media: YPA growth medium contains 10 g/liter potas- sium acetate, 20 g/liter Bacto-peptone and 10 g/liter Bacto yeast extract. SPII-22 sporulation medium contains 20 g/ liter potassium acetate (pH 7.0) supplemented with 75 mg/ liter each of adenine, histidine, leucine, lysine, methionine, tryptophan, tyrosine and uracil. AUCC medium (to select for the cloned SP014 gene) is synthetic complete medium lacking arginine and uracil and supplemented with 10 mg/ liter cycloheximide and 60 mg/liter canavanine. All other media have been described previously (KLAPHOLZ, WAD DELL and E~POSITO 1985). LB medium (SAMBROOK, FRITSCH and MANIATIS 1989) containing 100 rg/ml carbenicillin was prepared for bacterial cultures.

Cloning the SP014 gene: The SP014 gene was cloned by complementation of the spol4-1 sporulation defect. A spol4-1 diploid, H16, heterozygous for two recessive drug resistance markers, cyh2 and can1 (see Table 1 for complete genotype), was transformed with a yeast partial Sau3A ge- nomic library cloned into the BamHI site of YCp50 (library constructed by SCOTT HOUTTEMAN, University of Chicago). Freshly grown log-phase yeast cells were transformed by the method of Escherichia coli electroporation (DOWER, MILLER and RACSDALE 1988), resuspended in YPDA after electro- poration, washed in H20, and plated on synthetic media lacking uracil to select for transformants. Approximately 10,000 Ura+ transformants, derived from 5 pg of DNA, were replica plated to sporulation medium and incubated at 34". After 5 days, the sporulation plates were treated with ether to select against vegetative cells (DAWFS and HARDY 1974), and then replica plated to AUCC medium containing canavanine and cycloheximide. Unsporulated diploid cells, which are heterozygous for the recessive drug resistance markers, are sensitive to the drugs. In contrast, one-fourth of the haploid spore products are expected to be resistant to both drugs. From approximately 10,000 transformants screened in this manner, one transformant was isolated that grew on the selective medium. This transformant was ca- pable of wild-type levels of sporulation at either tempera- ture. Spontaneous Ura- segregants derived from this trans- formant regained the sporulation defect due to loss of the plasmid (data not shown).

Plasmids: The vector plasmids used in this study were YCp50 (BOTSTEIN et al. 1979) and RS306 (SIKORSKI and HIETER 1989). The original SPOI4-complementing plasmid, pS12, contains a 6.8-kb insert in YCp50 (Figure 1). pS14 was constructed by partial digestion of pS12 with SalI fol- lowed by ligation and is deleted for a 1-kb SalI fragment in the SP014 region. pS20 was constructed by cloning a 5.8- kb EcoRI fragment from the SP014 region of pS12 into the EcoRI site of YCp50. The disruption/duplication plasmid, pS25, was constructed by cloning the l-kb Sal1 fragment from pS12 into the SalI site of RS306.

Growth and sporulation: Cells (1 X lo6) from a freshly grown colony were inoculated into a 50-ml culture of YPA and grown for 20 hr at 30" to a density of 2-3 X lo' cells/ ml. The cell suspension was sonicated for 20 sec at 30% power with an Artek sonicator, the cells were pelleted, washed twice in distilled H20, and then resuspended in 25 ml of SPII-22. Sporulation cultures were incubated at 25", 30" or 34" with constant shaking at 200 rpm in a New Brunswick Aquatherm shaker.

Assays for cell viability, DNA synthesis, genetic recom- bination and meiotic divisions: At intervals during sporu-

Commitment to Meiosis

TABLE 1

Yeast strains

705

Strain Genotype Source

REE206

H16

H 2 3

H67

H 6 8

H 6 1

H36

K292

H46

MATa HO ade2 ade5 canl his7 leul 9 2 trp5 tyrl ura3 MATa HO ade2 ade5 canl his7 leul lys2 trp5 tyrl ura3 """- MATa ade.2-1 ADE5 canl cyh2 his7-2 lys2-1 leul-2 metl3-c trp5-c tyrl-1 ura3 spol4-1 M A T a ade2-1 ade5 C A N l CYHZ his7-1 lys2-1 leul-c MET13 trp5-d tyrl-2 ura3 spol4-1 ""

MATa ade2-1 ADES canl cyh2 his7-1 lys2-1 leul-12 metl3-c trp5-c tyrl-1 ura3 SP014 M A T a ade2-1 ade5 C A N l CYHZ his7-1 lys2-1 leul-c MET13 trp5-d tyrl-2 ura3 spol4-1 ""

MATa ADEZ ade5 can CYH2 HIS1 his7 HOM3 leu2-3 lysl LYSZ met14 pet8 T R P l u r a 3 M A T a ade2-1 ADES canl cyh2 his1 H I S 7 h o d - l o leu2-1 LYSl lys2 MET14 PET8 trpl ura3 """"-

MATa adeZ-IADE5 can CYHZHISI his7 HOM3 leu2-3lyslLYS2 met14 p e t B T R P l u r a 3 MATa ADE2 ade5 canl cyh2 his1 HlS7hom3-lOleu2-lLYSl lys2 MET14PET8 t rp l ura3 """"- " "

spol4:URA3:spol4 spol4:URA3:spol4

MATaadeZ-1 ADE5CANl cvh2 his4A-25 his7 LEU1 leu2-271~2-2 met13-ctrf15-ctvrl-2 ura3 MATaadeZ-1 ade5 canl CYHZ HIS4 HIS7 leul LEU2 lys2-1 MET13 trp-d tyrl-1 ura3-1

spoll-1 spoll-1

MATaadeZ-lADE5 canl cyh2his7-21eul-121ys2-I metl3-ctrp5-dtyrl-lura3spol1-1spol4-1 MATaade2-1 ade5 CAN1 CYH2his7-1 leul-c lys2-1 MET13 trp5-c tyr1-2ura3spoI 1-1 spol4-1 "" "

MATa ade.2-1 ade5 canl CYH2 his7 leul-12 lys2 metl3-d trp5-d tyrl-1 ura3 spol3-1 MATa ade2-1 ADE5 canl cyh2 his7 leul-c lys2 metl3-c trp5-c lyrl-2 ura3 spol3-1 """- MATaade2-1 ADE5 canl a h 2 his7-1 lvs2-1 leul-c metl3-ctr65-dtvrl-lura3sfIo13-1sbo14-1 MATaade2-1 ade5 CANl CYH2his7-2lys2-I leul-lZMETl3trp5-dtyrl-2ura3spol3-1spol4-l

This laboratory

This study

This study

This study

This study

This study

This study

S. KLAPHOLZ

This study

W303-1A MATa ade2 canl his3-11,15 leu2-3,112 trpl-1 ura3-1 R. ROTHSTEIN

WSQ3-1B MATa ade2 canl his3-11,15 leu2-3,112 trpl-1 ura3-1 R. ROTHSTEIN

SH219 --- MATa ade2 canl his3-11,15 leu2-3,112 trpl-l ura3-1 MATa ade2 canl his3-11.15 leu2-3,112 trpl-1 ura3-1

- This study

SH220 --- MATa ade2 canl his3-11,15 leu2-3,112 trpl-l ura3-1 MATa ade2 canl his3-11,15 leu2-3,112 trpl-1 ura3-1

- This study

lation, samples were removed, briefly sonicated, then washed and resuspended in distilled H20. Samples were removed for the DABA fluorimetric assay of DNA replica- tion (according to KLAPHOLZ, WADDELL and ESPOSITO 1985), and for plating to determine viability and recombi- nation frequency. Cells were maintained at the temperature of the sporulation culture during processing and plated on medium prewarmed to the same temperature. Viability was assayed on synthetic complete medium, and meiotic intra- genic recombination at his7 and trp5 was assayed on syn- thetic complete medium lacking either histidine or trypto- phan. His' colonies were patched to YPDA medium for analysis of intergenic recombination and the occurrence of the meiotic divisions (see below); selection for His' proto- trophs ensures that each colony is derived from a single cell (ESPOSITO and ESPOSITO 1974).

Cells that have not yet executed the meiosis I division can be distinguished from those that have undergone either the first or both meiotic divisions, since only the former class retains both alleles of a pair of close CEN-linked heteroal- leles, in this case leul. Heteroallelism at leul was determined by replica plating the His+ patches to sporulation medium at the permissive temperature, and subsequent replica plat- ing of the sporulation plates to Leu- medium and scoring the production of Leu+ meiotic recombinants. Furthermore, cells that have undergone only the first meiotic division can be distinguished from those that have undergone both di- visions since the former class are diploid maters (in the absence of CEN3-MAT recombination) while the latter are haploids. Diploid maters (a/a or a/a) vs. haploid maters (a or a) can be diagnosed by crosses to a/a and ala as well as a and a testers (KLAPHOLZ, WADDELL and E~POSITO 1985).

I I I I I I , I 1 .o 2.0 3.0 4.0 5.0 6.0

FIGURE 1 .-Restriction map of the region containing the SP014 gene. The restriction sites for BamHl (B), Bglll (Bg), EcoRl (E), Hindlll ( H ) and Sal1 (S) are shown above the line. The thick line represents the smallest region we have subcloned that retains complementing activity. The length of the fragment in kilobases is shown beneath the line. The thin line segment shown below the map reflects the region of the gene used to construct the disruption allele.

706 S. M. Honigberg, C. Conicella and R. E. Esposito

The ploidy of each isolate is inferred from the spore survival of such crosses, since triploid meiosis results in high aneu- ploidy and poor spore survival in contrast to diploid and tetraploid meiosis (ROMAN, PHILLIPS and SANDS 1955; MOR- TIMER and HAWTHORNE 1969). To rapidly assess spore viabiljty, each cross was made heterozygous for the recessive drug markers cyh2 and/or c a n l . Sporulation plates of these crosses were replicated to medium containing canavanine or cycloheximide to select for viable spores. We confirmed the results of this assay by choosing eight isolates categorized as diploids and eight categorized as haploids, crossing them to the testers, sporulating the zygotes and dissecting ten tetrads of each. The isolates judged to be haploids yielded much higher spore viability when crossed to other haploid strains (W303-1A and W303-1B), forming diploids, than when crossed to diploid strains isogenic to the corresponding haploid (SH219 and SH220), forming triploids (97 and 25%, respectively). Conversely, the isolates judged to be diploids gave the reverse result (5% viability when crossed to a haploid us. 94% when crossed to a diploid). The occurrence of the meiosis I division in the spol4:URA3:spol4 mutant was monitored utilizing the heterozygous close centromere- linked markers, t r p l , met14 and pet8. Expression of the recessive alleles for these markers indicates the occurrence of a reductional division (there is virtually no exchange between these markers and their centromeres).

Intergenic recombination was measured in the Eeul-trp5 (15 cM) and CEN3-MAT (25 cM) intervals in the various populations of cells described above, i e . , (1) those that return to growth upon plating without completing either of the meiotic divisions, (2) those that undergo only meiosis I , and (3) those that complete both divisions. In the cells that have not undergone meiosis I or I1 segregation (colonies heteroallelic at l eul ) , l eul - t rp5 exchange is determined from the frequency of colonies that are homoallelic at trp5 (the frequency of gene conversion at trp5 is two orders of mag- nitude lower than the frequency of recombination between these two markers, see Table 5). Similarly, CEN3-MAT exchange is based on the frequency of MAT homozygotes (MATaIMATa and MATaIMATa maters). Among cells that have executed just the meiosis I division ( leul -12/ leuI-12 or l ed -c / l eu l - c colonies), leul-trp5 recombinants are those in which a change of coupling has occurred between the leul and trp5 heteroalleles. Recombinants in the CEN-MAT in- terval are identified as leul homoallelic MATa/MATa non- maters. Finally, in cells that executed both divisions ( i e . , leul-c or leuI-12), the frequency of leul-trp5 exchange is determined in the standard way by analyzing the haploid products for a change of coupling between the parental alleles; CEN3-MAT exchange cannot be determined among these random haploid products (i.e., in the absence of te- trads). Allelism tests were performed by crosses to tester strains as described previously (KLAPHOLZ, WADDELL and ESPOSITO 1985). Fluorescence photomicroscopy: Nuclear segregation was

monitored using the fluorescent DNA stain 4',6'-diamidino- 2 phenylindole (DAPI) (Boehringer Mannheim). Samples (1 00 pl) were fixed with 900 p1 freshly diluted 4% paraform- aldehyde for at least 2 hr at room temperature and washed twice in 0.1 M NaHP04 (pH 8.0). Cells were stained in 1 mg/ml DAPI for 1-2 hr, washed twice in 0.1 M NaHP04 (pH 8.0) and resuspended on a slide in a drop of mounting medium (0.1 M NaHCOs, 90% glycerol, 0.1 mg/ml p-phen- ylene diamine). Immunofluorescent staining was performed as described (ROSE, WINSTON and HIETER 1989) with some modifications. The samples were fixed and then incubated in 30 mg/ml Zymolyase l O O T (ICN) according to DRESSER and GIROUX (1 988) until the majority of cells were sphero-

TABLE 2

Sporulation of spol4 strains

Percent Genotype Strain Temperature sporulation"

SP014 25" 63 SPO 14 59

spol4-1 25" 65

- REE206 34"

HI6 SpOI4-I 34" CO. 1

~ p 0 1 4 - 1 H23 25' 65

SPO 14 34" 63

SP014 25' 60 SP014 34" 53 - H67

spol4:uRA3:spol4 spol4:uRA3:spol4 34 CO. 1

H68 25" eo. 1

a The maximum level of sporulation is given after 72 hr in sporulation medium; approximately 80% of total asci contain three or fourspores.

plasts. Cells were applied to polylysine-coated multiwell slides and incubated with rat anti-yeast a-tubulin mono- clonal antibody YOL1/34 (generous gift ofJ. V. KILMARTIN; KILMARTIN, WRIGHT and MILSTEIN 1982) overnight at room temperature. Samples were washed with 0.1 M Na- HPO4 (pH 8.0) and then incubated with goat anti-rat anti- body coupled to fluorescein isothiocyanate (FITC) from Boehringer Mannheim, washed with 0.1 M NaHP04 (pH 8.0), and stained with DAPI as above. Cells were covered with the mounting medium and photographed with fluores- cence optics on a Leitz Orthoplan microscope.

RESULTS

Isolation of the sporulationdefective spol4-1 mu- tant: T h e spol4-1 mutant was identified by a strategy similar to that previously used to recover recessive spo mutants in our laboratory (ESPOSITO and ESPOSITO 1969). Briefly, a random spore population of a hom- othallic strain (REE206) was mutagenized to 30% survival with EMS. Following spore germination, di- ploidization (due to MAT switching, and mating of MATa and MATa haploids within each ascosporal colony), survivors were sporulated at various temper- atures and screened microscopically for the produc- tion of asci. The original spol4-1 isolate was out- crossed to related marked strains and heterothallic derivatives were obtained. The sporulation frequen- cies of the original parental strain and derivatives containing the mutation in heterozygous and homo- zygous conditions are shown in Table 2. T h e sP014-I mutant sporulates to levels similar to the parental strain (REE206) at 25" but fails to sporulate at 34". Mitotic growth rates at either temperature were in- distinguishable from a closely related wild-type strain (data not shown). Diploid strains heterozygous for the spol4-1 mutation sporulate to normal levels at both temperatures, indicating that spol4-1 is recessive (Table 2). The gene segregates in Mendelian fashion and maps to the right arm of chromosome X I (M.

Commitment to Meiosis 707

TOWNSEND, B. DIDOMENICO, S. KLAPHOLZ and R. E. ESPOSITO, cited in MORTIMER et al. 1990).

Cloning and disruption of the SP014 gene: The SP014 gene was isolated from a yeast CEN library by complementation of the sporulation defect (see MA- TERIALS AND METHODS). The original complementing plasmid, pS12, contained a 6.8-kb insert (Figure 1). A subclone containing a 5.3-kb EcoRI fragment (pS20) also complemented the spol4-1 strain but smaller sub- clones including a 1-kb deletion of a Sal1 fragment from pS12 failed to complement, suggesting that se- quences within this fragment might be internal to the SP014 gene. This 1-kb fragment was cloned into the Sal1 site of RS306, which contains the URA3 gene but no yeast origin of replication. The resulting plasmid, pS25, was cleaved with BglIl, which cuts twice in the insert, and integrated into the genome (ROTHSTEIN 1991). The resulting strain was crossed to a spol4-1 strain to verify by noncomplementation that the cloned gene was indeed SP014 . This diploid sporu- lates at the permissive temperature but not the restric- tive temperature, indicating that the integration in- activated the SP014 gene, and that the 1-kb fragment is internal to the region required for complementation (SHORTLE, HABER and BOTSTEIN 1982). The allele generated is a disruption/duplication with one copy of the gene containing a 5' deletion, another copy containing a 3' deletion, and the complete RS306 vector containing URA3 integrated between these two copies of spol4 . For convenience, we will refer to this allele as spol4:URM:spol4. A yeast strain homozygous for this allele fails to sporulate, while the otherwise isogenic SP014 strain sporulates to high levels (Table 2). As expected, Ura- revertants of the disrupted strain also reverted to sporulation competence, pre- sumably as a result of recombination between the duplicated spol4 genes to regenerate an intact SP014 gene. The cross between the spol4:URA3:spol4 mu- tant and the spol4-1 haploid was sporulated at the permissive temperature, and the spore products were dissected and crossed to spo14:URA3:spo14 tester strains. Among twelve tetrads dissected, the $014-1 gene always segregated from the disruption/duplica- tion allele, confirming that the cloned gene is the wild- type SP014 gene. The spol4:URA3:spol4 disruption had no effect on mitotic growth rates at 25", 30", 34" or 37" (data not shown).

Execution of the SP014 function occurs at the time of the meiotic divisions: T o ascertain the time in meiosis at which the SP014 function is required, we performed two types of temperature-shift experi- ments with the ts s p o l 4 - l / s ~ o l 4 - l mutant (ESPOSITO et al. 1970; HARTWELL 1974). In one experiment, aliquots of a sporulation culture of a spol4-1/sp014-1 diploid (H16) at the restrictive temperature were shifted at intervals to the permissive temperature and

34" -25" tl

I I I I I I 0 8 16 24

Time of Temp. Shift (hrs.)

FIGURE 2.-Temperature shift experiment on spol4-I cells in meiosis. H16 cells were placed in sporulation medium and then transferred at the times indicated from 34" to 25" (open circles) or from 25" to 34" (closed circles). Ascus production was measured by phase-contrast microscopy after a total of 72 hours in sporulation medium.

allowed to complete sporulation. In the second exper- iment, aliquots were shifted from the permissive to the restrictive temperature. After 72 hr, each sample was assayed for the presence of asci. The first temper- ature-shift study indicates the time at which SP014 function is initially required, while the second indi- cates the time at which SP014 is no longer needed. The results demonstrate that SP014 is required only during a narrow window in meiosis, from approxi- mately 6-12 hr (Figure 2), during the time that both meiotic divisions occur in this strain (see below). Fur- thermore, the rapid drop in ability to sporulate when cells are shifted from restrictive to permissive temper- ature suggests that the meiotic defect is irreversible. SP014 is required for the efficient induction of

late meiotic events: T o assess the ability of $014-1 mutants to execute meiotic chromosome segregation, H16 was incubated in sporulation medium at either 25" or 34", and samples were removed at various times and examined by DAPI fluorescence and phase contrast microscopy (Figure 3). Binucleate (indicated by arrows in Figure 3, C and D) and tetranucleate cells (Figure 3, E and F) were observed at both tem- peratures, demonstrating that both divisions occur in the mutant. By 24 hr at the permissive temperature, about 60% of the cells have formed asci (Figure 3, G and I). At the restrictive temperature, the nuclei of both binucleate and tetranucleate cells begin to frag- ment by 10 hr, and by 24 hr most nuclei appear to have completely disintegrated (Figure 3, H and J).

Both binucleate and tetranucleate cells were also seen in the disruption/duplication homozygote, H68. We measured the occurrence of the meiotic divisions over time in this latter strain and in the isogenic SPOl4 strain, H67 (Figure 4). Similar experiments with the conditional allele at the two temperatures, in the presence or absence of a plasmid bearing the SP014

708 S. M. Honigberg, C. Conicella and R. E. Esposito

1

FIGURE 3.--l,ight micrographs of. s p o l 4 - 1 strains undergoing meiosis stained with the DNA specific dye DAPI and visualized by fluorescence (A-H) or by phase contrast ( I and J) microscopy. !+mples from a 25” sporulation culture (left) or 34” sporulation culture (right) of strain HI6 were examined at various times in the course of meiosis: 5 hr (A and B). 7 hr (C and D), 10 hr (E and F) and 24 hr (G-J). I M T points out “immature” tetranucleate cells and M T a “mature” tetranucleate cell; T R indicates a trinucleate cell.

gene, are not shown, but gave identical results to the comparison between H67 and H68. Meiosis I initiates at approximately the same time in the disruption/ duplication strain as the wild-type isogenic control strain, but occurs in a somewhat smaller proportion of cells. One possible explanation for the 1.4-fold decrease in the fraction of cells executing meiosis I in the disruption/duplication strain relative to the con- trol strain is that the spo l4 mutation may have a small but real effect on the fraction of cells that undergo meiosis I. Alternatively, the nuclear degeneration de- scribed above may cause a (reproducible) underesti- mate in the number of binucleate cells.

0 10 20 30

Hours of Sporulation

FIGURE 4.-Meiotic divisions and sporulation. (A) SPOl4/SPO14 (strain H67). (B) spol4:URA3:spol4/spo14:URA3:spo14 cells (strain H68). The percentage of cells completing meiosis I (open squares) or meiosis I1 (filled circles) were monitored by DAPI fluorescence microscopy, and ascus formation (open triangles) was monitored by phase contrast microscopy. The percentage of cells having com- pleted meiosis I was calculated by dividing the sum of bi-, tri- and tetranucleate cells, and asci by the total number of cells. The percentage of cells having completed meiosis I1 was similarly cal- culated by dividing the sum of tri- and tetranucleate cells and asci by the total number of cells. Each time point is the average of three experiments; at least 300 cells were counted for each determination.

In contrast, the fraction of cells that have under- gone two divisions is approximately 10-fold lower in the spoI4:URA3:spo14 homozygote than in the wild- type isogenic control. Several lines of evidence sup- port the view that meiosis I1 in fact occurs less effi- ciently (in a smaller fraction of cells) in spo l4 strains. First, binucleate cells accumulate to a higher level in the spol4:URA3:spol4 homozygote than in the control (32 vs. 15%), as expected if meiosis I1 occurs less efficiently than meiosis I. Second, trinucleate cells, which presumably arise from a completed meiosis I1 division at only one pole of a binucleate cell (see Figure 3F), are much more common in the spol4-1 mutant at the restrictive temperature (8%) than at the per- missive temperature (0.5%). Third, the morphology of tetranucleate cells is different in spol4-1 strains at the restrictive compared to the permissive condition. In the latter case, most tetranucleate cells contain four distinct nuclei (an example is shown in Figure 3E), while at the restrictive temperature the tetranucleate cells usually contain DAPI stained material forming a cross or parallel lines between the nuclear bodies (an example is shown in Figure 3F). Since this internuclear

Commitment to Meiosis 709

FIGCHI.: 5.-Spindlc formation i n s p o l 4 - 1 straims A s p o l 4 - l strain (H16) w a s exposed to sporulation medium at 34" lor different lengths of time before being fixed and processed for immunofluo- rescence microscopy as described i n MATERIALS AND METHODS.

Arrows in each pair of photographs are directed at the same cell visualized by DAPI fluorescence (left) or the rat anti-tubulin anti- body (right). (A and B) The spindle pole body has just begun to separate in a cell that is still mononucleate. (C and D) A binucleate cell with the meiosis I spindle extending across the nuclei. (E and F) Tetranucleate cells containing the characteristic crossed spindle of the second meiotic division.

staining colocalizes with the spindle (see below) it apparently reflects an intermediate in the second meiotic division with some chromatin trailing along the meiosis I1 spindles. These intermediates are seen only transiently at the permissive temperature. There- fore, although both divisions initiate in spol4 strains, the second division is less efficient and is completed in only a small percentage of cells.

The meiotic spindle morphology of the same spol4- I diploid was examined using DAPI staining of the nuclei (Figure 5, A, C and E) and indirect immunoflu- orescent staining directed at a-tubulin (Figure 5, B, D and F). Doubly stained cells represented 65-70% of the populations at both permissive and restrictive temperatures. Three hundred doubly stained cells

FIGURE 6.-Return to growth of a S f 0 1 4 strain A wild-type S f 0 1 4 strain (H67) isogenic to the spol4:URA3:spol4 disruption strain was exposed to sporulation medium for 17 hr (A and B) before the cells were transferred to YPDA medium for either 4 hr (C and D) or 7 hr (E and F). After 17 hr in sporulation medium many multinucleate cells are visible but few asci are formed. At 4 hr after transfer to rich medium (YPDA), most of these multinu- cleate cells had been converted to asci; after an additional 3 hr in rich medium, spores are observed to germinate.

were scored for their nuclear phenotype at each tem- perature. Representative cells are shown from a spor- ulation culture under restrictive conditions, including mononucleate cells in which the duplicated spindle pole bodies (SPB) have begun to separate (Figure 5B), binucleate cells containing a spindle extending across the nucleus (Figure 5D) and tetranucleate cells con- taining crossed spindles (Figure 5F). Binucleate cells account for 13% of the total cells at 25" and 25% at 34", tri- and tetranucleate cells account for 24% of total cells at 25" and 7% at 34". The remaining cells are mononucleate, of which 50-55% contain SPBs that have begun to separate; the rest contain unsepa- rated SPBs. Thus, spindle pole body separation and spindle formation appear to be normal in this mutant. I t may be noted that in the double staining of meiotic cells the DAPI label was unusually diffuse, for reasons that were not clear.

Commitment to meiosis is defective in SP014 mu- tants: Both the temperature-shift and cytology exper- iments described above suggest that the SP014 func- tion is required at the approximate time of meiosis I and I1 chromosome segregation. This is also the period in which cells become irreversibly committed to the completion of meiosis. We therefore inquired

710 S. M. Honigberg, C . Conicella and R. E. Esposito

TABLE 3

Cell Budding of spol4:URA3:spol4 cells returned to growth

Nuclear cell types (%) Budded/total cells (W) Hours in growth

Genotype medium Mono Bi Tri/Tetra Asci Mono Bi Tri/Tetra Asci

SPO 14 URA3

0 21 13 64 2.5 c 2 5 5 2 7 66 CO. 1 eo . 1 34 78 32

90 14:URA3:spo 14 0 46 35 19 c0.2 14 14 8 spo 14:URA3:spo 14 7 76 19 5 <o. 1 86 78 74

Cells were removed from sporulation medium after 17 hr at 30", washed briefly with water and resuspended in YPDA. Cells were observed via phase-contrast microscope either before or after a 7-hr incubation at 30". The distribution of nuclear cell types is shown in columns 3-6, and the fraction of each cell type which contains buds are shown in columns 7-10. More than 500 cells were counted for each culture. The SP014 strain is H67, the spol4:URA?:spol4 strain is H68.

whether the SP014 function is required for this com- mitment. Commitment was assayed in spo l4 mutants at a time when these cells contained both bi- and tri- and tetranucleate cells (1 7 hr) with little or no nuclear degeneration. Cells were transferred from sporulation to growth medium and the production of asci deter- mined after 4 and 7 hr. Wild-type cells assayed in the same way behaved as expected; cells transferred from post-commitment stages (ie., binucleate and tetranu- cleate cells) first complete both meiotic divisions and form asci prior to resuming mitotic cell division (Fig- ure 6 and Table 3). In contrast, the isogenic spol4:URA3:spol4 disruption/duplication homozy- gotes appear defective in the commitment process in that binucleate and tetranucleate cells can directly return to growth without completion of meiosis. While a majority of cells were unbudded and multin- ucleate at 17 hr in sporulation medium, within seven hours after transfer to rich medium, most of these cells had at least one bud (Table 3). The buds from either binucleate or tetranucleate cells were unusually elongated (Figure 7, A-D) and, in most cases, several cycles of cell division occur without complete separa- tion of the cells. After several divisions, the nascent cells appeared to begin dividing normally (Figure 7, E and F). The lack of cell separation in the population of cells returning to mitotic growth enabled us to ask whether multinucleate cells can bud from more than one nucleus. The results indicate that approximately 20% of multinucleate cells do so (see multinucleate cells with more than one bud in Figure 7, C-E). In these cases, usually one of the nuclei appears to have undergone more divisions than the other, suggesting that the nuclei do not divide in synchrony.

The spol4-1 diploid behaved in a similar manner to the disruption/duplication strain: binucleate and te- tranucleate cells from the restrictive temperature ap- peared to return to growth when transferred to growth medium at the same temperature. However, because a high proportion of the sporulating cells in this strain background retain their buds, it was diffi- cult to precisely quantitate the fraction of meiotic ceiis

that bud upon return to growth. In summary, cyto- logical analysis indicates that spol4 mutants, unlike wild-type cells, do not irreversibly commit to meiosis and can reinitiate mitosis after either one or two meiotic divisions.

Genetic data (Table 4) confirm that binucleate and tetranucleate cells are not irreversibly committed to completing meiosis and spore development in spo14 mutants, and demonstrate that these cells return to growth as fully viable cells. The percentage of spo14- 1 cells committed to at least the meiosis I reductional division, that is diploids homoallelic for an initially heteroallelic close centromere-linked marker plus haploids (see MATERIALS AND METHODS), at ten hours of sporulation is significant at both the permissive and the restrictive temperature (55% and 30%, respec- tively). In the spo14-1 mutant at the permissive tem- perature, cells that have executed a reductional divi- sion are exclusively haploids. In contrast, at the re- strictive temperature, only 50% of the cells that have undergone a redu-ctional division are haploids; the remainder are diploids (Table 4; see MATERIALS AND METHODS for ploidy assay). The diploids have exe- cuted only meiosis I and not meiosis 11, a situation never observed at the permissive temperature or in the wild-type (reviewed in ESPOSITO and KLAPHOLZ 1981; MITCHELL 1988). Thus, viable cells can be formed after a reductional division despite the ab- sence of spore formation. Similarly, the spol4:URA3:spol4 homozygote also allows the pro- duction of high levels of viable, reductionally divided cells as monitored by segregation of close centromere- linked markers on three other chromosomes (Table 4). These results demonstrate that the execution of meiosis I is not sufficient for irreversible commitment to meiosis I1 and spore formation.

Precommitment events occur normally in spol4 strains: DNA synthesis, measured by the DABA fluo- rometric assay, demonstrates an approximate dou- bling of DNA content in H16 by 10 hr at both permissive and restrictive temperatures, similar to wild-type strains (Table 5; see also ESPOSITO and KLAP-

Commitment to Meiosis 71 1

i

" - .. . -

I

FIGURE i.-Return to growth of ;I spol4:URA3:spol4 strain. A sp014:Ul?A3:spol# strain (H68) w a s exposed t o sporul;ttion medium for 17 hr. hefore transfer to YPDA me- dium for 7 hr. Representative cells are slmwn visualized by DAPl fluo- rescence (A. <: and I;) or phase con- trast microscopy (R. D and E). Both binucleate cells (A and R) and tri/ tetranucleate cells (C-F) were ob- served to bud repeatedly and return to growth.

TABLE 4

Presence of reductional products after return to growth

Percentage of reductional

30" 8 .i

spo 14:URA3:spo I4 spo I4:liRA3:spo I4

30" 64

I50

450

64 52 48

21 16 20

Sanlplcs were removed from ;I spol4-I sporul;ltion culture a t I O hr ; u n d fro111 SPO14 or spo14:LiR,43:spo14 cultures a t I 7 hr. In both cases the times were chosen to ensure high frequewics of both Iinuclcate and ~ C I I . ; I I I I I C ~ C ; I I ~ cells ir l the culture. l h e spol4-I strain is H16, the wild-type S f 0 1 4 strain is 1367. and the spoI4:tiRA3:spo14 strain is 1~168.

* "Reductional products/cfu" is calrulated a s the f ~ x t i o n of viable cells t h a t k l v c lost heterozygosity at the centromere-linked gene, I P u I , which is present i n hrttroallrlir condition i n the spol4-1 strain, or a s twice the fraction of viable rrlls t h a t h;lve become auxotrophic for the indicated htttrozypus centronlc,re-linkrd genes present i n the SPO14 ; tnd the spoI4:1/RA3:spo14 strains.

-I'ests fc)r the ploidy of reduction;lI products is dcsrribetI i n M A W R I A I S A N I ) MFI'HODS.

HOLZ 1981). Meiotic recombination, monitored by retain viability at 34" better than cells that do enter heteroallelic exchange at his7 and t rp5 , also occurred meiosis. Analysis of heteroallelic recombination in the at wild-type levels at both permissive and restrictive spoI4:URA3:spo14 mutant at the leu2 locus occurred temperatures in the s p o l 4 - l / s p o l 4 - l strain (Table 5). at expected meiotic levels as well (data not shown). Moreover, the kinetics of recombinant formation in Finally, intergenic recombination in the spol4-1 ho- the mutant diploid and in the isogenic control bearing mozygote in the leuI-trp5 and CEN3-MAT intervals the SP014 gene on a CEN plasmid were nearly iden- was not significantly different at the two temperatures tical (Figure 8). The small difference in the final level or from the reported map distances of 16 and 25 cM, of recombinants at the two temperatures (1.5-2.5- respectively (Table 5). These frequencies are several fold lower at 34") may be because cells that do not hundredfold higher than mitotic exchange. Thus, pre- enter meiosis (and thus are always non-recornbin;tnts) meiotic DNA synthesis, intragenic and intergenic

712 S. M. Honigberg, C. Conicella and R. E. Esposito

TABLE 5

DNA content and recombination in spol4-llspol4-1 strains during sporulation

recombination* recombination (cM)'

content His+/104 Trp+/lOs leul- CEN3- Temp T,,/T0" cfu cfu trpSd MAT'

Intragenic Intergenic

DNA

25" 1.7 5.6 10.7 12.0 32 34" 2.2 3.0 5.7 10.5 25

The strain used for three experiments was H16, all tests were done on cells taken from 10 hr of sporulation.

a DNA/cell is expressed as the ratio of DNA content per cell at ten hours of sporulation to the content immediately after transfer to sporulation medium.

The frequency of intragenic recombinants is measured as the fraction of prototrophs formed among the total viable cells. ' Colonies tested for intragenic recombination were isolated as

His' prototrophs to ensure that each isolate arose from a single cell. The leul-trp5 map distance is the frequency of meiotic cells

that undergo a change in the coupling between the leu1 and trp5 heteroalleles (see MATERIALS AND METHODS); 304 and 383 colonies were tested at 25" and 34", respectively.

' The CEN3-MAT map distance is the frequency of diploid cells that have become homozygous at the MAT locus; 150 and 287 colonies were tested at 25" and 34". respectively.

meiotic recombination are unaffected by the spol4 mutation.

spol4 mutants rapidly lose their ability to return to growth late in meiosis: Wild-type cells retain via- bility throughout sporulation as measured by their ability to form colonies when plated on growth me- dium. Prior to irreversible commitment to the meiotic divisions, the cells return to growth as diploids, while after commitment they complete meiosis and sporu- lation before reentering the mitotic cycle after ger- mination (ESPOSITO and ESPOSITO 1974; SIMCHEN, PIGON and SALTS 1972). T o determine the viability of $014-1 mutants during meiosis, samples from a sporulation culture were removed at various times and plated on synthetic complete medium. When sporulated at the restrictive temperature, the strain displays a rapid loss in viability from 8 to 12 hr, slightly later than the time interval required for SP014 func- tion defined by temperature-shift experiments. As expected, at the permissive temperature this strain displays high viability throughout sporulation, while the isogenic control containing a CEN plasmid bearing the SPOI4 gene maintains high viability at both tem- peratures (Figure 8B). The spol4:URA3:spol4 disrup- tion/duplication diploid also rapidly loses viability after the meiotic divisions (data not shown).

Requirement for SP014 is independent of recom- bination or the meiosis I division: Mutants that are defective in a particular stage of meiosis but are ca- pable of proceeding through subsequent events have proven to be extremely useful for clarifying the de- fects in other meiotic mutants via double mutant, or epistasis, analysis (reviewed by ESPOSITO, DRESSER and BREITENBACH 1991). For example, if the phenotype

100 A.

.01 1

Hours of Sporulation

FIGURE 8.-Intragenic recombination at his7 and cell viability in spo l4 -1 cells during sporulation. HI6 cells containing or lacking a CEN plasmid, pS25, that bears the SP014 gene were exposed to sporulation medium at 25" or 34" for various times before plating on: (A) medium lacking histidine to detect prototrophic intragenic recombinants (25', no plasmid: open circles; 34". no plasmid: closed circles; 25", pS25: open squares; 34", pS25: closed squares) and (B) complete medium to determine the viability of the cells (25", no plasmid: open triangles; 34", no plasmid: closed triangles; 25", pS25: open squares; 34", pS25: closed squares). Prototroph values are given per colony forming units (cfu).

of a given mutant depends upon the execution of the wild-type function of another meiotic gene, then elim- ination of that other function (by mutation) should suppress the mutant defect. This type of analysis was carried out on the $014 mutant to determine if SP014 function is dependent upon the completion of recom- bination or reductional chromosome segregation. T o determine if the spol4 phenotype requires completion of recombination, we examined the behavior of dip- loids homozygous for spol4-1 and spol l-1. Diploids homozygous for the spo l l -1 mutant are Rec- at all temperatures and sporulate at a reduced level (KLA- PHOLZ, WADDELL and ESPOSITO 1985). Table 6 shows the phenotypes of a closely related spo l l - I homozy- gote, $1014-1 homozygote, and spol l -1 spol4-1 dou- bly mutant homozygote. The double mutant strain is Rec- (like s p o l l - I ) and completely fails to sporulate at the restrictive temperature (like spol4- I ) . DAPI staining indicates that both meiotic divisions occur (data not shown). The spol4-1 spol l -1 double mutant has lower viability at the restrictive temperature than the closely related spo14-I S P O l l single mutant, pre- sumably because spol I strains segregate chromosomes

Commitment to Meiosis 713

TABLE 6

Meiotic phenotype of spol4 spol l and spol4 spol3 strains

Percent Dyad/tetrad Percent Trp+/105 His'/lO" Reductional Genotype Strain Temp sporulation (X 100) viability cfu cfu division"

spol4 H16 25" 65 17 97 1070 780 55 ( n = 304) spo I 4 34 eo . 1 43 570 650 30 (n = 383)

H6 1 25" 21 42 46 2

spor I 34" 21 38 40 2

$ 1 0 1 1 s p o l 4 s p o l l spo l4 34" <o. 1

H36 25" 13

spol3 spo I3

25" 44 53 ~ 2 9 2 340

s p o l 3 spol4 sbo13 sbo14 34" eo. 1

H46 25" 52

46

>99 >99

99

48 15

67 56

95 50

6 7

1 .o 0.8

245 246

<5 <5

230 <3 (n = 250) 290 <3 (n = 250)

at the completion of sporulation.

randomly in the first division, generating inviable aneuploid spores. Thus, the spol4 phenotype is not rescued (suppressed) by the Rec- mutation. The sim- plest interpretation of this result is that the wild-type SP014 function is independent of meiotic recombi- nation.

We next inquired whether meiosis I segregation was required for the spol4 phenotype. In this case, the behavior of a spo14-1 ~$013-1 double mutant homozygote was examined. The spol3-1 nonsense mutation prevents meiosis I; only a meiosis 11-like equational division occurs, resulting in the formation of asci containing two diploid spores, termed "dyads" (KLAPHOLZ and ESPOSITO 1980). Table 6 shows that the spol4-1 mutation is not suppressed by the bypass of meiosis I. The $1014-1 $013-1 double mutant at the restrictive temperature undergoes a single-divi- sion meiosis similar to the spol3-1 single mutant, as revealed by DAPI staining (data not shown), but does not form spores. Genetic analysis of double mutant cells in return-to-growth studies indicates that the single division is equational (Table 6). These results provide evidence that the SP014 function acts inde- pendently of meiosis I segregation.

DISCUSSION

The genetic determination of cell fate from among two or more alternatives is a fundamental issue in development. In this paper we describe a gene that influences the choice between mitotic and meiotic fates during meiotic development. Yeast cells initiate meiosis in response to starvation for glucose and ni- trogen but retain the potential for either mitosis or meiosis through the early meiotic events of DNA

Asci formation was measured after 72 hr of sporulation, all other phenotypes were measured after 10 hr of sporulation, except where noted. ' The percentage of viable products that have undergone a reductional division was determined as described in MATERIALS AND METHODS.

The number of colony isolates tested in each experiment is given in parentheses. The data for reductional division in the spo l4 - l / spo l4 - I strain shown in the first row is taken from the experiment reported in Table 4; data for reductional division in the s p o l 3 - l / s p o l 3 - l s p o l 4 - l / sbo14-I strain shown in the fifth row is taken from KLAPHOLZ. WADDELL and ESPOSITO (1985), and was measured by dissection of dyad asci

replication, chromosome pairing and recombination. If glucose and nitrogen are supplied to these early meiotic cells, they terminate meiotic development and return directly to mitotic growth. In contrast, after "commitment to meiosis," the later meiotic events occur even in the presence of glucose and nitrogen, i .e. , the cells lose the capacity to return to growth and instead complete both meiotic divisions and sporulate before resuming mitotic division. Thus, commitment to meiosis has been operationally defined as that stage in meiotic development in which the cells' meiotic fate is irreversibly determined (KIRSOP 1954; GANESAN, HOLTER and ROBERTS 1958). The results presented here show that SP014 is required for the meiotic commitment process. In the absence of the SP014 function, both binucleate and tetranucleate meiotic cells (which normally are unable to return to growth prior to completing the sporulation process) can effi- ciently bud and produce viable mitotic cells when challenged with growth medium.

In both the original temperature-sensitive $014 mutant and the disruption/duplication strain, events that take place prior to commitment (DNA replication and recombination) occur normally. Meiosis I is largely unaffected (less than 1 .!%fold reduction), while meiosis I1 takes place at much lower frequency (10- fold reduction). No spore formation is observed and late in meiosis the nuclei of spol4 cells begin to degen- erate and the cells lose viability. Temperature-shift experiments aimed at defining the execution point of the SP014 gene function are consistent with the view that SP014 acts after the completion of recombina- tion, at approximately the same time that the meiotic divisions take place. Finally, double mutant analysis

714 S. M. Honigberg, C. Conicella and R. E. Esposito

with either the spoil or spo13 mutation, eliminating recombination or meiosis I segregation, respectively, indicates that the role of SP014 in ascus formation is independent of these two landmark events.

Because commitment and meiosis I segregation hap- pen concurrently, several groups have suggested that separation of the spindle pole body (SPB) at meiosis I may be the Commitment event (HORESH, SIMCHEN and FRIEDMANN 1979; ESPOSITO and KLAPHOLZ 198 1 ; MURRAY and SZOSTAK 1985). The phenotype of spol4 mutants clearly demonstrates that this is not the case, since SPB separation occurs in the absence of com- mitment. The phenotype of another mutant, cdc36, also demonstrates that SPB separation can be uncou- pled from commitment. In this case, the mutant ar- rests in meiosis with mononucleate cells that are ca- pable of returning to growth, even though they often contain separated spindle pole bodies (SHUSTER and BYERS 1989). The behavior of the $1014 mutant fur- ther indicates that neither meiosis I nor meiosis 11 irreversibly commits cells to completing the meiotic process, since commitment does not occur even among those .cells undergoing the first or both divi- sions.

Below we consider three mechanisms by which the commitment process may occur and how each relates to the requirement for SP014 in both meiosis I1 and subsequent spore formation, as well as in return to growth. The first two have been discussed previously (see Introduction). These mechanisms are: (1) repres- sion of cell division functions that allow the return to growth, (2) induction of a meiotic event whose exe- cution irreversibly prevents mitotic division, and (3) induction of a meiotic event+) that occurs at a more rapid rate than the return-to-growth response but does not specifically inhibit mitotic functions (for con- venience, this is referred to as the “kinetic choice” model). If SP014 acts via the first mechanism, ie., as a negative regulator of mitotic functions required for return to growth, then at least one other function of SP014 must be postulated to account for its positive role in meiosis I1 segregation and sporulation. In the second model, SP014 could be required for only a single meiotic event, but that event must also irre- versibly prevent return to growth. In the third model, the SP014 gene mediates the induction of a late meiotic event(s) that simply occurs more rapidly than does the response to added glucose and nitrogen. This model is the only one in which it is not necessary to postulate tw% roles for SP014; a positive one in meiosis and a negative one for mitotic functions. In this case, the kinetic difference alone causes the completion of meiosis to appear irreversible; at present, we favor this alternative for its simplicity. The absence of com- mitment in spof4 mutants follows from the failure to complete these later meiotic events, thereby allowing

the necessary time to return to growth. This type of kinetic choice mechanism may also operate during commitment to mitosis, since it has been proposed that a “positive feedback loop” leading to rapid induction of the G1 cyclins causes commitment to the comple- tion of the mitotic cycle at START (CROSS and TIN-

A striking feature of commitment to meiosis in yeast is that all of the late meiotic events (meiosis I and 11 segregation and spore formation) become committed at the same point. If the effect of the SP014 gene on commitment is through kinetic choice, this simulta- neous commitment implies that SP014 is involved in the induction of each of these events. In this regard, the increasingly severe phenotype of the spol4 muta- tions, ie., a slight decrease in meiosis I segregation, a more dramatic reduction in meiosis 11, and complete absence of spore formation, would reflect their differ- ent dependencies on SP014 for induction. Thus SP014 may play a critical role in coordinating the late stages of meiotic development.

Evidence for the coordination of events in meiosis invites comparison to the regulation of successive cel- lular processes during mitotic growth. Cell cycle con- trol in mitosis occurs via a chain of dependent events. When early mitotic events are blocked, later events do not initiate (PRINCLE and HARTWELL 1981), at least in part as a result of feedback or “checkpoint” controls such as the RAD9 gene, which functions to delay chromosome segregation until replication is complete (HARTWELL and WEINERT 1989). The find- ing that both binucleate and tetranucleate cells are capable of returning to growth in $1014 strains is consistent with previous conclusions from analysis of other meiotic mutants that meiosis I, meiosis 11, and spore formation are not on a single obligatory de- pendent pathway but rather can occur independently (ESPOSITO and KLAPHOLZ 198 1). If each meiotic land- mark event is independent of prior events, then the coordination of events in meiosis must depend on a somewhat different mechanism than the one proposed for the mitotic cell division cycle.

How might the coordination of genetically “inde- pendent” meiotic events be accomplished? Several po- tential mechanisms have been suggested to explain the coordination of mitotic events in mutants lacking the normal checkpoint functions (LI and MURRAY 1991). One additional possibility is suggested by the requirement for SP014 in several genetically inde- pendent events, and the increasingly severe effect of spol4 mutations on each subsequent event noted above. If each sequential event has an increased de- pendency on a common regulator such as SP014 , then an increasing level of SP014 activity during meiosis would ensure that the sequential events occur with precise order and timing. A system such as this may

KELENBERG 199 1 ; DIRICK and NASMYTH 199 1).

Commitment to Meiosis 715

be required to allow the rapid progression of cells through meiosis, and might be analogus to the rapid mitotic nuclear divisions during early embryogenesis in Drosophila melanogaster and Xenopus lamas, which lack the dependency relationships between cell cycle events characteristic of mitosis in somatic cells (re- viewed in O'FARRELL et al. 1989; MURRAY and KIRSCHNER 1989).

In summary, spol4 mutants do not become irre- versibly committed to meiosis during the late stages of meiosis when wild-type cells are fully committed. Additionally, although early meiotic events occur nor- mally in $014 mutants, increasingly severe pheno- types accumulate late in meiosis. Taken together, these results suggest that SP014 encodes a function involved in the coordinated induction of late meiotic events, and that this coordinate induction may be responsible for the phenomenon of commitment.

We are grateful to SUE KLAPHOLZ and MARK TOWNSEND for preliminary experiments on ~ $ 0 1 4 - 1 . We also thank SCOTT HOUTTE- MAN for providing the yeast library, J. KILMARTIN for providing the anti-tubulin antibodies, and CATHY ATCHESON, BRECK BYERS, LELA BUCKINGHAM, DAN GOTTSCHLING, BOB MCCARROLL, BRIAN SCHUTTE, BETH SHUSTER and RANDY STRICH for helpful comments on the manuscript. This work was supported by National Institutes of Health grants HD19252 and GM298182 (R.E.E.), Helen Hay Whitney postdoctoral fellowship (S.M.H.), and NATO-CNR fellow- ship (C.C.).

LITERATURE CITED

BOTSTEIN, D., S. C. FALCO, S. E. STEWART, M. BRENNAN, S. SCHERER, D. T. STINCHCOMB, K. STRUHL and R. W. DAVIS, 1979 Sterile host yeasts (SHY) a eukaryotic system of biolog- ical containment for recombinant DNA experiments. Gene 8:

BYERS, B., 1981 Multiple roles of the spindle pole bodies in the life cycle of Saccharomyces cerevisiae, pp. 1 19-1 33 in Molecular Genetics in Yeast (Alfred Benzon Symposium 16), edited by D. V. WETTSTEIN. Munksgaard, Copenhagen.

CROSS, F. R., and A. H. TINKELENBERG, 1991 A potential positive feedback loop controlling CLNl and CLN2 gene expression at the start of the yeast cell cycle. Cell 65: 875-883.

DAWES, I . W., and I . D. HARDIE, 1974 Selective killing of vege- tative cells in sporulated yeast cultures by exposure to diethyl ether. Mol. Gen. Genet. 131: 281-289.

DIRICK, L., and K. NASMYTH, 1991 Positive feedback in the activation of G1 cyclins in yeast. Nature 351: 754-757.

DOWER, W. J., J. F. MILLER and C. W. RAGSDALE, 1988 High efficiency transformation of E. coli by high voltage electropor- ation. Nucleic Acids Res. 16: 6 127-6 145.

DRESSER, M. E., and C. N. GIROUX, 1988 Meiotic chromosome behavior in spread preparations of yeast. J. Cell Biol. 106 567- 73.

ENGEBRECHT, J., and G. S. ROEDER, 1989 Yeast mer1 mutants display reduced levels of meiotic recombination. Genetics 121: 237-247.

ESPOSITO, R. E., M. DRESSER and M. BREITENBACH, 1991 Identifying sporulation genes, visualizing synaptonemal complexes, and large-scale spore and spore wall purification. Methods Enzymol. 194: 110-131.

ESPOSITO, R. E., and M. S. ESPOSITO, 1969 The genetic control of sporulation in Saccharomyces: I . The isolation of tempera-

17-24.

ture-sensitive sporulation-deficient mutants. Genetics 61: 79- 89.

ESPOSITO, R. E., and M. S. ESPOSITO, 1974 Genetic recombination and commitment to meiosis in Saccharomyces cerevisiae. Proc. Natl. Acad. Sci. USA 71: 3172-3176.

ESPOSITO, R. E., and S. KLAPHOLZ, 1981 Meiosis and ascospore development, pp. 2 1 1-287 in The Molecular Biology of the Yeast Saccharomyces: Lqe Cycle and Inheritance, edited by J. STRATH- ERN, E. JONES AND J. BROACH. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y.

ESPOSITO, M. S., R. E. ESPOSITO, M. ARNAUD and H. 0. HALVOR- SON, 1970 Conditional mutants of meiosis in yeast. J. Bacte- riol. 104 202-210.

GANESAN, A. T., H. HOLTER and C. ROBERTS, 1958 Some obser- vations on sporulation in Saccharomyces. C. R. Lab. Carlsberg 13: 1-6.

HARTWELL, L. H., 1974 Saccharomyces cerevisiae cell cycle. Bacte- rial. Rev. 38: 164-198.

HARTWELL, L. H., and T. WEINERT, 1989 Checkpoints: controls that ensure the order of cell cycle events. Science 246 629- 634.

HORESH, O., G. SIMCHEN and A. FRIEDMANN, 1979 Morphogenesis of the synaptonemal complex during yeast meiosis. Chromosoma 75: 101-1 15.

HOYT, M. A., L. TOTIS and B. T. ROBERTS, 1991 S. cerevisiae genes required for cell cycle arrest in response to loss of microtubule function. Cell 6 6 507-5 17.

KASSIR, Y., D. GRANOT and G. SIMCHEN, 1988 I M E l , a positive regulator of meiosis in S. cerevisiae. Cell 52: 853-862.

KILMARTIN, J. V., B. WRIGHT and C. MILSTEIN, 1982 Rat mono- clonal antitubulin antibodies derived by using a new nonse- creting cell line. J. Cell Biol. 93: 576-582.

KIRSOP, B. H., 1954 Studies in yeast sporulation. I. Some factors influencing sporulation. J. Inst. Brew. 6 0 393.

KLAPHOLZ, S., and R. E. ESPOSITO, 1980 Recombination and chromosome segregation during the single division meiosis in spol2-1 and spol3-1 diploids. Genetics 96: 567-572.

KLAPHOLZ, S., C. S. WADDELL and R. E. ESPOSITO, 1985 The role of the S P O l l gene in meiotic recombination in yeast. Genetics 110 187-216.

Lr, R., and A. W. MURRAY, 1991 Feedback control of mitosis in budding yeast. Cell 66: 5 19-53 1.

MALONE, R. E., and R. E. ESPOSITO, 1981 Recombinationless meiosis in Saccharomyces cerevisiae. Mol. Cell. Biol. 1: 891-901.

MENEES, T. M., and G. S. ROEDER, 1989 MEI4, a yeast gene required for meiotic recombination. Genetics 123: 675-682.

MITCHELL, A. P., 1988 Meiotic inhibition. Molecular control of meiosis, pp. 47-66 in Progress in Clinical and Biologzcal Re- search, edited by F. P. HASELTINE and N. L. FIRST. Alan R. Liss, New York.

MITCHELL, A. P., and I. HERSKOWITZ, 1986 Activation of meiosis and sporulation by repression of the M E 1 product in yeast. Nature 3 1 9 738-742.

MOENS, P. B., and E. RAPPORT, 1971 Spindles, spindle plaques and intranuclear meiosis in the yeast Saccharomyces cerevisiae. J. Cell Biol. 50: 344-361.

MORTIMER, R. K., and D. C. HAWTHORNE, 1969 Yeast genetics, pp. 386 in The Yeasts, edited by A. H. ROSE and J. S. HARRISON. Academic Press, New York.

MORTIMER, R. K., D. SCHILD, C. R. CONTOPOULOU and J. A. KANS, 1990 Genetic map of Saccharomyces cerevisiae, Edition 10. Yeast 5: 321-403.

MURRAY, A. W., and M. KIRSCHNER, 1989 Dominoes and clocks: the union of two views of the cell cycle. Science 246: 614-626.

MURRAY, A. W., and J. W. SZOSTAK, 1985 Chromosome segre- gation in mitosis and meiosis. Annu. Rev. Cell Biol. 1: 289- 315.

O'FARRELL, P. H., B. A. EDGAR, D. LAKICH and C. F. LEHNER,

716 S. M. Honigberg, C. Conicella and R. E. Esposito

1989 Directing cell division during development. Science

PADMORE, R., L. CAO and N. KLECKNER, 1991 Temporal com- parison of recombination and synaptonemal complex forma- tion during meiosis in S. cerevisiae. Cell 66: 1239-1 256.

PRINGLE, J., and L. HARTWELL, 1981 The Saccharomyces cerevisiae cell cycle, pp. 97-142 in Molecular Biology of the Yeast Saccha- romyces: Lije Cycle and Inheritance, edited by J. N. STRATHERN, E. W. JONES and J. R. BROACH. Cold Spring Harbor Labora- tory, Cold Spring Harbor, N.Y.

ROMAN, H., M. M. PHILLIPS and S. M. SANDS, 1955 Studies of polyploid Saccharomyces: I. Tetraploid segregation. Genetics 40: 546.

ROSE, M. D., F. WINSTON and P. HIETER, 1990 Methods in Yeast Genetics: A Laboratory Course Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y.

ROTHSTEIN, R., 1991 Targeting, disruption, replacement, and allele rescue: integrative DNA transformation in yeast. Meth- ods Enzymol. 194 281-301.

SAMBROOK, J., E. F. FRITSCH and T . MANIATIS, 1989 Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y.

SCHILD, D., and B. BYERS, 1978 Meiotic effects of DNA-defective cell division cycle mutations of Saccharomyces cerevisiae. Chro- mosoma 70: 109-130.

SHERMAN, F., and H. ROMAN, 1963 Evidence for two types of

SHILO, V., G. SIMCHEN and B. SHILO, 1978 Initiation of meiosis

246: 635-640.

allelic recombination in yeast. Genetics 4 8 255-261.

in cell cycle initiation mutants of Saccharomyces cerevisiae. Exp. Cell Res. 112: 241-248.

SHORTLE, D., J. E. HABER and D. BOTSTEIN, 1982 Lethal disrup- tion of the yeast actin gene by integrative DNA transformation. Science 217: 371-3.

SHUSTER, E. O., and B. BYERS, 1989 Pachytene arrest and other meiotic effects of the start mutations in Saccharomyces cerevisiae. Genetics 123: 29-43.

SIKORSKI, R. S., and P. HIETER, 1989 A system of shuttle vectors and yeast host strains designed for efficient manipulation of DNA in Saccharomyces cerevisiae. Genetics 122: 19-27.

SIMCHEN, G., 1974 Are mitotic functions required in meiosis? Genetics 76: 745-753.

SIMCHEN, G., R. PIGON and Y . SALTS, 1972 Sporulation in Sac- charomyces cerevisiae: Premeiotic DNA synthesis, readiness and commitment. Exptl. Cell Res. 75: 207-218.

SMITH, H. E., and A. P. MITCHELL, 1989 A transcriptional cascade governs entry into meiosis in Saccharomyces cerevisiae. Mol. Cell. Biol. 9: 2142-2152.

THOMAS, D., and D. BOTSTEIN, 1986 A gene required for the separation of chromosomes on the spindle apparatus in yeast. Cell 44: 65-76.

TODA, T., I. UNO, T . ISHIKAWA, S. POWERS, T. KATOKA, D. BROEEK, S. CAMERON, J. BROACH, K. MATSUMOTA and M. WIGLER, 1985 In yeast, RAS proteins are controlling ele- ments of adenylate cyclase. Cell 4 0 27-36.

WEINERT T . A,, and L. H. HARTWELL, 1988 The RAD9 gene controls the cell cycle response to DNA damage in Saccharo- myces cerevisiae. Science 241: 317-322.

Communicating editor: E. W. JONES