that the SPOI 1 gene is specifically required for meiotic recombination. First, ... meiosis I segregation, and hence, the production of viable haploid gametes,.
Copyright 0 1985 by the Genetics Society of America
T H E ROLE OF T H E S P O l l GENE I N MEIOTIC RECOMBINATION I N YEAST SUE KLAPHOLZ, CANDACE S. WADDELL' AND ROCHELLE EASTON ESPOSITO
Department of Molecular Genetics and Cell Biology, The University of Chicago, Chicago, Illinois 60637 Manuscript received October 5, 1984 Revised copy accepted February 23, 1985 ABSTRACT
Several complementary experimental approaches were used to demonstrate that the SPOI 1 gene is specifically required for meiotic recombination. First, sporulating cultures of $01 1-1 mutant diploids were examined for landmark biochemical, cytological and genetic events of meiosis and ascosporogenesis. Cells entered sporulation with high eficiency and showed a neardoubling of DNA content. Synaptonemal complexes, hallmarks of intimate homologous pairing, and polycomplex structures appeared during meiotic prophase. Although spontaneous mitotic intra- and intergenic recombination occurred at normal levels, no meiotic recombination was observed. Whereas greater than 50% of cells completed both meiotic divisions, packaging of the four meiotic products into mature ascospores took place in only a small subset of asci. Haploidization occurred in less than 1 % of viable colony-forming units. Second, the Rec- meiotic defect conferred by sfloll-1 was confirmed by dyad analysis of spores derived from spol3-1 singledivision meiosis in which recombination is not a requirement for viable ascospore production. Diploids homozygous for the spol3-1 mutation undergo meiotic levels of exchange followed by a single predominantly equational division and form asci containing two near-diploid spores. With the introduction of the g o 1 1-1 mutation, high spore viability was retained, whereas intergenic recombination was reduced by more than 100fold.
HE process of meiotic recombination has been investigated using a variety T of approaches, including the biochemical, cytological and genetic analysis of wild-type and mutant organisms (reviewed by BAKERand HALL1976; BAKER et al. 1976; STERNand HOTTA 1977; GOLUBOVSKAYA 1979; &POSIT0 and KLAPHOLZ 1981; DAWES1983). Such studies have demonstrated that proper meiosis I segregation, and hence, the production of viable haploid gametes, depends on meiotic recombination. When recombination is greatly reduced or absent, homologous chromosomes generally segregate randomly from one another at meiosis I, generating aneuploid gametes. Thus, in organisms in which a high level of aneuploidy leads to gametic inviability (e.g., plants, fungi), direct characterization of mutants defective in meiotic recombination becomes a difficult task.
'
Current address: Department of Biochemistry and Biophysics and the George W. Hooper Foundation, University of California, San Francisco, California 94143. Genetics 1 1 0 187-216, June, 1985.
188
S. KLAPHOLZ, C. S. WADDELL AND R . E. ESPOSITO
Two major approaches have been taken to diagnose recombination-defective (Rec-) mutations when adequate numbers of viable meiotic products cannot be obtained for genetic analysis. First, cytological examination of meiotic chromosome behavior in mutant organisms has been utilized to study the dependency of proper meiosis I segregation on chromosome pairing and recombination. For instance, studies in plants have demonstrated that defects in the establishment or maintenance of homologous chromosome pairing may result in complete failure of one of the two meiotic divisions (LEVAN1940; JOHNSSON 1944; STRINGHAM 1970). Unpaired chromosomes at metaphase I, present as a consequence of mutation, monosomy or haploidy, may undergo misdivision or equational centromere division, as well as random segregation (BELLING and BLAKESLEE 1927; LESLEY and FROST 1928; SEARS1952; SADASIVAIAH and KASHA 1971). Electron microscopic studies of meiotic chromosome behavior in a wide variety of organisms have further shown that chromosome pairing and exchange are highly correlated with the presence of synaptonemal complexes (SCs) (reviewed by MOSES 1968, 1977; WESTERGAARD and VON WETTSTEIN1972; GILLIES 1975). In yeast, in which it has not been possible to visualize discrete condensed chromosomes (BYERS1981), the analysis of SC structure has provided a means to explore the meiotic defects conferred by several cell division cycle (cdc) mutations (BYERSand GOETSCH1975; SCHILDand BYERS1978; and FRIEDMANN1979). HORESH, SIMCHEN A second approach to characterizing Rec- mutants involves the uncoupling of recombination from reductional segregation at meiosis I, thereby eliminating the dependence of viable product formation on meiotic exchange (Figure 1). Such an uncoupling occurs when cells are interrupted during meiotic development and returned to vegetative growth (SHERMAN and ROMAN 1963; R. E. ESPOSITOand M. S. ESPOSITO1974). Under these conditions, cells exhibiting meiotic levels of intra- and intergenic recombination in the absence of haploidization can be recovered. This observation has led to the conclusion that events committing cells to exchange occur prior to, and are separable from, those that commit cells to meiosis I chromosome segregation. The absence of recombinants among the mitotic cells obtained from return-to-growth experiments has been used as a diagnostic criterion for defining putative meiotic Rec- mutants. A significant limitation of this type of analysis, however, arises from the fact that many of the mutants analyzed become highly inviable during sporulation. In these cases it is difficult to distinguish mutations that allow commitment to recombination but prevent return to mitotic growth from those that directly eliminate exchange (6MALONEand ESPOSITO198 1). Another method that depends on the uncoupling of recombination from the meiosis I division employs the spol3-1 mutation (KLAPHOLZ 1980; MALONEand ESPOSITO1981; GAME1983; MALONE 1983). Diploids homozygous for spoI3I undergo meiotic levels of recombination, bypass reductional segregation, undergo a single, predominantly equational division and produce asci containing two diploid or near-diploid spores (KLAPHOLZand ESPOSITO1980a,b) (Figure 1). When certain Rec- mutations are analyzed in the presence of s p o l 3 - I , viable recombinationless two-spored asci are formed (KLAPHOLZ1980; MALONE
189
MEIOTIC RECOMBINATION return to growth
....... .,y diploid cell DNA SYNTHESIS
Wild type Rec' 2 divisions
PAIRING & EXCHANGE
MI DIVISION: REDUCTIONAL
MII DIVISION:
d
B
o_
8
~p013-IRet+
PRODUCTS
EQUATIONAL
2+: 2- tetrad
:.........
1 division
1+: 1- dyad
spof3-l Rec' 1 division
................................. 2+ 0-dyad
FIGURE1.-Schematic diagram of the major genetic events of meiosis showing two approaches that are used to uncouple meiotic exchange from reductional segregation. (1) Cells that are returned to vegetative growth following incubation in sporulation medium undergo meiotic levels of recombination followed by a single mitotic division. (2) The spol3-1 mutation permits full levels of meiotic exchange but eliminates the meiosis I (MI) division; cells undergo a single meiosis I1 (MII) division and form two-spored asci (dyads). (3) When certain Rec- mutations are coupled to s p o l 3 - 1 , both recombination and meiosis I are eliminated; a single meiosis I1 division ensues, leading to the production of viable recombinationless dyads (see text for further details).
and ESPOSITO1981; GAME1983; MALONE1983). This method has advantages over the return-to-growth procedure in that (1) Rec- cells d o not lose viability in sporulation medium because they do not become committed to reductional segregation and (2) all chromosomes derived from a single cell are recovered together in the two-spored ascus allowing reconstruction of each recombination and segregation event. T h e present study was aimed at defining the role of the S P O l l gene in meiosis. The s p o l l - 1 mutation, located on the left arm of chromosome VZZZ, was originally isolated in a large-scale mutant hunt aimed at identifying genes required for meiosis and spore formation (ESPOSITOand ESPOSITO 1969; EsPOSITO et al. 1972; M. S. ESPOSITOand R. E. ESPOSITO1974). T h e s p o l l - 1 mutant sporulates at a reduced level, forms immature and morphologically abnormal asci and exhibits extremely poor spore viability (MOENS et al. 1977; ESPOSITOand ESPOSITO19'78). T h e finding that rare, partially haploidized spol 1-1 meiotic products failed to contain recombinant phenotypes suggested that the wild-type S P O l l gene is required for genetic exchange during sporulation (KLAPHOLZ and ESPOSITO1982a,b). This paper describes an in-depth characterization of the sporulation phenotype conferred by the $01 1-1 mutation employing the complementary genetic and cytological approaches described above. We have attempted to quantitate its effect on intra- and inter-
190
S. KLAPHOLZ, C. S. WADDELL AND R. E. ESPOSITO
MAT
ode5
con/
WO3
metl3 cyhZ
I
l r p 5 /eu/
I
nn si Y
96.0
15.0
42.5
17.6
FIGURE2.-Genetic map of markers used in recombination studies; map distances are in centiMorgans (MORTIMER and SCHILD1980).
genic recombination and to evaluate the dependency of other events of meiosis and spore formation on recombination. MATERIALS AND METHODS
Strains: The primary strains of Saccharomyces cerevisiae used in these studies are listed with their complete genotypes in Table 1. Additional strains from our laboratory collection were used for complementation, ploidy and/or mating-type testing. The linkage relationships of genetic markers used in recombination experiments are shown in Figure 2. Media: The following recipes are for one liter of medium. Tetracycline (Squibb) is added at a final concentration of 20 pg/ml to prevent bacterial contamination. YPA-2 1 (growth medium) contains 10 g of potassium acetate, 20 g of Bacto-peptone and 10 g of Bacto-yeast extract, supplemented with mixture 21. SP11-21 (sporulation medium) contains 20 g of potassium acetate supplemented with mixture 21 titrated to pH 7.0. Mixture 21 consists of 75 mg of each of the following: adenine sulfate, L-arginine-HCI, L-aspartic acid, L-histidine HCI, L-leucine, L-lysine-HCI, L-methionine, L-phenylalanine, L-threonine, L-tryptophan, L-tyrosine and uracil; and 10 mg of PABA. All other media have been described previously (KLAPHOLZ and FSPOSITO 1982a). Genetic techniques: Standard procedures for the genetic manipulation of yeast were employed (MORTIMER and HAWTHORNE 1969; SHERMAN and LAWRENCE 1974; SHERMAN 1975). Mitotic recombination: Single colonies of freshly constructed K303 and K304 diploids were picked into 1 ml of distilled water (dHzO) and inoculated into two separate YPA-Ll-containing flasks at a concentration of 100 cells/ml. One culture was grown at 25" and the other at 34" to a density of 2-3 X lo7 cells/ml. Cells were then sonicated briefly, washed with dHpO and plated on (1) canavanine- or cycloheximide-containing media to monitor recombination between the centromere and can1 or the centromere and cyh2, respectively, and (2) tryptophanless or uracilless media to measure intragenic recombination resulting in prototrophy at trp5 or ura3, respectively. One hundred to 200 colonies of each drug-resistant (drug R) and prototrophic class were picked to YPD master plates for further genetic analysis. Sporulation: Liquid growth cultures of K303 and K304 were prepared by inoculating single colonies of freshly constructed diploids into YPA-21 medium at a density of 8 X lo4 cells/ml. Cultures were grown to -2 X IO' cells/ml, washed and resuspended in SPI 1-21 medium at - 5 x 1O7 cells/ml. Samples were taken at intervals during sporulation, sonicated briefly and washed twice in dHPO; aliquots were used for cell counts, platings, DNA measurements, fluorescence photomicroscopy and electron microscopy (see below). Strains sporulated on solid medium were incubated for 5 days at 25", 30" or 34". All liquid cultures were aerated at 250 rpm at 25" in a gyratory water bath shaker. Cell densities and sporulation frequencies were determined by hemacytometer counts of 200-300 cells. DNA measurements: Premeiotic DNA synthesis was measured by the procedure of DOI and Do1 (1976), as modified by FAST(1978), PLOTKIN(1978) and below. Quadruplicate I-ml aliquots of -5 X lo7 cells were removed from sporulating cultures of K303 and K304 at various times. Samples were spun in an Eppendorf microcentrifuge, the supernatants decanted and the pellets stored at -20". The frozen pellets were resuspended in 0.5 ml of 1.0 N NaOH and incubated at room temperature for 24 hr. Samples were then vortexed, placed in an ice bath and precipitated
191
MEIOTIC RECOMBINATION < A
s;
$1: q g1 1
33
3-
1 1
55 44.
215 .gg
gls ry
&. - 1
$12 $1:
ryry
y 0
%If 51%
3
hr= e - 5
~
22 21% 2 7
0
ui!
2.-
a $
3 3 . c5 a
E .z
C O
0
3%
ryfd
010
1 1
0
fdry
.kj !j
$ 5 zg 1 1 1 1
?': h o h
1:
1 1
a2
3
\o
31s
81s $13 u
U
1
$12 $12
?2
yz 2 2 qz k 2 212 U ' o
:E$& 4.^2z n. - 9
'D
gg
&
2.-
2;
d
U "
v1
'Sa 'i/p E.g ry o u 2E 2%
E
I & &
3
d '2z w
48 &
$12 81% 81% $13 ; 4 4 2- 2-
d
22
v, \o
21s
&tG
31s
$1:
U
v\v\
U
212
212
ryn)
v \ Y
'o
flf
$12
T
$1.
0
El% $5 0
t;ll 2
312 %
2
2-
2 212 2 212 *z
$19 -.-(
713 2 2
212
818 c:c:
- 1
2
1
e
1 1 g
s s
$ 3
21s U -
312
?Ig $1: & %
42 wry
PS sp 412
- 1
d.3 a13 ryry
1%
uu
h h
22
40
U 5 Qlj5 5 $ a or MATafMATafMATa if a> a . * In each dyad, the mating-capable sister spore was shown to be monosomic for chromosome 111 (see Table 7).
during sporulation. T h e relative proportions of 1:0:1, 0: 1:1 and 1:1:O dyads produced by the sporulated MATalMATa and trisomic N clones is a direct reflection of the ploidy of the MAT locus (e.g., MATa/MATa/MATa trisomic diploids produced 1:O:l > 1:l:O > 0:l:l dyads). The analysis demonstrated that (1) N clones from 0:0:2 dyads are MATaIMATa diploids, (2) N clones from 1:O:l dyads are MATa/MATa:/MATa: trisomics and ( 3 ) N clones from 0:l:l dyads are MATa/MATa/MATcr trisomics. T h e same type of dyad analysis was not feasible for the s p o l l - 1 ~ $ 0 1 3 - 1 double mutant because there is no appreciable recombination during sporulation; i.e., nearly all of the dyads derived from N clones are 0:0:2. Instead, we used a procedure exploiting the facts that the $1013-l mutation is ochre suppressible (KLAPHOLZand ESPOSITO 1980a) and that rare haploidized meiotic products can be selectively recovered from s p o l l - 1 diploids (KLAPHOLZand ESPOSITO1982a) (MATERIALS AND METHODS). Ochre suppressors were selected in nonmating ascosporal clones from 0:0:2, 1:O:l and 0:l:l dyads, and the resulting clones, which were S p o l l Spo 13+ in phenotype, were sporulated. Rare haploidized meiotic products were selected by plating for the expression of the recessive drug resistance marker, cyh2, and their mating types determined (Table 8). T h e ratio of a to a segregants was used to infer the MAT genotype of the parental N clone as follows. Parental clones that are MATaIMATa in genotype should segregate an equal number of a and a: meiotic products, whereas trisomic MATalMATalMATa or MATaIMATorlMATa: clones should produce an excess of a or a: segregants, respectively. T h e recovery of nonmating colonies is also expected by this procedure; these would include disomic and trisomic N meiotic products as well as diploid drug R colonies resulting from prior mitotic exchange or chromosome loss events that led to expression of the drug
206
S. KLAPHOLZ, C. S . WADDELL AND R. E. ESPOSITO
resistance allele. In previous studies 30% of colonies recovered from spoll-1 meiosis were nonmaters due to chromosome ZZZ disomy (KLAPHOLZand ESPOSITO 198%). The procedure used in this study yielded 5 0 4 0 % N colonies. Since ochre suppressors are known to depress the overall level of sporulation (ROTHSTEIN, ESPOSITOand ESPOSITO1977), the increased recovery of nonmaters may be due to a higher background of unsporulated diploid cells. Four N clones from 0:0:2 dyads generated equal numbers of a and a segregants as expected from MATaIMATa diploids. By criterion of unequal recovery of a and a segregants three of five N clones from 1:O:l and 0 : l : l dyads were classified as trisomic for chromosome ZZZ (Table 8). Since their sister spores were monosomic for chromosome ZZZ (see Table 7), the origin of these $01 11 $013-1 dyads is consistent with three-from-one chromatid segregation as has been observed in $1013-1diploids (KLAPHOLZand ESPOSITO1980b). T h e remaining two dyads in which one spore is monosomic and the other is disomic could have arisen by single chromatid loss or by three-from-one chromatid segregation with subsequent trisomic instability generating a disome. Segregation of heterozygous recessive markers: T h e segregation behavior of six heterozygous markers distributed over three chromosomes was examined in spol3-1 and spoll-1 $013-1 diploids. The majority class of dyads produced by the spol3-1 diploids was 2+:0- for all heterozygous markers, regardless of their distance from the centromere (Table 9; KLAPHOLZand ESPOSITO1980b). For centromere-unlinked markers we saw a range of 32-44% 1+:1- dyads. These values are consistent with the assumption that 1:l dyads arise from genecentromere exchange; a comparable measurement of gene-centromere exchange in tetrads (one-half of the second-division segregation frequency) gives a similar range (MORTIMER and HAWTHORNE1966). T h e only centromerelinked marker examined in this study, leul (3 cM from the centromere of chromosome VZZ, MORTIMER and SCHILD1980), exhibited 22% 1:1 segregation, a value much higher than expected from gene-centromere exchange alone (KLAPHOLZand ESPOSITO 1980b). Genotypic analysis of 1:1 clones from K 2 15 dyads has indicated that not all 1:l clones arise from gene-centromere exchange ($ KLAPHOLZand ESPOSITO1980b). For example, of four K 2 15 dyads that showed 1:1 segregation for leul only two contained one LEUlILEUl plus one 1eulIEeul spore. T h e other two dyads contained one LEUllleul and one leuZ/leul spore, consistent with either gene conversion of LEUl to leul or loss of a LEUl chromatid coupled with postmeiotic chromosome restitution in the Leu- spore. In the double mutant, the frequency of recombinant dyads was drastically reduced. Among the different gene-centromere intervals examined, the number of 1:l dyads ranged from zero of 171 dyads (
