JOURNAL OF BACTERIOLOGY, Dec. 1975, p. 1545-1557 Copyright 0 1975 American Society for Microbiology
Vol. 124, No. 3
Printed in U.S.A.
Sporulation in -Glucosamine Auxotrophs of Saccharomyces cerevisiae: Meiosis with Defective Ascospore Wall Formation WILLIAM L. WHELAN
AND
CLINTON E. BALLOU*
Department of Biochemistry, University of California, Berkeley, California, 94720 Received for publication 16 July 1975
Mutants that require exogenous D-glucosamine for growth were isolated from Saccharomyces cerevisiae X2180-1A after ethyl methane sulfonate mutagenesis. Class A auxotrophs fail to grow on yeast extract-peptone-dextrose and minimal media, whereas class B auxotrophs grow on minimal medium and readily revert to grow on yeast extract-peptone-dextrose medium. Class B auxotrophs are suppressible by a class of suppressors distinct from nonsense suppressors, and their properties suggest that they are defective in a regulatory function. All 23 mutants studied were recessive and allelic, and they define a new gene designated gcnl. An analysis of a class A auxotroph revealed that it lacked L-glutamine:D-fructose 6-phosphate amidotransferase (EC 2.6.1.16) activity and indicates that GCN1 codes the amino acid sequence of this enzyme. The finding that all mutants were allelic indicates that the amidotransferase is the only enzyme responsible for D-glucosamine synthesis in S. cerevisiae. The occurrence of allelic complementation and media-conditional mutants suggests that the amidotransferase is a multimeric enzyme with an activity subject to metabolic control. Diploids homozygous for gcnl fail to complete sporulation. They proceed through meiosis normally, as judged by the occurrence of meiotic recombination, the production of haploid nuclei, and the formation of multinucleate cells visible after Giemsa staining. However, the formation of glusulase-resistant ascospores is blocked, and deformed spores lacking the electron-dense outer layer characteristic of the normal spore wall are observed by electron microscopy. Cells that acquire the ability to synthesize D-glucosamine, because of gene conversion during meiosis, complete sporulation in a normal fashion. Thus, the GCN1 gene product appears to be synthesized late in sporulation and may prove to be a useful developmental landmark for the termination of ascospore development.
D-Glucosamine (as N-acetyl-D-glucosamine) occurs in two major components of the cell wall of the budding yeast Saccharomyces cerevisiae. It is present as chitin in the bud scars that are formed on separation of the mother and daughter cells (3, 4), and it also occurs in the mannan of the vegetative cell wall, where it links polysaccharide side chains to the protein backbone (27). We demonstrate here that D-glucosamine is essential for the growth of S. cerevisiae because mutants can be isolated that fail to grow in its absence. In addition, we find that the synthesis of this hexosamine is determined by a single gene (denoted GCN1), which probably codes the primary structure of the enzyme L-glutamine:D-fructose 6-phosphate amidotransferase (EC 2.6.1.16). Many of the D-glucosamine auxotrophs that we have isolated are defective in the ability to complete sporulation; they undergo meiosis and
produce haploid nuclei within deformed and glusulase-sensitive ascospores. Thus, we describe a novel type of sporulation-defective mutant differing from those previously described that were defective in the premeiotic round of deoxyribonucleic acid (DNA) synthesis (24) or in the meiotic divisions (5, 7).
MATERIALS AND METHODS Yeast strains. The isogenic wild-type strains S. cerevisiae X2180-1A and X2180-1B were provided by R. K. Mortimer. A strain bearing the alleles leu2-1, his4-4, and trp-1 was provided by S. Fogel. Media. The stated amounts of the following ingredients were dissolved in 1 liter of distilled water. Yeast extract-peptone-dextrose (YEPD) contained 10 g of Difco yeast extract, 20 g of Difco peptone, and 20 g of D-glucose. Minimal medium (MIN) contained 6.7 g of Difco yeast nitrogen base without amino acids and 20 g of D-glucose. Difco agar (20 g) was added when a solid medium was desired. When appropriate, D1545
1546
WHELAN AND BALLOU
glucosamine (Calbiochem A grade) or N-acetyl-Dglucosamine (Sigma Chemical Co.) was added to the autoclaved medium as a filter-sterilized solution. Sporulation of strains for tetrad analysis was performed on sporulation (SPOR) agar made from 2.5 g of yeast extract, 10 g of potassium acetate, and 20 g of agar per liter of distilled water. Incubations were at 30C. General genetic techniques. Crosses were performed by mixing 18-h cultures of the parent strains on YEPD containing 1 mg of D-glucosamine per ml (YEPD-GN). Diploid clones were obtained by isolating zygotes 3 to 6 h after mixing the haploids. Sporulation was induced in the mating mixture by replica plating the mixture to SPOR agar after incubation for 18 h or in an isolated diploid clone by replica plating after incubation for 2 to 3 days. Asci were observed after a 3-day incubation of the culture on SPOR. Tetrad analysis was performed by using glusulase (Endo Laboratories) to digest the ascus wall. Complementation was determined among Dglucosamine auxotrophs by mixing 18-h cultures of opposite mating type on YEPD-GN and replica plating the mating mixture to YEPD after overnight incubation. The YEPD replicas were incubated at 30 C for 7 days and were observed periodically for the appearance of growth. A spot test for allelic recombination was performed by replica plating a diploid culture to SPOR agar and then replica plating this culture, after a 3-day incubation, to the test medium on which only recombinants could grow. The mating type of isolated clones was determined by examining appropriate test crosses of the clone for the presence of zygotes 3 to 6 after mixing. Mutagenesis and mutant isolation. An 18-h YEPD culture of strain X2180-1A grown at 30 C was washed once with distilled water and suspended at a cell density of 5 x 107 cells/ml in 0.1 M potassium phosphate at pH 7.0. To 1 ml of this suspension was added 25 ,l of ethyl methane sulfonate (Eastman). After 30 min at room temperature (23 C) with continuous agitation, the suspension was diluted 1:100 in 5% sodium thiosulfate and then 1:100 in distilled water. The diluted suspension was spread on YEPD-GN, the plate was incubated at 30 C for 2 days, and the colonies were replica plated to YEPD. Colonies incapable of growth on YEPD at 30 C were picked from the master plate. About 50% of the cells survived the mutagenesis. Spontaneous revertants were picked from a YEPD replica of the strain grown on YEPD-GN. Ultravioletinduced revertants were obtained by irradiating a YEPD-GN culture with a General Electric germicidal lamp to yield about 30% survival and were selected by replica plating the cultures to YEPD after overnight incubation. Assessment of sporulation. Overnight cultures in liquid YEPD-GN were washed once with distilled water and suspended in 1% potassium acetate at a cell density of 5 x 107 cells/ml. The acetate cultures were shaken at room temperature and examined periodically during 7 days for ascus frequency, viable count, recombinant frequency, and frequency of haploid formation. The ascus frequency was determined by
J. BACTRIOL. the microscopy count of all cells in a hemocytometer grid (cells, visible buds, and asci were counted as cells), followed by a count of the asci in the same grid. No distinction was made between two-, three- and four-spored asci. Ascus frequencies in the wild-type strains were based on the examination of 100 to 200 cells, whereas ascus frequencies in the mutant strains were based on the examination of more than 500 cells. Spore frequencies were determined by treating the washed SPOR culture with glusulase (a 1:10 dilution of the commercial preparation) and examining the treated suspension for the presence of ascospores. Frequencies are expressed as spores per 1,000 cells, based on the examination of the debris from at least 2,000 cells. Viable counts were obtained from YEPDGN agar plates; the viability did not decline in the strains during the 7-day incubation. Prototrophic recombinants were detected on YEPD or MIN, as appropriate, and recombinant frequency is expressed as recombinants per viable cell. Assessment of haploid formation. Cells from diploid cultures previously incubated in SPOR medium for 7 days were spread on YEPD-GN to yield 100 to 200 colonies per plate. After a 2- to 3-day incubation, each plate was replica plated to a lawn of the his6 tester strain on YEPD-GN (a and a mating type on separate plates). After overnight incubation, the plates containing the testers and colony replicas were replica plated to MIN. The presence of cells capable of mating in the original colonies was revealed by growth on MIN in the area of contact of the colony and the tester. The plates were scored one day after replica plating to MIN for the presence of haploids in the colonies. Incubation for 2 to 4 days revealed the presence of a few nonhaploid cells capable of mating in colonies that were scored as negative the first day. For the determination of the genotype of the mater, the growth on MIN was streaked on MIN to obtain isolated colonies. These were sporulated and subjected to tetrad analysis. Assay for L-glutamine:D-fructose 6-phosphate amidotransferase. Overnight cultures (17 h in experiment 1 and 24 h in experiment 2) of the diploids XW290 and XW285 in liquid YEPD (supplemented with D-glucosamine as appropriate) were centrifuged and washed with about 150 ml of distilled water. The cells were transferred to weighed centrifuge tubes, washed with 30 ml of distilled water, and weighed. Both strains yielded about 4.5 g of wet cells per 300-ml culture. Each batch of cells was suspended in 10 ml of buffer (0.15 M potassium phosphate plus 0.0025 M sodium ethylenediaminetetraacetic acid, pH 7.0) and passed twice through an Aminco French pressure cell at 20,000 lb/in2. The suspensions was then centrifuged at 20,000 x g for 20 min, the sediment was washed with 2 ml of buffer, and the material extracted from the sediment was pooled with the supernatant fraction. Four volumes of Na2SO4 (saturated at 23 C) was added slowly and with constant gentle agitation to an aliquot of the supernatant that had been warmed to about 23 C. Insoluble material appeared immediately and was removed by centrifugation at 20,000 x g for 20 min. The sediment was dissolved in a volume of buffer equal to the volume of the initial sample
VOL. 124, 1975
subjected to Na2SO4 precipitation. Aliquots of the several fractions were assayed for amidotransferase activity at 30 C (8). The assay mixture contained 15 smol of D-fructose 6-phosphate (Calbiochem, converted from the barium salt to the potassium salt by passage through Dowex 50), 37.5 umol of potassium phosphate (pH 7.0), 2.5 Mmol of sodium ethylenediaminetetraacetic acid (pH 7.0), and cell extract or buffer to a final volume of 1.0 ml. Reactions were started by the addition of cell extract and were stopped by heating the reaction at 100 C for 3 min. Reactions were stopped immediately after the addition of extract to measure the background of D-glucosamine 6-phosphate and other interfering materials present at zero time. The heated solutions were cooled to room temperature and centrifuged to remove coagulated protein. To an aliquot of the supernatant fluid (diluted when necessary to 0.5 ml with distilled water) was added 0.1 ml of saturated NaHCO, solution. Freshly prepared acetic anhydride solution (0.1 ml of a cold 5% aqueous solution) was added and the tubes were shaken vigorously, left at room temperature for 3 min, and then heated at 100 C for 3 min. The tubes were cooled at room temperature, 0.2 ml of saturated sodium tetraborate was added, and the tubes were again heated at 100 C for 3 min. To the cooled solutions was added 6 ml of freshly prepared p-dimethylaminobenzaldehyde solution (1 g of p-dimethylaminobenzaldehyde in 1 ml of concentrated HCl and 100 ml of glacial acetic acid). After 20 min at 37 C, the absorbance was determined at 585 nm in a Zeiss spectrophotometer using a reference in which water had been substituted for the extract-substrate mixture in the color reaction. The absorbance (585 nm) values were converted to Mmol of D-glucosamine by reference to a standard curve that was linear in the range 0 to 0.15 Amol. Ghosh et al. (8) found that D-glucosamine 6-phosphate gave 85% of the absorbance given by D-glucosamine, and this correction factor was used. Protein was measured by the method of Lowry et al. (15) using bovine serum albumin as a reference. Giemsa stain. Cells were fixed in water containing 4% formaldehyde for 24 h or longer. The fixed cells were washed once with water, treated with 1 N NaOH for 1.5 h at 23 C, and washed twice with 0.05 M potassium phosphate (pH 7 buffer). The cells were then applied to cover slips previously coated with a 1% solution of bovine serum albumin. After immersion for 35 min in the staining solution, a 1:10 dilution of commercial Gurr Giemsa stain in the phosphate buffer, the cells were destained for 3 to 6 in 95% ethanol and rinsed with water. Photographs of wet mounts of the stained cells were taken shortly after the water rinse. Electron microscopy. Yeast cells were fixed in 2% glutaraldehyde for 18 h at 23 C. They were rinsed twice in water, postfixed in 1% osmium tetroxide for 2 h at 23 C, and then rinsed and suspended in warm 1% The solidified agar was cut into 1-mm blocks that were dehydrated in ethanol and then passed through three changes of propylene oxide. The blocks were infiltrated with increasing concentrations of epoxy resin (Araldite 6005) for several days, concludagar.
1547
SPORULATION IN D-GLUCOSAMINE AUXOTROPHS
ing with the pure resin at 30 C under vacuum. Sections 50 to 80 nm thick were cut and mounted on carbon-stabilized Parlodion films. The sections were poststained in saturated uranyl acetate for 20 min at 60 C and with lead citrate for 5 min at 23 C. The sections were observed and photographed in a Siemens 1A electron microscope at 80 kV and electron optical magnification of 8,000 and 12,000 times. Some sections were also observed without the poststaining.
RESULTS of mutants. characterization Isolation and The mutant strains (Table 1) were found at a frequency of about 0.001 per survivor of the ethyl methane sulfonate mutagenesis. Two classes of mutants were observed (Table 1). The majority (class A) failed to grow on YEPD or MIN, but grew well if D-glucosamine was added to either. A higher D-glucosamine concentration was required in MIN than in YEPD. There was essentially no growth of the class A mutants on MIN containing 1 mg of D-glucosamine per ml, whereas on MIN containing 10 mg of D-glucosamine per ml the growth rate was indistinguishable from that of the wild-type strain. The remaining mutants (class B) showed an irregular growth pattern on YEPD, suggestive of a high reversion rate. Class B mutants also grew on MIN at varying rates depending on the strain, but without reversion on that medium in contrast to their behavior on YEPD. Thus, the TABLE 1. Growth response of wild-type and mutant strains Medium
YEPD + 1 mg of Dglucosamine/ml YEPD MIN MIN + 1 mg of D-glucos-
Class
Class
Wild
A mu-
tantsa
B mutants
type
+
+
+
_
t
+
+ + +
+
+
+
-
-
+
-
4id
+
C
id
4
amine/ml MIN + 10 mg of D-
glucosamine/ml YEPD + 1 mg of N-acet-
yl-D-glucosamine/ml MIN + 10 mg of N-acet-
yl-D-glucosamine/ml a This class includes mutants 1, 5, 7-10, 12-16, 18, 19, 21, 23, and 24. "This class includes mutants 2-4, 11, 17, 20, and 22. c Growth appears 2 to 3 days after replica plating as papillae on a nongrowing background. d Growth is visible 2 to 3 days after replica plating and varies in extent among the strains.
1548
J. BACTERIOL.
WHELAN AND BALLOU
class B mutants are conditional in that they require D-glucosamine on YEPD but not on MIN. None of the mutants was temperature sensitive. N-acetyl-D-glucosamine did not satisfy the requirement for D-glucosamine of either class of mutant in either medium. Allelism. None of the 23 heterozygotes made from the strains of Table 1 crossed to the wild type tequired D-gltlcosamine. Tetrad analysis establtshed that the Eglucosamine requirement iili each inip~tant was imposed by a single recessive miutation. The complementation map, de' rived from all pairwise combinations, is shown in Fig. 1. The 23 mutants constitute a set of alleles of one gene. We designate the wild-type allele GCNI and the mutant alleles gcnl-l through gcnl-24. For convenience, a strain is assigned an allele number (Fig. 1) identical to its isolation number (Table 1). Mutant 6 (allele gcnl-6) is iot included betause the heterdzygous diploid had a low spore viability. Compleinmenting diploids containing gcnl-13 gave coi-i fluent growth one day after replica platihig to YEPD, the behavior observed for Wild-type strains, whereas complementing diploids colitaining gcnl-22 grew more slowly but gave confluent growth 2 days after replica plating. As an additional test of allelism, the diploid from the cross of gcnl-l with gcnl-13 was analyzed. Reciprocal recombination was not observed in any of the 45 asci analyzed. Three asci contained a single prototrophic spore; the genotypes of the other spores in these asci indicated that the prototrophic spores arose by the gene conversion of gcnl-1. We have not located GCN1 in the genome. The data summarized in Table 2 establish that this gene is not centromere linked or linked to the other loci specified. Reversion and suppression. Spontaneous revertants capable of growth on YEPD were isolated from the class B mutants. Tetrad analysis of the cross of revertant with wild type revealed that the D-glucosamine requirement of class B mutants was suppressible. Mutation in a gene other than GCN1 accounted for the reversion of seven strains tested (Table 3). The suppressor sup(gcnl)1, first recognized by its suppression of gcnl-ll, suppressed the expression of all of the class B alleles but had no effect on the expression of the class A alleles (Table 3)
or the wild-type allele. The mutant allele sup(gcn)l is recessive to the wild-type allele
SUP(gcnl)l. We tested the possibility that this suppressor of class B alleles was a nonsense suppressor (9, 10, 16). The suppressor had no effect on the expression of the known nonsense alleles leu2-1, his4-4, trpl-1. As a further test, we obtained a strain bearing a nonsense suppressor by isolating a derivative that no longer required leucine or histidine from a strain beating ochre mutations leu2-1 and his4-4 and the amber mutation trpl-1. The nonsense suppressor in the derivative, shown to be ochre specific in a separate cross to wild type, had no effect on the phenotype of the class B alleles. Thus, the mutant alleles of GCN1, which retain ability to grow on one of the media used, are suppressible by a suppressor digtinct from the classical nonsense suppressors. The hypdthesis that class B mtitants are defective in a regulatory functioh in D-glucosa'mine synthesis would accoufit for this behavior. The suppressor might act by altering the structure of a defective enzyme to a functional state or by altering the levels of cbntrolling fihetabolites. Thus, these mutants are of potential value in studies of the control of D-glucosamine synthesis in yeasts. Enzyme defect in D-glucosainine auxotrophs. Amidotransferase activity was assayed in extracts of the wild-type strain XW290 grown in liquid YEPD and YEPD-GN. The activity of the extracts of the mutant strain XW285 (homoallelic for gcnl-1) was compared with the activities found in the wild type (Fig. 2 and Table 4). By necessity, XW285 was grown in YEPD-GN. The rate of D-glucosamine 6-phosphate formation during the first 15 min of incubation was essentially constant and was proportional to the amount of extract added. No TABLE 2. Linkage data for the gcnl locus Segregation pattern (no.) Gene pair
Parental ditypes
Nonparental ditypes
Tetratypes
gcnl-adel gcnl-mt
15 48 26 30
26 36 17 40
67 175 63 140
gcnl-ade2 gcnl-lys2
(B) 1, 2, 3, 4, 7, 8, 9, 10, 11, 12, 14, 17, 18, 20, 23, 24 (A) 13, 22 (C)5, 15, 16, 19, 21 FIG. 1. Complementation map of gcnl. (Complementation does not occur within any of the three groups A, B, and C. Complementation occurs between all alleles of the three different groups that are not on overlapping lines. Alleles in different groups and on overlapping lines, such as 5 and 22 or 1 and 19, fail to complement.
VOL. 124, 1975
SPORULATION IN D-GLUCOSAMINE AUXOTROPHS
1549
TABLE 3. Suppression of mutant alleles of GCN1 No. with genotype of diploid Allele no.
2 3 4 11 17
aa gcnl-i GCNI
sup(gcnl)la SUP(gcnl)l
Parental ditype
Nonparental
1 1 0 1 2
1 1 1 1
ditype
GCNI aa gcnl-i
SUP(gcnl)lb sup (gcnl)1
Parental
Nonparental ditype
Tetratype
2 5 5 3 4 3 3
1 4 0 2 2 4 2
8 7 8 5 6 4 6
Tetratype Ttayeditype 2 4
6
4
7 4
20 22
2
1
11
Class Ac
72
0
0
a Parental ditype asci have two auxotrophic and two prototrophic spores, nonparental ditype asci have four protrotrophic spores, and tetratype asci have one auxotrophic and three prototrophic spores. " Parental ditype asci have four prototrophic spores, nonparental ditype asci have two auxotrophic and two prototrophic spores, and tetratype asci have one auxotrophic and three prototrophic spores. Class A alleles 1, 5, 7-10, 12-15, 18, and 24 were tested, three to nine asci per cross. c
0.10
o
E0.05
z
0
2040
60-
Time (minutes)
FIG. 2. Time course of D-glucosamine 6-phosphate formation catalyzed by the Na2SO4-precipitate from the 20,000 x g supernatant extracts. Symbols: 0, strain XW290 grown without D-glucosamine; A, strain XW290 grown with D-glucosamine; and 0, strain XW285 grown with D-glucosamine.
product was formed in the absence of L-glutamine, D-fructose 6-phosphate, or cell extract. When strain XW290 was grown without the added D-glucosamine, amidotransferase activity was present in the crude supernatant extract derived from the broken cells but was absent from the 20,000 x g sediment. The addition of four volumes of saturated Na2SO, to 20,000 x g supernatant extract precipitated the enzyme. Recovery of activity was 95% (Table 4). When strain XW290 was grown in the presence of D-glucosamine, amidotransferase activity was difficult to measure in the crude supernatant fraction because this fraction contained material that was indistinguishable from D-glucosamine 6-phosphate in the colorimetric assay. However, treatment of the crude supernatant fraction with saturated Na2SO, (4 volumes) and
dissolution of the precipitate in fresh buffer yielded the active enzyme with a low background of interfering material. The activity recovered per cell and the specific activity of the Na2SO,-precipitated material were independent of the presence of D-glucosamine in the growth medium. Amidotransferase activity was not detected in extracts of the mutant diploid XW285 (Fig. 2, Table 4), including the Na,SO, precipitate and the sediment derived from the broken cells. The minimum detectable D-glucosamine level (0.04 ,smol) was about 6% of the amount formed by the wild-type strain. Sporulation. The diploid, class A D-glucosamine auxotrophs reveal a marked deficiency in sporulation ability, whereas the class B mutants gave a range of ascus frequencies (Table 5). The wild-type ascus frequencies varied from 0.16 to 0.55, presumably as a result of differences in the conditions of presporulation growth (5, 25). When parallel cultures were used as inocula, ascus frequencies differed by only a few percent. The class A mutants, except 23 and 24, gave ascus frequencies less than 0.001. After treatment with glusulase to liberate ascospores and destroy unsporulated cells, spores were not detected in the debris. The few spores isolated from the glusulase-treated suspensions of mutants 23 and 24 were not revertant. However, since spores were obtained singly or in incomplete two- or three-spored asci, it is not certain that alleles 23 and 24 permit sporulation without reversion because segregation of all markers could not be determined. Viable ascospores were obtained from diploids homozygous for class B alleles, and the
1550
J. BACTERIOL.
WHELAN AND BALLOU
TABLE 4. L-Glutamine:D-fructose 6-phosphate amidotransferase activity of mutant (XW285) and wild-type (XW290) strains XW290 with 1 mg of
XW285 with 1 mg of
0.0 2.4 0.0
0.0 3.4 0.0
0.0 2.9 0.0
0.46, 0.52 23.8, 25.6 0.019, 0.020
NDd ND ND
ND ND ND
0.43, 0.49 10.2, 11.9 0.042, 0.041
0.48, 0.42 8.9, 11.7
0.0, 0.0 7.1, 10.5 0.0, 0.0
XW290 without
Fraction
D-glucosamine/ml
D-glucosamine
20,000 x g sediment Activitya Protein"
Sp actc 20,000 x g supernatant Activity Protein Sp act Na2SO4 precipitate Activity Protein Sp act
D-glucosamine/ml
0.054, 0.035
Expressed as micromoles of D-glucosamine 6-phosphate formed per milliliter of extract in 15 min at 30 C. Protein content in milligrams per milliliter of extract. c Expressed as micromoles of D-glucosamine 6-phosphate formed per milligram of protein in 15 min at 30 C. d ND, Not determined. a "
TABLE 5. Sporulation of glucosamine auxotrophs Allele
Class A 1, 5, 7-10, 12-14, 16, 18, 19, 21 23 24
Spore
Ascus frequency
frequency
0 50 7
Vol. 124, No. 3
Printed in U.S.A.
Sporulation in -Glucosamine Auxotrophs of Saccharomyces cerevisiae: Meiosis with Defective Ascospore Wall Formation WILLIAM L. WHELAN
AND
CLINTON E. BALLOU*
Department of Biochemistry, University of California, Berkeley, California, 94720 Received for publication 16 July 1975
Mutants that require exogenous D-glucosamine for growth were isolated from Saccharomyces cerevisiae X2180-1A after ethyl methane sulfonate mutagenesis. Class A auxotrophs fail to grow on yeast extract-peptone-dextrose and minimal media, whereas class B auxotrophs grow on minimal medium and readily revert to grow on yeast extract-peptone-dextrose medium. Class B auxotrophs are suppressible by a class of suppressors distinct from nonsense suppressors, and their properties suggest that they are defective in a regulatory function. All 23 mutants studied were recessive and allelic, and they define a new gene designated gcnl. An analysis of a class A auxotroph revealed that it lacked L-glutamine:D-fructose 6-phosphate amidotransferase (EC 2.6.1.16) activity and indicates that GCN1 codes the amino acid sequence of this enzyme. The finding that all mutants were allelic indicates that the amidotransferase is the only enzyme responsible for D-glucosamine synthesis in S. cerevisiae. The occurrence of allelic complementation and media-conditional mutants suggests that the amidotransferase is a multimeric enzyme with an activity subject to metabolic control. Diploids homozygous for gcnl fail to complete sporulation. They proceed through meiosis normally, as judged by the occurrence of meiotic recombination, the production of haploid nuclei, and the formation of multinucleate cells visible after Giemsa staining. However, the formation of glusulase-resistant ascospores is blocked, and deformed spores lacking the electron-dense outer layer characteristic of the normal spore wall are observed by electron microscopy. Cells that acquire the ability to synthesize D-glucosamine, because of gene conversion during meiosis, complete sporulation in a normal fashion. Thus, the GCN1 gene product appears to be synthesized late in sporulation and may prove to be a useful developmental landmark for the termination of ascospore development.
D-Glucosamine (as N-acetyl-D-glucosamine) occurs in two major components of the cell wall of the budding yeast Saccharomyces cerevisiae. It is present as chitin in the bud scars that are formed on separation of the mother and daughter cells (3, 4), and it also occurs in the mannan of the vegetative cell wall, where it links polysaccharide side chains to the protein backbone (27). We demonstrate here that D-glucosamine is essential for the growth of S. cerevisiae because mutants can be isolated that fail to grow in its absence. In addition, we find that the synthesis of this hexosamine is determined by a single gene (denoted GCN1), which probably codes the primary structure of the enzyme L-glutamine:D-fructose 6-phosphate amidotransferase (EC 2.6.1.16). Many of the D-glucosamine auxotrophs that we have isolated are defective in the ability to complete sporulation; they undergo meiosis and
produce haploid nuclei within deformed and glusulase-sensitive ascospores. Thus, we describe a novel type of sporulation-defective mutant differing from those previously described that were defective in the premeiotic round of deoxyribonucleic acid (DNA) synthesis (24) or in the meiotic divisions (5, 7).
MATERIALS AND METHODS Yeast strains. The isogenic wild-type strains S. cerevisiae X2180-1A and X2180-1B were provided by R. K. Mortimer. A strain bearing the alleles leu2-1, his4-4, and trp-1 was provided by S. Fogel. Media. The stated amounts of the following ingredients were dissolved in 1 liter of distilled water. Yeast extract-peptone-dextrose (YEPD) contained 10 g of Difco yeast extract, 20 g of Difco peptone, and 20 g of D-glucose. Minimal medium (MIN) contained 6.7 g of Difco yeast nitrogen base without amino acids and 20 g of D-glucose. Difco agar (20 g) was added when a solid medium was desired. When appropriate, D1545
1546
WHELAN AND BALLOU
glucosamine (Calbiochem A grade) or N-acetyl-Dglucosamine (Sigma Chemical Co.) was added to the autoclaved medium as a filter-sterilized solution. Sporulation of strains for tetrad analysis was performed on sporulation (SPOR) agar made from 2.5 g of yeast extract, 10 g of potassium acetate, and 20 g of agar per liter of distilled water. Incubations were at 30C. General genetic techniques. Crosses were performed by mixing 18-h cultures of the parent strains on YEPD containing 1 mg of D-glucosamine per ml (YEPD-GN). Diploid clones were obtained by isolating zygotes 3 to 6 h after mixing the haploids. Sporulation was induced in the mating mixture by replica plating the mixture to SPOR agar after incubation for 18 h or in an isolated diploid clone by replica plating after incubation for 2 to 3 days. Asci were observed after a 3-day incubation of the culture on SPOR. Tetrad analysis was performed by using glusulase (Endo Laboratories) to digest the ascus wall. Complementation was determined among Dglucosamine auxotrophs by mixing 18-h cultures of opposite mating type on YEPD-GN and replica plating the mating mixture to YEPD after overnight incubation. The YEPD replicas were incubated at 30 C for 7 days and were observed periodically for the appearance of growth. A spot test for allelic recombination was performed by replica plating a diploid culture to SPOR agar and then replica plating this culture, after a 3-day incubation, to the test medium on which only recombinants could grow. The mating type of isolated clones was determined by examining appropriate test crosses of the clone for the presence of zygotes 3 to 6 after mixing. Mutagenesis and mutant isolation. An 18-h YEPD culture of strain X2180-1A grown at 30 C was washed once with distilled water and suspended at a cell density of 5 x 107 cells/ml in 0.1 M potassium phosphate at pH 7.0. To 1 ml of this suspension was added 25 ,l of ethyl methane sulfonate (Eastman). After 30 min at room temperature (23 C) with continuous agitation, the suspension was diluted 1:100 in 5% sodium thiosulfate and then 1:100 in distilled water. The diluted suspension was spread on YEPD-GN, the plate was incubated at 30 C for 2 days, and the colonies were replica plated to YEPD. Colonies incapable of growth on YEPD at 30 C were picked from the master plate. About 50% of the cells survived the mutagenesis. Spontaneous revertants were picked from a YEPD replica of the strain grown on YEPD-GN. Ultravioletinduced revertants were obtained by irradiating a YEPD-GN culture with a General Electric germicidal lamp to yield about 30% survival and were selected by replica plating the cultures to YEPD after overnight incubation. Assessment of sporulation. Overnight cultures in liquid YEPD-GN were washed once with distilled water and suspended in 1% potassium acetate at a cell density of 5 x 107 cells/ml. The acetate cultures were shaken at room temperature and examined periodically during 7 days for ascus frequency, viable count, recombinant frequency, and frequency of haploid formation. The ascus frequency was determined by
J. BACTRIOL. the microscopy count of all cells in a hemocytometer grid (cells, visible buds, and asci were counted as cells), followed by a count of the asci in the same grid. No distinction was made between two-, three- and four-spored asci. Ascus frequencies in the wild-type strains were based on the examination of 100 to 200 cells, whereas ascus frequencies in the mutant strains were based on the examination of more than 500 cells. Spore frequencies were determined by treating the washed SPOR culture with glusulase (a 1:10 dilution of the commercial preparation) and examining the treated suspension for the presence of ascospores. Frequencies are expressed as spores per 1,000 cells, based on the examination of the debris from at least 2,000 cells. Viable counts were obtained from YEPDGN agar plates; the viability did not decline in the strains during the 7-day incubation. Prototrophic recombinants were detected on YEPD or MIN, as appropriate, and recombinant frequency is expressed as recombinants per viable cell. Assessment of haploid formation. Cells from diploid cultures previously incubated in SPOR medium for 7 days were spread on YEPD-GN to yield 100 to 200 colonies per plate. After a 2- to 3-day incubation, each plate was replica plated to a lawn of the his6 tester strain on YEPD-GN (a and a mating type on separate plates). After overnight incubation, the plates containing the testers and colony replicas were replica plated to MIN. The presence of cells capable of mating in the original colonies was revealed by growth on MIN in the area of contact of the colony and the tester. The plates were scored one day after replica plating to MIN for the presence of haploids in the colonies. Incubation for 2 to 4 days revealed the presence of a few nonhaploid cells capable of mating in colonies that were scored as negative the first day. For the determination of the genotype of the mater, the growth on MIN was streaked on MIN to obtain isolated colonies. These were sporulated and subjected to tetrad analysis. Assay for L-glutamine:D-fructose 6-phosphate amidotransferase. Overnight cultures (17 h in experiment 1 and 24 h in experiment 2) of the diploids XW290 and XW285 in liquid YEPD (supplemented with D-glucosamine as appropriate) were centrifuged and washed with about 150 ml of distilled water. The cells were transferred to weighed centrifuge tubes, washed with 30 ml of distilled water, and weighed. Both strains yielded about 4.5 g of wet cells per 300-ml culture. Each batch of cells was suspended in 10 ml of buffer (0.15 M potassium phosphate plus 0.0025 M sodium ethylenediaminetetraacetic acid, pH 7.0) and passed twice through an Aminco French pressure cell at 20,000 lb/in2. The suspensions was then centrifuged at 20,000 x g for 20 min, the sediment was washed with 2 ml of buffer, and the material extracted from the sediment was pooled with the supernatant fraction. Four volumes of Na2SO4 (saturated at 23 C) was added slowly and with constant gentle agitation to an aliquot of the supernatant that had been warmed to about 23 C. Insoluble material appeared immediately and was removed by centrifugation at 20,000 x g for 20 min. The sediment was dissolved in a volume of buffer equal to the volume of the initial sample
VOL. 124, 1975
subjected to Na2SO4 precipitation. Aliquots of the several fractions were assayed for amidotransferase activity at 30 C (8). The assay mixture contained 15 smol of D-fructose 6-phosphate (Calbiochem, converted from the barium salt to the potassium salt by passage through Dowex 50), 37.5 umol of potassium phosphate (pH 7.0), 2.5 Mmol of sodium ethylenediaminetetraacetic acid (pH 7.0), and cell extract or buffer to a final volume of 1.0 ml. Reactions were started by the addition of cell extract and were stopped by heating the reaction at 100 C for 3 min. Reactions were stopped immediately after the addition of extract to measure the background of D-glucosamine 6-phosphate and other interfering materials present at zero time. The heated solutions were cooled to room temperature and centrifuged to remove coagulated protein. To an aliquot of the supernatant fluid (diluted when necessary to 0.5 ml with distilled water) was added 0.1 ml of saturated NaHCO, solution. Freshly prepared acetic anhydride solution (0.1 ml of a cold 5% aqueous solution) was added and the tubes were shaken vigorously, left at room temperature for 3 min, and then heated at 100 C for 3 min. The tubes were cooled at room temperature, 0.2 ml of saturated sodium tetraborate was added, and the tubes were again heated at 100 C for 3 min. To the cooled solutions was added 6 ml of freshly prepared p-dimethylaminobenzaldehyde solution (1 g of p-dimethylaminobenzaldehyde in 1 ml of concentrated HCl and 100 ml of glacial acetic acid). After 20 min at 37 C, the absorbance was determined at 585 nm in a Zeiss spectrophotometer using a reference in which water had been substituted for the extract-substrate mixture in the color reaction. The absorbance (585 nm) values were converted to Mmol of D-glucosamine by reference to a standard curve that was linear in the range 0 to 0.15 Amol. Ghosh et al. (8) found that D-glucosamine 6-phosphate gave 85% of the absorbance given by D-glucosamine, and this correction factor was used. Protein was measured by the method of Lowry et al. (15) using bovine serum albumin as a reference. Giemsa stain. Cells were fixed in water containing 4% formaldehyde for 24 h or longer. The fixed cells were washed once with water, treated with 1 N NaOH for 1.5 h at 23 C, and washed twice with 0.05 M potassium phosphate (pH 7 buffer). The cells were then applied to cover slips previously coated with a 1% solution of bovine serum albumin. After immersion for 35 min in the staining solution, a 1:10 dilution of commercial Gurr Giemsa stain in the phosphate buffer, the cells were destained for 3 to 6 in 95% ethanol and rinsed with water. Photographs of wet mounts of the stained cells were taken shortly after the water rinse. Electron microscopy. Yeast cells were fixed in 2% glutaraldehyde for 18 h at 23 C. They were rinsed twice in water, postfixed in 1% osmium tetroxide for 2 h at 23 C, and then rinsed and suspended in warm 1% The solidified agar was cut into 1-mm blocks that were dehydrated in ethanol and then passed through three changes of propylene oxide. The blocks were infiltrated with increasing concentrations of epoxy resin (Araldite 6005) for several days, concludagar.
1547
SPORULATION IN D-GLUCOSAMINE AUXOTROPHS
ing with the pure resin at 30 C under vacuum. Sections 50 to 80 nm thick were cut and mounted on carbon-stabilized Parlodion films. The sections were poststained in saturated uranyl acetate for 20 min at 60 C and with lead citrate for 5 min at 23 C. The sections were observed and photographed in a Siemens 1A electron microscope at 80 kV and electron optical magnification of 8,000 and 12,000 times. Some sections were also observed without the poststaining.
RESULTS of mutants. characterization Isolation and The mutant strains (Table 1) were found at a frequency of about 0.001 per survivor of the ethyl methane sulfonate mutagenesis. Two classes of mutants were observed (Table 1). The majority (class A) failed to grow on YEPD or MIN, but grew well if D-glucosamine was added to either. A higher D-glucosamine concentration was required in MIN than in YEPD. There was essentially no growth of the class A mutants on MIN containing 1 mg of D-glucosamine per ml, whereas on MIN containing 10 mg of D-glucosamine per ml the growth rate was indistinguishable from that of the wild-type strain. The remaining mutants (class B) showed an irregular growth pattern on YEPD, suggestive of a high reversion rate. Class B mutants also grew on MIN at varying rates depending on the strain, but without reversion on that medium in contrast to their behavior on YEPD. Thus, the TABLE 1. Growth response of wild-type and mutant strains Medium
YEPD + 1 mg of Dglucosamine/ml YEPD MIN MIN + 1 mg of D-glucos-
Class
Class
Wild
A mu-
tantsa
B mutants
type
+
+
+
_
t
+
+ + +
+
+
+
-
-
+
-
4id
+
C
id
4
amine/ml MIN + 10 mg of D-
glucosamine/ml YEPD + 1 mg of N-acet-
yl-D-glucosamine/ml MIN + 10 mg of N-acet-
yl-D-glucosamine/ml a This class includes mutants 1, 5, 7-10, 12-16, 18, 19, 21, 23, and 24. "This class includes mutants 2-4, 11, 17, 20, and 22. c Growth appears 2 to 3 days after replica plating as papillae on a nongrowing background. d Growth is visible 2 to 3 days after replica plating and varies in extent among the strains.
1548
J. BACTERIOL.
WHELAN AND BALLOU
class B mutants are conditional in that they require D-glucosamine on YEPD but not on MIN. None of the mutants was temperature sensitive. N-acetyl-D-glucosamine did not satisfy the requirement for D-glucosamine of either class of mutant in either medium. Allelism. None of the 23 heterozygotes made from the strains of Table 1 crossed to the wild type tequired D-gltlcosamine. Tetrad analysis establtshed that the Eglucosamine requirement iili each inip~tant was imposed by a single recessive miutation. The complementation map, de' rived from all pairwise combinations, is shown in Fig. 1. The 23 mutants constitute a set of alleles of one gene. We designate the wild-type allele GCNI and the mutant alleles gcnl-l through gcnl-24. For convenience, a strain is assigned an allele number (Fig. 1) identical to its isolation number (Table 1). Mutant 6 (allele gcnl-6) is iot included betause the heterdzygous diploid had a low spore viability. Compleinmenting diploids containing gcnl-13 gave coi-i fluent growth one day after replica platihig to YEPD, the behavior observed for Wild-type strains, whereas complementing diploids colitaining gcnl-22 grew more slowly but gave confluent growth 2 days after replica plating. As an additional test of allelism, the diploid from the cross of gcnl-l with gcnl-13 was analyzed. Reciprocal recombination was not observed in any of the 45 asci analyzed. Three asci contained a single prototrophic spore; the genotypes of the other spores in these asci indicated that the prototrophic spores arose by the gene conversion of gcnl-1. We have not located GCN1 in the genome. The data summarized in Table 2 establish that this gene is not centromere linked or linked to the other loci specified. Reversion and suppression. Spontaneous revertants capable of growth on YEPD were isolated from the class B mutants. Tetrad analysis of the cross of revertant with wild type revealed that the D-glucosamine requirement of class B mutants was suppressible. Mutation in a gene other than GCN1 accounted for the reversion of seven strains tested (Table 3). The suppressor sup(gcnl)1, first recognized by its suppression of gcnl-ll, suppressed the expression of all of the class B alleles but had no effect on the expression of the class A alleles (Table 3)
or the wild-type allele. The mutant allele sup(gcn)l is recessive to the wild-type allele
SUP(gcnl)l. We tested the possibility that this suppressor of class B alleles was a nonsense suppressor (9, 10, 16). The suppressor had no effect on the expression of the known nonsense alleles leu2-1, his4-4, trpl-1. As a further test, we obtained a strain bearing a nonsense suppressor by isolating a derivative that no longer required leucine or histidine from a strain beating ochre mutations leu2-1 and his4-4 and the amber mutation trpl-1. The nonsense suppressor in the derivative, shown to be ochre specific in a separate cross to wild type, had no effect on the phenotype of the class B alleles. Thus, the mutant alleles of GCN1, which retain ability to grow on one of the media used, are suppressible by a suppressor digtinct from the classical nonsense suppressors. The hypdthesis that class B mtitants are defective in a regulatory functioh in D-glucosa'mine synthesis would accoufit for this behavior. The suppressor might act by altering the structure of a defective enzyme to a functional state or by altering the levels of cbntrolling fihetabolites. Thus, these mutants are of potential value in studies of the control of D-glucosamine synthesis in yeasts. Enzyme defect in D-glucosainine auxotrophs. Amidotransferase activity was assayed in extracts of the wild-type strain XW290 grown in liquid YEPD and YEPD-GN. The activity of the extracts of the mutant strain XW285 (homoallelic for gcnl-1) was compared with the activities found in the wild type (Fig. 2 and Table 4). By necessity, XW285 was grown in YEPD-GN. The rate of D-glucosamine 6-phosphate formation during the first 15 min of incubation was essentially constant and was proportional to the amount of extract added. No TABLE 2. Linkage data for the gcnl locus Segregation pattern (no.) Gene pair
Parental ditypes
Nonparental ditypes
Tetratypes
gcnl-adel gcnl-mt
15 48 26 30
26 36 17 40
67 175 63 140
gcnl-ade2 gcnl-lys2
(B) 1, 2, 3, 4, 7, 8, 9, 10, 11, 12, 14, 17, 18, 20, 23, 24 (A) 13, 22 (C)5, 15, 16, 19, 21 FIG. 1. Complementation map of gcnl. (Complementation does not occur within any of the three groups A, B, and C. Complementation occurs between all alleles of the three different groups that are not on overlapping lines. Alleles in different groups and on overlapping lines, such as 5 and 22 or 1 and 19, fail to complement.
VOL. 124, 1975
SPORULATION IN D-GLUCOSAMINE AUXOTROPHS
1549
TABLE 3. Suppression of mutant alleles of GCN1 No. with genotype of diploid Allele no.
2 3 4 11 17
aa gcnl-i GCNI
sup(gcnl)la SUP(gcnl)l
Parental ditype
Nonparental
1 1 0 1 2
1 1 1 1
ditype
GCNI aa gcnl-i
SUP(gcnl)lb sup (gcnl)1
Parental
Nonparental ditype
Tetratype
2 5 5 3 4 3 3
1 4 0 2 2 4 2
8 7 8 5 6 4 6
Tetratype Ttayeditype 2 4
6
4
7 4
20 22
2
1
11
Class Ac
72
0
0
a Parental ditype asci have two auxotrophic and two prototrophic spores, nonparental ditype asci have four protrotrophic spores, and tetratype asci have one auxotrophic and three prototrophic spores. " Parental ditype asci have four prototrophic spores, nonparental ditype asci have two auxotrophic and two prototrophic spores, and tetratype asci have one auxotrophic and three prototrophic spores. Class A alleles 1, 5, 7-10, 12-15, 18, and 24 were tested, three to nine asci per cross. c
0.10
o
E0.05
z
0
2040
60-
Time (minutes)
FIG. 2. Time course of D-glucosamine 6-phosphate formation catalyzed by the Na2SO4-precipitate from the 20,000 x g supernatant extracts. Symbols: 0, strain XW290 grown without D-glucosamine; A, strain XW290 grown with D-glucosamine; and 0, strain XW285 grown with D-glucosamine.
product was formed in the absence of L-glutamine, D-fructose 6-phosphate, or cell extract. When strain XW290 was grown without the added D-glucosamine, amidotransferase activity was present in the crude supernatant extract derived from the broken cells but was absent from the 20,000 x g sediment. The addition of four volumes of saturated Na2SO, to 20,000 x g supernatant extract precipitated the enzyme. Recovery of activity was 95% (Table 4). When strain XW290 was grown in the presence of D-glucosamine, amidotransferase activity was difficult to measure in the crude supernatant fraction because this fraction contained material that was indistinguishable from D-glucosamine 6-phosphate in the colorimetric assay. However, treatment of the crude supernatant fraction with saturated Na2SO, (4 volumes) and
dissolution of the precipitate in fresh buffer yielded the active enzyme with a low background of interfering material. The activity recovered per cell and the specific activity of the Na2SO,-precipitated material were independent of the presence of D-glucosamine in the growth medium. Amidotransferase activity was not detected in extracts of the mutant diploid XW285 (Fig. 2, Table 4), including the Na,SO, precipitate and the sediment derived from the broken cells. The minimum detectable D-glucosamine level (0.04 ,smol) was about 6% of the amount formed by the wild-type strain. Sporulation. The diploid, class A D-glucosamine auxotrophs reveal a marked deficiency in sporulation ability, whereas the class B mutants gave a range of ascus frequencies (Table 5). The wild-type ascus frequencies varied from 0.16 to 0.55, presumably as a result of differences in the conditions of presporulation growth (5, 25). When parallel cultures were used as inocula, ascus frequencies differed by only a few percent. The class A mutants, except 23 and 24, gave ascus frequencies less than 0.001. After treatment with glusulase to liberate ascospores and destroy unsporulated cells, spores were not detected in the debris. The few spores isolated from the glusulase-treated suspensions of mutants 23 and 24 were not revertant. However, since spores were obtained singly or in incomplete two- or three-spored asci, it is not certain that alleles 23 and 24 permit sporulation without reversion because segregation of all markers could not be determined. Viable ascospores were obtained from diploids homozygous for class B alleles, and the
1550
J. BACTERIOL.
WHELAN AND BALLOU
TABLE 4. L-Glutamine:D-fructose 6-phosphate amidotransferase activity of mutant (XW285) and wild-type (XW290) strains XW290 with 1 mg of
XW285 with 1 mg of
0.0 2.4 0.0
0.0 3.4 0.0
0.0 2.9 0.0
0.46, 0.52 23.8, 25.6 0.019, 0.020
NDd ND ND
ND ND ND
0.43, 0.49 10.2, 11.9 0.042, 0.041
0.48, 0.42 8.9, 11.7
0.0, 0.0 7.1, 10.5 0.0, 0.0
XW290 without
Fraction
D-glucosamine/ml
D-glucosamine
20,000 x g sediment Activitya Protein"
Sp actc 20,000 x g supernatant Activity Protein Sp act Na2SO4 precipitate Activity Protein Sp act
D-glucosamine/ml
0.054, 0.035
Expressed as micromoles of D-glucosamine 6-phosphate formed per milliliter of extract in 15 min at 30 C. Protein content in milligrams per milliliter of extract. c Expressed as micromoles of D-glucosamine 6-phosphate formed per milligram of protein in 15 min at 30 C. d ND, Not determined. a "
TABLE 5. Sporulation of glucosamine auxotrophs Allele
Class A 1, 5, 7-10, 12-14, 16, 18, 19, 21 23 24
Spore
Ascus frequency
frequency
0 50 7
