Curr Genet (1987)11:429-434 ~ ~ ~

© Springer-Verlag 1987

Mutations suppressing the effects of a deletion of the Phosphoglucose isomerase gene PGI1 in Saccharomyces cerevisiae

Andres Aguilera

Institut flit Mikrobiologie (Genetik), Technische Hochschule Darmstadt, Schnittspahnstrasse 10, D-6100 Darmstadt, Federal Republic of Germany

Summary. A mutant with a deletion covering the phos- phoglucose isomerase gene PGI1, allele pgilA, can only grow on a medium containing fructose and low concen- trations of glucose whereas growth is completely inhi. bited by glucose concentrations higher than 0.4%. This was used to select suppressor mutants restoring growth on synthetic media with 2% glucose as the sole carbon source. One complementation group, SPG1, was defined by recessive mutations. The ability to grow on glucose media was strictly dependent on functional mitochon- dria. The generation time of the selected mutants on YEP glucose was 6 - 8 h. No ethanol was formed from glucose and the levels of respiration were very high. These phenotypes were also observed in single pgilA mutants when growing on fructose media supplemented with 0.4% glucose. The other glycolytic enzymes, the enzy- mes of the glucose-6-phosphate oxidation pathway as well as catabolite repression were normal in suppressed pgilA mutants. The suppressor mutation alone caused no abnormal phenotype. The results suggest that the spgl suppressor mutations allow S. cerevisiae pgilA mutant strains to grow on glucose by using the Pentose-P cycle in combination with unusual strong respiration.

Key words: Phosphoglucose isomerase - pgilA suppres- sor mutations - Glucose-6-phosphate - Pentose phos- phate pathway - Saccaromyces cerevisiae

Introduction

Hexoses are metabolized in Saccharomyces cerevisiae via glycolysis with a concomitant production of ATP, NADH and intermediate metabolites necessary for cell

Present address: Department of Biochemistry, NYU Medical Center, 550 First Avenue, New York, NY 10016, USA

growth. Glucose is transported into the yeast cell by an as yet unidentified mechanism involving presumably two carriers and the hexokinase reactions to produce glucose-6-phosphate (Bisson and Fraenkel 1983, 1984). Phosphoglucose isomerase is the enyzme that converts glucose-6-P into fructose-6-P (for review, see Fraenkel 1982). The gene coding for phosphoglucose isomerase

PGI1 has been cloned (Aguilera and Zimmermann 1986). Strains with a deletion in this gene have been shown to be unable to grow either on glucose or on fructose as sole carbon source (Aguilera 1986), in contrast to the results obtained with the same type of mutants of Escherichia coli which are able to grow on both sugars (Vinopal et al. 1975).

Glucose-6-phosphate is required for the biosynthesis of inositol of membrane phospholipids (Henry 1982) and glucans of the cell wall (Ballou 1982) and thus is an essential metabolite inyeasts (Aguilera 1986). A small part of glucose-6-phosphate is metabolized by the glu- cose-6-phosphate dehydrogenase reaction. This reaction, required for the production of NADPH necessary for biosynthetic pathways, allows the conversion of hexo- ses to intermediates of the pentose-phosphate cycle. Part of the products of this cycle could reenter glyco- lysis at the level of furctose-6-phosphate and glyceral- dehyde-3-phosphate, but in yeasts this flux is not suffi- cient to support growth as shown by the inability of a pgilA deletion mutant to grow on glucose (Aguilera 1986).

Mutants in PGI1 are known to be glucose sensitive, but whether the cause for this sensitivity is glucose-6- phosphate accumulation (Maitra 1971) or ATP depletion (Ciriacy and Breitenbach 1979) is still unclear. In order to study this phenomenon a deletion mutant lacking most of the PG11 structural gene and unable to revert has been used to select for mutants resistant to glucose. Extragenic mutations suppressing the inability of the

430 A. Aguilera: pgil~ suppressor mutants in S. cerevisiae

pgi lA mutant to grow on glucose were isolated. They

were able to consume all the glucose from the media without production of ethanol and showed unusually high

respiration rates. The same phenotype was observed in

a single pgi12x deletion strain growing the 2% fructose and 0.4% glucose.

Materials and methods

Strains. The following strains of Saccharomyees cerevisiae were used: UTL-7A (MATa leu2-3,112 trpl ura3-52), DMA4 (MATa pgil 1eu2-3,112 trpl ura3-52), SS-4A (MATa adel SUC4), SS- l lA (MATa adel SUC4), AA28 (MATa leu2-3,112 ura3-52 trpl pgilA25:LEU2), AA30 (MATa GAL2 pgi12x25:LEU2 ura3), AA31 (MATa GAL2 pgilLx34:LEU2 his4 ura3), AA33 (MA Ta ura3 ade2 pgil A 34: LEU2 horn3), AA34 (MA Ta leu2-3, 112 ura3 ade2 pgil Lx34:LEU2) and XW5-1B (MA Ta ura3 kar2).

Media and growth conditions. YEP media containing 2% peptone (Difco, Detroit, Mich, USA), 1% yeast extract (Oxid, Basing- stoke, England) supplemented with 2% glucose (YEPD), 2% fructose + 0.1% glucose (YEPFd), 2% fructose + 2% glucose + 2% glycerol + 0.1% galactose (YEPFDGg) or 2% glycerol and 2% ethanol (YEPGE) as carbon sources were used as rich media. The synthetic medium was based on 1.7 g/1 Difco yeast nitrogen base without amino acids and ammonium sulphate, and was supplemented with 5 g/l (NH4)2SO 4, amino acids, adenine and uracil as required and 2% glucose (SD), 2% fructose + 0.1% glucose (SFd) or 2% fructose + 2% glucose + 2% glycerol and 0.1% galactose (SFDGg) as carbon sources. Media were solidi- fied with 15 g/1 agar (Oxoid). Liquid cultures were grown on a shaker at 28 °C. Sporulation was induced on plates with a potas- sium acetate medium supplemented with 0.05% glucose.

0 2 consumption and CO 2 production. They were determined manometrically in a standard Warburg respirometer under aero- bic conditions at 30 °C (Aguilera and Benitez 1985).

Ethanol, glucose and fructose determinations. Ethanol deter- minations were carried out using the kits from Boehringer (Mann- heim, FRG) by measuring the NADH formation at 340 nm in the oxidation of ethanol to acetic acid with alcohol dehydroge- nase and acetaldehyde dehydogenase. Fructose was determined in the media by measuring the NADPH formation at 340 nm using a coupled hexokinase - phosphoglucose isomerase - glu- cose-6-phosphate reaction whereas glucose was determined with the same reaction mix without the addition of phosPhoglucose isomerase. Glucose was also determined using the chemical method of glucose-oxidase/peroxidase reaction (Goldstein and Lampen 1975).

Preparation of crude extracts and enyzme assays. Crude extracts were prepared as described by Ciriacy and Breitenbach (1979) using ballotini glass beads (Sigma, St. Louis, Mo, USA) for breaking the cells. Enzymatic assays were performed according to Ciriacy and Breitenbach (1979), Entian and Zimmermann (1980) and Bergmeyer (1970). Specific activities were computed as mU based on protein determination according to Zahmenhoff (1957).

Assays of internal metabolities. Metabolites were extracted by the method of Ciriacy and Breitenbach (1979) from exponen- tial cultues in synthetic fructose media 2 h after the addition of

glucose to a final concentration of 2% and determined spectro- photometrically according to Bergmeyer (1970). Dry weights were determined by absorption measurements at 600 nm using a standard curve.

Mutagenesis and mutant selection, pgil Lx deletion mutants were grown overnight in YEP 2% Fructose-0.1% Glucose medium and mutagenized at 28 °C at a titer of 1 x 107 eells/ml in 0.1 M potassium phosphate buffer pH 7.0 containing 1% ethyl methane sulfonate (Serva, Heidelberg, FRG). After 1 h the cells were washed three times with buffer and incubated in YEP glycerol- ethanol for 8 h. Cells were then plated on synthetic media with 2% fructose or 2% fructose-2% glycerol with or without antirny- cin A. Colonies appearing were tested for growth on different carbon source supplemented media and those able to grow on SD in 4 days were selected for biochemical and genetical ana- lysis.

Genetical analysis. It was performed by published procedures (Sherman et al. 1979). Complementation grouPswere established using three criteria: growth in YEP glucose liquid media, growth on solid glucose synthetic media and genetic segregation analysis of the diploids formed by crossing the mutants with spore pro- geny derived from crosses of other mutants.

Results

Isolation and genetical characterization o f mutants suppresing the pgil A mutation

The pig lA deletion strain AA28 lacking most of the

structural gene coding for the phosphoglucose isomerase

(Aguilera 1986) was mutagenized with EMS 1% to get a

survival frequency between 10 and 50%. After incubation

under permissive growth conditions in YEP 2% fruc-

tose-0.1% glucose medium, cells were plated on selective

media YEP 2% fructose-2% glucose. Colonies growing

in spite of the high glucose concentration in the media

appeared after 4 days at frequencies of around to 10 -9 .

Twelve mutants isolated were crossed with the wild type PGII and tetrad analysis was performed in order to

study whether some of them were affectd in glucose

transport, as a possible cause of glucose-resistance. In this case PGI1 spores with a much reduced growth on glucose media should be recovered. But in all the cases

a 2 + : 2 - segregation for strong growth on glucose media was obtained, suggesting that these mutants were not affected in glucose uptake. On the contrary, after checking for growth on several media, it was found that most of the pgi lA suppressor double mutants were able to grow on 2% glucose as sole carbon source.

Further mutagenic treatments were performed with pgi lA deletion strains AA30 and AA31. Resistant colo- nies appeared after 6 days only on a synthetic medium

with 2% fructose and 2% glucose at a frequency of 1 0 - s - 1 0 -9. No mutants able to grow on fructose as

sole carbon source could be found, suggesting that there is no way of recovering glucose-6-phosphate from fruc-

A. Aguilera: pgi lA suppressor mutants in S. cerevisiae 431

Table 1. Genetical analysis of some of the mutants obtained. CT: complete viable tetrads/number of tetrads micromanipulated; SD: number of tetrads with 2+:2- and 4+0 segregation on synthetic glucose media;random + : - ; segregation of growth on glucose media in random spore analysis experiments. D" growth of the diploid on glucose media

Crosses a CT SD Random D

4+:0 - 2+:2 _ +:_

spgl-4 x SPG1 7/35 - 7 96 ;131 - spgl-6 x SPG1 3/31 - 3 104 : 78 - -

spgl-4 x spgl-4 8/19 8 b 0 61:0 b + + spgl-4 x spgl-1 - - - 42:0 + + spgl-4 x spgl-2 - - - 92:6 + + spgl-1 x spgl-6 - - - 49 : 0 + +

a all the strains carried the pgi lA deletion b segregation based of RHO + spores

Table 2. Enyzmatic acitivity (mU/mg protein) of mutant and wild type strains. PGI: phosphoglucose isomerase; HXK: hexokinases; ZWF: glucose-6-phosphate dehydrogenase; GND: 6-phosphogluconate dehydrogenase; MDH" malate dehydrogenase; ADH: alcohol dehydrogenase; gly: cells grown in glycerol media, D: cells grown in glucose media

Strains a PGI HXK ZWF GND MDH gly MDH D ADH gly ADH D

PGI1 1,400 1,300 - 2,459 483 - -

pgil A25 nd 60 70 3,859 998 - - pgi lA34 nd - 58 82 3,315 572 1,336 497

pgi l / ,25 spg1-1 nd 1,612 68 45 2,458 592 - 422 pgi1A25 spgl-2 nd 1,433 72 47 4,450 592 1,144 309 pgi1A25 spgl-3 nd 1,232 63 45 3,552 761 - 375 pgi lA25 spgl-4 nd 1,464 74 37 4,820 483 1,665 339 pgi1A25 spgl-6 nd 1,626 95 44 2,717 551 - 440

a all the other mutants selected gave the same results

nd not detectable

tose-6-phosphate other than through the phosphoglu-

cose isomerase reaction. Also, no mutants were obtain-

ed on a medium with high glucose concentrations in the presence o f antimycin A which blocks respiration. More than 30% of these mutants were able to grow on a synthetic medium with glucose as sole carbon sour- ce in 4 days while others appeared only after a long period o f up to 1 0 - 1 2 days.

Sixteen mutants growing well on the synthetic glu-

cose media (3 from strain AA28, 7 from AA30 and 6 from AA31) were selected and crossed to p g i l A strains AA33 and AA34. The diploid strains constructed failed

to grow in a liquid glucose medium in 2 days. This showed that the mutat ions were recessive. One comple- menta t ion group, called S P G 1 (suppression o f p g i l A ) ,

was clearly defined by crossing spore progenies from all the crosses made (Table 1). Six alleles o f this gene could

b e identified. Genetic analysis o f other mutants gave no simple Mendelian segregation or showed a papilla growth

on glucose o f the diploids constructed with them and were not further analysed. No abnormal properties were caused by the suppressor mutat ions in a PGI1 wild type background.

The mutan t carrying the allele spg l -1 was mitot ical ly instable with regard to respiratory competence. Only 40% of the cells were still able to grow on YEP gly- cerol + ethanol after growth on the permissive YEP 2% fructose-0.1% glucose medium and there the ability to grow on glucose as sole carbon source was lost simul- taneously. Mutants with instable respiratory efficiency were already reported (Sherman 1964). This instabili ty was not associated with the spg l mutat ion. Spore cul- tures with this allele recovered from several crosses showed a range of respiratory deficient cells between 5 and 100%. This instabili ty was also observed during meiosis in a p g i l A background where 50% of the spores were respiratory deficient. Also when two mitot ical ly stable spores spg l -4 were crossed and genetic analysis

432 A. Aguilera: pgil/, suppressor mutants in S. cerevisiae

Table 3. Metabolite concentration (nmol/mg dry weight), 02 consumption and CO 2 production rates (nmol/h x 107 C), glucose con- sumed (mg/ml) (D) and ethanol produced (#g/ml) (ETH) in mutant and wild type strains. G6P: glucose-6-phosphate; glycerol and D- glucose express the carbon source used in YEP media to culture the ceils

Strains a G6P ATP D ETH Glyerol D-Glucose

02 C02 0 2 C02

PGI1 1.3 1.2 19.0 5,847 576 420 195 3,070

pgil/`25 113 nd . . . . pgi1A34 91 nd - 632 620 - -

pgil/,25 spgl-i 56 nd 19.0 9 - - 558 443 pgil/,25 spgl-3 54 nd 19.0 3 839 - 616 542 pgi1/,25 spgl-4 96 nd 19.0 9 860 660 927 721 pgil A25 spg1-6 63 nd 19.0 17 762 - 558 444

a all the other mutants selected gave the same results

nd not detectable

was performed, only 4 of 8 complete tetrads gave an expected 4 + : 0 - segregation for growth on glucose and glycerol-ethanol media. The other 4 gave 3+:1 . (2), 2+:2 - (1) and 1+:3 . (1). By crossing with a PGI + strain it was shown that all "petite" spores were r h o - . Six tho-spores were crossed with the kar2mutant XW5- 1B which prevents nuclear fusion during conjugation (Polaina and Conde 1982) and heteroplasmons with the nucleus of pgilAx mutant and active mitochondria were selected on the appropriate selective media and tested for growth on glucose. All of them recovered the capacity to grow on glucose as sole carbon source (Table 2).

Physiological characterization of the spgl suppressor mutants

The fact that no suppressor mutant of the pgilA could be obtained on media with antimycin A and that rho- strains were unable to express the phenotype o f growth on glucose as sole carbon source suggested that mito- chondrial activity is essential for growth on glucose as sole carbon source in the absence of phosphoglucose isomerase. Table 2 shows different enzymatic activities tested in the mutants grown on YEP 2% fructose-0.1% glucose. They were not affected in hexokinases activi- ties, they did not have detectable phosphoglucose iso- merase acitvity and the two enzymes of the glucose-6- phosphate oxidation reactions, glucose-6-phosphate de- hydrogenase and 6-phosphogluconate dehydrogenase were at wild type levels. The levels of total malate de- hydrogenase and alcohol dehydrogenase were in the wild type range both under repressing and non-repressing conditions (Table 2). Also pyruvate decarbocylase (data not shown) was present at wild type levels. These re-

sults suggest that the mutations are not involved either in catabolite repression or in an overexpression of the glucose-6-phosphate oxidative enzymes which would allow a more efficient utilization of glucose through these reactions.

The suppressed deletion mutants still showed a large accumulation of glucose-6-phosphate, although lower than a single pgilA mutants, but ATP was still not de- tectable. In contrast to a wild type PGI1 strain, suppres- sed pgilA mutants were able to consume all the glucose from the media without production of ethanol (Table 3). Warburg experiments showed that no fermentation took place in the mutants when growing on glucose and that respiration rates were similar in media with glucose (repression conditions) and glycerol (derepres- sion conditions). On the other hand these mutatnts were not able to grow on glucose as sole carbon source when 2% ethanol was added to the media, while S. cere- visiae wild type strains grow with 10% ethanol (Agui- lera and Benitez 1986).

High glucose concentrations inhibit fermentation and stimulate respiration in a pgil A deletion mutant

Experiments on glucose and fructose consumption and ethanol production as well as CO 2 production and 0 2 consumption were performed with a single pgilAx mu- tant to check whether this mutation itself was respon- sible for the lack of ethanol production and high respira- tion rates in 2% glucose media. Figure 1 shows that after 4 h of transferring the ceils from permissive 2% fructose-0.1% glucose media to synthetic media with 2% fructose and increasing amounts of glucose a peak of fermentation appeared only in cultures supplemented

A. Aguilera: pgilZx suppressor mutants in S. cerevisiae 433

0 ×

J:

E

COt 4 h 3 O,

2 q o e4

9 h 3 o

2

I 0 £15 .1 .2 .4

g l u c o s e (%)

I l l glUCOSe fructose

[~e thano l

3c

iJ ,o! i! , i i

0 .05

7 2 h

Fi

E~

2 ~

ifl ,1 .2 . 4

g l u c o s e ( * / . )

Fig. 1. Left: CO 2 production and 02 comsumption rates from strain AA30 (pg/1A) growing on 2% fructose synthetic medium supplemented with increasing amounts of glucose as expressed in abscisa, after 4, 9 and 26 h of having transferred the cells from YEP 2% fructose-0.1% glucose. CO 2 data correspond only to fermentation. CO2 produced by respiration has been not con- sidered. The numbers besides the symbols represent the optical density of the cultures at 660 nm. On fructose media without glucose this optical density was 0.17 at each time. Right: Glu- cose and fructose consumed and ethanol produced after 72 h of the transfer. Open parts of boxes represent glucose or fruc- tose still present in the media

Fig. 2. Photomicrographs of a strain PGI1 (left) and a mutant pgilA spgl (right) growing on YEP glucose (400x)

with 0.05% glucose. At 26 h the peak appeared in cul- tures supplied with 0.1% glucose and then low fermen- tation levels could be detected with 0.2% glucose. On the other hand a uniform level of respiration was ob- served in all the cultures with glucose after 4 h. At 26 h respiration rates were dependent on the glucose

concentration. After 72 h cells in the presence of 0.4% glucose had metabolized fructose and glucose without concomitant production of ethanol. Normal fermenta- tive metabolism took place at glucose concentration lower than 0.1% (Fig. 1). With no glucose in the media, fermentation of fructose was marginal.

These results showed that no ethanol fermentation and high respiration rates in the pgilA suppressor mu- tants are phenotypes caused by the pgilA mutation it- self. The suppressor spgl mutation presumably allows those mutants to use efficiently the pentose-phosphate cycle to grow on glucose as sole carbon source. Figure 2 shows unusually large cells with multiple buds in the suppressed pgilA mutants growing on glucose.

Discussion

The capacity of the pentose phosphate pathway is in- sufficient to support growth on glucose in S. cerevisiae. pgi-mutants lacking phosphoglucose isomerase activity are not able to grow on glucose as sole carbon source (Maitra 1971; Clifton et al. 1978; Herrera and Pascual 1978; Ciriacy and Breitenbach 1979; Aguilera 1986). However, other organisms like E. coli are able to use this shunt to grow on glucose (Vinopal et al. 1975) PGI1 deletion mutants of S. cerevisiae (Aguilera 1986) were used to select secondary mutants which are able to grow on glucose as sole carbon source. One comple- mentation group SPG1 was defined.

The deletion mutant pgilA cannot grow on glucose or fructose alone. There is no fermentative CO 2 produc- tion on glucose (Aguflera 1986), but a slow and margi- nal formation of small amounts of ethanol from fruc- tose (Fig. 1). However, very low concentrations such as 0.05% glucose promote wild type growth on fructose and also formation of ethanol from fructose. Increasing concentrations of glucose beyond 0.2% inhibit growth (Aguilera 1986) and slow down fructose catabolism with less and less of the fructose being converted to ethanol (Fig. 1).

Carbon catabolite repression in the deletion mutant cannot be studied on media with only fructose or only glucose because these sugars do not support growth. Growth is possible on combinations of 2% fructose and low concentrations of glucose. Respiration in wild type cells is repressed under these conditions. However, no reduction of respiration was observed in the deletion mutant growing on a medium with 2% fructose and up to 0.4% glucose. In fact, the observed respiration rates were unusually high (Fig. 1) when compared to wild type (Table 3). At present it is not clear whether this is due to a lack of carbon catabolite repression because total malate dehydrogenase was still normally repressed (Table 2).

434 A. Aguilera: pgil ~ suppressor mutants in S. cerevisiae

The fact that very low concentrations of glucose pro- mote growth on and a normal alcoholic fermentation of fructose suggests that glucose-6-phosphate could be a signal required for opening up the glycolytic pathway. This has not been observed with previously isolated mu- tants (e.g. Ciriacy and Breitenbach 1979) probably be- cause the residual activity in those mutants was still sufficient to provide the required concentrations of glu- cose-6-phosphate. Rasenberger (1979) observed normal growth on fructose which was fermented at half the wild type rate in a pgil mutant with a residual activity of 5-6% of the wild type level.

The strong inhibition of growth on fructose medium and the redirection of fructose catabolism to a pre- ponderant oxidation and reduction of the ethanol yield could be explained by an inhibition of glycolysis. pgil mutants have been shown not to be affected in hexose uptake (Bisson and Fraenkel 1983). Maitra (1971) suggested that the inhibitory effect of glucose on a pgil mutant was caused by a toxicity of the high intraceliular concentrations of glucose-6-phosphate. On the other hand, Cifiacy and Breitenbach (1979) attri- buted this inhibitory effect to a depletion of ATP since ATP used in the phosphorylation of glucose is trapped in glucose-6-phosphate. The only catabolic reaction for glucose-6-phosphate in the deletion mutant is the "direct oxidation" through the pentose phosphate shunt and this pathway has only a very limited capacity in Saccha- romyces cerevisiae (Gancedo and Lagunas 1973).

It is conceivable that some of the glucose-6-phos- phate accumulating at very high concentrations could be cleaved by an intracellular unspecific phosphatase so that some of the ATP spent on phosphorylation is lost in addition to what is directly used in the formation of other cellular components. The trapping of ATP in a glucose-6-phosphate sink will become less prominent at lower glucose concentrations so that a reduced amount becomes available for phosphorylation of fruc- tose and a limited glycolytic flux. Pyruvate, a metabolite of the lower glycolytic pathway, can be metabolized by two competing reactions: Pyruvate decarboxylase lead- flag to the formation of acetaldehyde and finally ethanol whereas pyruvate dehydrogenase converts pyruvate to acetyl-CoA and the pyruvate carbokinase leading to oxa- lacetate. The activity of pyruvate decarboxylase in yeast is much higher than that of pyruvate dehydro- genase but the kM of the dehydrogenase is much lower and therefore more pyruvate is converted to acetyl-CoA at low pyruvate concentrations and consequently when the glycolytic flux is weak oxidation would be predo- minant over ethanol formation (Holzer 1961).

The suppressor mutation spgl allows a pgilA mutant to grow on 2% glucose. However, no ethanol is formed (Table 3). This degradation of glucose is completely oxidative because the respiratory inhibitor antimycin A completely blocks growth. Moreover, respiratory defi- cient derivatives of the pgilA deletion mutant do not

grow on glucose. Glucose catabolism is efficient enough to support growth because the high respiration levels in the spgl pgilA double mutant. The mutation spgl all by itself grows normally on glucose and this even in a respiratory deficient state. A hkely function of this gene is that it controls the flux through the pentose phosphate shunt.

Acknowledgements. I would like to thank F. K. Zimmermann for many fruitful discussions and carefut reading of the manus- cript. This investigation was supported by the Deutsche For- schungsgemeinschaft. A. A. was a "Fundacion Juan March" (Spain) Postdoctoral Fellow.

References

Agullera A (1986) Mol Gen Genet 204:310-316 Aguilera A, Benitez T (1985) Arch Microbiol 142:389-392 Agullera A, Benitez T (1986) Arch Microbiol 143:337-344 Aguilera A, Zimmermann FK (1986) Mol Gen Genet 202:83-

89 Ballou CE (1982) Yeast cell wall and cell surface. In: Strathern

N J, Jones EW, Broach JR (eds) The molecular biology of the yeast Saecharomyees. Metabolism and gene expression. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York, pp 335-360

Bergmeyer HU (1970) Methoden der enzymatischen Analyse. Verlag Chernie, Weinheim

Bisson LF, Fraenkel DG (1983) Proc Natl Acad Sci USA 80: 1730-1734

Bisson LF, Fraenkel DG (1984) J Bacteriol 159:1013-1017 Cifiacy M, Breitenbach I (1979) J Bacteriol 139:152-160 Clifton D, Weinstock SB, Fraenkel DG (1978) Genetics 88:1-11 Entian KD, Zimmermann FK (1980) Mol Gen Genet 177:345-

350 Fraenkel DG (1982) Carbohydrate metabolism. In: Strathern

N J, Jones EW, Broach JR (eds) The molecular biology of the yeast Saccharomyces. Metabolism and gene expression. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York, pp 1-37

Gancedo JM, Lagunas R (1973) Plant Sci Lett 1:193-198 Goldstein A, Larnpen JO (1975) Methods Enzymo142:504-511 Henry SA (1982) Membrane lipids of yeast: Biochemical and

genetic studies. In: Strathem N J, Jones EW, Broach JR (eds) The molecular biology of the yeast Saccharomyces. Meta- bolism and gene expression. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York, pp 101-158

Herrera LS, Pascual C (1978) J Gen Microbiol 108:305-310 Holzer H (1961) Cold Spring Harbor Symp Quant Biol 26:

277 -288 Maitra PK (1971) J Bacteriol 107:759-769 Polaina J, Conde J (1982) Mol Gen Genet 186:256-258 Rasenberger H (1979) Doctoral Thesis, Technische Hochschule

Darmstadt, FRG Sherman F (1964) Genetics 49:39-48 Sherman F, Fink GR, Lawrence CW (1979) Methods in yeast

genetics. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York

Vinopal RT, Hillman JD, Schulman H, Reznikoff WS, Fraenkel DG (1975) J Bacteriol 122:1172-1174

Zamenhoff S (1957) Methods Enyzmol 3:696-704

Communicated by K. Wolf

Received October 22, 1986