From: Flikweert M.T. (1999) Physiological roles of pyruvate
decarboxylase in Saccharomyces cerevisiae. PhD thesis, Delft University
of Technology, Delft. ISBN 90-9013020-9 (Yeast 12:247-257 (1996))
In Saccharomyces cerevisiae, the structural genes PDC1, PDC5 and
PDC6 each encode an active pyruvate decarboxylase. Replacement mutations in
these genes were introduced in a homothallic wild-type strain, using the dominant
marker genes APT1 and Tn5ble. A pyruvate-decarboxylase-negative (Pdc-) mutant
lacking all three PDC genes exhibited a three-fold lower growth rate in complex
medium with glucose than the isogenic wild-type strain. Growth in batch
cultures on complex and defined media with ethanol was not impaired in Pdc-
strains. Furthermore, in ethanol-limited chemostat cultures, the biomass yield
of Pdc- and wild-type S. cerevisiae were identical. However, Pdc- S. cerevisiae
was unable to grow in batch cultures on a defined mineral medium with glucose
as the sole carbon source. When aerobic, ethanol-limited chemostat cultures (D
= 0.10 1/h) were switched to a feed containing glucose as the sole carbon
source, growth ceased after approximately 4 hours and, consequently, the
cultures washed out. The mutant was, however, able to grow in chemostat
cultures on mixtures of glucose and small amounts of ethanol or acetate (5 % on
a carbon basis). No growth was observed when such cultures were used to
inoculate batch cultures on glucose. Furthermore, when the mixed-substrate
cultures were switched to a feed containing glucose as the sole carbon source,
wash-out occurred. It is concluded that the mitochondrial
pyruvate-dehydrogenase complex cannot function as the sole source of acetyl-CoA
during growth of S. cerevisiae on glucose, neither in batch cultures nor in
glucose-limited chemostat cultures.
Marcel T.Flikweert, Linda van der Zanden, Wouter M.T.M. Janssen, H.Yde
Steensma, Johannes P. van Dijken, and Jack T. Pronk.
In yeasts, respiratory dissimilation of pyruvate is initiated by its
conversion into acetyl-CoA. This can occur in two ways: via a direct reaction
catalysed by the mitochondrial pyruvate-dehydrogenase complex or via an
indirect route, involving pyruvate decarboxylase, acetaldehyde dehydrogenase
and acetyl-coenzyme A synthetase (Fig.1; Holzer and Goedde, 1957, Pronk et
Experiments with isogenic S. cerevisiae mutants defective in the
synthesis of an active pyruvate-dehydrogenase complex have demonstrated that,
during glucose-limited aerobic growth of wild-type cells, this enzyme is
predominantly or even exclusively responsible for respiratory pyruvate
dissimilation. Under these conditions, the indirect route apparently does not
play an important role in respiratory pyruvate metabolism (Pronk et al.,
To study the metabolic significance of the pyruvate-dehydrogenase bypass route,
it is of interest to investigate the physiology of mutants affected in pyruvate
decarboxylase (EC 18.104.22.168). S. cerevisiae contains three structural
genes that each encode an active pyruvate decarboxylase; PDC1, PDC5 and PDC6
(Hohmann, 1991a). Strains in which PDC1 and PDC5 or all three PDC genes have
been disrupted lack pyruvate-decarboxylase activity. Such pyruvate-decarboxylase-negative
(Pdc-) mutants showed a reduced growth rate in complex (yeast extract-peptone)
media supplemented with glucose (Hohmann, 1991a). Although, under all growth
conditions tested, expression of PDC6 was either very low or absent, revertants
of pdc1-pdc5 double mutants have been isolated, in which a recombination event
had caused a fusion of the PDC1 promotor and the PDC6 open-reading frame
(Hohmann et al., 1991b).
Enzymes of pyruvate metabolism in Saccharomyces cerevisiae. Numbered
reactions are catalysed by the following enzymes: 1, pyruvate decarboxylase; 2,
pyruvate dehydrogenase complex; 3, acetaldehyde dehydrogenase; 4, acetyl-coenzyme
A synthetase; 5, alcohol dehydrogenase.
Therefore, physiological studies on Pdc- S. cerevisiae mutants should
preferably be performed with stable strains in which all three PDC genes are
Physiological characterization of the S. cerevisiae Pdc- mutants
described in the literature has been restricted to growth studies in complex
media (Hohmann, 1991a). The auxotrophic markers present in these strains make
them unsuited for quantitative studies in defined media. In particular, the
addition of amino acids to growth media may lead to substantial changes in the
metabolism of the carbon source. For example, approximately 5% of yeast biomass
consists of leucine (Oura 1972). Furthermore, use of a genetic background
containing auxotrophic markers may obscure effects of gene disruptions on amino
acid metabolism: a partial leucine deficiency of a
pyruvate-dehydrogenase-negative mutant was initially overlooked because the
pda1 strain was constructed in a leu2 background, which required the inclusion
of leucine in the growth media (Wenzel et al., 1992a).
The aim of the present work was to construct isogenic,
pyruvate-decarboxylase-negative mutants of a homothallic, prototrophic S.
cerevisiae strain and to characterize their growth in mineral media.
The Saccharomyces cerevisiae strains used in this study are listed in
Table 1. PDC genes were disrupted in the homothallic, prototrophic, homozygous
diploid strain T2-3D (Wenzel et al., 1992a, Pronk et al., 1994,
de Jong-Gubbels et al., 1995). Escherichia coli, strain XL1 blue
(Bullock et al., 1987) was used for plasmid amplification.
Table 1. S. cerevisiae strains used in the present study. Strain T2-3D
(Wenzel et al., 1993, Pronk et al., 1994, de Jong-Gubbels et
al., 1995) is a homozygous diploid strain, derived from the heterozygous
strain CBS8066 (Centraal Bureau voor Schimmelcultures, Delft, The Netherlands).
Strains GG 562-GG 570 are isogenic mutants of T2-3D.
pdc1::Tn5ble/pdc1::Tn5ble PDC5/PDC5 PDC6/PDC6
pdc1::Tn5ble/pdc1::Tn5ble pdc5::Tn5ble/pdc5::Tn5ble PDC6/PDC6
pdc1::Tn5ble/pdc1::Tn5ble PDC5/PDC5 pdc6::APT1/pdc6::APT1
pdc1::Tn5ble/pdc1::Tn5ble pdc5::Tn5ble/pdc5::Tn5ble pcd6::APT1/pdc6::APT1
Wild-type S. cerevisiae and pdc mutants were grown to stationary
phase in shake-flask cultures on complex medium containing 2 % (v/v) ethanol.
After adding glycerol (15 % v/v), 2-ml aliquots were stored in sterile vials at
-70 oC. Prior to growth experiments, samples from a frozen stock culture were
streaked on complex medium-ethanol agar plates. Precultures were inoculated
directly from these plates.
Standard protocols were followed for plasmid isolation, restriction,
ligation, Southern blotting, hybridization and gel electrophoresis (Maniatis et
al., 1982). Yeast chromosomal DNA was isolated by the method of Holm et
al. (1986). S. cerevisiae and E. coli strains were transformed with
a Bio-Rad gene pulser (Dower et al., 1988). Sporulation, dissection and
mating of S. cerevisiae strains was performed according to published
Subclones of PDC1, PDC5 and PDC6 in pUC vectors were kindly provided by Dr.
S. Hohmann. The one-step gene-disruption method (Rothstein 1983) was used to
inactivate the PDC1, PDC5 and PDC6 genes in S. cerevisiae T2-3D (Fig.
PDC1 was disrupted by replacing an internal 1058 bp KpnI-BglII fragment with a
1.35 kb KpnI-BglII fragment from the plasmid pUT332 (Gatignol et al.,
1990) containing the marker gene Tn5ble under the control of the S.
cerevisiae TEF1 promoter and CYC1 terminator. A 1.15 kb HindIII-KpnI
fragment from pUT332, carrying the same marker gene, was used to replace an
internal 691 bp HindIII-KpnI fragment of PDC5. PDC6 was disrupted by replacing
an internal 1190 bp BclI fragment with a 3.2 kb BamHI-BclI fragment from the
plasmid pBEJ24 (Hadfield et al., 1990) containing the marker gene APT1
under the control of the S. cerevisiae PGK1 promoter and CYC1
terminator. After transformation of S. cerevisiae T2-3D with linear
restriction fragments containing the disrupted genes, transformants were
selected on YPD plates containing either phleomycin (strains expressing Tn5ble)
or G418 (transformants expressing APT1), as described by Wenzel et al.
(1992b) and Hadfield et al. (1990), respectively. Since the strains are
homothallic, spore-to-spore matings were used to obtain strains in which two or
three PDC genes were disrupted. The following combinations were used: GG 562 x
GG 564; GG 562 x GG 566, GG 564 x GG 566, and GG 562 x GG 567. The resulting
heterozygous diploid strains were again sporulated and dissected to obtain the
homozygous strains. Spore-to-spore matings were performed on CY plates with 2 %
(v/v) ethanol instead of glucose. Spore viability was low, probably due to the
pdc mutations (Hohmann 1991a). Dissection on glucose or galactose media did not
significantly improve spore viability. The genotype of all strains containing
single or multiple disrupted PDC genes was confirmed
by Southern analysis (Fig. 3).
The mineral medium contained per litre of demineralized water: (NH4)2SO4, 5
g; KH2PO4, 3 g; MgSO4·7H2O, 0.5 g; EDTA, 15 mg; ZnSO4·7H2O, 4.5 mg; CoCl2·6H2O,
0.3 mg; MnCl2·4H2O, 1 mg; CuSO4·5H2O, 0.3 mg; CaCl2·2H2O, 4.5 mg; FeSO4·7H2O,
3.0 mg; Na2MoO4·2H2O, 0.4 mg; H3BO3, 1.0 mg; KI, 0.1 mg and silicone antifoam
(BDH); 0.05 ml. After heat sterilization (120 °C) of the medium,
filter-sterilized vitamins were added, to final concentrations per litre of:
biotin, 0.05 mg; calcium pantothenate, 1.0 mg; nicotinic acid, 1.0 mg;
inositol, 25.0 mg; thiamin HCl, 1.0 mg; pyridoxine HCl, 1.0 mg and
para-aminobenzoic acid, 0.2 mg. The concentration of ethanol or glucose in the
reservoir medium was 5.75 g/l or 7.5 g/l respectively (0.25 Cmol/l). Complex
medium contained per litre: yeast extract (Difco), 10 g; peptone from casein
(Merck), 20 g; and 2 % (v/v) ethanol (YPE) or 20 g D-glucose (YPD). CY plates
contained per litre: yeast extract (Difco), 5 g; bactopeptone (Difco), 5 g;
agar (Difco), 20 g; and glucose, 20 g.
Fig. 2. Schematic representation of the gene disruptions in PDC1, PDC5 and
PDC6. Restriction sites are indicated by the following abbreviations: B =
BamHI, Bc = BclI, Bg = BglII, H = HindIII, K = KpnI, P = PstI.
Fig. 3. Southern analyses of genomic DNA restriction digests. Panel A:
HindIII digests. The probe contained the TEF1 promoter and the Tn5ble gene from
the phleomycin-resistance cassette. The largest hybridizing fragment in lanes
1-5 contains the native TEF1 promoter. Panel B: PstI digests. The probe
contained the 5' region of the PDC6 gene. Relevant restriction sites are
indicated in Figure 1. Lane 1: S. cerevisiae T2-3D (wild-type), lane 2:
GG 562 (pdc1::Tn5ble), lane 3: GG 564 (pdc5::Tn5ble), lane 4: GG 568
(pdc1::Tn5ble pdc5::Tn5ble), lane 5: GG 570 (pdc1::Tn5ble pdc5::Tn5ble
pdc6::APT1), lane 6: T2-3D (wild-type), lane 7: GG 566 (pdc6::APT1), lane 8: GG
570 (pdc1::Tn5ble pdc5::Tn5ble pdc6::APT1).
Precultures were prepared by inoculating 100 ml YPE (2% ethanol) with a few
colonies from a plate. Cultures were incubated on an orbital shaker (200 rpm)
at 30 °C for two days. For growth curves, 1 ml of the preculture was inoculated
in a 500 ml erlenmeyer with 100 ml YPE (2% ethanol) or 100 ml YPD (2% glucose)
and then shaken (200 rpm) at 30 °C. Optical-density measurements were performed
at appropriate intervals as described by Weusthuis et al. (1994). For
induction of pyruvate decarboxylase, 10 ml of a preculture was inoculated in a
100 ml shake flask with either 50 ml YPE (2% ethanol) or 50 ml YPD (8% glucose)
and shaken for 6 hours at 30 °C (Hohmann, 1991a).
Batch cultivation was performed at 30 °C in laboratory fermenters (Applikon,
Schiedam The Netherlands) with a working volume of 1.5 litre. The pH was
controlled at 5.0 ± 0.1 by automatic addition of 2 mol/l KOH and 1 mol/l H2SO4.
The fermenter was flushed with air at a flow rate of 1.5 l/min and stirred at
800 rpm. The dissolved-oxygen concentration was continuously monitored with an
oxygen electrode (Ingold, 34 100 3002) and remained above 60% of air
saturation. Cultures were grown on the mineral medium described above, with
glucose (25 g/l initial concentration) or ethanol (7.9 g/l initial
concentration) as the sole carbon source. 25 ml samples were withdrawn at
appropriate intervals for determination of dry weight and metabolite
Aerobic chemostat cultivation was performed at 30 °C in laboratory
fermenters (Applikon, Schiedam, The Netherlands), at a stirrer speed of 750 rpm
and at a dilution rate of 0.10 1/h. The working volume of the cultures was kept
at 1.0 l by a peristaltic effluent pump coupled to an electrical level sensor.
This set-up ensured that under all growth conditions, biomass concentrations in
samples taken directly from the cultures differed by less than 1% from biomass
concentrations in samples taken from the effluent line (Noorman et al.,
1991). The pH was kept constant at 5.0 by an ADI 1020 biocontroller, via the
automatic addition of 2 mol/l KOH. The fermenter was flushed with air at a flow
rate of 0.7 l/min using a Brooks 5876 mass-flow controller. The
dissolved-oxygen concentration was continuously monitored with an oxygen
electrode (Ingold, 34 100 3002) and remained above 50% air saturation.
Steady-state data refer to cultures without detectable oscillations. Chemostat
cultures were checked for purity using phase-contrast microscopy.
The dry weight of washed culture samples was determined using 0.45 µm
membrane filters and a microwave oven as described by Postma et al.
(1989). Parallel samples varied by less than 1%.
Organic acids, ethanol and glycerol in culture supernatants were determined
by HPLC analysis using a Phenomenex column (Rezex ROA Organic acid 00H-0138-KO)
at 60 °C. The column was eluted with 0.5 g/l sulphuric acid at a flow rate of
0.5 ml/min. Organic acids were detected by a Waters 441 UV-meter at 214 nm
coupled to a Waters 741 Data module. Ethanol and glycerol were detected by an
Erma ERC 7510 refractive-index detector coupled to a Hewlett Packard 3390A RI
integrator. 20 µl samples were injected using a Hamilton syringe. Glucose in
reservoir media and supernatants was determined enzymically using the GOD-PAP
method (Merck Systems kit 14144; detection limit ca. 5 µM). Ethanol was assayed
colorimetrically with an alcohol oxidase/peroxidase kit (Leeds Biochemicals;
detection limit ca. 100 µM).
For preparation of cell-free extracts, culture samples were harvested by
centrifugation, washed twice with 10 mM potassium-phosphate buffer, pH 7.5,
containing 2 mM EDTA, concentrated 4-fold and stored at -20 °C. Before
assaying, the samples were thawed at room temperature, washed and resuspended
in 100 mM potassium phosphate buffer, pH 7.5, containing 2 mM MgCl2 and 1 mM
dithiothreitol (DTT). Extracts were prepared by sonification with 0.7 mm
diameter glass beads at 0 °C for 2 min. at 0.5 min. intervals with an MSE
sonicator (150 W output, 7 µm peak-to-peak amplitude). Unbroken cells and
debris were removed by centrifugation at 4 °C (20 min. at 36,000 x g). The
supernatant was used as the cell-free extract.
Pyruvate-decarboxylase activity was assayed at 30 °C immediately after
preparation of the extracts, using a Hitachi model 100-60 spectrophotometer set
at 340 nm. Reaction rates were linearly proportional to the amount of cell-free
extract added. The assay mixture consisted of: 40 mM imidazole-HCl buffer (pH
6.5), 0.2 mM thiamine pyrophosphate (TPP), 0.15 mM NADH, alcohol dehydrogenase
88 U/ml (Boehringer), 5 mM MgCl2, and cell-free extract. The reaction was started
with 50 mM pyruvate.
Protein concentrations in cell-free extracts were determined by the Lowry
method. Bovine-serum albumin (BSA; fatty-acid-free, Sigma Chemical Co.) was
used as a standard. The protein content of whole cells was determined by a
modified biuret method (Verduyn et al., 1990).
Effects of the gene disruptions on pyruvate-decarboxylase expression were
investigated by measuring enzyme activities in cell-free extracts of wild-type S.
cerevisiae T2-3D and in homozygous mutant strains containing one, two or
three disrupted PDC genes. To discriminate between constitutive and
glucose-inducible pyruvate-decarboxylase activity, cells were pregrown in
complex medium with ethanol as the carbon source and then either incubated in
the ethanol medium used for growth or induced by incubation in glucose medium
(Hohmann, 1991a). The wild-type strain T2-3D exhibited a high
pyruvate-decarboxylase activity after induction in complex medium with glucose
(ca. 3 U/mg protein, Table 2). An approximately three-fold lower activity was
measured in extracts from non-induced wild-type cells grown on ethanol.
When strain GG562 carrying the pdc1::Tn5ble mutation was induced with glucose,
its pyruvate-decarboxylase activity, determined in cell-free extracts, was only
ca. 30 % lower than that of glucose-induced wild-type cells (Table 2). In
non-induced cells, disruption of PDC1 resulted in a ten-fold reduction of the
pyruvate-decarboxylase activity in comparison with the wild-type strain. Single
gene disruptions in either PDC5 or PDC6 did not significantly affect enzyme
activities, neither in induced nor in non-induced cells (Table 2).
Table 2. Specific pyruvate-decarboxylase activity and growth rates of
wild-type (T2-3D) and pdc mutant strains. For enzyme activity assays, cells
pregrown on complex medium with ethanol were induced on either 8% (w/v) glucose
or 2% (v/v) ethanol in complex medium. Growth rates were determined in complex
medium containing either 2% (v/v) ethanol or 2% (w/v) glucose.
1.0 ± 0.10
0.30 ± 0.02
0.1 ± 0.05
0.30 ± 0.01
1.0 ± 0.20
0.29 ± 0.01
1.0 ± 0.25
1.2 ± 0.10
0.29 ± 0.02
0.27 ± 0.00
Pyruvate-decarboxylase activities in strains which, in addition to a
disrupted PDC1 or PDC5 gene, contained a disruption in PDC6, were not
significantly different from the activities in strains carrying the
corresponding single gene disruptions (Table 2). When both PDC1 and PDC5 were
disrupted, and PDC6 was the only remaining intact PDC gene, no enzyme activity
was detected in cell-free extracts prepared from induced or non-induced cells.
A complete absence of pyruvate-decarboxylase activity was also observed in
extracts of a triple mutant (strain GG 570), in which all three PDC genes had
been disrupted (Table 2).
For an initial physiological characterization, and to enable comparison with
pdc mutations introduced in a different S. cerevisiae genetic background
(Hohmann 1991a), growth rates of the PDC mutant strains were determined in
shake-flask cultures on complex media with glucose or ethanol.
In complex medium with ethanol, growth rates of strains carrying one, two or
three disrupted PDC genes did not differ significantly from those of the
isogenic wildtype (Table 2). This result is consistent with the fact that
pyruvate decarboxylase is not involved in ethanol metabolism. Nevertheless, it
differs from the observation of Hohmann (1991a) that strains in which both PDC1
and PDC5 had been disrupted showed a 20 - 25 % reduction of the specific growth
rate on ethanol.
Disruption of any single PDC gene did not affect the growth rate in complex
medium with glucose. In double mutants, growth rates on glucose were not
significantly reduced when combinations of PDC6 and either PDC1 or PDC5 were
disrupted (Table 2). However, disruption of both PDC1 and PDC5 resulted in a 70
% decrease of the specific growth rate on glucose. This negative effect on
growth rate was not enhanced by the additional disruption of PDC6 (Table 2).
Our results confirm the conclusion of Hohmann (1991a) that, during growth in
ethanol- or glucose-containing media, PDC6 expression is either very low or
absent. However, it has been demonstrated that recombination events may lead to
the activation of PDC6 (Hohmann, 1991b). Since such instability is not
desirable in physiological studies, it was decided to use the triple mutant
strain GG 570 for further physiological investigations on the effects of
pyruvate-decarboxylase deficiency during growth of S. cerevisiae in
Quantitative analysis of yeast physiology requires the use of defined
mineral media. Therefore, aerobic growth of wild-type S. cerevisiae
T2-3D in a defined mineral salts medium supplemented with vitamins was compared
with growth of the isogenic pyruvate-decarboxylase-negative strain GG 570,
using pH-controlled fermenter cultures.
When grown on ethanol, there was no difference in growth rate between the
wild-type strain and the pyruvate-decarboxylase-negative mutant: both strains
grew exponentially with a specific growth rate of 0.13 ± 0.01 1/h. The
wild-type strain grew exponentially on glucose, with a specific growth rate of
0.45 ± 0.01 1/h (Fig. 4). Growth on glucose was accompanied by the formation of
ethanol and small amounts of pyruvate (3 mmol/l) and glycerol (0.3 mmol/l). In
contrast to the wild-type strain, the Pdc- strain GG 570 did not exhibit
exponential growth on glucose. Instead, growth ceased after less than one
biomass doubling (Fig. 4). No ethanol or acetate was detected, but
concentrations of pyruvate (8 mM) and glycerol (2 mM) attained higher values
than in wild-type cultures, even though the biomass concentrations in mutant
cultures were much lower.
Since a Pdc- mutant can not grow fermentatively, respiration is essential for
its growth on glucose. In S. cerevisiae, many enzyme activities involved
in respiratory sugar metabolism are subject to glucose catabolite repression
(Gancedo 1992). To investigate whether glucose repression of respiration might
be responsible for the mutant's impaired growth on glucose in batch cultures,
it was subsequently attempted to establish glucose-limited chemostat cultures.
The pyruvate-dehydrogenase complex, rather than the bypass via pyruvate
decarboxylase, is the predominant route of respiratory pyruvate metabolism
during glucose-limited growth at D=0.10 1/h (Pronk et al., 1994).
Furthermore, many key enzymes of glucose metabolism, including the
pyruvate-dehydrogenase complex, are expressed constitutively during growth of S.
cerevisiae T2-3D on ethanol (Wenzel et al., 1993; Pronk et al.,
1994; de Jong-Gubbels et al., 1995). It was therefore anticipated that
steady-state chemostat cultures growing on ethanol would readily adapt to
growth on glucose under glucose limitation.
Growth of wild-type S. cerevisiae T2-3D (l) and Pdc- triple mutant GG
570 (pdc1::Tn5ble pdc5::Tn5ble pdc6::APT1; (¡) triple mutant on a defined
mineral medium containing 25 g/l glucose as the sole carbon source. Batch
cultivation was performed in pH-controlled, aerobic fermenters.
In ethanol-limited chemostat cultures (D=0.10 1/h) grown on a defined
medium, the biomass yield of the pyruvate-decarboxylase-negative triple mutant
GG 570 was not significantly different from that of the isogenic wild-type
strain T2-3D (Table 3). In an attempt to avoid glucose repression,
ethanol-limited chemostat cultures (D = 0.10 1/h) of the
pyruvate-decarboxylase-negative triple mutant were switched to a medium
containing glucose as the sole carbon source.
During the first 4 h after the switch, the biomass concentration remained
approximately constant and the glucose concentration in the culture remained
below 0.2 g/l (Fig. 5). This suggested that indeed, the culture rapidly adapted
from ethanol-limited to glucose-limited growth. However, after this initial
period, the biomass concentration in the culture decreased and glucose
accumulated (Fig. 5). The observed decrease of the biomass concentration was
consistent with wash-out kinetics, indicating that growth had ceased completely.
The wash-out of biomass and the accumulation of glucose was accompanied by the
transient accumulation of pyruvate in the culture to a maximum concentration of
7 mM (Fig. 5).
The observation that, both in batch and chemostat cultures, growth of the Pdc-
strain on glucose continued for a number of hours before growth ceased, can in
theory be caused by a bottleneck in a biosynthetic pathway that requires
pyruvate decarboxylase. This would be consistent with the ability of Pdc-
strains to grow, albeit poorly, in complex media with glucose (Table 2), in
which precursors for biosynthesis can be obtained from yeast extract and/or
peptone. Since growth of the mutant strain on mineral medium with ethanol
appeared normal, formation of biosynthetic intermediates from ethanol was
apparently not affected.
Table 3. Steady-state biomass yields (YSX, g biomass/[mol substrate carbon]),
protein contents and pyruvate- decarboxylase activities in ethanol- and
glucose-limited, aerobic chemostat cultures of wild-type (T2-3D) and Pdc- (GG
570) S. cerevisiae. Relative concentrations of glucose and acetate in
mixed-substrate cultures are presented as a percentage of the total carbon
concentration (0.25 mol C/l) in the feed. Growth conditions: D = 0.10 1/h, pH
5, T = 30 oC, dissolved-oxygen concentration > 50 % air saturation (n.d.:
14.4 ± 0.4
41 ± 2
0.7 ± 0.3
GG 570 (Pdc-)
14.3 ± 0.3
42 ± 2
16.0 ± 0.3
40 ± 2
0.8 ± 0.1
95 % glucose-5 % acetate
16.5 ± 0.1
40 ± 2
95 % glucose-5 % acetate
16.2 ± 0.4
40 ± 1
To study whether growth was possible on mixtures of glucose and C2-compounds,
ethanol-limited chemostat cultures were switched to mineral medium containing a
mixture of glucose (237.5 mmol C/l) and ethanol (12.5 mmol C/l). This approach
resulted in steady-state cultures, in which no residual glucose or acetate
could be detected. Enzyme assays in cell-free extracts confirmed the absence of
pyruvate decarboxylase activity (Table 3). The biomass concentration in the
cultures did not differ significantly from that in similar cultures of the
wild-type strain (Table 3). The same results were obtained when, instead of
acetate, low concentrations of ethanol were added to the reservoir media (data
In the mixed-substrate cultures, glucose made up 95 % of the substrate carbon
fed to the cultures and was completely consumed. Nevertheless, when samples
from such glucose-limited cultures were used to inoculate batch cultures on
mineral medium with glucose, no growth was observed. The inability to grow on
glucose in batch cultures could not be relieved by the addition of low
concentrations of acetate or ethanol to the mineral media. Attempts to change
the medium feed of chemostat cultures from a 95 % glucose/5 % acetate mixture
to glucose as the sole carbon source reproducibly resulted in wash-out of the
cultures (data not shown). This indicated that even cells utilizing high
glucose-to-acetate ratios could not be readily adapted to growth on glucose as
the sole carbon source.
In batch cultures grown on complex media with ethanol or glucose, PDC1 was
expressed constitutively, whereas expression of PDC5 appeared to be induced by
glucose. PDC6 did not contribute significantly to the overall level of pyruvate
decarboxylase. The pattern of pyruvate-decarboxylase (Table 2) was generally
consistent with the pattern observed by Hohmann (1991a), who studied the effect
of PDC gene disruptions in a different S. cerevisiae genetic background.
It should be borne in mind, however, that the differential regulation of the
three PDC genes has so far only been studied during growth on complex media in
shake-flask cultures. The possibility that, under appropriate growth
conditions, PDC6 is transcribed at significant levels, can therefore not be
Growth experiments in defined and complex media with ethanol as the carbon
source gave no indications for pleiotropic effects of the PDC mutations. This
is in contrast with the results of Hohmann (1991a), who found a significantly
reduced growth rate of Pdc- strains on ethanol. Although, based on established
metabolic pathways, no effect of pyruvate-decarboxylase deficiency on ethanol
metabolism is to be expected, this discrepancy deserves further attention.
In the literature, the effect of a Pdc- phenotype, established either by random
mutagenesis (Schmitt & Zimmermann 1982) or by gene disruption (Hohmann
1991a) has only been studied in cultures grown on complex media. These studies
invariably demonstrated a reduction of the specific growth rate in
glucose-containing media. The residual growth rates in the mutant strains were
consistently at least 25 % of the wild-type rate (Schmitt & Zimmermann
1982; Hohmann 1991a). A similar effect was observed in the present study (Table
In the absence of pyruvate-decarboxylase activity, growth of S. cerevisiae
becomes critically dependent on respiration. Indeed, Hohmann (1991a)
demonstrated that growth of Pdc- mutants on glucose was completely arrested in
the presence of the respiratory inhibitor antimycin A. It is well-known that in
S. cerevisiae, many respiratory enzymes are repressed in the presence of
excess glucose (Entian 1986, Gancedo & Serrano 1989). However, glucose
catabolite repression of respiratory enzymes is generally not complete (Gancedo
1992). This is consistent with the observed growth, albeit at a reduced rate,
of the Pdc- strains in complex medium with glucose (Table 2).
Surprisingly, Pdc- S. cerevisiae completely failed to grow in batch
cultures on a mineral medium with glucose as the sole carbon source. Clearly,
if glucose repression of respiratory enzymes were the sole factor affecting the
growth rate of Pdc- mutants, a residual growth rate similar to that observed in
complex medium would be expected. Pdc- strains retained the ability to grow in
mineral medium with ethanol and were apparently able to convert glucose into
pyruvate (Fig. 5). It is therefore conceivable that the absence of growth on
glucose in a defined medium is due to a shortage of acetyl-CoA.
Fig. 5. Concentrations of biomass, glucose and pyruvate after switching a
chemostat culture (D = 0.10 1/h) of the Pdc- triple mutant S. cerevisiae
GG 570 (pdc1::Tn5ble pdc5::Tn5ble pdc6::APT1) from growth on a mineral medium
with ethanol (0.25 Cmol/l) to a medium containing glucose (0.25 Cmol/l) as the
sole carbon source. The dashed line drawn through biomass data points represent
wash-out kinetics, assuming a zero growth rate.
Inclusion of low concentrations of ethanol or acetate in the medium feed was
required to enable growth of Pdc- S. cerevisiae in glucose-limited
chemostat cultures. The most likely explanation for this dependency on
C2-compounds is a limitation in the synthesis of acetyl-CoA, an important
building block for the synthesis of TCA-cycle intermediates, lipids and some
amino acids (Oura 1972). An inability to synthesize one or more of these
biosynthetic precursors would be consistent with the phenotype of the Pdc-
mutants: growth in complex medium with glucose may still occur at the expense
of biosynthetic precursors available from yeast extract and peptone. Moreover,
the observation that growth in chemostat cultures continued for a short period
after transfer from an ethanol feed to a glucose feed (Figure 5) may be
explained from depletion of an intracellular metabolite pool.
Fig. 6. Hypothetical scheme of subcellular compartmentation of pyruvate and
acetyl-CoA metabolism in Saccharomyces cerevisiae, explaining the
requirement of a Pdc- mutant for C2-compounds. If acetyl-CoA export from the
mitochondria is restricted, glucose-grown cells depend on a source of cytosolic
acetyl-CoA. In the absence of pyruvate decarboxylase, cytosolic acetyl-CoA
cannot be synthesized from glucose, resulting in a requirement for exogenous
C2-compounds. Numbered arrows indicate the following pathways or enzymes: 1,
glycolysis; 2, pyruvate dehydrogenase complex; 3, TCA cycle; 4, pyruvate
decarboxylase; 5, acetaldehyde dehydrogenase; 6, acetyl-coenzyme A synthetase;
7, lipid synthesis.
In the Pdc- strain, the only reaction that can lead to the formation of
acetyl-CoA from pyruvate is the direct oxidative decarboxylation of pyruvate by
the mitochondrial pyruvate-dehydrogenase complex. As a result, formation of
acetyl-CoA in the Pdc- mutant is confined to the mitochondrial matrix (Figure
6). This necessitates export of acetyl-CoA to the cytosol, where lipid
synthesis occurs (Ratledge and Evans, 1989). Pyruvate decarboxylase,
acetaldehyde dehydrogenase and acetyl-CoA synthetase all have been reported to
be present in the cytosol of wild-type S. cerevisiae (van Urk et al.
1989; Jacobson and Bernofsky 1974; Kispal et al., 1991). Consequently,
the pyruvate-dehydrogenase bypass may act as the major source of cytosolic
acetyl-CoA in wild-type S. cerevisiae, which would preclude the
necessity of acetyl-CoA export from the mitochondrial matrix.
In yeasts, two enzyme systems may catalyse transport of acetyl-CoA across the
mitochondrial inner membrane. ATP-citrate lyase, a key enzyme of one of these
two systems, is absent in S. cerevisiae (Ratledge and Evans, 1989). A
second system, the acetyl-carnitine/carnitine translocase is generally assumed
to catalyse import of acetyl-CoA into the mitochondria (Kohlhaw and Tan-Wilson,
1977; Schmalix and Bandlow 1993). However, to what extent this system can also
catalyse the reverse reaction under physiological conditions is at present
The observation that small amounts of ethanol or acetate allow normal growth of
the Pdc- mutant in glucose-limited chemostat cultures strongly suggests that
pyruvate decarboxylase plays a crucial role in the supply of cytosolic
acetyl-CoA in wild-type cells (Figure 5) and that this function can not be
fulfilled by the mitochondrial pyruvate-dehydrogenase complex. Further studies
on the reversibility of the acetyl-carnitine shuttle under physiological
conditions is required to further substantiate this hypothesis. This
notwithstanding, it can be concluded that in addition to its key role in
alcoholic fermentation, pyruvate decarboxylase fulfils at least one other
essential role in glucose metabolism in S. cerevisiae. Furthermore, this
work exemplifies the necessity of defined growth conditions for studies on S.
cerevisiae mutants affected in central metabolic pathways. Only by using
mineral media and chemostat cultivation, the unexpected behaviour of Pdc-
mutants could be unveiled.
We thank Dr. Stefan Hohmann for providing us with subclones of the PDC genes
and our colleagues Gijs Kuenen, Lex Scheffers and Mike Jetten for critical
reading of the manuscript. J.T. P. thanks Marco van den Berg and Thibaut Wenzel
for their help and advice during the molecular genetic part of this study. This
research was carried out in the framework of the ABON program supported by the
Dutch Ministry of Economic Affairs.