Download TCA cycle activity in Saccharomyces cerevisiae is a function of the

Microbiology (2004), 150, 1085–1093
DOI 10.1099/mic.0.26845-0
TCA cycle activity in Saccharomyces cerevisiae
is a function of the environmentally determined
specific growth and glucose uptake rates
Lars M. Blank and Uwe Sauer
Institute of Biotechnology, ETH Zürich, Zürich, Switzerland
Correspondence
Lars M. Blank
[email protected]
Received 16 October 2003
Revised 19 December 2003
Accepted 22 December 2003
Metabolic responses of Saccharomyces cerevisiae to different physical and chemical
environmental conditions were investigated in glucose batch culture by GC-MS-detected
mass isotopomer distributions in proteinogenic amino acids from 13C-labelling experiments.
For this purpose, GC-MS-based metabolic flux ratio analysis was extended from bacteria to the
compartmentalized metabolism of S. cerevisiae. Generally, S. cerevisiae was shown to have low
catabolic fluxes through the pentose phosphate pathway and the tricarboxylic acid (TCA) cycle.
Notably, respiratory TCA cycle fluxes exhibited a strong correlation with the maximum specific
growth rate that was attained under different environmental conditions, including a wide range
of pH, osmolarity, decoupler and salt concentrations, but not temperature. At pH values of 4?0
to 6?0 with near-maximum growth rates, the TCA cycle operated as a bifurcated pathway to
fulfil exclusively biosynthetic functions. Increasing or decreasing the pH beyond this physiologically
optimal range, however, reduced growth and glucose uptake rates but increased the ‘cyclic’
respiratory mode of TCA cycle operation for catabolism. Thus, the results indicate that glucose
repression of the TCA cycle is regulated by the rates of growth or glucose uptake, or signals
derived from these. While sensing of extracellular glucose concentrations has a general influence
on the in vivo TCA cycle activity, the growth-rate-dependent increase in respiratory TCA cycle
activity was independent of glucose sensing.
INTRODUCTION
Genome-wide mRNA responses of the baker’s yeast
Saccharomyces cerevisiae to changes in pH, osmolarity or
temperature revealed differential expression of more than
1000 transcripts during adaptation (Causton et al., 2001;
Gasch et al., 2000). Since transcriptome or proteome
changes do not directly reveal cellular phenotypes, one
would like to connect these inventory data with the apparent cellular physiology (Bailey, 1999). One such approach
is metabolic flux analysis, which estimates material flow
through biochemical reaction networks, and thus provides
a direct link to the physiological phenotype (Hellerstein,
2003).
Different approaches for metabolic flux analysis based on
13
C-labelling experiments have been developed, allowing
precise quantification of central carbon metabolism (Sauer,
2004; Wiechert, 2001). Recent applications include the
bacteria Bacillus subtilis (Dauner et al., 2001; Zamboni &
Sauer, 2003), Corynebacterium glutamicum (Klapa et al.,
2003; Petersen et al., 2000; Wittmann & Heinzle, 2002) and
Abbreviations: MDV, mass distribution vector; METAFoR, metabolic
flux ratio; PP, pentose phosphate; TCA, tricarboxylic acid (for other
abbreviations see the legend of Fig. 1).
0002-6845 G 2004 SGM
Escherichia coli (Emmerling et al., 2002; Fischer & Sauer,
2003b; Jiao et al., 2003; Sauer et al., 2004). Although
conceptually more difficult, flux analysis has also been
applied successfully to compartmentalized microbes such
as S. cerevisiae (Christensen et al., 2002; Dos Santos et al.,
2003), Saccharomyces kluyveri (Möller et al., 2002), Kluyveromyces marxianus (Wittmann et al., 2002b) and Penicillium
chrysogenum (Van Winden et al., 2003). Often, metabolic
flux analysis combines 13C-labelling data with quantitative
physiology data to obtain a best-fit flux solution. A somewhat
different methodology is metabolic flux ratio (METAFoR)
analysis, which quantifies the relative contribution of converging pathways or reactions to a given intracellular metabolite (Fischer & Sauer, 2003a). Without data fitting, this
biochemical approach relies exclusively on 13C data and
has been used successfully with NMR data in the yeasts
S. cerevisiae (Maaheimo et al., 2001) and Pichia stipitis (Fiaux
et al., 2003).
Here, we extend METAFoR analysis by GC-MS from E. coli
(Fischer & Sauer, 2003a) to the compartmentalized S.
cerevisiae metabolism. The particular focus of this study
was to investigate the impact of different environmental
conditions such as pH, osmolarity and temperature on the
central carbon metabolism of S. cerevisiae during growth
on glucose.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 07:19:58
Printed in Great Britain
1085
L. M. Blank and U. Sauer
METHODS
Yeast strains and growth conditions. The haploid S. cerevisiae
strains used in this study are listed in Table 1. The hxk2, rgt2 and
snf3 mutants were constructed by using existing kanMX4 cassettes
(Winzeler et al., 1999). The cassettes were amplified using primers
located ~500 bp upstream and downstream of the start and stop
codon of the corresponding gene, respectively. Batch cultures of
20 ml were performed in 500 ml baffled Erlenmeyer flasks on a
rotary shaker at 30 uC and 225 r.p.m., ensuring fully aerated conditions, in yeast minimal medium (Verduyn et al., 1992) containing
(per litre) 5 g (NH4)2SO4, 3 g KH2PO4, 0?5 g MgSO4.7H2O,
4?5 mg ZnSO4.7H2O, 0?3 mg CoCl2.6H2O, 1?0 mg MnCl2.4H2O,
0?3 mg CuSO4.5H2O, 4?5 mg CaCl2.2H2O, 3?0 mg FeSO4.7H2O,
0?4 mg NaMoO4.2H2O, 1?0 mg H3BO3, 0?1 mg KI, 15 mg EDTA,
0?05 mg biotin, 1?0 mg calcium pantothenate, 1?0 mg nicotinic
acid, 25 mg inositol, 1?0 mg pyridoxine, 0?2 mg p-aminobenzoic
acid and 1?0 mg thiamin. To avoid pH changes due to ammonia
uptake and acetate production the medium was supplemented with
100 mM potassium hydrogen phthalate. For all experiments at pH
values of 6 or higher, 100 mM MOPS was used as the buffering
agent instead. The final pH of the cultures was in all experiments
within 0?05 of the pre-inoculation value. Filter-sterilized glucose was
added prior to the experiment at a final concentration of 5 g l21.
Analytical procedures and physiological parameters. Cell
growth was followed by monitoring the OD600. Samples for extracellular metabolite determination were centrifuged at 14 000 r.p.m.
in an Eppendorf tabletop centrifuge to remove cells. Glucose and
glycerol concentrations in the supernatant were determined with
commercial enzymic kits [Triglyceride 337-40A, Sigma; D-Glucose
Test (E0716251), R-Biopharm AG]. The physiological parameters
maximum specific growth rate, biomass yield on glucose, and specific glucose consumption rate were calculated during the exponential
growth phase (Sauer et al., 1999).
13
C-labelling experiments. All labelling experiments were done in
batch cultures assuming pseudo-steady-state conditions during the
exponential growth phase (Fischer & Sauer, 2003a; Sauer et al.,
1999; Wittmann & Heinzle, 2001). 13C-labelling of proteinogenic
amino acids was achieved by growth on 5 g glucose l21 as a mixture
of 80 % (w/w) unlabelled and 20 % (w/w) uniformly labelled
[U-13C]glucose (13C, >98 %; Isotech). Cells from an overnight
minimal medium culture were washed and used for inoculation
below an OD600 of 0?03. 13C-labelled biomass aliquots were harvested during the mid-exponential growth phase at an OD600 of ¡1.
At this point the residual glucose concentration was between 1
and 3 g l21, thus clearly above the reported 0?5 g l21 threshold of
invertase repression in S. cerevisiae strain CEN.PK 113-7D (Herwig
et al., 2001). The cells were harvested by centrifugation, washed once
with sterile water, and hydrolysed in 500 ml 6 M HCl at 105 uC for
24 h. The hydrolysate was dried in a heating block at 80 uC
under a constant airflow. The free amino acids were derivatized
at 85 uC for 1 h using 25 ml dimethylformamide and 25 ml
N-(tert-butyldimethylsilyl)-N-methyltrifluoroacetamide (Dauner &
Sauer, 2000; Wittmann et al., 2002a).
GC-MS analysis was carried out as reported recently (Fischer & Sauer,
2003a) using a previously described biochemical reaction network
(Maaheimo et al., 2001). The recently described cytosolic alanine
synthesis in glucose/acetate co-metabolism experiments was not seen
in our glucose experiments (Dos Santos et al., 2003). The labelling
pattern of phosphoenolpyruvate (PEP) derived from tyrosine and
phenylalanine was different from the labelling pattern of mitochondrial
pyruvate referred from valine. The labelling patterns of alanine and
valine, however, were highly similar, suggesting that alanine is
indeed synthesized from mitochondrial pyruvate in the experimental
conditions used in this study.
METAFoR analysis using amino acids mass isotopomer
data. The GC-MS data represent sets of ion clusters, each showing
the distribution of mass isotopomers of a given amino acid fragment. For each fragment a, one mass isotopomer distribution vector
(MDV) was assigned,
3
2
(m0 )
7
6
6 (m1 ) 7
X
7
6
6
mi ~1
ð1Þ
MDVa ~6 (m2 ) 7
7with
7
6
4 ... 5
(mn )
with m0 being the fractional abundance of the lowest mass and
mi>0 the abundances of molecules with higher masses. To obtain
the exclusive mass isotope distribution of the carbon skeleton,
corrections for naturally occurring isotopes in the derivatization
reagent and the amino acids were performed as described previously
(Fischer & Sauer, 2003a), followed by the calculations of the mass
distribution vectors for amino acids (MDVAA) and metabolites
(MDVM). Metabolic flux ratios were calculated from the MDVM as
described for E. coli (Fischer & Sauer, 2003a) with the following
exceptions, which account for the compartmentalized biochemical
reaction network of S. cerevisiae. Since the mitochondrial aspartate
aminotransferase is probably inactive under the conditions used
(Maaheimo et al., 2001), the mass distribution of mitochondrial
oxaloacetate was not directly accessible (Fig. 1). Instead, the MDV
of the four-carbon C-1–C-2–C-3–C-4 fragment of mitochondrial
oxaloacetate (OAA14mit) (see Fig. 1 for abbreviations) was accessed
by assuming that only anaplerosis and the TCA cycle contribute to
its labelling pattern. Thus, the MDVM of OAA14mit can be calculated
Table 1. Yeast strains used in this work
Strain
CEN.PK113-7D
LMB45
LMB179
LMB243
CEN.PK513-3A
MMB4
Genotype
Source or reference
MATa MAL2-8 SUC2
CEN.PK 113-7D hxk2 : : kanMX4
CEN.PK 113-7D snf3 : : kanMX4
CEN.PK 113-7D rgt2 : : kanMX4
CEN.PK 113-7D grr1 : : loxP-kanMX4-loxP
W303-1A MATa ade2-1 can1-100 his3-11,15 leu2-3,112 trp1-1
ura3-1 snf3 : : HIS3 rgt2 : : leu2 mth1 : : trp1 gpr1 : : kanMX4
Euroscarf*
This study
This study
This study
P. Kötter, pers. comm.
J. M. Gancedo, pers. comm.
c
*European Saccharomyces cerevisiae Archive for Functional Analysis ([email protected]).
1086
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 07:19:58
Microbiology 150
TCA cycle activity is a function of growth rate
OAA23mit+acetyl-CoA12 and OGA12 equals OAA34mit. The relative
contribution of the TCA cycle to OAA synthesis can then also be
quantified by:
OAA23mit {(GLU1U |GLU1U )
OAAmit
~1{
PEP23{(GLU1U |GLU1U )
through TCA cycle
ð4Þ
Since the independent equations 3 and 4 gave very similar results in
all analyses described here, only those obtained from equation 4 are
reported in the following.
The relative contribution of different synthesis pathways to mitochondrial and cytosolic acetyl-CoA could not be quantified, because the
MDVAA of LEU12 was not accessible with the derivatizing agent used
and LYS12 was not accessible with the procedure employed. Consequently, the MDVM of cytosolic acetyl-CoA was not available for
comparison with the calculated mitochondrial acetyl-CoA MDVM
from OGA.
From the MDVs of oxaloacetate one can calculate the cytosolic OAAcyt
derived from cytosolic pyruvate:
OAAcyt from pyruvatecyt ~
OAA24cyt {OAA24mit
PEP13{OAA24mit
ð5Þ
Furthermore one can determine the activities of the reversible exchange
fluxes between oxaloacetate and fumarate or succinate:
Fig. 1. Central carbon metabolism of S. cerevisiae during aerobic growth on glucose. Arrows with solid heads indicate the
reaction reversibilities considered. Arrows with open heads
highlight the biosynthesis pathways of amino acids that allow
for direct conclusions on the 13C-labelling pattern of their
precursor molecules. The superscripts cyt and mit refer to the
cytosolic and mitochondrial species of metabolites that are
considered in METAFoR analysis. Extracellular substrates and
products are given in capitals. Abbreviations: C1, one-carbon
unit from C1 metabolism; E4P, erythrose 4-phosphate; OAA,
oxaloacetate; OGA, 2-oxoglutarate; PEP, phosphoenolpyruvate;
P5P, pentose 5-phosphates; S7P, sedoheptulose 7-phosphate.
OAA14mit ~
OAA24mit {pyruvate23mit |MDVCO2
(0:5 pyruvate13mit z0:5 (pyruvate23mit |MDVCO2 ){pyruvate23mit |MDVCO2
ð6Þ
OAAcyt rev: from fumaratecyt ~
OAA24cyt {PEP23|MDVCO2
(0:5 PEP13z0:5 (PEP23|MDVCO2 ){PEP23|MDVCO2
ð7Þ
Since the malic enzyme is located in the mitochondria (Boles et al.,
1998), the mitochondrial pools of pyruvate and oxaloacetate are used
to calculate lower and upper bounds of malic enzyme activity:
pyruvate23mit {PEP23
Pyruvatemit from malate
~
(GLU1U |GLU1U ){PEP23
(lower bound)
ð8Þ
Pyruvatemit from malate (upper bound)~
with the following equation:
1{OAAmit
OAAmit rev: from fumaratemit ~
pyruvatemit from malate (lower bound)
!
mit
.(PEP13|MDVCO2 )z
through TCA cycle
OAA
ð9Þ
through TCA cycle
The fraction of OAAmit derived through the TCA cycle is obtained
from:
The contribution of the non-oxidative pentose phosphate (PP)
pathway via transketolase to the synthesis of the triose pool can be
estimated from the MDV of PEP12, revealing the fraction of trioses
rearranged between carbon C-1 and C-2 by the action of the
transketolase. The remaining PEP12 molecules originate from an
unbroken C2 unit of glucose derived through glycolysis.
OAAmit through TCA cycle~
PEP through transketolase~
OAAmit
ð2Þ
!
.OGA25
through TCA cycle
OGA25{(GLU2U |GLU1U |GLU1U )
1{
(GLU2U |GLU2U ){(GLU2U |GLU1U |GLU1U )
ð3Þ
13
where GLU1U and GLU2U are uniformly C-labelled 1- and 2carbon glucose fragments, respectively. This approach assumes that
the bond between carbon atoms 2 and 3 of every oxaloacetate is
broken per round through the TCA cycle.
mit
A second independent equation for calculating the fraction of OAA
through the TCA cycle can be formulated by assuming that OGA15
corresponds to OAA24mit+acetyl-CoA12. Thus, OGA25 equals
http://mic.sgmjournals.org
PEP12{GLU2U
GLU1U |GLU1U {GLU2U
ð10Þ
An upper bound for the contribution of the PP pathway to trioses
synthesis can than be calculated by assuming that five trioses are
produced from three pentoses and that at least two of these trioses are
rearranged by a transketolase:
PEP through PP pathway~5=2?PEP through transketolase
ð11Þ
Since the contribution of the PP pathway to PEP synthesis is an upper
bound, we present the results from equation 11 by including the error
interval.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 07:19:58
1087
L. M. Blank and U. Sauer
ND
ND
Fig. 2. Origin of metabolic intermediates in S. cerevisiae
during growth in batch cultures at 30 6C and pH 5?5 determined by NMR (grey bars) and GC-MS (white bars). The
NMR-based data were taken from Maaheimo et al. (2001). The
standard deviations were estimated from redundant mass distributions. ND, Not determined.
RESULTS
METAFoR analysis by GC-MS of S. cerevisiae
in glucose batch culture
To establish a rapid and robust methodology for flux
analysis, we extended METAFoR analysis by GC-MS from
bacterial metabolism (Fischer & Sauer, 2003a; Fischer et al.,
2004; Zamboni & Sauer, 2003) to the compartmentalized
metabolism of S. cerevisiae. Firstly, we found the mass
isotopomer distributions in the GC-MS-analysed proteinogenic amino acids to be generally consistent with previous
network analysis (Christensen et al., 2002; Maaheimo et al.,
2001; Van Winden, 2002) of amino acid biosynthesis in
S. cerevisiae (data not shown). Next, we compared GCMS-based METAFoR analysis of a shake-flask-grown
S. cerevisiae batch culture on glucose at 30 uC and pH 5?5
to NMR-based METAFoR analysis under almost identical
conditions (Maaheimo et al., 2001). Indeed, very similar
results such as low PP pathway and TCA cycle activities
were found (Fig. 2). Since the interpretation of 13Clabelling pattern in amino acids by MS and NMR analyses
is only connected by the assumed biochemical network,
this consistency provides evidence for faithful and reliable
quantification of metabolic flux ratios by either method.
Minor differences in flux ratios may be attributed to the
different minimal medium composition and to the somewhat lower growth rate of the NMR-analysed culture
(Maaheimo et al., 2001).
Impact of pH and temperature on the metabolic
flux profile
To investigate the influence of physical and chemical
environmental parameters on central carbon metabolism
1088
under aerobic fermentation conditions, we grew S. cerevisiae
at 25, 30 and 37 uC and at pH values of 3?5, 5?0 and 6?0.
The 16 determined metabolic flux ratios were surprisingly
stable under all conditions, and the relative fluxes through
the TCA cycle and PP pathway remained rather low (Fig. 3).
Generally, low TCA cycle activity and concomitant ethanol
production were expected for the respiro-fermentative
S. cerevisiae (Alexander & Jeffries, 1990). Previous estimates
for the relative catabolic PP pathway flux to trioses vary
between 0 and 4 % (Fiaux et al., 2003; Maaheimo et al.,
2001) and 14 % (Gombert et al., 2001). In the experiments
described here, we observe an upper bound of 7 % on the
fraction of PEP derived through the PP pathway from
uniformly labelled glucose experiments (Fig. 3). Thus, we
conclude that the catabolic PP pathway activity of S.
cerevisiae in batch cultures is indeed lower than seen for
example in E. coli (Fischer & Sauer, 2003a).
The growth rate of S. cerevisiae correlates with
the relative respiratory TCA cycle flux in batch
culture
The increased fraction of OAAmit derived through the TCA
cycle in the pH 3?5 culture indicates significantly increased
respiratory TCA cycle activity, when compared to the
biosynthetic TCA cycle activity (Fig. 3b). Since this low
pH also reduced the maximum specific growth rate by 10 %,
we could not distinguish between a growth rate and an
environmental condition effect on the TCA cycle activity.
To quantify the influence of environmental conditions
on the maximal specific growth rate and intracellular flux
response, we grew batch cultures under aerobic fermentation conditions, with extensive ethanol formation at
different pH values, osmolarities, temperatures, salt and
decoupler concentrations. As shown in Fig. 4, the contribution of the TCA cycle to the synthesis of OAAmit generally
increased with decreasing growth rate. This observation
appeared to be independent of the wide variety of environmental conditions that were used to reduce the growth
rate. The sole exception was temperature, which had only a
modest effect on TCA cycle fluxes (Figs 3 and 4), although
the growth rate was significantly reduced. Since the specific
glucose uptake rate is closely correlated with the specific
growth rate (Diderich et al., 1999; Van Hoek et al., 1998)
(Fig. 5a, b), it is not surprising that the glucose uptake
rate was likewise correlated with the relative TCA cycle
flux (data not shown). While we cannot exclude that the
TCA cycle activity was directly influenced by the environmental conditions chosen, the wide range of conditions
makes it rather unlikely that all of them have a similar
and specific effect on the TCA cycle flux. Moreover, it was
shown recently that the TCA cycle activity of C. glutamicum
does not change with the osmolarity of the medium (Varela
et al., 2003).
The OAAcyt pool that is required for anaplerosis of the
TCA cycle and biosynthesis of amino acids of the aspartate
family was synthesized exclusively via pyruvate carboxylase
(OAAcyt from pyruvatecyt) in most experiments. Only at
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 07:19:58
Microbiology 150
TCA cycle activity is a function of growth rate
(a)
ND
ND
(b)
ND
ND
growth rates below 0?2 h21 may a significant fraction of
7–22 % originate from the OAAmit pool, probably catalysed
by the action of the mitochondrial OAA transporter Oac1p,
since transport was shown to be bidirectional (Palmieri et al.,
OAAmit through TCA cycle
0.5
0.4
0.3
0.2
0.1
0.1
0.2
0.3
0.4
_1
Maximum specific growth rate (h )
0.5
Fig. 4. The fraction of OAAmit derived through the TCA cycle
as a function of the maximum specific growth rate of glucosegrown S. cerevisiae under different environmental conditions in
batch culture. The standard deviation of the TCA cycle activity
was estimated from redundant mass distributions and the
growth rate error was estimated using SigmaPlot 8.0 (SPSS).
DNP, 2,4-dinitrophenol.
http://mic.sgmjournals.org
Fig. 3. Origin of metabolic intermediates in
S. cerevisiae during growth in batch cultures. (a) At temperatures of 25 6C (black
bars), 30 6C (white bars) and 37 6C (grey
bars), at pH 5?0; (b) at pH values of 3?5
(black bars), 5?0 (white bars) and 6?0 (grey
bars), at 30 6C. The standard deviations
were estimated from redundant mass distributions. Malic enzyme activity could only be
determined in cultures with less than 5 %
TCA cycle activity, and was labelled not
determined (ND) in all other cases.
1999). This flux ratio change was also independent of the
environmental condition that reduced the growth rate.
In contrast to the above two flux ratios, however, none of
the 10 other detected flux ratios exhibited a distinct correlation with the growth rate (data not shown). The malic
enzyme for example was more active at acidic pH, but was
inactive at pH 7?5, although the growth rate was higher
(0?19 h21) than at pH 3?0 (0?07 h21). The reverse in vivo
activity pattern was determined for the gluconeogenic
reaction catalysed by PEP carboxykinase, which was only
detected at pH values 7?0 and 7?5.
To elucidate whether the growth rate was also correlated
with any other physiological property, we quantified the
growth physiology over a wide range of pH values (Fig. 5).
For standard conditions (high glucose concentration, 30 uC,
aerated batch culture, pH 5–6), these parameters (Fig. 5)
were in good agreement with previous reports for S.
cerevisiae CEN.PK strains (Smits et al., 2000; van Dijken
et al., 2000; van Maris et al., 2001). Akin to the maximum
growth rate (Fig. 5a), the specific glucose uptake rate
decreased significantly outside the optimal pH range of
4?0–6?0 (Fig. 5b). The biomass yield, in contrast, was rather
constant and exhibited no correlation with the growth rate
(Fig. 5c). The specific glycerol production rate showed a
generally positive correlation with the culture pH, i.e.
resulted in the highest production rate at pH 7?5 (Fig. 5d).
This positive correlation may reflect a similar response as
was reported for high osmolarity, where production of the
main osmolyte glycerol was increased in yeast (Hohmann,
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 07:19:58
1089
L. M. Blank and U. Sauer
(a)
(b)
(c)
(d)
Fig. 5. Influence of extracellular pH on maximum specific growth rate (a), specific glucose uptake rate (b), biomass yield
(c) and specific glycerol production rate (d) in batch cultures of S. cerevisiae. Standard deviations were estimated using
SigmaPlot 8.0 (SPSS) and the Gaussian law of error propagation.
2002). It should be noted that significant ethanol production was not reported because the high evaporation rate in
the shake flasks compromised a thorough quantitative
analysis.
Glucose sensing is not required for the increase
of relative TCA cycle flux during slow growth
Although the initial glucose concentrations were identical
in all the above experiments, we cannot exclude that
glucose sensing was relevant for modulating the relative
TCA cycle flux. We therefore constructed isogenic mutants
of the two glucose sensors Snf3p and Rgt2p (Ozcan et al.,
1996). If extracellular glucose sensing played a significant
role in modulating the TCA cycle flux, one would expect
no increase in flux upon an environmentally reduced
growth rate in both mutants. However, the TCA cycle
activity increased in both mutants at lower growth rates
(Fig. 6). To further exclude the possibility of a partial
functional overlap of the high and low glucose concentration sensors, Rgt2p and Snf3p, respectively, we used two
further mutants. Strain MMB4 cannot sense extracellular
glucose at all due to deletions in SNF3, RGT2, and the
plasma membrane G-protein coupled sensor GPR1. To
enable growth on glucose, this strain also carries the MTH1
deletion. The second strain was deleted in GRR1, which
is essential for glucose sensing from the Snf3p and Rgt2p
signal transduction pathway. Although both strains already
exhibit high respiratory TCA cycle flux under standard
conditions at growth rates of 0?39 and 0?26 h21 for MMB4
and the grr1 mutant, respectively, they still increased the
1090
relative TCA cycle flux upon an environmental modulation
of the specific growth rate with high osmolarity (Fig. 6).
These results strongly suggest that extracellular glucose
sensing is not involved in the relative TCA cycle flux increase
at slow growth rates in batch cultures. Nevertheless, the
higher TCA cycle flux in the grr1 and MMB4 mutants, when
Fig. 6. The fraction of OAAmit derived through the TCA cycle
in S. cerevisiae mutants impaired in glucose sensing. The
values were determined under standard batch conditions (white
bars) and at 0?75 M NaCl (grey bars), which reduced the
maximal specific growth rate by 40–60 %. The experimental
error of the TCA cycle activity was estimated from redundant
mass distributions.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 07:19:58
Microbiology 150
TCA cycle activity is a function of growth rate
compared to the wild-type at the same growth rate, shows
that glucose sensing has an additional repressive effect.
Is the TCA cycle flux correlated with the rate of
growth or glucose uptake?
The specific rates of growth and glucose uptake were
coupled in the batch experiments performed; thus we
cannot distinguish if one or both rates were correlated with
the relative TCA cycle flux. To address this question, we
used a hexokinase II-deficient mutant that exhibits alleviated glucose repression in glucose batch cultures (Diderich
et al., 2001). When compared to the wild-type, the specific
growth rate of the hxk2 mutant was only modestly lower,
but the specific glucose uptake rate was about 45 % lower.
The relative TCA cycle flux in this mutant exhibits no
correlation with the growth rate but a much better, albeit
weak, correlation with the glucose uptake rate (Fig. 7). This
would suggest that it is the specific glucose uptake rate
rather than the specific growth rate that correlates with the
relative TCA cycle flux in wild-type S. cerevisiae.
DISCUSSION
Using the newly developed GC-MS-based METAFoR analysis, we quantified intracellular flux responses of S. cerevisiae
to a wide range of environmental conditions in batch
culture. In contrast to all other monitored flux ratios, the
_
Specific growth rate (h 1)
.
01
0.2
0.3
OAAmit through TCA cycle
0.6
0.4
0.5
0.5
0.4
0.3
0.2
0.1
0
2
4
6
8
10
12
14
_1
Specific glucose uptake rate (mmol g
16
_1
18
h )
Fig. 7. The fraction of OAAmit derived through the TCA cycle
as a function of the maximum specific rates of growth ($) and
glucose uptake (#) in S. cerevisiae batch cultures with different pH (see also Fig. 4). Results from two batch experiments in
different media and in addition one experiment with the hxk2
mutant in standard medium are shown. The experimental error
of the TCA cycle activity was estimated from redundant mass
distributions and the growth and glucose uptake rates errors
were estimated using SigmaPlot 8.0 (SPSS).
http://mic.sgmjournals.org
relative respiratory activity of the TCA cycle increased with
decreasing growth rate and/or glucose uptake rate at extracellular glucose concentrations between 1 and 5 g l21. This
correlation was independent of the four different chemical
parameters that were used to modulate the growth rate.
As the sole exception, temperature-induced growth rate
changes gave a much less pronounced TCA cycle response.
The correlation observed here between the rates of growth
and/or glucose uptake and relative respiratory TCA cycle
flux contrasts with the generally held view of a cataboliterepressed TCA cycle in glucose-excess batch cultures of S.
cerevisiae (Gancedo, 1998; Rolland et al., 2002). Under
standard batch conditions, the TCA cycle operates as a
bifurcated pathway to sustain biomass precursor requirements (Gombert et al., 2001). Although expression of the
majority of the TCA cycle genes is subject to glucose
repression (DeRisi et al., 1997) at extracellular glucose
concentrations that may be as low 0?1 g l21 (Yin et al.,
2003), we show here that the relative in vivo respiratory
activity of the TCA cycle may increase even at high
glucose concentrations, provided the growth rate or the
glucose uptake rate are impaired by other environmental
parameters.
While the major regulation pathways of glucose repression are known (Rolland et al., 2002), the molecular
mechanisms that initiate repression are still elusive and
several metabolism-derived triggers have been discussed
(Carlson, 1999; Gancedo, 1998; Rolland et al., 2002). Our
glucose sensor mutant results exclude that glucose repression of the TCA cycle is exclusively mediated by sensing
of extracellular glucose concentrations, which is known
to repress several HXT genes (Rolland et al., 2002). The
relative TCA cycle flux increase in four different mutants
impaired in glucose sensing strongly suggests that the
metabolic trigger for TCA cycle repression must be an
intracellular, metabolism-derived signal. This view is consistent with increasing oxygen consumption rates upon
genetic reduction of growth rate and glucose uptake rate
(Ye et al., 1999). In addition, there is the concentrationdependent repression because mutants completely devoid
of glucose sensing (e.g. grr1 and MMB4) exhibited higher
TCA cycle fluxes than would be expected from their growth
rates.
Generally, one would expect the repression signal to be
related to the glucose uptake rate rather than to the growth
rate, but the two were coupled in all environmental
modulations. In the hxk2 mutant, however, the two
parameters were decoupled and the relative TCA-cycle
flux correlated better with uptake than with growth rate,
thus providing some evidence for a flux-related signal of
glucose repression of the TCA cycle. The imperfect uptake–
TCA cycle correlation in the hxk2 mutant, with a weaker
repression of the TCA cycle than expected from pH
experiments with similar growth rates, could be related to
the regulatory role of Hxk2p in glucose repression, which
would also influence the flux (Carlson, 1999; Gancedo,
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 07:19:58
1091
L. M. Blank and U. Sauer
1998; Rolland et al., 2002). Apparently, glucose repression
of the TCA cycle exhibits a different pattern and probably also uses different signals than the paradigm glucose
repression gene SUC2 (Meijer et al., 1998; Rolland et al.,
2002). The general dependence of relative TCA cycle fluxes
on the exact environmental conditions may also explain
minor differences in TCA cycle activity obtained from
previous 13C-labelling experiments (Christensen et al.,
2002; Fiaux et al., 2003; Gombert et al., 2001; Maaheimo
et al., 2001).
The methodology described for metabolic flux profiling
based on GC-MS data of proteinogenic amino acids in yeast
is robust, rapid, and also applicable to mini-cultivation
systems. Most of the reported flux ratios have varied only
within a rather narrow range. Importantly, the fluxes
through the PP pathway, the malic enzyme and the
gluconeogenic reaction catalysed by PEP carboxykinase
are low and change little when considering the severe
physiological impacts of the environmental conditions
used. The sole exceptions were the respiratory TCA cycle
flux and the mitochondrial exchange flux between oxaloacetate and fumarate. This suggests a general robustness of
intracellular metabolism to the environmental conditions
on a given substrate. While the rate at which glucose enters
the cell may vary over a wide range, the relative distribution
of carbon fluxes within the cell remains rather stable.
Dauner, M., Bailey, J. E. & Sauer, U. (2001). Metabolic flux
analysis with a comprehensive isotopomer model in Bacillus subtilis.
Biotechnol Bioeng 76, 144–156.
DeRisi, J. L., Iyer, V. R. & Brown, P. O. (1997). Exploring the
metabolic and genetic control of gene expression on a genomic
scale. Science 278, 680–686.
Diderich, J. A., Schepper, M., van Hoek, P. & 8 other authors
(1999). Glucose uptake kinetics and transcription of HXT genes in
chemostat cultures of Saccharomyces cerevisiae. J Biol Chem 274,
15350–15359.
Diderich, J. A., Raamsdonk, L. M., Kruckeberg, A. L., Berden, J. A.
& Van Dam, K. (2001). Physiological properties of Saccharomyces
cerevisiae from which hexokinase II has been deleted. Appl Environ
Microbiol 67, 1587–1593.
Dos Santos, M. M., Gombert, A. K., Christensen, B., Olsson, L. &
Nielsen, J. (2003). Identification of in vivo enzyme activities in the
cometabolism of glucose and acetate by Saccharomyces cerevisiae by
using 13C-labeled substrates. Eukaryot Cell 2, 599–608.
Emmerling, M., Dauner, M., Ponti, A., Fiaux, J., Hochuli, M.,
Szyperski, T., Wüthrich, K., Bailey, J. E. & Sauer, U. (2002).
Metabolic flux responses to pyruvate kinase knockout in Escherichia
coli. J Bacteriol 184, 152–164.
Fiaux, J., Cakar, Z. P., Sonderegger, M., Wüthrich, K., Szyperski, T.
& Sauer, U. (2003). Metabolic-flux profiling of the yeasts Saccharo-
myces cerevisiae and Pichia stipitis. Eukaryot Cell 2, 170–180.
Fischer, E. & Sauer, U. (2003a). Metabolic flux profiling of
Escherichia coli mutants in central carbon metabolism using GC-MS.
Eur J Biochem 270, 880–891.
Fischer, E. & Sauer, U. (2003b). A novel metabolic cycle catalyzes
glucose oxidation and anaplerosis in hungry Escherichia coli. J Biol
Chem 278, 46446–46451.
ACKNOWLEDGEMENTS
Fischer, E., Zamboni, N. & Sauer, U. (2004). High-throughput
The authors wish to thank Juana Maria Gancedo and Peter Kötter
for providing yeast strains. Lars M. Blank gratefully acknowledges
financial support by the Deutsche Akademie der Naturforscher
Leopoldina (BMBF-LPD/8-78).
Gancedo, J. M. (1998). Yeast carbon catabolite repression. Microbiol
metabolic flux analysis based on gas-chromatography-mass spectrometry derived 13C constraints. Anal Biochem (in press).
Mol Biol Rev 62, 334–361.
Gasch, A. P., Spellman, P. T., Kao, C. M., Carmel-Harel, O., Eisen,
M. B., Storz, G., Botstein, D. & Brown, P. O. (2000). Genomic
expression programs in the response of yeast cells to environmental
changes. Mol Biol Cell 11, 4241–4257.
REFERENCES
Alexander, M. A. & Jeffries, T. W. (1990). Respiratory efficiency and
Gombert, A. K., Moreira dos Santos, M., Christensen, B. &
Nielsen, J. (2001). Network identification and flux quantification
metabolite partitioning as regulatory phenomena in yeasts. Enzyme
Microb Technol 12, 2–19.
in the central metabolism of Saccharomyces cerevisiae under different
conditions of glucose repression. J Bacteriol 183, 1441–1451.
Bailey, J. E. (1999). Lessons from metabolic engineering for
Hellerstein, M. K. (2003). In vivo measurement of fluxes through
functional genomics and drug discovery. Nat Biotechnol 17, 616–618.
metabolic pathways: the missing link in functional genomics and
pharmaceutical research. Annu Rev Nutr 23, 379–402.
Boles, E., de Jong-Gubbels, P. & Pronk, J. T. (1998). Identifica-
tion and characterization of MAE1, the Saccharomyces cerevisiae
structural gene encoding mitochondrial malic enzyme. J Bacteriol
180, 2875–2882.
Herwig, C., Doerries, C., Marison, I. & von Stockar, U. (2001).
Quantitative analysis of the regulation scheme of invertase expression in Saccharomyces cerevisiae. Biotechnol Bioeng 76, 247–258.
Carlson, M. (1999). Glucose repression in yeast. Curr Opin Microbiol
Hohmann, S. (2002). Osmotic stress signaling and osmoadaptation
2, 202–207.
in yeasts. Microbiol Mol Biol Rev 66, 300–372.
Causton, H. C., Ren, B., Koh, S. S. & 7 other authors (2001).
Jiao, Z., Baba, T., Mori, H. & Shimizu, K. (2003). Analysis of
Remodeling of yeast genome expression in response to environmental changes. Mol Biol Cell 12, 323–337.
metabolic and physiological responses to gnd knockout in Escherichia
coli by using C-13 tracer experiment and enzyme activity
measurement. FEMS Microbiol Lett 220, 295–301.
Christensen, B., Gombert, A. K. & Nielsen, J. (2002). Analysis of flux
13
C-labelling experiments. Eur J Biochem 269,
Klapa, M. I., Aon, J. C. & Stephanopoulos, G. (2003). Systematic
Dauner, M. & Sauer, U. (2000). GC-MS analysis of amino acids
quantification of complex metabolic flux networks using stable
isotopes and mass spectrometry. Eur J Biochem 270, 3525–3542.
rapidly provides rich information for isotopomer balancing.
Biotechnol Prog 16, 642–669.
Maaheimo, H., Fiaux, J., Cakar, Z. P., Bailey, J. E., Sauer, U. &
Szyperski, T. (2001). Central carbon metabolism of Saccharomyces
estimates based on
2795–2800.
1092
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 07:19:58
Microbiology 150
TCA cycle activity is a function of growth rate
cerevisiae explored by biosynthetic fractional 13C labeling of common
amino acids. Eur J Biochem 268, 2464–2479.
Meijer, M. M., Boonstra, J., Verkleij, A. J. & Verrips, C. T. (1998).
Glucose repression in Saccharomyces cerevisiae is related to the
glucose concentration rather than the glucose flux. J Biol Chem 273,
24102–24107.
the distribution of fluxes among respiration and fermentation by
overexpression of HAP4 in Saccharomyces cerevisiae. FEMS Yeast Res
1, 139–149.
Van Winden, W. (2002). Verifying assumed biosynthetic pathways,
Möller, K., Christensen, B., Förster, J., Piskur, J., Nielsen, J. &
Olsson, L. (2002). Aerobic glucose metabolism of Saccharomyces
metabolic precursors and estimated measurement errors of amino
acids, trehalose and levulinic acid using redundant 2D [13C, 1H]
COSY NMR data. In Department of Bioprocess Technology,
pp. 229–245. Delft: Delft University of Technology.
kluyveri: growth, metabolite production, and quantification of
metabolic fluxes. Biotechnol Bioeng 77, 186–193.
Van Winden, W. A., Van Gulik, W. M., Schipper, D., Verheijen, P. J.,
Krabben, P., Vinke, J. L. & Heijnen, J. J. (2003). Metabolic flux and
Ozcan, S., Dover, J., Rosenwald, A. G., Wolfl, S. & Johnston, M.
(1996). Two glucose transporters in Saccharomyces cerevisiae are
1
glucose sensors that generate a signal for induction of gene
expression. Proc Natl Acad Sci U S A 93, 12428–12432.
Palmieri, L., Vozza, A., Agrimi, G., De Marco, V., Runswick, M. J.,
Palmieri, F. & Walker, J. E. (1999). Identification of the yeast
mitochondrial transporter for oxaloacetate and sulfate. J Biol Chem
274, 22184–22190.
Petersen, S., de Graaf, A. A., Eggeling, L., Möllney, M., Wiechert, W.
& Sahm, H. (2000). In vivo quantification of parallel and bidirec-
metabolic network analysis of Penicillium chrysogenum using 2D [13C,
H] COSY NMR measurements and cumulative bondomer simulation. Biotechnol Bioeng 83, 75–92.
Varela, C., Agosin, E., Baez, M., Klapa, M. & Stephanopoulos, G.
(2003). Metabolic flux redistribution in Corynebacterium glutamicum
in response to osmotic stress. Appl Microbiol Biotechnol 60, 547–555.
Verduyn, C., Postma, E., Scheffers, W. A. & Van Dijken, J. P. (1992).
Effect of benzoic acid on metabolic fluxes in yeasts: a continuousculture study on the regulation of respiration and alcoholic
fermentation. Yeast 8, 501–517.
13
C metabolic flux analysis. Metab Eng 3,
tional fluxes in the anaplerosis of Corynebacterium glutamicum. J Biol
Chem 275, 35932–35941.
Wiechert, W. (2001).
Rolland, F., Winderickx, J. & Thevelein, J. M. (2002). Glucose-sensing
Winzeler, E. A., Shoemaker, D. D., Astromoff, A. & 22 other authors
(1999). Functional characterization of the S. cerevisiae genome by
and -signalling mechanisms in yeast. FEMS Yeast Res 2, 183–201.
195–206.
Sauer, U. (2004). High-throughput phenomics: experimental
methods for mapping fluxomes. Curr Opin Biotechnol 15 (in press).
gene deletion and parallel analysis. Science 285, 901–906.
Sauer, U., Lasko, D. R., Fiaux, J., Hochuli, M., Glaser, R.,
Szyperski, T., Wüthrich, K. & Bailey, J. E. (1999). Metabolic flux
to lysine-producing Corynebacterium glutamicum: a novel approach
for metabolic flux analysis. Eur J Biochem 268, 2441–2455.
ratio analysis of genetic and environmental modulations of Escherichia
coli central carbon metabolism. J Bacteriol 181, 6679–6688.
Wittmann, C. & Heinzle, E. (2002). Genealogy profiling through
Sauer, U., Canonaco, F., Heri, S., Perrenoud, A. & Fischer, E.
(2004). The soluble and membrane-bound transhydrogenases
Wittmann, C. & Heinzle, E. (2001). Application of MALDI-TOF MS
strain improvement by using metabolic network analysis: metabolic
flux genealogy of several generations of lysine-producing corynebacteria. Appl Environ Microbiol 68, 5843–5859.
UdhA and PntAB have divergent functions in NADPH metabolism of Escherichia coli. J Biol Chem (in press). DOI 10.1074/
jbc.M311657200; PMID 14660605.
Wittmann, C., Hans, M. & Heinzle, E. (2002a). In vivo analysis of
Smits, P. H., Hauf, J., Muller, S., Hobley, T. J., Zimmermann, F. K.,
Hahn-Hagerdal, B., Nielsen, J. & Olsson, L. (2000). Simultaneous
Wittmann, C., Hans, M. & Bluemke, W. (2002b). Metabolic
overexpression of enzymes of the lower part of glycolysis can
enhance the fermentative capacity of Saccharomyces cerevisiae. Yeast
16, 1325–1334.
van Dijken, J. P., Bauer, J., Brambilla, L. & 20 other authors (2000).
An interlaboratory comparison of physiological and genetic properties of four Saccharomyces cerevisiae strains. Enzyme Microb Technol
26, 706–714.
Van Hoek, P., Van Dijken, J. P. & Pronk, J. T. (1998). Effect of
specific growth rate on fermentative capacity of baker’s yeast. Appl
Environ Microbiol 64, 4226–4233.
van Maris, A. J., Bakker, B. M., Brandt, M., Boorsma, A., Teixeira de
Mattos, M. J., Grivell, L. A., Pronk, J. T. & Blom, J. (2001). Modulating
http://mic.sgmjournals.org
intracellular amino acid labelings by GC/MS. Anal Biochem 307,
379–382.
physiology of aroma-producing Kluyveromyces marxianus. Yeast
19, 1351–1363.
Ye, L., Kruckeberg, A. L., Berden, J. A. & van Dam, K. (1999).
Growth and glucose repression are controlled by glucose transport
in Saccharomyces cerevisiae cells containing only one glucose transporter. J Bacteriol 181, 4673–4675.
Yin, Z., Wilson, S., Hauser, N. C., Tournu, H., Hoheisel, J. D. &
Brown, A. J. (2003). Glucose triggers different global responses in
yeast, depending on the strength of the signal, and transiently
stabilizes ribosomal protein mRNAs. Mol Microbiol 48, 713–724.
Zamboni, N. & Sauer, U. (2003). Knockout of the high-coupling
cytochrome aa3 oxidase reduces TCA cycle fluxes in Bacillus subtilis.
FEMS Microbiol Lett 226, 121–126.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 17 Jun 2017 07:19:58
1093