OUP user menu

Mixed and diverse metabolic and gene-expression regulation of the glycolytic and fermentative pathways in response to a HXK2 deletion in Saccharomyces cerevisiae

Sergio Rossell, Alexander Lindenbergh, Coen C. van der Weijden, Arthur L. Kruckeberg, Karen van Eunen, Hans V. Westerhoff, Barbara M. Bakker
DOI: http://dx.doi.org/10.1111/j.1567-1364.2007.00282.x 155-164 First published online: 1 February 2008

Abstract

In Saccharomyces cerevisiae the HXK2 gene, which encodes the glycolytic enzyme hexokinase II, is involved in the regulatory mechanism known as ‘glucose repression’. Its deletion leads to fully respiratory growth at high glucose concentrations where the wild type ferments profusely. Here we describe that deletion of the HXK2 gene resulted in a 75% reduction in fermentative capacity. Using regulation analysis we found that the fluxes through most glycolytic and fermentative enzymes were regulated cooperatively by changes in their capacities (Vmax) and by changes in the way they interacted with the rest of the metabolism. Glucose transport and phosphofructokinase were regulated purely at the metabolic level. The reduction of fermentative capacity in the mutant was accompanied by a remarkable resilience of the remaining capacity to nutrient starvation. After starvation, the fermentative capacity of the hxk2Δ mutant was similar to that of the wild type. Based on our results and previous reports, we suggest an inverse correlation between glucose repression and the resilience of fermentative capacity towards nutrient starvation. Only a limited number of glycolytic enzyme activities changed upon starvation of the hxk2Δ mutant and we discuss to what extent this could explain the stability of the fermentative capacity.

Keywords
  • glycolysis
  • regulation analysis
  • hexokinase II
  • fermentative capacity
  • starvation

Introduction

Glucose-limited cultures of the yeast Saccharomyces cerevisiae show different physiological characteristics and gene expression profiles as compared to cultures grown at glucose-excess. While glucose-limited cultures are characterized by a fully respiratory metabolism, glucose-excess cultures show a mixed respiro-fermentative metabolism (e.g. van Hoek et al., 2000). Glucose-repressed genes include genes encoding proteins involved in respiration, gluconeogenesis, the glyoxylate cycle, high-affinity glucose transport and the utilization of alternative carbon sources, as well as a large group of stress-response-element (STRE) controlled genes. On the other hand, the genes of some glycolytic enzymes are induced by high glucose concentrations (reviewed in Rolland et al., 2002).

In order to achieve a high biomass yield during bakers' yeast production, a completely respiratory metabolism is required. Fermentation (i.e. ethanol production) is avoided during the production phase by ensuring a high aeration rate and a low glucose influx. At the same time, fermentative capacity is an important quality parameter in the bakers' yeast industry. It is defined as the specific rate of carbon dioxide production under sugar-excess and anaerobic conditions (van Hoek et al., 1998). Cells harvested from glucose-limited cultures have a lower fermentative capacity than cells grown under glucose-excess conditions (van Hoek et al., 1998, 2000). Fermentative capacity has been reported to decrease in response to nutrient starvation (Nilsson et al., 2001b). The severity of this detrimental effect depended on the type of starvation (e.g. starvation for nitrogen vs. that for carbon) and on the physiological state of the cultures prior to starvation. Postdiauxic shift cultures respiring ethanol preserved their fermentative capacity better than respiro-fermentative cultures growing on glucose when challenged by nutrient starvation (Nilsson et al., 2001a, b). Nutrient starvation is a relevant phenomenon for the bakers' yeast industry, since at the final stages of production and during storage, cells are starved. In general, nutrient starvation is perhaps one of the most common stresses experienced by microorganisms.

Many of the adaptations to glucose excess conditions are regulated through the ‘main glucose repression pathway’. This signaling pathway senses the concentration of extracellular glucose and transmits this information to the transcription apparatus. Deletion of the HXK2 gene, which encodes the glucose phosphorylating enzyme hexokinase II, alleviated glucose repression (Zimmermann & Scheel, 1977; Entian & Zimmermann, 1980; Michels & Romanowski, 1980) as evidenced by fully respiratory growth at high glucose concentrations (Diderich et al., 2001), coconsumption of glucose with other sugars (Raamsdonk et al., 2001), derepression of high affinity hexose transporters (Petit et al., 2000) and increased plasma-membrane H+-ATPase activity (Diderich et al., 2001).

Although many components of the main glucose repression pathway are known and a sensing role of hexokinase II has been proposed, the mechanism through which hexokinase II triggers glucose repression is not fully understood. Although early studies suggested a correlation between glucose-phosphorylating capacity and glucose repression (Ma et al., 1989; Rose et al., 1991), specific point-mutations in the HXK2 gene had differential effects on phosphorylating capacity and glucose repression (Hohmann et al., 1999; Kraakman et al., 1999). Deletion of HXK1, encoding hexokinase I, in an hxk2Δ mutant further alleviated glucose repression while its overexpression restored it to some extent. However, overexpression of GLK1, encoding glucokinase (which also catalyzes glucose phosphorylation) in a hxk2Δ mutant, had no effect (Rose et al., 1991). Evidently, glucose-phosphorylating capacity is not the sole determinant of glucose repression. In addition, hexokinase II was reported to reside partly in the nucleus (Randez-Gil et al., 1998). This nuclear localization was shown to depend upon Mig1 (Ahuatzi et al., 2004), a transcription factor responsible for the repression of many glucose-repressible genes (Ostling & Ronne, 1998; DeVit & Johnston, 1999). These findings led to the suggestion that hexokinase II forms a repressor complex with Mig1 that is located in the nucleus during growth on glucose (Ahuatzi et al., 2004).

Most studies about the respiration-fermentation switch of yeast focus on the transcriptional regulation of the respiratory pathway. However, recent studies suggest that in yeast, gene expression and in particular transcription correlates poorly with glycolytic enzyme capacities and fluxes (Daran-Lapujade et al., 2004) and is often a minor component of the local regulation of glycolytic enzyme rates (Rossell et al., 2006; Daran-Lapujade, unpublished data). These studies showed that metabolic regulation of enzyme rates was just as important, or even more important, than their regulation by enzyme capacity changes. In the present study, we investigated the alleviation of glucose repression by the deletion of HXK2 from the point of view of the regulation of glycolytic and fermentative fluxes. Rather than focusing on transcriptional regulation alone, we investigated the extent to which the rates of glycolytic and fermentative enzymes were regulated by changes in enzyme capacities (brought about by the hierarchical cascade of gene expression) and the extent to which they were regulated through metabolic interactions. To this end we used regulation analysis (ter Kuile & Westerhoff, 2001). Its idea is as follows. Because enzymes are catalysts (and not substrates), enzyme rate equations are usually of the form: Embedded Image 1

in which v is the rate, f is a function of e, which is the concentration of the enzyme catalyzing the reaction, and g is a function of X and K, in which X is a vector of concentrations of substrates, products and other metabolic effectors, and K is a vector of constants parametrizing the strength with which the enzymes interact with their substrates, products and allosteric effectors. The important characteristic of the above equation is that f does not depend upon X and K, and g does not depend upon e. f(e) describes the dependency of the rate upon the enzyme concentration and can be taken to equal Vmax. Alterations of g(X,K) are regulated through the interaction of the enzyme with the rest of metabolism. Alterations of f reflect the regulation of the capacity of the enzyme of interest, brought about exclusively through gene expression. Stable covalent modification of the enzyme also falls in this category. The dissection and quantification of f and g is achieved by projecting Equation 1 into logarithmic space, considering a change between two steady states, and dividing both sides of the equation by the relative change in steady state flux J. Because at steady state the flux J equals the enzyme rate v, this results in: Embedded Image 2

Here ρh is the ‘hierarchical regulation coefficient’, quantifying the relative contribution of changes in enzyme capacity (Vmax) to the regulation of the enzyme's flux. ρm is the ‘metabolic regulation coefficient’, quantifying the relative contribution of changes in the interaction of the enzyme with the rest of metabolism to the regulation of the enzyme's flux. For a more elaborate description and discussion of the method see Rossell (2006). The term ‘hierarchical regulation coefficient’ was introduced by ter Kuile & Westerhoff (2001), because the Vmax depends on the complete gene-expression cascade of transcription, translation, posttranslational modification, and mRNA and protein degradation. The two regulation coefficients sum up to one (summation theorem for the regulation of flux) implying that determination of one coefficient will yield the other automatically (ter Kuile & Westerhoff, 2001; Rossell et al., 2005). In practice the hierarchical regulation coefficient is more readily determined, as f(e) usually can be taken to equal Vmax, and changes in the Vmax as well as in the flux J through the enzyme can often be measured or estimated.

In this study we investigate the regulation of the glycolytic and fermentative capacity upon deletion of HXK2. To our knowledge this is the first study in which the anaerobic metabolism of this mutant is considered. Using regulation analysis, we will first dissect the regulation of the rates of glycolytic and fermentative enzymes into the contributions of changes in enzyme capacities (ρh) and the contributions of changes in the enzymes' interaction with the rest of metabolism (ρm) in response to the deletion of the HXK2 gene. We shall do this for the regulation of fluxes through the individual glycolytic and fermentative enzymes. Subsequently we shall then examine how hexokinase II deletion impacts on the regulation of the fermentative capacity during starvation. We will show that the mutant exhibits a remarkable resilience towards nutrient starvation in terms of its fermentative capacity. Based on our results and previous reports, we suggest an inverse correlation between glucose repression and the resilience of fermentative capacity towards nutrient starvation. Furthermore, we discuss how the resilience of the hxk2Δ mutant towards nutrient starvation, in combination with its higher biomass yield in batch, makes this strain interesting for the bakers' yeast industry.

Materials and methods

Growth and starvations

The growth and starvation procedures have been described in detail elsewhere (Rossell et al., 2005). Briefly, S. cerevisiae strains CEN-PK 113-7D (MATa MAL2-8c SUC2) and KY116 (MATa MAL2-8c SUC hxk2Δ::KanMX4) (Diderich et al., 2001) were grown in well-aerated batch cultures at 30°C in defined mineral medium containing 101 mM glucose (Verduyn et al., 1992) kept at pH 5.0 by automatic addition of KOH. Cells were harvested at an OD600 nm of 1.0 (exponential phase) and concentrated by centrifugation. Raamsdonk (2000) reported that the hxk2Δ mutant (KY116) grown on mineral medium containing 20 μg L−1 of biotin showed biotin deficiency in late exponential growth phase (OD600 nm of 4.0) but not earlier. Our studies were carried out on cells grown on a mineral medium containing 50 μg L−1 of biotin (Verduyn et al., 1992) and harvested at an OD600 nm of 1.0. In these conditions, we confirmed that addition of higher concentrations of biotin did not affect the culture's growth rate (results not shown) and therefore we discarded the possibility of biotin deficiency.

For starvation experiments, the pellets were washed with equal volumes of ice-cold growth medium lacking either glucose or ammonium, and resuspended in the corresponding medium at 30°C to a cell density of 7.5 g L−1 wet weight (c. 1 g dry weight L−1) at pH 6.0. The suspensions, of c. 0.30 L, were kept in 2 L shake flasks on a rotary shaker at 30°C and 200 r.p.m. without pH control for 24 h. For the measurement of steady-state fluxes, the cells were harvested by centrifugation and resuspended in growth medium without a carbon source and kept on ice for at most 1 h prior to measurement. Similarly, for the measurement of zero-trans influx of glucose, cells were harvested by centrifugation and resuspended in growth medium lacking carbon and nitrogen sources, and kept on ice for at most 1 h prior to measurement.

Steady-state fluxes

For the measurement of fermentative capacity and other steady-state fluxes the cells were resuspended in medium lacking glucose at 30°C, kept anaerobic in a setup described by van Hoek (1998) with the modification that the headspace was flushed with N2 instead of CO2 as described in Rossell (2005). At time zero, 101 mM of glucose was added. Ethanol, glucose, glycerol, succinate, acetate and trehalose were monitored for 30 min by PCA extraction (Rossell et al., 2005) followed by HPLC [300 mm × 7.8 mm Ion exchange column Aminex-HPX 87H (Biorad), with 22.5 mM H2SO4 kept at 55°C as eluent at a flow rate of 0.5 mL min−1]. Glycogen was assayed according to Parrou & Francois (1997). The summed rates of production of acetate and succinate were always below 1% of the consumed glucose and are not reported. The fermentative capacity is defined as the specific rate of carbon dioxide production. Here the specific rate of ethanol production, which equals the specific rate of CO2 production if the minor (i.e. <1%) production of acetate is neglected, was measured. In a control experiment it had been verified that ethanol evaporation was negligible under these conditions.

The measured extracellular fluxes were used to calculate the fluxes through each enzyme in the glycolytic and fermentative pathways. The flux through the glucose transporter (GLT) was taken as equal to the measured glucose consumption flux. The fluxes through enzymes downstream of hexokinase (HK) were calculated from the steady-state rates of ethanol and glycerol production. The fluxes through HK, glucose-6-phosphate isomerase (PGI), 6-phosphofructokinase (PFK) and aldolase (ALD) were calculated by dividing the sum of the glycerol and ethanol fluxes by two. The flux through triose-phosphate isomerase (TPI) was calculated by subtracting the rate of glycerol production from the flux through the previous block (HK through ALD). The fluxes through glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and the enzymes downstream were taken as equal to the measured ethanol flux (Table3). The consumed carbon matched the produced carbon within experimental error (Fig. 1) and their means differed by 5% and 0.5 % in the wild type and hxk2Δ mutant, respectively. Our calculation of fluxes through individual enzymes implicitly assumes that these small gaps in the carbon balances are filled by synthesis or mobilization of storage carbohydrates. Such an assumption is justified because the measured glycogen flux was small, but inherently difficult to quantify accurately due to uncertainties of extraction and calibration.

View this table:
3

Steady-state rates of glycolytic and fermentative enzymes of the hxk2Δ mutant

EnzymeUnstarvedNitrogen starvedCarbon starved
GLT0.15 ± 0.030.15 ± 0.020.14 ± 0.02
HK–ALD0.14 ± 0.030.20 ± 0.020.13 ± 0.01
TPI0.11 ± 0.030.16 ± 0.020.10 ± 0.01
GAPDH–ADH0.25 ± 0.080.36 ± 0.040.23 ± 0.02
  • Steady-state rates of glycolytic and fermentative enzymes were calculated based on the measured fluxes reported in Table 1 as described in the ‘Materials and methods’ section. The steady-state rates in unstarved, nitrogen- and carbon-starved cultures are presented in mmol min−1 g−1 protein. Errors represent SEs of the mean of four independent experiments carried out on different batches of cells.

1

Carbon-flux balances in the wild type and hxk2Δ mutant. Carbon fluxes for each strain are represented with two columns: one depicting the consumed carbon (open column) and the other the produced carbon (dark shaded areas). Columns are divided into fluxes: glucose (open columns), glycerol (black areas), ethanol (light shaded areas), and CO2 (calculated from the production of ethanol; dark shaded areas). Other measured fluxes were negligible (see text). Error bars represent the SE of the mean of the sum of consumed or produced carbon fluxes of four independent experiments carried out with different batches of cells. The carbon-flux balance of the wild type has been published previously (Rossell et al., 2006).

Zero-trans influx of glucose

Zero-trans influx of 14C radio-labelled glucose was determined in a 5 s uptake assay at 30°C according to Walsh (1994), with the modifications introduced by Rossell (2005). The range of glucose concentrations was between 0.25 and 225 mM. Irreversible Michaelis–Menten equations were fitted to the results by nonlinear regression using sigmaplot 2001 version 7.0 (SPSS Inc.).

Enzyme activity measurements

Enzyme extractions and activity assays were performed as described by van Hoek (1998). Enzyme activities (Vmax) were measured in freshly prepared extracts through NAD(P)H-linked assays, using a COBAS BIO (Roche, Basel) automated analyzer for spectroscopic measurements. All enzyme assays were performed with four concentrations of cell extract to confirm that reaction rates were proportional to the amount of cell extract added.

Regulation analysis

Hierarchical regulation coefficients (ρh) to quantify the local flux regulation of glycolytic and fermentative enzymes (ter Kuile & Westerhoff, 2001; Rossell et al., 2005) in response to deletion of the HXK2 gene were calculated as follows: Embedded Image 3 in which the J refers to the in vivo flux through the enzyme (see section Steady-state fluxes) and the subscripts Mutant and WildType refer to the hxk2Δ mutant and the wild-type strain, respectively. We performed at least three independent measurements of the Vmax-values for each of the glycolytic and fermentative enzymes in the wild type and hxk2Δ mutant. The Vmax-values were translated into logarithmic space and the mean and SD of lnVmax were computed. For each enzyme, the mean lnVmax-value of the wild type was subtracted from that of the mutant yielding the numerator of Equation (3). The SD of the numerator of Equation (3) was computed as the square root of the sum of variances of the wild-type and mutant lnVmax-values. The denominator of Equation (3) and its SD were computed similarly, based on four independent determinations of the flux through the enzyme of interest (see above). The ratio of the numerator and denominator of Equation (1) equals the hierarchical regulation coefficient. The metabolic regulation coefficient was then calculated from ρm=1 –ρh. The SD of the hierarchical regulation coefficient was calculated by multiplying ρh and the square root of the sum of squared coefficients of variations (Cv=σ/μ) of the numerator and denominator. The SE of the mean (SEM) of ρh was computed by dividing the SD by the square root of 3. Note that ρh and ρm share the same SD in view of the summation theorem for the regulation of flux.

Results

The fermentative capacity is decreased in the hxk2Δ mutant, but stable during starvation

We first measured the overall steady-state fluxes of glucose, ethanol, glycerol, acetate, succinate, glycogen and trehalose under standardized conditions in the mutant and the wild type. To this end, S. cerevisiae strains CEN.PK 113-7D and the hxk2Δ mutant (KY116) were grown in well-aerated and pH-controlled batch cultures. In each starvation experiment, an aliquot of cells was harvested during exponential growth and split in three parts. One part (referred to as ‘unstarved’) was washed and transferred to an anaerobic vessel with a fresh and complete medium with excess of glucose (101 mM). This condition was meant to mimic the situation of bakers' yeast in dough (van Hoek et al., 1998). The above-mentioned fluxes were then measured over a period of 30 min. The other two batches of cells were washed and transferred to the same fresh medium, except that it lacked either ammonium (‘nitrogen-starved cells’) or glucose (‘carbon-starved cells’). After 24 h the starved cells were harvested and the fluxes were measured in a complete medium, in the same way as was performed for the unstarved cells.

Figure 1 depicts the measured carbon fluxes for the unstarved cultures of the wild type and mutant strains. In both strains glucose was converted predominantly to ethanol, glycerol and CO2. The production fluxes of acetate and succinate and the mobilization of storage carbohydrates were always below 1% of the rate of glucose consumption (not shown). The produced carbon could be accounted for by the consumed carbon within experimental error. In the unstarved hxk2Δ mutant the glucose consumption flux was decreased by 75% as compared to the flux in the unstarved wild type, and this was reflected in a proportional decrease of the production fluxes of ethanol, glycerol and CO2.

Table 1 shows the fermentative capacities of the wild type and mutant strains in the unstarved and starved cultures. The fermentative capacity is defined as the specific rate of CO2 production under the dough-like conditions specified above. Here it was measured as the specific ethanol flux, which should equal the CO2 flux under anaerobic conditions if we neglect the small acetate production. Deletion of the HXK2 gene resulted in a 75% reduction of the fermentative capacity. The responses of the fermentative capacities of the wild type and the hxk2 null mutant to nutrient starvation were very different. In the wild type, both types of nutrient starvation led to a substantial loss of fermentative capacity (between 50% and 70%), as was shown before (Rossell et al., 2006), while the hxk2Δ mutant showed no loss of fermentative capacity during 24 h of deprivation of either carbon or nitrogen. When expressed per unit protein, the fermentative capacity of nitrogen-starved cultures was even slightly increased in the mutant. The latter was not observed when fluxes were expressed per unit dry weight, as the cells accumulated carbohydrates and increased their dry-weight-to-protein ratio.

View this table:
1

Fermentative capacities of the wild type and hxk2Δ mutant

Wild typehxk2Δ
Fermentative capacity per unit proteinFermentative capacity per unit dry weightFermentative capacity per unit proteinFermentative capacity per unit dry weight
Unstarved1.04 ± 0.030.392 ± 0.0130.25 ± 0.040.081 ± 0.016
Nitrogen starved0.49 ± 0.050.072 ± 0.0030.36 ± 0.040.058 ± 0.005
Carbon starved0.33 ± 0.050.094 ± 0.0100.23 ± 0.020.072 ± 0.005
  • Fermentative capacities (measured as the specific rate of ethanol production under glucose-excess and anaerobic conditions) of both strains in unstarved, nitrogen- and carbon-starved cultures are presented in mmol min−1 g−1 protein or mmol min−1 g−1 dry weight. Errors represent SEs of the mean of four independent experiments carried out on different batches of cells.

Like the wild type (Rossell et al., 2006), the mutant accumulated storage carbohydrates during nitrogen starvation and mobilized this pool during the fermentative capacity assay (results not shown). Indeed, after nitrogen starvation, the mean rate of ethanol production in the fermentative capacity assay (0.36 μmol min−1 mg−1 protein; Table 1) exceeded what can be produced from the mean glucose influx (i.e. two times 0.15 μmol min−1 mg−1 protein).

Enzyme activities in the unstarved hxk2Δ mutant

Deletion of the HXK2 gene resulted in decreased fluxes through the glycolytic and fermentative enzymes. In order to investigate whether these flux reductions could be understood in terms of changes in the activities of the glucokinase and hexokinases and/or of the other enzymes in the glycolytic and fermentative pathways, we measured the maximum enzyme activities (Vmax) in extracts from unstarved cultures of the wild type and the mutant. The absolute values are shown in Fig. 2. Deletion of the HXK2 gene resulted in an 80% reduction of hexokinase capacity (Vmax), consistent with earlier results (Diderich et al., 2001). This is a surprisingly strong reduction, as HXK1 has been shown to be upregulated substantially at the mRNA level in a hxk2Δ mutant (Lin et al., 2002). Possibly the HXK1 gene is regulated also by posttranscriptional mechanisms which counteract the strong transcriptional regulation. This reduction of hexokinase capacity was accompanied by significant reductions in the Vmax values of PGI, ALD, TPI, PGK, PGM and PDC (Student's t test α=5%). The reduction of PDC activity was already reported by (Diderich et al., 2001). None of the enzyme activities was increased.

2

In vitro maximum activities of glycolytic and fermentative enzymes. The in vitro determined Vmax-values of the wild type (closed columns) are compared to those of the hxk2Δ mutant (open columns). Error bars represent the SEs of the mean of four (wild type) and three (mutant) independent enzyme activity measurements carried out on different batches of cells. Error bars of the GLT represent the SEs of the mean of two independent experiments carried out on different batches of cells.

Regulation analysis

Deletion of the HXK2 gene resulted in a reduced hexokinase activity that was accompanied by the reduction in the activities of PGI, ALD, TPI, PGK and PDC. These coincided with reduction of the rates of glycolytic and fermentative enzymes. In order to dissect to what extent the changes of Vmax were responsible for the changes of enzyme rates and to what extent these rates were rather regulated by changes in enzyme interactions with the rest of metabolism, we calculated the hierarchical (ρh) and metabolic (ρm) regulation coefficients for the comparison of the wild type and the mutant in unstarved cultures (cf. ‘Introduction’). These are summarized in Table2. All the hierarchical regulation coefficients have values below 1, meaning that the relative changes in the steady-state rates of all glycolytic and fermentative enzymes are larger than the relative changes in enzyme capacities (Table2). The highest ρh value (0.9) was, not surprisingly, obtained for hexokinase, of which one of the genes was deleted. Indeed the ‘gene-expression’ regulation reported by this coefficient in this case is more than the response by the organism; it comprises the regulatory act by the experimenter. Aldolase and enolase were largely hierarchically regulated (ρh=0.8). A number of enzymes were regulated cooperatively by enzyme capacity changes and changes in their interaction with the rest of metabolism. Notably GAPDH, PGK, PGM and PDC were regulated with almost equal contributions of enzyme capacity changes and changes in the way enzymes interacted with the rest of metabolism. One of the hierarchical regulation coefficients was negative (ADH), indicating that the flux and the enzyme activity changed in opposite directions. However, the value of this coefficient did not differ significantly from zero and we interpreted this as a predominantly metabolic regulation. Glucose transport and PFK were regulated exclusively at the metabolic level (ρm=1.0), without any significant hierarchical regulation. Also, PGI and PK were predominantly regulated by metabolism (ρm=0.8). Significant antagonistic regulation (cf. Rossell et al., 2006) was not observed (if we ignore the negative but statistically insignificant ρh of ADH).

View this table:
2

Hierarchical and metabolic regulation coefficients of the regulation ensuing from the deletion of the HXK2 gene

ρhSEMρm
GLT0.00.11.0
HK0.90.20.1
PGI0.20.10.8
PFK0.00.21.0
ALD0.80.40.2
TPI0.30.10.7
GAPDH0.40.20.6
PGK0.60.20.4
PGM0.50.10.5
ENO0.80.20.2
PK0.20.20.8
PDC0.50.10.5
ADH−0.20.31.2
  • Hierarchical (ρh) and metabolic (ρm) regulation coefficients were calculated as described in the ‘Materials and methods’ section. Errors were quantified with SEs of the mean (SEM) calculated for four (wild-type) and three (mutant) independent Vmax measurements and four independent flux estimations, all performed on independent batches of cells.

Enzyme activities of the hxk2Δ mutant upon starvation

Subsequently, we wondered why the fermentative capacity in the hxk2Δ mutant was so stable during carbon starvation. Figure 3 shows the Vmax-values of the glycolytic and fermentative enzymes of the hxk2Δ mutant as percentage of those in the unstarved condition. Only the capacity of the glucose transporters was significantly reduced after both types of starvation. After carbon starvation the activity of alcohol dehydrogenase was significantly increased (Student's t test α=5%). After nitrogen starvation the activity of enolase (ENO) increased significantly. In a previous study we have measured the enzyme activities in the wild type upon nitrogen and carbon starvation (Rossell et al., 2006) and we will discuss the differences between mutant and wild type in the ‘Discussion’ section below.

3

Vmax-values of the hxk2Δ mutant as a percentage of those in unstarved hxk2Δ cultures. Unstarved (black columns), nitrogen-starved (diagonally striped columns) and carbon-starved (grey columns). Error bars of glycolytic and fermentative enzymes represent the percentage SE of the mean, with respect to their corresponding unstarved mean Vmax-value, of three independent experiments carried out on different batches of cells. Exceptions are GLT and PK, for which two independent experiments were carried out.

Discussion

We investigated how the rates of the glycolytic and fermentative enzymes were regulated upon deletion of hxk2 and upon starvation. This is a part of a larger research programme in which we quantitatively investigate the interplay of metabolism and gene expression, with glycolysis in bakers' yeast as our model system. Previous studies focused on the regulation of flux in response to environmental perturbations (Rossell et al., 2006; Daran-Lapujade et al., submitted). These studies showed that existing paradigms of multi- or single-site regulation did not apply to the regulation of yeast glycolysis. In multisite modulation all Vmax values would be regulated proportionally, leading to metabolite homeostasis upon large flux changes (Fell & Thomas, 1995). Single-site modulation assumes that only one rate-controlling enzyme is regulated hierarchically.

The present study is the first example in which regulation analysis was applied to regulation in response to a gene deletion. HXK2 is even a special gene in that it encodes a protein that acts both as an enzyme catalyzing a glycolytic reaction and as a component of the main glucose-repression pathway. Upon deletion of HXK2 we found a broad range of combinations of metabolic and hierarchical regulation, from purely metabolic (ρh=0) to purely hierarchical (ρh=1). The majority of hierarchical regulation coefficients ranged between 0 and 1. This means that the regulation of the fluxes through those enzymes was brought about cooperatively by changes in the Vmax values of the enzymes and their interactions with the rest of metabolism (Rossell et al., 2005). The finding that another type of perturbation again resulted in a broad range of combinations of metabolic and hierarchical regulation corroborates our earlier conclusion that the regulation of glycolytic fluxes in bakers' yeast is variegated and may not be driven by single drives or constraints as suggested by single- or multi-site regulation hypotheses.

Based on our current knowledge of transcription regulation of glycolytic genes by the GCR1/GCR2/RAP2 system (Chambers et al., 1995), it may be unexpected that multisite modulation does not operate. The deletion of gcr1 and/or gcr2 results in decreased in vitro activities of the majority of glycolytic enzymes in cells that are grown on glucose but not in those grown on nonfermentable carbon sources (Clifton & Fraenkel, 1981; Uemura & Fraenkel, 1990). The fact that in our study the glycolytic enzyme activities were regulated differentially (Fig. 2), some remaining constant and others decreasing to variable extents in the mutant, suggests that the glycolytic genes are not simply coregulated at the transcriptional level. Moreover, if we compare the regulation of glycolysis upon different perturbations, such as the transition from aerobic to anaerobic conditions (Daran-Lapujade et al., submitted) or the starvation of the wild type for nitrogen or carbon (Rossell et al., 2006), it becomes clear that the distribution of regulation between metabolism and gene expression differs between conditions. This further suggests that the glycolytic genes are regulated more subtly by other mechanisms on top of or interacting with the GCR1/GCR2/RAP2 system. For instance, the gene SGC1 has been described as a suppressor of the GCR system (Sato et al., 1999). Furthermore, regulation of mRNA stability has been implicated in glucose repression and could therefore be involved in the response of the hxk2Δ mutant (de la Cruz et al., 2002). The interactions between the main glucose repression pathway in which Hxk2p functions, the GCR1/GCR2/RAP2 system, and possible other mechanisms such as regulation of mRNA stability, require further study, and in view of the above, a quantitative study.

The resilience of the hxk2 null mutant towards nutrient starvation is consistent with results obtained for other respiratory yeast cultures. Nilsson (2001b) reported that postdiauxic shift cultures respiring ethanol preserved their fermentative capacity better than cells growing on glucose when challenged by nutrient starvation. Analogously, the fermentative capacity of fully respiratory, glucose-limited chemostat cultures of S. cerevisiae at low dilution rates was also resilient to nitrogen or carbon starvation (K. van Eunen and B.M. Bakker, unpublished). These findings together suggest that the stability of the fermentative capacity upon nutrient starvation is inversely correlated with glucose repression.

It is surprising that the fermentative capacity is constant in the hxk2Δ mutant during starvation despite a strong decrease of the glucose-transport capacity. We compared these results to previously published results for the wild type under the same conditions. In contrast to the mutant, the wild type decreased not only the activity of glucose transport significantly upon nitrogen starvation, but also that of HXK, PGI, ALD, PGM, PYK and ADH (Rossell et al., 2006). The decrease of the latter set of enzymes may explain why the fermentative capacity of the wild type decreased upon nitrogen starvation, while that of the mutant remained constant. However, this argument only holds if the glucose transporter, of which the activity decreased in both strains, exerts no control on the fermentation flux under the applied conditions, at least in the mutant.

Upon carbon starvation of the wild type only the activity of the glucose transporter decreased, while other enzyme activities remained constant or even increased (Rossell et al., 2006). In this respect there is no difference between the mutant and the wild type. To explain the difference between the two strains it is again necessary to assume that the glucose transporter has no flux control, at least in the mutant. Moreover, the fact that in the wild type the fermentative capacity changed 2.5-fold more than the glucose-transport capacity suggested that some changes outside glycolysis partly caused the drop of fermentative capacity in this strain (Rossell et al., 2006). The latter changes may not occur in the mutant. However, since these changes were outside the scope of our studies, we can only speculate.

The flux control coefficient of glucose transport can be measured by inhibition of the glucose transporters by addition of maltose (Diderich et al., 1999). The inhibition of the glucose transporters by maltose is competitive with an inhibition constant between 32 and 42 mM (Reijenga et al., 2005a, b). As the mutant had a very low Km for glucose (5.4 mM, result not shown), we estimated that even addition of 250 mM of maltose would inhibit the transporter by only 23–28% at 101 mM glucose. This is less than the actual decrease of the transport capacity in the mutant upon either type of starvation (Fig. 3) and preliminary experiments to measure the control by the transporter in the mutant were inconclusive due to the relatively high error on the ethanol flux in comparison to the small percentage of inhibition of the transporters (not shown). The same experiment on the wild type, in which 250 mM of maltose should inhibit the transporter by 60–70% due to the lower affinity for glucose, showed no effect on the flux (not shown), suggesting that in the wild type the glucose transporter indeed had no flux control. As the mutant has a decreased hexokinase activity, it is tempting to speculate that if control has shifted in the mutant, it should have been towards hexokinase rather than towards the transporter.

At high sugar concentrations the hxk2Δ strain has an almost completely respiratory metabolism (Diderich et al., 2001) and therefore a much higher biomass yield than the wild type. This would alleviate the need for a restricted glucose influx and thorough mixing of glucose in the bakers' yeast production phase. The finding that the fermentative capacity of the mutant was much lower than that of the wild type at first sight seemed to disqualify the mutant for application purposes. However, also the wild type has a much lower fermentative capacity when grown under respiratory conditions. Moreover, after starvation, an integral part of the overall bakers' yeast production process, the wild type and the mutant had similar fermentative capacities. This makes the hxk2Δ an interesting starting point for growing yeast at faster rates in (fed-)batch cultures and with more efficient substrate utilization.

Acknowledgements

This work was supported by Netherlands Technology Foundation (STW) Grant DGC 5232 and E.U. Project BIO4-CT98-0562 (DG 12 - SSM1), as well as by the EU-FP6 Network of Excellence BioSim, and the Marie Curie network NucSys.

Footnotes

  • Editor: Merja Penttilä

References

View Abstract