FEMS Yeast Research 4 (2004) 633–641

www.fems-microbiology.org

Yeast cells display a regulatory mechanism in responseto methylglyoxal

Jaime Aguilera 1, Jose Antonio Prieto *

Department of Biotechnology, Instituto de Agroqu�ımica y Tecnolog�ıa de los Alimentos, Consejo Superior de Investigaciones Cient�ıficas,

Poligono de la Coma, s/n, P.O. Box 73, Burjassot Valencia 46100, Spain

Received 10 November 2003; received in revised form 12 December 2003; accepted 12 December 2003

First published online 28 January 2004

Abstract

Methylglyoxal (MG), a glycolytic by-product, is an extremely toxic compound. This fact suggests that its synthesis and deg-

radation should be tightly controlled. However, little is known about the mechanisms that protect yeast cells against MG toxicity.

Here, we show that in Saccharomyces cerevisiae, MG exposure increased the internal MG content and activated the expression of

GLO1 and GRE3, two genes involved in MG detoxification; GPD1, the gene for glycerol synthesis; and TPS1 and TPS2, the

trehalose pathway genes. This response was specific as demonstrated by the analysis of marker genes and effectors of the general

stress response. Physiological experiments with MG-treated cells showed that this compound triggers the overproduction of glyc-

erol. Furthermore, a gpd1 gpd2 double mutant showed enhanced MG contents compared with the wild-type. Overall, these results

appeared to indicate that up-variations in the intracellular content of the toxic compound are perceived by the cell as a primary

signal to trigger the transcriptional response. In agreement with this, MG-instigated GPD1 activation was enhanced in strains

lacking GLO1, and this effect correlated with the internal MG content. Finally, induction of GPD1, TPS1 and GRE3, and enhanced

MG contents were also observed in low-glucose-growing cells subjected to a sudden increase in glucose availability. The implications

of this regulatory mechanism on protection against MG are discussed.

� 2004 Federation of European Microbiological Societies. Published by Elsevier B.V. All rights reserved.

Keywords: Methylglyoxal; Saccharomyces cerevisiae; Yeast; Signalling; GRE3; GLO1; TPS1; TPS2

1. Introduction

Methylglyoxal (MG), an intrinsic intermediate ofglycolysis [1], is an extremely toxic compound, able to

react with, and modify, different molecular targets [2].

MG is formed by fragmentation of free solution forms

of both dihydroxyacetone phosphate (DHAP) and

glyceraldehyde-3-phosphate (GA3P) [3]. In bacteria,

MG can also be formed biosynthetically by the action of

MG synthase [4]. The homologous enzyme from Sac-

* Corresponding author. Tel.: +34-96-3900022;

fax: +34-96-3636301.

E-mail address: prieto@iata.csic.es (J. Antonio Prieto).1 Present address: Kluyver Laboratory of Biotechnology, Delft

University of Technology, Julianalaan 67, Delft 2628 BC, The

Netherlands.

1567-1356/$22.00 � 2004 Federation of European Microbiological Societies

doi:10.1016/j.femsyr.2003.12.007

charomyces cerevisiae has been isolated and character-

ised [5], but its sequence has not been reported, nor a

gene encoding the enzyme has been localised.Cells of different origin accumulate MG under

physiological conditions resulting in a loss of control

over carbohydrate metabolism. Thus, Escherichia coli

cells overproduce MG by deregulation of the transport

and metabolism of glucose [6]. In bovine endothelial

cells, incubation in 30 mM glucose increases MG [7].

Enhanced MG levels upon osmotic and oxidative stress

in mammalian [8,9] and yeast [10] cells have also beenreported. Consistent with this, increased consumption of

glucose during osmotic adaptation in yeast has been

demonstrated [11].

In yeast, there are three established pathways for MG

detoxification. First, MG conversion into LL-lactaldehyde

by the action of MG reductase [12]. Second, the glyoxa-

. Published by Elsevier B.V. All rights reserved.

634 J. Aguilera, J.A. Prieto / FEMS Yeast Research 4 (2004) 633–641

lase system, consisting in two enzymes, glyoxalase I and

glyoxalase II (encoded by the GLO1 and GLO2 genes,

respectively), whose activity converts MG into DD-lactic

acid in the presence of glutathione [4]. Finally, the yeast

aldose reductase, encoded by the GRE3 gene [10], whichtransforms MG into 1,2-propanediol in a two-step reac-

tion, dependent on NADPH [13].

Expression of GLO1 is induced by osmotic stress and

regulated by the high-osmolarity glycerol (HOG) re-

sponse pathway [11], the main signalling cascade that

contributes to the up-regulation of osmotically induced

genes [14,15]. Transcription of GRE3 is up-regulated

under a variety of stress conditions, and several factorshave been found to control its expression, between them

Msn2p and Msn4p [10], the two factors mediating the

general stress response pathway in Saccharomyces [16].

Stress-triggered induction of GRE3 increases aldose re-

ductase activity, leading to a drop in the intracellular

level of MG, a circumstance that is not observed in cells

of gre3 and glo1 gre3 mutants [10].

The stress-instigated transcriptional activation ofMG degradative pathways suggests that this regulatory

mechanism could function in other physiological situa-

tions that result in MG overproduction. Thus, there is

evidence that yeast cells activate MG-degradative

pathways in response to exogenous MG, since a glo1

mutant shows a clear phenotype of MG sensitivity [4].

The strong toxicity of MG also suggests that its syn-

thesis should be tightly controlled. In bacteria, thiscontrol is established by allosteric regulation of the MG

synthase activity [6]. However, this control level remains

unclear in eukaryotic cells because, in them, non-

enzymatic formation appears to be the primary source

of MG [17]. Neither the overall protective mechanisms

involved in the yeast response to increased MG levels,

nor the metabolic signals mediating such response have

still been elucidated.In this paper, we show that expression of GRE3 and

GLO1 genes from Saccharomyces is specifically up-

regulated in response to increased intracellular MG

levels. We also show that the same stimulus provokes

the induction of GPD1, TPS1 and TPS2, three genes

involved, directly or indirectly, in control of the level of

glycolytic intermediates. Evidence suggesting that yeast

Table 1

Saccharomyces cerevisiae strains used in this study

Strain Genotype

CEN.PK113-11A MATa URA3 LEU2 trp1-289 his3-D1 MAL2-8cS

W303-1A MATa ade2-1 his3-11,15 leu-2-3,112 trp1-1 ura3-1

W303 msn2 msn4 MATa ade2-1 his3-11,15 leu-2-3,112 trp1-1 ura3-1

msn2-D 3::HIS3 msn4 1::TRP1

W303 gpd1 gpd2 MATa ade2-1 his3-11, 15 leu-2-3, 112 trp1-1 ura3

gpd1 ::TRP1 gpd2 ::URA3

YPH250 MATa ura 3-52 his3-D200 leu2-D1 trp1-D1 lys2-8

YGL1 MATa ura 3-52 his3-D200 leu2-D1 trp1-D1 lys2-8

cells sense intracellular MG levels is presented, and

the implication of the nuclear factors Msn2p/Msn4p is

discussed.

2. Materials and methods

2.1. Strains and culture conditions

The S. cerevisiae strains used in this work are listed in

Table 1. Yeast cells were cultured in YNB minimal

media [0.67% yeast nitrogen base without amino acids

(DIFCO) plus 2% glucose] supplemented with the ap-propriate concentrations of essential nutrients [18]. Cells

were grown routinely in 250- or 1000-ml Erlenmeyer

flasks at 30 �C on an orbital shaker (200 rpm).

For MG experiments, cells were grown to a mid-ex-

ponential phase (OD600 ¼ 0.3–0.5) and MG (Sigma) was

added at various final concentrations. For osmotic-

shock treatment, cells were collected and shifted to

0.3 M NaCl-containing medium. Samples were taken atthe indicated times.

2.2. RNA purification and Northern blot analysis

Total RNA from yeast cells was prepared as

described [18]. Equal amounts of RNA (30 lg) were

separated in 1% (w/v) agarose gels, containing formal-

dehyde (2.5% v/v), transferred to a Nylon membraneand hybridised with 32P-labelled probes. Probes were

obtained as follows: PCR-amplified DNA fragments

containing sequences of GRE3 (whole ORF), GPD1

(+23 to +848), GLO1 (+124 to +506) and TPS1 (whole

ORF) genes; 2.75 kb NotI fragment of plasmid pTPS2

[19] containing TPS2 sequence; 1.1 and 1.3 kb fragments

of the CTT1 gene sequence obtained by EcoRI restric-

tion of plasmid pRB322-5109 [20]; 1.4 kb fragment ofthe HSP26 sequence obtained from a SphI/BglII re-

striction of the plasmid pVZ26 [21]; and HSP104 1.2 kb

EcoRI fragment of plasmid pUZ1 [22]. Probes were

radiolabelled with the Random-primer Kit Ready-to-

Go (Pharmacia Biotech) and [a32P]dCTP (Amersham).

For control, filters were hybridised with a PCR-ampli-

fied probe of the IPP1 (whole ORF) gene [23]. We used

Reference or source

UC2 K.-D. Entian

can1-100 GAL mal SUC2 [45]

can1-100 GAL mal SUC2 [46]

-1 can1-100 GAL mal SUC2 [31]

01 ade2-101 Yeast Genetic Stock Center

01 ade2-101 glo1D::HIS3 Y. Inoue

Fig. 1. Expression of GRE3 and GLO1 is up-regulated by MG. YNB-

grown cells of the CEN.PK113-11A strain were exposed to a final

concentration of 0.5 (s), 1 (M) or 2 (�) mM MG (a), or 1 mM MG

(b,c). Samples were taken at the indicated times and analysed for

GRE3 (a) and GLO1 (b) transcript levels, or MG content (c). Northern

blots and MG determinations were performed as described in Sec-

tion 2. Values in (c) are given as means�SD of three independent

experiments. The graph in panel (a) represents quantification of the

mRNA levels of GRE3, relative to those of IPP1.

J. Aguilera, J.A. Prieto / FEMS Yeast Research 4 (2004) 633–641 635

this gene, instead of ACT1 or rDNA, because pre-

liminary experiments had indicated that its expression is

not altered by addition of MG. The filters were analysed

by autoradiography. For quantification, films were

scanned and the images quantified with the Quantity-One software (BioRad) with no modification. Spot in-

tensities were evaluated with respect to the mRNA level

of IPP1 and presented as percentages of the maximal

amount of induction.

2.3. Cell sampling and determination of metabolites

For MG assays, 40 ml of yeast culture (OD600 ¼ 0.4–0.5) were filtered and the cell cake was resuspended

immediately in 1.0 ml of acid extraction buffer (2.3 M

HClO4, 90 mM imidazole). Then, the mixture was

poured into a screw-cap Eppendorf tube containing

0.3 g of glass beads (acid-washed, 0.4-mm diameter) and

frozen in liquid nitrogen. This protocol was found to be

faster in obtaining the cell extracts than the traditional

methanol-quenching procedure [24], especially for thelarge amount of sample volume (40 ml). Moreover, no

significant differences were found in the MG measure-

ments of control cells sampled with both methods (data

not shown). Consequently, the latter was only chosen

when removal of medium from the cells was required

(MG-shocked cells). In this case, the yeast culture was

centrifuged at 3000g for 2 min ()10 �C) and sprayed into

35 ml of buffered 60% methanol solution [24] kept at)40 �C. The mixture was centrifuged and the pellet was

washed again under the same conditions. Finally, the

cell cake was resuspended as above and frozen in liquid

nitrogen. The suspension was allowed to thaw on ice

and extraction was carried out by three freeze-thaw cy-

cles with shaking in a Mini-bead beater (30 s at maxi-

mum speed) between each cycle. After centrifugation

(5 min at 8000g, 0 �C), the supernatant was analysedimmediately for MG content. For glycerol measure-

ments, cells were collected by filtration. The filter was

transferred quickly to a cold (4 �C) tube containing 1 ml

of distilled water and the yeast suspension was boiled for

10 min, cooled on ice and centrifuged at 15,300g for

10 min (4 �C). Finally, the supernatant was collected and

used for the assays.

Glycerol was measured using a standard commercialkit (Roche), following manufacturer�s instructions.

Methylglyoxal was determined by HPLC as 2-methyl-

quinoxaline, according to the conditions described by

Chaplen et al. [25]. Samples (0.5 ml) were derivatised

using 1,2-diaminobenzene, as described by Cordeiro and

Ponces Freire [26], except that the reaction mixture was

incubated for 3 h at room temperature. 5-Methylqui-

noxaline was used as internal standard. For quantifica-tion, a linear methylglyoxal calibration curve was

obtained by plotting known methylglyoxal concentra-

tions against ratios of analyte peak height to internal-

standard peak height. An intracellular water value of

1.4 ll ml�1 per unit of OD600 [27] was used to calculate

the molar concentration of internal MG. For the tre-

halose assay, samples were taken and treated with tre-

halase as described in [28]. The glucose liberated wasmeasured colorimetrically, by the glucose oxidase and

peroxidase method, in the presence of 2,20-azino-bis(3-ethylbenzothiazoline-6-sulfonic acid) diammonium

salt (ABTS, Sigma), following manufacturer�s instruc-

tions.

3. Results

3.1. MG triggers transcriptional activation of GRE3 and

GLO1

We analysed by Northern blot the expression of

GRE3 in cells of the wild-type yeast strain CEN.PK113-

11A exposed to MG. As shown in Fig. 1(a), expression

of GRE3 was induced by sub-lethal concentrations ofMG. Like for many other genes activated by stress or

chemical stimuli [23], the GRE3 response to MG was

time- and dose-dependent, reaching a maximum level of

mRNA after 60 min of exposure at 1 mMMG under the

assay conditions (Fig. 1(a)). Furthermore, the MG-im-

posed transcriptional response was also observed for

GLO1 (Fig. 1(b)), the gene for glyoxalase I.

We further investigated whether externally addedMG was internalised. For this purpose, we analysed the

kinetics of intracellular MG variations after addition of

1 mM MG (final concentration). As shown in Fig. 1(c),

the internal concentration of MG showed an immediate

increase when cells were exposed to the toxic compound,

636 J. Aguilera, J.A. Prieto / FEMS Yeast Research 4 (2004) 633–641

reaching a maximum level, around 60 lM, after 10–20

min of exposure. Later, the amount of intracellular MG

decreased, with values at 40 min only threefold higher

than the basal value (6.9� 1.4 lM). In contrast, external

MG levels did not vary significantly during the timecourse of the experiment: 0.98� 0.3 mM just after the

addition to 0.95� 0.2 mM at 40 min. Therefore, a shift

to MG-containing medium causes a transient increase in

the internal content of the toxic agent and the activation

of genes implied in its degradation.

3.2. MG activates a specific transcriptional response

MG is a toxic compound able to cause cell death at

high doses and to delay growth at sublethal concentra-

tions. Therefore, the up-regulation of GRE3 and GLO1

by MG could be explained as a result of the activation

by this metabolite of the general stress response. In or-

der to investigate this possibility, RNA samples from

cells exposed to 1 mM MG were hybridised with probes

of HSP26, HSP104 and CTT1, three marker geneswidely used in studies on the general stress response [16].

As expected from previous reports, transcription of

these genes was found to be fully derepressed upon os-

motic- or heat-shock (Fig. 2(a)). However, their ex-

pression was completely indifferent to a 30-min

treatment with 1 mM MG. Neither was induction found

Fig. 2. MG exposure does not trigger the general stress response. (a)

Cell cultures of the CEN.PK113-11A strain were heat-shocked (39 �C)or shifted to 0.7 M NaCl- or 1 mM MG-containing medium. (b) YNB-

grown cells of the W303-1A wild-type (}) and msn2 msn4 mutant (M)

strain, were transferred to 4 mM MG-containing medium. Samples

were taken after 30 min or at the indicated times and analysed by

Northern blot as described in Section 2. Filters were probed for

HSP26, HSP104 and CTT1 (a) or GRE3 (b) mRNA. The graph in

panel (b) represents quantification of the mRNA levels of GRE3, rel-

ative to those of IPP1. (c) Cells from mid-exponential phase

(OD600 ¼ 0.3–0.4) were diluted from 1.0 to 10�3, spotted (5 ll) ontoYNB plates containing MG at the indicated concentrations and in-

cubated at 30 �C for 2–4 days.

when longer times or higher MG concentrations were

tested (data not shown).

This result suggested that MG does not activate the

S. cerevisiae general stress response. In order to confirm

this idea, we analysed the induction of GRE3 in a dou-ble-disruption mutant msn2 msn4 and the parental

W303-1A strain. The transcriptional factors MSN2/

MSN4 have been reported to mediate the general stress

response in S. cerevisiae [16] and in particular the acti-

vation by different stress conditions of GRE3 and GLO1

[10,29,30]. As for the reference strain CEN.PK113-11A,

MG-triggered activation of GRE3 was observed for cells

of the W303-1A background, although maximal mRNAlevels in this strain were found when cells were chal-

lenged to 4 mMMG (Fig. 2(b)). Under these conditions,

induction of GRE3 was delayed, but not abolished, in

the msn2 msn4 double mutant. Thus, the nuclear factors,

Msn2p and Msn4p, appear to play an indirect role in the

response to MG.

To further analyse this possibility, we tested the effect

of the MSN2/MSN4 disruption on MG resistance. Asshown in Fig. 2(c), no obvious phenotype was found

when wild-type and msn2 msn4 mutant cells were spot-

ted on plates containing 4 mM MG, a concentration

that impairs growth in S. cerevisiae [10]. Similar results

were observed when a range of MG concentrations at

4–8 mM were tested (data not shown). Hence, the

MSN2/MSN4 gene function is not essential to trigger

the protective response to MG in S. cerevisiae.

3.3. Genes involved in glycerol and trehalose pathways

respond to MG

The results described above indicated that yeast cells

respond to MG by activating the expression of genes

which provide high MG-degradative capacity. We were

also interested to know if this protective response in-cludes the control of MG formation. In S. cerevisiae, the

synthesis of this metabolite appears to depend on the

internal content of triose phosphates. Therefore, we

analysed the MG-induced response of GPD1, the gene

encoding glycerol-3-phosphate dehydrogenase [31,32].

This activity transforms DHAP in glycerol and, thus,

redirects a part of the carbon flux away from glycolysis.

As depicted in Fig. 3(a), addition of 4 mM MG to aW303-1A yeast culture induced the expression of GPD1

to a level similar to that observed for GRE3 (Fig. 2(b)).

We also tested if the addition of MG could affect the

transcript level of TPS1 and TPS2, the genes coding for

the enzymes of the trehalose biosynthetic pathway

[33,34]. Indeed, the expression of TPS1 and TPS2 was

up-regulated by transferring cells to a MG-containing

medium (Fig. 3(b)). Finally, we studied if the effectsobserved in the GRE3 expression by deletion of MSN2

and MSN4 could be extended to other MG-respon-

sive genes. As shown in Fig. 3(a), the MG-induced

Fig. 4. The levels of MG depend on the activity of the glycerol path-

way. (a) Cells of the wild-type (W303-1A) strain were grown in YNB

and exposed to 4 mM MG, final concentration (black bars). Untreated

samples were used as control (grey bars). At the indicated times,

samples were taken and analysed for intracellular glycerol. (b) YNB-

grown cells of the W303-1A (grey bars) and gpd1 gpd2 mutant (black

bars) strains were subjected to osmotic stress (0.3 M NaCl) and

analysed for intracellular MG content. MG and glycerol assays were

performed as described in Section 2. Values in panels correspond to a

representative experiment. Three independent repetitions displayed the

same tendencies.

Fig. 3. The glycerol and trehalose pathway genes are induced by MG.

Cells of the W303-1A wild-type (}) and msn2 msn4 mutant (M) strain,

were transferred to 4 mM MG-containing medium. Samples were ta-

ken at the indicated times and analysed by Northern blot for mRNA

levels of GPD1 (a), TPS1 (b) and TPS2 (c) as described in Section 2.

Graphs represent mRNA levels relative to those of IPP1.

J. Aguilera, J.A. Prieto / FEMS Yeast Research 4 (2004) 633–641 637

up-regulation of GPD1 was again impaired, but not

abolished, in the msn2 msn4 mutant strain. Similar re-

sults were observed for TPS1 (Fig. 3(b)) and TPS2 (Fig.

3(c)). In the latter, a full induction in the mutant strain

was even observed after 60 min of MG exposure. Thus,Msn2p/Msn4p appears to affect the timing of the in-

duction but not the magnitude of the response.

3.4. The level of MG depends on the activity of the

glycerol pathway

The transcriptional activation of GPD1, TPS1 and

TPS2 in response to MG suggested that the activity ofglycolytic sinks at the upper part of glycolysis could

regulate the formation of this compound. To test this

hypothesis, we first examined the pattern of glycerol

production following a shift to 4 mM MG-containing

medium (Fig. 4(a)). As can be seen, the transfer trig-

gered the overproduction of glycerol after 3 h of MG

exposure. This observation is consistent with a role of

the glycerol pathway in the control of MG synthesis. To

further reinforce this idea, we analysed the levels of in-

tracellular MG in a yeast strain lacking GPD1 andGPD2. GPD2, the gene homologous to GPD1, is re-

sponsible for the synthesis of glycerol under anaerobic

conditions [32]. Thus, no glycerol production takes place

in a gpd1 gpd2 double mutant. As shown in Fig. 4(b),

internal MG levels were significantly higher, 24.2� 1.8

versus 9.6� 1.3 lM, in the gpd1 gpd2 mutant than in the

reference strain grown in YNB medium. These differ-

ences were much more pronounced after an osmotic-shock, a condition that stimulates MG formation [10].

Indeed, osmotic stress caused a moderate and transient

increase of MG in cells of the wild-type (Fig. 4(b)). This

variation was also observed in NaCl-stressed cells of the

gpd1 gpd2 mutant but, in this case, the overproduction

of MG was maintained longer (Fig. 4(b)).

We also tested if MG exposure could affect the level

of trehalose. This disaccharide is produced from glu-cose-6-phosphate, so its synthesis could function, like

that of glycerol, as a metabolic sink for intermediates at

the upper part of glycolysis. However, overproduction

of trehalose was not detected in MG-treated cells, even

after 3 h of MG exposure (data not shown). Attempts to

determine MG levels in a tps1 strain were also discon-

tinued due to the inability of mutant cells to grow on

glucose.

3.5. The MG response correlates with the intracellular

content of the toxic agent

The results shown above indicated that MG addition

enhances the intracellular levels of this compound,

suggesting that this variation could be the signal to ac-

tivate the transcriptional response. Further evidenceabout this was obtained by analysing GPD1 mRNA

levels and MG content in mutant cells of MG pathways

638 J. Aguilera, J.A. Prieto / FEMS Yeast Research 4 (2004) 633–641

(YPH250 background), exposed to the same concen-

tration of external MG. As expected, addition of MG to

YNB-grown cells of the YPH250 wild-type strain caused

a rapid increase in the intracellular MG content

(Fig. 5(a)), although the maximum value reached,around 20 lM, was clearly much lower than that found

for other reference strains (Fig. 1(c)). In consonance

with this, the induction level of GPD1 was also weaker

(Fig. 5(b)) than that depicted above (Fig. 2(a)). Inter-

estingly, lack of GLO1 resulted in a more pronounced

transcriptional activation of GPD1 (Fig. 5(b)). This ef-

fect correlated again with the intracellular level of MG,

which in these cells reached values eightfold higher thanthose observed for the wild-type strain after 40 min of

MG stimulation (Fig. 5(a)). Similar results were also

found for the glo1 gre3 double disruption mutant (data

not shown). Therefore, the increase of the intracellular

MG level and not the mere presence of exogenous MG

appeared to trigger the MG-induced transcriptional re-

sponse. These results also suggested that the metabolism

Fig. 5. Internal MG levels modulate the genetic response to the toxic

agent. YNB-grown cells of the wild-type (YPH250, grey bars, }) and

glo1 (YGL1, black bars, �) mutant strains were exposed to 4 mM MG

(final concentration) and analysed for intracellular MG (a) and GPD1

mRNA levels (b). Samples were taken at the indicated times. MG

determinations, total RNA preparation and mRNA quantification

were carried out as described in Section 2. Values in (a) are given as

means�SD of three independent experiments. The graphs in (b)

represent quantification of the mRNA levels of GPD1, relative to those

of IPP1.

of MG was not required to activate this regulatory

mechanism. To confirm this point, we tested if acetol or

1,2-propanediol, the main products of the aldose re-

ductase pathway, were able to induce the expression of

MG-responsive genes. None of these compounds causedany transcriptional activation, even when concentra-

tions of 4–16 mM were analysed (data not shown).

3.6. Enhanced glucose uptake triggers the MG response

Overproduction of MG has been reported to occur

under conditions resulting in a loss of control over

carbohydrate metabolism. According to this, it would beexpected that the MG response could also take place in

such situations, and not only after the addition of ex-

ogenous MG. To obtain further evidence about this

idea, we measured the levels of MG and mRNA of MG-

responsive genes in a yeast culture shifted from 0.2% to

2% glucose. It is well known from previous studies that

glucose concentrations below 20 mM (i.e., 0.36%), ac-

tivate the expression of the high-affinity glucose trans-port system [35] and that a sudden increase in glucose

availability induces an accelerated glucose uptake [36].

As shown in Fig. 6(a), the level of intracellular MG of

low-glucose-grown cells displayed a sharp increase after

the transfer, achieving its highest value, 17.3� 1.9 lM,

Fig. 6. Enhanced glucose uptake triggers the MG-dependent response.

A YNB-(0.2% glucose) growing W303-1A yeast culture was pulsed

with glucose (2%, final concentration). At the indicated times, cell

samples were taken and analysed for MG content (a) or GPD1 (}),

TPS1 (�), GRE3 (s) and GLO (M) mRNA levels (b). MG determi-

nations and Northern-blot analysis were performed as described in

Section 2. Values in (a) are means� SD of three independent experi-

ments. The graph in panel (b) represents quantification of the mRNA

levels of each gene, relative to those of IPP1.

J. Aguilera, J.A. Prieto / FEMS Yeast Research 4 (2004) 633–641 639

at 10 min. Later, the amount of MG started to decrease

reaching values around 10–11 lM after 20 min. This

amount is similar to that observed for 2%-glucose-

grown cells of the CEN.PK113-11A strain (Fig. 1(c),

control). Correlated with this variation, we observed aninduction in the expression of GPD1, TPS1 and GRE3

(Fig. 6(b)), suggesting again that the elevation in the

MG levels triggers the transcriptional response. How-

ever, the three genes followed different induction kinet-

ics. GPD1 up-regulation was early, around fivefold at

5 min, and remained high, threefold, during the time-

course of the experiment. TPS1 showed a gradual in-

duction, with a maximum, fourfold, at 40 min. GRE3was scarcely induced, and no variation was found for

the GLO1 mRNA.

4. Discussion

In this work, we show that MG exposure triggers a

specific cell response, focused on a particular set of geneswith functions in MG protection. GRE3 and GLO1 are

two key elements of the MG detoxification mechanism

in yeast, and lack of them produces MG sensitivity

[4,10].

MG exposure activated also the expression of GPD1,

TPS1 and TPS2. The up-regulation of GPD1 is logical if

the main purpose of this activation is to control the level

of triose-phosphates, from which MG is formed. As wedemonstrated, lack of glycerol production in a gpd1 gpd2

double mutant resulted in enhancedMG contents in cells

growing under non-stress conditions. Furthermore, this

effect was more pronounced in cells subjected to osmotic

shock, a stress situation that activates glucose metabo-

lism [11] and glycerol formation [15]. Consistent with

this, overproduction of glycerol was stimulated in

MG-exposed cells. Thus, our results suggest that up-regulation of glycerol synthesis provides a basic mecha-

nism to control the intracellular content of MG.

We noted, however, that MG exposure activated the

change in the glycerol pools only 3–4 h after MG ad-

dition, whereas GPD1 induction peaks at 30 min.

Analysis of internal MG levels in MG-exposed cultures

(Fig. 1) also demonstrated that yeast cells control the

MG content within 30 min of MG exposure. Thus, theseresults cast doubts on the functional role of the GPD1

activation. However, we rationalise that this response

could obey to the need to maintain a long-term potential

of defence against MG. Indeed, similar features can be

found in the well-known osmotic response of yeast.

First, the maximum level of glycerol synthesis is ob-

served around 2 h after the shift to high-osmolarity

environments, whereas the highest induction of GPD1 isfound much earlier, 15–30 min [23,31]. Second, although

the osmotic equilibrium is reached in a relatively short

time, the overproduction of the osmolyte is kept high till

the osmo-stress condition is suppressed [37]. As we

showed, extracellular MG levels remained almost in-

variable after the addition of MG to a yeast culture.

Thus, the induction of the glycerol pathway presumably

enhances the cell�s ability to manage a hostile high-MGenvironment after adaptation to growth under these

conditions.

Nevertheless, the induction by MG of GPD1 could

also provide additional mechanisms to control the level

of this compound. Indeed, the onset of GPD1 expres-

sion, and in particular of TPS1 and TPS2, could sug-

gest the existence of glycerol and trehalose cycles, since

MG exposure did not trigger the biosynthesis of thedisaccharide. Although the existence of these futile cy-

cles is unclear, activation of glycerol and trehalose cy-

cles has been claimed to play a protective role under

different stress conditions by down-regulating carbon

flux through the upper part of glycolysis [38–41]. Like

in these situations, the activation of futile cycles in

MG-exposed cells could provide a mechanism to con-

trol carbon flux by reducing the rate of MG formation.More work is, however, required to demonstrate the

function of futile cycles in MG-shocked cells.

The existence in Saccharomyces of a protective re-

sponse against MG suggested that dangerous internal

MG levels could be perceived as a signal to trigger the

transcriptional regulation. As we show, the exposition of

yeast cells to exogenous MG transiently increased the

internal content of the toxic agent. However, none of themain degradation products of MG, acetol or 1,2-pro-

panediol, was able to induce the expression of MG-

responsive genes. This suggested that no further

metabolism of MG was required to activate the tran-

scriptional response. In agreement with this, glo1 mu-

tant cells showed enhanced expression of GPD1 when

compared with the wild-type under the same stimulus.

Moreover, this response correlated with higher levels ofMG in the strain lacking GLO1 than in the reference

strain. Thus, our results suggest that intracellular rather

than extracellular MG is the signal to trigger the tran-

scriptional response. In this respect, it is interesting to

note that MG is not a substance that cells can easily

encounter in the environment, at least at the concen-

trations assayed in this study. It seems unlikely, there-

fore, that a specific protective mechanism againstexternal MG has been evolutionary developed.

Additional data evidencing that yeast cells could

monitor alterations in the internal MG level were

obtained by analysis of MG-responsive genes in low-

glucose-growing cells pulsed with 2% glucose, a situa-

tion that sharply enhances glucose consumption [42],

leading to a glycolytic overflux [36]. As we demon-

strated, a sudden increase in glucose availability causedMG over-accumulation. This is consistent with previ-

ous reports showing enhanced MG levels by overex-

pression of glycolytic genes, like PGI1 and PFK1 [43].

Fig. 7. The yeast�s protective response against MG. An increase in the

intracellular concentration of MG is perceived by the cell as a primary

signal to trigger a specific response, focused on a subset of genes,

GLO1 and GRE3, two key elements of the MG-detoxifying cellular

machinery, and GPD1, TPS1 and TPS2, the genes for the glycerol and

trehalose pathways. These could indirectly control de novo synthesis of

MG by eliminating glycolytic intermediates or by down-regulating

carbon flux (see text for details).

640 J. Aguilera, J.A. Prieto / FEMS Yeast Research 4 (2004) 633–641

Again, this variation activated the expression of TPS1,

TPS2 and GRE3. Moreover, this result indicates that

the MG response is triggered not only by the external

addition of MG, but also by small internal MG chan-

ges that do not detectably impair viability and growth.

Indeed, the maximal value of internal MG reachedafter addition of 1 mM MG was around 65 lM for the

CEN.PK113-11A strain, whereas a concentration not

higher than 12–13 lM was detected after pulsing with

glucose. Thus, these differences could account for the

lack of induction of GLO1 under such conditions. Like

for other stimuli and environmental conditions [44], the

MG-response shows a dose- and gene-dependency.

Overall, the results reported here led us to propose amodel for the MG-triggered defensive response (Fig. 7).

High internal MG levels are interpreted by the cell as a

signal that MG synthesis is too high and should be de-

creased. This triggers a dual response: first, enhancing

the cell machinery to eliminate the excess of the toxic

agent, and second, preventing new formation of MG by

activating the glycerol and trehalose pathways. How the

MG level is perceived as a signal and how it is transducedto induce the expression of genes are open questions. Our

results indicate that MG exposure does not activate the

general stress mechanism and that MSN2/MSN4, the

two nuclear factors mediating this response in Saccha-

romyces [16], have only an indirect effect on the induction

of MG-responsive genes. Similar results have been found

for subsets of genes induced by different stress condi-

tions, like heat shock or H2O2 [44]. In addition, no effectson MG resistance were found byMSN2/MSN4 deletion.

Thus, our results implicate additional signal transduc-

tion pathways and regulators of the MG response.

Whether MG is able to interact directly with elements

of a signal transduction pathway, or whether the

mechanism is indirect is also unknown. Nevertheless, theability of MG to modify proteins at the molecular level

[2] make this metabolite an obvious candidate to mod-

ulate putative targets of signal transduction pathways.

Acknowledgements

We thank F. Randez-Gil and F. Estruch for the

critical reading of this manuscript. We are also very

grateful to Y. Inoue, K.-D. Entian and F. Estruch for

providing the plasmids and yeast strains. This work wassupported by the Comisi�on Interministerial de Ciencia y

Tecnolog�ıa (Project AGL2001-1203).

References

[1] Kalapos, M.P. (1999) Methylglyoxal in living organisms: chem-

istry, biochemistry, toxicology and biological implications. Tox-

icol. Lett. 110, 145–175.

[2] Vander Jagt, D.L., Robinson, B., Taylor, K.K. and Hunsaker,

L.A. (1992) Reduction of trioses by NADPH-dependent aldo-keto

reductases, aldose reductase, methylglyoxal, and diabetic compli-

cations. J. Biol. Chem. 267, 4364–4369.

[3] Phillips, S.A. and Thornalley, P.J. (1993) The formation of

methylglyoxal from triose phosphates. Investigation using a

specific assay for methylglyoxal. Eur. J. Biochem. 212, 101–105.

[4] Inoue, Y. and Kimura, A. (1995) Methylglyoxal and regulation of

its metabolism in microorganisms. Adv. Microb. Physiol. 37, 177–

227.

[5] Murata, K., Fukuda, Y., Watanabe, K., Saikusa, T., Shimosaka,

M. and Kimura, A. (1985) Characterization of methylglyoxal

synthase in Saccharomyces cerevisiae. Biochem. Biophys. Res.

Commun. 131, 190–198.

[6] Ferguson, G.P., T€otemeyer, S., MacLean, M.J. and Booth, I.R.

(1998) Methylglyoxal production in bacteria: suicide or survival.

Arch. Microbiol. 170, 209–219.

[7] Shinohara, M., Thornalley, P.J., Giardino, I., Beisswenger, P.,

Thorpe, S.R., Onorato, J. and Brownlee, M. (1998) Overexpres-

sion of glyoxalase-I in bovine endothelial cells inhibits intracellu-

lar advanced glycation endproduct formation and prevents

hyperglycemia-induced increases in macromolecular endocytosis.

J. Clin. Invest. 101, 1142–1147.

[8] Brownlee, M. (1995) Advanced protein glycosylation in diabetes

and aging. Annu. Rev. Med. 46, 223–234.

[9] Abordo, E.A., Minhas, H.S. and Thornalley, P.J. (1999) Accu-

mulation of a oxaldehydes during oxidative stress: a role in

cytotoxicity. Biochem. Pharmacol. 58, 641–648.

[10] Aguilera, J. and Prieto, J.A. (2001) The Saccharomyces cerevisiae

aldose reductase is implied in the metabolism of methylglyoxal in

response to stress conditions. Curr. Genet. 39, 273–283.

[11] Inoue, Y., Tsujimoto, Y. and Kimura, A. (1998) Expression of the

glyoxalase I gene of Saccharomyces cerevisiae is regulated by high

osmolarity glycerol mitogen-acivated protein kinase pathway in

osmotic stress response. J. Biol. Chem. 273, 2977–2983.

J. Aguilera, J.A. Prieto / FEMS Yeast Research 4 (2004) 633–641 641

[12] Murata, K., Fukuda, Y., Shimosaka, M., Watanabe, K., Saikusa,

T. and Kimura, A. (1985) Metabolism of 2-oxoaldehyde in yeasts.

Purification and characterization of NADPH-dependent methyl-

glyoxal-reducing enzyme from Saccharomyces cerevisiae. Eur. J.

Biochem. 151, 631–636.

[13] Kuhn, A., van Zyl, C., van Tonder, A. and Prior, B.A. (1995)

Purification and partial characterization of an aldo-keto reductase

from Saccharomyces cerevisiae. Appl. Environ. Microbiol. 61,

1580–1585.

[14] O�Rourke, S.M., Herskowitz, I. and O�Shea, E.K. (2002) Yeast go

the whole HOG for the hyperosmotic response. Trends Genet. 18,

405–412.

[15] Hohmann, S. (2002) Osmotic stress signalling and osmoadapta-

tion in yeasts. Microbiol. Mol. Biol. Rev. 66, 300–372.

[16] Estruch, F. (2000) Stress-controlled transcription factors, stress-

induced genes and stress tolerance in budding yeast. FEMS

Microbiol. Rev. 24, 469–486.

[17] Margarida, A., Martins, T.B.S., Cordeiro, C.A. and Ponces-

Freire, A.M. (2001) In situ analysis of methylglyoxal metabolism

in Saccharomyces cerevisiae. FEBS Lett. 499, 41–44.

[18] Sherman, F., Fink, G.R. and Hicks, J.B. (1986) Methods in Yeast

Genetics. Cold Spring Harbor Laboratory, Cold Spring Harbor,

NY.

[19] Vogel, G., Aeschbacher, R.A., Muller, J., Boller, T. and Wiem-

ken, A. (1998) Trehalose-6-phosphate phosphatases from Arabid-

opsis thaliana: identification by functional complementation of the

yeast tps2 mutant. Plant J. 13, 673–683.

[20] Spevak, W., Fessl, F., Rytka, J., Traczy, K.A., Skoneczny, M. and

Ruis, H. (1983) Isolation of the catalase T structural gene of

Saccharomyces cerevisiae by functional complementation. Mol.

Cell. Biol. 3, 1545–1551.

[21] Susek, R.E. and Lindquist, S. (1990) Transcriptional derepression

of the Saccharomyces cerevisiae HSP26 gene during heat shock.

Mol. Cell. Biol. 10, 6362–6373.

[22] Sanchez, Y., Taulien, J., Borkovich, K.A. and Lindquist, S. (1992)

Hsp104 is required for tolerance to many forms of stress. EMBO

J. 11, 2357–2364.

[23] Rep, M., Albertyn, J., Thevelein, J., Prior, B.A. and Hohmann, S.

(1999) Different signalling pathways contribute to the control of

GPD1 gene expression by osmotic stress in Saccaromyces cerevi-

siae. Microbiology 145, 715–727.

[24] Gonzalez, B., Franc�ois, J. and Renaud, M. (1997) A rapid and

reliable method for metabolite extraction in yeast using boiling

buffered ethanol. Yeast 13, 1347–1356.

[25] Chaplen, F.W.R., Fahl, W.E. and Cameron, D.C. (1996) Method

for determination of free intracellular and extracellular methylgly-

oxal in animal cells grown in culture. Anal. Biochem. 238, 171–178.

[26] Cordeiro, C. and Ponces Freire, A. (1996) Methylglyoxal assay in

cells as 2-methylquinoxaline using 1,2-diaminobenzene as deriv-

atizing reagent. Anal. Biochem. 234, 221–224.

[27] Gaxiola, R., de Larrinoa, I.F., Villalba, J.M. and Serrano, R.

(1992) A novel and conserved salt-induced protein is an important

determinant of salt tolerance in yeast. EMBO J. 11, 3157–3164.

[28] Hounsa, C., Brandt, E.V., Thevelein, J., Hohmann, S. and Prior,

B.A. (1998) Role of trehalose in survival of Saccharomyces

cerevisiae under osmotic stress. Microbiology 144, 671–680.

[29] Garay-Arroyo, A. and Covarrubias, A.A. (1999) Three genes

whose expression is induced by stress in Saccharomyces cerevisiae.

Yeast 15, 879–892.

[30] Rep, M., Krantz, M., Thevelein, J.M. and Hohmann, S. (2000)

The transcriptional response of Saccharomyces cerevisiae to

osmotic shock. Hot1p and Msn2p/Msn4p are required for

the induction of subsets of high osmolarity glycerol pathway-

dependent genes. J. Biol. Chem. 275, 8290–8300.

[31] Albertyn, J., Hohmann, S., Thevelein, J.M. and Prior, B.A. (1994)

GPD1, which encodes glycerol-3-phosphate dehydrogenase, is

essential for growth under osmotic stress in Saccharomyces

cerevisiae, and its expression is regulated by the high-osmolarity

glycerol response pathway. Mol. Cell. Biol. 14, 4135–4144.

[32] Ansell, R., Granath, K., Hohmann, S., Thevelein, J.M. and Adler,

L. (1997) The two isoenzymes for yeast NAD+-dependent glycerol

3-phosphate dehydrogenase encoded by GPD1 and GPD2 have

distinct roles in osmoadaptation and redox regulation. EMBO J.

16, 2179–2187.

[33] Gonz�alez, M.I., Stucka, R., Bl�azquez, M.A., Feldmann, H. and

Gancedo, C. (1992) Molecular cloning of CIF1, a yeast gene

necessary for growth on glucose. Yeast 8, 183–192.

[34] De Virgilio, C., Burckert, N., Bell, W., Jeno, P., Boller, T. and

Wiemken, A. (1993) Disruption of TPS2, the gene encoding the

100-kDa subunit of the trehalose-6-phosphate synthase/phospha-

tase complex in Saccharomyces cerevisiae, causes accumulation of

trehalose-6-phosphate and loss of trehalose-6-phosphate phos-

phatase activity. Eur. J. Biochem. 212, 315–323.

[35] Diderich, J.A., Schepper, M., van Hoek, P., Luttik, M.A., van

Dijken, J.P., Pronk, J.T., Klaassen, P., Boelens, H.F.M., Teixeira

de Mattos, M.J., van Dam, K. and Kruckeberg, A.L. (1999)

Glucose uptake kinetics and transcription of HXT genes in

chemostat cultures of Saccharomyces cerevisiae. J. Biol. Chem.

274, 15350–15359.

[36] Teusink, B., Walsh, M.C., van Dam, K. and Westerhoff, H.V.

(1998) The danger of metabolic pathways with turbo design. TIBS

23, 162–169.

[37] Blomberg, A. and Adler, L. (1992) Physiology of osmotolerance in

fungi. Adv. Microb. Phys. 33, 145–212.

[38] Blomberg, A. (2000) Metabolic surprises in Saccharomyces

cerevisiae during adaptation to saline conditions: questions, some

answers and a model. FEMS Microbiol. Lett. 182, 1–8.

[39] €Olz, R., Larsson, K., Adler, L. and Gustafsson, L. (1993) Energy

flux and osmoregulation of Saccharomyces cerevisiae grown in

chemostats under NaCl stress. J. Bacteriol. 175, 2205–2213.

[40] Winderickx, J., de Winde, J.H., Crauwels, M., Hino, A.,

Hohmann, S., van Dijck, P. and Thevelein, J.M. (1996) Regula-

tion of genes encoding subunits of the trehalose synthase complex

in Saccharomyces cerevisiae: novel variations of STRE-mediated

transcription control? Mol. Gen. Genet. 252, 470–482.

[41] Parrou, J.L., Teste, M.A. and Franc�ois, J. (1997) Effects of

various types of stress on the metabolism of reserve carbohy-

drates in Saccharomyces cerevisiae: genetic evidence for a stress-

induced recycling of glycogen and trehalose. Microbiology 143,

1891–1900.

[42] Theobald, U., Mailinger, W., Reuss, M. and Rizzi, M. (1993) In

vivo analysis of glucose-induced fast changes in yeast adenine

nucleotide pool applying a rapid sampling technique. Anal.

Biochem. 214, 31–37.

[43] Inose, T. and Murata, K. (1995) Enhanced accumulation of toxic

compound in yeast cells having high glycolytic activity: a case

study on the safety of genetically engineered yeast. Int. J. Food

Sci. Technol. 30, 141–146.

[44] Gasch, A.P., Spellman, P.T., Kao, C.M., Carmel-Harel, O., Eisen,

M.B., Storz, G., Botstein, D. and Brown, P.O. (2000) Genomic

expression programs in the response of yeast cells to environmen-

tal changes. Mol. Biol. Cell 11, 4241–4257.

[45] Thomas, B.J. and Rothstein, R. (1989) Elevated recombination

rates in transcriptionally active DNA. Cell 56, 619–630.

[46] Estruch, F. and Carlson, M. (1993) Two homologous zinc finger

genes identified by multicopy suppression in a SNF1 protein

kinase mutant of Saccharomyces cerevisiae. Mol. Cell. Biol. 13,

3872–3881.