- Email: [email protected]
Glucose Signaling in Yeast Is Partially Mimicked by Galactose and Does Not Require the Tps1 Protein
Molecular Cell Biology Research Communications 1, 52–58 (1999) Article ID mcbr.1999.0112, available online at http://www.idealibrary.com on
Glucose Signaling in Yeast Is Partially Mimicked by Galactose and Does Not Require the Tps1 Protein Cristina Rodrı´guez and Juana M. Gancedo 1 Instituto de Investigaciones Biome´dicas “Alberto Sols” CSIC-UAM, Madrid 28029, Spain
Received February 22, 1999
pathway(s). The fact that distinct metabolic intermediates are required to trigger the induction of different glycolytic enzymes (4) supports the idea that there are multiple elements in glucose signaling. An approach to unravel the question of the existence of single or multiple signaling pathways and to identify putative molecules involved in them is to explore the effects of a different sugar, such as galactose, which is metabolized through initial steps distinct from those used by glucose and may therefore mimic only some of the effects produced by glucose. Regarding the elements involved in glucose signaling, it has been advanced that the product of the TPS1 gene, encoding trehalose-6P synthase, could play a role in glucose sensing (11), although the proposal of an involvement of Tps1 in glucose signaling was weakened by the results of Hohmann et al. (12) that suggested that the signaling defects observed in a tps1 mutant were a secondary effect of metabolic problems. However since these results were obtained in a hxk2 mutant, it could be argued that they were not conclusive because Hxk2 itself is involved in glucose signaling. To avoid this problem we have examined the consequences of eliminating Tps1 in a diploid HXK2/hxk2 strain which is still responsive to glucose. Our results show that galactose is able to produce many of the effects triggered by glucose; however in some cases both the rate and intensity of the response were lower. We show also that Tps1 is dispensable for glucose signaling.
Glucose produces multiple effects in Saccharomyces cerevisiae, as it controls the expression of many genes and the activity of various enzymes. However, the elements involved in glucose signaling are not well characterized. In this work the capacity of galactose to bring about the same effects than glucose has been assessed. Galactose mimics glucose only partially; it is suggested that it does not interact with a “sensor” in the plasma membrane and that it produces a weaker intracellular signal than glucose. To examine whether trehalose-6P synthase (Tps1) is required to transduce the glucose signal, we have constructed a tps1 hxk2/ tps1 HXK2 strain which, at difference of a tps1 strain, grows on glucose, and, at difference of a tps1 hxk2 strain, still possess the Hxk2 protein, possibly involved in glucose repression. From the response of this strain to glucose, we conclude that Tps1 does not play a prominent role in glucose signaling. © 1999 Academic Press
Glucose affects drastically the metabolism of Saccharomyces cerevisiae. The most conspicuous effects produced by the addition of glucose to a derepressed yeast culture are a transient increase in cAMP (1, 2), an increase in the rate of transcription rate of several genes (3, 4) and the repression of the transcription of a large number of genes (5, 6). Inactivation of gluconeogenic enzymes (7) and activation of others such as trehalase (8, 9) or plasma membrane ATPase (10) are also well documented. Although much information has accumulated on the mechanisms underlying the different glucose effects, it is not yet clear whether there is a single signaling pathway for all the effects produced and what are the elements involved in the signaling
MATERIALS AND METHODS Plasmids and yeast strains. Plasmid pBM284 contains a 3.3 kb HindIII-HindIII GAL80 fragment with TRP1 replacing the 1 kb EcoR1-EcoR1 fragment from GAL80 (13). From pBM284 we constructed plasmid pCR1 as follows: theBglII-BglII fragment containing the TRP1 gene and short GAL80 flanking sequences was replaced by the HIS3 gene and the resulting HindIII-HindIII gal80D::HIS3 fragment was placed in a Bluescript vector. Plasmid pMR124 has a 3.9 kb
Abbreviations used: FbPase, fructose-1,6-bisphosphatase; Gal, galactose; Gly, glycerol; Glu, glucose; GlutDH (NAD), NAD-dependent glutamate dehydrogenase. 1 Corresponding author. Instituto de Investigaciones Biome´dicas, Arturo Duperier 4, 28029 Madrid, Spain. Fax: 34-91-5854587. E-mail: [email protected].
1522-4724/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.
52
Vol. 1, No. 1, 1999
MOLECULAR CELL BIOLOGY RESEARCH COMMUNICATIONS TABLE 1
Yeast Strains Used in This Study Strain
Genotype
Reference
W303-1A W303-1B W303D WG80-1A WDC-3B WH1D WTH2D
MATa ade2 his3 leu2 ura3 trp1 MATa ade2 his3 leu2 ura3 trp1 MATa/MATa ade2/ade2 his3/his3 leu2/leu2 ura3/ura3 trp1/trp1 MATa ade2 his3 leu2 ura3 trp1 gal80::HIS3 MATa ade2 his3 leu2 ura3 trp1 tps::HIS3 MATa/MATa ade2/ade2 his3/his3 leu2/leu2 ura3/ura3 trp1/trp1 HXK2/hxk2::LEU2 MATa/MATa ade2/ade2 his3/his3 leu2/leu2 ura3/ura3 trp1/trp1 HXK2/hxk2::LEU2 tps1::HIS3/tps1::HIS3
35 35 This work This work 17 This work This work
To test FbPase inactivation by sugars, cells were first derepressed in a mineral medium with potassium acetate as described by Gancedo (16). Catabolite inactivation was triggered by adding to the derepressed cells glucose or galactose at a final concentration of 110 mM. To measure acidification of the medium in reponse to sugar addition, yeast cells at the exponential phase of growth were suspended in distilled water at 20 mg wet weight/ml and left overnight at 4°C. The yeast suspension was then centrifuged and the cells resuspended at the same concentration in 0.1 M KCl, 10 mM glycylglycine buffer pH4, at room temperature. The pH of the suspension was measured for a few minutes to establish a basal value before addition of the sugar. After addition of glucose or galactose (10 mM final concentration) pH readings were taken every minute.
EcoRI-EcoRI fragment containing the HXK2 gene cloned into pBR328 (14). Plasmid pMR226, provided by M. Rose (Universita¨t Frankfurt), is derived from pMR124 by substitution of the SphI-XbaI fragment containing the complete reading frame and flanking sequences of the HXK2 gene by the LEU2 gene. The yeast strains used in this work are described in Table 1. To disrupt the GAL80 gene in strain W303-1A, the yeast was transformed with the 4kb EcoR1-XhoI fragment from pCR1 (which contains GAL80 disrupted by HIS3) and transformants were selected in a medium lacking histidine. We checked that the chosen transformant, WG80-1A, expressed high levels of galactokinase when grown on YPglycerol. To construct a series of isogenic diploid strains we crossed first strains W303-1A and WDC-3B. In the diploid obtained we disrupted one copy of HXK2 using a 2.9 kb HindIII-PvuII fragment of plasmid pMR226 (which contains the HXK2 gene disrupted by LEU2). The construction was verified by Southern analysis, the TPS1 HXK2/tps1D hxk2D diploid was sporulated and segregants tps1D hxk2D and TPS1 hxk2D were isolated. Strain WH1D was obtained by crossing a TPS1 hxk2D segregant with strain W303-1A, and strain WTH2D was the product of a cross between a tps1D hxk2D segregant and strain WDC-3B. The diploids were rechecked by Southern analysis. The control strain W303D is the product of a cross between W303-1A and W303-1B. Transformation of S.cerevisiae was performed as described by Ito et al. (15).
Preparation of extracts and enzymatic tests. Yeast extracts were prepared by shaking with glass beads as described (17). FbPase and NAD-dependent glutamate dehydrogenase [GlutDH (NAD)] were tested spectrophotometrically (18, 19); invertase was assayed as Goldstein and Lampen (20), but using whole cells and measuring the glucose formed in the reaction with hexokinase and glucose-6P dehydrogenase. Protein was determined by the method of Lowry et al. (21) using the BCA protein-assay reagent (Pierce). Determination of cAMP and other metabolites. To measure intracellular metabolites, yeast cells were collected by rapid filtration, frozen in liquid nitrogen and extracted according to Gamo et al. (22). cAMP was determined in the extracts as in Eraso and Gancedo (23) and other metabolites were assayed spectrophotometrically as described by Bergmeyer (24).
Growth conditions, invertase derepression, fructose1,6-bisphosphatase (FbPase) inactivation and measure of acidification of the medium. The yeasts were grown in a rich medium (YP) containing 1% yeast extract and 2% peptone; 2% glucose, 2% galactose, 2% glycerol or 2% ethanol were used as carbon sources. The yeasts were collected at the exponential phase of growth. To measure derepression of invertase, the yeasts were grown in YPglucose and resuspended in either YP2%glucose or YP0.05% glucose and incubated for 3 h, or grown in YPgalactose and resuspended in either YP2%galactose or YP0.05% galactose.
RESULTS AND DISCUSSION When comparing the effects of adding glucose or galactose to yeast cells grown on a non-sugar carbon source, it should be taken into account that such cells metabolize glucose readily but cannot use galactose 53
Vol. 1, No. 1, 1999
MOLECULAR CELL BIOLOGY RESEARCH COMMUNICATIONS
FIG. 1. Effects of the addition of glucose (E) or galactose (F) to derepressed cells of S. cerevisiae WG80-1A ( gal80D). (A) Changes in cAMP levels (B) Two-step inactivation of FbPase (C) Induction of pyruvate carboxylase. See Materials and Methods for details.
until the proteins required for galactose catabolism have been induced. This hinders the production of short-term effects by galactose and can affect the conclusions derived from experiments of this kind (25). To circumvent this problem we made use of a gal80 strain, in which the GAL genes no longer require galactose to be expressed, although they remain glucose repressible (26). Using this strain we looked at the changes in cAMP levels caused by the addition of glucose or galactose to derepressed yeast cells. As shown in Fig. 1A, the rapid increase in cAMP produced by galactose is not as marked as that triggered by glucose, confirming earlier results (27). We have studied also the inactivation of FbPase and the activation of plasma membrane ATPase produced by the two sugars. Catabolite inactivation of FbPase consists of two consecutive processes: a short-term reversible inactivation due to phosphorylation, which occurs in a few minutes and a long-term irreversible inactivation due to proteolysis of the protein which takes 1 to 2 hours (7). Short-term inactivation of FbPase is elicited to a similar degree by glucose or galactose (Fig. 1B), the subsequent proteolytic inactivation, however, proceeds at a slower rate in the presence of galactose.
An indirect measure of the plasma membrane ATPase activation is the rate of acidification of the medium caused by the sugar. An interesting difference was found between glucose and galactose: while glucose triggered acidification in cells grown in fermentable or non-fermentable carbon sources, galactose acted on cells grown in galactose or glycerol but not on cells grown in glucose or ethanol (Fig. 2). The results observed with glucose-grown cells can be explained by the repression of the GAL genes by glucose but the behaviour of the ethanol-grown cells has no straightforward explanation. The transcription of several genes encoding glycolytic enzymes is activated by glucose (4); a typical example is the increased transcription of the PDC1 gene which causes a large increase in the activity of pyruvate decarboxylase. Galactose caused pyruvate decarboxylase induction, but both the rate of synthesis of the enzyme and the final level reached were lower that upon glucose addition (Fig. 1C). The repression of different enzymes by glucose and galactose has been also compared. As shown in Table 2, FbPase is repressed by galactose as strongly as by glucose and there is also a marked repression of glutamate dehydrogenase by galactose, only slightly weaker 54
Vol. 1, No. 1, 1999
MOLECULAR CELL BIOLOGY RESEARCH COMMUNICATIONS
FIG. 2. Effects of the addition of glucose (E) or galactose (F) to derepressed cells of S. cerevisiae WG80-1A ( gal80D) on proton excretion to the medium. Yeast was grown on YPglucose (A) YPgalactose (B) YPglycerol (C) or YPethanol (D) until early exponential phase (3-4 mg wet weight/ml) and starved at 4°C as described in Materials and Methods. Glucose or galactose were then added to aliquots of the yeast suspension (10 mM final concentration) and the pH of the medium was followed along time.
than that produced by glucose. In the case of invertase, while the enzyme is partly derepressed in yeast grown on 2% galactose compared with yeast grown on 2% glucose, the enzyme cannot be derepressed at all by
incubation on 0.05% galactose, in conditions where 0.05% glucose allows a strong derepression. How can we interpret the similarities and differences in the effects produced by glucose and galactose? There are two basic possibilities, not mutually exclusive, for the mode of action of the sugars: they may interact with a membrane protein which would then modify its conformation or that of a neighboring protein, triggering some type of regulatory cascade; they may change the levels of intermediate metabolites which would allosterically activate or inhibit some regulatory protein(s). Regarding the first possibility there is now ample evidence for the glucose transporter homologs Snf3 and Rgt2 acting as glucose sensors (28, 29). These proteins are required for the induction by glucose of different HXT genes and for catabolite repression of GAL2, SUC2, FBP1, ICL1 or GDH2 (29, Lafuente, M. J., Gancedo, C., and Gancedo, J. M., unpublished). In addition the Ras/cAMP pathway which is activated by glucose (2) appears to play a role in activating the transcription of PFK26 and PFK27 (25) and in repressing the transcription of a number of genes subject to
TABLE 2
Repression of Some Enzymes by Glucose and Galactose in S. cerevisiae FbPase a GlutDH (NAD) a (nmol/min/mg protein) Glucose Galactose Glycerol
,1 ,1 52
17 26 166
Invertase b (nmol/min/mg yeast) 2% 0.05% 2% 0.05%
Glucose Glucose Galactose Galactose
12 227 49 29
a
The S. cerevisiae strain WG80-1A ( gal80D) was grown on YP with the carbon source indicated, collected at the exponential phase of growth and enzymes were tested as described in Materials and Methods. Data are averages of 2 experiments. b The yeast was grown in YPglucose or YPgalactose and derepression and testing of invertase were carried out as described in Materials and Methods, in the media indicated. Data are averages of at least 3 experiments.
55
Vol. 1, No. 1, 1999
MOLECULAR CELL BIOLOGY RESEARCH COMMUNICATIONS
intermediary metabolites of the glycolytic pathway (32). It has been shown that different intermediates are required for the activation of transcription of different genes (4, 25) and that the expression of gluconeogenic genes increases in fbp1 mutants where the concentration of hexose phosphates is lower than in a wild-type strain when the yeast utilizes a non-sugar carbon source (Mc Cammon, personal communication, J. M. Gancedo, unpublished) As we have shown in this paper and as reported recently by others (25), galactose may mimic some of the effects of glucose. Galactose, however, is unable to induce invertase at low concentration (Table 2) or to induce glucose transporter genes such as HXT1 (3). Galactose is therefore unlikely to interact with Snf3 or Rgt2; this is further supported by the observation that in a snf3rgt2 mutant repression by glucose is relieved, while repression by galactose is not (Lafuente, M. J., Gancedo, C., and Gancedo, J. M., unpublished). The existence of a galactose receptor, equivalent to Snf3 or Rgt2, cannot be discarded but it is made unlikely by the fact that, after the sequencing of the S.cerevisiae genome has been completed, it has been found that all proteins with sugar transporter features have short cytoplasmic C-tails (33), in contrast with the long
TABLE 3
Changes in Metabolite Levels upon Addition of Glucose or Galactose to Derepressed Yeast Cells a Glucose-6P Addition
Time
— Glucose Glucose Galactose Galactose
0 1.5 3 1.5 3
h h h h h
Fructose-6P
Fructose-1,6P 2
(mM) 0.64 1.63 1.07 1.46 1.65
0.32 0.37 0.27 0.40 0.57
0.1 6 4 2 1.5
a S. cerevisiae WG80-1A ( gal80D) was grown on YPethanol; glucose or galactose were added to the culture and samples were taken at the times indicated (see Materials and Methods for details). Data are averages of 2 experiments.
catabolite repression (Zaragoza, O., Lindley, C., and Gancedo, J. M., submitted) although cAMP does not appear involved in the control of other genes regulated by glucose (25, 30, 31, Zaragoza, O., Lindley, C., and Gancedo, J. M., submitted). On the other hand when glucose is available to the cell there is a marked increase in the concentration of
FIG. 3. Effects of the addition of glucose to a set of isogenic strains of S. cerevisiae, W303D (wt, F), WH1D (HXK2/hxk2, ‚) and WTH2D (hxk2/hxk2, E). (A) Changes in cAMP levels. (B) Two-step inactivation of FbPase. (C) Induction of pyruvate carboxylase. (D) Proton excretion to the medium. See Materials and Methods for details.
56
Vol. 1, No. 1, 1999
MOLECULAR CELL BIOLOGY RESEARCH COMMUNICATIONS TABLE 4
Repression by Glucose of Some Enzymes in a Set of Isogenic Yeast Strains a FbPase
GlutDH (NAD) (nmol/min/mg protein)
Relevant genotype
Wild-type hxk2/HXK2 tps1hxk2/tps1HXK2
Invertase (nmol/min/mg yeast)
Glu
Gly
Glu
Gly
2% Glu
0.05% Glu
,1 ,1 1
45 45 30
12 14 10
200 210 170
20 110 30
410 440 620
a Enzyme levels were measured in repressed or derepressed conditions as described in Materials and Methods. Data are averages of at least 3 independent experiments.
strain (Fig. 3D). In addition FbPase and glutamate dehydrogenase were repressed to the same extent in the different strains (Table 4). Curiously, while repression of invertase by glucose was partially relieved in the diploid strain with a single copy of HXK2, in the strain which lacked also Tps1 the repression by glucose was recovered. From our results we may conclude that there is at present no decisive evidence for Tps1 playing a role in glucose signaling.
C-tails of Snf3 and Rgt2 which are thought to interact with further regulatory proteins (29). Galactose signaling is, therefore, more likely to take place through an increase in the levels of glycolytic intermediates. We have compared the concentration of glucose6P, fructose-6P and fructose-1,6P 2 in yeast cells in the presence of glucose or galactose (Table 3). The levels of glucose-6P and fructose-6P are very similar in both cases while fructose-1,6P 2 is 2-3 fold higher with glucose than with galactose. Although it is still premature to propose that this difference accounts for the different behaviour of both sugars, it is likely that galactose produces smaller changes than glucose in the concentration of the still unidentified regulatory metabolite(s) and produces therefore a weaker signal. In the case of glucose the question remains open regarding the identity of the proteins which relay the glucose signal from Snf3 and Rgt2 although there is preliminary evidence for Mth1 and Std1/Msn3 being involved (Zhang, X., Solimeo, H. T., and Schmidt, M. C., personal communication; Lafuente, M. J., Gancedo, C., and Gancedo, J. M., unpublished). It has been hypothesized that the protein Tps1, a subunit of the trehalose synthase complex which catalyzes the synthesis of trehalose-6P, could participate in glucose signaling (34), although the fact that this signaling was restored in a tps1hxk2 double mutant indicated that Tps1 could be dispensable in a genetic background where catabolite repression of some genes had been eliminated (12). To reexamine to what extent the Tps1 protein could be involved in the effects produced by glucose, we have compared the behaviour of 3 diploid isogenic strains (wt, hxk2/HXK2 and tps1 hxk2/tps1 HXK2) upon glucose addition. We have found that Tps1 is not required for the rapid increase in cAMP (Fig. 3A) nor for induction of pyruvate decarboxylase (Fig. 3B) or for acidification of the medium (Fig. 3C). The short-term inactivation of FbPase is also independent of Tps1 and there is only some slowing down of the proteolytic degradation of FbPase in the tps1
ACKNOWLEDGMENTS We thank M. Johnston for plasmid pBM284 and M. Rose for plasmid pMR226. This work was supported by Grant PB094-0091CO2-01 from the Direccio´n General de Investigacio´n Cientı´fica y Te´cnica. C.R. had a fellowship from the Fundacio´n Ramo´n Areces.
REFERENCES 1. Mazo´n, M. J., Gancedo, J. M., and Gancedo, C. (1982) Eur. J. Biochem. 127, 605– 608. 2. Thevelein, J. M. (1994) Yeast 10, 1753–1790. ¨ zcan, S., and Johnston, M. (1995) Mol. Cell Biol. 15, 1564 – 3. O 1572. 4. Boles, E., Mu¨ller, S., and Zimmermann, F. K. (1996) Mol. Microbiol. 19, 641– 642. 5. Johnston, M., and Carlson, M. (1992) in The Molecular and Cellular Biology of the Yeast Saccharomyces (Jones, E. W., Pringle, J. R., and Broach, J. R., Eds.), Vol. 2, pp. 193–281 Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. 6. Gancedo, J. M. (1998) Microbiol. Mol. Biol. Rev. 62, 334 –361. 7. Gancedo, J. M., and Gancedo, C. (1997) in Yeast Sugar Metabolism (Zimmermann, F. K., and Entian, K.-D., Eds.), pp. 359 –377. Technomic, Lancaster, PA. 8. Van der Plaat, J. B. (1974) Biochem. Biophys. Res. Commun. 56, 580 –587. 9. Uno, I., Matsumoto, K., Adachi, K., and Ishikawa, T. (1983) J. Biol. Chem. 258, 10867–10872. 10. Serrano, R. (1983) FEBS Lett. 156, 11–14. 11. Van Aelst, L., Hohmann, S., Bulaya, B., de Koning, W., Sierkstra, L., Neves, M. J., Luyten, K., Alijo, R., Ramos, J., Coccetti, P., Martegani, E., de Magelhaes, N. M., Brandao, R. L., Van Dijck, P., Vanhalewyn, M., Durnez, P., Jans, A. W. H., and Thevelein, J. M. (1993) Mol. Microbiol. 8, 927–943.
57
Vol. 1, No. 1, 1999
MOLECULAR CELL BIOLOGY RESEARCH COMMUNICATIONS
12. Hohmann, S., Neves, M. J., de Koning, W., Alijo, R., Ramos, J., and Thevelein, J. M. (1993) Curr. Genet. 23, 281–289. 13. Yocum, R. R., and Johnston, M. (1984) Gene 32, 75– 82. 14. Balbas, P., Soberon, X., Merino, E., Zurita, M., Lomerli, H., Valle, F., Flores, N., and Bolivar, F. (1986) Gene 50, 3– 40. 15. Ito, H., Fukuda, Y., Murata, K. and Kimura, A. (1983) J. Bacteriol. 153, 163–168. 16. Gancedo, C. (1971) J. Bacteriol. 107, 401– 405. 17. Bla´zquez, M. A., Lagunas, R. Gancedo, C., and Gancedo, J. M. (1993) FEBS Lett. 329, 51–54. 18. Funayama, S., Gancedo, J. M., and Gancedo, C. (1980) Eur. J. Biochem. 109, 61– 66. 19. Doherty, D. (1970) Methods Enzymol. 17A, 850 – 856. 20. Goldstein, A., and Lampen, J. O. (1975) Methods Enzymol. 42, 504 –511. 21. Lowry, O. H., Rosebrough, N. J., Farr, A. L., and Randall, R. S. (1951) J. Biol. Chem. 193, 265–275. 22. Gamo, F.-J., Portillo, F., and Gancedo, C. (1983) FEMS Microbiol. Lett. 106, 233–238. 23. Eraso, P., and Gancedo, J. M. (1984) Eur. J. Biochem. 141, 195–198.
24. Bergmeyer, H. U. (1983) Methods of Enzymatic Analysis. Verlag Chemie, Weinheim, Germany. 25. Gonc¸alves, P. M., Griffioen, G., Bebelman, J. P., and Planta, R. J. (1997) Mol. Microbiol. 25, 483– 493. 26. Torchia, T. E., Hamilton, R. W., Cano, C. L., and Hopper, J. E. (1984) Mol. Cell. Biol. 4, 1521–1527. 27. Eraso, P., and Gancedo, J. M. (1985) FEBS Lett. 191, 51–54. ¨ zcan, S., Dover, J., Rosenwald, A. G., Wo¨lfl, S., and Johnston, 28. O M. (1996) Proc. Natl. Acad. Sci. USA 93, 12428 –12432. ¨ zcan, S., Dover, J., and Johnston, M. (1998) EMBO J. 17, 29. O 2566 –2573. 30. Matsumoto, K., Uno, I., Toh-e, A., Ishikawa, T., and Oshima, Y. (1982) J. Bacteriol. 150, 277–285. 31. Matsumoto, K., Uno, I., Ishikawa, T., and Oshima, Y. (1983) J. Bacteriol. 156, 898 –900. 32. Gancedo, J. M., and Gancedo, C. (1973) 55, 205–211. 33. Boles, E., and Hollenberg, C. P. (1997) FEMS Microbiol. Rev. 21, 85–111. 34. Thevelein, J. M. (1992) Antonie van Leeuwenhoek 62, 109 – 130. 35. Thomas, B. J., and Rothstein, R. (1989) Cell 56, 619 – 630.
58
Glucose Signaling in Yeast Is Partially Mimicked by Galactose and Does Not Require the Tps1 Protein Cristina Rodrı´guez and Juana M. Gancedo 1 Instituto de Investigaciones Biome´dicas “Alberto Sols” CSIC-UAM, Madrid 28029, Spain
Received February 22, 1999
pathway(s). The fact that distinct metabolic intermediates are required to trigger the induction of different glycolytic enzymes (4) supports the idea that there are multiple elements in glucose signaling. An approach to unravel the question of the existence of single or multiple signaling pathways and to identify putative molecules involved in them is to explore the effects of a different sugar, such as galactose, which is metabolized through initial steps distinct from those used by glucose and may therefore mimic only some of the effects produced by glucose. Regarding the elements involved in glucose signaling, it has been advanced that the product of the TPS1 gene, encoding trehalose-6P synthase, could play a role in glucose sensing (11), although the proposal of an involvement of Tps1 in glucose signaling was weakened by the results of Hohmann et al. (12) that suggested that the signaling defects observed in a tps1 mutant were a secondary effect of metabolic problems. However since these results were obtained in a hxk2 mutant, it could be argued that they were not conclusive because Hxk2 itself is involved in glucose signaling. To avoid this problem we have examined the consequences of eliminating Tps1 in a diploid HXK2/hxk2 strain which is still responsive to glucose. Our results show that galactose is able to produce many of the effects triggered by glucose; however in some cases both the rate and intensity of the response were lower. We show also that Tps1 is dispensable for glucose signaling.
Glucose produces multiple effects in Saccharomyces cerevisiae, as it controls the expression of many genes and the activity of various enzymes. However, the elements involved in glucose signaling are not well characterized. In this work the capacity of galactose to bring about the same effects than glucose has been assessed. Galactose mimics glucose only partially; it is suggested that it does not interact with a “sensor” in the plasma membrane and that it produces a weaker intracellular signal than glucose. To examine whether trehalose-6P synthase (Tps1) is required to transduce the glucose signal, we have constructed a tps1 hxk2/ tps1 HXK2 strain which, at difference of a tps1 strain, grows on glucose, and, at difference of a tps1 hxk2 strain, still possess the Hxk2 protein, possibly involved in glucose repression. From the response of this strain to glucose, we conclude that Tps1 does not play a prominent role in glucose signaling. © 1999 Academic Press
Glucose affects drastically the metabolism of Saccharomyces cerevisiae. The most conspicuous effects produced by the addition of glucose to a derepressed yeast culture are a transient increase in cAMP (1, 2), an increase in the rate of transcription rate of several genes (3, 4) and the repression of the transcription of a large number of genes (5, 6). Inactivation of gluconeogenic enzymes (7) and activation of others such as trehalase (8, 9) or plasma membrane ATPase (10) are also well documented. Although much information has accumulated on the mechanisms underlying the different glucose effects, it is not yet clear whether there is a single signaling pathway for all the effects produced and what are the elements involved in the signaling
MATERIALS AND METHODS Plasmids and yeast strains. Plasmid pBM284 contains a 3.3 kb HindIII-HindIII GAL80 fragment with TRP1 replacing the 1 kb EcoR1-EcoR1 fragment from GAL80 (13). From pBM284 we constructed plasmid pCR1 as follows: theBglII-BglII fragment containing the TRP1 gene and short GAL80 flanking sequences was replaced by the HIS3 gene and the resulting HindIII-HindIII gal80D::HIS3 fragment was placed in a Bluescript vector. Plasmid pMR124 has a 3.9 kb
Abbreviations used: FbPase, fructose-1,6-bisphosphatase; Gal, galactose; Gly, glycerol; Glu, glucose; GlutDH (NAD), NAD-dependent glutamate dehydrogenase. 1 Corresponding author. Instituto de Investigaciones Biome´dicas, Arturo Duperier 4, 28029 Madrid, Spain. Fax: 34-91-5854587. E-mail: [email protected].
1522-4724/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.
52
Vol. 1, No. 1, 1999
MOLECULAR CELL BIOLOGY RESEARCH COMMUNICATIONS TABLE 1
Yeast Strains Used in This Study Strain
Genotype
Reference
W303-1A W303-1B W303D WG80-1A WDC-3B WH1D WTH2D
MATa ade2 his3 leu2 ura3 trp1 MATa ade2 his3 leu2 ura3 trp1 MATa/MATa ade2/ade2 his3/his3 leu2/leu2 ura3/ura3 trp1/trp1 MATa ade2 his3 leu2 ura3 trp1 gal80::HIS3 MATa ade2 his3 leu2 ura3 trp1 tps::HIS3 MATa/MATa ade2/ade2 his3/his3 leu2/leu2 ura3/ura3 trp1/trp1 HXK2/hxk2::LEU2 MATa/MATa ade2/ade2 his3/his3 leu2/leu2 ura3/ura3 trp1/trp1 HXK2/hxk2::LEU2 tps1::HIS3/tps1::HIS3
35 35 This work This work 17 This work This work
To test FbPase inactivation by sugars, cells were first derepressed in a mineral medium with potassium acetate as described by Gancedo (16). Catabolite inactivation was triggered by adding to the derepressed cells glucose or galactose at a final concentration of 110 mM. To measure acidification of the medium in reponse to sugar addition, yeast cells at the exponential phase of growth were suspended in distilled water at 20 mg wet weight/ml and left overnight at 4°C. The yeast suspension was then centrifuged and the cells resuspended at the same concentration in 0.1 M KCl, 10 mM glycylglycine buffer pH4, at room temperature. The pH of the suspension was measured for a few minutes to establish a basal value before addition of the sugar. After addition of glucose or galactose (10 mM final concentration) pH readings were taken every minute.
EcoRI-EcoRI fragment containing the HXK2 gene cloned into pBR328 (14). Plasmid pMR226, provided by M. Rose (Universita¨t Frankfurt), is derived from pMR124 by substitution of the SphI-XbaI fragment containing the complete reading frame and flanking sequences of the HXK2 gene by the LEU2 gene. The yeast strains used in this work are described in Table 1. To disrupt the GAL80 gene in strain W303-1A, the yeast was transformed with the 4kb EcoR1-XhoI fragment from pCR1 (which contains GAL80 disrupted by HIS3) and transformants were selected in a medium lacking histidine. We checked that the chosen transformant, WG80-1A, expressed high levels of galactokinase when grown on YPglycerol. To construct a series of isogenic diploid strains we crossed first strains W303-1A and WDC-3B. In the diploid obtained we disrupted one copy of HXK2 using a 2.9 kb HindIII-PvuII fragment of plasmid pMR226 (which contains the HXK2 gene disrupted by LEU2). The construction was verified by Southern analysis, the TPS1 HXK2/tps1D hxk2D diploid was sporulated and segregants tps1D hxk2D and TPS1 hxk2D were isolated. Strain WH1D was obtained by crossing a TPS1 hxk2D segregant with strain W303-1A, and strain WTH2D was the product of a cross between a tps1D hxk2D segregant and strain WDC-3B. The diploids were rechecked by Southern analysis. The control strain W303D is the product of a cross between W303-1A and W303-1B. Transformation of S.cerevisiae was performed as described by Ito et al. (15).
Preparation of extracts and enzymatic tests. Yeast extracts were prepared by shaking with glass beads as described (17). FbPase and NAD-dependent glutamate dehydrogenase [GlutDH (NAD)] were tested spectrophotometrically (18, 19); invertase was assayed as Goldstein and Lampen (20), but using whole cells and measuring the glucose formed in the reaction with hexokinase and glucose-6P dehydrogenase. Protein was determined by the method of Lowry et al. (21) using the BCA protein-assay reagent (Pierce). Determination of cAMP and other metabolites. To measure intracellular metabolites, yeast cells were collected by rapid filtration, frozen in liquid nitrogen and extracted according to Gamo et al. (22). cAMP was determined in the extracts as in Eraso and Gancedo (23) and other metabolites were assayed spectrophotometrically as described by Bergmeyer (24).
Growth conditions, invertase derepression, fructose1,6-bisphosphatase (FbPase) inactivation and measure of acidification of the medium. The yeasts were grown in a rich medium (YP) containing 1% yeast extract and 2% peptone; 2% glucose, 2% galactose, 2% glycerol or 2% ethanol were used as carbon sources. The yeasts were collected at the exponential phase of growth. To measure derepression of invertase, the yeasts were grown in YPglucose and resuspended in either YP2%glucose or YP0.05% glucose and incubated for 3 h, or grown in YPgalactose and resuspended in either YP2%galactose or YP0.05% galactose.
RESULTS AND DISCUSSION When comparing the effects of adding glucose or galactose to yeast cells grown on a non-sugar carbon source, it should be taken into account that such cells metabolize glucose readily but cannot use galactose 53
Vol. 1, No. 1, 1999
MOLECULAR CELL BIOLOGY RESEARCH COMMUNICATIONS
FIG. 1. Effects of the addition of glucose (E) or galactose (F) to derepressed cells of S. cerevisiae WG80-1A ( gal80D). (A) Changes in cAMP levels (B) Two-step inactivation of FbPase (C) Induction of pyruvate carboxylase. See Materials and Methods for details.
until the proteins required for galactose catabolism have been induced. This hinders the production of short-term effects by galactose and can affect the conclusions derived from experiments of this kind (25). To circumvent this problem we made use of a gal80 strain, in which the GAL genes no longer require galactose to be expressed, although they remain glucose repressible (26). Using this strain we looked at the changes in cAMP levels caused by the addition of glucose or galactose to derepressed yeast cells. As shown in Fig. 1A, the rapid increase in cAMP produced by galactose is not as marked as that triggered by glucose, confirming earlier results (27). We have studied also the inactivation of FbPase and the activation of plasma membrane ATPase produced by the two sugars. Catabolite inactivation of FbPase consists of two consecutive processes: a short-term reversible inactivation due to phosphorylation, which occurs in a few minutes and a long-term irreversible inactivation due to proteolysis of the protein which takes 1 to 2 hours (7). Short-term inactivation of FbPase is elicited to a similar degree by glucose or galactose (Fig. 1B), the subsequent proteolytic inactivation, however, proceeds at a slower rate in the presence of galactose.
An indirect measure of the plasma membrane ATPase activation is the rate of acidification of the medium caused by the sugar. An interesting difference was found between glucose and galactose: while glucose triggered acidification in cells grown in fermentable or non-fermentable carbon sources, galactose acted on cells grown in galactose or glycerol but not on cells grown in glucose or ethanol (Fig. 2). The results observed with glucose-grown cells can be explained by the repression of the GAL genes by glucose but the behaviour of the ethanol-grown cells has no straightforward explanation. The transcription of several genes encoding glycolytic enzymes is activated by glucose (4); a typical example is the increased transcription of the PDC1 gene which causes a large increase in the activity of pyruvate decarboxylase. Galactose caused pyruvate decarboxylase induction, but both the rate of synthesis of the enzyme and the final level reached were lower that upon glucose addition (Fig. 1C). The repression of different enzymes by glucose and galactose has been also compared. As shown in Table 2, FbPase is repressed by galactose as strongly as by glucose and there is also a marked repression of glutamate dehydrogenase by galactose, only slightly weaker 54
Vol. 1, No. 1, 1999
MOLECULAR CELL BIOLOGY RESEARCH COMMUNICATIONS
FIG. 2. Effects of the addition of glucose (E) or galactose (F) to derepressed cells of S. cerevisiae WG80-1A ( gal80D) on proton excretion to the medium. Yeast was grown on YPglucose (A) YPgalactose (B) YPglycerol (C) or YPethanol (D) until early exponential phase (3-4 mg wet weight/ml) and starved at 4°C as described in Materials and Methods. Glucose or galactose were then added to aliquots of the yeast suspension (10 mM final concentration) and the pH of the medium was followed along time.
than that produced by glucose. In the case of invertase, while the enzyme is partly derepressed in yeast grown on 2% galactose compared with yeast grown on 2% glucose, the enzyme cannot be derepressed at all by
incubation on 0.05% galactose, in conditions where 0.05% glucose allows a strong derepression. How can we interpret the similarities and differences in the effects produced by glucose and galactose? There are two basic possibilities, not mutually exclusive, for the mode of action of the sugars: they may interact with a membrane protein which would then modify its conformation or that of a neighboring protein, triggering some type of regulatory cascade; they may change the levels of intermediate metabolites which would allosterically activate or inhibit some regulatory protein(s). Regarding the first possibility there is now ample evidence for the glucose transporter homologs Snf3 and Rgt2 acting as glucose sensors (28, 29). These proteins are required for the induction by glucose of different HXT genes and for catabolite repression of GAL2, SUC2, FBP1, ICL1 or GDH2 (29, Lafuente, M. J., Gancedo, C., and Gancedo, J. M., unpublished). In addition the Ras/cAMP pathway which is activated by glucose (2) appears to play a role in activating the transcription of PFK26 and PFK27 (25) and in repressing the transcription of a number of genes subject to
TABLE 2
Repression of Some Enzymes by Glucose and Galactose in S. cerevisiae FbPase a GlutDH (NAD) a (nmol/min/mg protein) Glucose Galactose Glycerol
,1 ,1 52
17 26 166
Invertase b (nmol/min/mg yeast) 2% 0.05% 2% 0.05%
Glucose Glucose Galactose Galactose
12 227 49 29
a
The S. cerevisiae strain WG80-1A ( gal80D) was grown on YP with the carbon source indicated, collected at the exponential phase of growth and enzymes were tested as described in Materials and Methods. Data are averages of 2 experiments. b The yeast was grown in YPglucose or YPgalactose and derepression and testing of invertase were carried out as described in Materials and Methods, in the media indicated. Data are averages of at least 3 experiments.
55
Vol. 1, No. 1, 1999
MOLECULAR CELL BIOLOGY RESEARCH COMMUNICATIONS
intermediary metabolites of the glycolytic pathway (32). It has been shown that different intermediates are required for the activation of transcription of different genes (4, 25) and that the expression of gluconeogenic genes increases in fbp1 mutants where the concentration of hexose phosphates is lower than in a wild-type strain when the yeast utilizes a non-sugar carbon source (Mc Cammon, personal communication, J. M. Gancedo, unpublished) As we have shown in this paper and as reported recently by others (25), galactose may mimic some of the effects of glucose. Galactose, however, is unable to induce invertase at low concentration (Table 2) or to induce glucose transporter genes such as HXT1 (3). Galactose is therefore unlikely to interact with Snf3 or Rgt2; this is further supported by the observation that in a snf3rgt2 mutant repression by glucose is relieved, while repression by galactose is not (Lafuente, M. J., Gancedo, C., and Gancedo, J. M., unpublished). The existence of a galactose receptor, equivalent to Snf3 or Rgt2, cannot be discarded but it is made unlikely by the fact that, after the sequencing of the S.cerevisiae genome has been completed, it has been found that all proteins with sugar transporter features have short cytoplasmic C-tails (33), in contrast with the long
TABLE 3
Changes in Metabolite Levels upon Addition of Glucose or Galactose to Derepressed Yeast Cells a Glucose-6P Addition
Time
— Glucose Glucose Galactose Galactose
0 1.5 3 1.5 3
h h h h h
Fructose-6P
Fructose-1,6P 2
(mM) 0.64 1.63 1.07 1.46 1.65
0.32 0.37 0.27 0.40 0.57
0.1 6 4 2 1.5
a S. cerevisiae WG80-1A ( gal80D) was grown on YPethanol; glucose or galactose were added to the culture and samples were taken at the times indicated (see Materials and Methods for details). Data are averages of 2 experiments.
catabolite repression (Zaragoza, O., Lindley, C., and Gancedo, J. M., submitted) although cAMP does not appear involved in the control of other genes regulated by glucose (25, 30, 31, Zaragoza, O., Lindley, C., and Gancedo, J. M., submitted). On the other hand when glucose is available to the cell there is a marked increase in the concentration of
FIG. 3. Effects of the addition of glucose to a set of isogenic strains of S. cerevisiae, W303D (wt, F), WH1D (HXK2/hxk2, ‚) and WTH2D (hxk2/hxk2, E). (A) Changes in cAMP levels. (B) Two-step inactivation of FbPase. (C) Induction of pyruvate carboxylase. (D) Proton excretion to the medium. See Materials and Methods for details.
56
Vol. 1, No. 1, 1999
MOLECULAR CELL BIOLOGY RESEARCH COMMUNICATIONS TABLE 4
Repression by Glucose of Some Enzymes in a Set of Isogenic Yeast Strains a FbPase
GlutDH (NAD) (nmol/min/mg protein)
Relevant genotype
Wild-type hxk2/HXK2 tps1hxk2/tps1HXK2
Invertase (nmol/min/mg yeast)
Glu
Gly
Glu
Gly
2% Glu
0.05% Glu
,1 ,1 1
45 45 30
12 14 10
200 210 170
20 110 30
410 440 620
a Enzyme levels were measured in repressed or derepressed conditions as described in Materials and Methods. Data are averages of at least 3 independent experiments.
strain (Fig. 3D). In addition FbPase and glutamate dehydrogenase were repressed to the same extent in the different strains (Table 4). Curiously, while repression of invertase by glucose was partially relieved in the diploid strain with a single copy of HXK2, in the strain which lacked also Tps1 the repression by glucose was recovered. From our results we may conclude that there is at present no decisive evidence for Tps1 playing a role in glucose signaling.
C-tails of Snf3 and Rgt2 which are thought to interact with further regulatory proteins (29). Galactose signaling is, therefore, more likely to take place through an increase in the levels of glycolytic intermediates. We have compared the concentration of glucose6P, fructose-6P and fructose-1,6P 2 in yeast cells in the presence of glucose or galactose (Table 3). The levels of glucose-6P and fructose-6P are very similar in both cases while fructose-1,6P 2 is 2-3 fold higher with glucose than with galactose. Although it is still premature to propose that this difference accounts for the different behaviour of both sugars, it is likely that galactose produces smaller changes than glucose in the concentration of the still unidentified regulatory metabolite(s) and produces therefore a weaker signal. In the case of glucose the question remains open regarding the identity of the proteins which relay the glucose signal from Snf3 and Rgt2 although there is preliminary evidence for Mth1 and Std1/Msn3 being involved (Zhang, X., Solimeo, H. T., and Schmidt, M. C., personal communication; Lafuente, M. J., Gancedo, C., and Gancedo, J. M., unpublished). It has been hypothesized that the protein Tps1, a subunit of the trehalose synthase complex which catalyzes the synthesis of trehalose-6P, could participate in glucose signaling (34), although the fact that this signaling was restored in a tps1hxk2 double mutant indicated that Tps1 could be dispensable in a genetic background where catabolite repression of some genes had been eliminated (12). To reexamine to what extent the Tps1 protein could be involved in the effects produced by glucose, we have compared the behaviour of 3 diploid isogenic strains (wt, hxk2/HXK2 and tps1 hxk2/tps1 HXK2) upon glucose addition. We have found that Tps1 is not required for the rapid increase in cAMP (Fig. 3A) nor for induction of pyruvate decarboxylase (Fig. 3B) or for acidification of the medium (Fig. 3C). The short-term inactivation of FbPase is also independent of Tps1 and there is only some slowing down of the proteolytic degradation of FbPase in the tps1
ACKNOWLEDGMENTS We thank M. Johnston for plasmid pBM284 and M. Rose for plasmid pMR226. This work was supported by Grant PB094-0091CO2-01 from the Direccio´n General de Investigacio´n Cientı´fica y Te´cnica. C.R. had a fellowship from the Fundacio´n Ramo´n Areces.
REFERENCES 1. Mazo´n, M. J., Gancedo, J. M., and Gancedo, C. (1982) Eur. J. Biochem. 127, 605– 608. 2. Thevelein, J. M. (1994) Yeast 10, 1753–1790. ¨ zcan, S., and Johnston, M. (1995) Mol. Cell Biol. 15, 1564 – 3. O 1572. 4. Boles, E., Mu¨ller, S., and Zimmermann, F. K. (1996) Mol. Microbiol. 19, 641– 642. 5. Johnston, M., and Carlson, M. (1992) in The Molecular and Cellular Biology of the Yeast Saccharomyces (Jones, E. W., Pringle, J. R., and Broach, J. R., Eds.), Vol. 2, pp. 193–281 Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. 6. Gancedo, J. M. (1998) Microbiol. Mol. Biol. Rev. 62, 334 –361. 7. Gancedo, J. M., and Gancedo, C. (1997) in Yeast Sugar Metabolism (Zimmermann, F. K., and Entian, K.-D., Eds.), pp. 359 –377. Technomic, Lancaster, PA. 8. Van der Plaat, J. B. (1974) Biochem. Biophys. Res. Commun. 56, 580 –587. 9. Uno, I., Matsumoto, K., Adachi, K., and Ishikawa, T. (1983) J. Biol. Chem. 258, 10867–10872. 10. Serrano, R. (1983) FEBS Lett. 156, 11–14. 11. Van Aelst, L., Hohmann, S., Bulaya, B., de Koning, W., Sierkstra, L., Neves, M. J., Luyten, K., Alijo, R., Ramos, J., Coccetti, P., Martegani, E., de Magelhaes, N. M., Brandao, R. L., Van Dijck, P., Vanhalewyn, M., Durnez, P., Jans, A. W. H., and Thevelein, J. M. (1993) Mol. Microbiol. 8, 927–943.
57
Vol. 1, No. 1, 1999
MOLECULAR CELL BIOLOGY RESEARCH COMMUNICATIONS
12. Hohmann, S., Neves, M. J., de Koning, W., Alijo, R., Ramos, J., and Thevelein, J. M. (1993) Curr. Genet. 23, 281–289. 13. Yocum, R. R., and Johnston, M. (1984) Gene 32, 75– 82. 14. Balbas, P., Soberon, X., Merino, E., Zurita, M., Lomerli, H., Valle, F., Flores, N., and Bolivar, F. (1986) Gene 50, 3– 40. 15. Ito, H., Fukuda, Y., Murata, K. and Kimura, A. (1983) J. Bacteriol. 153, 163–168. 16. Gancedo, C. (1971) J. Bacteriol. 107, 401– 405. 17. Bla´zquez, M. A., Lagunas, R. Gancedo, C., and Gancedo, J. M. (1993) FEBS Lett. 329, 51–54. 18. Funayama, S., Gancedo, J. M., and Gancedo, C. (1980) Eur. J. Biochem. 109, 61– 66. 19. Doherty, D. (1970) Methods Enzymol. 17A, 850 – 856. 20. Goldstein, A., and Lampen, J. O. (1975) Methods Enzymol. 42, 504 –511. 21. Lowry, O. H., Rosebrough, N. J., Farr, A. L., and Randall, R. S. (1951) J. Biol. Chem. 193, 265–275. 22. Gamo, F.-J., Portillo, F., and Gancedo, C. (1983) FEMS Microbiol. Lett. 106, 233–238. 23. Eraso, P., and Gancedo, J. M. (1984) Eur. J. Biochem. 141, 195–198.
24. Bergmeyer, H. U. (1983) Methods of Enzymatic Analysis. Verlag Chemie, Weinheim, Germany. 25. Gonc¸alves, P. M., Griffioen, G., Bebelman, J. P., and Planta, R. J. (1997) Mol. Microbiol. 25, 483– 493. 26. Torchia, T. E., Hamilton, R. W., Cano, C. L., and Hopper, J. E. (1984) Mol. Cell. Biol. 4, 1521–1527. 27. Eraso, P., and Gancedo, J. M. (1985) FEBS Lett. 191, 51–54. ¨ zcan, S., Dover, J., Rosenwald, A. G., Wo¨lfl, S., and Johnston, 28. O M. (1996) Proc. Natl. Acad. Sci. USA 93, 12428 –12432. ¨ zcan, S., Dover, J., and Johnston, M. (1998) EMBO J. 17, 29. O 2566 –2573. 30. Matsumoto, K., Uno, I., Toh-e, A., Ishikawa, T., and Oshima, Y. (1982) J. Bacteriol. 150, 277–285. 31. Matsumoto, K., Uno, I., Ishikawa, T., and Oshima, Y. (1983) J. Bacteriol. 156, 898 –900. 32. Gancedo, J. M., and Gancedo, C. (1973) 55, 205–211. 33. Boles, E., and Hollenberg, C. P. (1997) FEMS Microbiol. Rev. 21, 85–111. 34. Thevelein, J. M. (1992) Antonie van Leeuwenhoek 62, 109 – 130. 35. Thomas, B. J., and Rothstein, R. (1989) Cell 56, 619 – 630.
58