Molecular Genetics and Genomics

, Volume 274, Issue 4, pp 384–393

Replacement of a conserved tyrosine by tryptophan in Gal3p of Saccharomyces cerevisiae reduces constitutive activity: implications for signal transduction in the GAL regulon


    • Laboratory of Molecular Genetics, School of Biosciences and BioengineeringIndian Institute of Technology Bombay
  • Paike Jayadeva Bhat
    • Laboratory of Molecular Genetics, School of Biosciences and BioengineeringIndian Institute of Technology Bombay
Origina Paper

DOI: 10.1007/s00438-005-0031-6

Cite this article as:
Lakshminarasimhan, A. & Bhat, P.J. Mol Genet Genomics (2005) 274: 384. doi:10.1007/s00438-005-0031-6


The ability of Saccharomyces cerevisiae to utilize galactose is regulated by the nucleo-cytoplasmic shuttling of a transcriptional repressor, the Gal80 protein. Gal80 interacts with the transcriptional activator Gal4 in the nucleus and inhibits its function, preventing induction of the GAL genes. In response to galactose, the relative amounts of Gal80 in the cytoplasm and the nucleus are modulated by the action of a signal transducer, Gal3. Although it has been speculated that Gal3 binds galactose, this has not been experimentally demonstrated. In this study, we show that replacement of a conserved tyrosine in Gal3 by tryptophan leads to a reduction of its constitutive activity in the absence of galactose. In addition, this mutant protein was fully functional in vivo only when high concentrations of galactose were present in the medium. When overexpressed, the mutant was found to activate the genes GAL1 and GAL7/10 differentially. The implications of these findings for the fine regulation of GAL genes, and its physiological significance, are discussed.


GHMP kinasesSignal transductionFine regulationGalactose recognitionConserved tyrosine


In eukaryotes, compartmentalization of the transcriptional machinery in the nucleus means that regulatory signals must be transmitted from the cytoplasm. Mechanisms such as phosphorylation/dephosphorylation, multimerization, and ligand-dependent sequestration have been reported to be responsible for transmitting signals to the transcriptional machinery (Ferrell 1998). These mechanisms can be grouped into two broad classes, depending on whether the transcriptional activator is normally localized in the cytoplasm or the nucleus. In the former scenario, the transcriptional activator normally resides in the cytoplasm, and enters the nucleus to activate transcription in response to a signal (Evans 1988). In the latter case, a repressor constitutively blocks the activator function of a nuclear transcription factor. Here, the activating stimulus has to traverse the nucleocytoplasmic membrane and inactivate the repressor (Lohr et al. 1995). It is obvious that such mechanisms can provide additional avenues for controlling the expression of genes, but need to be controlled either temporally or quantitatively. While different mechanisms can generate different patterns of biological responses, the underlying regulatory feature that permits a particular mechanism to generate a specific response is not always clear (Komeili and O’Shea 2000).

In Saccharomyces cerevisiae, the expression of genes involved in galactose metabolism (GAL genes) is tightly regulated by the interplay between a signal transducer (Gal3), a transcriptional activator (Gal4) and a repressor (Gal80) (Johnston 1987; Lohr et al. 1995; Bhat and Murthy 2001). Under non-inducing conditions, Gal4 remains bound to its cognate upstream activating sequence(s) (UASG) in the promoters of GAL genes, but its function is masked by the Gal80 protein. Gal80 itself shuttles between the cytoplasm and the nucleus (Peng and Hopper 2000). In the presence of galactose, Gal3 sequesters the Gal80 protein in the cytoplasm, resulting in the net translocation of Gal80 from the nucleus to the cytoplasm (Peng and Hopper 2002; Verma et al. 2003). This leads to a reduction in the effective concentration of Gal80 protein in the nucleus, thus de-repressing Gal4. Recently, it has been hypothesized that binding of Gal3 to Gal80 competes with dimerization of Gal80 protein, thus reducing the nuclear concentration of dimeric Gal80 protein, which is the form that inhibits Gal4 (Pilauri et al. 2005). How the Gal3 protein is activated by galactose is not clearly understood. Overexpression of GAL1 also leads to activation of Gal4 (Bhat and Hopper 1992), suggesting that Gal1 may also have the potential to function as a signal transducer.

Members of the GAL gene family respond differentially to galactose (Melcher and Xu 2001). Those that have a single UASG, such as MEL1 (which codes for α-galactosidase), GAL3 and GAL80, show a high basal level of expression. On the other hand, those genes that have multiple UASG, such as GAL1, 10, 7 and 2 (coding for galactokinase, epimerase, transferase and permease, respectively), do not show any detectable basal expression. Based on experimental analysis, it has been suggested that the tight regulation of GAL1, 10, 7 and 2 is due to the interaction between Gal80 and Gal80 dimers bound to adjacent GAL4 binding sites (Melcher and Xu 2001). This dimer–dimer interaction is speculated to repress these GAL genes completely. Thus, the status of the GAL genes is not only dependent on the nuclear concentration of Gal80 protein, but also on the number and arrangement of the UASG associated with a given target gene. Members of the GAL family are generally dispersed through the genome, but GAL1, 10 and 7 form a tight cluster: GAL1 and GAL10 are divergently transcribed, while GAL10 and GAL7 are organized in tandem (Citron and Donelson 1984). Based on genome analysis it has been suggested that clustering of co-expressed genes is evolutionarily selected and may provide an efficient mechanism for coordinated expression of functionally coupled genes such as the GAL genes (Citron and Donelson 1984).

The Gal3 protein is thought to exist in active and inactive forms (Bhat and Hopper 1992). This two-state model postulates that galactose shifts the equilibrium of the Gal3 protein into the active form, which can bind the Gal80 protein (Bhat and Hopper 1992). While the interaction of Gal3 with Gal80 has been directly demonstrated (Suzuki-Fujimoto et al. 1996; Zenke et al. 1996; Yano and Fukasawa 1997; Platt and Reece 1998), neither direct binding of galactose to Gal3 nor any galactose-dependent conformational change in Gal3 has been reported. If galactose is recognized by Gal3, it should be possible to isolate mutants that are impaired in galactose binding. Analysis of the behavior of such a mutant Gal3 protein is expected to yield insights into the molecular mechanism of galactose-dependent signal transduction. For this purpose, mutants of Gal3 with diminished ability to recognize galactose were sought. In the course of this search, we replaced a conserved tyrosine present in the putative galactose-binding motif of the Gal3 protein (Platt et al. 2000). We focused on this residue since it is present in galactokinases, but not in other members of the GHMP (Galactokinase, Homoserine kinase, Mevalonate kinase, Phosphomevalonate kinase) kinase superfamily (Zhou et al. 2000). Secondly, it is well documented that recognition of galactose is generally mediated by stacking interactions between galactose and aromatic amino acid residues (Adar and Sharon 1996). However, analysis of this mutant suggested that an altered conformation was responsible for its properties. Overexpression of the mutant Gal3 protein led to differential activation of the GAL1 and GAL7/10 genes, suggesting that different members of the GAL regulon differ in their inducibility. The mechanism and the physiological significance of this fine regulation are discussed.

Materials and methods


Bactopeptone, bacto-agar and nitrogen base without amino acids were purchased from Difco (Detroit, MI.). Yeast extract and tryptone were purchased from Hi-Media Laboratories (Mumbai, India). TRIS, EDTA, BSA, galactose, glucose, amino acids, ATP, PMSF (phenylmethylsulfonylfluoride), PEI (polyethyleinime), DTT (dithiothreitol), MgCl2,NBT (nitroblue tetrazolium chloride), BCIP (5-bromo-4-chloro-3-indolylphosphate), agarose and other chemicals and reagents were purchased from Sigma (St Louis, MO). The rest of the chemicals used in this study were of analytical grade. T4 DNA ligase, Taq polymerase, calf intestinal alkaline phosphatase (CIAP), restriction enzymes, molecular weight standards and dNTPs were purchased from MBI Fermentas (Amherst, NY) Vent DNA polymerase and DpnI were obtained from New England Biolabs (Beverly, MA). Custom-made oligonucleotides for PCR amplification were obtained from Microsynth (Balgach, Switzerland).

Strains, media and growth conditions

The yeast strains used in this study are listed in Table 1. Synthetic complete medium was supplemented with the appropriate carbon source, 2% (w/v) glucose or 2% (w/v) galactose, 0.5% (w/v) 2-deoxygalactose or 0.5% (v/v) glycerol + 2% (v/v) lactate, as indicated. YEPD media contained 2% (w/v) dextrose, 1% (w/v) Bactopeptone and 0.5% (w/v) yeast extract (Adams et al. 1989). When required, ethidium bromide was added to the medium to a final concentration of 20 μg/ml, in order to eliminate mitochondria. To determine the induction activity of wild type and mutant Gal3 proteins, transformants were grown on glycerol+lactate medium to an OD600 of 0.6 and induced with 0.02, 0.2, 2 or 10% galactose. Cells were harvested for enzyme activity determination 3 h after the addition of galactose. The level of constitutive activity was measured in extracts of cells grown in glycerol+lactate to an OD600 of 0.6.
Table 1

List of Saccharomyces strains used for this study





MATα ade1 ile leu 2-3, 112 ura 3-52trp1-HIII his3- Δ1

T.V.S. Murthy


MEL1 LYS2:: GAL1UAS- GAL1TATA- HIS3 gal3Δgal1ΔMATα ura3-52 leu2-3, 112 ade1 ile MEL1 GAL3::LEU2

J. E. Hopper


MATα ura 3-52 trp 1-289 gal 1Δ

M. Johnston


MATα ade1 ile leu 2-3, 112 ura 3-52 trp1-HIII his3- Δ1 MEL1 LYS2:: GAL1UAS- GAL1TATA- HIS3

J. E. Hopper

Escherichia coli strains were grown in L-Broth (Luria complete medium), with ampicillin at a final concentration of 75 μg/ml when required for plasmid maintenance. Yeast cells were grown at 30°C and E. coli cells were grown at 37°C.


pAP-45, a multicopy plasmid derived from pYEX-BX (Clontech), was obtained from R. J. Reece (Platt and Reece 1998). pYJM-3 is a multicopy plasmid containing GAL3 under the control of the CYC1 promoter element (Murthy 2000). To obtain pYCplac33- GAL3 and pYCplac33- GAL3Y57W, pT1-3B (Torchia and Hopper 1986) was digested with SalI and HindIII, and the fragment containing the GAL3UAS- GAL3TATA-GAL3 cassette was isolated. This was ligated to SalI+ HindIII-digested YCplac33. pAP-45 Y57W, pYCplac33-Y57W GAL3 and pYJM-3Y57W were generated from pAp-45, pYCplac33- GAL3 and pYJM-3, respectively using the protocol supplied with the Quik-Change site-directed mutagenesis kit. Overlapping primers (5′-GGTGAGCATATAGAT TGGTGCGATTTTTCAGTT-3′ and 5′-AAACTGAAAAATCGCA CCAATCTATATGCTCACCT-3′) were used which resulted in the substitution of Tyr57 with Trp (codons underlined). The mutations were confirmed by sequencing the complete wild type and Y57W genes (Microsynth). pHISGAL80 is a multicopy plasmid which expresses the Gal80 protein with N-terminal hexahistidine tag is under the control of the GPD promoter (Murthy 2000).

Phenotypic analysis to determine Gal3 activity

Phenotypic analyses were carried out to determine the ability of Gal3 and Gal1 (galactokinase) to transduce the galactose signal by using appropriate strains and growth conditions as described below.

Complementation analysis of GAL3

When grown on ethidium bromide, S. cerevisiae cells lose their mitochondria (ρ). Sc385 (gal3Δ) does not grow on galactose+ethidium bromide (Douglas and Pelroy 1963). To analyze the function of mutant GAL3, Sc385 transformants were replica plated onto plates containing galactose+ethidium bromide and allowed to grow at 30°C.

Reporter enzyme assays used to determine Gal3 activity

The genes of the GAL regulon can be used as reporters for the galactose response, such that enzyme activity measured in vitro is a direct reflection of the signal transducing activity of Gal3 (Gal1). Reporter gene activity was measured in gal3 gal1 or gal3 transformants harboring wild-type or mutant GAL3 or GAL1 plasmids that had been grown in glycerol+lactate broth and induced with different concentrations of galactose. The cells were harvested, broken with glass beads and centrifuged at 15,000 rpm for 30 min. The protein level in the supernatant was estimated by the Bradford (1976) assay, and the extract was used for galactokinase (Blume and Beutler 1975), galactose-1-phosphate uridyltransferase, UDP-glucose-4-epimerase (Fukasawa et al. 1980) and α-galactosidase assays (Post-Beittenmiller et al. 1984) or for Western analysis (Bhat and Hopper 1990).

Transferase assay

The protocol used to assay galactose transferase (Gal7) activity was adapted from that for the UDP-galactose pyrophosphorylase assay described by Abraham and Howell (1969), with appropriate modifications. The premix contained 250 mM TRIS-HCl pH 8.5, 40 mM galactose-1-phosphate, 7.5 mM UDP-glucose, 0.25 mM DTT, 12.5 mM MgCl2 and 14 C-labeled galactose-1-phosphate. The reaction was started by adding 10 μl of the enzyme source to 10 μl of the premix and incubated at 30°C for 1 h. The reaction was stopped by boiling the tubes for 2 min and then cooling in ice. After adding CIAP (1 U in 2 μl of buffer), the reaction mixture was incubated at 37°C for 30 min to hydrolyze the galactose-1-phosphate remaining in the solution. The phosphatase reaction was stopped by boiling the tubes and cooling on ice. Each sample was then applied to a DEAE cellulose disk (DE-81, Whatman). The disks were dried, then washed three times with double distilled water to remove all unbound components, and dried again before measuring the bound radioactivity. The radioactivity on the disk, which was due to the labeled UDP-galactose formed during the reaction, was counted using an LKB liquid scintillation counter.

Gal3-Gal80 binding assay

Sc385 (gal3Δ) bearing pHISGAL80 (which expresses a 6′ His-tagged Gal80 protein) was grown in uracil dropout medium containing glycerol+lactate to an OD600of 0.5. Cells were harvested and disrupted with an equal amount of glass beads in breaking buffer (10 mM NaH2PO4,protease inhibitors with or without 100 mM NaCl). Cell extracts containing wild-type or mutant Gal3 protein were obtained from Sc385 transformants bearing pYJM-3 or pYJM-3Y57W, respectively. The cell extracts were mixed in a buffer containing 50 mM galactose, 2 mM ATP and 100 mM NaCl, and allowed to stand overnight at 4°C. The mixture then incubated with 40 μl of Ni2+ NTA-agarose (Qiagen) in a 1.5-ml microfuge tube in the presence of 10 mM imidazole, to allow the His-tagged gal80 to bind, and the matrix was washed three times with the same buffer used for binding but containing 30 mM imidazole. The matrix with the bound proteins was then heated in boiling water bath for 5 min in 2′SDS buffer containing 250 mM imidazole. The supernatant was subjected to SDS-PAGE and analyzed for the presence of Gal3 on a Western blot with an antiserum raised against the wild-type protein. To monitor constitutive binding the whole assay was done in the absence of 100 mM NaCl and in the absence of galactose and ATP (Yano and Fukasawa 1997).


Y57W GAL3 complements a gal3 mutant

Y57W GAL3 was cloned in the single-copy plasmid vector YCplac33, together with the wild-type upstream elements (GAL3UAS). It is therefore expected that addition of galactose will induce its expression three to fivefold (Bajwa et al. 1988). To test this, Sc385 (gal3) transformants bearing YCplac33-Y57W GAL3 was patched on glucose medium and replica plated onto a medium containing 2% galactose plus ethidium bromide (Fig. 1). Sc385 transformants bearing YCplac33- GAL3 (wild type) and the empty vector were used as controls. Tranformants containing YCplac33-Y57W GAL3 or YCplac33- GAL3 were able to grow on a medium containing galactose plus ethidium bromide, while transformants bearing the vector alone did not. This indicates that the Y57W Gal3 protein can activate the galactose-metabolizing genes and thus complements the chromosomal gal3 mutation in the presence of 2% galactose. Ethidium bromide was included in the medium in order to eliminate mitochondrial function, so as to ensure that the transformants cannot utilize amino acids as a carbon source (Murthy 2000).
Fig. 1

Ability of Y57W GAL3 to complement a gal3 mutant when expressed at low levels. Patches of Sc385 (gal3) cells transformed with YCplac33-Y57W GAL3 (Y57W), YCplac33- GAL3 (WT) or YCplac33 (V) as indicated were grown on uracil drop-out glucose medium (left panel) and replica plated onto uracil drop out galactose (2%) plus ethidium bromide medium (right panel)

The Y57W Gal3 protein is an effective signal transducer at high but not at low concentrations of galactose

Complementation analysis was carried out at galactose concentrations ten times higher than that required for maximal induction of GAL genes (Kew and Douglas 1976). Consequently, any minor defect in the signal transduction activity of the mutant Gal3 would not be detectable in the plate assay. Extracts of Sc385 transformants bearing pYCplac33-Y57W GAL3 and pYCplac33- GAL3 were therefore assayed for galactokinase activity. It is clear that the mutant Gal3 protein does not activate GAL1 at low galactose concentrations (Fig. 2a). However, when 2% galactose was used for induction, Y57W Gal3 was just as effective as the wild type. Note that, at low galactose concentrations, the Y57W Gal3 protein is unable to activate GAL1 efficiently, although Gal1p itself could serve as an alternative signal transducer in the Sc385 strain.
Fig. 2

Signal transducing activity of Y57W Gal3p. The ability of Y57W GAL3 to activate GAL7, MEL1 and GAL1 genes when expressed at low levels was measured as follows. ScTEB756UΔ (gal1, gal3) cells transformed with YCplac33- GAL3 or YCplac33-Y57W GAL3 were induced with 0.02, 0.2, 2 and 10% galactose, as indicated and assayed for transferase a and α-galactosidase b activities, respectively. c Sc385 (gal3) cells transformed with YCplac33- GAL3 or YCplac33-Y57W GAL3 were induced with 0.02, 0.2, 2 and 10% galactose and assayed for kinase activity

Thus, due to the presence of GAL1, the reporter gene activity in Sc385 may not reflect the real signal transducing capacity of the Y57W GAL3 protein. We therefore tested Sc756UΔ (gal1, gal3) transformants harboring pYCplac33-Y57W GAL3. These cells were grown on glycerol+lactate and induced with different concentrations of galactose (0.02, 0.2 and 2%). Cell extracts of these transformants were then assayed for transferase and α-galactosidase activity (Fig. 2b, c). Cell-free extracts obtained from transformants bearing YCplac33-Y57W GAL3 induced with 0.02 and 0.2% galactose showed diminished transferase and α-galactosidase activity as compared to extracts of transformants bearing YCplac33- GAL3. However, extracts obtained from transformants bearing YCplac33-Y57W GAL3 induced with 2% galactose showed levels of enzyme activity comparable to those in cells carrying YCplac33- GAL3. These results clearly show that Y57W Gal3 protein is impaired in its ability to respond to low concentrations of galactose.

Y57W GAL3 supports growth at 2% but not at 0.02% galactose

To further evaluate the signal transduction activity of the Y57W Gal3 protein, the ability of Sc385 (gal3) transformants harboring YCplac33-Y57W GAL3 to grow on 0.02% and 2% galactose was monitored. No significant difference is observed in the growth profiles of Sc385 transformants bearing YCplac33- GAL3 and YCplac33-Y57W GAL3 when they were grown in liquid medium containing 2% galactose (Fig. 3a). However, there was a significant difference between the growth profiles on 0.02% galactose (Fig. 3b). It has been reported that the GAL2-encoded permease is required for growth on galactose concentrations below 0.5% (Bhat et al. 1990). It is therefore possible that the growth defect of transformants expressing Y57W Gal3 protein on 0.02% galactose could be due to the low expression of GAL2—transcription of which is itself controlled by Y57W Gal3.
Fig. 3

Ability of Y57W GAL3 to support growth on galactose. Sc385 (gal3) transformants bearing YCplac33- GAL3 or YCplac33-Y57W GAL3 were grown on uracil drop-out medium supplemented with 2% (a) or 0.02% galactose (b)

Y57W Gal3 protein is impaired in its ability to constitutively activate GAL1

Overexpression of Gal3 has been shown to activate the GAL genes constitutively (Bhat and Hopper 1992; Murthy 2000). The ability of overexpressed Y57W Gal3 to activate GAL1 in the absence of galactose was therefore determined. Sc385 (gal3) transformants harboring pYJM-3Y57W were grown in glycerol+lactate, and induced with 0, 0.02, 2 or 10% galactose as described. Cell extracts obtained from the transformants grown in the absence of galactose showed 10–15% of the galactokinase (Gal1) activity found in cell extracts from Sc385 transformants bearing pYJM-3 (expressing the wild type GAL3 gene; Fig. 4a). Thus, the mutant Gal3 protein is also much less effective than the wild type in mediating activation of GAL1 in the absence of galactose. Western analysis (Fig. 4b) followed by densitometry (data not shown) indicated that the wild type and Y57W Gal3 proteins were expressed at comparable levels in both transformant strains. The addition of galactose increased the levels of galactokinase activity. However, extracts obtained from transformants bearing pYJM-3Y57W induced with 10% galactose showed only 50% of the activity displayed by cells carrying pJYM-3 under the same conditions.
Fig. 4

Ability of Y57W GAL3 to constitutively activate GAL1 upon overexpression. a Galactokinase (Gal1p) activity in Sc385 transformants bearing pYJM-3Y57W. Sc385 transformants bearing pYJM-3 or pYJM-3Y57W were induced with 0, 0.02, 2 and 10% galactose and assayed for galactokinase activity. b Expression levels of wild-type Gal3p and Y57W Gal3 protein. Total cell extracts were prepared from Sc385 transformants bearing pYJM-3 (lanes 1 and 3), pYJM-3Y57W (lanes 2 and 4) or YEp24 (lane 5) after growth in glycerol+lactate with (lanes 3 and 4) or without (lanes 1, 2 and 5) galactose, and subjected to Western analysis with anti-Gal3p antiserum (upper panel). This experiment was repeated three times with similar results. The blot was also probed with antiserum directed against glucose-6-phosphate dehydrogenase (Zwf1p; lower panel)

The Y57W Gal3 protein binds Gal80 as efficiently as wild-type Gal3

To determine whether the difference in the constitutive induction activity between the wild type and Y57W Gal3 proteins is due to the difference in their ability to bind Gal80, a binding assay was performed. Equal amounts of total cell extracts containing wild-type Gal3 or Y57W Gal3 were allowed to bind 6′His-tagged Gal80 in the presence and absence of galactose and ATP. Gal80 was then recovered from the mixture by affinity chromatography, and tested for the presence of bound Gal3 (Y57W-Gal3) by Western analysis. In all tests, Gal3 and Y57W Gal3 were able to bind His-tagged Gal80 equally well, as seen in the Western blots shown in Fig. 5a, b. These results indicate that the defect in the ability of the mutant Gal3 protein to constitutively activate the GAL switch is not due to the defect in recognizing Gal80.
Fig. 5

Western analysis showing the interaction of Y57W Gal3 with Gal80. Wild type (lanes 1 and 5) and Y57W Gal3 protein (lanes 2 and 6) were allowed to bind to 6′His-tagged Gal80 protein in the absence (a) and presence (b) of galactose and ATP. The tagged Gal80p was recovered by affinity chromatography, and the bound fractions were screened for the presence of Gal3 by Western analysis. Crude extracts containing Gal3 protein were loaded in lanes 3 and 7. Extracts subjected to the binding assay in the absence of Gal80 protein were loaded in lanes 4 and 8

Constitutive activation of GAL7 by Y57W Gal3 protein

In order to estimate the constitutive signal transducing activity of Y57W Gal3 in the absence of GAL1, ScTEB756UΔ (gal1, gal3) transformants harboring pYJM-3Y57W were grown in glycerol+lactate, and induced with 0, 0.02, 0.2 and 2% galactose. Cell extracts of ScTEB756UΔ transformants were assayed for transferase and α-galactosidase activity (Fig. 6a, b). Cell extracts obtained from ScTEB756UΔ transformants bearing pYJM-3Y57W showed 40–60% of the transferase (Gal7)/α-galactosidase (Mel1) activity displayed by transformants bearing pYJM-3 (expressing wild-type GAL3). When cell extracts obtained from transformants bearing the pYJM-3Y57W were induced with 2% galactose, the reporter gene activities was comparable to those of transformants bearing the pYJM-3 (wild type). The expression levels of wild-type and Y57W Gal3 proteins were comparable (Fig. 6c).
Fig. 6

Ability of Y57W GAL3 to constitutively activate GAL7 and MEL1 when expressed at high levels. a, b ScTEB756UΔ transformants bearing pYM-3 and pYJM-3Y57W were induced with 0, 0.02, 0.2 and 2% galactose and assayed for transferase (a) and α-galactosidase (b) activities, respectively. c Total extracts of ScTEB756UΔ cells transformed pYJM-3 (lanes 1 and 3), pYJM-3Y57W (2 & 4) or pYJM (lane 5), and grown in glycerol+lactate with (lanes 3 and 4) or without (lanes 1, 2 and 5) galactose, were subjected to Western blot analysis. A 15-μg aliquot of total protein was loaded in each lane

Y57W Gal3 shows a difference in its ability to constitutively activate GAL1 and GAL7/10

The difference in the ability of Y57W Gal3 protein to constitutively activate GAL1 (15% of wild type) and GAL7 (60%) just described was observed in strains with different genetic backgrounds (gal3 and gal3 gal1, respectively). Therefore, we wanted to determine whether this difference is intrinsic to Y57W Gal3 protein or not. For this purpose, kinase (Gal1), transferase (Gal7), epimerase (Gal10) and α-galactosidase (Mel1) assays were performed using cell extracts obtained from Sc385 transformants bearing pYJM-3Y57W grown in the absence of galactose. Cell extracts obtained from Sc385 (gal3) transformants bearing pYJM-3Y57W showed 10–15% galactokinase activity and 40–60% α-galactosidase/transferase and epimerase activity compared to Sc385 transformants bearing the pYJM-3 (wild type) (Fig. 7). Western analysis, together with the fact that Y57W Gal3 protein induces comparable transferase activity to the wild type when grown on 2% galactose indicates that the reduction in constitutive activity seen in the presence of Y57W Gal3 is not due to a decrease in the protein level. Thus, the Y57W Gal3 protein differs quantitatively in its ability to activate kinase and transferase/epimerase genes in the absence of added galactose.
Fig. 7

Differential ability of Y57W GAL3 to activate GAL1 and GAL7 genes when expressed at high levels. Sc385 transformants bearing pYJM-3 or pYJM-3Y57W were grown in glycerol+lactate and assayed for galactokinase, transferase, α-galactosidase and epimerase activities. The level of enzyme activity in the wild type was considered to be 100%, and corresponds to a specific activity of 13.80 nmol/μg/h for kinase, 9.31 nmol/μg/h for transferase, 216.5 nmol/μg/min for α-galactosidase, and 32.86 nmol/μg/min for epimerase


Based on genetic and biochemical experiments, it has been hypothesized that the primary step in the signal transduction pathway that controls the GAL/MEL regulon is the recognition of galactose by Gal3. In this study, we analyzed the behavior of a mutant form of Gal3, which is defective in galactose recognition, in order to obtain further insights into the mechanism of signal transduction. Upon overexpression, the Y57W-Gal3 protein showed a reduced ability to activate GAL genes in the absence of galactose compared to the wild-type Gal3 protein (Fig. 4a). This difference is due to the difference in the transcription of GAL genes—assuming that the Y57W and wild-type Gal3 proteins do not affect post-transcriptional events. However, overexpressed Y57W was less effective in activating GAL1 than GAL7 (Fig. 7), which suggests that the GAL1 promoter is less sensitive to the removal of Gal80 protein than the GAL7 promoter. This difference in promoter sensitivity could be attributable to a difference in the upstream elements of GAL1 and GAL7. For example, GAL1 has four UASG motifs, while GAL7 has only two. However, the constitutive expression of GAL1 was also less than that of GAL10, despite the fact that these two genes share the same upstream elements. This paradox is difficult to explain based on the promoter structure, suggesting that other unknown mechanisms could be responsible for the observations reported here. Nevertheless, these studies bring out a hitherto unknown feature of the GAL system in that the promoters associated with GAL1, 7 and 10 respond differently when Gal80 protein is partially sequestered. Accordingly, when a wild-type strain is induced with galactose, dissociation of Gal80 protein might occur first from the Ga4 protein bound to the UASG of GAL7/GAL10, followed by that at GAL1. Independent evidence in support of the hypothesis that GAL1 and GAL7 are differentially regulated comes from studies on the induction kinetics of GAL genes. Thus, it has been shown that GAL7 transcription precedes that of GAL10 and GAL1 (John and Davis 1981; Greger and Proudfoot 1998; Yarger 1980). What might be the physiological significance of the difference we observe with respect to GAL1 and GAL10/7? Based on our results we suggest that, as the sequestration of Gal80 protein proceeds, GAL7 and GAL10 transcription ensues first, followed by GAL1. This temporal expression pattern ensures that toxicity due to the accumulation of galactose-1-phosphate (De Atauri et al. 2004) does not occur at any time during induction.
Fig. 8

Schematic representation of the kinetic model. According to this model, galactose stabilizes the otherwise weak interaction between Gal3 and Gal80, thereby increasing the half-life of the Gal3-Gal80 complex. The Gal3-Gal80 protein complex can be stabilized either by binding of galactose to a pre-existing complex, or by the interaction of galactose-bound Gal3 protein with Gal80. The stable Gal3-Gal80 protein complex is indicated in bold (circle and rectangle). The bold arrows represent the galactose dependent stabilization of the Gal3-Gal80 protein complex

Mutant derivatives of Gal3 and/or Gal1 that are devoid of signal transduction activity in the presence of galactose have been isolated previously (Vollenbroich et al. 1999; Menezes et al. 2003; Pilauri et al. 2005), as have forms of Gal3 which are able to constitutively activate GAL genes in the absence of galactose (Blanck et al. 1997). However, mutants of Gal3 protein with lowered constitutive activity upon overexpression have not been isolated before. In this study, upon overexpression, Y57W Gal3 showed a reduced ability to constitutively activate GAL genes as compared to the wild type. Based on the structures of galactokinases (Thoden and Holden 2003; Thoden et. al 2005), it is clear that this conserved tyrosine at position 57 of Gal3 is not among the residues identified as being involved in substrate binding and/or catalysis. However, the Nε2 of a conserved histidine present in the galactokinase signature sequence is hydrogen bonded to the hydroxyl group of the conserved tyrosine, which is also present in the signature sequence. The main-chain amide nitrogen of the conserved histidine interacts with the 6-hydroxyl group of galactose, suggesting that the conserved tyrosine is required for the proper orientation of active-site residues. Therefore, it appears that the absence of the hydrogen bond due to the Y57W substitution could lead to impaired galactose binding. However, high concentrations of galactose overcome this defect. Since this mutation also leads to lowered constitutive activity, it is plausible that the conformation of the mutant protein is altered conformation as compared to wild type. The results presented above suggest that Y57W Gal3 is defective in Gal80 protein binding (in the absence of galactose), which would lead to an increase in the steady-state concentration of the Gal80 protein in the nucleus relative to that seen in the wild type. However, we were not able to detect any difference between wild-type Gal3 and Y57W Gal3 in binding to Gal80 in vitro (Fig. 5). If wild-type Gal3 and Y57W Gal3 bind Gal80 equally well, it would be difficult to explain our in vivo observations with the current understanding of the GAL system. It is possible that in vivo, a minute difference in the ability of wild type and mutant proteins to bind Gal80 might get reflected in a detectable change in GAL gene expression. Such a minute difference might be undetectable under the experimental conditions used in binding assay (Fig. 5). These considerations led us to develop an alternative model to explain the mechanism of Gal3 protein mediated signal transduction.

As opposed to the two-state model (Bhat and Hopper 1992), we propose that Gal3 exists in a conformation, which has the ability to interact with Gal80 even in the absence of galactose. However, this interaction is weak and therefore unable to sequester sufficient Gal80 protein to cause activation of GAL genes. Accordingly, the role of galactose is to stabilize the interaction between Gal3 and Gal80, thereby increasing the half-life of the Gal3-Gal80 protein complex (Fig. 8). The galactose induced stabilization could involve a conformational change in the Gal3 protein. In light of this model, we suggest that the half-life of the Y57W Gal3-Gal80 protein complex is less than that of the wild-type Gal3-Gal80 complex, leading to a lower level of constitutive activation of GAL genes. Also, high concentrations of galactose were required to stabilize the Y57W Gal3-Gal80 interaction. We favor this model over the two-state model for the following reasons. First, this model is parsimonious in that it does not require the existence of two Gal3 states in the absence of galactose. Moreover, no experimental evidence in support of the two-state model has yet been reported. Second, the model presented here is supported by the observation that, in the absence of NaCl, wild-type Gal3 protein is associated with Gal80 irrespective of the carbon source in which the cells are grown (Yano and Fukasawa 1997). Also, the fact that galactose counteracts the NaCl induced destabilization of the Gal3-Gal80 complex in the presence of NaCl provides further circumstantial evidence for this model (Yano and Fukasawa 1997). Third, the observation that the Gal3-Gal80 complex can be recovered from reaction mixtures by gel filtration only in the presence of galactose and ATP, supports the above model (Timson et al. 2002). Direct experimental evidence for this model could come from a determination of the half-life of the Gal3-Gal80 protein complex in the presence and absence of the ligands. Experiments are underway to analyze the above kinetic model.


We would like to thank Prof. P. V. Balaji and M. S. Sujatha for their help in preparing the manuscript. We also acknowledge financial support from the Board of Research in Nuclear Sciences, India

Copyright information

© Springer-Verlag 2005