Skip to main content
NCBI home page
Genetics logoLink to Genetics
. 2005 Oct;171(2):469–476. doi: 10.1534/genetics.105.045237

Mutational Hypersensitivity of a Gene Regulatory Protein: Saccharomyces cerevisiae Gal80p

Karsten Melcher 1,1
PMCID: PMC1456764  PMID: 15998719

Abstract

The inhibitor of galactose catabolic (GAL) gene expression in Saccharomyces cerevisiae, Gal80p, interacts with the activator Gal4p and the signal transducer Gal3p and self-associates. Selection for loss of Gal80p inhibitor function yielded gal80 mutants at an extremely high rate. Out of these, 21 nonoverlapping point mutants were identified; each were due to a single-amino-acid exchange in conserved residues. Semiquantitative biochemical analysis of the corresponding mutant proteins revealed that each of the 21 amino acid alterations caused simultaneous defects in every single protein-protein interaction and in Gal80's structural integrity. Thus, Gal80 provides an unprecedented example for a protein's structural sensitivity to minimal sequence alterations.

CONTROL of the galactose catabolic (GAL) genes in Saccharomyces cerevisiae has long served as a model for eukaryotic gene regulation. Transcription of the GAL genes depends on an intricate relationship among three regulatory proteins: the transcriptional activator Gal4p, the inhibitor Gal80p, and the signal transducer Gal3p (for reviews see Johnston and Carlson 1992; Melcher 1997; Bhat and Murthy 2001). Gal4p binds as dimers to UASGAL elements upstream of the galactose catabolic genes and activates their transcription by direct interaction of the Gal4p activation domain with the Spt-Ada-Gcn5-acetyltransferase (SAGA) and mediator coactivator complexes (e.g., Koh et al. 1998; Park et al. 2000; Bhaumik and Green 2001; Larschan and Winston 2001; Bryant and Ptashne 2003; Klein et al. 2003; Kuras et al. 2003; Bhaumik et al. 2004). In the absence of the inducer galactose, Gal4p remains bound to UASGAL elements, but its activation domain is masked by direct binding of Gal80p dimers (Johnston et al. 1987; Ma and Ptashne 1987a; Yun et al. 1991). The third player, Gal3p, is a catalytically inactive paralog of the galactokinase Gal1p, the enzyme that catalyzes the committed step of galactose catabolism, the conversion of galactose and ATP into galactose-1-phosphate and ADP. In the presence of the galactokinase substrates, galactose and ATP, Gal3p binds Gal80p and relieves the inhibition of Gal4p by Gal80p (Zenke et al. 1996; Yano and Fukasawa 1997). Gal80p shuttles between nucleus, where it binds Gal4p, and cytoplasm, where Gal3p is localized. Inactivation of Gal80 under inducing conditions involves its sequestering and potentially also modification by Gal3 (Peng and Hopper 2000, 2002). In addition, Gal80p may also be removed from the activation domain to another part of Gal4p (Leuther and Johnston 1992; Platt and Reece 1998; Sil et al. 1999; K. Melcher, unpublished results).

An additional level of control is mediated by the number of Gal4p binding sites. Genes with multiple Gal4p binding sites, such as GAL1, have extremely low basal activity under noninducing conditions, whereas genes regulated by single Gal4p binding sites, such as MEL1, have significant and biologically important basal activity (Bram et al. 1986). The mechanism of differential GAL gene expression involves an inherently transient (Gal80)2-(Gal80)2 interaction that is stabilized by adjacent (Gal80)2-(Gal4p)2-DNA complexes. This interaction appears to be a prerequisite for complete repression since spacings of Gal4p binding sites that support (Gal80)2-(Gal80)2 interaction strictly correlate with complete repression and higher-order Gal80p complexes may be necessary for a complete shielding of the Gal4p activation domain (Melcher and Xu 2001).

Finally, galactokinase Gal1p, like its paralog Gal3p, can also interact with Gal80p and inactivate Gal80p under inducing conditions (Zenke et al. 1996; Platt and Reece 1998). Accordingly, constitutively expressed GAL1 suppresses a gal3 induction defect (Bhat and Hopper 1991). However, under physiological conditions GAL1 is under the control of four Gal4p binding sites and hence has extremely low basal activity. Therefore, when galactose first enters the cell, Gal1p is not present and induction depends on the high basal level of Gal3p.

The Gal80p interaction site of Gal4p has been mapped to a 28-amino-acid region that overlaps with Gal4p's core activation domain (Johnston et al. 1987; Ma and Ptashne 1987a). A second weak Gal80p interaction site has been implicated by two-hybrid analysis, but location and significance of this site have been unknown (Sil et al. 1999). In contrast, even an extensive Gal80p deletion analysis failed to pinpoint the Gal4p interaction site within Gal80p. All deletions with the exception of a truncation of the 12 carboxy-terminal amino acids and of a region encompassing amino acids 321–340, which is involved in Gal3p interaction, resulted in loss of repression (Nogi and Fukasawa 1989).

In this study, I have analyzed a set of 21 independent Δgal3 suppressing mutants that each change a single Gal80p amino acid and provide evidence that each mutation causes structural perturbations rather than exclusive inactivation of specific domains or interaction functions.

MATERIALS AND METHODS

Yeast strains, media, and enzyme assays:

Yeast strains used were derivatives of strains YJ0 (Δgal4 Δgal80 MEL1 ura3-52 leu2-3,112 his3 ade1) (Leuther and Johnston 1992) and 21R (GAL4 GAL80 MEL1 ura3-52 leu2-3,112 ade1) (Johnston and Hopper 1982) in which either GAL3 (YJ0Δgal3) or GAL4 and GAL80 (21RΔgal4Δgal80) were deleted by the loxP-kanMX-loxP marker rescue approach (Güldener et al. 1996). Media were supplemented with either 3% glycerol + 2% lactic acid (noninducing condition) or 3% glycerol + 2% lactic acid + 2% galactose (inducing conditions). Extracts were prepared and α-galactosidase and β-galactosidase assays were performed as described (Melcher et al. 2000).

Construction of mutant libraries:

First, GAL4 and GAL80 with their endogenous regulatory regions were cloned into the centromeric plasmid pRS316 (Sikorski and Hieter 1989). Second, an XbaI site was introduced just upstream of the GAL80 consensus translation initiation (“Kozak”) site by site-directed mutagenesis using the QuickChange system (Stratagene, La Jolla, CA). Next, three different fragments of GAL80 were amplified by mutagenic PCR: (i) a 534-bp 5′ fragment starting 96 bp upstream of the introduced XbaI site and ending 75 bp downstream of the GAL80 EcoRI site, (ii) an 857-bp fragment starting 78 bp upstream of the EcoRI site and ending 82 bp downstream of the NheI site, and (iii) a 612-bp 3′ fragment starting 77 bp upstream of the NheI site and ending 98 bp downstream of the GAL80 SpeI site. Mutagenic conditions were initially as described (Leung et al. 1989; Cadwell and Joyce 1992). For generation of all mutants with single-point mutations, PCR was performed with Taq polymerase using standard “nonmutagenic” conditions. PCR fragments were cotransformed with pRS316-GAL4-GAL80 deleted for the GAL80 (i) XbaI-EcoRI fragment, (ii) EcoRI-NheI fragment, or (iii) NheI-BamHI fragment into strain YJ0Δgal3. Δgal3 suppressors were selected on medium with galactose as sole carbon source.

Engineering of GAL80 clones with heterologous dimerization domains:

The IINI GCN4 mutant zipper open reading frame (I9V, I16N, N23V, I30N quadruple altered specificity mutant; Zeng et al. 1997) was PCR amplified from plasmid pXZ1130 (a kind gift of Jim Hu) with primers with introduced BamHI and BglII sites. This fragment was introduced into an engineered BamHI site replacing the stop codon of GAL80(298–435) (a kind gift of Kerstin K. Leuther). In addition, the region encoding the linker of λ-repressor was PCR amplified from plasmid MVP1 (Emami and Carey 1992) with primers with introduced BamHI and SalI sites and inserted between the stop codon and IINI zipper of pUC119-GAL80(298–435)-zipper. Zipper and linker-zipper fusions to wild-type and mutant GAL80 open reading frames were constructed in the in vitro transcription/translation vector pTL37N (Leuther and Johnston 1992) and the yeast centromeric vector pRS316 (Sikorski and Hieter 1989) by standard cloning procedures (details will be given by request). All PCR amplifications were performed with proofreading polymerase and verified by DNA sequencing.

Recombinant and in vitro translated proteins:

Recombinant Gal4p(1-147+34) (Melcher and Xu 2001) and His6-tagged GST-Gal3p (Melcher 2000) were expressed and purified as published previously. GAL80 mutant open reading frames were isolated as PCR fragments with introduced NcoI sites at the start-ATG and cloned into pTL37N (Leuther and Johnston 1992) or into pSPUTK (Stratagene) for in vitro transcription. Partially Δgal3 suppressing mutants (Figure 4) were PCR amplified with primers that incorporate a T7 promoter and an avian myeloblastosis virus (AMV) 5′-untranslated region (details will be given upon request) for direct in vitro transcription. RNAs were translated in rabbit reticulocyte lysate (Promega, Madison, WI). All interaction assay reactions containing in vitro translated proteins were adjusted with unprogrammed reticulocyte lysate to equal amounts of total lysate.

Gal80-Gal4, Gal80-Gal3, and Gal80-Gal80 interaction assays:

See supplementary data at http://www.genetics.org/supplemental/ for details.

RESULTS

Construction of GAL80 mutant libraries and selection of gal80 mutants:

The goal of this project has been to generate a panel of repression-defective gal80 point mutants and then to analyze the biochemical defect(s) of the corresponding mutant proteins. In a previous study (Nogi and Fukasawa 1989) three repression-defective mutants had been isolated, but the interactions affected in these mutants had not been determined (but see discussion and Zenke et al. 1999). I wished to use a selection system that would allow me to isolate and map defects in the Gal80p-Gal4p, Gal80p-Gal80p, and (Gal80p)2-(Gal80p)2 interactions since none of the corresponding interaction sites have been identified in Gal80p. In contrast, mutations that disrupt the Gal80p-Gal3p interaction produce constitutively repressed GAL genes and have been identified and analyzed previously (Nogi and Fukasawa 1984; Zenke et al. 1996; Yano and Fukasawa 1997).

To create random mutations within three overlapping regions spanning the complete GAL80 open reading frame, each of these regions was amplified using mutagenic PCR. Mutagenic fragments were then cotransformed with centromeric GAL80 gapped plasmids, each of which lacked a region of GAL80 slightly smaller than the corresponding mutagenic PCR fragments to allow in vivo gap repair (Muhlrad et al. 1992) (Figure 1).

Figure 1.

Figure 1.

Open in a new tab

Cartoon depicting construction of GAL80 mutant libraries. See text for details.

A Δgal80 Δgal3 strain was cotransformed with the mutagenic PCR fragments plus the corresponding GAL80 gapped plasmids and transformants were selected for growth on a medium with galactose as sole carbon source. In the absence of the signal transducer Gal3p, only transformants that lacked Gal80p or that produced Gal80p defective in tight repression would escape the inhibition of Gal4p and express the GAL genes necessary for galactose utilization. Defects both in the Gal80p-Gal4p interaction and in Gal80p dimerization would be expected to compromise or abolish repression. Importantly, this selection scheme also allowed me to identify mutants affected in differential GAL gene expression (Gal802-Gal802 defects). In these mutants, the GAL1 gene, that is normally tightly repressed because it is under the control of four Gal4p binding sites, would be expected to be expressed at significant basal levels and constitutively expressed GAL1 suppresses the induction defect of a gal3 mutant (Bhat and Hopper 1991). To validate that a low constitutive Gal1p level is sufficient for suppression, I transformed a gal3 deletion strain with Gal4pΔ(238–411)Δ(425–679), a variant of Gal4p that is not completely inhibited by wild-type levels of Gal80p and therefore allows basal expression of GAL1 (Ding and Johnston 1997). As predicted, and in accordance with an earlier study (Ding and Johnston 1997), partial escape of Gal80p's inhibition of Gal4pΔ(238–411)Δ(425–679) was sufficient to suppress the induction defect (data not shown).

To avoid selection for loss of the GAL80-containing reconstituted plasmid, chromosomal GAL4 was also deleted and wild-type GAL4 was cloned into the gapped plasmids. Thus, cells that lose the reconstituted plasmid would be unable to grow on the selective galactose medium.

Gal80p function is highly susceptible to mutational assault:

Using the mutagenesis protocol by Leung et al. (1989), PCR fragments were generated with ∼1.2% misincorporations, resulting in an average of six amino acid-changing mutations per fragment (data not shown). Essentially 100% of the ∼7000 corresponding cotransformants from all three pools, but no cells transformed with the wild-type GAL80 fragment, suppressed the Δgal3 growth defect. When using the less mutagenic PCR conditions described by Cadwell and Joyce (1992), I found on average 3–4 mutations per PCR fragment and still almost 80% of all transformants were able to grow on medium with galactose as sole carbon source (data not shown). Finally, to create GAL80 pools containing mostly 0 or 1 point mutations, GAL80 fragments were amplified using Taq polymerase and standard nonmutagenic PCR conditions. About 5% of the cotransformants were able to suppress the Δgal3 growth defect. Plasmids from 40 of these transformants covering all three pools were isolated and sequenced. Five of them had double mutations, 26 were single missense mutations resulting in single-amino-acid substitutions, 3 were frameshift mutations, and 6 were nonsense mutations. Thus, the majority of GAL+ transformants appear to carry gal80 loss-of-function mutations resulting from single-amino-acid exchanges.

Mutant alleles representing changes in 21 different conserved single amino acids were analyzed further. First, Mel1p (α-galactosidase) activities were determined under noninducing conditions (see Figure 2). MEL1 expression is under the control of the Gal4p-Gal80p system and hence is a direct measure for Gal80 repression defects in vivo. Each of the Δgal3 suppressing mutants was at least partially defective in MEL1 repression. Expression levels of gal80 variants were determined by immunoblotting and most mutant proteins were expressed at levels comparable to or higher than those of wild-type Gal80p (Figure 2). Therefore, the single-amino-acid substitutions of the Gal80p variants analyzed compromised primarily protein activity, rather than protein abundance or stability.

Figure 2.

Figure 2.

Open in a new tab

Summary of repression and interaction defects of Gal80 mutant proteins. The top indicates positions of amino acid exchange mutations within the GAL80 open reading frame and within the three mutagenic fragments. Locations of three N-terminal mutants (K34E, Y53C, and S84P, shaded type) were determined, but their interaction defects were not analyzed. For the remaining unique 21 mutants relative basal Mel1p activities (a correlate for in vivo repression defects) are indicated by the height of bars relative to wild type and to a Δgal80 strain. Visibility of the corresponding mutant proteins in native gels (as an indicator for structural integrity) and their affinities for Gal4(1-147+34) (“Gal4p”), Gal80VP16 (“Gal80p”), and GST-Gal3 (“Gal3p”) are indicated by the height of bars relative to wild-type Gal80p.

Substitutions of single conserved amino acids of Gal80p cause defects in all Gal80 protein-protein interactions:

To determine specific protein defects, the same 21 gal80 alleles were cloned into expression vectors and, together with wild-type GAL80, were in vitro translated and labeled with [35S]methionine. First, interactions with the Gal4 activation domain were analyzed by supershifting a complex consisting of a recombinant Gal4p DNA-binding domain-activation domain fusion protein and a 32P-labeled consensus Gal4p binding site oligonucleotide (supplementary Figure S1 at http://www.genetics.org/supplemental/). Second, the ability of Gal80 mutant proteins to homodimerize was tested by an increase in native gel mobility upon dimerization with a highly charged Gal80 derivative (supplementary Figure S2 at http://www.genetics.org/supplemental/). Third, interactions of wild-type and mutant Gal80p with Gal3p were monitored by GST pulldown assays in the presence of galactose and ATP (supplementary Figure S3 at http://www.genetics.org/supplemental/). Each assay was performed with increasing concentrations of wild-type and mutant Gal80 proteins to estimate relative affinities (see supplementary data at http://www.genetics.org/supplemental/ and summary in Figure 2). Importantly, although equal amounts of full-length Gal80p species were loaded onto native gels, almost all mutant forms of Gal80p migrated aberrantly and/or as diffused bands, indicating that these variants were misfolded (supplementary Figure S2).

Remarkably, none of the 21 amino acid exchanges affected exclusively a single one of the protein-protein interactions analyzed. Rather MEL1 repression defects strongly correlated with simultaneous defects in each of the Gal80-Gal4, Gal80-Gal3, and Gal80-Gal80 interactions as well as with altered Gal80p mobilities in native PAGE. The only three exceptions were D260N, G282D, and H297D. These exchanges only moderately affected Gal80's structural integrity and Gal80-Gal3 and Gal80-Gal80 interactions, but severely compromised interactions with Gal4 (see discussion).

Mutants with mild MEL1 repression defects are not selectively compromised in Gal80 dimer-dimer interaction:

On the basis of mutational sensitivity, Gal80p can be separated into four regions. Amino acid exchanges in the N terminus of Gal80p as well as between amino acids 224 and 369 appear to be extremely detrimental to Gal80p's overall structure. In contrast, changes in amino acids 118–215 as well as in Gal80's C terminus were underrepresented in the mutant pool and caused less severe defects. For instance, Gal80-I176T, and even more Gal80-Q392H and Gal80-F394S showed very mild defects in all functions analyzed. These proteins are potential candidates for variants with specific defects in Gal80p dimer-dimer interaction, which would be expected to suppress Δgal3 by relieving tight GAL1 repression (four Gal4p binding sites) without affecting MEL1 repression (one Gal4p binding site).

To probe these mutants for Gal80p dimer-dimer interaction in vivo, I used reporter constructs with Gal4p binding sites either on the same side of the DNA helix (spaced 10 bp apart) or on opposite sides of the DNA helix (spaced 6 bp apart). Only the 10-bp spacing supports Gal80p dimer-dimer interaction and allows complete reporter repression (Melcher and Xu 2001). As shown in Figure 3, basal expression of the I176T, Q392H, and F394S variants remained higher when Gal4p binding sites were spaced 6 bp, rather than 10 bp, apart. This indicates that these mutants were not severely defective in dimer-dimer interaction and suggests that Δgal3 suppression is due to cumulative interaction defects caused by a relatively small change in Gal80p overall structure.

Figure 3.

Figure 3.

Open in a new tab

Differential repression in gal80 mutants. β-Galactosidase activities were determined from Δgal80 strains cotransformed with plasmids expressing the indicated Gal80p variants together with the reporter constructs depicted at the bottom. Centromeric reporter plasmids contain fusions between the lacZ open reading frame and the MEL1 5′-untranslated region, in which the MEL1 Gal4p binding site was replaced by three consensus Gal4p binding sites spaced either 6 or 10 bp apart.

Analysis of gal80 mutants that migrate normally on native PAGE:

Next, I attempted to identify partial mutants, which are more likely to be not grossly misfolded (as judged by their migration on native PAGE). First, I isolated plasmids from 21 small (i.e., presumably partial) gal3 suppressor colonies. Plasmids were then used as templates to PCR amplify the respective gal80 open reading frames with primers that incorporated a T7 promoter and a translational enhancer sequence for efficient in vitro transcription and translation. PCR fragments were transcribed and translated in the presence of [35S]methionine, separated by both native and denaturing PAGE, and visualized by autoradiography (Figure 4). All gal80 alleles were efficiently translated (see Figure 4, SDS-PAGE gels at bottom) and most of the translated proteins were indistinguishable from wild-type Gal80p on native gels, consistent with the prediction that small suppressor colonies are more likely to be not grossly misfolded and to retain significant partial function. However, protein-protein interaction assays of all variants tested demonstrated that all had only slight defects in Gal80-Gal4 and Gal80-Gal80 interaction (data not shown). Thus, partial suppression of Δgal3 appears to be predominantly due to minor defects rather than to the selective inactivation of specific functions.

Figure 4.

Figure 4.

Open in a new tab

Migration of partially Δgal3-suppressing gal80 mutant proteins in native and denaturing gels. In vitro translated labeled Gal80 wild-type and mutant proteins were separated by native (top) and denaturing (bottom) PAGE and visualized by autoradiography. In contrast to supplemental Figure S2 (http://www.genetics.org/supplemental/), Gal80 proteins migrate as doublet bands in native gels. This difference is solely due to the loading of programmed reticulocyte lysates rather than translated proteins in buffer (data not shown).

A heterologous dimerization domain is insufficient to restore normal migration on native gels:

Aberrant migration on native gels and defects in several protein-protein interactions may, at least in some cases, be a consequence of a primary defect in Gal80p dimerization. Gal80p monomers are likely to adopt a structure different from that of dimers and thus might well migrate as diffused bands on native gels. In addition, dimerization appears to stabilize the interaction with Gal4p dimers (Melcher and Xu 2001) and potentially with Gal3p, as well. To further explore this possibility, a functional heterologous dimerization domain was genetically fused to selected gal80 mutant proteins. The yeast Gcn4 leucine zipper is well established as a heterologous dimerization domain (for review see Rieker and Hu 2000) and a variant of this zipper, “IINI,” has been generated that efficiently homodimerizes, but that does not dimerize with endogenous Gcn4p (Zeng et al. 1997).

To determine whether fusion of the heterologous dimerization domain interferes with normal Gal80p function, I engineered two different hybrids: (i) the zipper variant fused directly to the C terminus of wild-type Gal80 and (ii) Gal80p separated from the heterologous dimerization domain by the flexible linker of λ-repressor (Emami and Carey 1992) (Figure 5A). Hybrids were expressed under the control of the wild-type GAL80 promoter and 3′-UTR in a strain deleted for chromosomal GAL80 and Mel1 activity was determined under inducing (galactose) and noninducing (glycerol/lactic acid) conditions. As seen in Figure 5B (left), C-terminal fusion of the Gcn4p variant zipper or of the linker-zipper cassette to Gal80p had relatively mild effects on MEL1 expression, indicating that zipper and linker-zipper appendixes do not interfere strongly with Gal80p function.

Figure 5.

Figure 5.

Open in a new tab

Phenotypes of Gal80 hybrids with heterologous dimerization domains. (A) Schematic of the Gal80-zipper and Gal80-linker-zipper hybrids (see text for details). (B) α-Galactosidase (Mel1p) activities of strains expressing Gal80 and Gal80 fusion proteins. Left, Gal80-zipper and Gal80-linker-zipper can functionally substitute wild-type Gal80p. Right, basal α-galactosidase activities in strains expressing gal80 mutant hybrids. (C) Mobilities of in vitro translated Gal80 hybrids in denaturing (SDS-PAGE) and native (native PAGE) gels. Equal molar amounts of all proteins were separated on each gel and subjected to autoradiography.

Zipper and linker-zipper cassettes were fused to Gal80pL116P, L262P, and L319P. These mutant proteins were selected because they represent the three mutant pools and because each migrates as a diffused band on native gels and is strongly defective in Gal80-Gal80, Gal80-Gal4, and Gal80-Gal3 interactions. Fusion of zipper- and zipper-linker cassettes to these mutant proteins only slightly affected, if at all, basal MEL1 expression (Figure 5B, right), migration on native gels (Figure 5C), and the in vitro interaction with Gal4p (data not shown). These data are most easily interpreted in that L116P, L262P, and L319P substitutions primarily cause folding defects in Gal80p that interfere with protein-protein interactions, including Gal80p dimerization, rather than causing specific Gal80-Gal80 interaction defects.

DISCUSSION

Many proteins are organized in a modular way; i.e., they consist of different, physically separable, functional domains. Functional domains are typically delineated by deletion analyses, an approach that led to the identification of separable DNA binding and activation domains in Gal4p (Ma and Ptashne 1987b), a finding that served as a basis for the yeast two-hybrid system. In many deletion analyses it is assumed that a stably produced truncated protein defective in an interaction lacks part of, or the entire, interaction domain.

Gal80p provides a stunning lesson to what degree this interpretation may fail, not just with truncated proteins but even with single-amino-acid substitutions. Taking advantage of Gal80p functioning in at least four different protein-protein interactions, I was able to test randomly chosen Δgal3-suppressing gal80 mutants for each of three interactions in a semiquantitative way. All of 21 single-amino-acid exchange mutants spanning the complete GAL80 open reading frame were defective in each interaction analyzed. Clearly not all of the amino acids affected can be involved in all interactions simultaneously, suggesting that instead these point mutations compromised Gal80p's structural integrity. In support of this hypothesis, mutant proteins migrated aberrantly on native gels, strongly implying that they were incorrectly folded. With the high rate of random amino acid changes that caused gal80 mutant phenotypes, I conclude that Gal80p has an unusually complex structure that is highly susceptible to mutational assault.

The almost perfect correlation between the severities of the in vitro determined defects of Gal80p mutants and their in vivo repression defects is a strong indication that these assays correctly measure interaction and folding defects and that these defects are indeed the underlying causes for mutant phenotypes. Only three mutant proteins, Gal80 D260N, G282D, and H297D, exhibited a partial selectivity in interaction defects. These three proteins are only partially compromised in Gal80-Gal80 and Gal80-Gal3 interactions and migrate as visible bands on native gels, yet are severely defective in binding Gal4p. They all change the net charge of Gal80p and I suggest that these changes specifically affect the Gal80-Gal4 interaction while at the same time moderately compromise the structural integrity of Gal80p. Three repression-defective gal80 mutants isolated by Nogi and Fukasawa also all caused altered Gal80p net charges (Nogi and Fukasawa 1989). One of the underlying amino acid changes in a position next to the three residues that I have identified above (Gal80 G310E) was introduced at its conserved position in Kluyveromyces lactis Gal80p. This K. lactis mutant protein was also shown to be severely defective in Gal4p interaction, but not or only moderately affected in Gal1p interaction (Zenke et al. 1999). It should be noted that in a separate study on the mechanism of GAL gene induction I have identified a second Gal80p region involved in Gal4p interaction (K. Melcher, unpublished results).

The Gal80-Gal3 interaction, which has not been targeted in the selection strategy of this study, is different. Three gal80 mutants have been isolated previously that are defective in the Gal80p-Gal3p interaction (Hashimoto et al. 1983; Nogi and Fukasawa 1984; Nogi et al. 1984; Yano and Fukasawa 1997). These mutants were isolated by their ability to inhibit Gal4p even in the presence of galactose (“superrepressors”) and hence require, in contrast to the repression-defective mutants analyzed here, overall structural integrity and intact Gal80p-Gal4p and Gal80p-Gal80p interactions. Mutants that are selectively defective in the Gal80p-Gal3p interaction are therefore the only ones that can be directly selected for interaction specificity. The underlying mutations are clustered between amino acids 301 and 351, the center of which is protected by Gal3p against protease attack (Timson and Reece 2002). Moreover, deletion of amino acids 322–340 converts Gal80p into a supperrepressor (Nogi and Fukasawa 1989). This region is strikingly different from the otherwise highly homologous Gal80 paralog from K. lactis (Zenke et al. 1993), a related yeast that lacks Gal3p. These data indicate that the Gal3p interaction site is likely to function as a largely independent module acquired late in evolution.

Selection for Δgal3 suppressors made it possible to also identify the 7 of the 21 gal80 mutants analyzed that have only slight repression defects. These mutants have amino acid changes in two distinct regions of Gal80p, clearly pointing toward regions of different structural fragility within Gal80p that appear to be colinear with primary sequence. However, even the phenotypes of all amino acid substitutions within the two structurally less vulnerable regions analyzed here seem to be due to changes in Gal80p conformation rather than to specific interaction defects. Attempts to solve the crystal structure of Gal80p are under way to hopefully allow an understanding of the structural basis of the mutational sensitivity of this unusual protein.

Acknowledgments

I thank Stefanie Lamberth and Matthias Hammelmann for technical assistance; Toshio Fukasawa for antibodies against Gal80p; Jim Hu, W. Vivianne Ding, Kerstin K. Leuther, and Michael Carey for plasmids; and Paul Thompson for critically reading the article. This work was supported by grants from the Deutsche Forschungsgemeinschaft (Me 1575 and SFB 474).

References

  1. Bhat, P. J., and J. E. Hopper, 1991. The mechanism of inducer formation in gal3 mutants of the yeast galactose system is independent of normal galactose metabolism and mitochondrial respiratory function. Genetics 128: 233–239. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Bhat, P. J., and T. V. Murthy, 2001. Transcriptional control of the GAL/MEL regulon of yeast Saccharomyces cerevisiae: mechanism of galactose-mediated signal transduction. Mol. Microbiol. 40: 1059–1066. [DOI] [PubMed] [Google Scholar]
  3. Bhaumik, S. R., and M. R. Green, 2001. SAGA is an essential in vivo target of the yeast acidic activator Gal4p. Genes Dev. 15: 1935–1945. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Bhaumik, S. R., T. Raha, D. P. Aiello and M. R. Green, 2004. In vivo target of a transcriptional activator revealed by fluorescence resonance energy transfer. Genes Dev. 18: 333–343. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Bram, R. J., N. F. Lue and R. D. Kornberg, 1986. A GAL family of upstream activating sequences in yeast: roles in both induction and repression of transcription. EMBO J. 5: 603–608. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Bryant, G. O., and M. Ptashne, 2003. Independent recruitment in vivo by Gal4 of two complexes required for transcription. Mol. Cell 11: 1301–1309. [DOI] [PubMed] [Google Scholar]
  7. Cadwell, R. C., and G. F. Joyce, 1992. Randomization of genes by PCR mutagenesis. PCR Methods Appl. 2: 28–33. [DOI] [PubMed] [Google Scholar]
  8. Ding, W. V., and S. A. Johnston, 1997. The DNA binding and activation domains of Gal4p are sufficient for conveying its regulatory signals. Mol. Cell. Biol. 17: 2538–2549. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Emami, K. H., and M. Carey, 1992. A synergistic increase in potency of a multimerized VP16 transcriptional activation domain. EMBO J. 11: 5005–5012. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Güldener, U., S. Heck, T. Fiedler, J. Beinhauer and J. H. Hegemann, 1996. A new efficient gene disruption cassette for repeated use in budding yeast. Nucleic Acids Res. 24: 2519–2524. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Hashimoto, H., Y. Kikuchi, Y. Nogi and T. Fukusawa, 1983. Regulation of expression of the galactose gene cluster in Saccharomyces cerevisiae. Mol. Gen. Genet. 191: 31–38. [DOI] [PubMed] [Google Scholar]
  12. Johnston, M., and M. Carlson, 1992. Regulation of carbon and phosphate utilization, pp. 193–281 in The Molecular Biology of the Yeast Saccharomyces cerevisiae, edited by J. R. Broach, J. R. Pringle and E. W. Jones. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.
  13. Johnston, S. A., and J. E. Hopper, 1982. Isolation of the yeast regulatory gene GAL4 and analysis of its dosage effects on the galactose/melibiose regulon. Proc. Natl. Acad. Sci. USA 79: 6971–6975. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Johnston, S. A., J. J. Salmeron and S. S. Dincher, 1987. Interaction of positive and negative regulatory proteins in the galactose regulon of yeast. Cell 50: 143–146. [DOI] [PubMed] [Google Scholar]
  15. Klein, J., M. Nolden, S. L. Sanders, J. Kirchner, P. A. Weil et al., 2003. Use of a genetically introduced cross-linker to identify interaction sites of acidic activators within native transcription factor IID and SAGA. J. Biol. Chem. 278: 6779–6786. [DOI] [PubMed] [Google Scholar]
  16. Koh, S. S., A. Z. Ansari, M. Ptashne and R. A. Young, 1998. An activator target in the RNA polymerase II holoenzyme. Mol. Cell 1: 895–904. [DOI] [PubMed] [Google Scholar]
  17. Kuras, L., T. Borggrefe and R. D. Kornberg, 2003. Association of the Mediator complex with enhancers of active genes. Proc. Natl. Acad. Sci. USA 100: 13887–13891. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Larschan, E., and F. Winston, 2001. The S. cerevisiae SAGA complex functions in vivo as a coactivator for transcriptional activation by Gal4. Genes Dev. 15: 1946–1956. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Leung, D., E. Chen and D. Goeddel, 1989. A method for random mutagenesis of a defined DNA segment using a modified polymerase chain reaction. Technique 1: 11–15. [Google Scholar]
  20. Leuther, K. K., and S. A. Johnston, 1992. Nondissociation of GAL4 and GAL80 in vivo after galactose induction. Science 256: 1333–1335. [DOI] [PubMed] [Google Scholar]
  21. Ma, J., and M. Ptashne, 1987. a The carboxy-terminal 30 amino acids of GAL4 are recognized by GAL80. Cell 50: 137–142. [DOI] [PubMed] [Google Scholar]
  22. Ma, J., and M. Ptashne, 1987. b Deletion analysis of GAL4 defines two transcriptional activating segments. Cell 48: 847–853. [DOI] [PubMed] [Google Scholar]
  23. Melcher, K., 1997. Galactose metabolism in Saccharomyces cerevisiae: a paradigm for eukaryotic gene regulation, pp. 235–269 in Yeast Sugar Metabolism, edited by F. K. Zimmermann and K.-D. Entian. Technomic Publishing, Lancaster, PA.
  24. Melcher, K., 2000. A modular set of prokaryotic and eukaryotic expression vectors. Anal. Biochem. 277: 109–120. [DOI] [PubMed] [Google Scholar]
  25. Melcher, K., and H. E. Xu, 2001. Gal80-Gal80 interaction on adjacent Gal4 binding sites is required for complete GAL gene repression. EMBO J. 20: 841–851. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Melcher, K., B. Sharma, W. V. Ding and M. Nolden, 2000. Zero background yeast reporter plasmids. Gene 247: 53–61. [DOI] [PubMed] [Google Scholar]
  27. Muhlrad, D., R. Hunter and R. Parker, 1992. A rapid method for localized mutagenesis of yeast genes. Yeast 8: 79–82. [DOI] [PubMed] [Google Scholar]
  28. Nogi, Y., and T. Fukasawa, 1984. Nucleotide sequence of the yeast regulatory gene GAL80. Nucleic Acids Res. 12: 9287–9298. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Nogi, Y., and T. Fukasawa, 1989. Functional domains of a negative regulatory protein, GAL80, of Saccharomyces cerevisiae. Mol. Cell. Biol. 9: 3009–3017. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Nogi, Y., H. Shimada, Y. Matsuzaki, H. Hashimoto and T. Fukasawa, 1984. Regulation of expression of the galactose gene cluster in Saccharomyces cerevisiae. Mol. Gen. Genet. 195: 29–34. [DOI] [PubMed] [Google Scholar]
  31. Park, J. M., H. S. Kim, S. J. Han, M. S. Hwang, Y. C. Lee et al., 2000. In vivo requirement of activator-specific binding targets of mediator. Mol. Cell. Biol. 20: 8709–8719. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Peng, G., and J. E. Hopper, 2000. Evidence for Gal3p's cytoplasmic location and Gal80p's dual cytoplasmic-nuclear location implicates new mechanisms for controlling gal4p activity in Saccharomyces cerevisiae. Mol. Cell. Biol. 20: 5140–5148. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Peng, G., and J. E. Hopper, 2002. Gene activation by interaction of an inhibitor with a cytoplasmic signaling protein. Proc. Natl. Acad. Sci. USA 99: 8548–8553. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Platt, A., and R. J. Reece, 1998. The yeast galactose genetic switch is mediated by the formation of a Gal4p-Gal80p-Gal3p complex. EMBO J. 17: 4086–4091. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Rieker, J., and J. Hu, 2000. Molecular applications of fusions to leucine zippers. Methods Enzymol. 328: 282–296. [DOI] [PubMed] [Google Scholar]
  36. Sikorski, R. S., and P. Hieter, 1989. A system of shuttle vectors and yeast host strains designed for efficient manipulation of DNA in Saccharomyces cerevisiae. Genetics 122: 19–27. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Sil, A. K., S. Alam, P. Xin, L. Ma, M. Morgan et al., 1999. The Gal3p-Gal80p-Gal4p transcription switch of yeast: Gal3p destabilizes the Gal80p-Gal4p complex in response to galactose and ATP. Mol. Cell. Biol. 19: 7828–7840. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Timson, D. J., and R. J. Reece, 2002. Kinetic analysis of yeast galactokinase: implications for transcriptional activation of the GAL genes. Biochimie 84: 265–272. [DOI] [PubMed] [Google Scholar]
  39. Yano, K., and T. Fukasawa, 1997. Galactose-dependent reversible interaction of Gal3p with Gal80p in the induction pathway of Gal4p-activated genes of Saccharomyces cerevisiae. Proc. Natl. Acad. Sci. USA 94: 1721–1726. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Yun, S. J., Y. Hiraoka, M. Nishizawa, K. Takio, K. Titani et al., 1991. Purification and characterization of the yeast negative regulatory protein GAL80. J. Biol. Chem. 266: 693–697. [PubMed] [Google Scholar]
  41. Zeng, X., A. M. Herndon and J. C. Hu, 1997. Buried asparagines determine the dimerization specificities of leucine zipper mutants. Proc. Natl. Acad. Sci. USA 94: 3673–3678. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Zenke, F. T., W. Zachariae, A. Lunkes and K. D. Breunig, 1993. Gal80 proteins of Kluyveromyces lactis and Saccharomyces cerevisiae are highly conserved but contribute differently to glucose repression of the galactose regulon. Mol. Cell. Biol. 13: 7566–7576. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Zenke, F. T., R. Engels, V. Vollenbroich, J. Meyer, C. P. Hollenberg et al., 1996. Activation of Gal4p by galactose-dependent interaction of galactokinase and Gal80p. Science 272: 1662–1665. [DOI] [PubMed] [Google Scholar]
  44. Zenke, F. T., L. Kapp and K. D. Breunig, 1999. Regulated phosphorylation of the Gal4p inhibitor Gal80p of Kluyveromyces lactis revealed by mutational analysis. Biol. Chem. 380: 419–430. [DOI] [PubMed] [Google Scholar]

Articles from Genetics are provided here courtesy of Oxford University Press

RESOURCES