MCB
Home Help [Feedback] [For Subscribers] [Archive] [Search] [Contents]
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowReprints and Permissions
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Bhoite, L. T.
Right arrow Articles by Stillman, D. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Bhoite, L. T.
Right arrow Articles by Stillman, D. J.

Molecular and Cellular Biology, November 1998, p. 6436-6446, Vol. 18, No. 11
0270-7306/98/$04.00+0
Copyright © 1998, American Society for Microbiology. All rights reserved.

Residues in the Swi5 Zinc Finger Protein That Mediate Cooperative DNA Binding with the Pho2 Homeodomain Protein

Leena T. Bhoite and David J. Stillman*

Division of Molecular Biology and Genetics, Department of Oncological Sciences, Huntsman Cancer Institute, University of Utah Health Sciences Center, Salt Lake City, Utah 84132

Received 29 January 1998/Returned for modification 30 March 1998/Accepted 20 August 1998

    ABSTRACT
Top
Abstract
Introduction
Materials & Methods
Results
Discussion
References

The Swi5 zinc finger and the Pho2 homeodomain DNA-binding proteins bind cooperatively to the HO promoter. Pho2 (also known as Bas2 or Grf10) activates transcription of diverse genes, acting with multiple distinct DNA-binding proteins. We have performed a genetic screen to identify amino acid residues in Swi5 that are required for synergistic transcriptional activation of a reporter construct in vivo. Nine unique amino acid substitutions within a 24-amino-acid region of Swi5, upstream of the DNA-binding domain, reduce expression of promoters that require both Swi5 and Pho2 for activation. In vitro DNA binding experiments show that the mutant Swi5 proteins bind DNA normally, but some mutant Swi5 proteins (resulting from SWI5* mutations) show reduced cooperative DNA binding with Pho2. In vivo experiments show that these SWI5* mutations sharply reduce expression of promoters that require both SWI5 and PHO2, while expression of promoters that require SWI5 but are PHO2 independent is largely unaffected. This suggests that these SWI5* mutations do not affect the ability of Swi5 to bind DNA or activate transcription but specifically affect the region of Swi5 required for interaction with Pho2. Two-hybrid experiments show that amino acids 471 to 513 of Swi5 are necessary and sufficient for interaction with Pho2 and that the SWI5* point mutations cause a severe reduction in this two-hybrid interaction. Analysis of promoter activation by these mutants suggests that this small region of Swi5 has at least two distinct functions, conferring specificity for activation of the HO promoter and for interaction with Pho2.

    INTRODUCTION
Top
Abstract
Introduction
Materials & Methods
Results
Discussion
References

Specific interactions between multiple transcription factors are often required to achieve complex patterns of gene regulation. The importance of cooperative DNA binding by transcription factors containing identical or homologous subunits has long been recognized, but only more recently has cooperative binding of proteins with heterologous DNA-binding domains been studied. In vitro DNA binding experiments have shown that the Swi5 and Pho2 DNA-binding proteins bind cooperatively to the HO promoter (6). The SWI5 gene was first identified by its requirement for expression of the HO gene that encodes an endonuclease that initiates mating type switching in yeast. The PHO2 gene was originally identified as a transcriptional activator of the PHO5 acid phosphatase gene, and activation of PHO5 requires the cooperative binding of the Pho2 homeodomain and the Pho4 basic helix-loop-helix protein (2). PHO2 (also known as BAS2 or GRF10) was subsequently shown to activate HIS4 and various ADE genes, and at these target genes Pho2 interacts with Bas1, a Myb-like DNA-binding protein (9, 39, 42). Thus, Pho2, a homeodomain protein, interacts with at least three different partners, the Swi5 zinc finger protein, the Pho4 basic helix-loop-helix protein, and the Bas1 Myb-like protein.

Transcriptional regulation of the HO gene by SWI5 is highly complex (16, 30). Swi5 recognizes two sites in the HO promoter, called site A and site B, located approximately 1,800 and 1,300 nucleotides, respectively, upstream from the transcription start site (24, 37). Swi5 binds to both of these sites with relatively low affinity, and binding by Pho2 is extremely weak. In vitro binding studies have shown that Swi5 and Pho2 bind each of these sites cooperatively, leading to the production of high-affinity ternary complexes. Mutations that eliminate Swi5 binding at either of these sites sharply reduce HO expression, indicating that both sites are required for HO transcription (24). Although PHO2 is required for activation of either an HO-lacZ reporter or a heterologous reporter gene containing only the Swi5 and Pho2 binding sites from site B [the HO(site B)-lacZ reporter], a pho2 mutation does not affect expression of the endogenous HO gene (6, 24). However, mutations in the Swi5 binding sites that reduce, but do not eliminate, Swi5 binding render the HO promoter completely PHO2 dependent (24). These results suggest a complex role for Pho2 in activation of HO gene expression. The genetic data also suggests that a physical interaction between proteins bound at site A (-1800) and site B (-1300) is required for activation of HO transcription (24).

The nuclear localization of Swi5 is cell cycle regulated and has been shown to play an important role in the transcriptional regulation of HO (27, 37). Swi5 accumulates in the cytoplasm during G2, enters the nucleus only during anaphase after the inactivation of Clb/Cdc28 protein kinase, and is then rapidly degraded during G1. More recently, SWI5 has been shown to be responsible for activating a wide variety of early G1-specific genes such as EGT2 (22), ASH1 (3), CDC6 (32), RME1 (40), and SIC1 (21, 41). However, HO is specifically transcribed in mother cells only in late G1, when it requires the additional transcription factor complex SBF (16, 30).

Yeast has another zinc finger transcription factor, Ace2, which is very similar to Swi5 (8, 10). Despite the fact that the DNA-binding domains of Swi5 and Ace2 are nearly identical and the two proteins recognize the same DNA sequences in vitro, SWI5 and ACE2 can activate different genes in vivo (10, 11). SWI5, but not ACE2, activates HO expression, while ACE2 activates expression of the CTS1 gene. It is likely that the cooperative binding of Swi5 and Pho2 contributes to the specific activation of HO by SWI5 but not by ACE2.

In vitro characterization of the cooperative interaction between Swi5 and Pho2 has revealed several interesting features. First, in vitro experiments do not detect Swi5-Pho2 interaction in the absence of DNA (5). Second, although DNA-binding modules of both proteins are sufficient for DNA binding, they are insufficient for cooperative DNA binding (4). Swi5 requires a region N-terminal to the DNA-binding domain for synergistic interactions with Pho2, and Pho2 requires a region C-terminal to the homeodomain to interact with Swi5. This interaction is likely to be flexible since promoter mutations that alter the spacing between the protein binding sites are tolerated (4). To better understand the interactive surfaces of the two proteins, in this paper we describe a genetic screen to identify residues in Swi5 that are specifically defective in cooperative binding with Pho2. By both in vitro and in vivo assays, we show that specific mutations within a 24-amino-acid region (positions 482 to 505) preceding the zinc finger DNA-binding domain of Swi5 alter cooperative binding with Pho2 at the HO promoter.

    MATERIALS AND METHODS
Top
Abstract
Introduction
Materials & Methods
Results
Discussion
References

Strains. The yeast strains used in this study are listed in Table 1, and all are isogenic either in the W303 or K765 background. Strains DY2406 with the a1 mutation in the HO promoter [HO(a1)] (24) and DY1641 with a lexA-lacZ reporter integrated at the URA3 locus (20) have been described. The strains with the integrated SWI5* mutants (expressing point mutations) were constructed by first constructing strains with a swi5::URA3 disruption by using BamHI-cleaved M3405. Transformation was then performed to replace the disrupted allele with the various SWI5* mutations, with 5-fluoro-orotic acid used to screen for loss of the swi5::URA3 allele. Standard genetic methods were used for strain construction and gene replacement (33, 35), and gene replacements were confirmed by Southern analysis.

                              
View this table:
[in this window]
[in a new window]
  		TABLE 1.   Yeast strains

Plasmids. The plasmids used in this study are listed in Table 2. In many cases, multiple steps were involved in plasmid construction, and details of plasmid construction are available on request. The HO(site B)-lacZ reporter plasmid M1853 (4) and the CTS1(46)-lacZ reporter plasmid M1912 (11) have been previously described, and M3403 and M3404 are YIp versions of these reporters. Plasmid M3202 is a pRS313 (36) derivative with the BamHI site in the polylinker destroyed and contains the SWI5 gene (-1031 to +2435) with two BamHI sites introduced by site-directed mutagenesis using primers F373 (5' GTATTATTTACGGATCCAGGAATTG 3') and F378 (5' TTAATGTGGGATCCGAATTGAGG 3'). The first BamHI site at codons 396 and 397 is translationally silent, and the second BamHI site at residues 509 and 510 introduces a serine-to-threonine change at residue 510. 

                              
View this table:
[in this window]
[in a new window]
  		TABLE 2.   Plasmids

Plasmids M2024 and M2025 that express His-tagged Swi5 and Pho2 proteins, respectively, in Escherichia coli have been described previously (6). Plasmid M3113 is identical to plasmid M2024, except that the ClaI site in the vector backbone has been destroyed with T4 DNA polymerase. All of the SWI5* point mutants from the pRS313 vector backbone were cloned into the M3113 E. coli expression vector as SalI-ClaI fragments, with the exception of M3637 (the R484G mutant) and M3638 (the R484S mutant), which were generated by site-directed mutagenesis. Oligonucleotides F435 (5' GGAAATAACAAATCCACTTTCAGGCTC 3') and F435 (5' GGAAATAACAAAGCTACTTTCAGGCTC 3') were used to introduce mutations R484G and R484S, respectively.

Alanine substitutions in Swi5 (positions 482 to 505) were introduced by site-directed mutagenesis. Oligonucleotides F454 (5' ATTTCCGAAGCGCCTTCTCCC 3'), F455 (5' GAAACGCCTGCTCCCGTTCTT 3'), F456 (5' CGAAGGAAGAGCTCCTCAATTC 3'), and F457 (5' GTTATTTCCGAAGCGCCTGCTCCCGTTC 3') were used to introduce the single mutations T490A, S492A, and S505A and the double mutation T490A,S492A, respectively. To introduce multiple alanine mutations, combinations of the above oligonucleotides were used in the site-directed mutagenesis reactions.

Plasmid M3895 expressing LexA-Swi5(WT, 471-513) (LexA-wild-type Swi5 region of amino acids 471 to 513) was constructed by PCR amplification of the region with oligonucleotides F542 (5' ACAGACAATGAATTCGATGATAATGAGG 3') and F543 (5' TGTGTGGATCCCCTTAATGTGTGTG 3'), cleavage of the product with EcoRI and BamHI at sites created with the primers, and cloning of the fragment into the YEp(HIS3) plasmid pLexA202+PL (34). The same PCR strategy was used to create LexA-Swi5(471-513) fusion constructs with single point mutations [e.g., LexA-Swi5(E482K,471-513)]. Plasmid M3913, which expresses a GAD-Pho2 fusion protein with amino acids 5 to 559 of Pho2, contains an EcoRI-SalI PHO2 fragment from M3570 (PHO2 in pRS316) cloned into the YEp(LEU2) plasmid pGAD-C3 (19).

Isolation of SWI5* mutants. Residues 358 to 530 of SWI5 were PCR amplified with oligonucleotides F283 (5' TATTCAGAGAAACCTTTGGGCCTGG 3') and F375 (5' GAGTTTTCTTGTGATTTTTGAGGG 3') by using an error-prone mutagenesis protocol (23). The reaction mixture contained 16.6 mM (NH4)2SO4, 67 mM Tris (pH 8.8), 170 µg of bovine serum albumin per ml, 100 mM beta -mercaptoethanol, 1 mM TTP, 1 mM dGTP, 1 mM dCTP, 200 µM dATP, 5 ng of KpnI-linearized M2667 DNA (SWI5 in YEplac112) template per ml, and 5 U of Taq polymerase. Reactions were performed in a 100-µl volume by using amplification conditions of 15 cycles of 94°C for 1 min, 55°C for 1 min, and 72°C for 1.5 min.

Typically, the PCR products from several reaction mixtures were pooled, and then approximately 300 ng of product was directly transformed, along with the gapped plasmid (BamHI-digested M3202, with a HIS3 marker), into yeast strain DY4174 (MATa swi5 ace2) carrying the M1853 HO(site B)-lacZ reporter plasmid (URA3 marker). His+ Ura+ transformants were screened for white colony color by using the chromogenic substrate 5-bromo-4-chloro-3-indolyl-beta -D-galactopyranoside (X-Gal) in a blue versus white colony lift assay (11). Rho- mutations that affect blue colony color were eliminated by screening for growth on glycerol medium. White colonies, deficient in activating the HO(site B)-lacZ reporter, were grown on medium containing 5-fluoro-orotic acid to select for loss of the URA3 [HO(site B)-lacZ] reporter and then mated to strain DY1133 (MATalpha swi5 ace2) carrying the M1912 CTS1(46)-lacZ reporter. A blue versus white colony assay for lacZ activity was performed on the resulting diploids, and SWI5 mutants that were unable to activate the CTS1(46)-lacZ reporter were considered to contain swi5 null mutations and were discarded.

In vitro DNA binding assays. The Pho2, wild-type Swi5, and various mutant Swi5 proteins were expressed in E. coli as histidine-tagged fusion proteins and purified by HiTrap (Pharmacia) nickel column chromatography as described earlier (4). Gel retardation assays were performed with an HO(site B) probe from plasmid M1403 as described previously (24). For band shifts involving Swi5 and Pho2, a range of Swi5 concentrations was used such that a linear relationship existed between the amount of Swi5 added and the amount of binary complex. For comparative purposes with the mutant Swi5 proteins, each set of band shifts also included wild-type Swi5 as a standard control. The amount of complex formed by each Swi5 mutant was determined with ImageQuant software on a Molecular Dynamics phosphorimager and converted to a percentage relative to wild-type Swi5 binding, which was normalized to 100%.

Quantitation of RNA levels. Cells were grown in yeast extract-peptone-dextrose (YEPD) medium and harvested in early log phase, and total RNA was isolated as described previously (10). S1 nuclease protection assays were performed essentially as described previously (18) with oligonucleotides specific for HO (F376, 5' GCCCTGTGTGACATTTATGACGCGGGCAGCGGAGCCATCTGCGCACATAACGTAAGAGTTAGCCCACCGC 3'), SIC1 (F444, 5' CGACCCAATGGTTCCTGCTCTTCCCTTACTGTTCCATTATCATGACTTTCAAATTGGAATAGTGTCCTCTGACAGT 3'), and CMD1 (F393, 5' GGGCAAAGGCTTCTTTGAATTCAGCAATTTGTTCTTCGGTGGAGCC 3'). Quantitative analysis was performed with ImageQuant software and a Molecular Dynamics phosphorimager. Radioactivity in each band was measured, the background level from the corresponding position of the no-RNA lane was subtracted, and the value for HO or SIC1 was normalized by dividing by the value for the CMD1 internal control.

Other methods. Site-directed mutagenesis was performed as described previously (1), and all mutations were confirmed by dideoxy sequencing. Extracts were prepared and quantitative assays for beta -galactosidase activity were performed with the chromogenic reagent o-nitrophenyl-beta -D-galactopyranoside (ONPG) as described earlier (7).

    RESULTS
Top
Abstract
Introduction
Materials & Methods
Results
Discussion
References

Isolation of Swi5 point mutants defective for interaction with Pho2. Deletion analysis demonstrated that the DNA-binding domains of Swi5 and Pho2 are not sufficient for cooperative DNA binding at the HO promoter (4). Amino acids 537 to 632 comprise the Swi5 zinc finger DNA-binding domain, and an N-terminal region that includes part of the first zinc finger (amino acids 394 to 609) was shown to be required for interaction with Pho2 in vitro. To more precisely identify the Pho2-interacting region of Swi5, we decided to isolate Swi5 point mutations that interfered in vivo with the cooperative interaction with Pho2 at the HO promoter. Our strategy combined PCR-mediated random mutagenesis with plasmid gap repair (28) to generate a mutagenized plasmid library of SWI5.

The screen used the HO(site B)-lacZ reporter, which contains 31 nucleotides from the site B region of the HO promoter. This reporter is expressed in a SWI5 PHO2 strain but not in strains with either a swi5 or pho2 mutation (Fig. 1). We reasoned that an amino acid change in the Pho2 interaction domain of Swi5 would disrupt the ability of Swi5 to activate this reporter, and we thus screened for colonies that were white in the presence of the chromogenic substrate X-Gal. Clearly, the vast majority of these SWI5 mutations would be null alleles, so we devised a secondary screen to identify SWI5* mutants that were capable of binding DNA and activating transcription in vivo. For this secondary screen, we used the CTS1(46)-lacZ reporter, which contains a 46-bp region of the CTS1 promoter inserted into the CYC1 promoter (Fig. 1). This reporter can be activated by Swi5 but in a PHO2-independent manner (11). [The CTS1(46)-lacZ reporter can also be activated by Ace2, so the experiments were conducted with an ace2 mutant to make the reporter completely SWI5 dependent.]


View larger version (29K):
[in this window]
[in a new window]
  		FIG. 1.   Activity of reporters in SWI5* mutants. Yeast strains were independently transformed with the HO(site B)-lacZ and the CTS1(46)-lacZ reporters, and transformants were grown in selective medium and assayed for beta -galactosidase activity. Three independent transformants were assayed, and standard errors are shown. The normalized levels are also shown, as percentages of the wild-type level. All of the strains are ace2 mutants, and the SWI5* mutant alleles are present at the native SWI5 locus. The following strains were used: DY4854, DY1923, DY1143, DY1985, DY4684, DY4686, DY4852, DY4688, DY4690, DY4692, DY4695, DY4696, and DY4698.

A preliminary screen in which residues 358 to 680 of Swi5 were mutagenized was conducted; this large region (approximately 1 kb) was chosen because of available restriction sites and the deletion analysis that indicated that the Pho2 interaction region was within amino acids 394 to 609 of Swi5. Several mutations that were unable to activate the HO(site B)-lacZ reporter were isolated (e.g., F485S, Q498R, and S505P), and these residues mapped outside the zinc finger DNA-binding domain. As expected, mutations in the DNA-binding domain that failed to activate the HO(site B)-lacZ reporter also failed to activate the CTS1(46)-lacZ reporter, suggesting that the DNA-binding domain of Swi5 may be distinct from the Pho2 interaction region of Swi5 (data not shown). Based on these results, and the fact that Swi5 derivatives consisting of just the zinc finger region of Swi5 (amino acids 523 to 632) do not bind cooperatively with Pho2 (12), we decided to limit our mutagenesis to a smaller region. Site-directed mutagenesis was performed to introduce two restriction sites needed for gap repair (28), allowing us to limit our mutagenesis to a region containing amino acids 358 to 530.

A total of 20,000 yeast transformants containing plasmids with potential mutations in amino acids 358 to 530 of Swi5 were screened, and 1,000 candidates with decreased expression of the HO(site B)-lacZ reporter were identified. As described in Materials and Methods, a large number of colonies with apparent swi5 null mutations were eliminated after screening with the CTS1(46)-lacZ reporter. We selected 25 clones that were specifically defective in activation of the HO(site B)-lacZ reporter and retained at least 50% activity at the CTS1(46)-lacZ reporter. These clones were subjected to DNA sequence analysis. Some clones had single mutations, while other clones had multiple amino acid substitutions. Among the clones with multiple mutations, those with single amino acid substitutions were generated either by subcloning or by site-directed mutagenesis. We focused on nine unique mutations (derived from 15 clones, since some mutations were recovered multiple times) clustered in a 24-amino-acid region spanning residues 482 to 505, preceding the DNA-binding domain of Swi5, as shown in Fig. 1. Most of the mutations were transitions and a few were transversions, which was consistent with the conditions of mutagenesis employed (23). The high fraction of transition mutations, coupled with the nature of the genetic code, limits the spectrum of amino acid substitutions. For example, the Q498R mutation was recovered four times as the result of a CAA-to-CGA change. The other single-nucleotide transition mutations result in a TAA stop codon or a synonymous CAG glutamine codon.

In vivo analysis of Swi5 point mutants. We next analyzed the in vivo activity of the putative Pho2 interaction-defective Swi5 mutants. To avoid any possible complications due to copy number of the SWI5* mutants present on YCp plasmids, gene replacement methods were used to introduce the various SWI5* mutants at the SWI5 locus in place of the wild-type allele. These strains were transformed with the HO(site B)-lacZ reporter, and extracts were prepared for quantitative beta -galactosidase assays to measure promoter activity. As shown in Fig. 1, this reporter is dependent on both SWI5 and PHO2, since mutation of either of these genes results in a 20-fold drop in promoter activity. Importantly, the activity of all the SWI5* mutants is similar to that of the SWI5 pho2 mutant, or, as in the case of the S483G, R484G, and R484S mutants, only marginally higher. Although this phenotype is consistent with mutations that debilitate Swi5-Pho2 interaction, mutations leading to an unstable protein or transcriptionally inactive Swi5 could also cause a similar phenotype. Western immunoblot analysis showed that the various Swi5 mutants accumulated to approximately the same level as wild-type Swi5 (data not shown). We used the PHO2-independent CTS1(46)-lacZ reporter to determine whether the SWI5* mutants were transcriptionally active. Quantitative measurements showed that all the mutants activated CTS1(46)-lacZ as efficiently as wild-type SWI5 did, with the exception of the S505P mutant, which exhibited less than 50% activity (Fig. 1). This data suggests that none of the substitutions drastically alter the stability or the transcriptional ability of Swi5 and suggests that these nine residues within a 24-amino-acid patch are specifically involved in interactions with Pho2.

We next examined the ability of the SWI5* mutants to activate the native chromosomal HO gene. A quantitative S1 nuclease protection assay was used to measure HO expression, and the level of CMD1 mRNA was used as an internal control. A swi5 mutation reduces HO expression nearly 100-fold (Fig. 2A, lane 2). A pho2 mutation has little effect on expression from the native HO promoter (lane 3), as described previously (24). Interestingly, four of the nine SWI5* mutants, the V494A, S497P, Q498R, and S505P mutants, show a marked reduction in HO expression, with levels reduced to less than 50% of the wild-type SWI5 levels (Fig. 2A, compare lanes 13, 14, 15, and 16 with lane 4; all strains contain an ace2 mutation).


View larger version (84K):
[in this window]
[in a new window]
  		FIG. 2.   SWI5* mutations specifically affect HO expression. S1 nuclease protection assays using probes specific for HO (A) and SIC1 (B) were performed with the CMD1 probe as the internal control by using yeast strains which contained the wild-type SWI5 gene or Pho2 interaction-defective SWI5* mutants at the SWI5 locus. RNAs from the following strains were prepared: DY150 (lane 1), DY161 (lane 2), DY1921 (lane 3), DY4854 (lane 4), DY1923 (lane 5), DY1936 (lane 6), DY1143 (lane 7), DY4684 (lane 8), DY4686 (lane 9), DY4852 (lane 10), DY4688 (lane 11), DY4690 (lane 12), DY4692 (lane 13), DY4695 (lane 14), DY4696 (lane 15), and DY4698 (lane 16). The numbers below lanes 1 to 7 indicate HO mRNA levels, normalized to that of wild type (lane 1, 100%). For lanes 4 to 16, the strains with the SWI5* mutations all contained an ace2 mutation, and the HO mRNA levels shown below these lanes are normalized to that of the ace2 strain (lane 4, 100%). The ACE2 and PHO2 genotypes are indicated at the top of the lanes. WT, wild type.

We analyzed SIC1 gene expression in these mutants to determine whether these SWI5* mutations specifically altered HO expression or whether they also affected regulation of another SWI5-regulated G1 phase gene. The SIC1 gene is expressed in early G1 phase of the cell cycle, with both Swi5 and Ace2 contributing to activation (21, 41). This can be seen in Fig. 2B, where swi5 (lane 2) and ace2 (lane 4) mutations reduce SIC1 expression, and SIC1 expression is further reduced in the swi5 ace2 double mutant (lane 7). Since the SWI5* mutants were integrated in an ace2 mutant background, most of the residual SIC1 expression observed is SWI5 dependent. As shown in Fig. 2B, all of the SWI5* mutants, with the possible exception of the V494A mutant, were able to activate SIC1 to levels comparable to that of wild-type SWI5 alone (compare lanes 8 to 16 with lane 4).

Reduced expression of the HO(a1) promoter by the Pho2 interaction-defective Swi5 mutants. The HO promoter contains two binding sites for Swi5, site A at -1800 and site B at -1300. Although a pho2 mutation has little effect on the native HO promoter, mutations that modestly reduce Swi5 binding to either of these sites render the promoter entirely PHO2 dependent (24). These results suggest that Pho2 promotes Swi5 binding to the compromised sites via cooperative interactions and that interaction between these sites is needed for HO expression (24). This model predicts that mutations interfering with the cooperative binding between Swi5 and Pho2 would fail to activate HO expression in a strain with a mutant HO promoter that is PHO2 dependent. To confirm the nature of Pho2 interaction-defective Swi5 mutants, we integrated all the point mutations at the native SWI5 locus in a strain with the a1 mutation in the HO promoter.

The HO(a1) mutant promoter has a 2-nucleotide substitution that reduces Swi5 binding in vitro, but Pho2 is able to stimulate Swi5 binding due to the cooperative interactions (24). A quantitative S1 nuclease protection assay was used to measure expression from the HO(a1) mutant promoter (Fig. 3). HO(a1) expression is turned off in a swi5 mutant strain (lane 11) and reduced to about 5% of the wild-type level in a pho2 mutant (lane 12). Importantly, all of the SWI5* point mutants show a major defect in activation of the HO(a1) promoter despite the presence of Pho2 (lanes 2 to 10), except for those with substitutions at residue 484 (lanes 4 and 5). The R484G and R484S mutant Swi5 proteins partially activated the HO(a1) promoter, suggesting that these Swi5 proteins may retain some ability to interact with Pho2. These results corroborate a critical role for residues 482 to 505 of Swi5 in cooperative interactions with Pho2.


View larger version (32K):
[in this window]
[in a new window]
  		FIG. 3.   SWI5* mutations reduce expression from the HO(a1) promoter. Yeast strains which contained the wild-type SWI5 gene or Pho2 interaction-defective SWI5* mutants at the SWI5 locus along with the PHO2-dependent HO(a1) mutant promoter were constructed. S1 nuclease protection assays using probes specific for HO and the CMD1 probe (internal control) were performed on RNA extracted from the following strains: DY5031 (lane 1), DY4905 (lane 2), DY4907 (lane 3), DY4909 (lane 4), DY4911 (lane 5), DY4913 (lane 6), DY4915 (lane 7), DY4917 (lane 8), DY4919 (lane 9), DY4921 (lane 10), DY4843 (lane 11), and DY2406 (lane 12). The numbers below the lanes indicate the HO mRNA levels, normalized to that of a wild-type strain with the HO(a1) promoter (lane 1, 100%). The PHO2 genotype is given at the top of the lanes. WT, wild type.

In vitro defects in cooperative DNA binding. Two different assays, the lacZ reporter assay and S1 nuclease analysis, have shown that at least some of the Swi5 point mutants fit the criteria of being specifically defective for cooperative interaction with Pho2 in vivo. Although these mutations do not map to the zinc finger DNA-binding domain of Swi5, subtle alterations in the DNA-binding ability of the Swi5 mutants could also impair cooperative interactions with Pho2 in vivo. To examine this possibility, we used quantitative band shift analysis to study the DNA-binding properties of the Swi5 mutants in vitro. All of the mutants were purified from E. coli as His-tagged fusion proteins as described previously (4). First, the independent binding of each Swi5 mutant protein to the HO(site B) probe, without Pho2, was examined (Fig. 4A, C, and E). Several protein concentrations were tested for each mutant Swi5 protein, and each gel included similar concentrations of wild-type Swi5 as a standard. Quantitative analysis showed that the nine mutant proteins yield levels of protein-DNA complex similar to that seen with the wild-type Swi5 protein. This indicates that there is no apparent defect in the ability of the Swi5 mutants to bind HO DNA, at least in the absence of Pho2.


View larger version (75K):
[in this window]
[in a new window]
  		FIG. 4.   DNA binding by mutant Swi5 proteins, without and with Pho2. (A to C) DNA binding to HO promoter DNA by Swi5 alone. (D to F) DNA binding to HO promoter DNA by Swi5 in the presence of 8.1 ng of Pho2. Each panel illustrates an independent gel retardation assay using mutant Swi5 proteins, with wild-type Swi5 included in each assay as a standard internal control. The experiments in panels A and B, C and D, and E and F were conducted in pairs, at the same time with the same probe and the same preparations of purified proteins. Thus, direct comparisons can be made within each set of paired panels. (A and B) The following amounts of Swi5 were added to each binding reaction mixture: 73, 145, 290, and 580 ng of wild-type Swi5 (lanes 2 to 5, respectively); 88, 175, 350, and 700 ng of Swi5(E482K) (lanes 6 to 9, respectively); 38, 75, 150, and 300 ng of Swi5(S483G) (lanes 10 to 13, respectively); 83, 165, 330, and 660 ng of Swi5(R484G) (lanes 14 to 17, respectively). (C and D) The following amounts of Swi5 were added to each binding reaction mixture: 20, 40, 80, and 160 ng of wild-type Swi5 (lanes 2 to 5, respectively); 17, 34, 75, and 150 ng of Swi5(R484S) (lanes 6 to 9, respectively); 22, 43, 85, and 170 ng of Swi5(F485S) (lanes 10 to 13, respectively); 32, 63, 125, and 250 ng of Swi5(V494A) (lanes 14 to 17, respectively). (E and F) The following amounts of Swi5 were added to each binding reaction mixture: 73, 145, 290, and 580 ng of wild-type Swi5 (lanes 2 to 5, respectively); 55, 110, 220, and 440 ng of Swi5(S497P) (lanes 6 to 9, respectively); 62, 123, 245, and 490 ng of Swi5(Q498R) (lanes 10 to 13, respectively); 42, 83, 165, and 330 ng of Swi5(S505P) (lanes 14 to 17, respectively). Lane 1 in all panels has no added protein. The reaction mixtures in lane 18 (panels A, C, D, and E) and lane 19 (panels B and F) contained 73 ng of Swi5 protein only. WT, wild type.

We next determined whether the cooperative binding of the Swi5 mutants with Pho2 was altered (Fig. 4B, D, and F). We used the same protein concentrations that gave a linear relationship between the amount of Swi5 added and the amount of Swi5-DNA complex formed and the same labeled HO DNA probe. Quantitative analysis revealed differences in the amount of ternary complex (Swi5-Pho2-DNA) formed by the Swi5 mutants, and we divided the mutants into three broad categories (classes A to C) (Table 3). Two adjacent mutants in class A, the S497P and Q498R mutants, showed the strongest defect in interaction with Pho2, with 95 and 85% reductions, respectively, in the amount of ternary complex formed (Fig. 4F, lanes 6 to 13). Three mutants in class B, the E482K, S483G, and F485S mutants, showed a modest reduction (about 30%) in cooperative binding of Pho2 (Fig. 4B, lanes 6 to 13, and 4D, lanes 10 to 13). The last category of Swi5 mutants, class C, showed no apparent change in the cooperative DNA binding with Pho2 in vitro. Two of these mutants (the R484G and R484S mutants) have substitutions at the same residue, and these two mutations were the least defective in activating the HO(a1) promoter (Fig. 3). We also noted that the other two class C mutants, the V494A and S505P mutants, both overlap potential cyclin-dependent kinase (CDK) phosphorylation sites (see below).

                              
View this table:
[in this window]
[in a new window]
  		TABLE 3.   Quantitative analysis of in vitro DNA binding by Swi5 mutantsa

Mutations in potential CDK phosphorylation sites of Swi5. The V494A and S505P mutations of Swi5 are unusual in that they cause significant reduction in expression of both the HO(a1) promoter and the HO(site B)-lacZ reporter, yet the in vitro DNA binding experiments show no defect in cooperative interactions between the V494A and S505P mutants with Pho2. There are at least two ways to explain these differences. One is that these mutations cause a specific defect in activation of HO that is independent of Pho2 interaction. The other possibility is that in vivo modifications of Swi5, such as phosphorylation, might influence the ability of Swi5 to interact with Pho2 in vivo, while the lack of phosphorylation of E. coli-expressed Swi5 may permit interaction with Pho2 under in vitro conditions.

Earlier studies have shown that Swi5 phosphorylation regulates its subcellular localization in vivo, and three residues (S522, S646, and S664) can be phosphorylated by Cdc28 CDKs in vitro (27). More recently, Measday et al. (26) have shown that Swi5 is also a target for phosphorylation by the Pho85 CDK. Various consensus sequences have been proposed for phosphorylation sites for CDKs, including S/T-P-X-K/R for Cdc28 (27) and S/T-P-X-N (where N is any hydrophobic amino acid) for Pho85 CDK (21, 28). Swi5 contains several potential Pho85 CDK phosphorylation sites dispersed throughout the length of the protein. Interestingly, three of these putative Pho85 CDK sites are located in the Pho2-interacting region of Swi5, including two sites that overlap the V494A and S505P mutations (Fig. 5). Although the V494A mutation is within the SPVL sequence, the valine-to-alanine substitution does not alter the nature of this site as a potential phosphorylation site, based on the current consensus sequence. The S505P substitution, however, alters the phosphorylatable serine within this site.


View larger version (18K):
[in this window]
[in a new window]
  		FIG. 5.   Swi5 residues important for interaction with Pho2. The diagram shows the Swi5 protein, with residues 545 to 632 comprising the three zinc fingers of Swi5 indicated. The region of amino acids 471 to 511 is expanded, and the 482-to-505 region important for Pho2 interaction is shaded. The primary amino acid sequence of Swi5 from residues 471 to 511 is shown, amino acid substitutions that reduce interaction with Pho2 are in boldface type, and those residues that affect expression of the native HO locus are indicated with a star. The putative phosphorylation sites for the Pho85 CDK are overlined, and putative phosphorylated residues (T490, S492, and S505) that were mutated to alanines have a box beneath them.

Thus it was reasonable to investigate whether mutations at these potential phosphorylation sites alter the ability of Swi5 to activate both PHO2-dependent and PHO2-independent reporters in vivo. We used site-directed mutagenesis to convert the phosphorylatable serine or threonine residues to alanines in the three consensus sites and expressed them on YCp vector backbones either as single- or multiple-mutant combinations. Immunoblots showed that these proteins with alanine mutations accumulated to approximately the same level as the wild type did (data not shown). Table 4 shows a comparison of the Swi5 alanine mutants that were analyzed for their ability to activate HO(site B)-lacZ and CTS1(46)-lacZ reporters. We noted two intriguing features of these mutants. First, although all three substitutions (at T490, S492, and S505) map within the Pho2 interaction region of Swi5, none of the alanine mutations reduced the activity of the PHO2-dependent reporter HO(site B)-lacZ indicating that they did not interfere with Pho2 interaction. In contrast, the S492A mutation, either alone or in combination with the T490A and S505A mutations, significantly stimulated expression from both reporters. This experiment also demonstrates that alanine substitutions at all three putative phosphorylation sites lead to a striking fourfold increase in the level of the HO(site B)-lacZ reporter in comparison with a twofold increase in activity of the CTS1(46)-lacZ reporter. Interestingly, the HO(site B)-lacZ reporter is also PHO2 dependent whereas the latter reporter is not. Finally, we note that the S505P and S505A mutants have quite different phenotypes, demonstrating that alanine substitution mutations do not always indicate the importance of a particular amino acid residue.

                              
View this table:
[in this window]
[in a new window]
  		TABLE 4.   Effect of alanine substitution mutations at potential phosphorylation sites on reporter activation

Two-hybrid analysis of Swi5-Pho2 interactions. The previous experiments show that mutations in specific residues of Swi5 between amino acids 480 and 505 alter the ability of Swi5 to bind DNA cooperatively with Pho2, but they do not address whether this region is sufficient for interaction with Pho2. We used two-hybrid interactions (15) to address this question. We generated a plasmid that expresses the LexA DNA-binding domain fused in frame to a 42-amino-acid region of wild-type Swi5 (amino acids 471 to 513). This LexA-Swi5(471-513) fusion protein is unable to activate transcription of a lacZ reporter containing LexA binding sites in the promoter, either in a PHO2 strain (Fig. 6, line 2) or in a pho2 strain (data not shown). This suggests that the amino acid 471-to-513 region of Swi5 does not contain an activation domain. This experiment also suggests that although this region can interact with Pho2 (see below), native Pho2 does not provide the activation domain function for the Swi5-Pho2 heterodimer. It has been previously shown that Pho2 lacks an activation domain (17), and we have identified an activation domain present near the N terminus of Swi5 (unpublished observations).


View larger version (33K):
[in this window]
[in a new window]
  		FIG. 6.   Two-hybrid analysis of Swi5-Pho2 interactions. Strain DY1641 was transformed with bait and prey plasmids (HIS3 and LEU2 markers, respectively), and transformants were grown in selective medium and assayed for beta -galactosidase activity. Three independent transformants were assayed, and standard errors are shown. The normalized levels are also shown as a percentage of the wild-type level. The LexA-Swi5 plasmids contain amino acids 471 to 513 of Swi5 fused in frame to the LexA DNA-binding domain. The following plasmids were used: M1921, M3895, M3896, M3897, M3899, M3900, M3901, M3902, M3903, M3917, M3918, M3931, M3466, and M3913.

Since Pho2 apparently lacks an endogenous activation domain, we constructed a plasmid that expresses a Gal4 activation domain (GAD) fused to Pho2 for use in two-hybrid experiments. As shown in Fig. 6, expression of the LexA-Swi5(WT,471-513) bait along with the GAD-Pho2 prey leads to strong activation of the two-hybrid reporter (line 3). Expression of either the LexA-Swi5(WT,471-513) (line 2) prey alone or the GAD-Pho2 prey alone (line 1) does not lead to activation. This experiment clearly demonstrates that this 42-amino-acid region of Swi5 is sufficient for interaction with Pho2. In these experiments, Swi5 lacking its own DNA-binding domain is able to interact with Pho2. In contrast, in vitro experiments were unable to detect any in vitro interaction between Swi5 and Pho2 in the absence of DNA (5). The success of the present experiment may reflect the greater sensitivity of two-hybrid assays (14), or it is possible that another DNA-binding domain, LexA, can substitute for that of Swi5. In summary, this two-hybrid experiment clearly demonstrates that this 42-amino-acid region of Swi5 is sufficient for interaction with Pho2.

We next determined whether specific mutations in LexA-Swi5(471-513) affect the two-hybrid interaction with GAD-Pho2. To that end, we constructed plasmids expressing LexA-Swi5 fusions containing each of the nine single amino acid substitutions and tested them in the two-hybrid assay with GAD-Pho2 (Fig. 6). The V494A, S497P, and Q498R substitutions resulted in a dramatic drop in reporter activity to levels comparable to those of LexA-Swi5 without GAD-Pho2, suggesting a complete loss of interaction between LexA-Swi5(471-513) and GAD-Pho2. The E482K, S483G, and F485S mutations activate the reporter at slightly higher levels, indicating weak interactions with GAD-Pho2 in the two-hybrid assay. Interestingly, the three substitution mutants that retain cooperative interactions with Pho2 in vitro, the R484G, R484S, and S505P mutants, show two-hybrid interaction with GAD-Pho2 nearly as strong as that of wild-type LexA-Swi5(471-513). This implies that these three mutations only marginally affect the ability of LexA-Swi5(471-513) to interact with GAD-Pho2 in vivo. Western immunoblot analysis (data not shown) shows that the three mutant LexA fusions (the V494A, S497P, and Q498R mutants) with the most severely reduced two-hybrid interaction are expressed at the same level as that of wild-type LexA-Swi5(471-513). This demonstrates that these three amino acid residues play an important role in interactions of Swi5 with the Pho2 homeodomain protein. In summary, the two-hybrid experiments support our contention that a small region of Swi5 between residues 471 and 513 is both necessary and sufficient for interaction with Pho2.

In a previous experiment (Table 4), we showed that alanine substitutions in the three potential phosphorylation sites in Swi5(471-513) causes a fourfold increase in activity of the PHO2-dependent reporter HO(site B)-lacZ reporter, with only a modest increase in CTS1(46)-lacZ, a PHO2-independent reporter. We thus considered the possibility that changes in the phosphorylation state of this region of Swi5 might alter its interaction with Pho2. To test the hypothesis, we constructed a plasmid expressing LexA-Swi5(T490A,S492A,S505A,471-513), with alanine substitutions in all three putative phosphorylation sites. In the two-hybrid assay with GAD-Pho2, activation by he LexA-Swi5(T490A,S492A,S505,471-513) fusion protein was 1.4 times higher than that by wild-type LexA-Swi5(471-513) (Fig. 6, lines 3 and 13), although both proteins were expressed at similar levels (data not shown). This result supports the idea that mutation of the phosphorylation sites may increase the ability of Swi5 to interact with Pho2 and thus alter the activation of the HO(site B)-lacZ reporter in vivo.

    DISCUSSION
Top
Abstract
Introduction
Materials & Methods
Results
Discussion
References

The Swi5 zinc finger protein and the Pho2 homeodomain bind cooperatively to the HO promoter at both Swi5 binding sites in the promoter, site A at -1800 and site B at -1300 (6, 24). We have previously shown that cooperative interactions at the HO promoter require additional regions of each protein in addition to the DNA-binding domains. Deletion analysis mapped the interaction domain of Swi5 to a region N-terminal to the zinc fingers and that of Pho2 to a region C-terminal to the homeodomain (4). Interestingly, the two proteins do not interact in solution in the absence of DNA (5), and promoter mutation studies indicate that there is flexibility in the binding of the two proteins (4). In this study, we have explored the binding interface of Swi5 for Pho2. Genetic screens were used to identify a short stretch of Swi5, residues 482 to 505 preceding the DNA-binding domain, that is required for interaction with Pho2. Both in vitro and in vivo analyses show that some of these residues are critical for the Swi5-Pho2 interaction. Two-hybrid assays, using wild-type and mutant versions of LexA-Swi5(471-513), demonstrate that this region of Swi5 is necessary and sufficient for interaction with Pho2. We believe that these mutations in Swi5 change amino acid residues that either make critical contacts with Pho2 or disrupt the integrity of the Swi5 surface that interacts with Pho2.

The key to our strategy in mapping the Pho2 interaction-specific residues of Swi5 was the use of two SWI5-dependent reporter constructs that differ in their requirement for PHO2 (Fig. 1). The HO(site B)-lacZ reporter requires both SWI5 and PHO2 for activation, but the CTS1(46)-lacZ reporter is efficiently expressed in a SWI5 pho2 strain and is thus PHO2 independent. This allowed us to distinguish between SWI5* mutants that were specifically defective in Pho2 binding from the mutants that were transcriptionally defective, unstable, or unable to bind DNA.

We identified nine unique mutations that clustered between residues 482 and 505 of Swi5. The location of these residues is consistent with our earlier deletion analysis, in which Swi5(384-709), but not Swi5(496-709), was able to bind DNA cooperatively with Pho2 (4). The nuclear magnetic resonance solution structure of a Swi5 fragment showed that the first zinc finger has a beta -strand and an alpha -helix that are not observed in other zinc finger structures (13, 31). Amino acids 471 to 513 of Swi5, lacking the zinc finger DNA-binding domain or the additional structural elements in the first zinc finger, is sufficient for interaction with Pho2 in a two-hybrid assay. This demonstrates that the DNA-binding and Pho2 interaction domains of Swi5 are functionally and structurally distinct.

Yeast has two zinc finger proteins, Swi5 and Ace2, that have nearly identical DNA-binding domains. Although both proteins bind in vitro to site B within the HO promoter with the same affinity, only Swi5 activates HO transcription (10, 11). Chimeras containing portions of Ace2 and Swi5 have been constructed, and these experiments show that amino acids 394 to 521 of Swi5 are required for activation of HO (25). Thus, the Pho2 interaction region of Swi5 (amino acids 482 to 505), defined in this study, lies within the region of Swi5 (amino acids 394 to 521), mapped by the chimeric analysis, that is required for promoter-specific activation of HO. Because of the overlap of the Pho2 interaction region and the HO specificity region of Swi5, it is possible that mutations in this region could also affect promoter specificity of the Swi5 transcription factor (see below). Although the screen was designed to identify mutations that affect interaction with Pho2, one could expect to recover mutations that specifically affect activation of the HO gene since an HO UAS fragment was used in the primary screen.

We have classified the SWI5* mutations based upon in vitro DNA binding studies (Table 5). We first examined the ability of the Swi5 mutant proteins purified from E. coli to bind to the HO promoter in the absence of Pho2. All of the mutants showed normal DNA-binding activity. However, there were major differences in the ability of the Swi5 mutant proteins to bind DNA cooperatively with Pho2. Table 5 shows that for most Swi5 mutants there is a good correlation between the ability to interact with Pho2, either in the in vitro DNA binding assay or the in vivo two-hybrid assay, and the ability to activate transcription of PHO2-dependent promoters.

                              
View this table:
[in this window]
[in a new window]
  		TABLE 5.   Summary of Pho2 interaction-defective SWI5* allelesa

The class A mutations, S497P and Q498R, are at adjacent positions, and caused the strongest defect in cooperative binding with Pho2 in vitro. These two mutations also caused a significant drop in two-hybrid interaction with Pho2, as well as loss of activation of the PHO2-dependent promoters, HO(site B)-lacZ and HO(a1). Although the S497P mutation is a structurally severe mutation in comparison to the Q498R mutation, both substitutions have comparably severe effects in both in vitro and in vivo assays. These results suggest that residues S497 and Q498 are critical components of the Pho2-interactive surface of Swi5.

The class B mutations, E482K, S483G, and F485S, cause a moderate reduction in cooperative DNA binding with Pho2 in vitro. This apparent defect in Pho2 interaction is enhanced in the in vivo assays (Table 5). These mutations caused a strong defect in the two-hybrid assay and in activation of HO(site B)-lacZ and HO(a1), although the defect is not as pronounced as that caused by the class A mutations.

The four class C mutations R484G, R484S, V494A, and S505P result in mutants that retain most of their cooperative interactions with Pho2 in the in vitro assay. However, there are some striking differences among these mutants in the in vivo assays (Table 5). First, for the R484G and R484S mutants, the in vitro phenotype is consistent with a strong two-hybrid interaction with Pho2 and activation of the PHO2-dependent HO(a1) promoter. Based on these results, the poor activation of the HO(site B)-lacZ reporter used in the initial screen is surprising. We suggest that the HO(site B)-lacZ reporter assay is the most sensitive in vivo assay because this reporter has a single Swi5 binding site. In contrast, the HO(a1) promoter has two Swi5 binding sites, and although one Swi5 binding site has substitution mutations, interactions with Pho2 promote strong binding (24). Thus it is possible that a mutation at residue R484 very modestly reduces interaction with Pho2 and that the different assays have different degrees of sensitivity.

The other class C mutations, V494A and S505P, caused significantly reduced expression of the PHO2-dependent promoters, HO(site B)-lacZ and HO(a1), despite resulting in mutants with normal DNA binding with Pho2 in the in vitro assay (Table 5). The V494A and S505P class C mutants have one very different phenotype, since the V494A mutant fails to interact with Pho2 in the two-hybrid assay while the S505P mutant shows a strong two-hybrid interaction. As described above, the DNA-binding domain of Swi5 is not sufficient to activate HO, and regions of Swi5 overlapping the Pho2 interaction domain are required for promoter-specific activation of HO (11, 25). We suggest that the S505P substitution mutant probably does not fit the criterion of being defective in interacting with Pho2, and phosphorylation of the S505 residue may not be a crucial component of the Pho2-interactive surface. Serine 505 might be required for activating the HO promoter, and thus the S505P mutation affects activation of both native HO and HO(a1). The V494A mutant also affects expression of native HO. However, the defect of the V494A mutant in interacting with Pho2 in the two-hybrid assay suggests a dual role, with valine 494 being required for both interaction with Pho2 and for specific activation of the HO promoter. This result suggests that this region of Swi5 has multiple functions, conferring specific activation of the HO promoter and interacting with Pho2, and that these two distinct functions may overlap in one region of the Swi5 protein.

How can we reconcile the observation that for some of the SWI5* mutants, such as the V494A mutant, the in vitro DNA-binding activity does not fully correlate with the in vivo phenotype? We have previously shown that the Swi5-Pho2-DNA ternary complex is significantly more stable in vitro than either the Swi5-DNA or Pho2-DNA binary complex (6). Moreover, additional modifications such as phosphorylation may influence Swi5-Pho2 interaction in vivo, and this sensitivity might be partly lost in in vitro assays using proteins purified after expression in E. coli. Swi5 is heavily phosphorylated in vivo (22), and more recently it has been shown to be phosphorylated in vitro by Pho85 CDK (21). The Pho2 interaction region of Swi5(482-505) contains three potential phosphorylation sites for the Pho85 CDK (Fig. 5), and the S505P mutation alters one of these phosphorylatable serine residues. Additionally, conversion of the three phosphorylatable residues to alanines caused a striking increase in activity of Swi5 as a transcriptional activator. Specifically, the S492A mutation, singly or in combination with the T490A and S505A mutations, caused a significantly greater increase in activation of the PHO2-dependent HO(site B)-lacZ reporter than of the CTS1 (46)-lacZ reporter. It is conceivable that the phosphorylation status of this region of Swi5 might influence its ability to interact with Pho2.

Combinatorial control, involving cooperative interactions between DNA-binding proteins, is an important mechanism in transcriptional regulation. Specific interactions between two DNA-binding proteins allow different combinations of transcription factors to act at different genes. Pho2 interacts with at least three different partner proteins, the Swi5 zinc finger protein, the Pho4 basic helix-loop-helix protein, and the Bas1 Myb-like protein. Acting with these different partner proteins, Pho2 activates transcription of many different genes. We are unable to find any significant sequence similarities with either Pho4 or Bas1 to the region of Swi5 required for interaction with Pho2. Distinct regions of Pho2 may interact with these three proteins, or the interaction motif may be sufficiently degenerate that it cannot be identified by inspection. It is also possible that the interaction regions of Swi5, Pho4, and Bas1 may have similar structures without obvious sequence similarities. The in vivo and in vitro analyses presented here show that it is possible to identify single amino acid residues in Swi5 that are critical for Pho2 interaction and to provide new tools for the role of protein-protein interactions in combinatorial control of gene expression.

    ACKNOWLEDGMENTS

We thank members of the Stillman lab for helpful discussions and Rob Brazas, Bob Dutnall, and Helen McBride for comments on the manuscript.

Oligonucleotide synthesis and DNA sequencing were performed at the Huntsman Cancer Institute DNA/Peptide and DNA Sequencing Facilities, respectively, which are supported in part by NCI grant 5 P30 CA42014. The present work was supported by grants from the National Institutes of Health.

    FOOTNOTES

* Corresponding author. Mailing address: Division of Molecular Biology and Genetics, Department of Oncological Sciences, Huntsman Cancer Institute, University of Utah Health Sciences Center, 50 N. Medical Dr., Room 5C334 SOM, Salt Lake City, UT 84132. Phone: (801) 581-5429. Fax: (801) 581-3607. E-mail: stillman{at}genetics.utah.edu.

    REFERENCES
Top
Abstract
Introduction
Materials & Methods
Results
Discussion
References

1. Ausubel, F. M., R. Brent, R. E. Kingston, D. E. Moore, J. G. Seidman, J. A. Smith, and K. Struhl. 1987. Current protocols in molecular biology. John Wiley & Sons, Inc., New York, N.Y.
2. Barbaric, S., M. Munsterkotter, J. Svaren, and W. Horz. 1996. The homeodomain protein Pho2 and the basic-helix-loop-helix protein Pho4 bind DNA cooperatively at the yeast PHO5 promoter. Nucleic Acids Res. 24:4479-4486[Abstract/Free Full Text].
3. Bobola, N., R. P. Jansen, T. H. Shin, and K. Nasmyth. 1996. Asymmetric accumulation of Ash1p in postanaphase nuclei depends on a myosin and restricts yeast mating-type switching to mother cells. Cell 84:699-709[Medline].
4. Brazas, R. M., L. T. Bhoite, M. D. Murphy, Y. Yu, Y. Chen, D. W. Neklason, and D. J. Stillman. 1995. Determining the requirements for cooperative DNA binding by Swi5p and Pho2p (Grf10p/Bas2p) at the HO promoter. J. Biol. Chem. 270:29151-29161[Abstract/Free Full Text].
5. Brazas, R. M., and D. J. Stillman. 1993. Identification and purification of a protein that binds DNA cooperatively with the yeast SWI5 protein. Mol. Cell. Biol. 13:5524-5537[Abstract/Free Full Text].
6. Brazas, R. M., and D. J. Stillman. 1993. The Swi5 zinc finger and the Grf10 homeodomain proteins bind DNA cooperatively at the Yeast HO promoter. Proc. Natl. Acad. Sci. USA 90:11237-11241[Abstract/Free Full Text].
7. Breeden, L., and K. Nasmyth. 1987. Cell cycle control of the yeast HO gene: cis- and trans-acting regulators. Cell 48:389-397[Medline].
8. Butler, G., and D. J. Thiele. 1991. ACE2, an activator of yeast metallothionein expression which is homologous to SWI5. Mol. Cell. Biol. 11:476-485[Abstract/Free Full Text].
9. Daignan-Fornier, B., and G. R. Fink. 1992. Coregulation of purine and histidine biosynthesis by the transcriptional activators BAS1 and BAS2. Proc. Natl. Acad. Sci. USA 89:6746-6750[Abstract/Free Full Text].
10. Dohrmann, P. R., G. Butler, K. Tamai, S. Dorland, J. R. Greene, D. J. Thiele, and D. J. Stillman. 1992. Parallel pathways of gene regulation: the homologous regulators SWI5 and ACE2 differentially control transcription of HO and chitinase. Genes Dev. 6:93-104[Abstract/Free Full Text].
11. Dohrmann, P. R., W. P. Voth, and D. J. Stillman. 1996. Role of negative regulation in promoter specificity of the homologous transcriptional activators Ace2p and Swi5p. Mol. Cell. Biol. 16:1746-1758[Abstract].
12. Dutnall, R. N. 1995. Structural and functional studies of a zinc finger DNA-binding domain. Ph.D. Dissertation. University of Cambridge, Cambridge, United Kingdom.
13. Dutnall, R. N., D. Neuhaus, and D. Rhodes. 1996. The solution structure of the first zinc finger domain of SWI5: a novel structural extension to a common fold. Structure 4:599-611[Medline].
14. Estojak, J., R. Brent, and E. A. Golemis. 1995. Correlation of two-hybrid affinity data with in vitro measurements. Mol. Cell. Biol. 15:5820-5829[Abstract].
15. Fields, S., and O. Song. 1989. A novel genetic system to detect protein-protein interactions. Nature 340:245-246[Medline].
16. Herskowitz, I., B. Andrews, W. Kruger, J. Ogas, A. Sil, C. Coburn, and C. Peterson. 1992. Integration of multiple regulatory inputs in the control of HO expression in yeast, p. 949-974. In S. L. McKnight, and K. R. Yamamoto (ed.), Transcriptional regulation. Cold Spring Harbor Laboratory Press, Plainview, N.Y.
17. Hirst, K., F. Fisher, P. C. McAndrew, and C. R. Goding. 1994. The transcription factor, the Cdk, its cyclin and their regulator: directing the transcriptional response to a nutritional signal. EMBO J. 13:5410-5420[Medline].
18. Iyer, V., and K. Struhl. 1996. Absolute mRNA levels and transcriptional initiation rates in Saccharomyces cerevisiae. Proc. Natl. Acad. Sci. USA 93:5208-5212[Abstract/Free Full Text].
19. James, P., J. Halladay, and E. A. Craig. 1996. Genomic libraries and a host strain designed for highly efficient two-hybrid selection in yeast. Genetics 144:1425-1436[Abstract].
20. Kasten, M. M., D. E. Ayer, and D. J. Stillman. 1996. SIN3-dependent transcriptional repression by interaction with the Mad1 DNA-binding protein. Mol. Cell. Biol. 16:4215-4221[Abstract].
21. Knapp, D., L. Bhoite, D. J. Stillman, and K. Nasmyth. 1996. The transcription factor Swi5 regulates expression of the cyclin kinase inhibitor p40-SIC1. Mol. Cell. Biol. 16:5701-5707[Abstract].
22. Kovacech, B., K. Nasmyth, and T. Schuster. 1996. EGT2 gene transcription is induced predominantly by Swi5 in early G1. Mol. Cell. Biol. 16:3264-3274[Abstract].
23. Leung, D. W., E. Chen, and D. V. Goeddel. 1989. A method for random mutagenesis of a defined DNA segment using a modified polymerase chain reaction. Technique J. Methods Cell Mol. Biol. 1:11-15.
24. McBride, H. J., R. M. Brazas, Y. Yu, K. Nasmyth, and D. J. Stillman. 1997. Long-range interactions at the HO promoter. Mol. Cell. Biol. 17:2669-2678[Abstract].
25. McBride, H. J., and D. J. Stillman. Regions of the homologous Swi5p and Ace2p transcription factors required for specific activation of the yeast HO and CTS1 genes identified with chimeric fusions. Submitted for publication.
26. Measday, V., H. J. McBride, J. Moffat, D. J. Stillman, and B. Andrews. Interactions between the Pho85p/Pcl2p cyclin dependent kinase and the Swi5p transcription factor. Submitted for publication.
27. Moll, T., G. Tebb, U. Surana, H. Robitsch, and K. Nasmyth. 1991. The role of phosphorylation and the CDC28 protein kinase in cell cycle-regulated nuclear import of the S. cerevisiae transcription factor SWI5. Cell 66:743-758[Medline].
28. Muhlrad, D., R. Hunter, and R. Parker. 1992. A rapid method for localized mutagenesis of yeast genes. Yeast 8:79-82[Medline].
29. Nasmyth, K. 1985. At least 1400 base pairs of 5'-flanking DNA is required for the correct expression of the HO gene in yeast. Cell 42:213-223[Medline].
30. Nasmyth, K. 1993. Regulating the HO endonuclease in yeast. Curr. Opin. Genet. Dev. 3:286-294[Medline].
31. Neuhaus, D., Y. Nakaseko, J. W. R. Schwabe, and A. Klug. 1992. Solution structures of two zinc-finger domains from SWI5, obtained using two-dimensional 1H NMR spectroscopy; a zinc finger structure with a third strand of beta -sheet. J. Mol. Biol. 228:637-651[Medline].
32. Piatti, S., C. Lengauer, and K. Nasmyth. 1995. Cdc6 is an unstable protein whose de novo synthesis in G1 is important for the onset of S phase and for preventing a `reductional' anaphase in the budding yeast Saccharomyces cerevisiae. EMBO J. 14:3788-3799[Medline].
33. Rothstein, R. 1991. Targeting, disruption, replacement, and allele rescue: integrative DNA transformation in yeast. Methods Enzymol. 194:281-302[Medline].
34. Ruden, D. M., J. Ma, K. Li, K. Wood, and M. Ptashne. 1992. Generating yeast transcriptional activators containing no yeast protein sequences. Nature 350:250-252.
35. Sherman, F. 1991. Getting started with yeast. Methods Enzymol. 194:1-21[Medline].
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[Abstract/Free Full Text].
37. Tebb, G., T. Moll, C. Dowser, and K. Nasmyth. 1993. SWI5 instability may be necessary but is not sufficient for asymmetric HO expression in yeast. Genes Dev. 7:517-528[Abstract/Free Full Text].
38. Thomas, B. J., and R. Rothstein. 1989. Elevated recombination rates in transcriptionally active DNA. Cell 56:619-630[Medline].
39. Tice-Baldwin, K., G. R. Fink, and K. T. Arndt. 1989. BAS1 has a Myb motif and activates HIS4 transcription only in combination with BAS2. Science 246:931-935[Abstract/Free Full Text].
40. Toone, W. M., A. L. Johnson, G. R. Banks, J. H. Toyn, D. Stuart, C. Wittenberg, and L. H. Johnston. 1995. Rme1, a negative regulator of meiosis, is also a positive activator of G1 cyclin gene expression. EMBO J. 14:5824-5832[Medline].
41. Toyn, J. H., A. L. Johnson, J. D. Donovan, W. M. Toone, and L. H. Johnson. 1996. The Swi5 transcription factor of Saccharomyces cerevisiae controls exit from mitosis by induction of the cdk-inhibitor Sic1 in telophase. Genetics 145:85-96[Abstract].
42. Zhang, F., M. Kirouac, N. Zhu, A. G. Hinnebusch, and R. J. Rolfes. 1997. Evidence that complex formation by Bas1p and Bas2p (Pho2p) unmasks the activation function of Bas1p in an adenine-repressible step of ADE gene transcription. Mol. Cell. Biol. 17:3272-3283[Abstract].

Molecular and Cellular Biology, November 1998, p. 6436-6446, Vol. 18, No. 11
0270-7306/98/$04.00+0
Copyright © 1998, American Society for Microbiology. All rights reserved.


This article has been cited by other articles: