key: cord-0006590-vrcewef4
authors: O'Conalláin, C.; Doolin, M.-T.; Taggart, C.; Thornton, F.; Butler, G.
title: Regulated nuclear localisation of the yeast transcription factor Ace2p controls expression of chitinase (CTS1) in Saccharomyces cerevisiae
date: 1999
journal: Mol Gen Genet
DOI: 10.1007/s004380051084
sha: 33631c8c2b890531aef76ec18490179f5c186d62
doc_id: 6590
cord_uid: vrcewef4

The yeast transcription factor Ace2p regulates expression of the chitinase gene CTS1 in a cell cycle-dependent manner. Nuclear localisation of Ace2p is restricted to late M and early G1 phases of the mitotic cell cycle. We show here that this nuclear localisation is directly associated with regulation of CTS1 expression. Using a version of Ace2p tagged with a c-myc epitope, we show that the protein is excluded from the nucleus of cells during most phases of the mitotic cell cycle. A mutant derivative in which one threonine and two serine residues, which are candidate phosphorylation sites, were replaced by alanine (to mimic constitutive dephosphorylation) is localised in the nucleus throughout the cell cycle. The mechanism of localisation of Ace2p therefore involves regulation of its phosphorylation state, and closely resembles that used by the homologous transcription factor Swi5p. The wild-type Ace2 protein associates with Cdc28p in vivo, suggesting this may be the kinase that mediates the phosphorylation event. The stability of the protein is greatly reduced in a mutant that is constitutively localised to the nucleus, but is restored in a deletion derivative which remains in the cytoplasm. Ace2p is therefore controlled throughout the cell cycle at three levels: transcription, nuclear localisation, and proteolysis.

Endochitinase activity in the yeast Saccharomyces cerevisiae is encoded by a single chitinase gene, CTS1, with two allelic variants (Kuranda and Robbins 1987) . Chitinase is responsible for breaking down the chitin septum between mother and daughter cells after cell division. The enzyme is a secretory protein, and is stored in vesicles in the periplasmic space (Elango et al. 1981) . The secretory vesicles fuse to the plasma membrane and release the enzyme, some of which apparently binds to the septum via chitin-binding domains. Chitin represents only 1% of the components of the cell wall, and is localised speci®cally in the septum. De®ciencies in chitinase production therefore lead to defects in cell separation (Kuranda and Robbins 1991; Dohrmann et al. 1992) .

Expression of chitinase is controlled by the transcription factor Ace2p, ®rst identi®ed as a regulator of the metallothionein gene CUP1 (Butler and Thiele 1991; Dohrmann et al. 1992) . In strains carrying a deletion of ace2, expression of CTS1 is greatly reduced. Ace2p is very similar, particularly in the C-terminal region that includes the DNA binding domains, to the yeast transcription factor Swi5p, which regulates expression of the gene for the HO endonuclease (Nasmyth et al. 1987; Butler and Thiele 1991) . Both proteins recognise the same DNA sequence (Dohrmann et al. 1996) . ACE2 and SWI5 are regulated throughout the mitotic cell cycle in a similar manner ± both are expressed in G2, and the proteins remain in the cytoplasm until late M phase (Nasmyth et al. 1990; Dohrmann et al. 1992) . Their target genes (CTS1 and HO, respectively) are expressed in G1.

Despite these similarities, the proteins have very different functions. Swi5p does not normally function as a regulator of CTS1 expression, though overexpression of SWI5 can partially compensate for the chitinase Mol Gen Genet (1999) 262: 275±282 de®ciency phenotype of an ace2 deletion mutant (Dohrmann et al. 1992) . As well as regulating the expression of the HO endonuclease (Nasmyth et al. 1987) , Swi5p is also required for maximal expression of SIC1, EGT2, ASH1, CDC6, RME1, PCL2 and PCL9 (Piatti et al. 1995; Toone et al. 1995; Bobola et al. 1996; Kovacech et al. 1996; Toyn et al. 1997; Aerne et al. 1998 ). Overexpression of ACE2 can lead to expression of HO in a swi5-deleted strain, and ACE2 plays a role in regulation of some of the other genes. Ace2p and Cts1p (but not Swi5p) are also involved in pseudohyphal growth (King and Butler 1998) . There is also a dierence in timing between Ace2p-and Swi5p-speci®c regulation, because in strains deleted for swi5, Ace2p-driven expression of both SIC1 and EGT2 occurs later in G1 than in the wild type (Kovacech et al. 1996; Toyn et al. 1997) .

Regulation of the nuclear localisation of Swi5p has been well characterized. Phosphorylation of three serine residues by the cell cycle kinase Cdc28p causes the protein to remain in the cytoplasm, and dephosphorylation is associated with translocation to the nucleus (Nasmyth et al. 1990; Moll et al. 1991) . We show here that entry of Ace2p into the nucleus is regulated in a similar manner, and that constitutive nuclear localisation results in reduced stability of Ace2p, and increased expression of CTS1. The stability of a deletion derivative of the protein that remains in the cytoplasm is not affected, however.

The yeast strains used were DTY59 (MATa, his6, leu2-3, -112, ura3-52, ace1D-225, CUP1 R-3 ; Butler and Thiele 1991), LKY6 (MATa, his6, leu2::CTS1-lacZ, ura3-52, ace1D-225, ace2 ::hisG, CUP1 R-3 ) (King and Butler 1998) , W303/CDC28-HA (MATa, leu2, ura3, CDC28-HA, his3) (gift from N. Lowndes, ICRF), CG378 (MATa, ade5, canR, ura3 ) (Craig Giroux) and PSY580 (MATa, ura3-52, leu2D1, trp1D63) (Winston et al. 1995) . In W303/CDC28-HA the endogenous CDC28 gene has been replaced with an HA-tagged version (Sorger and Murray 1994) . All yeast strains were grown in rich (YEPD) medium or synthetic complete (SC) medium lacking nutrients as speci®ed (Sherman et al. 1986 ) at 30°C or 25°C.

To tag Ace2p with an epitope marker, a NotI site was introduced immediately upstream of the termination codon by site-directed mutagenesis, using the pALTER mutagenesis kit (Promega) and the oligonucleotide 5¢-GAAACTGATGCTGCGGCCGCCTCT GACGAACA-3¢. The myc tag sequence was isolated from pUC119 encoding three tandem copies of the c-myc 9E10 epitope (gift from S. Kron and D. Kornitzer) by PCR, using the oligonucleotides 5¢-GGGGGCGGCCGCTCCTCTAGAGGTGAACAAAAGT-3¢ and 5¢-GGGGGCGGCCGCCTATCCGTTCAAGTCTTCT-3¢, which place NotI sites at either end of the epitope in the same reading frame as the site in ACE2. The inserts were veri®ed by sequencing. A HindIII site was introduced immediately upstream of the ACE2 start codon to allow the generation of a GAL1-ACE2 fusion, using the oligonucleotide 5¢-TAAAGAAAGCTTTAGG-CCTAAAAACGG-3¢. The GAL1-10 promoter was isolated from pBM272 using HindIII and EcoRI, and cloned into the centromeric vector pRS316 (Sikorski and Heiter 1989) at the same sites. The tagged ACE2 gene was introduced at the HindIII site in this plasmid, and is expressed from the GAL1 promoter. Plasmid pGAL-ACE2myc3 expresses a version of the wild-type Ace2p marked with three copies of the myc tag, under the control of the GAL1 promoter. Plasmid pGAL-AAAmyc3 is a similar construct, but contains one threonine to alanine, and two serine to alanine substitutions as described below. Plasmid pGAL-HA contains a triple HA tag inserted at the introduced NotI site. Plasmid pMM2 contains the entire ACE2 ORF, together with promoter and terminator sequences, isolated from YEpACEr on a HpaI-HindIII fragment (Butler and Thiele 1991) and cloned into pRS316. pMM2-TRIP is a similar construct, containing the three alanine substitutions.

The wild-type ACE2 sequence contains two ClaI sites, at codons 62 and 414. To generate an N-terminal truncation, a PCR reaction was performed using the oligonucleotides TRUI (5¢-TACTATAACATCGATGAC-3¢), which correspond to the region surrounding the ®rst ClaI site, and TRU2 (5¢-CCCCATCGATT-TTCCCTTATATGGTCAA-3¢), which introduces a ClaI site at codon 181. The PCR product was used to replace the ClaI fragment in the wild-type ACE2 gene in pGAL-AAAmyc3, resulting in a product that lacks amino acids 181±414. The plasmid is called pGAL-DAAAmyc3.

Directed mutagenesis of threonine/serine codons Threonine 575 and serines 701 and 714 of Ace2p were replaced by alanines, using the site-directed mutagenesis procedure of Kunkel (1985) , and the oligonucleotides 5¢-AACAAGAACTTACAG-CGCCCAA-3¢, 5¢-CATGACACAGCTCCCGTAAAAG-3¢, and 5¢-CCTAGGACTGCGCCAATG-3¢. The mutagenesis procedure was carried out on a 770-bp EcoRI-HindIII fragment, which was later sequenced completely and used to replace the same fragment in the wild-type ACE2 gene.

Cells were grown to an OD 600 of approximately 0.8 in synthetic medium containing ranose as sole carbon source, induced by the addition of 2% galactose, and allowed to grow for up to 3 h. CG378 cells were arrested with a-factor at 1 lg/ml for 3 h where indicated (see below). Lysates were prepared from up to 100 ml of cells. Cells were pelleted and resuspended in 100 ll of lysis buer (50 mM TRIS-HCl pH 7.5, 150 mM NaCl, 1 mM MgCl 2 , 5 mM EDTA, 10 mM NaF, 50 mM b-glycerophosphate, 1 mM DTT, 0.5 mM PMSF, 25 lg/ml chymostatin, leupeptin and antipain, 5 lg/ml pepstatin) per 10 ml of cells, and lysed with glass beads. The crude extract was clari®ed by centrifugation at 13,000 rpm at 4°C for 30 min, and the protein concentration was determined using a protein assay (Bio-Rad). For detection of myc-tagged Ace2p, 100 lg of total protein was fractionated on a 7.0% SDSpolyacrylamide gel, and transferred to nitrocellulose membranes using a semi-dry blotter (C.B.S.). The anti-c-myc monoclonal antibody 9E10 (1 mg/ml) was obtained from Noel Lowndes (ICRF) or purchased from Boehringer Mannheim, and used at a dilution of 1/1000 in 0.5´blocking solution (Boehringer Mannheim). HRPlinked anti-mouse IgG (Sigma) was used at a dilution of 1/700 and detected by luminol-based chemiluminesence (Boehringer Mannheim; Pierce). Anti-HA monoclonal antibodies (12CA5 at 0.4 mg/ ml, Boehringer Mannheim) were used at a dilution of 1/1000 to detect HA-tagged Cdc28 and Ace2 proteins. To verify protein concentrations, membranes were stripped following the manufacturer's instructions, and tubulin levels were measured with the monoclonal antibody YOL1/34 (0.5 mg/ml, Harlan Sera-Labs), at a dilution of 1/500, and HRP-conjugated anti-rat IgG (Sigma). For immunoprecipitation, 250 lg of clari®ed extract was incubated with 4 lg of 9E10 antibody at 4°C for 1 h. 25 ll of Protein G beads (Sigma) was added and the incubation continued for a further hour at 4°C. The immunoprecipitated material was washed ®ve times with lysis buer, resuspended in SDS sample buer and fractionated by electrophoresis on a 10% SDS polyacyrlamide gel. HA-tagged Cdc28p and c-myc-tagged Ace2p were detected as above.

The method used was adapted from that of Adams and Pringle (1994) . Cells were grown overnight in selective synthetic medium, and expression of the GAL1 promoter was induced by growth in YP containing 2% galactose for 3 h. CG378 cells were arrested with alpha factor (1 lg/ml) for 45 min. The cells from 5 ml of culture were harvested and ®xed in 1 ml 0.1 M potassium phosphate (pH 6.5) containing 4% formaldehyde. Cells were incubated at room temperature for 2 h, washed and resuspended in 250 ll of 0.1 M potassium phosphate pH 6.5, 1.2 M sorbitol. Cell walls were digested for times varying up to 15 min with 38 lg of lyticase (in 50% glycerol) and mercaptoethanol (1.25 ll). The cells were washed and resuspended in 100 ll of 0.1 M potassium phosphate pH 6.5, 1.2 M sorbitol. Then 15 ll of cell suspension was applied to polylysine-coated slides (Sigma) for 3 min, and plunged into methanol at A20°C for 6 min and acetone at A20°C for 30 s. Non-speci®c antibody binding was blocked by incubation for 30 min in 1% BSA in phosphate-buered saline pH 7.5. The primary anti-c-myc monoclonal antibody 9E10 was applied at a concentration of 5 lg/ml in blocking solution and a¯uoroisothiocyanate (FITC)-conjugated anti-mouse IgG secondary antibody was applied at a concentration of 10 lg/ml in blocking solution (Jackson Immunoresearch Laboratories). The cells were incubated for a minimum of 1 h with 9E10 and 45 min with anti-mouse IgG, and washed with blocking solution between antibody treatments. The cells were also stained with 4¢,6-diamidino-2-phenylindole (DAPI; Sigma) at a concentration of 0.5 lg/ml in phosphatebuered saline pH 7.5, to visualise cell nuclei. The mounting solution used contained 90% glycerol and 10% phosphate-buered saline pH 7.5. To reduce photobleaching, 1,4-diazabicyclo-[2.2.2]octane (DABCO; Sigma) was added to the mounting solution at a concentration of 2.5%. Cells were visualized under 100´magni®cation using an Olympus BX60 microscope with a UV light source, and were photographed with Ilford HP400 ®lm at an ISO of 1600.

We have previously demonstrated, using an ACE2-bgalactosidase fusion, that the location of the Ace2 protein varies during the cell cycle (Dohrmann et al. 1992) . In G2 or early M phase cells, Ace2p is found in the cytoplasm, and it translocates to the nucleus in late M phase, remaining there into G1. The similarities in sequence between Ace2p and Swi5p suggested that their localisation might be regulated in similar fashion (Dohrmann et al. 1992) . Cell cycle-speci®c phosphorylation of Swi5p at three serine residues by the kinase Cdc28 causes retention in the cytoplasm; translocation to the nucleus in late M phase is associated with dephosphorylation (Nasmyth et al. 1990; Moll et al. 1991) . Two of these serines are conserved in Ace2 (at positions 701 and 714), and the third is replaced by a threonine (at position 575), which is also part of a Cdc28p recognition site (Fig. 1) . To investigate the role of these residues in the localisation of Ace2p, a mutant protein was constructed in which the threonine and the serines were replaced by alanines, to mimic constitutive dephosphorylation (Ace2-AAAp). b-Galactosidase fusions proved to be markedly unstable (data not shown), so the wild-type and mutant proteins were tagged with a triple c-myc epitope (Bobola et al. 1996) . Expression was driven from the GAL1 promoter to prevent any eects due to cell cycle-dependent regulation of expression of the constructs. The fusions are biologically active, as they complement the chitinase de®ciency phenotype of an Fig. 1 Comparison of the Ace2p and Swi5p proteins. The two proteins were aligned with the aid of the Gap program (GCG Wisconsin package); identical amino acids are indicated by vertical lines, and similar amino acids by colons or periods. The conserved potential Cdc28p recognition sites are boxed. One site overlaps part of the nuclear localization signal in Swi5p (Tebb et al. 1993 ) that is also conserved in Ace2p (indicated by the stippled boxes). The region deleted to generate the truncated DAAAp protein is indicated by the arrows ace2 deletion mutant (data not shown). The tagged proteins were detected by immuno¯uorescence.

The subcellular localisation of Ace2p varies with cell cycle phase, even when the gene is ectopically expressed from the GAL1 promoter ( Fig. 2A) . In cells in G1 phase (characterised by round unbudded cells with a single nucleus), Ace2p is localised to the nucleus, and no cytoplasmic staining is evident ( Fig. 2A) . This is more clearly seen in Fig. 2B , where cells have been arrested in G1 with alpha-factor. In most other cells the cytoplasm, but not the nucleus, is heavily stained (Fig. 2A) . Ace2p is therefore excluded from the nucleus in these cells, even when expressed from an ectopic promoter, suggesting an active mechanism is required for nuclear entry. In the mutant protein containing the triple alanine substitutions, staining is predominantly nuclear (Fig. 2C) . Although not all cells express the protein, when present it is found in the nucleus at all stages of the cell cycle; the cytoplasmic localisation and stage-dependent exclusion from the nucleus described for the wild-type protein is never observed. Substitution of alanine at these positions is therefore associated with increased nuclear localisation of Ace2p, as observed with the corresponding derivative of Swi5p (Nasmyth et al. 1990 ).

Both wild-type and triple-alanine mutant myc fusion proteins were expressed from the GAL1 promoter, and detected by Western analysis (Fig. 3A) . The wild-type Ace2 protein (Fig. 3A, lane 2 ) was detected at much higher levels than the triple alanine mutant version (lane 3), although screening with an anti-tubulin antibody demonstrates that approximately equal amounts of total protein were loaded in both lanes. This suggests the mutant protein is much less stable than the wild type, and this reduced stability is correlated with its cellular location. It is likely that the change in stability is a direct result of entry into the nucleus rather than an eect of the alanine substitutions on protein structure, as the wild-type Ace2p is also unstable when cells are arrested with a-factor (Fig. 3B ). In these cells, Ace2p is also localised in the nucleus ( Fig. 2B ; Dohrmann et al. 1992 ). The protein used in this experiment was tagged with a triple HA-epitope, which allows detection at very low levels.

Swi5p is also an unstable protein, and the instability is associated with a region towards the N-terminal end of the protein (Tebb et al. 1993 ). Ace2p and Swi5p are not very similar in this region, but do share a sequence which may form an a-helical structure (Fig. 1) . To determine the role of the N-terminal region in regulating the stability of Ace2p, we generated a protein derivative that contained the three alanine substitutions, but was also deleted for amino acids 181±414 (DAAAp). As shown in Fig. 3C , the levels of protein detected are much greater than for the equivalent alanine mutant, which is almost undetectable (Fig. 3C ; compare lane 4 with lane 2). The protein levels are also approximately 25% higher than that of wild-type Ace2p. The size of the protein is reduced by 26 kDa, as a result of the deletion. Rather surprisingly however, this mutant derivative is not constitutively localised in the nucleus, but rather is excluded from the nucleus (Fig. 2D ). It is therefore apparent that nuclear localisation requires N-terminal sequences, and that stability is directly associated with cellular localisation.

Because approximately 2% of yeast genes encode protein kinases (Hunter and Plowman 1997) it is important to determine which kinase is responsible for phosphorylation of Ace2p in vivo. As Swi5p is phosphorylated by Cdc28p (Moll et al. 1991) we investigated potential interactions of Ace2p with this kinase. A co-immunoprecipitation experiment was carried out using an Ace2 protein tagged with a triple c-myc epitope, and a yeast strain in which the endogenous Cdc28 had been replaced by Cdc28 tagged with an HA epitope (Sorger and Murray 1994) . The wild-type and mutant Ace2 proteins were immunoprecipitated using the anti c-myc antibody (9E10). Subsequent screening of the immunoprecipitate with anti-HA antibody (12CA5) revealed the presence of a protein of the same size as Cdc28-HA, which coimmunoprecipitated with Ace2p (Fig. 4, lane 3) . The protein was also detected, though in lower amounts, when Ace2-AAAp was immunoprecipitated (Fig. 4, lane  2) . This may be due to the reduced stability of the triple alanine construct, or may re¯ect a biologically relevant dierence. The fact that the wild-type Ace2p and Cdc28p co-immunoprecipitate however suggests the two proteins interact in vivo, and the resulting phosphorylation event is associated with retention of Ace2p in the cytoplasm.

To determine the eect of nuclear localisation of Ace2p on the expression of chitinase, a CTS1-lacZ fusion was integrated in the genome of a yeast strain carrying an ace2 disruption. Expression of chitinase could therefore be monitored by measuring b-galactosidase levels. The wild-type or triple alanine mutant form of Ace2p was supplied on a single-copy plasmid, under the control of Fig. 2A±D A mutant version of the Ace2p (Ace2-AAAp) is constitutively localised to the nucleus. Yeast cultures were transformed with plasmids expressing Ace2p, Ace2-AAAp or a deletion variant (DAAAp) from the GAL1 promoter. All three proteins are tagged with the c-myc epitope, and detected with anti-c-myc (9E10) and FITC-conjugated secondary antibodies. The panels on the right (FITC) show detection of the Ace2 proteins, and those on the left (DAPI) show the location of the nucleus containing DAPI-stained DNA. In A, Ace2p is clearly excluded from the nucleus in most cycling cells, and is found in nucleus only in round unbudded (G1) cells. Panel B shows the localisation of Ace2p in a-factor arrested cells. Panel C shows Ace2-AAAp, which is predominantly nuclear when present, and is never excluded from the nucleus. Panel D shows the localisation of a version of Ace2-AAAp (DAAAp), which is deleted for 234 amino acids in the N terminal region. This derivative is always excluded from the nucleus. The host strains used were DTY59 (A), CG378 (B) and PSY580 (C and D). The host strain chosen has no eect on the patterns seen the endogenous promoter. As reported previously, expression of CTS1 is greatly reduced in the absence of the Ace2 protein (Dohrmann et al. 1992 (Dohrmann et al. , 1996 . In the presence of a wild-type Ace2p, expression of CTS1-lacZ is reduced in cells arrested in early M phase by treatment with nocodazole, as compared to cycling cells (Fig. 5) . When expression is driven from the triple alanine mutant however, CTS1-lacZ levels are higher than wild type in cycling cells, and remain high in nocodazolearrested cells. High-level expression of CTS1 therefore correlates with increased amounts of Ace2p in the nucleus.

Our results suggest that cell cycle-regulated entry of Ace2p into the nucleus is associated with dephosphorylation of one threonine and two serine residues, and the mechanism closely resembles that previously described for Swi5p. The protein is associated with Cdc28 kinase in vivo, suggesting that this may be the kinase responsible for the phosphorylation event. Alteration of phosphorylation levels is a common mechanism for regulating transport of diverse proteins to the nucleus Vandromme et al. 1996) . In many cases phosphorylation is associated with transport into the nucleus (Mosialos et al. 1991; Rihs et al. 1991; Vancurova et al. 1995) , and phosphorylation of the SV40 large T antigen increases the binding anity of the nuclear localisation signal for the nuclear import machinery (HuÈ bner et al. 1997; Xiao et al. 1998 ). However, phosphorylation by a cell cycle-dependent kinase (CDK) has been shown to correlate with retention in the cytoplasm (Moll et al. 1991; Sidorova et al. 1995; O'Neill et al. 1996) , and dephosphorylation of the NF-AT family of transcription factors by calcineurin is associated with nuclear localisation (Beals et al. 1997; Scott et al. 1997) . Our results suggest that the dephosphorylation event is a positive signal for nuclear transport, as the wild-type Ace2 protein is actively excluded from the nucleus in most cycling cells, even when the protein is overexpressed from the GAL1 promoter ( Fig. 2A) . This exclusion was not previously described for Swi5p. A mutant version of Ace2p, containing three alanine substitutions, is constitutively localised to the nucleus (Fig. 2C) .

Although Swi5p is a major activator of expression of HO, constitutive nuclear entry of Swi5p does not result in constitutive expression of HO. The transcription factor Swi6p also regulates HO expression, and exhibits phosphorylation-dependent cell cycle-regulated nuclear localisation (Sidorova et al. 1995) . Constitutive nuclear localisation of Swi6p also does not aect cell cycle regulation of HO (Sidorova et al. 1995) . We demonstrate here that a mutant form of Ace2p which is localised to the nucleus throughout the cell cycle (Fig. 2C ) causes an associated increase in expression of CTS1 (Fig. 5) . Expression of CTS1 directed by the Ace2p triple alanine mutant remains high in cells arrested in early M phase (Fig. 5) . The localisation of Ace2p therefore plays an important role in regulating expression of its target gene. Ace2p belongs to a growing number of transcription factors whose localisation is controlled, and this is an important element in the regulation of their activity (Jans 1995; Vandromme et al. 1996) . It is likely that retention of transcription factors in the cytoplasm, owing to phosphorylation by CDKs, is a major mechanism of transcriptional regulation in eukaryotes, as the Swi5 protein is localised in a similar manner when expressed in mammalian cells .

The stability of Ace2p varies with cell cycle phase, and dramatically decreases when the protein is localised Fig. 4 Ace2p and Cdc28p form a complex. Ace2-AAAp (lane 2) and Ace2p (Lane 3) were immunoprecipitated from W303 cells expressing Cdc28-HA using anti-c-myc antibody (9E10) and Protein G beads. The immunoprecipitate was then screened with anti-HA (12CA5) antibody to detect Cdc28-HA. Lane 4 shows direct precipitation of Cdc28-HA using the 12CA5 antibody, and lane 1 is a mock precipitation using untagged Ace2p. The heavy chains of the primary antibody (Ab) are indicated The ACE2 derivatives were expressed from the endogenous promoter. b-Galactosidase levels were measured in cycling cells (empty bars), and in cells arrested in early M phase by treatment with nocodazole (®lled bars). The values are averages from at least three independent transformants, and the standard errors are indicated in the nucleus (Fig. 3) . Swi5 protein is also extremely unstable when localised in the nucleus (Tebb et al. 1993) . Cell cycle-regulated proteolysis of speci®c proteins has been implicated in driving cell cycle progression (King et al. 1997) . Many cyclins contain PEST sequences (Rechsteiner and Rogers 1996) , which target them to a ubiquitin-dependent conjugation pathway leading to degradation by the 26S proteasome (Hilt and Wolf 1996) . Proteolysis at the metaphase-anaphase transition is directed through a dierent ubiquitin-dependent pathway, represented by the Anaphase-Promoting Complex (King et al. 1997; Zachariae et al. 1997) . The degradation signal is a 9-amino acid motif called the destruction box (D box). Swi5p contains a PEST sequence that is detected using the program PESTFIND (Rechsteiner and Rogers 1996) , but this motif is not in the region determined to be responsible for instability (Tebb et al. 1993 ). The Ace2 protein sequence does not contain any regions with close matches to either PEST sequences or D-boxes. The region of Swi5p that confers instability has been localised to a region between amino acids 182 and 326 (Tebb et al. 1993 ). Comparison of this sequence with Ace2p reveals the presence of two conserved regions, a glutamine-rich region which may form an amphipathic a-helix between residues 186 and 234 of Ace2p, and a second region between amino acids 290 and 322 (Fig. 1) . Both of these regions were deleted in the stable derivative described here. It is therefore possible that degradation of Ace2p and Swi5p is regulated through one or both of these regions.

Ace2p and Swi5p are very similar, both in sequence and in function. Both proteins bind to the same DNA sequence in vitro (Dohrmann et al. 1996) and we have now demonstrated that regulation of nuclear localisation of both is similar. However, nuclear localisation of Ace2p requires sequences in the N-terminal segment of the protein, which has not been reported for Swi5p (Fig. 2D ). The two transcription factors also activate expression of dierent target genes: Swi5p, but not Ace2p, regulates expression of the genes for the endonuclease HO and the cyclin-like proteins Pcl2 and Pcl9 (Aerne et al. 1998) , and Ace2p alone regulates expression of CTS1 (Dohrmann et al. 1992 ). Disruption of ACE2, but not SWI5, allows production of pseudohyphae in some genetic backgrounds (King and Butler 1998) . Both proteins are required for maximal expression of RME1 (which encodes a negative regulator of meiosis) (Toone et al. 1995) , ASH1 (required for mother cell-speci®c expression of HO; Bobola et al. 1996) , the cyclin kinase inhibitor Sic1 (Knapp et al. 1996; Toyn et al. 1997 ) and the early G1 transcript EGT2 (Kovacech et al. 1996) . Interestingly, expression of the Ace2pdriven CTS1 gene occurs later in G1 than that of the Swi5p-driven SIC1 (Toyn et al. 1997) . In a swi5-deleted strain, Ace2p-driven expression of SIC1 is later then Swi5p-regulated expression, and coincides with expression of CTS1. SIC1 is also expressed during telophase, while CTS1 is not. Expression of EGT2, which is primarily controlled by Swi5p, is also delayed in a swi5 deletion background (Kovacech et al. 1996) . These differences may be due to a dierence in the timing of nuclear localisation of Ace2p and Swi5p which is not detectable in our experiments, or may be due to dierent interactions between Swi5p and Ace2p with ancillary proteins. Swi5p binding to the HO promoter is enhanced by the Pho2p homeodomain factor (Brazas and Stillman 1993; Brazas et al. 1995) , but Ace2p does not interact with Pho2p (Dohrmann et al. 1996) . It also appears that some factors prevent either Ace2p or Swi5p binding to the wrong promoters. A series of mutations in the CTS1 promoter (NCE mutations) allow Swi5-driven expression of CTS1 in the absence of Ace2 (Dohrmann et al. 1996) . Mutations in SIN5 allow Ace2p-driven expression of HO, suggesting Sin5p may normally prevent Ace2p binding to the HO promoter (Stillman et al. 1994; Dohrmann et al. 1996) . It will be interesting to determine the range of genes regulated by the two transcription factors and their degree of overlap, and experiments are underway to investigate the role of the timing of nuclear entry and proteolysis on the functions of these two proteins.

Relationship of actin and tubulin distribution to bud growth in wild-type and morphogeneticmutant Saccharomyces cerevisiae

Swi5 controls a novel wave of cyclin synthesis in late mitosis

Nuclear localization of NF-ATc by a calcineurin-dependent, cyclosporin-sensitive intramolecular interaction

Asymmetric accumulation of Ash1p in postanaphase nuclei depends on a myosin and restricts yeast mating-type switching to mother cells

The Swi5 zinc ®nger and the Grf10 homeodomain proteins bind DNA cooperatively at the yeast HO promoter

Determining the requirements for cooperative DNA binding by Swi5p and Pho2p (Grf10p/Bas2p) at the HO promoter

ACE2, an activator of yeast metallothionien expression which is homologous to SWI5

Parallel pathways of gene regulation: homologous regulators SWI5 and ACE2 dierentially control transcription of HO and chitinase

Role of negative regulation in promoter speci®city of the homologous transcriptional activators Ace2p and Swi5p

Secretory character of yeast chitinase

Proteasomes: destruction as a programme

The protein kinase CK2 site (Ser 111/112 ) enhances recognition of the simian virus 40 large T-antigen nuclear localization sequence by importin

The protein kinases of budding yeast: six score and more

The regulation of protein transport to the nucleus by phosphorylation

Cyclin-dependant kinase site-regulated signal-dependant nuclear localization of the SWI5 yeast transcription factor in mammalian cells

Ace2p, a regulator of CTS1 (chitinase) expression, aects pseudohyphal production in Saccharomyces cerevisiae

How proteolysis drives the cell cycle

The transcription factor Swi5 regulates expression of the cyclin kinase inhibitor p40 SIC1

EGT2 gene transcription is induced predominantly by Swi5 in Early G1

Rapid and ecient site-speci®c mutagenesis without phenotypic selection

Cloning and heterologous expression of glycosidase genes from Saccharomyces cerevisiae

Chitinase is required for cell separation during growth of Saccharomyces cerevisiae

The role of phosphorylation and the CDC28 protein kinase in cell cycle-regulated nuclear import of the S. cerevisiae transcription factor SWI5

A protein kinase-A recognition sequence is structurally linked to transformation by p59v-rel and cytoplasmic retention of p68c-rel

Cell cycle regulation of SWI5 is required for mother-cell-speci®c HO transcription in yeast

The identi®ca-tion of a second cell cycle control on the HO promoter in yeast: cell cycle regulation of SWI5 nuclear entry

Regulation of PHO4 nuclear localization by the PHO80-PHO85 Cyclin-CDK complex

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

PEST sequences and regulation by proteolysis

The rate of nuclear cytoplasmic protein transport is determined by the casein kinase II site¯anking the nuclear localization sequence of the SV40 T-antigen

Dynamic equilibrium between calcineurin and kinase activities regulates the phosphorylation state and localization of the nuclear factor of activated T-cells

Methods in yeast genetics

Cell cycle-regulated phosphorylation of Swi6 controls its nuclear localization

A system of shuttle vectors and yeast host strains designed for ecient manipulation of DNA in Saccharomyces cerevisiae

S-phase feedback control in budding yeast independent of tyrosine phosphorylation of p34cdc28

Epistasis analysis of suppressor mutations that allow HO expression in the absence of the yeast SWI5 transcriptional activator

SWI5 instability may be necessary but is not sucient for asymmetric HO expression in yeast

Rme1, a negative regulator of meiosis, is also a positive activator of G1 cyclin gene expression

The Swi5 transcription factor of Saccharomyces cerevisiae has a role in exit from mitosis through induction of the cdk-inhibitor Sic1 in telophase

Nucleoplasmin associates with and is phosphorylated by casein kinase II

Regulation of transcription factor localization: ®ne tuning of gene expression

Construction of a set of convenient Saccharomyces cerevisiae strains that are isogenic to S288C

Negative charge at the protein kinase CK2 site enhances recognition of the SV40 large T-antigen NLS by importin: eect of conformation

Identi®cation of subunits of the anaphase-promoting complex of

Acknowledgements We gratefully acknowledge the assistance and advice of Dr Jeremy Toyn and Dr Lee Johnson at the National Institute for Medical Research, and Dr Noel Lowndes at the Imperial Cancer Research Foundation. We thank Ray Deshaies for the Cdc28-HA construct, and Dennis Thiele for the ace2 disruption plasmid. C. O'Conallain and M-T. Doolin were partially supported by the Irish Science Agency, Forbairt. This work was supported by grants from Forbairt and from the Wellcome Trust (038606/Z/93/ Z) to G. Butler.