Yeast ASF1 Protein Is Required for Cell Cycle Regulation of Histone Gene Transcription

Corresponding author: Rolf Sternglanz, Department of Biochemistry and Cell Biology, State University of New York, Stony Brook, NY 11794-5215., rolf{at}life.bio.sunysb.edu (E-mail)

Communicating editor: F. WINSTON

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Transcription of the four yeast histone gene pairs (HTA1-HTB1, HTA2-HTB2, HHT1-HHF1, and HHT2-HHF2) is repressed during G1, G2, and M. For all except HTA2-HTB2, this repression requires several trans-acting factors, including the products of the HIR genes, HIR1, HIR2, and HIR3. ASF1 is a highly conserved protein that has been implicated in transcriptional silencing and chromatin assembly. In this analysis, we show that HIR1 interacts with ASF1 in a two-hybrid analysis. Further, asf1 mutants, like hir mutants, are defective in repression of histone gene transcription during the cell cycle and in cells arrested in early S phase in response to hydroxyurea. asf1 and hir1 mutations also show very similar synergistic interactions with mutations in cac2, a subunit of the yeast chromatin assembly factor CAF-I. The results suggest that ASF1 and HIR1 function in the same pathway to create a repressive chromatin structure in the histone genes during the cell cycle.

ASSEMBLY of genes into nucleosomes leads to transcriptional repression by preventing access of key components of the transcriptional machinery to the DNA. Factors that contribute to formation of repressive chromatin include histone deposition complexes, chromatin-remodeling complexes, histone deacetylases, and, in the case of the silenced loci, specific silencing proteins (reviewed in TYLER and KADONAGA 1999 ). Transcriptional activation is often brought about by chromatin-remodeling complexes that lessen the interaction between DNA and the histone octomers in nucleosomes (KADONAGA 1998 ). The importance of the nucleosome in gene regulation is supported by the finding that mutations that alter intracellular levels of histones lead to transcriptional derepression of a variety of genes (NORRIS and OSLEY 1987 ; CLARK-ADAMS et al. 1988 ; HAN and GRUNSTEIN 1988 ). Phenotypes resulting from histone imbalances in the cell include slow growth, disruption of silencing, and altered chromosome transmission fidelity (MEEKS-WAGNER and HARTWELL 1986 ; HAN and GRUNSTEIN 1988 ; KAUFMAN et al. 1998 ).

Core histones in Saccharomyces cerevisiae are encoded by four divergently transcribed gene pairs, HTA1-HTB1, HTA2-HTB2, HHT1-HHF1, and HHT2-HHF2. Expression of these genes is under very tight cell cycle control such that their mRNA accumulates only during S phase. This ensures that histones are present when they are most needed, as new DNA is being made and readied to be packaged into chromatin. For three of the four histone gene pairs (all except HTA2-HTB2), this periodic RNA accumulation results from transcriptional activation at G1/S and cell cycle repression in early G1, G2, and M (OSLEY et al. 1986 ). The repression requires several trans-acting factors, including HIR1 and HIR2, highly conserved proteins that interact with each other (OSLEY and LYCAN 1987 ; SHERWOOD et al. 1993 ; SPECTOR et al. 1997 ).

ASF1 was originally identified in a screen for genes that, when overexpressed, cause a defect in silencing at the mating-type loci in S. cerevisiae (LE et al. 1997 ). Subsequent analyses revealed that overexpression of ASF1 also reduces silencing at telomeres (LE et al. 1997 ; SINGER et al. 1998 ) and rDNA (SINGER et al. 1998 ). The ASF1 protein is highly conserved, with homologs in Drosophila, Caenorhabditis elegans, mice, humans, and plants. While asf1 null mutants are viable, the cells grow slowly, are sensitive to DNA-damaging and replication-blocking agents (LE et al. 1997 ; TYLER et al. 1999 ), and show an elevated chromosome loss rate relative to wild-type cells (LE et al. 1997 ). These phenotypes suggest a role for ASF1 in chromatin structure and/or DNA damage repair. ASF1 mRNA fluctuates during the cell cycle, peaking at the G1/S transition (LE et al. 1997 ; CHO et al. 1998 ; SPELLMAN et al. 1998 ). This regulation is most likely achieved through transcriptional activation via two mluI cell-cycle box (MCB) boxes located upstream of the ASF1 coding region.

Very recently the Drosophila ASF1 homolog was identified as a factor required for the in vitro assembly of nucleosomes onto newly replicated DNA (TYLER et al. 1999 ). In this analysis, ASF1 was found in a complex with histones H3 and H4. The histones that were associated with ASF1 had an acetylation pattern characteristic of newly synthesized histones. The data suggested that ASF1 is a histone chaperone involved in chromatin assembly in vitro, and phenotypes of asf1 mutants in yeast are consistent with such a role in vivo (TYLER et al. 1999 ). Subsequent to this analysis, a human homolog of ASF1 was also shown to bind histones H3 and H4 and to promote the formation of nucleosomes onto DNA (MUNAKATA et al. 2000 ). Assembly of chromatin onto newly replicated DNA in vitro is facilitated not only by ASF1, but also by chromatin assembly factor 1 (CAF-I), which also acts as a chaperone for histones H3 and H4 (SMITH and STILLMAN 1989 ; KAMAKAKA et al. 1996 ). It was further shown that, in yeast, mutations in ASF1 are synergistic with mutations in CAC1, which encodes a subunit of CAF-I. Strains containing both mutations grow more slowly and show defects in silencing and sensitivity to radiation that are much more severe than in strains with either single mutation (TYLER et al. 1999 ).

Mutations in genes encoding CAF-I subunits (CAC1, 2, and 3) also show interactions with mutations in HIR1 and HIR2 (KAUFMAN et al. 1998 ; QIAN et al. 1998 ). hir1 mutations have no effect on cell growth rate or silencing. However, when combined with mutations in CAC1, CAC2, or CAC3, cells have a slower growth rate, increased sensitivity to methyl methanesulfonate, and greater silencing defects than seen in cac mutants alone (KAUFMAN et al. 1998 ; QIAN et al. 1998 ). These results led to the hypothesis that the HIR proteins might be involved in histone deposition and/or heterochromatin formation at loci other than the histone genes (KAUFMAN et al. 1998 ). However, it is also possible that the synergistic effects between cac and hir mutations result from a sensitivity of cac mutants to the histone imbalances that result from loss of repression of the histone genes in hir strains (KAUFMAN et al. 1998 ; QIAN et al. 1998 ).

Here we report the identification of HIR1 in a two-hybrid screen with ASF1 as bait. We further show that ASF1, like HIR1, is required for the cell cycle repression of histone gene transcription. Finally, an analysis of the genetic interactions among ASF1, CAC2, and HIR1 suggests that ASF1 and HIR1 function together in a pathway to establish the normal chromatin structure required for repression of the histone genes.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Strains and media:
Strains used in this study are listed in Table 1. Unless otherwise indicated, all strains were grown in YPD medium supplemented with adenine at 25 µg/ml. For fluorescent microscopy, cells were grown in SC medium lacking methionine and uracil and supplemented with adenine at 150 µg/ml.

Plasmids:
pCALA1 is a CEN3-LEU2 plasmid with lacZ under control of the HTA1 promoter (OSLEY et al. 1986 ). ASB25 (pGFP-ASF1) was created in a three-step process. First, PCR was used to amplify the amino-terminal two-thirds of the ASF1 coding region such that the ATG was replaced with a HindIII site. The PCR fragment was cloned as a HindIII-Sty1 fragment into plasmid pLS67 (ASF1 in pUC18) cut with the same enzymes so that the 5' flanking sequences and ATG of ASF1 were replaced with a HindIII site. The ASF1 gene was then isolated from pUC18 as a HindIII-SnaBI fragment and cloned into pGFP-N-fus (NIEDENTHAL et al. 1996 ) so that ASF1 was fused at its amino terminus to green fluorescent protein (GFP) sequences and under control of the repressible MET25 promoter. pASB41 (ASF1-[HA]₃) was created as follows. First, PCR mutagenesis was used to replace the stop codon at the 3' end of the ASF1 open reading frame in pLS67 with a NotI site to create pASB31. A fragment encoding a triple tandem repeat of the HA epitope flanked by Not1 restriction sites was inserted into the Not1 site of pASB31 to create pASB35. An XbaI-EcoRI fragment encoding the carboxy terminus of ASF1 in plasmid pASB5 (full-length ASF1 in YCp50) was replaced with an XbaI-EcoRI fragment from pASB35 to create ASF1-[HA]₃ in YCp50. A DNA fragment used to replace the ASF1 open reading frame with the Schizosaccharomyces pombe his5⁺ gene was synthesized by PCR using synthetic oligonucleotides and plasmid pME3 as described by WACH et al. 1997 .

RNA analysis:
Total RNA was isolated from 1–2 x 10⁸ cells as previously described by ARNDT et al. 1987 . Northern analysis was performed as described by FERNANDEZ-SARABIA et al. 1992 . The probes used were a 2.3-kb EcoRI/BamHI fragment from the HTA1 locus that also contains a region of the constitutively transcribed PRT1 gene, a 900-bp EcoRI/HindIII fragment containing HHF2, an 866-bp XhoI/HindIII fragment containing CLN2, a 600-bp EcoRI/HindIII fragment containing ACT1, and a 700-bp AccI/HindIII fragment containing HTB2. The Northerns were processed using high stringency conditions that prevent cross-hybridization of the histone probes. Quantitative S1 analysis was performed using end-labeled probes specific for lacZ and RP51A as described by MORAN et al. 1990 . To create the graphs in Fig 1, Northerns and S1 gels were scanned using a Molecular Dynamics (Sunnyvale, CA) Storm 840 Phosphoimager and Image Quant software. The values shown were obtained by normalizing to the loading controls (RP51A or ACT1). The value for wild type in the absence of hydroxyurea was set at 1.0.

View larger version (43K): In this window In a new window Download PPT slide   Figure 1. asf1 mutants have a Hir- phenotype. RNA was isolated from exponentially growing strains before (-) or after (+) a 45-min incubation in 0.2 M hydroxyurea. RNA levels for the indicated genes were measured either by S1 protection (A) or by Northerns (C and E). The results shown in A, C, and E were quantitated and plotted in B, D, and F after normalizing for loading as described in MATERIALS AND METHODS. The strains used in this analysis all contained pCALA1 (HTA1-lacZ) and were the following: SY561 = WT, SY563 = hir1, SY565 = asf1, SY567 = hir1asf1, SY569 = asf1(YCp50), SY571 = asf1 (pASF1). (G) asf1 mutants arrest normally in S phase in response to hydroxyurea. Exponentially growing cultures of W3031a (WT), SY501 (asf1), and W3031 (hir1) were left untreated (-HU) or were treated for 3 hr with 0.2 M hydroxyurea (+HU). The DNA content of the cells was determined by flow cytometry. Asterisks show position of the peak of 1C and 2C DNA in untreated cells.

Western analysis:
Cell extracts were prepared from 10⁸ cells as previously described by SUTTON et al. 1991 and proteins were separated on a 10% SDS-PAGE gel. Following electrophoresis, proteins were transferred to Immobilon-P membrane (Millipore, Bedford, MA) using a Bio-Rad (Richmond, CA) Trans-Blot SD electrophoretic cell. Antibodies were the anti-HA monoclonal antibody from 12CA5 and the mouse monoclonal against ribosomal protein L3 (gift of J. Warner). Secondary antibody was horseradish peroxidase-conjugated anti-mouse from Amersham (Buckinghamshire, UK). Antibodies were detected using ECL Plus (Amersham) according to manufacturer's instructions.

FACS analysis:
Flow cytometry was performed essentially as described by DAVIS and ENGEBRECHT 1998 . Cells (10⁷) from a mid-log phase culture were pelleted, washed once in H₂O, pelleted, resuspended in cold 70% ethanol, and vortexed. Cells were stored at 4° until processing. Cells were pelleted, washed with 1 ml of 50 mM sodium citrate, resuspended in 500 µl of 50 mM sodium citrate containing 0.1 mg/ml RNase A, and incubated at 37° for 2 hr. Five hundred microliters of 50 mM sodium citrate and 20 µg/ml propidium iodide were added and cells were incubated in the dark at 4° overnight. Immediately prior to sorting, the cells were sonicated briefly. Flow cytometry was performed on a Becton-Dickinson (Franklin Lakes, NJ) FACScan.

Cell synchronization:
Cells were synchronized with -factor (Diagnostic Chemical Ltd., Oxford, CT) at a final concentration of 3.0 or 6.0 µM as described by FERNANDEZ-SARABIA et al. 1992 .

Protein localization:
Cells expressing the GFP protein or GFP-ASF1 were grown to early exponential phase in SC-uracil-methionine with adenine added at a final concentration of 150 µg/ml. 4'-6'-diamidino-2-phenylindole (DAPI) was added to 1 µg/ml and cells were grown for an additional hour. Cells were washed in H₂O and resuspended in mounting medium (0.1% p-phenylenediamine, 0.1 x PBS, 90% glycerol). Cells were examined using a Zeiss axioskop microscope with a x63 objective. A computer-generated merge was used to confirm that the DAPI and GFP-ASF1 signals colocalize (Fig 4; data not shown).

View larger version (77K): In this window In a new window Download PPT slide   Figure 2. Effect of deletion of ASF1 on the periodic transcription of HTA1. Wild-type (WT = SY561), asf1 (SY565), and hir1 (SY563) strains were arrested for 3 hr with 3 µM -factor. Following release from the -factor, aliquots were taken at 15-min intervals and RNA was isolated. Total RNA was examined by Northern analysis using probes specific for the indicated genes.

View larger version (23K): In this window In a new window Download PPT slide   Figure 3. ASF1 protein persists throughout the cell cycle. Strain SY514, which has a chromosomal deletion of ASF1 and an HA epitope-tagged version of ASF1 on a URA3/cen vector, was grown to exponential phase in SC-uracil medium. Cells were arrested for 3 hr with 6 µM -factor. Following release from the -factor arrest, aliquots were taken at 15-min intervals and cell extracts prepared. ASF1-HA protein levels were determined by Western analysis using the 12CA5 antibody against the HA epitope. The filter was stripped and reprobed with an antibody specific for ribosomal protein L3 (RPL3). Asterisks indicate the G1/S transition for each cell cycle as defined by the sample in which 50% of the cells had new buds. UT contains an extract from a sample in which ASF1 is not tagged (SY513).

View larger version (102K): In this window In a new window Download PPT slide   Figure 4. ASF1 is localized to the nucleus. Cells containing GFP (SY505) or GFP-ASF1 (SY504) under control of the methionine-repressible MET25 promoter were grown in the absence of uracil and methionine to early exponential phase. Cells were incubated for an additional 45 min with 1.0 µg/ml DAPI. GFP and DAPI staining were viewed by fluorescence microscopy.

Two-hybrid analysis:
To identify proteins that interact with ASF1, we used the version of the two-hybrid system previously described by HOLLENBERG et al. 1995 and a yeast GAL4 activation domain (GAD) library provided by P. James (JAMES et al. 1996 ). PCR mutagenesis was used to create an EcoRI site at the amino terminus of the ASF1 coding sequences in plasmid ASB8 (ASF1 in pUC118) to create ASB18. An EcoRI-SnaBI fragment containing the entire ASF1 open reading frame was cloned into plasmid pSTT91 [pBTM116 (BARTEL et al. 1993 ) with the ADE2 gene inserted] such that LexA sequences are upstream of and in-frame with ASF1 sequences to create pJBB1. A truncated version, in which only sequences encoding the amino-terminal 143 amino acids of ASF1 (out of 279 total) were fused in-frame with LexA sequences, was created by cloning an EcoRI-StyI fragment from ASB18 into pSTT91 to create pJBB9. The interacting clone obtained from the library encoded an in-frame fusion between GAD and amino acids 384–840 of HIR1. Filter assays for ß-galactosidase for the two-hybrid analysis were as described by BARTEL et al. 1993 except that nitrocellulose filters were used. Quantitative ß-galactosidase assays were done on three independent transformants for each strain, using the permeabilized cell assay (KAISER et al. 1994 ).

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

HIR1 interacts with ASF1 in the yeast two-hybrid system:
To identify proteins that interact with ASF1, we carried out a two-hybrid search. We screened a yeast genomic GAD library using as bait a construct in which the entire ASF1 open reading frame was fused to LexA coding sequences. We screened 10⁶ colonies and identified only one interacting clone specific for the lexA-ASF1 plasmid. When sequenced, this clone was found to encode GAD fused to the carboxy-terminal half of HIR1 (amino acids 384–840; Table 2). A frameshift mutation was created at the junction between GAL4 and HIR1 sequences, and this mutation abolished the two-hybrid interaction with LexA-ASF1. The ASF1 protein contains a highly acidic domain in its carboxy terminus (77% acidic residues within 72 amino acids). We created a fusion between LexA and the amino-terminal half of ASF1, which lacks this acidic region. The truncated LexA-ASF1 protein showed an even stronger interaction with GAD-HIR1 in a two-hybrid assay (Table 2). This shows that the HIR1-ASF1 interaction was not mediated via the acidic stretch of ASF1.

In a separate search, we used the truncated lexA-ASF1 clone as bait and again obtained only one interacting plasmid. This plasmid encoded a fusion of the carboxy-terminal half of SAS4 to GAD. SAS4 (something about silencing) was originally identified because sas4 mutations restore silencing in strains containing a mutated HMR locus (XU et al. 1999 ). Identification of a silencing protein, SAS4, in a two-hybrid search with ASF1, which is also implicated in silencing, was gratifying. The in vivo significance of this interaction is under investigation.

asf1 mutants have a Hir^- phenotype:
In wild-type cells, the four histone gene loci (HTA1-HTB1, HTA2-HTB2, HHT1-HHF1, and HHT2-HHF2) are only transcribed late in G1 and early in S phase. This pattern of transcription results from a combination of activation during early S phase and repression during early G1, G2, and M phases (OSLEY et al. 1986 ; FREEMAN et al. 1992 ). Strains deleted for HIR1 or HIR2 are unable to repress transcription of three of the four histone gene loci (excluding HTA2-HTB2) during the cell cycle (OSLEY and LYCAN 1987 ). Furthermore, when DNA synthesis is blocked by hydroxyurea or by mutation, wild-type cells strongly repress transcription of the histone genes, while hir mutants do not (OSLEY and LYCAN 1987 ). Because of the two-hybrid interaction between ASF1 and HIR1, we tested whether asf1 mutants also have a Hir^- phenotype as defined by failure to repress histone gene transcription following inhibition of DNA replication. We compared the levels of RNA made from an HTA1-lacZ fusion in wild-type and mutant cells grown in the absence or presence of hydroxyurea. The HTA1-lacz fusion is used to identify hir mutants because HTA1 mRNA is also subject to post-transcriptional controls that partially obscure transcriptional effects (LYCAN et al. 1987 ). Exponentially growing cells were treated for 45 min with hydroxyurea, after which RNA was isolated and an S1 nuclease protection assay was performed. As expected, in the wild-type strain HTA1-lacZ mRNA levels were substantially reduced in the hydroxyurea-treated sample compared to untreated cells (Fig 1A and Fig B). This was not the case for the hir1 mutant, which showed elevated RNA levels both in the absence and presence of hydroxyurea. Like the hir1 mutant, the asf1 cells also failed to repress transcription of HTA1-lacZ after incubation with hydroxyurea (Fig 1A and Fig B). This Hir^- phenotype of the asf1 mutant is recessive, because when ASF1 was introduced into the asf1 strain on a low copy number plasmid, HTA1-lacZ RNA levels were repressed in hydroxyurea-treated cells (Fig 1A and Fig B; compare lane 12 with lanes 6 and 10). Thus, asf1 strains have a Hir^- phenotype and fail to repress HTA1 transcription in response to hydroxyurea.

Because HIR1 and HIR2 are required for the cell cycle repression of HHT1-HHF1 and HHT2-HHF2 as well as HTA1-HTB1 (OSLEY and LYCAN 1987 ), we used Northern analysis to determine the effect of mutation of ASF1 on HHF2 transcription. The results (Fig 1C and Fig D) show that, like hir1 mutants, asf1 mutants also fail to repress HHF2 transcription in response to hydroxyurea treatment. HTA2-HTB2 transcription is also under cell cycle control and is repressed in response to hydroxyurea. However, this repression is independent of HIR1 and HIR2 (OSLEY and LYCAN 1987 ; Fig 1E and Fig F). Repression of HTA2-HTB2 is similarly unaffected by mutation of asf1 (Fig 1E and Fig F). This analysis shows that the pattern of transcriptional repression of the histone gene pairs is similar in asf1 and hir1 mutants. Both mutants are defective in repression of transcription of HTA1-HTB1 and HHT2-HHF2, and neither is defective in repression of HTA2-HTB2.

However, Fig 1 also shows that the steady-state levels of mRNA (in the absence of hydroxyurea) are different in the two mutants, with levels of HTA1-HTB1 and HHT2-HHF2 RNA being higher than wild type for the hir1 mutant and lower than wild type for the asf1 strain. The level of HTB2 RNA in the asf1 mutant is also significantly lower than in wild-type and hir1 strains. For all three gene pairs tested, combining the hir1 and asf1 mutations results in a pattern of transcription in the absence of hydroxyurea that is similar to that of the hir1 mutant alone. These results suggest that when HIR1 is present, cells lacking ASF1 are not only defective in repression of histone gene transcription, but may also be somewhat defective in activation.

The defect of the asf1 strain in repression of HTA1-lacZ and HHF2 transcription after treatment with hydroxyurea could result from a requirement for ASF1 in transcriptional repression or from a defect in the response of asf1 mutants to hydroxyurea. To test the latter possibility, we treated wild-type, hir1, and asf1 cells with hydroxyurea for 3 hr. The DNA of these strains was then analyzed by flow cytometry. The results (Fig 1G) show that asf1 and hir1 strains are indistinguishable in their response to hydroxyurea; the mutant strains show a uniform arrest with an intermediate DNA content. The asf1 and hir1 profiles are somewhat shifted compared to that of wild type. It is not known, however, whether these strains undergo slightly more replication before hydroxyurea-induced arrest than wild-type cells. In the absence of hydroxyurea, asf1 mutants show a slight decrease in number of cells in G1 relative to those in G2 or M (Fig 1G and TYLER et al. 1999 ).

To determine whether the defect in the asf1 mutant in repression of HTA1 transcription was specific to hydroxyurea-treated cells, we looked at the pattern of HTA1 RNA accumulation as cells proceeded synchronously through the cell cycle. In this experiment, cells were arrested in G1 with -factor and then released from the block and RNA was isolated from samples obtained during the subsequent cell cycle. HTA1 RNA was detected by Northern analysis and compared to levels of PRT1 RNA that do not fluctuate during the cell cycle and to CLN2 RNA, whose accumulation peaks late in G1. Wild-type cells show a cell cycle-dependent pattern of HTA1 RNA expression with a peak of accumulation at G1/S (Fig 2). In contrast, both the hir1 mutant and the asf1 mutant express HTA1 RNA at all stages of the cell cycle, indicating that the asf1 mutation, like the hir1 mutation, relieves repression in G1, G2, and M (Fig 2).

ASF1 protein persists throughout the cell cycle:
ASF1 is itself periodically transcribed during the cell cycle, with maximal levels at the G1/S transition (LE et al. 1997 ; CHO et al. 1998 ; SPELLMAN et al. 1998 ). If ASF1 protein also only accumulates at G1/S, then whatever the role that ASF1 has in regulation of histone gene transcription, it must perform this function during early S phase. To investigate this, we analyzed the pattern of ASF1 protein accumulation during the cell cycle. We tagged ASF1 at its carboxy terminus with three tandem copies of the HA epitope. This epitope-tagged protein suppresses the growth defect of asf1 strains (data not shown). We transformed a MATa asf1 strain with a low copy number plasmid containing ASF1-HA. Protein extracts were prepared from cells as they proceeded synchronously through the cell cycle following release from an -factor block. The extracts were analyzed for ASF1-HA levels by Western immunoblotting. The results (Fig 3) show that although there is a drop in ASF1 levels following release from the -factor block, ASF1 protein is present throughout the cell cycle. A similar result was obtained when levels of ASF1-HA were examined in cdc15 cells synchronized by release from a temperature block in mitosis (data not shown).

ASF1 protein is located in the nucleus at all stages of the cell cycle:
ASF1's function in control of histone gene transcription might be regulated through cell cycle-specific changes in cellular localization. To test this, ASF1 was tagged at its amino terminus with GFP and its expression put under control of the methionine-repressible MET25 promoter. When examined by fluorescence microscopy, it was clear that most, if not all, of the GFP-ASF1 protein is localized to the nucleus and that this nuclear localization occurs in cells at all stages of the cell cycle (Fig 4). Cells expressing GFP alone, in contrast, have a diffuse staining throughout the cell. Although this GFP-tagged version of ASF1 can complement the growth defect of asf1 strains when cells are grown in the absence of methionine (data not shown), it should be noted that the cells are larger than those with wild-type ASF1 (Fig 4).

ASF1 and HIR1 show genetic interactions with CAC2:
hir1 mutations show genetic interactions with mutations in CAC1, CAC2, and CAC3, genes encoding subunits of the yeast chromatin assembly factor CAF-I. The double mutants have a growth defect and more severe silencing defects than seen in the single mutants (KAUFMAN et al. 1998 ; QIAN et al. 1998 ). One model proposed for these genetic interactions is that HIR1 is part of a second chromatin assembly complex that functions when CAF-I is absent (KAUFMAN et al. 1998 ). Alternatively, cac mutants may be sensitive to perturbations in histone stoichiometry that result from mutations in HIR genes (KAUFMAN et al. 1998 ). The two-hybrid interaction we found between ASF1 and HIR1, as well as the similar effects of hir1 and asf1 mutations on repression of histone gene transcription, suggest that ASF1 and HIR1 function together in a common pathway. If ASF1 and HIR1 function in the same pathway, certain predictions about genetic interactions can be made. First, asf1 mutations should have genetic interactions with cac mutations, just as hir1 mutations do. Second, strains with mutations in both ASF1 and HIR1 should not have phenotypes more severe than those with a single mutation. Third, strains with mutations in hir1, asf1, and cac1 should have no more severe defects than strains with mutations in cac2 and either asf1 or hir1 alone. These predictions were confirmed by the data in Fig 5. Fig 5A shows growth of strains with combinations of deletions of ASF1, CAC2, and HIR1. asf1 strains have a growth defect, while cac2 and hir1 strains do not. It can also be seen [and has previously been reported in TYLER et al. 1999 ] that the asf1 cac2 strain grows more slowly than the asf1 strain, just as the hir1 cac2 strain grows more slowly than the hir1 or cac2 strain. Furthermore, the asf1 hir1 strain does not grow more slowly than the asf1 strain. Finally, the asf1 hir1 cac2 strain does not grow more slowly than the asf1 cac2 strain (Fig 5A).

View larger version (53K): In this window In a new window Download PPT slide   Figure 5. Effects of combinations of asf1, cac2, and hir1 mutations on cell growth and telomeric silencing. (A) Cell growth. Cells were streaked onto YPD plates and incubated at 30° for 2 days. Strains used were the following: W3031a, CDY1-3 (asf1), W3031 (hir1), YB0151 (cac2), SY577 (asf1 hir1), SY578 (asf1 hir1 cac2), SY579 (asf1 cac2), and SY580 (hir1 cac2). (B) Telomeric silencing. Cells were grown in YPD to late log phase at 30°. Tenfold serial dilutions were spotted onto SC or 5-FOA plates and incubated at 30° for 2 days and 3 days, respectively. Strains contained a URA3 gene near the telomere of chromosome VII and were the following: YDS631 (WT), SY546 (asf1), SY603 (hir1), SY604 (cac2), SY605 (asf1 cac2), SY606 (asf1 hir1), SY607 (hir1 cac2), SY608 (asf1 hir1 cac2).

We further investigated genetic interactions among ASF1, HIR1, and CAC2 by testing the effects of deletions of these genes on telomeric silencing. To do this, we created a series of strains with various combinations of mutations, all containing a URA3 reporter gene inserted near the telomere of chromosome VII. The fraction of cells that form colonies on plates containing fluoroorotic acid (FOA) reflects the level of silencing of the URA3 reporter at the telomere; strains with defective silencing grow poorly on FOA, while those with intact silencing grow well. The results in Fig 5B show that neither asf1 nor hir1 strains have substantial defects in silencing; they appear with a frequency similar to wild type on FOA (the asf1 colonies are small, because asf1 strains have a growth defect, but the number of colonies on the FOA plates are similar in both cases). In contrast, and as previously shown (KAUFMAN et al. 1997 ), cac2 mutants have a severe silencing defect. This defect is exacerbated when cac2 mutations are combined with either hir1 mutations (Fig 5B and KAUFMAN et al. 1998 ) or asf1 mutations (Fig 5B and TYLER et al. 1999 ). However, combining the hir1 deletion with the asf1 deletion does not cause a silencing defect (Fig 5B). Taken together, the growth and silencing data suggest that ASF1 and HIR1 are in the same pathway affecting chromatin structure or assembly and that CAF-I functions in a separate pathway.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

HIR1 is required for cell cycle repression of transcription of three of the four histone gene pairs. Several lines of evidence suggest that HIR1 interacts with ASF1 to bring about this repression. First, using the yeast two-hybrid system, we showed that HIR1 and ASF1 interact in yeast. ASF1 also interacts with HIR1 and HIR2 in vitro (P. D. KAUFMAN, personal communication). Second, the pattern of repression of histone gene transcription is similar in asf1 mutants and hir1 mutants. For both strains, repression of HTA1-HTB1 and HHT2-HHF2 transcription is defective both after hydroxyurea treatment and during a synchronized cell cycle, and repression of HTA2-HTB2 is not affected. Third, combining a hir1 mutation or an asf1 mutation with a cac2 mutation causes similar synergistic reductions in growth rate and telomeric silencing. However, no synergistic effects were seen when the asf1 and the hir1 mutations were combined. hir1 and asf1 also show similar interactions with mutations in SPT10. SPT10 encodes a putative acetyltransferase that is required for transcription of HTA2-HTB2 (DOLLARD et al. 1994 ). Mutations in SPT10 cause lethality when combined with mutations in HIR1 (D. HESS and F. WINSTON, personal communication). We have found that a mutation in SPT10 also causes lethality when combined with deletion of ASF1 (data not shown).

The mechanism of repression of histone gene transcription by HIR1/ASF1 is not yet clear. There is no evidence that either HIR1 or ASF1 binds directly to the cis-acting site required for repression of HTA1 (data not shown). However, chromatin immunoprecipitation shows that HIR1 and HIR2 associate with the HTA1-HTB1 regulatory region (M. A. OSLEY, unpublished result). HIR2 (and probably HIR1, which interacts strongly with HIR2) binds to histones H3, H4, and H2B in vitro (J. RECHT and M. A. OSLEY, unpublished results) as does the vertebrate equivalent, HIRA (LORAIN et al. 1998 ). Drosophila and human ASF1 bind to histones H3 and H4 (TYLER et al. 1999 ; MUNAKATA et al. 2000 ). These results suggest that ASF1 and HIR1 may both function at the histone genes to create the proper chromatin structure for repression. This repressive chromatin might be achieved via histone deposition, as suggested by in vitro analysis of Drosophila ASF1 (TYLER et al. 1999 ) or by recruitment of chromatin remodeling complexes (TYLER and KADONAGA 1999 ). In the absence of ASF1 or HIR1, the chromatin structure of the HTA1 regulatory locus cannot prevent access to the transcriptional machinery during early G1, G2, and M and the gene is constitutively expressed.

Transcription of the histone loci is regulated not only by repression, but also by activation in late G1/early S phase (OSLEY et al. 1986 ). In our analysis, the hir1 mutant had levels of HTA1 and HHF2 RNA that were higher than those of wild-type cells, due to the loss of cell cycle repression coupled with normal activation (Fig 1). In contrast, although asf1 mutants also showed no cell cycle repression, the steady state levels of HTA1 and HHF2 RNA were not elevated relative to wild type (Fig 1). The simplest explanation for this is that asf1 mutants are defective not only in repression, but also in activation of these histone genes. It has previously been shown that maximal transcription of HTA1-HTB1 requires the SWI/SNF chromatin remodeling complex (DIMOVA et al. 1999 ). Perhaps asf1 mutants are defective in the recruitment of this complex to the histone genes, either because of a direct role of ASF1 in its recruitment, or because the chromatin structure of asf1 strains is not permissive for the SWI/SNF interaction. SWI/SNF is not required for maximal activation of histone gene transcription when HIR1 is absent (DIMOVA et al. 1999 ). This could explain why the asf1 hir1 double mutant shows elevated levels of histone gene transcription similar to the hir1 mutant. Although repression of HTA2-HTB2 transcription was not impaired in asf1 mutants, transcript levels were lower than in wild-type cells. Perhaps ASF1 is also required for maximal transcription of this locus, either by recruiting SWI/SNF, or SPT10, and/or SPT21, which are necessary for activation of HTA2-HTB2 transcription (DOLLARD et al. 1994 ).

asf1 mutants have phenotypes in addition to those seen for hir1 mutants. These include a reduced growth rate, increased sensitivity to DNA-damaging agents, and enrichment of cells in G2 and M (LE et al. 1997 ; TYLER et al. 1999 ). The phenotypic differences between asf1 and hir1 strains most likely result from an involvement of ASF1 in formation or modulation of chromatin structure at loci that are not regulated by HIR1.

While ASF1 RNA is maximally transcribed during the G1/S transition, the protein persists throughout the cell cycle and is even present in cells that have been arrested in G1 with -factor for 3 hr. We cannot rule out the possibility that the HA-epitope tag used in our analysis affects the synthesis or stability of the ASF1 protein. However, it is not uncommon for proteins to persist throughout the cell cycle, even when their genes are periodically transcribed. Examples include MCM2 and MCM3 (YOUNG and TYE 1997 ), which are transcribed in M, and RFA2, which, like ASF1, contains MCB regulatory sequences and is transcribed at G1/S (DIN et al. 1990 : BRILL and STILLMAN 1991 ). ASF1 is also localized to the nucleus at all stages of the cell cycle. This argues that regulation of ASF1 for the cell cycle control of histone gene expression must be at a post-translational level. Perhaps, as has been proposed for HIR proteins (SPECTOR et al. 1997 ), ASF1 is modified by a cell-cycle specific signal so that its activity is confined to discrete stages.

CAF-I, HIR1, and ASF1 have all been suggested to function in histone deposition (SMITH and STILLMAN 1989 ; KAMAKAKA et al. 1996 ; KAUFMAN et al. 1998 ; TYLER et al. 1999 ; MUNAKATA et al. 2000 ). If true, then we would propose that ASF1 and HIR1 function together to form repressive chromatin at the histone genes. CAF-I would then be in a separate pathway required for formation of silenced chromatin at telomeres and HM loci. The effect of asf1 and hir1 mutations on telomere silencing in cac2 mutants might result from a direct role of ASF1 and HIR1 in histone deposition near telomeres. Alternatively, the synergistic loss of silencing in asf1 cac2 and hir1 cac2 strains may be an indirect consequence of the histone imbalances in these mutant strains. It has previously been shown that telomeric silencing in cac mutants is very sensitive to changes in histone levels (KAUFMAN et al. 1998 ). CAF-I probably does not function in histone deposition at the histone genes, because cac mutants are not defective in histone gene expression. If all three proteins are involved in histone deposition, there must be additional factors that have this role in the cell. While the asf1 cac2 hir1 mutant is slow growing, it is alive, while mutations that completely abolish histone deposition would most likely cause lethality. Alternatively, even though ASF1 can assist in histone deposition in vitro, this may not be its (only) role in vivo. For example, it may function to both positively and negatively regulate transcription of histone and other genes by chromatin remodeling. ASF1 may do this directly, as part of a chromatin remodeling complex, or it may recruit other factors. Further identification of proteins that interact with ASF1 may address this.

	FOOTNOTES

¹ Present address: Department of Molecular Genetics and Microbiology, University of New Mexico Health Sciences Center, Albuquerque, NM 87131.

	ACKNOWLEDGMENTS

We thank R. Heller for excellent technical assistance and P. Kaufman and F. Winston for communicating unpublished results. This work was supported by National Institutes of Health grant GM-28220 to R.S. and GM-40118 to M.A.O.

Manuscript received January 3, 2001; Accepted for publication March 5, 2001.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

ARNDT, K. T., C. STYLES, and G. R. FINK, 1987  Multiple global regulators control HIS4 transcription in yeast. Science 237:874-880[Abstract/Free Full Text].

BARTEL, P. L., C. CHIEN, R. STERNGLANZ and S. FIELDS, 1993 Using the two-hybrid system to detect protein-protein interactions, pp. 153–179 in Cellular Interactions in Development: A Practical Approach, edited by D. A. HARTLEY. IRL Press, Oxford.

BRILL, S. J. and B. STILLMAN, 1991  Replication factor-A from Saccharomyces cerevisiae is encoded by three essential genes coordinately expressed at S phase. Genes Dev. 5:1589-1600[Abstract/Free Full Text].

CHIEN, C. T., S. BUCK, R. STERNGLANZ, and D. SHORE, 1993  Targeting of SIR1 protein establishes transcriptional silencing at HM loci and telomeres in yeast. Cell 75:531-541[Medline].

CHO, R. J., M. J. CAMPBELL, E. A. WINZELER, L. STEINMETZ, and A. CONWAY et al., 1998  A genome-wide transcriptional analysis of the mitotic cell cycle. Mol. Cell 2:65-73[Medline].

CLARK-ADAMS, C. D., D. NORRIS, M. A. OSLEY, J. S. FASSLER, and F. WINSTON, 1988  Changes in histone gene dosage alter transcription in yeast. Genes Dev. 2:150-159[Abstract/Free Full Text].

DAVIS, L. and J. ENGEBRECHT, 1998  Yeast dom34 mutants are defective in multiple developmental pathways and exhibit decreased levels of polyribosomes. Genetics 149:45-56[Abstract/Free Full Text].

DIMOVA, D., Z. NACKERDIEN, S. FURGESON, S. EGUCHI, and M. A. OSLEY, 1999  A role for transcriptional repressors in targeting the yeast Swi/Snf complex. Mol. Cell 4:75-83[Medline].

DIN, S., S. J. BRILL, M. P. FAIRMAN, and B. STILLMAN, 1990  Cell-cycle-regulated phosphorylation of DNA replication factor A from human and yeast cells. Genes Dev. 4:968-977[Abstract/Free Full Text].

DOLLARD, C., S. L. RICUPERO-HOVASSE, G. NATSOULIS, J. D. BOEKE, and F. WINSTON, 1994  SPT10 and SPT21 are required for transcription of particular histone genes in Saccharomyces cerevisiae.. Mol. Cell. Biol. 14:5223-5228[Abstract/Free Full Text].

FERNANDEZ-SARABIA, M. J., A. SUTTON, T. ZHONG, and K. T. ARNDT, 1992  SIT4 protein phosphatase is required for the normal accumulation of SWI4, CLN1, CLN2, and HCS26 RNAs during late G1. Genes Dev. 6:2417-2428[Abstract/Free Full Text].

FREEMAN, K. B., L. R. KARNS, K. A. LUTZ, and M. M. SMITH, 1992  Histone H3 transcription in Saccharomyces cerevisiae is controlled by multiple cell cycle activation sites and a constitutive negative regulatory element. Mol. Cell. Biol. 12:5455-5463[Abstract/Free Full Text].

HAN, M. and M. GRUNSTEIN, 1988  Nucleosome loss activates yeast downstream promoters in vivo. Cell 55:1137-1145[Medline].

HOLLENBERG, S. M., R. STERNGLANZ, P. F. CHENG, and H. WEINTRAUB, 1995  Identification of a new family of tissue-specific basic helix-loop- helix proteins with a two-hybrid system. Mol. Cell. Biol. 15:3813-3822[Abstract].

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].

KADONAGA, J. T., 1998  Eukaryotic transcription: an interlaced network of transcription factors and chromatin-modifying machines. Cell 92:307-313[Medline].

KAISER, C., S. MICHAELIS and A. MITCHELL, 1994 Methods in Yeast Genetics. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

KAMAKAKA, R. T., M. BULGER, P. D. KAUFMAN, B. STILLMAN, and J. T. KADONAGA, 1996  Postreplicative chromatin assembly by Drosophila and human chromatin assembly factor 1. Mol. Cell. Biol. 16:810-817[Abstract].

KAUFMAN, P. D., R. KOBAYASHI, and B. STILLMAN, 1997  Ultraviolet radiation sensitivity and reduction of telomeric silencing in Saccharomyces cerevisiae cells lacking chromatin assembly factor-I. Genes Dev. 11:345-357[Abstract/Free Full Text].

KAUFMAN, P. D., J. L. COHEN, and M. A. OSLEY, 1998  Hir proteins are required for position-dependent gene silencing in Saccharomyces cerevisiae in the absence of chromatin assembly factor I. Mol. Cell. Biol. 18:4793-4806[Abstract/Free Full Text].

LE, S., C. DAVIS, J. B. KONOPKA, and R. STERNGLANZ, 1997  Two new S-phase-specific genes from Saccharomyces cerevisiae.. Yeast 13:1029-1042[Medline].

LORAIN, S., J. P. QUIVY, F. MONIER-GAVELLE, C. SCAMPS, and Y. LECLUSE et al., 1998  Core histones and HIRIP3, a novel histone-binding protein, directly interact with WD repeat protein HIRA. Mol. Cell. Biol. 18:5546-5556[Abstract/Free Full Text].

LYCAN, D. E., M. A. OSLEY, and L. M. HEREFORD, 1987  Role of transcriptional and posttranscriptional regulation in expression of histone genes in Saccharomyces cerevisiae.. Mol. Cell. Biol. 7:614-621[Abstract/Free Full Text].

MEEKS-WAGNER, D. and L. H. HARTWELL, 1986  Normal stoichiometry of histone dimer sets is necessary for high fidelity of mitotic chromosome transmission. Cell 44:43-52[Medline].

MORAN, L., D. NORRIS, and M. A. OSLEY, 1990  A yeast H2A H2B promoter can be regulated by changes in histone gene copy number. Genes Dev. 4:752-763[Abstract/Free Full Text].

MUNAKATA, T., N. ADACHI, N. YOKOYAMA, T. KUZUHARA, and M. HORIKOSHI, 2000  A human homologue of yeast anti-silencing factor has histone chaperone activity. Genes Cells 5:221-233[Abstract].

NIEDENTHAL, R. K., L. RILES, M. JOHNSTON, and J. H. HEGEMANN, 1996  Green fluorescent protein as a marker for gene expression and subcellular localization in budding yeast. Yeast 12:773-786[Medline].

NORRIS, D. and M. A. OSLEY, 1987  The two gene pairs encoding H2A and H2B play different roles in the Saccharomyces cerevisiae life cycle. Mol. Cell. Biol. 7:3473-3481[Abstract/Free Full Text].

OSLEY, M. A. and D. LYCAN, 1987  Trans-acting regulatory mutations that alter transcription of Saccharomyces cerevisiae histone genes. Mol. Cell. Biol. 7:4204-4210[Abstract/Free Full Text].

OSLEY, M. A., J. GOULD, S. KIM, M. Y. KANE, and L. HEREFORD, 1986  Identification of sequences in a yeast histone promoter involved in periodic transcription. Cell 45:537-544[Medline].

QIAN, Z., H. HUANG, J. Y. HONG, C. L. BURCK, and S. D. JOHNSTON et al., 1998  Yeast Ty1 retrotransposition is stimulated by a synergistic interaction between mutations in chromatin assembly factor I and histone regulatory proteins. Mol. Cell. Biol. 18:4783-4792[Abstract/Free Full Text].

SHERWOOD, P. W., S. V. TSANG, and M. A. OSLEY, 1993  Characterization of HIR1 and HIR2, two genes required for regulation of histone gene transcription in Saccharomyces cerevisiae.. Mol. Cell. Biol. 13:28-38[Abstract/Free Full Text].

SINGER, M. S., A. KAHANA, A. J. WOLF, L. L. MEISINGER, and S. E. PETERSON et al., 1998  Identification of high-copy disruptors of telomeric silencing in Saccharomyces cerevisiae.. Genetics 150:613-632[Abstract/Free Full Text].

SMITH, S. and B. STILLMAN, 1989  Purification and characterization of CAF-I, a human cell factor required for chromatin assembly during DNA replication in vitro. Cell 58:15-25[Medline].

SPECTOR, M. S., A. RAFF, H. DESILVA, K. LEE, and M. A. OSLEY, 1997  Hir1p and Hir2p function as transcriptional corepressors to regulate histone gene transcription in the Saccharomyces cerevisiae cell cycle. Mol. Cell. Biol. 17:545-552[Abstract].

SPELLMAN, P. T., G. SHERLOCK, M. Q. ZHANG, V. R. IYER, and K. ANDERS et al., 1998  Comprehensive identification of cell cycle-regulated genes of the yeast Saccharomyces cerevisiae by microarray hybridization. Mol. Biol. Cell 9:3273-3297[Abstract/Free Full Text].

SUTTON, A., D. IMMANUEL, and K. T. ARNDT, 1991  The SIT4 protein phosphatase functions in late G1 for progression into S phase. Mol. Cell. Biol. 11:2133-2148[Abstract/Free Full Text].

TYLER, J. K. and J. T. KADONAGA, 1999  The "dark side" of chromatin remodeling: repressive effects on transcription. Cell 99:443-446[Medline].

TYLER, J. K., C. R. ADAMS, S. R. CHEN, R. KOBAYASHI, and R. T. KAMAKAKA et al., 1999  The RCAF complex mediates chromatin assembly during DNA replication and repair. Nature 402:555-560[Medline].

WACH, A., A. BRACHAT, C. ALBERTI-SEGUI, C. REBISCHUNG, and P. PHILIPPSEN, 1997  Heterologous HIS3 marker and GFP reporter modules for PCR-targeting in Saccharomyces cerevisiae.. Yeast 13:1065-1075[Medline].

XU, E. Y., S. KIM, K. REPLOGLE, J. RINE, and D. H. RIVIER, 1999  Identification of SAS4 and SAS5, two genes that regulate silencing in Saccharomyces cerevisiae.. Genetics 153:13-23[Abstract/Free Full Text].

YOUNG, M. R. and B. K. TYE, 1997  Mcm2 and Mcm3 are constitutive nuclear proteins that exhibit distinct isoforms and bind chromatin during specific cell cycle stages of Saccharomyces cerevisiae.. Mol. Biol. Cell 8:1587-1601[Abstract].