Identification of SAS4 and SAS5, Two Genes That Regulate Silencing in Saccharomyces cerevisiae

Corresponding author: David H. Rivier, Department of Cell and Structural Biology, University of Illinois, 601 S. Goodwin Ave., Urbana, IL 61801., rivier{at}uiuc.edu (E-mail)

Communicating editor: F. WINSTON

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

In Saccharomyces cerevisiae, chromatin-mediated silencing inactivates transcription of the genes at the HML and HMR cryptic mating-type loci and genes near telomeres. Mutations in the Rap1p and Abf1p binding sites of the HMR-E silencer (HMRa-e**) result in a loss of silencing at HMR. We characterized a collection of 15 mutations that restore the -mating phenotype to MAT HMRa-e** strains. These mutations defined three complementation groups, two new groups and one group that corresponded to the previously identified SAS2 gene. We cloned the genes that complemented members of the new groups and identified two previously uncharacterized genes, which we named SAS4 and SAS5. Neither SAS4 nor SAS5 was required for viability. Null alleles of SAS4 and SAS5 restored SIR4-dependent silencing at HMR, establishing that each is a regulator of silencing. Null alleles of SAS4 and SAS5 bypassed the role of the Abf1p binding site of the HMR-E silencer but not the role of the ACS or Rap1p binding site. Previous analysis indicated that SAS2 is homologous to a human gene that is a site of recurring translocations involved in acute myeloid leukemia. Similarly, SAS5 is a member of a gene family that included two human genes that are the sites of recurring translocations involved in acute myeloid leukemia.

TRANSCRIPTION is regulated by factors that act locally at promoters and enhancers, as well as by factors that influence the chromatin structure of genes. There are now five well-described ATP-dependent chromatin remodeling complexes, SWI/SNF, RSC, NURF, CHRAC, and AFC, that use the energy of ATP hydrolysis to alter the relationship between DNA and core histone proteins and activate (CAIRNS 1998 ; VARGA-WEISZ and BECKER 1998 ) or repress (HOLSTEGE et al. 1998 ; SCHNITZLER et al. 1998 ) transcription. Histone acetylation also influences chromatin structure and plays a role in the activation of transcription. Several proteins that possess histone acetylase activity activate transcription of specific genes, whereas others are components of the general transcription machinery (BANNISTER and KOUZARIDES 1996 ; BROWNELL and ALLIS 1996 ; BROWNELL et al. 1996 ; MIZZEN et al. 1996 ; KUO et al. 1998 ).

Reciprocal translocations that form in-frame gene fusions are common in human leukemias (CLEARY 1991 ; RABBITTS 1994 ). The translocation partners identified to date encode proteins that are known or suspected to activate transcription, including DNA binding proteins that are transcription activators and histone acetylases (CLEARY 1991 , CLEARY 1992 ; RABBITTS 1994 ; ROWLEY et al. 1997 ; WARING and CLEARY 1997 ; CARAPETI et al. 1998 ). Thus, altered patterns of transcription, possibly as a result of altered chromatin structures at promoters, have been implicated in the etiology of human leukemia.

At the other end of the spectrum, there are a growing number of proteins that block the expression of genes by causing the formation of an inactive chromatin structure that contains those genes. Well-characterized examples include the proteins that mediate heterochromatin formation and cause the classically defined position effects on gene expression. In Saccharomyces, heterochromatin formation is responsible for silencing the mating-type genes at HML and HMR and for silencing reporter genes inserted near telomeres (GRUNSTEIN 1998 ; LUSTIG 1998 ). A related form of silencing inactivates reporter genes inserted into the rDNA (BRYK et al. 1997 ; SMITH and BOEKE 1997 ). Silencing at HML, HMR, and at the telomeres involves the assembly of a repressive chromatin structure that contains the Sir2, Sir3, and Sir4 proteins that bind to each other and to core histones (MORETTI et al. 1994 ; HECHT et al. 1995 , HECHT et al. 1996 ; STRAHL-BOLSINGER et al. 1997 ). In addition, silenced chromatin contains hypoacetylated nucleosomes, implicating a role for acetylation and deacetylation in regulating regional effects on gene expression (BRAUNSTEIN et al. 1996 ; RUNDLETT et al. 1998 ). An understanding of how particular patterns of histone acetylation are established and how repressive chromatin structures are assembled in particular chromosomal regions is lacking.

The work presented here extends our dissection of silencing in Saccharomyces. Silencing of HML and HMR is mediated by flanking regulatory sites known as silencers (ABRAHAM et al. 1984 ; FELDMAN et al. 1984 ; BRAND et al. 1985 ). The best-characterized silencer, HMR-E, consists of binding sites for three proteins: the two transcription factors, Abf1p and Rap1p, and ORC, the replication initiator protein. The combination of these proteins is thought to recruit the SIR proteins to the silencer (CHIEN et al. 1993 ; MORETTI et al. 1994 ; FOX et al. 1997 ). Mutations in any one of the HMR-E binding sites, in an otherwise wild-type cell, have little effect on silencing (BRAND et al. 1987 ; KIMMERLY et al. 1988 ). In contrast, HMRa-e**, a double mutant silencer with lesions in both the Rap1p and Abf1p binding sites, has a substantial defect in HMR silencing (KIMMERLY et al. 1988 ). Previous work demonstrated that mutations in some genes, such as SAS2, can suppress the silencing defect caused by this mutant silencer (AXELROD and RINE 1991 ; REIFSNYDER et al. 1996 ; EHRENHOFER-MURRAY et al. 1997 ). SAS2 encodes an acetylase homolog, suggesting that analysis of SAS2 and functionally related genes may provide a genetic entrée to genes that regulate the modification of histones and the assembly/disassembly of particular chromatin structures (REIFSNYDER et al. 1996 ; EHRENHOFER-MURRAY et al. 1997 ). In addition, SAS2 is highly similar to MOZ, a human gene involved in acute myeloid leukemia (REIFSNYDER et al. 1996 ).

In this study, we present the analysis of additional mutations that restore silencing of HMR flanked by the HMRa-e** silencer. This work identified two new genes, SAS4 and SAS5, and established that these genes were regulators of silencing. SAS4 lacked any recognizable homolog. SAS5 had similarity to ANC1, a yeast gene implicated in transcriptional activation and chromatin remodeling, and to AF-9 and ENL, two human genes that are the sites of recurring translocations that contribute to leukemia.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Dominance tests:
The original sas mutant strains (DRY22-29 and DRY31-DRY42) were of the -mating type and contained the HMRa-e** allele. Two tests were used to determine which mutants contained dominant mutations and which contained recessive mutations. In the first test, the sas mutants were mated to a matap HMRa-e** strain (DRY1351) in which the promoter region of the MATa1 gene is deleted, and the mating phenotype of the diploid was tested. Among the 20 mutants tested, 2 were dominant mutants that could suppress the silencing defect of both HMRa-e** alleles and had the -mating phenotype. The remaining 18 mutants produced diploids that were nonmating, and thus were either recessive, cis-dominant, or weakly dominant. In a second test, the remaining 18 mutants were mated to a matap strain (DRY1352) that contained a null allele of HMR in which the entire locus was replaced with the URA3 gene (matap hmr::URA3). Of the 18 diploids, 15 were unable to mate and hence were judged to be recessive.

Complementation analysis:
As described in the text, complementation analysis was performed by crossing each of the original sas mutants to matap HMRa-e** strains harboring a deletion of SAS2 (DRY1356), SAS4 (DRY1354), or SAS5 (DRY1358). The resulting matap/MAT HMRa-e**/HMRa-e** sas/sas- diploids were tested for the -mating phenotype. Diploids with the -mating phenotype indicated that the original mutation being tested did not complement the SAS null allele, whereas diploids with the nonmating phenotype indicated that the original mutation being tested complemented the SAS null allele. By this criterion, the original mutant strains that comprised the SAS2 complementation group were DRY23, DRY26, DRY28, DRY29, DRY30, DRY33, DRY34, DRY35, and DRY41 (Table 1). The mutant strains that comprised the SAS4 complementation group were DRY25, DRY27, DRY32, and DRY38. The mutant strains that comprised the SAS5 complementation group were DRY24 and DRY40.

Allelism tests:
Allelism between the sas4::kanMX allele and the original sas4-1 allele was tested in 29 tetrads from a cross between a MATa HMRa-e** sas4::kanMX strain (DRY1360) and the original MAT HMRa-e** sas4-1 mutant strain (DRY24). Each of the tetrads from this cross contained two segregants with the -mating phenotype, indicating that sas4-1 and the sas4::kanMX mutations were allelic. Similarly, MATa/MAT diploids homozygous for HMRa-e** and sas5-1/sas5::HIS3 (derived from DRY24 crossed to DRY1391) segregated two -mating competent and two a-mating segregants in each of 39 tetrads. Thus sas5::HIS3 and sas5-1 were allelic.

Cloning of SAS4 and SAS5 genes:
A yeast genomic library in a LEU2-CEN vector was transformed into DRY601 (sas4-1) or DRY342 (sas5-1) (SPENCER et al. 1990 ). Approximately 3000 colonies were screened for the nonmating phenotype. Plasmids were isolated from nonmating colonies and retransformed into DRY601 or DRY342. Plasmids that conferred the nonmating phenotype upon retransformation were mapped and partially sequenced. The partial sequence was used to identify the complete sequence from GenBank.

Disruption of SAS4 and SAS5:
The entire coding regions of the SAS4 and SAS5 genes were deleted by PCR-mediated gene disruption (BAUDIN et al. 1993 ). Disruption of SAS4 was as follows. The kanMX4 gene of plasmid pDR760 was amplified by PCR using the 5'-ccgaaaatttctacagcattaaaagcatatgagagttcatcacgttgtaaaacgacggcc-3' and 5'-atattgaatttcatttacaccatcgcattatattagttcaaaggctcgtatgttgtgtgg-3' primers. pDR760 contains an EcoRI-BamHI kanMX4 fragment from pFA6 (WACH et al. 1994 ) inserted into EcoRI-BamHI-cleaved pUC18. sas4::kanMX strains were constructed by transformation of the PCR products and confirmed by DNA blot analysis. Disruption of SAS5 was as follows: the HIS3 gene of plasmid pJJ217 (JONES and PRAKASH 1990 ) was amplified by PCR using the 5'-tctatgttttcaggcattgtttaatttcatgatggctgtccggcctcctctagtacactc-3' and 5'-ccttttttttttttttggtgccatataatagacgctcttttgcgcgcctcgttcagaatg-3' primers. sas5::HIS3 strains were constructed by transformation of the PCR products and confirmed by DNA blot analysis.

PCR protocol:
PCR reactions for gene disruption were carried out using the high-fidelity Elongase kit (GIBCO, Grand Island, NY) under the conditions recommended by the manufacturer.

Yeast strain construction:
Two isogenic sets of strains were used in this work. The first was derived from JRY2069, the second from W303-1a. SAS4 and SAS5 were disrupted in JRY2069 to generate DRY1373 and DRY1374, respectively. The W303 derivatives containing disruptions of SAS4 or SAS5 were generated as follows. SAS4 was disrupted in DRY439 to generate DRY1364, in CAF23 to generate DRY1370, in CAF68 to generate DRY1366, in CAF176 to generate DRY1369, in CAF179 to generate DRY1365, and in CAF396 to generate DRY1368. One copy of SAS4 was disrupted in the diploid strain DRY1338 to generate DRY1361 (MATa/MAT HMRa-e**/HMR-ssabf1-::ADE2 ade2::HIS3/ade2::LEU2). DRY1322 and DRY1360 were segregants derived from DRY1361. The HMR-ssabf1-::ADE2 allele of DRY1361 gives rise to a pink colony color in an otherwise ade2- background, allowing the alleles of HMR to be unambiguously assigned in segregants of DRY1361. SAS5 was disrupted in JRY5273 to generate DRY1314, in DRY439 to generate DRY2109, in CAF23 to generate DRY2112, in CAF68 to generate DRY2111, in CAF176 to generate DRY2114, in CAF179 to generate DRY2110, and in CAF396 to generate DRY2113.

All other strains isogenic with W303 were derived by cross. DRY1351, DRY1352, and DRY1354 were segregants from a cross between DRY1322 and JRY4186, a matap hmr::URA3 derivative of W303-1a described previously (LOO et al. 1995 ). DRY1358 was a segregant from a cross between JRY4186 and DRY1314. DRY1356 was a segregant from a cross between JRY4186 and JRY5274, a MAT HMRa-e** sas2-1 derivative of W303-1a described previously (EHRENHOFER-MURRAY et al. 1997 ). DRY1391 is a MATa HMRa-e** sas5::HIS3 strain derived from a cross between DRY1314 and DRY1803 (MATa HMR-ssabf1-::ADE2 ade2::LEU2). As described above, HMR-ssabf1::ADE2 allows assignment of the alleles of HMR in the segregants.

sir4::LEU2 sas strains were generated by cross to a MAT sir4::LEU2 HMR-SS::ADE2 strain. Since this strain lacks the HMRa genes it has the -mating phenotype. In addition, the presence of the ADE2 gene at HMR allows unambiguous assignment of HMR alleles in segregants. DRY1397 was a segregant from a cross between DRY1360 and DRY1804 (W303-1a; MAT sir4::LEU2 HMR-SS::ADE2 lys2). Similarly, DRY1398 was a segregant from a cross between DRY1391 and DRY1804. The CAF strains were provided by C. Fox.

DRY601 and DRY342, the strains used to clone SAS4 and SAS5, were segregants derived from crosses between YAA87 (mata1 HMRa-e** ade2-101^oc leu2-3,112 ura3-52; AXELROD and RINE 1991 ) and DRY25 (sas4-1) or DRY24 (sas5-1), respectively.

Quantitative and patch mating assays:
Quantitative matings were performed essentially as described previously (EHRENHOFER-MURRAY et al. 1997 ). Cells were grown to an OD₆₀₀ of 0.5–1.0 in rich medium supplemented with adenine. Serial dilutions of test strains were mixed with 1.2 x 10⁷ cells of a MATa lawn (JRY2726) or a MAT lawn (JRY2728) and plated onto YM medium supplemented with adenine. Equivalent dilutions of test strains were plated onto solid rich medium to determine the number of viable cells/dilution. Mating efficiencies were calculated as the number of diploids formed per viable cell plated and were normalized to the efficiency of an isogenic wild-type strain. Values reported are the average of two to eight independent trials performed with at least two independent isolates of each strain tested.

Sequence comparison:
Proteins with similarity to Sas5p were identified using the tblastn program against the nonredundant sequences in GenBank. Alignment of the proteins with similarity to Sas5p was carried out using Blockmaker, ClustalW, and Multishade. Alignment and comparison were carried out using the resources provided at the NCSA Biology Workbench (http://biology.ncsa.uiuc.edu) using default parameters.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

sas mutations define three complementation groups:
To identify genes that regulate position effect silencing in yeast, we analyzed mutations that potentially restored silencing at an HMR locus flanked by an HMR-E silencer containing mutations in two domains. This mutant silencer is known as HMRa-e**, with the lowercase e designating a loss of function and the two asterisks indicating the mutations in the Rap1p and Abf1p binding sites. MAT strains containing the HMRa-e** allele display the nonmating phenotype characteristic of a/ diploids due to the simultaneous expression of both the -genes at MAT and the a-genes at HMR (KIMMERLY et al. 1988 ). A previous report described a collection of mutations that restore the -mating phenotype to HMRa-e** strains and thus potentially alter the function of genes that regulate silencing (AXELROD and RINE 1991 ). Two of these mutations, in fact, restore the -mating phenotype by suppressing the silencing defects of the HMRa-e** silencer. Analysis of these mutations led to the identification of two genes not previously known to play a role in silencing. One of these genes, CDC7, encodes a protein kinase required for cell-cycle progression, and the other, SAS2, encodes a homolog of a human gene involved in leukemia, as described above (AXELROD and RINE 1991 ; EHRENHOFER-MURRAY et al. 1997 ). In an effort to identify novel regulators of silencing and stimulated by the connection between SAS2 and human leukemia, we performed a systematic analysis of the remaining mutations. We found that 15 mutant strains from 10 independently mutagenized cultures contained recessive mutations responsible for restoring the -mating phenotype to MAT HMRa-e** strains (see MATERIALS AND METHODS). The genes responsible for the -mating phenotype of these mutants were referred to generically as SAS genes, as before, to reflect that they had Something to do About Silencing.

To determine whether the sas phenotype of the mutants was due to a mutation in a single nuclear gene, three mutants were chosen for initial characterization (DRY23, DRY24, and DRY25). Each mutant was mated to a mata1 HMRa-e** strain, forming a MAT/matap diploid homozygous for HMRa-e** and heterozygous for the mutation of interest. Tetrad analysis showed that in each case the suppressor of the HMRa-e** mutation segregated as a single nuclear mutation (see MATERIALS AND METHODS). To determine the number of mutant genes represented among the sas mutants, a complementation analysis was performed. Each of the 15 mutants contained mutations that fell into one of three complementation groups. One group corresponded to the SAS2 gene. The other mutations fell into two new complementation groups that corresponded to the newly identified genes SAS4 and SAS5. The complementation analysis was confirmed with null alleles of SAS4 and SAS5, as discussed below.

Identification of the SAS4 and SAS5 genes:
To clone wild-type copies of the SAS4 and SAS5 genes, a MAT HMRa-e** sas4-1 strain (DRY601) and a MAT HMRa-e** sas5-1 strain (DRY342) were transformed with a yeast genomic library in a centromere-containing vector. Transformants were screened for clones that could complement the sas phenotype. Complementation restored the -mating phenotype of the sas4 and sas5 mutants to the nonmating phenotype of SAS strains. In the case of the sas4-1 mutant, two overlapping and complementing clones each contained a 2.0-kb SalI-HindIII fragment of genomic DNA that, when subcloned into a Cen vector, could complement sas4-1. This fragment contained only a single open reading frame from chromosome IV previously known only by the systematic name of YDR181c. In the case of sas5-1, a single complementing plasmid clone was recovered. Subcloning analysis of the insert in this plasmid established that a 1.5-kb XbaI-SmaI fragment could complement the sas5-1 mutation. This fragment contained only a single open reading frame from chromosome XV previously known by two names, YOR213c and SC33KB 3. Allelism tests confirmed that the genes that complemented SAS4 and SAS5 were indeed the SAS4 and SAS5 structural genes, respectively (see MATERIALS AND METHODS).

SAS4 and SAS5 are nonessential genes:
Silencing in Saccharomyces is not an essential function and cells completely defective in silencing have normal growth rates and survival qualities. Silencing, however, is mediated by a combination of proteins some of which are essential for life, such as ORC, Rap1p, and Abf1p, and others that are nonessential, such as the SIR proteins (SHORE and NASMYTH 1987 ; FOSS et al. 1993 ; LOO et al. 1995 ). To determine whether SAS4 was essential for life, the entire coding region of SAS4 was replaced with the kanMX coding region, which confers G418 resistance, on one chromosome of an a/ diploid strain heterozygous for the HMRa-e** allele. Analysis of 28 tetrads from this diploid, upon sporulation, revealed that each tetrad contained four viable spores and the sas4::kanMX-containing spores showed no obvious growth defect. Thus SAS4, like the SIR genes, encoded a protein dispensable for growth. Among the segregants from this diploid, each of the MAT HMRa-e** sas4::kanMX segregants had the -mating phenotype. Thus, suppression of the HMR-E silencer defect reflected the null phenotype of SAS4 (Figure 1).

View larger version (57K): In this window In a new window Download PPT slide   Figure 1. Mating phenotypes of strains with sas4-1, sas5-1, sas4, and sas5 alleles in two strain backgrounds. (A) Comparison of -mating phenotype in MAT HMRa-e** strains with SAS4 or sas4 mutant alleles in JRY2069 (top) and in W303 (bottom). Results of quantitative mating analysis are presented in parentheses. (B) Comparison of -mating phenotype in MAT HMRa-e** strains with SAS5 or sas5 mutant alleles in JRY2069 (top) and in W303-1a (bottom). Results of quantitative mating analysis are presented in parentheses. Isogenic strains shown are HMRa-e** (JRY2069), sas4-1 (DRY24), sas4 (DRY1374), sas5-1 (DRY25), and sas5 (DRY1373). Strains isogenic with JRY3009 (MAT W303-1a) are HMRa-e** (JRY5273), sas4 (DRY1322), and sas5 (DRY1314).

The same strategy was used to test whether SAS5 was an essential gene. As with sas4 mutants, sas5 mutants were viable and had a normal growth rate. Moreover, MAT HMRa-e** sas5::HIS3 segregants were mating proficient. Thus, SAS5 was not essential for viability, and suppression of the HMRa-e** silencer defect reflected the null phenotype of SAS5 (Figure 1).

Complementation analysis with some of the original sas mutants indicated that SAS4 and SAS5 were newly characterized genes. To test more rigorously the assignment of mutants to complementation groups, complementation analysis was repeated using null alleles of SAS2, SAS4, and SAS5. A matap HMRa-e** sas4::kanMX strain (matap indicates a deletion of the MATa1 promoter; LOO and RINE 1994 ) (DRY1354) was mated to each of the original 15 recessive sas mutants to determine which contained lesions in the SAS4 gene. Similar experiments were performed with a strain containing a sas2::TRP1 allele (DRY1356) and with a strain containing the sas5::HIS3 allele (DRY1358). The results from these complementation tests clearly revealed that 9 mutants contained a sas2 mutant allele, 4 contained a sas4 mutant allele, and 2 contained a sas5 mutant allele (see MATERIALS AND METHODS). Transformation experiments revealed that each mutant could be complemented only by plasmids containing a wild-type copy of the corresponding SAS gene (data not shown). Based upon these multiple lines of evidence, we have renamed YDR181c as SAS4 and YOR213c as SAS5.

SAS4 and SAS5 are regulators of silencing:
The experiments described above established that mutations in SAS4 and SAS5 restored the -mating phenotype in MAT cells containing the HMRa-e** mutation. There are two ways of restoring the -mating phenotype: the SAS4 and SAS5 mutations could block a1 function in some way such that the a1/2 repressor fails to repress expression of 1; alternatively, the SAS4 and SAS5 mutations could restore silencing of the mutant HMR locus. We distinguished between these models by determining whether the -mating phenotype in the sas mutants depended upon the function of SIR4, which is required for silencing.

The SIR4 dependence of the sas4 and sas5 phenotypes was tested by crossing both a MATa HMRa-e** sas4::kanMX strain (DRY1360) and an isogenic MATa HMRa-e** sas5::HIS3 strain (DRY1391) to an isogenic MAT HMR-SS::ADE2 sir4::LEU2 strain (DRY1376) in which the natural HMR-E silencer was replaced by a synthetic silencer and the MATa genes normally found at HMR were replaced by ADE2 (MCNALLY and RINE 1991 ; RIVIER et al. 1999 ). Thus, the HMR allele present in all segregants from these crosses could be unambiguously identified. Nine MAT HMRa-e** sas4::kanMX sir4::LEU2 segregants were identified from the first cross and 12 MAT HMRa-e** sas5::HIS3 sir4::LEU2 segregants were identified from the second cross. All of these segregants were unable to mate, whereas all the MAT HMRa-e** sas4::kanMX SIR4 segregants and all the MAT HMRa-e** sas5::HIS3 SIR4 segregants were able to mate (Figure 2). The SIR4 dependence of the sas mutant phenotypes established that sas mutants restored silencing per se.

View larger version (66K): In this window In a new window Download PPT slide   Figure 2. SAS4 and SAS5 mutants restore silencing of HMRa-e**. The -mating phenotype of HMRa-e** sas4 strains was abolished by deletion of SIR4 (top). Similarly, the -mating phenotype of HMRa-e** sas5 strains is abolished by deletion of SIR4 (bottom). Hence, SAS4 and SAS5 mutations restore SIR-dependent silencing. The strains shown were sas4 SIR4 (DRY1322), sas4 sir4 (DRY1397), sas5 SIR4 (DRY1314), and sas5 sir4 (DRY1398).

The previous experiments established that the sas4 and sas5 phenotypes were dependent on silencing functions. Nevertheless, these experiments did not eliminate the formal possibility that sas4 or sas5 mutations might also affect MATa1 function. Therefore, two a/ diploids were constructed, one homozygous for sas4 (DRY1426) and one for sas5 (DRY1428). Both diploids had the nonmating phenotype of a wild-type a/ diploid. Thus, the effect of sas4 and sas5 on mating phenotype was exclusively through a silencing mechanism (Figure 3).

View larger version (86K): In this window In a new window Download PPT slide   Figure 3. SAS4 and SAS5 are not required for MATa-gene expression. Deletion of SAS4 or SAS5 results in the -mating phenotype in haploid MAT HMRa-e** strains and does not result in the -mating phenotype in MATa/MAT diploids. Strains shown are MAT (JRY3009), HMRa-e** (JRY5273), sas4 (DRY1322), sas5 (DRY1314), MATa/MAT (DRY1568), sas4/sas4 (DRY1426), and sas5/sas5 (DRY1428).

Efficient silencing by null alleles of SAS4 and SAS5 depends on the ACS and Rap1p binding site of a synthetic HMR-E silencer:
In the context of the wild-type HMR-E silencer the Rap1p and Abf1p binding sites and the ARS consensus sequence element (ACS) appear to have redundant functions; mutation of any individual element does not disrupt silencing, whereas mutation of any pairwise combination of elements does (BRAND et al. 1987 ; KIMMERLY et al. 1988 ). In principle, null alleles of SAS4 or SAS5 could restore silencing to the HMRa-e** silencer by bypassing the role of the Rap1p binding site, the Abf1p binding site, or both. Alternatively, null alleles of SAS4 or SAS5 could restore silencing by increasing the activity of the silencer elements that remain in HMRa-e** strains, namely, the ACS of HMR-E or the HMR-I silencer. To explore these possibilities we systematically tested which silencer elements were required for silencing in sas4 and sas5 strains. To make these experiments simpler to interpret, we used mutant forms of a synthetic silencer (HMR-SS) that lack some of the apparent functional redundancy that complicates analysis of mutant forms of the natural HMR-E silencer (MCNALLY and RINE 1991 ; RIVIER et al. 1999 ). Previous analysis indicates that restoration of silencing by null alleles of SAS2 does not depend on either the Abf1p binding site of the synthetic silencer or the HMR-I silencer. In contrast, null alleles of SAS2 do not suppress mutations in the ACS of the synthetic silencer and only partially suppress defects in the Rap1p binding site. Thus, the silencing that results from null alleles of SAS2 depends on the ACS and Rap1p binding sites of the synthetic silencer.

Deletion of the HMR-I silencer from a strain containing the synthetic silencer (HMR-SS I) (DRY439) resulted in a 10-fold loss of silencing as judged by decreased mating efficiency relative to a strain that contained the synthetic silencer and HMR-I (HMR-SS) (DRY874; Figure 4 and Figure 5). Deletion of either SAS4 (DRY1364; Figure 4) or SAS5 (DRY2109; Figure 5) restored silencing in an HMR-SS I strain to wild-type levels. Therefore, silencing did not depend on HMR-I in either sas4 or sas5 strains. We next investigated the role of the Abf1p binding site in strains lacking SAS4 or SAS5. Mutation of the Abf1p binding site of the synthetic silencer in a strain lacking HMR-I (JRY4889) (HMR-SS abf1- I) resulted in a 2- to 3-fold decrease in mating efficiency. Deletion of either SAS4 (DRY1365; Figure 4) or SAS5 (DRY2110; Figure 5) in an HMR-SS abf1- I strain restored silencing to wild-type levels. Therefore silencing did not depend on the Abf1p binding site in sas4 or sas5 strains. Collectively, these and previous data revealed that neither the Abf1p binding site of the synthetic silencer or HMR-I is required for silencing in sas2, sas4, or sas5 strains.

View larger version (26K): In this window In a new window Download PPT slide   Figure 4. The ACS and Rap1p binding sites of the synthetic HMR-E silencer (HMR-SS) contribute to silencing in sas4 strains. The logarithmic values of quantitative mating assays for strains containing the indicated alleles of HMR-E are shown. Open bars indicate the values of SAS4 strains, solid bars the value of sas4 strains. The strains shown are DRY439 and DRY1364 [HMR-SS I (ssI)], DRY879 and DRY1365 [HMR-SS abf1- I (ssbI)], DRY875 and DRY1366 [HMR-SS rap1- (ssr)], DRY882 and DRY1370 [HMR-SS rap1- I (ssrI)], DRY881 and DRY1368 [HMR-SS acs- (ssa)], DRY878 and DRY1369 [HMR-SS acs- I (ssaI)].

View larger version (25K): In this window In a new window Download PPT slide   Figure 5. The ACS and Rap1p binding sites of the synthetic HMR-E silencer (HMR-SS) contribute to silencing in sas5 strains. The logarithmic values of quantitative mating assays for strains containing the indicated alleles of HMR-E are shown. Open bars indicate the values of SAS5 strains, solid bars the value of sas5 strains. The strains shown are DRY439 and (DRY2109) [HMR-SS I (ssI)], DRY879 and (DRY2110) [HMR-SS abf1- I (ssbI)], DRY875 and (DRY2111) [HMR-SS rap1- (ssr)], DRY882 and (DRY2112) [HMR-SS rap1- I (ssrI)], DRY881 and (DRY2113) [HMR-SS acs- (ssa)], DRY878 and (DRY2114) [HMR-SS acs- I (ssaI)].

We next investigated the role of the ACS of the synthetic silencer in silencing in sas4 and sas5 strains. Deletion of the ACS of the synthetic silencer (DRY881) (HMR-SS acs-) in an otherwise wild-type strain resulted in an approximately 10-fold decrease in silencing as judged by mating efficiency. Deletion of SAS4 (DRY1368; Figure 4) or SAS5 (DRY2113; Figure 5) from an HMR-SS acs- strain did not increase mating efficiency more than 2-fold and did not restore mating to wild-type levels. Thus, the ACS of the synthetic silencer was required for the efficient restoration of silencing by null alleles of SAS4 and SAS5. Strains lacking the ACS of the synthetic silencer and HMR-I (HMR-SS acs- I) (DRY878) have a mating efficiency that is approximately five orders of magnitude less than strains with a wild-type HMR allele. Deletion of SAS4 (DRY1369) or SAS5 (DRY2114) from an HMR-SS acs- I strain resulted in an increase in silencing, but only to a level that was approximately four orders of magnitude less than wild type. Thus, the ACS was required for efficient restoration of silencing by null alleles of SAS4 and SAS5, both in the presence and absence of HMR-I. Previous analysis indicated that null alleles of SAS2 were not capable of even slight suppression of silencing in either HMR-SS acs- or HMR-SS acs- I strains as seen here for null alleles of SAS4 and SAS5. To determine whether null alleles of SAS2 were phenotypically distinct from null alleles of SAS4 or SAS5, we compared the mating efficiency of HMR-SS acs- I strains lacking SAS2, SAS4, or SAS5. By our assays, deletion of SAS2 resulted in the same slight suppression of the silencing defect of the HMR-SS acs- I allele as did deletion of SAS4 or SAS5 (data not shown). Thus, the dependence of silencing on the ACS by null alleles of SAS2, SAS4, or SAS5 was indistinguishable by the assays used here.

Finally, we investigated the contribution of the Rap1p binding site to silencing in sas4 and sas5 strains. Deletion of the Rap1p binding site of the synthetic silencer results in a reduction of mating efficiency by three to four orders of magnitude in the presence of HMR-I (HMR-SS rap1-) (DRY875). Deletion of either SAS4 (DRY1366) or SAS5 (DRY2111) from an HMR-SS rap1- strain resulted in an increase in mating efficiency, but only to a level that was two to three orders of magnitude less than wild type. Therefore, the Rap1p binding site of the synthetic silencer was required for efficient restoration of silencing by null alleles of SAS4 and SAS5. These results were similar to previous results that deletion of SAS2 partially restores silencing to the HMR-SS rap1- allele, indicating that null alleles of SAS2, SAS4, and SAS5 have similar phenotypes in this context. To further explore the role of the Rap1p binding site in silencing in sas4 and sas5 strains, we analyzed the HMR-SS rap1- allele in the absence of HMR-I (HMR-SS rap1- I). A strain containing this HMR-SS rap1- I allele (DRY882) mated approximately five orders of magnitude less well than a strain containing the wild-type allele of HMR (Figure 4 and Figure 5). Deletion of SAS4 from this HMR-SS rap1- I strain (DRY1370) resulted in an increase in mating efficiency, but only to a level that was three to four orders of magnitude less than wild type (Figure 4). Comparable levels of silencing were previously reported for a sas2 HMR-SS rap1- I strain (EHRENHOFER-MURRAY et al. 1997 ). In contrast, deletion of SAS5 from an HMR-SS rap1- I strain (DRY2112) did not result in an increase in mating efficiency (Figure 5). Thus, by these criteria, the Rap1p binding site was required for efficient restoration of silencing by null alleles of SAS2, SAS4, and SAS5, and furthermore, the Rap1p binding site made a more significant contribution to restoration of silencing by null alleles of SAS5 than by null alleles of SAS2 or SAS4. Collectively, the results presented here and previously indicated that the suppression of silencing defects at HMR in sas2, sas4, and sas5 strains depends on the ACS and Rap1p binding sites of the synthetic silencer, and not on the Abf1p binding site or HMR-I.

Sas5p was a family member of a protein implicated in human leukemia:
Comparison of the predicted protein sequence of SAS5 with other proteins encoded by the yeast genome revealed one strong paralog, ANC1. The Blastp comparison of Sas5p and Anc1p resulted in a score of 10^-26, with two regions of similarity that together span the majority of the length of both proteins. ANC1 was originally identified as a potential regulator of the actin cytoskeleton, but more recent evidence indicates that ANC1 encodes a protein intimately connected with transcription (HENRY et al. 1994 ; KIM et al. 1994 ; WELCH and DRUBIN 1994 ; CAIRNS et al. 1996 ). Previous analysis of the predicted protein encoded by ANC1 revealed that it has significant sequence similarity to proteins encoded by two human genes, AF-9 and ENL, that are the sites of reciprocal translocations that contribute to acute myeloid leukemia and that it also has significant sequence similarity to a putative protein encoded by the uncharacterized yeast open reading frame SC33KB 3 (WELCH and DRUBIN 1994 ; CAIRNS et al. 1996 ). As revealed here, SC33KB 3 is identical to SAS5. As described previously, the region of highest similarity among these four proteins is a 42-amino-acid region corresponding to amino acids 52–93 of Sas5p. In addition to ANC1, our analysis identified a second yeast paralog of SAS5, the YNL107w gene, whose function was unknown. Comparisons of the SAS5 sequence with the non-Saccharomyces entries of GenBank revealed similarities to the YD67 gene of Schizosaccharomyces pombe and the M04B2.3 gene of Caenorhabditis elegans, in addition to AF-9 and ENL.

The alignment of SAS5 to its related genes indicated that the region of 42 amino acids that is similar among SAS5, ANC1, AF-9, and ENL also corresponds to the region of highest similarity with YNL107w, YD67, and MO4B23 (Figure 6). Within this region, SAS5 was 38–51% identical and 56–65% similar to each of the related proteins, suggesting that these proteins are members of a family of proteins that contain a region of conserved function. Although the similarity among the Sas5p-related proteins implies that each has related functions, the region of similarity does not extend over the entire length of these proteins; therefore the family members may not carry out the exact same function. In contrast to Sas5p, the sequence of Sas4p was not highly similar to other known proteins and thus defined a pioneer protein.

View larger version (69K): In this window In a new window Download PPT slide   Figure 6. Sequence similarity among Sas5p-related proteins. Shown is a 76-amino-acid region that spans the 42-amino-acid region of highest similarity among family members. Black shading around letters indicates that at least half of the aligned proteins had identical amino acids. Gray shading around letters indicates that at least half of the family members had similar amino acids. The 42-amino-acid region of highest similarity among the proteins shown is bracketed by the carets. Sequences were aligned and similarity was scored as indicated in MATERIALS AND METHODS. Gene names are indicated to the left of the aligned sequence: YD67 (S. pombe), M04B2.3 (C. elegans), AF-9 and ENL (human), YNL107w, ANC1, and SAS5 (S. cerevisiae).

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

We characterized 15 recessive mutations that suppressed the silencing defect associated with a mutant HMR-E silencer. Nine of the mutations were in the previously characterized SAS2 gene, which encodes an acetylase homolog. Of the remaining mutations, 4 were in SAS4 and 2 were in SAS5. Cells bearing null alleles of either SAS4 or SAS5 were viable, and the phenotypes of the null alleles were indistinguishable from those of the original mutants.

The sas4 and sas5 mutations restored the -mating phenotype by restoring silencing of HMRa-e** rather than by interfering with the function of the MATa1-encoded protein. This conclusion was based upon the requirement of SIR4 function for the suppression caused by null alleles of SAS4 or SAS5 and by the requirement for the ACS and a Rap1p binding site of the synthetic HMR-E silencer for efficient suppression of the silencing defect by a null allele of SAS4 and SAS5. Furthermore, MATa/MAT diploids homozygous for null alleles of SAS4 or SAS5 had the nonmating phenotype, ruling out the possibility that SAS4 or SAS5 was required for a1 function.

The function of SAS4 and SAS5 in silencing:
There are at least two ways of thinking about how sas4 and sas5 mutations increased silencing mediated by the mutant HMRa-e** silencer. One view is that the proteins encoded by these genes directly inhibited the function of ORC or Rap1p at HMR-E. Mutation of either gene would then relieve the inhibitory effect, allowing Rap1p or ORC to have increased function at the HMR-E silencer. An alternative model is that neither SAS4 nor SAS5 had a direct effect at HMR-E. Rather, these proteins might have a direct effect on the assembly of silenced chromatin at telomeres. Previous studies have revealed a competition between telomeric silencing and silencing of HMR (BUCK and SHORE 1995 ). In particular, some mutations that increase silencing at the telomeres result in a decrease in silencing at HMR (BUCK and SHORE 1995 ; WOTTON and SHORE 1997 ). By relieving silencing at telomeres, the sas4 and sas5 mutations could favor restoration of silencing at HMR, despite the presence of the mutant silencer. Our recent analysis demonstrates that both SAS4 and SAS5 are required for the telomeric position effect and, hence, play a positive role in the formation of repressive chromatin (XU et al. 1999 ). Certainly, the requirement for SAS4 and SAS5 in telomeric silencing lends favor to the latter model.

A link between SAS genes and human leukemia:
Many of the mutations that are known to contribute to human leukemia are reciprocal chromosomal translocations that result in the formation of chimeric genes. Previous work established that SAS2 is highly similar to the human MOZ gene, which is a site of recurring reciprocal translocations that form a chimeric gene with CBP in one subtype of acute myeloid leukemia (BORROW et al. 1996 ; REIFSNYDER et al. 1996 ). SAS2 and MOZ are members of the MYST gene family whose members have similarity to acetylases (REIFSNYDER et al. 1996 ). Furthermore, two MYST family members, ESA1 and Tip60, are known histone acetylases (YAMAMOTO and HORIKOSHI 1997 ; SMITH et al. 1998 ). As described here, mutations in SAS4 and SAS5 restore silencing to defective alleles of HMR-E as do mutations in SAS2, raising the possibility that the functions of these three proteins are related. Remarkably, SAS5 family members in humans are also found as chimeric genes created by other chromosomal rearrangements in different subtypes of acute myeloid leukemias (TKACHUK et al. 1992 ; NAKAMURA et al. 1993 ). These data extend the connections between the SAS genes and genes that contribute to acute myeloid leukemias.

The relationship between yeast silencing genes and human leukemia genes is further extended by studies of a family of proteins that share a SET domain. Set1p contains a block of ~130–140 amino acids, known as a SET domain, which is shared among a variety of proteins throughout eukaryotes (NISLOW et al. 1997 ). The SET-domain family members have disparate effects on transcription. For instance, SET1 activates transcription of some genes in yeast and represses transcription of others, including silencing of genes near telomeres (NISLOW et al. 1997 ). Similarly, Drosophila proteins with a SET domain include trithorax and enhancer of zeste, which are responsible for establishing stable activated and repressed states of gene expression in development (JONES and GELBART 1993 ). The human trithorax homolog, known variously as HRX, MLL, and ALL-1, is the site of recurring chromosomal translocations that contribute to a variety of human leukemias, including acute myeloid leukemia. Strikingly, the SAS5-related genes AF9 and ENL are fused to the SET1-related human trithorax homolog in specific types of acute myeloid leukemia (TENEN et al. 1997 ; WARING and CLEARY 1997 ). Hence, both partners in HRX-AF9 and HRX-ENL fusions are related to yeast genes involved in silencing. Based on the parallels described for SAS2 and SAS5, any SAS4 homolog discovered in humans would be a logical candidate to evaluate for association with chromosomal breakpoints in human leukemias.

	ACKNOWLEDGMENTS

We thank B. Cairns and L. Pillus for discussions including unpublished results. We also thank C. Fox, L. Pillus, and A. Wach for generously providing plasmids and strains. This work was supported by National Institutes of Health (NIH) grant GM-52103 (D.R.), by NIH grant GM-31105 (J.R.), by a March of Dimes Basil O'Connor Starter Scholar Award (D.R.), and by NIH predoctoral training award 5T32-GM07283 (S.K.). The initial stages of this work were funded by a postdoctoral fellowship from the California Division of the American Cancer Society (D.R.).

Manuscript received October 5, 1998; Accepted for publication May 4, 1999.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

ABRAHAM, J., K. A. NASMYTH, J. N. STRATHERN, A. J. KLAR, and J. B. HICKS, 1984  Regulation of mating-type information in yeast. Negative control requiring sequences both 5' and 3' to the regulated region. J. Mol. Biol. 176:307-331[Medline].

AXELROD, A. and J. RINE, 1991  A role for CDC7 in repression of transcription at the silent mating-type locus HMR in Saccharomyces cerevisiae.. Mol. Cell. Biol. 11:1080-1091[Abstract/Free Full Text].

BANNISTER, A. J. and T. KOUZARIDES, 1996  The CBP co-activator is a histone acetyltransferase. Nature 384:641-643[Medline].

BAUDIN, A., O. OZIER-KALOGEROPOULOS, A. DENOUEL, F. LACROUTE, and C. CULLIN, 1993  A simple and efficient method for direct gene deletion in Saccharomyces cerevisiae.. Nucleic Acids Res. 21:3329-3330[Free Full Text].

BORROW, J., V. P. STANTON, JR., J. M. ANDRESEN, R. BECHER, and F. G. BEHM et al., 1996  The translocation t(8;16)(p11;p13) of acute myeloid leukaemia fuses a putative acetyltransferase to the CREB-binding protein. Nat. Genet. 14:33-41[Medline].

BRAND, A. H., L. BREEDEN, J. ABRAHAM, R. STERNGLANZ, and K. NASMYTH, 1985  Characterization of a "silencer" in yeast: a DNA sequence with properties opposite to those of a transcriptional enhancer. Cell 41:41-48[Medline].

BRAND, A. H., G. MICKLEM, and K. NASMYTH, 1987  A yeast silencer contains sequences that can promote autonomous plasmid replication and transcriptional activation. Cell 51:709-719[Medline].

BRAUNSTEIN, M., R. E. SOBEL, C. D. ALLIS, B. M. TURNER, and J. R. BROACH, 1996  Efficient transcriptional silencing in Saccharomyces cerevisiae requires a heterochromatin histone acetylation pattern. Mol. Cell. Biol. 16:4349-4356[Abstract].

BROWNELL, J. E. and C. D. ALLIS, 1996  Special HATs for special occasions: linking histone acetylation to chromatin assembly and gene activation. Curr. Opin. Genet. Dev. 6:176-184[Medline].

BROWNELL, J. E., J. ZHOU, T. RANALLI, R. KOBAYASHI, and D. G. EDMONDSON et al., 1996  Tetrahymena histone acetyltransferase A: a homolog to yeast Gcn5p linking histone acetylation to gene activation. Cell 84:843-851[Medline].

BRYK, M., M. BANERJEE, M. MURPHY, K. E. KNUDSEN, and D. J. GARFINKEL et al., 1997  Transcriptional silencing of Ty1 elements in the RDN1 locus of yeast. Genes Dev. 11:255-269[Abstract/Free Full Text].

BUCK, S. W. and D. SHORE, 1995  Action of a RAP1 carboxy-terminal silencing domain reveals an underlying competition between HMR and telomeres in yeast. Genes Dev. 9:370-384[Abstract/Free Full Text].

CAIRNS, B. R., 1998  Chromatin remodeling machines: similar motors, ulterior motives. Trends Biochem. Sci. 23:20-25[Medline].

CAIRNS, B. R., N. L. HENRY, and R. D. KORNBERG, 1996  TFG/TAF30/ANC1, a component of the yeast SWI/SNF complex that is similar to the leukemogenic proteins ENL and AF-9. Mol. Cell. Biol. 16:3308-3316[Abstract].

CARAPETI, M., R. C. AGUIAR, J. M. GOLDMAN, and N. C. CROSS, 1998  A novel fusion between MOZ and the nuclear receptor coactivator TIF2 in acute myeloid leukemia. Blood 91:3127-3133[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].

CLEARY, M. L., 1991  Oncogenic conversion of transcription factors by chromosomal translocations. Cell 66:619-622[Medline].

CLEARY, M. L., 1992  Transcription factors in human leukaemias. Cancer Surv. 15:89-104[Medline].

EHRENHOFER-MURRAY, A. E., D. H. RIVIER, and J. RINE, 1997  The role of Sas2, an acetyltransferase homologue of Saccharomyces cerevisiae, in silencing and ORC function. Genetics 145:923-934[Abstract].

FELDMAN, J. B., J. B. HICKS, and J. R. BROACH, 1984  Identification of sites required for repression of a silent mating type locus in yeast. J. Mol. Biol. 178:815-834[Medline].

FOSS, M., F. J. MCNALLY, P. LAURENSON, and J. RINE, 1993  Origin recognition complex (ORC) in transcriptional silencing and DNA replication in S. cerevisiae.. Science 262:1838-1844[Abstract/Free Full Text].

FOX, C. A., A. E. EHRENHOFER-MURRAY, S. LOO, and J. RINE, 1997  The origin recognition complex, SIR1, and the S phase requirement for silencing. Science 276:1547-1551[Abstract/Free Full Text].

GRUNSTEIN, M., 1998  Yeast heterochromatin: regulation of its assembly and inheritance by histones. Cell 93:325-328[Medline].

HECHT, A., T. LAROCHE, S. STRAHL-BOLSINGER, S. M. GASSER, and M. GRUNSTEIN, 1995  Histone H3 and H4 N-termini interact with SIR3 and SIR4 proteins: a molecular model for the formation of heterochromatin in yeast. Cell 80:583-592[Medline].

HECHT, A., S. STRAHL-BOLSINGER, and M. GRUNSTEIN, 1996  Spreading of transcriptional repressor SIR3 from telomeric heterochromatin. Nature 383:92-96[Medline].

HENRY, N. L., A. M. CAMPBELL, W. J. FEAVER, D. POON, and P. A. WEIL et al., 1994  TFIIF-TAF-RNA polymerase II connection. Genes Dev. 8:2868-2878[Abstract/Free Full Text].

HOLSTEGE, F. C., E. G. JENNINGS, J. J. WYRICK, T. I. LEE, and C. J. HENGARTNER et al., 1998  Dissecting the regulatory circuitry of a eukaryotic genome. Cell 95:717-728[Medline].

JONES, J. S. and L. PRAKASH, 1990  Yeast Saccharomyces cerevisiae selectable markers in pUC18 polylinkers. Yeast 6:363-366[Medline].

JONES, R. S. and W. M. GELBART, 1993  The Drosophila Polycomb-group gene Enhancer of zeste contains a region with sequence similarity to trithorax. Mol. Cell. Biol. 13:6357-6366[Abstract/Free Full Text].

KIM, Y. J., S. BJORKLUND, Y. LI, M. H. SAYRE, and R. D. KORNBERG, 1994  A multiprotein mediator of transcriptional activation and its interaction with the C-terminal repeat domain of RNA polymerase II. Cell 77:599-608[Medline].

KIMMERLY, W., A. BUCHMAN, R. KORNBERG, and J. RINE, 1988  Roles of two DNA-binding factors in replication, segregation and transcriptional repression mediated by a yeast silencer. EMBO J. 7:2241-2253[Medline].

KUO, M. H., J. ZHOU, P. JAMBECK, M. E. CHURCHILL, and C. D. ALLIS, 1998  Histone acetyltransferase activity of yeast Gcn5p is required for the activation of target genes in vivo. Genes Dev. 12:627-639[Abstract/Free Full Text].

LOO, S. and J. RINE, 1994  Silencers and domains of generalized repression. Science 264:1768-1771[Abstract/Free Full Text].

LOO, S., P. LAURENSON, M. FOSS, A. DILLIN, and J. RINE, 1995  Roles of ABF1, NPL3, and YCL54 in silencing in Saccharomyces cerevisiae.. Genetics 141:889-902[Abstract].

LUSTIG, A. J., 1998  Mechanisms of silencing in Saccharomyces cerevisiae.. Curr. Opin. Genet. Dev. 8:233-239[Medline].

MCNALLY, F. J. and J. RINE, 1991  A synthetic silencer mediates SIR-dependent functions in Saccharomyces cerevisiae.. Mol. Cell. Biol. 11:5648-5659[Abstract/Free Full Text].

MIZZEN, C. A., X. J. YANG, T. KOKUBO, J. E. BROWNELL, and A. J. BANNISTER et al., 1996  The TAF(II)250 subunit of TFIID has histone acetyltransferase activity. Cell 87:1261-1270[Medline].

MORETTI, P., K. FREEMAN, L. COODLY, and D. SHORE, 1994  Evidence that a complex of SIR proteins interacts with the silencer and telomere-binding protein RAP1. Genes Dev. 8:2257-2269[Abstract/Free Full Text].

NAKAMURA, T., H. ALDER, Y. GU, R. PRASAD, and O. CANAANI et al., 1993  Genes on chromosomes 4, 9, and 19 involved in 11q23 abnormalities in acute leukemia share sequence homology and/or common motifs. Proc. Natl. Acad. Sci. USA 90:4631-4635[Abstract/Free Full Text].

NISLOW, C., E. RAY, and L. PILLUS, 1997  SET1, a yeast member of the trithorax family, functions in transcriptional silencing and diverse cellular processes. Mol. Biol. Cell 8:2421-2436[Abstract/Free Full Text].

RABBITTS, T. H., 1994  Chromosomal translocations in human cancer. Nature 372:143-149[Medline].

REIFSNYDER, C., J. LOWELL, A. CLARKE, and L. PILLUS, 1996  Yeast SAS silencing genes and human genes associated with AML and HIV-1 Tat interactions are homologous with acetyltransferases. Nat. Genet. 14:42-49[Medline].

RIVIER, D. H., J. L. EKENA, and J. RINE, 1999  HMR-I is an origin of replication and a silencer in Saccharomyces cerevisiae.. Genetics 151:521-529[Abstract/Free Full Text].

ROWLEY, J. D., S. RESHMI, O. SOBULO, T. MUSVEE, and J. ANASTASI et al., 1997  All patients with the T(11;16)(q23;p13.3) that involves MLL and CBP have treatment-related hematologic disorders. Blood 90:535-541[Abstract/Free Full Text].

RUNDLETT, S. E., A. A. CARMEN, N. SUKA, B. M. TURNER, and M. GRUNSTEIN, 1998  Transcriptional repression by UME6 involves deacetylation of lysine 5 of histone H4 by RPD3. Nature 392:831-835[Medline].

SCHNITZLER, G., S. SIF, and R. E. KINGSTON, 1998  Human SWI/SNF interconverts a nucleosome between its base state and a stable remodeled state. Cell 94:17-27[Medline].

SHORE, D. and K. NASMYTH, 1987  Purification and cloning of a DNA binding protein from yeast that binds to both silencer and activator elements. Cell 51:721-732[Medline].

SMITH, E. R., A. EISEN, W.