This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowReprints and Permissions
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Safi, A.
Right arrow Articles by Rusche, L. N.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Safi, A.
Right arrow Articles by Rusche, L. N.

 Previous Article  |  Next Article 

Molecular and Cellular Biology, April 2008, p. 2567-2578, Vol. 28, No. 8
0270-7306/08/$08.00+0     doi:10.1128/MCB.01785-07
Copyright © 2008, American Society for Microbiology. All Rights Reserved.

Evolution of New Function through a Single Amino Acid Change in the Yeast Repressor Sum1p{triangledown}

Alexias Safi, Kelley A. Wallace,{dagger} and Laura N. Rusche*

Institute for Genome Sciences and Policy and Biochemistry Department, Duke University, Durham, North Carolina 27710

Received 28 September 2007/ Returned for modification 21 October 2007/ Accepted 2 February 2008


arrow
ABSTRACT
 
The SUM1-1 mutation is an example of a single amino acid change that results in new function. Wild-type Sum1p in Saccharomyces cerevisiae is a DNA-binding repressor that acts locally, whereas mutant Sum1-1p forms an extended repressive chromatin structure. By characterizing a panel of mutations in which various amino acids replaced the critical residue, threonine 988, we found that threonine was required for wild-type function and that in the absence of threonine the association of Sum1p with DNA was reduced. Isoleucine, the amino acid in mutant Sum1-1p, was required for the novel spreading property. Thus, the SUM1-1 mutation results in both a loss and a gain of function. The presence of isoleucine caused Sum1-1p to self-associate, a property that may promote spreading. In addition, isoleucine enabled Sum1-1p to associate with the origin recognition complex (ORC) and accumulate near ORC binding sites. Thus, both threonine and isoleucine at position 988 enable Sum1p to form intermolecular interactions. We propose that interaction domains may be hotspots for gain-of-function mutations because alterations in such domains have the potential to redirect a protein to new sets of binding partners. In addition, self-association of chromatin proteins may promote the formation of extended chromatin structures.


arrow
INTRODUCTION
 
Neomorphic, or gain-of-function, mutations are likely to be initiating events in the evolution of new functions, particularly because such mutations can result from only one or a few amino acid substitutions (10, 27) and therefore can arise spontaneously. However, despite the potential importance of neomorphic mutations in the process of evolution, little is known of the molecular mechanisms by which this class of mutations acts. The SUM1-1 mutation in Saccharomyces cerevisiae is an example of a single amino acid change (T988I) that results in new function (6). Wild-type Sum1p is a DNA-binding protein that acts locally to repress promoters of midsporulation and other genes (2, 39). In contrast, the mutant Sum1-1p forms an extended repressive chromatin structure at the mating-type loci (32, 35). Thus, the mutation changes two properties of the protein—its genomic location and its ability to spread. A simple model explaining the gain of two properties through a single point mutation is that the primary effect of the mutation is to direct the protein to new genomic locations, which happen to be more permissive for spreading. In this scenario, the increased ability to spread is a secondary effect of the altered location. However, when this model was tested by experimentally recruiting wild-type Sum1p to the mating-type locus HMR, where the mutant Sum1-1p spreads, wild-type Sum1p remained unable to spread (21). Therefore, the ability to spread is a property of the mutant protein and not the genomic location. Thus, in addition to being a model neomorphic mutation, Sum1-1p provides a useful case study for understanding what limits the spreading of silenced chromatin.

It is important that the spreading of silenced chromatin be properly regulated. In most eukaryotic cells, silenced chromatin, known as heterochromatin, is found constitutively at pericentromeric and telomeric regions, and facultative heterochromatin forms at other locations (11, 36). Heterochromatin serves a structural role in maintaining integrity of the genome and also regulates expression of genes within the heterochromatin domain. However, heterochromatin must be restricted to the appropriate regions of the genome so that it does not interfere with expression of nonheterochromatic genes. Recent work has investigated how factors, such as boundary sequences and histone-modifying enzymes, limit the spreading of heterochromatin proteins (37, 38). Less attention has been given to how the intrinsic properties of chromatin proteins themselves dampen the spreading process. The example of Sum1-1p, which appears to be at a tipping point between spreading and not spreading, suggests that the extent of spreading is, in part, determined by properties of chromatin proteins. Therefore, we have investigated how the SUM1-1 mutation increases the ability of the protein to spread.

The SUM1-1 mutation was initially isolated as a suppressor of the mating defect of a sir2 mutation (17). In S. cerevisiae, the Sir complex constitutively silences the mating-type loci HML{alpha} and HMRa through the formation of an extended repressive chromatin structure (30). In the absence of this silencing, the cells express both a and {alpha} mating-type genes and consequently are unable to mate. The SUM1-1 mutation restores silencing, and hence mating, to sir mutant strains through the formation of an alternative repressive chromatin at the mating-type loci (32). In wild-type cells, the Sir proteins are recruited to the silenced mating-type loci through sequence elements called silencers, which flank each locus. Each silencer has binding sites for at least two of three cellular factors, including the origin recognition complex (ORC), which binds to all four silencers. The silencer binding proteins, in turn, recruit the Sir proteins. Finally, the Sir proteins propagate along the chromosome away from the silencers through a sequential deacetylation mechanism in which Sir2p deacetylates histone tails and Sir3p and Sir4p bind to the newly deacetylated histones (14, 20, 31).

Like the Sir complex, the mutant Sum1-1 complex is recruited to the silent mating-type loci through the silencers. In particular, ORC, but not the other two silencer binding proteins, is required for the association of Sum1-1p with the silencers (21, 32), and it has been proposed that the SUM1-1 mutation increases the affinity of the protein for ORC (32, 35). Also like the Sir complex, the Sum1-1 complex propagates along the chromosome away from the silencer. This spreading requires the deacetylase activity of Hst1p (21), a deacetylase that is related to Sir2p and that associates with both the wild-type and mutant Sum1p (22, 32, 39). It is predicted that Sum1-1p or another component of the Sum1-1 complex associates with unacetylated histones to enable spreading.

It remains unclear how a single amino acid substitution results in both the relocalization of the protein and an increased ability to spread. To understand how the SUM1-1 mutation exerts its effect, we generated a panel of mutations in which various amino acids substituted for threonine 988 in Sum1p. Threonine was required for wild-type function, and isoleucine, the amino acid present in Sum1-1p, was required for the novel spreading property. Thus, the SUM1-1 mutation results in both a loss and a gain of function.


arrow
MATERIALS AND METHODS
 
Yeast strains and plasmids. Plasmids used in this study are described in Table 1. Mutations of threonine 988 in SUM1 were constructed by site-directed mutagenesis using a published procedure (3). Briefly, the plasmid to be mutated was amplified by PCR using Pfu Turbo (Stratagene) and two complementary oligonucleotide primers containing the desired mutation. The final product of this reaction, a mutated plasmid containing staggered nicks, was incubated with DpnI, which cuts the methylated template plasmid but not the unmethylated PCR product, and then transformed into Escherichia coli. Mutations were confirmed by sequencing.


View this table:
[in this window]
[in a new window]

  		TABLE 1. Plasmids used in this study

To generate plasmids for expression of Sum1p in E. coli, the desired region of SUM1 was amplified by PCR with primers containing NdeI (5') and BamHI (3') restriction sites. The resultant PCR product was ligated into the NdeI and BamHI sites of the expression vector pET15b. To generate hemagglutinin (HA)-tagged alleles of SUM1, an AvrII site was created at the 5' end of the open reading frame as previously described (32). A multimerized HA tag flanked by AvrII sites was ligated into this site. The resultant HA-tagged proteins were stably expressed in yeast.

S. cerevisiae strains used in this study were all derived from W303-1a (Table 2). The sum1{Delta}::LEU2, sum1{Delta}::URA3, sir2{Delta}::HIS3 (32), and TRP1::SUM1 (21) alleles and the pGAS2-HIS3 and pYJL038C-URA3 reporters (13) were previously described. Alleles of SUM1 were integrated at the SUM1 locus by replacing the sum1{Delta}::URA3 allele. Standard genetic crosses were performed to generate the desired combinations of alleles.


View this table:
[in this window]
[in a new window]

  		TABLE 2. Strains used in this study

Yeast mating and reporter assays. For quantitative mating assays, overnight cultures of cells were diluted and plated on complete (yeast extract-peptone-dextrose [YPD]) plates to determine total cell number and on minimal (yeast nitrogen base without amino acids and glucose) plates with approximately 2 x 107 mating-type tester cells (LRY1021) to determine the number of mating events. (The prototrophic diploids grow on minimal medium, but the auxotrophic haploids cannot.) The fraction of cells that mated was calculated by dividing the number of colonies on the mating plate by the number of colonies on the complete plate.

To assess the expression of the pGAS2-HIS3 reporter construct, overnight cultures of yeast were suspended in minimal medium at a concentration of 10 optical density (OD) units/ml, and 10-fold serial dilutions were prepared. Three microliters of each dilution was plated on complete (YPD) plates to verify that all the strains were similarly diluted and on plates lacking histidine and adenine to assess expression of the pGAS2-HIS3 reporter. The plates lacked adenine to ensure that the cells retained the plasmids expressing the SUM1 genes.

Genetic screen for additional SUM1-1-like mutations. The genetic screen for identifying additional mutations in SUM1 with silencing phenotypes was based on a published procedure (24). The C-terminal portion of SUM1, encoding amino acids 836 to 1062, was mutagenized by error-prone PCR using Taq polymerase, 0.5 mM MnCl2, and unbalanced deoxynucleotides (three deoxynucleoside triphosphates at 0.5 mM and one at 0.05 mM). The template for the PCR was plasmid DMC326, which contains wild-type SUM1 and a LEU2 marker, and the primers are listed in Table 3. To generate a library of mutagenized plasmids, yeast strain LRY532 (MAT{alpha} sum1{Delta} sir2{Delta}) was cotransformed with the PCR product and a gapped plasmid (pLR18) containing the SUM1 gene and an ADE2 reporter. The plasmid was cut with BsiWI and SphI, removing 312 base pairs at the 3' end of SUM1 and leaving overlaps of 380 and 230 base pairs with the mutagenized PCR product. Repair of the gapped plasmid by homologous recombination with the PCR product led to stable transformants on plates lacking adenine. Approximately 50,000 colonies were screened by replica plating to a lawn of the opposite mating type (LRY1021). Twenty-one plasmids conferred mating when retransformed into a fresh tester strain. These 21 plasmids were sequenced.


View this table:
[in this window]
[in a new window]

  		TABLE 3. Oligonucleotides used in this study

RNA isolation and quantitative reverse transcriptase PCR (RT-PCR). RNA was isolated from log-phase cells as described previously (33). For reverse transcription, cDNA was synthesized as described (13). The cDNA was quantified by real-time PCR analysis using Sybr green to detect the reaction products. The cycling parameters were denaturation at 95°C for 3 min and 42 cycles of 10 seconds at 95°C, 20 seconds at 55°C, and 2.5 min at 68°C, followed by a melting curve from 55°C to 95°C to verify that a single product was generated. The cDNA samples were compared to a standard curve prepared with genomic DNA and were normalized to a control mRNA (ACT1). Reported values represent averages of at least two independent cDNA preparations, each analyzed in two separate PCRs. Sequences of the oligonucleotides used for PCR are in Table 3.

Preparation of whole-cell protein samples. To prepare whole-cell protein samples, 0.1 volume of 100% trichloroacetic acid was added to a culture of yeast grown to late log phase, and the cells were incubated on ice for 20 min. Cells (2.5 OD equivalents) were then collected by centrifugation and suspended in 75 µl sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis sample buffer. The sample was heated for 3 min at 95°C, vortexed for 3 min with glass beads, heated again for 3 min at 95°C, and spun for 2 min in a microcentrifuge at the maximum speed. The supernatant was collected and used for immunoblot assays.

Chromatin and protein IP. Chromatin immunoprecipitations (IPs) were performed as previously described (32) using 10 OD equivalents of cells and 3 µl anti-myc tag (Upstate Biotechnology; 06-549). Cells were treated with 1% formaldehyde for 20 min (or 3 hours for Fig. 4B and C) to cross-link proteins to DNA. The DNA was sheared by sonication to an average size of 600 to 700 base pairs in all experiments. For quantitative PCR, a standard curve was prepared using input DNA. The standard curve and immunoprecipitated samples were amplified with primers for a control locus (ATS1) and the locus of interest in separate reactions, which enabled the relative amount of each locus in the IP sample to be determined. The cycling parameters were the same as described above for RT-PCR analysis. Reported values represent averages of at least two independent IPs, each analyzed in two separate PCRs. Sequences of the oligonucleotides used for PCR are in Table 3.


Figure 4
View larger version (36K):
[in this window]
[in a new window]

  		FIG. 4. Isoleucine is required at position 988 for association with ORC binding sites. (A) Association of myc-Sum1p with the E silencer at HMR. DNA coprecipitated with myc-Sum1p was quantified by RT-PCR. A strain of the genotype MAT{alpha} sir2{Delta} sum1{Delta} (LRY532) was transformed with plasmids expressing alleles of 3myc-Sum1p or untagged Sum1-T988Ip. (B) Association of myc-Sum1p with the ARS1 origin of replication. A strain of the genotype MAT{alpha} sum1{Delta} (LRY142) was transformed with plasmids expressing alleles of 3myc-Sum1p or untagged Sum1-T988Ip. (C) Association of myc-Sum1p with the ARS309 origin or replication. The same chromatin IP samples as those analyzed for panel B were examined for the relative enrichment of ARS309 compared to ATS1. (D) Two-hybrid interaction between LexA-Orc5p and GAD-Sum1(775-1062). Yeast strains (L40 and LRY1729) were transformed with plasmids expressing LexA-Orc5 (pTT93) and the GAD fused to the C terminus of Sum1p (pRH01, pRH02, pLR539, and pLR540). Yeast cells were grown on filters placed on plates selecting for the plasmids. The filters were then developed with X-Gal to detect the expression of the LacZ reporter gene.

Protein co-IP was performed as previously described (32). Cells were lysed in a buffer containing 50 mM Tris, pH 7.5, 250 mM NaCl, 5 mM EDTA, 0.1% NP-40, 1.5 mM dithiothreitol, and protease inhibitors (32). Cleared lysates were incubated with 5 µl rabbit polyclonal antibodies (Upstate Biotechnology; 06-549 and 05-902) at 4°C overnight and then 60 µl protein A agarose beads (Upstate Biotechnology) for 1 h. Samples were electrophoretically fractionated on 7.5% polyacrylamide-SDS gels, transferred to membranes, and probed using mouse monoclonal antibodies (Upstate Biotechnology; 05-724 and 05-904). To determine whether the coprecipitation of myc- and HA-tagged Sum1-T988Ip required DNA, a cell lysate was prepared in a similar buffer containing 1 mM MgCl2 and no EDTA. The lysate was incubated with 500 units DNase I (Invitrogen) and the appropriate antibody overnight. To assess whether the DNase I was active, identical reaction mixtures were set up containing 5 µg plasmid DNA, with or without DNase I. After incubation overnight, nucleic acids were isolated from the samples by phenol-chloroform extraction followed by ethanol precipitation. The entire samples were loaded on an agarose gel and visualized with ethidium bromide. In a separate experiment to assess whether chromatin was digested by the DNase I, endogenous nucleic acids were isolated from samples after overnight digestion. This isolated DNA was then diluted serially 10-fold and analyzed by PCR for the presence of two genomic loci, ACT1 and HMR.

Protein purification and gel shift assays. His-tagged proteins were expressed in Rosetta II E. coli cells (Novagen) and induced with 1 mM IPTG (isopropyl-β-D-thiogalactopyranoside) at 37° for 3 h. E. coli cells were lysed in buffer containing 300 mM NaCl, 10 mM imidazole, 50 mM sodium phosphate, pH 8.0, 0.1% Triton, 10 µM β-mercaptoethanol, and 1x BugBuster (Novagen). Soluble protein was purified from the cleared lysate over a nickel-nitrilotriacetic acid matrix (Qiagen) and eluted in 300 mM NaCl, 250 mM imidazole, 50 mM sodium phosphate, pH 8.0. Protein from the peak fractions was further purified using an S200 size-exclusion column in buffer containing 10 mM Tris, 100 mM NaCl, 1 mM EDTA, 10% glycerol using an AKTA fast protein liquid chromatography system. The peak fractions were concentrated using Amicon Ultra centrifugal filter devices (Millipore).

Gel shift conditions were based on published procedures (28, 39). The probe was generated by end labeling a single-stranded oligonucleotide corresponding to the Sum1p binding site in the SMK1 promoter (5'-GCAAGTGTCACAAATTAGTGGGTCGA). The labeling reaction mixture contained 30 pmol oligonucleotide, 100 µCi [{gamma}-32P]ATP (6,000 Ci/mmol), and 10 units T4 polynucleotide kinase (New England Biolabs) in the manufacturer's buffer in a total volume of 20 µl. After incubation for 1 hour at 37°C, the unincorporated counts were removed using Mini Quick Spin columns (Roche). A complementary oligonucleotide was added in threefold molar excess in a total volume of 50 µl, and the annealing of the two oligonucleotides was promoted by heating to 75°C for 5 minutes and then slowly cooling to room temperature. For the gel shift assay, purified proteins were diluted in 20 mM Tris, pH 7.5, 1 mM EDTA, 50 mM NaCl, 10% glycerol, 5 mM β-mercaptoethanol, 1 mM phenylmethylsulfonyl fluoride, 0.1 mg/ml bovine serum albumin to three times the desired concentration. Five microliters of diluted protein was combined with 10 µl binding mix (10 mM Tris, pH 7.5, 50 mM NaCl, 4 mM MgCl2, 6% glycerol, 20 µg/ml sonicated salmon sperm DNA, and 1.5 µl [roughly 0.9 pmol] probe) and incubated for 20 min at room temperature prior to the addition of 3 µl sucrose loading dye (40% sucrose, 0.25% bromophenol blue, 0.25% xylene cyanol). Samples were run on a 5% acrylamide gel with a cross-linking ratio of 19:1 in 0.5x Tris-borate-EDTA at 200 V at 4°C. The gel was dried and exposed to film.

Assay of β-galactosidase activity. Two-hybrid interactions were detected through activation of a LacZ reporter construct. To assay for β-galactosidase activity, yeast were grown for 18 h directly on a Magna Graph Nylon membrane (GE; NJOHY08250) placed on a YPD plate. The filter was then flash frozen in liquid nitrogen, placed onto filter paper soaked in developing solution (65 mM sodium phosphate, pH 7.0, 10 mM KCl, 1 mM MgSO4, 0.6% β-mercaptoethanol, 1% X-Gal stock [5-bromo-4-chloro-3-indolyl-β-D-galactopyranoside; 100 mg/ml in N,N-dimethyl formamide]), and developed for 2 h at 37°C in a closed container.


arrow
RESULTS
 
Only isoleucine at position 988 confers a robust silencing phenotype. The SUM1-1 mutation (T988I) has been assumed to be a neomorphic mutation rather than a loss-of-function (hypomorphic) mutation based on observations that the complete deletion of SUM1 does not confer a SUM1-1-like phenotype (6) and that the mutation is dominant (18). However, it is possible that the replacement of threonine 988 with isoleucine disrupts just one property of Sum1p and that the loss of this particular function results in a protein that silences the mating-type loci rather than the midsporulation genes. To gain further insight into the nature of the SUM1-1 mutation, a panel of additional mutations was created, in which various amino acids were placed at position 988. If the T988I mutation is indeed a neomorphic mutation, then the only other amino acids that should result in a similar phenotype are hydrophobic amino acids similar to isoleucine, such as leucine and valine. On the other hand, if the T988I mutation disrupts a property of wild-type Sum1p, then any amino acid that is not similar to threonine, such as lysine or alanine, should have a phenotype like SUM1-T988I. The conservative substitution of serine for threonine was also included in the panel of substitutions.

Expression plasmids encoding Sum1p with and without an N-terminal Myc tag were modified by site-directed mutagenesis to incorporate the selected amino acids at position 988. The mutations were confirmed by sequencing, and expression of the Myc-tagged variants was assessed by immunoblotting. All of the proteins were stably expressed at levels slightly greater than the wild-type Sum1p (Fig. 1A).


Figure 1
View larger version (34K):
[in this window]
[in a new window]

  		FIG. 1. Only isoleucine at position 988 confers a robust silencing phenotype. (A) Immunoblot of total proteins from yeast expressing Myc-tagged Sum1p bearing the indicated amino acid at position 988. Yeast of the genotype MAT{alpha} sum1{Delta} (LRY144) was transformed with plasmids expressing alleles of 3myc-Sum1p or an empty vector (pRS412). For each sample, the second lane contains four times more material than the first lane. As a loading control, the blots were also probed for 3-phosphoglycerate kinase (PGK). (B) Quantitative mating assays of yeast of the genotype MAT{alpha} sir2{Delta} sum1{Delta} (LRY532) expressing untagged Sum1p bearing the indicated amino acid at position 988. The fraction of cells that mated to a MATa tester strain (LRY1021) is indicated. (C) Quantitative RT-PCR analysis of a1 mRNA transcribed from HMR. RNA was isolated from yeast of the genotype MAT{alpha} sir2{Delta} sum1{Delta} (LRY532) expressing alleles of 3myc-Sum1p or an empty vector (pRS412). Cells were grown in minimal medium prior to RNA isolation. (D) Quantitative RT-PCR analysis of a1 mRNA transcribed from HMR. RNA was isolated from yeast of the genotype MAT{alpha} sir2{Delta} expressing alleles of 3myc-Sum1p integrated at the SUM1 locus (LRY257 and LRY2068 to LRY2073). Cells were grown in rich medium (YPD) prior to RNA isolation.

The ability of the variants of Sum1p to silence HMRa was assessed using a quantitative mating assay. When HMRa is not silenced, MAT{alpha} cells express both a and {alpha} mating-type genes and hence are unable to mate. Thus, the fraction of cells that mate reflects the fraction of cells in which HMRa is effectively silenced. MAT{alpha} yeast cells bearing sum1{Delta} and sir2{Delta} mutations were transformed with plasmids expressing SUM1 with various substitutions at position 988. Due to the deletion of SIR2, Sir-mediated silencing does not occur at HMRa, and these cells are unable to mate. If the plasmid-derived variant of Sum1p is able to silence HMRa, the cells will gain the ability to mate. As expected, when the yeast expressed Sum1-T988Ip, bearing the original SUM1-1 mutation, approximately 15% of the cells mated (Fig. 1B), indicating that silencing was restored to HMRa. Also as expected, very few cells (<0.0004%) expressing wild-type Sum1p mated, indicating that wild-type Sum1p did not restore silencing. Three of the substitutions—serine, alanine, and lysine—had phenotypes similar to that of wild-type Sum1p and did not confer silencing. The other two substitutions—valine and leucine—had intermediate phenotypes, enabling 0.08% and 0.02% of the cells to mate, respectively. Thus, only the substitution of isoleucine for threonine (the original SUM1-1 mutation) resulted in a robust silencing phenotype, whereas variants of Sum1p with the amino acids most similar to isoleucine—leucine and valine—had modest silencing phenotypes.

To examine transcription at HMRa directly, the amount of a1 mRNA was measured by quantitative RT-PCR. This analysis was performed using cells in which alleles of SUM1 were on plasmids (Fig. 1C) or integrated into the genome (Fig. 1D). Consistent with the mating assay, expression of Sum1-T988Ip resulted in the repression of a1 transcription, compared to cells expressing either no Sum1p or wild-type Sum1p. Sum1-T988Vp also resulted in significant repression, although not quite as much as Sum1-T988Ip. Modest repression was also noted with Sum1-T988Lp. Apparently, in the cases of the valine and leucine mutations, the reduction in a1 mRNA was not sufficient in most cells to allow mating. The serine, alanine, and lysine mutations had even lower levels of repression. Therefore, the T988I mutation is a neomorphic mutation, conferring a new property on the protein.

The T988I mutation is unique. Of the amino acids tested at position 988, isoleucine was the only one that enabled Sum1p to silence HMRa sufficiently for mating. Furthermore, given that the amino acids that were not tested are less similar to isoleucine than are valine and leucine, which achieved only modest silencing, there are probably not other amino acids that would confer silencing when placed at position 988. However, it is possible that substitutions at other positions would have a similar phenotype. For example, if T988I enhances the affinity of Sum1p for another protein, other substitutions in the same region of Sum1p might also improve the affinity of Sum1p for that protein. To identify such mutations, a genetic screen was conducted. Single nucleotide substitutions in the 3' portion of wild-type SUM1 were generated by error-prone PCR (24) and transformed into a sir2{Delta} strain, which was initially unable to mate due to the absence of Sir2p. Approximately 50,000 transformants were screened, and ultimately 21 plasmids that restored mating were isolated and sequenced. All 21 of these plasmids contained the T988I mutation, indicating that this mutation is most likely the only single-nucleotide substitution in this portion of SUM1 that can confer a silencing phenotype.

Many of the sequenced plasmids were clearly independently derived, because 10 of the plasmids contained mutations in addition to T988I (Table 4). To determine whether these additional mutations contributed to silencing, the mating efficiencies of yeast expressing Sum1p with these additional mutations were examined. In no case was the mating efficiency greater than that of T988I. Therefore, these additional mutations probably do not contribute positively to the mating phenotype. In fact, one of the mutations, a premature stop codon at position 1021, resulted in an extremely low mating efficiency (Table 4). The isolation of this plasmid strengthens, in two ways, the conclusion that T988I is a gain-of-function mutation. First, the portion of Sum1p that is missing in this plasmid must contribute positively to the silencing phenotype, and therefore, the silencing phenotype is not caused by inactivating this domain. Second, the fact that this plasmid was recovered from the screen indicates that if other mutations with similarly weak mating abilities existed, they too should have been recovered. The failure to identify any other mutations with a SUM1-1-like phenotype strengthens the conclusion that SUM1-T988I is a gain-of-function mutation, as gain-of-function mutations are more rare than loss-of-function mutations.


View this table:
[in this window]
[in a new window]

  		TABLE 4. Mutations of SUM1 that suppress the mating defect of sir2{Delta} cellsa

Threonine is required at position 988 for repression of midsporulation genes. In addition to enabling Sum1p to silence the mating-type loci, the SUM1-T988I mutation reduces the ability of Sum1p to repress midsporulation genes (32). This decreased repression could result from the protein being drawn away from midsporulation genes to the mating-type loci. Alternatively, threonine 988 may be critical for repression. To distinguish these two possibilities, we examined the repressive potential of the panel of substitutions at position 988. If the loss of repression results from a competition between the mating-type loci and midsporulation genes, then only the isoleucine substitution should result in reduced repression because it is the only substitution which confers robust silencing of HMRa. In contrast, if threonine is important for wild-type function, then all of the substitutions, with the possible exception of serine, should perturb function.

The repressive ability of Sum1p was assessed in two ways. First, the expression of a HIS3 reporter gene driven by the GAS2 promoter, which is repressed by Sum1p, was evaluated by growth on plates lacking histidine (Fig. 2A). In the presence of the wild-type Sum1p (second row) no cells were able to grow in the absence of histidine, indicating that the HIS3 reporter was completely repressed. In contrast, replacement of threonine 988 with other amino acids resulted in the derepression of the HIS3 reporter, indicating that threonine 988 was crucial for function. Consistent with the properties of threonine being important for function, the placement of serine at position 988 was the least disruptive for function.


Figure 2
View larger version (31K):
[in this window]
[in a new window]

  		FIG. 2. Threonine is required at position 988 for repression of midsporulation genes. (A) Expression of pGAS2-HIS3 reporter in yeast of the genotype MAT{alpha} sum1{Delta} pGAS2-HIS3 (LRY1386) expressing untagged Sum1p bearing the indicated amino acid at position 988 or an empty vector (pRS412). Tenfold serial dilutions were spotted on minimal medium lacking histidine and adenine (-His) or complete medium (Growth). The lack of adenine ensures that the cells retain the plasmids expressing the SUM1 genes. (B) Quantitative RT-PCR analysis of SMK1 mRNA from the same yeast as that described for panel A. (C) Quantitative RT-PCR analysis of SMK1 mRNA from the yeast of the genotype MAT{alpha} sir2{Delta} expressing alleles of 3myc-Sum1p integrated at the SUM1 locus (LRY257 and LRY2068 to LRY2073). (D) Quantitative RT-PCR analysis of SMK1 mRNA from yeast expressing myc-Sum1-T988Ip (LRY459), Sum1p (LRY1295), or myc-Sum1-T988Ip and Sum1p (LRY1260).

A second approach to assessing the repressive potential of the Sum1p variants was to measure the amount of steady-state mRNA from another Sum1p-repressed promoter, SMK1, using quantitative RT-PCR. RNA was isolated from cells in which alleles of SUM1 were on plasmids (Fig. 2B) or integrated into the genome (Fig. 2C). As with the pGAS2-HIS3 reporter, the wild-type Sum1p bearing threonine at position 988 achieved the most repression. Serine, which is similar to threonine, had modest activity but clearly resulted in less repression than threonine. These results indicate that threonine 988 is crucial for function. It is important to note that some of the substitutions, such as lysine and alanine, resulted in proteins that had neither the wild-type function (Fig. 2) nor the mutant function (Fig. 1). Therefore, the SUM1-T988I mutation results in both the loss of the wild-type function and the gain of a new function.

SUM1-1 is classified as a dominant allele with respect to the mutant function, i.e., repression of a1 at HMR (6, 18). However, the results above suggest that with respect to the wild-type function, SUM1-T988I is a loss-of-function allele and likely to be recessive. This prediction is, indeed, borne out experimentally. Coexpression of Sum1-T988Ip along with wild-type Sum1p did not disrupt repression of SMK1 (Fig. 2D).

Threonine 988 contributes to DNA binding. There are several events which must occur for Sum1p to repress a gene, including binding of Sum1p to the promoter, recruitment of the deacetylase Hst1p, and additional, poorly defined downstream events. It is not likely that the loss of threonine 988 affects the ability of Sum1p to interact with Hst1p, because Sum1-T988Ip interacts with Hst1p as efficiently as wild-type Sum1p does (32). Therefore, we investigated whether substitutions at position 988 affect the association of Sum1p with the promoter. To do so, chromatin IPs were performed to examine the association of myc-tagged Sum1 proteins with the promoters of the GAS2 and SMK1 genes. If threonine 988 plays a role in the association of Sum1p with the promoter, then in the absence of threonine at position 988 the association of Sum1p with the repressed promoters should be reduced. On the other hand, if threonine 988 contributes to a later step in the process of repression, then the substituted proteins should still be associated with the repressed promoters. At both promoters, the association of wild-type Sum1p was significantly greater than the associations of any of the substituted Sum1 proteins (Fig. 3A and B). Therefore, threonine 988 contributes to the stability of the protein-DNA complex.


Figure 3
View larger version (37K):
[in this window]
[in a new window]

  		FIG. 3. Threonine 988 contributes to DNA binding. (A) Association of myc-Sum1p with the SMK1 promoter. DNA coprecipitated with myc-Sum1p was quantified by RT-PCR. The y axis represents the relative enrichment of the SMK1 promoter compared to the ATS1 promoter, which is not regulated by Sum1p or Sum1-T988Ip. If the recovery of a particular region is not enhanced compared to ATS1, the relative enrichment is 1. A strain of the genotype MAT{alpha} sum1{Delta} (LRY142) was transformed with plasmids expressing alleles of 3myc-Sum1p or untagged Sum1p. (B) Association of myc-Sum1p with the GAS2 promoter. The same chromatin IP samples as those analyzed for panel A were examined for the relative enrichment of GAS2 compared to ATS1. (C) DNA gel shift assay using recombinant proteins corresponding to the C-terminal half of Sum1p (amino acids 523 to 1062). Approximately 0.9 pmol of a radiolabeled double-stranded oligonucleotide probe corresponding to the Sum1p binding site from the SMK1 promoter was incubated with 16, 12, or 8 pmol protein in a volume of 15 µl. (D) Twenty picomoles of the recombinant proteins used for panel C was separated on a denaturing acrylamide gel and stained with Coomassie blue.

One way in which threonine 988 could contribute to the association of Sum1p with target promoters is by enhancing the affinity of Sum1p for its DNA-binding sites. In fact, Sum1-T988Ip was found to bind DNA less well than wild-type Sum1p (M. Pierce and A. Vershon, personal communication). To further examine the importance of threonine 988 in DNA binding, gel shift experiments were performed. A double-stranded oligonucleotide probe corresponding to the Sum1p binding site at the SMK1 promoter (39) was radiolabeled and incubated with a recombinant fragment of the Sum1 protein that binds specifically to midsporulation elements (28). When equal amounts of protein were used, a greater fraction of the probe was shifted by the wild-type protein than by any of the substituted proteins (Fig. 3C). Examination of the purified proteins by SDS-polyacrylamide gel electrophoresis revealed that all of the proteins were intact and used in equal amounts (Fig. 3D). In addition, the relative affinities of the substituted proteins were confirmed using several different preparations of proteins (data not shown). Thus, the presence of threonine at position 988 increased the affinity of the protein for its DNA-binding site.

Isoleucine is required at position 988 for association with ORC binding sites. The results outlined above reveal that the association of Sum1p with midsporulation genes is reduced in the absence of threonine at position 988. It is possible that when the protein is no longer associated with its normal binding sites it accumulates at other sites in the genome. One potential binding site for these mutant Sum1 proteins could be ORC, which has been suggested to have some affinity for wild-type Sum1p (15, 21). In this scenario, any substitution of threonine 988 which reduces the association of the protein with midsporulation promoters would result in the accumulation of the protein at ORC binding sites, as has been reported for Sum1-T988Ip. Alternatively, the function gained in the presence of isoleucine at position 988 could be required for the association of Sum1p with ORC. For example, the T988I mutation could increase the affinity of the protein for ORC itself (35) or for another protein that is important to anchor the protein at ORC binding sites. In such a case, only the T988I mutation should result in the accumulation of the protein with ORC binding sites. To distinguish between these two models, the association of Sum1p with the E silencer at HMRa and two other ORC binding sites (ARS1 and ARS309) was examined by chromatin IP (Fig. 4A to C). As previously observed (21), Sum1-T988Ip associates with all three of these loci, whereas wild-type Sum1p does not. The valine and leucine substitutions were modestly associated with HMRa, but the other substitutions were not. Furthermore, none of the substituted proteins other than T988I was associated with the ARS1 and ARS309 ORC binding sites. These results indicate that the association of Sum1p-T988Ip with ORC binding sites is the result of a new property conferred by the T988I mutation and is not due to the loss of wild-type function that occurs in the absence of threonine 988.

It has been proposed that the SUM1-1 mutation increases the affinity of Sum1p for ORC based on a two-hybrid interaction between Orc5p and the C-terminal portion of Sum1-1p (positions 775 to 1062) but not wild-type Sum1p (35). If isoleucine does indeed increase the affinity of Sum1p for ORC, the two-hybrid interaction should occur specifically when isoleucine is present. To test this model, the two-hybrid assay was repeated using plasmids expressing Sum1775-1062 fused to the Gal4 activation domain (GAD) and containing threonine, isoleucine, alanine, or valine at position 988. The reporter strain contained a LacZ gene with LexA binding sites in its promoter and a LexA-Orc5 fusion protein. As previously reported, LacZ was activated when the Sum1-GAD fusion contained isoleucine but not threonine at position 988 (Fig. 4D). LacZ was not activated when alanine was at position 988, consistent with the inability of this substitution to accumulate at ORC binding sites. However, activation of LacZ did occur when valine was at position 988, suggesting that the interaction between this variant of Sum1p and ORC is more stable than that of the threonine and alanine variants. This result is consistent with the moderate ability of Sum1-T988V to confer silencing (Fig. 1).

The T988I mutation results in increased self-association. As shown above, the T988I mutation contributes to the silencing phenotype in two ways. First, it results in reduced association of the protein with its native binding sites, freeing it to act elsewhere. Second, the presence of isoleucine creates a new property which both causes the protein to accumulate at ORC binding sites (Fig. 4) and enables the protein to spread along the chromosome (21). It has been suggested that the new property is an increased affinity for ORC itself (35). However, it is not immediately apparent how an increased affinity for ORC would lead to the ability to spread along the chromosome. An alternative model is that both wild-type and mutant Sum1p have a low affinity for ORC, and the T988I mutation causes the protein to accumulate at ORC binding sites by increasing the affinity of the protein for another component of chromatin, thereby providing a second anchor in addition to ORC. Such a stabilizing interaction with chromatin could also enable the protein to spread. One way in which the T988I mutation could stabilize Sum1p at ORC binding sites and increase the ability of Sum1p to spread is by increasing the affinity of Sum1p for itself.

To determine whether Sum1-T988Ip interacts with itself, a coprecipitation experiment was conducted from yeast expressing Sum1p with myc and HA tags. When HA-Sum1-T988Ip was precipitated, myc-Sum1-T988Ip was coprecipitated. Similarly, when myc-Sum1-T988Ip was precipitated, HA-Sum1-T988Ip coprecipitated (Fig. 5A and B, lanes 4). In contrast, the wild-type Sum1p did not self-associate (lanes 1), nor did wild-type Sum1p associate with Sum1-T988Ip (lanes 2 and 3). Thus, the T988I mutation did increase the self-association of Sum1p.


Figure 5
View larger version (32K):
[in this window]
[in a new window]

  		FIG. 5. Sum1-T988Ip self-associates. (A) HA-Sum1p was immunoprecipitated from yeast of the genotype MAT{alpha} sum1{Delta} (LRY144) transformed with plasmids expressing HA- and myc-tagged Sum1p with threonine or isoleucine at position 988, as indicated. (B) myc-Sum1p was immunoprecipitated from yeast of the genotype MAT{alpha} sum1{Delta} (LRY144) transformed with plasmids expressing myc- and HA-tagged Sum1p with threonine or isoleucine at position 988, as indicated. (C) Lysates from the yeast used for panel B, lane 4, were treated with DNase I (lane 2) or left untreated (lane 1) during the IP of myc-Sum1-1p. The precipitated material was then evaluated by immunoblotting for the presence of HA-Sum1-1p. (D) The same lysates as those used in the upper portion of panel C were spiked with 5 µg plasmid DNA. Total DNA was extracted from the lysate, separated on an agarose gel, and stained with ethidium bromide. (E) DNA was extracted from the same lysates as those used in the lower portion of panel C. A 10-fold serial dilution of this DNA was analyzed by PCR amplification of HMR or ACT1. (F) myc-Sum1p was immunoprecipitated from yeast of the genotype MAT{alpha} sum1{Delta} (LRY144) transformed with plasmids expressing myc- and HA-tagged Sum1p with threonine, isoleucine, alanine, or valine at position 988, as indicated.

The observed self-association of Sum1-T988Ip could simply reflect the precipitation of the repressive chromatin structure. If this were the case, the interaction would involve DNA. To test this possibility, we determined whether the coprecipitation still occurred after the lysate was incubated with DNase I. No decrease in coprecipitation was observed after treatment with DNase I (Fig. 5C). Controls indicated that the DNase I was active because a plasmid added to the lysate was digested (Fig. 5D) and DNA corresponding to two genomic loci was greatly reduced (Fig. 5E). These data are consistent with the self-association of Sum1-T988Ip being independent of DNA.

If the observed self-association is related to the ability to spread, it might also occur with Sum1-T988V, which causes modest repression, but not with Sum1-T988A, which does not cause repression. To test this prediction, the coprecipitation assay was repeated with tagged proteins in which alanine or valine was incorporated at position 988 (Fig. 5F). Coprecipitation did not occur when alanine was present, consistent with the inability of this variant to spread. In contrast, the two tagged versions of Sum1-T988V did coprecipitate.

The results presented in Fig. 4 and 5 are consistent with the SUM1-1 mutation increasing the affinity of the protein both for itself and for ORC. Alternatively, one of these interactions could be dependent on the other. It seems unlikely that ORC facilitates the self-association of Sum1-1p because the interaction between ORC and Sum1-1p appears less stable than the self-association of Sum1-1p. We have been unable to coprecipitate Sum1-1p with ORC (data not shown), and a published report of such a coprecipitation indicates that Orc5p and Sum1-1p coprecipitated only when the silenced HMR locus was present on a multicopy plasmid (35), implying that the coprecipitation was mediated by DNA. To investigate whether self-association facilitates the interaction of Sum1-1p with ORC, the two-hybrid assay was conducted in a reporter yeast strain in which SUM1 had been deleted. We reasoned that if self-association was involved in the two-hybrid interaction, it would be likely that either the ORC-interacting domain of Sum1p or the second site necessary for self-association would not be present in the GAD-Sum1-1(786-1062) construct, and therefore, the two-hybrid interaction would no longer occur in a sum1{Delta} strain. Nevertheless, even in the absence of full-length Sum1p, the mutant GAD-Sum1-1 construct interacted more strongly with LexA-Orc5 than did wild-type GAD-Sum1 (Fig. 4B). This result does not eliminate the possibility that the interaction between the ORC and Sum1-1p is mediated by the self-association of Sum1-1p, but it does indicate that, if it is, the C-terminal portion of Sum1 (positions 788 to 1062) is sufficient for both interactions.


arrow
DISCUSSION
 
A single amino acid substitution results in both a loss and a gain of function. In this study, we found that the substitution of isoleucine for threonine 988 of Sum1p results in both the loss of wild-type function and the gain of a new function. Any substitution at position 988 other than serine, the amino acid most similar to threonine, reduced the ability of Sum1p to repress the GAS2 and SMK1 promoters and also reduced the association of Sum1p with these promoters. Furthermore, these mutations reduced the affinity of the protein for its specific DNA-binding site in a gel shift assay. Therefore, threonine is essential for wild-type function. In keeping with this conclusion, threonine 988 is perfectly conserved in all identifiable homologs of Sum1p and lies in a highly conserved portion of the protein. When this conserved domain of Sum1p from S. cerevisiae (positions 813 to 1054) was compared with the three most distantly related identifiable homologs (from Kluyveromyces lactis, Ashbya gossypii, and Kluyveromyces waltii) using FASTA pairwise alignment (26), the average percent identity was 68% and the average similarity was 90%. In contrast, comparison of the full-length proteins yielded an average identity of 35% and similarity of 61%. Thus, the C-terminal portion of the protein is particularly well conserved. A reasonable hypothesis, given that mutation of T988 reduces the association of the protein with DNA, is that this conserved domain participates in binding DNA.

In the absence of threonine at position 988, the Sum1 protein no longer associates with promoters and is consequently released to associate with other sites. This reduced affinity for the usual binding site may be important for enabling Sum1-1p to relocalize to the mating-type loci. However, loss of threonine 988 alone was not sufficient for Sum1p to associate with ORC binding sites and to spread. Instead, the association with ORC binding sites was observed only with isoleucine at position 988, indicating that a new function is created by the presence of isoleucine. Two-hybrid and chromatin IP data suggest that the mutation increases the affinity of the protein for ORC. The dual nature of the SUM1-1 mutation (loss and gain of function) may explain why only mutations at T988 were recovered in the genetic screen for additional mutations like SUM1-1.

It should be noted that both threonine and isoleucine at position 988 enable Sum1p to form intermolecular interactions, suggesting that this position is on the surface of the protein. The structure of Sum1p is unknown. However, the sequence around position 988 is predicted to form an amphipathic alpha-helix by top-performing servers that predict secondary structure (16, 23, 25, 34). The minimum region that is predicted to form an alpha-helix by all programs utilized spans from S981 to Y996, placing threonine 988 at the center of the hydrophilic face. If such a helix does form, the hydrophobic side most likely faces inward, whereas the hydrophilic side is exposed to the solvent and could interact with other proteins or DNA. When isoleucine, a hydrophobic amino acid, is substituted for threonine, a hydrophilic amino acid, the face of this proposed alpha-helix is clearly altered, potentially disrupting interactions normally formed by the helix. However, this substitution probably does not disrupt the formation of the alpha-helix itself, as the helix is still predicted to form when isoleucine is at position 988.

These observations suggest a model to explain how the SUM1-1 mutation alters the properties of the Sum1 protein. We propose that the structure of Sum1p is organized such that T988 is part of a domain that is presented on the surface of the protein and is well positioned to form intermolecular interactions. In the wild-type protein, this "interacting domain" may play a role in DNA binding. The mutation of threonine to isoleucine alters the molecular properties of this domain such that its affinity for the original interacting partners is reduced and, simultaneously, the affinity for new partners, such as ORC and Sum1p itself, is increased. Consistent with this model, a fortuitous mutation revealed that when M992, which is predicted to be next to T988 on the hydrophilic face, is also mutated to isoleucine, Sum1-1p-mediated silencing is disrupted (Edward Winter, personal communication).

Implications for the evolution of new function. Studies of the mechanisms by which new genes originate (1, 19) have focused primarily on the source of new genes. For example, gene duplication, retrotransposition, exon shuffling, and horizontal transfer have all been documented to give rise to new genes. However, less attention has been given to the issue of how new function evolves once an "extra" copy of a gene is available as a substrate for evolutionary tinkering. Clearly, just as there are multiple ways in which additional copies of genes arise, there are multiple ways in which amino acid substitutions lead to novel function. The model proposed above to explain the effect of the SUM1-1 mutation can be generalized and could be one mechanism by which neomorphic mutations act. Specifically, interaction domains could be hotspots for neomorphic mutations because mutations in these locations have the potential to redirect the protein to a new set of binding partners, potentially leading to altered function. Because this mechanism involves loss as well as gain of function, it would be most likely to occur in the context of gene duplication or another event that generated an "extra" copy of a gene.

In principle there are three stages in the process of neofunctionalization. The first two stages, outlined above, involve (i) generating an "extra" copy of a gene that can be a substrate for further mutagenesis without resulting in loss of an existing gene function and (ii) the acquisition of a novel function, potentially through a neomorphic mutation. It is likely that, although a single amino acid change can create new function, the protein will not perform this new function optimally. Therefore, the final stage involves (iii) the refinement of the new gene through positive selection for improved variants, resulting in a period of accelerated evolution. Due to this process of refinement, even young genes of novel function will have multiple mutations that contribute positively to the new function, and it may be difficult to identify single mutations in naturally occurring genes that represent the key event in the neofunctionalization process. Consequently, laboratory-generated neomorphic mutations, such as SUM1-1, should prove useful in elucidating the types of mechanisms that enable small genetic changes to give rise to significant functional changes.

Role of self-association in facilitating spreading. The spreading of silenced chromatin is thought to occur by a sequential modification mechanism (12, 31). The key events in this process are (i) the posttranslational modification of a nucleosome, (ii) the binding of a chromatin protein to the modified nucleosome, and (iii) the recruitment by the chromatin protein of an enzyme that will catalyze the same modification on the next nucleosome. In this model, the two features that a repressive complex must possess to spread are (i) an enzyme to modify nucleosomes and (ii) a protein domain that binds to nucleosomes modified by the enzyme. These two features are most likely present in the wild-type Sum1 complex as well as other repressive complexes that do not spread. For example, the yeast corepressor Tup1p and mammalian corepressor SMRT/N-CoR interact with histones but do not spread (8, 40). The presence of a histone-binding domain in these nonspreading repressive complexes raises a paradox. Why don't they spread? The observation that the SUM1-1 mutation results in the self-association of Sum1p suggests that it may be the ability to self-associate that facilitates spreading.

The most likely way in which self-association contributes to the spreading process is by stabilizing the extended chromatin structure. Measured affinities between histone binding domains and optimally modified histone peptides are modest (low µM range) (5, 9) and inadequate to be the sole tether retaining chromatin proteins on a chromosome. Therefore, other stabilizing interactions, such as self-association, must exist to enable an extended chromatin structure to persist. Consistent with this model, the Sir proteins from S. cerevisiae and HP1 from mouse and Schizosaccharomyces pombe, proteins which are known to form extended regions of repressive chromatin, all self-associate (4, 7, 29).


arrow
ACKNOWLEDGMENTS
 
We thank Johannes Rudolph for advice and facilities for protein purification; Andrew Vershon, Michael Pierce, and Edward Winter for communicating unpublished results; David Shore and Rolf Sternglanz for plasmids and yeast strains; Ashley Bushey and Valerie Vaughn for technical assistance; and Patrick Lynch, Kristin Scott, and Joshua Warren for critical reading of the manuscript.

This work was supported by a grant from the National Institutes of Health (GM073991).


arrow
FOOTNOTES
 
* Corresponding author. Mailing address: Institute for Genome Sciences and Policy, Box 3382, Duke University, Durham, NC 27710. Phone: (919) 684-0354. Fax: (919) 668-0795. E-mail: lrusche{at}biochem.duke.edu Back

{triangledown} Published ahead of print on 11 February 2008. Back

{dagger} Present address: College of Osteopathic Medicine of the Pacific, Western University of Health Sciences, Pomona, CA 91766. Back


arrow
REFERENCES
 
    1
  1. Babushok, D. V., E. M. Ostertag, and H. H. Kazazian, Jr. 2007. Current topics in genome evolution: molecular mechanisms of new gene formation. Cell. Mol. Life Sci. 64:542-554.[CrossRef][Medline]
    2. 2
  2. Bedalov, A., M. Hirao, J. Posakony, M. Nelson, and J. A. Simon. 2003. NAD+-dependent deacetylase Hst1p controls biosynthesis and cellular NAD+ levels in Saccharomyces cerevisiae. Mol. Cell. Biol. 23:7044-7054.[Abstract/Free Full Text]
    3. 3
  3. Braman, J., C. Papworth, and A. Greener. 1996. Site-directed mutagenesis using double-stranded plasmid DNA templates. Methods Mol. Biol. 57:31-44.[Medline]
    4. 4
  4. Brasher, S. V., B. O. Smith, R. H. Fogh, D. Nietlispach, A. Thiru, P. R. Nielsen, R. W. Broadhurst, L. J. Ball, N. V. Murzina, and E. D. Laue. 2000. The structure of mouse HP1 suggests a unique mode of single peptide recognition by the shadow chromo domain dimer. EMBO J. 19:1587-1597.[CrossRef][Medline]
    5. 5
  5. Carmen, A. A., L. Milne, and M. Grunstein. 2002. Acetylation of the yeast histone H4 N terminus regulates its binding to heterochromatin protein SIR3. J. Biol. Chem. 277:4778-4781.[Abstract/Free Full Text]
    6. 6
  6. Chi, M. H., and D. Shore. 1996. SUM1-1, a dominant suppressor of SIR mutations in Saccharomyces cerevisiae, increases transcriptional silencing at telomeres and HM mating-type loci and decreases chromosome stability. Mol. Cell. Biol. 16:4281-4294.[Abstract]
    7. 7
  7. Cowieson, N. P., J. F. Partridge, R. C. Allshire, and P. J. McLaughlin. 2000. Dimerisation of a chromo shadow domain and distinctions from the chromodomain as revealed by structural analysis. Curr. Biol. 10:517-525.[CrossRef][Medline]
    8. 8
  8. Davie, J. K., R. J. Trumbly, and S. Y. Dent. 2002. Histone-dependent association of Tup1-Ssn6 with repressed genes in vivo. Mol. Cell. Biol. 22:693-703.[Abstract/Free Full Text]
    9. 9
  9. Fischle, W., Y. Wang, S. A. Jacobs, Y. Kim, C. D. Allis, and S. Khorasanizadeh. 2003. Molecular basis for the discrimination of repressive methyl-lysine marks in histone H3 by Polycomb and HP1 chromodomains. Genes Dev. 17:1870-1881.[Abstract/Free Full Text]
    10. 10
  10. Golding, G. B., and A. M. Dean. 1998. The structural basis of molecular adaptation. Mol. Biol. Evol. 15:355-369.[Abstract]
    11. 11
  11. Grewal, S. I., and S. Jia. 2007. Heterochromatin revisited. Nat. Rev. Genet. 8:35-46.[CrossRef][Medline]
    12. 12
  12. Grewal, S. I., and J. C. Rice. 2004. Regulation of heterochromatin by histone methylation and small RNAs. Curr. Opin. Cell Biol. 16:230-238.[CrossRef][Medline]
    13. 13
  13. Hickman, M. A., and L. N. Rusche. 2007. Substitution as a mechanism for genetic robustness: the duplicated deacetylases Hst1p and Sir2p in Saccharomyces cerevisiae. PLoS Genet. 3:e126.[CrossRef][Medline]
    14. 14
  14. Hoppe, G. J., J. C. Tanny, A. D. Rudner, S. A. Gerber, S. Danaie, S. P. Gygi, and D. Moazed. 2002. Steps in assembly of silent chromatin in yeast: Sir3-independent binding of a Sir2/Sir4 complex to silencers and role for Sir2-dependent deacetylation. Mol. Cell. Biol. 22:4167-4180.[Abstract/Free Full Text]
    15. 15
  15. Irlbacher, H., J. Franke, T. Manke, M. Vingron, and A. E. Ehrenhofer-Murray. 2005. Control of replication initiation and heterochromatin formation in Saccharomyces cerevisiae by a regulator of meiotic gene expression. Genes Dev. 19:1811-1822.[Abstract/Free Full Text]
    16. 16
  16. Karplus, K., R. Karchin, J. Draper, J. Casper, Y. Mandel-Gutfreund, M. Diekhans, and R. Hughey. 2003. Combining local-structure, fold-recognition, and new fold methods for protein structure prediction. Proteins 53(Suppl. 6):491-496.[CrossRef][Medline]
    17. 17
  17. Klar, A. J., S. N. Kakar, J. M. Ivy, J. B. Hicks, G. P. Livi, and L. M. Miglio. 1985. SUM1, an apparent positive regulator of the cryptic mating-type loci in Saccharomyces cerevisiae. Genetics 111:745-758.[Abstract/Free Full Text]
    18. 18
  18. Laurenson, P., and J. Rine. 1991. SUM1-1: a suppressor of silencing defects in Saccharomyces cerevisiae. Genetics 129:685-696.[Abstract]
    19. 19
  19. Long, M., E. Betran, K. Thornton, and W. Wang. 2003. The origin of new genes: glimpses from the young and old. Nat. Rev. Genet. 4:865-875.[Medline]
    20. 20
  20. Luo, K., M. A. Vega-Palas, and M. Grunstein. 2002. Rap1-Sir4 binding independent of other Sir, yKu, or histone interactions initiates the assembly of telomeric heterochromatin in yeast. Genes Dev. 16:1528-1539.[Abstract/Free Full Text]
    21. 21
  21. Lynch, P. J., H. B. Fraser, E. Sevastopoulos, J. Rine, and L. N. Rusche. 2005. Sum1p, the origin recognition complex, and the spreading of a promoter-specific repressor in Saccharomyces cerevisiae. Mol. Cell. Biol. 25:5920-5932.[Abstract/Free Full Text]
    22. 22
  22. McCord, R., M. Pierce, J. Xie, S. Wonkatal, C. Mickel, and A. K. Vershon. 2003. Rfm1, a novel tethering factor required to recruit the Hst1 histone deacetylase for repression of middle sporulation genes. Mol. Cell. Biol. 23:2009-2016.[Abstract/Free Full Text]
    23. 23
  23. McGuffin, L. J., K. Bryson, and D. T. Jones. 2000. The PSIPRED protein structure prediction server. Bioinformatics 16:404-405.[Abstract/Free Full Text]
    24. 24
  24. Muhlrad, D., R. Hunter, and R. Parker. 1992. A rapid method for localized mutagenesis of yeast genes. Yeast 8:79-82.[CrossRef][Medline]
    25. 25
  25. Ouali, M., and R. D. King. 2000. Cascaded multiple classifiers for secondary structure prediction. Protein Sci. 9:1162-1176.[Medline]
    26. 26
  26. Pearson, W. R., and D. J. Lipman. 1988. Improved tools for biological sequence comparison. Proc. Natl. Acad. Sci. USA 85:2444-2448.[Abstract/Free Full Text]
    27. 27
  27. Perutz, M. F. 1983. Species adaptation in a protein molecule. Mol. Biol. Evol. 1:1-28.[Abstract]
    28. 28
  28. Pierce, M., K. R. Benjamin, S. P. Montano, M. M. Georgiadis, E. Winter, and A. K. Vershon. 2003. Sum1 and Ndt80 proteins compete for binding to middle sporulation element sequences that control meiotic gene expression. Mol. Cell. Biol. 23:4814-4825.[Abstract/Free Full Text]
    29. 29
  29. Rudner, A. D., B. E. Hall, T. Ellenberger, and D. Moazed. 2005. A nonhistone protein-protein interaction required for assembly of the SIR complex and silent chromatin. Mol. Cell. Biol. 25:4514-4528.[Abstract/Free Full Text]
    30. 30
  30. Rusche, L. N., A. L. Kirchmaier, and J. Rine. 2003. The establishment, inheritance, and function of silenced chromatin in Saccharomyces cerevisiae. Annu. Rev. Biochem. 72:481-516.[CrossRef][Medline]
    31. 31
  31. Rusche, L. N., A. L. Kirchmaier, and J. Rine. 2002. Ordered nucleation and spreading of silenced chromatin in Saccharomyces cerevisiae. Mol. Biol. Cell 13:2207-2222.[Abstract/Free Full Text]
    32. 32
  32. Rusche, L. N., and J. Rine. 2001. Conversion of a gene-specific repressor to a regional silencer. Genes Dev. 15:955-967.[Abstract/Free Full Text]
    33. 33
  33. Schmitt, M. E., T. A. Brown, and B. L. Trumpower. 1990. A rapid and simple method for preparation of RNA from Saccharomyces cerevisiae. Nucleic Acids Res. 18:3091-3092.[Free Full Text]
    34. 34
  34. Simossis, V. A., and J. Heringa. 2004. Integrating protein secondary structure prediction and multiple sequence alignment. Curr. Protein Pept. Sci. 5:249-266.[CrossRef][Medline]
    35. 35
  35. Sutton, A., R. C. Heller, J. Landry, J. S. Choy, A. Sirko, and R. Sternglanz. 2001. A novel form of transcriptional silencing by Sum1-1 requires Hst1 and the origin recognition complex. Mol. Cell. Biol. 21:3514-3522.[Abstract/Free Full Text]
    36. 36
  36. Talbert, P. B., and S. Henikoff. 2006. Spreading of silent chromatin: inaction at a distance. Nat. Rev. Genet. 7:793-803.[Medline]
    37. 37
  37. Valenzuela, L., and R. T. Kamakaka. 2006. Chromatin insulators. Annu. Rev. Genet. 40:107-138.[CrossRef][Medline]
    38. 38
  38. van Leeuwen, F., and D. E. Gottschling. 2002. Genome-wide histone modifications: gaining specificity by preventing promiscuity. Curr. Opin. Cell Biol. 14:756-762.[CrossRef][Medline]
    39. 39
  39. Xie, J., M. Pierce, V. Gailus-Durner, M. Wagner, E. Winter, and A. K. Vershon. 1999. Sum1 and Hst1 repress middle sporulation-specific gene expression during mitosis in Saccharomyces cerevisiae. EMBO J. 18:6448-6454.[CrossRef][Medline]
    40. 40
  40. Yoon, H. G., Y. Choi, P. A. Cole, and J. Wong. 2005. Reading and function of a histone code involved in targeting corepressor complexes for repression. Mol. Cell. Biol. 25:324-335.[Abstract/Free Full Text]

Molecular and Cellular Biology, April 2008, p. 2567-2578, Vol. 28, No. 8
0270-7306/08/$08.00+0     doi:10.1128/MCB.01785-07
Copyright © 2008, American Society for Microbiology. All Rights Reserved.




This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowReprints and Permissions
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Safi, A.
Right arrow Articles by Rusche, L. N.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Safi, A.
Right arrow Articles by Rusche, L. N.

Copyright © 2009 by the American Society for Microbiology.   For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»