![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
Department of Cell Biology,1 Department of Biological Chemistry and Molecular Pharmacology, Harvard Medical School, 240 Longwood Ave., Boston, Massachusetts 021152
Received 25 August 2004/ Returned for modification 21 October 2004/ Accepted 22 February 2005
 |
ABSTRACT |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
INTRODUCTION |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The assembly and inheritance of heterochromatin are thought to be governed primarily by changes in histone modifications (21, 44). The assembly of silent chromatin in budding yeast requires histones H3 and H4. Deletion of the N-terminal tail of histone H3 or H4 compromises silencing at both the HM loci and telomeres (22, 25, 64). Mutation of lysine 16 of histone H4, a major site of acetylation in budding yeast, to glutamine or glycine also causes severe defects in silencing (23, 33).
In addition to histones, the protein deacetylase Sir2 and the histone binding proteins Sir3 and Sir4 are required to assemble silent chromatin. Sir2 is the founding member of a conserved family of NAD-dependent protein deacetylases (20, 30, 52). The primary target of budding yeast Sir2 is thought to be the N-terminal tails of histones H3 and H4; therefore, Sir2 may be the enzyme responsible for creating the regions of hypoacetylated nucleosomes that are observed in silent chromatin (4, 57). Sir3 and Sir4 bind to the N-terminal tails of histones H3 and H4 in vitro, with a preference for the hypoacetylated rather than the acetylated tail (6, 14). In addition, mutations in histone H4 that disrupt silencing can be suppressed by second-site suppressors in SIR3, suggesting that Sir3 and histone H4 may interact directly in vivo (23).
In other eukaryotes, the assembly of silent chromatin requires modifications of histones which then also are bound by specific silencing proteins. For example, in the fission yeast Schizzosaccharomyces pombe, lysine 9 of histone H3 is deacetylated by Sir2 and then methylated by Clr4 (40, 42, 49). This modified nucleosome forms a binding site for the chromodomain proteins Swi6 and Chp1 which, although not homologous to Sir3 and Sir4, play functionally similar roles (2, 13, 41).
In addition to modifying and binding to histones, the Sir proteins engage in extensive protein-protein interactions with one another. Sir4 tightly binds to Sir2, and this interaction is thought to be essential for the recruitment of Sir2 to the HM loci and telomeres (9, 18, 36, 55). Sir4 also interacts with itself, suggesting that it functions as a dimer in vivo (7, 8, 39). Lastly, Sir4 binds to Sir3 in the absence of Sir2 and independently of chromatin (15, 55). These observations suggest that the three Sir proteins form a soluble complex, which has been named the SIR complex (15, 36, 37, 55). The existence of a stable SIR complex in solution, however, has not been demonstrated. Native purification of these proteins from yeast has yielded separate Sir4/Sir2 and Sir3 fractions (18, 62).
The assembly of silent chromatin is a stepwise process (18, 32, 47). First, recruitment proteins, including Sir1, Abf1, Rap1, Ku dimers, and the ORC complex, bind to regions called the E and I silencers, which initiate the assembly of silent chromatin at the HM loci, or to telomeres (46). These proteins then recruit the SIR complex by binding to Sir4 and perhaps Sir3 (32, 38). Once initiated, the spreading of silent chromatin requires all three Sir proteins as well as the enzymatic activity of Sir2 (18, 32, 47). Therefore, it has been proposed that successive cycles of histone deacetylation by Sir2 and recruitment of additional SIR complexes cause the spread of silent chromatin emanating from silencers and telomeres (18, 34, 46). This spreading is thought to involve the polymerization of the SIR complex, yet little is known about which Sir-Sir interactions are required for nucleation and growth of the SIR polymer. Similar stepwise models for heterochromatin assembly have been proposed in other organisms. For example, the histone binding proteins Swi6 and HP1, from fission yeast and mammals, respectively, may also polymerize as they spread along chromatin (13), although even less is known about this process.
Although Sir3 interacts with Sir4 in the SIR complex, a number of experiments have suggested that Sir3 can function independently of Sir4 and Sir2. First, the overexpression of Sir3 extends regions of silent chromatin, yet only Sir3 is found in these regions, suggesting that it can spread along chromatin alone (15, 43, 55). Second, a fragment of Sir3 binds to the nonacetylated N-terminal tail of histone H4 in vitro with a Kd of 35 nM (6), an affinity that may allow Sir3 to bind to hypoacetylated chromatin independently of Sir4 in vivo. Accordingly, deletion of the acetylase gene SAS2 results in hypoacetylation of lysine 16 on histone H4 and allows significant spreading of Sir3 in the absence of Sir2 at telomeres (26, 56). It has been difficult, however, to assess the importance of these interactions in vivo, because mutations that selectively disrupt the pairwise interactions of Sir proteins have not been reported.
Here we show that large portions of the cellular Sir4, Sir3, and Sir2 proteins form a stable and soluble SIR complex and that the integrity of this complex is required for the proper assembly of silent chromatin. To directly test the physiological significance of the interaction between Sir4 and Sir3 in vivo, we created mutations in Sir4 that selectively abolish the binding of Sir4 to Sir3. These mutations interfere with silencing at both the HM loci and telomeres by specifically blocking the recruitment of Sir3 to chromatin, despite significant deacetylation of histones near silencers. Our results therefore demonstrate that the association of Sir3 with silent chromatin depends on both the direct interaction of Sir3 with Sir4 and Sir3-histone interactions.
|
MATERIALS AND METHODS |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
was used for the amplification of DNA.
|
strains that were either hml::TRP1 or complemented by SIR-CEN plasmids. HML was deleted with pKK294, SIR2 was complemented by pRS315-SIR2, and SIR4 was complemented by pRS315-SIR4 (kindly provided by Jasper Rine, University of California, Berkeley). SIR2-H364Y was made by two-step gene replacement. A C-terminal fragment of SIR2 was cloned into pRS306 to create pDM638 (kindly provided by Jason Tanny). pDM638 was cut with BglII and transformed into yeast. Transformants were screened for nonmaters and were assumed to contain SIR2-H364Y followed by URA3 and a C-terminal fragment of SIR2. Nonmaters were grown on plates containing 5-fluoroorotic acid (5-FOA), and the resultant Ura– cells were again screened for nonmaters. Nonmaters were screened by Western blotting for the expression of full-length Sir2 and by PCR for the presence of the H364Y substitution (which destroys an NcoI restriction site).
sir4 coiled-coil mutants were also made by two-step gene replacement. The BamHI/XhoI restriction fragment from pGST-Sir4-C2-M1307N, pGST-Sir4-C2-I1311N, pGST-Sir4-C2-E1310R, pGST-Sir4-C2-K1324D, and pGST-Sir4-C2-K1325D (7) was cloned into pRS306 to create pAR417, pAR416, pAR423, pAR424, and pAR425, respectively. These plasmids were cut with MfeI and transformed into yeast. Transformants were assumed to contain a mutated full-length SIR4 gene followed by URA3 and a short C-terminal duplication of SIR4. URA+ transformants were grown on 5-FOA to select for recombination between the duplicated C-terminal regions, and the resultant Ura– strains were screened by PCR for the presence of the mutation and by Western blotting for the expression of full-length Sir4. Screening by PCR was done with oligonucleotides that have different abilities to hybridize with wild-type SIR4 and mutant SIR4. SIR4-M1307N and SIR4-I1311N, which are mating defective, were crossed by transforming complementing plasmid pRS314-SIR4 into the appropriate strains.
pDM538 (pRS313-SIR4) was made by cloning the XhoI/SacII fragment of pRS315-SIR4 into pRS313. The BsrGI/XhoI fragment of pDM538 was replaced with a BsrGI/XhoI SIR4 fragment that was amplified by PCR from SIR4 (ADR21), SIR4-M1307N (ADR2972), sir4-E1310R (ADR3060), SIR4-I1311N (ADR2973), and sir4-K1324D (ADR3062) to create pAR450 to pAR454, respectively. The BsrGI/XhoI fragment contains the SIR4 sequence from the natural BsrGI site to 216 nucleotides downstream of the stop codon, followed by an XhoI site introduced during PCR. The correct sequence of this region of SIR4 (and each of the sir4 mutants) was confirmed by sequencing of all five plasmids. The image of the Sir4 coiled coil (see Fig. 2) was made by using WebLab Viewer Lite 3.2 (protein database code 1NYH).
|
The hmrE (58) strains were constructed by crossing derivatives of CCFY100 (kindly provided by Kurt W. Runge, The Lerner Research Institute) (45) to the appropriate strains.
The sir2::HIS3 (JRY3433), sir3::HIS3 (JRY3289), and sir4::HIS3 (JRY3411) strains were kindly provided by Jasper Rine, University of California, Berkeley.
Immunoprecipitation and Western blotting. Yeast extracts for immunoprecipitation and Western blotting were made by bead beating (multitube bead beater; Biospec) frozen cell pellets in 350 µl lysis buffer (50 mM HEPES-KOH [pH 7.6], 500 mM sodium acetate, 5 mM magnesium acetate, 0.1 mM EDTA, 0.25% NP-40, 5% glycerol and, added just before use, 1 mM dithiothreitol [DTT], 1 mM phenylmethylsulfonyl fluoride [PMSF], leupeptin at 1 µg/ml, pepstatin at 1 µg/ml, bestatin at 1 µg/ml, and 1 mM benzamidine) with an excess of glass beads (Biospec) for two pulses of 30 s, with incubation on ice for 5 min between the pulses. A typical extract was made from cells harvested from a 50-ml culture at an optical density at 600 nm (OD600) of 1.0 to 1.5. The resulting lysate was separated from the glass beads and centrifuged at 14,000 rpm for 5 min to remove insoluble material. The protein concentration of each lysate was determined by using Bradford reagent (Bio-Rad), and samples were normalized based on these measurements. A portion of the lysate was mixed with an equal volume of 2 x sodium dodecyl sulfate (SDS) sample buffer (4% SDS, 160 mM Tris-HCl [pH 6.8], 20% glycerol, 20 mM EDTA, 0.04% bromophenol blue, and 1 mM PMSF) for Western blotting.
Standard methods were used for polyacrylamide gel electrophoresis and protein transfer to nitrocellulose (Schleicher & Schuell). Typically, samples were run on 10% or 12.5% polyacrylamide gels (80:1 and 120:1 acrylamide:bis, respectively, with no added SDS). Blots were stained with Ponceau S to confirm transfer and equal loading of samples and then were blocked for 30 min in blocking buffer (4% nonfat dried milk [Carnation] in TBST [20 mM Tris-HCl {pH 7.5}, 150 mM NaCl, 0.1% Tween 20]). All antibodies were incubated overnight at 4°C or for 2 h at 25°C. After being washed with TBST, the blots were incubated with horseradish peroxidase-conjugated anti-rabbit or anti-mouse antibodies (Amersham) at a 1:5,000 dilution in blocking buffer for 30 min at 25°C, washed again, incubated with Western Lightning reagents (Perkin-Elmer) according to the manufacturer's instructions, and then exposed to X-Omat film (Kodak).
The following antibodies were used for Western blotting. Anti-myc antibody 9E10 (Covance) was used at a dilution of 1:1,000 in TBST-0.02% NaN3. Affinity-purified rabbit polyclonal anti-Sir2 and anti-Sir4 antibodies (36) were used at a dilution of 1:5,000 in antibody storage buffer (autoclaved 4% nonfat dried milk, TBST, 5% glycerol, 0.02% NaN3).
The remaining lysate was used for immunoprecipitation with 9E10 or Sir4. A total of 0.5 to 3 µg of antibody was added to the lysate, and the mixture was incubated on ice for 30 min. Samples then were centrifuged at 14,000 rpm for 5 min at 4°C and transferred to 10 to 15 µl of protein A-Sepharose CL-4B beads (Pharmacia) which had been equilibrated in lysis buffer. The beads were rotated end over end at 4°C for 2 to 4 h. After immunoprecipitation, the beads were washed quickly three times with lysis buffer (the beads were transferred to fresh tubes after the second wash) and two times with low-salt wash buffer (50 mM HEPES-KOH [pH 7.6], 150 mM sodium acetate, 5 mM magnesium acetate, 5% glycerol and, added just before use, 1 mM DTT and 1 mM PMSF). All washes were performed on ice. After the final wash, residual buffer was removed from the beads with a Hamilton syringe, and the beads were resuspended in 1.5x SDS sample buffer, heated to 65°C for 15 min, and processed for Western blotting as described above.
Silver staining was performed as described previously (18).
Centrifugation at 100,000 x g was performed with a TLA 100.3 rotor in a tabletop ultracentrifuge (Beckman).
In some experiments, the lysis buffer contained an additional 0.5 M, 1.0 M, or 1.5 M sodium acetate or 100 µg/ml ethidium bromide and 0.1 µg/ml DNase I (750 U). When the beads were washed more extensively, the third wash with lysis buffer was extended, and a fourth wash with tube transfer was added after the extended wash.
Calf intestinal phosphatase (New England Biolabs [NEB]) was used for phosphatase treatment. After lysis buffer washes, the beads were washed with phosphatase buffer (NEB buffer 3) one time, incubated with 10 U of phosphatase in NEB buffer 3 (or no addition) at 37°C for 15 min, and then washed two times with low-salt wash buffer before being processed for Western blotting.
ChIP. Typically, 200 ml of yeast was grown to an OD600 of 1.5 at 30°C. Yeast lysates were prepared as described above, except that before harvesting of yeast cells, cultures were fixed with 1% formaldehyde for 15 min at 25°C. Cross-linking was quenched by the addition of glycine to 125 mM followed by 5 min of incubation. The cells were harvested, washed with 20 mM Tris-HCl (pH 7.5)-150 mM NaCl, and split into three tubes; yeast pellets were frozen in liquid nitrogen. The cells were lysed with 600 µl chromatin immunoprecipitation (ChIP) lysis buffer (50 mM HEPES-KOH [pH 7.6], 500 mM NaCl, 1 mM EDTA, 1% Triton X-100, 0.1% sodium deoxycholate, 0.1% SDS and, added just before use, 1 mM PMSF, leupeptin at 1 µg/ml, pepstatin at 1 µg/ml, bestatin at 1 µg/ml, and 1 mM benzamidine). Prior to normalization of various lysates, the chromatin in the lysates was sheared by sonication with three 20-s pulses at 40% power (Branson digital Sonifier 450 with a microtip), with 5 min on ice between pulses. The DNA was sheared to an average size of 400 to 500 bp. After sonication, the lysates were processed as described above. Sir3, Sir4, and histones H4 and H3 were immunoprecipitated with 0.5 µl affinity-purified rabbit anti-Sir3 polyclonal antibody, 1.0 µl affinity-purified rabbit anti-Sir4 antibody (36), 2 µl antibody to histone H4 acetylated at K5, K8, K12, and K16 [histone H4-Ac(K5, K8, K12, K16)], and 2 µl rabbit anti-histone H3-Ac(K9, K14) antibody (the latter two from Upstate Biotechnology). Beads were washed three times with ChIP lysis buffer (the beads were transferred to fresh tubes after the second wash), one time with Li buffer (10 mM Tris-HCl [pH 8.0], 0.25 M LiCl, 0.5% NP-40, 0.5% sodium deoxycholate, 1 mM EDTA), and one time with 50/10-TE (50 mM Tris-Cl [pH 8.0], 10 mM EDTA). Protein and cross-linked DNA were eluted in 100 µl 50/10-TE-1% SDS at 65°C for 15 min. After the eluate was removed from the beads, the beads were washed with 150 µl 50/10-TE-0.67% SDS; the wash was pooled with the first eluate. The mixture was incubated for 6 h to overnight at 65°C to reverse the cross-links, incubated with 50 µg/ml proteinase K for 2 h at 55°C, extracted with an equal volume of 1:1 unbuffered phenol:chloroform, and extracted with chloroform alone. The remaining associated DNA was precipitated by the addition of 400 mM LiCl, 10 µg glycogen, and 2 volumes of ethyl alcohol, washed with 75% ethyl alcohol, air dried, and resuspended in 50 µl of 10/1-TE (10 mM Tris-Cl [pH 8.0], 1 mM EDTA) containing 2 µg/ml RNase A.
Input DNA was prepared by mixing 50 µl of the starting lysate (after sonication) with 200 µl 50/10-TE-1% SDS. The lysate was processed in a fashion similar to that of the immunoprecipitates, except that a second phenol:chloroform extraction was performed and the final DNA pellet was resuspended in 50 µl 10/1-TE containing 20 µg/ml RNase A.
The precipitated DNA and the input DNA were analyzed by PCR with either 5 µl of a 1:10 dilution of the immunoprecipitated material or 5 µl of a 1:10,000 dilution of the input material. PCR amplifications were performed with 12.5-µl reaction mixtures containing PCR buffer (10 mM Tris-HCl [pH 8.3], 1.5 mM MgCl2, 50 mM KCl), 0.1 mM deoxynucleoside triphosphates, 0.1 mCi of [-32P]dCTP (3,000 Ci/mmol; Amersham), 1 µM each JH301 and JH302 internal ACT1 primers, and 1 µM each silent-locus primer. Reaction conditions were 1 cycle of 95°C for 2 min, 55°C for 30 s, and 72°C for 1 min; 25 or 26 cycles of 95°C for 30 s, 55°C for 30 s, and 72°C for 30 s; and a final step of 72°C for 4 min. The primers used for ChIP are listed in Table 2 and were described previously (18, 19, 32); they yielded products ranging from 150 to 400 bp. After completion of the PCR amplifications, 1.5 µl of loading dye (50 mM Tris-HCl [pH 8.0], 50% sucrose, 4 M urea, 0.05% bromophenol blue) was added to each reaction mixture, and reactions were run on 6% acrylamide (30:1 acrylamide:bis)-Tris-borate-EDTA gels for 45 min at 100 V. Gels were dried and subjected to autoradiography and quantification by phosphorimaging (Personal FX and QuantityOne; Bio-Rad).
|
in the Sir3 ChIP and of sir4 in the Sir4 ChIP were set to 1; the average values were scaled (see Fig. 8) so that the wild-type values in both histone H3 and histone H4 ChIPs were set to 1.
|
|
|
Quantitative mating.
Tested strains and tester strains were placed on YEP-2% glucose plates supplemented with 0.005% adenine and tryptophan and grown overnight at 30°C. On the following morning, the cells were scraped from the plates, inoculated into YEP-2% glucose liquid medium containing 0.005% adenine and tryptophan, and grown for 2 to 4 h at 25°C to an OD600 of 1 to 2 for the tested strains and to an OD600 of 4 to 6 for the tester strains. Quantities of 106, 105, 104, 103, and 102 cells of the tested strain (in 100 µl) were mixed with 107 to 5 x 107 cells of the tester strain (in 300 µl). The mixture was plated directly on synthetic minimal-2% glucose plates and grown at 25°C for 3 days, and then counting was done. A quantity of 102 cells of the tested strain was also plated on synthetic complete-2% glucose plates to accurately determine the number of viable cells that were mated in the experiment. Each mating was performed in triplicate. Mating efficiencies were typically between 50 and 100% for wild-type cells (ADR21 and ADR22), and the mating defect of sir4 was reproducibly more severe for MATa cells than for MAT cells (see Fig. 4A).
|
|
RESULTS |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In order to resolve the discrepancy between these observations and to test whether Sir2, Sir3, and Sir4 form a soluble SIR complex, we quantified the relative amounts of Sir4 and Sir2 present in Sir3 complexes that could be immunoprecipitated from whole-cell extracts by a different affinity purification strategy. We first created a strain that replaces the endogenous SIR3 with a fusion of SIR3 and 13 myc epitopes (SIR3-myc13). Sir3-myc13 is fully functional for silencing and is expressed in yeast at wild-type levels (see Fig. S2 in the supplemental material). Precipitation of Sir3-myc13 with an anti-myc antibody resulted in coprecipitation of Sir4 and Sir2 (Fig. 1A), consistent with the results of earlier studies with affinity-tagged Sir3 (15, 55). Large portions of Sir4 and Sir2 coprecipitated with Sir3-myc13 (Fig. 1B) and, based on serial dilutions of the precipitates, we estimated that 25% of total Sir4 and 13% of total Sir2 were precipitated (Fig. 1C). Although the Sir2 protein appeared to be present in the precipitate at lower levels than Sir4 (Fig. 1B), it has been shown that virtually all Sir4 is bound to Sir2 (36) and that Sir2 stains poorly with silver (data not shown). It is therefore likely that equal amounts of Sir2 and Sir4 coprecipitate with Sir3.
|
Why do TAP purifications of Sir3-containing complexes apparently behave differently from the anti-myc precipitates? Our current and previous TAP purifications of both Sir3 and Sir4 did not yield a SIR complex that contained stoichiometric amounts of each protein (18, 62). We found that although the complete SIR complex was present in the starting TAP lysate, it failed to elute from the affinity resin during the first step of the purification (see Fig. S1C in the supplemental material), perhaps as a result of aggregation or extensive multimerization of one of its components. This was not an issue in the anti-myc immunoprecipitations because the bound material was eluted from the affinity resin by SDS (Fig. 1B).
Using coprecipitation experiments, we examined the architecture of the SIR complex. As was reported previously by Grunstein and colleagues (15, 55), Sir3 interacts with Sir4 independently of Sir2 (see Fig. S3 in the supplemental material); because silent chromatin does not form in sir2 cells, these data confirm that the interaction between Sir3 and Sir4 occurs independently of silent chromatin. In addition, the SIR complex can form independently of Sir2 enzymatic activity. SIR2-H364Y, which renders Sir2 inactive and destroys silencing in vivo (20, 61), does not affect the stability of the SIR complex (see Fig. S3 in the supplemental material). As shown previously, the interaction between Sir2 and Sir3 depends on the presence of Sir4 (55), suggesting that in yeast extracts, Sir2 and Sir3 cannot form a stable complex in the absence of Sir4.
Sir4 coiled-coil mutants cannot bind to Sir3 in vivo. Recently, two crystal structures of C-terminal fragments of Sir4 (Sir4-C) were reported (7, 39); these showed that the extreme C terminus (amino acids 1272 to 1335) self-associates to form a parallel coiled coil. This domain lies within the region of Sir4 that binds to Sir3 (36, 37), and C-terminal coiled-coil fragments of Sir4 bind tightly to C-terminal fragments of Sir3 in vitro (7). Three amino acids (M1307, E1310, and I1311) located on the exposed surface of the coiled coil were shown to be critical for the interaction between the C-terminal fragments of Sir3 and Sir4 in vitro, while neighboring surface sites (K1324 and K1325) had more modest effects on this interaction (7). Although these substitutions blocked the interaction between fragments of Sir3 and Sir4-C, they did not affect the ability of Sir4-C to self-associate into a coiled coil. The cluster of three residues was proposed to be the specific binding site for two Sir3 proteins on the Sir4 coiled coil (7).
We were interested in understanding the physiological significance of the interaction between Sir3 and Sir4. We therefore tested if mutations in the coiled coil of Sir4 tested in vitro also disrupted the interaction between full-length Sir4 and Sir3 in vivo and if this interaction played a role in the assembly of silent chromatin. We replaced individual surface residue in the Sir4 coiled coil to create five mutants, sir4-M1307N, sir4-E1310R, sir4-I1311N, sir4-K1324D, and sir4-K1325D (Fig. 2A). These sir4 mutants replaced the endogenous SIR4 gene. Four of the mutants were expressed at wild-type levels (Fig. 2B, lanes 18 to 23, middle row), whereas sir4-K1324D consistently produced less Sir4 protein than wild-type SIR4 (Fig. 2B, lane 22; see also Fig. 5B, lane 18). These mutations did not affect the overall structure of Sir4 because each of these mutant proteins interacted normally with Sir2 (Fig. 2B, lanes 10 to 15, bottom row), and the Sir2 associated with the mutant Sir4 proteins had wild-type levels of deacetylase activity (Fig. 3).
|
|
The interaction between Sir3 and Sir4 is required for silencing. Using sir4 mutants that are selectively deficient in binding to Sir3, we were able to test whether the integrity of the SIR complex affects the assembly of silent chromatin. Silencing at HML and HMR can be monitored by the ability of haploid strains to mate (16). The presence of silent chromatin at HML and HMR is required to maintain haploid cell identity and wild-type mating efficiencies, while a loss of silencing causes cells to adopt a pseudodiploid cell identity and lose the ability to mate. Using quantitative mating assays, we determined that sir4-M1307N and sir4-I1311N are completely defective for silencing at both HML and HMR loci (Fig. 4A). sir4-E1310R cells, despite a strong defect in coprecipitation assays (Fig. 2B, lanes 4 and 12), maintained wild-type levels of mating. sir4-K1324D cells had minor mating defects, showing a two- to threefold decrease in mating efficiency. Previous studies showed that small changes in Sir4 dosage can cause a loss of silencing (59); therefore, the small mating defect could be attributed to the reduced expression of Sir4-K1324D (Fig. 2B, lane 22, and Fig. 5B) and not to the inability of Sir4-K1324D to bind to Sir3 (Fig. 2B, lanes 6 and 14). Although the HM loci are able to tolerate a weakening in the binding of Sir3 to Sir4, the loss of this interaction in sir4-M1307N and sir4-I1311N cells leads to a complete disruption of silencing.
Because Sir4-E1310R and Sir4-K1324D have detectable defects in their binding to Sir3, we wondered if a stronger defect in HM silencing in sir4-E1310R and sir4-K1324D might be revealed in a sensitized strain background. We therefore monitored the expression of TRP1 gene that had been integrated at hmrE, at which the E element of the HMR-E silencer has been mutated and can no longer bind to the Rap1 protein (58). The TRP1 gene is effectively silenced in SIR4 cells, and these cells grow poorly on plates lacking tryptophan (Fig. 4B, top row). In contrast, sir4-M1307N, sir4-E1310R, sir4-I1311N, and sir4-K1324D are unable to silence TRP1 at hmrE.
Silencing at telomeres can be monitored by assaying the repression of a reporter gene placed at a telomeric location. We monitored the expression of a URA3 gene that had been integrated at the left telomere of chromosome VII, compared to an identical URA3 gene that is integrated at the ADH4 locus, 20 kb away (12). When the telomeric URA3 is silenced in wild-type cells, cells can grow on 5-FOA, which is toxic to Ura+ cells. The URA3 gene integrated at ADH4, however, is expressed, and cells are unable to grow on 5-FOA. Like silencing at hmrE, all four sir4 mutants that are defective in the interaction with Sir3 were completely defective in telomeric silencing (Fig. 4C). sir4-K1325D cells showed no defects in either HM locus or telomeric silencing, consistent with this mutation having no effect on the interaction between Sir3 and Sir4 in coprecipitation experiments (Fig. 2B, lanes 7 and 15). Thus, the ability of Sir3 to bind to Sir4 is required for silencing at both the HM loci and telomeres. Furthermore, silencing at telomeres is more sensitive to mutations in Sir4 that only weaken its interaction with Sir3.
Because the sir4 mutants are defective in binding Sir3, but not Sir2, we tested if they cause a dominant loss of silencing. We constructed low copy CEN plasmids containing either wild-type SIR4, or each of the four mutations that had telomeric silencing defects. The sir4 on the plasmids was expressed at a similar level to the endogenous SIR4 (Fig. 5B, except sir4-K1324D, which expressed at a lower level) and when transformed into sir4 cells, each of the mutants was unable to complement the silencing defect of sir4, behaving identically to the strains tested in Fig. 4C. These plasmids were transformed into wild-type cells and the effect of simultaneous expression of SIR4 and the sir4 mutations on telomeric silencing was examined. sir4-M1307N and sir4-I1311N caused a partial dominant loss of silencing (Fig. 5A, approximately 100-fold less 5-FOA resistance than the wild type), while sir4-E1310R and sir4-K1324D showed no dominance. Because of the dominant phenotype of the M1307N and I1311N substitutions in SIR4, we will refer to these mutations as SIR4-M1307N and SIR4-I1311N in the sections that follow. The dominance of these SIR4 alleles is similar to what was previously observed for SIR2 alleles that produce an enzymatically inactive protein (61), although SIR2-H364Y causes a complete loss of silencing at telomeres when expressed simultaneously with wild-type SIR2 (Fig. 5A).
The Sir3-Sir4 interaction is required for the recruitment of Sir3 to silencers.
We next asked if the sir4 mutants affect the recruitment of Sir4 or Sir3 to silencers or telomeres, and if the interaction of Sir3 with Sir4 is required for the spreading of silent chromatin away from these sites of initiation. Using ChIP, we examined the localization of Sir3 and Sir4 to different regions of silent chromatin. It was previously shown that low levels of Sir4 bind to silencers in the absence of the Sir3 protein (18, 32, 47). Consistent with these results, Sir4-M1307N and Sir4-I1311N, which cannot bind to Sir3, were recruited to the silencers HMR-E and HML-E to the same extent as Sir4 was recruited in a sir3 strain (Fig. 6A and B, HMR-E and HML-E, compare lanes 6 and 7 to lane 4 in Sir4 ChIP). A similar result also was obtained at a telomeric site on the right arm of chromosome VI (Fig. 6C, TEL0.07, compare lanes 6 and 7 to lane 4). Sir3 in SIR4-M1307N and SIR4-I1311N cells, however, did not localize to HMR-E and TEL0.07, and was present only at low levels at HML-E (Fig. 6, HMR-E, HML-E, and TEL0.07, lanes 6 and 7 in Sir3 ChIP). sir4-E1310R, one of the mutants that weaken the interaction between Sir3 and Sir4 (Fig. 2B), also failed to recruit Sir3 to telomeres (Fig. 6C, TEL0.07, lane 8 in Sir3 ChIP), but was able to recruit wild-type levels of Sir3 to HMR-E and HML-E (Fig. 6A and B, lane 8 in Sir3 ChIP). In addition, the spreading of both Sir3 and Sir4 to sites within extended silent chromatin was greatly impaired in SIR4-M1307N and SIR4-I1311N cells (Fig. 7, HMRa, HML, and TEL0.6, lanes 6 and 7), and was impaired at telomeres in sir4-E1310R (Fig. 7, TEL0.6, lane 8).
These results demonstrate that the binding of Sir3 to Sir4 is required for the localization of Sir3 to silent chromatin regions. The absence of the entire SIR complex at silencers and the telomeres results in a lack of spreading of the Sir proteins to sites within extended silent chromatin. In addition, the ChIP data (Fig. 6 and 7) was entirely consistent with our phenotypic analysis (Fig. 4). When Sir3 was not recruited to telomeres and silencers, and Sir3 and Sir4 did not spread, no silencing was observed. Although we did not measure the localization of Sir2 in this experiment, it is likely that its localization is identical to that of Sir4 because previous ChIP experiments showed no differences in the localization of Sir2 and Sir4 (18, 47), and the tight binding of Sir2 to Sir4 is unaffected in the sir4 coiled-coil mutants (Fig. 2B).
Previous experiments showed that the Sir2 protein that is recruited to silencers in sir3 cells is sufficient to cause partial deacetylation of the N-terminal tails of histone H3 and H4 in adjacent nucleosomes (18). Because wild-type Sir3 protein is present in SIR4-M1307N and SIR4-I1311N cells but cannot associate with Sir4, we were able to ask whether partial deacetylation of histones by silencer-bound Sir2/Sir4 is sufficient for the recruitment of Sir3 to chromatin in the absence of the Sir3-Sir4 interaction. In both SIR4-M1307N and SIR4-I1311N cells, the partial deacetylation of histone H3 and H4 occurs (Fig. 8A and B, HMR-E and HML-E, lanes 6 and 7), confirming previous results that some Sir3-independent deacetylation can be observed after the recruitment of Sir2 and Sir4 to silencers. Despite this deacetylation, little Sir3 localizes to either HMR-E or HML-E (Fig. 6A and B, lanes 6 and 7, Sir3 ChIP).
We compared the localization of Sir3 and diacetylated histone H3 at the HMR-E silencer and of Sir3 and tetraacetylated histone H4 at the HML-E silencer (Fig. 9). Approximately one-fourth as much acetylated histone H3 was found at HMR-E in SIR4-M1307N and SIR4-I1311N cells and one-half as much acetylated histone H4 was found at HML-E in SIR4-M1307N and SIR4-I1311N cells relative to sir4. Despite this change, only small amounts (10 to 20% of the wild-type level) of Sir3 localized to HMR-E and HML-E in these mutants. The amount of Sir3 recruited also was less than the amount recruited in sir2 and SIR2-H364Y cells, in which histone H4 is nearly fully acetylated. These results demonstrate that the binding of Sir3 to Sir4 is critical for the recruitment of Sir3 and suggest that deacetylation of the nucleosome is not sufficient for efficient Sir3 recruitment.
|
|
DISCUSSION |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Assembly of the SIR complex. We have shown that Sir2, Sir4, and Sir3 form a stable and soluble SIR protein complex. Although the interactions between these three proteins were previously shown by coprecipitation and two hybrid assays (15, 36, 37, 55), earlier native biochemical purifications did not detect a complex containing all three Sir proteins (10, 18). Previous experiments suggested that this was due to the N-terminal region of Sir4 inhibiting its interaction with Sir3 (36), but because we only examined the interactions of full-length proteins, our data does not address this model. Our data show that the SIR complex is soluble and contains a large portion of the Sir4 in the cell, but does not efficiently elute from affinity resins. Nonetheless, our results are consistent with previous findings that virtually all Sir4 is bound to Sir2, and a significant fraction of Sir3 is not associated with the Sir2/Sir4 subcomplex in whole-cell extracts (18, 36, 62).
Point mutations in the coiled-coil domain of Sir4 that preclude its assembly into the SIR complex disrupt silencing at both the HM loci and telomeres. The specific defects in binding to Sir3 are likely the sole source of the phenotypes seen in these sir4 mutations. The amino acid substitutions in these mutations are in the coiled-coil domain of Sir4 that fall within the previously defined Sir3 binding region and cause similar in vitro and in vivo disruptions of Sir3 binding to Sir4 (7). Although there is good agreement between previous in vitro pull-down experiments with purified proteins (7) and our immunoprecipitation of complexes formed in vivo, the in vitro pull-down experiments used only C-terminal fragments of both Sir3 and Sir4 and therefore could not rule out the presence of a second interacting surface between the two proteins. The results presented here show that the coiled-coil domain of Sir4 is necessary for the interaction between full-length Sir4 and full-length Sir3 in vivo.
The C terminus of Sir4 has also been shown to bind Rap1, Yku70, and Ubp3 (35, 37, 38, 65). The interactions of these proteins with full-length Sir4 are difficult to detect; therefore, it is difficult to rule out whether the sir4 coiled-coil mutants also affect the binding of Sir4 to these proteins. We believe it is unlikely, however, that a defect in the interaction between these proteins and Sir4 contributes to the phenotype of our sir4 mutants for the following reasons. (i) As much Sir4-M1307N and Sir4-I1311N is recruited to silencers and telomeres as wild-type Sir4 is recruited to these loci in sir3 cells (Fig. 6, compare lanes 6 and 7 to lane 4, Sir4 ChIP). We therefore conclude that the binding of Sir4-M1307N and Sir4-I1311N to recruitment factors (like Rap1 and Yku70) is normal. (ii) yku70 mutants only have telomeric silencing defects (3), while the sir4 coiled-coil mutants also have defects at HMR (Fig. 4A and B). (iii) ubp3 cells have been shown to improve silencing (35), so a defect in Ubp3 binding in the sir4 coiled-coil mutants would be predicted to improve silencing, not weaken it, as the sir4 coiled-coil mutants do. (iv) The E element of the HMR-E silencer is a binding site for Rap1 (58), so if a defect in the binding of Rap1 to Sir4 was the cause of the phenotypes of the sir4 coiled-coil mutants, than we would predict that the silencing defects of these mutants in a hmrE strain would be no more severe than hmrE in a SIR4 strain. The sir4 coiled-coil mutants, however, completely derepress silencing at hmrE (Fig. 4B).
Previously we had suggested that Sir3 and the Sir4/Sir2 complex meet on chromatin either at silencers or telomeres or as they bind to deacetylated histones during spreading (34). Another possibility suggested from this study is that the SIR complex forms in solution and loads onto chromatin as a single complex, initially by interactions with DNA binding proteins and then during spreading of silent chromatin by interactions with other SIR complexes and deacetylated nucleosomes (Fig. 10).
|
mutation, in which lysine 16 of histone H4 is hypoacetylated, does not cause any increase in the amount of Sir3 recruited to silencers in SIR4-M1307N or SIR4-I1311N cells relative to sir4 cells, nor does it suppress any of the silencing defects in SIR4-M1307N and SIR4-I1311N cells (data not shown). These results suggest that although Sir3 has a high affinity for deacetylated N-terminal tails of histones in vitro (6), the association of Sir3 with deacetylated histone tails is not sufficient for the full recruitment of Sir3 to chromatin in vivo. Sir3 binding to deacetylated histones, however, may contribute to its recruitment to silencers. Similar amounts of Sir3 were recruited to HML-E in SIR4-M1307N and SIR4-I1311N cells as in sir2 and SIR2-H364Y cells (Fig. 6B, lanes 2, 3, 6, and 7), suggesting that Sir3 binding to either Sir4 or deacetylated histone H4 can recruit some Sir3 to this site. Although there is some Sir3 recruitment to HML-E in SIR4-M1307N and SIR4-I1311N cells, there is no recruitment to HMR-E or to TEL0.07 (Fig. 6A and C, lanes 6 and 7, Sir3 ChIP). This difference may be caused by the greater deacetylation of histone H4 at HML-E compared to HMR-E and TEL0.07 (Fig. 8B, lanes 6 and 7 compared to lane 4; Fig. 8A and C, lanes 6 and 7 compared to lane 4). The assembly of silent chromatin therefore requires both the interaction of Sir3 with Sir4 and the deacetylation of histones (Fig. 10A). When either of these two steps are defective, silent chromatin cannot assemble. In SIR4-M1307N and SIR4-I1311N cells, although the Sir4/Sir2 subcomplex is recruited to silencers and telomeres and histones proximal to these sites are deacetylated, the nucleation of silent chromatin does not occur (Fig. 10B). In SIR2-H364Y cells, although the full SIR complex can be recruited to initiation sites, enzymatically inactive Sir2 is unable to deacetylate histones and the nucleation of silent chromatin also does not occur (Fig. 10C).
Spreading of silent chromatin is thought to require successive recruitment of additional SIR complexes followed by deacetylation of nucleosomes (Fig. 10A). However, because deacetylation alone is not sufficient to promote Sir3 binding to chromatin (Fig. 6 to 8), we propose that the recruitment of all subsequent SIR complexes requires both histone deacetylation and lateral interactions between SIR complexes (Fig. 10A). In this model, the growth of the SIR polymer along the chromatin fiber requires two distinct interactions, one between the SIR complex and deacetylated histone tails and one between adjacent SIR complexes. We speculate that the interaction between Sir3 and the Sir4 coiled coil does not bridge adjacent SIR complexes, rather that only intact SIR complexes are able to interact during spreading of silent chromatin.
Is the Sir3-Sir4 interaction required for the spreading of silent chromatin? The dominant phenotype of SIR4-M1307N and SIR4-I1311N supports the model that the binding of Sir3 to Sir4 is required for the spreading of silent chromatin (Fig. 10B). When these mutants are expressed simultaneously with wild-type SIR4, silencing at telomeres is reduced by 100-fold (Fig. 5A). Under these coexpression conditions, 25% of the Sir4 dimers should contain two wild-type Sir4 molecules, 25% should contain two mutant molecules, and 50% should be dimers that contain one each wild-type and mutant Sir4 molecules. We would expect that the population containing two mutant Sir4 proteins to have little or no dominant phenotype as these mutant dimers will be unable to bind to Sir3 and incorporate into the SIR complex, and should therefore be unable to cap the spreading SIR polymer. Assuming that each Sir4 dimer binds to two Sir3 proteins (7), the dominant phenotype of SIR4-M1307N and SIR4-I1311N is likely to stem from the association of the mixed wild-type-mutant dimers with Sir3 during spreading of silent chromatin. Because these mixed dimers can only interact with one Sir3 molecule, their incorporation into silent chromatin would cap the growing SIR polymer and cause a dominant-negative phenotype.
The dominant-negative phenotype of SIR4-M1307N and SIR4-I1311N is less potent than that previously reported for SIR2-H364Y (61). Because SIR2-H364Y can still incorporate into the SIR complex (see Fig. S3 in the supplemental material), SIR complexes containing Sir2-H364Y are likely to be recruited into spreading silent chromatin with wild-type efficiency, but their lack of deacetylase activity would block the recruitment of additional SIR complexes and cap the growing SIR polymer. In contrast, the ability of mixed wild-type Sir4-M1307N or Sir4-I1311N dimers to associate with the growing SIR polymer may be impaired due to the loss of one Sir3 binding surface on the coiled coil dimer. As a result, their ability to poison the spreading of silent chromatin would be compromised compared to SIR complexes containing Sir2-H364Y.
Inheritance of silent chromatin. The N-terminal tails of histones can be acetylated, methylated, ubiquitinated, and phosphorylated (21). This wide array of possible modifications of histones has led to the proposal that different combinations of histone modifications form a "histone code" that determines the identity and function of a given chromatin domain. Because histone H3 and H4 tetramers remain associated with daughter DNA strands during DNA replication (27, 53), their modification states are also inherited and therefore may form the basis for the epigenetic inheritance of chromatin domains. In silent chromatin, the hypoacetylation of histones may be expected to form the primary epigenetic mark that ensures the stability of silent chromatin and its inheritance after DNA replication (4, 57). Our data, however, show that in addition to histone deacetylation, the tight binding of Sir3 to Sir4 at silencers and telomeres is required to recruit Sir3, suggesting that both histone modifications and interactions between silencing proteins are essential for the assembly and inheritance of silent chromatin domains.
|
ACKNOWLEDGMENTS |
|---|
This work was supported by grants from the NIH (GM61641 to D.M. and GM52504 to T.E.) and the Ellison Medical Foundation (to D.M.). A.D.R. is a fellow of the Jane Coffin Childs Memorial Fund for Medical Research, B.E.H. is supported by a fellowship (GM06835001) from the NIH, and D.M. is a scholar of the Leukemia and Lymphoma Society.
|
FOOTNOTES |
|---|
Supplemental material for this article may be found at http://mcb.asm.org/.
|
REFERENCES |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
2. Bannister, A. J., P. Zegerman, J. F. Partridge, E. A. Miska, J. O. Thomas, R. C. Allshire, and T. Kouzarides. 2001. Selective recognition of methylated lysine 9 on histone H3 by the HP1 chromo domain. Nature 410:120-124.[CrossRef][Medline]
3. Boulton, S. J., and S. P. Jackson. 1998. Components of the Ku-dependent non-homologous end-joining pathway are involved in telomeric length maintenance and telomeric silencing. EMBO J. 17:1819-1828.[CrossRef][Medline]
4. Braunstein, M., A. B. Rose, S. G. Holmes, C. D. Allis, and J. R. Broach. 1993. Transcriptional silencing in yeast is associated with reduced nucleosome acetylation. Genes Dev. 7:592-604.
5. Bryk, M., M. Banerjee, M. Murphy, K. E. Knudsen, D. J. Garfinkel, and M. J. Curcio. 1997. Transcriptional silencing of Ty1 elements in the RDN1 locus of yeast. Genes Dev. 11:255-269.
6. 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.
7. Chang, J. F., B. E. Hall, J. C. Tanny, D. Moazed, D. Filman, and T. Ellenberger. 2003. Structure of the coiled-coil dimerization motif of Sir4 and its interaction with Sir3. Structure (Cambridge) 11:637-649.
8. Chien, C. T., P. L. Bartel, R. Sternglanz, and S. Fields. 1991. The two-hybrid system: a method to identify and clone genes for proteins that interact with a protein of interest. Proc. Natl. Acad. Sci. USA 88:9578-9582.
9. Cuperus, G., R. Shafaatian, and D. Shore. 2000. Locus specificity determinants in the multifunctional yeast silencing protein Sir2. EMBO J. 19:2641-2651.[CrossRef][Medline]
10. Ghidelli, S., D. Donze, N. Dhillon, and R. T. Kamakaka. 2001. Sir2p exists in two nucleosome-binding complexes with distinct deacetylase activities. EMBO J. 20:4522-4535.[CrossRef][Medline]
11. Goldstein, A. L., and J. H. McCusker. 1999. Three new dominant drug resistance cassettes for gene disruption in Saccharomyces cerevisiae. Yeast 15:1541-1553.[CrossRef][Medline]
12. Gottschling, D. E., O. M. Aparicio, B. L. Billington, and V. A. Zakian. 1990. Position effect at S. cerevisiae telomeres: reversible repression of Pol II transcription. Cell 63:751-762.[CrossRef][Medline]
13. Grewal, S. I., and D. Moazed. 2003. Heterochromatin and epigenetic control of gene expression. Science 301:798-802.
14. 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.[CrossRef][Medline]
15. Hecht, A., S. Strahl-Bolsinger, and M. Grunstein. 1996. Spreading of transcriptional repressor SIR3 from telomeric heterochromatin. Nature 383:92-96.[CrossRef][Medline]
16. Herskowitz, I. 1989. A regulatory hierarchy for cell specialization in yeast. Nature 342:749-757.[CrossRef][Medline]
