![[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}
Jeffrey D. Wuitschick,1,
Cheryl D. Givens,1
Aaron M. Kosinski,1 and
Janet F. Partridge1*
Department of Biochemistry, St. Jude Children's Research Hospital, Memphis, Tennessee1
Received 12 October 2004/ Returned for modification 8 December 2004/ Accepted 21 December 2004
 |
ABSTRACT |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
INTRODUCTION |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In the fission yeast Schizosaccharomyces pombe, heterochromatin assembles at the mating type locus, centromeres, and telomeres (1). The chromodomain protein Swi6, the fission yeast homolog of heterochromatin protein 1 (HP1), localizes to and is required for maintenance of the repressed nature of all sites of heterochromatin (2, 11). The localization of Swi6 to these sequences is dependent on the activity of the Clr4 histone H3 K9-methyltransferase [the homolog of Su(var)39 enzymes] (5, 12) and is lost in cells where lysine 9 methylation is blocked by mutation (25).
Other components required for the establishment and maintenance of heterochromatin at all these loci are not yet resolved. At centromeres, heterochromatin maintenance and establishment additionally requires components of the RNA interference (RNAi) apparatus, which in fission yeast include a single Argonaute protein (Ago1) and a single Dicer enzyme (Dcr1) (30, 40, 41). Centromeric double-stranded transcripts are processed by Dicer, which generates short interfering centromeric RNAs (siRNAs) (31) that then somehow direct the complex to the centromere, via either siRNA-DNA or siRNA-nascent RNA interactions, most likely following their association with the Argonaute PAZ domain-containing protein, Ago1. This process is required for recruitment of the K9-histone H3 methyltransferase Clr4 to the locus (41), and this K9 H3 methylation then attracts additional K9-MeH3 binding chromodomain proteins, which contribute to the repressive architecture of centromeric heterochromatin (15).
We have previously demonstrated that another chromodomain protein, Chp1, is also essential for the formation of centromeric heterochromatin (28, 29). Recently, Chp1 has been shown to be a component of the RITS (RNA-induced initiation of transcriptional gene silencing) complex, which contains centromeric siRNAs, Ago1, Chp1, and a previously uncharacterized protein, Tas3 (39). RITS localization to centromeric sequences is required for the recruitment of Clr4 and Dicer activity and, hence, siRNA production is required for the association of the RITS complex with centromeres (39).
Despite their importance in the formation and maintenance of centromeric heterochromatin, deletions of components of the RNAi machinery have little impact on the maintenance of silent heterochromatin at the other well-characterized heterochromatic locus in fission yeast, the mating type region (mat2/3) (20). Recently, however, it has been demonstrated that multiple mechanisms ensure the formation and maintenance of heterochromatin at mat2/3. Therefore, loss of one mode of heterochromatin formation can be compensated for by redundancy within the system (20, 21, 37). The silent mat2/3 locus encompasses a 4.3-kb region of DNA (cenH) that is almost identical to the sequence found at the outer repeats of the centromere (17). Like the outer repeats, this cenH region is assembled into heterochromatin through an RNAi-dependent pathway (20). However, proximal to the mat3 locus, heterochromatin assembly occurs independent of the RNAi pathway as, for example, Atf1/Pcr1 transcription factors can directly recruit Clr6 histone deacetylase and Clr4 methyltransferase to these mat2/3 unique sequences (see Fig. 1 and 9, below) (21, 22).
|
|
|
MATERIALS AND METHODS |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Strain generation.
Strains used in this study are listed in Table 1. The chp1
his3+, rik1 LEU2, and clr4 LEU2 deletion strains were previously described (12, 28, 29). ago1 ura4+ and tas3 ura4+ were generated by replacing the complete open reading frames (ORFs) with the ura4+ reporter gene, and dcr1 Kanr was generated by replacement of the dicer ORF with KanMX6 (4). chp1-6xmyc+ has been described previously (28). tas3+ was tagged at its C terminus with 13xmyc epitopes (4). Chp1-TAP was generated by C-terminal tagging of the Chp1 ORF with the tandem affinity purification tag (36). Strains were generated in a ura4D18, leu1-32, and his3D background and outcrossed three times after construction and prior to introduction of additional marked loci. The function of tagged proteins was tested by the introduction, by crossing, of a centromeric ura4+ marker gene [otr1R(Sph1)::ura4+] (2) and assessing silencing of the centromeric ura4+ transgene by a serial dilution assay of cells on PMG-URA and PMG plus 5-fluoroorotic acid (FOA) media. FOA is toxic to cells that express ura4+ (7).
|
CD-chp1-6xmyc and RRM-chp1-6xmyc were generated by removing amino acids 2 to 79 and 311 to 407 of chp1-6xmyc+, respectively. These deletions were performed using a two-stage transformation process. Initially, the chromodomain was replaced with the ura4+ marker gene by transforming Chp1-6xmyc cells (PY 189) by electroporation with a PCR product containing ura4+ flanked with 
80 bases of homology to sequences flanking the chromodomain. ura4+ transformants were selected by growth of cells on medium lacking uracil, and cells in which the chromodomain of Chp1 had been replaced by ura4+ (CD::ura4+chp1-6xmyc) were detected by PCR (PY 282). Removal of ura4+ and generation of CDchp1-6xmyc cells was achieved by retransformation with a second chimeric PCR product that only contained sequences that flank the chp1 chromodomain and selecting for growth of cells on medium containing 5-FOA. PCR and DNA sequencing were used to confirm the generation of CDchp1-6xmyc cells after three outcrosses (PY 387). RRMchp1-6xmyc cells were derived in a similar fashion (PY 338, giving rise to PY 450). Strains were crossed to introduce the centromeric ura4+ marker at otr1R(Sph1) and the ura4+ minigene (ura4 DS/E) at the endogenous ura4+ locus.
chp1 cells bearing the pREP81-3xHA vector series were generated by transformation of PY 225 with vector, and the episomal ARS plasmid was maintained under selection for leucine. Expression from the vector was maintained by growth of cells on medium lacking thiamine, as the hemagglutinin (HA)-tagged protein is under control of the weakest derivative of the no message in thiamine (nmt1) promoter (6).
DNA constructs. Yeast two-hybrid vectors were generated by PCR amplification of the Chp1 ORF with primers bearing NcoI and BamHI sites, or of the Tas3 ORF with primers containing EcoRI and BamHI sites, digestion of the purified products, and cloning into NcoI/BamHI-digested pGBKT7 (Clontech) (Chp1) or EcoRI/BamHI-digested pGADT7 (Tas3). Truncated chp1 vectors were generated in a similar fashion, using primers that amplify DNA fragments encoding amino acids 1 to 409 and cloning into pGBKT7. A pGADT7-tas3 derivative construct was also generated by digesting pGADT7-tas3 with SacI (SacI cuts within tas3 and in the 3' polylinker region of the vector) and religation. The resulting plasmid contained coding sequence for amino acids 1 to 282 of tas3.
The pREP81-3xHA construct for expression in fission yeast was generated by cloning of annealed oligonucleotides with NdeI- and SalI-compatible ends that encode a 3xHA epitope tag and an NcoI site into NdeI/SalI-digested pREP81 (6). Release of NcoI/BamHI fragments from the pGBKT7 series and insertion into the NcoI/BamHI sites of the pREP81-3xHA vector placed the Chp1 ORF in frame with the N-terminal 3xHA tag and allowed simple shuttling of two-hybrid vector inserts into the S. pombe expression vector. The integrity of cloned DNA was verified by sequence analysis.
Protein interaction assays. (i) Yeast two-hybrid.
All budding yeast manipulations were performed according to the Clontech matchmaker yeast two-hybrid manual. pGADT7 and pGBKT7 derivative plasmids were cotransformed into AH109 budding yeast. Cultures were grown overnight at 30°C in SD-L-T, diluted back to an optical density at 600 nm of 1, and serially diluted (1:10) in distilled water. These dilutions were spotted on SD-L-T, SD-L-T-H-A plus (2.5 to 10 mM) 3-amino-1,2,4 triazole (3-AT) and 5-bromo-4-chloro-3-indolyl-
-D-galactopyranoside (X--Gal; Clontech). Plates were incubated for 4 days at 30 °C prior to being photographed.
(ii) Fission yeast. A total of 1.3 x 109 cells of PY 862, PY 906, and PY 974 were washed in phosphate-buffered saline (PBS) and resuspended in 1 ml of extraction buffer (50 mM HEPES [pH 7.6], 300 mM KO-acetate, 10% glycerol, 1 mM EDTA, 1 mM EGTA, 0.2% NP-40, 5 mM MgO-acetate, 1 mM dithiothreitol, 1 mM Na3VO4, 1 mM NaF, 1 mM benzamidine, and 1 mM phenylmethylsulfonyl fluoride, with Complete [EDTA-free] protease inhibitors [Roche]) prior to grinding in liquid N2 with a pestle and mortar until >70% lysis was achieved (assessed by microscopic examination). Cell lysates were centrifuged for 25 min at 30,000 rpm at 4°C in a Beckman TLA 120-2 rotor. Twenty microliters of supernatant was retained on ice as input material, and 40 µl of immunoglobulin G (IgG)-Sepharose beads (Amersham Biosciences; preblocked with 30 µg of bovine serum albumin) was added to the remainder of supernatant and incubated with rotation for 1 h at 4°C. Beads were washed five times with 1 ml of 80% extraction buffer-20% radioimmunoprecipitation assay buffer prior to boiling in 2x sample buffer and loading onto 8% prosieve (FMC) sodium dodecyl sulfate (SDS) gels and electrophoresis. Separated proteins were transferred onto nitrocellulose, and the blot was blocked in 5% milk-PBS-0.2% Tween (PBST) prior to probing with a 1:200 dilution of anti-myc antibody (9E10; Santa Cruz) in PBST and detection of bound antibody with horseradish peroxidase-linked anti-mouse antibody, followed by enhanced chemiluminescence and exposure to film.
ChIP.
Chromatin immunoprecipitation (ChIP) on fixed chromatin by using anti-Myc antibody (9E10; Santa Cruz) was performed as previously described (28). mat unique primers 5'-GGCAATACAACTTTGGCGATCATTTAC-3' and 5'-TGTTTAGCGCACTTTGATTTTCCAGTC-3' or Tel-specific primers 5'-CGATGCTCTCGACAAAGCCGTTCT-3' and 5'-CCATCTCAAACTTCTGTTCAACATT-3' (33) were used in multiplex PCR with cnp1+ primers 5'-GCCTGGAGATCCTATTCCACGGCC-3' and 5'-GAACGCTTCAGCCGCTTCCTGAAGAC-3' under conditions of exponential amplification (26 cycles). [-32P]dCTP was used to spike the PCR, and uptake of 32P into the PCR products was quantified by use of ImageQuant software after separation of PCR products on 4% polyacrylamide gels, drying of gels, and imaging on a Storm PhosphorImager (Molecular Dynamics Inc.). Relative IP values represent averages from at least two independent experiments.
Immunofluorescence. Fission yeast were grown overnight at 25°C to 5 x 106/ml and fixed with freshly prepared formaldehyde (3.8%) for 25 min. Permeabilized cells (18) were incubated with the following mouse monoclonal antibodies: anti-myc (1/100 dilution of 9E10; Santa Cruz), anti-HA (1/500; 12CA5; Roche), 1/15 anti-TAT1 (gift from K. Gull [42]) or 1/500 sheep anti-Cnp1 (gift from R. Allshire [23]), and secondary antibodies were all used at 1/100 (Texas red-conjugated anti-sheep or fluorescein isothiocyanate-conjugated anti-mouse antibody; Jackson Immunologicals). 4',6'-Diamidino-2-phenylindole (DAPI) was used to identify nuclei. Cells were cytospun onto polylysine-coated slides and photographed using a Zeiss Axioskop II microscope fitted with a Ludl filter wheel and chroma filters and a Coolsnap HQ camera (Photometrics). All images were taken at maximum resolution, using a 100x objective, with 100-ms exposure for fluorescein isothiocyanate, 10 ms for DAPI, and 100 ms for Texas red, and using IPLab software (Scanalytics). Images were imported into Adobe Photoshop and saved in CMYK format as Tiff files at 300-pixels/in. resolution. No manipulation of images was performed.
Defects in chromosome segregation were assessed in cells that had been stained with anti-TAT1 antibody (42) to reveal tubulin and DAPI staining of nuclei. Cells which exhibited spindle lengths of greater than 5 µm were analyzed for the presence of lagging chromosomes on the spindle (11). On average, 100 mitotic cells were analyzed.
RNA analysis. RNA was prepared from 20-ml cultures of yeast grown in YES at 32°C to 5 x 106 cells/ml. Strains bearing centromeric ura4+ insertions were grown at 25°C. cDNA was prepared by oligo(dT)-primed reverse transcription-PCR (RT-PCR), and competitive PCR of ura4+ and ura4DS/E was performed as described previously (13). ura4+ (U) levels were quantified relative to ura4-DS/E (L) and relative to wild-type strains. tas3-13xmyc+ transcript levels were measured using multiplex PCR with primers derived from cnp1+ (above) and tas3+ (5'-TCATTTTTCGTTTTTATGCTCTAAAATTGAAGCC-3' and 5'-CAACCTTTGCCTGAGCAGCTTATG-3') under exponential PCR conditions, and tas3-13xmyc+ transcript levels were normalized to cnp1+ transcript levels and quantified relative to wild-type strains.
Cell growth assays: serial dilution assay of cell growth of S. pombe. Serial (1:5) dilutions of cells grown overnight in PMG-leucine medium at 25°C to 5 x 106 cells/ml and washed in PMG with no supplements were spotted onto PMG-leucine medium, PMG-leucine medium supplemented with 2 mg of FOA/ml, or PMG medium lacking both uracil and leucine. Cultures were grown for 5 days at 25°C. A total of 1.2 x 104 cells were contained in the first spot.
To assess the maintenance of silencing at the mat2/3 locus, serial (1:5) dilutions of cells grown overnight in YES medium at 25°C to 5 x 106 cells/ml and washed in PMG with no supplements were spotted onto PMG medium containing reduced adenine (10%) and incubated for 5 days at 25°C. A total of 1.2 x 104 cells were contained in the first spot.
TSA treatment. Cells were grown to 5 x 106 cells/ml in YES medium at 32°C, prior to dilution to 5 x 103 cells/ml in YES containing 35 µg of trichostatin A (TSA; A.G. Scientific)/ml and growth for 10 generations at 32°C. Cells were then allowed to recover for an additional 10 generations in YES alone, prior to washing in PMG with no supplements, serial (1:5) dilution of cells, spotting onto PMG medium lacking adenine or with reduced adenine (10% adenine), and incubation at 25°C for 5 days. A total of 1.2 x 104 cells were contained within the first spot.
|
RESULTS |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
To confirm that Chp1 indeed associates with sites of heterochromatin other than centromeres, we performed ChIP with cells expressing Chp1-6xmyc and probed the immunoprecipitates for other heterochromatic DNA sequences. The silent mating type locus (mat2/3) is enriched in lysine 9-methylated histone H3 chromatin between mat2 and mat3, and within this domain there is an extensive region of very high homology to sequences from the outer repeats of the centromere (the cenH [for centromere homology] region) (17). To determine whether Chp1 associates with sequences unique to the mating type locus, primers from sequences more than 800 bp distal of cenH were used for PCR amplification of ChIP material. These unique sequences (Mat unique) were enriched in the anti-myc immunoprecipitates of chp1-6xmyc+ strains relative to the euchromatic control (cnp1+), or to immunoprecipitates prepared from non-myc-tagged strains (Fig. 1B). Similarly, sequences from the telomeric region were enriched in the anti-myc immunoprecipitates of chp1-6xmyc+ strains (Fig. 1B, lower panel). Thus, Chp1 associates with unique sequences from the mat2/3 region, with telomeres, and with the outer repeat sequences of the centromere (28, 39), as recently suggested by others (27, 32).
The chromodomain, but not the RNA recognition motif (RRM), is required for Chp1 function.
To define regions of Chp1 necessary for its activity and localization, we tested whether the two recognizable domains, the chromodomain (amino acids 2 to 79) and a putative RRM (amino acids 311 to 407) are essential for Chp1 function (Fig. 2A). We constructed S. pombe strains in which either the RRM or the chromodomain was deleted from chp1-6xmyc+ (28). These deletions were performed using a two-stage transformation process, which relies on the high efficiency of homologous recombination in fission yeast. Initially the chromodomain was replaced with the ura4+ marker gene, and then removal of ura4+ and the generation of CDchp1-6xmyc cells was achieved by retransformation with a chimeric PCR product containing only sequences that flank the chp1 chromodomain. A centromeric marker gene (cen::otr-ura4+) (2) was subsequently introduced into CDchp1-6xmyc cells. RRMchp1-6xmyc cells were derived in a similar fashion. Levels of the mutant Chp1 proteins expressed in these strains were similar to those present in wild-type Chp1-6xmyc yeast (Fig. 2B), indicating that the mutations do not alter the stability or expression of Chp1 in vivo.
|
) exhibited 80% derepression of transcription of the normally silent centromeric marker gene (cen::otr-ura4+), suggesting that outer repeat heterochromatin is largely disrupted in this mutant background (28, 29, 39). This loss of centromeric silencing was accompanied by a high incidence of cells displaying mitotic defects, as approximately 30% of late-anaphase cells exhibited lagging chromosomes (Fig. 2D) (10). Importantly, cells lacking only the Chp1 chromodomain (CDchp1-6xmyc) behaved similar to the chp1 null strain, with elevated transcription of the centromeric marker gene (Fig. 2C) and numerous mitotic chromosome segregation defects (Fig. 2D). In contrast, cells lacking the RRM domain of Chp1 (RRMchp1-6xmyc) exhibited no loss of Chp1 function.
To address whether the chromodomain or RRM might direct Chp1 localization (Fig. 2E), we compared the localization of Chp1 in these mutant strains to that of wild-type yeast. In wild-type cells, Chp1-6xmyc localizes to several spots at the nuclear periphery when assessed by indirect immunofluorescence with anti-myc antibodies (Fig. 1A and 2E). This pattern of Chp1-6xmyc localization is disrupted in a rik1 background, in which Chp1 chromatin association is lost (28) and where Chp1-6xmyc appears to accumulate in the nucleolus (Fig. 2E). RRMchp1-6xmyc showed a normal pattern of localization, with several discrete spots at the nuclear periphery (Fig. 2E). In contrast, CDchp1-6xmyc exhibited a diffuse faint spotty staining pattern throughout the nucleoplasm and was not associated with chromatin (Fig. 2E). Thus, the chromodomain, but not the RRM, is essential for Chp1 localization to all sites of heterochromatin and for normal Chp1 function at centromeric sequences.
The C-terminal domain of Chp1 is also essential for Chp1 function.
While these findings established that the chromodomain of Chp1 is necessary for targeting Chp1 to chromatin, it was unclear whether it was alone sufficient to direct Chp1 function and localization. To answer this question, we transformed chp1-null cells with plasmids encoding an N-terminally 3xHA epitope-tagged version of full-length or C-terminally truncated Chp1 (Fig. 3A). A plasmid encoding full- length Chp1 (pREP81-3xHA-Chp1) fully complemented the chp1 phenotype, as it reestablished silencing of a centromeric marker gene (conferring growth on FOA medium, which is toxic to ura4+-expressing cells) (Fig. 3B), accurate segregation of chromosomes (Fig. 3C), and proper targeting of 3xHA-Chp1 to typical spots at the nuclear periphery (Fig. 3D, left panel). Using this plasmid-based system, we then asked if sequences C-terminal to the RRM of Chp1 are important for Chp1 function. In contrast to the plasmid bearing full-length Chp1, expression of residues 1 to 409 of Chp1 (3xHA-chp11-409) completely failed to complement chp1 function. These cells exhibited high levels of chromosome segregation defects (Fig. 3C) and defects in centromeric silencing (Fig. 3B). Consistent with this result, immunolocalization of 3xHA-chp11-409 with anti-HA antibody revealed that the truncated protein was diffusely localized throughout the chromatin and nucleolus (Fig. 3D). Staining was not observed in all cells, and the high levels of immunofluorescence in others may have been caused by differential plasmid accumulation. Steady-state levels of the full-length and truncated Chp1 proteins were similar within the cell population (Fig. 3E). Therefore, the C-terminal domain of Chp1 is required for heterochromatin localization and function of Chp1. However, in the absence of its chromodomain, expression of the C terminus of Chp1 is not sufficient to drive heterochromatin localization or function of Chp1 (Fig. 3); thus, both the chromodomain and the C-terminal domain are essential for Chp1 localization and function.
|
|
Since reagents allowing for a direct evaluation of the association of Ago with Chp1 and Tas3 are not available, we addressed what effect loss of Ago1 had on the localization of Tas3 and Chp1 (Fig. 5A and B, middle panels). As expected (39), in ago1 cells there is a loss of centromere-associated Chp1 and Tas3, as revealed by loss of costaining of Chp1-6xmyc (Fig. 5A) and Tas3-13xmyc (Fig. 5B) with Cnp1-stained centromeres. Surprisingly, however, other foci of Chp1-6xmyc and Tas3-13xmyc persisted in ago1 cells. Similar results were obtained in cells from which Dicer (dcr1) was deleted (Fig. 5A and B, right panels). Thus, in either the absence of Ago1 or of siRNAs, Tas3 and Chp1 association with centromeres is lost, but Tas3 and Chp1 appear to still bind other sites of heterochromatin.
|
and dcr1 backgrounds reflected the maintenance of association of these proteins with other sites of heterochromatin, we performed ChIPs. Indeed, the unique sequences from mat2/3 and telomeric DNA sequences remained clearly enriched in anti-myc-immunoprecipitated material from ago1 or dcr1 strains expressing either Chp1-6xmyc or Tas3-13xmyc (Fig. 5C and D), whereas their association with centromeric heterochromatin was lost (Fig. 5A and B) (39). Therefore, two pools of Chp1 and Tas3 are present within fission yeast: one that is recruited to centromeres in an Ago1- and siRNA-dependent fashion, and a second that associates with noncentromeric heterochromatin independently of the RNAi pathway.
We also assessed the localization of Chp1-6xmyc and Tas3-13xmyc in cells lacking tas3 and chp1, respectively (Fig. 6A). In cells with tas3 deleted, Chp1 was delocalized, with a cloud of small foci of staining that mainly accumulated over the nucleolus (Fig. 6A); therefore, in addition to the known requirement of Tas3 for Chp1 localization to centromeres, Tas3 is also required for localization of Chp1 to all other sites of heterochromatin. Strikingly, Tas3-13xmyc staining was completely absent in cells in which chp1 was deleted (Fig. 6A). To address whether this was due to a loss of Tas3 protein in chp1 cells, we determined the steady-state levels of the Tas3-13xmyc protein in various deletion backgrounds. In accordance with the immunolocalization data, we found that Tas3-13xmyc protein was essentially lost in chp1 cells, whereas it was present at normal levels in the ago1 cells (Fig. 6B). Chp1-6xmyc levels were unaffected by deletion of tas3+ or of ago1+.
|
cells was due to suppression of tas3-13xmyc+ transcription, we performed quantitative RT-PCR using RNA prepared from wild-type, chp1, and ago1 backgrounds. Deletion of chp1+ had no effect on tas3-13xmyc+ transcript levels (Fig. 6C); therefore, the loss of Tas3 protein in chp1 null cells is a posttranscriptional effect.
Tas3 and Chp1 interact independently of Ago1.
The RNAi-independent association of Chp1 and Tas3 with noncentromeric heterochromatin suggests that Chp1 and Tas3 may exist as a complex that acts independently of Ago1 and siRNAs. To test this notion, yeast two-hybrid analyses using a Gal4-DNA binding domain (GBD) fusion of Chp1 and GAL4 activation domain (GAD) fusion of Tas3 were performed in AH109 budding yeast bearing integrated GAL4 site-dependent reporters. All three reporter genes were strongly induced, allowing growth of the yeast on medium lacking adenine and histidine and blue coloration of the colonies from metabolism of X--Gal, included in the selective medium (Fig. 7A). The induction of HIS3 expression was sufficient to allow growth of the yeast on media containing up to 10 mM 3-AT (data not shown), which is a competitive inhibitor of the His3 protein. Specificity of the reporter gene induction was confirmed by the absence of growth on selective media when either Chp1-GBD or Tas3-GAD protein fusions were tested in combination with the empty GAD or GBD vectors, respectively. It is unlikely that endogenous budding yeast proteins facilitate the interaction of Tas3 and Chp1, as budding yeast lack all known components of the RNAi pathway. Finally, interaction mapping studies demonstrated that the C-terminal portion of Chp1 (residues 407 to 960), which is essential for Chp1 function (Fig. 3), was sufficient for interaction with Tas3, whereas the N-terminal domain of Tas3 (residues 1 to 282) was sufficient for interaction with Chp1 (Fig. 7A).
|
strains. Protein extracts were prepared from strains expressing both a functional C-terminal TAP-tagged genomic Chp1 fusion protein (Chp1-TAP) and Tas3-13xmyc. IgG-Sepharose was used to immunoprecipitate Chp1-TAP complexes, and immunoprecipitated material was probed for Tas3-13xmyc by blotting with anti-myc antibodies (Fig. 7B). Immunoprecipitation of lysate from cells expressing Tas3-13xmyc and nontagged Chp1 was performed as a control for specificity of the IgG-Sepharose pull down (Fig. 7B, lane 4). Chp1-TAP coimmunoprecipitated with Tas3-13xmyc (lane 5), and this interaction was also evident in extracts from ago1 cells (lane 6). Therefore, Tas3 and Chp1 are also present in an RNAi-independent complex in S. pombe and bind noncentromeric heterochromatin regions in the absence of Ago1. Chp1 is required for establishment, and to a lesser extent maintenance, of mat2/3 silencing. To assess the role of Chp1 at the mat2/3 locus, we first examined the impact of deletion of chp1+on silencing of the ade6+ marker gene inserted close to mat3 (at the EcoRV site) (Fig. 8A) (37). The efficiency of maintenance of silencing was determined by plating serial dilutions of cells onto plates limiting in adenine. Wild-type cells efficiently silence mat3::ade6+ expression, giving rise to red colonies on plates limiting for adenine (37), whereas yeast bearing deletions in clr4+ or rik1+ alleviate silencing at the mat2/3 locus (14) and produce white colonies. Nonetheless, deletion of chp1+ or ago1+ only slightly alleviated silencing (20, 38), as revealed by the presence of pink colonies compared with the red, fully repressed colonies (Fig. 8A).
|
cells bearing the mat3::ade6+ marker on medium lacking adenine was observed than in similarly treated chp1+ cells. Similar findings were evident in ago1 cells, and no further enhancement of growth on media lacking adenine was observed for ago1 chp1 cells; therefore, deletion of chp1+ has no additional impact on the assembly of heterochromatin at mat2/3 sequences compared to the removal of Ago1 alone. We conclude that the RNAi-independent association of Chp1 with mat2/3 has little consequence on the establishment of heterochromatin on these sequences. |
DISCUSSION |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Firstly, our observation that the RNAi-independent form of the Chp1-Tas3 complex localizes to both telomeres and the mat2/3 locus suggests that this complex plays a role in regulating heterochromatin at these sites. The mat2/3 locus includes the 4.3-kb cenH region and here, like at centromeres, the establishment of heterochromatin is dependent upon the RNAi pathway (20). However, the establishment of heterochromatin at mat3 proximal sequences (Mat unique) is RNAi independent and results from direct recruitment of histone-modifying enzymes by the Atf1/Pcr1 transcription factors (Fig. 9) (20, 21). Importantly, our finding that Chp1 binds to telomeres and to the mating type locus independent of the RNAi apparatus suggests that Chp1 can be recruited to genomic regions that are methylated on lysine 9 of histone H3 through RNAi-independent mechanisms. The mating type locus therefore provided a useful system to probe the function of Chp1 independent of the RNAi pathway and allowed us to address whether localization of Chp1 (and Tas3) to these sequences is directly responsible for silencing.
Previous work testing the role of Chp1 in silencing of the mat2/3 locus showed a minor role for Chp1 in the maintenance of silencing at mating type sequences (38) and, similarly, a minor role for components of the RNAi pathway (20). By use of an ago1 strain, we confirmed that deletion of RNAi pathway components has little effect on the maintenance of silencing of the mat2/3 locus (Fig. 8A) (20). Moreover, there were no significant differences in the maintenance of silencing in ago1, chp1, and ago1chp1 strains, and in these cells treated with TSA there were no significant differences in the establishment of silencing at the mating type locus (Fig. 8B). Therefore, Chp1 that binds independently of the RNAi pathway to unique sequences at mat2/3 plays no additional role in either the maintenance or establishment of heterochromatin at the mating type locus, in contrast to the essential RNAi-dependent role of Chp1 in the maintenance and establishment of heterochromatin and silencing at cenH and centromeres.
In principle our studies agree with those of others that have examined the role of Chp1 in the establishment of heterochromatin at multiple loci and which suggest that Chp1 recruitment promotes silencing at mating type locus sequences inserted upstream of the ura4+ marker gene at a euchromatic site (32). Wild-type cells efficiently nucleate heterochromatin at this ectopic location and silence ura4+ expression, but chp1 null or RNAi mutants fail to establish heterochromatin on these sequences (3, 20, 32). The sequences used in this assay were, however, derived from the cenH region within mat2/3 that shares 96% identity with centromeric sequences (3, 17). Our demonstration that the RNAi-independent binding of Tas3 and Chp1 to unique sequences at mat2/3 does not support silencing, whereas association of RITS with centromeric or cenH sequences is required for establishment of silencing (20, 29, 32, 39, 40), suggests that the distinguishing factors, Ago1 and siRNAs, are key to understanding the mechanism by which RITS promotes silencing of centromeric sequences, through recruitment of the Clr4 histone methyltransferase.
The data that we have presented complement a very recent publication by Noma et al. (27). In accordance with our data, they demonstrated association of RITS components with heterochromatic loci and Tas3 and Chp1 association with the mat2/3 locus in the absence of dicer. In addition, they identified cenH transcripts which are elevated in swi6 and dcr1 mutant backgrounds and further increased in the compound mutant background in which mat2/3 silencing is lost. They attributed the maintenance of silencing in the single mutant backgrounds to posttranscriptional, as well as transcriptional, silencing mediated by RITS tethered in cis at the mat2/3 locus (27). An alternative explanation might be that only when both RNAi-dependent and RNAi-independent silencing are lost do transcript levels rise sufficiently at mat2/3 to cause loss of heterochromatin formation.
In agreement with our model for K9-MeH3-mediated recruitment of Chp1 and Tas3 to mat unique sequences, Noma et al. demonstrated a correlation between the binding activity of Chp1 and Tas3 with the extent of K9-MeH3 at mat2/3 in clr4 mutant backgrounds. Chp1 copurified with siRNAs from clr4 mutant strains in which Chp1 and Tas3 maintained some association with mat2/3, but not from clr4 deletion strains in which histone H3 K9 methylation and Chp1 chromatin association were lost. These and other data derived from use of a mutant Chp1 that does not bind chromatin led to the conclusion that RITS association with chromatin is required for efficient processing of transcripts at heterochromatic loci and for generation and recruitment of siRNAs into the RITS complex (27). However, our study suggests that RITS localization is not the sole determinant of silencing efficiency at mat2/3, as the silencing defect in strains lacking multiple RITS components is not as profound as in strains lacking clr4+ (Fig. 8 and data not shown). Deletion of clr4+ perturbs both the RNAi-dependent and RNAi-independent heterochromatin formation at mat2/3, whereas loss of RITS components affects only RNAi-dependent silencing.
Whether the Chp1-Tas3 complex plays a passive or active role in heterochromatin regulation remains to be resolved, as although deletion of chp1+ has little consequence on the RNAi-independent formation of heterochromatin at the mat2/3 locus, it is feasible that it plays an important role in the establishment and/or maintenance of heterochromatin at other sites, for example, at the telomeres or rRNA gene repeats. Specifically, deletion of chp1+ perturbs silencing of a polymerase II-transcribed marker inserted within the rRNA locus and, although chp1+ loss or mutation of the RNAi pathway does not influence the maintenance of silencing at telomeric sequences (19, 38), a role for Chp1 in the formation of heterochromatin at telomeric sequences has recently been described (32).
A second possibility is that the Chp1-Tas3 complex represents an intermediate complex that is stockpiled at noncentromeric heterochromatic loci and is thus readily available for interaction with siRNA-loaded Ago1. After binding, this forms a functional RITS complex, which can then be recruited to centromeric DNA sequences by siRNA-mediated targeting. At the mat2/3 locus, the close proximity of the cenH sequences linked in cis to the store of Tas3-Chp1 complex suggests an appealing model for a rapid switch of the Tas3-Chp1 complex to silencing competent RITS on binding to Ago1 and recruitment, perhaps by spreading, onto cenH sequences. By extrapolation, the accumulation of the Chp1-Tas3 complex at telomeres may provide a source for loading of RITS onto centromeric sequences during the switch in chromosomal positioning between mitotic and meiotic cell cycles. During mitotic interphase the centromeres are clustered at the spindle pole body, and at the onset of meiotic prophase their position at the spindle pole is taken by telomeres and the centromeres dissociate (8). After meiotic prophase, the chromosomes resume the mitotic interphase configuration, with centromeres associated with the spindle pole. Depending on the timing of centromeric transcription, and how siRNA formation is regulated during meiosis, such chromosomal movements may facilitate transfer of RITS to centromeric sequences.
Our model predicts that Ago1 is not an obligate component of the Chp1 and Tas3 complex and, recently, it has been found that Ago1, but not Tas3 or Chp1, is required for posttranscriptional gene silencing (34). Our data support a model whereby Ago1 can associate with the nuclear Chp1-Tas3 complex to form RITS and direct transcriptional silencing, or it can associate with cytoplasmic factors to drive posttranscriptional gene silencing. If Ago1 were limiting within the cell, such flexibility would allow for a rapid response to changes in signaling pathways.
The mechanism by which Chp1 binds to either centromeres or other heterochromatic loci is poorly understood. The chromodomain of Chp1 is not required for the association of Tas3 with Chp1, yet it is essential for localization of Chp1 to heterochromatin. The chromodomain of Chp1 binds to K9-MeH3 peptides in vitro (29), and Chp1 localization is dependent on the Clr4 histone methyltransferase (28). This, together with the colocalization of Chp1 with Swi6 at sites of heterochromatin (this study and references 27, 28, and 32), implies that the K9-MeH3 interaction mediated by Chp1's chromodomain contributes to its association with heterochromatin. However, as recently suggested (24), the interaction of the chromodomain with K9-MeH3 in native chromatin may not be of high enough affinity to tether proteins to heterochromatin, and now clearly in the case of Chp1, interaction with Tas3 also plays a crucial role in Chp1's ability to associate with heterochromatin.
|
ACKNOWLEDGMENTS |
|---|
This work was supported in part by Cancer Center Support Grant Developmental funds 2 P30 CA 21765-25 and 2 P30 CA 21765-26 and the American Lebanese Syrian Associated Charities of St. Jude Children's Research Hospital.
|
FOOTNOTES |
|---|
V.J.P. and J.D.W. contributed equally to this work.
|
REFERENCES |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
2. Allshire, R. C., E. R. Nimmo, K. Ekwall, J. P. Javerzat, and G. Cranston. 1995. Mutations derepressing silent centromeric domains in fission yeast disrupt chromosome segregation. Genes Dev. 9:218-233.
3. Ayoub, N., I. Goldshmidt, R. Lyakhovetsky, and A. Cohen. 2000. A fission yeast repression element cooperates with centromere-like sequences and defines a mat silent domain boundary. Genetics 156:983-994.
4. Bahler, J., J. Q. Wu, M. S. Longtine, N. G. Shah, A. McKenzie III, A. B. Steever, A. Wach, P. Philippsen, and J. R. Pringle. 1998. Heterologous modules for efficient and versatile PCR-based gene targeting in Schizosaccharomyces pombe. Yeast 14:943-951.[CrossRef][Medline]
5. 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 chromodomain. Nature 410:120-124.[CrossRef][Medline]
6. Basi, G., E. Schmid, and K. Maundrell. 1993. TATA box mutations in the Schizosaccharomyces pombe nmt1 promoter affect transcription efficiency but not the transcription start point or thiamine repressibility. Gene 123:131-136.[CrossRef][Medline]
7. Boeke, J. D., J. Trueheart, G. Natsoulis, and G. R. Fink. 1987. 5-Fluoroorotic acid as a selective agent in yeast molecular genetics. Methods Enzymol. 154:164-175.[Medline]
8. Chikashige, Y., D. Q. Ding, H. Funabiki, T. Haraguchi, S. Mashiko, M. Yanagida, and Y. Hiraoka. 1994. Telomere-led premeiotic chromosome movement in fission yeast. Science 264:270-273.
9. Dernburg, A. F., and G. H. Karpen. 2002. A chromosome RNAissance. Cell 111:159-162.[CrossRef][Medline]
10. Doe, C. L., G. Wang, C. Chow, M. D. Fricker, P. B. Singh, and E. J. Mellor. 1998. The fission yeast chromo domain encoding gene chp1+ is required for chromosome segregation and shows a genetic interaction with alpha-tubulin. Nucleic Acids Res. 26:4222-4229.
11. Ekwall, K., J. P. Javerzat, A. Lorentz, H. Schmidt, G. Cranston, and R. Allshire. 1995. The chromodomain protein Swi6: a key component at fission yeast centromeres. Science 269:1429-1431.
12. Ekwall, K., E. R. Nimmo, J. P. Javerzat, B. Borgstrom, R. Egel, G. Cranston, and R. Allshire. 1996. Mutations in the fission yeast silencing factors clr4+ and rik1+ disrupt the localisation of the chromodomain protein Swi6p and impair centromere function. J. Cell Sci. 109(Pt. 11):2637-2648.[Abstract]
13. Ekwall, K., T. Olsson, B. M. Turner, G. Cranston, and R. C. Allshire. 1997. Transient inhibition of histone deacetylation alters the structural and functional imprint at fission yeast centromeres. Cell 91:1021-1032.[CrossRef][Medline]
14. Ekwall, K., and T. Ruusala. 1994. Mutations in rik1, clr2, clr3 and clr4 genes asymmetrically derepress the silent mating-type loci in fission yeast. Genetics 136:53-64.[Abstract]
15. Elgin, S. C., and S. I. Grewal. 2003. Heterochromatin: silence is golden. Curr. Biol. 13:R895-R898.[CrossRef][Medline]
16. Funabiki, H., I. Hagan, S. Uzawa, and M. Yanagida. 1993. Cell cycle-dependent specific positioning and clustering of centromeres and telomeres in fission yeast. J. Cell Biol. 121:961-976.
17. Grewal, S. I., and A. J. Klar. 1997. A recombinationally repressed region between mat2 and mat3 loci shares homology to centromeric repeats and regulates directionality of mating-type switching in fission yeast. Genetics 146:1221-1238.[Abstract]
18. Hagan, I. M., and J. S. Hyams. 1988. The use of cell division cycle mutants to investigate the control of microtubule distribution in the fission yeast Schizosaccharomyces pombe. J. Cell Sci. 89(Pt. 3):343-357.
19. Hall, I. M., K. Noma, and S. I. Grewal. 2003. RNA interference machinery regulates chromosome dynamics during mitosis and meiosis in fission yeast. Proc. Natl. Acad. Sci. USA 100:193-198.
20. Hall, I. M., G. D. Shankaranarayana, K. Noma, N. Ayoub, A. Cohen, and S. I. Grewal. 2002. Establishment and maintenance of a heterochromatin domain. Science 297:2232-2237.
21. Jia, S., K. Noma, and S. I. Grewal. 2004. RNAi-independent heterochromatin nucleation by the stress-activated ATF/CREB family proteins. Science 304:1971-1976.
22. Kim, H. S., E. S. Choi, J. A. Shin, Y. K. Jang, and S. D. Park. 2004. Regulation of Swi6/HP1-dependent heterochromatin assembly by cooperation of components of the mitogen-activated protein kinase pathway and a histone deacetylase Clr6. J. Biol. Chem. 279:42850-42859.
23. Kniola, B., E. O'Toole, J. R. McIntosh, B. Mellone, R. Allshire, S. Mengarelli, K. Hultenby, and K. Ekwall. 2001. The domain structure of centromeres is conserved from fission yeast to humans. Mol. Biol. Cell 12:2767-2775.
24. Meehan, R. R., C. F. Kao, and S. Pennings. 2003. HP1 binding to native chromatin in vitro is determined by the hinge region and not by the chromodomain. EMBO J. 22:3164-3174.[CrossRef][Medline]
25. Mellone, B. G., L. Ball, N. Suka, M. R. Grunstein, J. F. Partridge, and R. C. Allshire. 2003. Centromere silencing and function in fission yeast is governed by the amino terminus of histone H3. Curr. Biol. 13:1748-1757.[CrossRef][Medline]
26. Moreno, S., A. Klar, and P. Nurse. 1991. Molecular genetic analysis of fission yeast Schizosaccharomyces pombe. Methods Enzymol. 194:795-823.[Medline]
27. Noma, K. I., T. Sugiyama, H. Cam, A. Verdel, M. Zofall, S. Jia, D. Moazed, and S. I. Grewal. 2004. RITS acts in cis to promote RNA interference-mediated transcriptional and post-transcriptional silencing. Nat. Genet. 36:1174-1180.[CrossRef][Medline]
28. Partridge, J. F., B. Borgstrom, and R. C. Allshire. 2000. Distinct protein interaction domains and protein spreading in a complex centromere. Genes Dev. 14:783-791.
29. Partridge, J. F., K. S. Scott, A. J. Bannister, T. Kouzarides, and R. C. Allshire. 2002. cis-acting DNA from fission yeast centromeres mediates histone H3 methylation and recruitment of silencing factors and cohesin to an ectopic site. Curr. Biol. 12:1652-1660.[CrossRef][Medline]
30. Provost, P., R. A. Silverstein, D. Dishart, J. Walfridsson, I. Djupedal, B. Kniola, A. Wright, B. Samuelsson, O. Radmark, and K. Ekwall. 2002. Dicer is required for chromosome segregation and gene silencing in fission yeast cells. Proc. Natl. Acad. Sci. USA 99:16648-16653.
31. Reinhart, B. J., and D. P. Bartel. 2002. Small RNAs correspond to centromere heterochromatic repeats. Science 297:1831.
32. Sadaie, M., T. Iida, T. Urano, and J. Nakayama. 2004. A chromodomain protein, Chp1, is required for the establishment of heterochromatin in fission yeast. EMBO J. 23:3825-3835.[CrossRef][Medline]
33. Sadaie, M., T. Naito, and F. Ishikawa. 2003. Stable inheritance of telomere chromatin structure and function in the absence of telomeric repeats. Genes Dev. 17:2271-2282.
34. Sigova, A., N. Rhind, and P. D. Zamore. 2004. A single Argonaute protein mediates both transcriptional and posttranscriptional silencing in Schizosaccharomyces pombe. Genes Dev. 18:2359-2367.
35. Takahashi, K., E. S. Chen, and M. Yanagida. 2000. Requirement of Mis6 centromere connector for localizing a CENP-A-like protein in fission yeast. Science 288:2215-2219.
36. Tasto, J. J., R. H. Carnahan, W. H. McDonald, and K. L. Gould. 2001. Vectors and gene t
