A Novel Mechanism of Antagonism between ATP-Dependent Chromatin Remodeling Complexes Regulates RNR3 Expression

Strains and media.The Saccharomyces cerevisiae strains used in this study are listed in Table 1. Cells were grown at 30°C in YPD (1% yeast extract, 2% peptone, and 2% dextrose) medium supplemented with 0.05 mg/ml adenine. Where indicated, methyl methanesulfonate (MMS) was added to give a final concentration of 0.03% for 2.5 h. Gene deletion and epitope tagging were carried out by homologous recombination using PCR-generated cassettes (30). For the ISW2 reexpression studies, cells were grown to an optical density at 600 nm (OD₆₀₀) of 1.0 in YP (1% yeast extract, 2% peptone) containing 2.5% raffinose, and then galactose was added to give a final concentration of 3% for the times indicated in Fig. 4.

Northern blotting and chromatin mapping.Cells from 10 ml of yeast culture (OD₆₀₀ = 0.7) were harvested for total RNA extraction. Fifteen micrograms of total RNA was separated on 1.2% formaldehyde-containing agarose gels and transferred to a Hybond-XL membrane (GE Biosciences, Piscataway NJ) by capillary blotting. Probes for RNR3 or Scr1 were prepared by PCR. The signal of scR1 (small cytoplasmic RNA) in each sample was used to correct for recovery and loading of RNA. The details of this method have been published elsewhere (32). MNase mapping of nucleosomes by indirect end labeling was carried out using a published protocol (50).

ChIP.The chromatin immunoprecipitation (ChIP) assay was performed essentially as described in previous publications (35, 48). One hundred milliliters of yeast culture (OD₆₀₀ = 0.7 to 1.0) was treated with formaldehyde (1% [vol/vol]) for 15 min at room temperature, and cross-linking was quenched by the addition of glycine to give 125 mM. Whole-cell extracts were prepared by glass bead disruption and sheared into fragments averaging 300 to 600 bp in size by using a Bioruptor (Diagenode, Philadelphia PA). Whole-cell extracts were immunoprecipitated with antibodies indicated throughout the text. RNAPII antibody was obtained from commercial sources (8WG16; Covance, Berkeley CA). The following antibodies were raised in rabbits and were described in previous publications (51): TBP (full-length), TAF1 (amino acids 1 to 225), Tup1 (full-length), and Swi2 (amino acids 1 to 851). The protein-immune complexes were recovered using 30 μl of protein A-Sepharose CL-4B beads (GE Biosciences, Piscataway, NJ). The immunoprecipitated DNA and input DNA were analyzed by semiquantitative PCR with primers directed toward RNR3 (36, 48). The percent immunoprecipitation was calculated, and data are expressed relative to the cross-linking observed in untreated wild-type cells. Data are presented as the means and standard deviations of results from at least three independent experiments. For detecting nucleosome eviction over the RNR3 promoter, cells were processed as described above except that the chromatin was sheared extensively into fragments averaging 200 base pairs in length. Chromatin was precipitated using an antibody to the core domain of H3 (Abcam, Cambridge, MA). The oligonucleotides used in this study are listed in Table 2.

Examination of the functional interaction between ISW2 and SWI/SNF.Two ATP-dependent chromatin remodeling complexes regulate the chromatin structure at RNR3. ISW2 is required for placement of nucleosomes into precise translational positions and contributes to the repression of transcription, and SWI/SNF is required for the eviction of nucleosomes and activation of transcription (35, 52, 53). Since these two complexes have opposing actions on chromatin, we examined the functional relationship between these two complexes. Strains containing isw2Δ, swi2Δ, and double isw2Δ swi2Δ mutations were constructed. The double mutant grew significantly less well than either of the single mutants, indicating that the combined mutations cause severe “synthetic sickness” (Fig. 1A). Synthetic lethality between swi2Δ mutants and components of the RSC complex and CHD1 has been noted previously, suggesting that functional interactions between chromatin remodeling complexes extend beyond SWI/SNF and ISW2 (5, 43).

• Open in new tab
• Download powerpoint

FIG. 1.

Interplay between ISW2 and SWI/SNF at RNR3. (A) Spot test for strain growth. Cells were spotted and grown at 30°C on YPD plates for 24 and 48 h. (B) Northern blotting for RNR3 mRNA. Wild-type (WT) and mutant cells were treated (+) or not treated (−) with 0.03% MMS for 2.5 h, and mRNA levels were detected by Northern blotting. scR1 is a loading control. The level of mRNA is indicated below the panel and is expressed relative to the signal of that from untreated wild-type cells observed after correction for loading.

Next, we examined the expression of RNR3 in the mutants under repressing (−MMS) and induced (+MMS) conditions in the mutants. Deleting ISW2 caused a small increase in the level of mRNA under the uninduced condition but had no effect on the induction of RNR3 by MMS (Fig. 1B, lanes 3 and 4). These results are consistent with its role in repression and nucleosome positioning. In contrast, the activation of RNR3 was essentially eliminated in the swi2Δ strain. Interestingly, the double mutant displayed the same phenotype as the swi2Δ mutant, indicating that deleting ISW2 cannot suppress the requirement for SWI/SNF in the activation of RNR3. The slight increase in the level of MMS-induced mRNA in the double mutant, compared to the level in the single swi2Δ mutant, is not seen consistently. This is in contrast to the results for the GAL1 and INO1 genes, where deleting ISW2 suppressed the transcription defects of a swi2Δ mutant (22). Thus, the nature of the functional interaction between ISW2 and SWI/SNF is gene specific.

The chromatin structure over RNR3 was then examined. MNase mapping and histone H3 cross-linking were used to determine the effects of the mutations on nucleosome positioning and eviction, respectively. MNase mapping cannot distinguish random nucleosome placement from eviction, and examination of remodeling by the ChIP assay using antibodies to core histones cannot definitely detect random positions of nucleosomes, because the overall signal across a region may diminish somewhat but will remain high. Thus, the combined use of MNase mapping and H3 cross-linking provides a clearer picture of chromatin structure. RNR3 is packaged into precisely positioned nucleosomes in the repressed state, which is obvious from the internucleosomal hypersensitive sites regularly spaced over the promoter and the 5′ end of the open reading frame (Fig. 2A). As described previously (53), deleting ISW2 caused a digestion pattern that is significantly different from that of wild-type cells; the internucleosomal hypersensitive sites were diminished, indicating that nucleosome positioning was lost. On the other hand, deleting swi2Δ had no effect on nucleosome positions in untreated cells (Fig. 2A). We next examined if the loss of nucleosome positioning in the isw2Δ mutant is dependent upon SWI/SNF. The double mutant had a digestion pattern indistinguishable from that of isw2Δ, indicating that SWI/SNF is not required for depositioning of nucleosomes in the isw2Δ cells under normal growth conditions.

• Open in new tab
• Download powerpoint

FIG. 2.

Chromatin structure and nucleosome density in ATP-dependent chromatin remodeling complex mutants. (A). MNase mapping of nucleosome positions at RNR3. The approximate positions of nucleosomes are indicated on the left. The URSs, which contain the binding sites for Crt1, are indicated within the graphic. The TATA box resides within nucleosome −1 and is at position −75 relative to the start site of transcription (25). ND, digested naked DNA; WT, wild type. (B) ChIP analysis of histone H3 cross-linking over the core promoter of RNR3. Wild-type and mutant cells were treated (+) or not treated (−) with 0.03% MMS for 2.5 h and then cross-linked with formaldehyde. Data are represented as percentages of immunoprecipitated (IP) signal versus input. Data are presented as the means and standard deviations of results from at least three independent experiments. (C) Same as in panel B, except primers directed to a subtelomeric region (TEL) or the ISW2-regulated gene REC104 were used to amplify the DNA.

H3 cross-linking to the promoter of RNR3 was measured using primers directed within nucleosome −1, which resides over the TATA box (48). Activation of gene expression leads to a three- to fourfold reduction in H3 cross-linking in wild-type cells, indicating a loss of the promoter nucleosome (Fig. 2B). Cross-linking was only slightly reduced in the isw2Δ mutant in uninduced cells, indicating that even though the MNase digestion pattern indicates that nucleosome positioning is disrupted, no histone eviction occurs under this condition. Treating the mutant with MMS led to nucleosome eviction over the promoter, however. Thus, ISW2 is required for maintenance of nucleosome positions at RNR3 but is not involved in eviction. This explains why the level of derepression and preinitiation complex (PIC) formation is low in the isw2Δ mutant even though the MNase digestion pattern suggests complete disruption of nucleosome positioning. Deletion of SWI2 likewise led to a slight but not statistically significant decrease in H3 cross-linking in untreated cells, and as expected, MMS-induced nucleosome eviction was impaired. Thus, SWI/SNF is required for the removal of a nucleosome from the core promoter under the activated condition (+MMS). The level of cross-linking of H3 in the double isw2Δ swi2Δ mutant was essentially the same as that in the isw2Δ strain when the cells were untreated. However, deleting SWI2 in the isw2Δ background blocked the MMS-induced nucleosome eviction. Thus, ISW2 cannot suppress the activation defect of the swi2Δ mutant, because the core promoter nucleosome, while adopting a random position, is not evicted in the double mutant. Next, we examined the cross-linking of H3 in the mutants to a subtelomeric region (TEL) and REC104 as controls (Fig. 2C). ISW2 is required for the sliding of the promoter nucleosome upstream toward the upstream repression sequence (URS) of REC104 (11, 13, 14), and the chromatin mapping pattern at this gene indicates that nucleosomes adopt new positions rather than being evicted in the isw2Δ mutant (11). ChIP analysis detected no nucleosome eviction over REC104 (Fig. 2C), which is supported also by recent genome-wide mapping of nucleosome positions in an isw2Δ mutant (45). A subtelomeric region was also examined as a control region, and there was not a significant change in the H3 cross-linking over this region of the genome in any of the mutants.

ISW2 excludes SWI/SNF from RNR3.We next examined the recruitment of Swi2 to RNR3 in the mutants to determine if ISW2 regulates SWI/SNF binding. ISW2 associates with RNR3 across the entire gene and regions upstream of the promoter, suggesting that it is broadly associated with the gene in an untargeted manner (13, 45, 53). Interestingly, deleting ISW2 caused a very significant increase in the cross-linking of Swi2 to RNR3 in untreated cells (Fig. 3A). The level of cross-linking was very close to that observed in MMS-treated wild-type cells. Treatment of the isw2Δ cells with MMS led to a slight increase in Swi2 cross-linking, however. Thus, SWI/SNF is constitutively associated with RNR3 in the isw2Δ mutant. As a control, we examined Swi2 cross-linking to another ISW2 target gene, REC104. Swi2 cross-linking was not increased at this ISW2-regulated gene, suggesting that the increase in SWI/SNF association is gene specific (also see Fig. 5).

• Open in new tab
• Download powerpoint

FIG. 3.

Analysis of SWI/SNF and transcription factor cross-linking to RNR3. (A) ChIP assays were conducted as described for Fig. 2B. Antibodies to the N terminus of Swi2 were used to immunoprecipitate (IP) chromatin, and primers directed to RNR3 (left) and REC104 (right) were used to amplify DNA. Data are presented as percentages of IP. WT, wild type. (B) ChIP assay for TFIID components. Cross-linking data are presented relative to levels for untreated wild-type cells. XL, cross-linking. (C). ChIP analysis for the activator Rap1 and the repressor Tup1. Rap1 was immunoprecipitated using antibodies to the myc epitope that was incorporated into its C terminus.

To determine if the disrupted chromatin structure caused by the deletion of ISW2 led to a broad and nonspecific increase in the cross-linking of transcription factors known to be recruited to RNR3, we expanded our ChIP analysis to monitor the cross-linking of two components of the TFIID complex, Rap1 and Tup1. Deleting ISW2 did not lead to an increase in TBP binding, nor did it cause an increase in the cross-linking of TAF1, a TFIID-specific TAF_II (Fig. 3B). Rap1 is a sequence-specific DNA binding protein that we recently identified as a regulator of the RNR genes, and this protein is recruited in a checkpoint-dependent manner to RNR3 (42). Furthermore, Rap1 is required for SWI/SNF recruitment to RNR3 (42), and thus, Rap1 may be responsible for the recruitment of SWI/SNF in uninduced isw2Δ cells. If this were true, we would expect Rap1 cross-linking to change when ISW2 is deleted. However, deletion of ISW2 failed to cause Rap1 binding in untreated cells (Fig. 3C). Therefore, while Rap1 is required for recruitment of SWI/SNF to RNR3 in wild-type cells, the constitutive recruitment of SWI/SNF in the isw2Δ mutant occurs by a different mechanism. Next, we examined the recruitment of Tup1. We previously showed that deleting ISW2 led to an increase in Tup1 cross-linking over the URS of RNR3 (53), and this was observed again here (Fig. 3C). Thus, while deletion of ISW2 caused increased recruitment of SWI/SNF and Tup1, it does not result in the nonspecific association of all transcription factors to RNR3. Surprisingly, the ChIP data indicate that SWI/SNF is recruited in the isw2Δ mutant in untreated cells, yet nucleosomes are not evicted efficiently (Fig. 2B and 3A). This suggests either that SWI/SNF is more effective at evicting nucleosomes from promoters configured in the repressed state or that eviction requires the recruitment of Rap1. We believe that the latter is more likely because the DNA binding domain of Rap1 is required for the eviction of nucleosomes at the RNR3 and HIS4 promoters, preventing the reassembly of nucleosomes back to the repressed state (42, 47).

Restoration of ISW2 suppresses SWI/SNF recruitment.The constitutive recruitment of SWI/SNF in the Δisw2 mutant was unexpected. To provide a more direct correlation between ISW2 occupancy and the suppression of SWI/SNF recruitment, we used a strategy developed by Fazzio and Tsukiyama (11). A strain containing ISW2 under the control of the GAL1 promoter was grown in raffinose to reduce ISW2 levels, and expression was restored upon the addition of galactose to the medium. Switching the cells from raffinose to galactose led to an increase in the level of Isw2p within 45 min, and this increase leveled off by 90 min (Fig. 4A). The increase in Isw2p in the cells correlated with its cross-linking to RNR3 (Fig. 4B) and the reestablishment of nucleosome positioning over the 5′ end of the gene (Fig. 4C). The restoration of the repressive chromatin structure is indicated by the reappearance of the internucleosomal hypersensitive sites. Importantly, the levels of Swi2 and Swi3 do not change over the time course of ISW2 reexpression.

• Open in new tab
• Download powerpoint

FIG. 4.

Conditional reexpression of Isw2 causes nucleosome positioning. A strain containing ISW2-FLAG under the control of the GAL1 promoter (11) was grown in raffinose to deplete Isw2 and then supplemented with galactose. Aliquots were removed prior to galactose addition (0) and 45, 90, or 120 min afterwards. (A) Western blotting of extracts from cells. Anti-Flag antibodies were used to detect Isw2, and polyclonal antibodies to SWI/SNF and TFIID subunits were used for loading controls. (B) ChIP assay using anti-Flag antibodies to detect Isw2 cross-linking over the RNR3 promoter. The amount of DNA in immunoprecipitates from an untagged strain (ISW2) was set to 1.0, and cross-linking relative to that value is presented as the means and standard deviations of results from three experiments. (C) MNase mapping of the RNR3 promoter was carried out as described for Fig. 2A. Multiple concentrations of MNase were used in the mapping experiment, but a panel of one concentration is shown to allow a better side-by-side comparison of the patterns at each time point. ND, digested naked DNA.

Next, we used this strain to monitor Swi2 recruitment to RNR3 after the cells were shifted to galactose to correlate the association of ISW2 and the return of nucleosome positioning to changes in the level of SWI/SNF recruitment. Results for a representative experiment are shown in Fig. 5A, left panel, indicating that the cross-linking of Swi2 to RNR3 was reduced within 45 min after the addition of galactose and that the effect was complete by 90 min. A graph showing the averages and standard errors of results from multiple experiments is shown in the right panel, indicating that the results are highly reproducible. The kinetics of the reduction in Swi2 cross-linking matches that of ISW2 association and the reestablishment of nucleosome positions at RNR3 (Fig. 4). Since the changes are observed rapidly and correlate with the appearance of ISW2 at RNR3, this suggests that the effects that we observe are direct. Furthermore, the level of Swi2 remained constant throughout (Fig. 4A), so the effect is not due to changes in protein levels. On the other hand, Swi2 cross-linking was not decreased at REC104 or iYGLWδ6, two other ISW2-regulated regions, or at subtelomeric DNA (Fig. 5A, B, and C). It should be noted that the level of cross-linking of Swi2 to REC104 and iYGLWδ6 represents background levels, but the data nonetheless indicate that restoring ISW2 expression does not lead to a change in the cross-linking of SWI/SNF at these ISW2-regulated loci. Thus, disruption of nucleosome positioning at other ISW2-dependent genes does not lead to SWI/SNF recruitment. As additional controls, we examined the cross-linking of RNAPII and Tup1 to the promoter. As noted before, cells lacking ISW2 have elevated Tup1 recruitment (Fig. 3C) (53), and reexpression of ISW2 leads to reduced cross-linking of this factor (Fig. 5C). There was no change in RNAPII cross-linking, as expected (Fig. 5C). These results indicate that restoring ISW2 to the cell specifically suppresses the elevated level of SWI/SNF recruitment to RNR3. Thus, we provide solid evidence that ISW2 antagonizes the recruitment of SWI/SNF to RNR3.

• Open in new tab
• Download powerpoint

FIG. 5.

ISW2-dependent nucleosome positioning excludes SWI/SNF from RNR3. Results are shown for a ChIP assay performed after reexpression of Isw2. (A) Cross-linking of Swi2. A representative gel from a single experiment is shown on the left. Averages and standard deviations of results from three experiments measuring the cross-linking of Swi2 to RNR3 are shown on the right. (B) Cross-linking of Swi2 at a subtelomeric region (TEL) and REC104. (C) Cross-linking of RNAPII (8WG16) and Tup1 over RNR3.

The N-terminal tail of histone H4 restricts SWI/SNF recruitment.ISW2 can antagonize the recruitment of SWI/SNF by removing it from the promoter directly or by regulating the chromatin structure over RNR3. We addressed these two possibilities by altering the chromatin over RNR3 without changing the status of ISW2 within the cell. The H4 tail is required for ISW2 function in vivo and in vitro (7, 10). Furthermore, the tails of histones, in particular that of H4, have been implicated in the folding of nucleosomes into higher-order structures (8, 9, 38). We therefore examined if removing the histone tails individually results in constitutive recruitment of SWI/SNF. Treating wild-type cells with MMS caused a 2- to 2.5-fold increase in Swi2 cross-linking, and the level of Swi2 binding under the uninduced and induced conditions in the H2A, H2B, and H3 N-terminal tail deletion mutants were essentially equal (Fig. 6A). However, there was an obvious increase in the cross-linking of Swi2 to RNR3 in the untreated histone H4 tail mutant, and the level of cross-linking was not increased significantly by MMS treatment (Fig. 6A). The constitutive SWI/SNF recruitment in the H4 tail mutant was specific for RNR3, as SWI/SNF association was not significantly increased at REC104 (Fig. 6B). Also similar to the isw2Δ mutant, cross-linking of TBP or RNAPII was not elevated in untreated H4 tail mutant cells (Fig. 6C); however, the MMS-induced levels of recruitment were reduced somewhat.

• Open in new tab
• Download powerpoint

FIG. 6.

N-terminal tails of H4 specifically are required for SWI/SNF exclusion. (A) Swi2p cross-linking to RNR3 in histone tail mutants was examined. Assays were conducted as described for Fig. 3A. Gray bars and black bars show data from untreated and MMS-treated cells, respectively. WT, wild type. (B) Cross-linking of SWI/SNF to REC104 in untreated cells. (C) As in panel A, except the cross-linking of TBP (left) and RNAPII (right) to RNR3 was examined. (D) MNase mapping of nucleosome positioning in the histone H4 tail mutant. The circle marks the doublet that appears upon exposure of the TATA box to nuclease. The bar highlights the broadening of the hypersensitive site between nucleosomes +1 and +2 that occurs upon activation of the gene. (E) Cross-linking of histone H3 to the promoter of RNR3 in the H4 tail mutant. ND, digested naked DNA; DRE, damage response element.

We further characterized the changes in chromatin structure at RNR3 in the H4 tail mutant. First, we examined the positioning of nucleosomes over RNR3 in the H4 tail mutant. Deleting the tail of H4 caused a partial disruption of chromatin structure, particularly in the region downstream of the promoter (nucleosome +1). The hypersensitive site downstream of nucleosome +1 is altered (Fig. 6D). Specifically, the hypersensitive site between nucleosomes +1 and +2 broadens in appearance and is reduced in intensity, indicating a weakening of positioning in that region. A similar, although more pronounced, change occurs when the promoter nucleosomes are remodeled. Surprisingly, changes in the positioning of nucleosome −1 in the H4 tail mutant are subtle. A slight increase in digestion over the TATA box is observed (doublet within nucleosome −1), but clearly, it is not as pronounced as what occurs when cells are treated with MMS or when isw2Δ is deleted. However, treating the H4 mutant with MMS led to a pattern resembling that of the fully remodeled state observed in wild-type cells, suggesting an additional loss of nucleosome positioning. Then, we examine the density of nucleosomes at the core promoter by using an antibody to the core domain of H3 in ChIP assays. In the uninduced state, the density of nucleosomes is not affected by the H4 tail mutation (Fig. 6E). Thus, similar to that in the isw2Δ mutant, the disrupted nucleosome positioning detected by MNase mapping in the H4 tail mutant is caused by a disruption in positioning rather than an eviction of nucleosomes. However, there was a reduction in the MMS-induced eviction of the promoter nucleosome in this mutant. The reduction in MMS-induced histone density loss correlated with reduced PIC formation (Fig. 6C) and Tup1 release (see below). It was a surprise that no single histone tail is required for SWI/SNF recruitment. Thus, if the histone tails play a role in SWI/SNF recruitment, there is redundancy among the tails for this function. In addition, even though the levels of SWI/SNF recruitment in the H4 mutant in the absence of DNA damage are similar to those in induced wild-type cells, the extents and natures of the remodeling under the two conditions are not equal. Stimulated wild-type cells show complete disruption of positioning and nucleosome eviction. In contrast, deleting the H4 tail leads to partial loss of positioning but no eviction. This suggests that DNA damage signals regulate steps in the activation process in addition to SWI/SNF recruitment, such as Rap1 recruitment and PIC formation, which are required for the remodeling of chromatin at RNR3 (42). Interestingly, the data also suggest that complete disruption of nucleosome positioning is not required for the constitutive association of SWI/SNF with RNR3. Deleting the tail of H4 had a significantly lesser effect on nucleosome positioning than deleting ISW2, but SWI/SNF recruitment levels were very similar under these conditions. Collectively, the results suggest that disrupting chromatin structure by deleting ISW2 or deleting the tail of H4 leads to the constitutive association of SWI/SNF with RNR3.

Constitutive SWI/SNF recruitment is not dependent on Crt1 or Tup1.Crt1 and Tup1 are the dominant regulators of RNR3 (25). Although they were once thought to function only in repression, a growing body of evidence suggests that they have roles in activation as well (26). For example, Tup1 has been implicated in the recruitment of SWI/SNF to stress-induced genes (31), and we recently found that Crt1 plays a transient role in recruiting SWI/SNF to RNR3 to open up the promoter and can physically interact with SWI/SNF in vitro (51). Although both Crt1 and Tup1 leave the RNR3 promoter in activated cells, it is possible they are responsible for the constitutive recruitment of SWI/SNF in the isw2Δ or H4 tail mutant. First, we examined Tup1 cross-linking in the H4 tail mutant. Tup1 cross-linking is increased in the isw2Δ mutant (Fig. 3C), and if the increase in Tup1 association in these cells is responsible for SWI/SNF recruitment, it would be expected that Tup1 would likewise be elevated in the histone H4 mutant. This is not the case. The level of Tup1 cross-linking in this mutant is equivalent to that in wild-type cells, yet this mutant still displayed elevated SWI/SNF levels (Fig. 7A). Second, we examined the recruitment of SWI/SNF to RNR3 in a Δtup1 mutant. Deleting TUP1 causes chromatin remodeling and PIC formation (25, 52), although the status of SWI/SNF at the promoter is not known. Here, we found that deleting TUP1 actually led to a very robust SWI/SNF recruitment in the absence of DNA damage signals (Fig. 7B). This result, together with the analysis of Tup1 recruitment in the H4 tail mutant, suggests that Tup1 is not responsible for recruiting SWI/SNF in the isw2Δ and histone H4 mutant and also provides another line of evidence that organizing RNR3 into a repressive chromatin structure suppresses SWI/SNF recruitment.

• Open in new tab
• Download powerpoint

FIG. 7.

Crt1 and Tup1 are not responsible for constitutive SWI/SNF recruitment to RNR3. (A) Tup1 cross-linking to RNR3 in the wild type (WT) and the histone H4 mutant. Assays were conducted as described for Fig. 3A. Gray bars and black bars show data from untreated and MMS-treated cells, respectively. (B) SWI/SNF recruitment in a tup1Δ mutant. Cells were not treated with MMS. IP, immunoprecipitation. (C) SWI/SNF recruitment in wild-type, crt1Δ, and crt1Δ isw2Δ cells.

Next, we examined the requirement for Crt1 in constitutive SWI/SNF recruitment. We examined the recruitment of SWI/SNF in a crt1Δ mutant and a crt1Δ isw2Δ double mutant. Similar to the phenotype caused by the deletion of TUP1, deleting CRT1 caused constitutive SWI/SNF recruitment (Fig. 7C), as we noted previously (35). In addition, SWI/SNF recruitment was constitutively high in the crt1Δ isw2Δ double mutant (Fig. 7C). Since the SWI/SNF recruitment was high and constitutive in the absence of Crt1, the data indicate that Crt1 is not required for maintenance of SWI/SNF at the promoter under these conditions.