A novel mechanism of antagonism between ATP-dependent chromatin remodeling complexes regulates RNR3 expression

Abstract

Gene expression depends upon the antagonistic actions of chromatin remodeling complexes. While this has been studied extensively for the enzymes that covalently modify the tails of histones, the mechanism of how ATP-dependent remodeling complexes antagonize each other to maintain the proper level of gene activity is not known. The gene encoding a large subunit of ribonucleotide reductase, RNR3, is regulated by ISW2 and SWI/SNF, complexes that repress and activate transcription, respectively. Here, we studied the functional interactions of these two complexes at RNR3. Deletion of ISW2 causes constitutive recruitment of SWI/SNF, and conditional reexpression of ISW2 causes the repositioning of nucleosomes and reduced SWI/SNF occupancy at RNR3. Thus, ISW2 is required for restriction of access of SWI/SNF to the RNR3 promoter under the uninduced condition. Interestingly, the binding of sequence-specific DNA binding factors and the general transcription machinery are unaffected by the status of ISW2, suggesting that disruption of nucleosome positioning does not cause a nonspecific increase in cross-linking of all factors to RNR3. We provide evidence that ISW2 does not act on SWI/SNF directly but excludes its occupancy by positioning nucleosomes over the promoter. Genetic disruption of nucleosome positioning by other means led to a similar phenotype, linking repressed chromatin structure to SWI/SNF exclusion. Thus, incorporation of promoters into a repressive chromatin structure is essential for prevention of the opportunistic actions of nucleosome-disrupting activities in vivo, providing a novel mechanism for maintaining tight control of gene expression.

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.

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.

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.

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.

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.

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.

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.