(A) Schematic of the chromatin architecture found at the GAL1/10 locus prior to induction as described in (Bryant et al., 2008). Nucleosomes are represented by green ovals, and two kinds of hypersensitive sites (HS and hs) are indicated by hatched bars. The white oval represents the unknown factor(s) that protect the UASg. The numbers between the arrows show the length of each segment protected from MNase digestion, and the numbers at the bottom show the percent of the population protected. The transcription start sites of the GAL1 and GAL10 genes are indicated by grey bars and the beginnings of the ORFs by black bars. The four Gal4 sites comprising the UASg are in cyan. The TATA box in the GAL1 promoter lies approximately coincident with the rightmost site marked hs.
(B) RSC binding to the GAL1/10 locus. Yeast bearing TAP-tagged RSC were cross-linked, sonicated, and the isolated chromatin digested with MNase to an extent that yielded primarily mononucleosomes. RSC bound DNA-fragments were isolated on IgG-beads. The purified DNA was then subjected to paired-end high throughput (Illumina) DNA sequencing. The resulting fragments were mapped to the S. cerevisiae genome to determine their sizes and positions. The number of fragments that cross any given base pair in the GAL1/10 locus is shown by the blue line (the fragment density). The data is represented as fold enrichment of RSC bound fragments over a random distribution. Cells were grown in glucose at 25°C, and the TAP-tag was added to the RSC subunit RSC8.
(C) Inactivation of the RSC DNA binding subunit Rsc3. Cells bearing the ts mutant RSC (rsc3-1), also TAP-tagged, were grown in glucose at 25°C and then shifted to 37°C for various times as indicated. Cells were treated as described in (B) and the recovered DNA analyzed by QPCR. RSC binding is presented as the percentage immunoprecipitated.
(D) RSC binding in the absence of Gal4. Cells deleted for gal4 and bearing TAP-tagged wt RSC were analyzed for RSC binding and MNase sensitivity as described in (C). RSC binding is shown as fold over a control locus in the PHO5 gene. Supplemental Figure S1 shows that the MNase protection pattern in and around the UASg is not altered by deletion of gal4.
(E) Gal4 and putative RSC binding sites in the UASg. The Gal4 binding sites are indicated in blue. Gal4 sites each bind a dimer of the protein, and each of the three strong binding sites has the sequence CGG-N₁₁-CCG. Putative binding sites for Rsc3/30 (Badis et al., 2008) are indicated in red. The rightmost putative RSC site overlaps Gal4 binding site 4, which differs from the Gal4 consensus in one base pair as shown. The arrow indicates the site of truncation of the UASg in the mutant strain created for the experiments in (F). Sequences to the right of the arrow are deleted in the truncation mutant.
(F) RSC binding to a wt and a truncated UASg. Cells bearing TAP-tagged RSC and grown in raffinose were probed as described in (C).

(A) Histone H2B, H3 and H4 at and around the UASg. Crosslinked chromatin from wt cells (blue line), cells bearing myc-tagged histone H4 (red line) and cells expressing FLAG-tagged H2B (green line) was digested with MNase and then probed with the respective antibodies. DNA was analyzed by QPCR with results presented as fold over a control locus (in the PHO5 promoter). Cells were grown in media containing glucose and similar results were found for cells grown in raffinose (data not shown).
(B) Histone H2A and its variant H2A.Z at and around the UASg. An experiment as the one described in (A) was performed with wt cells or cells bearing HA-tagged H2A. MNase digested chromatin from wt cells was precipitated with an antibody against H2A.Z and that from cells bearing HA-tagged H2A was precipitated with an antibody recognizing the HA-epitope. The resulting DNA was analyzed as in (A). Supplemental Figure S2 shows that the MNase protection pattern in and around the UASg is not affected by deletion of htz1, the gene that encodes H2A.Z.
(C) Pattern of MNase protection at the GAL1/10 locus assayed by paired-end DNA sequencing. Cells bearing the rsc3-1ts mutation were grown at the permissive temperature in media containing raffinose, crosslinked, digested with MNase, and the resulting DNA subjected to paired-end sequencing (Illumina) as described in the legend of Figure 1B. The blue line indicates the density of the resulting MNase protected fragments over the GAL1/10 locus, and the fragments bound by RSC and protected from MNase digestion of Figure 1B are shown for comparison (red line). The distributions of the sizes of the mapped fragments at the UASg (left inset) and at the neighboring nucleosome are shown (right inset).
(D) Size distribution of MNase protected fragments over the GAL1/10 locus. MNase protected fragments were determined as described in (C) and the number of fragments of the sizes indicated is represented by the color saturation (see Experimental Procedures). The fragment density curve of (C) is superimposed for reference. The boxed areas indicate fragments of sizes corresponding to ordinary nucleosomes - in the ORFs and the one to the right of the UASg (labeled ‘normal’), - and fragments of smaller sizes that are associated with the RSC/nucleosome complex at the UASg (labeled ‘smaller’).

(A) Histone H3 binding to the UASg in the presence and absence of bound RSC. Cells bearing rsc3-1ts (a mutation in a DNA binding subunit) were grown in glucose at 25°C (black) and then shifted to 37°C for the times indicated (red and blue). Cells were probed for histone H3 as described for Figure 2A.
(B) Histone H3 binding to the UASg in the absence of active RSC. Cells bearing RSC sth1-3ts (a mutation in RSC’s catalytic subunit) grown in glucose at the permissive (black) and non-permissive (red) temperature were probed for H3 as in (A).
(C) Quantitation of the loss of an HS site as RSC is inactivated. A MNase protection experiment was performed as described (Bryant et al., 2008). Cells bearing the rsc3-1ts mutation were grown in raffinose at 25°C and then shifted to 37°C, and MNase protection assayed at various times. The black bar above the schematic shows the position analyzed, and the dots show the increasing protection of the HS site as the cells were grown at 37°C. Similar results were found for the HS site to the right of the UASg in the schematic (not shown).
(D) Effect of truncating the UASg on an HS site. A MNase protection experiment was performed as in (C) with cells bearing the truncated UASg of Figure 2D.
(E) Nucleosome disposition at and around a UASg inserted at an ectopic position. A MNase protection experiment was performed using cells bearing a UASg inserted 551 bp downstream from the GAL1 translation start site. In this mutant the DNA spanning the GAL1-GAL10 promoters was deleted. Protection was analyzed after growth of cells in non-inducing medium (2% raffinose, blue curve) and 30 min following addition of 2% galactose (red curve). The numbers below the figure describe the percent protection of the hatched bars, indicating that HS sites flanking the UASg were created by the insertion.
(F) Gal4 binding to the UASg inserted at an ectopic position. Cells bearing the ectopically positioned UASg described in (E) were grown in raffinose and a ChIP experiment detecting Gal4 was performed as described (Floer et al., 2008) except that chromatin was digested with MNase prior to immunoprecipitation. Gal4 binding is shown as fold over a control location in the PHO5 gene.
(G) RSC binding to the UASg inserted at an ectopic position. Cells bearing the ectopically positioned UASg and TAP-tagged RSC were grown in raffinose, chromatin was treated with MNase and a ChIP experiment recognizing the TAP-tag was performed as in Figure 1C.

(A) Effect of inactivating RSC on induction: Raff ⇒ Gal. Galactose was added to wt (blue) or rsc3-1ts mutant cells (red) growing in raffinose at 37°C for 3h. At the times indicated GAL1 mRNA levels were determined as described (Floer et al., 2008).
(B) Effect of deleting putative RSC binding sites on induction: Raff ⇒ Gal. Galactose was added to raffinose-grown cells that bore wt RSC and either the wt (blue) or the truncated UASg (red) of Figure 2D.
(C) Effect of inactivating RSC on induction: Glu ⇒ Gal. wt (blue) and rsc3-1ts cells (red) were grown in glucose for 3h at 37°C and then shifted to media lacking glucose but containing galactose and raffinose. GAL1 mRNA levels were measured as described in (A).
(D) Effect of removing RSC binding sites on induction: Glu ⇒ Gal. An experiment was performed as that described in (C) except that cells bearing wt RSC and either the wt (blue) or truncated UASg (red) were grown at 25°C.
(E) Effect of inactivating RSC on Gal4 bound to the UASg: Raff. wt cells (blue) and rsc3-1ts cells (red) were grown in raffinose at 25°C and then shifted for 3h to 37°C. Gal4 binding was analyzed as described for Figure 3F and is shown normalized to a control locus in the PHO5 gene.
(F) Effect of inactivating RSC on Gal4 binding to the UASg: Glu ⇒ Gal. wt (blue) and rsc3-1ts cells (red) were grown in glucose for 3h at 37°C and shifted to medium containing galactose. Gal4 binding to the UASg was analyzed as described in (E).
(G) Effect of inactivating RSC on GAL4 mRNA production. GAL4 mRNA levels were measured for cells grown as described for the experiment of (C).
(H) Effect of deleting RSC binding sites on binding of a hormone regulated Gal4-fusion protein to the UASg. Cells were deleted for gal4 but contained a plasmid expressing a myc-tagged Gal4DBD-ER-VP16 fusion (Nalley et al., 2006), and either the wt UASg (blue) or its truncated derivative (red). Gal4 binding to the UASg was determined at the times indicated, following addition of estrogen to cells growing in glucose, by probing for myc. The data was normalized to a control locus in the PHO5 gene.

The distribution of fragments protected from MNase digestion at four promoters found on chromosome II. The data is displayed as in Figure 2D, except that the disposition of RSC (taken from data of the experiment of Figure 1B) is overlayed in red. The names of the genes, and the coordinates along the genome are shown in the figure.

HeLa cells were transfected with an integrating plasmid bearing a 700 bp DNA segment spanning the UASg and the flanking GAL1/10 promoters, and integrant were selected. A MNase protection experiment was performed with yeast wt for RSC or bearing the rsc3-1ts mutation, and with HeLa cells bearing the inserted UASg segment. The blue bars indicate the protection of the DNA segment in the UASg as indicated in the schematic above the figure. The red bars indicate protection of the sequence just to the left of the UASg that is HS in wt yeast.

(A) Cryo-EM structure of RSC bound to a nucleosome (taken from (Chaban et al., 2008)). RSC interacts closely with nucleosomal DNA at three different positions (grey-colored RSC density labeled 1-3). Some DNA density (solid black line) is apparent in the structure, but large portions of the DNA (hatched black lines) were not detected, suggesting that they were highly mobile/disordered as a result of interaction of the nucleosome with RSC.
(B) Projection of the UASg onto a nucleosome. The UASg was modeled onto a single turn of a nucleosome (corresponding to 80 bp) with the dyad axis placed in the center of the UASg. Gal4 sites are shown in blue and putative RSC binding sites in red. Alignment of this projection with the structure in (A) places Gal4 sites 1 and 4 on the unwrapped ends of the nucleosome and sites 2 and 3 on a part of the nucleosome that is largely accessible in the structure. The positions of the putative RSC binding sites in the UASg correspond closely to the three RSC densities shown to contact nucleosomal DNA.
(C) Model of Gal4 binding to sites 2 and 3 in a UASg bound by a RSC/nucleosome complex. The UASg was positioned in the RSC/nucleosome structure of (A) as described in (B). Two Gal4 dimers (as represented by their DNA binding and dimerization domains, shown in red and purple) were positioned on Gal4 binding sites 2 and 3. The orientation of Gal4 dimers on DNA was taken from the Gal4/DNA structure of (Hong et al., 2008). The model shows that sites 2 and 3 are exposed along a surface of the RSC/UASg/nucleosome complex and that Gal4 can bind these sites without disrupting the structure.