The nucleosome positioning activity for the PHO8 promoter could be enriched from a yeast whole-cell extract (WCE) over four sequential fractionation steps. (A) Top panel: schematics of nucleosome positions at the KRE2-CWC21-PHO8 locus, according to Barbaric et al (1992) and Jiang and Pugh (2009). Nucleosomes are numbered relative to NDR1. Middle panel: mapped Pho4 (Barbaric et al, 1992) or predicted Rsc3 (Badis et al, 2008) binding sites (Supplementary Figure S8A). Lower panel: KRE2, CWC21 and PHO8 open reading frames (rectangular bars with large broken arrows), TATA box (T; Basehoar et al, 2004) and transcriptional start sites (TSS, small broken arrows; Miura et al, 2006). Scale bar: distance in base pairs from PHO8 ORF start. All panels drawn to scale. (B) Extract fractionation scheme. Fractions positive for the PHO8 promoter nucleosome positioning activity are labelled in bold. SN, supernatant. (C) DNaseI indirect end labelling analysis of the PHO8 promoter region in vivo or in vitro after salt gradient dialysis assembly and incubation with either WCE in the presence or absence of ATP, or with one of the indicated fractions (see B) in the presence of ATP. Black dots: diagnostic bands, which are characteristic for the in vivo pattern and seen in vitro only in the presence of ATP and the nucleosome positioning activity. Black dots in parentheses: hypersensitive site within the lacZ ORF of the pUC19 backbone specific for the in vitro pattern that always co-occurred with the in vivo-like PHO8 promoter pattern. The yeast sequence terminates close to the top marker band. Schematics on the left analogous to (A). Position of marker bands is labelled relative to the PHO8 ORF start. Ramps and boxes: relative DNaseI concentrations. All samples were electrophoresed alongside in the same gel, but the in vivo samples migrated slightly faster, probably because of different total DNA concentration.

Purified RSC could rescue the nucleosome positioning activity of an extract generated from an rsc3-ts mutant grown under restrictive conditions. DNaseI indirect end labelling analysis of the (A) PHO8, (B) RIM9, (C) CHA1, and (D) SNT1 promoter regions as in Figure 2, but with WCEs generated from wild-type (BY4741) grown logarithmically at 30°C, or from rsc3-ts strain (TH8239) grown logarithmically at 25°C with or without an overnight shift to 37°C. Addition of RSC corresponded to the 1:5 ratio in Figure 2.

Loss of essential RSC subunits at elevated temperature altered chromatin structure at the PHO8, RIM9 and CHA1, but not at the SNT1 promoter. DNaseI indirect end labelling analysis of the (A) PHO8, (B) RIM9, (C) CHA1 and (D) SNT1 promoter regions in vivo. Nuclei were isolated from wild type (wt; BY4741) and strains carrying a temperature-sensitive (rsc3-ts (TH8247) and arp9-ts (YBC1536)) or temperature-sensitive degron (sth1-td (YBC2191)) allele of the indicated RSC subunits. Strains were grown logarithmically at 25°C and then shifted to the non-permissive temperature (37°C) overnight. Wt nuclei were also prepared from cells grown logarithmically at 30°C. Bars in-between lanes mark the intensity (bar width) and extent (bar length) of DNaseI hypersensitive sites. A stippled line separates samples that were not electrophoresed alongside on the same gel but combined in the figure. Asterisks indicate artefact bands. Ramps, markers and schematics as in Figure 2.

Purified RSC repositioned nucleosomes in salt gradient dialysis chromatin, but only in few cases, resulting in in vivo-like positions. DNaseI indirect end labelling analysis of the (A) PHO8, (B) RIM9, (C) CHA1 and (D) SNT1 promoter regions in vitro after assembly by salt gradient dialysis and incubation with WCE or purified RSC complex in the presence or the absence of ATP as indicated. The amount of RSC is given as the molar ratio of RSC to nucleosomes. In each panel, lanes 1 and 2 show the wt in vivo DNaseI pattern. Free DNA samples correspond to the respective non-assembled plasmids in the absence of WCE, RSC and ATP but under otherwise identical conditions. Bars in between lanes mark hypersensitive regions that correspond, at least to some degree, to NDRs of the in vivo patterns. The arrow between lanes 12 and 13 in D marks a nuclease-sensitive region that becomes inaccessible because of RSC activity. Ramps: increasing DNaseI concentrations. Position of marker bands is labelled relative to the ORF start of the respective locus. Schematics on the left are analogous to Figure 1A for the respective locus. Predicted Rsc3 binding sites (Supplementary Figure S8) are indicated by black dots.

RSC was specifically required for nucleosome positioning in vitro as both SWI/SNF and Isw2 failed to rescue the rsc3-ts 37°C extract. DNaseI indirect end labelling analysis of the (A) PHO8, (B) RIM9, (C) CHA1, and (D) SNT1 promoter regions as in Figures 2 and 3 but with addition of purified SWI/SNF or Isw2 remodelling enzymes as indicated. All remodelling enzymes were added at the same molar concentrations, corresponding to the 1:5 ratio in Figure 2.