Interaction of RSC Chromatin Remodeling Complex with Nucleosomes Is Modulated by H3 K14 Acetylation and H2B SUMOylation In Vivo.

Chromatin remodeling complexes are multi-subunit nucleosome translocases that reorganize chromatin in the context of DNA replication, repair, and transcription. To understand how these complexes find their target sites on chromatin, we use genetically encoded photo-cross-linker amino acids to map the footprint of Sth1, the catalytic subunit of the RSC complex, on nucleosomes in living yeast. We find that H3 K14 acetylation induces the interaction of the Sth1 bromodomain with the H3 tail and mediates the interaction of RSC with neighboring nucleosomes rather than recruiting it to chromatin. RSC preferentially resides on H2B SUMOylated nucleosomes in vivo and shows a moderately enhanced affinity due to this modification in vitro. Furthermore, RSC is not ejected from chromatin in mitosis, but changes its mode of nucleosome binding. Our in vivo analyses show that RSC recruitment to specific chromatin targets involves multiple histone modifications likely in combination with histone variants and transcription factors.

Introduction

Storage and accessibility of genetic information are two conflicting requirements that a cell must balance. While the DNA must be compacted to meet the space limitations of the nucleus, access to its information content for transcription, repair, and replication processes must be ensured. Eukaryotes accomplish this by packaging their DNA in chromatin; however, densely packed chromatin territories restrict the accessibility of the underlying DNA (Padeken and Heun, 2014). Hence, to facilitate access to DNA, chromatin has evolved highly malleable properties to meet the demands for dynamic changes (Seeber and Gasser, 2017; Talbert and Henikoff, 2017; Tessarz and Kouzarides, 2014).

Chromatin remodeling enzymes use ATP hydrolysis to rearrange nucleosomes to enable other factors to access DNA (Clapier et al., 2017). Posttranslational modifications (PTMs) of histones and histone variants either modulate the stability and DNA-binding properties of the nucleosome or signal the recruitment of machinery that initiates the transcription, replication, or repair of DNA (Tessarz and Kouzarides, 2014). The turnover of most histone PTMs is rapid, making these processes very dynamic.

Common to all families of chromatin remodelers are an affinity for nucleosomes, the ability to recognize histone PTMs via specialized domains, and a DNA-dependent ATPase domain that translocates DNA relative to the histone octamer. Apart from that, remodelers differ significantly in subunit composition, specificity for histone modifications, the processes in which they are involved, and whether they promote chromatin opening or closing. How these enzymes work, how their activity is regulated, and how they are recruited to specific loci are currently being actively investigated.

The RSC complex is an abundant, essential chromatin remodeling complex of the SWI/SNF family in budding yeast (Cairns et al., 1996). RSC is involved in transcription (Brahma and Henikoff, 2018; Floer et al., 2010; Krietenstein et al., 2016; Musladin et al., 2014; Spain et al., 2014), chromosome segregation (Hsu et al., 2003), replication (Niimi et al., 2012), and the response to DNA damage (Rowe and Narlikar, 2010; Shim et al., 2007).

Recently, the high-resolution structure of the RSC complex bound to a nucleosome has been solved by cryo-electron microscopy, revealing three flexibly connected domains (Patel et al., 2019; Wagner et al., 2020; Ye et al., 2019). The motor domain of the Sth1 subunit binds at superhelical location (SHL) +2, from where it translocates the DNA one base pair at a time into the direction of the dyad, possibly creating a loop that propagates around the nucleosome (Zhang et al., 2006). The ARP module, which couples DNA translocation and ATPase activity (Clapier et al., 2016), connects the motor domain to the substrate recognition module (SRM). The latter contains DNA-binding Zn-cluster domains, five bromodomains, a histone-tail-binding BAH domain, and the nucleosome-binding C-terminal tail of Sfh1 (Patel et al., 2019; Wagner et al., 2020; Ye et al., 2019). Owing to the flexible tethering of these domains, their structure, substrate preference, and interaction with the nucleosome remained largely unresolved. The only established lysine acetylation site on histones recognized by RSC is H3 K14ac by the bromodomains of Sth1 and Rsc4 (Chen et al., 2020; Kasten et al., 2004; VanDemark et al., 2007).

Here, we use quantitative in vivo cross-linking with genetically encoded photo-activatable cross-linker amino acids to reveal the footprint of the Sth1 subunit of RSC on the nucleosome. The interaction of Sth1 with the N-terminal H3 tail depends on the presence of H3 K14, which is recognized by the C-terminal bromodomain of Sth1 upon acetylation (Chen et al., 2020). We further show that Sth1 preferentially cross-links to the SUMOylated form of H2B, suggesting that H2B SUMOylation acts in the context of RSC remodeling.

Results

In Vivo Cross-linking Survey of the Nucleosome

Genetic code expansion allows for the incorporation of unnatural amino acids (UAAs) in response to amber (UAG) stop codons in a variety of cells and organisms (Neumann-Staubitz and Neumann, 2016). In yeast, this is achieved by transforming the cells with a plasmid encoding an evolved aminoacyl-tRNA synthetase specific for the desired UAA and its cognate amber suppressor tRNA. A second plasmid is introduced encoding the gene of interest with an amber codon replacing the codon for the amino acid that shall be converted to the UAA. In these cells, the evolved aminoacyl-tRNA synthetase charges its cognate tRNA with the UAA, which is subsequently incorporated at the site specified by the amber codon in the protein of interest.

We have established the incorporation of photo-activatable cross-linker amino acids in histones and chromatin-interacting proteins to study the dynamics of chromatin in living yeast (Hoffmann and Neumann, 2015; Wilkins et al., 2014). The cross-linking reaction follows a long-wavelength UV light (365 nm)-inducible radical mechanism that results in the formation of binary covalent adducts that can be quantitated by western blot (Dorman and Prestwich, 1994).

To map the interactome of the nucleosome in living yeast, we created a library of more than one hundred amber mutants covering the surface-exposed residues of the nucleosome. We incorporated p-benzoyl-l-phenylalanine (pBPA) in response to amber stop codons by genetic code expansion for each individual mutant and analyzed cross-link products formed upon irradiation by SDS-PAGE and western blot against the HA-epitope on the histone (Figures S1 and S2).

Cross-link scans of the nucleosomal surface revealed differential binding patterns across each of the histones, highlighting the viability of interactome mapping in the living nucleus. We then asked if this approach could be employed to characterize individual chromatin binding proteins, specifically nucleosome bound chromatin remodeling complexes.

Footprint of Sth1, the Catalytic Subunit of RSC, on the Nucleosome

In order to probe the footprint of the catalytic subunit of RSC, Sth1, on the nucleosome surface we selected 58 histone pBPA mutants from the cross-linking survey that had produced a band of appropriate combined mass for a cross-link product of a histone with Sth1 (approximately 170 kDa). To test whether these bands indeed result from cross-linking to Sth1 we performed the same cross-linking reaction in strains with or without a C-terminal 3xmyc-tag on Sth1 (Figure 1 and S3–S6). The epitope tag leads to a slower migration of the corresponding band of the cross-link product in western blots. This analysis identified nine positions on the nucleosome core and H3 tail that interact with Sth1 (Figure 1A). We detected two further positions in the H3 tail (T6 and T11), when we precipitated the cross-link products with anti-myc antibody beads (against myc-tagged Sth1) prior to western blot analysis with anti-HA antibodies (Figure 1B). This confirmed the ability of pBPA at these positions to cross-link to Sth1.

Figure 1

Mapping the Interaction Surface of Sth1 on the Nucleosome In Vivo

(A) Yeast cells (wild-type or Sth1-3myc) expressing histones with pBPA at the indicated position were UV irradiated and cross-link products analyzed by SDS-PAGE and western blot using anti-HA antibodies recognizing the pBPA-containing histone. The shift in mobility resulting from the myc-tag identifies a cross-link product with Sth1. H2B-Sth1 cross-links appear as a double band due to H2B SUMOylation (arrow heads). See Figures S1–S6 for full-sized western blots.

(B) Cross-link reactions from positions in the H3 tail performed in Sth1-3myc cells were subjected to immunoprecipitation with anti-myc antibody beads prior to analysis by western blot with anti-HA antibodies. The Sth1-histone cross-link is indicated by the arrowhead. A cross-reactive band is marked by an asterisk.

(C) Graphical representation of the positions identified in (A) and (B) on the structure of the nucleosome. Figure was prepared using PDB: 1ID3 and PyMol v1.7.6.6.

When we measured the growth rates of cells expressing the pBPA-containing histones, we observed a moderate reduction, indicating that the amber suppressor system, the fragment produced by termination at the amber stop codon, or the mutated histones stress the cells to some extent (Figure S7). However, we expect that the biological role of RSC and its interaction with the nucleosome is comparable with the native state because the mutant histone represents only a small fraction of the total amount. To visualize the footprint of Sth1 on the nucleosome, we mapped these positions on the structure of the nucleosome core particle (Figure 1C).

Binding of the H3 Tail by Sth1 Depends on H3 K14

We hypothesized that the C-terminal bromodomain of Sth1 might mediate the interaction with the H3 tail because H3 K14 acetylation enhances nucleosome binding by RSC (Duan and Smerdon, 2014) and a recent co-crystal structure revealed an extensive interface between the Sth1 bromodomain and an H3 tail peptide (Chen et al., 2020). We therefore mutated H3 K14 to alanine on the same H3 copy containing pBPA. The mutation interfered with cross-linking of H3 T6pBPA and T11pBPA to Sth1, whereas cross-linking from H3 S22pBPA was only partially affected by K14A, suggesting that K14 is required for the interaction of Sth1 with the tip of the H3 tail (Figure 2A).

Figure 2

Cross-linking of the H3 Tail to Sth1 Is Regulated by H3 K14ac

(A) Cross-linking from positions in the H3 tail in Sth1-3myc cells. Anti-myc immunoprecipitates were analyzed by western blot with anti-HA antibodies.

(B) Deletion of gcn5 interferes with H3 T6pBPA cross-linking to Sth1. WCEs of cross-linked samples were irradiated with UV light and whole-cell extracts analyzed by western blot using anti-HA antibodies.

(C) Effect of mutating endogenous H3 K14 to alanine on cross-linking to Sth1. Yeasts (wild-type or H3 K14A) expressing H3 T6pBPA, H3 S22pBPA, or H2A A61pBPA with or without K14A mutation were analyzed as in (B).

(D) Quantitative comparison of cross-linking efficiencies from H3 T6pBPA in wild-type and H3 K14A yeasts. Error bars are standard deviations of five independent experiments.

In order to demonstrate that acetylation of H3 K14 is essential for this interaction, we deleted the gene encoding lysine acetyltransferase Gcn5, the enzyme responsible for the deposition of H3 K14ac (Kuo et al., 1996; Zhang et al., 1998). This indeed abolished cross-linking between the H3 tail and Sth1, much like the H3 K14A mutation (Figure 2B).

Next, we asked whether H3 K14ac serves to recruit RSC to nucleosomes. Therefore, we performed cross-linking experiments from H2A A61pBPA-nucleosomes in yeast with or without a genomic H3 K14A mutation (Figure 2C, left panel) (Dai et al., 2008). We observed that cross-linking of Sth1 was only slightly reduced by the K14A mutation. Hence, recruitment of RSC to chromatin does not require H3 K14ac, otherwise the cross-linking efficiency from this position would have been reduced in the mutant background.

This is consistent with the micromolar concentration of nucleosomes in the yeast nucleus (approximately 60,000 nucleosomes [Oberbeckmann et al., 2019] in a volume of 3 fL [Jorgensen et al., 2007]) being more than one thousand times greater than the dissociation constant (KD) of RSC-nucleosome complexes (Lorch et al., 1998). Therefore, increasing the affinity of RSC for nucleosomes by histone modifications is not expected to enhance the level of saturation (Θ) of RSC with nucleosomes because Θ is hardly affected by changes in ligand concentration if their concentration is greater than ten times KD.

However, H3 K14ac may control which nucleosomes are bound by RSC. In this case, mutation of H3 K14 in nucleosomes without the cross-linker (i.e., in the genomic copy of the H3 gene) should shift RSC binding to cross-linker-containing nucleosomes that still possess H3 K14ac.

Therefore, we performed cross-linking experiments from H3 S22 in the background of a yeast strain bearing the K14A mutation in the genomic copy of H3 (Figure 2C, middle panel). If H3 K14ac recruits RSC to nucleosomes, this mutation should increase the cross-linking efficiency because the cross-linker-containing nucleosomes are the only ones with an intact H3 K14 residue. However, cross-linking from H3 S22pBPA was not affected by genomic H3 K14A mutation; hence, H3 K14ac does not control recruitment of RSC.

Finally, we asked whether the Sth1 bromodomain exclusively interacts with H3 tails that are part of the nucleosome bound by RSC or whether H3 tails from neighboring nucleosomes are also substrates. Therefore, we performed cross-linking experiments from H3 T6pBPA in the genetic background of H3 K14A cells (Figure 2C, right panel). If the bromodomain of Sth1 only interacts with H3 tails that are part of the same nucleosome that the complex is bound to, the mutation of the endogenous H3 should not affect the cross-linking efficiency. However, if Sth1 interacts with H3 tails of neighboring nucleosomes, the mutation would abrogate the competition with their histone tails and therefore increase cross-linking. Indeed, the K14A mutation of the endogenous H3 allele increased cross-linking between Sth1 and H3 T6pBPA 15-fold (Figure 2D), strongly indicating that the bromodomain of Sth1 is able to interact with acetylated H3 tails of other nucleosomes.

RSC Preferentially Interacts with SUMOylated H2B In Vivo

Cross-link products of Sth1 to histones H2A and H3 each migrated as a single band in western blots with an apparent molecular mass of about 180 kDa (Figure 1). H2B-Sth1 cross-link products (from positions T51 and S115), however, showed a second band shifted by approximately 10 kDa to a higher apparent mass (Figure 1). Initially, we hypothesized that this mass shift is a result of H2B K123 ubiquitination, the major site of H2B ubiquitination in budding yeast (Robzyk et al., 2000). However, the cross-linking pattern of ubiquitination-deficient mutant H2B T51pBPA K123R was indistinguishable from that of H2B T51pBPA (Figure 3A). Hence, we tested additional H2B sites known to be ubiquitinated (K46, K49, and K111) (Swaney et al., 2013), which also did not change the cross-linking pattern (Figure 3A). Next, we analyzed the impact of mutating combinations of K6, K7, K16, and K17 in H2B T51pBPA to arginine on cross-linking to Sth1 (Figure 3B). These are the major sites of H2B SUMOylation in S. cerevisiae (Nathan et al., 2006). Indeed, in the absence of these lysine residues, we observed only a single H2B-Sth1 cross-link product. Interestingly, mutating either pair of lysine residues in the H2B N terminus (K6/7 or K16/17) was sufficient to abolish the slower migrating band. Accordingly, mutations of the same sites have previously been shown to abolish H2B SUMOylation (Nathan et al., 2006). Finally, when we analyzed a strain lacking E3 SUMO ligases Siz1 and Siz2 required for H2B SUMOylation (Nathan et al., 2006), the cross-link reaction produced only a single H2B-Sth1 band (Figure 3C), unambiguously confirming the upper band to be the SUMOylated form of H2B.

Figure 3

RSC Prefers Binding of H2B SUMOylated Nucleosomes

(A) Cross-link pattern of H2B T51pBPA is not affected by mutation in lysine residues reported to be subject to ubiquitinylation. Cross-links to H3 S22pBPA are used as reference to identify Sth1-H2B cross-link.

(B) Effect of mutating H2B SUMOylation sites on H2B T51pBPA cross-link pattern.

(C) Effect of deletion of siz1 and siz2 on H2B T51pBPA cross-link pattern. Intensities of upper and lower cross-link bands were quantified by densitometry. Error bars are standard deviations of five independent experiments. In all panels, yeasts expressing H2B T51pBPA with the indicated mutations were UV-irradiated and whole-cell lysates analyzed by SDS-PAGE and western blot using anti-HA antibodies. Full blots in Figure S8.

Densitometric quantification of both bands indicates that 20%–30% of H2B that interacts with Sth1 is SUMOylated.

Impact of H2B SUMOylation on RSC In Vitro

In order to reveal the impact of H2B SUMOylation on the interaction of RSC with nucleosomes, we produced nucleosomes containing an in-frame fusion of SUMO to H2B (Xenopus sequence truncated by the first five N-terminal residues, Figure S9). RSC complex purified from yeast (see Transparent Methods and Figure S10) bound to SUMOylated nucleosomes with approximately 2-fold higher affinity than to unmodified nucleosomes in the presence of the non-hydrolysable ATP analog AMP-PNP in biolayer interferometry (BLI) experiments (Figures 4A, 4B, and S11). To further analyze RSC action on nucleosomes, we compared the rate of ATP hydrolysis by RSC in the presence of SUMOylated and unmodified nucleosomes (Figures 4C and 4D). We observed a slightly increased activity on SUMOylated nucleosomes without a change in catalytic efficiency (Vmax/c(RSC)/KM). Similarly, the rate of nucleosome remodeling by RSC in electromobility shift assays was not affected by H2B SUMOylation (Figures 4E and 4F). However, we observed a reduced amount of free DNA in remodeling reactions with SUMOylated nucleosomes, suggesting that the modification may have an influence on the ejection of the octamer during remodeling (Figure 4G). Altogether our data suggest that SUMOylation of H2B per se only has a modest role in RSC affinity and activity in vitro and most likely synergizes with other factors in the recruitment of RSC to chromatin in vivo.

Figure 4

In Vitro Characterization of RSC-nucleosome Affinity and Activity

(A) Biotinylated RSC complex was immobilized on streptavidin biolayer tips and nucleosome (167-bp Widom-601 DNA) binding was analyzed by BLI at different concentrations of nucleosomes (0, 22, 66, 200 nM). See Figure S11 for raw data file.

(B) KD values were determined by fitting the amplitudes of each binding kinetic with GraphPad.

(C) V0 values of RSC remodeling reactions (determined with an ADP-Glo Assay) at different concentrations of unmodified and SUMOylated nucleosomes (0.8, 2.5, 7.5, 22, 66, 200 nM; 197bp Widom-601 DNA) fit to a Michaelis-Menten equation.

(D) Vmax and KM determined from data shown in (C) for unmodified (2.79 ± 0.13 μM/min; 2.25 ± 0.56 nM) and SUMOylated nucleosomes (3.14 ± 0.16 μM/min; 3.52 ± 0.87 nM).

(E) RSC nucleosome remodeling activity (in the presence of Nap1) by EMSA. The relative amounts of unremodeled (top part of the bar) and two remodeled species (middle and bottom) were quantified for unmodified (WT) (197bp Widom-601 DNA).

(F) Same as E but for SUMOylated nucleosomes.

(G) The amount of ejected DNA in EMSA was determined densitometrically and normalized to the initial amount of free DNA. All error bars are standard deviations of the means of at least three independent experiments. A typical EMSA gel is shown in Figure S12.

Modulation of Sth1-Nucleosome Interactions during the Cell Cycle

Chromatin compaction in mitosis is thought to counteract transcription by preventing access of transcription factors, RNA polymerase, and chromatin remodelers to DNA (Yokoyama and Gruss, 2013). To test whether RSC binding to chromatin is influenced by chromatin structure, we analyzed the cross-linking efficiency of histones to Sth1 during the cell cycle (Figure 5). Therefore, we synchronized temperature-sensitive cdc15-2 yeasts harboring plasmids to produce the pBPA-containing histones using a temperature shift protocol. We then sampled over time, cross-linked, and analyzed the cross-link products by SDS-PAGE and western blot. Of the four positions studied, two (H3 S22 and K56) displayed a constant cross-linking efficiency between mitosis and interphase, indicating that the RSC remodeling complex remains bound to nucleosomes throughout the cell cycle. For the other two positions (H3 T80 and H2B T51), however, we observed a reciprocal change in intensity between mitosis and interphase, indicating that chromatin structure has a subtle influence on how RSC binds nucleosomes.

Figure 5

Effect of the Cell Cycle Stage on Histone-Sth1 Cross-link Efficiency

(A) Yeasts (cdc15-2) expressing the indicated pBPA-containing histones were synchronized using the illustrated temperature-shift protocol. Samples were irradiated and whole-cell lysates analyzed by SDS-PAGE and western blot using anti-HA antibodies.

(B) Band intensities of three to four independent replicates of the experiment shown in (A) were quantified by densitometry using Fuji software (see Transparent Methods for details). Error bars are standard errors of the mean. Full blots in Figure S13.

(C) Fluorescence-activated cell sorting (FACS) analysis of synchronized yeast populations analyzed in (A) and (B).

Discussion

We analyzed more than one hundred sites in core histones for their suitability for pBPA incorporation and cross-linking. Many sites gave rise to abundant and diverse cross-link products. Identification of the cross-link partners is the limiting step to proteome-wide mapping of chromatin interacting proteins. Ideally, this could be done by mass spectrometry, which has been established for mammalian cells (Kleiner et al., 2018) and should also be feasible in yeast. However, because of the large number of different cross-link products formed from the same position on histone proteins, the amount of material of each individual cross-link product is very low. This restricts quantitative measurements by mass spectrometry, which are essential for the investigation of the impact of mutations or the cell cycle on the interactions.

Here, we explored whether cross-link products to Sth1 were present in the cross-linking patterns using electrophoretic mobility shift assays (EMSAs) to reveal the footprint of the protein on the nucleosome in vivo. Drawing on structural information of the nucleosome-bound RSC complex (Wagner et al., 2020; Ye et al., 2019), cross-link reactions from the H3 αN-helix (R52, K56) most likely target the motor domain of Sth1, consistent with the observation that mutations in this helix have a strong effect on RSC remodeling activity (Somers and Owen-Hughes, 2009).

Positions on the nucleosome surface (H2A A61, E65; H2B T51, S115; H3 T80) most likely cross-link to the SnaC domain of Sth1 (Wagner et al., 2020), whereas positions in the H3-tail probably target its bromodomain in agreement with recent structural studies (Chen et al., 2020). The RSC complex therefore contacts the acidic patches on both sides of the nucleosome simultaneously, with Sth1 SnaC on one side and with Sfh1 (Wagner et al., 2020) on the other.

Our observations further show that H3 K14 acetylation is required for the interaction of the H3 tail with Sth1. Binding of the RSC complex to nucleosomes, however, is little affected by removing this mark indicating that its recruitment is controlled by additional mechanisms, e.g., general regulatory factors of transcription and DNA sequence motifs (Krietenstein et al., 2016). Our data suggest that the bromodomain of Sth1 binds the K14ac mark of H3 tails of neighboring nucleosomes. This property would be well compatible with the idea that RSC contributes to the formation of a nucleosome-free region at yeast promoters (Krietenstein et al., 2016; Wippo et al., 2011).

Interestingly, position H3 T80 and H2B T51 show reciprocal changes in cross-linking intensities between mitosis and interphase (Figure 5). We speculate that the RSC-nucleosome interaction is modulated by changes in chromatin structure during the cell cycle. Since the cross-linking efficiencies from two other sites (H3 S22 and K56) displayed hardly any changes at different cell cycle stages, we conclude that RSC activity is not controlled by eviction of the remodeler from chromatin by condensation in mitosis, as has been observed for the homologous human chromatin remodeler BRG-1 (Muchardt et al., 1996).

Alternatively, the changes in cross-linking efficiencies with the cell cycle stages might be the result of cell cycle variations in histone PTMs. For example, H3 K56ac displays pronounced cell cycle variation (Ozdemir et al., 2006). Although H3 K56ac does not affect RSC binding and remodeling, it may affect RSC binding at the αN-helix of H3 (Neumann et al., 2009). Incorporation of pBPA at this position ablates acetylation and would be consistent with a loss of cell cycle-dependent cross-linking.

Our cross-linking experiments revealed a previously unreported preference of RSC for H2B SUMOylated nucleosomes. The cross-linking reactions from H2B positions produced a double band with the upper band being about 20%–30% of the lower band intensity. In contrast, SUMOylation affects only about 5% of H2B molecules (Nathan et al., 2006), implying that RSC has a strong thermodynamic preference for these nucleosomes. However, our biochemical analyses do not support this conclusion. Alternatively, RSC may be trapped kinetically at such sites by mediating the deposition of the modification through recruitment of the SUMOylation machinery. Indeed, RSC subunits were identified in a Siz1 pull-down (Srikumar et al., 2013). Future experiments should address whether H2B SUMOylation modulates RSC activity in vivo or whether H2B SUMOylation depends on RSC activity.

Limitations of the Study

Our approach may be useful to study the preference of protein complexes for particular histone variants. For example, it has been shown that RSC preferentially remodels H2A.Z-containing nucleosomes at promoters (Cakiroglu et al., 2019). By comparing cross-linking efficiencies from equivalent positions between two histone variants, it should be possible to identify such preferences. Specifically, cross-linking from positions close to the acidic patch of canonical H2A and H2A.Z could be used to test whether RSC preferentially resides on nucleosomes with the latter variant. However, in order to allow such comparisons, expression levels of both isoforms must be carefully balanced.

Resource Availability

Lead Contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the Lead Contact, Heinz Neumann (heinz.neumann@mpi-dortmund.mpg.de or heinz.neumann@h-da.de).

Materials Availability

All unique/stable reagents generated in this study are available from the Lead Contact with a completed Materials Transfer Agreement, if there is potential for commercial application.

Data and Code Availability

This study did not generate datasets or code.

Methods

All methods can be found in the accompanying Transparent Methods supplemental file.