H2A.Z deposition requires each nucleosomal acidic patch.

(A) Schematic of SWR1C mediated dimer exchange. Cylinder is representative of a nucleosome; light grey area of cylinder represents H3/H4 tetramer, dark grey area of cylinder represents AB heterodimer, yellow star represents Cy5, pink star represents Cy3, and solid black line represents DNA. The nucleosome undergoes two rounds of SWR1C mediated dimer eviction where ZB heterodimers (orange) replace AB heterodimers. (B) 77N0-Cy3 nucleosomes were remodeled by SWR1C under single turnover conditions (30 nM SWR1C, 10 nM 77N0-Cy3 nucleosomes, 70 nM ZB dimers, 1mM ATP or AMP-PNP). Dimer eviction is monitored by measuring Cy5 emission at 670 nm. SWR1C was able to perform dimer exchange on an AB-Cy5/AB-Cy5 nucleosome (red line) but was unable to exchange dimers on an AB-apm-Cy5/AB-apm-Cy5 nucleosome (black line). The AMP-PNP reaction contained the AB-apm-Cy5/AB-apm-Cy5 nucleosomal substrate. (C) Asymmetrically assembled 77N0-Cy3 and Cy3-0N77 nucleosomes were remodeled by SWR1C under single turnover conditions (50 nM SWR1C, 10 nM nucleosomes, 50 nM ZB dimers, 1mM ATP) and one-phase eviction rates were calculated in Prism 9 and plotted. The dimer labels are listed in linker proximal (LP), linker distal (LD) order (see inset for template nucleosome represented as a cylinder with H3/H4 tetramer in light grey, LP dimer in green, LD dimer in dark blue, DNA as a solid black line with Cy3 (pink star) on the 0bp-linker side). Nucleosomes with a ZB dimer contralateral to the dimer being evicted had significantly faster rates for both linker distal (first two bars) and linker proximal (second two bars). Linker distal eviction was also significantly faster than linker proximal eviction. Rates for nucleosomes containing an AB-apm dimer contralateral to the dimer being evicted (5th and 6th bars) were not significantly different from no enzyme controls (last bar) in either orientation. SWR1C was also unable to evict a dimer with an acidic patch mutant (7th bar). (D) Asymmetrically assembled 0N0-Cy3 nucleosomes were remodeled by SWR1C under single turnover conditions (50 nM SWR1C, 10 nM nucleosomes, 50 nM ZB dimers, 1mM ATP) and one-phase eviction rates were calculated in Prism 9 and plotted. Nucleosomes contained a single AB-Cy5 dimer and a contralateral AB or ZB dimer (see inset for template nucleosome represented as a cylinder with H3/H4 tetramer in light grey, AB-Cy5 dimer in dark grey, contralateral AB or ZB dimer represented by dark grey to orange gradient, 0N0-Cy3 DNA as a solid black line with Cy3 (pink star) label on one side). AB-Cy5/ZB 0N0-Cy3 nucleosomes had significantly faster eviction than AB-Cy5/AB 0N0-Cy3 nucleosomes. At least 3 independent nucleosome preparations were used for substrate, and error bars reflect 95% confidence intervals from at least 3 replicates.

Nucleosomal acidic patches are required for SWR1C nucleosome binding.

Asymmetrically assembled 77N0-Cy3 nucleosomes were incubated with SWR1C serially diluted in concentration from approximately 1 uM to 1 nM. Fluorescence polarization values were collected and plotted by concentration for each nucleosome type. Each graph has a representative nucleosome at the top left (represented by a cylinder with H3/H4 tetramer in light grey, AB dimer in dark grey, AB-apm dimer in light blue, 77N0-Cy3 DNA as a solid black line with Cy3 (pink star) label on the 0bp-linker side). FP data was normalized between experiments by translation, which does not effect the binding curve fit so that plots can be compared at a glance. (A) AB/AB 77N0-Cy3 nucleosomes showed a Kd as determined by a fit to the Morrison equation (blue line). The Kd calculated for each individual replicate was 13.55, 13.32, 13.94, and 13.75. The average of those individual values is 13.64. The standard deviation is 0.27. A Kd was unable to be calculated for nucleosomes with a linker proximal acidic patch mutant (B), a linker distal acidic patch mutant (C), or both acidic patches mutated (D) at the range of concentrations of SWR1C tested suggesting at least a 10-fold reduction in binding affinity. Error bars reflect 95% confidence intervals from at least 3 replicates.

Swc5 contains a key arginine-rich motif.

(A) Alignment of SWR1C subunit Swc5 homologs reveal a conserved arginine in a region we named the RRKR motif in S. cerevisiae adjacent to the essential BCNT region and conserved LDW residues (asterisks are used to denote residues mutated in RRKR-4A and LDW-3A mutants). A linear schematic of this gene with the acidic N-terminal domain in orange, gene body in grey, RRKR motif in red and BCNT region in purple shows relative locations of these domains. (B) Swc5 and Swc5 derivatives were reconstituted into SWR1CSwc5Δ complexes to measure their activity on dimer exchange using FRET at single turnover conditions (30 nM SWR1C, 10 nM 77N0-Cy3 AB-Cy5/AB-Cy5 nucleosomes, 70 nM ZB dimers, 1 mM ATP). The RFU for each reaction was normalized and plotted as a function of time. SWR1CSwc5Δ was unable to carry out dimer exchange (black triangles), and recombinant Swc5 (rSwc5) rescued activity (red circles). SWR1CSwc5Δ reconstituted with Swc5 containing mutations in the LDW (rSwc5LDW-3A) or RRKR motifs (rSwc5RRKR-4A) showed no dimer eviction activity (purple and blue circles respectively), while an N-terminal truncation of Swc5 (rSwc579-303) had a reduction in activity compared to full length Swc5 (yellow triangles). (C) Asymmetrically assembled 77N0-Cy3 nucleosomes were incubated with SWR1CSwc5Δ serially diluted in concentration from approximately 1 uM to 1 nM. Fluorescence polarization values were collected and plotted by concentration. A representative nucleosome is shown at the top left (represented by a cylinder with H3/H4 tetramer in light grey, AB dimer in dark grey, 77N0-Cy3 DNA as a solid black line with Cy3 (pink star) label on the 0bp-linker side). (D) Nucleosomal stimulation of SWR1C ATPase activity with (purple bars) or without (green bars) ZB dimers added was measured for complexes containing SWR1C, SWR1CSwc5Δ, SWR1CSwc5Δ +rSwc5, and SWR1CSwc5Δ +rSwc5RRKR-4A using a phosphate sensor assay. Calculated rates were normalized to basal SWR1C activity (blue bar). Stimulation of ATPase activity was lost in the SWR1CSwc5Δ complex and reduced in the SWR1CSwc5Δ +rSwc5RRKR-4A complex. Error bars reflect 95% confidence intervals from at least 3 replicates.

Swc5 is a nucleosome binding subunit.

Gel mobility shift assays were performed with 5 nM 0N0 nucleosomes and recombinant Swc5 ranging in concentration from 0 to 500 nM, in 50 nM increments. Swc5 was able to bind AB/AB nucleosomes (A) but showed reduced binding of AB-apm/AB-apm nucleosomes (B). Swc5LDW-3A bound AB/AB nucleosomes (C) whereas Swc5RRKR-4A showed reduced binding (D). Percent of nucleosomes bound for each condition was calculated using ImageQuant and plotted by Swc5 concentration (E) and specific binding Kd values were predicted along with 95% confidence intervals (CI) predicted in Prism 9 (F).

TR-FRET assay confirms Swc5 nucleosome binding activity.

(A) Schematic of the TR-FRET assay. Recombinant 6His-Swc5 is labelled with a ULight alpha 6His acceptor antibody, and biotinylated nucleosomal DNA is labelled with an Eu-streptavidin acceptor fluorophore. The FRET signal increases as 6His-Swc5 binds to the nucleosome. (B) TR-FRET assay was performed with 2nM 31N1 nucleosomes and recombinant Swc5 or Swc5RRKR-4A ranging in concentration from 0.6-6000 nM. Recombinant wildtype Swc5 bound with an apparent Kd of ∼133nM, while the Swc5RRKR-4A derivative bound with an apparent Kd of ∼592nM. Note that the Swc5RRKR-4A assays did not reach full saturation, and thus the measured Kd is likely an under-estimate of the reported value. Kd values were determined from triplicate titrations of each Swc5 variant and are reported as means ± standard error of the mean.

Swc5 interacts with the nucleosome acidic patch.

(A) Schematic of Oregon green based fluorescence quenching assay. Top shows a nucleosome (represented as a cylinder with H3/H4 tetramer in light grey, AB dimers in dark grey, and 0N0 DNA as a solid black line) with an Oregon green label (green star) on H4. Oregon green is exquisitely sensitive to changes in its local environment. Binding of an acidic patch interacting protein (pink) does not interact with Oregon green label on H4, resulting in no change in fluorescence intensity. A label on H2B near the acidic patch (bottom) is affected by an acidic patch interacting protein, resulting in decreased fluorescence intensity. (B) Swc5 was added to 10 nM H4-OG AB/AB 0N0 nucleosomes (orange squares), H2B-OG AB/AB 0N0 nucleosomes (blue circles) and H2B-OG AB-apm/AB-apm 0N0 nucleosomes (black circles) resulting in quenching of the AB/AB H2B-OG 0N0 nucleosomes but not the H4-OG AB/AB 0N0 nucleosomes or the H2B-OG AB-apm/AB-apm 0N0 nucleosomes. (C) Swc5LDW-3A similarly quenched H2B-OG AB/AB 0N0 nucleosomes with no change in H4-OG AB/AB 0N0 nucleosomes while Swc5RRKR-4A (D) did not show significant quenching in either condition. Error bars reflect standard deviations from at least 3 replicates.

Cryo-EM structure of Swc5/nucleosome complex.

(A-D) Reconstructions of selected particle subsets obtained from 3D classification that was focused on the Swc5 moiety (purple) illustrate the dynamic nature of Swc5 binding to the nucleosome. Three interactions persist among the 3D classes: (E) Swc5 interaction with the N-terminal tail of histone H4 (green), (F) Swc5 interaction with the C-terminal helix of histone H2B including K123, and (G) an Arginine anchor bound into the acidic patch of the histone core. The gray circles and arrows denote the reconstructions and locations from where the cut-out panels originate.

Swc5 arginine-rich region is key for SWR1C activity in vivo.

Genetic complementation using a swc5Δ strain (MATa his3Δ1 leu2Δ0 met15Δ0 ura3Δ0 swc5!1::kanMX4) transformed with an empty pRS416 CEN/ARS URA3 vector (top row in both panels) or the same vector with various alleles of SWC5. Cultures were grown to an OD of 1 and spotted with serial dilution onto plates with synthetic complete media lacking uracil without (left) or with (right) 2% formamide.