|
Extended Data Fig. 2. Data processing scheme of the ADP-BeFx sample.a, A representative raw micrograph (out of 7,769 total images) of the SWI/SNF-nucleosome complex assembled in the presence of ADP-BeFx. b, Flow chart of the cryo-EM data processing procedure. The particle stack of class 3 (198,543 particles; blue dashed box) after the first sorting was chosen for the combined processing (Extended Data Fig. 4). c, Fourier Shell Correlation curves of the complex showing a final average resolution of 8.96Ã
(FSC=0.143).
|
|
Extended Data Fig. 3. Data processing scheme of the ATPγS sample.a, A representative raw micrograph (out of 6,903 total images) of the SWI/SNF-nucleosome complex assembled in the presence of ATPγS. b, Flow chart of the cryo-EM data processing procedure. The particle stack of class 5 (192,029 particles; red dashed box) after the first sorting was chosen for the combined processing (Extended Data Fig. 4). c, Fourier Shell Correlation curves of the complex showing a final average resolution of 10Ã
(FSC=0.143).
|
|
Extended Data Fig. 4. Data processing scheme of the combined dataset.a, Flow chart of the data processing procedure by combining ADP-BeFx (Extended Data Fig. 2) and ATPγS (Extended Data Fig. 3) datasets. b, Fourier Shell Correlation curves of the body showing a final average resolution of 4.7Ã
(FSC=0.143). c, model-map FSC of SWI/SNF body.
|
|
Extended Data Fig. 5. Comparing the structures of chromatin remodelers from different families.All but the INO80 complex have their ATPase module binding to SHL2/−2. INO80 engages the nucleosomal DNA at SHL −6 to −7. SWI/SNF is different from the INO80/SWR1 family remodelers in that its Arp module (Arp7/9) is sandwiched between the Body and the ATPase modules. The Snf2 ATPase module is connected through the long HSA domain to the rest of the complex, whereas the ATPases INO80 and Swr1 directly contact the main body of the corresponding complexes. All remodelers are aligned based on histone proteins. The ATPase in each complex is colored red, whereas Arp proteins are colored green. PDB codes of the other chromatin remodelers are shown in parentheses.
|
|
Extended Data Fig. 6. Structural features of the Spine sub-module of the SWI/SNF complex.a, The Swi3 Coiled-coil dimer (left) resembles the structure of the dominant-negative allele of MYC (OmoMYC, shown in blue in the middle panel; PDB ID 5I4Z). The OmoMYC structure was rigid body docked in the Spine density corresponding to the Swi3 Coiled-coil and then compared with the Swi3 Coiled-coil. b, Swi3 forms an asymmetric dimer in the SWI/SNF complex. c, The density at the tip of the Spine shows features of β-sheet and is therefore assigned to Snf12 based on closed proximity to Snf12 SWIB domain and secondary structure prediction.
|
|
Extended Data Fig. 7. Structural features of the Arm sub-module of the SWI/SNF complex.a, The Snf5 RPT1 and Swi3 SWIRMA heterodimer was aligned with the human BAF47/BAF155 crystal structure (PDB ID 5GJK). b, The Snf5 RPT2 and Swi3 SWIRMB heterodimer was aligned with the human BAF47/BAF155 crystal structure (PDB ID 5GJK). c, RPT1/SWIRMA interface shows slight difference with the RPT2/SWIRMB interface. RPT1 and 2 was aligned, resulting in the SWIRM domains slightly shifting from each other. d, Comparing the interfaces between the SWIRM domains and Swi1. The two SWIRM domains was aligned, resulting in Swi1 H4 (yellow; contacting SWIRMB) occupying a similar position as Swi1 H1 (gold) and Snf5 H-N on SWIRMA. In all panels, structural elements related to RPT1/SWIRMA is depicted using darker colors, whereas RPT2/SWIRMB associated structures is shown in lighter colors.
|
|
Extended Data Fig. 8. Structural features of the Core sub-module of the SWI/SNF complex.a, The Swi1 ARM repeat domain is aligned with β-catenin (gray; PDB ID 3BCT). The insertions of the Swi1 ARM repeat domain are depicted in magenta. b, Detailed interaction between the Swi1 ARM repeat domain with the Arm and Hinge sub-modules. c, Detailed interaction between the Swi1 ARM repeat domain with the Spine sub-module. The EM density of Swi1 is also shown in b and c as mesh.
|
|
Extended Data Fig. 9. Interactions between the Snf2 Anchor domain and the rest of the SWI/SNF complex.a, The Snf2 Anchor linker region interacts with the Swi1 ARM repeat domain. b, Snf2 Anchor helices 1 and 2 are sandwiched between the two SANT domains of Swi3 in the Hinge region. The EM density of Snf2 Anchor is shown as mesh.
|
|
Extended Data Fig. 10. Locations of yeast specific subunits and the extranucleosomal DNA.a and b, Snf6 (a) and Swp82 (b) are positioned at peripheral locations within SWI/SNF. Map and structural models are shown with Snf6 and Swp82 highlighted in a and b, respectively. Swp82 is in close proximity to the nucleosomal DNA near SHL â2 (b). c, The extranucleosomal DNA density is close to Snf6 and is indicated as dashed lines. The N-termini of Swi1 and Snf5 are also labeled. The N-terminal regions of Swi1 and Snf5, which are highly flexible and therefore not resolved in the structure, could take trajectories close to the extranucleosomal DNA.
|
|
Fig. 1:. Cryo-EM structure of the SWI/SNF-nucleosome complex.a, Front (left) and back (right) views of the cryo-EM composite map (see Methods) of the SWI/SNF-nucleosome complex assembled in the presence of ADP-BeFx. b, Same views of structural model of the SWI/SNF-nucleosome complex as in a. Subunits are colored same as in a.
|
|
Fig. 2:. Structural organization of the body of SWI/SNF.a, A cartoon depicting the molecular architecture of the SWI/SNF-nucleosome complex. b-e, Close-up view of detailed interactions within the Spine (b), the Hinge (c), the Arm (d) and the Core (e) modules, respectively. Highlighted elements are shown as cartoon and others as transparent surface representation. Blue spheres depict the locations of a subset of invariant residues harboring cancer patient mutations that occur at interfaces between these conserved subunits. Yeast residue number is shown in parentheses. Subunits are colored the same as in Fig. 1.
|
|
Fig. 3:. SWI/SNF-nucleosome interactions.a, An overview of the SWI/SNF-nucleosome complex depicting how the HSA and Anchor domains of Snf2 load the ATPase onto the nucleosome. Blue spheres indicate the positions of a subset of invariant residues harboring cancer patient mutations in Snf2/SMARCA4/BRG1 that reside between conserved SWI/SNF subunits. Yeast residue number is shown in parentheses. Dotted lines indicate regions enlarged in c. b, Sequence alignment of the C-terminal extension of Snf5 RPT2. Sc, Saccharomyces cerevisiae; Cg, Candida glabrata; Td, Torulaspora delbrueckii; Zr, Zygosaccharomyces rouxii; Hs, Homo sapiens; Mm, Mus musculus. The yeast Sfh1 subunit of the RSC complex is also included. Domain organization of Snf5/SMARCB1/INI1/BAF47 is also shown. c, Close-up view showing the C-terminal extension (dotted line) of Snf5 RPT2 contacting the acidic patch of the nucleosome (dotted red circle). d, Model for chromatin remodeling by SWI/SNF. Prior to engagement with nucleosome, the ATPase module of SWI/SNF is flexible. Nucleosome binding involves the ATPase and the HSA of Snf2 recognizing SHL2 and SHL-6, respectively. The Arm of SWI/SNF interact with the nucleosome surface near the acidic patch, likely serving as an anchor during active remodeling to keep the octamer from moving. Upon ATP hydrolysis, a bulge in the nucleosomal DNA is introduced at SHL 211, which is then propagated to the Exit side of the nucleosome when ADP is released and the next ATP molecule is bound, resulting in nucleosomal DNA translocation.
|
