pubmed.ncbi.nlm.nih.gov

Nucleosome-Chd1 structure and implications for chromatin remodelling - PubMed

  • ️Sun Jan 01 2017

. 2017 Oct 26;550(7677):539-542.

doi: 10.1038/nature24046. Epub 2017 Oct 11.

Affiliations

Nucleosome-Chd1 structure and implications for chromatin remodelling

Lucas Farnung et al. Nature. 2017.

Abstract

Chromatin-remodelling factors change nucleosome positioning and facilitate DNA transcription, replication, and repair. The conserved remodelling factor chromodomain-helicase-DNA binding protein 1(Chd1) can shift nucleosomes and induce regular nucleosome spacing. Chd1 is required for the passage of RNA polymerase IIthrough nucleosomes and for cellular pluripotency. Chd1 contains the DNA-binding domains SANT and SLIDE, a bilobal motor domain that hydrolyses ATP, and a regulatory double chromodomain. Here we report the cryo-electron microscopy structure of Chd1 from the yeast Saccharomyces cerevisiae bound to a nucleosome at a resolution of 4.8 Å. Chd1 detaches two turns of DNA from the histone octamer and binds between the two DNA gyres in a state poised for catalysis. The SANT and SLIDE domains contact detached DNA around superhelical location (SHL) -7 of the first DNA gyre. The ATPase motor binds the second DNA gyre at SHL +2 and is anchored to the N-terminal tail of histone H4, as seen in a recent nucleosome-Snf2 ATPase structure. Comparisons with published results reveal that the double chromodomain swings towards nucleosomal DNA at SHL +1, resulting in ATPase closure. The ATPase can then promote translocation of DNA towards the nucleosome dyad, thereby loosening the first DNA gyre and remodelling the nucleosome. Translocation may involve ratcheting of the two lobes of the ATPase, which is trapped in a pre- or post-translocation state in the absence or presence, respectively, of transition state-mimicking compounds.

PubMed Disclaimer

Conflict of interest statement

The authors declare no competing financial interest.

Figures

Extended Data Figure 1
Extended Data Figure 1. Cryo-EM structure determination and analysis.

a. Formation of the nucleosome-Chd1-FACT-Paf1C complex. SDS-PAGE of peak fraction used for cryo-EM grid preparation containing Chd1, FACT subunits, Paf1C subunits and histones. Identity of the bands was confirmed by mass spectrometry. For gel source data, see Supplementary Figure 1. b. Representative cryo-EM micrograph of data collection. c. 2D class averages contain nucleosome-like shapes. d. Sorting and classification tree used to reconstruct the nucleosome-Chd1 particle at 4.8 Å resolution. Steps 1 and 2 of batch 1 global classification are shown representatively for all three batches.

Extended Data Figure 2
Extended Data Figure 2. Quality of the nucleosome-Chd1 structure.

a. Overall fit of the nucleosome-Chd1 structure to the electron density. Two views are depicted as in Fig. 1b, c. b-f. Electron density (grey mesh) for various Chd1 domains reveals secondary structure and a good fit for DNA (SHL -4 to SHL +7). g. Superposition of the histone octamer core with canonical octamer core (PDB code 3LZ0). The canonical octamer core is rendered in grey. h. Nucleosome-Chd1 reconstruction colored according to local resolution. i. Angular distribution of particles. Red dots indicate the presence of at least one particle image assigned within ±1°. Shading from white to black indicates the density of particle images at a given orientation. j. Estimation of the average resolution. The dark blue line indicates the Fourier shell correlation between the half maps of the reconstruction. The dotted light blue line indicates the Fourier shell correlation between the derived model and the reconstruction. Resolutions are given for the FSC 0.143 and the FSC 0.5 criterion. The dotted lines show the Fourier shell correlation between the derived Chd1 domains and the corresponding masked regions.

Extended Data Figure 3
Extended Data Figure 3. Chd1-DNA interactions and Chd1 interaction interfaces.

a. Overview of Chd1-DNA interactions. b. Contact of chromo-wedge with DNA at SHL +1. c. Secondary DNA contacts of ATPase. Contact of motif Ib with first DNA gyre around SHL -6. d. Modeling linear B-DNA (orange) onto the ATPase motor in the nucleosome-Chd1 structure leads to a clash with the double chromodomain (purple). B-DNA was superimposed onto nucleosomal DNA at SHL +2. e. ADP·BeF3 binds in the active site of the Chd1 ATPase motor. Electron density is shown for ADP·BeF3, motif I (Walker A, P-loop, residues 403-410), motif II (Walker B, residues 510-515), and the arginine fingers (R804 + R807). Motifs I and II are shown in ribbon representation. ADP·BeF3 and the arginine finger residues are shown as sticks. Density for ADP is strong, whereas density for BeF3- is weaker and thus we cannot formally rule out that BeF3- is not bound or shows only partial occupancy. f. Contact of W793 with the phosphate backbone of the guide strand at SHL +2. Electron density is shown as a grey mesh. Side chain of W793 is shown as a stick representation. g. Interface between the double chromodomain and the SANT/SLIDE domains of the DNA binding region. Chd1 domains are colored as in Fig. 1a. h. Sequence of the Widom 601 sequence with 63 bp of extranucleosomal DNA.

Extended Data Figure 4
Extended Data Figure 4. ATPase conservation and histone H4 tail binding.

a. Chd1 binds the N-terminal tail of histone H4 (green) with ATPase lobe 2 (surface representation coloured according to electrostatic surface potential; red, negative, white, neutral, blue, positive). The view is the inverse of that in Fig. 1b, i.e. after a 180° rotation. b. Chd1 ATPase activity results in DNA translocation towards the octamer dyad, loosening DNA gyre 1 and triggering nucleosome remodelling. c. Sequence alignment of ATPase regions in ScChd1 (356-883), ScIsw1 (177-689), ScSnf2 (746-1270), HsChd4 (703-1233), DmMi-2 (707-1231), and SsoRad54 (423-802). Arginine ‘fingers’ of ScChd1 (R804+R807) are indicated and ATPase motifs are underlined. Sequence coloured according to identity. Darker shades of blue indicate higher conservation, whereas lighter shades of blue indicate less conservation. Alignment was generated with MAFFT and visualized using JalView.

Figure 1
Figure 1. Structure of nucleosome-Chd1 complex.

a. Chd1 domain architecture. Residues at domain boundaries are indicated. b-d. Three views of the structure. Chd1 domains are colored as in (a). H2A, H2B, H3, H4, tracking strand, and guide strand are in yellow, red, light blue, green, dark blue, and cyan, respectively. The histone octamer dyad axis is indicated as black line or black oval circle. SHL, superhelical location.

Figure 2
Figure 2. Chd1-DNA interactions.

a. Detachment of nucleosomal DNA from the histone octamer at SHL -7 to -5. Extranucleosomal DNA rotates by ~60º with respect to its location in the absence of Chd1 (orange, modelled by extending nucleosomal DNA with B-DNA). The position of Chd1 is indicated in grey color. b. Primary ATPase-DNA interactions. Location of ATPase motifs on lobe 1 and lobe 2 are highlighted in red and green, respectively. The view is from the center of the histone octamer onto nucleosomal DNA. DNA register is indicated by numbering next to DNA bases. Color code is as in Fig. 1. ADP·BeF3 is shown as grey spheres. The model of lobe 2 in the pre-translocated position (grey) was derived from superposition of the nucleosome-Snf2 structure (PDB code 5X0Y).

Figure 3
Figure 3. Chd1 structural changes and ATPase activation.

a. Swinging of double chromodomain (open state, light pink; closed state, purple) onto DNA liberates ATPase lobe 2 (grey). The structure of free Chd1 in its inactive state (PDB code 3MWY) was placed by superimposing ATPase lobe 1 (orange). In the inactive state, the chromo-wedge binds to a basic patch on lobe 2. View as in Fig. 1c. b. ATPase closure and activation. Lobe 2 (sea green) rotates by ~40º to allow for binding of ADP·BeF3 (grey spheres). BeF3- was modeled in a tetrahedral conformation for simplicity but is likely planar when it mimics part of the pentavalent transition state of ATP hydrolysis.

Similar articles

Cited by

References

    1. Narlikar GJ, Sundaramoorthy R, Owen-Hughes T. Mechanisms and Functions of ATP-Dependent Chromatin-Remodeling Enzymes. Cell. 2013;154:490–503. - PMC - PubMed
    1. Delmas V, Stokes DG, Perry RP. A mammalian DNA-binding protein that contains a chromodomain and an SNF2/SWI2-like helicase domain. PNAS. 1993;90:2414–2418. - PMC - PubMed
    1. Lieleg C, et al. Nucleosome spacing generated by ISWI and CHD1 remodelers is constant regardless of nucleosome density. Mol Cell Biol. 2015;35:1588–1605. - PMC - PubMed
    1. Hughes AL, Rando OJ. Comparative Genomics Reveals Chd1 as a Determinant of Nucleosome Spacing in Vivo. G3 (Bethesda) 2015;5:1889–1897. - PMC - PubMed
    1. Lusser A, Urwin DL, Kadonaga JT. Distinct activities of CHD1 and ACF in ATP-dependent chromatin assembly. Nat Struct Mol Biol. 2005;12:160–166. - PubMed

Publication types

MeSH terms

Substances