1. Volume 12, Issue 2, Pages 449-460 (August 2003)
Crystal Structure and Functional Analysis of a Nucleosome Recognition Module of the Remodeling Factor ISWI 
Tim Grüne, Jan Brzeski, Anton Eberharter, Cedric R Clapier, Davide F.V Corona, Peter B Becker, Christoph W Müller 
Molecular Cell 
Volume 12, Issue 2, Pages (August 2003)
DOI: /S (03)

2. Figure 1
DNA and Nucleosome Binding Properties of Full-Length ISWI, ISWI-N, and ISWI-C
(A) Schematic representation of the domain structure of the ATPase ISWI. Starting residues of proteolytic fragments A, B, C1, and C2 with their estimated molecular weight and boundaries of constructs ISWI-N and ISWI-C are depicted.
(B) Recombinant ISWI derivatives. Full-length ISWI (FL), the N- and C-terminal fragments, were expressed in E. coli, purified, resolved by SDS-PAGE, and stained with Coomassie blue.
(C) ISWI binds four-way-junction (4WJ) DNA. Binding reactions contained 4WJ DNA and ISWI-FL (31, 62, 125, 250, 500, 1000 fmols), ISWI-N, or ISWI-C (0.5, 1, 2, 4, 8, 16 pmols each). The resulting complexes were resolved by native gel electrophoresis in 4.5% polyacrylamide and stained with SYBR-GOLD.
(D) ISWI binds nucleosomes with overhanging DNA. Binding reactions contained mononucleosomes assembled on a radiolabeled 248 bp DNA fragment and ISWI-FL (16, 31, 62, 125, 250, 500, 1000 fmols), ISWI-N, or ISWI-C (0.25, 0.5, 1, 2, 4, 8, 16 pmols each). The resulting complexes were resolved by electrophoresis on a 1.4% agarose gel and visualized by autoradiography.
(E) ISWI binds nucleosome core particles. Binding reactions contained mononucleosomes assembled on a radiolabeled 146 bp DNA fragment and either of ISWI-FL, ISWI-N, or ISWI-C (8, 16, 31, 62, 125 fmols; 0.25, 0.5, 1, 2, 4, 8 pmols). The resulting complexes were resolved by electrophoresis on a 1.4% agarose gel and visualized by autoradiography.
Molecular Cell  , DOI: ( /S (03) )

3. Figure 2 C-Terminal Deletion of ISWI Abolishes Substrate Recognition
The ATPase activity of 0.84 pmols of either ISWI-FL or ISWI-N was assayed during a time course in reactions containing no effector (blue lines), or 100 ng of plasmid DNA (green lines) or the same amount of DNA assembled into nucleosomes (red lines). The number of ATP molecules hydrolyzed per ISWI molecule are approximate numbers calculated from the percentage of hydrolysis of a known amount of ATP in the reaction.
Molecular Cell  , DOI: ( /S (03) )

4. Figure 3 Structure of ISWI-C
(A) Two orthogonal views of ISWI-C (691–991). The HAND domain is depicted in blue, SANT domain in green, SLIDE domain in yellow, and the spacer helix connecting the SANT and SLIDE domains in red. Disordered loops are depicted in gray. (A), (C), and Figures 4A and 4C were produced with the programs Molscript (Kraulis, 1991) and Raster3D (Merritt and Bacon, 1997).
(B) Sequence alignment of ISWI homologs from Drosophila melanogaster (dmISWI), human (hSNF2H), yeast (scIsw1p), the Arabidopsis thaliana homolog At5g18620 (atISWI), and the SANT domain of yeast Ada2p. Conserved and conservatively substituted residues are highlighted in blue. The secondary structure elements of the dmISWI structure are indicated. SANT (796–845) and SLIDE (898–962) domains as predicted by program SMART are underlined; brackets correspond to the deletion mutants ΔSANT and ΔSLIDE. Dashed lines indicate disordered regions.
(C) Electrostatic surface representation of ISWI-C. Depicted are the surfaces pointing away from the nucleosome (I) and facing it (II) in the hypothetical nucleosome/ISWI-C model described in the Discussion.
Molecular Cell  , DOI: ( /S (03) )

5. Figure 4
Comparison of SANT and SLIDE Domains with DNA Binding Modules of c-Myb
(A) Stereo diagram of the CA-backbones of SANT (green, rmsd27CA = 1.2 Å) and SLIDE (yellow, rmsd34CA = 1.1 Å) domain superimposed with DNA-bound repeat R3 of c-Myb (red). For the SANT domain only helixes SA1 and SA3 were used in the superposition.
(B) Structural sequence alignment of SANT and SLIDE domain with DNA binding modules R2 and R3 of c-Myb. Residues in SANT and SLIDE which allow similar contacts as in c-Myb modules R2 and R3 are indicated by green dots. Residues which do not allow similar contacts are indicated by red squares. Residues in c-Myb modules R2 and R3 contacting DNA bases and DNA backbone are highlighted in blue and yellow, respectively. Conserved and conservatively substituted residues are depicted on a light blue background.
(C) Models of hypothetical complexes of SANT (left) and SLIDE (right) bound to DNA based on the superposition of both domains onto c-Myb repeat R3. Residues compatible and incompatible with DNA binding are depicted in green and red, respectively. In (B) the corresponding residues are marked by red squares or green spheres.
Molecular Cell  , DOI: ( /S (03) )

6. Figure 5 Roles of SANT and SLIDE Domains in Substrate Binding
(A) Full-length ISWI (ISWI-FL) and derivatives characterized by individual deletions of either the SANT or SLIDE domains or simultaneous deletion of both domains (ΔSANT, ΔSLIDE, ΔSANTΔSLIDE, respectively) were expressed in E. coli, purified, and resolved by SDS-PAGE.
(B) Binding of ISWI derivatives to 4WJ DNA. Binding reactions contained 4WJ DNA, and either full-length ISWI (ISWI-FL) or the deleted proteins as indicated (62, 125, 250, 500, and 1000 fmols). Complexes were analyzed as in Figure 1C.
(C) Binding of ISWI derivatives to nucleosomes. Binding reactions contained mononucleosomes reconstituted on a 248 bp DNA fragment, and either full-length ISWI (ISWI-FL) or the deleted proteins as indicated (16, 31, 62, 125, 250, 500, and 1000 fmols). Complexes were analyzed as in Figure 1D.
Molecular Cell  , DOI: ( /S (03) )

7. Figure 6 Roles of SANT and SLIDE Domains in ISWI Activity
(A) The ATPase activity of the indicated ISWI derivatives was tested analogously to Figure 2 during a 40 min time course in reactions containing either no effector (blue lines), 100 ng of plasmid DNA (green lines), or the same amount of DNA assembled into nucleosomes (red lines). Note the different scale in the lower part of the figure. The figure summarizes three independent experiments.
(B) Nucleosome sliding assay. As indicated, 35, 70, 140, and 280 fmols of full-length ISWI or of the ISWI derivatives were tested for their ability to move a nucleosome positioned centrally on a 248 bp fragment to a peripheral position. The bands corresponding to central and peripheral nucleosome positions are indicated to the left. The sliding reaction was stopped by the addition of competitor DNA and samples were separated by native PAGE. The radiolabeled nucleosomal DNA was detected by autoradiography. The asterisk marks the positions of nucleosomes moved by ISWIΔSANT.
(C) Nucleosome spacing assay. One pmol of each ISWI derivative was tested for nucleosome spacing activity in a NAP-1 assisted nucleosome assembly reaction. The chromatin was partially digested with three increasing amounts of micrococcal nuclease (from left to right), and the resulting DNA fragments were purified, resolved on an agarose gel, and stained with SYBR Gold. The size marker is a 123 bp ladder. Note the DNA fragment ladders indicative of regularity of the nucleosome array in the reactions containing intact ISWI or the ΔSANT derivative.
Molecular Cell  , DOI: ( /S (03) )