Dependency of ISW1a chromatin remodeling on extranucleosomal DNA

Abstract

The nucleosome remodeling activity of ISW1a was dependent on whether ISW1a was bound to one or both extranucleosomal DNAs. ISW1a preferentially bound nucleosomes with an optimal length of approximately 33 to 35 bp of extranucleosomal DNA at both the entry and exit sites over nucleosomes with extranucleosomal DNA at only one entry or exit site. Nucleosomes with extranucleosomal DNA at one of the entry/exit sites were readily remodeled by ISW1a and stimulated the ATPase activity of ISW1a, while conversely, nucleosomes with extranucleosomal DNA at both entry/exit sites were unable either to stimulate the ATPase activity of ISW1a or to be mobilized. DNA footprinting revealed that a major conformational difference between the nucleosomes was the lack of ISW1a binding to nucleosomal DNA two helical turns from the dyad axis in nucleosomes with extranucleosomal DNA at both entry/exit sites. The Ioc3 subunit of ISW1a was found to be the predominant subunit associated with extranucleosomal DNA when ISW1a is bound either to one or to both extranucleosomal DNAs. These two conformations of the ISW1a-nucleosome complex are suggested to be the molecular basis for the nucleosome spacing activity of ISW1a on nucleosomal arrays. ISW1b, the other isoform of ISW1, does not have the same dependency for extranucleosomal DNA as ISW1a and, likewise, is not able to space nucleosomes.

FIG. 1.

ISW1a binding to nucleosomes was enhanced by ∼33 bp of extranucleosomal DNA at one entry/exit of the nucleosome. (A) Affinity-purified ISW1a and ISW1b from strain YTT449 were resolved on 4 to 20% gradient SDS-PAGE stained with Coomassie blue. (B) Mononucleosomes (1.3 pmol) with different lengths of extranucleosomal DNA at one entry/exit site were incubated with 320 fmol (lanes 2, 6, 10, 14, and 18), 640 fmol (lanes 3, 7, 11, 15, and 19) and 1.3 pmol (lanes 4, 8, 12, 16 and 20) of ISW1a. Samples were analyzed on a 4% native polyacrylamide gel. (C) The percentage of nucleosomes bound in panel B was calculated based on the amount of nucleosomes remaining unbound, and this percentage was plotted versus the length of extranucleosomal DNA. Filled diamonds, squares, and triangles indicate values for 0.32, 0.64, and 1.3 pmol of ISW1a, respectively. (D) Nucleosomes (53N0) with DNA end-labeled on the 5′ end of the bottom strand were incubated with two different concentrations of ISW1a (650 fmol, lanes 4 and 5; 1.3 pmol, lanes 6 and 7) and treated with ExoIII (2 U, lanes 2, 4, and 6; 0.2 U, lanes 3, 5, and 7 at 30°C). No ExoIII was added to the sample shown in lane 1. The site at which ExoIII digestion was stalled or halted is indicated in terms of the nucleotides away from the dyad axis.

FIG. 2.

ISW1a preferentially bound nucleosomes with at least 33 bp of extranucleosomal DNA at both entry/exit sites. (A) Mononucleosomes reconstituted with end-labeled 253-bp 601 DNA (53N53) were incubated with (lanes 7 to 10) or without (lanes 1 and 3 to 6) 10 U of NotI. Next, nucleosomes were incubated with 0.12 (lanes 4 and 8), 0.37 (lanes 5 and 9), and 1.1 pmol (lanes 6 and 10) of ISW1a at 30°C and analyzed on a 4% native polyacrylamide gel. The * (panel A) represents ISW1a binding to a negligible amount of 53N53 nucleosomes that were not cut by NotI. (B) Mononucleosomes (1.3 pmol) having 53 bp of extranucleosomal DNA at one entry/exit site, with different lengths of extranucleosomal DNA: 0 (lanes 1 and 7), 10 (lanes 2 and 8), 20 (lanes 3 and 9), 33 (lanes 4 and 10), 40 (lanes 5 and 11), and 53 (lanes 6 and 12) bp at the other entry/exit site were incubated with 200 fmol of ISW1a at 30°C for 30 min. Samples were analyzed by 4% native gel electrophoresis. Lanes 1 to 6 are nucleosomes alone, with no ISW1a.

FIG. 3.

ISW1a bound to 70N70 nucleosomes protected 35 to 38 bp of extranucleosomal DNA. Nucleosomes (70N70) with DNA end-labeled on the 5′ end of the top or bottom strand were incubated with either ISW1a (lanes 10 to 12 and 26 to 28) or ISW1b (lanes 14 to 16 and 30 to 32) and treated with ExoIII (0.22 to 2 U at 30°C). No ExoIII was added to lanes 1, 5, 9, 13, 17, 21, 25, and 29. The site at which ExoIII digestion was stalled or halted is indicated in terms of the nucleotides away from the dyad axis. The nucleotide positions were determined by comparison with the DNA sequencing ladders from the same DNA.

FIG. 4.

ISW2 did not preferentially bind nucleosomes with extranucleosomal DNA at both entry/exit sites. (A) Mononucleosomes, as described in the Fig. 2 legend, were incubated with either 100 (lanes 2, 4, 6, 8, 10, and 12) or 315 (lanes 3, 5, 7, 9, 11, and 13) fmol of ISW2 and analyzed as described previously. (B) The percentages of nucleosomes bound in panel A and those with 0.64 fmol of ISW1a with different lengths of extranucleosomal DNA were plotted. These results were repeated, with the standard deviations shown. The filled squares and circles indicate values for 100 and 315 fmol of ISW2, respectively, and the open circles indicate values for 640 fmol of ISW1a.

FIG. 5.

ISW1a was shown to be a heterodimer of Isw1 and Ioc3 by cross-linking. The concentrations of ISW1a and ISW1b were varied while cross-linking with a constant final concentration of 1% formaldehyde and then separated by SDS-PAGE. Lanes 1, 2, and 7 have no formaldehyde, and the sizes of the molecular mass standards are indicated on the left in kDa.

FIG. 6.

ISW1a bound to nucleosomes with both extranucleosomal DNAs did not interact with nucleosomal DNA two helical turns from the dyad axis. ISW1a bound to 70N0 or 53N53 nucleosomes was footprinted using hydroxyl radical generated from Fe-EDTA. The quantification of the cleavage pattern of nucleosomes with (N+ISW1a, gray) and without (N, black) ISW1a is shown. On the x axis is indicated the relative positions of the different nucleosomal superhelical locations (SHL).

FIG. 7.

ISW1a was unable to mobilize 33N33 nucleosomes. ISW1a (A) and ISW1b (C) remodeling of 0N33 or 33N33 nucleosomes was monitored by native PAGE. The concentration of ISW1a or ISW1b was varied from 1 to 27 nM, and lanes 1 and 6 did not contain either ISW1a or ISW1b. Changes in nucleosome translational positions due to ISW1a (B) and ISW1b (D) remodeling of 33N33 were determined by site-directed mapping of histone-DNA interactions. Quantification of the cleaved sites in DNA is shown for nucleosomes before (black) and after (gray) remodeling. The numbering above the peaks indicates the position of the cut site relative to that of the translation position before remodeling being referred to as 0. Illustrations at the bottom of each profile show the different directions of nucleosome movement observed.

FIG. 8.

The ATPase activity of ISW1a was not stimulated by 33N33 nucleosomes. The ATPase activities of ISW1a (A) and ISW1b (B) were measured in the presence of end-positioned (⧫, 70N0) or centrally positioned (○, 33N33) nucleosomes, free DNA (▴), and enzyme alone (▪) with different incubation times. These assays were done in triplicate with standard deviations shown.

FIG. 9.

The Ioc3 subunit of ISW1a was associated with extranucleosomal DNA. ISW1a was DNA photoaffinity labeled at different sites in the extranucleosomal DNA while bound to either 75N0 or 28N32 nucleosomes. Nucleosomes (75N0, lanes 1 to 5; 28N32, lanes 6 to 9) assembled with a series of photoreactive DNA probes were incubated with either 1.3 pmol (lanes 1 to 5) or 430 fmol (lanes 6 to 9) of ISW1a, UV irradiated, and digested with DNase I and S1 nuclease as described previously. The probe position numbers indicate the site where photoreactive nucleotides were incorporated relative to the dyad axis.