The INO80 ATP-dependent chromatin remodeling complex is a nucleosome spacing factor

Abstract

The mobilization of nucleosomes by the ATP-dependent remodeler INO80 is quite different from another remodeler (SWI/SNF) that is also involved in gene activation. Unlike that recently shown for SWI/SNF, INO80 is unable to disassemble nucleosomes when remodeling short nucleosomal arrays. Instead, INO80 more closely resembles, although with notable exceptions, the nucleosome spacing activity of ISW2 and ISW1a, which are generally involved in transcription repression. INO80 required a minimum of 33 to 43 bp of extranucleosomal DNA for mobilizing nucleosomes, with 70 bp being optimal. INO80 prefers to move mononucleosomes to the center of DNA, like ISW2 and ISW1a, but does so with higher precision. Unlike ISW2/1a, INO80 does not require the H4 tail for nucleosome mobilization; instead, the H2A histone tail negatively regulates nucleosome movement by INO80. INO80 moved arrays of two or three nucleosomes with 50 or 79 bp of linker DNA closer together, with a final length of ∼30 bp of linker DNA or a repeat length of ∼177 bp. A minimum length of >30 bp of linker DNA was required for nucleosome movement and spacing by INO80 in arrays.

FIG. 1.

INO80 preferentially binds to nucleosomes with extranucleosomal DNA. (A) The affinities of INO80 for end-positioned nucleosomes with various lengths of extranucleosomal DNA were measured by gel shift analysis on a 4% native PAGE with 1× TE. The binding reaction mixtures had 30 nM nucleosomes with end-labeled PCR DNA and increasing amounts of INO80 (2.5, 5, 10, 20, and 40 nM). These reaction mixtures did not contain any salmon sperm DNA. (B) The observed K_d values of INO80 for nucleosomes with different extranucleosomal DNA lengths was determined from the data shown in panel A and are plotted with respect to the length of extranucleosomal DNA. These data are from three independent binding titration experiments, and the standard deviations are shown.

FIG. 2.

INO80 optimally requires 70 bp of extranucleosomal DNA to move nucleosomes to the center of DNA. (A to H) Nucleosome movement was tracked by site-directed mapping, with nucleosomes having a photoreactive group attached to amino acid residue 53 of histone H2B. Nucleosomes with various lengths of extranucleosomal DNA were used as described for Fig. 1, except that in these reaction mixtures salmon sperm DNA was added in place of unlabeled PCR DNA. The reaction mixtures contained 33 nM nucleosomes (based on octamer concentration), 40 nM INO80, and 800 μM ATP. The samples were analyzed as described in Materials and Methods, and the phosphorimages before (black) and after (gray) INO80 remodeling are overlaid to illustrate changes in nucleosome position. The number 0 refers to the original nucleosome position, and the other numbers to how many base pairs the nucleosome has shifted after INO80 remodeling. The lengths of extranucleosomal DNA in these experiments were as follows: no extranucleosomal DNA (0N0) (A), 20 bp DNA (0N20) (B), 33 bp (0N33) (C), 43 bp (0N43) (D), 53 bp (0N53) (E), 70 bp (0N70) (F), 109 bp (0N109) (G), and 168 bp (ON168) (H). The extent of nucleosomes shifted from the original position is indicated below each plot as a percentage of the initial nucleosomes that were moved. (I) The same remodeling reactions were stopped with γ-S-ATP and competitor DNA before loading on to a 5% PAGE gel (60:1, high resolution). The − and + symbols above each lane indicate whether INO80 was added, and the nucleosome nomenclature is the same as for panels A to H. (J) The extent of remodeling, based on changes in electrophoretic mobility indicated in panel I, was quantified and plotted relative to the length of extranucleosomal DNA.

FIG. 3.

INO80 does not remodel centrally positioned nucleosomes. (A to C) Site-directed mapping of nucleosomes with extranucleosomal DNA at both entry sites before and after INO80 remodeling was performed as described for Fig. 2A to H. In panels A and B, the nucleosomes start in the center of DNA with either 53 or 70 bp of extranucleosomal DNA, respectively, at both entry sites. In panel C nucleosomes are asymmetrically placed on DNA with 30 and 70 bp of extranucleosomal DNA at either entry site. (D and E) The samples shown in panels A to C were analyzed by gel shift as described for Fig. 2I.

FIG. 4.

The ATPase activity of INO80 is enhanced with increasing lengths of extranucleosomal DNA. (A) The rate of ATP hydrolysis by INO80 with 0N70 nucleosomes was measured with different concentrations of ATP (10 to 900 μM). Nucleosomes (33 nM) were prebound with INO80 (6.67 nM) for 15 min at 30°C before addition of ATP. The amount of ATP hydrolyzed at different time points was determined using thin-layer chromatography, and ATP hydrolyzed (in μM) versus time (in seconds) was plotted for the different ATP concentrations. (B) The K_m and K_cat values for INO80 were determined by plotting the rate of ATP hydrolyzed (in μM min⁻¹) versus the concentration of ATP, with nonlinear fitting to the Michaelis-Menten equation using GraphPad. (C) The rates of ATP hydrolysis of INO80 with DNA or nucleosomes were determined as for panel A with 80 μM ATP. The free DNA used was the same DNA used to reconstitute the 0N70 nucleosomes. (D) The effects of extranucleosomal DNA length on the rate of ATP hydrolysis by INO80 were examined using nucleosome core particle (0N0), or 53 or 109 bp of extranucleosomal DNA at only one entry site (0N53 and 0N109), or with 53 bp of extranucleosomal DNA at both entry sites (53N53). The assays were performed as described for panel A, except with a fixed concentration of 80 μM ATP. In all of these reactions, only PCR-generated DNA was used.

FIG. 5.

INO80 cannot mobilize dinucleosomes with only 6 or 30 bp of linker DNA, but it can with 79 bp. (A to C) Dinucleosomal substrates containing 6 (A), 30 (B), or 79 bp (C) of linker DNA between the two nucleosomes were remodeled with INO80. All three nucleosomes contained 9 and 30 bp of flanking DNA on either side. Dinucleosome positions were mapped as for Fig. 2, before and after INO80 remodeling, and are displayed in the same way. On the right side is the schematic view of the nucleosomes moved by INO80. The direction of nucleosome movement and the number of bp moved are indicated by arrows and numbers, respectively. The thickness of the arrow indicates the relative amount of nucleosomes moved. In panel C, the scale on the left side has been expanded to allow better visualization of the location of the remodeled nucleosome positions.

FIG. 6.

INO80 spaces di- and trinucleosomes ∼30 bp apart. Site-directed mapping of INO80 was carried out with di- (A) and trinucleosomal (B) substrates containing 50 bp of linker DNA between the nucleosomes, and the findings are displayed as described for Fig. 5. The dinucleosomes had 19 and 30 bp of flanking DNA, while the trinucleosomes had 5 and 30 bp of flanking DNA. The scale on the left side has been expanded for ease in mapping changes in nucleosome position.

FIG. 7.

INO80 does not have an H4 N-terminal histone tail requirement for binding or ATP hydrolysis. (A) The affinity of INO80 for nucleosomes with and without particular histone tails were measured by gel shift analysis. Increasing amounts of INO80 (5, 10, 20, and 40 nM) were bound with 33 nM nucleosomes for 30 min at 30°C and analyzed on a 4% native PAGE gel in 1× TE. (B) The rate of ATP hydrolysis was measured for the same nucleosome substrates as described in panel A, with INO80 (6.7 nM) prebound to nucleosomes (33 nM) for 15 min at 30°C. ATP was added to a final concentration of 80 μM and incubated for the indicated times. Reactions were stopped with SDS and EDTA as described in Materials and Methods and analyzed by thin-layer chromatography.

FIG. 8.

The N-terminal histone H2A tail has a negative effect on INO80 movement of nucleosomes. The rates of nucleosome movement by INO80 with nucleosomes missing one or more histone tails were measured by gel shift analysis as described in Materials and Methods. INO80 (40 nM) was prebound to nucleosomes (33 nM) for 15 min at 30°C, followed by addition of 800 μM ATP. After the desired incubation time at 30°C, reactions were stopped by the addition of γ-S-ATP and sonicated salmon sperm DNA to final concentrations of 1.5 mM and 0.5 mg/ml, respectively. Samples were analyzed on a 5% native PAGE gel, and the percentage of nucleosomes with an altered mobility relative to the original was determined. Nucleosomes were missing the N-terminal tails of H2A (A), H2B (B), H3 (C), H4 (D), H3 and H4 (E), or all histones (F). In the case of the experiment shown in panel F, the extent of nucleosome movement could only be determined using the restriction enzyme accessibility assay.

FIG. 9.

INO80 does not require any of the histone tails for efficiently mobilizing nucleosomes. (A) The location of the NotI restriction endonuclease site is shown in terms of the number of bp from the dyad axis of the 601 nucleosome before and after INO80 remodeling. End-positioned nucleosomes with 70 bp of extranucleosomal DNA (0N70) were used for the experiment. The locations of mobilized nucleosomes were previously determined by site-directed mapping. (B) Nucleosomes (33 nM) with (WT) or without (gAll) all the histone tails were remodeled with INO80 (40 nM) and ATP (800 μM). Samples were incubated with NotI and deproteinized, and DNA was analyzed on a 6% native PAGE gel. (C) The nucleosomes were mobilized by INO80 for different lengths of time ranging from 5 s to 21 min and analyzed as described for panel B.