Nap1 regulates proper CENP-B binding to nucleosomes

Abstract

CENP-B is a widely conserved centromeric satellite DNA-binding protein, which specifically binds to a 17-bp DNA sequence known as the CENP-B box. CENP-B functions positively in the de novo assembly of centromeric nucleosomes, containing the centromere-specific histone H3 variant, CENP-A. At the same time, CENP-B also prevents undesired assembly of the CENP-A nucleosome through heterochromatin formation on satellite DNA integrated into ectopic sites. Therefore, improper CENP-B binding to chromosomes could be harmful. However, no CENP-B eviction mechanism has yet been reported. In the present study, we found that human Nap1, an acidic histone chaperone, inhibited the non-specific binding of CENP-B to nucleosomes and apparently stimulated CENP-B binding to its cognate CENP-B box DNA in nucleosomes. In human cells, the CENP-B eviction activity of Nap1 was confirmed in model experiments, in which the CENP-B binding to a human artificial chromosome or an ectopic chromosome locus bearing CENP-B boxes was significantly decreased when Nap1 was tethered near the CENP-B box sequence. In contrast, another acidic histone chaperone, sNASP, did not promote CENP-B eviction in vitro and in vivo and did not stimulate specific CENP-B binding to CENP-A nucleosomes in vitro. We therefore propose a novel mechanism of CENP-B regulation by Nap1.

Figure 1.

Specific and non-specific CENP-B DBD binding to the CENP-A and H3 nucleosomes. (A) Schematic representations of 192-bp α-satellite DNAs. The box and the dashed circle indicate the location of the CENP-B box and the nucleosome positioning, respectively, as revealed by the previous MNase mapping (42). (B) CENP-B DBD binding to the CENP-A nucleosomes with or without the CENP-B box DNA. The CENP-A nucleosomes (140 nM) with the CENP-B box sequence (Cb+) (lanes 2–6) or without the CENP-B box sequence (Cb−) (lanes 8–12) were incubated with the CENP-B DBD for 20 min at 37°C. CENP-B DBD concentrations were 0 µM (lanes 2 and 8), 0.7 µM (lanes 3 and 9), 1.4 µM (lanes 4 and 10), 2.1 µM (lanes 5 and 11) and 2.8 µM (lanes 6 and 12). The samples were analyzed by non-denaturing 5% polyacrylamide gel electrophoresis, followed by ethidium bromide staining. Lanes 1 and 7 indicate the 192-bp naked DNA. (C) CENP-B DBD binding to the H3 nucleosomes with or without the CENP-B box DNA. The H3 nucleosomes (140 nM) with the CENP-B box sequence (Cb+) (lanes 2–6) or without the CENP-B box sequence (Cb−) (lanes 8–12) were incubated with the CENP-B DBD for 20 min at 37°C. CENP-B DBD concentrations were 0 µM (lanes 2 and 8), 0.7 µM (lanes 3 and 9), 1.4 µM (lanes 4 and 10), 2.1 µM (lanes 5 and 11) and 2.8 µM (lanes 6 and 12). The samples were analyzed by non-denaturing 5% polyacrylamide gel electrophoresis, followed by ethidium bromide staining. Lanes 1 and 7 indicate the 192-bp naked DNA.

Figure 2.

Nap1 stimulates specific binding of the CENP-B DBD to the CENP-A and H3 nucleosomes. (A) CENP-B DBD binding to the CENP-A nucleosomes with the CENP-B box DNA, in the presence of Nap1 or sNASP. The CENP-A nucleosomes (Cb+) (140 nM) were incubated with the CENP-B DBD in the presence of Nap1 (2.8 µM, lanes 7–11) or sNASP (2.8 µM, lanes 18–22) for 20 min at 37°C. Lanes 2–6 and 13–17 indicate control experiments in the absence of Nap1 or sNASP. CENP-B DBD concentrations were 0 µM (lanes 2, 7, 13 and 18), 0.7 µM (lanes 3, 8, 14 and 19), 1.4 µM (lanes 4, 9, 15 and 20), 2.1 µM (lanes 5, 10, 16 and 21)s and 2.8 µM (lanes 6, 11,17 and 22). The samples were analyzed by non-denaturing 5% polyacrylamide gel electrophoresis, followed by ethidium bromide staining. Lanes 1 and 12 indicate the 192-bp naked DNA. (B) Graphic representation of the specific complex formation. The relative band intensities of the specific CENP-B DBD-CENP-A nucleosome complexes were plotted with the standard deviations (n = 3). (C) CENP-B DBD binding to the H3 nucleosomes with the CENP-B box DNA, in the presence of Nap1 or sNASP. The H3 nucleosomes (Cb+) (140 nM) were incubated with the CENP-B DBD in the presence of Nap1 (2.8 µM, lanes 7–11) or sNASP (2.8 µM, lanes 18–22) for 20 min at 37°C. Lanes 2–6 and 13–17 indicate control experiments in the absence of Nap1 or sNASP. CENP-B DBD concentrations were 0 µM (lanes 2, 7, 13 and 18), 0.7 µM (lanes 3, 8, 14 and 19), 1.4 µM (lanes 4, 9, 15 and 20), 2.1 µM (lanes 5, 10, 16 and 21) and 2.8 µM (lanes 6, 11,17 and 22). The samples were analyzed by non-denaturing 5% polyacrylamide gel electrophoresis, followed by ethidium bromide staining. Lanes 1 and 12 indicate the 192-bp naked DNA. (D) Graphic representation of the specific complex formation. The relative band intensities of the specific CENP-B DBD-H3 nucleosome complexes were plotted with the standard deviations (n = 3).

Figure 3.

Nap1 dissociates non-specifically bound CENP-B DBD from the CENP-A nucleosome. (A) CENP-B DBD binding to the CENP-A nucleosomes without the CENP-B box DNA, in the presence of Nap1 or sNASP. The CENP-A nucleosomes (Cb−) (140 nM) were incubated with the CENP-B DBD in the presence of Nap1 (2.8 µM, lanes 7–11) or sNASP (2.8 µM, lanes 18–22) for 20 min at 37°C. Lanes 2–6 and 13–17 indicate control experiments in the absence of Nap1 or sNASP. CENP-B DBD concentrations were 0 µM (lanes 2, 7, 13 and 18), 0.7 µM (lanes 3, 8, 14 and 19), 1.4 µM (lanes 4, 9, 15 and 20), 2.1 µM (lanes 5, 10, 16 and 21) and 2.8 µM (lanes 6, 11,17 and 22). The samples were analyzed by non-denaturing 5% polyacrylamide gel electrophoresis, followed by ethidium bromide staining. Lanes 1 and 12 indicate the 192-bp naked DNA. (B) Graphic representation of the experiments shown in panel A. The relative band intensities of the CENP-A nucleosomes were plotted with the standard deviations (n = 3). (C) CENP-B DBD binding to the H3 nucleosomes without the CENP-B box DNA, in the presence of Nap1 or sNASP. The H3 nucleosomes (Cb−) (140 nM) were incubated with the CENP-B DBD in the presence of Nap1 (2.8 µM, lanes 7–11) or sNASP (2.8 µM, lanes 18–22) for 20 min at 37°C. Lanes 2–6 and 13–17 indicate control experiments in the absence of Nap1 or sNASP. CENP-B DBD concentrations were 0 µM (lanes 2, 7, 13 and 18), 0.7 µM (lanes 3, 8, 14 and 19), 1.4 µM (lanes 4, 9, 15 and 20), 2.1 µM (lanes 5, 10, 16 and 21) and 2.8 µM (lanes 6, 11,17 and 22). The samples were analyzed by non-denaturing 5% polyacrylamide gel electrophoresis, followed by ethidium bromide staining. Lanes 1 and 12 indicate the 192-bp naked DNA. (D) Graphic representation of the experiments shown in panel C. The relative band intensities of the H3 nucleosomes were plotted with the standard deviations (n = 3).

Figure 4.

The CENP-B and CENP-A assembly levels are reduced by the tethering of Nap1, but not sNASP, to the centromere. (A) A schematic drawing of the tetR-fusion/alphoid^tetO array tethering system on the HAC. A HeLa cell line containing a stable alphoid^tetO HAC (61, HeLa-HAC-2-4) was transfected with the tetR-EYFP-fusion protein expressing plasmids (tetR-EYFP alone, -Nap1 or -sNASP). Immunofluorescence analysis was performed 2 days after transfection. (B) Cells were co-stained with antibodies against CENP-B and CENP-A. The HAC centromere signals were determined by the EYFP signals (arrowhead), and DNA was visualized with Qnuclear^™ Deep Red Stain (Invitrogen). The scale bars represent 10 µm (light gray). (C and D) Immunofluorescence signals of CENP-B (C) and CENP-A (D) on the HAC centromere, against those of all centromeres on host chromosomes within the same single nucleus, were quantified and plotted as relative arbitrary fluorescence units (AFU). Solid lines indicate the median. Asterisks indicate significant differences, with P < 0.001 (Mann–Whitney test).

Figure 5.

Nap1 tethering reduces the CENP-B assembly level on non-centromeric alphoid chromatin in vivo. (A) A schematic drawing of the alphoid^tetO ectopic integration site. (B) An alphoid^tetO ectopic integration cell line (61, HeLa-Int-03) was transfected with a set of tetR-EYFP-fusion expressing plasmids. The CENP-B signals on the ectopic site were apparently reduced by Nap1 tethering, but not by sNASP tethering. Arrowheads indicate alphoid^tetO DNA integration sites. The scale bars represent 10 µm.

Figure 6.

Turnover of tetR-EYFP-fusion proteins on the tetO site. HeLa-Int-03 cells were transfected with a set of tetR-EYFP-fusion expressing plasmids. A focus of the tetR-EYFP-fusion proteins on the tetO array was bleached, and the recovery rate was measured. (A) FRAP examples. A 3-µm diameter area containing the tetO array was bleached, and the fluorescence intensity in the bleached area was measured. Arrowheads indicate tetO arrays. (B) FRAP results. The relative intensity of the bleached area for the indicated tetR-EYFP-fusion protein was plotted (averages of 10–15 cells with the standard deviations), and each halftime of recovery t_1/2, obtained by fitting the curve to single exponential association kinetics, is shown. (C) FRAP results for EYFP-alone, EYFP-NAP1 and EYFP-sNASP. The same diameter in the nucleus was bleached, and the recovery was measured (averages of 15–20 cells). Note that the x-axis unit is ms. The FRAP assay was performed 1 day after transfection, using a confocal microscope (FV-1000; Olympus) with a PlanSApo 60 × (NA = 1.35) oil-immersion lens. For panel (A) and (B), images (0.5 s/frame) were collected at 10 s intervals, and a 3-µm diameter spot was bleached (100% 488-nm laser transmission; two iterations) after three images. For panel (C), 10 images were collected (65 ms/frame) before bleaching, and 90 images were further collected. ImageJ (National Institutes of Health) was used for intensity measurements and fitting analysis.

Figure 7.

Nap1 decreases non-specific CENP-B binding in cells. (A) Schematic diagram of the ChIP analysis. HeLa cells were transfected with alphoid^tetO DNA plus or minus each expression plasmid. The ChIP analysis was performed using an anti-CENP-B C-terminus antibody [5E6C1 (66)]. (B) CENP-B specifically binds to wild-type canonical CENP-B box on the alphoid^tetO DNA. HeLa cells were transfected with alphoid^tetO containing wild-type or mutant CENP-B boxes (66). The ChIP analysis was performed on the introduced alphoid^tetO DNAs. The rate of recovery of immunoprecipitates using the anti-CENP-B antibody was normalized to the control IP without the antibody (beads). (C) CENP-B overexpression increased its non-specific binding to the introduced alphoid^tetO-containing mutant CENP-B boxes and its specific binding to endogenous alphoid DNA containing canonical CENP-B boxes on the chromosome 21 centromere (21-I alphoid). The ChIP analysis was performed by co-transfecting the alphoid^tetO DNA-containing mutant CENP-B boxes and the Halo-CENP-B expression plasmid (0, 500 and 1500 ng). Real-time PCR analysis was performed on the introduced alphoid^tetO DNA and 21-I alphoid DNA. (D) Non-specific binding of CENP-B to the introduced alphoid^tetO-containing mutant CENP-B boxes was decreased by Nap1 overexpression. The ChIP analysis was performed by co-transfecting the alphoid^tetO DNA-containing mutant CENP-B boxes and the Halo-CENP-B expression plasmid (500 ng) with the EYFP, EYFP-Nap1 or EYFP-sNASP expression plasmid. Real-time PCR analysis was performed as in panel C. Error bars, s.d. (n = 3). P-values (t-test) are indicated in the figure. (E) Expression of Halo-CENP-B and EYFP-fusion proteins in HeLa cells. HeLa cells co-transfected with each plasmid set were analyzed by western blotting, using antibodies against CENP-B, GAPDH and GFP. Asterisks indicate non-specific signals with the anti-GFP antibody.