Chromatin decondensation and nuclear reprogramming by nucleoplasmin

Abstract

Somatic cell nuclear cloning has repeatedly demonstrated striking reversibility of epigenetic regulation of cell differentiation. Upon injection into eggs, the donor nuclei exhibit global chromatin decondensation, which might contribute to reprogramming the nuclei by derepressing dormant genes. Decondensation of sperm chromatin in eggs is explained by the replacement of sperm-specific histone variants with egg-type histones by the egg protein nucleoplasmin (Npm). However, little is known about the mechanisms of chromatin decondensation in somatic nuclei that do not contain condensation-specific histone variants. Here we found that Npm could widely decondense chromatin in undifferentiated mouse cells without overt histone exchanges but with specific epigenetic modifications that are relevant to open chromatin structure. These modifications included nucleus-wide multiple histone H3 phosphorylation, acetylation of Lys 14 in histone H3, and release of heterochromatin proteins HP1beta and TIF1beta from the nuclei. The protein kinase inhibitor staurosporine inhibited chromatin decondensation and these epigenetic modifications with the exception of H3 acetylation, potentially linking these chromatin events. At the functional level, Npm pretreatment of mouse nuclei facilitated activation of four oocyte-specific genes from the nuclei injected into Xenopus laevis oocytes. Future molecular elucidation of chromatin decondensation by Npm will significantly contribute to our understanding of the plasticity of cell differentiation.

FIG. 1.

Decondensation of sperm and F9 chromatin by egg extract and Npm. (A) Decondensation of sperm chromatin and centromeric heterochromatin of F9 nuclei in egg extract. Sperm and F9 nuclei prior to incubation (0 h) and after incubation in buffer B and egg extract are shown. Incubation time was 2 h for F9 and 30 min for sperm, since swollen sperm nuclei became fragile after incubation for 30 min. ERS was included in the reactions indicated by +E. Sperm DNA was stained with Hoechst 33342, and F9 nuclei were stained with anti-tm-H3K9 antibody and Topro 3 (for DNA). F9 nuclei were also applied to FISH using the major satellite probe. Dotted lines in the FISH panels show nuclear contour. Bars, 10 μm for sperm and 3 μm for F9 in all images in the figure. (B) Central portions of the images marked by # in panel A were magnified threefold. (C) Silver-stained SDS-polyacrylamide gel electrophoresis gel loaded with Npm purified from eggs (eNpm) and from oocytes (oNpm) and rNpm. Hyperphosphorylation of eNpm and six-His tag on rNpm explain slower migration of these Npm compared with oNpm. (D) ERS-dependent centromeric heterochromatin decondensation in F9 nuclei by purified eNpm. F9 nuclei were incubated with 1 μM eNpm with ERS (+E) and apyrase (+A) for 2 h, followed by immunostaining and FISH. (E) Immunoblotting of egg extract immunodepleted with Npm antibody (Npm-depleted), with normal rabbit immunoglobulin G (Mock-depleted) and with anti-Npm antibody plus Npm add-back (Npm-depleted and add-back). The remaining Npm in these egg extracts was detected by anti-Npm antibody. HP1β was used as a loading control. (F) Immunofluorescence of F9 nuclei incubated in Npm-depleted egg extract followed by staining with tm-H3K9 antibody and Topro 3. (G) Immunofluorescence of F9 nuclei stained with tm-H3K9 antibody and Topro 3 after incubation with 1 μM oNpm and 5 μM rNpm in the presence of ERS for 2 h.

FIG. 2.

Decondensation of F9 chromatin by Npm and egg extract. (A) eNpm moderately decondensed sperm chromatin regardless of ERS but did not decondense centromeric heterochromatin in NIH 3T3 nuclei even with ERS. Both nuclei were incubated with 1 μM eNpm for 2 h and stained with Hoechst 33342 (sperm) and tm-H3K9 antibody plus Topro 3 (NIH 3T3). Bars, 3 μm for NIH 3T3 and 10 μm for sperm. (B) Electron microscopy of F9 nuclei after incubation in the described solutions. Bar, 1 μm. Arrowheads indicate nucleoli. (C) Npm facilitated sensitivity of F9 chromatin to the MNase digestion. F9 nuclei were incubated in each solution for 2 h and digested with MNase for the indicated time. Genomic DNA was isolated from the nuclei and applied to Southern hybridization. Ethidium bromide staining (top) and hybridization signal using the probes encoding minor satellite and 18S rRNA are shown. Arrowheads indicate mononucleosomes.

FIG. 3.

Histone was not released but phosphorylated by Npm. (A) Histone released from F9 nuclei into the supernatant (s) by various concentrations of NaCl and histone that remained in the nuclear pellet (p) were analyzed by Coomassie staining of the SDS gel (CBB) and immunoblotting with antibodies against histone H1 (H1) and H2B (H2B). The arrow indicates the height of each core histone confirmed by immunoblotting (not shown). (B) ³²P was primarily incorporated into what appeared to be histone H3, H2A, and rNpm (reacted with anti-Npm antibody; not shown) when F9 nuclei were incubated with Npm. Autoradiography (top) and Coomassie staining (CBB) of the same gel are shown. Decondensation of centromeric heterochromatin was monitored by tm-H3K9 antibody and scored as + (with obvious decondensation) or − (without decondensation) at the bottom (Dec). (C) Higher magnification of the gel images to compare the sizes of the proteins incorporating ³²P (Autorad) and core histones stained by the Coomassie dye (CBB). F9 nuclei were incubated with 5 μM rNpm and ERS. The gel used in panel C is different from that shown in panel B.

FIG. 4.

Histone H3 phosphorylation and acetylation by Npm. (A) Immunoblotting demonstrating histone H3 phosphorylation and acetylation in F9 nuclei by 5 μM rNpm. Total histone H3 was used as a loading control. The figures were prepared from more than one membrane in panels A and E because of the significantly diminished signal due to the repeated stripping of the membrane. (B) Immnofluorescence shows that H3S10 and H3S28 were globally phosphorylated when F9 nuclei were incubated with 5 μM rNpm. Bar, 3 μm. (C) Time course of histone H3 phosphorylation in F9 nuclei incubated with 5 μM rNpm. (D) Okadaic acid (OA) unmasked p-H3S10 in F9 nuclei incubated in egg extract. (E) Immunoblotting shows inhibition of histone H3 phosphorylation and chromatin decondensation by staurosporine. ac-H3K14 was not inhibited by staurosporine. (F) PKA phosphorylated histone H3 when incubated with F9 nuclei but could not induce centromeric heterochromatin decondensation without rNpm.

FIG. 5.

Npm releases heterochromatin proteins from F9 nuclei. (A) HP1β and TIFβ were lost from F9 nuclei by Npm as shown by immunoblotting. (B) Immunofluorescence confirmed loss of HP1β and TIF1β from F9 nuclei by rNpm. Bar, 5 μm. (C) Dose-dependent histone H3 modifications by various forms of rNpm. Each rNpm was used at 5, 1.5, and 0.5 μM (black triangles). (D) Summary of epigenetic modification in F9 nuclei incubated with various forms of rNpm. The scores (− to ++) are intended to show overall tendency, not exactly quantitative values.

FIG. 6.

Expression of oocyte-specific genes from Npm and PGA-pretreated F9 nuclei injected into Xenopus oocytes. RT-PCR of Msy2 and H1foo (both unspliced forms) comparing intact F9 cells, Xenopus oocytes, and mouse ovary and Xenopus oocytes injected with buffer B-pretreated F9 nuclei (buffer B), injected with eNpm-pretreated F9 (eNpm), injected with PGA-pretreated F9 (PGA), or injected with Npm-pretreated F9 plus 30 μg/ml α-amanitin (eNpm+α amanitin). In eNpm+RNase, RNA isolated from Xenopus oocytes injected with Npm-pretreated F9 was treated with RNase prior to RT-PCR. Sizes of the amplified bands are shown in parentheses. Npm and major satellite were used as controls. The post-oocyte injection period is shown in days. Note that unspliced H1foo was undetectable in mouse ovary (#), the positive control. Only spliced H1foo was detectable in that sample (not shown).

FIG. 7.

Model depicting the epigenetic modifications induced by Npm and gene activation. See Discussion for the details.