Suppression of Oct4 by germ cell nuclear factor restricts pluripotency and promotes neural stem cell development in the early neural lineage

Abstract

The earliest murine neural stem cells are leukemia inhibitory factor (LIF)-dependent, primitive neural stem cells, which can be isolated from embryonic stem cells or early embryos. These primitive neural stem cells have the ability to differentiate to non-neural tissues and transition into FGF2-dependent, definitive neural stem cells between embryonic day 7.5 and 8.5 in vivo, accompanied by a decrease in non-neural competency. We found that Oct4 is expressed in LIF-dependent primitive neural stem cells and suppressed in FGF-dependent definitive neural stem cells. In mice lacking germ cell nuclear factor (GCNF), a transcriptional repressor of Oct4, generation of definitive neural stem cells was dramatically suppressed, accompanied by a sustained expression of Oct4 in the early neuroectoderm. Knockdown of Oct4 in GCNF(-/-) neural stem cells rescued the GCNF(-/-) phenotype. Overexpression of Oct4 blocked the differentiation of primitive to definitive neural stem cells, but did not induce the dedifferentiation of definitive to primitive neural stem cells. These results suggested that primitive neural stem cells develop into definitive neural stem cells by means of GCNF induced suppression of Oct4. The Oct4 promoter was methylated during the development from primitive neural stem cell to definitive neural stem cell, while these neural stem cells lose their pluripotency through a GCNF dependent mechanism. Thus, the suppression of Oct4 by GCNF is important for the transition from primitive to definitive neural stem cells and restriction of the non-neural competency in the early neural stem cell lineage.

Figure 1.

Oct4 is suppressed by germ cell nuclear factor (GCNF) in the early mouse neural tube. A, B, Oct4 expression in ES cells, pNS- and dNSC-derived neurospheres generated from ES cells were analyzed by RT-PCR (A) and immunoblotting (B). β-actin was used as an internal control. C, Transverse sections of E8.5 mouse spinal cord (top) or midbrain (bottom) were double-immunostained using Nestin and Oct4 antibodies. Scale bar, 250 μm. D, Expression of GCNF mRNA in E8.5 and E9.5 mouse neural tubes by in situ hybridization. E, Transverse sections from the E8.5 mouse forebrain were immunostained by using Oct4 antibody. pNS, pNSC-derived neurospheres; dNS, dNSC-derived neurospheres.

Figure 2.

Characterization of primitive and definitive neural stem cell derived neurospheres from GCNF^+/+ and GCNF^−/− embryos. A, The numbers of primary pNSC-derived neurosphere colonies from E6.5 and E7.5 GCNF^−/− embryos and littermate controls were not significantly different (t₍₉₎ = 0.27; p > 0.05 for E6.5, t₍₂₄₎ = 0.02; p > 0.05 for E7.5). B, The numbers of primary pNSC-derived neurosphere colonies generated from E8.5 were significantly increased in GCNF^−/− compared with GCNF^+/− and GCNF^+/+ embryos (t₍₁₄₎ = 2.2; p < 0.05 vs GCNF^+/+). C, The numbers of primary dNSC-derived neurosphere colonies were reduced significantly in E8.5 GCNF^−/− compared with GCNF^+/− and GCNF^+/+ embryos (t₍₈₎ = 7.4; p < 0.05 vs WT). D, Passageable secondary clonal pNSC-derived neurospheres were derived from GCNF^−/− but not GCNF^+/+ embryos. Growth factors were used as indicated. E, Quantitative RT-PCR analysis of Oct4 expression in pNSC-derived neurospheres and dNSC-derived neurospheres from GCNF^+/+ and GCNF^−/− E8.5 embryos. Oct4 is almost absent in GCNF^+/+ dNSC-derived neurospheres, while GCNF^−/− dNSC-derived neurospheres expressed 10-fold greater amounts of Oct4 (t₍₂₎ = 15.1; p < 0.05 vs WT). The amount of Oct4 was not significantly different between GCNF^+/+ and GCNF^−/− pNSC-derived neurospheres (t₍₂₎ = 1.31; p > 0.05). F, Qualitative RT-PCR analysis of the expression of GATA-4 and Brachyury-T. Total RNA isolated from the E9.5 embryonic head was used for the positive control. G, Differentiation of pNSC-derived neurospheres from E7.5 GCNF^+/+, GCNF^+/− and GCNF^−/− embryos. The ratios of neurons in total cells were not changed in GCNF^−/− cells. pNS; pNSC-derived neurospheres, dNS; dNSC-derived neurospheres

Figure 3.

Characterization of primitive and definitive neural stem cell derived neurospheres from GCNF^+/+ and GCNF^−/− ES cells. A, From GCNF^+/+ and GCNF^−/− ES cells, clonal pNSC-derived neurospheres were directly generated in serum-free floating culture with various growth factors. GCNF^−/− ES cells generated slightly but significantly increased numbers of neurospheres compared with GCNF^+/+ ES cells (t₍₁₄₎ = 3.60; p < 0.05). GCNF^−/− ES cells were still able to generate neurospheres even in the presence of SU5402 (t₍₁₄₎ = 18.9; p < 0.05). B, Quantitative RT-PCR analysis of Oct4 expression in undifferentiated ES cells, pNSC-derived neurospheres and secondary dNSC-derived neurospheres from GCNF^+/+ and GCNF^−/− ES cells. A two way ANOVA on genotypes and types of stem cells revealed a main effect of genotype F_(1,12) = 6.316, p < 0.05). C, Relative GCNF mRNA amounts in GCNF^+/+ ES cells, pNSC-derived neurospheres and dNSC-derived neurospheres were quantified using real-time quantitative RT-PCR. D–F, GCNF^+/+ and GCNF^−/− ES cells or pNSC-derived neurospheres colonies from ES cells were dissociated and plated in wells of 24-well dishes at a low-density (10 cells/μl;500 μl/well). Viable cells in the wells were stained by Hoechst 33342 and counted 4 and 24 h later. Oct4 expression was examined using an anti-Oct4 antibody. D, Proportions of Oct4 positive cells. The numbers of cells that express Oct4 are significantly increased in the GCNF^−/− compared with GCNF^+/+ cells both 4 and 24 h after plating (t₍₆₎ = −2.41; p < 0.05 in at 4 h, t₍₆₎ = −5.48; p < 0.05 at 24 h). E, Relative numbers of viable cells from ES cells (GCNF^+/+, 4 h = 100%). F, Relative numbers of viable cells from primary pNSC-derived neurospheres (GCNF^+/+, 4 h = 100%). In each of ES cells and dissociated primary pNSC-derived neurospheres, GCNF^−/− cells exhibit higher cell viability compared with GCNF^+/+ cells 24 h after plating (t₍₆₎ = −3.61; p < 0.05 for ES cells 24 h, t₍₆₎ = −3.61; p < 0.05 for pNSC 24 h). pNS; pNSC-derived neurospheres, dNS; dNSC-derived neurospheres

Figure 4.

Suppression of Oct4 rescues impaired neural stem cell development in GCNF ^−/− neural stem cells but the transition from primitive to definitive neural stem cells is not reversible by Oct4 overexpression. A, Structure of siRNA expressing lentiviral vectors. Dissociated primary pNSC-derived neurospheres from ES cells were cultured and infected with these lentiviruses in 96well plates (200 μl). B, C, All secondary spheres generated from dissociated primary spheres were confirmed as lentiviral infected by GFP fluorescence throughout the spheres, indicating infection of the cell of origin (B; C, bright field). D, Mechanically dissociated cells from single primary pNSC-derived neurospheres were immediately infected by lentiviral vectors which express GFP/GFP-siRNA (control), GFP/Nanog-siRNA, GFP/Oct4 siRNA, GFP/Oct4-m1 siRNA (one mutation in Oct4-siRNA) and GFP/Oct4-m6 (6 mutations in Oct4-siRNA), and cultured in serum-free medium with LIF or FGF. Double infections of Oct4-siRNA and Oct4-overexpressing virus were also tested. E, Quantitative RT-PCR analysis of Oct4 expression in LIF-dependent secondary pNSC-derived neurospheres (with/without lentiviral vectors) from GCNF^−/− ES or WT cells. Oct4 was suppressed to 43% of that seen in control GCNF^−/− spheres. F, The numbers of secondary colonies from dissociated single primary pNSC-derived neurospheres from GCNF^+/+ ES cells infected with control (GFP) or Oct4 lentivirus. Growth factors were added as indicated. Oct4 overexpression increased the numbers of secondary colonies seen with LIF but decreased the numbers seen with FGF. G, The numbers of primary colonies from dissociated forebrain germinal zone tissue from WT E14.5 CD1 embryos infected with control (GFP) or Oct4 lentivirus. Growth factors were added as indicated. No LIF-dependent colonies were observed. H, The numbers of secondary colonies from dissociated primary dNSC-derived neurospheres from WT E8.5 CD1 embryos infected with control (GFP) or Oct4 lentivirus. Growth factors were added as indicated.

Figure 5.

GCNF induces de novo methylation of the Oct4 promoter during the transition from primitive to definitive neural stem cells. A, The DNA methylation profile of 16 CpG sites located in the Oct4 proximal promoter from −469 to the ATG start codon in ES cells, primary pNSC-derived neurospheres, secondary or tertiary LIF- and FGF-dependent spheres and secondary dNSC-derived neurospheres from GCNF^+/+ and GCNF^−/− ES cells. The open circles represent unmethylated CpG dinucleotides, and the black circles represent methylated CpG sites. B, Ratio of methylated to unmethylated CpG sites at 16 CpGs in the Oct4 proximal promoter. Methylation in GCNF^+/+ dNSC-derived neurospheres (FGF-second) is significantly increased compared with LIF+FGF neurospheres from GCNF^+/+ ES cells (t₍₁₆₎ = −2.69; p < 0.05) or dNSC-derived neurospheres from GCNF^−/− cells (t₍₁₅₎ = 3.53; p < 0.05). C, DNA methylation profile in the Oct4 proximal promoter in dNSC-derived neurospheres from WT E8.5, E14 and adult brain. Ratios indicate average ratios of methylated to unmethylated sites at the 16 CpG sites ± SEM. D, Binding of GCNF to the Oct4 promoter in ES cells, pNSC-derived neurospheres and dNSC-derived neurospheres was detected by ChIP assays. Normal rabbit IgG was used as a control antibody.

Figure 6.

GCNF^−/− dNSCs generate chimeras with early mouse embryos. A, Strategy of the experiment. Dissociated primary neurospheres derived from GCNF^+/+ or ^−/− ES cells were infected with a lentivirus that express GFP. Clonal secondary neurosphere cultured with LIF and FGF were GFP fluorescent. Dissociated secondary spheres were cultured with FGF and without LIF to make small tertiary neurospheres for 2 or 3 d. Then, these neurospheres were cocultured with E2.5 host morula and cultured overnight to test aggregation. B–G, Representative pictures of aggregated embryos, which did not integrate (B–D) or did integrate (E–G) dNSCs into the inner cell mass of blastocysts after overnight incubation. dNSCs derived from ES cells were green fluorescent. H, The percentages of embryos with integrated dNSCs. Left two bars show the data with GCNF^+/+ or ^−/− tertiary neurospheres cultured for 2 d with FGF and without LIF (t₍₂₎ = 19.9; p < 0.05). Right two bars show the data with similar tertiary neurospheres cultured for 3 d with FGF and without LIF(t₍₂₎ = 5.72; p < 0.05). Both 2 and 3 d GCNF^−/− spheres showed an increased frequency of integration into host embryos compared with wild-type dNSC-derived colonies.

Figure 7.

GCNF^−/− dNSCs have ability to contribute to non-neural tissue. A, B, Aggregated embryos were transplanted into host animals and analyzed at E9.0. Bright field (A) and fluorescent (B) images of the same embryo show that GCNF^−/− dNSC contributed broadly in the recovered embryos. C–M, Sections of the heart (C–E), endoderm (F–I), and neural tube (J–M) immunostained with GFP (D, G, K), AFP (F), and Pax6 (J) or nuclear-stained by Hoechst 33342 (H, L) and merged images (E, I, M). GFP-positive donor cells were found in the cardiac wall (D), αFP-positive endoderm (I) and neural tube as Pax6 positive neural cells (M). nt, Neural tube.

Figure 8.

A model depicting the role of GCNF/Oct4 signaling in early NSC development. During neural differentiation from pluripotent ES or inner cell mass (ICM) cells, GCNF binds to the Oct4 promoter in pNSCs and induces Oct4 suppression that leads to the transition to dNSCs. GCNF induce de novo methylation of the Oct4 promoter (Gu et al., 2006). Notch signaling also may be involved in the pNSC/dNSC transition, but more likely has an important role in the self-renewal of FGF-dependent dNSCs and their transition into EGF-dependent dNSCs. Oct4 remains suppressed by methylation even after GCNF is no longer expressed. Further details are mentioned in the text. DR0; Direct repeat element with 0 base pair spacing between the half sites.