HDAC3 controls gap 2/mitosis progression in adult neural stem/progenitor cells by regulating CDK1 levels

Abstract

The maintenance of the resident adult neural stem/progenitor cell (NSPC) pool depends on the precise balance of proliferation, differentiation, and maintenance of the undifferentiated state. Identifying the mechanisms that regulate this balance in adult hippocampal NSPCs can provide insight into basic stem cell self-renewal principles important for tissue homeostasis and preventing tumor formation. Pharmacological inhibition of histone deacetylases (HDACs), a class of histone-modifying enzymes, have promising effects in cancer cells, yet the specific roles of individual HDACs in stem cell proliferation is unclear. Here using conditional KO (cKO) mice and in vitro cell culture, we show that histone deacetylase 3 (HDAC3) is required for the proliferation of adult NSPCs. Detailed cell cycle analysis of NSPCs from Hdac3 cKO mice reveals a defect in cell cycle progression through the gap 2/mitosis (G2/M) but not the S phase. Moreover, HDAC3 controls G2/M phase progression mainly through posttranslational stabilization of the G2/M cyclin-dependent kinase 1 (CDK1). These results demonstrate that HDAC3 plays a critical role in NSPC proliferation and suggest that strategies aimed at pharmacological modulation of HDAC3 may be beneficial for tissue regeneration and controlling tumor cell growth.

Keywords: acetylation; adult hippocampal neurogenesis; epigenetic; malignancy; ubiquitination.

Conflict of interest statement

The authors declare no conflict of interest.

Figures

Fig. 1.

Loss of HDAC3 results in decreased proliferation in vitro. (A) Immunocytochemical staining of BrdU followed by 1 h BrdU labeling in HCN cells infected with control (Cont) or HDAC3-shRNA-IRES-EGFP lentivirus for 3 d. Quantification of GFP+BrdU+ cells out of the total GFP+ cells is shown. (B) BrdU staining of 1 h BrdU labeling in HCN cells treated with either vehicle (DMSO) or HDAC3i for 1 d. Quantification of BrdU+ cells out of the total DAPI+ cells is shown. (C) HDAC3 WT or F/F neurospheres were infected with Ad-Cre-GFP virus for 2 d. Immunocytochemical staining of GFP and BrdU following 1 h BrdU labeling revealed a decrease in the percentage of BrdU+GFP+ double-positive cells out of the total GFP+ cells in HDAC3 F/F neurospheres. In A–C, *P < 0.05. (Scale bars: 200 µm in A and B; 100 µm in C.) All experiments were performed at least three times independently.

Fig. 2.

HDAC3 controls G2/M phase progression. (A and B) Flow cytometry analysis of HDAC3 WT and F/F neurospheres infected with Ad-Cre-GFP virus for 2 d before being harvested and fixed in 70% ethanol. Quantification of the percentage of cells in S and G2/M phase is shown. (C) Cell cycle analysis of HCN cells treated with HDAC3i or VPA for 1 d. The percentage of each cell cycle stage from one representative experiment is shown in the bar graph. (D) HCN cells were synchronized in the G1/G0 phase by adding BMP4 to the proliferation medium for 1 d before switching to proliferation medium containing either DMSO or HDAC3i. Cells were collected at 6 and 12 h and analyzed by flow cytometry. The percentage of each cell cycle stage from one representative experiment is shown in the bar graph. (E) HCN cells were synchronized in M phase by addition of nocodazole in proliferation medium for 16 h before switching to proliferation medium containing either DMSO or HDAC3i. Cells were collected at 6 and 12 h and analyzed by flow cytometry. The percentage of each cell cycle stage from one representative experiment is shown in the bar graph. In A and B, *P < 0.05. All cell cycle experiments were replicated at least three times independently.

Fig. 3.

G2/M phase-specific protein CDK1 is down-regulated after HDAC3 deletion. (A) Western blotting against G2/M phase cell cycle proteins in WT and HDAC3 F/F neurospheres infected with Ad-Cre-GFP virus for 2 d and in (B) HCN cells treated with DMSO, VPA, or HDAC3i for 1 d. Levels of CDK1, p-CDK1, and cyclin B1 were quantified. (C) qPCR of several cell cycle genes in WT and HDAC3 F/F neurospheres infected with Ad-Cre-GFP virus. These qPCR assays were repeated three times independently. (D and E) Western blotting against CDK1 in HCN cells treated with DMSO or HDAC3i together with protein synthesis inhibitor CHX for different lengths of time. (F) Western blot of lysates from HCN cells treated with vehicle (DMSO), MG132, HDAC3i, or HDAC3i+MG132. (G) Immunoprecipitation of CDK1 and Western blotting against ubiquitin in HCN cells treated with MG132 and DMSO or HDAC3i. Western blotting of CDK1 from cell lysate is shown (Lower). (H, Left) Western blotting against Flag in HCN cells coelectroporated with HA-CDK1, Flag-ubiquitin (Ub), and HDAC3. Western blotting against HDAC3 from cell lysate is shown (H, Lower Left). (H, Right) Immunoprecipitation of CDK1 and Western blotting against ubiquitin in HCN cells coelectroporated with HA-CDK1, Flag-Ub, and HDAC3 overexpression plasmids. Levels of ubiquitinated CDK1 were quantified. The asterisk denotes a nonspecific band. All Western blot experiments were replicated three times independently.

Fig. 4.

HDAC3 is broadly expressed in adult dentate gyrus. (A) Immunostaining of HDAC3, GFAP (a marker of RGLs), and Sox2 (a marker of RGLs and TAPs) of brain sections from P30 WT mice. HDAC3 is expressed in GFAP+Sox2+ RGL cells (arrow). (B) Immunostaining of HDAC3, DCX (a marker of immature neurons), and NeuN (a marker of mature neurons). HDAC3 is expressed in both DCX+NeuN− immature neurons (arrows) and DCX−NeuN+ mature neurons. (C) Immunostaining of HDAC3 and MCM2 (a marker of proliferation). Arrows indicate colocalization of HDAC3 and MCM2. (Scale bar: 50 µm.) One representative image from three independent WT mice is shown.

Fig. 5.

Loss of HDAC3 results in decreased proliferation in vivo. (A) Schematic of TAM injection and collection of brain tissue. (B) Immunostaining of YFP, Ki67, and GFAP in Hdac3 WT and cKO mice 10 dpt and quantification of YFP+Ki67+ proliferating cells (arrows) and YFP+Ki67+GFAP+ RGL cells. (C) Immunostaining of YFP, Ki67, and DCX in WT and cKO mice 10 dpt and quantification of YFP+Ki67+DCX− TAPs (arrows) at 10 and 30 dpt. (D) Immunostaining of YFP and NeuN in WT and cKO mice 60 dpt and quantification of YFP+NeuN+ cells (arrows) at 30, 60, and 90 dpt. In B–D, *P < 0.05. (Scale bars: 50 µm in B–D.) In B–D, at least six mice of each genotype were used. For the 90 dpt time point in D, three HDAC3 WT mice were used. *P < 0.05. DG, dentate gyrus.