Esophageal cancer-related gene 4 is a secreted inducer of cell senescence expressed by aged CNS precursor cells

Abstract

Mammalian aging is thought to be partially caused by the diminished capacity of stem/precursor cells to undergo self-renewing divisions. Although many cell-cycle regulators are involved in this process, it is unknown to what extent cell senescence, first identified as irreversible growth arrest in vitro, contributes to the aging process. Here, using a serum-induced mouse oligodendrocyte precursor cell (mOPC) senescence model, we identified esophageal cancer-related gene 4 (Ecrg4) as a senescence inducer with implications for the senescence-like state of postmitotic cells in the aging brain. Although mOPCs could proliferate indefinitely when cultured using the appropriate medium (OPC medium), they became senescent in the presence of serum and maintained their senescent phenotype even when the serum was subsequently replaced by OPC medium. We show that Ecrg4 was up-regulated in the senescent OPCs, its overexpression in OPCs induced senescence by accelerating the proteasome-dependent degradation of cyclins D1 and D3, and that its knockdown by a specific short hairpin RNA prevented these phenotypes. We also show that senescent OPCs secreted Ecrg4 and that recombinant Ecrg4 induced OPC senescence in culture. Moreover, increased Ecrg4 expression was observed in the OPCs and neural precursor cells in the aged mouse brain; this was accompanied by the expression of senescence-associated beta-galactosidase activity, indicating the cells' entrance into senescence. These results suggest that Ecrg4 is a factor linking neural-cell senescence and aging.

Fig. 1.

A high concentration of FCS induces mOPC senescence. (A) Morphology of mOPCs cultured in low FCS, high FCS, and reversion medium. (Scale bar, 10 μm.) (B–D) Data from cells cultured as in A. (B) SA-β-gal staining of mOPCs. (Scale bar, 100 μm.) (Insets) DAPI staining (white) of the same field. (C) Percentage of BrdU+ mOPCs shown as the mean ± SD of three independent experiments. (D) Western blot analysis for phosphorylated Rb (p-Rb). GAPDH is a loading control.

Fig. 2.

Ecrg4 induces senescence accompanied by the proteasomal degradation of cyclins D1 and D3. (A–C) Ecrg4 expression in mOPCs cultured in low FCS, high FCS, and reversion medium and analyzed by RT-PCR (A), Western blotting (B), and immunolabeling (C). (Scale bar, 10 μm.) Arrowheads in B indicate bands with probable protein modifications. (D and E) SA-β-gal staining (green) of CG4 cells overexpressing either Ecrg4 (D) or Ecrg4-shRNA (E). (Scale bar, 50 μm.) (F) Percentage of BrdU+ CG4 cells overexpressing Ecrg4 (Left) or Ecrg4-shRNA (Right). (G) Cell-cycle analysis of CG4 cells overexpressing Ecrg4 (Left) or Ecrg4-shRNA (Right). (H) Western blot analysis of cell-cycle regulatory proteins in CG4 cells overexpressing Ecrg4 (Left) or Ecrg4-shRNA (Right). (I) Levels of cyclin D1 and D3 in Ecrg4-overexpressing CG4 cells with or without lactacystin. (J) RT-PCR analysis of cell-cycle regulatory proteins in Ecrg4-expressing CG4 cells. *P < 0.05 and **P < 0.01 compared with control. Values denote the mean ± SD of three independent experiments.

Fig. 3.

Ecrg4 is secreted, and recombinant Ecrg4 protein induces senescence in mOPCs. (A) Immunoblot analysis of the culture medium from mOPCs grown in all three conditions and probed for Ecrg4. (B–D) Experiments were performed with mOPCs treated with rEcrg4 or a control cell extract. (B) SA-β-gal staining. (Scale bar, 100 μm.) (C) Percentage of BrdU+ mOPCs. (D) Immunoblot analysis of p-Rb, cyclin D1, and cyclin D3. (E) Costaining of SA-β-gal, O4, and GFAP in either rEcrg4- or FCS-treated senescent OPCs. (Scale bar, 50 μm.) **P < 0.01 compared with control. ns, not significant. Values denote the mean ± SD of three independent experiments.

Fig. 4.

Increased expression of Ecrg4 in senescent NPCs and OPCs in the aged mouse brain. (A) Schematic diagram showing the location of Ecrg4+ cells in the brain of aged (15–21 months old) mice. Green, NPCs and OPCs; yellow, neurons. (B–F) Comparison of the labeling for various markers in the brain of young (2 months old) and aged mice. (B) Immunolabeling of the CC for Ecrg4 and Olig1 (Left) or Ecrg4 and NG2 (Right). (Scale bar, 30 μm.) (C) SA-β-gal staining (blue) of the CC that is counterstained with Nuclear Fast Red (pink). (Scale bar, 30 μm.) (D) Immunolabeling of the dentate gyrus for Ecrg4 and Sox2 (Left) or Ecrg4 and Musashi1 (Right). (Scale bar, 50 μm.) (E) SA-β-gal staining (blue) of the dentate gyrus as in C. (Scale bar, 100 μm.) (F) Colocalization of SA-β-gal (green) and Ecrg4 (red) in the aged dentate gyrus. (Scale bar, 50 μm.) Arrowheads point to Ecrg4+SA-β-gal+ cells. All nuclei were counterstained with DAPI (blue).