Downregulation of activating transcription factor 5 is required for differentiation of neural progenitor cells into astrocytes

Abstract

The mechanisms that regulate neural progenitor cell differentiation are primarily unknown. The transcription factor activating transcription factor 5 (ATF5) is expressed in neural progenitors of developing brain but is absent from mature astrocytes and neurons. Here, we demonstrate that ATF5 regulates the conversion of ventricular zone (VZ) and subventricular zone (SVZ) neural progenitors into astrocytes. Constitutive ATF5 expression maintains neural progenitor cell proliferation and blocks their in vitro and in vivo differentiation into astrocytes. Conversely, loss of ATF5 function promotes cell-cycle exit and allows astrocytic differentiation in vitro and in vivo. CNTF, a promoter of astrocytic differentiation, downregulates endogenous ATF5, whereas constitutively expressed ATF5 suppresses CNTF-promoted astrocyte genesis. Unexpectedly, constitutive ATF5 expression in neonatal SVZ cells both in vitro and in vivo causes them to acquire properties and anatomic distributions of VZ cells. These findings identify ATF5 as a key regulator of astrocyte formation and potentially of the VZ to SVZ transition.

Figure 1.

ATF5 is expressed by VZ and SVZ progenitor cells but not in nuclei of mature astrocytes. A, B, ATF5 is highly expressed in the VZ but not in neurons of developing cortex (CX) of E14 rat brain. Coronal sections of E14 rat telencephalon in area of lateral ventricle (dark area in center) were stained for ATF5 (A; red) and tubulin β (III) (B; green). C, D, ATF5-positive cells (D; red) in the VZ of E15 telencephalon do not express GFAP (C; green). Coronal section, lateral ventricle is dark area at left. E-G, ATF5 is expressed in the postnatal SVZ. Saggital sections of rat P9 brains were stained with anti-ATF5 (E; red) or with 4′, 6′-diamidino-2-phenylindole (DAPI) to reveal nuclei (F; blue). G, Merged images. Note the nuclear localization of ATF5 in clusters of cells in SVZ. H-J, ATF5 is either undetectable or in cytoplasm of cortical astrocytes in postnatal brain. Cortical sections of P9 rat brain were stained with either anti-GFAP (green), anti-ATF5 (red), or anti-DAPI (blue). J shows merged image. Note absence of ATF5 staining in nuclei of GFAP+ cells with morphology of mature astrocytes. The top arrow shows GFAP+ cell with no nuclear and apparent small amount of cytoplasmic ATF5. The middle arrow shows GFAP+ cell with cytoplasmic ATF5 staining. The bottom arrow shows GFAP+ cell without apparent ATF5 staining. Scale bars: (in A, C) A-D, 100 μm; (in G) E-G, 50 μm; (in H) H-J, 25 μm.

Figure 2.

ATF5 is expressed by cells in transition between neural progenitors and astrocytes, but not by mature astrocytes, in attached neurosphere cultures. Cultures of attached neurospheres were coimmunostained for ATF5 (red) and GFAP (green). A, Progenitor cells near the center of the culture show strong staining for ATF5. Farther from the periphery of the cultures, cells resembling immature astrocytes (B) show coexpression of ATF5 and GFAP, whereas cells resembling mature astroglia (C) express GFAP but not ATF5. Scale bar, 50 μm.

Figure 3.

ATF5 represses and Azip-ATF5 and ATF5 siRNA accelerate astrocyte differentiation in cultures of E15 telencephalic neural progenitor cells. A, Effects of ATF5 and Azip-ATF5 on morphology and GFAP expression of cultured E15 telencephalic cells. After 1 d in culture, cells were infected with retroviruses expressing eGFP, FLAG-ATF5 and eGFP, or eGFP-Azip-ATF5. On the 10th day after infection, the cultures were immunostained for expression of eGFP, GFAP, nestin, and ATF5. The panel shows examples of cells costained for eGFP and GFAP expression under each indicated condition. The inset in c shows a magnified version of bottom cell cluster. B, CNTF downregulates ATF5 expression, and ATF5 represses CNTF-promoted genesis of astrocytes. Experimental conditions were as in A except that CNTF was added to the medium as indicated. Values represent means ± SEM for three cultures in which at least 200 infected cells were evaluated per culture for expression of ATF5 (left) or GFAP (right). Student's t test analysis of infected cells for ATF5 expression (left) without versus with CNTF: eGFP alone, p < 0.03; GFAP expression (right) without versus with CNTF: eGFP alone, p < 0.02. There were no other significant differences. Comparable results were achieved in three independent experiments. C, ATF5 represses and Azip-ATF5 (Azip) promotes astroglial differentiation in both the absence and presence of CNTF. Experimental conditions were as in A except that CNTF was added to the medium as indicated. Values represent the mean ± SEM for three cultures in which at least 300 transfected cells were evaluated per culture. Student's t test analysis of infected cells: GFAP expression without versus with CNTF: eGFP alone, p < 0.005; ATF5 retrovirus, no significant difference; Azip retrovirus, p < 0.04; TUJ1 expression without versus with CNTF: eGFP alone, p < 0.04; ATF5 retrovirus, no significance; Azip retrovirus, p < 0.04. Left panel (no CNTF) for GFAP expression: eGFP alone versus ATF5 retrovirus, p < 0.003; eGFP alone versus Azip retrovirus, p < 0.003. Left panel (no CNTF) for TUJ1 expression: eGFP alone versus ATF5 retrovirus, p < 0.001; eGFP alone versus Azip retrovirus, p < 0.005. Right panel (plus CNTF) for GFAP expression: eGFP alone versus ATF5 retrovirus, p < 0.006; eGFP alone versus Azip retrovirus p < 0.06. Right panel (plus CNTF) for TUJ1 expression: eGFP alone versus ATF5 retrovirus, p < 0.002; eGFP alone versus Azip retrovirus, not significant. Comparable results were achieved in three independent experiments. D, Downregulation of ATF5 expression by ATF5 siRNA promotes astrocyte differentiation. Cultured E15 telencephalic cells were transiently transfected with pCMS-eGFP with or without ATF5 siRNA. Seven days after transfection, GFP+ cells were immunostained for expression of ATF5 or GFAP as indicated. Values represent the mean ± SEM for six cultures in which at least 200 transfected cells were evaluated per culture. Comparable results were achieved in 10 independent experiments. Student's t test analysis of transfected cells: GFAP expression for eGFP alone versus GFP plus ATF5 siRNA, p < 0.03; ATF5 expression for eGFP alone versus eGFP plus ATF5 siRNA, p < 0.007. E, ATF5 increases the number of nestin+ cells and of dividing cells in cultures of E15 telencephalic neural progenitors. Experimental conditions were as in A with infected cells evaluated for expression of nestin or Ki67 antigen. Values represent the mean ± SEM for three cultures in which at least 200 transfected cells were evaluated per culture. Comparable results were achieved in three independent experiments. Student's t test analysis of infected cells, nestin expression: eGFP alone versus ATF5 retrovirus, p < 0.001; eGFP versus Azip retrovirus, p < 0.01; Ki67 expression: eGFP versus ATF5, p = 0.002; eGFP versus Azip, not significant. F, The nestin+ cells generated in cultures of E15 telencephalic neural progenitors by ATF5 do not coexpress GFAP. Experimental conditions were as in A with infected cells evaluated for expression of GFAP and/or nestin. Values represent the mean ± SEM for three cultures in which at least 200 transfected cells were evaluated per culture. Student's t test analysis of infected cells, GFAP expression: eGFP alone versus ATF5 retrovirus, p < 0.02; eGFP alone versus Azip retrovirus, p < 0.02; nestin expression: eGFP alone versus ATF5 retrovirus, p < 0.0001; eGFP alone versus Azip retrovirus, not significant; colocalized GFAP and nestin: eGFP versus ATF5 retrovirus, p < 0.005; eGFP versus Azip retrovirus, not significant.

Figure 4.

ATF5 represses and Azip-ATF5 promotes neuronal and astrocytic differentiation of PSA-NCAM+ SVZ cells. A, Morphology of PSA-NCAM+ SVZ cells infected with retroviral constructs expressing eGFP, FLAG-ATF5 and eGFP, or eGFP-Azip-ATF5. Immediately after plating, the cells were cocultured (separated by a 0.4 μm membrane) with virus-producing GP2 293 cells as described in Materials and Methods. Infected SVZ cells were stained 5 d later for expression of eGFP. Scale bar, 100 μm. B, ATF5 represses and Azip-ATF5 promotes neuronal differentiation of cultured PSA-NCAM+ SVZ cells. Experimental conditions were as in A. Infected eGFP+ cells were evaluated for expression of tubulin β (III). Values represent the mean ± SEM for three cultures in which at least 100 transfected cells were evaluated per culture. Student's t test analysis of infected cells: eGFP alone versus ATF5 retrovirus, p < 0.006; eGFP alone versus Azip, p < 0.0001. C, ATF5 promotes and Azip-ATF5 represses proliferation of cultured PSA-NCAM+ SVZ cells. Experimental conditions were as in A except that cultures were exposed to BrdU for 5 h before fixing and were evaluated for BrdU incorporation. Values represent the mean ± SEM for three cultures in which at least 100 transfected cells were evaluated per culture. Student's t test analysis of infected cells: eGFP versus ATF5 retrovirus, p < 0.0001; eGFP versus Azip retrovirus, p < 0.0002. D, Regulation of GFAP and nestin expression in cultured PSA-NCAM+ SVZ cells by ATF5 and Azip-ATF5. Experimental conditions were as in A. Values represent the mean ± range for two cultures in which at least 100 transfected cells were evaluated per culture. E, ATF5 promotes expression of β-catenin in cultured PSA-NCAM+ SVZ cells. Experimental conditions were as in A. Values represent the mean ± range for two cultures in which at least 100 transfected cells were evaluated per culture. Data at far right show percentage of infected GFAP+ cells that also express β-catenin.

Figure 5.

Constitutive expression of ATF5 and Azip-ATF5 affects fate of SVZ neural progenitor cells in vivo. SVZ neural progenitor cells in vivo. A, Quantification of the effects of ATF5 and Azip-ATF5 on the migratory behavior of SVZ neural progenitor cells. The SVZ of P1 rats was stereotactically injected with retrovirus expressing eGFP (control), eGFP plus ATF5, or eGFP-Azip-ATF5. One week later, the animals were killed, and brain sections were immunostained for GFP to reveal the locations of infected cells. The numbers of infected neurons assessed per brain ranged from 112 to 156. Comparable results were found in two additional brains. CC, Corpus callosum; RMS, rostral migratory stream; OB, olfactory bulb. B, ATF5 and Azip-ATF5 affect the expression of markers by SVZ neural progenitor cells in vivo. Experimental conditions are as in A. Data represent mean ± SEM for data from three brains; at least 140 cells were assessed per brain.

Figure 6.

Constitutive expression of ATF5 and Azip-ATF5 in SVZ neural progenitor cells affects their morphology, localization, proliferation, and expression of β-catenin. Experimental conditions are as in Figure 5 except for M and N, in which animals were killed and brains examined 3.5 months after injection with retrovirus expressing eGFP alone (M; showing mature glial cells in the white matter) or eGFP and Flag-tagged ATF5 (N; showing a hyperplastic mass at the ventricular surface). The diagrams to the right of each row depict the treatment and brain areas at which the photographs were taken. Cells were stained as indicated for expression of GFP (green) and β-catenin (red). The arrows in E-G show positions of the same cells, each with the characteristic morphology of VZ neural progenitors. Scale bars: A, 25 μm; D, 100 μm; (in N) M, N, 40 μm. CC, Corpus callosum; V, ventricle.

Figure 7.

Constitutive expression of ATF5 and Azip-ATF5 in SVZ neural progenitor cells affects their morphology, localization, and expression of nestin and GFAP. Experimental conditions are as in Figure 5. Diagrams below each row depict the treatment and brain areas at which the photographs were taken. Cells were stained as indicated for expression of GFP, nestin, or GFAP. The third horizontal row shows merged images; the fourth horizontal row shows 4× magnifications to highlight cell morphology. Cells with constitutive ATF5 expression display a simple morphology in which a single long process (arrows) extends from a round cell body. In contrast, cells expressing Azip-ATF5 have a complex morphology in which several long, thick processes extend from irregularly shaped cell bodies (arrowheads). Scale bars: (in D) D, H, L, P, T, X, 10 μm; (in A) A-C, E-G, I-K, M-O, Q-S, U-W, 25 μm. CC, Corpus callosum.