Gfap-positive radial glial cells are an essential progenitor population for later-born neurons and glia in the zebrafish spinal cord

Abstract

Radial glial cells are presumptive neural stem cells (NSCs) in the developing nervous system. The direct requirement of radial glia for the generation of a diverse array of neuronal and glial subtypes, however, has not been tested. We employed two novel transgenic zebrafish lines and endogenous markers of NSCs and radial glia to show for the first time that radial glia are essential for neurogenesis during development. By using the gfap promoter to drive expression of nuclear localized mCherry we discerned two distinct radial glial-derived cell types: a major nestin+/Sox2+ subtype with strong gfap promoter activity and a minor Sox2+ subtype lacking this activity. Fate mapping studies in this line indicate that gfap+ radial glia generate later-born CoSA interneurons, secondary motorneurons, and oligodendroglia. In another transgenic line using the gfap promoter-driven expression of the nitroreductase enzyme, we induced cell autonomous ablation of gfap+ radial glia and observed a reduction in their specific derived lineages, but not Blbp+ and Sox2+/gfap-negative NSCs, which were retained and expanded at later larval stages. Moreover, we provide evidence supporting classical roles of radial glial in axon patterning, blood-brain barrier formation, and locomotion. Our results suggest that gfap+ radial glia represent the major NSC during late neurogenesis for specific lineages, and possess diverse roles to sustain the structure and function of the spinal cord. These new tools will both corroborate the predicted roles of astroglia and reveal novel roles related to development, physiology, and regeneration in the vertebrate nervous system. GLIA 2016;64:1170-1189.

Keywords: genetic ablation; neural stem cell; neurogenesis; radial glia; spinal cord; zebrafish.

FIGURE 1

The gfap:nls-mCherry transgene labels radial glial nuclei. (A–I) Maximum intensity projections showing cross sections of 24, 48, and 72 hpf gfap:nls-mCherry embryos, immunolabeled with anti-Gfap (white) for radial glia and Hoechst (blue) for nuclei. (A′, D′, G′) Magnified, single optical slices of individual cells in the yellow boxed area in A, D, G. Colored asterisks match nucleus with radial body. (J) Transverse section of a 96 hpf spinal cord from a gfap:nls-mCherry embryo colabeled for anti-Blbp (green), anti-Gfap (white), and nuclei (Hoechst, blue). (K–M) Higher magnification of boxed area in K with separated fluorescent channels (K, all channels; L, mCherry, Gfap, Blbp; M, mCherry and Gfap). Scale bar in G = 5 μm; scale bar in I = 10 μm.

FIGURE 2

gfap+ radial glia coexpress neural stem markers. (A–C) Cross sections of 24, 48, and 72 hpf double transgenic gfap:nls-mCherry;- 3.9nestin:GFP embryos labeled with anti-Gfap (white) and anti-GFP to visualize nestin+, gfap+ radial glial populations. (D–F) Cross sections of 24, 48, and 72 hpf gfap:nls-mCherry embryos immunolabeled with anti-Sox2 (green) and anti-Gfap (white). Scale bar = 10 μm. (D′–E″) Enlarged views of yellow-boxed areas comparing ventral Sox2+ nuclei in panels D′, E′, F′ with ventral nls-mCherry+ nuclei in panels D″, E″, F″. Sox2+, mCherry negative nuclei were outlined for clarity. Scale bar = 5 μm (G) Quantification of the total number of Sox2+ cells present at all time points. (H) Quantification of the percentage of total Sox2+ nuclei that are nls-mCherry+ (red) throughout neural development. Numbers above each population represent the total number of cells assayed per time point. (I) Cross section of a gfap:nls-mCherry embryo treated with BrdU at 24 hpf, fixed at 48 hpf, and labeled for anti-Sox2 (green) and anti-BrdU (cyan). Scale bar = 10μm. (I′-I‴) Enlarged view of yellow-boxed area in panel I, with ventral Sox2+, nls-mCherry negative cells outlined for clarity. Asterisks denote ventral Sox2+/nls-mCherry- cells that are BrdU+. Scale bar = 5μm. (J) Quantification of the percentage of Sox2+ populations positive for BrdU at all time points. Sox2+/nls-mCherry+ cells are shown in red, and Sox2+/nls-mCherry negative cells are depicted in white.

FIGURE 3

Fate mapping of gfap-expressing radial glia in the spinal cord. (A) Cross section of a 30 hpf gfap:nls-mCherry embryo immunolabeled with anti-Islet-1 (ISL), showing nls-mCherry is absent in RB sensory neurons but present in a subpopulation of MNs (asterisks). (B) Cross section of a 48 hpf gfap:nls-mCherry;gata2:eGFP embryo showing secondary motorneurons (SMN) expressing GFP. Cells absent for nls-mCherry are denoted (arrowheads). Scale bar = 10 μm. (C) Quantification for the percentage of sensory (RB), total ISL+ MNs, primary motorneurons (PMN), and secondary motorneurons (SMN) retaining nls-mCherry+ nuclei (red). (D) Cross section of a 30 hpf gfap:nls-mCherry embryo immunolabeled with anti-GABA to identify DoLA (D), CoSA (C), VeLD (V), and KA (K) interneurons. Subset Z-stacks and maximum intensity projections were generated for each neuronal population for clarity. Scale bar = 10 μm. (E) Quantification of the percentage of each interneuron population expressing nls-mCherry+ nuclei (red) at 30 hpf. (F, G) Cross sections of 24 and 72 hpf double transgenic gfap:nls-mcherry;olig2:eGFP embryos showing ventral Olig2+ populations as well as dorsal OPCs (asterisks). Scale bar = 10 μm. (H) Quantification of the percentage of ventral Olig2+ cells along with dorsal migrating OPCs that retain nls-mCherry+ nuclei (red) throughout oligodendroglial development. In all graphs, nls-mCherry negative cells are depicted in gray, and the numbers below each population represent the total number of cells assayed.

FIGURE 4

Metronidazole (Mtz) treatment induces radial glial cell death during late neurogenesis. (A–F) Gross morphology of transgenic gfap:nfsb-mCherry^sc059 embryos following treatment. Enlargements comparing brain hemorrhaging compared with control in insets. Scale bar = 100 μm. (G, H) Lateral views of cell death within the spinal cord visualized by immunolabeling with antiactivated caspase-3. Scale bar = 20 μm. (I–M) Images of live gfap:nfsb-mCherry^sc129 embryos at 28 and 38 hpf expressing secreted Annexin-V YFP labeling. (M) Example of NTR-mCherry+ cell with radial glial morphology expressing Annexin-V (YFP) labeling undergoing apoptosis over the course of 72 min.

FIGURE 5

Spinal cord radial glial debris is cleared by microglia. (A–B″) Lateral views of 72 hpf double transgenic gfap:nfsb-mCherry^sc059;- pU1:egfp embryos to depict macrophages (arrows) and microglia (arrowheads). (C–D″) Lateral views of 72 hpf double transgenic gfap:nfsb-mCherry^sc059;mpeg:eGFP embryo showing the presence of macrophages. (E–F″) Lateral views of a 72 hpf gfap:nfsb-mCherry ^sc059;pU1:eGFP embryo injected with control or spi1b morpholino (spi1b MO). (G) Quantification depicting percent of the total spinal cord with immune cells present in control (untreated) and Mtz (treated) embryos expressing either the mpeg or pU1 promoter and injected with spi1b MO. (H) Quantification of the levels of NTR-mCherry+ expression through average pixel intensity in control embryos and embryos treated in Mtz that were either uninjected or injected with spi1b MO. Scale bar = 20 μm.

FIGURE 6

Radial glial ablation reduces the neural progenitor population. (A–F) Cross sections of 24, 48, and 72 hpf double transgenic gfap:nfsb-mCherry^sc059;-3.9nestin:GFP embryos treated with (A–C) vehicle control or (D–F) Mtz. All sections were immunolabeled with anti-GFP to enhance the signal of the −3.9nestin:gfp transgene. (G–L) Cross section views of 24, 48, and 72 hpf gfap:nfsb-mCherry^sc059 embryos immunolabeled with anti-Sox2 (green) to visualize neural stem cell populations following treatment with (G–I) vehicle control or (J–L) Mtz. Ventral populations not lost following ablation are denoted (arrowheads). (M–P) Two examples of transverse sections of the spinal cord of 96 hpf gfap:nfsB-mCherry control (M, N) and Mtz-treated (O, P) embryos labeled for Blbp (green), Gfap (white), and nuclei (Hoechst). (M′–P′) mCherry and Blbp channels for the corresponding sections seen in M–P. A single Blbp+ midline cells can be seen with minute mCherry expression in a small lateral membrane process (M′ arrowhead); however, more laterally positioned Blbp cells were consistently negative for mCherry (M′, N′) as was the vast majority of Blbp labeling following gfap-mediated ablation (O′,P′). Scale bar = 10 μm.

FIGURE 7

Radial glial ablation and its effect on cell proliferation. (A–H′) Cross section views of proliferative cells within spinal cords of (A–D′) vehicle control and (E–H′) Mtz-treated embryos throughout neurogenesis visualized by immunolabeling with anti-BrdU (green), anti-Gfap (white), and Hoechst nuclear dye (blue). (A′–H′) represent the same images in (A–H) but only showing anti-Gfap and anti-BrdU labeling. BrdU+ nuclei outside of the spinal cord (white asterisks) or straddling the border (yellow arrowheads) are denoted. Scale bar = 10 μm. (I) Quantification of all BrdU+ populations visualized within the spinal cords between control (black line) and Mtz (green line) treatment groups. (J) Quantification of all elongated BrdU+ populations found outside of the spinal cord between vehicle control (gray) and Mtz (green) treatment groups. (K, L) Cross section views of (K) control and (L) Mtz-treated embryos immunolabeled for BrdU (cyan) and Sox2 (green) at 96 hpf. Scale bar = 10 μm. (K′–L″) Enlarged view of ventral Sox2+/NTR-mCherry negative cells (outlined) to denote cycling cells within this population (yellow asterisk). Scale bar = 5 μm. (M, N) Quantification of the two cycling Sox2+ populations present in the spinal cord following ablation. In all graphs, control averages are depicted in gray and Mtz averages are in black. Significant increases are depicted as black asterisks, and significant decreases are depicted as red asterisks.

FIGURE 8

Reductions in later-born neurons and glia following radial glial ablation. (A, B) Lateral views of spinal cords showing interneurons immunolabeled with anti-GABA within control and Mtz-treated gfap:nfsb-mCherry^sc129 embryos and quantified in (C). CoSA interneurons are highlighted in green. (D, E) Lateral views of secondary motorneurons within control and Mtz-treated gfap:nfsb-mCherry ^sc059;gata2:eGFP embryos and quantified in (F). (G, H) Lateral views of oligodendrocyte progenitor cells (OPCs, yellow asterisks) in control and Mtz-treated gfap:nfsb-mCherry^sc059;olig2:eGFP embryos and quantified in (I). Black asterisks denoted significance. Scale bar = 50 μm.