Fig 2: SOX5 and Sox6 are required for the generation of newborn neurons and are not essential for astrogliogenesis(A) Confocal images showing YFP, Dcx, and Tbr2 in the SGZ at 30 dpi in 3-month-old control, Sox5icKO, and Sox6icKO mice. White arrows indicate Dcx+ YFP+ cells.(B) Quantitation of Dcx+ cells number among YFP+ cells in the indicated mice.(C) Comparison of the number of Dcx+ cells among YFP+ cells in control, Sox5icKO, and Sox6icKO mice at 7 dpi (n = 8, 6, and 7 for control, Sox5icKO, and Sox6icKO, respectively), 14 dpi (n = 7, 7, and 6), and 30 dpi (n = 8, 6, and 6).(D) Analysis of astrogliogenesis in YFP+ cells of 3-month-old control, Sox5icKO, and Sox6icKO mice at 30 dpi using immunohistochemistry for S100 and Sox2 to identify astrocytes. Yellow arrows indicate YFP+S100+Sox2+ astrocytes. White boxes correspond to single z planes to better visualize S100 expression in some cells.(E) Quantitation of the number of S100+Sox2+ astrocytes in the recombined YFP+ cells in the SGZ (top) or the hilus (Hi) and stratum moleculare (Smol) zones (bottom) in the indicated mice. SGZ, subgranular zone.Data are mean value ± SEM. *p < 0.05, **p < 0.01, and ***p < 0.001 by unpaired Student’s t test. Scale bars represents 15 μm (A), 50 μm (D, left and center), and 40 μm (D, right). See also Figure S4.

Fig 3: Loss of Sox5 favors quiescent state in RGLs and BMP4 inhibit Sox5 expression(A) Confocal images showing immunostaining for YFP, BrdU, and GFAP of control and Sox5icKO mice brain sections at P90 at 30 dpi. Yellow arrows indicate recombined RGL cells that retain BrdU after 1 month (YFP+ BrdU+ rGFAP+ cells).(B) Experimental design for BrdU label retention experiment.(C) Quantification of BrdU+ cells among YFP+ recombined cells and BrdU+ rGFAP+ YFP+ cells among rGFAP+ YFP+ cells. Each symbol represents an independent biological replicate.(D) Scheme of adult NSC preparation grown as neurospheres.(E) Immunohistochemistry in hippocampal neurospheres for Sox2 and Sox5.(F) Neurospheres dissociated and seeded as attached cells and cultured with FGF2 alone or in combination with BMP4. A 40 min BrdU pulse was given before fixation, and BrdU incorporation was analyzed using immunohistochemistry.(G) Quantitation of Sox5 protein levels in NSCs in FGF2 and in FGF2 plus BMP4 conditions expressed in arbitrary units relative to the FGF2 condition. p = 0.0001 by Student’s t test (n = 300 cells).(H) qPCR analysis for the indicated transcripts in NSCs grown with FGF2 plus BMP4 with respect to those grown with FGF2 alone. Results are shown as 2−ΔΔCt normalized with respect to two housekeeping genes (Pgk1 and Gapdh) and relative to values in cells grown in FGF2 alone (dashed line on y axis = 1). A total of three or four independent neurospheres cultures were used, and three technical replicates were performed.(I) Immunohistochemistry of nucleofected NSCs (YFP+ cells; green), using either pCIG or pCIG-Sox5 vectors, to analyze proliferating Ki67+ cells in each condition.(J) Quantitation of the percentage of Ki67+ cells in each indicated condition.In all graphs, data are mean value ± SEM. *p < 0.05, **p < 0.01, and ***p < 0.001 by unpaired (C) or paired (H and J) Student’s t test. Scale bars represent 50 μm (A), 30 μm (E), and 80 μm (F and I).

Fig 4: Sox5 loss hinders RGL activation in response to environmental enrichment(A) Confocal images showing YFP, MCM2, and GFAP at 14 dpi in control and in Sox5icKO mice exposed to conventional (basic) or enriched (EE) environmental conditions. White arrows indicate YFP+ rGFAP+ MCM2− cells and yellow arrows YFP+ rGFAP+ MCM2+ cells.(B) Experimental design for control and Sox5icKO mice in which TAM i.p. injection was performed at P60 on 5 consecutive days, and after 7 days animals were exposed to an enriched environment for 8 days, changing objects every other day.(C) Quantitation of the number of radial GFAP in YFP+ cells in control and Sox5icKO mice.(D) Quantitation of the number of rGFAP+ MCM2+ in rGFAP+ YFP+ cells.Results are shown as mean value ± SEM. *p < 0.05, **p < 0.01, and ***p < 0.001 by unpaired Student’s t test. Scale bar, 25 μm.

Fig 5: SoxD factors control Ascl1 expression in RGLs and NSCs of the hippocampal neurogenic niche(A) Confocal image showing immunostaining for YFP, Ascl1, and GFAP in the SGZ of 3-month-old control, Sox5icKO, and Sox6icKO mice at 14 dpi. Cyan arrows indicate recombined Ascl1+ rGFAP− cells, white arrows rGFAP+ cells, and yellow arrows double rGFAP+ Ascl1+ cells.(B) Quantitation of the number of rGFAP+ in the population of YFP+ cells and of rGFAP+ Ascl1+ cells in the rGFAP+ YFP+ group in the indicated mice.(C) Quantitation in Sox5icKO mice of the proportion of YFP+ RGLs maintaining or losing Sox5 expression (Sox5+ or Sox5−, respectively). The percentage of Ascl1+ cells in the subset of Sox5+/YFP+ or Sox5−/YFP+ RGLs is indicated.(D) Transcription levels for the indicated genes analyzed using qPCR in hippocampal NSCs transfected with Sox5-pCIG or Sox6-pCIG, represented as 2−ΔΔCt normalized with respect to the housekeeping gene Gapdh and relative to values in NSCs transfected with control pCIG (dashed line on y axis = 1).(E) Representative images of NSCs transduced with lentiviral particles expressing shRNA-control or shRNA-Sox5-A and GFP as a reporter protein. Graphs showing transcriptional levels by qPCR for each lentiviral transduced condition. Data are represented as 2−ΔΔCt normalized with respect to the housekeeping gene Gapdh and relative to values in NSCs transfected with shRNA-control.(F) Genomic sequence alignment comparing human ASCL1 locus with homologous D. rerio, G. gallus, X. tropicalis, M. domestica, R. norvegicus, M. musculus, and M. rhesus Ascl1 loci. Presence of conserved Sox-binding sites (red squares) within a 1,304 bp conserved enhancer, 6.6 kb upstream of the human ASCL1 coding region.(G) Quantitative analysis of transcriptional activity of Sox5, Sox6, and Sox2 on a 1 kb mouse Ascl1 enhancer in adult NSCs. Graph shows normalized luciferase activity units relative to the pCIG control. Dots in bars correspond to independent experiments.(H) Quantitation of three independent experiments of ChIP assays in rat adult hippocampal NSCs. DNA fragments for Ascl1 immunoprecipitated with Ac-H3, IgG, or anti-SOX6 antibodies were analyzed using qPCR with primers specific for a 214 bp fragment containing Sox-binding sites depicted in (F). Data are expressed as percentage of input = 100 × 2[adjusted input – Ct (IP)].Results are shown as mean value ± SEM. *p < 0.05, **p < 0.01, and ***p < 0.001 by unpaired Student’s t test. Scale bars, 25 μm (A) and 50 μm (D). See also Figure S6.