Fig 2: Rescue of the postnatal persistence of radial glia and cortical neurogenesis phenotypes in Prdm16-Vcam1 DKO mice.(A) Increased Vcam1 immunofluorescence in the V-SVZ of juvenile and adult Prdm16 cKO mice. The dashed lines delineate the boundaries between the V-SVZ and the striatum and the boundary between the V-SVZ and the ventricle. Scale bar: 50 µm.(B) The persistence of GFAP+ radial processes is partially rescued in Prdm16-Vcam1 DKO mice. Scale bar: 50 µm.(C) Whole-mount en-face preparations showing neural stem cells with single ?-tubulin+ dots and ependymal cells with clusters of ?-tubulin+ dots. Scale bar: 50 µm. The cell composition defect in Prdm16 cKO mice is rescued in Prdm16-Vcam1 DKO.(D) Quantification of Vcam1 immunofluorescence positive area in V-SVZ. N=4–6 images per mouse, 3 mice per genotype. Welch’s t-test, p=0.0287.(E) Quantification of the number of radial processes at one month of age. N=4 images per mouse, 4–8 mice per genotype. One-way ANOVA with Tukey’s multiple comparisons test, control vs. Prdm16 cKO, p=0.002, Prdm16 cKO vs. Prdm16-Vcam1 DKO, p=0.019.(F) Quantification of neural stem and ependymal cells. N=4–5 images per mouse, 3 mice per genotype. One-way ANOVA with Tukey’s multiple comparisons test. Ependymal cells, control vs. Prdm16 cKO, p=0.0017, Prdm16 cKO vs. Prdm16-Vcam1 DKO, p=0.0097. Neural stem cells, control vs. Prdm16 cKO, p=0.0124. Prdm16 cKO vs. Prdm16-Vcam1 DKO, p=0.0262.(G) The postnatal persistence of newly born DCX+ neurons in the cerebral cortex seen in Prdm16 cKO mice is rescued in Prdm16-Vcam1 DKO mice. The dashed lines delineate the boundaries between the white matter and the grey matter. Scale bar: 100 µm.(H) Quantification of DCX+ cells at P21 in the cerebral cortex. Only DCX+ cells in the grey matter were counted. N=5–8 images per mouse, 3 mice per genotype. One-way ANOVA with Tukey’s multiple comparisons test. Control vs. Prdm16 cKO, p=0.0041. Prdm16 cKO vs. Prdm16-Vcam1 DKO, p=0.0055.

Fig 3: Impact of LPC surgery on mice subjected to running wheel. (A) Workflow and representative eriochrome cyanine-stained sections with lesion in the ventrolateral white matter delineated by dashed lines. Figure was drawn using BioRender. (B) Representative images of LPC lesion (left) and contralateral normal appearing white matter (right) 4 days post injury stained for mature oligodendrocytes (CC1) in red, OPCs (PDGFRa) in green, oligodendrocyte lineage cells (OLIG2) in white, and DNA (DAPI) in blue. (C) Representative images of LPC lesion 4 days post injury stained for myelin and myelin debris (MBP) in red, axons (NFH) in green, and astrocytes (GFAP) in white, and DNA (DAPI) in blue. In both (B) and (C), the lesion is outlined by the irregular dashed line while the non-involved contralateral site is denoted by the rectangle dashed line. Scale bar represents 100 µm.

Fig 4: Persistence of cells with radial processes in juvenile and adult Prdm16 cKO mice(A) Prdm16 immunofluorescence in the V-SVZ and striatum of Prdm16 cKO and control mice at P16. The dashed lines delineate the boundaries between the V-SVZ and striatum. Scale bar: 50 µm.(B) A diagram shows the area of lateral V-SVZ used in the quantification and shown in (D). Quantification of the number of radial processes at one month of age in the striatum. Images were taken between -0.1 mm posterior and 0.74 mm anterior to Bregma. Only radial processes that extend beyond 200 µm away from the ventricular walls were counted. N=4 images per mouse, 4 mice per genotype. All statistical tests were performed using data from each mouse (biological replicate) as an individual observation in all figures except for scRNAseq. Welch’s t-test. p=0.0276. In all figures: *, p<0.05. **, p<0.01. ***, p<0.001. N.S., not significant.(C) A diagram shows the area of the cerebral cortex used in the quantification and shown in (E). Quantification of the number of radial processes at P21 in the cerebral cortex. Images were taken between 0.1 and 1.18 mm anterior to Bregma. Only radial processes that extend beyond the white matter were counted. N=5–8 images per mouse, 3 mice per genotype. Welch’s t-test. p=0.0021.(D) Persistence of GFAP+/Nestin+ basal processes in juvenile and adult Prdm16 cKO mice. Scale bar: 50 µm. The dashed lines indicate ventricular walls.(E) Persistence of GFAP+ basal processes in P21 Prdm16 cKO cortex. Scale bar: 100 µm.

Fig 5: Immunofluorescence shows redistribution of AQP4 from perivascular to non-perivascular domains in the border zone of the infarct. (A) Illustration showing the infarct core in the cortex, ipsilateral to the dMCAO. (B) Scanned overview of a coronal section with AQP4 (red) and GFAP (green) immunofluorescence co-staining. Dotted lines show the limits of the infarct. Location of the high magnification images shown in (C–F) is indicated in the scanned overview. (C–F) Confocal z-projections (2 µm thickness) of areas indicated in B, taken with a 63 × objective. Arrows indicate vessels, asterisk indicate strong AQP4 immunofluorescence staining in neuropil. (C) In the border zone closest to the infarct core, GFAP-positive astrocytic processes rectified towards the core (dotted line), are accompanied by diffuse AQP4 immunofluorescence staining throughout the area, with relatively modest immunostaining of the perivascular processes (arrows). (D) Moving further away from the core into the peri-infarct area, strong perivascular AQP4 staining (arrows) is accompanied by pronounced neuropil expression (asterisk). Strong GFAP immunofluorescence is present in astrocytic processes. (E,F). In the ipsilateral cortex remote from the infarct (E), and in the contralateral cortex (F), AQP4 immunofluorescence is mainly localized around the vessels (arrows). Strong GFAP staining, indicative of astrogliosis is more pronounced in the ipsilateral cortex. Pseudo-colors: red for AQP4 and green for GFAP. Scale Bars: (A); 2 mm, (C–F); 50 µm.