Fig 2: Astrocytes form limitans borders that isolate hydrogels and non-neural FBR components from viable neural tissue.a, b Progression of astrocyte limitans border formation from 1 week a to 6 weeks b for DCHMO, DCHK, and L-NIO-induced stroke. c, d OPC identified by NG2-targeted reporter (tdT) c and Olig2 d intermingle with GFAP-positive astrocytes and do not migrate into CD13-positive regions. e Comparison of extent of blood–brain barrier (BBB) disruption and repair as measured by IgG staining around DCHMO and DCHK deposits at 1 and 6 weeks after injection. f Quantification of the percentage increase in IgG levels in the hydrogel injected caudate putamen (CP) normalized to the non-injected contralateral side. **P = 0.0012 and ***P = 0.0001 for DCHK versus DCHMO at 1 week and 6 week, respectively, and not significant (NS) for DCHK samples between the two timepoints, two-way ANOVA with Bonferroni. Graph shows mean ± s.e.m with individual data points superimposed showing n = 5 and 6 mice for DCHMO and DCHK, respectively.

Fig 3: Accelerated ischemic brain injury in pericyte-deficient Pdgfrb + ?/- mice and protection by 3K3A-APC. (A) Scheme of the experimental set-up. Mice underwent 45-min transient middle cerebral artery occlusion (tMCAo) followed by reperfusion. 3K3A-APC (0.2 mg/kg i.v.) or vehicle were administered 10 min and 4 h after tMCAo. Cerebral blood flow (CBF) was monitored for the first 75 min. Blood-brain barrier (BBB) breakdown was evaluated by tissue analysis 8 h after tMCAo. Motor neurological score, neuropathological, and other tissue analyses (i.e., pericyte coverage, capillary length, neurodegeneration) were performed 24 h after tMCAo. (B) Immunostaining for endothelial (CD31) and pericyte (CD13) markers in non-ischemic cortex of Pdgfrb+/+ control mice and Pdgfrb + ?/- pericyte-deficient mice. (C–F) CBF determined by laser Doppler flowmetry during first 75 min of tMCAo in Pdgfrb+/+ controls (C) and Pdgfrb + ?/- mice treated with vehicle (D) or 3K3A-APC (E). (F) CBF quantification at 75 min after tMCAo initiation in mice studied in (C–E). (G) Representative images of cresyl violet staining of brain sections (at + 1.6, + 0.8, 0, –0.8, or –1.6 mm from the bregma) 24 h after tMCAo from Pdgfrb+/+ mice and Pdgfrb + ?/- mice treated with vehicle or 3K3A-APC. Infarct areas are outlined. Volumes of infarct (H) and edema (I), and motor neurological score (J) 24 h after tMCAo in Pdgfrb+/+ mice and Pdgfrb + ?/- mice treated with vehicle or 3K3A-APC. In (C–F,H–J), mean ± SEM, n = 5 mice per group. The p-values were calculated by one-way ANOVA followed by Tukey’s multiple comparisons test.

Fig 4: Glucose uptake in brain pericytes via SGLT2.a Representative example of the expression of SLC5A2 (SGLT2) in cultured human brain microvascular pericytes, HEK-293T, and HK2 cells assessed using RT-PCR and quantification using qPCR (n = 3). The mRNA expression was normalized to that of RNA18S. b Representative immunoblottings for SGLT2, CD13, and ß-actin in cultured pericytes of total lysates or membrane fractions obtained by ultracentrifugation, which were transfected with pcDNA3 or pcDNA3-SLC5A2. Human kidney lysates were used as a positive control for SGLT2 expression. c A representative immunoblotting for SGLT2 in cultured brain pericytes after SLC5A2 overexpression with or without PNGase F treatment. The black arrowhead indicates glycosylated SGLT2 and the white arrowhead indicates unglycosylated SGLT2. d Representative images of immunofluorescence for SGLT2 (green), CD13 (red), and DAPI in a cultured brain pericyte (scale bar = 20 µm). e, f (e) Brain pericytes were incubated in Na+-containing HEPES-buffered saline, Na+-free solution, or Na+-containing solution with 100 or 1,000 nmol/L luseogliflozin for 15 min, followed by incubation with additional 200 µmol/L of 2-NBDG (scale bar = 20 µm). f Fluorescence intensity of 2-NBDG was measured using a microplate reader and normalized to that of Hoechst 33342 (n = 8). Data are presented as dot-plots of individual experiments and mean values ± SD. n.s. (not significant) P > 0.05, ***P < 0.001, ****P < 0.0001 by One-way ANOVA followed by Bonferroni’s post hoc test. SGLT2 sodium-glucose cotransporter 2, Luseo luseogliflozin, pMCAO permanent middle cerebral artery occlusion, 2-NBDG 2-deoxy-2-[(7-nitro-2,1,3-benzoxadiazol-4-yl) amino]-d-glucose.

Fig 5: Altered pericyte coverage in mTBI mice. a–b Representative images (a) and quantification (b) showing extravasation of plasma IgG (red) in brain parenchyma 3 days after mTBI. n = 6 mice. Dash line indicates an average value from sham-operated group (n = 5 mice). Scale bar = 50 µm. c Representative images showing CD13-positive pericyte coverage (green), lectin-positive endothelial vascular profiles (blue) and extravasation of fibrin deposits (red) in brain parenchyma of mTBI-affected ipsilateral side or contralateral side 3 days after mTBI. Scale bar = 50 µm. Insert on right: high magnification images of boxed regions in (c1–c3), Scale bar = 10 µm. d–e Quantification of CD13-positive pericyte cell bodies per mm2 (d) and pericyte coverage on lectin microvessel profiles (e) in brain parenchyma of mTBI-affected ipsilateral side and contralateral side 3 days after mTBI. n = 9 mice per group. f Pearson’s correlation plot between BBB impairment based on fibrin deposits and loss of pericyte coverage in brain parenchyma 3 days after mTBI. n = 18 hemispheres. r = Pearson's coefficient. P, significance. g Quantification of microvascular density in the cortexes of mTBI-affected ipsilateral side and contralateral side 3 days after mTBI. n = 9 mice per group. Mean ± SD in b, d–e, g; **, P < 0.01; ***, P < 0.001; NS, non-significant (P > 0.05) by Student’s t-test