Fig 2: Astrocytes dynamically respond to sub-chronic stress, knockdown of aquaporin-4 leads to astrocyte dystrophy in the prefrontal cortex.A) Representative images of vascular tomato lectin (magenta) and immunohistology for astrocyte GFAP (cyan) and AQP4 (yellow) in the PFC (20×, scale bar = 100 μm). B) Left: Representative images of immunohistology for ALDH1L1 in the PFC (20×, scale bar = 100 μm). Right: Number of ALDH1L1+ astrocytes in the PFC per mm2. C) Area of GFAP+ material in the PFC. D) Area of GFAP+ material associated with blood vessels in the PFC. E) Total area of AQP4+ material in the PFC. F) Area of astrocyte AQP4+ material relative to vessel area. G) Area of tomato lectin+ vessels in the PFC. # p<0.05 main effect (no specific contrasts detected). * p<0.05 planned comparison indicated (Sidak’s test).

Fig 3: Chronic stress (14-days) leads to astrocyte atrophy and vascular-associated aquaporin-4 loss in the prefrontal cortex in male, but not female, mice.A) Representative images of vascular tomato lectin (TL) staining (magenta) and immunohistology for astrocyte GFAP (cyan) and AQP4 (yellow) in the PFC (20× + zoom, scale bar = 100 μm). Dashed quadrilaterals indicate TL+ vessel segments with little- to no AQP4 coverage. B) Number of ALDH1L1+ astrocytes in the PFC per mm2. C) Area of GFAP+ material in the PFC. D) Area of GFAP+ material associated with blood vessels in the PFC. E) Area of astrocyte AQP4+ material relative to vessel area. F) Area of tomato lectin+ vessels in the PFC. G) Left: Representative image of Alexa Fluor 488-conjugated dextran (10,000 MW, pale yellow) and vascular tomato lectin (magenta) in the PFC. Image features three extravasation sites, with one site highlighted in XZ-YZ dimensions. Middle: Number of dextran extravasation sites in the PFC per image. Right: Mean fluorescence intensity of dextran extravasation sites in the PFC (normalized to control males). Bars represent mean ± S.E.M. # p<0.05 main effect (no specific contrasts detected). * p<0.05 planned comparison indicated (Sidak’s test).

Fig 4: Knockdown of prefrontal aquaporin-4 dysregulates astrocyte structure yet has little effect on behavioral stress susceptibility in female mice.Adult female mice received infusions of either AAV5-scrambled-shRNA or AAV5-Aqp4-shRNA into the PFC (see Fig. 4A for schematic). Following three weeks of recovery, animals were exposed to 7-days of sub-CUS or were left unstressed. Post-behavioral testing (n = 8–10/group), a subset of mice received intravenous injections of tomato lectin, after which brains were processed for confocal microscopy (n = 5–6/group). A Total amount of distance traveled in the open field test (OFT). B Time spent in the center during the OFT. C Total amount of time spent exploring objects in the test phase of the temporal object order recognition task (TOR). D Discrimination index in the TOR. E Time spent exploring both the novel and familiar arm in the T-maze task. F Total number of alternations in the T-maze task. G Proportion of spontaneous alternations in the T-maze task. H Number of ALDH1L1+ astrocytes in the PFC per mm2. I Representative images of vascular tomato lectin (TL) staining (magenta) and immunohistology for astrocyte GFAP (cyan, top panel) and AQP4 (yellow, bottom panel) in the PFC (20× + zoom, scale bar = 100 µm). J Area of GFAP+ material in the PFC. K Area of GFAP+ material associated with blood vessels in the PFC. L Total area of AQP4+ material in the PFC. M Area of astrocyte AQP4+ material relative to vessel area. N Area of tomato lectin+ vessels in the PFC. Bars represent mean ± S.E.M. Graphed points indicate data derived from individual animals. †p < 0.05 main effect (specific group difference not detected by planned comparison). *p < 0.05 planned comparison indicated (Sidak’s test following a significant main effect or interaction).

Fig 5: Chronic stress (14-days) leads to astrocyte atrophy and vascular-associated aquaporin-4 loss in the prefrontal cortex in male, but not female, mice.A Representative images of vascular tomato lectin (TL) staining (magenta) and immunohistology for astrocyte GFAP (cyan) and AQP4 (yellow) in the PFC (20× + zoom, scale bar = 100 µm). Dashed quadrilaterals indicate TL+ vessel segments with little- to no AQP4 coverage. B Number of ALDH1L1+ astrocytes in the PFC per mm2. C Area of GFAP+ material in the PFC. D Area of GFAP+ material associated with blood vessels in the PFC. E Area of astrocyte AQP4+ material relative to vessel area. F Area of tomato lectin+ vessels in the PFC. G Left: Representative image of Alexa Fluor 488-conjugated dextran (10,000 MW, pale yellow) and vascular tomato lectin (magenta) in the PFC. Image features three extravasation sites, with one site highlighted in XZ-YZ dimensions. Middle: Number of dextran extravasation sites in the PFC per image. Right: Mean fluorescence intensity of dextran extravasation sites in the PFC (normalized to control males). Bars represent mean ± S.E.M. Graphed points indicate data derived from individual animals. †p < 0.05 main effect (specific group difference not detected by planned comparison). *p < 0.05 planned comparison indicated (Sidak’s test following a significant main effect or interaction).