Caveolin-1 Regulates Ultrastructural Alterations of Astrocytes in Chronic Cerebral Hypoperfusion

Abstract

Cerebral white matter exhibits heightened susceptibility to chronic hypoperfusion. Increasing evidence implicates glial cells, notably astrocytes, in mediating chronic ischemic demyelination. To elucidate the involvement of astrocytes in ischemic white matter pathologies, we conducted an ultrastructural characterization of intracellular contents in reactive astrocytes of mice following bilateral carotid artery stenosis (BCAS), a model of chronic cerebral hypoperfusion. BCAS triggered robust activation of astrocytes, with electron-dense dark astrocytes demonstrating cytoplasmic/nuclear hypercondensation via transmission electron microscopy, in the corpus callosum. These astrocytes exhibited markedly elevated cellular stress hallmarks, including mitochondria alteration, Golgi cisternal vesiculation/fragmentation, and endoplasmic reticulum dilation. Chronic hypoperfusion enhanced phagocytic activity and increased the lysosomal pathway in dark astrocytes. The conditional knockout of astrocytic Caveolin-1 (Cav-1) prompted adaptive cellular remodeling, characterized by condensed nucleoplasm and increased organelle abundance without structural alterations. Following BCAS, astrocyte-specific Cav-1 ablation significantly attenuated ultrastructural indicators of ischemia-related cellular stress, indicating enhanced astrocytic tolerance to chronic hypoperfusion. Most importantly, astrocytic Cav-1 deficiency ameliorated demyelination in the corpus callosum. Overall, our study provides the quantitative ultrastructural analysis of astrocytes in ischemic white matter and identifies astrocytic Cav-1 as a regulatory checkpoint for chronic ischemic demyelination.

Supplementary Information

The online version contains supplementary material available at 10.1007/s10753-025-02431-0.

Introduction

Chronic cerebral hypoperfusion, disrupting blood flow from distal parts of long deep arteries, is a critical contributor to white matter injury (WMI) [1]. WMI is linked with a series of symptoms, such as cognitive decline, gait disturbances, and movement disorders, increasing the risk of dementia and stroke [2]. WMI after chronic hypoxia-ischemia is characterized by diffuse myelin loss, axon injury, and glial cell proliferation at micro-structural levels [3, 4]. Astrocytes represent the major components of glial cells, constituting 19%−40% of those in the central nervous system (CNS) [5, 6]. They fulfill well-established functions, including participating in synapse development, neuronal metabolism [7–9], as well as remyelination through astrocyte-oligodendrocyte crosstalk [10]. In different brain regions, astrocytes exhibit remarkable morphological features, molecular expression, and functional heterogeneity [7, 9, 11]. In grey matter, astrocytes are known as protoplasmic astrocytes, exhibiting short and largely branched tertiary processes, interacting with the neuronal synapse to maintain nervous homeostasis [9]. In white matter, astrocytes are visualized as fibrous with long unbranched processes and close proximity to the Ranvier node, oligodendrocytes, and vasculature to support network activity and myelination [9, 10, 12]. Under pathological context, astrocytes become reactive and dramatically change their characteristic to respond to brain damage [13]. A recent study has found that astrocytes became hypertrophic with many mitochondria, clear cytoplasm, and some processes to enwrap cellular debris in the ischemic penumbra after transient ischemic insult [14, 15]. In an aged mouse model of Alzheimer’s disease, electron-dense astrocytes with cell stress markers, for altered mitochondria and dilated endoplasmic reticulum (ER), were observed in the hippocampal parenchyma [16]. However, the morphological alteration and functional heterogeneity of astrocytes, as well as the underlying mechanisms in chronic cerebral hypoperfusion, remain poorly understood.

Caveolin-1 (Cav-1), a fundamental membrane scaffolding protein, orchestrates diverse cellular processes, including functionally integrating signal transduction [17], oxidative stress response [18], and DNA repair [19]. Recently, increasing Cav-1 activity with the downstream Src-kinase and Akt pathways has been proposed to prevent cell death by alleviating oxidative stress and DNA damage [20]. Cav-1 has also been reported to be associated with cell morphology through Tyr14 phosphorylation-dependent conformational change [20]. Moreover, Cav-1, as an integral membrane protein, has a crucial role in inter-organelle communication, like the ER, Golgi, and mitochondria, through cholesterol homeostasis [21]. Our previous study has demonstrated that endothelial Cav-1 deficiency compromises oligodendrogenesis and myelin regeneration through disrupting vascular-oligodendrocyte precursor cell interaction [22]. Astrocytes in the CNS also abundantly express Cav-1 [23]. However, it remains unclear whether Cav-1 could modulate astrocytic adaptive responses to chronic ischemic injury and subsequent demyelination in the corpus callosum.

Thus, we performed an ultrastructural analysis for the distinct structural properties and functional characteristics of astrocytes in the corpus callosum using bilateral carotid artery stenosis (BCAS), a well-accepted animal model of chronic cerebral hypoperfusion [4]. We found that after BCAS, astrocytes exhibited marked activation, characterized by electron-dense cell bodies and prominent cellular stress responses, such as mitochondria alteration, Golgi apparatus vesiculation, and ER dilation. Additionally, chronic hypoxic-ischemic injury led to significant upregulation of lysosomal pathway activity. Notably, astrocyte-specific Cav-1 conditional knockout promoted adaptive cellular remodeling, thereby enhancing astrocytic resilience to chronic hypoxia-ischemia induced by BCAS.

Materials and methods

Antibodies

Anti-GFAP (ab53554; 1:300 for immunofluorescence; RRID: AB_880202), anti-Galectin-3 (ab2785; 1:200 for immunofluorescence; RRID: AB_303298), anti-Cav-1 (ab2910; 1:300 for immunofluorescence; RRID: AB_303405), Anti-MBP (ab40390; 1:300 for immunofluorescence; RRID: AB_1141521), and anti-GRP78 (ab21685; 1:200 for immunofluorescence; RRID: AB_2119834) were purchased from Abcam, Cambridge, UK. Anti-GFAP (3670; 1:300 for immunofluorescence; RRID: AB_561049) was purchased from Cell Signaling Technology, Massachusetts, USA. Anti-Tom20 (11802-1-AP; 1:200 for immunofluorescence; RRID: AB_2207530) was purchased from Proteintech, Wuhan, China. Anti-GM130 (AP90199; 1:200 immunofluorescence; RRID: AB_3714574) was purchased from Beyotime Biotechnology, Wuhan, China. Anti-LAMP1 (AF4320; 1:300 for immunofluorescence; RRID: AB_2296826) was purchased from R&D Systems, Minneapolis, Minnesota, USA. Alexa Fluor 594-AffiniPure Donkey Anti-Goat IgG (H + L) (705-585-003; 1:400 immunofluorescence; RRID: AB_2340432), Alexa Fluor 488-AffiniPure Donkey Anti-Goat IgG (H + L) (705-545-003; 1:400 immunofluorescence; RRID: AB_2340428), Alexa Fluor 594-AffiniPure Donkey Anti-Mouse IgG (H + L) (715-585-150; 1:400 immunofluorescence; RRID: AB_2340854), Alexa Fluor 488-AffiniPure Donkey Anti-Mouse IgG (H + L) (715-545-150; 1:400 immunofluorescence; RRID: AB_2340846), Alexa Fluor 488-AffiniPure Donkey Anti-Rabbit IgG (H + L) (711-545-152; 1:400 immunofluorescence; RRID: AB_2313584), and Alexa Fluor 594-AffiniPure Donkey Anti-Rabbit IgG (H + L) (711-585-152; 1:400 immunofluorescence; RRID: AB_2340621) were purchased from Jackson, West Grove, USA.

Animals

Male C57BL/6 mice of 8–10 weeks, Gfap-Cre mice (T004857), and Cav-1^fl/fl mice (T008506) were purchased from Gempharmatech CO., Ltd (Nanjing, Jiangsu, China). Cav-1-cKO mice were generated by crossing Gfap-Cre mice with Cav-1^fl/fl, Cre-neg littermates as controls. Animals were housed under conventional conditions on a 12 h light/dark cycle, with 20–25 °C room temperature, humidity 45–60%, as well as food and water available ad libitum. The mice were kept in groups of no more than 5 animals per cage.

BCAS Surgery

The BCAS model was established as reported previously [4]. Briefly, mice were deeply anesthetized with 5% isoflurane and maintained with 2% isoflurane in oxygen (RWD Life Science Co., LTD). Both the bilateral common carotid arteries (CCAs) were surgically exposed. A 0.18 mm inner diameter microcoil (Sawane Spring Co., Japan) was twined around the CCA, with an identical microcoil applied to the contralateral CCA following a 30-minute interval. The sham group underwent the same procedure without planting the microcoil around their CCAs. Rectal temperature was maintained between 36.5 and 37.5 °C during surgery.

Cerebral Blood Flow (CBF) Measurements

CBF at the baseline, 2 weeks, and 4 weeks after BCAS surgery was assessed by the Laser speckle flowmetry (PeriCam PSI Normal Resolution with PIMSoft, Perimed, Sweden). Anesthetized by 2% isoflurane in oxygen, mice were placed in the prone position to expose the skull. Color-coded blood flow images in high-resolution mode were captured by a CMOS camera positioned above the head. The mean value of CBF was quantified based on the color image program incorporated in the flowmetry system.

Tissue Processing and Immunohistochemistry

After anesthesia with isoflurane, mice from the sham-operated group, as well as those at 2 weeks and 4 weeks after BCAS surgery, were transcardially perfused with phosphate-buffered saline and 4% paraformaldehyde (PFA). Brains were extracted from the skull and immersed in 4% PFA for 4 h, followed by 15% and 30% sucrose in phosphate buffer for 24 h each. Embedded in Tissue-Tek O.C.T compound (Sakura Finetek, USA) and stored at −80 °C, frozen brains were cut into 20 μm-thick sections using a cryostat microtome (Leica, Wetzlar, Germany). For immunofluorescence staining, the frozen sections were fixed with 4% PFA for 10 min, followed by blocking with 5% normal donkey serum, 1% BSA, and 0.1% Triton X-100 for 1 h at room temperature. These sections were incubated with the primary antibodies overnight at 4 °C, followed by appropriate fluorophore-conjugated secondary antibodies (1:400) and DAPI (Sigma, USA). Images were captured using the Olympus BX53F2 microscope (Olympus, Japan) at constant microscope exposure times and zoom levels. The 3D reconstruction images were processed by Imaris software.

Quantification for Immunofluorescence Images

Positive signals were quantified using ImageJ software. For each group (n = 5 mice), five randomly selected fields within the region of interest were analyzed across five non-adjacent sections by investigators blinded to group allocation. Single- or double-labeled cells were manually counted using the cell counter tool of ImageJ software, and the counts were normalized to the total cell population. To quantify MBP staining and intra-astrocytic immunofluorescence signals, images were binarized in ImageJ using the Auto Threshold function with a fixed threshold to separate foreground from background pixels. The data were then averaged to generate a single value per animal and normalized to the sham group. Astrocytic hypertrophy was quantified through cell body area/process length in GFAP-stained sections.

Transmission Electron Microscope (TEM)

Following euthanasia with isoflurane, fresh corpus callosum tissue samples (approximately 1 mm³) were rapidly dissected on ice from both sham-operated mice and BCAS mice at the 4-week time point after the surgery. Samples were immediately immersed in primary fixative consisting of 2.5% (v/v) glutaraldehyde (Sigma-Aldrich, G5882) in 0.1 M phosphate buffer (pH 7.4) and fixed overnight at 4 °C. Subsequently, samples were rinsed three times (10 min each) with 0.1 M phosphate buffer (pH 7.4). Post-fixation was performed in 1% (w/v) osmium tetroxide solution in 0.1 M phosphate buffer (pH 7.4) for 2 h at 4 °C. Following three additional washes (10 min each) with 0.1 M phosphate buffer, samples were en bloc stained with 2% (w/v) aqueous uranyl acetate (SPI-Chem, GZ02625) for 2 h at 4 °C. Samples were then dehydrated through a graded ethanol series (30%, 50%, 70%, 80%, 90%, 95%, and 100% ethanol; 15 min each step), followed by two changes of anhydrous acetone (15 min each). Dehydrated samples were infiltrated and embedded in PON-812 epoxy resin (SPI-Chem, GS02660). Polymerization was carried out at 60 °C for 48 h. Ultrathin sections (approximately 60 nm thick) were cut using an ultramicrotome (Leica EM UC7) and collected onto 150-mesh formvar-coated copper grids. Sections were then stained with lead citrate (SPI-Chem, GA10701) for 5 min. Images were acquired using a JEOL JEM-1400Flash TEM operating at an acceleration voltage of 120 kV.

Ultrastructural Analysis of Astrocytes

Typical astrocytes were characterized by the electron-lucent cytoplasm and nucleoplasm, a granular nuclear pattern, and the presence of intermediate filaments. Furthermore, they exhibited angular cellular processes that intimately interact with parenchymal elements, including end-foot terminals enveloping blood vessels, direct contact with myelinated axons, and close associations with neighboring glial cells [24]. Activated astrocytes exhibited a large amount of chromatin condensation along the nuclear membrane and a large nucleolus [25].

Ultrastructural features of cellular contents were investigated, such as the presence of mitochondria, Golgi apparatus, as well as ER. Mitochondria were characterized by the electron-dense matrix, double membrane, and numerous cristae formed by inner membrane invaginations [26]. Altered mitochondria displayed a deterioration of the inner and/or outer membrane, electron-lucent space due to the degradation of cristae, or a “holy shape” for mitochondria wrapping around themselves [16, 27]. The elongated mitochondria were identified if their length was over 1000 nm [28]. The ultrastructure of Golgi apparatus was shown as stacks of five to seven flattened cisternae overlaying one another, with vesicles at the ends of cisternae [29]. The ER was characterized by its long and narrow stretches and proximity to the mitochondria [24]. Dilated Golgi and ER apparatus were identified by the swollen electron-lucent appearance, measuring at least 100 nm in diameter [16, 26]. Lysosomes were identified by their circular and homogenous (primary) or heterogeneous appearance (secondary and tertiary). In brief, the primary lysosome displayed a round and homogenous shape with electron-dense granules found near the membrane. Secondary lysosomes were round and heterogeneous, determined by the target substrates, such as disintegrated intracellular organelles and phagosomes [30]. Tertiary lysosomes exhibited morphological heterogeneity, containing vacuolar structures, lipofuscin granules, and lipid droplets [30, 31]. Multivesicular bodies were formed through the fusion of primary lysosomes with endocytic vesicles, functioning primarily in the recycling of plasma membrane components [32].

To assess the astrocytic area, including the area of cytoplasm and nucleoplasm, the outline of each astrocyte cell body was traced using the freehand tool in ImageJ software. Meanwhile, the width of Golgi apparatus and ER cristae, as well as the length of mitochondria, were calculated by ImageJ software.

Images of 8 astrocytes from each animal (n = 5) in the corpus callosum were acquired. These images were blind to the experimental condition to avoid bias. We analyzed 40 astrocytic cell bodies in each group, which were considered to be sufficient to obtain statistical power, based on the software G*Power V3.1 (effect size of 0.9; power of 0.9 estimated at 27 astrocytes in each group) [16, 33].

Statistical Analysis

All parameters were presented as mean ± standard deviation (SD) and analyzed by the two-tailed unpaired t-test, or one-way ANOVA, followed by Dunnett post hoc test as appropriate. Statistical analyses were performed with GraphPad Prism software, version 8.0 (GraphPad Software, Inc., USA). P < 0.05 was considered statistically significant (*P < 0.05, **P < 0.01).

Results

Chronic Cerebral Hypoperfusion Induces Demyelination and Astrocytic Activation in the Corpus Callosum

To induce chronic cerebral hypoperfusion, we subjected mice to BCAS surgery with a microcoil of 0.18 mm inner diameter. The CBF was assessed by laser speckle contrast imaging at baseline, 2 weeks, and 4 weeks post-surgery. Quantitative analysis revealed a significant reduction in CBF to 80.8% of baseline levels at 2 weeks, followed by partial recovery to 84.5% of baseline levels by 4 weeks post-surgery (354.50 ± 32.18 vs. 286.60 ± 24.57 vs. 299.60 ± 15.73; compared with the sham group, P < 0.01, respectively; Fig. 1A, B). Immunofluorescence analysis revealed a reduction in the expression of myelin basic protein (MBP), a marker for myelinated fibers, from 2 to 4 weeks post-injury (compared with the sham group, P < 0.01, respectively; Fig. 1C, D), providing consistent evidence with our previous reports [22, 34, 35] for substantial corpus callosum demyelination resulting from chronic cerebral hypoperfusion.

Fig. 1.

Corpus callosum demyelination and astrocytic activation under chronic hypoperfusion. (A, B) Representative CBF images and quantitative analysis at baseline, 2 weeks, and 4 weeks post-BCAS surgery (n = 5 in each group; mean ± SD; **P < 0.01; one-way ANOVA, Dunnett post hoc test). (C, D) Immunofluorescence images and quantification of MBP staining (n = 5 in each group; mean ± SD; **P < 0.01; one-way ANOVA, Dunnett post hoc test). Scale bar, 100 μm. (E) Representative images for time-course changes in GFAP⁺ astrocytes across the entire corpus callosum. Scale bar, 100 μm. (F) Quantification of the number of GFAP⁺ astrocytes across the entire corpus callosum (n = 5 in each group; mean ± SD; **P < 0.01; one-way ANOVA, Dunnett post hoc test). (G) Quantification of astrocytic hypertrophy through cell body area/process length in GFAP-stained tissue (n = 5, 40 astrocytes in each group; mean ± SD; **P < 0.01; one-way ANOVA, Dunnett post hoc test). Abbreviations: BCAS, bilateral carotid artery stenosis; CBF, cerebral blood flow; MBP, myelin basic protein; GFAP, glial fibrillary acidic protein; SD, standard deviation.

To determine the prevalence and morphology of astrogliosis, we immunostained the main constituent of astrocyte intermediate filaments (glial fibrillary acidic protein, GFAP), also a reactive astrocytic marker, in the corpus callosum. In the sham group, GFAP⁺ astrocytes kept a quiescent state with fibrous processes, while in the BCAS group, astrocytes became hypertrophic. The number of GFAP⁺ astrocytes, along with the cell body area/process length, was markedly enhanced at 2 weeks and remained at an elevated level at 4 weeks following BCAS (compared with the sham group, all P < 0.01; Fig. 1E–G). These data suggest that chronic hypoxic-ischemic stress promotes demyelination and reactive astrogliosis.

Reactive Astrocytes Undergo Increasing Cellular Stress in the Corpus Callosum after Chronic Hypoperfusion

Previous studies have suggested that oxidative stress, concomitant with disturbed mitochondrial dysfunction [36] and ER stress [37], is one of the pathological pathways for chronic cerebral hypoperfusion-induced WMI [38, 39]. Based on these findings, we performed co-immunostaining of GFAP with specific organelle markers: Tom20 for mitochondria, GM130 for the Golgi apparatus, and GRP78 for the ER. Immunofluorescence analysis revealed an increased abundance of mitochondria, Golgi apparatus, and ER in the BCAS group (Supplementary Fig. 1A–F, all P < 0.01). To elucidate the ultrastructural characteristics of astrocytes in the corpus callosum under a hypoxia-ischemia context, we conducted detailed morphological analyses using the TEM. Representative images highlighting the distinguishing ultrastructural features of astrocytes, oligodendrocytes, and microglia are presented in Supplementary Fig. 2. Astrocytes in the BCAS group exhibited electron-dense morphology, characterized by darker cytoplasm and nucleoplasm, along with significant cellular enlargement (Fig. 2A). Quantitative analysis revealed marked increases in total cell area (101.60 ± 29.51 μm² vs. 48.50 ± 23.02 μm² in the sham group; P < 0.01; Fig. 2B), cytoplasmic area (56.69 ± 17.11 μm² vs. 15.63 ± 10.85 μm² in the sham group; P < 0.01; Fig. 2C), and the cytoplasmic-to-cell area ratio (55.76 ± 5.76% vs. 31.28 ± 9.40% in the sham group; P < 0.01; Fig. 2D), indicating substantial hypertrophic remodeling under chronic hypoxic-ischemic conditions. As well, the astrocytes in the BCAS group showed a significant increase in the number of mitochondria (15.75 ± 4.30 vs. 3.65 ± 2.52 in the sham group; P < 0.01; Fig. 2A, E). Additionally, these astrocytes contained more ultrastructurally altered mitochondria (14.80 ± 4.33 vs. 0.38 ± 0.77 in the sham group; P < 0.01; Fig. 2F, G) and typically elongated mitochondria (1.65 ± 1.05 vs. 0.13 ± 0.40; P < 0.01 in the sham group; Fig. 2H, I).

Fig. 2.

Mitochondria characterization of astrocytes in the corpus callosum with chronic hypoperfusion. (A) Representative TEM images of mitochondria within astrocytes. Left, the image of mitochondria within astrocytes in the intact corpus callosum; Right, the image in the demyelinated corpus callosum. Scale bar, 2 μm; high-magnification image, 500 nm. (B–D) Morphometric quantification of cellular parameters: (B) Cell body area, (C) Cytoplasmic area, and (D) Cytoplasmic-to-cell body area ratio. (E) Quantitative graphs representing the total number of mitochondria per astrocyte. (F, G) The number of altered mitochondria and the corresponding ratio over all mitochondria. (H, I) The number of elongated mitochondria and the corresponding ratio over all mitochondria. All data were presented as mean ± SD and compared by unpaired t-test; **P < 0.01 (n = 5, 40 astrocytes in each group). Red outline, cytoplasmic membrane; Green outline, nuclear membrane; Purple outline, non-altered mitochondria; Yellow outline, altered mitochondria; Red asterisk, elongated mitochondria. White arrow, intermediate filaments. Abbreviations: BCAS, bilateral carotid artery stenosis; TEM, transmission electron microscope; SD, standard deviation.

Moreover, the total number of Golgi apparatus was remarkably increased (4.85 ± 2.62 vs. 1.23 ± 0.48 in the sham group; P < 0.01; Fig. 3A, B). This increase was specifically attributed to the dilated Golgi apparatus (3.48 ± 1.91 vs. 0.00 ± 0.00 in the sham group; P < 0.01; Fig. 3C, D) because there was no significant difference in the number of non-dilated Golgi apparatus between the two groups (1.38 ± 1.13 vs. 1.23 ± 0.48 in the sham group; P = 0.44; Fig. 3E). We also observed an increase in the ER (15.85 ± 4.80 vs. 4.65 ± 2.48 in the sham group; P < 0.01; Fig. 4A, B) with more dilated cisternae (8.37 ± 3.81 vs. 0.18 ± 0.45 in the sham group; P < 0.01; Fig. 4C, D). Given the close relationship between organelles and their structural integrity, these findings indicate that astrocytes subjected to chronic hypoperfusion required enhanced energy production and metabolic activity.

Fig. 3.

Ultrastructure of the Golgi apparatus within astrocytes after chronic hypoperfusion. (A) Representative images of the Golgi apparatus within astrocytes. Scale bar, 2 μm; high-magnification image, 500 nm. (B) Quantitative analysis of the total number of Golgi apparatus per astrocyte. (C, D) The number of dilated Golgi apparatus and the corresponding ratio over all Golgi. (E) The number of non-dilated Golgi apparatus. All data were presented as mean ± SD and compared by unpaired t-test; **P < 0.01 (n = 5, 40 astrocytes in each group). Red outline, cytoplasmic membrane; Green outline, nuclear membrane; Blue pseudo-coloring, Golgi apparatus; Red asterisk, dilated Golgi apparatus. Abbreviations: BCAS, bilateral carotid artery stenosis; SD, standard deviation.

Fig. 4.

ER features of astrocytes in the corpus callosum with chronic hypoperfusion. (A) TEM images of ER within astrocytes in the sham and BCAS group, respectively. Scale bar, 2 μm; high-magnification image, 500 nm. (B) Quantitative graphs representing the total number of ER per astrocyte. (C, D) The number of dilated ER and the corresponding ratio over all ER. All data were presented as mean ± SD and compared by unpaired t-test; **P < 0.01 (n = 5, 40 astrocytes in each group). Red outline, cytoplasmic membrane; Green outline, nuclear membrane; Purple pseudo-coloring, ER; Red asterisk, dilated ER. Abbreviations: BCAS, bilateral carotid artery stenosis; ER, endoplasmic reticulum; TEM, transmission electron microscope; SD, standard deviation.

Reactive Astrocytes Contain More Phagocytic Inclusions and Exhibit Augmented Lysosomal Response after Chronic Hypoperfusion

Functioning as non-prototypical phagocytes, astrocytes can engulf various neural debris, such as degenerated axons, apoptotic neurons, and myelin breakdown products, into lysosomes for degradation in many pathological situations [14, 15, 40]. In the present study, the expression of LAMP1, a lysosomal-associated membrane protein, was higher in the BCAS group (Supplementary Fig. 3 A, B; P < 0.01). TEM revealed that the astrocyte could enwrap myelin-like structures in the cytoplasm in the BCAS group, while few or no phagosomes were detected in the sham group (Fig. 5A). Intracellular analysis of astrocytes showed that the lysosomal pathway was upregulated under hypoxic-ischemic conditions, with an increased number of lysosomes at various stages (Fig. 5B–G).

Fig. 5.

Phagocytic characteristics and the lysosomal compartments of astrocytes in the corpus callosum under chronic hypoperfusion. (A) Representative TEM images of astrocytic phagocytosis in the sham and BCAS groups. Left, the image of an astrocyte in the intact corpus callosum; Middle, the image of an astrocyte under hypoperfusion with a distinct phagosome; Right, the high-magnification image shown in the middle panel. Scale bar, 2 μm; high-magnification image, 500 nm. (B) The lysosomes at different stages within astrocytes in the intact and injured corpus callosum, respectively; Scale bar, 2 μm; high-magnification image, 500 nm. (C–G) Quantitative analysis of representative lysosome compartments between the two groups. (H–J) Proportions of primary, secondary, and tertiary lysosomes relative to total lysosome population (n = 5, 40 astrocytes in each group; mean ± SD; **P < 0.01; unpaired t-test). White outline, cytoplasmic membrane; Green pseudo-coloring, nucleus; Red arrow, myelin-like phagosome; Purple pseudo-coloring, primary lysosome; Orange pseudo-coloring, secondary lysosome; Blue pseudo-coloring, tertiary lysosome; Yellow pseudo-coloring, multivesicular bodies. Abbreviations: BCAS, bilateral carotid artery stenosis; TEM, transmission electron microscope; SD, standard deviation.

Primary lysosomes have been described as round and homogeneous structures containing electron-dense granules. In contrast, secondary lysosomes are larger and frequently contain empty phagosomes [30]. Tertiary lysosomes exhibit morphologically heterogeneous and often contain large lipid bodies [30, 31]. Based on these characteristics, the proportions of primary (32.43 ± 8.25% vs. 30.71 ± 28.55% in the sham group; P = 0.72; Fig. 5H) and secondary lysosomes (40.54 ± 11.54% vs. 39.54 ± 29.04% in the sham group; P = 0.84; Fig. 5I) remained comparable between groups, whereas the BCAS group displayed a higher ratio of tertiary lysosomes compared to the sham group (11.35 ± 8.14% vs. 4.42 ± 12.09% in the sham group; P < 0.01; Fig. 5J).

These findings indicate that chronic hypoxia-ischemia might induce compensatory lysosomal expansion in reactive astrocytes without commensurate functional enhancement, potentially resulting in compromised clearance efficiency.

Conditional knockout of astrocytic Cav-1 prompts adaptive cellular remodeling

To investigate whether Cav-1 mediates astrocytic stress responses, we analyzed the cellular contents of Cav-1-cKO mice ultrastructurally. Immunostaining conducted on the corpus callosum of Cav-1-cKO mice confirmed significant knockout of astrocytic Cav-1 (Supplementary Fig. 4 A). Knockout efficiency was as high as 94.6% (Supplementary Fig. 4B; P < 0.01). Under physiological conditions, astrocytes of Cav-1-cKO mice presented hypertrophic cell bodies (Supplementary Fig. 4 C). The enlarged cell body was possibly caused by increased cytoplasm because the cytoplasmic-to-cell area ratio was higher in Cav-1-cKO astrocytes (Supplementary Fig. 4D–F). Accordingly, the organelles, such as mitochondria, ER, and lysosomes within the Cav-1-cKO astrocytic cytoplasm were more than those in the wildtype (WT) astrocytes (Supplementary Fig. 4 C, G–J). These alterations suggest that astrocyte-specific knockout of Cav-1 may trigger compensatory cellular remodeling.

Conditional knockout of astrocytic Cav-1 decreases ultrastructural markers of cellular stress in astrocytes after chronic hypoperfusion

We next extended our ultrastructural analysis for stress markers in Cav-1^−/− astrocytes following BCAS. The quantitative evaluation revealed a significant increase in mitochondria of Cav-1-cKO BCAS astrocytes compared to the WT BCAS group (20.90 ± 7.24 vs. 16.40 ± 4.47; P < 0.01; Fig. 6A, B). This difference was primarily attributed to a marked elevation in non-altered mitochondria (16.25 ± 6.35 vs. 1.55 ± 1.04 in the WT BCAS group; P < 0.01; Fig. 6C–H).

Fig. 6.

Intracellular contents of astrocytes in WT and Cav-1-cKO mice under chronic hypoperfusion. (A) Representative images of the mitochondria and Golgi apparatus in the WT and Cav-1-cKO mice after BCAS surgery. Scale bar, 2 μm; high-magnification image, 500 nm. (B–H) Quantitative analysis of the total number of mitochondria and the difference in their classification between the two groups. (I–M) Quantitative graphs representing the total number of Golgi apparatus (I), the non-dilated Golgi apparatus (J), the ratio of non-dilated overall Golgi apparatus (K), the dilated Golgi apparatus (L), and the ratio of dilated overall Golgi apparatus (M). (n = 5, 40 astrocytes in each group; mean ± SD; **P < 0.01; unpaired t-test). White outline, cytoplasmic membrane; Green pseudo-coloring, nucleus; Orange pseudo-coloring, mitochondria; Purple pseudo-coloring, Golgi apparatus. Abbreviations: WT, wildtype; BCAS, bilateral carotid artery stenosis; Cav-1, caveolin-1; SD, standard deviation.

Although the total number of Golgi apparatus showed no significant difference between the two groups (6.20 ± 3.17 vs. 5.08 ± 2.19 in the WT BCAS group; P = 0.07; Fig. 6A, I), Cav-1-cKO BCAS astrocytes exhibited a higher number (4.37 ± 2.18 vs. 1.43 ± 1.01 in the WT BCAS group; P < 0.01; Fig. 6J) and percentage of non-dilated Golgi apparatus (72.00 ± 18.56% vs. 25.13 ± 14.45% in the WT BCAS group; P < 0.01; Fig. 6K). Conversely, the number of dilated Golgi apparatus (1.81 ± 1.40 vs. 3.65 ± 1.44 in the WT BCAS group; P < 0.01; Fig. 6L) and its percentage (28.00 ± 18.56% vs. 74.87 ± 14.45% in the WT BCAS group; P < 0.01; Fig. 6M) were significantly lower in Cav-1-cKO BCAS group.

We also quantified the number of ER within the cytoplasm of astrocytes and found that there was a significant increase in ER density in Cav-1-cKO BCAS astrocytes compared to the WT BCAS group (28.03 ± 5.69 vs. 16.58 ± 4.00, P < 0.01; Fig. 7A, B). Both the non-dilated (16.18 ± 3.49 vs. 7.95 ± 2.44 in the WT BCAS group, P < 0.01; Fig. 7C) and dilated ER (11.85 ± 3.91 vs. 8.63 ± 2.91 in the WT BCAS group; P < 0.01; Fig. 7D) were elevated in the Cav-1-cKO BCAS group. To determine if this increase was due to a larger cytoplasmic area in Cav-1-cKO BCAS astrocytes, we further investigated the ratio of corresponding ER profiles. These analyses demonstrated a marked shift in ER subpopulation distribution, with Cav-1-cKO BCAS astrocytes exhibiting a higher proportion of non-dilated ER (58.24 ± 8.83% vs. 48.75 ± 11.52% in the WT BCAS group, P < 0.01; Fig. 7E) and a corresponding reduction in dilated ER components (41.76 ± 8.83% vs. 51.25 ± 11.52% in the WT BCAS group, P < 0.01; Fig. 7F). Overall, these changes in the ultrastructural feature of cellular stress indicate that the adaptive remodeling for Cav-1 knockout under physiological conditions may improve astrocytic tolerance to chronic ischemic stress.

Fig. 7.

Ultrastructural characteristics of ER in WT and Cav-1-cKO mice under chronic hypoperfusion. (A) TEM images of ER within astrocytes of the WT and Cav-1-cKO mice after BCAS surgery. Scale bar, 2 μm; high-magnification image, 500 nm. (B–F) Quantitative graphs representing the difference in ER between the two groups (n = 5, 40 astrocytes in each group; mean ± SD; **P < 0.01; unpaired t-test). White outline, cytoplasmic membrane; Green pseudo-coloring, nucleus; Orange pseudo-coloring, ER. Abbreviations: WT, wildtype; BCAS, bilateral carotid artery stenosis; Cav-1, caveolin-1; ER, endoplasmic reticulum; TEM, transmission electron microscope; SD, standard deviation.

Conditional knockout of astrocytic Cav-1 enhances lysosomal degradation in astrocytes

Our analysis of intracellular contents in astrocytes following BCAS revealed a significant reduction in the total lysosomal counts in Cav-1-cKO mice compared to WT controls (7.98 ± 2.11 vs. 9.75 ± 2.84, respectively; P < 0.01; Fig. 8A–C). Further characterization of lysosomal subpopulations demonstrated a distinct shift in the lysosomal distribution of Cav-1-cKO astrocytes (Fig. 8D–G), marked by a decreased proportion of primary (19.78 ± 7.52% vs. 31.48 ± 9.48% in the WT BCAS group; P < 0.01; Fig. 8H) and tertiary lysosomes (7.14 ± 6.10% vs. 11.74 ± 8.35% in the WT BCAS group; P < 0.01; Fig. 8J) but an increased proportion of secondary lysosomes (46.68 ± 6.56% vs. 41.69 ± 8.35% in the WT BCAS group; P < 0.01 Fig. 8I) and multivesicular bodies (26.40 ± 9.09% vs. 15.09 ± 8.64% in the WT BCAS group; P < 0.01; Fig. 8K). Collectively, these data may imply an upregulation of lysosomal degradation, potentially driven by enhanced phagosome-lysosome fusion.

Fig. 8.

The lysosomal profiles of astrocytes in WT and Cav-1-cKO mice under chronic hypoperfusion. (A, B) Representative images of lysosome compartments of astrocytes in the WT BCAS and Cav-1-cKO BCAS group. Scale bar, 2 μm; high-magnification image, 500 nm. (C–G) Quantitative analysis of the number of lysosome compartments of astrocytes between the two groups. (H–K) Quantitative analysis of the ratio of lysosome compartments of astrocytes between the two groups. (n = 5, 40 astrocytes in each group; mean ± SD; **P < 0.01; unpaired t-test). White outline, cytoplasmic membrane; Green pseudo-coloring, nucleus; Purple pseudo-coloring, primary lysosome; Orange pseudo-coloring, secondary lysosome; Blue pseudo-coloring, tertiary lysosome; Yellow pseudo-coloring, multivesicular bodies. Abbreviations: WT, wildtype; BCAS, bilateral carotid artery stenosis; Cav-1, caveolin-1; SD, standard deviation.

Conditional knockout of astrocytic Cav-1 does not affect astrocytic activation and phagocytic transformation

To assess whether Cav-1 knockout influences astrocyte activation and phagocytic activity, we performed immunostaining of GFAP with Galectin-3 (a phagocytic marker) for astrocytes in the corpus callosum. Quantitative analysis revealed that there was no significant difference in the proportion of GFAP⁺ astrocytes between the two genotypes following chronic hypoperfusion (Supplementary Fig. 5 A, B). As well, Cav-1 deficiency showed minimal impact on astrocytic phagocytic transformation (Supplementary Fig. 5 A, C).

Conditional knockout of astrocytic Cav-1 ameliorates demyelination after chronic hypoperfusion

We then asked whether astrocytic ultrastructural alterations were correlated with myelin integrity. As for myelination in the corpus callosum, immunofluorescence analysis revealed a marked increase in MBP intensity in Cav-1-cKO mice subjected to BCAS (Supplementary Fig. 5D, E), indicating astrocytic Cav-1 deficiency mitigated demyelination after chronic cerebral hypoperfusion.

Discussion

Sustained cerebral hypoperfusion, along with the subsequent progression of WMI, has been identified as a key contributor to the decline in executive function and memory [41]. As the most numerous glial cells in the CNS, astrocytes have been reported to play an important role in the pathogenesis of WMI [15, 42]. However, research examining astrocytic ultrastructure under pathophysiological conditions remains limited. In the present study, we investigated the ultrastructural features of astrocytes in the corpus callosum following chronic cerebral hypoperfusion. We found that astrocytes were activated under the hypoxia-ischemia context, with a darker cytoplasm and nucleoplasm, as well as significant cellular hypertrophy. Additionally, these cells displayed markers of cellular stress, such as structurally altered mitochondria, expanded Golgi apparatus, and dilated ER. Notably, we also observed enhanced lysosomal activity after hypoperfusion. Interestingly, conditional knockout of astrocytic Cav-1 promoted adaptive cellular remodeling, marked by an increase in intracellular organelles without ultrastructural abnormalities, which further improved astrocytic tolerance to chronic hypoxic stress. Meanwhile, the phagosome-lysosome fusion was increased in Cav-1-deficient astrocytes, which may indicate an upregulation of degradation function, as the activation and phagocytosis activity were not affected.

Mitochondrial and metabolic dysfunction have been found to be associated with neuropathologies, such as Alzheimer’s disease and vascular dementia [36, 43]. In a rat model of carotid artery occlusion, mitochondria in microglia were swollen with abnormal cristae [36], which were similar to our study in that the number of altered and elongated mitochondria of astrocytes in mice was significantly increased after BCAS surgery. However, the effects of Cav-1 in regulating mitochondrial numbers are contradictory. Supported by our data, both the number of mitochondria and the percentage of non-dilated mitochondria were elevated in the Cav-1-deficient astrocytes. However, another recent study suggested that the knockdown of Cav-1 increased the fission process and decreased the ER-mitochondrial distance, without changing the metabolic demand and mitochondrial biogenesis in hepatic stellate cells [44]. This discrepancy may partially originate from cell-type specificity and distinct in vivo versus in vitro knockout paradigms.

Our findings also revealed the alteration of morphology in the Golgi apparatus of astrocytes, with dilated cisternae and enhanced vesicular budding subjected to chronic hypoperfusion. In Cav-1-deficient astrocytes, the number of non-dilated Golgi apparatus was increased without ultrastructural alteration. Accounting for the energy and nutrient deprivation, the structure and function of the Golgi apparatus could be impaired [29], which may be a key contributor to the pathological alteration of the Golgi apparatus in chronic hypoxia-ischemia. Furthermore, the Golgi apparatus is a membrane-bound organelle that maintains cellular homeostasis and communication by vesicle transport [29, 45]. The enhanced vesicular budding may be a compensatory mechanism to maintain secretory functions despite cellular stress. Cav-1 is a scaffolding protein that constitutes the major structural protein of caveolae as well as the transport vesicles derived from the trans-Golgi network [46, 47]. The morphology of the Golgi apparatus was not affected in the Cav-1-deficient astrocytes, possibly due to the adaptive cellular remodeling.

In addition, ER is the largest, continuous membrane-bound organelle within the cell, mainly responsible for Ca²⁺ handling and protein synthesis and trafficking [48]. The morphology of ER is tightly coupled to its function, because it needs to maintain a dynamic structure to contact with other organelles, such as mitochondria [49]. In the ultrastructural insight, the earliest ER dilation occurs in cytoplasmic tubules, which then develop into membrane blebs and swell, eventually affecting the ER segments connected to the nuclear envelope [50]. In our study, the number of ER with a higher ratio of dilated cisternae was remarkably increased following chronic hypoperfusion, which may be an indicator of the severity of neuroinflammation. Though ER stress is reported as a compensatory adaptive response to ischemic precondition, chronic ER stress can play a detrimental role by triggering apoptotic pathways [51, 52]. Notably, the ratio of non-dilated ER was higher in the Cav-1-cKO astrocytes, even though the number of both the non-dilated and dilated ER was increased after chronic hypoxia-ischemia. While previous studies have established Cav-1 biosynthesis within the ER and its modulation of cholesterol homeostasis through ER interactions [21], our findings demonstrate the Cav-1-mediated regulation of ER ultrastructural architecture, significantly expanding the functional spectrum of Cav-1/ER interplay.

Importantly, we found a significant upregulation of the lysosome pathway in astrocytes following chronic ischemic injury. This is in line with previous studies that the lysosome pathway is enhanced to degrade various components, including degenerated axons and myelin debris internalized by astrocytes under pathological conditions [14, 15]. In Cav-1-cKO astrocytes, lysosomal subpopulations were shifted with a lower level of primary and tertiary lysosomes and a higher proportion of secondary lysosomes. As well, Cav-1 deficiency showed minimal impact on astrocytic activation and phagocytic function. A recent study suggests the involvement of Cav-1 in the lysosomal degradation pathway, where Cav-1 deficiency enhanced lysosomal activity in breast cancer models [53]. Considering secondary lysosomes resulting from the fusion of primary lysosomes with target substrates, including disintegrated intracellular organelles and phagosomes [31], and their incomplete degradation formed tertiary lysosomes [54], we further inferred that both phagosome-lysosome fusion and substrate degradation were enhanced in Cav-1-cKO astrocytes. Uncovering the mechanism of Cav-1 in modulating the lysosome function would be a potential therapeutic target for chronic WMI.

Our ultrastructural analysis unveiled chronic hypoperfusion-induced astrocytic stress hallmarks, including altered mitochondrial structure, marked ER/Golgi dilation, and lysosomal pathway upregulation. Astrocyte-specific Cav-1 knockout induced adaptive organellar remodeling that enhanced ischemic tolerance through bolstered lysosomal functional capacity. Collectively, the present study defines quantitative ultrastructural signatures of ischemic astrocytes and identifies Cav-1 as a critical regulator of chronic WMI pathogenesis.

Supplementary Information

Below is the link to the electronic supplementary material.

Author Contributions

Conceptualization: Y.X., X.L., and Z.W.Methodology: M.W., X.Z., and D.W.Experiments: M.W., X.Z., X.S.Data analysis: D.C. and Y.Z.Supervision: Q.L., Y.L., and Z.H.Writing manuscript: M.W., X.Z., and J.G.Manuscript review/edits: Y.X., X.L., Z.W., and W.Z.

Funding

This work was supported by National Natural Science Foundation of China (NO. 82471359 to Y.X., 82371311 to X.Z., U22A20341 and 82471341 to X.L.), Natural Science Foundation of Jiangsu Province (BK20221553 to Y.X.), China Postdoctoral Science Foundation (NO. 2023M731746 to X.Z.), Nanjing Medical Science and Technology Development Foundation for Distinguished Young Scholars (NO. JQX23006 to X.Z.).

Data Availability

All data associated with this study are present in the paper or the Supplementary Materials. Requests for materials should be addressed to X.Y. (xieyi@njmu.edu.cn) or X.L. (xfliu2@vip.163.com).

Declarations

Ethics Approval

All animal experiments were performed according to the 2011 Eighth Edition of the National Institute of Health Guide for the Care and Use of Laboratory Animals and Institutional Animal Care and Use Committee of Jinling Hospital.

Competing interests

The authors declare no competing interests.