Endothelial PAI-1 Drives Lead-Induced Cerebral Amyloid Angiopathy via Activation of C3⁺ Decorin⁺ A1-like Astrocytes

Simple Summary

Lead (Pb) exposure is increasingly recognized as a risk factor for Alzheimer’s disease (AD) and cerebral amyloid angiopathy (CAA), yet how Pb contributes to these diseases is not fully understood. In this study, we discovered that Pb exposure activates endothelial cells in the brain to produce high levels of plasminogen activator inhibitor-1 (PAI-1), a molecule involved in vascular inflammation and extracellular matrix remodeling. We found the endothelial PAI-1 could trigger astrocytes to be transformed into C3⁺ decorin⁺ A1-like astrocytes, a subtype A1-like reactive astrocyte linked to CAA, using both an in vitro cell culture and an in vivo APP/PS1 mouse model. Importantly, the inhibition of PAI-1 significantly reduced the formation of these reactive astrocytes and mitigated Pb-induced pathology. Our newly identified PAI-1-endothelial-astrocytic pathway offers a promising therapeutic target for Pb-induced CAA and other types of CAA, and even AD.

Abstract

Environmental lead (Pb) exposure remains a significant public health concern, and its association with cerebrovascular injury and Alzheimer’s disease (AD) is increasingly recognized. In this study, we demonstrated using an in vitro system that Pb exposure significantly increased the expression and release of endothelial plasminogen activator inhibitor-1 (PAI-1). A conditioned medium collected from Pb-treated endothelial cells induced the formation of complement component 3 (C3)⁺ decorin⁺ A1-like astrocytes, which had been shown to be specifically associated with vascular amyloid. Immunoprecipitation with the PAI-1 antibody to remove PAI-1 from the culture medium, or treatment of endothelial cells with PAI-1 inhibitors, significantly inhibited the formation of C3⁺ decorin⁺ A1-like astrocytes. Furthermore, in vivo studies further supported this finding, indicating that lead does indeed increase the number of perivascular C3⁺ decorin⁺ A1-like astrocytes, and that the PAI-1 inhibitor blocked this induction. Building upon our previous findings, we demonstrate that lead exposure may induce cerebral amyloid angiopathy (CAA) pathology through the formation of C3⁺ decorin⁺ A1-like astrocytes mediated by endothelial cell PAI-1. Our results strongly suggest that PAI-1 is a key mediator linking endothelial stress and lead-induced vascular amyloidosis pathology.

1. Introduction

Lead (Pb) exposure remains a persistent environmental and public health threat, with widespread presence in air, drinking water [1,2], cigarettes [3], household products, paint, and industrial emissions [4]. Over 60 million people over 65 in 2024 [5] are affected by excessive historical Pb exposure. Recently, epidemiological studies [6,7,8,9,10] have shown that bone Pb (with a half-life of 30 years) [11,12] is associated with accelerated cognitive decline and an increased risk of dementia, especially in individuals carrying the APOE-ε4 allele [13]. Additionally, adults who lived in regions with historically higher atmospheric Pb levels, or near Pb-emitting industrial facilities, show a significantly higher prevalence of memory impairment decades later [14]. Furthermore, Pb has been detected within diffuse neurofibrillary tangles and senile plaques in brains with Alzheimer’s disease (AD) [15,16,17]. However, how Pb exposure participates in the pathogenesis of dementia and AD remains unclear. Previous studies have shown that early Pb exposure predetermines the regulation of amyloid precursor protein (APP) expression, and prenatal Pb exposure may increase APP gene expression and amyloid formation in old age [16,18]. In addition, we and other researchers have demonstrated that Pb exposure can impair the blood–brain barrier (BBB) function in young animals, thereby inducing lifelong amyloid pathology [19]. Most recently, we found that Pb exposure specifically accelerates the formation of vascular amyloid deposits in young mice, which in turn leads to cerebral amyloid angiopathy (CAA) pathology [20]. It is increasingly recognized that CAA is a contributing factor to the pathology of AD, as CAA can be observed in more than 80% of AD cases [21,22]. In particular, CAA is closely associated with a heavier burden of parenchymal amyloid and tau pathology in AD [23], and the coexistence of CAA with AD leads to more severe cognitive impairment than AD alone [24]. However, the pathogenesis of vascular amyloid and CAA is not fully understood. Most recently, consistent with a previous study showing that the plasminogen activator inhibitor type 1 (PAI-1) mechanism is specifically involved in CAA formation [25], we have found that inhibiting PAI-1 by its specific inhibitor, tiplaxtinin (TIP), markedly reduces Pb-induced vascular amyloid deposition and pathology [22]. PAI-1 is a key regulator of the plasminogen/plasmin system and is synthesized by a wide range of cell types, including vascular smooth muscle cells (VSMCs), endothelial cells, and brain parenchymal cells under pathological conditions [26,27,28]. Studies have found that PAI-1 levels are elevated in the brain tissue of humans with AD, and pharmacological inhibition or genetic ablation of PAI-1 reduces amyloid burden in mouse models [25,29,30,31,32,33]. However, how PAI-1 mechanistically drives the amyloid and CAA pathological process remains to be determined.

Complement component 3 (C3) A1 astrocytes are significantly induced in neurodegenerative disorders, including AD [34,35,36,37]. It has been reported that activated C3⁺ astrocytes are highly associated with vascular amyloid deposits without major microglial reactivity in CAA mice [38,39,40]. Importantly, a recent report indicates that the C3⁺ decorin⁺ A1-like astrocyte subtype activated by endothelial cells is specifically associated with vascular amyloids in the brains of APP mice and AD/CAA patients but not with parenchymal amyloids [41]. However, it remains unclear whether C3⁺ decorin⁺ A1-like astrocytes activated by endothelial cells are involved in Pb-mediated vascular amyloid and CAA pathology. Therefore, this study aimed to investigate whether Pb exposure, in addition to accelerating the formation of vascular amyloid, could directly activate endothelial cells, thereby inducing C3⁺ decorin⁺ A1-like astrocytes, and whether this process is mediated by the PAI-1 mechanism.

As previously described [20], to study the pathogenesis of AD and related dementias in specific high-risk populations (individuals over 75 years of age with a history of Pb exposure by 2050), we used 8-week-old mice exposed to 50 mg/kg Pb acetate (PbAc) for 8 consecutive weeks or in vitro cultured cells treated with 1 µM PbAc. The blood lead levels (BLL) in these mice were 29.5 ± 13.9 µg/dL (approximately 1.4 µM), which is consistent with the reference blood lead level of 30 µg/dL recommended by the U.S. Centers for Disease Control and Prevention (CDC) in 1975. Our data indicated that Pb exposure stimulated vascular amyloid deposition and significantly induced the number of C3⁺ decorin⁺ A1-like astrocytes in APP/PS1 mice. In vitro experiments showed that the Pb-treated endothelial cell culture medium could transform astrocytes into the C3⁺ decorin⁺ A1-like astrocytes. In addition, we found that Pb exposure stimulated the overexpression and release of PAI-1 in endothelial cells, and that removing PAI-1 from the Pb-treated cell culture medium by immunoprecipitation with a PAI-1 antibody significantly inhibited the formation of C3⁺ decorin⁺ A1-like astrocytes. Furthermore, the PAI-1 inhibitor TIP significantly suppressed the number of Pb-induced C3⁺ decorin⁺ A1-like astrocytes in vitro and in vivo.

2. Materials and Methods

2.1. Cultures of Endothelial and Astroglial Cells and Conditioned Media Collected from Endothelial Cell Culture

This study used human umbilical vein endothelial cells (HUVECs; PCS-100-013, <20 passages) and human SVG p12 astrocytes (CRL-8621) [42], as well as cells obtained from the mouse endothelial cell line bEnd.3 (CRL-2299) and the mouse astrocyte cell line C8-D1A (CRL-2541) (all purchased from ATCC, Manassas, VA, USA). In brief, cells were maintained in Dulbecco’s Modified Eagle’s Medium (DMEM), supplemented with 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin in a humidified incubator with 5% CO₂ at 37 °C. Cultures were passaged twice per week to ensure consistent growth. All supplies were purchased from Thermo Fisher Scientific Inc. (Waltham, MA, USA). Same amount HUVEC were cultured with either culture media (control group), 3 µM TIP (TIP group), 1 µM Pb (Pb group), or 1 µM PbAc in combination with 3 µM TIP (Pb-TIP group) to confluence in 100 mm culture dishes. Each dish was rinsed and replenished with 3 mL of DMEM for 24 h. The supernatant was then collected into a centrifuge tube and centrifuged at 1000× g for 10 min to remove cellular debris. HUVEC-conditioned media (HUVEC-CM) were collected and stored at −80 °C for use in the experiments. To measure the cell expression level of PAI-1, 2 × 10⁵ HUVECs were seeded in 12-well cell culture plates (n = 6 per condition) for 24 h and then treated with or without 1 μM Pb for 24 h (n = 6). At the end of the exposure period, cells were washed with PBS and lysed in RIPA buffer supplemented with protease inhibitors (1:25, Roche, Basel, Switzerland). Scraped cell extracts using cold plastic cell scrapers together with the RIPA buffer were collected in microcentrifuge tubes and incubated on ice for an additional 30 min. After centrifugation at 16,000× g for 10 min at 4 °C, levels of PAI-1 in the supernatant were assayed by Western blot. SVG p12 astrocyte viability was measured with MTT (3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyl-tetrazolium bromide, Sigma, St. Louis, MO, USA) assay. Briefly, cultured cells were incubated with MTT a final concentration of 0.25 mg/mL for 1 h at 37 °C to convert MTT into purple crystals. The cell culture media were then discarded and the cells were lysed with an equal volume of 100% DMSO. Absorbance was subsequently measured using a microplate reader (Molecular Device, Sunnyvale, CA, USA). Data were presented as a percentage of control [43].

2.2. Animals and Treatments

Heterozygous APP/PS1 transgenic mice on a mixed C57BL/6 and C3H background (Stock No. 004462, Jackson Laboratory, Bar Harbor, ME, USA) were maintained in the Laboratory Animal Resource Center (LARC) at Indiana University School of Medicine. Animals were group-housed (3–5 per cage) under standard laboratory conditions (12 h light/dark cycle, controlled temperature and humidity) with unrestricted access to food and water. The standard rodent diet (2018SX, Envigo, Indianapolis, IN, USA) was used for all groups. All reagents were prepared using double-distilled water, and solutions were adjusted to pH 7.2 prior to oral administration. Lead acetate (PbAc; 467863, Sigma–Aldrich, St. Louis, MO, USA) was stored at 4 °C and freshly prepared weekly. A custom-formulated diet supplemented with the PAI-1 inhibitor tiplaxtinin (12 mg/kg; BOC Sciences, Shirley, NY, USA) was used in a treatment group. The medicated diet was stored at 4 °C and replaced twice per week to ensure stability. As previously described and justified [20], APP/PS1 female mice at 8 weeks of age were administered either 50 mg/kg PbAc (i.e., 27 mg Pb/kg) or an equivalent molar concentration of Na-acetate (NaAc) via oral gavage once daily. Three groups of mice were exposed for 8 weeks for measurement of C3⁺ decorin⁺ A1-like astrocytes around brain vessels and vascular-bound decorin levels, NaAc exposed (Na, n = 5), PbAc exposed (Pb, n = 5), and PbAc exposed with TIP treatment (Pb + TIP, n = 5).

2.3. Immunodepletion of PAI-1 from HUVEC-CM

Immunodepletion of PAI-1 from HUVEC-CM was performed using the Pierce Immunoprecipitation Kit (26146, Thermo Fisher Scientific, Waltham, MA, USA) according to the manufacturer’s instructions. Protein concentrations of HUVEC-CM were quantified using a BCA assay (23227, Thermo Fisher Scientific, Waltham, MA, USA). For each immunodepletion reaction, 500 µg of total protein was used. Samples were first pre-cleared by incubation with settled control agarose resin for 30–60 min at 4 °C with gentle end-over-end mixing to minimize non-specific binding. After centrifugation at ~1000× g for 1 min, the supernatant was collected for immunodepletion. Pre-cleared HUVEC-CM was incubated with 5 µg of either a rabbit monoclonal IgG control antibody (ab172730, Abcam, Cambridge, UK) or anti-PAI-1 antibody (ab222754, Abcam, Cambridge, UK) overnight at 4 °C to allow immune complex formation. The antibody–antigen complex was then captured by incubation with protein A/G plus agarose in a spin column for 1 h with gentle mixing. After centrifugation, the flow-through was collected as the PAI-1–depleted fraction. PAI-1 depletion efficiency was confirmed by ELISA.

2.4. Quantification of PAI-1 by ELISA

The concentration of PAI-1 in HUVEC- HUVEC-CM was determined using a human PAI-1 sandwich ELISA kit (E-EL-H2104, Elabscience, Houston, TX, USA) according to the manufacturer’s instructions. Equal volumes (100 µL) of standards and samples were loaded into 96-well plates pre-coated with anti-PAI-1 antibody and allowed to bind for 90 min at room temperature (RT). Wells were then washed and sequentially incubated with biotin-conjugated secondary antibodies for 1 h and horseradish peroxidase (HRP)-linked reagents for an additional 30 min at RT. Colorimetric development was achieved using 3,3′,5,5′-tetramethylbenzidine (TMB) substrate. Absorbance was measured at 450 nm with a FlexStation 3 Multi-Mode Microplate Reader (Molecular Devices, San Jose, CA, USA). PAI-1 concentrations were calculated from a standard curve generated within the same assay plate.

2.5. Western Blot Analysis

HUVECs were lysed in RIPA buffer supplemented with protease inhibitors (1:25, Roche, Basel, Switzerland), followed by a 30 min incubation on ice to ensure complete protein extraction. Lysates were clarified by centrifugation at 16,000× g for 10 min at 4 °C, and total protein concentrations in the resulting supernatants were determined using a BCA Protein Assay Kit (Thermo Fisher Scientific, Waltham, MA, USA). Equal amounts of protein (10 µg per lane) were resolved on 4–12% Bis–Tris SDS-PAGE gels and transferred onto nitrocellulose membranes. After blocking with 5% skim milk for 2 h at RT, membranes were incubated overnight at 4 °C with primary antibodies against PAI-1 (rabbit, 1:1000, Abcam, ab222754) or GAPDH (mouse, 1:1000, Abcam, sc-32233). Following three washes in TBST, blots were incubated with IRDye 680RD goat anti-mouse or IRDye 800CW goat anti-rabbit secondary antibodies (1:5000; LI-COR Biosciences, Lincoln, NE, USA) for 1 h at RT. Fluorescent signals were visualized using an Odyssey Imaging System (LI-COR), and band intensities were quantified using Image J software (Fiji, NIH, Bethesda, MD, USA). PAI-1 expression was normalized to GAPDH and expressed as a protein ratio for statistical comparison.

2.6. Histology and Immunohistochemistry

Brains harvested from treated APP/PS1 mice were rapidly frozen in liquid nitrogen. The left hemisphere was sectioned sagittally using a vibratome, with a section thickness of 30 μm. Three sections from the left hemisphere (sections 13–15 based on the Allen Mouse Brain Sagittal Atlas), spaced 210 μm apart, were selected for analysis. No significant differences in plaque distribution were observed among these sections in our previous study [20]. Brain vessels were stained with an anti-collagen IV antibody (1:80 Invitrogen Waltham, MA, USA) for 24 h at 4 °C followed by an alkaline phosphatase (AP) goat anti-rabbit IgG antibody (1:1000, Abcam, Cambridge, UK) for 1 h at room temperature, and detected by AP staining kit (red) (Abcam, Cambridge, UK) following the manufacturer’s instructions. Vascular amyloid deposition was double stained with 1% Thioflavin S for 8 min, washed 2 times in 80% and 95% ethanol for 3 min, and then washed with distilled water 3 times. Slides were imaged using a high-content imaging system (ImageXpress Micro Confocal) equipped with Nikon objectives (10×/NA 0.45 and 60×/NA 0.94, Nikon, Melville, NY, USA). Confocal image stacks were acquired at 3 µm z-intervals and rendered using maximum intensity projections [20]. For each section, we randomly selected nine image regions in the cerebral cortex (cortical layer 2–5, Figure S1) of a control mouse, and then applied these same cortical regions to all mice to quantify vascular amyloid deposition and others. The total plaque number was calculated for each section, and values from three sections were averaged to generate a single data point for each mouse. A total of five mice per group were examined. All analyses were performed by an observer blinded to the treatment paradigm.

Brain sections and cultured SVG were also triple-stained overnight at 4 °C with primary antibodies: mouse anti-GFAP (1:400, MAB360, Sigma-Aldrich, St. Louis, MO, USA), goat anti-complement C3 (1:500, PA1-29715, Thermo Fisher Scientific), and rabbit anti-decorin (1:50, 14667, Thermo Fisher Scientific). After thorough washing, sections were incubated for 1 h at RT with species-specific secondary antibodies, including Alexa Fluor 488-conjugated anti-mouse (1:2000), Alexa Fluor 64-conjugated anti-goat (1:200), and Cy3-conjugated anti-rabbit (1:200) (all Thermo Fisher Scientific, Waltham, MA, USA). Triple-stained fluorescent images were captured using a Zeiss microscope (AX10, Carl Zeiss AG, Oberkochen, Germany), employing appropriate filters for different fluorescent dyes. Each dye was assigned to a separate channel to prevent bleed-through. Cells exhibiting simultaneous fluorescence signals of GFAP, C3, and decorin were identified as C3⁺ decorin⁺ A1-like astrocytes. Quantitative analysis was performed using ImageJ software (National Institutes of Health, Bethesda, MD, USA). For quantification of GFAP, C3, and decorin triple staining, mouse brain sections and image regions (Figure S2) were selected as described above. The average percentage was calculated for each section, and values from three sections were averaged to generate a single data point for each mouse. A total of five mice per group were examined. For in vitro experiments, three randomly selected fields per culture well were analyzed, and the mean value of these fields was used for quantification. All image acquisition and analyses were performed by an observer blinded to treatment conditions. A total of five mice per group were examined, and immunofluorescence experiments were independently repeated at least once.

2.7. Statistical Analysis

Data were analyzed using one-way ANOVA followed by Dunnett’s post hoc test for multiple group comparisons. Results are presented as mean ± standard deviation (SD). Statistical significance was defined as p ≤ 0.05. All analyses were performed using GraphPad Prism 10.6.1 software.

3. Results

3.1. Pb Exposure Increases Vascular Aβ Deposition in APP/PS1 Mice

To evaluate whether Pb exposure promotes vascular amyloid accumulation, we performed double labeling of cortical sections with Thioflavin S and collagen IV to visualize amyloid accumulation and brain vasculature, respectively (Figure 1A). In the control group APP/PS1 mice, vascular-associated amyloid-beta was present, but at very low levels. In contrast, mice exposed to Pb for two months showed a marked elevation in vascular Aβ deposition. Quantitative analysis revealed that the vascular amyloid burden in Pb-treated mice was significantly more than double compared with the Na-treated controls (26 ± 4.69 vs. 12.3 ± 3.1, p < 0.01) (Figure 1B). These findings indicate that Pb exposure accelerates vascular amyloid deposition in APP/PS1 mice.

3.2. Pb Exposure Stimulates Perivascular C3⁺ Decorin⁺ A1-like Astrocytes in APP/PS1 Mice

In CAA, amyloid progressively deposits along cerebral vasculature [44]. Studies suggest that the ensuing cerebrovascular dysfunction is driven primarily by endothelial signaling alterations rather than by the physical presence of amyloid alone [45]. Therefore, to determine whether Pb-associated vascular amyloid is accompanied by astrocytic reactivity, we performed triple immunofluorescence staining for GFAP (astrocytic marker), C3 (A1 reactive astrocyte marker), and decorin—an extracellular matrix (ECM) proteoglycan molecule involved in matrix remodeling that acts as a ligand for multiple cytokines and growth factors, enabling it to modulate signaling cascades and play a key role in proinflammatory processes [46] (Figure 2A). Under control conditions, only a very small proportion of astrocytes surrounding brain vessels expressed C3 or decorin. Pb exposure, however, resulted in a pronounced increase in astrocytes co-expressing C3 and decorin, forming a perivascular reactive glial niche. The proportion of C3⁺ decorin⁺ astrocytes increased nearly fivefold in Pb-treated animals relative to controls (25 ± 6.9% vs. 5 ± 2.1%, p < 0.01) (Figure 2B). These results suggest that Pb can stimulate perivascular C3⁺ decorin⁺ A1-like astrocytes and this process may be associated with vascular stress such as endothelial cell activation.

3.3. Endothelial Cells Exposed to Pb Promote C3⁺ Decorin⁺ A1 Astrocytic Phenotype in SVG Astrocytes

To investigate whether endothelial-derived signals stimulated by Pb are required to drive astrocytic activation, we treated human SVG astrocytes with conditioned media collected from HUVECs exposed to Pb for 24 h. The viability of HUVECs was measured and no difference was observed between control and Pb-treated cells (100 ± 2.3% VS. 99.7 ± 2.9%). We performed triple immunostaining for GFAP, C3, and decorin in human SVG astrocytes following 24 h exposure to HUVEC-CM. As expected, HUVEC-CM from Pb-treated HUVEC cells markedly increased the proportion of C3⁺ decorin⁺ A1-like astrocytes (24.2 ± 3.7% vs. 6.4 ± 0.7% in control, p < 0.05) (Figure 3A,B). We also used mouse endothelial cell line bEnd.3 and the astroglial cell line C8-D1A, and obtained similar results. HUVEC-CM collected from Pb-treated HUVEC cells markedly increased the proportion of C3⁺ decorin⁺ A1-like astrocytes from 6.2 ± 1.3% to 20.2 5.2% (p < 0.05). Cell viability in SVG cells treated with conditioned medium from Pb-treated HUVECs was 99.3 ± 1.7%, compared with 100.0 ± 4.5% in control cells. These data indicate that Pb-exposed endothelial cells indeed release soluble factors capable of reprogramming astrocytes toward an A1-like, decorin-enriched state similar to that observed in vivo.

3.4. Pb Exposure Elevates PAI-1 Levels in HUVEC Cells and Their Conditioned Media

Studies have shown that levels of PAI-1, a key regulator of fibrinolysis and ECM remodeling, are elevated in the brains of individuals with AD, and that either pharmacological inhibition or genetic deletion of PAI-1 reduces amyloid pathology in multiple mouse APP models [25,29,30,31,32,33]. Our previous work further demonstrated that PAI-1 played a specific and critical role in promoting CAA formation, and we have also shown that the pharmacological blockade of PAI-1 with its selective inhibitor tiplaxtinin (TIP) markedly reduces Pb-induced vascular amyloid deposition and associated pathology [20]. Thus, we investigated whether Pb exposure drove the upregulation of PAI-1 in endothelial cells. Using HUVEC cells, a well-characterized mouse endothelial cell line, we found that Pb exposure significantly elevated PAI-1 expression levels in HUVEC cells relative to control (1.29 ± 0.11 vs. 1.00 ± 0.09, p < 0.05) (Figure 4A,B). Consistently, HUVEC-CM collected from cells exposed to 1 µM PbAc also exhibited significantly higher PAI-1 levels—an increase of approximately 36.7%—compared with HUVEC-CM collected from Na-treated controls (76.33 ± 6.27 vs. 55.85 ± 3.58, p < 0.05) (Figure 4C). These results indicate that endothelial PAI-1 is a Pb-responsive factor.

3.5. PAI-1 Mediates Pb-Induced Activation of C3⁺ Decorin⁺ SVG Astrocytes

To test whether PAI-1 directly mediates Pb-induced activation of A1-like astrocytes, SVG cells were treated with HUVEC-CM collected from Pb-exposed endothelial cells, either after PAI-1 immunodepletion or in the presence of a PAI-1 inhibitor. Removal of PAI-1 by immunoprecipitation almost fully abolished the ability of Pb–HUVEC-CM to stimulate C3⁺ decorin⁺ astrocyte formation, reducing their proportion from 24.2 ± 3.7% to 4.5 ± 1.0% (p < 0.01) (Figure 5A,B). To confirm that PAI-1 was effectively removed, we immunoprecipitated HUVEC-CM using an anti-PAI-1 antibody, which eliminated more than 90% of PAI-1 from both control and Pb-treated samples (Figure 4C). Species- and isotype-matched rabbit monoclonal IgG was used as a negative control to account for non-specific binding during the immunoprecipitation procedure. IgG immunoprecipitation did not affect PAI-1 levels (Figure 4C) and no change in protein levels was observed. Similarly, treatment with the PAI-1 inhibitor TIP also significantly attenuated astrocytic activation, decreasing the population of C3⁺ decorin⁺ A1-like astrocytes from 24.2 ± 3.7% to 11.9 ± 2.5% (p < 0.05) (Figure 5A,B). Treatment with PAI-1-depleted or PAI-1-inhibited control HUVECs-CM did not change astrocytic activation (6.9 ± 1.5% and 6.7 ± 1.0%). Cell viability of SVG cells treated with PAI-1-depleted or PAI-1-inhibited Pb-treated HUVECs-CM was 97.8 ± 2.7% and 97.1 ± 1.1%, respectively. These findings demonstrate that endothelial-derived PAI-1 is a key player in regulating Pb-induced C3⁺ decorin⁺ astrocyte formation.

3.6. TIP Treatment Reduces Pb-Induced Perivascular C3⁺ Decorin⁺ A1-like Astrocytes in APP/PS1 Mice

Finally, to further confirm the crucial role of PAI-1 in Pb-induced C3⁺ decorin⁺ A1-like astrocyte formation, we exposed two-month-old female APP/PS1 mice to Pb and simultaneously treated them with the PAI-1 inhibitor TIP. Two months later, we performed triple immunofluorescence staining for GFAP, C3, and decorin on cortical sections obtained from these APP/PS1 mice. As expected, compared with the Pb-treated group, TIP treatment significantly reduced the percentage of perivascular C3⁺ decorin⁺ A1-like astrocytes (15 ± 3.6% vs. 25 ± 6.9% in Pb group, p < 0.05) (Figure 6A,B). These in vivo results confirmed our in vitro observations, indicating that Pb-induced C3⁺ decorin⁺ A1-like astrocyte activation is at least partially (if not entirely) mediated by the PAI-1 signaling pathway, and that this activation can be mitigated by pharmacological inhibition of PAI-1.

4. Discussion

This study reveals a novel pathogenic mechanism by which Pb induces the transformation of astrocytes into C3⁺ decorin⁺ A1-like astrocytes through the overexpression of PAI-1 in endothelial cells, thereby stimulating the formation of CAA. Our findings further confirm our previous research that Pb exposure increases CAA in APP/PS1 mice through the PAI-1 signaling pathway in cerebral blood vessels [20]. It is worth noting that we observed that most vascular amyloid depositions occurred in the cortex, while there were very little depositions in the hippocampus and other brain regions; therefore, we could not draw any conclusions about these areas. Given this, this paper only presents data from the cortex.

PAI-1 is a multifaceted regulator of vascular homeostasis that inhibits tissue-and urokinase-type plasminogen activators (tPA/uPA) and thereby limits plasmin formation. Excess PAI-1 expression leads to fibrinolytic impairment [47], endothelial dysfunction [48], and perivascular fibrosis [20]. In the CNS, elevated PAI-1 levels have been observed in the brains of human patients with AD and transgenic mouse models, and pharmacologic inhibition or genetic deletion of PAI-1 mitigates amyloid burden and improves vascular function [25,29,33]. In a previous study, we found that a Pb-induced elevation of cerebral vascular PAI-1 was accompanied by increased vascular Aβ binding affinity, reduced perivascular drainage, and reduced microglial endocytosis of Aβ, and that inhibiting PAI-1 with TIP significantly reduced Pb-induced vascular amyloid formation [20]. Interestingly, in the absence of Pb exposure, we observed no significant difference in PAI-1 expression levels between 4-month-old APP/PS1 mice and control mice. This is likely because the mice used in our study were relatively young and had minimal vascular amyloid deposition. In this report, we further expand on these studies, showing that Pb exposure significantly increases the expression of PAI-1 in vascular endothelial cells, and that endothelial PAI-1 plays a key role in the Pb-induced formation of C3⁺ decorin⁺ A1-like astrocytes (a subset of neurotoxic A1-like astrocytes characterized by Liddelow et al. [34,41,49]). In a CAA study, it was reported that C3⁺ decorin⁺ A1-like astrocytes were specifically associated with vascular amyloid deposits, but not with parenchymal amyloid plaques in mouse models and patients with AD/CAA [41]. In our study, using an in vitro system, we observed that the conditioned medium collected from Pb-treated human endothelial cells significantly induced the transformation of human astrocytes into C3⁺ decorin⁺ A1-like astrocytes. Notably, Pb exposure could stimulate the overexpression of PAI-1 and its release into the conditioned medium of endothelial cells. We therefore tried to investigate the role of endothelial PAI-1 in the formation of C3⁺ decorin⁺ A1-like astrocytes. As expected, the immunodepletion of PAI-1 from the culture medium almost completely blocked C3⁺ decorin⁺ A1-like astrocyte formation. Additionally, co-treatment with the PAI-1 inhibitor TIP also significantly inhibited this process. Furthermore, our in vivo studies confirmed the in vitro results, demonstrating that Pb exposure significantly increased the number of perivascular C3⁺ decorin⁺ A1-like astrocytes in APP/PS1 mice, and that PAI-1 inhibitors significantly suppressed this increase. Therefore, our results indicate that PAI-1 released by activated endothelial cells is a key paracrine mediator of CAA-associated astrocyte activation, even if not only during Pb exposure.

It is currently unclear whether endothelial-induced C3⁺ decorin⁺ A1-like astrocytes are the cause or the result of CAA formation. Our results show that Pb exposure can induce C3⁺ A1 decorin⁺ astrocytes independently of Aβ or amyloid, and inhibition of PAI-1 can block the formation of both C3⁺ decorin⁺ A1-like astrocytes and CAA. Pb is considered a toxic metal ion to the cardiovascular cellular system, especially because it can specifically target vascular endothelial cells in peripheral blood vessels and the blood–brain barrier (BBB). Therefore, it is reasonable to assume that Pb directly activates endothelial cells in the BBB, releasing PAI-1 along the perivascular space, which then transforms astrocytes into C3⁺ decorin⁺ A1-like astrocytes, thereby leading to CAA formation. Decorin is a leucine-rich proteoglycan and plays a key structural and regulatory role in the extracellular matrix (ECM) [46,50]. Elevated decorin has been detected in cerebrospinal fluid (CSF) from APP transgenic AD mouse models and human patients at the preclinical stage of AD [51]. Furthermore, since decorin contributes to ECM remodeling and vascular fibrosis [52,53,54,55], and vascular fibrosis and ECM thickening impede perivascular Aβ drainage [56,57], it is necessary to investigate whether perivascular C3⁺ decorin⁺ A1-like astrocytes participate in the formation of CAA by affecting perivascular Aβ clearance.

5. Conclusions

In summary, our research findings reveal a novel pathogenic mechanism: lead exposure enhances the paracrine signaling of PAI-1 in endothelial cells, thereby mediating the formation of C3⁺ decorin⁺ A1-like astrocytes associated with CAA. Combined with our previous research [20], these findings further confirm that PAI-1 is a key signal mediating Pb-induced CAA, and the resulting demyelination and cognitive decline. Therefore, identifying the PAI-1 endothelial-astrocytic pathway provides a novel therapeutic target for Pb-induced CAA and other types of CAA, and even AD.