Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2025 Aug 12.
Published in final edited form as: Am J Physiol Cell Physiol. 2025 Jun 23;329(2):C341–C354. doi: 10.1152/ajpcell.00615.2024

Glucose transporter 1 is essential to maintain brain endothelial cell homeostasis under hyperglycemia condition

Benjamin Miao 1,#, Mohammad Sarif Mohiuddin 1,#, Rashu Barua 1, Md Wahiduzzaman 1, Zhi Fang 1,2,3, Wenquan Hu 1,2, Munichandra Babu Tirumalasetty 1, Xiaoran Sun 1, Mayank Choubey 1, Qing Robert Miao 1,2
PMCID: PMC12341840  NIHMSID: NIHMS2094565  PMID: 40549566

Abstract

Diabetic patients are prone to developing cerebrovascular disease (CVD) due to a multitude of factors. Particularly, the hyperglycemic environment is a key contributor to the progression of diabetes-associated complications. However, there is a dearth of knowledge regarding glucose transporter 1 (GLUT1, also known as SLC2A1)-dependent mechanisms responsible for these adverse effects. Here, we revealed the importance of glucose transporter 1 in preserving brain endothelial cell homeostasis beyond regulating glucose uptake. To elucidate the GLUT1-mediated protective mechanism, we used bulk RNA-seq to analyze transcriptomic alterations under hyperglycemia and GLUT1 deficiency conditions and validated the critical gene changes in cultured human brain endothelial cells and diabetic mouse models. We found that GLUT1 downregulation is linked to increased expression levels of podocalyxin (PODXL) and decreased thioredoxin-interacting protein (TXNIP) within healthy brain endothelial cells incubated with high glucose, demonstrating an anti-stress response mechanism. Interestingly, brain endothelial cells isolated from diabetic mice no longer showed a similar protection mechanism. Instead, the diabetic endothelial cells are characterized by considerably enriched GLUT1 and TXNIP expression under a hyperglycemic state. GLUT1 overexpression recaptures the diabetic features, such as elevated expression of TXNIP and NOD-like receptor pyrin domain-containing 3 (NLRP3) inflammasome, along with increased IL-1β production and permeability. Our findings of a GLUT1-dependent regulatory mechanism for the endothelium provide a potentially deeper insight into mechanistic shifts that occur due to the diabetic disease state and the pathogenesis of diabetes-associated vascular complications.

Short Summary:

Glucose transporter-1 is known for regulating glucose uptake in brain endothelial cells. This study used global transcriptome analysis and diabetic mouse models to reveal the novel role of Glucose transporter-1 in regulating brain endothelial cell homeostasis by reducing the inflammation response and increasing the protection mechanism. Importantly, the Glucose transporter-1-dependent protection mechanism is compromised in diabetic conditions, which explains why diabetic patients have a high risk of cerebrovascular diseases.

Graphical Abstract

graphic file with name nihms-2094565-f0008.jpg

Introduction

Type 2 diabetes mellitus (T2DM) is a widespread epidemic in the world, accounting for around 90% of the total cases of diabetes (1). Further, T2DM is associated with a vast array of other severe complications, such as cerebrovascular diseases, nephropathy, thrombosis, and atherosclerosis, which presents clinical demands for further understanding of T2DM-associated complications (2). Specifically, in characterizing its environment, diabetes is known to be characterized by hyperglycemic conditions (3–5). These increased glucose concentrations have been associated with increases in inflammation and oxidative stress, which play critical roles in the comorbidities presented in T2DM patients (5, 6). Although progress has been made with regard to treatments for and overall knowledge of T2DM, much remains to be further investigated, with particular regard to the adverse effects of hyperglycemia-caused cerebrovascular diseases.

Endothelial cells (ECs) line the insides of blood vessels, serving as the first barrier between circulation and the vasculature (7). ECs are involved in many different functions, such as angiogenesis and hemostasis (7, 8). More importantly, the endothelium preserves vascular homeostasis by mediating glucose uptake and stress responses (9, 10). In particular, GLUT1 is responsible for transporting glucose within human brain microvascular endothelial cells (HBMVEC) and has been shown to play a vital role in maintaining the homeostasis of endothelial cells (10–12). Of note, in response to hyperglycemia, ECs protect the intracellular environment against a harmful surge of glucose by reducing expression levels of the integrated membrane protein, GLUT1, thwarting increases in oxidative stress and thrombotic activity (3, 4). These mechanistic shifts experienced under high glucose conditions illustrate the protective roles of GLUT1 in mediating stress responses.

Despite the seemingly protective effect involving the regulation of GLUT1 under hyperglycemic conditions, the diabetic model is interestingly characterized by increased oxidative stress and cytokine response (3). One key factor contributing to the stressed landscape is the hyperglycemic conditions associated with T2DM, resulting in a surplus of Advanced Glycation End Products (AGEs) and Reactive Oxygen Species (ROS) production (5). As a result, the hyperglycemic conditions of T2DM contribute to endothelial dysfunction, compromising endothelium homeostasis (5, 9). In particular, endothelial dysfunction triggers the activation of inflammatory responses and characteristic of oxidative stress, leading to the development and increased risk of diabetes-associated cerebrovascular disease (5, 6, 13, 14). However, the brain endothelial dysfunction-prone state attributed to hyperglycemia has yet to be characterized extensively.

In this study, we seek to further characterize the association of GLUT1 with anti-stress responses under high glucose (HG) conditions. Our findings demonstrate a relationship between GLUT1 and key anti-stress regulators in healthy endothelial cells under HG conditions, which is lost in the brain endothelial cells of diabetic mice. Thus, our data suggest that within brain microvascular endothelial cells, the GLUT1-mediated protective mechanism is essential to mitigate endothelial dysfunction, which ultimately may be compromised under diabetic conditions, leading to endothelial dysfunction via increases in inflammation and oxidative stress.

Materials and Methods

Cell culture, transfection, and reagents

Human brain microvascular endothelial cells (HBMVEC) were purchased from Cell Biologics (H-6023, Chicago, IL) and grown in complete human endothelial cell medium from Cell Biologics (H-1168, Chicago, IL). Glut-1 siRNA (sc-35493) and control siRNA (sc-37007) were purchased from Santa Cruz Biotechnology (Dallas, TX). Glut1 siRNA is a pool of 3 target-specific 19–25 nt siRNAs designed to knock down human GLUT1 gene expression. Control siRNA is a non-targeting 20–25 nt siRNA designed as a negative control. Cells were transfected with RNAiMax or Lipofectamine 2000 (Invitrogen) following the manufacturer’s protocol.

Seahorse

Human brain microvascular endothelial cells were seeded in seahorse XFe 96-well microplates (Agilent Technologies, Sana Clara, USA). Cells were washed and incubated in a base medium (Agilent Technologies) at 37 °C for 1 h. The basal oxygen consumption rate (OCR) and the OCR following the addition of oligomycin, FCCP, and rotenone/antimycin A were measured in real-time with Seahorse XF Cell Mito Stress Test Kit using the Seahorse XFe96 Analyzer following the manufacturer’s instructions.

RNA-seq analysis

Total RNA was isolated from HBMVECs using the RNeasy Mini Kit (74106, Qiagen), and quality was assessed by fragment analysis (Agilent). The qualified RNA samples of HBMVECs in four groups (incubated with either basal medium or plus 25 mM glucose, transfected with either control siRNA or Glut1 siRNA) were sent to the Genomic Science and Precision Medicine Center (GSPMC) at the Medical College of Wisconsin for RNA-seq on an Illumina HiSeq. Sequencing reads were processed through the MAPR-Seq workflow (https://bioinformaticstools.mayo.edu/research/maprseq/ ) with differential expression analysis completed with EdgeR software (http://www.bioconductor.org/packages/release/bioc/html/edgeR.html ) (RRID:SCR_012802). Data were deposited in the NCBI’s Gene Expression Omnibus database (GSE271637) (RRID:SCR_005012).

Reverse-transcription quantitative Polymerase Chain Reaction (RT-qPCR)

Total RNA was extracted from cells using the RNeasy kit (QIAGEN, Hilden, Germany). The cDNA was reverse-transcribed from 1 μg total RNA using the iScript cDNA synthesis kit (Bio-Rad, Los Angeles, CA, USA) according to the manufacturer’s instructions. The quantitative PCR was run on the Bio-Rad CFX Real-Time PCR Detection System using iTaq Universal SYBR Green Supermix (Bio-Rad, Los Angeles, CA, USA). The relative messenger RNA (mRNA) expression of each gene was normalized to the housekeeping gene (GAPDH) mRNA levels. THBD primers (sc-36686-PR and sc-36687-PR) were purchased from SCBT (Santa Cruz Biotechnology, Dallas TX, USA). The real-time PCR primers were synthesized by Integrated DNA Technologies (Coralville, IA, USA). Primer sequences are listed in Supplementary Table 1.

Gene Set Enrichment Analysis (GSEA)

Utilizing a gene-set enrichment analysis (GSEA) signal-to-noise ratio ranking metric, we ranked the enriched genes by their association in HBMVEC cells (siGlut1 vs. NS and HG+siGlut1 vs HG+NS) (15). NS is the control siRNA without silencing functions. Pre-ranked GSEA was performed using “Hallmark ‘‘ gene sets for a broad analysis of shifts within HBMVEC as well as in terms of gene expression regulation (16). GSEA analysis was conducted using GSEA V4.1.0 (RRID:SCR_005012). Geneset was considered substantial when the false discovery rate (FDR) was less than 0.25. The gene set enrichment analysis (GSEA) program and the R package fast gene set enrichment analysis (fgsea) (15, 17) were applied in our analysis. Visualizations were generated through GSEA (version 4.1.0), ggplot2 (version 3.4.4) (RRID:SCR_014601), ggcorrplot (version 0.1.4.1), and EnhancedVolcano (version 1.18.0) (15).

Diabetic Mouse models

Mice homozygous for diabetes spontaneous mutation (Leprdb) (db/db) (RRID:IMSR_JAX:000697) become obese at 3 to 4 weeks, with blood glucose rising at 4 to 8 weeks. Streptozotocin (STZ) dissolved in sodium citrate buffer (pH 4.5), was administered intraperitoneally to C57BL/6J male mice (RRID:IMSR_JAX:000664, 12 weeks old) at a daily dose of 75 mg/kg body weight for five consecutive days. Subjects were randomly assigned to experimental groups to minimize selection bias. The animals were then maintained for four weeks. Control animals received injections of citrate buffer. Body weight and fasting blood glucose concentration were monitored weekly throughout the experimental period. Mouse became diabetic one week after the STZ injections. Brain endothelial cells were isolated when the fasting blood glucose reached 350mg/dl following the method described in our previous publication (18). To reduce experimental bias, investigators were blinded to group assignments during data collection and analysis. Group sizes were determined based on the power analysis of our prior published studies (18).

Western blotting

Cellular protein was extracted using RIPA lysis buffer (Thermo Scientific) containing protease and phosphatase inhibitors tablets (Pierce-Thermo Scientific). The protein concentrations were determined using a detergent-compatible colorimetric protein assay (Bio-Rad, California, USA). Cellular protein was separated by 10% SDS-PAGE gel and transferred onto a nitrocellulose membrane. The membranes were further blocked by Tris Buffered Saline (TBS) with 1% (w/v) Casein buffer (Biorad, USA) and incubated with primary antibodies at 4 °C overnight. Washes were performed with TBS containing 0.1% Tween-20 (TBST) before the addition of the secondary antibody. Primary antibodies were detected using highly cross-adsorbed IRDye 800CW Goat (polyclonal) anti-Mouse IgG (H + L) (LI-COR #925–32210, RRID:AB-2687825) and IRDye 800CW Goat (polyclonal) anti-Rabbit IgG (H + L) (LI-COR #926–32211, RRID:AB_621843). All blots were imaged wet using the Odyssey® CLx imaging system, utilizing the 680 nm and 780 nm channels.

Cell Viability

Cell viability was assessed using an MTS assay (Cell Titer 96 Aqueous One Solution Cell Proliferation Assay, cat #3580, Promega Corp., USA), according to the manufacturer’s instructions. Briefly, at the end of the experiment, 20 μl of MTS reagent was added to the culture medium and incubated for 3 hours. Absorbance was measured at 490 nm with a microplate reader (Synergy LX Multi-Mode Microplate Reader, USA) to determine cell viability, which was expressed as a percentage of the control group.

Immunostaining

Primary HBMVEC cells were fixed with 1% paraformaldehyde in PBS for 15 minutes. Next, the cells were permeabilized with 0.1% Triton-X 100 in PBS for another 15 minutes and blocked with 5% BSA in PBS for 1 hour. The cells were incubated overnight at 4°C with the VE-Cadherin antibody (1:100) (Invitrogen, Cat#36–1900, RRID:AB_148487) diluted in 5% BSA/PBS. Following a wash with PBS, the cells were incubated for 1 hour at room temperature with Alexa Fluor™ Plus 488 conjugated goat anti-rabbit IgG antibody (1:1000) (Invitrogen, Cat#A32731, RRID:AB_2633280), diluted in 5% BSA/PBS. Subsequently, the cells were counter-stained with DAPI (Southern Biotech, Birmingham, AL, USA). All slides were examined using a Keyence (BZ-X800 series).

HBMVEC monolayer permeability assay

The permeability of the HBMVEC monolayer cells was assessed by measuring the penetration rate of FITC-dextran (FD40S, Sigma-Aldrich) through the monolayer cells, following a previous report (19). Briefly, HBMVECs from different groups were seeded to confluence on Transwell inserts (24-well format; 0.4 μm pore; 3470, Corning). Fresh culture medium (200 μL) containing 100 μg/mL FITC-dextran was added to the upper chamber, while 1 mL of fresh culture medium without FITC-dextran was added to the lower chamber. Samples (100 μL) were collected from the lower chamber at 0 and 60 minutes, and the medium was replaced with fresh culture medium. The samples were then analyzed using a fluorescence microplate reader set to a wavelength of 488/510 nm (ex/em). The permeability coefficient (cm/min) was determined using the formula V/(SA × Cd) × (Cr/T), where V represents the medium volume in the receiver chamber, SA is the surface area of the cell monolayer, Cd is the concentration of FITC-dextran in the donor chamber at time 0, and Cr is the concentration of FITC-dextran in the receiver chamber at sampling time T. Changes in permeability are expressed as a percentage of control.

Lentiviral overexpression

Lentivirus expressing Glut1 (NM_006516.4) (EX-C0124-Lv242, Genecopoeia) and control lentivirus lacking a transgene were generated using a second-generation lentivirus packaging system with the packaging plasmid psPAX2 (12260, Addgene; RRID:Addgene_12260) and the envelope plasmid pVSVG (12259, Addgene; RRID:Addgene_12259). Briefly, lentiviral plasmids were co-transfected with psPAX2 and pVSVG virus-packaging plasmids in Lenti-X™ 293T Cells (Takara, Japan, cat# 632180) using Lipofectamine 3000. After 12 hours, the culture media were replaced with a high FBS-containing medium without antibiotics. After 24 and 48 hours, the lentivirus supernatant was collected, centrifuged at 12,000 rpm for 5 minutes, and filtered using a 0.45 μm filter. After filtration, the virus was concentrated using the LentiX concentrator (Takara, cat# 631231) according to the manufacturer’s protocol. Subsequently, the HBMVEC cells were infected with the lentiviral supernatants. After 24 hours, cells were selected with 1μg/ml puromycin for 3–4 days. The selected cell populations were allowed to grow back to confluency.

Superoxide detection by fluorescent microscopy

Primary human brain endothelial cells (HBMVECs) were seeded onto collagen-coated 24-well plates and incubated at 37°C with 5% CO2 overnight until they reached 70–80% confluence. After incubation, the cells were treated with 5 μM Dihydroethidine (DHE) (Invitrogen, Cat#D1168) for 30 minutes at 37°C to assess superoxide production. Following this, the cells were washed with PBS to remove any excess dye. The cells were then examined using Keyence (BZ-X800 series).

Enzyme-Linked Immunosorbent Assay (ELISA)

Interleukin-1 beta (IL-1β) released by HBMVECs into the culture medium was quantified using the Human IL-1β Sandwich ELISA Kit (Proteintech, Cat#KE00021) following the manufacturer’s protocol. Briefly, 100 μL of standards and samples were added to the designated wells of a 96-well microplate and incubated at 37°C for 2 hours. After incubation, the wells were washed four times, followed by the addition of 100 μL of 1× Detection Antibody Solution. The plate was then incubated at 37°C for 1 hour, followed by another wash step. Next, 100 μL of 1× Streptavidin-HRP Solution was added and incubated at 37°C for 40 minutes, followed by a final wash step. For signal development, 100 μL of TMB Substrate Solution was added to each well, and the plate was incubated in the dark for 15–20 minutes. The reaction was stopped by adding 100 μL of Stop Solution, and the plate was gently tapped to ensure proper mixing. Absorbance was immediately measured at 450 nm.

Statistical analysis

All data are presented as means ± SD. The Student’s t-test was used to compare differences between the two groups. For experiments involving more than two groups, one-way ANOVA was employed, followed by Tukey’s post hoc test to identify specific group differences. All analyses were conducted using GraphPad Prism 8 (GraphPad Software, San Diego, CA) (RRID:SCR_002798), which ensured rigorous evaluation of the experimental data and accounted for potential multiple comparisons.

Results

GLUT1 expression is downregulated under high glucose conditions and results in substantial transcriptional alterations in brain endothelial cells

When culturing human brain microvascular endothelial cells (HBMVECs) within a hyperglycemic medium, we observed a notable decrease in cell viability in a glucose concentration-dependent manner, indicating the presence of a cytotoxic environment (Fig 1A). Furthermore, GLUT1 expression decreases accordingly to increased glucose concentration (Fig 1B). To determine the effects of GLUT1 loss on cell viability, validated small interfering RNAs (siRNAs) targeting human GLUT1 were implemented to knock down GLUT1 expression in HBMVECs. The efficacy of siRNAs in knocking down GLUT1 was demonstrated in Supplementary Figure S1. The quantification of GAPDH housekeeping is shown in Supplementary Figure S2. High glucose and GLUT1 knockdown do not affect the expression of GAPDH (Fig. S2). As expected, GLUT1 deficiency-reduced glucose consumption resulted in the associated reduction of mitochondrial respiration determined by Seahorse Real-time Cell Metabolic Analysis (Fig 1C). However, of importance, as shown in (Fig 1A), GLUT1 knockdown does not affect the cell viability. Rather, high glucose is responsible for the decreased viability through GLUT1-independent mechanisms.

Figure 1:

Figure 1:

Open in a new tab

High glucose incubation decreases HBMVEC cell viability and GLUT1 transcription. (A) An assessment of cell viability under high glucose conditions and GLUT1 knockdown (siGlut1) reveals that reduced viability correlates with increased glucose concentrations, whereas GLUT1 knockdown does not influence cell viability. (B) GLUT1 transcription levels decline in a manner dependent on glucose concentration. (C) Seahorse metabolic analysis indicated that siGlut1 knockdown decreased respiratory productivity. Data are presented as mean ± SD, with n = 5 samples per group in Figure 1A and n = 3 samples per group in Figures 1B and 1C. *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.005, ****P ≤ 0.0001.

Bulk RNA-seq analysis reveals GLUT1 is essential for maintaining homeostasis of HBMVEC

To further examine the effects of high glucose and GLUT1 loss on transcriptional alterations in HBMVECs, total RNA was isolated for bulk RNA-sequencing analysis following the method described in our previous publication (18). We compared the HBMVECs transfected with non-silencing control siRNA (siControl) to GLUT1 knockdown group (transfected with Glut-1 siRNA, siGlut1) (siControl vs. siGlut1) under normal culture conditions and the high glucose (HG) incubation settings (HG+siControl vs. HG+siGlut1). The RNA-seq results demonstrate that a decrease in GLUT1 considerably affects the transcriptional profile of human brain endothelial cells, as shown in the volcano plot (Fig. 2A), resulting in distinct and robust differentiation from the respective controls (Fig. 2B).

Figure 2: RNA-seq analysis reveals the effects of high glucose and GLUT1 knockdown on the transcriptome and functional alterations of HBMVECs.

Figure 2:

Open in a new tab

(A and B) GLUT1 was knocked down in HBMVECs using siRNA (SiGlut1) and treated with high glucose (HG), with bulk RNA-seq data demonstrating significant differentiation in gene transcription. (C) Gene Set Enrichment Analysis (GSEA) of RNA-seq data shows that the coagulation pathway diminishes under GLUT1 deficiency conditions. (D) Fast Gene Set Enrichment Analysis (FGSEA) in R provides a detailed characterization of the pathway shifts in GLUT1-deficient HBMVECs under normal and high glucose conditions, with pathways arranged by normalized enrichment scores (NES) in descending order. (A-D) Each group contains n = 3, except for HGNS, which has n = 2 due to poor sequencing in one sample. (A) was filtered based on a significance cutoff of P ≤ 0.05. (C) Enrichment plots for siGlut1 and HG+siGlut1 show downregulation with NES values of −1.30 and −1.32, respectively. The significance of the results was P= 0.048 for siGlut1 and P = 0.050 for HG+siGlut1, with false discovery rates (FDR, q-value) of 0.136 and 0.143, respectively.

To characterize these transcriptional shifts, Gene Set Enrichment Analysis (GSEA) was used to generate a comprehensive report of cellular pathway shifts within the hallmark pathway gene sets (15). Specifically, the coagulation pathway for both comparisons, siControl vs. siGlut1 and HG+siControl vs. HG+siGlut1 was negatively enriched, indicating potential antithrombotic properties by decreasing coagulability (Fig. 2C). Deeper computational analysis using Fast Gene Set Enrichment Analysis (fgsea) on R revealed broader pathway implications. Interestingly, associated with GLUT1 knockdown, there was a substantial decrease in the enrichment of inflammatory pathways such as TNF signaling via NF-κB and reactive oxygen species, suggesting a correlation between GLUT1 expression and regulation of stress-induced inflammatory responses (Fig 2D).

Next, to identify key players, specific marker genes involved in the oxidative stress pathway and positively correlated with endothelial dysfunction were analyzed within the context of the experimental conditions (20). It was found that several key markers listed in Figures 3A and 3B were downregulated after GLUT1 knockdown under both normal glucose levels and high glucose conditions. In particular, our findings show notable enrichments in podocalyxin (PODXL) and thrombomodulin (THBD), which have been implicated as key regulators for maintaining the integrity of the endothelium (21). Of note, linear regression analysis results indicate that the downregulation of thioredoxin interacting protein (TXNIP) is positively correlated with GLUT1 expression under high glucose conditions (Figs. 3C and 3D). TXNIP is integral to maintaining glucose homeostasis (22). Moreover, it has been established that upregulation of TXNIP increases the severity of endothelial dysfunction as well as insulin resistance (22–24). Our results demonstrate positive shifts within genes (such as THBD, PODXL, and HMOX2) that improve and maintain endothelial cell integrity. Overall, the shifts of gene expression patterns after GLUT1 knockdown imply that GLUT1 expression is essential in maintaining the homeostasis of endothelial cells.

Figure 3: RNA-seq reveals a GLUT1 knockdown potentiated mechanism aimed at protecting the endothelium.

Figure 3:

Open in a new tab

(A and B) Enhanced volcano plots illustrate expression profiles of specific genes within the siGlut1 and HG+siGlut1 groups. (C and D) The correlation plot shows the relative association of gene expressions under siGlut1 and HG+siGlut1 conditions, with SLC2A representing Glut1. (A-D) Each group contains n = 3, except for HG+siControl, which has n = 2 due to a sequencing error in one sample. (A and B) Significance cutoff values were set at log2FC ≥ 0.5 and P ≤ 0.05 for both siGlut1 and HG+siGlut1 groups.

Empirical results demonstrate strong anti-stress response within HBMVEC under HG conditions

To further determine the extent to which GLUT1 regulates the high glucose response, we first knocked down GLUT1 using validated siRNA. As shown in Figure 4A, GLUT1 knockdown results in a decrease in TXNIP and an increase in PODXL. When further treating the respective HBMVEC samples with 25 mM and 50 mM high glucose for 48 hours, consistent with RNA-seq data, TXNIP is decreased, but PODXL and THBD are increased in GLUT1 knockdown HBMVECs (Figs. 4B and 4C). The subsequent responses in gene expression changes are more substantial when incubated with 50 mM glucose (Fig. 4C). Moreover, following glucose-dependent reduced GLUT1 expression, a more prominent response is most associated with the GLUT1 knockdown groups (Figs. 4B and 4C). The expression of genes responsible for preserving the endothelium integrity (PODXL) and anti-thrombosis (THBD) are enriched in the GLUT1 knockdown groups (Figs 4B and 4C). Lastly, our results consistently show a positive correlation between GLUT1 and TXNIP expression levels, as both decrease under GLUT1 knockdown conditions (Figs 4B and 4C). This holds significance as TXNIP is involved in the induction of endothelial cell dysfunction, and thus, by down-regulating its expression, the integrity of the endothelium can be potentially preserved (23). These results are also reflected in the protein levels determined by Western blot analysis (Fig. 4D).

Figure 4. Effects of GLUT1 knockdown and high glucose on the expression of THBD, TXNIP, and PODXL in human brain microvascular endothelial cells.

Figure 4

Open in a new tab

(A) GLUT1 knockdown leads to a significant downregulation of TXNIP transcription in HBMVECs under normal culture conditions. (B and C) High glucose and GLUT1 knockdown enhance the expression of THBD and PODXL while suppressing TXNIP transcription in HBMVECs in a glucose concentration-dependent manner (25 mM and 50 mM glucose). The gene transcription levels were determined using real-time PCR. (A-C) Data are presented as mean ± SD, n = 3 samples per group. *P ≤ 0.05, **P≤ 0.01, ***P≤ 0.0005, **** P≤ 0.0001. (D) The effects of high glucose (25 mM) and GLUT1 knockdown on the expression of THBD, PODXL, TXNIP, and NLRP3 were further analyzed by Western blot analysis.

Loss of Glut1-mediated gene regulation in diabetic brain endothelial cells

We isolated brain endothelial cells following the method described in our previous publication (18) from wild-type or diabetic mice, respectively, and examined their responses to high glucose. As shown in Figure 5A, brain endothelial cells of wild-type mice show a similar response to high glucose as human brain endothelial cells (Fig. 4), decreased transcriptional levels of Glut1 and Txnip, along with increased transcription of Thbd and Podxl.

Figure 5: The Glut1-dependent regulation of Thbd, Txnip, and Podxl expression in brain endothelial cells was compromised in diabetic mice.

Figure 5:

Open in a new tab

(A) Effects of high glucose (25 mM and 50 mM) on the expression of Glut1, Thbd, Txnip, and Podxl in brain endothelial cells isolated from C57BL/6 wild-type mice (12 weeks old, male). (B-D) Effects of high glucose (25 mM) on Glut1, Thbd, Txnip, and Podxl expression in brain endothelial cells isolated from diabetic mice (male, 4/10 weeks for db/db mice, 12 weeks for C57BL/6 mice). Gene transcription levels were determined by real-time PCR. The data are presented as mean ± SD, n = 3 samples per group. * P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.0005, ****P ≤ 0.0001.

The db/db mice are well-accepted type 2 diabetic mouse models, and their blood glucose increases with age and reaches diabetic levels at ten weeks old (Fig. S3A). Interestingly, we observed opposite responses in brain endothelial cells isolated from diabetic mice. High glucose treatment significantly increased the expression of Glut1 and Txnip while decreasing Thbd in brain endothelial cells isolated from ten-week-old (10W) db/db mice (Fig. 5C). In contrast, little change was noted in brain endothelial cells of the 4W db/db mice (Fig. 5B). Our observations of diabetes-related deterioration of endothelial health align with established findings of glucose-dependent progression of vascular damage. The contrast observations of Glut1 expression were noted within diabetic and non-diabetic states, demonstrating a potential impairment of the previously described GLUT1-dependent regulatory mechanism in healthy brain endothelial cells (25, 26). In addition to the db/db model traditionally used to model T2DM, Streptozotocin (STZ)-induced diabetic mice are used as a compatible equivalent to type 1 diabetes mellitus (T1DM) conditions (27). Blood glucose reaches diabetic levels three weeks after STZ injections (Fig. S3B). The brain endothelial cells isolated from STZ-induced diabetic mouse models were also analyzed for their transcriptional composition under high glucose conditions. Brain endothelial cells of STZ-induced diabetic mice exhibit consistent alterations in Glut1-related gene expression under high glucose conditions (Fig. 5D). Notably, increased Podxl expression is a shared feature, but other genes such as Glut1 and Txnip show divergent responses, suggesting model-specific impairments in the Glut1 response mechanisms that contributes to further complications such as CVD (27, 28).

While Podxl experiences an enrichment in diabetic brain endothelial cells (Figs. 5C and 5D), the response is dampened with respect to transcript fold change in wild-type brain endothelial cells (Fig. 5A), compared to under similar conditions. Importantly, our observations suggest increased levels of Txnip, which have been established to be responsible for endothelial dysfunction and progressing the pathogenesis of CVD (24, 29). Txnip enrichment indicates that endothelial cells in the brain of diabetic mice show signs of a deteriorating endothelium, which directly contrasts with the observed stabilizing characterization given to the response of the healthy endothelium under high glucose conditions. These results reveal novel insights into the functional alterations contributing to the severity of the diabetic state.

GLUT1 overexpression induces oxidative stress-mediated inflammation and subverts endothelial permeability

To verify the sufficient conditions implicating GLUT1 as responsible for the previously observed results, we used a lentivirus vector to overexpress GLUT1 in human brain microvascular endothelial cells. Treatment with 25/50 mM HG further validated our proposed theory regarding the regulatory response crucial for maintaining endothelial integrity, with 50 mM HG showing more significant expression shifts (Fig. 6A and 6B). Consistently, as illustrated in Figures 6A and 6B, the reduction of GLUT1 in response to HG incubation induces certain markers (PODXL and THBD) involved in protecting endothelial integrity, while important regulators of inflammasome activation (TXNIP and NLRP3) are significantly reduced (21, 30). However, overexpressing GLUT1 reverses the relationships described in wild-type HBMVECs at both the transcript and protein levels (Fig. 6A–B). This is important as it implicates GLUT1 as a critical regulator of a potent protective mechanism for wild-type, non-disease state cells under both necessary and sufficient conditions. Furthermore, this response potentially illuminates the opposite regulatory relationships in diabetic disease states, as well as key distinctions between the disease and healthy states. We previously demonstrated that the diabetic state in 10-week db/db mice suggested significantly increased levels of Txnip expression. By amplifying GLUT1 expression here, we can replicate the expression profile of the selected biomarkers under the diabetic state (Fig. 6A and 6B). Ultimately, this indicates that GLUT1 may serve as the primary inducer of the observed downstream effects regarding endothelial permeability and the regulation of oxidative stress-mediated inflammation.

Figure 6: GLUT1 overexpression induces oxidative stress-mediated inflammation.

Figure 6:

Open in a new tab

(A/B) Effects of GLUT1 overexpression (GLUT1 OE) on gene expression in HBMVECs under high glucose (25 and 50 mM) conditions for 24 hours, determined by real-time PCR (A) and Western blot (B). (C) Effects of GLUT1 overexpression (GLUT1 OE) on IL-1β production, assessed using ELISA. HBMVEC cells were treated with high glucose (25 mM) for 24 hours. Data are presented as mean ± SD. *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.005, ****P ≤ 0.0001. n = 3 samples per group. (D) Dihydroethidium (DHE, 5 μM) staining was conducted to detect superoxide levels. High glucose (HG) treatment reduced superoxide levels compared to the control. However, GLUT1 overexpression significantly increased superoxide levels, while HG combined with GLUT1 overexpression resulted in a decrease in superoxide compared to GLUT1 OE alone. Scale bars: 100 μM.

To further characterize the state induced by GLUT1 upregulation, we also demonstrated the relationship between GLUT1 expression and the mediation of oxidative stress. GLUT1 overexpression leads to increased levels of TXNIP and its interacting factor, NOD-like receptor protein NLRP3 (Fig. 6A and 6B). It has been shown that TXNIP-NLRP3 interactions are involved in the assembly of the NLRP3 inflammasome, primarily due to the heightened production of reactive oxygen species (ROS) (31). Under conditions of GLUT1 overexpression (Fig. 6C–D), we further show elevated levels of superoxide and inflammatory cytokine production, such as interleukin 1β (IL-1β), indicating a detrimental cellular expression profile upon inflammasome assembly and activation (30).

In Figure 6C, we illustrate the activity of the NLRP3 inflammasome through the increased production of downstream secreted IL-1β, which facilitates the inflammatory reactions responsible for mediating vascular permeability (30). Due to GLUT1 overexpression and IL-1β induction, the endothelial adherens junction protein VE-Cadherin was reduced (Fig. 6B) (32). This finding demonstrates that GLUT1 upregulation compromises endothelial integrity, reflecting similar effects observed in diabetic mice (33). Further validating our proposed theories on a functional level, Figure 7A shows that GLUT1 overexpression leads to decreased fluorescent staining of VE-Cadherin in HBMVECs. The loss of VE-Cadherin is detrimental to endothelial integrity and significantly increases permeability (Fig. 7B), which is a characteristic co-morbidity present in the diabetic disease state (34). Importantly, GLUT1 overexpression combined with co-incubation under high-glucose (HG) conditions also exhibits the previously described glucose-mediated restoration of endothelial integrity on a functional level through the preservation of VE-Cadherin expression (Fig 7A). Ultimately, by rescuing the HG-mediated protective mechanism, endothelial permeability is not severely compromised and is significantly improved compared to the GLUT1 overexpressed group (Fig. 7B).

Figure 7: High glucose treatment and GLUT1 overexpression affect adherens junctions (VE-cadherin) and endothelial permeability in HBMVEC cells.

Figure 7:

Open in a new tab

HBMVEC cells were treated with 25 mM glucose (high glucose, HG) for 24 hours. (A) After HG treatment, immunofluorescence staining was conducted to evaluate the expression of VE-cadherin (Ve-cad, green) under different conditions. DAPI (blue) was used to stain nuclei. High glucose (25 mM) increased VE-cadherin expression; however, lentiviral overexpression of GLUT1 (GLUT1 OE) resulted in reduced VE-cadherin expression. Scale bars: 20 μm. (B) HBMVEC permeability was evaluated under various conditions using a FITC-Dextran EC-monolayer permeability assay. The bar graphs illustrate permeability (%) at 0 minutes and 60 minutes post-treatment. At 0 minutes, no significant difference in permeability was observed across the groups. At 60 minutes, the HG (25 mM) treatment reduced permeability compared to the control (***P ≤ 0.005). Overexpression of GLUT1 significantly increased permeability (****P ≤ 0.0001), while co-treatment with HG and GLUT1 overexpression (HG + GLUT1 OE) mitigated the rise in permeability. Data are presented as mean ± SD (*P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.005, ****P ≤ 0.0001), n = 3 samples per group.

Discussion

Within this study, our results characterize the differing roles and importance of the GLUT1-dependent high glucose response mechanism within brain endothelial cells between healthy and diabetic conditions. In particular, it was shown that under a healthy state, GLUT1 downregulation in response to high glucose is required for a stabilizing environment, preserving the endothelium homeostasis. However, under diabetic conditions, this regulatory mechanism may be compromised.

Regulating glucose transport across the plasma membrane, GLUT1 is a critical factor in the maintenance of endothelial cells (3, 35). In response to hyperglycemic conditions, GLUT1 expression decreases, limiting the excessive transport of glucose across the membrane in an attempt to prevent the induction of oxidative stress (3). Previous studies have shown that reactive oxygen species and TXNIP result in endothelial dysfunction, compromising the function of the endothelium and leading to further complications such as CVD (14, 29, 36, 37). Thus, by lowering GLUT1 expression in the setting of high glucose conditions, endothelial cells protect their intracellular environment against endothelial dysfunction via a cascade of expression shifts. Based on the literature (38–40), the known transcriptional regulation of GLUT1 is hypoxia-inducible factor 1-alpha (HIF-1α) under hypoxia conditions and peroxisome proliferator-activated receptor (PPAR). Our current study demonstrated that GLUT1 is an important regulator for protecting brain endothelial cell homeostasis. We need further investigation to reveal a comprehensive regulatory mechanism.

Our findings demonstrate this phenomenon via shifts of integral genes within non-diabetic endothelial cells. In particular, with respect to protecting the endothelium, our results indicate key enrichments in the expression of THBD and PODXL. THBD is an essential gene in preventing thrombosis by binding with thrombin to produce activated protein C (41, 42). Furthermore, THBD has been shown to hold much broader implications, such as mediating leukocyte trafficking and inhibiting inflammation via the activation of advanced glycation end products (41, 43). PODXL is a key membrane protein that mediates the adhesion interaction and barrier formation (44). PODXL is essential to maintaining endothelial barrier integrity, and its loss has been linked to endothelium dysfunction (21, 44). Additionally, our results indicate a downregulation in thioredoxin (TXNIP) expression. TXNIP is a stress-responsive gene and remarkably glucose-sensitive. Moreover, TXNIP has been suggested to hold an essential role in the pathogenesis of CVD, especially within diabetes, by promoting endothelial dysfunction (23, 24). Reduction of TXNIP levels decreases inflammatory responses, which is essential for the health of the endothelium. Overall, these shifts in gene expression portray a larger picture of a protective mechanism in reaction to high glucose by a reduction in GLUT1 expression.

Contrarily, our findings from diabetic models provided evidence that presented a stark contrast to those observed under non-diabetic conditions. First, it is of note that GLUT1 expression is upregulated in diabetic brain endothelial cells when incubated with high glucose, suggesting this change could indicate a compromise in the conventional protective cellular response to hyperglycemia, namely reducing Glut1 to decrease glucose uptake. The expression of key regulators THBD and TXNIP was also opposite of the observed pattern within the non-diabetic condition, suggesting a high-stress and endothelial dysfunction-prone environment in diabetic brain endothelial cells. This kind of endothelial dysfunction within the diabetic condition contributes to the progression of harmful effects, providing clinical significance to this phenomenon. For instance, endothelial dysfunction has been extensively linked with hyper-coagulopathy and thrombosis, as dysfunction plays an important role in thrombin generation and downregulation of THBD (5, 45, 46). Brain endothelial dysfunction has also been linked to the pathobiology of cerebrovascular disease linked to the progression of Alzheimer’s disease as well (45, 47, 48). Ultimately, loss of integrity within the endothelium results in the progression of CVD, which is more severe and pronounced under the diabetic disease state (49). Four-week db/db mice just have hyperglycemic conditions in a short time period, which may not impair the regulation system of downregulating Glut1 expression completely. The response to hyperglycemia in four-week db/db mice suggests that it may be at the tipping point of losing the capability to adapt to hyperglycemia. Forty-eight hours of glucose incubation consistently does not impair the regulation system of downregulating Glut1 expression.

Alternatingly, GLUT1 overexpressing HBMVEC recaptures many features of diabetic endothelial cells isolated from the brains of diabetes mice. Through GLUT1 overexpression, the expression profile related to the maintenance of brain endothelial cell homeostasis is inverted. Furthermore, enrichment in GLUT1 expression results in an expression pattern that closely mirrors that of the demonstrated diabetic disease state expression profile of 10-week db/db mice. On a functional level, our data demonstrates weakened endothelial junctions, as demonstrated by the lowered expression of VE-Cadherin. These results closely support established findings implicating the diabetic disease state to compromised endothelial integrity (33, 34). Additionally, further supporting our observed results, GLUT1 overexpression promotes the secretion of IL-1β, a potent cytokine for compromised endothelial cell permeability, through TXNIP-NLRP3 inflammasome activity (50, 51). Ultimately, these results, when interpreted together, demonstrate the potential implication for a severe disease state due to GLUT1-mediated permeability. Unexpectedly, GLUT1 overexpression does not change the PODXL induction under high glucose conditions. This phenomenon may reflect metabolic saturation-caused stress response, which may be regulated by a GLUT1-independent mechanism that requires further investigation.

Our findings provide important insights into the characterization of shifts that occur within brain endothelial cells when incubated in high glucose conditions both in the diabetic disease state and healthy conditions. Moreover, our results provide new avenues of investigation, such as looking deeper into the regulatory mechanisms of GLUT1, which hold importance in serving as a hopeful means of mitigating the progression of adverse effects resulting from endothelial dysfunction. However, it would be remiss to ignore limitations such as a lack of patient data pertaining to the GLUT1-led response, preventing comprehensive verification of our findings in humans. Nevertheless, our in vivo wild-type and db/db mouse models verify the consistent result shown through in vitro cell culture, providing scientific rigor for our discoveries.

Overall, our results reveal a novel mechanism involving GLUT1 expression levels as the main proponent in potentiating a stabilizing mechanism within brain endothelial cells. Under high glucose conditions, the downregulation of GLUT1 promotes exceptional shifts in gene expression in a manner that promotes the preservation of endothelial homeostasis, which is essential for alleviating the progression of diabetes-associated vascular complications. Interestingly, we also demonstrate that under diabetic conditions, within diabetic mice, the hyperglycemic disease state comprises the GLUT1-led anti-stress response, instead promoting an endothelial dysfunction-prone state. Our investigation offers new insights into the previously unrecognized role of glucose transporters in brain endothelial cells both under healthy and diabetic conditions, potentially providing clinical value by opening avenues for future investigation as to how to restore endothelial homeostasis in diabetic brain endothelial cells.

Supplementary Material

Supplemental Figs. S1-S3: DOI.10.6084/m9.figshare.29307098

Supplemental Tables S1: DOI.10.6084/m9.figshare.29315387

Acknowledgments

Funding:

National Institutes of Health grant HL141733 (QRM)

Startup funds from the New York University Langone Health-Long Island Hospital (QRM)

RNA-seq service at the Genomic Science and Precision Medicine Center (GSPMC) at the Medical College of Wisconsin

The graphical abstract was generated by the BioRender under the license of NYU Langone Health (WZ282D3QO8).

Footnotes

Competing interests: Authors declare that they have no competing interests.

Data and materials availability:

All data are available in the main text or the supplementary materials. Supplementary figures are available at DOI: 10.6084/m9.figshare.29307098. The bulk RNA-seq data in this study are publicly available in Gene Expression Omnibus (GEO) at GSE271637. Plasmid DNA and cell lines generated in this study are available to share under materials transfer agreements (MTAs).

References

  • 1.Chatterjee S, Khunti K, and Davies MJ. Type 2 diabetes. Lancet 389: 2239–2251, 2017. [DOI] [PubMed] [Google Scholar]
  • 2.Schalkwijk CG, and Stehouwer CD. Vascular complications in diabetes mellitus: the role of endothelial dysfunction. Clin Sci (Lond) 109: 143–159, 2005. [DOI] [PubMed] [Google Scholar]
  • 3.Cohen G, Riahi Y, Alpert E, Gruzman A, and Sasson S. The roles of hyperglycaemia and oxidative stress in the rise and collapse of the natural protective mechanism against vascular endothelial cell dysfunction in diabetes. Arch Physiol Biochem 113: 259–267, 2007. [DOI] [PubMed] [Google Scholar]
  • 4.Vazzana N, Ranalli P, Cuccurullo C, and Davi G. Diabetes mellitus and thrombosis. Thromb Res 129: 371–377, 2012. [DOI] [PubMed] [Google Scholar]
  • 5.Kaur R, Kaur M, and Singh J. Endothelial dysfunction and platelet hyperactivity in type 2 diabetes mellitus: molecular insights and therapeutic strategies. Cardiovasc Diabetol 17: 121, 2018. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Poredos P, and Jezovnik MK. Endothelial Dysfunction and Venous Thrombosis. Angiology 69: 564–567, 2018. [DOI] [PubMed] [Google Scholar]
  • 7.Sturtzel C Endothelial Cells. Adv Exp Med Biol 1003: 71–91, 2017. [DOI] [PubMed] [Google Scholar]
  • 8.Trimm E, and Red-Horse K. Vascular endothelial cell development and diversity. Nat Rev Cardiol 20: 197–210, 2023. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Gavriilaki E, Anyfanti P, Gavriilaki M, Lazaridis A, Douma S, and Gkaliagkousi E. Endothelial Dysfunction in COVID-19: Lessons Learned from Coronaviruses. Curr Hypertens Rep 22: 63, 2020. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Patching SG. Glucose Transporters at the Blood-Brain Barrier: Function, Regulation and Gateways for Drug Delivery. Mol Neurobiol 54: 1046–1077, 2017. [DOI] [PubMed] [Google Scholar]
  • 11.Koch H, and Weber YG. The glucose transporter type 1 (Glut1) syndromes. Epilepsy Behav 91: 90–93, 2019. [DOI] [PubMed] [Google Scholar]
  • 12.Barros LF, San Martin A, Ruminot I, Sandoval PY, Fernandez-Moncada I, Baeza-Lehnert F, Arce-Molina R, Contreras-Baeza Y, Cortes-Molina F, Galaz A, and Alegria K. Near-critical GLUT1 and Neurodegeneration. J Neurosci Res 95: 2267–2274, 2017. [DOI] [PubMed] [Google Scholar]
  • 13.Tran ND, Wong VL, Schreiber SS, Bready JV, and Fisher M. Regulation of brain capillary endothelial thrombomodulin mRNA expression. Stroke 27: 2304–2310; discussion 2310–2301, 1996. [DOI] [PubMed] [Google Scholar]
  • 14.Wardlaw JM, Smith C, and Dichgans M. Small vessel disease: mechanisms and clinical implications. Lancet Neurol 18: 684–696, 2019. [DOI] [PubMed] [Google Scholar]
  • 15.Subramanian A, Tamayo P, Mootha VK, Mukherjee S, Ebert BL, Gillette MA, Paulovich A, Pomeroy SL, Golub TR, Lander ES, and Mesirov JP. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A 102: 15545–15550, 2005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Liberzon A, Birger C, Thorvaldsdottir H, Ghandi M, Mesirov JP, and Tamayo P. The Molecular Signatures Database (MSigDB) hallmark gene set collection. Cell Syst 1: 417–425, 2015. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Soneson C, Love MI, and Robinson MD. Differential analyses for RNA-seq: transcript-level estimates improve gene-level inferences. F1000Res 4: 1521, 2015. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Fang Z, Sun X, Wang X, Ma J, Palaia T, Rana U, Miao B, Ragolia L, Hu W, and Miao QR. NOGOB receptor deficiency increases cerebrovascular permeability and hemorrhage via impairing histone acetylation-mediated CCM1/2 expression. J Clin Invest 132: 2022. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Song L, Ge S, and Pachter JS. Caveolin-1 regulates expression of junction-associated proteins in brain microvascular endothelial cells. Blood 109: 1515–1523, 2007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Higashi Y Roles of Oxidative Stress and Inflammation in Vascular Endothelial Dysfunction-Related Disease. Antioxidants (Basel) 11: 2022. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Cait J, Hughes MR, Zeglinski MR, Chan AW, Osterhof S, Scott RW, Canals Hernaez D, Cait A, Vogl AW, Bernatchez P, Underhill TM, Granville DJ, Murphy TH, Roskelley CD, and McNagny KM. Podocalyxin is required for maintaining blood-brain barrier function during acute inflammation. Proc Natl Acad Sci U S A 116: 4518–4527, 2019. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Zhou R, Tardivel A, Thorens B, Choi I, and Tschopp J. Thioredoxin-interacting protein links oxidative stress to inflammasome activation. Nat Immunol 11: 136–140, 2010. [DOI] [PubMed] [Google Scholar]
  • 23.Lam YT, Tan RP, Michael P, Yang N, Dunn LL, Cooke JP, Celermajer DS, Wise SG, and Ng MKC. Endothelial thioredoxin interacting protein (TXNIP) modulates endothelium-dependent vasorelaxation in hyperglycemia. Microvasc Res 143: 104396, 2022. [DOI] [PubMed] [Google Scholar]
  • 24.Li X, Kover KL, Heruth DP, Watkins DJ, Guo Y, Moore WV, He LG, Zang M, Clements MA, and Yan Y. Thioredoxin-interacting protein promotes high-glucose-induced macrovascular endothelial dysfunction. Biochem Biophys Res Commun 493: 291–297, 2017. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Liao YJ, Ueno M, Nakagawa T, Huang C, Kanenishi K, Onodera M, and Sakamoto H. Oxidative damage in cerebral vessels of diabetic db/db mice. Diabetes Metab Res Rev 21: 554–559, 2005. [DOI] [PubMed] [Google Scholar]
  • 26.Li Q, Atochin D, Kashiwagi S, Earle J, Wang A, Mandeville E, Hayakawa K, d’Uscio LV, Lo EH, Katusic Z, Sessa W, and Huang PL. Deficient eNOS phosphorylation is a mechanism for diabetic vascular dysfunction contributing to increased stroke size. Stroke 44: 3183–3188, 2013. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Fidler TP, Marti A, Gerth K, Middleton EA, Campbell RA, Rondina MT, Weyrich AS, and Abel ED. Glucose Metabolism Is Required for Platelet Hyperactivation in a Murine Model of Type 1 Diabetes. Diabetes 68: 932–938, 2019. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Rahadian A, Fukuda D, Salim HM, Yagi S, Kusunose K, Yamada H, Soeki T, Shimabukuro M, and Sata M. Thrombin inhibition by dabigatran attenuates endothelial dysfunction in diabetic mice. Vascul Pharmacol 124: 106632, 2020. [DOI] [PubMed] [Google Scholar]
  • 29.Nasoohi S, Ismael S, and Ishrat T. Thioredoxin-Interacting Protein (TXNIP) in Cerebrovascular and Neurodegenerative Diseases: Regulation and Implication. Mol Neurobiol 55: 7900–7920, 2018. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Choi EH, and Park SJ. TXNIP: A key protein in the cellular stress response pathway and a potential therapeutic target. Exp Mol Med 55: 1348–1356, 2023. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Zhou R, Yazdi AS, Menu P, and Tschopp J. A role for mitochondria in NLRP3 inflammasome activation. Nature 469: 221–225, 2011. [DOI] [PubMed] [Google Scholar]
  • 32.Giannotta M, Trani M, and Dejana E. VE-cadherin and endothelial adherens junctions: active guardians of vascular integrity. Dev Cell 26: 441–454, 2013. [DOI] [PubMed] [Google Scholar]
  • 33.Alshammari MA, Alshehri AO, Alqahtani F, Khan MR, Bakhrebah MA, Alasmari F, Alshammari TK, and Alsharari SD. Increased Permeability of the Blood-Brain Barrier in a Diabetic Mouse Model (Lepr(db)(/db) Mice). Int J Mol Sci 25: 2024. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Ward R, Li W, Abdul Y, Jackson L, Dong G, Jamil S, Filosa J, Fagan SC, and Ergul A. NLRP3 inflammasome inhibition with MCC950 improves diabetes-mediated cognitive impairment and vasoneuronal remodeling after ischemia. Pharmacol Res 142: 237–250, 2019. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Koepsell H Glucose transporters in brain in health and disease. Pflugers Arch 472: 1299–1343, 2020. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Yamawaki H, Pan S, Lee RT, and Berk BC. Fluid shear stress inhibits vascular inflammation by decreasing thioredoxin-interacting protein in endothelial cells. J Clin Invest 115: 733–738, 2005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Ouyang H, Wang Y, Wu J, and Ji Y. Mechanisms of pulmonary microvascular endothelial cells barrier dysfunction induced by LPS: The roles of ceramides and the Txnip/NLRP3 inflammasome. Microvasc Res 147: 104491, 2023. [DOI] [PubMed] [Google Scholar]
  • 38.Zelzer E, Levy Y, Kahana C, Shilo BZ, Rubinstein M, and Cohen B. Insulin induces transcription of target genes through the hypoxia-inducible factor HIF-1alpha/ARNT. EMBO J 17: 5085–5094, 1998. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Gleadle JM, and Ratcliffe PJ. Induction of hypoxia-inducible factor-1, erythropoietin, vascular endothelial growth factor, and glucose transporter-1 by hypoxia: evidence against a regulatory role for Src kinase. Blood 89: 503–509, 1997. [PubMed] [Google Scholar]
  • 40.Riahi Y, Sin-Malia Y, Cohen G, Alpert E, Gruzman A, Eckel J, Staels B, Guichardant M, and Sasson S. The natural protective mechanism against hyperglycemia in vascular endothelial cells: roles of the lipid peroxidation product 4-hydroxydodecadienal and peroxisome proliferator-activated receptor delta. Diabetes 59: 808–818, 2010. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Conway EM. Thrombomodulin and its role in inflammation. Semin Immunopathol 34: 107–125, 2012. [DOI] [PubMed] [Google Scholar]
  • 42.Rochfort KD, and Cummins PM. Thrombomodulin regulation in human brain microvascular endothelial cells in vitro: role of cytokines and shear stress. Microvasc Res 97: 1–5, 2015. [DOI] [PubMed] [Google Scholar]
  • 43.Loghmani H, and Conway EM. Exploring traditional and nontraditional roles for thrombomodulin. Blood 132: 148–158, 2018. [DOI] [PubMed] [Google Scholar]
  • 44.Horrillo A, Porras G, Ayuso MS, and Gonzalez-Manchon C. Loss of endothelial barrier integrity in mice with conditional ablation of podocalyxin (Podxl) in endothelial cells. Eur J Cell Biol 95: 265–276, 2016. [DOI] [PubMed] [Google Scholar]
  • 45.Gimbrone MA Jr., and Garcia-Cardena G. Endothelial Cell Dysfunction and the Pathobiology of Atherosclerosis. Circ Res 118: 620–636, 2016. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Vinayagam S, and Sattu K. SARS-CoV-2 and coagulation disorders in different organs. Life Sci 260: 118431, 2020. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Sweeney MD, Sagare AP, and Zlokovic BV. Blood-brain barrier breakdown in Alzheimer disease and other neurodegenerative disorders. Nat Rev Neurol 14: 133–150, 2018. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Parsamanesh N, Asghari A, Sardari S, Tasbandi A, Jamialahmadi T, Xu S, and Sahebkar A. Resveratrol and endothelial function: A literature review. Pharmacol Res 170: 105725, 2021. [DOI] [PubMed] [Google Scholar]
  • 49.Henning RJ. Type-2 diabetes mellitus and cardiovascular disease. Future Cardiol 14: 491–509, 2018. [DOI] [PubMed] [Google Scholar]
  • 50.Fetsko AR, Sebo DJ, Budzynski LB, Scharbarth A, and Taylor MR. IL-1beta disrupts the initiation of blood-brain barrier development by inhibiting endothelial Wnt/beta-catenin signaling. iScience 27: 109651, 2024. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 51.Versele R, Sevin E, Gosselet F, Fenart L, and Candela P. TNF-alpha and IL-1beta Modulate Blood-Brain Barrier Permeability and Decrease Amyloid-beta Peptide Efflux in a Human Blood-Brain Barrier Model. Int J Mol Sci 23: 2022. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

All data are available in the main text or the supplementary materials. Supplementary figures are available at DOI: 10.6084/m9.figshare.29307098. The bulk RNA-seq data in this study are publicly available in Gene Expression Omnibus (GEO) at GSE271637. Plasmid DNA and cell lines generated in this study are available to share under materials transfer agreements (MTAs).

RESOURCES