Bupleuri Radix Ameliorates Vascular Inflammation in Human Umbilical Vein Endothelial Cells via Modulation of Tight Junction Protein Expression and Inhibition of Nuclear Factor-κB Activation.

Cardiovascular diseases (CVD) are associated with high morbidity and mortality rates worldwide [ 1 ]. The mortality rate is estimated to reach 2.36 million persons per year by 2030 [ 2 ]. The main pathogenesis of CVD is endothelial cell damage caused by inflammation [ 3 , 4 ]. Hence, to prevent the development and progression of CVD, there is a need for new drugs that can suppress endothelial cell inflammation, manage controllable prognostic factors, and alleviate endothelial cell damage. Bupleuri Radix (BR) is obtained by drying the roots of Bupleurum falcatum Linne (belonging to the family Umbelliferae) [ 5 ]. BR has traditionally been used to harmonize the liver and stomach and to alleviate chest pressure and pain [ 6 ]. In vitro and in vivo studies have revealed that BR exhibits anti-inflammatory [ 7 ], antioxidant [ 8 ], antiobesity [ 9 ], and anticancer [ 10 ] properties. Consequently, it can be used as a therapeutic agent for various inflammatory diseases [ 11 ]. However, limited research has been conducted on the direct effects of BR on vascular inflammation and CVD. CVD are associated with chronic inflammatory conditions, such as atherosclerosis, and regulating these inflammatory responses is crucial for cardiovascular health. Therefore, further studies are warranted to explore the impact of the anti-inflammatory properties of BR on vascular inflammation and CVD. Interleukin (IL)-1β-induced endothelial inflammation and dysfunction play key roles in the etiology of CVD by reducing vascular homeostasis [ 12 ]. Considering the anti-inflammatory effects of BR, this study assessed whether BR can alleviate endothelial cell dysfunction in human umbilical vein endothelial cells (HUVECs) with IL-1β-induced inflammation.

Vascular cell inflammation increases cell migration and neovascularization [ 13 ]. In our study, BR inhibited the migration of IL-1β-stimulated HUVECs without inducing toxicity ( Fig. 1 ). We performed a tube formation assay to determine whether BR could suppress angiogenesis in IL-1β-stimulated HUVECs. IL-1β treatment increased tube formation in HUVECs. However, BR pretreatment resulted in lower tube formation than IL-1β treatment alone ( Fig. 2A ). Analysis of neovascularization images using the Gilles Carpentier Research Web Site revealed markedly lower neovascularization ( Fig. 2B-H ) in cells pretreated with BR than in those treated with IL-1β alone. Cell barrier defects caused by changes in tight junction (TJ) protein expression in the epithelium facilitate the onset of vascular inflammation that is associated with various diseases [ 14 ]. Therefore, we determined whether BR could regulate TJ protein expression in IL-1β-stimulated HUVECs. The expression level of zonula occludens (ZO)-2, a protein related to TJs, markedly decreased after 24 h of treatment with IL-1β; however, its level was restored to that in the control group in cells pretreated with BR for 1 h ( Fig. 3A, B ). Assessment of changes in ZO-2 expression using fluorescence staining yielded the same result ( Fig. 3C , G). Among claudins, another group of proteins related to TJs, the expression level of only claudin-1 distinctly increased after IL-1β treatment. Western blotting and immunofluorescence staining revealed that the expression level decreased to the level in the control group after BR treatment ( Fig. 3D-G ). We also examined the changes in endothelial cell permeability after IL-1β and BR treatments by measuring the level of FITC–dextran (70 KD) infiltration through the HUVEC monolayer. The permeability of cells to FITC–dextran that had been increased by IL-1β treatment tended to decrease after BR pretreatment ( Fig. 3H ). Moreover, BR treatment restored the TEER level that had been markedly decreased by IL-1β treatment ( Fig. 3I ). These results indicate that BR regulates TJ protein expression altered during vascular inflammation, thereby contributing to the protection of vascular barrier functions. Under inflammatory conditions, the expression levels of adhesive molecules, such as ICAM-1 and VCAM-1, increase. Moreover, the migration of monocytes in vascular endothelial cells is enhanced [ 15 ]. Therefore, we examined the changes in the expression levels of ICAM-1 and VCAM-1 in IL-1β-stimulated HUVECs. As shown in Fig. 4A and B, the expression levels of ICAM-1 and VCAM-1 markedly increased in HUVECs after 24 h of treatment with IL-1β. However, the expression levels decreased after pretreatment with BR for 1 h ( Fig. 4C, D ). Furthermore, the immunofluorescence assay for ICAM-1, which showed a marked decrease in expression following BR treatment, revealed expression patterns identical to those observed in western blotting ( Fig. 4E, F ). Next, we performed an adhesion assay to evaluate the migration of white blood cells to vascular endothelial cells. As shown in Fig. 4G and H , the adhesion of U937 cells to HUVECs was markedly lower in the group pretreated with BR than in that treated with only IL-1β. These results indicate that BR downregulates the expression of adhesive molecules and reduces monocyte migration to suppress vascular inflammation. ROS are closely associated with various inflammatory diseases, including CVD [ 16 ]. Therefore, we assessed whether BR could control IL-1β-induced ROS generation in HUVECs. As shown in Fig. 5A and B , BR treatment markedly reduced the level of ROS generation in IL-1β-stimulated HUVECs. NAC (a ROS scavenger) and mitoTEMPO (a mitochondrial superoxide dismutase mimetic and peroxyl scavenger) were used as positive controls [ 17 ]. The antioxidant effect in BR-treated cells was similar to that in the positive controls. Assessment of DCF-DA-stained cells under a fluorescence microscope revealed that the ROS levels that had increased after IL-1β treatment decreased after BR treatment ( Fig. 5C, D ). These results indicate that BR significantly reduces IL-1β-induced oxidative stress. NF-κB is an important transcription factor that regulates the expression of proteins associated with inflammation [ 18 ]. As shown in Fig. 5E and F , the nuclear translocation level of the p65 subunit of NF-κB that had been increased after 24 h of treatment with IL-1β was restored to the level in the control group after 1 h of pretreatment with BR. Moreover, the cytoplasmic expression of NF-κB inhibitor α (IκB-α) that had been decreased after IL-1β treatment increased after BR treatment. The immunofluorescence assay also revealed the same trend ( Fig. 5G ). These results indicate that BR inhibits IL-1β-induced NF-κB activation, suppressing vascular inflammation. We performed a network pharmacology analysis to further investigate the potential of BR as a preventive and therapeutic agent for atherosclerosis and to elucidate the underlying mechanisms. We generated a list of BR target genes using TCMID. We used these BR target genes in the network analysis with the JEPPETTO plug-in of Cytoscape to assess the correlation between genetic and biological functions. This KEGG-based gene enrichment analysis revealed that the peroxisome proliferator-activated receptor (PPAR) signaling pathway was most significantly related to the BR target genes ( Fig. 6A ). Moreover, steroid hormone biosynthesis, galactose metabolism, glycerol lipid metabolism, VEGF signaling, and arachidonic acid metabolism were correlated with cellular and pathological changes related to atherosclerosis. Furthermore, we verified whether BR could target the genes associated with atherosclerosis listed in DisGeNet. Based on the results, we identified 17 genes as shared target genes for BR and endometriosis ( Fig. 6B ). All genes, except for F10 , exhibited strong PPIs ( Fig. 6C ).

We investigated the protective effects of BR against vascular inflammation and elucidated the underlying mechanisms. BR reduced cell migration, suppressed angiogenesis, and decreased adhesive molecule expression, preventing the adhesion of monocytes to endothelial cells in IL-1β-stimulated HUVECs. Furthermore, BR maintained the integrity of vascular endothelial cells by regulating TJs through the suppression of intracellular ROS generation and nuclear translocation of NF-κB p65. Although further studies are warranted to validate the clinical efficacy of BR in suppressing vascular inflammation and maintaining TJ integrity and to elucidate the underlying mechanisms, our findings suggest that BR protects against inflammation-induced vascular injury and could help prevent CVD.

In this study, we used HUVECs with IL-1β-induced inflammation to assess the preventive and therapeutic effects of BR. In vitro analyses revealed that BR can effectively suppress endothelial inflammation ( Fig. 7 ). Inflammatory cytokines and excessive ROS activate vascular endothelial cells; this is the main cause of endothelial abnormalities that result in ischemic heart diseases, such as myocardial infarction and stroke [ 19 ]. IL-1 is a representative proinflammatory cytokine similar to tumor necrosis factor (TNF). The IL-1 family consists of 11 distinct members. Of these, IL-1β exhibits major biological activities [ 20 ]. It induces the expression of E-selectin in vascular endothelial cells. Moreover, it increases the expression levels of adhesive molecules, such as ICAM-1 and VCAM-1, and integrin ligands in white blood cells, thereby promoting the adhesion of white blood cells and vascular endothelial cells [ 12 , 21 ]. ICAM-1 is expressed in the early stages of atherosclerosis, whereas VCAM-1 is expressed in the late stages [ 22 ]. Consistently, ICAM-1 expression markedly increased within 30 min of IL-1 treatment in our study. Previous studies have revealed that the expression levels of ICAM-1 and VCAM-1 are closely associated with angiogenesis [ 22 , 23 ]. In this study, IL-1β-stimulated HUVECs exhibited a marked increase in cross-linkages and connections with endothelial cells, along with increased expression levels of ICAM-1 and VCAM-1. Such angiogenesis-related processes were suppressed by BR treatment, suggesting the ability of BR to control vascular inflammation. Vascular endothelial cells exhibit strong cell–cell binding to ensure response to blood flow and inflammatory reactions. The cell–cell connections that play a crucial role in the maintenance of the endothelial cell barrier include TJs, adherens junctions, and gap junctions. The TJ family includes ZO-1, ZO-2, and ZO-3, which are connected to the cytoplasm via occludins and claudins in the cell membrane [ 24 ]. Ni et al. [ 25 ] and Clark et al. [ 26 ] reported that claudin-5 and occludins play important roles in TJ formation in endothelial cells. In this study, IL-1β-stimulated HUVECs exhibited significantly altered expression of ZO-2, claudin-1, and claudin-5 after BR treatment. As BR contains various components, it is challenging to interpret the exact effect of specific components. Future studies should therefore assess the specific effect of BR components on cell barrier integrity. In this study, we demonstrated that BR maintains the integrity of vascular endothelial cells by regulating TJ expression in vitro . Our results indicate the protective effect of BR on endothelial cells during the inflammatory response. Previous empirical studies on endothelial cell dysfunction induced by oxidative stress, TNF-α, LPS, or diabetes have revealed the key role of the activation of the NF-κB pathway [ 27 - 29 ]. NF-κB signaling for transcriptional activation occurs via classical (canonical) and alternative (noncanonical) pathways; both these pathways involve the dimerization of five proteins [p50, p52, p65 (RelA), RelB, and c-Rel] that bind with DNA to induce the expression of target proteins [ 20 ]. The canonical pathway is activated in most cases involving IL-1β stimulation. In particular, IκB-α is phosphorylated for degradation, and a dimer of p65 is translocated to the nucleus [ 30 , 31 ]. Consistent with these reports, in our study, IL-1β treatment led to the activation of the canonical pathway and the nuclear translocation of a dimer of NF-κB p65. This phenomenon may regulate the expression of adhesive molecules (e.g., ICAM-1 and VCAM-1) and TJs (e.g., ZO-2, claudin-1, and claudin-5). BR contains saikosaponins, volatile oils, polysaccharides, fatty acids, and sterols. Of these, saikosaponins serve as the most important physiologically active component [ 32 ]. In particular, saikosaponin a exhibits a strong anti-inflammatory effect and can alleviate the inflammatory response by regulating the NF-κB pathway and inhibiting ROS generation [ 33 ]. Further mechanistic studies are warranted to assess the protective effect of BR on vascular inflammation. Nevertheless, our results indicate that NF-kB may be an important mediator of the anti-inflammatory effects of BR. Interestingly, we predicted the potential therapeutic effects of BR based on a network pharmacology analysis. We analyzed potential target genes obtained from TCMID and found that several signaling pathways, including PPAR signaling pathways, were associated with BR ( Fig. 6 ). To confirm whether BR can effectively prevent arteriosclerosis, it is necessary to establish an animal model and assess the expression of related proteins and the underlying mechanisms in future research.

BR samples were purchased from Seogyeong Herbal Medicine Co., Ltd. (Busan, Republic of Korea). The samples were placed in 70% ethyl alcohol for subsequent ultrasonic extraction (Powersonic 405; Hwashin Technology Co., Ltd., Seoul, Republic of Korea). BR powder was dissolved in sterile distilled water and filtered using a Minister ® syringe filter (0.2 μm; Sartorius AG, Weender Landstraß, Germany). The resulting solution was diluted to appropriate concentrations (0.25-2 μg/mL) for subsequent use. HUVECs and human leukemia cells (U937 cells) were purchased from the American Type Culture Collection (Manassas, VA, USA). HUVECs were cultured in Vascular Endothelial Growth Factor (VEGF) Endothelial Medium (VascuLife VEGF LifeFactors Kit; Lifeline Cell Technology, Frederick, MD, USA). U937 cells were cultured in Roswell Park Memorial Institute Medium (RPMI-1640; Welgene, Daegu, Republic of Korea) containing 10% fetal bovine serum (FBS; Welgene) and 100 units/mL penicillin/streptomycin (P/S) at 37℃ under 5% CO 2 conditions. The cytotoxicity of BR was assessed using the MTT assay. In brief, the cells were treated with appropriate concentrations (0.25-2.00 μg/mL) of BR extract and cultured for 24 h. These cells were then treated with 0.5 mg/mL MTT solution (Invitrogen, Carlsbad, CA, USA) and incubated for 2 h to induce the reaction. Subsequently, the medium containing MTT was removed, and dimethyl sulfoxide (Sigma-Aldrich Chemical, St. Louis, MO, USA) was added to completely dissolve the formazan crystals produced by the reaction. The dissolved formazan was transferred to a 96-well plate at 200 μL/well. Subsequently, the absorbance was measured at 540 nm using a microplate reader (Beckman Coulter Inc., Brea, CA, USA). The cytotoxicity is expressed as a percentage based on the control. HUVECs were seeded in a six-well plate (4 × 10 5 cells/well) and cultured at 37℃ for 24 h. A scratch wound was created using a 200-μL tip, and the cells were then washed twice with phosphate-buffered saline (PBS). Following this, a medium without VascuLife VEGF LifeFactors was added. After pretreatment with BR for 1 h, the cells were treated with IL-1β extract for 24 h. Cell migration distance was quantified using a wound healing size tool (ImageJ software, v1.8.0; NIH, Bethesda, MD, USA). HUVECs were seeded on Matrigel (BD Biosciences, San Jose, CA, USA). After 1 h, when cell adhesion could be observed, the cells were treated with BR extract for 1 h and then treated with IL-1β for 24 h. Subsequently, fluorescence staining was performed using calcein acetoxymethyl ester (Calcein-AM; Biotium Inc., Fremont, CA, USA). The suppression of tube formation was assessed using a Carl Zeiss TM Axio Vert.A1 inverted microscope (Oberkochen, Germany) and an EVOS FL Auto 2 imaging system (Thermo Fisher Scientific, Waltham, MA, USA). Quantitative evaluation of tube formation was performed using the ImageJ Angiogenesis Analyzer tool on the Gilles Carpentier Research Web Site ( http://image.bio.methods.free.fr ). Samples were prepared using the Bio-Rad Protein Assay Kit (Bio-Rad Laboratories, Hercules, CA, USA). Proteins were then separated and transferred onto a nitrocellulose membrane (Thermo Fisher Scientific). The membrane was incubated with specific primary antibodies, and the bands were enhanced for analysis using chemiluminescence reagents (Thermo Fisher Scientific). Antibodies were purchased from Santa Cruz Biotechnology, Inc. (Dallas, TX, USA), Cell Signaling Technology, Inc. (Beverly, MA, USA), Thermo Fisher Scientific, and Abcam (Cambridge, UK) ( Table 1 ). The cells were fixed with methanol for 15 min at 4℃. Subsequently, the nuclear membrane was permeabilized using 0.2% Triton X-100. The cellular reaction was blocked by incubating the cells with 5% BSA/PBS-T for 1 h. After overnight treatment with primary antibodies ( Table 2 ) at 4℃, the cells were incubated with a secondary antibody for 1 h at 25℃. The nucleus was stained with 4’,6-diamidino-2-phenylindole (DAPI) and fixed by crystal mounting. Images were captured using a fluorescence microscope (EVOS FL Auto 2 imaging system; Thermo Fisher Scientific). U937 cells (1.2 × 10 5 cells/well) were stained with DAPI for 15 min and seeded in wells containing HUVECs treated with IL-1β and BR for 12 h. Subsequently, the cells were washed thrice with the culture medium to remove any detached cells. Finally, the U937 cells adhering to HUVECs were observed using a Carl Zeiss TM Axio Vert.A1 inverted microscope (Oberkochen) and an EVOS FL Auto 2 imaging system (Thermo Fisher Scientific). HUVECs (1.5 × 10 4 cells/well) were seeded in Transwell ® polycarbonate membrane well inserts (6.5-mm diameter, 0.4-µm pore size; Corning Inc., NY, USA) and allowed to stabilize for 24 h. The cells were pretreated with BR for 1 h and then treated with IL-1β for 24 h. Fresh medium was added to the lower chamber of the Transwell insert, and the medium in the upper chamber was replaced with medium containing FITC–dextran (Invitrogen Life Technologies); the samples were incubated in the dark for 30 min. The medium in the lower chamber was transferred into a 96-well plate at 200 μL/well. Fluorescence was then measured using an ELISA reader (BioTek) at an excitation wavelength of 490 nm and emission wavelength of 520 nm. HUVECs (1.5 × 10 4 cells/well) were seeded in Transwell ® polycarbonate membrane well inserts (6.5-mm diameter, 0.4-µm pore size; Corning Inc.) and allowed to stabilize for 24 h. The cells were pretreated with BR for 1 h and then treated with IL-1β for 24 h. Saturation in the chamber was reached after 72 h. TEER of the exposed cells was then measured using Epithelial Volt-Ohm Meter 2 (EVOM2; World Precision Instruments, Sarasota, FL, USA). The effect of BR on ROS generation after IL-1β treatment was determined by staining the cells with 2’,7’-dichlorodihydrofluorescein diacetate (H2-DCF; Thermo Fisher Scientific). In brief, HUVECs were seeded in a 12-well plate (1.2 × 10 5 cells/well), cultured for 24 h, and then treated with 10 μM H2-DCF in a 5% CO 2 incubator at 37℃ for 20 min. The changes in ROS generation were measured using a BD Accuri C 6 flow cytometer (BD Biosciences). Moreover, the cells cultured under identical conditions were treated with H2-DCF in the dark for 20 min, washed twice with PBS, and then fixed with 3.7% paraformaldehyde. The fixed cells were again washed twice with PBS, and fluorescence intensity was assessed using EVOS FS Auto (Thermo Fisher Scientific). N-acetylcysteine (NAC) and mitoTEMPO were used as appositive controls to estimate ROS generation. The cells were pretreated with BR extract for 1 h and subsequently treated with IL-1β for 24 h. Then, they were harvested and centrifuged at 842 × g for 5 min at 4℃. The nuclei and cytoplasm of the collected cells were isolated using the NE-PER Nuclear and Cytoplasmic Extraction Kit (Thermo Fisher Scientific). Proteins were quantified using Bio-Rad reagent (Bio-Rad Laboratories). The potential target genes involved in the association between BR and atherosclerosis were identified using the Traditional Chinese Medicine Integrative Database (TCMID) ( https://bidd.group/TCMID ) and DisGeNet ( https://www.disgenet.org ). In total, 53 BR target genes were further analyzed using the Java Enrichment of Pathways Extended To TOpology (JEPPETTO) plug-in for gene enrichment based on the Kyoto Encyclopedia of Genes and Genomes (KEGG) platform of Cytoscape (Ver. 3.9.1, https://cytoscape.orgβ ). The potential target genes involved in the association between BR and atherosclerosis were used to analyze protein–protein interactions (PPIs) based on the STRING plug-in of Cytoscape. Differences between groups were analyzed by one-way analysis of variance with Tukey’s test using GraphPad Prism (version 8.4.2; GraphPad Software Inc., La Jolla, CA, USA). Statistical significance was set at p < 0.05.