Introduction
Endothelial cells (ECs) form a semi-permeable monolayer that separates the bloodstream from underlying tissues and regulate the infiltration of blood cells and proteins through the vessel wall (1,2). A number of studies have previously suggested that vascular endothelial dysfunction caused by increased junctional permeability is a key event in the development and progression of atherosclerosis and atherosclerotic cardiovascular diseases (3,4). Chronic metabolic disorders, such as hyperglycemia and dyslipidemia, promote endothelial dysfunction and eventually leakage, resulting in increased monocyte diapedesis (5-7). Endothelial dysfunction during diabetes is closely associated with increased oxidized low-density lipoprotein (oxLDL) levels mediated by high glucose (HG) conditions (8). HG and oxLDL induce oxidative stress by increasing the production of reactive oxygen species (ROS) (9), which drive inflammatory processes to mediate monocyte activation and recruitment to the inflamed endothelium (10). Subsequently, endothelial inflammation enhances the expression of intercellular adhesion molecule-1 (ICAM-1) and vascular cell adhesion molecule-1 (VCAM-1) on ECs to potentiate monocyte adhesion prior to diapedesis into the sub-endothelial space, which precedes atherosclerotic plaque development (11).
Endothelial permeability is regulated by the close cooperation between inter-endothelial cell adherens and tight junctions, which form an endothelial barrier to control the paracellular passage of blood constituents into the subendothelial space (2,12-16). Endothelial integrity is maintained by the highly coordinated opening and closing of endothelial adherens and tight junctions (17). Vascular endothelial (VE)-cadherin, a major constituent of adherens junctions, interacts with the tight junction molecule zonula occludens-1 (ZO-1) to link together adjacent ECs to regulate diapedesis (2,13). There is evidence that vascular permeability and leukocyte invasion are promoted by the phosphorylation of VE-cadherin (16,18) and the downregulation of ZO-1 expression (16).
Pharmacological compounds that can maintain the homeostasis of endothelial junctions have attracted attention due to their potential clinical applicability in alleviating vascular diseases (19-21). Rosmarinic acid (RA; also known as α-O-caffeoyl-3,4-dihydroxyphenyl lactic acid) is a polyphenolic antioxidant that can be found in the Lamiaceae family of herbs and mint, such as Rosmarinus officinalis (22). RA has been reported to exert antioxidant and anti-inflammatory activities. In particular, the antidiabetic and cardioprotective activities of RA have been previously high-lighted (23,24). It was demonstrated that oxLDL-mediated oxidative stress can upregulate thioredoxin interacting protein (TXNIP) expression, which then binds to nucleotide-binding domain and leucine-rich repeat containing family, pyrin domain-containing 3 (NLRP3) to activate the NLRP3 inflammasome in ECs under HG conditions (25). RA was found to reverse this form of oxLDL-mediated oxidative stress and NLRP3 inflammasome activation (25). In addition, the subsequent oxLDL-mediated IL-1β secretion in ECs was reversed by RA treatment by downregulating ROS production, p38 phosphorylation and forkhead box O1 (FOXO1) protein expression in ECs (25). However, the potential effect of oxLDL-triggered NLRP3 activation on endothelial integrity, permeability and monocyte transmigration remains poorly understood. In addition, the potential effect of RA on these oxLDL-mediated pathophysiological processes along with its underlying mechanism remain unknown.
Therefore, in the present study, the effects of oxLDL-triggered NLRP3 activation on the interaction between ECs and monocytes, in addition to endothelial integrity and permeability, under HG conditions, were evaluated. Specific research emphasis was placed on the possible effects of RA on these oxLDL-mediated events in the endothelium.
Materials and methods
Materials
Low-glucose DMEM (cat. no. SH30021.01), high-glucose DMEM (cat. no. SH30243.01), RPMI-1640 (cat. no. SH30027.01), fetal bovine serum (FBS; cat. no. SH30070.03), penicillin-streptomycin 100X solution (cat. no. SV30010) and 0.05% trypsin-EDTA (cat. no. SH30236.01) were purchased from HyClone; Cytiva. RA (cat. no. R4033) and human low-density lipoprotein (LDL; cat. no. 437644) were provided by MilliporeSigma. Primary antibodies against ICAM-1 (cat. no. ab109361), VCAM-1 (cat. no. ab134047), phosphorylated (p-)-VE cadherin (Y685; cat. no. ab119785), ZO-1 (cat. no. ab96587) and TXNIP (cat. no. ab188865) were purchased from Abcam. Antibodies against phosphorylated (p-)p-38 (cat. no. 9211s) and FOXO1 (cat. no. 2880s) were purchased from Cell Signaling Technology, Inc. Antibodies against p38 (cat. no. sc-535) and total VE-cadherin (cat. no. sc-6458) were obtained from Santa Cruz Biotechnology, Inc. TurboFect transfection reagent (cat. no. R0531) was purchased from Thermo Fisher Scientific, Inc. Clarity Western Enhanced chemiluminescence western blotting detection reagent (cat. no. BR170-5061) was obtained from Bio-Rad Laboratories, Inc. BCECF (cat. no. C3411) was obtained from Sigma-Aldrich; Merck KGaA. SB203580 (cat. no. 1202) was obtained from Tocris Bioscience. AS1842856 (cat. no. 16761) was procured from Cayman Chemical Company. All other reagents, including MitoTEMPO (cat. no. SML0737), Ac-YVAD-cmk (cat. no. SML0429), 2-mercaptoethanol, protease inhibitor cocktail (cat. no. P8340), 2′,7′-dichlorodihy-drofluores-cein diacetate (cat. no. D6883), N-acetyl-L-cysteine (NAC; cat. no. A9165), SB203580 (cat. no. S8307) and the β-actin antibody (cat. no. a2066), were all obtained from Sigma-Aldrich; Merck KGaA.
Preparation of oxLDL
oxLDL was prepared as described previously by Jin et al (26). Briefly, human LDL (2 mg/363 µl of 150 mM NaCl, pH 7.5, 0.01% EDTA) was added with 1xPBS to make final concentration of 2 mg/1 ml. Then, human LDL was dialyzed against PBS for 16 h at 4°C to remove EDTA and then oxidized with 5 µM CuSO4 for 16 h at 37°C. The extent of LDL oxidation was assessed by measuring the formation of thiobarbituric acid-reactive substances (TBARS; cat. no. 10009055; Cayman Chemical Company) at a wavelength of 540 nm using a microplate reader (VersaMax Microplate Reader; Molecular Devices, LLC). The level of TBARS for oxLDL and native LDL was 13.5±0.5 and 2.5±0.4, respectively.
Cell culture and treatment
The human umbilical vein endothelial cell line EA.hy926 (cat. no. ATCC-CRL-2922) and human monocyte cell line THP-1 were originally obtained from the American Type Culture Collection. EA.hy926 cells and THP-1 cells were grown in DMEM and RPMI-1640 medium, both supplemented with 10% FBS, 100 IU/ml penicillin and 10 µg/ml streptomycin, respectively. All cells were maintained in a humidified incubator at 37°C with 5% CO2. ECs were cultured under HG conditions (25 mM) for 48 h at 37°C and pre-treated with Ac-YVAD-cmk (20 µM), NAC (3 mM), mito-TEMPO (20 µM), AS1842856 (50 nM), SB203580 (0.5 µM) or RA (1, 10, 50 and 100 µM) prior to oxLDL treatment for 1 h at 37°C. Subsequently, 50 µl oxLDL stock solution (2 mg/ml) per 1 ml media was added for additional 24 h at 37°C.
Gene silencing with siRNA
Cells were transfected with 100 nM TXNIP-targeting small interfering RNA (siTXNIP; sense, 5′-GCC GUU AGG AUC CUG GCU UTT-3′ and antisense, 5′-AAG CCA GGA UCC UCC UAA CGG CTT-3′; Bioneer Corporation) or control small interfering RNA (siCTRL; sense, 5′-CCU ACG CCA CCA AUU UCG U-3′ and antisense, 5′-ACG AAA UUG GUG GCG UAG G-3′; Bioneer Corporation) by using the Turbofect transfection reagent (cat. no. R0531; Thermo Fisher Scientific, Inc.) according to the manufacturer's protocols. Briefly, ECs were transfected with control siRNA (100 nM) or TNXIP siRNA (100 nM) in the HG conditions (25 mM) for 48 h and changed with fresh medium before being stimulated with oxLDL (100 µg/ml) for an additional 12 h. After incubation for total 60 h at 37°C, gene silencing efficiency was determined by western blot analysis.
Adhesion assay
EA.hy926 cells (4×105 cells/ml) were seeded in six-well plates and cultured until ~90% confluence under HG conditions (25 mM) for 48 h at 37°C before being treated with oxLDL in the presence or absence of caspase-1, ROS, p38 MAPK and TXNIP inhibitors or RA at the indicated concentrations (1, 10, 50 and 100 µM) for an additional 24 h at 37°C. THP-1 cells (7×105 cells per well) were labeled with the BCECF fluorescent dye for 30 min at 37°C, before being added onto the ECs and incubated for 30 min at 37°C. Following incubation, non-adherent cells were washed three times with PBS and the number of THP-1 cells adhered onto the EC cells was counted using a fluorescence microscope (Eclipse Ti-U, Nikon Corporation). Images were acquired at ×200 magnification from five randomly selected fields per well. Adhered cells were counted using ImageJ version 5.2.0 software (National Institutes of Health).
Transmigration assay
To study the transmigration of THP-1 cells through an EC monolayer, a Transwell system (Falcon® Permeable Support for 24-well Plate with 8.0-µm Transparent PET Membrane; Corning, Inc.) was used. ECs (1×105 cells) cultured in DMEM supplemented with 10% FBS, 100 IU/ml penicillin and 10 µg/ml streptomycin were first seeded into the upper chambers of the Transwell chamber and placed into a 24-well plate. After 4 h at 37°C, THP-1 cells (5×105 cells) were then added to the upper chambers of the Transwell chamber before RPMI medium supplemented with 10% FBS was added into the lower chambers. The migration chambers were then incubated for 24 h in a 37°C cell culture incubator. The number of cells that migrated across the EC monolayer into the lower chamber was immediately counted using 0.4% tryptophan blue staining at room temperature and a Countess II automated cell counter (Thermo Fisher Scientific, Inc.).
Proteins extraction and western blotting
ECs were lysed in RIPA buffer (25 mM Tris-HCl, pH 7.4, 150 mM NaCl, 1% NP-40, 1% sodium deoxycholate and 0.1% SDS) containing a protease inhibitor cocktail. The supernatant was collected after centrifugation at 12,000 × g for 20 min at 4°C and the protein concentration was quantified by using Bradford assay. Aliquots of 20-60 µg total protein from the different experimental groups were then analyzed. The cell homogenates were denatured with 4X SDS sample buffer containing 5% 2-mercaptoethanol, boiled for 5 min at 95°C, loaded into a 8-10% SDS-polyacrylamide gel and separated by electrophoresis for 200 min at 100 V. The protein samples were then transferred onto PVDF membranes for 90 min at 190 mA using a wet transfer system (Bio-Rad Laboratories, Inc.). The membranes were blocked with 5% fat-free dried milk in 1X Tris buffered saline-containing 0.05% Tween-20 (TBS-T) for 1 h at room temperature and then probed with ICAM-1 (1:1,000), VCAM-1 (1:1,000), p-VE-cadherin (1:1,000), VE-cadherin (1:2,000), ZO-1 (1:1,000), p-p38 (1:1,000), p38 (1:2,000), FOXO-1 (1:500), TXNIP (1:1,000) or β-actin (1:5,000) primary antibodies overnight at 4°C. Following primary antibody incubation, the membranes were washed with 1X TBS-T and incubated with HRP-conjugated anti-rabbit, anti-mouse or anti-goat secondary antibodies (1:5,000) for 1 h at room temperature. The immunoreactive bands were visualized by Clarity Western ECL reagent according to the manufacturer's protocol. The relative level of each protein was normalized to the level of β-actin as a loading control. Protein densities were quantified using ChemiDOC™ XRS+ Systems with Image Lab™ Software Version 5.2 (Bio-Rad Laboratories, Inc.).
Statistical analysis
The data were analyzed using GraphPad Prism 7 software (GraphPad Software, Inc.). One-way analyses of variance followed by Tukey's multiple-range test or two-way analyses of variance followed by Bonferroni's test were used to detect significant differences among the mean values. Data are presented as the means ± standard from ≥ three independent experimental repeats. P<0.05 was considered to indicate a statistically significant difference.
Results
oxLDL-mediated expression of ICAM-1 and VCAM-1 in ECs under HG conditions is ROS-, p38-, FOXO1/TXNIP-dependent but RA reduces oxLDL-mediated ICAM-1 and VCAM-1 expression
In a previous study (25), RA reversed oxLDL-mediated NLRP3 inflammasome activation by downregulating ROS production, p38 phosphorylation and FOXO1 protein expression in ECs. Therefore, in the present study, the effects of oxLDL-mediated NLRP3 activation on the expression of adhesion molecules ICAM-1 and VCAM-1 under HG conditions were first examined. Afterwards, the effects of RA on oxLDL-mediated ICAM-1 and VCAM-1 expression were examined. After the ECs were stimulated with oxLDL (100 µg/ml) under low glucose or HG conditions, the expression of ICAM-1 and VCAM-1 were found to be significantly increased, though the magnitude of increase was markedly greater following stimulation under HG conditions (Fig. S1). However, HG conditions without oxLDL stimulation did not result in any changes in the expression of ICAM-1 and VCAM-1 (Fig. S1). oxLDL-mediated ICAM-1 and VCAM-1 expression under HG conditions were found to be significantly reversed by the selective and irreversible inhibitor of caspase-1 Ac-YVAD-cmk (Fig. 1A). Next, the effects of ROS production on oxLDL-mediated ICAM-1 and VCAM-1 expression in ECs under HG conditions were investigated. ROS scavengers mitoTEMPO and NAC were found to significantly reverse oxLDL-induced ICAM-1 and VCAM-1 upregulation in the ECs under HG conditions (Fig. 1B). In addition, the p38 and FOXO1 pathway was potently activated in ECs treated with oxLDL under HG conditions, compared to low glucose conditions (Fig. S2). Therefore, the effects p38 and FOXO1 pathway activation on the upregulation of ICAM-1 and VCAM-1 in ECs stimulated with oxLDL under HG conditions were next tested. The results showed that the p38 inhibitor SB203580 and the FOXO1 inhibitor AS1842856 significantly reversed the oxLDL-induced expression of ICAM-1 and VCAM-1 in ECs under HG conditions (Fig. 1C). Subsequently, transfection with siTXNIP, which significantly knocked down TXNIP expression induced by oxLDL under HG conditions (Fig. S3), also significantly reversed the oxLDL-induced expression of ICAM-1 and VCAM-1 in ECs under HG conditions (Fig. 1D). Finally, oxLDL-mediated ICAM-1 and VCAM-1 upregulation in ECs under HG conditions was found to be significantly reduced by RA in a dose-dependent manner (Fig. 1E). Therefore, the RA-mediated decrease in ICAM-1 and VCAM-1 expression in ECs was possibly dependent on the inhibition of ROS production, p38 phosphorylation, FOXO1 and TXNIP induction.
RA reverses oxLDL-induced endothelial junction permeability under HG conditions
Next, to determine if oxLDL-induced inflammasome activation can increase EC permeability under HG conditions, VE-cadherin phosphorylation and ZO-1 expression were measured in ECs following stimulation with oxLDL under HG conditions. Consistent with the ICAM-1 and VCAM-1 expression results, oxLDL stimulation under HG conditions significantly increased the phosphorylation of VE-cadherin and decreased the expression of ZO-1 in ECs, which was in turn significantly reversed by pretreatment with Ac-YVAD-cmk (Fig. 2A), ROS scavengers mitoTEMPO and NAC (Fig. 2B) or p38 and FOXO1 inhibitors SB203580 and AS1842856, respectively (Fig. 2C). TXNIP knockdown also significantly decreased VE-cadherin phosphorylation whilst restoring ZO-1 downregulation in the presence of oxLDL under HG conditions (Fig. 2D). Furthermore, RA was found to reverse EC dysfunction caused by oxLDL treatment under HG conditions by restoring endothelial junction integrity under HG conditions. As shown in Fig. 2E, RA significantly reversed the oxLDL-induced increases in p-VE-cadherin levels under HG conditions and oxLDL-induced decreases in ZO-1 expression under HG conditions at 50 and 100 µM.
Inhibitory effects of RA on HG- and oxLDL-induced THP-1 monocyte adhesion to ECs is dependent on the ROS-TXNIP/NLRP3 signaling pathway
Next, the effects of inhibiting the expression of adhesion molecules in ECs on THP-1 adhesion to ECs were next investigated. Consistent with the expression data of adhesion molecules aforementioned, Ac-YVAD-cmk (Fig. 3A), the ROS scavengers MitoTEMPO and NAC (Fig. 3B) and both p38 and FOXO1 inhibitors (Fig. 3C) significantly reversed the oxLDL-induced adhesion of THP-1 monocytes to ECs under HG conditions. In addition, Fig. 3D revealed that TXNIP knockdown significantly reversed oxLDL-triggered THP-1 cells adhesion to ECs under HG conditions. RA was also found to significantly diminish oxLDL-induced THP-1 monocyte adhesion to ECs under HG conditions (Fig. 3E). These findings suggest that RA effectively inhibits oxLDL-mediated THP-1 adhesion to ECs under HG conditions through downregulation of adhesion molecule (ICAM-1 and VCAM-1) expression.
oxLDL increases THP-1 monocyte diapedesis under HG conditions in a manner that can be reversed by RA
Subsequently, the effects of oxLDL on THP-1 transmigration through the EC monolayer under HG conditions and the effects of RA on THP-1 transmigration through ECs induced by oxLDL and HG were all examined (Fig. 4A). oxLDL-treated ECs under HG conditions resulted in significantly higher THP-1 transmigration levels through ECs compared with those in untreated control cells, which were in turn significantly reversed by Ac-YVAD-cmk (a caspase-1 inhibitor), MitoTEMPO and NAC (ROS scavengers), SB203580 (a p38 MAPK inhibitor) and AS1842856 (a FOXO1 inhibitor) treatment (Fig. 4B). THP-1 transmigration was significantly reduced on TXNIP siRNA-transfected ECs treated with oxLDL under HG conditions compared with that on CTRL siRNA-transfected ECs (Fig. 4C). RA also significantly reversed oxLDL-induced monocyte transmigration through ECs under HG conditions at 50 and 100 µM (Fig. 4D). These findings suggest that RA can alleviate oxLDL-induced endothelial hyperpermeability and consequent THP-1 transmigration under HG conditions.
Discussion
Endothelial dysfunction is considered to be one of the first stages of atherosclerosis under hyperglycemic and dyslipidemic conditions (27). Endothelial dysfunction increases the expression of adhesion molecules ICAM-1 and VCAM-1 by ECs, which increases the adhesion of monocytes to ECs to facilitate their diapedesis into the subendothelial space, which activates atherosclerotic plaque generation (11). Various pathological conditions, including diabetes and obesity, perturb the integrity of cell junctions, resulting in the increased movement of monocytes and plasma lipoproteins into the intima (20,28). Therefore, regulating the interaction between monocytes and ECs, in addition to endothelial permeability, would greatly limit the instigation of atherogenesis. The overexpression of endothelial adhesion molecules has been previously reported to be key for the pathogenesis of vascular diseases, such as atherosclerosis (29,30). In addition, HG conditions and oxLDL further contribute to the development of atherosclerosis by increasing the expression of ICAM-1 and VCAM-1 (31,32). Previous studies have documented that inflammatory cytokines, such as TNF-α, IL-33 and IL-1β, was positively associated with the increased expression of VCAM-1 and ICAM-1 in vascular and cardiac cells, leukocyte adhesion to ECs and endothelial hyperpermeability (33-35).
RA has been previously reported to exert beneficial effects on the cardiovascular system because of its antioxidant and anti-inflammatory properties. It reduced adriamycin-induced cardiotoxicity in H9C2 cells by inhibiting ROS production (36), in addition to inhibiting atherosclerotic plaque formation in Apolipoprotein E−/− mice fed on a high-cholesterol diet, by reducing the serum levels of proinflammatory cytokines TNF-α and IL-1β (37). In a previous study, it was shown that RA protected ECs by inhibiting oxLDL-mediated NLRP3 inflammasome activation and resultant IL-1β production under HG conditions (25). Therefore, for further study, the present study investigated the effects of RA on oxLDL-mediated interactions between monocytes and ECs, endothelial permeability and transmigration of monocytes through endothelial monolayers. ECs stimulated with oxLDL under HG conditions were found to exhibit significantly higher expression levels of adhesion molecules ICAM-1 and VCAM-1, which resulted in increased THP-1 monocyte adhesion onto EC monolayer. oxLDL-induced endothelial inflammation under HG conditions has been reported to be orchestrated by ROS-mediated p38 phosphorylation, FOXO1/TXNIP induction and IL-1β production (25,38,39). Correspondingly, ECs pretreated with the irreversible inhibitor of caspase-1 (an IL-1β converting enzyme), ROS scavengers and p38 and FOXO1 inhibitors all showed significant reductions in oxLDL-induced ICAM-1 and VCAM-1 protein expression under HG conditions. Notably, TXNIP knockdown also reversed HG- and oxLDL-mediated overexpression of ICAM-1 and VCAM-1 in ECs. In particular, inhibitors of ROS, p38 MAPK, FOXO1 and TXNIP, all of which are involved in the expression of the adhesion molecules ICAM-1 and VCAM-1, significantly reduced oxLDL-induced THP-1 monocyte adhesion to ECs under HG conditions. RA, which was previously found to downregulate the activity of the p38/FOXO1/TXNIP pathway and inhibit inflammasome activation in oxLDL- and HG-stimulated ECs (22), diminished oxLDL-mediated upregulation of ICAM-1 and VCAM-1 expression in ECs under HG conditions whilst significantly reducing THP-1 monocyte adhesion to ECs in a dose-dependent manner. oxLDL-induced expression of adhesion molecules (ICAM-1 and VCAM-1) was found to be markedly higher under HG conditions compared with that stimulated under low glucose conditions. In addition, oxLDL treatment increased the levels of NLRP3 activation pathway mediators p-p38, FOXO-1 and TXNIP under low glucose conditions, which were enhanced under HG conditions. However, HG without oxLDL stimulation did not exert any changes in the expression of ICAM-1 and VCAM-1 whilst weakly increasing p-p38, FOXO-1 and TXNIP levels. These results suggest that oxLDL enhances the inflammatory response under HG conditions but not under low glucose conditions or HG conditions without oxLDL.
Endothelial inflammation alters junction integrity and increases endothelial permeability (40,41). Vascular permeability and monocyte diapedesis are increased by the phosphorylation of VE-cadherin, a major component of adherens junctions (18,42), and by the downregulation of ZO-1, a key molecule of tight junctions (16). HG and oxLDL were found to increase VE-cadherin phosphorylation but decrease ZO-1 expression in ECs in the present study, which in turn significantly induced THP1 monocyte transmigration through ECs. Pretreatment of the ECs with ROS scavengers, inhibitors of caspase-1, p38 MAPK and FOXO1 and transfection with TXNIP siRNA all reversed the oxLDL-induced phosphorylation of VE-cadherin and ZO-1 downregulation under HG conditions. In addition, inhibiting ROS/p38 MAPK/TXNIP inflammasome pathway activation significantly reduced monocyte diapedesis through the EC monolayers. RA also reversed the oxLDL-increased phosphorylation of VE-cadherin and downregulation of ZO-1 expression in a dose-dependent manner, which resulted in the blockage of monocyte diapedesis through EC monolayers.
Taken together, results of the present study revealed that oxLDL can trigger the overexpression of adhesion molecules in ECs under HG conditions, which leads to the adhesion of monocytes to ECs. In addition, oxLDL stimulation increases endothelial permeability, promoting monocyte diapedesis. Based on previous findings and the results of this study, RA effectively prevented the expression of adhesion molecules and THP-1 monocyte adhesion to ECs whilst inhibiting endothelial monolayer leakage and THP-1 monocyte diapedesis, by regulating oxLDL-mediated ROS/p38 MAPK/FOXO1/TXNIP/inflammasome activation under HG conditions (Fig. 5). These findings suggest that RA exerts protective effects against hyperlipidemia- and hyperglycemia-induced cardiovascular disease by modulating the interaction between monocytes and ECs, as well as monocyte diapedesis.