Human umbilical vein endothelial cells (HUVECs) and THP-1 monocytes were acquired from Type Culture Collection of the Chinese Academy of Sciences (Shanghai, China). HUVECs were nurtured in DMEM medium comprising 1% penicillin/streptomycin and 10% fetal bovine serum (FBS). When HUVECs were fused into monolayer cells, they were passaged with 0.125% pancreatin-0.02% EDTA digestion. The 3-5th generations of HUVECs were used for experiments. HUVECs were starved for 10 h in DMEM medium containing 1% FBS before being pre-treated with tumor necrosis factor-α (TNF-α, 50 ng/mL, Peprotech, Rocky Hill, NJ, USA) for 6 h [19, 20]. THP-1 monocytes were incubated in RPMI-1640 medium, supplemented with 1% penicillin and streptomycin, 10% FBS, 1% L-glutamine, 1% HEPES buffer and 0.05 mM β-mercaptoethanol. Both HUVECs and THP-1 monocytes were nurtured at 37 °C in a 5% CO₂ incubator.

An eight-folds of S1P (Cayman, Ann Arbor, MI, USA) was added to ApoM (Abnova, Taiwan, China) and incubated for 30 min at room temperature to load ApoM with S1P. Unbound S1P was removed using a desalting PD10 column (GE Healthcare, Beijing, China) after the mixture was incubated with serum-free M200 medium. Then ApoM-S1P was diluted to 0.5 μM and stored at  -20 °C.

Before adhesion assays, HUVECs were pretreated with 50 ng/mL of TNF-α for 6 h. Then HUVECs were treated with ApoM (1.0 μM), S1P (1.0 μM), or ApoM-S1P for 24 h [21] and grouped into TNF-α group, Control group, TNF-α + ApoM group, TNF-α + S1P group, and TNF-α + ApoM + S1P group. Antagonists for S1P receptors (S1PR1, S1PR2 and S1PR3) were added in TNF-α stimulated HUVECs for 30 min before cells were subjected to ApoM-S1P treatment. According to antagonist treatment, cells were grouped into TNF-α + ApoM + S1P + W146 group, TNF-α + ApoM + S1P + JTE-013 group, and TNF-α + ApoM+S1P + CAY10444 group, respectively.

THP-1 monocytes were labeled with 5 μmol/L calcein-acetoxymethyl ester (Thermo Fisher Scientific, Waltham, MA, USA) and cultured in the incubator for 30 min. The labeled cells were washed twice with PBS and re-suspended in serum-free medium. Then culture medium of HUVECs was removed. The resuspensions of labeled THP-1 monocytes were added onto monolayers of HUVECs. After 1 h incubation, the plates were rinsed twice with medium without serum. The adherent THP-1 monocytes on HUVECs were counted with a fluorescence microscope (Olympus IX71, Tokyo, Japan). The culture dish was divided into four quadrants according to its diameter. For each quadrant, one 200 × visual field in the middle of the quadrant was chosen for analysis. The average number of the four quadrants was identified as the number of adherent THP-1 monocytes.

Total RNA of HUVECs was extracted by RNeasyPlus Micro (Qiagen, dusseldorf, Germany). The ratio of OD_{260 nm} to OD_{280 nm} was measured to quantify and test the purity of RNA. Reverse Transcription of RNA into cDNA was conducted by Reverse Transcription System (Promega, Madison, WI, USA). The mRNA expression levels of E-selectin, ICAM-1, VCAM-1, IL-1β, IL-18, caspase-1, ASC, NLRP3, IL-6, and IL-10 were detected by qRT-PCR using SYBR Premix ExTaq (Takara Bio, Inc., Otsu, Japan). GAPDH was regarded as internal reference. All tests were conducted in triplicates. All data and figures were depicted as the mean ± standard deviation (SD), and data analysis was employed the 2^-ΔΔCt method [22]. The amplified primer sequences of each gene and its primers are shown in Table 1.

Cell lysates were made from HUVECs using RIPA buffer comprising phosphatase inhibitors (Cayman Chemical, Ann Arbor, MI, USA) and Halt protease inhibitors (Pierce, Rockford, IL, USA). Bradford assay was adopted to test the concentration of proteins. The cell lysates were mixed with loading buffer and separated on a 10% SDS polyacrylamide gels. The proteins on gels were then transferred onto PVDF membrane. The membranes were blocked in TBST containing 0.05 g/mL BSA for 1 h. After blocking, the membranes were incubated overnight at 4 °C with primary antibodies (abcam, san francsico, USA), rabbit-derived GAPDH (1:10000, ab181602), E-selectin (1:200, ab18981), VCAM-1 (1:2000, ab134047), ICAM-1 (1:2000, ab53013), caspase-1(1:500, ab138483), NLRP3 (1:500, ab214185), ASC (1:1000, ab155970), PI3K (1:1000, ab191606), p-PI3K (1:1000, ab182651), AKT (1:10000, ab179463), p-AKT (T308, 1:1000, ab38449) and p-AKT (Ser473,1:2000, 4060 T, Cell Signaling Technology, Boston, USA). The membranes were washed in TBST and then incubated at room temperature for 2 h with goat anti-rabbit IgG (1:5000, Beijing ComWin Biotech Co., Ltd., Beijing, China). The membranes were then rinsed three times with TBST. ECL luminescent solution was used before a chemiluminecence imaging analysis system (GE Healthcare, USA) was applied for observation.

The supernatants from untreated and treated HUVECs were centrifuged (300 g, 5 min) and cell-free supernatants were frozen at  -80 °C until usage. The assays were conducted utilizing commercial ELISA kits (abcam, san francsico, USA) in duplicates, according to the manufacturers’ instructions. The microplate absorbance reader (Tecan, Groening, Austria) was employed to measure the absorbance at 450 nm. Standard curves were applied to calculate the concentrations of analytes.

HUVECs were seeded in 24-well plates at the density of 5 × 10³ cells/well, followed by fixation in 4% paraformaldehyde for 15 min, permeation with 0.5% triton-X-100 for 20 min, and incubation with appropriate concentration of primary antibody of caspase-1 (1:500, ab1872, abcam, san francsico, USA) overnight at 4 °C. After that, sections were incubated with secondary antibody of Alexa Fluor 488-labeled goat anti-rabbit IgG (ab150077, 1:2000, abcam, MA, USA) for 45 min avoiding light, prior to PBS rinsing. Then cells were stained with 0.5 μg/mL of DAPI for 15 min avoiding light, and washed with PBS. Subsequently, sections were sealed and observed under a fluorescence microscope (Olympus IX71, Tokyo, Japan).

The ROS was measured using DCF-DA kit (BestBio, Shanghai, China) according to the manufacturer’s instructions. Cells (1 × 10⁷/mL) were labeled with 10 μM of fluorescent probe DCFH-DA, followed by incubation in an incubator for 20 min at 37 °C, and serum-free cell culture solution washing for three times. The fluorescence activity was analyzed by fluorescence microscopy with an excitation wavelength of 488 nm and an emission wavelength of 520 nm. The amount of intracellular ROS was proportional to the fluorescence intensity of DCF, and the fluorescence intensity indicated the ROS level.

All data were analyzed using SPSS 17.0 software (SPSS, Chicago, IL, USA) and GraphPad Prism 6.0. T test was used for comparison between two groups, and differences among groups were analyzed using One-way analysis of variance. P < 0.05 was regarded as a statistically significant difference.

In the early stages of AS, endothelial cell function is impaired, and the expression of adhesion molecules is significantly increased, thereby recruiting pro-inflammatory immune cells to adhere to the vascular wall [23]. In this study, we discovered that the number of THP-1 monocytes adhering to the surface of HUVECs in the TNF-α group was significantly increased compared with the Control group (Fig. 1a), and up-regulate the mRNA and protein levels of adhesion molecules E-selectin, ICAM-1, VCAM-1 (P < 0.001) (Fig. 1b-c). ELISA illustrated that the contents of three adhesion molecules in TNF-α group were higher than those in the Control group (P < 0.05) (Fig. 1d). All these stated that TNF-α could enhance the adhesion function of HUVECs and induce the generation of adhesion molecules.

Pyroptosis is a more recently identified pathway of programmed cell death and accompanied with the severe inflammation response [24]. The detection on effect of TNF-α treatment on pyroptosis of HUVECs displayed that the mRNA levels of caspase-1, ASC, NLRP3, IL-1β and IL-18 in TNF-α group were higher than those in Control group, while anti-inflammatory cytokine were lower than Control group (P < 0.01) (Fig. 1e). Consistent expression pattern was found on protein expressions of pyroptosis and inflammation related factors (P < 0.05) (Fig. 1f). ELISA results manifested that compared with Control group, the contents of IL-1β, IL-18, and pro-inflammatory cytokine IL-6 in TNF-α group were increased, while the content of anti-inflammatory cytokine IL-10 was decreased (P < 0.01) (Fig. 1g). Subsequently, immunofluorescence was utilized to inspect the expression of caspase-1 in HUVECs. The results manifested that TNF-α group had higher caspase-1 expression than the Control group (Fig. 1h). Meanwhile, DCF-DA kit detection results described that there was elevated ROS level in TNF-α group in comparison to the Control group (P < 0.001) (Fig. 1i). These indicated that TNF-α could enhance inflammation and promote pyroptosis of HUVECs.

ApoM, a specific S1P chaperone, is a novel apolipoprotein associated with anti-inflammatory functions [25]. Study has illustrated the effect of ApoM-S1P in mediating anti-inflammatory effects on endothelial cells [23]. Hence, we investigated whether ApoM-S1P treatment could affect TNF-α-induced adhesion and pyroptosis of HUVECs. Fluorescence microscope observed that compared with the TNF-α + ApoM group or TNF-α + S1P group, TNF-α + ApoM + S1P group had suppressed adhesion quantity of THP-1 monocytes (P < 0.001) (Fig. 2a). In addition, ApoM-S1P treatment could substantially decrease the protein expression levels of VCAM-1, E-selectin, and ICAM-1 (P < 0.01). While there were no significant changes of adhesion quantity of THP-1 monocytes in both TNF-α + ApoM group and TNF-α + S1P group in comparison to that in TNF-α group (P > 0.05) (Fig. 2b-d). In addition, qRT-PCR results explained that TNF-α + ApoM + S1P group had a suppressed the mRNA levels of IL-1β, IL-18, caspase-1, ASC, NLRP3 and IL-6, and an elevated mRNA level of IL-10 than TNF-α + ApoM group or TNF-α + S1P group (P < 0.01) (Fig. 2e). The results obtained by Western blot and ELISA supported the protective effect of ApoM + S1P on pyroptosis and inflammation of HUVECs (P < 0.01) (Fig. 2f-g). In addition, compared with TNF-α group, no significant difference was revealed in TNF-α + ApoM group and TNF-α + S1P group (P > 0.05) (Fig. 2e-g). Then immunofluorescence was employed to assay caspase-1 expression in HUVECs. The results presented that TNF-α + ApoM+S1P group had heightened caspase-1 expression compared with the TNF-α + ApoM group or TNF-α + S1P group (Fig. 2h), whereas caspase-1 expression showed no obvious difference in cells treated with ApoM or S1P alone when compared with cells in TNF-α group (TNF-α + ApoM group or TNF-α + S1P group vs TNF-α group; P > 0.05) (Fig. 2h). Meanwhile, DCF-DA kit detection results displayed that TNF-α + ApoM+S1P group had elevated ROS level in comparison to TNF-α + ApoM group or TNF-α + S1P group (P < 0.001) (Fig. 2i), and there was no significant difference of ROS level in either TNF-α + ApoM group, or TNF-α + S1P group when compared with TNF-α group (P > 0.05). These results indicated that ApoM-S1P can observably reduce TNF-α-induced adhesion and cell pyroptosis.

To further investigate the mechanism of ApoM+S1P inhibition on inflammatory factors and adhesion molecules, TNF-α-induced HUVECs were pretreated with specific S1PR1 antagonist W146, S1PR2 antagonist JTE-013, or S1PR3 antagonist CAY10444 for 30 min before ApoM-S1P treatment. Compared with TNF-α + ApoM + S1P group, TNF-α + ApoM + S1P + JTE-013 group had an elevated adhesion quantity of THP-1 monocytes, while TNF-α + ApoM + S1P + W146 group and TNF-α + ApoM + S1P + CAY10444 group showed no difference with TNF-α + ApoM+S1P group (Fig. 3a). Detection on adhesion molecules demonstrated that E-selectin, ICAM-1 and VCAM-1 were increased after JTE-013 treatment in comparison to TNF-α + ApoM + S1P group (Fig. 3b-c). Results also demonstrated that JTE-013 could reverse the protective impact of ApoM + S1P on pyroptosis of HUVECs, evidenced by increased IL-1β, IL-18 and IL-6, and decreased expression of IL-10 (P < 0.001) (Fig. 3d). No substantial difference was uncovered for TNF-α + ApoM + S1P + W146 group and TNF-α + ApoM + S1P + CAY10444 group when compared with TNF-α + ApoM + S1P group. Subsequently, DCF-DA kit detection results manifested that JTE-013 group had elevated ROS level in comparison to TNF-α + ApoM + S1P group (P < 0.001) (Fig. 3e), and there was no markedly difference of ROS level between W146 group, CAY10444 group, and TNF-α + ApoM + S1P group (P > 0.05). These results indicate that ApoM-S1P inhibits TNF-α-induced adhesion and cell pyroptosis by binding to S1PR2.

Previous evidence has demonstrated that PI3K/AKT is a downstream signaling pathway of S1P [26]. Therefore, we hypothesized that ApoM+S1P may exert its protective role in AS though PI3K/AKT signaling pathway. In this regards, we detected the effect of ApoM, S1P or/and SIPR2 antagonist JTE-013 on phosphorylation level of PI3K and AKT with the application of Western blot. The analysis revealed that TNF-α group had a lower phosphorylation level of PI3K and AKT than Control group. TNF-α + ApoM + S1P could partially restore the phosphorylation level of PI3K and AKT when compared with TNF-α group, while SIPR2 antagonist JTE-013 could inhibit the phosphorylation level of PI3K and AKT in comparison to ApoM-S1P treatment alone (P < 0.001). The phosphorylation level of PI3K and AKT in TNF-α + ApoM + S1P group was substantially higher than those in TNF-α + ApoM group and TNF-α + S1P group (P < 0.001) (Fig. 4). These indicated that ApoM-S1P could reverse the inhibitory effect of TNF-α on PI3K/AKT signaling pathway via mediating SIPR2.