Substrates with different stiffness were prepared as described previously by adjusting the ratio of 40% acrylamide to 2% bis-acrylamide (w/v) (Sangon Biotech, China) (Chen et al., 2019; Tian et al., 2019). Tetramethylethylenediamine (TEMED, Klamar, China) and 10% ammonium persulfate solution (APS, Sinopharm, China) were used as cross-linking reagent. Before polymerization, 200 μl mixture was added onto a glass slide pre-treated with gel slick (Lonza, Switzerland), and a piece of silicified cover slip was put upon the gel solution to make a sandwich. After 5 min, the substrates were treated with sulfo-SANPAH (Thermo Fisher, USA) and coated with 1 mg/mL collagen I (Corning, USA) to facilitate the cell adhesion. Finally, the substrate was sterilized by 75% alcohol and UV before cell culture.

Human aortic smooth muscle cells (SMCs) were purchased from ScienCell Research Laboratories (ScienCell, USA) and used for experiments at passages 5–10. Cells were cultured in SMC medium (ScienCell, USA) supplemented with 2% fetal bovine serum, 1% penicillin/streptomycin, and 1% smooth muscle cell growth supplement (SMCGS) in a 37°C incubator (Corning, USA) with 5% CO₂.

To induce the foam cell state, SMCs were primed by Chol: MβCD complex (20 μg/ml) (Sigma, USA), a “water-soluble cholesterol” for 72 h.

In the inhibition experiments, SMCs were treated with GSK2033 (MedChemExpress, China), a liver X receptor (LXR) inhibitor, at a final concentration of 1 μM in the culture medium for 72 h. DMSO was used as control.

In the agonist experiments, SMCs were treated with GW3965 (Sigma, USA), a selective agonist of LXR, at a final concentration of 1 μM in the culture medium for 72 h. DMSO was used as control.

To optimize the concentration of Chol: MβCD, SMCs were seeded in a 96-well plate and treated with a concentration gradient of Chol: MβCD (0, 10, 20, 40, and 80 μg/ml) for 72 h at standard culture condition. Then, the culture media was removed and the cell were incubated with CCK-8 reagent (DOJINDO, Japan) for 1 h. Optical density (OD) at 450 nm of each well was determined using a plate-reader (Bio Tek, Synergy 2, USA). The OD of the experiment and the control groups were both corrected by subtracting the OD of the blank. Cell viability was expressed as the percentage OD_{450 nm} of experimental group to that of the control group.

Oil Red O staining was used to quantify the amount of cholesterol in the SMCs. Briefly, after cholesterol loading, SMCs were fixed in 4% (w/v) paraformaldehyde (PFA) and incubated in 60% isopropanol for 3 min. Then, cells were stained with Oil Red O and Harris' hematoxylin. Images were captured by a light-microscope (DMI3000B, Leica, Germany). The cellular cholesterol was extract by isopropanol and OD value at 520 nm was determined to represent the cholesterol concentration.

The cholesterol uptake by SMCs was identified by BODIPY 493/503 staining (Qiu and Simon, 2016). After 20 μg/ml cholesterol treatment for 72 h, SMCs were washed 3 times with Dulbecco's phosphate buffered saline (DPBS) and 2 μM BODIPY 493/503 was added (Cayman, USA) for 15 min at 37°C in dark. Then, cells were fixed in 4% (w/v) PFA for 15 min, and incubated with DAPI for 5 min to stain the nucleus. Images were captured by a laser scanning confocal microscope (Leica, TCS SP5 II, Germany). Cholesterol uptake was quantified by relative fluorescence intensity (FI).

Cellular total cholesterol (TC) was determined by TC assay kit (NanJing Jiancheng Bioengineering Institute, China) according to manufacturer's instructions. The cellular cholesterol concentration was normalized by total protein content determined by BCA Protein Assay Kit (Thermo Scientific, USA).

Cholesterol concentration is proportional to the intensity of the 20S-methyl-21-[(7-nitro-2,1,3β-benzoxadiazol-4-yl) amino]-pregn-5-en-3-ol (NBD cholesterol) (J&K Scientific, China) fluorescence signal. After cholesterol loading, the SMCs were incubated with NBD cholesterol for 8 h and Apoa1 for additional 8 h to induce cholesterol efflux. The culture medium comprised of the cholesterol efflux was collected and the SMCs were lysed with 2% Triton X-100 to release the cholesterol remained in the cells. Both the efflux and the supernatant from the lysed cells (non-efflux) were subject for fluorescence intensity (FI) measuring at 528±20 nm (with 485±20 nm excitation wavelength). Cholesterol efflux rate = efflux FI / (efflux FI + non-efflux FI) × 100%.

Total RNA was extracted with the RNA simple Total RNA Extraction kit (TIANGEN, China) and reverse transcribed into cDNA by using the Fastking RT kit (with gDNase) (TIANGEN, China), following the manufacturer's instruction. The mRNA level was determined by RT-qPCR (Thermo Fisher Scientific, USA) using the Talent qPCR PreMix Kit (SYBR Green) (TIANGEN, China) and the specific primer pairs was listed in Table 1. All data were normalized to GAPDH and expressed as fold change over the controls.

Total protein of SMCs was extracted by the plus RIPA lysis buffer (Beyotime Biotechnology, China) with protease inhibitor. BCA Protein Assay Kit (Thermo Scientific, USA) was used to quantify the protein concentration. Then, equal amount of the total cell lysates was loaded for SDS-PAGE electrophoresis, and transferred to PVDF membrane. After 10% BSA blocking, the membrane was incubated with primary antibody and incubated overnight at 4°C. The primary antibodies used in this research are ACTA2 (Abcam, UK), CD68 (Abcam, UK), ABCA1 (Novus, USA), α-Tubulin (Cell Signaling Technology, USA), and GAPDH (Cell Signaling Technology, USA). After removing the primary antibody, HRP-conjugated secondary antibodies (Cell Signaling Technology, USA) were added and incubated with the membrane for 1 h at room temperature and visualized by Immobilon western chemiluminescent HRP substrate (Merck Millipore, Germany). The protein expression was normalized by the expression of α-Tubulin or GAPDH in respective samples.

To characterize the SMCs phenotype, cells were fixed in 4% (w/v) PFA, and incubated with primary antibodies of ACTA2 or CD68 overnight at 4°C. Then, cells were washed by DPBS and incubated with the fluorescent labeled secondary antibodies accordingly for 1 h at room temperature. Finally, cell nucleus was stained with DAPI for 5 min. Images were captured under a laser scanning confocal microscope (Leica, TCS SP5 II).

SMCs were seeded on substrates with different stiffness overnight to reach 50–70% confluency before transfection. The negative siRNA control (NC) or siRNA targeting ABCA1 (siABCA1) (Guangzhou RiboBio, China) was transfected into SMCs at a concentration of 50 nM in the culture medium according to manufacturer's instructions. Twenty-four hours after the transfection, cholesterol was added into the cell culture medium and the cells were treated by cholesterol for 24 h.

Unless otherwise indicated, the results were shown as mean ± SEM (n = 3), and student t-test was used for pairwise comparisons for the same stiffness between groups and the different stiffness within groups. For all cases, p < 0.05 was considered statistically significant. Data were analyzed by GraphPad Prism 6.0 software.

To confirm the effects of loading on SMCs phenotype switching, Chol: MβCD, a “water-soluble cholesterol” that can be efficiently and quickly uptaken by cells, was used to treat SMCs (Christian et al., 1997; Qin et al., 2006). The cholesterol concentration was titrated to show that little cytotoxicity was detected at 20 μg/ml (or lower) after 72 h (Figure 1A). Meanwhile, the cellular uptake of cholesterol was confirmed by Oil Red O staining and quantitation. After Chol: MβCD treatment for 72 h, 20 μg/ml group showed more cholesterol droplets than the other two groups (Figure 1C). In addition, Oil Red O stained cellular cholesterol was extracted by isopropanol and the OD values were compared (Figure 1B).

Furthermore, to characterize the SMCs phenotype switching from SMCs to foam cell-like state, the gene expression level of SMC maker (ACTA2) and macrophage marker gene (CD68 and LGALS3) were analyzed by RT-qPCR. The continuous cholesterol accumulation significantly reduced ACTA2 expression and promoted CD68 and LGALS3 expression (Figures 1D,E). Moreover, ACTA2 and CD68 expression was confirmed by immunostaining and Western blotting analysis (Figures 1F,G).

To explore the possible influence of substrate stiffness on SMCs cholesterol level, a lipophilic fluorescent probe, BODIPY 493/503, was used to label cellular neutral lipid contents. SMCs were seeded on substrate of different stiffness and treated by cholesterol at indicated times points. The cholesterol accumulation in SMCs showed a time-dependent increase during 72 h and significant difference at different substrate stiffness (Figures 2A,B). At the substrate stiffness of 1 and 100 kPa, cholesterol levels in SMCs were significantly higher than other groups at all time points. This was further confirmed by the quantitation of total cholesterol (TC) level where 1, 40, and 100 kPa were selected as representative stiffnesses (Figure 2C). These results suggested that the stiffness of the substrate affected the accumulation of cholesterol in SMCs.

We further determined whether the changes in substrate stiffness would affect the SMCs phenotype. ACTA2 was chosen to represent the contractile phenotype of the SMCs and CD68 as a marker for macrophage phenotype. RT-qPCR results showed that there was a decrease in gene expression of SMC contractile maker ACTA2 and an increase in macrophage marker CD68 (Figures 3A,B) upon cholesterol treatment. While ACTA2 expression increased with substrate stiffness, the suppressive effect of cholesterol was consistent for all stiffness (Figure 3A). The treatment of cholesterol demonstrated a very strong external force to drive SMCs to express stronger macrophage phenotype. However, consistent with the low cholesterol accumulation on 40 kPa substrate, cholesterol treatment demonstrated a significantly lower CD68 expression on 40 kPa compared with 1 and 100 kPa (Figure 3B), implying that substrate stiffness exerted a regulatory effect on SMC phenotype. The protein expression level of both ACTA2 and CD68 was consistent with the mRNA level (Figures 3C,D). SMCs on the 40 kPa substrate showed a relatively smaller change in macrophage marker expression (Figures 3E,F), suggesting less phenotypic change occurred on 40 kPa substrate.

Although cellular cholesterol homeostasis is a combined outcome of the uptake, synthesis, and efflux of cholesterol, cholesterol efflux is the key event of RCT to remove excess cholesterol. Cholesterol uptake by SMCs is largely mediated by non-specific pinocytosis (Rivera et al., 2013); therefore, we focused on the effect of substrate stiffness on the regulation of cholesterol efflux. Cholesterol efflux is mediated by ABCA1, an active transmembrane transporter related to the transmembrane transport of cholesterol and phospholipids. Interestingly, we found that ABCA1 mRNA expression on 40 kPa substrate was significantly higher than 1 and 100 kPa substrates (Figure 4A). Immunostaining showed that the expression of ABCA1 is higher on 40 kPa substrate compared to that on 1 and 100 kPa ones (Figure 4B). Western blotting analysis confirmed the protein level of ABCA1 being consistent with the gene expression level (Figures 4C,D). Therefore, the regulatory effect of substrate stiffness on cholesterol accumulation in SMC is related to ABCA1 expression and cholesterol efflux.

To further understand the influence of substrate stiffness on SMCs cholesterol efflux, we modulated the LXR signaling pathway, the upstream regulator of ABCA1, to regulate the expression of ABCA1. LXR has been reported as a sterol-responsive nuclear receptor. GW3965 is a selective agonist to LXR and GSK2033 is a selective LXR antagonist (Collins et al., 2002; Delvecchio et al., 2007; Helder et al., 2020). The activation effect of GW3965 and the inhibition effect of GSK2033 on LXR expression was confirmed (Figure 5A). Activation of LXR by GW3965 treatment increased both the mRNA and protein expression of ABCA1 in SMCs (Figures 5B,C). Moreover, the expression level of LXR and ABCA1 on 40 kPa was higher than that on 1 and 100 kPa substrates. In contrast, inhibition of LXR by GSK2033 treatment almost abolished both the mRNA level and protein expression level of ABCA1 on all substrates with different stiffnesses (Figures 5B,D). As a result, with the upregulation of LXR and cholesterol treatment for 72 hours, SMC cholesterol efflux increased significantly (p < 0.05). In addition, compared with 1 and 100 kPa, 40 kPa stiffness induced the highest cholesterol efflux rate in SMCs, suggesting the additive effect of substrate stiffness and LXR signaling. In contrast, downregulating LXR by GSK2033 treatment blocked cholesterol efflux of SMCs, regardless of substrate stiffness (Figure 5E). Furthermore, change in cholesterol efflux resulted in the change of the cellular cholesterol content. BODIPY 493/503 staining demonstrated that LXR agonist GW3965 treatment led to very low cholesterol accumulation in SMCs, and there is no observable difference among the different substrate stiffness. However, inhibition of LXR resulted in the inhibition of cholesterol efflux. Consequently, the cellular cholesterol level was increased on all substrate stiffness (Figure 5F).

To further verify that ABCA1 is a key mediator of substrate stiffness regulating cholesterol accumulation in SMCs, we knocked down cholesterol efflux gene ABCA1. The results showed that siABCA1 effectively reduced ABCA1 expression (Figure 6A). BODIPY 493/503 staining results showed that, compared with the control group, the cholesterol accumulation level in SMCs of each substrate stiffness was higher in the siABCA1 group (Figures 6B,C).