**3. Results**

#### *3.1. Non-Cytotoxic Doses of HLP Inhibits TNF-*α*-Induced Cell Viability Loss and MMP Activation*

A7r5 cell viability was investigated following incubation with a range of concentrations (from 1 to 20 ng/mL) of TNF-α for 24 h, and it was found that TNF-α at low concentrations (lower than 10 ng/mL) dose-dependently increased the cell viability. However, above the dose of 10 ng/mL, TNF-α reduced about 10% of cell viability (Figure 1a). Because MMPs break down components of ECM, which is a crucial role in the process of VSMC migration [7,8], the effect of TNF-α on MMP activities was then tested by gelatin zymography in serum-free conditioned medium to identify the contribution of MMP-2 or MMP-9 to the pro-migratory ability of TNF-<sup>α</sup>. As shown in Figure 1b, MMP-9 activity was tremendously increased by TNF-α in a concentration-dependent manner, whereas MMP-2 activity was less affected. According to the results, to provide a maximum dynamic range for quantifying the VSMC proliferative and pro-migratory responses, cell incubation with 10 ng/mL of TNF-α for 24 h was chosen in all subsequent experiments.

In our previous study, HLP at concentrations of > 0.05 mg/mL was demonstrated to be an antioxidant agent, as tested by its 1,1-diphenyl-2-picrylhydrazyl (DPPH) scavenging effect and ability to inhibit LDL oxidation in standard antioxidant evaluation [14], as shown in Table S4. Next, a preliminary screening was performed to study the effect of HLP alone (Figure S1) or together with TNF-α at 10 ng/mL (Figure 1c) on A7r5 cell growth for 24 h, using the MTT assay. The viability of A7r5 was significantly decreased by 0.25, 0.50, and 1.0 mg/mL of HLP in the absence or presence of TNF-α in a dose-dependent manner, when receptively compared to the control and TNF-α alone group. In order to study whether HLP is an inhibitor of cell migration and MMP-9 activation in the TNF-α-treated VSMC, the effect of HLP on A7r5 cell viability by MTT assay, showing cell growth, was significantly altered by the treatments of above the dose of HLP at 0.25 mg/mL, and was excluded in further studies (Figure 1c). In subsequent experimental migration research, the concentration range was used to avoid the influence of cell viability on the observed parameters. As shown in Figure 1d, it is worth noting the TNF-α-induced increase in MMP-9 activity was significantly inhibited in the cells incubated with the combinations of TNF-α together with this dose range of HLP (between 0.01 and 0.10 mg/mL).

**Figure 1.** Effect of tumor necrosis factor-alpha (TNF-α) or/and *Hibiscus* leaf polyphenol (HLP) on cell viability and matrix metalloproteinase (MMP) activities in vascular smooth muscle cells (VSMCs). (**a**) A7r5 cells were treated with various concentrations (0–20 ng/mL) of TNF-α for 24 h. Cell viability was analyzed by MTT assay. (**b**) A7r5 cells in serum-free medium were treated with various concentrations of TNF-α for 24 h. The culture medium of cells after treatment was subjected to gelatin zymography to analyze the MMP activity. (**c**) A7r5 cells were treated with TNF-α (10 ng/mL) in the absence or presence of various concentrations (0.01, 0.05, 0.10, 0.25, 0.50 and 1.00 mg/mL) of HLP for 24 h. Cell viability was analyzed by MTT assay. The quantitative data are presented as mean ± standard deviation (SD) (*n* = 3) from three independent experiments. # *p* < 0.05, ## *p* < 0.01 compared with the control. \* *p* < 0.05, \*\* *p* < 0.01 compared with the TNF-α group. (**d**) A7r5 cells in serum-free medium were treated with TNF-α in the absence or presence of various concentrations of HLP for 24 h. The culture medium of cells after treatment was subjected to gelatin zymography to analyze the MMP activity. The result is representative of at least three independent experiments. +, added; −, non-added.

#### *3.2. HLP Downregulated TNF-*α*-Increased Protein and mRNA Levels of MMPs*

To understand further the downregulatory effects of HLP on the TNF-α-activated MMP-9, Western blotting was performed. As shown in Figure 2a, TNF-α elevated the protein levels of MMP-2 and MMP-9, and TNF-α together with the indicated concentrations of HLP (0.01, 0.05, and 0.10 mg/mL) caused a marked decreased level of MMP-9, but not MMP-2. The HLP-mediated decrease in the protein level of MMP-9 coincided well with its mRNA level, as evidenced by quantitative RT-PCR results (Figure 2b), indicating that HLP might downregulate the expression of MMP-9 majorly, but that of MMP-2 partially, at the transcriptional level.

**Figure 2.** Effect of HLP on TNF-α-induced protein and mRNA levels of MMPs in VSMCs. A7r5 cells were treated with TNF-α (10 ng/mL) in the absence or presence of various concentrations (0, 0.01, 0.05, and 0.10 mg/mL) of HLP for 24 h. (**a**) Western blot analysis and (**b**) real-time quantitative RT-PCR of protein and mRNA levels of MMP-2 and MMP-9 in the treated cells. β-actin was served as an internal control of protein level. The quantitative data are presented as mean ± SD (*n* = 3) from three independent experiments. # *p* < 0.05, ## *p* < 0.01 compared with the control. \* *p* < 0.05, \*\* *p* < 0.01 compared with the TNF-α group. +, added; −, non-added.

## *3.3. HLP Inhibits TNF-*α*-Induced Akt*/*AP-1 Signaling*

MAPK and Akt have been shown to be involved in MMP-9 induction in various tumor types and migratory cell phenotypes [5,23]. To examine whether the activities of these protein kinases are downregulated by HLP, we analyzed their phosphorylation in A7r5 cells after being exposed to 10 ng/mL of TNF-α in the presence or absence of HLP at the indicated concentrations for 24 h. Immunoblot analysis with anti-phospho-specific antibodies was then performed. As shown in Figure 3a, the TNF-α-induced phosphorylated level of Akt was tremendously reduced by HLP in a concentration-dependent manner, whereas that of ERK was little affected. MMP-9 promoter was shown to have several transcription-factor-binding motifs, including binding sites for AP-1 and NF-κB [23], indicating that the AP-1 and NF-κB signal pathway may play a key role in the regulation of MMP-9 expression. Therefore, whether HLP could interfere the translocation of AP-1 or NF-κB into the nucleus in TNF-α-stimulated VSMC by immunoblotting analysis of the nucleus extracts prepared from the treated cells was then tested. The data in Figure 3b demonstrate that stimulation with 10 ng/mL of TNF-α for 24 h induced significantly the nuclear levels of c-Jun, c-Fos, and NF-κB (p65), compared to that of the control group. After exposure to TNF-α for 24 h, HLP treatments inhibited nuclear levels of c-Jun and c-Fos, components of transcription factor AP-1, in a dose-dependent manner, with the higher concentrations (0.10 mg/mL) being more effective. In contrast, there was no noticeable change in the translocation of nuclear NF-κB in the same condition of HLP treatments. Furthermore, to confirm that HLP could affect the DNA-binding activities of the translocated AP-1 and NF-κB in the TNF-α model VSMC, EMSA was carried out. The nuclear extracts of the above-treated cells were incubated with a DNA probe specific for AP-1, and the binding was analyzed by mobility shift (Figure 3c). A decrease in the DNA binding activity of AP-1 (left panel), but not NF-κB (right panel), was presented in the cells treated with TNF-α in the presence of HLP at various concentrations for 24 h.

**Figure 3.** Effect of HLP on TNF-α-induced Akt/AP-1 signaling in VSMCs. A7r5 cells were treated with TNF-α (10 ng/mL) in the absence or presence of various concentrations (0, 0.01, 0.05 and 0.10 mg/mL) of HLP for 24 h; (**a**) the cytoplasmic fraction was analyzed for the expressions of p-Akt, Akt (protein kinase PKB, also known as Akt), p-ERK, and ERK (extracellular signal-regulated kinase), and (**b**) the nuclear fraction was analyzed for the expressions of NF-κB, c-Jun, and c-Fos, two components of activator protein-1 (AP-1). These protein levels were determined by Western blotting. β-actin and C23 served as a cytoplasmic and nuclear internal control, respectively. (**c**) The nuclear extracts were analyzed for AP-1 (left panel) and NF-κB (right panel) DNA-binding activities using biotin-labeled AP-1 and NF-κB specific oligonucleotide by electrophoretic mobility shift assay (EMSA). Lane 1 represents nuclear extracts incubated with unlabeled oligonucleotide (free probe) to confirm the specificity of binding. Results are representative of at least three independent experiments. +, added; −, non-added.

### *3.4. HLP Inhibits TNF-*α*-Induced Abnormal VSMC Migration*

To evaluate whether HLP reversed A7r5 cells from the TNF-α stimulation, a set of well-established and classical methods, wound-healing and Boyden chamber assays, was used to determine VSMC migration. The effect of HLP on abnormal VSMC migration was analyzed by wound-healing assay, in which A7r5 cells were induced to migrate by physical wounding of cells plated on fibronectin-precoated 6-well plates. Under light microscopy, an apparent and gradual increase of cells in the denude zone was observed at the cells exposed to TNF-α more than control for 24 and 48 h (Figure 4a). A7r5 cells treated with TNF-<sup>α</sup>, together with the indicated doses of HLP, showed a reduced capacity to heal the wounded area, compared to the TNF-α alone. The quantitative results demonstrate that HLP could dose- and time-dependently inhibit TNF-α-stimulated VSMC migration. Subsequently, the effect of HLP on VSMC invasion was examined by a Boyden chamber coated with Matrigel under light microscopy. After a 24-h incubation period, TNF-α promoted a marked increase in the amount of cell invasion. The results further show that the number of cells invaded to the lower chamber was significantly reduced by HLP treatments. The data in Figure 4b indicate that such decrease was dose-dependent, with a 70% decrease when the TNF-α model cells were treated with HLP at 0.01 mg/mL. Therefore, it is possible that the anti-VSMC migratory/invasive effect of HLP was conducted by inactivating Akt/AP-1, subsequently leading to a reduction in MMP-9 expression and activation in TNF-α stimulation.

**Figure 4.** Effect of HLP on TNF-α-induced A7r5 cell motility and invasion. (**a**) Monolayers of A7r5 cells treated with TNF-α (10 ng/mL) in the absence or presence of various concentrations (0, 0.01, 0.05 and 0.10 mg/mL) of HLP were scraped and the number of cells in the denuded zone was photographed and quantified after indicated times (0, 24, and 48 h). Quantitative assessment of the mean number of cells in the denuded zone was presented as mean ± SD (*n* = 3) from three independent experiments. (**b**) A7r5 cells were treated with TNF-α in the absence or presence of various concentrations of HLP for 24 h. Invasion assay was performed using Boyden chamber. Representative photomicrographs of the membrane-associated cells were assayed by Giemsa stain. The purple parts indicate the membrane-associated cells. "% of control" denotes the mean number of cells in the membrane expressed as a proportion of that control group. Images were taken at 200× magnification; scale bar, 30 μm. The quantitative data are presented as mean ± SD (*n* = 3) from three independent experiments. # *p* < 0.05, ## *p* < 0.01 compared with the control. \* *p* < 0.05, \*\* *p* < 0.01 compared with the TNF-α group. +, added; −, non-added.

#### *3.5. HLP Inhibits TNF-*α*-Induced Abnormal VSMC Proliferation*

In the following experiment, the cytotoxic effect of HLP at dosages above 0.10 mg/mL and TNF-α (10 ng/mL) was also detected using cell growth curve analysis. As shown in Figure 5a, the TNF-α-induced proliferation of A7r5 cells under the uses of TNF-α and HLP at 0.2 and 0.5 mg/mL was significantly lower than that under TNF-α alone. Importantly, the cell growth curve confirmed the anti-proliferative effect was more pronounced when HLP at the doses of > 0.10 mg/mL were used in the TNF-α-exposed cells. We then investigated whether the HLP effect against TNF-α was attributed by induction of cell death or/and inhibition DNA synthesis. For this purpose, the level of DNA synthesis through BrdU incorporation in the treated cells grown under low-serum conditions was measured. As shown in Figure 5b, TNF-α caused an increase in BrdU incorporation, and TNF-α together with higher doses of HLP had a marked decreased level in BrdU incorporation.

To further hypothesize that HLP may be involved in the VSMC cell death, flow cytometry was used to examine whether the number of hypodiploid cells (apoptotic cells), which are stained less intensely with PI dye, can be unequivocally detected in the subG1 phase (left panel, Figure 5c). When A7r5 cells were treated with TNF-α at 10 ng/mL in the presence of HLP at 0.2 mg/mL for 24 h, it was not observed that an apparent accumulation of cells in the subG1 phase. Here, the cell cycle distribution of

TNF-α-treated VSMC affected by HLP was also evaluated. The 24-h TNF-α-stimulated cells showed a marked increase in S phase with fewer cells in G0/G1 phase, after TNF-α alone compared with control. When compared with the TNF-α alone group, the combination group had fewer cells in S phase and more cells in G0/G1 phase, indicating the 24-h HLP treatments could significantly lead to cell cycle block at G0/G1 phase in a dose-dependent manner (right panel, Figure 5c). In addition, when the cells were exposed to 0.5 mg/mL of HLP, a concomitant time-dependent slight and significant increase in apoptotic rates, compared to the TNF-α-treated group, was observed. Since the combination of HLP (0.2 mg/mL) and TNF-α (10 ng/mL) has the best antagonistic action of cell cycle regulation, the doses of combination were selected for further mechanistic studies of anti-VSMC proliferation, especially in G0/G1 arrest.

**Figure 5.** Effect of HLP on TNF-α-treated A7r5 cell growth curve, DNA synthesis, and cell cycle progression. A7r5 cells were treated with TNF-α (10 ng/mL) in the absence or presence of various concentrations (0, 0.2 and 0.5 mg/mL) of HLP for indicated time (0, 24 and 48 h). (**a**) The cell growth curve was evaluated using the Corning Cell Counter. (**b**) DNA synthesis was assayed by BrdU assay. (**c**) Cell cycle distribution was detected by flow cytometery. Quantitative assessment of the percentage of the cells in the cell cycle distribution (subG1, G0/G1, S, and G2/M phase) was indicated by PI dye. The proportion of cells in G0/G1 phase was quantitatively presented as mean ± SD (*n* = 3) of three independent experiments ± SD. # *p* < 0.05, ## *p* < 0.01 compared with the control. \* *p* < 0.05, \*\* *p* < 0.01 compared with the TNF-α group. +, added; −, non-added.

#### *3.6. HLP Induces Cell Cycle Arrest in the Present of TNF-*α

To investigate further the mechanism of the effect HLP on cell cycle arrest at G0/G1 phase, A7r5 cells treated with TNF-α (10 ng/mL) in the presence or absence of HLP (0.2 mg/mL) for 24 and 48 h were subjected to immunoblot analysis. We first analyzed the expressions of phospho-p53 (p-p53), p53, and cki, including p16, p21, and p27. Among them, p-p53, p21, and p27 levels were significantly induced by a 24-h HLP treatment (Figure 6a). To further investigate whether the inhibitory effect of HLP on TNF-α occurred because it blocked A7r5 cell cycle progression, the changes in protein levels of PCNA, E2F, and p-Rb, regulators of cell cycle G0/G1 arrest, were also studied (Figure 6b). Stimulation with TNF-α at 10 ng/mL for not only 24 h, but also 48 h, promoted time-dependently the expressions of PCNA, E2F, and p-Rb, compared to the receptive control group. After exposure to TNF-α for 48 h, the HLP treatment significantly inhibited three expressions (Figure 6b).

**Figure 6.** Effect of HLP on TNF-α-regulated the expressions of cell cycle regulatory proteins in VSMCs. A7r5 cells were treated with TNF-α (10 ng/mL) in the absence or presence of 0.2 mg/mL of HLP for 24 and 48 h. The protein levels of cdi, including p-p53, p53, p21, p27, and p16 (**a**), anti-proliferating cell nuclear antigen (PCNA), E2F, and p-Rb (**b**) were determined by Western blotting. β-actin was served as an internal control. (**c**) The expressions of cyclin E/cdk2 and Rb/E2F complexes were further analyzed. The cell extracts were immunoprecipitated (IP) with cdk2 or E2F. The precipitated complexes were examined for immunoblotting (IB) using anti-cyclin E or Rb antibody. Results are representative of at least three independent experiments. +, added; −, non-added.

Using immunoprecipitation, we confirmed that the addition of TNF-α upregulated the formation of cyclin E/cdk2 complex without noticeable change in cyclin D/cdk4 complex (data not shown) in A7r5 cells at 24 to 48 h, but HLP reversed the increases (line 1, Figure 6c). Moreover, there was a more significant increase in expression of Rb/E2F complex in the TNF-α combined with HLP treatments group (Figure 6c). As shown in Figure 6c (line 3), an increase in Rb/E2F complex was correlated with a decrease in p-Rb at 48 h of cell cycle (line 3, Figure 6b). The HLP-increased expression of Rb/E2F complex prevented the release of E2F transcription factor, and then reduced the transcription of the genes required for the cell cycle progression (Figure 6b,c). These data show that HLP regulated the association of cyclin E/cdk2, and Rb/E2F, inducing the cell cycle arrest at G0/G1 phase of A7r5 cells in the presence of TNF-<sup>α</sup>.

#### *3.7. HLP Reduced Atherosclerotic Lesions, and the Abnormal Migration and Proliferation of VSMC in a Rabbit Model*

Oxidant stress is a major cause of VSMC dysfunction and inflammation through various pathways [6,7]. To investigate the antioxidant action of HLP resulting from VSMC dysfunction, the ROS generation (DCF fluorescence) following the HLP treatments in the TNF-α-stimulated cells was examined (left panel, Figure 7a). The results showed that TNF-α significantly increased the fluorescence of intracellular ROS generation at not only 24 h, but also 48 h, whereas HLP at 0.2 mg/mL inhibited production of intracellular ROS (right panel, Figure 7a), implicating its antioxidant effects. In the same condition, the inhibitory effect of HLP on the amount of hydrogen peroxide (H2O2), the major form of ROS, was similar to the result of ROS production upon TNF-α stimulation (Figure S2). Collectively, these results sugges<sup>t</sup> that cyclin E/cdk2-dependent Rb phosphorylation and Akt/AP-1/MMP-9 signaling pathway mediated the in vitro action of HLP against to TNF-α-induced ROS production, controlling the balance of VSMC proliferation and migration (Figure 7b).

**Figure 7.** Effect of HLP on TNF-α-induced ROS production in VSMC. (**a**) A7r5 cells were treated with TNF-α (10 ng/mL) in the absence or presence of 0.2 mg/mL of HLP for 24 and 48 h. The treated cells were then labeled with fluorescent probe, dichlorofluorescin diacetate (DCFH-DA), and reactive oxygen species (ROS) production was measured using Muse™ Cell Analyzer. M1: DCF-negative cells. M2: DCF-positive cells. The results are presented as mean ± SD (*n* = 3) from three independent experiments. ## *p* < 0.01 compared with the control. \*\* *p* < 0.01 compared with the TNF-α group. +, added; −, non-added. (**b**) Schematic representation of TNF-α-antagonist potential of HLP on VSMCs. TNF-α induces intracellular ROS production, leading cell migration and proliferation through Akt/AP-1/MMP-9 signaling and cyclin E/cdk2-mediated Rb phosphorylation in A7r5 cells. HLP functions against TNF-α via downregulation of Akt/MMP-9 and upregulation of p53 signals that subsequently inhibit VSMC migration and proliferation. Red arrows represent the changes in response to TNF-α stimulation; blue arrows represent changes in TNF-α-exposed VSMCs receiving HLP intervention.

#### *3.8. HLP Reduced Atherosclerotic Lesions and the Abnormal Migration and Proliferation of VSMC in a Rabbit Model*

Because abnormal VSMC migration and proliferation contribute significantly in the pathogenesis of cardiovascular diseases, improvements in VSMC dysfunction will prevent the development of atherosclerosis [24]. For the clinical use of HLP for atherosclerosis, we investigated the HLP effect against VSMC dysfunction, using an atherosclerotic rabbit model. As shown in Figure 8a,b, HLP can significantly reduce the elevation of the concentrations of serum triglycerides (TG), total cholesterol (CHO), and LDL cholesterol (LDL-c) (Figure 8a), and the ratio of LDL-c and high-density lipoprotein cholesterol (HDL-c) (Figure 8b) enhanced by a HFD treatment. Past reports have shown that the decrease of LDL-c/HDL-c ratio, not just the LDL-c level alone, is of a lot of importance for reducing the atheroma burden [24,25]. In addition, the serum level of TNF-α was also significantly reduced after HFD-fed rabbits were treated with HLP (Figure 8c), confirming that HLP has a TNF-α antagonistic effect. Our study showed, in addition to possessing benefits to serum lipids, HLP can effectively decrease serum LDL/HDL ratio and TNF-α level, thus improving atherosclerosis.

To evaluate the in vivo atheroprotective effect of HLP against the extent of atherosclerosis in the aorta, the area of fatty region in the atherosclerotic lesions was analyzed using oil Red-O staining. The data in Figure 8d reveal that the subintimal lipid deposition in the HFD-treated rabbits was improved after HLP treatment. In addition, IHC staining indicated the expressions of α-SMA (upper panel) and PCNA (lower panel), served receptively as markers of VSMC migration and proliferation, were showed in the intima of atherosclerotic lesions from aortic roots of the rabbits treated with HFD (Figure 8e). As shown in Figure 8e, VSMC dysfunction was significantly observed in the atherosclerotic lesions in the HFD-treated rabbits, but very few expressions of α-SMA and PCNA in the HFD plus HLP-fed rabbits, which was consistent with the HLP-reduced the cell migration and proliferation in TNF-α-treated A7r5 cells in vitro (Figures 4 and 5). In the treatment process, HLP administration did not have any adverse effects on body weight or liver and renal function, compared to those of control (data not shown). Additionally, Western blotting of tissue extracts in the aortic arch showed the expressions of active-MMP-9, p-Akt, and E2F were markedly decreased, but the phosphorylation of p53 was increased in the group of HFD plus HLP, when compared with HLP or HFD-fed groups (Figure 8f). These results indicate that HLP can significantly improve VSMC dysfunction of HFD-treated rabbits by inhibiting cell migratory and proliferative signal pathways in vivo, as well as in vitro (Figure 8g).
