Apigenin inhibits fibroblast proliferation and reduces epidural fibrosis by regulating Wnt3a/β-catenin signaling pathway

Apigenin was provided by Sigma (St. Louis, MO, USA) and the purity of apigenin is 98%.

Human fibroblasts were provided by Shanghai Cell Repository of Chinese Academy of Sciences. Fibroblasts were cultured in DMEM (Gibco, Grand Island, NY) containing 15% fetal bovine serum (Gibco) at 37 °C under 5% CO₂. API was dissolved in dimethyl sulfoxide to prepare various concentrations of API solution. After the fibroblasts were seeded into dishes until reaching the density of about 60–80%, API in different concentrations was used to treat fibroblasts for 24 h.

The Wnt3a overexpression (Wnt3a-OE) plasmid was purchased from Ruinan biotechnology center (Shanghai, China). For lentivirus production, 293T cells were plated into a 6-well plate at a confluence of 70–90%. p-Receiver-Lv185-Wnt3a, VSVG, Rre, and Rev plasmids were co-transfected into 239T cells by the Lipofectamine 3000 reagent (Invitrogen, USA) according to the manufacturer’s instructions. The supernatant containing Wnt3a-OE lentiviruses was collected and filtered at 48 h after transfection. For lentivirus infection, the lentiviral supernatant was mixed with 10 ng/ml polybrene (Genechem, China) and applied to the fibroblasts. The second round of infection was performed after the first infection for 24 h. After the second infection for 48 h, the infected fibroblasts were selected with 1 μg/ml puromycin (Sigma, USA). The efficiency of infection was evaluated by Western blotting analysis and the fibroblasts which overexpressing Wnt3a were used for subsequent experiments.

The cell viability was measured by Cell Counting Kit-8 (CCK-8) (Dojindo, Tokyo, Japan) following the instructions of the manufacturer. When the density of cells was about 60–80%, API in different concentrations (0, 10, 30, 60, 90, 120, 150 μM) was added into 96-well plates and maintained for 24 h. The fibroblasts were further incubated with 0.01 ml CCK-8 reagent at 37 °C. The absorbance values were measured with a microplate reader (Tecan) at 450 nm after 2 h, and the cell survival rate was calculated.

To detach the proliferation of fibroblasts, Cell-Light KFluor555 EdU kit (KeyGEN, Jiangsu, China) was used. Fibroblasts were co-cultured with API in various concentrations (0, 30, 60, 90 μM) on 6-well plates for 24 h at 37 °C. Fifty micromolars per liter of EdU operating fluid was added into each well for 2 h. These cells were then stained with DAPI for 10 min after being treated with osmotic enhancers and fixatives. The inverted fluorescence microscope was used to obtain images at × 200 magnification.

After the treatment with API in different concentrations, fibroblasts were collected and lysed using the RIPA buffer (Beyotime, Hangzhou, China) on ice. Ultrasonic wave was used to treat the cell lysate before collection of the supernatants. The concentration of protein in the supernatants was measured with BCA Protein Assay Kit (Thermo, USA). Equivalent protein was separated by SDS electrophoresis gel and then transferred onto polyvinylidene difluoride membranes. After being blocked in 5% Bull Serum Albumin (BSA) in TBST for 2 h, the membranes were incubated with the primary antibodies at 4 °C overnight. The membranes were then incubated with the secondary antibodies at 4 °C for 2 h after being washed 3 times in 1× TBST for 15 min.

This study was approved by the Animal Research Committee of Yangzhou University. Twenty-four healthy SD male rats weighted 220–250 g were randomly divided into four groups and six rats in each group. After 1% pentobarbital sodium (40 mg/kg) was intraperitoneally injected, laminectomy was performed and the L1 vertebral plate was removed according to a previous study [4]. API was dissolved in 5% sodium carboxyl methylcellulose (CMC-Na). The rats were gavaged with API in different doses (1 mg/kg, 2 mg/kg, and 4 mg/kg) or 5% CMC-Na solution (control group) once a day for 1 month. After 1 month, the rats were sacrificed by a fatal dose of pentobarbital sodium and underwent the intracardial perfusion using 4% paraformaldehyde. The entire L1 spinal specimen, including the epidural fibrotic tissue and paraspinal muscles, was collected and fixed in the buffered formalin. The spinal specimen was then embedded in paraffin after decalcification using ethylenediaminetetraacetic acid (EDTA). The paraffin-embedded tissue blocks were used for the subsequent experiments.

Transverse sections of 4 μm were prepared using paraffin blocks from each group. After the use of a series of dewaxing methods, the sections were stained with hematoxylin and eosin (H&E) and Masson’s trichrome. Six odd-numbered sections with H&E staining were used to assess epidural fibrosis and calculate fibroblast number using a light microscope at the magnification of 40 and 400, respectively. Six even-numbered sections with Masson’s trichrome staining were observed to evaluate the density of collagen fiber using a light microscope at a magnification of 400. The Image Pro Plus 6.0 image analysis software was applied to measure the value of collagen optical density.

After the process of deparaffinization and rehydration, the antigenicity of tissue sections was activated in sodium citrate solution for 15 min at 95 °C. The endogenous peroxidase activity was blocked by 3% H₂O₂. The sections were then incubated with diluent anti-PCNA in a wet slides box at 4 °C for 12 h, and followed by the incubation with biotinylated anti-rabbit IgG for 2 h after being washed with PBS for 3 times. The diaminobenzidine solution was used for staining and hematoxylin was applied for counterstaining these sections.

The data were statistically analyzed using the SPSS software (version 22.0). Student’s t test was used for the comparison between two groups, and one-way analysis of variance was used for the comparison between multiple groups. Data are presented as mean ± standard deviation. The value of P < 0.05 was regarded as statistically significant in all analyses.

To detect the effect of API on fibroblast proliferation, API in various concentrations was used to treat the cells for 24 h. The CCK-8 detection was used to measure cell viability. As shown in Fig. 1a, API induced cell viability reduction in a concentration-dependent manner. The EdU incorporation assay was used to examine the proliferation capacity of fibroblasts. The percentage of EdU-positive cells decreased after the cells were treated with API, and the 90 μM API-treated group reached the lowest level compared with those of the other groups (Fig. 1b, c). Moreover, the Western blotting analysis indicated that API inhibited the expressions of PCNA and cyclinD1 (Fig. 2a). The results revealed that API inhibited fibroblast proliferation in a concentration-dependent manner.

To verify the effect of API on regulating Wnt3a/β-catenin pathway, Western blotting was performed to detect the expressions of related proteins after fibroblasts were treated with API in various concentrations (0 μM, 30 μM, 60 μM, and 90 μM) for 24 h (Fig. 2a). The results demonstrated that API downregulated the Wnt3a expression in a concentration-dependent manner. Moreover, API also downregulated its downstream proteins such as p-GSK3β, β-catenin, and cyclinD1 as well as the value of p-GSK3β/GSK3β, which was in accord with the reduction of cell proliferation. These results indicated that API could inhibit Wnt3a/β-catenin pathway in a concentration-dependent manner (Fig. 2b, c).

To further explore the underlying mechanism of how API inhibits fibroblast proliferation, Wnt3a gene was overexpressed in fibroblasts and API was then used to treat overexpressed fibroblasts. The Western blotting analysis showed that the level of Wnt3a expression was increased in Wnt3a-OE fibroblast lines, and Wnt3a-OE promoted expressions of PCNA and cyclinD1, indicating that Wnt3a could stimulate fibroblast proliferation (Fig. 2d). Upregulated expressions of p-GSK3β and β-catenin and the value of p-GSK3β/GSK3β were also demonstrated in Wnt3a-OE fibroblast lines. Moreover, the inhibitory effect of API on PCNA, Wnt3a, and its downstream proteins was partially reversed by overexpression of Wnt3a (Fig. 2e, f). Therefore, these results indicated that API inhibited fibroblast proliferation by suppressing Wnt3a/β-catenin signaling pathway.

To clarify the effect of API on reducing the formation of epidural fibrosis, the laminectomy model was established and treated with API. The H&E images demonstrated that the fibrotic tissue in the surgical field was decreased after the treatment with API. Marked fibrotic tissues were observed in laminectomy sites in the control group, while loose fibrotic tissues were demonstrated in the API-treated groups (Fig. 3a). Fibroblasts were further counted from H&E images at × 400 magnification (Fig. 3b). The fibroblasts numbers in the API-treated groups were less than that in the control group, showing a dose-dependent manner. The fibroblast number in fibrotic tissue in the 4 mg/kg API group was significantly less than those in other API-treated groups and control group (P < 0.05) (Fig. 3c).

Moreover, Masson’s trichrome staining images at × 400 magnification were selected to evaluate the destiny of collagen (Fig. 4a). Dense collagen tissue was found in laminectomy sites in the control group, while weak collagen in API-treated groups. The results of collagen density in API-treated groups were coincidence with H&E staining, indicating that API decreased the destiny of collagen. The values of collagen optical density in the API-treated groups were significantly less than that in the control group, showing a dose-dependent manner. The values of collagen optical density in the 4 mg/kg API-treated group was significantly less than those of other API-treated groups and control group (P < 0.05) (Fig. 4b). The above results showed that API decreased fibroblast number and collagen density, which further reduced epidural fibrosis.

Proliferating cell nuclear antigen (PCNA) is a key factor in regulating cell proliferation. The immunohistochemical staining of fibrotic tissues from each treated group is shown in Fig. 5. The expressions of PCNA in the API-treated groups were lower than that in the control group. These results indicated that API suppressed fibroblast proliferation in epidural fibrotic tissue.