Chlorogenic acid (≥95% titration) was obtained from Sigma–Aldrich China (Shanghai, China). Human recombinant TGF-β1was purchased from Peprotech (Rocky Hill, NJ, United States). A rat TGF-β1 ELISA kit was purchased from Elabscience Biotechnology Co., Ltd. (Wuhan, China). Foetal bovine serum (FBS) and RPMI 1640 basic were purchased from Gibco (Grand Island, NY, United States). Cell counting kit-8 (CCK-8) was purchased from Dojindo Company (Japan). CCl4 was purchased from Sigma (St. Louis, MO, United States). Rabbit anti-rat Smad2, p-Smad2, Smad3, p-Smad3, Smad2/3, p-Smad2/3 and TGF-β1 antibodies were purchased from Cell Signaling Technology (CST, Boston, MA, United States). Smad7, CTGF, α-SMA, MMP-9, TIMP-1, collagen I, glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and horseradish peroxidase (HRP)-labeled secondary antibody were acquired from Wuhan Boster Biotechnology Co., Ltd. (Wuhan, China). The Trizol reagent, PeproTech RNAiso Plus and realtime polymerase chain reaction (PCR) kit were purchased from TaKaRa (Dalian, China). The miR-21 and negative control lentiviral vectors were synthesized by Gene Chem Co., Ltd. (Shanghai, China).

The LX2 cells were cultured in RPMI 1640 medium with 10% foetal bovine serum (FBS) at 37°C in 5% CO2 and 95% humidified air. Cell morphological changes after treatments were observed by a regular phase contrast microscope. Real-time PCR and western blot analysis were employed to evaluate the effect of CGA on LX2 cells. The cells were passaged in 6-well plates overnight and pretreated with CGA at various concentrations (20 μg/ml, 40 μg/ml and 80 μg/ml) for 24 h. For the last 6 h, TGF-β1 (10 ng/ml) was added to the wells for modeling but not to those for the normal group. After 24 h, the supernatants and cells were collected.

The procedures were carried out according to our past study (Li et al., 2016). The cell counting kit-8 (CCK-8) assay was employed to examine cell cytotoxicity. LX2 cells were seeded in 96-well plates at a density of 5.0 × 10⁴ cells/ml. CGA at different concentrations (20 μg/ml, 40 μg/ml and 80 μg/ml) was added. After 24 h, CCK-8 with 10 μl solution was added and the cells were incubated at 37°C for 2 h before meaturing the absorbance at 450 nm on a microplate reader. The experiments were repeated in duplicate.

To over-express the miR-21 gene, GV273-miR-21/NC-EGFP was transfected into the 293 T cell line. Viral supernatant were collected from the transfected 293 T cells after 48 h (3 × 10⁸ TU/ml) and were used to infect the LX2 cells. To knockdown miR-21 expression, the specific sequence of hsa-miR-21-3p-inhibition (16129-1) was 5′-TCGAGAAAAAACAGCCCATCGACTGGTGTTGTTTTTC-3′, and the reference sequence was 5′-TTCTCCGAACGTGTCACGT-3′. GV369-miR-21/NC-EGFP was transfected into the 293T cell line, and the viral supernatant was collected after 48 h (5 × 10⁸ TU/ml). LX2 cells were transfected with lentivirus at a multiplicity of infection (MOI) of 10 plaque-forming units (PFU)/cell according to the manual. The medium was replaced 12 h later, and then the cells continued to be incubated for 72 h. overexpression and knockdown of miR-21 was confirmed by Real-time PCR analysis.

Fifty male Sprague-Dawley rats (180 to 220 g) were obtained from the Hubei Provincial Centers for Disease Control and Prevention (Wuhan, China). All rats were housed in a room with controlled temperature (22–25°C), a 12:12 h light-dark cycle and free access to water and food. All study protocols abided by internationally accepted principles and the Guidelines for the Care and Use of Laboratory Animals of Huazhong University of Science and Technology and was approved by the Ethics Committee of Union Hospital, Tongji Medical College, Huazhong University of Science and Technology. The rats were randomly divided into 5 groups (n = 10): CGA low dosage (15 mg/kg), CGA middle dosage (30 mg/kg), CGA high dosage (60 mg/kg), and the experiment and normal groups. All groups except the normal group were given a CCl4 intra-peritoneal injection (i.p.) to induce liver fibrosis (CCl4: olive oil, 2:3 (vol/vol) per kg body weight, 4 ml/kg for the first dose and then 2 ml/kg twice per week) for 8 weeks. After 4 weeks, the rats in the low, middle, and high-dose CGA groups were administered CGA at 15 mg/kg, 30 mg/kg, or 60 mg/kg (at a concentration of 1.5, 3, or 6 mg/ml, the feeding volume was 1 ml/100 g) for four consecutive weeks, and the rats in the experiment and normal groups were administered normal saline at the same volume of 1 ml/100 g.

The rats were sacrificed to collect specimens. The procedure adhered to our past experiments (Huang et al., 2013; Jin et al., 2015). The liver tissues were collected and fixed in 4% paraformaldehyde solution, embedded in paraffin for histological examinations and stained with haematoxylin-eosin (HE) and Masson’s trichrome to assess liver damage and fibrosis development. Histological changes were observed at magnifications of ×100 and ×200. Five non-consecutive and random histological fields were analyzed to obtain the mean value.

The procedures were referred to in our previous study (Du et al., 2016). Expression of TGF-β1 in serum was determined by sandwich ELISA. The serum was collected and assayed for TGF-β1 by a rat TGF-β1 ELISA kit. The procedure conformed to the directions in the instruction manual from the kit.

The procedures were carried out according to our previous study (Zhou et al., 2013; Li et al., 2017). Total RNA was extracted from LX2 cells and liver tissue using TRIzol reagent (Invitrogen) and the cDNA was produced by using a PrimeScript^TM RT reagent kit according to the manufacturer’s protocol. Real-time qPCR was carried out in a 10 μl reaction containing 0.2 μl of cDNA and incubated at 37°C for 15 min and 85°C for 5 s. Real-time PCR reactions were performed at 95°C for 10 s, followed by 40 cycles of 95°C for 5 s and 60°C for 20 s according to the instructions of the SYBR Premix Ex Taq kit. The data were analyzed by the 2^-ΔΔC_T method. All primers were synthesized by GenScript (Nanjing, China), and the sequences are shown as Table 1.

The procedures followed our past experimental steps (Guo et al., 2015b; Ding et al., 2016). For liver tissue lysate preparation, liver tissue was homogenized with a hand-held homogenizer in lysis buffer. For cell lysate preparation, the LX2 monolayer cells were rinsed with PBS and lysed in RIPA buffer with a cocktail of protease inhibitors on ice. Total protein was extracted from the liver tissue and LX2 cells. The protein concentration was determined with a BCA protein assay kit. In each protein sample, an equivalent volume of 2× sodium dodecyl sulphate (SDS) loading buffer (100 mM Tris-HCl, pH 6.8; 4% SDS; 20% glycerine; 10% β-mercaptoethanol; and 0.2% bromophenol blue) was added and mixed again. The mixtures were then denatured at 95°C for 10 min, and the protein was loaded into each well and separated with 10% SDS-PAGE gel. After separation for approximately 80 min, the proteins were transferred to PVDF membranes, and the PVDF membranes were saturated with 5% non-fat milk containing in PBS for 1 h at room temperature and then probed with specific antibodies overnight at 4°C. The membrane was washed after incubation and then incubated for 1 h with the HRP-labeled secondary antibody. After further washing the membrane with TBST three times, bands were identified with enhanced chemiluminescence (ECL) reagent, and the signals were detected by exposing the membranes to X-ray films. The dilutions of the primary and secondary antibody were as follows: Smad2, 1:1000; Smad3, 1:1000; Smad2/3, 1:1000; p-Smad2, 1:1000; p-Smad3, 1:1000; p-Smad2/3, 1:1000; TGF-β1, 1:500; α-SMA, 1:1000; Smad7, 1:1000; CTGF, 1:800; MMP-9, 1:1000; TIMP-1, 1:1000; GAPDH, 1:5000; and secondary antibody 1: 5000. The Fuji ultrasonic-Doppler velocity profile (UVP) system and the ImageJ program were used for the densitometry analysis of the immunoreactive bands.

The procedure adhered to our previous study (Yang et al., 2016). The liver tissue (10 μm) were deparaffinized and hydrated, then the soaked slides were inactivated in 5% H2O2/methanol to block endogenous peroxidase activity. Next, the slides were incubated with normal goat serum for 10 min and incubated with α-SMA antibody (dilution, 1:500) or collagen I antibody (dilution, 1:200) at 4°C overnight. The slides were washed the next day and incubated with biotinylated secondary antibody at 37°C for 60 min. Then, the slides were washed again and incubated with horseradish peroxidase-labeled streptavidin at 37°C. The samples were developed with diaminobenzidine (DAB) and stained with haematoxylin. After the slides were washed with distilled water and dehydrated, they were made transparent and mounted under a microscope for examination. Image-Pro Plus software was employed to evaluate the mean optical density value of the images after immunohistochemical analysis.

The results are reported as the mean ± standard deviation (SD). Comparisons of the data between groups were expressed with Student’s t-test and one-way ANOVA followed by Tukey’s test. Statistical significance was considered significant when the P-value < 0.05 (Jin et al., 2013; Ding et al., 2015). All statistical analyses were performed with SPSS 12.0. Graphs were drawed with GraphPad Prism software (version 6).

CCK8 assay showed that pretreatment on unstimulated LX2 cells with a series of concentrations of CGA at 20 μg/ml, 40 μg/ml and 80 μg/ml for 24 h did not significantly affect cell viability (Figure 1A). The same result was shown for the cell morphology observation (Figure 1B). In addition, when assayed with real-time PCR, the difference of changes of miR-21 between CGA-treated cells and the normal group was not significant (Figure 1C). Therefore, we chose CGA at 20 μg/ml, 40 μg/ml, and 80 μg/ml to treat LX2 cells for 24 h.

There was a dose-dependent increase and time-dependent change in the protein expression of CTGF and in the mRNA expression of miR-21 and CTGF in LX2 cells treated with TGF-β1 (Figure 2). The protein expression of CTGF was measured after treatment with TGF-β1 at a range of 0–10 ng/ml for 6 h (Figure 2A) and treated with TGF-β1 at 10 ng/ml for 0–24 h in LX2 cells (Figure 2B). As shown in Figure 2C, the protein level of CTGF was significantly increase after treatment with TGF-β1 at 10 ng/ml compared with the 0 ng/ml group. In Figure 2D, when treated with 10 ng/ml TGF-β1 for 0–24 h. The protein expression of CTGF was highest at time point of 6 h compared with the 0 h group. The same result was shown for the mRNA expression of miR-21 and CTGF. As shown in Figures 2E,F, TGF-β1 treatment at 10 ng/ml for 6 h significantly increased the mRNA expression of miR-21 and CTGF (P < 0.01), so we chose TGF-β1 treatment at 10 ng/ml for 6 h for subsequent experiments.

As shown in Figures 3A–C, compared with the normal group, the mRNA levels of miR-21, CTGF, TIMP-1 and α-SMA in the experimental group were significantly increased (P < 0.01), and the mRNA levels of Smad7 and MMP-9 were notably decreased (P < 0.01). After treatment with CGA at a series of concentrations for 24 h, the mRNA levels of miR-21, CTGF, α-SMA and TIMP-1 were decreased, and the mRNA levels of Smad7 and MMP-9 were increased compared with the experimental group (P < 0.05 or 0.01). As shown in Figure 3D, western blot analysis was used to assess the protein levels. Compared with the normal group, the protein expression of CTGF, α-SMA, TIMP-1, p-Smad2, p-Smad3 and p-Smad2/3 were significantly elevated and the protein levels of Smad7 and MMP-9 were decreased in TGF-β1-stimulated LX2 cells. However, the protein expression of CTGF, α-SMA, TIMP-1, p-Smad2, p-Smad3 and p-Smad2/3 was effectively inhibited after treatment with CGA, and CGA could improve the protein expression of Smad7 and MMP-9 compared with the experimental group. The protein expression of Smad2, Smad3 and Smad2/3 had no significant changes.

We investigated the expression of downstream signaling molecules after transfecting LX2 cells in vitro with the miR-21 lentiviral vector GV369. The green fluorescent protein (GFP) was observed with a fluorescence microscope after transfection for 48 and 72 h (Figure 4A). Meanwhile, to confirm the transduction efficiency, the mRNA expression of miR-21, Smad7 and CTGF, and the protein levels of p-Smad2, p-Smad3, p-Smad2/3, CTGF, Smad7, α-SMA, TIMP-1 and MMP-9 were detected at 72h after transfection (Figures 4B,C). We observed that there was no difference in the mRNA levels of miR-21, Smad7 and CTGF (P > 0.05, Figure 4B) and the protein levels of p-Smad2, p-Smad3, p-Smad2/3, CTGF, Smad7, α-SMA, TIMP-1 and MMP-9 between the normal and lentivirus-NC group (lentivirus negative control group) (P > 0.05, Figure 4C). However, compared with the normal group, after miR-21 overexpression, the mRNA expression of miR-21 and CTGF increased, and the mRNA expression of Smad7 decreased (P < 0.01, Figure 4B). The protein expression of p-Smad2, p-Smad3, p-Smad2/3, CTGF, α-SMA and TIMP-1 increased, and the protein levels of Smad7 and MMP-9 decreased (P < 0.01, Figure 4C).

The miR-21 in LX2 cells was overexpressed with the lentiviral vector for 72 h, and then, the cells were treated with a series of concentrations of CGA (20 μg/ml, 40 μg/ml, and 80 μg/ml) for 24 h. TGF-β1 was added to the LX2 cells for the last 6 h before harvest. As shown in Figures 5A–C, compared with the lentivirus-up group, the mRNA levels of miR-21, CTGF, α-SMA and TIMP-1 in the lentivirus-up/TGF-β1 group were increased (P < 0.01 or 0.05), and the mRNA levels of Smad7 and MMP-9 were decreased (P < 0.01 or 0.05). The protein expression of p-Smad2, p-Smad3, p-Smad2/3, CTGF, α-SMA and TIMP-1 were also increased, and the protein levels of Smad7 and MMP-9 were decreased (Figure 5D). After treatment with CGA at a series of concentrations, compared to that in lentivirus-up/TGF-β1 group, the mRNA levels of miR-21, CTGF, α-SMA and TIMP-1 were decreased (Figures 5A–C, P < 0.05 or 0.01), and the mRNA levels of Smad7 and MMP-9 were increased relatively (Figures 5A–C, P < 0.01 or 0.05). The protein levels of CTGF, α-SMA, TIMP-1, p-Smad2, p-Smad3, and p-Smad2/3 were decreased, and the protein levels of Smad7 and MMP-9 were increased (Figure 5D).

We investigated the expression of downstream signaling molecules after transfecting LX2 cells in vitro with the miR-21 lentiviral vector GV273. The green fluorescent protein (GFP) was observed with a fluorescence microscope after transfection for 48 and 72 h (Figure 6A). Meanwhile, to confirm the transduction efficiency, the mRNA expression of miR-21, Smad7 and CTGF, and the protein levels of p-Smad2, p-Smad3, p-Smad2/3, CTGF, Smad7, α-SMA, TIMP-1 and MMP-9 were detected at 72h after transfection (Figure 6B,C). We observed that there was no difference between the normal and lentivirus-NC group (lentivirus negative control group) in the mRNA levels of miR-21, Smad7 and CTGF (Figure 6B, P > 0.05) and the protein levels of p-Smad2, p-Smad3, p-Smad2/3, CTGF, Smad7, α-SMA, TIMP-1 and MMP-9 (Figure 6C). However, compared with the normal group, after miR-21 knockdown, the mRNA expression of miR-21 and CTGF decreased, and the mRNA expression of Smad7 increased (P < 0.01, Figure 6B). The protein expression of p-Smad2, p-Smad3, p-Smad2/3, CTGF, α-SMA and TIMP-1 decreased, and the protein levels of Smad7 and MMP-9 increased (Figure 6C, P < 0.01).

The miR-21 in LX2 cells was knocked-down with a lentiviral vector for 72 h, and then, the cells were treated with a series of concentrations of CGA (20 μg/ml, 40 μg/ml, and 80 μg/ml) for 24 h. TGF-β1 was added to the LX2 cells for the last 6 h before harvest. As shown in Figures 7A–C, compared with the lentivirus-down group, the mRNA levels of miR-21, CTGF, α-SMA and TIMP-1 in the lentivirus-down/TGF-β1 group were increased (P < 0.01), and the mRNA levels of Smad7 and MMP-9 were decreased (P < 0.01). The protein expression levels of p-Smad2, p-Smad3, p-Smad2/3, CTGF, α-SMA and TIMP-1 were also decreased, and the protein expression levels of Smad7 and MMP-9 were increased (Figure 7D). After treatment with CGA at a series of concentrations, the mRNA levels of miR-21, CTGF, α-SMA and TIMP-1 were decreased (Figures 7A,B, P < 0.05 or 0.01), and the mRNA levels of Smad7 and MMP-9 increased relatively (Figure 7C, P < 0.05 or 0.01) compared to that in the lentivirus-down/TGF-β1 group. The protein levels of CTGF, α-SMA, TIMP-1, p-Smad2, p-Smad3, and p-Smad2/3 were decreased, and the protein levels of Smad7 and MMP-9 were increased (Figure 7D).

As shown in Figure 8, the expression level of TGF-β1 in the serum in the experimental group was markedly increased when compared with that in the normal group (P < 0.01). The level of TGF-β1 was decreased significantly compared with that in the experimental group when the rats were treated with CGA at different concentrations (P < 0.05 or 0.01). The inhibitory effect was enhanced when the concentration of CGA increased (P < 0.01), which suggested that CGA could inhibit the expression of TGF-β1.

Haematoxylin-eosin and Masson’s trichrome staining were used to evaluate the anti-fibrosis role of CGA. As shown in Figure 9A, slices from the liver of the normal group showed liver structural integrity without inflammatory cell infiltration. However, CCl4-induced liver injury in the experimental group showed fibrosis and inflammatory cell infiltration and the loss of structural integrity. Nevertheless, in the CGA group, the slices showed less fibrosis, less inflammatory cell infiltration and less liver cell necrosis compared with the experimental group. As shown in Figure 9B, collagen is blue in fibrotic tissue when stained by Masson’s trichrome, slices from the liver of the normal group showed no fibrosis. However, excess blue collagen fibers were observed in the experiment group. After treatment with CGA at different concentrations, the area of collagen fibers decreased obviously and high concentration of CGA showed stronger effect of anti-liver fibrosis.

As shown in Figures 9C,D, the staining of α-SMA and collagen I in the normal group was not remarkable (P < 0.01). After induction with CCl4, the positive staining of α-SMA and collagen I in the experimental group significantly increased, showing as dark brown. The staining in the CGA groups showed a smaller area and weaker staining, and fewer positive cells were observed compared with the experimental groups (P < 0.05 or 0.01). The inhibitory effect was strengthened when the concentration of CGA increased (P < 0.05 or 0.01).

As shown in Figures 10A–C, compared with the normal group, the mRNA levels of miR-21, CTGF, α-SMA, TIMP-1, and TGF-β1 in the experimental group were markedly increased (P < 0.01 or 0.05), and the mRNA levels of Smad7 and MMP-9 was markedly decreased (P < 0.01). However, treatment with CGA at different concentrations decreased the mRNA levels of miR-21, CTGF, α-SMA, TIMP-1, and TGF-β1 and elevated the mRNA levels of Smad7 and MMP-9 compared with that in the experimental group (P < 0.05 or 0.01). As shown in Figure 10D, western blot analysis was carried out to determine the protein expression. Compared with the normal group, the protein expression of p-Smad2, p-Smad3, p-Smad2/3, CTGF, TGF-β1, TIMP-1 and α-SMA were significantly increased, and the protein levels of Smad7 and MMP-9 were decreased in the experimental group. However, the protein expression of p-Smad2, p-Smad3, p-Smad2/3, CTGF, TGF-β1, TIMP-1 and α-SMA was significantly inhibited after CGA treatment, and CGA could elevate Smad7 and MMP-9 expression compared with the experimental group (P < 0.05 or 0.01). There were no changes in the protein levels of Smad2, Smad3 and Smad2/3.