Fuzheng Huayu Recipe Ameliorates Liver Fibrosis by Restoring Balance between Epithelial-to-Mesenchymal Transition and Mesenchymal-to-Epithelial Transition in Hepatic Stellate Cells.

Activation of hepatic stellate cells (HSCs) depending on epithelial-to-mesenchymal transition (EMT) reflects the key event of liver fibrosis. Contrastively, mesenchymal-to-epithelial transition (MET) of HSCs facilitates the fibrosis resolution. Here we investigated the effect of Fuzheng Huayu (FZHY) recipe, a Chinese herbal decoction made of Radix Salviae Miltiorrhizae, Semen Persicae, Cordyceps sinensis, Pollen Pini, and Gynostemma pentaphyllum, on liver fibrosis concerning the balance of EMT and MET in HSCs. In contrast to the increased TGF-β 1/BMP-7 ratio in activated HSCs, FZHY administration induced significant upregulation of BMP-7 and downregulation of TGF-β 1 at both transcription and translation levels. Restoration of TGF-β 1/BMP-7 ratio inhibited the expression of p38 MAPK and phosphorylated p38 MAPK, resulting in the reversal of epithelial-to-mesenchymal transition (EMT) to mesenchymal-to-epithelial transition (MET) as characterized by the abolishment of EMT markers (α-SMA and desmin) and reoccurrence of MET marker (E-cadherin). In vivo treatment of FZHY recipe also demonstrated the statistical reduction of activated HSCs with EMT phenotype, which attenuated the carbon tetrachloride- (CCl4-) induced liver fibrosis in a dose-dependent manner. These findings may highlight a novel antifibrotic role of FZHY recipe on the basis of rebalancing EMT and MET in HSCs.

1. Introduction

Hepatic fibrosis represents the common outcome of chronic liver diseases, such as chronic viral infections, alcoholism, fatty liver diseases, and cholestasis, and associates with both hepatocellular carcinoma (HCC) and hepatic failure. Activation of hepatic stellate cells (HSCs) has been well established to contribute to most fibrosis-related characteristics [1].

Nowadays, Fuzheng Huayu (FZHY) recipe, a decoction of Radix Salviae Miltiorrhizae, Semen Persicae, Cordyceps sinensis, Pollen Pini, and Gynostemma pentaphyllum, is reported to exert antifibrosis actions in several clinical trials [2-5]. Three-month treatment of FZHY recipe gives rise to the decreased serum levels of procollagen peptides and laminin in 63 patients with chronic hepatitis B (CHB). Liver biopsy uncovers the remarkable amelioration of fibrosis grade in most of them [2]. Similar results are detected in another multicentral clinical trial for 180 patients with CHB. Significant improvements in serum level of collagen type IV and fibrosis scoring confirm the therapeutic role of FZHY recipe in hepatic fibrosis [3]. Although FZHY recipe exhibits the prospect for fibrosis prevention, mechanisms underlying its actions still remain to be clarified.

Recent studies uncover that TGF-β 1, the key mediators of liver fibrosis, induces typical characteristics of epithelial-to-mesenchymal transition (EMT), also known as activation, in HSCs [6, 7]. Accompanied by the loss of epithelial markers and gain of mesenchymal markers, HSCs undergoing EMT demonstrate collagen overproduction and formation of fibrous tissue, which further initiates and perpetuates the liver fibrosis [6]. In contrast, reverse process of EMT, namely, mesenchymal-to-epithelial transition (MET), has recently been proved to counteract TGF-β 1-induced fibrotic response in different organs [8, 9]. BMP-7, one of the members of TGF-β surperfamily, acts as the critical mediator of MET [10]. Therefore, balance between EMT and MET of HSCs is suggested to determine the liver fibrosis on a basis of, at least to a large extent, TGF-β 1/BMP-7 ratio. Surprisingly, there are growing evidences that FZHY recipe exerts intensive impact on the TGF-β 1/BMP-7 ratio, and also the downstream signaling. The FZHY recipe has already been identified to downregulate the hepatic expression of TGF-β 1 [11] and strongly inhibits the EMT-specific signal in renal interstitial cells [4]. Thus FZHY recipe is hypothesized to effect a pattern of EMT/MET rebalancing by modulating the TGF-β 1/BMP-7 ratio.

To verify the role of FZHY recipe in the balance between EMT and MET, primary rat HSCs were activated and incubated with FZHY-containing serum. The expression of both TGF-β 1 and BMP-7 in HSCs was evaluated according to their transcription and translation levels. Analysis of p38 MAPK, with and without phosphorylation, reflected the downstream signaling depending on TGF-β 1/BMP-7 ratio. Markers specific to epithelial and mesenchymal cells further characterized the status of EMT and MET. Finally, experimental liver fibrosis was subjected to FZHY treatment so as to reveal the consequence of EMT/MET rebalancing.

2. Materials and Methods

2.1. Cell Preparation

HSCs were separated from adult male Sprague-Dawley (SD) rats (400–500 g, Experimental Animal Center of Chinese Academy of Science, Shanghai, China) by in situ serial infusion with D-Hanks' and perfusion medium (Hanks' medium containing 0.05% collagenase IV and 0.1% Pronase E), redigestion with collagenase IV and DNase, and single-step density gradient centrifugation with 18% Nycodenz (W/V) [4]. Their living percentage was up to 95% as defined by staining of trypan blue, while their purity was over 90% when assessed by Oil Red O staining and real-time PCR for α-smooth muscle actin (α-SMA), desmin, albumin (ALB), CK19, CD31, and CD68, respectively [1] (Figures 1(a)–1(d), 1(g)). Immunocytochemical staining of α-SMA (Santa Cruz, USA) denoted the spontaneous activation of nearly all HSCs on the 10th day (Figures 1(e) and 1(f)).

Figure 1

Characterization of HSCs isolated from rat liver. (a, b) Cell morphology of quiescent (a) and activated HSCs (b) (400x). (c, d) Oil Red O staining of quiescent (c) and activated HSCs (d) (400x). (e, f) Immunocytochemistry staining of quiescent (e) and activated HSCs for α-SMA (f) (400x). (g) Real-time PCR for ALB, α-SMA, CD31, CD68, CK-19, and desmin, respectively, demonstrated the purity of quiescent and activated HSCs. P < 0.05 compared with quiescent HSCs.

The rats received humane care in accordance with the Declaration of Helsinki (1964). The study was approved by the ethical committee of Shanghai Jiaotong University School of Medicine.

2.2. FZHY Administration

Adult male SD rats were treated by FZHY recipe (Pearl Ocean Pharmaceutical Holdings Limited, Beijing, China) intragastrically (10-fold of the maximum daily human dosage of 65 kg, every 2 hours for 3 times). Blood was collected from inferior vena cava 2 hours after the last administration. FZHY-containing serum was then obtained by centrifugation (3000 g for 10 minutes) and inactivation (56°C for 30 min) [12].

Activated HSCs in logarithmic growth phase were randomized into groups of model, 5% FZHY treatment, 10% FZHY treatment, and 20% FZHY treatment. Quiescent HSCs were used as the control group. All FZHY-treated groups were incubated with FZHY-containing serum for 48 h, while saline at the same volume was added into the control and model groups.

2.3. Real-Time PCR

Total RNA, being extracted from HSCs, was subjected to RT reaction by ExScript RT reagent kit (TAKARA, Kusatsu, Japan). Real-time PCR was then performed using SYBR Premix Ex Taq (TAKARA, Kusatsu, Japan) on a Light Cycler (Roche Diagnostics GmbH, Penzberg, Germany) according to the manufacturer's instructions. Primer sequences and endogenous control for these reactions were exhibited in Table 1. Relative gene expression levels were calculated by the 2−ΔΔCT method [13].

Table 1

Primers for real-time PCR.

Gene	Primer sequence (5′-3′)	Annealing temperature (°C)	Product length (bp)
ALB	Forward primer: GACTGCCCTGTGTGGAAGACReverse primer: CGAAGTCACCCATCACCGTC	56	321
α-SMA	Forward primer: GCGGGGATCCATGAGACCACCTACReverse primer: GAGCCGCCGATCCACACTGA	60	218
CK19	Forward primer: CACTACGCAGATCCAGATAAACAReverse primer: GAAGTCGCACTGGTAGCAAG	56	388
CD31	Forward primer: CTTCACCATCCAGAAGGAAGAGACReverse primer: CACTGGTATTCCATGTCTCTGGTG	56	360
CD68	Forward primer: TCATGGGAATGCCACAGTTTCReverse primer: GAGGGCCAACAGTGGAGAA	56	62
Desmin	Forward primer: GGCCGACGCAGTGAACCAGReverse primer: CGCCGCAGCTCTCGCATCT	60	213
E-cadherin	Forward primer: TGAAGCCCCCATTTTTGTGCReverse primer: GCTCCGAATCTTCTCTGTCCATCT	58	221
TGF-β 1	Forward primer: GCGCCTGCAGAGATTCAAGTCAReverse primer: AAAGCCCTGTATTCCGTCTCCT	56	187
GAPDH	Forward primer: CACGGCAAGTTCAACGGCACAGTReverse primer: AGCGGAAGGGGCGGAGATGAT	60	222

2.4. Western Blot

Total protein of each sample was prepared by standard procedures and quantified by the bicinchoninic acid method (Pierce, Rockford, USA). These protein samples were loaded on 10% polyacrylamide gels and electrophoretically separated. Subsequently, the protein was transferred to polyvinilidene difluoride membranes (Millipore, USA), blocked with 0.1% Tween 20 (TBST) in 5% nonfat dry milk (NFDM) based on Tris-buffered saline (TBS), reacted with primary antibodies (anti-TGF β 1 (1 : 500), anti-BMP-7 (1 : 1000), anti-p38 MAPK (1 : 1500), anti-phosphorylated-p38 MAPK (anti-p-p38 MAPK, 1 : 1000), anti-E-cadherin (1 : 500), anti-α-SMA (1 : 1000), and anti-desmin (1 : 600), Santa Cruz, USA) overnight at 4°C and then HRP-conjugated goat anti-mouse IgG (1 : 3000; Jackson ImmunoResearch Laboratories, Inc., USA) for 2 h at room temperature. After washing, the membrane was processed using Super-Signal West Pico Chemiluminescent Substrate (Pierce), and antibody against GAPDH (Santa Cruz, USA) (1 : 1000) as an internal control [14].

2.5. Animals and Grouping

Adult male SD rats were randomly divided into normal control (n = 10), model (n = 10), and FZHY-treated group (n = 10). Except for those in the normal control group, all rats were subcutaneously injected with 40% CCl4 (0.3 mL/100 g) every 3 days for 6 weeks. Rats of the normal control group received same volume of olive oil in the same way [15]. During the exposure to CCl4, FZHY recipe dilution (75 mg/mL of FZHY recipe) was intragastrically administered to the low-dose (2.5 mL/kg/d), medium-dose (5 mL/kg/d), and high-dose (10 mL/kg/d) FZHY-treated rats for 4 weeks, while saline was administered to rats in the model group [16].

2.6. Immunohistochemistry

Liver specimens of all groups were prepared and subjected to the following procedure: (1) inhibition of endogenous peroxidase activity by 3% H2O2 for 15 minutes; (2) digestion by 0.01% trypsin for 10 minutes; (3) blocking nonspecific binder by bovine albumin for 10 minutes; (4) incubation with anti-α-SMA (1 : 100, Santa Cruz, USA) overnight at 4°C, (5) reaction with biotin-conjugated secondary antibody (1 : 200; Jackson ImmunoResearch Laboratories, Inc., USA) for 30 minutes; (6) visualization by DAB kit [17].

2.7. Hydroxyproline Assay

Liver samples were hydrolyzed in HCl (6 mol/L) at 120°C for 16 h. After being neutralized with NaOH (6 mol/L), the supernatant was oxidized with Chloramine T (Sigma-Aldrich Corp., USA) in acetate/citrate buffer and mixed with Ehrlich's solution. The mixture was then incubated at 60°C for 30 min, and its absorbance was determined at 560 nm [18].

2.8. Histology Assay

Histological characteristics, including hepatocyte steatosis, inflammation, and fibrosis, of liver tissue were investigated by both hematoxylin-eosin (H&E) and Van Gieson (VG) staining. Liver fibrosis was evaluated according to Metavir stage [19]. Obtained results were independently interpreted by 2 pathologists, who were not aware of the study.

2.9. Statistical Analysis

Data are expressed as means ± standard deviation (SD). All groups were compared statistically by one-way ANOVA using SPSS 16.0 statistical package. Differences with P < 0.05 were considered statistically significant.

3. Results

3.1. FZHY Recipe Recovered the TGF-β 1/BMP-7 Ratio in Activated HSCs

When compared to those of quiescent HSCs (control group), significant increase in the mRNA and protein levels of TGF-β 1 was detected in activated HSCs (model group) (P < 0.05, Figures 2(a)–2(c)). Meanwhile, the mRNA and protein levels of BMP-7 were dramatically downregulated in activated HSCs (control group versus model group, P < 0.05). In contrast, treatment of FZHY-containing serum dose-dependently reduced the mRNA level of TGF-β 1, whereas it restored the mRNA level of BMP-7 (Figure 2(a)). Western blot further confirmed that the changes in TGF-β 1 and BMP-7 transcription were paralleled by 3.29-fold decrease in TGF-β 1 protein expression (model group versus 20% FZHY-treated group, P < 0.05) and 7.71-fold increase in BMP-7 protein expression (model group versus 20% FZHY-treated group, P < 0.05) after the treatment of FZHY recipe (Figures 2(b) and 2(c)).

Figure 2

Normalization of TGF-β 1/BMP-7 ratio by FZHY recipe inhibits p38 MAPK signaling pathway in activated HSCs. (a) The mRNA expression levels of TGF-β 1, BMP-7, and p38 MAPK in the control, model, 5% FZHY-treated, 10% FZHY-treated, and 20% FZHY-treated group, respectively. (b) Western blot of TGFβ 1, BMP-7, p38 MAPK, and p-p38 MAPK in the control, model, 5% FZHY-treated, 10% FZHY-treated, and 20% FZHY-treated group, respectively. (c) Relative protein levels of TGFβ 1, BMP-7, p38 MAPK, and p-p38 MAPK. (d) The TGF-β 1/BMP-7 ratio at transcriptional and translational levels. P < 0.05 compared with control group. # P < 0.05 compared with model group.

According to the expression of TGF-β 1 and BMP-7, there was statistical upregulation of TGF-β 1/BMP-7 ratio in activated HSCs at both transcriptional and translational levels (control group versus model group, P < 0.05, Figure 2(d)). However, administration of FZHY recipe, no matter the 5% FZHY-, 10% FZHY-, and 20% FZHY-treated groups, resulted in the normalization of TGF-β 1/BMP-7 ratio (Figure 2(d)), suggesting a regulatory role in the balance of EMT and MET.

3.2. Restoration of TGF-β 1/BMP-7 Ratio Attenuated EMT and Induced MET of HSCs via p38 MAPK Signaling

To assess the effect of TGF-β 1/BMP-7 ratio on EMT/MET balance, we evaluated the downstream signaling of p38 MAPK, and then epithelial marker (E-cadherin) and mesenchymal markers (α-SMA and desmin) in quiescent, activated, and FZHY-treated HSCs, respectively. In contrast to the greatly upregulated transcription of p38 MAPK, α-SMA, and desmin in activated HSCs (control group versus model group, P < 0.05), their mRNA levels reduced after FZHY administration in a dose-dependent manner (Figures 2(a), 3(a)). On the contrary, the E-cadherin expression of activated HSCs was over 64.51% lower than the normal level (control group versus model group, P < 0.05), while treatment of FZHY recipe led to about 6.48-fold upregulation in its mRNA level (model group versus 20% FZHY-treated group, P < 0.05) (Figure 3(a)). Similarly to the changes in transcription, declined protein level of E-cadherin and increased protein levels of p38 MAPK, p-p38 MAPK, α-SMA, and desmin characterized the activation of HSCs (control group versus model group, P < 0.05) (Figures 2(b) and 2(c), 3(b) and 3(c)). However, FZHY-treated HSCs, with exposure to 10% and 20% FZHY-containing serum for 48 hours, demonstrated p38 MAPK, p-p38 MAPK, E-cadherin, α-SMA, and desmin levels comparable with those of quiescent HSCs (Figures 2(b) and 2(c), 3(b) and 3(c)). Thus the p38-dependent recovery from EMT to MET was indicated in FZHY-treated HSCs.

Figure 3

Inactivation of p38 MAPK signaling pathway prevents EMT and induced MET of HSCs. (a) The mRNA expression levels of E-cadherin (MET specific gene), α-SMA (EMT specific gene), and desmin (EMT specific gene) in the control, model, 5% FZHY-treated, 10% FZHY-treated, and 20% FZHY-treated group, respectively. (b) Western blot of E-cadherin, α-SMA, and desmin in the control, model, 5% FZHY-treated, 10% FZHY-treated, and 20% FZHY-treated group, respectively. (c) Relative protein levels of E-cadherin, α-SMA, and desmin. P < 0.05 compared with control group. # P < 0.05 compared with model group.

3.3. Rebalance of EMT and MET by FZHY Administration Abolished HSCs with Activated Phenotype In Vivo

No α-SMA positive cell was found in hepatic lobules of normal control group, indicating no activated HSC existed. However, numerous α-SMA positive cells with spindle or asteroid shape, most of which located in sinusoidal space, central vein, and fibrous tissue, were observed in rats exposure to CCl4 for 6 weeks (model group, Figure 4). At higher magnification of microscopy, the number of α-SMA positive cells rised up to 36.7 ± 11.4 per field (control group versus model group, P < 0.05). Dramatically, only 16.9 ± 8.9  α-SMA positive cells per field could be detected in the high-dose FZHY-treated group, which was significantly lower than that in the model group (P < 0.05, Figure 4).

Figure 4

FZHY recipe abolishes the EMT-dependent activation of HSCs. Arrows indicate the immunohistochemical staining of α-SMA in the control group (a), model group (b), low-dose FZHY-treated group (c), medium-dose FZHY-treated group (d), and high-dose FZHY-treated group (e), respectively (×200). (f) Quantification of α-SMA-positive HSCs in different groups. P < 0.05 compared with control group. # P < 0.05 compared with model group.

3.4. Treatment of FZHY Recipe Inhibited CCl4-Induced Liver Fibrosis

When compared with the normal control group, hepatic micro- and macrovesicular steatosis characterized the model group after CCl4 exposure. Other pathological disorders, including ballooning degeneration of hepatocytes, infiltration of inflammatory cells in lobules, and active proliferation of myofibroblast, were also visualized (Figure 5). Contrastively, most of these alternations were attenuated after different dosage of FZHY treatment (Figure 5).

Figure 5

Administration of FZHY recipe ameliorates the CCl4-induced pathological disorders in liver. (a–e) H&E staining of the control group (a), model group (b), low-dose FZHY-treated group (c), medium-dose FZHY-treated group (d), and high-dose FZHY-treated group (e), respectively (×100). (f) Hepatic hydroxyproline content of different groups. P < 0.05 compared with control group. # P < 0.05 compared with model group.

The VG staining uncovered similar phenomenon that little fibrosis existed in the liver of normal control rats, while excessive collagen deposition and well-delineated fibrosis septa emerged in the model group. Fibrous tissue is widely distributed in sinusoidal space and lobule, sometimes even bridging central vein and portal regions (Figure 6). After the treatment of FZHY recipe for 4 weeks, Metavir scores revealed the great improvement of liver fibrosis from 2.49 ± 0.61 (model group) to 1.57 ± 0.49 (medium-dose FZHY-treated group, P < 0.05), and 1.08 ± 0.33 (high-dose FZHY-treated group, P < 0.01) (Figure 6).

Figure 6

FZHY treatment attenuates CCl4-induced liver fibrosis. (a–e) VG staining of the control group (a), model group (b), low-dose FZHY-treated group (c), medium-dose FZHY-treated group (d), and high-dose FZHY-treated group (e), respectively (×100). (f) Metavir score of liver fibrosis in different groups. P < 0.05 compared with model group. # P < 0.01 compared with model group.

4. Discussion

Recently, increased studies have revealed the potential of FZHY recipe in prevention and treatment of organic fibrosis, such as hepatic, renal interstitial, and myocardial fibrosis [2–5, 11]. Similar to other traditional Chinese prescriptions, a variety of mechanisms are now recognized to underlie the effect of FZHY recipe, which includes (1) preventing HSCs from activation, as well as promoting their apoptosis; (2) inhibiting the proliferation of HSCs; (3) downregulating the expression and secretion of collagen; (4) inducing the activation of matrix metalloproteases; (5) protecting hepatocytes from apoptosis and facilitating their proliferation; (6) improving the liver function test; and (7) normalizing the proteome of liver tissue [2, 3, 5, 11]. However, these complicated mechanisms make it a difficult task to identify the comprehensive effect of FZHY recipe.

Balance of EMT and MET takes a central place in the activation of HSCs. HSCs which underwent EMT demonstrate most fibrosis-inducing characteristics, including migration towards the lesions, secretion of cytokines (TGF-β 1, PDGF, etc.), and synthesis of ECM (collagen, laminine, hyaluronic acid, etc.) [20, 21]. MET represents the counteracting process against EMT-related fibrotic response [9]. Nowadays, transdifferentiation between EMT and MET is indicated to be tightly controlled by cytokines, especially TGF-β 1 and BMP-7 [6]. After being incubated with FZHY-containing serum for 48 hours, HSCs showed dose-dependent reduction of TGF-β 1 expression in the 10% and 20% FZHY-treated groups. On the other hand, both the mRNA and protein levels of BMP-7, which experienced expression loss during EMT, upregulated significantly in the HSCs of 5%, 10%, and 20% FZHY-treated groups. TGF-β 1/BMP-7 ratio was finally normalized in the groups of FZHY administration, mimicking that of the control group. Therefore, FZHY recipe may exert its effects via regulating TGF-β 1/BMP-7 ratio in a dose-dependent manner.

MAPK signaling acts as the common pathway downstream to TGF-β 1 and BMP-7 stimulation [22]. p38 MAPK among MAPK family has been proved to mediate the TGF-β 1-induced synthesis of procollagen I in murine glomerular mesangial cells and seems to be indispensible for the glomerular sclerosis [23, 24]. Moreover, activation of p38 MAPK plays an essential role in the TGF-β 1-induced expression of mesenchymal marker (α-SMA) in human vascular smooth muscle cells [25]. Thus activation of p38 MAPK signaling, partially in the form of phosphorylated p38 MAPK, may reflect the key step from high TGF-β 1/BMP-7 ratio to EMT/MET unbalance. In contrast to their limited expression in the control group, elevated p38 MAPK and p-p38 MAPK in the model group indeed characterized the activation of p38 MAPK signaling during EMT of HSCs in our experiments. Dramatically, treatment of 10% and 20% FZHY-containing serum effectively inactivated p38 MAPK signaling by downregulating the expression and phosphorylation of p38 MAPK.

To shed light on the p38-mediated effect on phenotypic regulation, markers specific to mesenchymal and epithelial cells were subjected to investigation. E-cadherin, an epithelial cell specific transmembrane glycoprotein, forms complex with α-catenin, β-catenin, and catenin p120, which maintains the polarity of quiescent HSCs by binding to skeleton proteins. Significant decrease of E-cadherin contributes to the cytoskeleton rearrangement and polarity loss, and then EMT of HSCs [26]. In the present study, both the mRNA and protein levels of E-cadherin were significantly higher in the 10% and 20% FZHY-treated groups rather than those in the model group. Opposite to its promotional action on E-cadherin, treatment of FZHY-containing serum led to the inhibition of mesenchymal markers in HSC [27]. When determined with QPCR and Western blot, the expression levels of α-SMA and desmin were significantly lowered in the 5%, 10%, and 20% FZHY-treated groups than the controls. Moreover, the level of α-SMA and desmin proteins in the 10% and 20% FZHY-treated groups were in parallel to the normal control group, while the mRNA of α-SMA and desmin in these two groups were even lower than those in the normal control group. These observations qualify FZHY for a potential recipe to abrogate EMT and induce MET in HSCs.

Facilitating the transdifferentiation from EMT to MET, FZHY recipe is likely to ameliorate liver fibrosis by means of removing HSCs with EMT characteristics. In vivo treatment of FZHY recipe was then performed in the rodent model of CCL4-induced liver fibrosis. When compared to that of the control group, there were much more α-SMA-positive HSCs in the fibrous septum and sinusoid of the model group. FZHY therapy, however, statistically reduced these EMT signals specific to fibrosis-inducing HSCs. Being consistent with previous reports [2, 3, 11], attenuation of liver fibrosis was resultantly achieved by FZHY recipe. The Metavir score in high-dose FZHY-treated group was only 1.08 ± 0.33, which was significantly lower than that in the model group (2.49 ± 0.61, P < 0.05).

5. Conclusions

FZHY recipe may downregulate the TGF-β 1/BMP-7 ratio and, subsequently, result in the reversal of EMT to MET in activated HSCs via the inhibition of p38 MAPK signaling pathway. The rebalance of EMT and MET by FZHY recipe serves as a novel resolution of liver fibrosis on the basis of reducing HSCs with EMT characteristics.