Mallotus furetianus extract protects against ethanol‐induced liver injury via the activation of the cAMP‐PKA pathway

Abstract The protective effects of Mallotus furetianus extract (MF) on liver fibrosis induced with ethanol were examined using in vivo and in vitro model. MF treatment suppressed plasma alanine aminotransferase and aspartate aminotransferase activities in ethanol plus carbon tetrachloride (CCl4)‐induced cirrhosis rat model. MF also suppressed the increase in type l collagen and α‐smooth muscle actin expression in the livers of ethanol plus CCl4‐induced rat by the maintenance of intracellular glutathione levels. Furthermore, we evaluated the effect of MF on the alcohol‐induced activation of hepatic stellate cells (HSCs), which are responsible for the increased production and deposition of the extracellular matrix in liver injury. Here, we observed the enhancement of the intracellular reactive oxygen species (ROS) levels and the increase in type I collagen and a‐SMA expression in HSCs activated with ethanol. However, the enhanced ROS levels were suppressed with the treatments of MF or diphenyleneiodonium (DPI). Furthermore, the treatment of MF or DPI suppressed the increase in type I collagen and a‐SMA expression activated with ethanol. We also observed that the treatment of MF or LY194002 suppressed the increase in type I collagen expression in HSCs activated with ethanol, suggesting that ethanol induced type I collagen expression via the PI3K‐Akt signaling pathway. On the other hand, the suppression of the synthesis of type I collagen in ethanol and MF‐treated HSCs was inhibited by H‐89. From these results, MF may suppress the increase in the activity of NADPH oxidase in HSCs activated with ethanol through the cAMP‐PKA pathway.

by timely treatment. Therefore, it is important to find the treatment by which hepatic steatosis returns to normal liver.

Mallotus furetiamus is a tropical plant observed in the Hainan Island,
China (Lin & Zhou, 1992). Its leaves, commonly called "Shan Ku Cha," have been drunk as popular aromatic beverage and used as a folk medicine for cholecystitis. The extract also has activities of anti-oxidation and anti-atherosclerosis (Liu, Wang, Wu, Qu, & Lin, 2008). Recently, Huang et al. (2017) have shown that the extract of Mallotus furetiamus (MF) with hot water decreased the intracellular lipid accumulation in oleic acid-induced steatosis in hepatocellular carcinoma cells, a cellular hepatic steatosis model. From these results, it is suggested that MF is effective for the treatment of hepatic steatosis in ALD.
The animal models of ALD are important for research of the evaluation for the treatment employed for this disease. For animal models to study ALD rodents are the most suitable model and are the most commonly used (Brandon-Warnere, Schrum, Scmidt, & McKillop, 2012).
Currently, two animal models for the administration of alcohol, the Lieber-DeCarli liquid diet model (Lieber, De Carli, & Sorrel, 1989) and the Tsukamoto-French gastric model (Tsukamoto et al., 1995), have been used. However, both models do not bring about cirrhosis in rats. Moreover, Tipoe et al. examined the combined these two diets and showed that an increase in mediators of profibrogenesis was not equally with the histological evidence of fibrosis (Tipoe et al., 2008). On the other hand, Siegers, Pauli, Korb, and Younes (1986) showed that the intrapritonealy injection of low-dose carbon tetrachloride (CCl 4 ) and drinking of a 5% ethanol solution induced experimental fibrosis in rats within 4 weeks. The hepatic histological changes in the ethanol plus CCl 4 -induced fibrosis rat model and human alcoholic cirrhosis were similar. We also reported the histological change of liver in the ethanol plus CCl 4 -induced fibrosis model (Kojima-Yuasa et al., 2003Tamura et al., 2013).
In the progression of ALD, reactive oxygen species (ROS) play a key role (Crosas-Molist & Fabregat, 2015). ROS promotes necrosis and/or apoptosis of hepatocytes and also contributes to liver fibrosis by the increase in the release of pro-fibrotic cytokines and the expression of collagen gene in hepatic stellate cells (HSCs). When liver damage occurs, HSCs are activated and produce α-smooth muscle actin and type I collagen (Cui et al., 2011). Then, HSCs change the phenotype to differentiate to the myofibroblasts (Friedman, 2008).
Recently, we established an in vitro model of ethanol-induced injury with HSC and reported that the treatment of Ecklonia cava polyphenol prevents HSC activation (Takahashi et al., 2012).
In this study, the protective effect of MF against ALD was examined in in vivo ethanol plus CCl 4 -induced cirrhosis rat model and in vitro alcohol-injury model of HSCs.

| Extraction of Mallotus furetianus
Dried Mallotus furetianus (Lot No. 20110424044)  After filtration, the solvent was evaporated and then lyophilized to dryness.

| Animals
The study was accepted by the Osaka City University animal experiment committee (approval number: 1905017) and conducted in accordance with the regulations on animal experiments in Osaka City University. Male Wistar rats from Japan SLC, Inc., Shizuoka, Japan, were housed at 24 ± 1°C with humidity of 40%-60% with 12/12-hr light and dark cycle. The rats were given water and standard rat chow (LaboMR stock, Japan SLC, Inc.) ad libitum.

| Animal experiments
Male Wistar rats (180-210 g body weight) were fed an AIN93Gbased control diet for 5 days. After then, the 30 rats were divided into 5 groups based on body weight by persons not directly related to the experiment. The CCl 4 solution was given by intraperitoneal injection twice a week (Monday and Thursday), and 5% ethanol in water was given ad libitum for three weeks. In the morning of after three weeks, all the five groups of rats were sacrificed with isoflurane inhalation anesthesia and blood removal with a syringe from the heart.

| Histological analysis
Liver tissues of all rats were fixed with 10% buffered formalin fixative. Paraffin-embedded pathological tissue sections were stained with Weigert's elastic Van Gieson stain. The pathologist was not informed of the detail of the groups.

| Liver damage biomarkers
Alanine aminotransferase (ALT) and aspartate aminotransferase (AST) activities in serum of all rats ware assayed with a Transaminase CII-Test Kit (Fujifilm Wako Pure Chemical Co.).

| Preparation and culture of HSCs
Hepatic stellate cells isolated from male Wistar rats (300-350 g body weight) were incubated in Dulbecco's modified Eagle's medium (DMEM) with 10% fetal bovine serum for 3 days (Kojima-Yuasa et al., 2003). Then, HSCs were incubated in DMEM containing 100 mM of ethanol and with or without 12.5 µg/ml MF.

| Lipid peroxidation
The lipid peroxidation was evaluated by the measurement of the thiobarbituric acid-reactive species (TBARS) (Ohkawa, Onishi, & Yagi, 1979) as following. One hundred microliters of liver homogenate of all rats was incubated with 750 µl of thiobarbituric acid (TBA) (0.8%), 100 µl of sodium dodecyl sulfate (SDS, 8.1%), 750 µl of acetic acid buffer (2.5 M, pH 3.4), and 300 µl of distilled water for 20 min in boiling water. The absorbance was measured spectrophotometrically at 532 nm. The data were showed as nmol equivalent of malondialdehyde (MDA) using tetramethoxypropane as a standard.

| Hepatic glutathione (GSH) level
The livers of all rats were homogenated with 9 volumes of 25 mM of Tris-HCl buffer. After homogenization with Polytron PT 1600E and centrifugation, the supernatant treated with 5% perchloric acid (PCA) was analyzed by HPLC. Hepatic GSH level was assayed using 300 µl of the supernatant using the method of Sack, Willi, and Hunziker (2000). The amount of protein was assayed using the Bradford method (Bradford, 1976).
Hydrogen peroxide solution and normal goat serum were used for the blocking. Then, the specimens were incubated with an α-SMA monoclonal antibody or an antitype I collagen polyclonal antibody for primary antibody, and then biotinylated anti-mouse or anti-rabbit goat immunoglobulin for second antibody. The specimens were finally incubated with a horseradish peroxidase-labeled streptavidin-biotin complex and developed color by 3,3-diaminobenzidine tetrahydrochloride/nickel chloride. The immunocytochemical intensity of type I collagen or α-SMA was measured by ImageJ.

| Fluorescent immunostaining
HSCs were fixated with 4% paraformaldehyde fixative and then were incubated with anti-phospho-Akt (Ser473) polyclonal antibody for primary antibody and goat ant-rabbit IgG H&L (Alexa Fluor ® 488) antibody for secondary antibody. The cells were observed using FSX100 Bio Imaging Navigator. The immunofluorescence intensity of phospho-Akt was measured by ImageJ.

| Statistical analysis
The results are presented as the mean ± SE. Significant difference in assay values was performed using analysis of variance followed by Tukey's test. A value of p < .05 was considered statistically significant.

| Effect of MF treatment on plasma AST and ALT activities and liver fibrosis
The body weights of ethanol plus CCl4-treated rats were significantly lowered compared with control group. However, the body weights among ethanol plus CCl 4 -treated rats, ethanol plus CCl 4, and 0.012% MF group and ethanol plus CCl 4 and 0.04% MF group were not different significantly (Table 1).
After 3 weeks, the plasma AST and ALT activities in rats given with ethanol plus CCl 4 increased by 1.72-and 4.76-fold, respectively.
These enzyme activities in rats given with the ethanol plus CCl 4 and 0.012% MF showed a decreasing tendency from those in rats given with the ethanol plus CCl 4. Furthermore, these enzyme activities in the rats given with the ethanol plus CCl 4 and 0.04% MF were significantly reduced to the levels of the control group rats (Figure 1) suggesting that the protective effect of MF against liver injury with ethanol plus CCl 4 is dose-dependent.
Livers of the control rats, MF-treated rats, and CCl 4 -treated rats were not observed histological abnormalities. However, lipid accumulation and liver fibrosis were observed in the livers of the ethanol plus CCl 4 -treated rats. However, ethanol plus CCl 4 -induced liver steatosis and fibrosis were not observed by the treatment of MF (Figure 2).

| Effect of MF treatment on hepatic TBARS and GSH levels in ethanol plus CCl 4 -treated rats
The hepatic TBARS levels of rats treated with ethanol plus CCl 4 markedly increased as compared to that of the control group rats.  Note: Data are presented as the mean ± S.E. Values without a common letter are significantly different (p < .05).

TA B L E 1 The body weight of each group
However, MF treatment maintained the TBARS level at the control level (Figure 3a). On the other hand, treatment of ethanol plus CCl 4 significantly lowered the GSH level as compared to control, but MF treatment maintained the GSH level to the control level ( Figure 3b).

| Effect of MF on the type I collagen and α-SMA expression in ethanol-treated HSCs
The effect of MF on the type I collagen and α-SMA expression was examined by immunohistochemistry in ethanol-treated HSCs. As shown in Figure 4, the ethanol-induced increases in type I collagen and α-SMA expression were suppressed to near-control levels by the treatment of MF.

| Effect of MF on intracellular ROS levels in ethanol-treated HSCs
The intracellular ROS levels were measured using DCF-DA, which is

| Effect of diphenyleneiodonium (DPI) on intracellular ROS levels and on the expressions of type I collagen and α-SMA in ethanol-treated HSCs
We examined whether ethanol-induced ROS were mainly produced by nicotinamide adenine dinucleotide phosphate (NADPH) oxidase in HSCs. Cells were treated with NADPH oxidase inhibitor DPI (0.1 µM) and ethanol (100 mM) for 9 hr. The treatment of DPI suppressed the intracellular ROS levels enhanced with ethanol to the levels of the control cells ( Figure 6). These results indicated that NADPH oxidase inhibitor DPI inhibited the generation of ROS induced by ethanol, suggesting that NADPH oxidase could be a source of ROS in HSCs.
Furthermore, we treated HSCs with DPI and ethanol for 24 hr, and we examined the effect of DPI on the expression of type I collagen and α-SMA by immunohistochemistry. As shown in Figure 7, the ethanol-induced increases in type I and α-SMA expression were suppressed by the treatment of DPI to near-control levels. These results suggested that NADPH oxidase regulates the activation of HSCs through the ROS formation. Son, Hines, Lindquist, Schrum, and Rippe (2009) reported that the inhibition of PI3K signaling during the activation of HSCs inhibited the synthesis of type I collagen and reduced the expression of profibrogenic factors. Lu et al. (2010) have also shown that the PI3K-Akt pathway regulates proliferation and collagen production in bleomycininduced fibroblast. Therefore, the effect of LY294002, an inhibitor F I G U R E 5 Effect of MF on intracellular ROS levels in ethanol-treated HSCs. HSCs were incubated for 3, 6, or 9 hr with 100 mM of ethanol, with or without 12.5 µg/ml of MF. The fluorescence intensity of intracellular ROS was measured by ImageJ. Data are presented as the mean ± SE. Values without a common letter are significantly different (p < .01) Furthermore, the effect of MF and DPI on phosphorylation of AKT was investigated. As shown in Figure 9, the increase in the levels of phosphorylated Akt in ethanol-treated HSCs was suppressed with the treatment of MF or DPI. Mallat et al. (1998)    Control 100 mM EtOH + 12.5 µg/ml MF extract 100 mM EtOH + 0.1 µM DPI 100 mM EtOH CCl 4 -induced rats were suppressed with the treatment of MF by keeping intracellular ROS and glutathione levels. Furthermore, we evaluated the effect of MF on the HSCs activation, which is responsible for the increased production and deposition of the extracellular matrix in liver injury to determine the detailed mechanism of the protective effect of MF. Activation of HSCs includes the increased the expression of type I collagen, the expression of cytoskeleton markers, such as α-SMA, and increased proliferation (Friedman, 2000).

| D ISCUSS I ON
ROS are an important trigger for HSC activation (Svegliati-Baroni et al., 2001). In the present study, we also observed the enhancement of the intracellular ROS levels and the increase in the expressions of type I collagen and α-SMA in HSCs activated with ethanol.
Furthermore, the enhancement of ROS levels was suppressed with the treatments of MF or DPI, an inhibitor of NADPH oxidase. These results suggest that NADPH oxidase could be a source of ROS in HSCs activated with ethanol. MF or DPI treatments also suppressed the ethanol-induced increase in the expressions of type I collagen and α-SMA.
The PI3K-Akt signaling pathway may regulate the activation of HSCs (Son et al., 2009). We observed that the increase in the expression of type I collagen in HSCs activated with ethanol was sup- In conclusion, these results strongly suggest that MF is beneficial in the prevention of ethanol-induced liver injury. However, further studies should examine the clinical effect of MF to more fully understand its potential.

| E THI C AL RE VIE W
The study was approved by the Osaka City University animal experiment committee (approval number: 1905017) and conducted F I G U R E 1 0 Effect of MF and H-89, an inhibitor of PKA, on the type I collagen expression in ethanol-treated HSCs. HSCs were incubated for 24 hr with 100 of mM ethanol and with or without 12.5 µg/ml of MF or 1 μM of H-89. The immunocytochemical intensity of type I collagen was measured by ImageJ. Data are presented as the mean ± SE. Values without a common letter are significantly different (p < .01)

ACK N OWLED G M ENTS
We thank to Mr. Hideo Nagae, Ms. Mayu Goto, and Ms. Sumiko Ohnishi for assistance of animal experiments.

CO N FLI C T O F I NTE R E S T
The authors declare that they have no conflict of interest.