The Pharmacological Targets and Clinical Evidence of Natural Products With Anti-hepatic Inflammatory Properties

Inflammation contributes heavily to the pathogenesis of liver fibrosis, cirrhosis, and even hepatocellular carcinoma. Inflammation is probably a promising target for treatment of liver diseases. The natural products are considered as the potential source of new drug discovery and their pharmacological effects on hepatic inflammation have been widely reported. In this review, the natural products with anti-hepatic inflammatory properties are summarized based on their pharmacological effects and mechanisms, which are related to the suppression on the inflammation mediators including cytokines and chemokines, pattern recognition receptors, the activated transcriptional factors, and the potential regulatory factors. The clinical evidence is also summarized.


INTRODUCTION
Hepatic inflammation can be triggered by microbial infection, metabolic disorders, or exposure to drugs and toxic substances (Strazzabosco et al., 2005) and almost exists in every form of liver diseases. The acute hepatitis is usually short and self-limited, while the chronic is characterized with continuing inflammation, tissue injury, and healing. During chronic hepatitis, hepatic stellate cells (HSCs), the fibrogenic cells in the liver, are activated and initiate collagen deposition, which ultimately cause liver fibrosis and cirrhosis. Hence, targeting hepatic inflammation is an important strategy to block the progression to the end state of liver disease.
The anti-inflammatory effects of natural products are widely reported, which mostly target the inflammatory mediators including cytokines, chemokines, the receptors of cytokine or chemokine, the activated transcription factors, and the additional regulatory factors such as adhesion molecules, nitric oxide, carbon monoxide, and hydrogen sulfide (Kmieć, 2001;Anuar et al., 2006). In this review, the natural compounds with anti-hepatic inflammatory properties are summarized based on their pharmacological mechanisms and described according to their chemical classification ( Table 1, Figures 1,2). On the other hand, the clinical evidence from randomized controlled trials (RCTs) (Tables 2, 3) is also presented to visualize the entire profile of the studies on anti-hepatic inflammatory natural products.

SUPPRESSION OF CYTOKINES/CHEMOKINES SECRETION AND LEUKOCYTES INFILTRATION
The cytokines and chemokines are produced by the hepatic macrophages (Kupffer cells, KCs), natural killer (NK) cells, and NKT cells. KCs can release tumor necrosis factor alpha (TNFα), interleukin (IL)-6, IL-1β, or leukotrienes, which attract T cells and induce apoptosis of hepatocytes and activation of HSCs (Bilzer et al., 2006). NK cells produce interferon gamma (IFN-γ), IL-8 and apoptosis-inducing TNF-related apoptosis ligand and even directly promote hepatocyte death (Dunn et al., 2007). NKT cells are activated by the glycolipid antigens from bacteria (Kinjo et al., 2005) and involved in antiviral defense mechanisms in hepatitis B (Kakimi et al., 2000).
Recruitment of leukocytes consists of rolling on endothelium mediated by selectins, firm attachment to endothelium mediated by integrins, and migration through interendothelial spaces. TNF-α and IL-1 promote the expression of selectins and integrin ligands on endothelium. Chemokines produced by tissue macrophages increase the avidity of integrins for their ligands and promote directional migration of leukocytes. Monocytes are the largest type of leukocytes. Gr1 (hi) monocytes express high levels of C-C chemokine receptor type 2 (CCR2) but lack CX3C chemokine receptor 1 (CX3CR1). In inflammation, Gr1 (hi) monocytes actively enter inflamed tissue and are considered as the precursors for macrophages and dendritic cells. Gr1 (lo) monocytes lack CCR2 but express high levels of CX3CR1 existing in non-inflamed tissues, representing steady-state precursor cells for tissue macrophages (Tacke et al., 2007). CCR2 mediates entry of Gr1 (hi) monocytes into the inflamed tissues (Boring et al., 1997). Enhanced hepatic expression of C-C chemokine ligand 2 (CCL2) contributes to the formation and maintenance of inflammatory infiltration during chronic liver disease (Marra et al., 1998).
Matrine, a kind of alkaloid, was reported to ameliorate the hepatic infiltration of Gr1 (hi) monocytes and the expression

Inactivation of Inflammasome Pathway
Inflammasome is a cytoplasmic complex composed of NODlike receptor (NLR), the adapter, apoptosis associated speclike protein containing CARD (ASC), and the effecter, caspase-1 protein (Martinon et al., 2002). Several members of inflammasomes family have been identified (Negash and Gale, 2015). NOD-like receptor protein 3 (NLRP3) is the most thoroughly studied to date.

Prevention on HMGB1 Release
High mobility group box 1 (HMGB1) is one of the first identified members of the DAMP molecular family . HMGB1 release occurs during tissue injury or microbial invasion via passive and active ways. Passive release is nearly instantaneous, which occurs in the context of necrotic cell death. The active HMGB1 secretion depends on acetylation of nuclear localization sequences sites, which prevents the continuous bidirectional shuttle of HMGB1 between the cytoplasm and the nucleus, and leads to cytoplasmic accumulation of hyperacetylated HMGB1. Caspase-1 activated by the inflammasome system is required in pyroptosis, a gradual induction of programmed, proinflammatory cell death, which allows cytoplasmic HMGB1 to reach the extracellular space (Lamkanfi et al., 2010;Lu et al., 2012). HMGB1 can interact with TLR2, TLR4, TLR9, and the receptor for advanced glycation endproducts, in which, TLR4 is the dominant one (Andersson and Tracey, 2011). HMGB1 binds to TLR4 and activates macrophages (Yang et al., 2010) through nuclear factor κB (NF-κB) pathway (Park et al., 2004). HMGB1 is critical for neutrophil recruitment, injury amplification, and lethal liver injury (Huebener et al., 2015).
The alkaloid, betaine, decreased serum levels of alanine aminotransferase (ALT), aspartate aminotransferase (AST) and histological scores for steatosis, inflammation, and necrosis, as well as serum and hepatic HMGB1 in non-alcoholic fatty liver disease (NAFLD) induced by high-fat diet in rats (Zhang W. et al., 2013).
Anethole, a compound of phenylpropanoids, was found to attenuate liver injury and pro-inflammatory cytokines secretion in hepatic ischemia/reperfusion (I/R) mouse, and to inhibit the release of HMGB1 by prevention on nuclear translocation of interferon regulatory factor and interaction to histone acetyltransferase p300 (Cho et al., 2013).

Down-Regulation of TLR4 and CD14 Expression
TLRs, especially TLR4, are receptors of LPS, the component of outer membrane of Gram-negative bacteria. The gut-derived LPS is involved in the pathogenesis of inflammation in several kinds of liver diseases, such as NAFLD and alcoholic liver disease. LPS is recognized by the complex of CD14, TLR4, and myeloid differentiated protein-2 (Fujihara et al., 2003). A serum LPS binding protein (LBP) transfers LPS to CD14. CD14 concentrates LPS and presents it to TLR4 to activate the down-stream signaling cascade and ultimately initiate transcription of pro-inflammatory factors. Leptin was recently found to induce CD14 expression via activation of signal transducer and activator of transcription 3 (STAT3) signaling in KCs, resulting in enhanced responsivity against low-dose LPS in the liver (Imajo et al., 2012), which contributes to the progression of NAFLD.
The alkaloid, betaine, resulted in significant amelioration of serum ALT, AST, endotoxin, TNF-α, IFN-γ, and IL-18 and histology in liver, as well as down-regulation of the expression of hepatic TLR4 mRNA and protein in chronic alcoholic liver injury induced by high fat diet plus ethanol and fish oil in rats (Shi et al., 2010).
The flavonoids, oroxylin A and puerarin, were found to down-regulate the expression of TLR4 and CD14. Oroxylin A inhibited hepatic TLR4 expression and the downstream NF-κB activation in LPS/D-GalN-induced liver injury (Huang et al., 2015). Puerarin decreased hepatic inflammation in chronic alcohol-intake rats, and inhibited the protein expression of CD14, TLR2, and TLR4 (Peng et al., 2013).
The polyphenolic compounds, resveratrol and curcumin were reported to be effective on CD14 and TLR4. Resveratrol dramatically inhibited inflammation in a low-dose LPS-induced model of NASH through inhibition of the STAT3-CD14 pathway in KCs (Kessoku et al., 2016). Curcumin also reduced hepatic TLR2, TLR4, and TLR9 in Con A-stimulated liver tissues in mouse (Tu et al., 2012).
Aloin, a compound of anthraquinone, reduced liver injury in alcoholic hepatitis mice, and simultaneously, decreased serum LPS and the protein expression of hepatic TNF-α, TLR4, and MyD88 (Cui et al., 2014).

Inactivation of TLR4-Downstream Signaling
The downstream signaling of TLR4 involves myeloid differentiation primary response gene 88 (MyD88)-dependent and MyD88-independent pathway. In MyD88-dependent pathway, MyD88 associates with TLR and IL-1 receptor and recruits interleukin-1-receptor associated kinase (IRAK) to the receptor complex. IRAKs are subsequently phosphorylated and dissociated from the receptor complex and interact with  TNF-receptor associated factor 6 (TRAF6) (Li and Verma, 2002). TRAF6 activates mitogen-activated protein kinases (MAPK) and NF-κB pathway, and initiates pro-inflammatory genes expression. The alkaloid, betaine, ameliorated hepatitis and decreased hepatic mRNA and protein levels of TLR4 and NF-κB in NAFLD rats induced by high-fat diet (Zhang W. et al., 2013).
Pogostone, a compound of lactones, reduced liver injury and mortality induced by LPS in mice by inhibition on phosphorylation of p38 MAPK and NF-κB .
The polyphenolic compound, curcumin, was reported to inhibit the activation of p38 MAPK/JNK cascade, which correlated to its amelioration on LPS-induced liver injury (Zhong et al., 2016).
Esculentoside A, a compound of saponins, reduced liver injury, F4/80 (+) and CD11b (+) cells infiltration and activation of NF-κB and MAPK in the liver stimulated by CCl 4 .
IRAK-4, a member of the IRAK family, functions upstream of the other IRAKs and involves in signaling of innate immune responses from TLRs and T-cell receptors.
Taurine, a sulfur-containing β-amino acid, the major constituent of bile, protected against hepatic I/R injury and inhibited TNF-α expression in KCs partially by down-regulation of IRAK-4 and the downstream NF-κB activation (Sun et al., 2012).
The glucosides, paeoniflorin and salidroside, were found to inactivate NF-κB. Paeoniflorin ameliorated liver injury and inhibited serum TNF-α in NASH in rats and simultaneously inhibited the activity of ROCK and activation of NF-κB in liver . Paeoniflorin was also reported to inhibit the activation of NF-κB in liver tissue in Con A-induced hepatitis . Salidroside (Hu et al., 2014) inhibited Con A-induced hepatits, proinflammatory cytokines, hepatic infiltration of CD4 (+), CD8 (+) by regulating interferoninducible CXCL10 and NF-κB activation in liver tissue (Hu et al., 2014).
The polyphenol, curcumin, down-regulated the protein expression of NF-κB in the liver of db/db mouse (Jiménez-Flores et al., 2014).
Silybin, a compound of flavonolignans, reduced plasma levels of transaminases and liver content of pro-inflammatory cytokines, inhibited hepatic NF-κB activation, and increased plasma and tissue levels of IL-10 in hepatitis induced by Con A in mouse (Schümann et al., 2003).
The terpenoid, triptolide suppressed neutrophil infiltration, pro-inflammatory cytokine level and NF-κB activation in I/R liver in mice (Wu et al., 2011).
Synthesis of GSH is regulated by catalytic and modifier subunit of glutamate-cysteine ligase, which are characteristic target genes of nuclear factor (erythroid-derived 2)-like 2 (Nrf2) (Solis et al., 2002).
The alkaloids, berberine and tetrandrine, were reported to target the production of ROS. Berberine dramatically attenuated the hepatic histopathologic damage, restored the liver function, and decreased the oxidative stress level in I/R liver in rat (Sheng et al., 2015). Tetrandrine ameliorated I/R liver injury by suppressing oxidative stress, including decreasing malondialdehyde (MDA), myeloperoxidase (MPO), and increasing SOD (Cheng et al., 2008).
The flavonoids, including scutellarin, baicalin, hesperidin, quercetin, wogonoside, and oroxylin A, were reported to inhibit ROS. Scutellarin (Tan et al., 2007) decreased the production of ROS and the expression of inducible nitric oxide synthase (iNOS) in Con A-induced hepatitis. Baicalin  reduced MPO activity and lipid peroxidation, and increased the anti-oxidative SOD expression in liver tissue in Con A-injected mice. Hesperidin down-regulated the expression of nitric oxide, hydroperoxides, and thiobarbituric acid reactive substances, and increased GSH, glutathione reductase, GPx and glutathione Stransferases (GST) in the liver of rats treated with LPS (Rotimi et al., 2016). Quercetin reduced secretion of AST, MDA, and increased levels of GSH and SOD in rat primary hepatocytes stimulated by ethanol (Liu et al., 2010). Wogonoside (Gao et al., 2016) decreased inflammatory factors accompanied with inhibition on the production of MDA by activating Nrf2 and increasing heme oxygenase-1 (HO-1) and catalytic subunit of glutamate-cysteine ligase in LPS/D-GalN-induced liver injury.
The glucosides including paeoniflorin, geniposide, pinocembrin-7-O-β-D-glucoside and forsythiaside A, were found to inhibit ROS. Paeoniflorin (Tao et al., 2016) ameliorated I/R liver injury by decreasing MDA content and enhancement of the activities of hepatic SOD, GSH, and GPx. Geniposide  ameliorated I / R liver injury by decreasing MDA and increasing the ratio of GSH / glutathione disulfide and the protein expression of HO-1 in liver. In chronic ethanolchallenged mouse, pinocembrin-7-O-β-D-glucoside significantly reduced hepatic ROS and MDA, and restored the activity of GSH, SOD, and GPx and increased the hepatic expression of Nrf2 and the downstream anti-oxidant HO-1 . Forsythiaside A (Pan et al., 2015) protected against LPS/D-GalNinduced liver injury by up-regulated the expression of Nrf2 and HO-1 in the liver.
The polyphenolic compound, curcumin, was reported to reduce hepatic oxidative stress in Con A-induced hepatitis (Wang et al., 2012), to decrease the levels of MDA and 4-hydroxy nonyl alcohol in the liver of NASH rats  and to accelerate liver antioxidant enzymes levels, including SOD, catalase, GSH and GPx, in LPS-induced liver injury (Zhong et al., 2016).
The saponnins, glycyrrhizin and esculentoside A, inhibited the production of ROS. glycyrrhizin (Tsuruoka et al., 2009) reduced hepatitis in Con A-stimulated mice and inhibited the expression of iNOS. Esculentoside A  inhibited MDA release and increased GPx activity in liver induced by CCl4 and LPS/D-GalN.

CLINICAL EVIDENCE
The most powerful evidence of a drug candidate comes from RCTs. In this part, the natural compounds verified in RCTs are summarized.
The alkaloids, matrine and berberine, were verified in RCTs. Matrine combined with INF exhibited better clinical efficacy including negative conversion rate of hepatitis B virus e-antigen, hepatitis B virus DNA, and AST level, and fewer adverse effects than did INF monotherapy in patients with chronic hepatitis B (Wang et al., 2017). Berberine improved serum ALT, AST, hepatic fat content, and insulin resistance in NAFLD patients (Yan et al., 2015;Wei et al., 2016).
Taurine, the sulfur-containing β-amino acid, was reported to decrease serum ALT and AST activities and levels of cholesterol and triglyceride in hepatitis C patients (Hu et al., 2008).
The flavonoid, quercetin, exhibited safety (up to 5 g daily) and a potential for antiviral activity in some hepatitis C patients in a phase I dose escalation study .
Silybin, a compound of flavonolignans, has also been reported to improve liver enzymes in alcoholic or viral hepatitis (Vailati et al., 1993). In a randomized trial on NASH, patients in silybin group had reductions in fibrosis based on histology and AST than did the placebo group (Chan et al., 2017).
The polyphenols, resveratrol and curcumin, were verified in RCTs. Resveratrol combined with pegylated-INF-α2b and Ribavirin was reported to improve AST, viremia, histological activity index and C-reactive protein in hepatitis C patients comparing to that in pegylated-INF-α2b, Ribavirin and placebo group (Pennisi et al., 2017). Curcumin significantly reduced the liver fat content, serum levels of ALT, AST, total cholesterol, triglycerides and glucose in NAFLD patients compared with the placebo group (Rahmani et al., 2016;Panahi et al., 2017).
The saponin, glycyrrhizin, is used intravenously or orally in chronic hepatitis B and C patients. And its preparation under the name of Stronger Neo-Minophagen C decreased ALT and AST levels in patients with chronic hepatitis in multiple double-blind studies (van Rossum et al., 1999;Iino et al., 2001;Zhang and Wang, 2002). It was suggested that glycyrrhizin had a preventive effect on the development of hepatocellular carcinoma in patients with chronic hepatitis C (Arase et al., 1997).
It's interesting that the effects of curcumin and resveratrol, the two controversial polyphenolic compounds, were both supported by the results of RCTs. Curcumin targets multiple anti-inflammatory mechanisms, which indicates that curcumin is probably a powerful anti-inflammatory compound. But recently, the effects of curcumin were questioned because it was classified as a candidate of panassay interference compounds and invalid metabolic panaceas (Nelson et al., 2017). Although the RCTs on many diseases based on its anti-inflammatory effects have been conducted (Derosa et al., 2016;Sahebkar et al., 2016), there was rare report about the effect of curcumin on human hepatitis under RCT design until the reports of Panahi et al. and Rahmani et al. come out. It is encouraging even if the histological evidence were absent. The anti-inflammatory potential of resveratrol was demonstrated in metabolic syndrome in animal studies, while the results from trials in metabolic syndrome were not as promising as the pre-clinical data. It was also reported that resveratrol did not decrease the levels of ALT and AST, as well as the insulin resistance and steatosis in NAFLD patients compared with baseline (Chachay et al., 2014). To confirm the effects of resveratrol, more RCTs designed with histological checkpoints needed to be developed.
On the other hand, several natural products with anti-hepatic inflammatory properties are the predominant active compounds in the traditional Chinese formulas, Xiao Chai Hu Tang (in Kampo name of Sho-saiko-to) and Yin chen hao tang (in Kampo name of Inchin-ko-to) which are popular medicine used to treat liver diseases in ancient China and have been verified in RCTs up to now. Xiao Chai Hu Tang [containing baicalin, baicalein, scutellarin (Shimizu, 2000), oroxylin A (Liu et al., 2002), wogonin and ginsenoside Rg1 (Ohtake et al., 2004)] was reported to increase INFγ and hepatitis B virus core antibody and hepatitis B virus e-antibody in peripheral blood mononuclear cells from patients of chronic hepatitis B (Kakumu et al., 1991), which partly contributed to its promotion on clearance of hepatitis B virus eantigen in the children with chronic hepatitis B (Tajiri et al., 1991) and to adjust the decreased IL-10 production and the increased IL-4 and IL-5 production of mononuclear cells from patients with hepatitis C (Yamashiki et al., 1997). Yin chen hao tang [containing geniposide, genipin (Inao et al., 2004), and emodin (Imanishi et al., 2004)] was reported to improve serum ALT, AST, gamma-glutamyl transferase, and fibrosis indicators (Hyaluronic acid, type III procollagen N paptide, and type IV collagen) in postoperative biliary atresia patients (Kobayashi et al., 2001;Tamura et al., 2007).

CONCLUDING REMARKS
The anti-inflammatory effects of natural compounds in liver have been widely demonstrated in different models of liver injury in vitro and in vivo. It has been demonstrated that the inflammation mediators including cytokines, chemokines, pattern recognition receptors, the activated transcriptional factors and the regulatory factor, ROS, are the potential targets of these compounds. The anti-hepatic inflammatory activity is not limited to certain class of compounds, since the alkaloids, quinine, flavonoids, glucosides, phenylpropanoids, polyphenols, sapoins, terpenoids, etc., all present activity on hepatitis. It is also obvious that the flavonoids are the majority of these compounds and mostly target the inhibition of oxidative stress.
But, most of the basic researches are limited to observations of the changes of inflammation parameters and the relative pathways. Only berberine was confirmed to inhibit inflammasome by inactivation of P2x7 by using P2x7knockdown cell line. To identify the pharmacological targets, it is necessary to employ the transgenic models of the potential target molecules in studies in the future. On the other hand, using the ideal animal models with the characters of human hepatitis in pre-clinic state probably prevent the controversial results from the clinical trials. For example, con A-induced acute hepatitis by a massive cytokine storm is not a reliable animal model of autoimmune hepatitis which is characterized by persistent chronic inflammation. While, the cytochrome P450 2D6 (CYP2D6) humanized mouse targeting the human autogantigen CYP2D6 is a valid model to study autoimmune mediated liver damage (Christen et al., 2007).
In most of the RCTs, ALT and AST were measured as hallmark of hepatitis. The histological evidence was only provided in few clinical trials, such as the RCTs of silybin and resveratrol. Since liver biopsy is still the golden diagnosis standard for the most types of hepatitis, the histological examination should be conducted to confirm the effects of natural compounds on inflammation in liver tissue.
Finally, as we see in the clinical evidence, the traditional Chinese medicine formulas which contain groups of compounds present positive results on hepatitis. It seems like a potential strategy for hepatitis treatment to develop new compound formulas consisting of natural compounds with clear chemical structures and action targets, since the pathogenesis of hepatitis is more like an orchestra of pathological mechanisms instead of single target.
The available evidence from basic and clinical research suggest the natural compounds with anti-hepatic inflammatory properties are potential resource for new drug development for liver diseases. The underlying mechanisms and safety are deserved to be investigated thoroughly by optimized animal and clinical studies.

AUTHOR CONTRIBUTIONS
JP conceived and designed the project and wrote the manuscript.