Nerve growth factor induced farnesoid X receptor upregulation modulates autophagy flux and protects hepatocytes in cholestatic livers

Ming-Shian Tsai, Hui-Ming Lee, Shih-Che Huang, Cheuk-Kwan Sun, Ting-Chia Chiu, Po-Han Chen, Yu-Chun Lin, Tzu-Min Hung, Po-Huang Lee, Ying-Hsien Kao

Abstract

Upregulation of nerve growth factor (NGF) in parenchymal hepatocytes has been shown to exert hepatoprotective function during cholestatic liver injury. However, the modulatory role of NGF in regulation of liver autophagy remains unclear. This study aimed to scrutinize the regulatory role of NGF in hepatic expression of farnesoid X receptor (FXR), a bile acid (BA)-activated nuclear receptor, and to determine its cytoprotective effect on BA-induced autophagy and cytotoxicity. Livers of human hepatolithiasis and bile duct ligation (BDL)-induced mouse cholestasis were used for histopathological and molecular detection. The regulatory roles of NGF in autophagy flux and FXR expression, as well as its hepatoprotection against BA cytotoxicity were examined in cultured hepatocytes. FXR downregulation in human hepatolithiasis livers showed positive correlation with hepatic NGF levels. NGF administration upregulated hepatic FXR levels, while neutralization of NGF decreased FXR expression in BDL-induced cholestatic mouse livers. In vitro studies demonstrated that NGF upregulated FXR expression, increased cellular LC3 levels, and exerted hepatoprotective effect in cultured primary rat hepatocytes. Conversely, autophagy inhibition abrogated NGF-driven cytoprotection under BA exposure, suggesting involvement of NGF-modulated auophagy flux. Although FXR agonistic GW4064 stimulation did not affect auophagic LC3 levels, FXR activity inhibition significantly potentiated BA-induced cytotoxicity and increased cellular p62/SQSTM1 and Rab7 protein in SK-Hep1 hepatocytes. Moreover, FXR gene silencing abolished the protective effect of NGF under BA exposure. These findings support that NGF modulates autophagy flux via FXR upregulation and protects hepatocytes against BA-induced cytotoxicity. NGF/FXR axis is a novel therapeutic target for treatment of cholestatic liver diseases.

Keywords: Autophagy; Bile acids; Cholestasis; Farnesoid X receptor; Hepatocytes.

1. Introduction

Cholestasis is clinically one of the most and severe liver diseases, in which disruption and dysregulation of hepatobiliary transport systems often result in both systemic and intrahepatic accumulation of bile acids (BAs) and relevant toxic metabolite compounds [1]. BAs are amphipathic molecules synthesized from cholesterol in a reaction mediated by the hepatic enzyme cholesterol-7α-hydroxylase [2]. In cholestatic livers, accumulated BAs lead to chronic inflammation that in turn deteriorates liver function and provides a pro-oncogenic microenvironment in biliary tract and the liver [1, 3-5]. It has been found that the accumulated BAs contribute to parenchymal cell death and mesechymal cell activation, thereby leading to periductular fibrosis, biliary fibrosis, and eventual cirrhosis [6]. BA exposure has been linked with mitochondrial dysfunction and hepatobiliary malignancy in cholestatic livers [7, 8].
Emerging evidence indicates that BAs play an important role in the regulation of liver physiological functions, including their own synthesis, fatty acid, lipid, and lipoprotein synthesis, and glucose metabolism in the liver [9]. The BA-modulated functions in various types of cells have been discovered to be mediated through different receptors, including a nuclear hormone receptor farnesoid X receptor (FXR) [10] and a cell surface G protein-coupled receptor TGR5 [11]. Both FXR and TGR5 are found to possess pleiotropic functions, including liver regeneration [12-14] and immune modulation [15, 16]. A growing number of studies showed that FXR is downregulated in obstructively cholestatic human and animal livers [17, 18]. Genetic studies suggest that FXR gene mutation contributes to progression of familial intrahepatic cholestasis [18, 19], while combined deletion of FXR and small heterodimer partner genes experimentally induces juvenile onset of cholestasis in mice [20]. Other lines of evidence reveals that FXR functionally regulates hepatic autophagy, thereby exerting protective effect against BA-induced cytotoxicity [21-23]. Consistent to the TGR5-driven anti-tumor effect, FXR also promotes liver regeneration after partial hepatectomy and alleviates aging-related liver regenerative defects [12], whereas treatment with dual TGR5/FXR agonist experimentally improves BA homeostasis and decreases hepatic inflammation and cholangiopathic injury in mice [24].
Nerve growth factor (NGF) is the prototype member of NGF family, whose physiological function is originally identified in nervous system. Increasing evidence suggests that NGF not only pathophysiologically involves in liver tumor formation [25], but also participates in modulation of liver metabolism [26]. Our previous studies have identified that expression of NGF in parenchymal hepatocytes exerts hepatoprotective effect partly through upregulating sirtuin 1 expression and anti-oxidative activity in the cholestatic livers [27, 28]. Despite hepatoprotection of sirtuin 1 and FXR signaling axis against cholestatic liver injury [29-31], the molecular link between NGF and FXR expression in the cholestatic livers, however, remains unclear. To test this hypothesis, this study scrutinized the relationship between hepatic NGF and FXR levels in both human hepatolithiasis and mouse cholestatic livers, and determined the role of NGF/FXR axis in BA-induced autophagy and cytotoxicity using cultured hepatocytes. This study unravels a possible therapeutic target in the treatment of obstructive cholestasis.

2. Materials and Methods

2.1. Chemicals

BA chemicals, including deoxycholic acid (DCA) and chenodeoxycholic acid (CDCA), and synthetic FXR agonist GW4064 and antagonist Z-guggelsterone (Z-GS) were purchased from Sigma-Aldrich Chemicals (St. Louis, MA) and dissolved in DMSO as stock solutions. Autophagy flux inhibitors, including 3-methyladenine (3-MA), bafilomycin A (BMA), and chloroquine (CQ) were purchased from Sigma-Aldrich Chemicals.

2.2. Human hepatolithiasis specimens

Liver tissues were collected from four patients with hepatocellular carcinoma (HCC) and five patients with hepatolithiasis undergoing liver resection with surgical procedures as previously reported [28]. All collection procedures were approved by the Institutional Review molecular and histological studies.

2.3. Western blotting

It was quantified using a BCA-based protein assay kit (Pierce Biotechnology, Rockford, IL, USA) and an equal amount of total protein for each lane was subjected to SDS-PAGE and immunoblotting as previously described [32]. Immunodetecting antibodies raised against actin, TGR5, and FXR were purchased from Santa Cruz Technology (Santa Cruz, CA), while those against LC3I/II, Rab7, and p62/SQSTM1 were from Cell Signaling (Beverly, MA). Secondary antibodies horseradish peroxidase (HRP)-conjugated anti-rabbit and mouse IgGs were from Jackson ImmunoResearch Laboratories (West Grove, PA). The immunoreactive signals were visualized by enhanced chemiluminescence detection system (Millipore, Billerica, MA), documented on a digital imaging system (UVP), and densitometrically analyzed by using ImageJ software (NIH). Relative protein levels were expressed as the ratios of densities between proteins and actin for each sample.

2.4. Histopathology and immunohistochemistry (IHC) staining

Formalin-fixed and paraffin-embedded human liver tissues were sectioned and subjected to regular hematoxylin and eosin (H&E) as well as FXR immunohistochemical (IHC) staining as previously described [27, 28]. In brief, sections were deparaffinized and rehydrated, followed by antigen retrieval treatment and incubation with primary antibodies at 4 °C overnight. Anti-FXR antibody was purchased from Santa Cruz Technology. The antigen location in tissue sections was visualized with an HRP-linked polymer Envision detection system (DAKO, Glostrup, Denmark) followed by counterstaining with hematoxylin. NHC sections were treated with equimolar concentrations of isotype-matched normal IgG as negative controls.

2.5. Animal model of bile duct ligation-induced cholestasis

ICR male mice were raised with ad libitum access to food at 20-22 °C with a 12 hour light-dark cycle in the Animal Center of I-Shou University. All procedures were carried out in accordance with the National Institutes of Health guide for the care and use of laboratory animals (NIH Publications No. 8023, revised 1978), and approved by the Institute of Animal Care and Use Committee at E-Da Hospital (Approval No. IACUC-102033 and 103030). Eight-week-old mice were randomly assigned to experimental groups. Common bile duct ligation (BDL) was performed to induced cholestatic liver injury as previously described. [27] For NGF therapy and antibody neutralization treatments, recombinant NGF (1 mg/kg), rabbit anti-NGF neutralizing antibody (2 mg/kg; AB1526SP, Millipore), and equivalent doses of PBS and specie- and isotype-matched normal IgG (N-IgG) were intraperitoneally injected twice weekly after BDL surgery. Fourteen days after BDL surgery, Livers were dissected and subjected to protein expression quantification.

2.6.Cell culture

For NGF stimulation experiments, primary hepatocytes were isolated from adult male Sprague Dawley rats using a two-step collagenase perfusion method as previously described [33]. For mechanistic study, human HCC cell lines, SK-Hep1 cells (BCRC67005) were purchased from Bioresource Collection and Research Center (Hsin-Chu, Taiwan). SK-Hep1 cells were maintained in DMEM (Invitrogen, Logan, UT) supplemented with 10% heat-inactivated FBS (Invitrogen) and standard antibiotics. Cells were incubated at 37ºC in a humidified atmosphere of 5% CO2 in air, and the nutrient medium was renewed twice a week. Before experiments, cells were trypsinized and seeded on culture dishes and plates for protein expression and cell cytotoxicity assays, respectively.

2.7. Immunofluorescent staining

Cellular distribution of autophagy LC3 marker was visualized using immunofluorescent staining method as previously described [34]. In brief, adherent primary hepatocytes on chamber slide after treatment were immediately fixed with ice-cold methanol/acetone mixture, followed by membrane permeabilization. Anti-LC3I/II antibody purchased from Cell Signaling was used as detecting antibody. After all staining procedures, slides were mounted in DAPI Fluoromount-G medium (SounthernBiotech, Birmingham, AL), which contains 4′,6-diamidino-2-phenylindole (DAPI) for nuclear visualization. and observed under a fluorescent microscope (Axiovert, Zeiss, Germany).

2.8.Cytotoxicity assay

For determining viability of BA-exposed cells, 1×104 cells per well were seeded onto 96-well plate 24 hours before experiment. After being pretreated with NGF, FXR activity modulators, or autophagy inhibitors for 24 hours, the cells were exposed to BAs for another 24 hours and subjected to an MTT-based cellular assay which detects the overall NAD- and NADP-dependent LDH activity in viable cells was used as previously described [35]. The

2.9. FXR gene silencing by siRNA another 24 hours, followed by cell viability assay.

2.10. Statistical analysis

In vivo and in vitro results were presented as mean ± standard error of mean (SEM) and mean ± standard deviation (SD) as indicated. Significant differences of parametric data were analyzed one-way ANOVA followed by the Bonferroni post hoc test. Spearman correlation analysis was performed using PASW statistics software (ver. 18, SPSS Inc. Chicago, IL). p values less than 0.05 indicate statistical significance.

3. Results

3.1. Hepatic expression of FXR and autophagy markers in human hepatolithiasis livers

As FXR has been well identified to regulate hepatic autophagy and metabolism [21-23], we first sought to determine changes in expression levels of bile acid receptors and autophagy markers in human cholestatic livers. For this purpose, five patients with hepatolithiasis were enrolled and four paralesional liver tissues were collected from HCC patients without hepatolithiasis as NHC and used for molecular and histological studies. Previous serum biochemistry study showed no significant differences of liver injury markers between NHC and lithiasis cases [28]. Western blot detection of whole liver protein extracts was used to measure FXR, TGR5 and autophagy marker expression (Fig. 1A). Subsequent densitometry indicated FXR expression markedly elevated in contralateral part and significantly downregulated in lithiasis part of livers (Fig. 1B). By contrast, TGR5 protein expression significantly decreased in lithiasis tissues compared to that in non-hepatolithiasis control livers, but was not different from contralateral levels. Both LC3II/I ratios and p62/sequestosome1 (SQSTM1) protein contents markedly increased in both lithiasis and contralateral regions, suggesting autophagy flux interruption and autophagosomal accumulation therein. Conversely, as an important endolysosome trafficking regulator to avoid autophgosome accumulation, Rab7 levels reduced in lesions of human lithiasis livers (Fig. 1C), also implicating autophagy flux retardation in the diseased livers. Despite limited number of human samples, Spearman’s correlation analysis using previously reported NGF data [28] indicated that protein expression of FXR (Figs. 1D), but not TGR5 (Fig. 1E), showed positive relationship with hepatic NGF levels in hepatolithiasis. Moreover, hepatic NGF and FXR expression showed marginal relationship with autophagic marker expression in human hepatolithiasis livers (supplemental Fig. S1). These findings strongly suggest a role for NGF in regulation of hepatic FXR expression and autophagy.

3.2. Localization of FXR in parenchymal hepatocytes of human hepatolithiasis livers

To scrutinize the histological changes and FXR distribution in cholestasis-injured human livers, the paraffin sections were subjected to H&E and IHC stainings, respectively. . The H&E staining results indicated that NHC livers collected from normal HCC tissues showed mild fibrotic and fatty changes and the diseased lesion of hepatolithiasis livers displayed typical bile duct metaplasia, one of histopathology features in cholestatic livers (Fig. 2). The FXR IHC staining patterns clearly showed that FXR protein was intensely located in the nuclei of parenchymal hepatocytes of normal liver tissues, including NHC and contralateral region of lithiasis livers. Intense immunodetected FXR signal of lithiasis livers was only seen in the parenchymal cytoplasms, biliary ductular epithelial cells as well as infiltrated inflammatory cells. The lack of nuclear FXR localization in hepatocytes implicated insufficient FXR activity in diseased lithiasis livers.

3.3. NGF experimentally upregulated FXR expression in mouse cholestatic livers

To observe the expression changes of TGR5 and FXR and the biomodulatory effect of NGF in cholestatic livers, we next experimentally conducted BDL-induced cholestasis in a mouse model. The liver protein extracts of normal mice and those simultaneously receiving treatment of recombinant NGF peptides, neutralizing antibodies, or respective solvent controls were subjected to western blotting detection (Fig. 3A, 3C). The densitometry analysis revealed that FXR expression was significantly suppressed in BDL-injured mouse livers, while NGF supplementation remarkably restored hepatic FXR expression in injured livers (Fig. 3B). By contrast, administration of NGF neutralizing antibodies remarkably potentiated the FXR downregulaton in diseased livers as compared to corresponding N-IgG controls (Fig. 3D). Despite no change of autophagy LC3 expression in NGF supplemented and depleted cholestatic mouse livers, NGF therapy significantly reduced hepatic p62 and Rab7 contents (Fig. 3B), while NGF neutralization dramatically increased p62 expression in BDL-injured mouse livers (Fig. 3D). These findings strongly suggest that NGF might modulate not only FXR expression, but also autophagy during the progression of liver cholestasis.

3.4. NGF treatment upregulated FXR expression in cultured primary rat hepatocytes

To further clarify the regulatory role of NGF in FXR expression, the protein lysates of primary rat hepatocytes with NGF stimulation were analyzed by western blot (Fig. 4A). The densitometrial analysis clearly demonstrated that NGF stimulation at concentrations equal to and higher than 1 ng/ml significantly upregulated expression of FXR (Fig. 4B), but not that of TGR5 protein (Fig. 4C) in rat primary hepatocytes. To examine the modulatory effect of NGF on autophagy flux in hepatocytes, western blotting detection was used to measure autophagy marker levels in cultured rat primary hepatocytes (Fig. 4D). Densitometry confirmed that exogenous NGF treatment significantly increased Rab7 protein contents (Fig. 4F), while it merely induced significant elevation of LC3II/LC3I ratio at 10 ng/ml in cultured hepatocytes (Fig. 4E), although p62 protein was undetectable in rat primary hepatocytes (data not shown). Immunofluorescent staining further evidenced intense punctuate pattern of LC3 expression in the cytoplasms of NGF-treated primary hepatocytes (Fig. 4G, 4H). These findings suggest that NGF may regulate autophagy flux in hepatocytes via increasing FXR expression and its activity.

3.5. NGF stimulation and FXR antagonism modulated BA-suppressed autophagy flux in SK-Hep1 hepatocytes

As both primary and secondary BAs have been previously reported to elicit hepatotoxicity via suppression of autophagic flux [8, 36], we sought to observe the in vitro effects of NGF stimulation and FXR activation on autophagy flux in cultured SK-Hep1 hepatocytes. The NGF-pretreated SK-Hep1 cells were further exposed to either CDCA, a primary BA, or DCA, a secondary BA, for 24 hours and the lysates were subjected to western blotting detection (Fig. 5A, 5E). Subsequent densitometry analysis results showed that both DCA and CDCA potently increased cellular content of mature autophagosome markers, as shown by increased LC3II/I ratios in all BA-treated groups. It is notable that 24-hour NGF pretreatment significantly potentiated the elevation of LC3II/I ratios in the cells exposed to high-dose CDCA, but not DCA(Fig. 5B, 5F). Meanwhile, NGF increased p62 contents (Fig. 5C, 5G) at both high-dose BAs. In addition, NGF markedly potentiated DCA-elicited Rab7 downregulation at lower BA doses (Fig. 5D), but did not affect Rab7 expression in CDCA-exposed hepatocytes, (Fig. 5H).
We next examined the effects of FXR activity on autophagy flux in the cells treated BAs in the presence of either GW4064, a selective FXR agonist, or Z-GS, a synthetic FXR antagonist (Fig. 6A). Western blotting and densitometry revealed that FXR agonistic treatment with GW4064 did not change the DCA- and CDCA-elevated LC3II/I ratios, while FXR inactivation by Z-GS significantly ameliorated both BA-increased auophagy flux (Fig. 6B). Similarly, FXR agonistic treatment did not affect the DCA- and CDCA-elevated p62 and Rab7 contents, whereas FXR inactivation prominently potentiated the increases of cellular p62 and Rab7 protein levels in the BA-treated cells (Fig. 6C, 6D). These findings strongly suggest that NGF/FXR are both able to modulate autophagy flux in parenchymal hepatocytes.

3.6. Involvement of NGF-enhanced autophagy flux in its protection against BA-induced SK-Hep1 cytotoxicity

To verify the contribution of NGF and upregulated FXR to its hepatoprotective effect in vitro, cellular viability assay was performed by using the cultured hepatocytes exposed to either BA species. The cytotoxicity data demonstrated that 24-hour NGF pretreatment significantly enhanced the viability of SK-Hep1 cells under exposure to DCA or CDCA at higher doses (Fig. 7A, 7B). Next, to test the involvement of autophagy flux in the NGF-driven cytoprotective effect, the cells were pretreated with NGF and different autophagy inhibitors for 24 hours, followed by another 24-hour exposure to high-dose DCA or CDCA. The viability assay data indicated that autophagy inhibition by 3-MA, BMA and CQ prominently abolished NGF-driven cytoprotection against insults from both high-dose BAs (Fig. 7C, 7D), suggesting that NGF-enhanced autophagy flux is involved in its protective effect against the BA-exerted hepatotoxicity.

3.7. Involvement of FXR expression and its activity in NGF-enhanced autophagy and protection against BA-induced cytotoxicity

To examine the role of FXR expression in the NGF-enhanced autophagy and hepatoprotection, constitutive FXR expression in SK-Hep1 hepatocytes was silenced by specific siRNA transfection, which effectively reduced FXR expression after 48-hour of siRNA delivery (Fig. 8A). Subsequent exposure to NGF and BAs and viability assay demonstrated that FXR gene silencing treatment significantly abolished the NGF-driven hepatoprotection against both DCA- and CDCA-induced hepatotoxicity (Fig. 8B, 8C). Meanwhile, the lysates collected from the cells with FXR silence and NGF stimulation were further subjected to western blotting detection (Fig. 8D, 8E). The densitometry analyses indicated that FXR knockdown significantly prevented the NGF-elevated LC3II/LC3I ratios and p62 protein expression (Fig. 8F, 8G), while it markedly augmented constitutive and NGF-stimulated cellular Rab7 levels (Fig. 8H). To confirm the involvement of FXR activity in the BA-induced hepatotoxicity, SK-Hep1 cells were exposed to DCA or CDCA in the presence of FXR agonist GW4064 or antagonist Z-GS. The cell viability data showed that FXR agonistic treatment partially reversed DCA- and CDCA-induced hepatocyte cell death, while FXR antagonistic treatment significantly potentiated both BA-evoked cytotoxicity in SK-Hep1 hepatocytes (Fig. 8I, 8J). These findings strongly suggest that NGF/FXR signaling axis is involved in cellular autophagy and plays a hepatoprotective role in the BA-exerted hepatotoxicity.

4. Discussion

The present study is the first to provide molecular link between hepatic NGF and FXR expression in parenchymal hepatocytes, particularly emphasizing the significance of NGF/FXR signaling axis in the setting of liver cholestasis. Our results indicate that hepatic NGF contents clearly showed positive correlation with FXR expression in human cholestatic livers obtained from patients with hepatolithiasis. To further confirm the regulatory role of NGF/FXR in auophagy in injured livers, we evidenced that intraperitoneal injection of NGF recombinant peptides significantly increased FXR expression and simultaneously modulated autophagy flux in BDL-injured mouse livers. Interestingly, not only FXR but also autophagy marker LC3 have positive relationship with hepatic NGF contents, strongly suggesting the hepatoprotective role of NGF/FXR axis through regulation of hepatic autophagy flux. Herein we propose that manipulation of NGF/FXR signaling axis could be a therapeutic target for cholestatic liver injury.
Autophagy is an evolutionarily conserved intracellular machinery, through which cells process aged or damaged organelles by delivering cytoplamsic substrates to lysosomes for degradation [37]. Autophagy not only orchestrates cellular homeostasis but also plays a cytoprotective role against various pathogenic insults, including chemoresistance during anti-tumor treatment [38, 39]. Aside from apoptosis, autophagy status is involved in endoplasmic reticulum stress-induced cell death of rat hepatocytes [40]. The regulatory loop between FXR and autophagy has been implicated in the pathogenesis of cholestatic disease. In this regard, hepatic autophagy deficiency leads to reduced expression of FXR [22] and conversely FXR gene depletion attenuates feeding-mediated inhibition of macroautophagy in the livers [21]. Moreover, BA-induced FXR activation is believed to enhance progression of liver regeneration [12-14], while SIRT1-activated FXR signaling pathway reportedly attenuates triptolide-induced hepatotoxicity in rats [41] and controls liver regeneration through regulation of BA metabolism via mTOR activation [31]. The present in vitro mechanistic study demonstrated that treatments with exogenous NGF, FXR agonistic agent, and FXR gene silence also gave rise to protection against BA-induced hepatotoxicity in cultured hepatocytes. Taken with our previous finding that the NGF-upregulated hepatic SIRT1 expression in hepatocytes is mechanistically involved in its hepatoprotective effects [28], our findings support that NGF/SIRT1/FXR signaling axis may be key to autophagy regulation and its hepatoprotective effect in cholestasis.
Similar to that seen in human NHC livers, mild fatty morphological change was rarely noted in limited percentage of human hepatolithiasis tissues (around 20%, 1 out of 5 cases), which may in part reflect the notion that balloon degeneration of hepatocytes is one of histological features frequently seen in fibrosing cholestatic hepatitis [42]. Due to limited case number of this study and unavailability of normal human tissues for better control, we cannot rule out the possibility that the fibrosing state of NHC livers might not be representative for the normal liver control. However, the lower and less nuclear localization of FXR in human lithiasis liver than contralateral region could be explained by its regulatory role in regulation of lipid metabolism, which was indirectly supported by the facts that coordinative activation of BA receptor FXR/TGR5 alleviates non-alcoholic fatty liver development [15] and ameliorates steatosis and inflammation in alcoholic steatohepatitis mice [43]. Similar evidence in a rat model of alcoholic liver disease supports that drug-restored FXR expression and activity alone may mitigates hepatocyte steatosis and alcoholic liver injury [44]. This study evidenced the positive correlation of hepatic NGF with FXR, but not TGR5 expression in human cholestatic livers, suggesting that FXR might play a predominant role in the hepatoprotection mechanism in the progression of cholestasis.
In this study, we tested two different types of BAs (i.e., DCA and CDCA) and intriguingly noted that they exhibited differential effects on Rab7 levels in the hepatocytes under low-dose exposure condition. CDCA, as a primary BA originated in the liver, did not suppress Rab7 at doses equal to or lower than 200 µM. By contrast, the DCA, as a secondary BA that is formed in large intestine by enteric bacterial enzymes, passively absorbed from colon, and returned to livers via portal vein, apparently has the ability to downregulate Rab7 expression in the hepatocytes with low-dose treatment. The differential effects of CDCA and DCA on Rab7 expression in hepatocytes supports that secondary BAs generate higher geno- and cytotoxicity than primary BAs, which involves auphagic stress-survival pathway and eventually leads to carcinogenesis [4, 45]. The higher cytotoxicity of secondary BA could be reflected by the fact that NGF was able to increase LC3II/I ratios in the CDCA-treated cells, but not in those with DCA treatment. In addition, the BA-associated Rab7 downregulation is consistent with the reduced Rab7 levels seen in human lithiasis and mouse BDL livers. In fact, the small guanosine triphosphatase Rab7 has been implicated in late endocytic pathway [46] and is known to regulate lipid droplets in hepatocytes [47, 48]. Rab7 activation is essential for trafficking of both multivesicular bodies and lysosomes to lipid droplet surface during lipophagy, while experimental Rab7 inactivation results in impaired fusion of autophagy machinery with lipid droplets, leading to hepatocellular lipid accumulation and hepatic steatosis [49]. The BA-suppressed hepatic Rab7 expression in cholestatic liver might contribute to the fatty morphology change in human hepatolithiasis livers.
In the mechanistic context of NGF-induced hepatoprotection, this study evidenced that anti-NGF neutralization treatment dramatically increased hepatic p62/SQSTM1 levels in BDL-treated mouse liver, whereas FXR antagonistic treatment coincidentally elevated p62 expression in DCA-treated hepatocytes, implicating that p62 might play an important role in the NGF-elicited hepatoprotection. Despite no prominent response of cellular p62 and Rab7 levels found in the FXR agonist-treated cells, p62 sequestosome protein has been regarded an oxidative stress-inducible protein regulated by redox-sensitive transcription factor Nrf2 [50] as well as FXR [51], and also an autophagy substrate colocalized with LC3 and degraded by autophagy machinery and ubiquitin proteasome system [52]. In addition, previous studies on neuron cells demonstrated that NGF stimulated colocalization of p62 and Rab7 is essential in neurite outgrowth and differentiation of PC12 cells [53] and the recruitment of p62 adapter protein to p75 and TrkA neurotrophin receptors plays a pivotal role in NGF-driven NF-κB activation and neuron survival [54, 55]. Moreover, NGF modulates energy homeostasis and mitochondrial remodeling during neuron differentiation [56]. Whether hepatic NGF expression also contributes to hepatoprotection via p62/SQSTM1-involved mechanism is rarely addressed. It is noteworthy that FXR agonistic agent, GW4064, simultaneously activates Nrf2-associated anti-oxidants and expression of anti-apoptotic molecules, while FXR-p62/SQSTM1 pathway has been recently demonstrated to improve survival of mouse hepatocytes with steatosis [57]. The present study provides the first evidence revealing that hepatic NGF might participate in regulation of both hepatic autophagy and hepatoprotection via FXR-p62/SQSTM1 pathway (Fig. 8K). We reasonably propose that NGF/FXR signaling blockade might interrupt autophagy flux and lead to intracellular p62 accumulation and the NGF-sustained p62 expression in the hepatocytes under high-dose BA exposure might contribute to its hepatoprotective mechanism. However, the detailed molecular mechanism remains further investigation.
In conclusion, this study demonstrates that NGF upregulates FXR expression in parenchymal hepatocytes and exerts hepatoprotection against BA injury. NGF/FXR signal activation may alleviate cell death of hepatocytes under BA exposure through modulation of autophagy flux. These findings strongly suggest that NGF/FXR axis may be a potential therapeutic target for treating cholestatic liver disease.

References

[1] A. Kosters, S.J. Karpen, The role of inflammation in cholestasis: clinical and basic aspects, Semin Liver Dis 30 (2) (2010) 186-194.
[2] C. Degirolamo, S. Modica, G. Palasciano, A. Moschetta, Bile acids and colon cancer: Solving the puzzle with nuclear receptors, Trends Mol Med 17 (10) (2011) 564-572. [3] X. Wang, X. Fu, C. Van Ness, Z. Meng, X. Ma, W. Huang, Bile acid receptors and liver cancer, Curr Pathobiol Rep 1 (1) (2013) 29-35.
[4] C. Bernstein, H. Holubec, A.K. Bhattacharyya, H. Nguyen, C.M. Payne, B. Zaitlin, H. Bernstein, Carcinogenicity of deoxycholate, a secondary bile acid, Arch Toxicol 85 (8) (2011) 863-871.
[5] G. Xie, X. Wang, F. Huang, A. Zhao, W. Chen, J. Yan, Y. Zhang, S. Lei, K. Ge, X. Zheng, J. Liu, M. Su, P. Liu, W. Jia, Dysregulated hepatic bile acids collaboratively promote liver carcinogenesis, Int J Cancer 139 (8) (2016) 1764-1775.
[6] G.M. Hirschfield, E.J. Heathcote, M.E. Gershwin, Pathogenesis of cholestatic liver disease and therapeutic approaches, Gastroenterology 139 (5) (2010) 1481-1496.
[7] A.P. Rolo, C.M. Palmeira, J.M. Holy, K.B. Wallace, Role of mitochondrial dysfunction in combined bile acid-induced cytotoxicity: the switch between apoptosis and necrosis, Toxicol Sci 79 (1) (2004) 196-204.
[8] G. Zhang, M.A. Park, C. Mitchell, T. Walker, H. Hamed, E. Studer, M. Graf, M. Rahmani, S. Gupta, P.B. Hylemon, P.B. Fisher, S. Grant, P. Dent, Multiple cyclin kinase inhibitors promote bile acid-induced apoptosis and autophagy in primary hepatocytes via p53-CD95-dependent signaling, J Biol Chem 283 (36) (2008) 24343-24358.
[9] P.B. Hylemon, H. Zhou, W.M. Pandak, S. Ren, G. Gil, P. Dent, Bile acids as GW4064 regulatory molecules, J Lipid Res 50 (8) (2009) 1509-1520.
[10] B.M. Forman, E. Goode, J. Chen, A.E. Oro, D.J. Bradley, T. Perlmann, D.J. Noonan, L.T. Burka, T. McMorris, W.W. Lamph, R.M. Evans, C. Weinberger, Identification of a nuclear receptor that is activated by farnesol metabolites, Cell 81 (5) (1995) 687-693.
[11] Y. Kawamata, R. Fujii, M. Hosoya, M. Harada, H. Yoshida, M. Miwa, S. Fukusumi, Y. Habata, T. Itoh, Y. Shintani, S. Hinuma, Y. Fujisawa, M. Fujino, A G protein-coupled receptor responsive to bile acids, J Biol Chem 278 (11) (2003) 9435-9440.
[12] M. Fan, X. Wang, G. Xu, Q. Yan, W. Huang, Bile acid signaling and liver regeneration, Biochim Biophys Acta 1849 (2) (2014) 196-200.
[13] W. Huang, K. Ma, J. Zhang, M. Qatanani, J. Cuvillier, J. Liu, B. Dong, X. Huang, D.D. Moore, Nuclear receptor-dependent bile acid signaling is required for normal liver regeneration, Science 312 (5771) (2006) 233-236.
[14] L. Ding, Y. Yang, Y. Qu, T. Yang, K. Wang, W. Liu, W. Xia, Bile acid promotes liver regeneration via farnesoid X receptor signaling pathways in rats, Mol Med Rep 11 (6) (2015) 4431-4437.
[15] R.H. McMahan, X.X. Wang, L.L. Cheng, T. Krisko, M. Smith, K. El Kasmi, M. Pruzanski, L. Adorini, L. Golden-Mason, M. Levi, H.R. Rosen, Bile acid receptor activation modulates hepatic monocyte activity and improves nonalcoholic fatty liver disease, J Biol Chem 288 (17) (2013) 11761-11770.
[16] N. Pean, I. Doignon, T. Tordjmann, Bile acids and liver carcinogenesis: TGR5 as a novel piece in the puzzle?, Clin Res Hepatol Gastroenterol 37 (3) (2013) 226-229.
[17] Y. Xiao, K. Zhou, Y. Lu, W. Yan, W. Cai, Y. Wang, Administration of antibiotics contributes to cholestasis in pediatric patients with intestinal failure via the alteration of FXR signaling, Exp Mol Med 50 (12) (2018) 155.
[18] M. Cariello, E. Piccinin, O. Garcia-Irigoyen, C. Sabba, A. Moschetta, Nuclear receptor FXR, bile acids and liver damage: Introducing the progressive familial intrahepatic cholestasis with FXR mutations, Biochim Biophys Acta Mol Basis Dis 1864 (4 Pt B) (2018) 1308-1318.
[19] N. Gomez-Ospina, C.J. Potter, R. Xiao, K. Manickam, M.S. Kim, K.H. Kim, B.L. Shneider, J.L. Picarsic, T.A. Jacobson, J. Zhang, W. He, P. Liu, A.S. Knisely, M.J. Finegold, D.M. Muzny, E. Boerwinkle, J.R. Lupski, S.E. Plon, R.A. Gibbs, C.M. Eng, Y. Yang, G.C. Washington, M.H. Porteus, W.E. Berquist, N. Kambham, R.J. Singh, F. Xia, G.M. Enns, D.D. Moore, Mutations in the nuclear bile acid receptor FXR cause progressive familial intrahepatic cholestasis, Nat Commun 7 (2016) 10713.
[20] S. Anakk, M. Watanabe, S.A. Ochsner, N.J. McKenna, M.J. Finegold, D.D. Moore, Combined deletion of Fxr and Shp in mice induces Cyp17a1 and results in juvenile onset cholestasis, J Clin Invest 121 (1) (2011) 86-95.
[21] S. Seok, T. Fu, S.E. Choi, Y. Li, R. Zhu, S. Kumar, X. Sun, G. Yoon, Y. Kang, W. Zhong, J. Ma, B. Kemper, J.K. Kemper, Transcriptional regulation of autophagy by an FXR-CREB axis, Nature 516 (7529) (2014) 108-111.
[22] B. Khambu, T. Li, S. Yan, C. Yu, X. Chen, M. Goheen, Y. Li, J. Lin, O.W. Cummings, Y.A. Lee, S. Friedman, Z. Dong, G.S. Feng, S. Wu, X.M. Yin, Hepatic autophagy deficiency compromises FXR functionality and causes cholestatic injury, Hepatology (2018).
[23] J.M. Lee, Transcriptional coordination of hepatic autophagy by nutrient-sensing nuclear receptor PPARalpha and FXR, Ann Pediatr Endocrinol Metab 21 (4) (2016) 193-198. [24] A. Baghdasaryan, T. Claudel, J. Gumhold, D. Silbert, L. Adorini, A. Roda, S. Vecchiotti, F.J. Gonzalez, K. Schoonjans, M. Strazzabosco, P. Fickert, M. Trauner, Dual farnesoid X receptor/TGR5 agonist INT-767 reduces liver injury in the Mdr2-/- (Abcb4-/-) mouse cholangiopathy model by promoting biliary HCO(-)(3) output, Hepatology 54 (4) (2011) 1303-1312.
[25] G. Rasi, A. Serafino, L. Bellis, M.T. Lonardo, F. Andreola, M. Zonfrillo, G. Vennarecci, P. Pierimarchi, P. Sinibaldi Vallebona, G.M. Ettorre, E. Santoro, C. Puoti, Nerve growth factor involvement in liver cirrhosis and hepatocellular carcinoma, World J Gastroenterol 13 (37) (2007) 4986-4995.
[26] Y.H. Kao, P.H. Lee, T.C. Chiu, Y.C. Lin, C.K. Sun, P.H. Chen, M.S. Tsai, Transcriptome analysis reveals a positive role for nerve growth factor in retinol metabolism in primary rat hepatocytes, Cytokine 107 (2018) 74-78.
[27] M.S. Tsai, Y.C. Lin, C.K. Sun, S.C. Huang, P.H. Lee, Y.H. Kao, Up-regulation of nerve growth factor in cholestatic livers and its hepatoprotective role against oxidative stress, PLoS One 9 (11) (2014) e112113.
[28] M.S. Tsai, P.H. Lee, C.K. Sun, T.C. Chiu, Y.C. Lin, I.W. Chang, P.H. Chen, Y.H. Kao, Nerve growth factor upregulates sirtuin 1 expression in cholestasis: a potential therapeutic target, Exp Mol Med 50 (1) (2018) e426.
[29] S.Y. Wu, S.C. Cui, L. Wang, Y.T. Zhang, X.X. Yan, H.L. Lu, G.Z. Xing, J. Ren, L.K. Gong, 18beta-Glycyrrhetinic acid protects against alpha-naphthylisothiocyanate-induced cholestasis through activation of the Sirt1/FXR signaling pathway, Acta Pharmacol Sin 39 (12) (2018) 1865-1873.
[30] L. Ding, B. Zhang, J. Li, L. Yang, Z. Wang, Beneficial effect of resveratrol on alphanaphthyl isothiocyanateinduced cholestasis via regulation of the FXR pathway, Mol Med Rep 17 (1) (2018) 1863-1872.
[31] J.L. Garcia-Rodriguez, L. Barbier-Torres, S. Fernandez-Alvarez, V. Gutierrez-de Juan, M.J. Monte, E. Halilbasic, D. Herranz, L. Alvarez, P. Aspichueta, J.J. Marin, M. Trauner, J.M. Mato, M. Serrano, N. Beraza, M.L. Martinez-Chantar, SIRT1 controls liver regeneration by regulating bile acid metabolism through farnesoid X receptor and mammalian target of rapamycin signaling, Hepatology 59 (5) (2014) 1972-1983.
[32] Y.H. Kao, Y.C. Lin, M.S. Tsai, C.K. Sun, S.S. Yuan, C.Y. Chang, B. Jawan, P.H. Lee, Involvement of the nuclear high mobility group B1 peptides released from injured hepatocytes in murine hepatic fibrogenesis, Biochim Biophys Acta 1842 (9) (2014) 1720-1732.
[33] G.G. Pinkse, M.P. Voorhoeve, M. Noteborn, O.T. Terpstra, J.A. Bruijn, E. De Heer, Hepatocyte survival depends on beta1-integrin-mediated attachment of hepatocytes to hepatic extracellular matrix, Liver Int 24 (3) (2004) 218-226.
[34] Y.C. Chang, M.C. Hsieh, H.J. Wu, W.C. Wu, Y.H. Kao, Methylglyoxal, a reactive glucose metabolite, enhances autophagy flux and suppresses proliferation of human retinal pigment epithelial ARPE-19 cells, Toxicol In Vitro 29 (7) (2015) 1358-1368.
[35] C.C. Hsiao, P.H. Chen, C.I. Cheng, M.S. Tsai, C.Y. Chang, S.C. Lu, M.C. Hsieh, Y.C. Lin, P.H. Lee, Y.H. Kao, Toll-like receptor-4 is a target for suppression of proliferation and chemoresistance in HepG2 hepatoblastoma cells, Cancer Lett 368 (1) (2015) 144-152. [36] S. Manley, H.M. Ni, B. Kong, U. Apte, G. Guo, W.X. Ding, Suppression of autophagic flux by bile acids in hepatocytes, Toxicol Sci 137 (2) (2014) 478-490.
[37] N. Mizushima, M. Komatsu, Autophagy: renovation of cells and tissues, Cell 147 (4) (2011) 728-741.
[38] J. Song, Z. Qu, X. Guo, Q. Zhao, X. Zhao, L. Gao, K. Sun, F. Shen, M. Wu, L. Wei, Hypoxia-induced autophagy contributes to the chemoresistance of hepatocellular carcinoma cells, Autophagy 5 (8) (2009) 1131-1144.
[39] Y. Xu, Y. An, Y. Wang, C. Zhang, H. Zhang, C. Huang, H. Jiang, X. Wang, X. Li, miR-101 inhibits autophagy and enhances cisplatin-induced apoptosis in hepatocellular carcinoma cells, Oncol Rep 29 (5) (2013) 2019-2024.
[40] J. Zhang, M.W. Morris, Jr., W.A. Dorsett-Martin, L.C. Drake, C.D. Anderson, Autophagy is involved in endoplasmic reticulum stress-induced cell death of rat hepatocytes, J Surg Res 183 (2) (2013) 929-935.
[41] J. Yang, L. Sun, L. Wang, H.M. Hassan, X. Wang, P.B. Hylemon, T. Wang, H. Zhou, L. Zhang, Z. Jiang, Activation of Sirt1/FXR signaling pathway attenuates triptolide-induced hepatotoxicity in rats, Front Pharmacol 8 (2017) 260.
[42] C. Lackner, M. Gogg-Kamerer, K. Zatloukal, C. Stumptner, E.M. Brunt, H. Denk, Ballooned hepatocytes in steatohepatitis: the value of keratin immunohistochemistry for diagnosis, J Hepatol 48 (5) (2008) 821-828.
[43] A. Iracheta-Vellve, C.D. Calenda, J. Petrasek, A. Ambade, K. Kodys, L. Adorini, G. Szabo, FXR and TGR5 agonists ameliorate liver injury, steatosis, and inflammation after binge or prolonged alcohol feeding in mice, Hepatol Commun 2 (11) (2018) 1379-1391.
[44] W. Xu, C. Lu, L. Yao, F. Zhang, J. Shao, S. Zheng, Dihydroartemisinin protects against alcoholic liver injury through alleviating hepatocyte steatosis in a farnesoid X receptor-dependent manner, Toxicol Appl Pharmacol 315 (2017) 23-34.
[45] C.M. Payne, C. Crowley-Skillicorn, H. Holubec, K. Dvorak, C. Bernstein, M.P. Moyer, H. Garewal, H. Bernstein, Deoxycholate, an endogenous cytotoxin/genotoxin, induces the autophagic stress-survival pathway: implications for colon carcinogenesis, J Toxicol 2009 (2009) 785907.
[46] T. Wang, Z. Ming, W. Xiaochun, W. Hong, Rab7: role of its protein interaction cascades in endo-lysosomal traffic, Cell Signal 23 (3) (2011) 516-521.
[47] B. Schroeder, R.J. Schulze, S.G. Weller, A.C. Sletten, C.A. Casey, M.A. McNiven, The small GTPase Rab7 as a central regulator of hepatocellular lipophagy, Hepatology 61 (6) (2015) 1896-1907.
[48] F. Guerra, C. Bucci, Multiple roles of the small GTPase Rab7, Cells 5 (3) (2016).
[49] R.J. Schulze, K. Rasineni, S.G. Weller, M.B. Schott, B. Schroeder, C.A. Casey, M.A. McNiven, Ethanol exposure inhibits hepatocyte lipophagy by inactivating the small guanosine triphosphatase Rab7, Hepatol Commun 1 (2) (2017) 140-152.
[50] T. Ishii, E. Warabi, R.C.M. Siow, G.E. Mann, Sequestosome1/p62: a regulator of redox-sensitive voltage-activated potassium channels, arterial remodeling, inflammation, and neurite outgrowth, Free Radic Biol Med 65 (2013) 102-116.
[51] J.A. Williams, A.M. Thomas, G. Li, B. Kong, L. Zhan, Y. Inaba, W. Xie, W.X. Ding, G.L. Guo, Tissue specific induction of p62/Sqstm1 by farnesoid X receptor, PLoS One 7 (8) (2012) e43961.
[52] S. Manley, J.A. Williams, W.X. Ding, Role of p62/SQSTM1 in liver physiology and pathogenesis, Exp Biol Med (Maywood) 238 (5) (2013) 525-538.
[53] I.S. Samuels, M.L. Seibenhener, K.B. Neidigh, M.W. Wooten, Nerve growth factor stimulates the interaction of ZIP/p62 with atypical protein kinase C and targets endosomal localization: evidence for regulation of nerve growth factor-induced differentiation, J Cell Biochem 82 (3) (2001) 452-466.
[54] V. Mamidipudi, X. Li, M.W. Wooten, Identification of interleukin 1 receptor-associated kinase as a conserved component in the p75-neurotrophin receptor activation of nuclear factor-kappa B, J Biol Chem 277 (31) (2002) 28010-28018.
[55] M.W. Wooten, M.L. Seibenhener, V. Mamidipudi, M.T. Diaz-Meco, P.A. Barker, J. Moscat, The atypical protein kinase C-interacting protein p62 is a scaffold for NF-kappaB activation by nerve growth factor, J Biol Chem 276 (11) (2001) 7709-7712.
[56] F. Martorana, D. Gaglio, M.R. Bianco, F. Aprea, A. Virtuoso, M. Bonanomi, L. Alberghina, M. Papa, A.M. Colangelo, Differentiation by nerve growth factor (NGF) involves mechanisms of crosstalk between energy homeostasis and mitochondrial remodeling, Cell Death Dis 9 (3) (2018) 391.
[57] S. Haga, Yimin, M. Ozaki, Relevance of FXR-p62/SQSTM1 pathway for survival and protection of mouse hepatocytes and liver, especially with steatosis, BMC Gastroenterol 17 (1) (2017) 9.