Article Text
Abstract
Background: Endogenous opioids modulate the growth of nervous and non-nervous cells. Hepatic stellate cells (HSCs) are the main cell phenotype involved in liver fibrogenesis, display molecular markers of neuronal cells and respond to neurotransmitters.
Aim: To evaluate the role of endogenous opioids on liver fibrogenesis.
Methods: Activated rat HSCs (passage 1–3) were used to evaluate cell proliferation and intracellular signalling pathway activation. Liver fibrosis was induced in rats by dimethylnitrosamine (DMN) administration.
Results: Opioid receptors showed a different pattern of expression when measured in quiescent and activated (in vitro and in vivo) HSCs. The activation of opioid receptors increased HSC proliferation and collagen accumulation. Opioid receptor stimulation induced a calcium-dependent protein kinase Cα (PKCα)/extracellular regulated kinase (ERK)/phosphatidylinositol 3-kinase (PI3K) pathway activation that mediated the effect of endogenous opioids on HSC proliferation and collagen synthesis. In DMN-treated rats, the opioid antagonist naloxone reduced α-smooth muscle actin expression (as a marker of HSC activation) and collagen deposition, both measured by morphometry after 5 weeks of treatment. In both DMN-treated rats and human liver biopsies from chronic liver diseases, opioid receptors were observed in HSCs in area of active fibrogenesis. The endogenous opioid met-enkephalin increased its expression in zone 3 hepatocytes close to the area of necrosis after DMN administration and in the cellular target of chronic liver injury in human biopsies, and stimulated HSC proliferation and collagen synthesis.
Conclusions: Endogenous opioids released during chronic liver injury participate in the process of liver fibrogenesis by stimulating HSC proliferation and collagen production in a paracrine manner.
Statistics from Altmetric.com
Hepatic cirrhosis is considered a common response to chronic liver injury, regardless of the aetiological agents (virus, toxic or metabolic agents, or autoimmune mechanisms).1 Chronic liver diseases are characterised by the co-existence of parenchymal (hepatocytes) or non-parenchymal (ie, cholangiocytes) injury (ie, necrosis or apoptosis) that represents the basic mechanism leading to activation of hepatic stellate cells (HSCs).1 In response to chronic liver injury, HSCs undergo a process of activation, acquiring a myofibroblast-like phenotype. Such a process is characterised by increased proliferation close to the areas of necrosis, excessive synthesis of extracellular matrix components and contractile behaviour due to α-smooth muscle actin (αSMA) expression.2 In experimental conditions and in human liver diseases, liver fibrosis develops from areas of hepatocellular necrosis and inflammation, suggesting that paracrine signals derived from damaged hepatocytes or recruited inflammatory cells affect HSC activation.3–7
The finding that HSCs express neuroendocrine features during fibrosis development led researchers to investigate the role of the central nervous system in modulating liver fibrogenesis.8–10 In this regard, it has been shown that the adrenergic autonomic nervous system is necessary for the proliferative and functional response to parenchymal liver injury, whereas nerve growth factor (NGF) induces HSC apoptosis.11–14 A recent observation correlates the role of opioid receptors (ORs) in growth regulation not only in neuronal but also in non-neuronal cells and tissues.15 Endogenous opioids (EOs) bind to ORs belonging to the G-protein-coupled receptor (GPCR) family: δORs, μORs and κORs.16 It has been shown that during chronic cholestatic conditions, the opioidergic neurotransmission, together with plasma and hepatic levels of EOs, is markedly increased and regulates the cell response to cholestatic liver injury.17–22 In addition, EO plasma levels are increased in patients with acute and chronic liver diseases, and their levels correlate with the degree of liver injury.17–24
Therefore, this study aimed to answer the following questions. (1) Do HSCs express δORs, μORs and κORs in the different phases of their activation process? (2) Does the activation of these receptors affect HSC proliferation and collagen production? (3) Which are the intracellular pathways that mediate the OR signal in HSCs? (4) Could EOs participate in the neuroendocrine peptide loop that regulates the mechanisms of wound healing during non-cholestatic chronic liver injury?
METHODS
In vitro study
HSC isolation and culture
Rat and human HSCs were isolated by the pronase and collagenase method and then subcultured as previously described.6 25
Expression of ORs in HSCs
The expression of ORs was evaluated by immunoblots and reverse transcription-PCR (RT-PCR) in freshly isolated (quiescent) and activated HSCs, and by immunohistochemistry in liver sections, as described below. Rat brain and normal human colon (in excess for histological purposes after surgical resection), were used as positive controls, for western blotting and RT-PCR
Experimental design
To evaluate the OR-mediated effect on HSC proliferation and type I collagen accumulation, activated HSCs were cultured for 48 h in serum- and insulin-free medium (SIFM) and then incubated for the indicated period of time with: SIFM (control), (d-pen2,5)-enkephalin (DPDPE, 0.1 μM, a δOR selective agonist), (d-Ala2, N-Me-Phe4, Gly5-ol)-enkephalin (DAMGO, 0.1 μM, a μOR selective agonist), dynorphin A (0.1 μM, a κOR selective agonist), platelet-derived growth factor (PDGF)-BB (25 ng/ml) or transforming growth factor β (TGFβ; 1 ng/ml).26–28 In some experiments, the three EOs were also used simultaneously to stimulate HSCs. In parallel experiments, HSCs were incubated with the non-specific OR agonist met-enkephalin (100 μM) in the presence or absence of ICI 174,864 (a δOR antagonist, 300 nM), CTAP (a μOR antagonist, 10 nM) and nor-binaltorphimine (a κOR antagonist, 10 nM).29–32 To demonstrate that the HSC response, observed after OR activation, is mediated by mitogen-activated protein kinase (MAPK) and phosphatidylinositol 3-kinase (PI3K) pathways, the experiments were also performed by preincubating cells for 30 min at 37°C with either PD98059 (50 μM, a MAPK kinase (MEK) inhibitor) or wortmannin (100 nM, a PI3K inhibitor), followed by the incubation with OR agonists.25 Furthermore, to investigate OR transduction through the activation of Ca2+ signalling, cells were also pre-incubated with BAPTA/AM (1,2-bis(o-aminophenoxyethane)-N,N,N',N'-tetraacetic acid acetoxymethyl ester, an intracellular Ca2+ chelator, 5 μM), followed by the incubation with OR agonists.33 Finally, to confirm further the direct agonist–receptor interaction, we evaluated HSC proliferation and type I collagen accumulation in cells preincubated with naloxone (1 μM, a general OR antagonist) for 30 min and then stimulated with the indicated agonists.34
Determination of HSC proliferation
Activated HSCs (passage 1–3) were incubated in SIFM for 48 h. After this time, the medium was removed and the cells were incubated in SIFM alone, or containing the indicated OR agonists and/or 10% serum for up to 24 h. HSC proliferation was determined as previously described both by immunoblots for proliferating cell nuclear antigen (PCNA) protein expression and by the measurement of bromodeoxyuridine (BrdU) incorporation.35 Evaluation of PCNA protein expression was performed by normalising the intensity of the bands of immunoblots for β-actin. BrDU incorporation was measured employing the Cell Proliferation ELISA BrdU assay according to the instructions provided by the vendor (Roche, Monza, Italy).
Western blot
Cell extracts were obtained as previously described.28 Nitrocellulose sheets were then incubated with antibodies against pERK (1:5000 final dilution), pAkt (1:1000), pJNK (1:4000), PCNA (1:1000), β-actin (1:5000), δOR (Santa Cruz Biotechnology Inc., sc-9111), μOR (sc-15310) and κOR (sc-9112) (1:1000). The antigen–antibody complexes were visualised using the ECL (enhanced chemiluminescence) detection system as recommended by the manufacturer (Amersham, Arlington Heights, IL). The intensity of the bands was determined by scanning video densitometry using the ChemiIimager 4000 low-light imaging system (Alpha Innotech, San Leandro, CA).
Reverse transcription-PCR
Total RNA was extracted from freshly isolated and cultured HSCs, from frozen brain of male Sprague–Dawley rats and from endoscopic biopsies from human colon in excess for diagnostic purposes using TRIzol Reagent (Invitrogen, Carlsbad, CA). A 2 μg aliquot of RNA was reverse-transcribed in a 25 μl reaction volume by using 200 U of Moloney murine leukaemia virus reverse transcriptase (Promega Corporation, Madison, WI) and 1 μl of random hexamer primer (p(dN)6), at 42°C for 60 min. Reverse transcription was stopped by a denaturing enzyme at 72°C. To measure the efficiency of reverse transcription, 1 μl of the same reverse transcriptase reaction was amplified for rat β-actin-specific primers and for human glyceraldehyde phosphate dehydrogenase (GADPH) primers (26 cycles of amplification).
PCR for rat and human ORs was performed with the following primers, designed on published sequences: rat μOR (NM_013071) sense 5′-GCC CTC TAC TCT ATC GTG TGT-3′, antisense 5′-GCA ATG TAG CGG TCC AC-3′, PCR product size 287 bp; rat δ1OR receptor (NM_012617), sense 5′-GCT CCT CGC CAA CGT CTC-3′, antisense 5′-CCG TCT TCA GCT TAG TGT ACC G-3′, PCR product size 206 bp; rat κOR (NM_017167), sense 5′-GTG TGG GCT CCG AGG AC-3′, antisense 5′-TCT GGA AGG GCA TAG TGG TA-3′, PCR product size 225 bp; human μOR (NM_001008504), sense 5′-TTG GCG TAC TCA AGT TGCTC-3′, antisense 5′-GGA AGG GCA GGGTAC TGG T-3′, PCR product size 328 bp; human δOR (NM_000911), sense 5′-GAT GAG TGT TGA CCG CTA CA-3′, antisense 5′-ATG GGC ACC ACG AAG G-3′, PCR product size 255 bp; human κOR (NM_000912), sense 5′-TCA TCA TCA CGG CGG TCT-3′, antisense 5′-CAC GGC AAT GTA GCGGTC CA-3′, PCR product size 302 bp.
Hot Start PCRs were performed with 2 μl of cDNA in a total reaction volume of 25 μl, containing 2.5 U of Taq (GoTaq Flexi DNA Polymerase, Promega Corporation), 5 μl of 5× MgCl2-free buffer, 0.2 μg of sense and antisense primers, 0.2 mmol/l dNTPs and 1, 5 or 3 mM MgCl2. For amplification of ORs, all reactions were carried out for 40 cycles, consisting of 30 s at 95°C, 50 s at 60°C (55°C for rat μOR, and 58°C for human μOR and κOR), and 50 s at 72°C. The amplified PCR products were subjected to electrophoresis through a 1.5% agarose gel and then stained with 0.5 mg/ml ethidium bromide TAE buffer.
Collagen assay
Type I collagen concentration was measured in HSC culture medium by ELISA as previously described.6 Results are expressed as micrograms of collagen per milligram of total protein.
In vivo study
Induction of acute and chronic liver injury
Fifty-five male Sprague–Dawley rats (140–150 g body weight) were used for the in vivo study. Hepatic injury was induced by intraperitoneal injections of 10 mg/kg dimethylnitrosamine (DMN) dissolved in saline to obtain a 1% solution, administered on three consecutive days a week up to a total of 5 weeks.2 Control animals received saline. To verify the in vivo effects of EOs on liver fibrosis, treatment with naloxone (0.4 mg/kg/day) was begun simultaneously with DMN administration.36 Animals were sacrificed after 1 (n = 5 for each treatment group) and 5 (n = 10 for each treatment group) weeks of treatment. To evaluate met-enkepahlin and αSMA (as a marker of activated rat HSCs28) expression in the early stages of hepatic fibrogenesis, rats were also sacrificed 24 and 48 h after the third DMN injection (n = 3 for each time point). In parallel experiments, rats were also sacrificed after 1 week of DMN treatment for HSC isolation (n = 2).37 At the time of sacrifice, portal blood was obtained for routine alanine transaminase (ALT) determination. To determine the degree of necroinflammatory liver injury, histological grading was performed blind by an independent pathologist, as previously described.5 28 38
Immunohistochemistry
OR and met-enkephalin expression and localisation were visualised in formalin-fixed paraffin-embedded liver sections from DMN-treated rats, and from 12 human hepatic biopsies. Liver biopsy specimens were randomly selected from the histological archives of the Department of Gastroenterology of Ancona, from a group of patients that underwent liver biopsy for diagnostic purposes: three subjects without liver injury at biopsy, three patients with primary biliary cirrhosis (PBC), three patients with chronic hepatitis C (HCV) and three patients with chronic alcoholic liver injury. For OR visualisation, liver sections were treated with 0.05% pronase in Tris-buffered saline at 37°C for 10 min. Liver sections were then incubated with the primary antibodies according to the manufacturer’s instructions (1:50 dilution overnight at 4°C). After incubation with the appropriate peroxidase-conjugated secondary antibody, the reaction was visualised by Tris-buffered saline containing 0.06% diaminobenzidine and 0.01% H2O2.25
For the simultaneous detection of αSMA, as a marker of activated rat HSCs, and of PCNA expression, as a marker of S-phase nuclei in methanol-fixed sections, a sequential double immunoenzymatic reaction was performed as previously described.28
For the simultaneous detection of glial fibrillary acidic protein (GFAP, as a general marker of rat HSCs) or αSMA (as a general marker of human HSCs) and OR, double immunostaining was performed on cryostat sections as previously described.9 10
Met-enkephalin expression and localization in formalin-fixed paraffin-embedded liver sections was determined using a Met-Enkephalin-Immunohistochemistry Staining Kit from Peninsula Laboratories Inc. (San Carlos, CA).
Morphometric determinations
Quantitative analysis of αSMA/PCNA-positive HSCs was performed by two independent observers at 100× final magnification using a computerised image analysis system connected to an Olympus microscope (Olympus Vanox AHBT3, Olympus Optical Co. Ltd, Tokyo, Japan) as previously described.2 Quantitative analysis of fibrotic tissue deposition was performed on Sirius Red-stained sections by two independent observers at 10× final magnification using the computerised image analysis system indicated above.2
Hydroxyproline determination
Hydroxyproline content in liver samples was determined as previously described.39
Statistical analysis
Results are expressed as mean (SD). Group means were compared by analysis of variance (ANOVA) followed by the Student–Newman–Keuls test if the former was significant. A p value of <0.05 was considered statistically significant.
RESULTS
In vitro study
Heterogenic expression of the ORs in quiescent and activated HSCs
In control liver, a positive immunostaining for δOR was observed in the wall of the centrilobular vein, while µOR was not detected (fig 1A,C). In contrast, after DMN administration for 1 week, a positive immunostaining for both receptors was observed in stellate-shaped perisinusoidal cells, close to a necrotic area and therefore considered to be HSCs (fig 1B,D).10 Only weak reactivity to the κOR antibody became apparent in stellate-shaped perisinusoidal cells at dilutions lower than 1:20, and thus this approach was not used firther in the present study (data not shown).
To confirm OR expression in rat HSCs, four different experimental methods were additionally performed. First, double immunostaining for GFAP and ORs showed co-localization of these antigens in HSCs (fig 1E–G and data not shown). Secondly, to confirm that the reactive elements were HSCs, we isolated HSCs from control and DMN-treated animals.37 By western blot, a reduced protein expression of δOR after DMN treatment was observed, while both µOR and κOR expression were increased (fig 1H). Thirdly, mRNA expression for all three receptors was observed in both rat and human HSCs by RT-PCR (fig 2A). Fourthly, the in vitro model of HSC culture was used to confirm the different protein expression of ORs during the activation process. HSC activation was associated with reduced δOR, and increased µOR and κOR expression in comparison with quiescent HSCs (fig 2B–E). Finally, by western blot, all three ORs were detected in in vitro activated human HSCs (fig 2B).
Effect of EOs on HSC proliferation and type I collagen accumulation in the culture medium
A previous report from our group has shown that EOs downregulate cholangiocyte proliferation.20 To evaluate the potential effect on HSC proliferation, cells were incubated with OR agonists in the presence or absence of 10% serum. While serum-induced HSC proliferation was not modified by the agonists (data not shown), OR agonists were able to increase proliferation of HSCs (fig 3), as assessed by western blot for PCNA expression (fig 3A) and ELISA for BrdU incorporation in S-phase nuclei (fig 3B). Finally, HSC incubation with OR agonists induced a significant accumulation of type I collagen in the culture medium (fig 3C).
To confirm that EO effects were mediated by agonist–receptor interaction, HSCs were pretreated with the general OR antagonist naloxone before incubation with the specific OR agonists, showing reduced HSC proliferation (PCNA expression and BrdU incorporation) and type I collagen accumulation down to control levels (fig 3A–C). No effects were observed in HSCs in the presence of SIFM (fig 3A–C). The specificity of naloxone in blocking the EO–OR axis was confirmed by the lack of any effect on PDGF-induced HSC proliferation and TGFβ-induced type I collagen accumulation (fig 3A–C). No additive effects on HSC proliferation and type I collagen synthesis were observed when the three EOs were incubated simultaneously to stimulate HSCs (data not shown).
Characterisation and role of intracellular OR signalling
In recent years, it has been shown that several intracellular kinases regulate HSC behaviour, such as extracellular regulated kinase (ERK) 1/2, PI3K and Jun N-terminal kinase (JNK).37 Activation of all three ORs, using the selective agonists, did not produce any increase in JNK phosphorylation (fig 4A), but induced a significant and sustained (up to 60 min) increase in both ERK1/2 and Akt phosphorylation (fig 4B, C). In fig 5A,B it is shown that these pathways are independently activated by the three OR agonists, since wortmannin (a PI3K inhibitor) did not affect ERK phosphorylation, while PD98059 (a MEK inhibitor) did not affect Akt phosphorylation. Despite their independent activation, both of these pathways are needed to sustain the fibrogenetic effects of EOs, since both wortmannin and PD98059 inhibited HSC proliferation and collagen accumulation in the culture medium, induced by OR activation (Fig 5C,D).
The three different ORs belong to the GPCR superfamily.16 Since it is known that GPCRs can activate conventional protein kinase C (PKC) isoforms by inducing Ca2+ release in the cytoplasm, we evaluated the effect of the three EOs on the activation of the α isoform of PKC. All three OR agonists induced PKCα phosphorylation up to 60 min of incubation (fig 6A). As a confirmation of the relevance of Ca2+ signalling in transducing the OR message, we pretreated HSCs with the intracellular Ca2+ chelator BAPTA/AM. OR agonist-induced HSC proliferation (BrdU incorporation) and collagen production were neutralised by BAPTA/AM (fig 6B, C).
In vivo study
Hepatic met-enkephalin expression during rat liver injury
To verify eventual modification of EOs during the development of hepatic fibrosis, met-enkephalin expression was studied at different time points after DMN administration. In control liver, only a faint cytoplasmic immunostaining for met-enkephalin was observed (fig 4B). Already 24 h after DMN administration, strong expression was mostly evident as coarse “granuli” in the cytoplasm of hepatocytes, while after 1 week the immunostaining showed both a cytoplasmic and a pericanalicular localisation (fig 7C,D). At the same time points after DMN administration, αSMA (as a marker of activated HSCs in rats) immunostaining was performed in methanol-fixed paraffin-embedded liver sections. While in control livers αSMA-positive cells were not detected in the hepatic lobule (data not shown),37 40 these cells appeared as scattered and isolated elements 24–48 h after DMN administration (fig 7E and data not shown), while the formation of initial septa could be detected 1 week after DMN (fig 7F).
Role of OR blockage in HSC activation and proliferation, and liver fibrosis development in vivo
To confirm in vivo a role for the EO–OR axis in the development of hepatic fibrosis, rats were divided into two groups to receive DMN alone or in association with naloxone (0.4 mg/kg/day) for up to 5 weeks.36 To quantitate in vivo the process of HSC activation and proliferation, we evaluated by morphometry the number of cells exhibiting one or more αSMA-positive cytoplasmic process with a PCNA-positive nucleus, and the percentage of αSMA-positive parenchyma.28 In vivo administration of naloxone to DMN-treated rats reduced the process of HSC activation in vivo, as shown by the reduced number of αSMA/PCNA-positive cells and by the decreased percentage of αSMA-positive hepatic parenchyma (fig 8A,B). This resulted, at the end of the fifth week of treatment, in complete αSMA-positive septa delimiting parenchymal nodules in DMN-treated rats, while only thin and incomplete septa from thickened centrilobular veins were observed in animals treated with DMN + naloxone (fig 8C,D). As expected, this was accompanied by modifications in terms of extracellular matrix deposition. After 5 weeks of treatment, routine examination of Sirius Red-stained liver sections showed micronodular cirrhosis in all animals receiving DMN (fig 9A). In contrast, only thin and incomplete fibrotic septa were observed in DMN + naloxone-treated rats (fig 9B). Fibrosis was quantified by morphometry for Sirius Red-positive parenchyma and hydroxyproline content, showing reduced collagen deposition in DMN + naloxone-treated animals at the fifth week of treatment (fig 9C,D). No differences in terms of ALT values were observed in DMN- compared with DMN + naloxone-treated rats (55.5 (3.5) vs 51.1 (4.2) and 125.8 (12.9) vs 118.7 (15.3) at 1 and 5 weeks in DMN and DMN + naloxone, respectively, p<0.05 vs controls, 31.0 (2.7)). Furthermore, no differences in terms of the histological degree of liver injury were detected between the two treatment groups (4.7 (1.2) vs 4.9 (0.8) and 2.2 (0.5) vs 2.5 (0.3) at 1 and 5 weeks in the DMN and DMN+naloxone groups, respectively).
Effect of Met-enkephalin on HSC proliferation and type I collagen accumulation in the culture medium
Given the increased met-enkephalin expression in hepatocytes in the early phase of liver injury (fig 7), and the effect of naloxone in reducing liver fibrosis in vivo (fig 9), HSCs were incubated with the non-specific OR agonist met-enkephalin in the presence or absence of specific OR antagonists to confirm the plausible role of a paracrine stimulation by EO. Met-enkephalin significantly increased HSC proliferation (measured by BrdU incorporation) and type I collagen accumulation in the culture medium (fig 10A,B). This effect was abolished by the specific blockage of δORs, while μOR and κOR antagonists produced only a non-significant decrease (fig 10A,B).
Met-enkephalin and OR expression in human chronic liver diseases
The plausibility of an EO–OR interaction was finally studied in nine patients with three different active non-cirrhotic chronic liver diseases: PBC, chronic HCV and alcoholic hepatitis. In morphologically normal human liver biopsies, faint met-enkephalin expression was observed in the cytoplasm of hepatocytes (fig 11A). In chronic alcoholic and HCV liver injury, a stronger expression was observed as coarse granuli or diffuse staining in the cytoplasm of hepatocytes close to the area of necrosis, while in PBC the strongest expression was observed in bile duct elements as coarse granuli (fig 11B,C and data not shown).
Concerning the OR expression, parenchymal and inflammatory cells were negative in all human sections, while a strong signal was detected in perisinusoidally located, stellate-shaped cells in areas of active fibrogenesis independently from the aetiology of the chronic liver injury (fig 11D–F). To confirm, as in rats, that the reactive elements were HSCs, double immunostaining was performed. Co-localisation of αSMA (as a general marker of human HSCs10) with OR confirmed that HSCs express ORs during the development of hepatic fibrosis in humans (fig 11G–I and data not shown).
DISCUSSION
This study provides evidence that EO peptides play an important role in the fibrogenic response to chronic liver injury. In particular, this study shows that: (1) HSCs express ORs in vitro and in vivo, showing that δOR is the only OR that is downregulated during the activation process; (2) activation of OR by EO increases HSC proliferation and type I collagen accumulation in the culture medium; (3) these changes in cell growth and collagen accumulation are mediated by activation of the calcium-dependent PKC/ERK/PI3K pathway; and (4) EO stimulates HSCs in a paracrine manner.
Liver fibrosis is the final result of all forms of chronic liver injury and is due to excessive accumulation of extracellular matrix proteins by HSCs.1 2 Several studies demonstrated that HSCs acquire a neuroendocrine phenotype in the course of activation, and that they produce and respond to autonomic nervous system neurotransmitters to promote hepatic fibrosis, leading to the discovery that the neuroendocrine system is involved in the process of fibrogenesis.9–13 Recent studies showed the correlation between EO peptides and specific forms of chronic liver injury. In particular, it has been demonstrated that opioid peptide plasma levels and activity, opioid peptide hepatic concentration and central opioidergic neurotransmission are increased during both human and experimental cholestasis, leading to the finding that EOs play a critical role in the regulation of cholangiocyte biology.17–20 Moreover, a previous observational report has shown increased fibrosis of the space of Disse in intravenous drug abusers compared with their respective controls.41
By RT-PCR, western blot and immunohistochemistry, we demonstrated the expression of “classic” ORs in rat and human HSCs, in vitro and in vivo. More importantly, these receptors are functionally active, since incubation of HSCs with OR selective agonists resulted in a marked increase in cell proliferation and type I collagen accumulation in the culture medium, mediated by a calcium-dependent PKCα/ERK/PI3K and JNK-independent pathway. The finding of reduced δOR expression in activated compared with quiescent HSCs was surprising. However, like many other GPCRs, it has been shown that ORs are susceptible to changes in expression and affinity in the presence of a high concentration of ligands in the microenvironment. In particular it has been demonstrated that, upon activation, OR expression is downregulated as a result of higher intracellular degradation that occurs as a sort of negative feedback.42 Analogous observations have been made in models of chronic liver disease both in the central nervous system43 and in other liver cell phenotypes, such as cholangiocytes.20 Although we have not studied EO production by HSCs, our current data are thus in accordance with such a concept. In agreement with this, reduced δOR expression was found in HSCs isolated 1 week after DMN treatment, which is associated with high met-enkephalin expression in the damaged liver. Furthermore, in vitro, met-enkephalin (as a non-specific OR agonist) induced an increase in HSC proliferation and collagen synthesis presumably mostly acting on δORs, since only the specific δOR antagonist was able to block this process in vitro.
Based on the in vitro findings, we wanted to verify whether the EO–HSC interaction has a role in the development of liver fibrosis in vivo. To avoid the confounding effect of EOs on cholangiocyte proliferation in liver fibrosis after bile duct ligation (BDL),20 we used DMN administration as a model of chronic liver injury which leads to centrilobular fibrosis and cirrhosis.2 37 40 Our data corroborate the hypothesis of a paracrine stimulation by EO on HSCs based on three major findings: (1) the expression of the general EO met-enkephalin increases early in hepatocytes during experimental liver injury; (2) in vivo administration of the general OR antagonist naloxone to DMN-treated rats reduces liver fibrosis; and (3) in vitro, met-enkephalin induces HSC proliferation and collagen synthesis.
To our knowledge, this is the first study to demonstrate met-enkepahlin expression in hepatocytes in a model of experimental hepatic fibrosis not associated with chronic cholestasis. In addition, our findings expand previous observations on the role of EO in hepatic fibrogenesis. Recently, Ebrahimkhani et al showed that naloxone reduced collagen deposition in the BDL model of hepatic fibrosis, and this was associated with reduced αSMA expression and matrix metalloproteinase-2 (MMP-2) activity in the liver. They found that activated HSCs express δOR only, that its activation is associated with increased tissue inhibitor of metalloproteinase-1 (TIMP-1) and type I procollagen expression depending on the experimental conditions in vitro, and that naltrexone presumably blocks these effects by acting on S-adenosylmethionine.22 44 It is well known that hepatic fibrosis during BDL is driven by cholangiocyte proliferation.45 Although Ebrahimkhani et al showed that naltrexone did not modify ductular proliferation during BDL, our recent observation that the increase in opioid peptide synthesis in the course of cholestasis limits the excessive growth of the biliary tree by the interaction with the OR expressed by cholangiocytes, prompted us to investigate a different model of hepatic fibrosis.20 In our hands, no differences in terms of liver injury were observed by measuring ALT levels and necroinflammatory liver injury. Concerning the technical approach of RT-PCR for the detection of ORs in HSCs, we used different primers for the three types of OR compared with those listed in the study of Ebrahimkhani et al. Moreover, we also perfomed the experiment with a higher amount of cDNA for each sample of HSC compared with Ebrahimkhani et al and we used specific thermal cycles, as described in detail in the Methods section. In addition, western blot and immunohistochemistry confirmed the presence of all three functionally active ORs in rat and human HSCs, while the in vitro and in vivo data confirm a direct effect of EOs on HSCs.
Thus, during liver injury, EO peptides produced and released by damaged cells interact with their receptors in HSCs, stimulating cell proliferation and collagen deposition. It is well known that the hepatic wound healing response represents a complex process in which multiple factors participate,1 but our data provide novel perspectives in understanding the pathophysiology of liver fibrosis and may open up additional options for the experimental medical therapy or prevention of chronic liver injury.
Acknowledgments
This work was supported by a research grant from Ministero della Salute (Fattori Nutrizionali e sviluppo di malattie metaboliche), Italy. SDM is a recipient fellow of Dottorato di Ricerca in Alimenti e Salute, V ciclo. This work was supported by MIUR grant 2005067975_004 to SM, and by MIUR grant 2006068958_001 to AB.
REFERENCES
Footnotes
Competing interests: None
Linked Articles
- Digest