These authors contributed equally to this study.
Hepatitis C virus receptor expression in normal and diseased liver tissue†
Article first published online: 12 DEC 2007
Copyright © 2007 American Association for the Study of Liver Diseases
Volume 47, Issue 2, pages 418–427, February 2008
How to Cite
Reynolds, G. M., Harris, H. J., Jennings, A., Hu, K., Grove, J., Lalor, P. F., Adams, D. H., Balfe, P., Hübscher, S. G. and McKeating, J. A. (2008), Hepatitis C virus receptor expression in normal and diseased liver tissue. Hepatology, 47: 418–427. doi: 10.1002/hep.22028
Potential conflict of interest: Nothing to report.
- Issue published online: 26 JAN 2008
- Article first published online: 12 DEC 2007
- Manuscript Accepted: 15 AUG 2007
- Manuscript Received: 13 JUL 2007
- United States Public Health Service. Grant Numbers: AI50798, AI40034
- Wellcome Trust (United Kingdom)
The principal site of hepatitis C virus (HCV) replication is the liver. HCV pseudoparticles infect human liver derived cell lines and this suggests that liver-specific receptors contribute to defining HCV hepatotropism. At least three host cell molecules have been reported to be important for HCV entry: the tetraspanin CD81, scavenger receptor class B member I (SR-BI), and the tight junction (TJ) protein Claudin 1 (CLDN1). Hepatocytes in liver tissue coexpress CD81, SR-BI, and CLDN1, consistent with their ability to support HCV entry. CLDN1 localized at the apical-canalicular TJ region and at basolateral-sinusoidal hepatocyte surfaces in normal tissue and colocalized with CD81 at both sites. In contrast, CLDN1 appeared to colocalize with SR-BI at the basolateral-sinusoidal surface. CLDN1 expression was increased on basolateral hepatocyte membranes in HCV-infected and other chronically inflamed liver tissue compared with normal liver. In contrast, CLDN4 hepatocellular staining was comparable in normal and diseased liver tissue. Conclusion: HCV infection of Huh-7.5 hepatoma cells in vitro significantly increased CLDN1 expression levels, consistent with a direct modulation of CLDN1 by virus infection. In HCV infected livers, immunohistochemical studies revealed focal patterns of CLDN1 staining, suggesting localized areas of increased CLDN1 expression in vivo which may potentiate local viral spread within the liver. (HEPATOLOGY 2007.)
Hepatitis C virus (HCV) has emerged as the major etiological agent of liver disease in many regions of the world. Approximately 170 million individuals are infected worldwide, and the majority are at risk of developing serious progressive liver disease, with HCV being the leading indication for liver transplantation. The principal site of HCV replication is the liver; however, the cell types targeted by the virus and the frequency of infected cells are poorly defined.1, 2 The selective association of a virus with a target cell is initially defined by interactions between the viral glycoproteins and specific cell surface molecules. Recent advances have allowed us to study the mechanism of HCV entry: firstly, the development of the retrovirus pseudoparticle system [hepatitis C virus pseudoparticle (HCVpp)], which measures glycoprotein-dependent particle entry,3, 4 and secondly, the ability of the JFH-1 strain of HCV to replicate and release infectious particles in cell culture [hepatitis C virus, cell culture produced (HCVcc)].5–7
The observation that HCVpp infects human liver–derived cell lines in vitro suggests that liver-specific receptors contribute to defining HCV tropism for the liver. Recent evidence suggests that at least three host cell molecules are important for HCV entry in vitro: the tetraspanin CD81,4, 7–9 the scavenger receptor class B member I (SR-BI),9–12 and the tight junction (TJ) protein Claudin 1 (CLDN1).13 HCV glycoproteins have been reported to interact with CD81 and SR-BI.14 Mutagenesis and antibody blocking studies suggest that the first extracellular loop of CLDN1 interacts with HCV13; however, the exact role(s) played by each of the receptors and whether they interact to form receptor complexes are unclear.15
CD81 is expressed in most tissues and localizes to the basolateral surface of polarized epithelial cell lines.16, 17 SR-BI is expressed within steroidogenic tissue, macrophages, and the liver,18 localizing to the apical surface in a variety of epithelial cells.19 CLDN1 is expressed in many tissues, but at a high concentration in the liver.20 Studies with polarized epithelial cell lines suggest that TJ proteins, including CLDN1, localize exclusively to the upper apical region of the cell and form protein interactions between cells.21 To understand how CD81, SR-BI, and CLDN1 coordinate viral entry, it is important to compare receptor expression pattern(s) and localization in liver tissue and hepatoma cell lines.
In the current study, we demonstrate that hepatocytes in tissue and ex vivo express CD81, SR-BI, and CLDN1, consistent with their ability to support HCV entry.22 CLDN1 localized at the apical-canalicular TJ region and basolateral-sinusoidal hepatocyte surface in normal tissue and colocalized with CD81 at both sites. In contrast, CLDN1 appeared to colocalize with SR-BI at the basolateral-sinusoidal surface. These observations are consistent with HCV entry into the liver via the sinusoidal blood, allowing access to receptor complexes at the sinusoidal hepatocyte surface.
Hepatocellular CLDN1 levels were increased in HCV-infected and other chronically inflamed liver tissue compared with normal liver, with discrete focal regions of basolateral expressed CLDN1 noted in the HCV-infected livers. To ascertain if HCV can modulate CLDN1 levels, we quantified expression in HCVcc-infected Huh-7.5 hepatoma cells by confocal microscopy. HCV-infected cells expressed significantly more CLDN1 than uninfected cells, consistent with a direct modulation of CLDN1 by virus infection.
Materials and Methods
Huh-7.523 and 293T cells from the American Type Culture Collection were propagated in Dulbecco's modified Eagle medium supplemented with 10% fetal bovine serum and 1% nonessential amino acids.
Human Liver Tissue and Primary Cell Isolation.
Formalin-fixed paraffin-embedded tissue was obtained from hepatectomy specimens of patients undergoing liver transplantation for cirrhosis due to HCV, alcoholic liver disease (ALD), autoimmune hepatitis (AIH), primary biliary cirrhosis (PBC), and primary sclerosing cholangitis (PSC). Liver disease in each of the above categories was diagnosed according to standard clinical, biochemical, virological, immunological, radiological, and histological criteria. Normal liver was obtained from surplus donor tissue used for reduced-size liver transplantation. Informed consent from each patient/donor was obtained together with regional ethics committee approval.
Sinusoidal endothelial cells (SECs) and biliary epithelial cells (BECs) were enriched from explanted tissue by collagenase digestion and immunomagnetic selection with antibodies specific for CD31 (Dako, United Kingdom) or epithelial cell adhesion molecule (EP-CAM) (Progen-Biotechnik, GmbH), respectively. Cells were cultured in complete media containing human serum and growth factors.24, 25 SECs were transduced with lentiviral vectors expressing SR-BI and CLDN1, as previously reported.12, 13 Hepatocytes were isolated from explanted tissue by a modified collagenase perfusion technique and cultured in serum-free complete Williams E media, as previously described.26
Representative 3-μm sections were cut from paraffin blocks of formalin-fixed tissue, placed onto charged slides, and incubated for 1 hour at 60°C. Sections were dewaxed and rehydrated, and an endogenous peroxidase block performed with 0.3% hydrogen peroxide in distilled water. Tissue was subjected to an agitated low-temperature epitope retrieval technique as previously described.27 Sections were mounted onto a Shandon sequencer and incubated with primary antibodies or nonimmune sera (summarized in Table 1) in trishydroxymethylaminomethane-buffered saline/0.1% Tween (TBS/Tween) applied for 1 hour. After a TBS/Tween wash, sections were visualized with ChemMate Envision (Dako, United Kingdom) and NovaRed (Vector Labs, Burlingame, CA). Microscopic examination was performed by two observers, and expression was assessed semiquantitatively on a scale of 0 to +++ according to the intensity of staining (0 = absent, +/− = faint/equivocal, + = weak, ++ = moderate, +++ = strong). Acetone-fixed frozen sections of normal and HCV-infected tissues were used to confirm that the pattern and staining intensity of CD81, SR-BI, and CLDN1 expression had not been altered during formalin fixation and processing.
|Antibody Group||Specificity||Clone||Source||Working Concentration (μg/mL)*|
|HCV receptors||CD81||M38||F. Berditchevski†||2.0|
|HCV||NS5A||9E10||C. M. Rice‡||1.0|
|Liver cell markers||CD10: biliary canaliculi||56C6||Nova Castra||1:20|
|CD68: macrophages (Kupffer cells)||PG-M1||Dako||1:100|
|Cytokeratin 7: bile duct epithelium||OV-TL12/30||Dako||1:200|
|Cytokeratin 8: hepatocyte, bile duct epithelium||Cam 5.2||BD Biosciences||1:10|
Laser Scanning Confocal Microscopy.
Laser scanning confocal microscopy was performed on a Zeiss meta head confocal microscope. Background and autofluorescence of tissue samples were corrected throughout. Following agitated low-temperature epitope retrieval pretreatment of sections, primary antibodies were applied for 1 hour. After a TBS/Tween wash, anti-mouse Alexafluor-488 and anti-rabbit Alexafluor-633 (Invitrogen, California) were applied at 1/200 dilution for 30 minutes. Following a further wash, sections were counterstained with hematoxylin and mounted. Anti-CLDN1 fluorescent intensity was calculated by linear profiling of uninfected and HCV-infected Huh-7.5 cells.
HCVpp and HCVcc Genesis and Infection.
Pseudoviruses were generated by the transfection of 293T cells with plasmids encoding human immunodeficiency virus provirus expressing luciferase and the HCV envelope or murine leukemia virus glycoproteins.4 Supernatants were harvested 48 hours post transfection, pooled, and filtered. Virus-containing media were added to target cells plated at 1.5 × 104 cells/cm2 and incubated for 8 hours. Unbound virus was removed and cells were replaced with respective growth media. Infections were terminated after 72 hours by lysing cells and firefly luciferase activity was measured.4
HCV-JFH-1 was generated as previously described.7 Briefly, RNA was transcribed in vitro from full-length genomes with the Megascript T7 kit (Ambion, Austin, TX) and electroporated into Huh-7.5 cells. Seventy-two hours and 96 hours post electroporation, supernatants were collected, pooled, and stored immediately at −80°C. Virus-containing media were added to target cells plated at 1.5 × 104 cells/cm2 and incubated for 8 hours. Unbound virus was removed and media replaced with 3% fetal bovine serum/Dulbecco's modified Eagle medium and incubated at 37°C. Infected cells were detected by methanol fixation and staining for nonstructural protein 5A (NS5A) with anti–NS5A 9E10 and Alexa-conjugated anti-mouse immunoglobulin G (Invitrogen, Carlsbad, CA).
CD81, SR-BI, and CLDN1 Expression in Normal Liver Tissue.
To ascertain the CD81, SR-BI, and CLDN1 localization in normal liver tissue, formalin-fixed tissue was stained with antibodies specific for each of the receptors. Antibody specificity was confirmed by the demonstration of reactivity with Chinese hamster ovary cells engineered to express each of the human receptors. CD81 was primarily expressed on sinusoidal endothelium and hepatocytes (Fig. 1A and Table 2). Hepatocellular CD81 expression was mainly basolateral with some canalicular localization. Staining was also present in the stroma of portal tracts, but there was no detectable staining of bile ducts. SR-BI was expressed on sinusoidal endothelium and hepatocytes, with minimal staining of bile ducts (Fig. 1B and Table 2). Hepatocellular SR-BI expression was mainly located at the basolateral membrane with minimal staining at the canaliculi.
|Receptor||Bile Ducts||Vessels||Inflammatory Cells||Hepatocytes||Sinusoidal Endothelium|
CLDN1 was detected on bile ducts and hepatocytes, with low-level staining of sinusoidal endothelium (Fig. 1C). Hepatocyte CLDN1 expression was apparent at the basolateral and canalicular membranes (Figs. 1C and 2). This localization was confirmed by costaining with a broad-spectrum anti-cytokeratin (Cam5.2; Fig. 2C), which is expressed on the basolateral membrane of hepatocytes, and a canalicular marker, CD10 (Fig. 2F). Dual staining of CLDN1 with the endothelial marker CD31 confirmed low-level expression of CLDN1 in SECs (Fig. 2I).
Formalin fixation and paraffin processing did not alter CD81, SR-BI, or CLDN1 expression pattern(s), with acetone-fixed frozen tissue showing a similar pattern and intensity of receptor distribution (not shown).
HCVpp Infection of Primary Liver Cells.
To define which cell types within the liver support HCV entry, primary human hepatocytes (PHHs), SECs, and BECs were isolated from normal tissue and enriched populations screened for viral receptor expression and HCVpp infection. Hepatocytes and BECs express low levels of all three receptors, whereas SECs express CD81, low-level CLDN1, and no detectable SR-BI (Fig. 3). SECs were transduced with retroviral vectors to express SR-BI or CLDN1 (Fig. 3C) and evaluated for their ability to support HCVpp entry. Only hepatocytes and Huh-7.5 hepatoma cells supported HCVpp infection, with parental and transduced SECs showing no evidence of HCVpp infection (Fig. 3D). All primary cells supported the entry and replication of pseudoparticles bearing MLV glycoprotein (Fig. 3D). These infectivity data suggest that primary hepatocytes express functionally active receptors that facilitate HCV entry.
Does HCV Infection Modulate Receptor Expression?
We compared the pattern and staining intensity of anti-CD81, anti–SR-BI, and anti–CLDN1 antibodies for normal, HCV-infected, and diseased liver tissue. No significant changes were noted in CD81 and SR-BI expression levels between normal and diseased tissue (Table 2). Hepatocellular CLDN1 expression was increased in HCV-infected and other diseased tissue compared with normal liver (Table 2 and Fig. 4). In contrast, CLDN1 staining of sinusoidal endothelium was not increased in diseased tissue compared with normal tissue (Table 2). In diseased livers, CLDN1 was highly expressed on reactive ductules in cirrhotic septa, whereas native bile ducts were stained with an intensity similar to that seen in normal liver. In HCV-infected liver, CLDN1 expression was increased on the basolateral hepatocyte membrane, with a tendency to be strongest in periseptal region(s) (Fig. 4).
To determine whether antibodies to other CLDN family members showed altered staining, we compared the reactivity of antibodies specific for CLDN2, CLDN4, and CLDN5 for normal, diseased, and HCV-infected livers (Table 2). CLDN antibodies stained bile ducts in all samples with comparable intensity (Table 2). CLDN2 was present on inflammatory cells, which were more prominent in diseased liver than normal liver. Hepatocytes in normal and diseased liver showed minimal expression of CLDN2 and CLDN5. Hepatocytes expressed CLDN4 with no discernible increase in staining intensity in HCV-infected livers compared with normal or diseased livers (Table 2).
To semiquantify the increased CLDN1 expression observed in the HCV-infected livers, we determined the endpoint dilution at which an anti–CLDN1 antibody would stain normal and HCV-infected tissue. Anti–CLDN1 (5 μg/mL) stained basolateral CLDN1 more intensely in the HCV-infected livers than in the normal livers (Fig. 4D). Reducing the antibody concentration abolished basolateral membranous staining in the normal liver while staining the HCV-infected liver (Fig. 4B-F). No hepatocyte staining was observed in the two normal livers at a concentration of 0.3 μg/mL (Fig. 4C), whereas hepatocytes demonstrated a focal dotlike CLDN1 pattern at both the basolateral and canalicular membranes in the HCV-infected livers (Fig. 4F), suggesting localized domains of high CLDN1 expression. Quantitative confocal imaging of NL1 and HCV5 tissue demonstrated a significant increase in plasma membrane CLDN1 staining in the HCV-infected liver (Fig. 5).
To address whether HCV infection can directly modulate CLDN1 protein levels, we studied co-receptor expression in HCVcc JFH-infected Huh-7.5 hepatoma cells by confocal microscopy. The frequency of cells with intense anti–CLDN1 staining was increased in the HCV-infected cell population compared with naïve cells (Fig. 6). The expression of NS5A and CLDN1 in HCV-infected cells were analyzed by linear fluorescent profiling of individual NS5A+ and NS5A− cells (Fig. 6E,F). The average anti–CLDN1 fluorescent intensity was significantly increased in JFH-infected cells compared to uninfected cells; this was more apparent for cells expressing NS5A within the infected culture (Fig. 6F; P < 0.0001). In contrast, HCV infection had a modest effect on the level of CLDN1 messenger RNA with respect to the housekeeping gene glyceraldehyde 3-phosphate dehydrogenase in Huh-7.5 cells, demonstrating a 28% ± 32% increase. These data demonstrate that HCV infection increases plasma membrane CLDN1 protein levels in Huh-7.5 hepatoma cells.
HCV Receptor Colocalization.
To address whether the viral receptors colocalize in hepatocytes, normal liver tissue and HCV infected liver tissue were costained with antibodies specific for CLDN1/CD81 and CLDN1/SR-BI. CLDN1 colocalized with CD81 in normal and HCV-infected tissue. In normal liver, CLDN1/CD81 colocalization appeared to be strongest in the apical-canalicular region (Fig. 7A-D). In contrast, in the HCV-infected liver tissue, colocalization was more prominent in the basolateral region (Fig. 5A and Fig. 7E-H), presumably reflecting the increased basolateral expression of CLDN1 in HCV-infected liver samples. CLDN1 colocalized with SR-BI at the basolateral membrane in normal and HCV-infected tissue, with minimal costaining at the canalicular region (Fig. 7I-P). Because the majority of published reports have studied HCV entry into Huh-7 hepatoma cells, we were interested in comparing their pattern of receptor expression to those in the liver. CLDN1 localized predominantly to the plasma membrane (Fig. 6E) with minimal intracellular staining and showed discrete regions of colocalization with CD81 (not shown). Because of the low level of SR-BI expression in Huh-7.5 cells, we were unable to study its localization with CLDN1. Because of incompatibilities in the isotype and species of the anti-CD81 and anti–SR-BI antibodies available, we were unable to perform a three-way labeling reaction. Our data support a model in which CD81, SR-BI, and CLDN1 colocalize within the hepatocellular basolateral membrane, with CD81 and CLDN1 showing evidence of colocalization at the canaliculi.
We demonstrate that hepatocytes within normal tissue coexpress CD81, SR-BI, and CLDN1, consistent with their ability to support HCV entry and replication.22, 28 CLDN1 localized at both the apical-canalicular TJ region and basolateral-sinusoidal hepatocyte surfaces in normal tissue, showing colocalization with CD81 at both sites. In contrast, CLDN1 appeared to colocalize with SR-BI at basolateral-sinusoidal surfaces. Because HCV is likely to enter the liver through the sinusoidal blood, the observation that CD81, SR-BI, and CLDN1 are expressed at the sinusoidal surface of hepatocytes is consistent with physiological routes of viral access. Sinusoidal endothelium coexpressed CD81 and SR-BI with low-level CLDN1 expression (Fig. 1 and Table 2); however, isolated SECs failed to express detectable levels of SR-BI (Fig. 3). Transduction of SECs to express SR-BI and increased CLDN1 levels had no effect on their ability to support HCVpp entry (Fig. 3). Comparable data were obtained with HCVcc (J6/JFH-1 and JFH-1) only infecting PHH and failing to infect BECs or SECs (data not shown). We previously reported that sinusoidal endothelium expressed DC-SIGN and L-SIGN lectins known to interact with HCV,29 supporting a model in which SECs concentrate viral particles from the circulating blood and potentiate access to the basolateral surface of the permissive hepatocyte membrane.
Hepatocytes, the major epithelial cells within the liver, are highly polarized, and their plasma membranes are separated by TJs into apical-canalicular and basolateral-sinusoidal domains.30 CLDNs and other TJ transmembrane proteins are reported to localize at the most apical surface of polarized epithelial cells.31, 32 In this study, we demonstrate CLDN1 expression at both the canalicular and basolateral hepatocyte surface, independent of the fixative used for preparation of liver tissue (Fig. 2). Variable patterns of CLDN1 expression were observed in normal tissue stained with different concentrations of anti–CLDN1 (Fig. 4). At a high antibody concentration, hepatocellular staining was detected at canalicular and basolateral regions, whereas at low concentrations, the staining was largely confined to the canalicular region(s) (Fig. 4B); this suggests that CLDN1 is predominantly expressed at the apical-canalicular surface in normal liver. Rahner and colleagues32 reported that a large proportion of CLDNs localize not in strands at TJs but in intracellular vesicles or in unassembled states on lateral or basolateral cell surface(s). Our data support a model in which tight-junctional (apical-canalicular) and nonjunctional (basolateral-sinusoidal) forms of CLDN1 exist in hepatocytes. At the present time, the form of CLDN1 that functions with CD81 and SR-BI to facilitate HCV entry is unknown. Evans and colleagues13 demonstrated that a CLDN1 variant lacking the C-terminal region supported HCV entry into Huh-7.5 hepatoma cells. Because this region of CLDN1 mediates interactions with cytoplasmic and signaling components of the TJ complex,33 these data support a model in which HCV may use nonjunctional forms of CLDN1.
Hepatocytes within HCV-infected and diseased liver tissue showed increased basolateral staining with anti–CLDN1, whereas sinusoidal endothelium demonstrated comparable staining patterns regardless of disease status (Table 2). In contrast, hepatocellular CLDN4 staining intensity was comparable in normal and diseased livers (Table 2). Altered CLDN expression has been linked to several human inflammatory diseases.34, 35 Utech and colleagues36 reported that gamma interferon treatment of the T84 epithelial cell line led to an endocytosis of TJ proteins, including CLDN1, from the apical membrane. It is interesting to note that increased hepatocellular CLDN1 expression was more prominent in HCV-infected livers in areas adjacent to cirrhotic inflamed septa. Experiments to study the effect(s) of gamma interferon on CLDN1 expression have shown no modulation in Huh-7.5 cells, consistent with the treated cells supporting HCVpp entry (M. Farquhar, unpublished data, 2007). The observation that HCV-infected Huh-7.5 cells expressed greater levels of CLDN1 at the plasma membrane (Fig. 6) demonstrates that infection can directly increase CLDN1 levels without altering messenger RNA levels. The focal regions of increased CLDN1 expression observed in the HCV-infected liver may be a direct consequence of viral infection and may represent areas of ongoing HCV replication (Fig. 4). Several reports suggest that CLDN expression and localization may be regulated at the level of protein phosphorylation and palmitoylation,37, 38 and future experiments will study the effect(s) of HCV proteins on CLDN1 posttranslational modification and processing.
The functional consequences of increased hepatocellular CLDN1 expression in late-stage HCV infection are unclear. It will be important to compare hepatocellular CLDN1 expression levels during HCV disease progression. Higashi and colleagues39 reported that reduced CLDN1 expression correlated with the malignancy of hepatocellular carcinoma. Most cholestatic disorders are associated with changes in TJs and the hepatocyte cytoskeleton; it will be interesting to study HCV-infected subjects with cholestatic disorders following liver transplantation. In vitro experiments to study the interplay between HCV replication and TJ protein localization and function will require the development of polarized cell systems that support HCV infection.
We thank Takaji Wakita for JFH-1; Charles Rice for J6/JFH-1, anti–NS5A 9E10, and Huh-7.5 cells; and Michelle Farquhar for citing unpublished data.
- 27Interleukin 6 expression by Hodgkin/Reed-Sternberg cells is associated with the presence of 'B' symptoms and failure to achieve complete remission in patients with advanced Hodgkin's disease. Br J Haematol 2002; 118: 195–201., , , , , , et al.