Final Report Summary - HCV-POLAR (MECHANISMS OF HEPATITIS C VIRUS ENTRY IN POLARIZED CELLS)
Currently, there is no strong evidence indicating that HCV particle directly binds to CLDN1 or OCLN. However, CLDN1 and CD81 interact as shown by fluorescence resonance energy transfer (FRET) and fluorescence intensity ratio (FIR) assays, suggesting that CLDN1 might be a partner of CD81 in a HCV receptor complex. It has been suggested that CD81 might be responsible for transporting the virus to cell-cell contact regions where the virus could make contact with tight junction proteins. Epidermal growth factor and ephrin receptor A2, by regulating CLDN1-CD81 co-receptor association, also act as host cofactors for HCV entry. One of the most recent cellular factor involved in HCV entry described is the Niemann-Pick C1-like (NPC1L1) cholesterol uptake receptor that seems to act at the fusion step. Finally, it has been recently shown that the transferrin receptor 1 (TsfR1) is also an entry factor for HCV, acting at a post-binding step after CD81 interaction. HCV infects primarily hepatocytes that are highly polarized cells with a complex organization. In recent years, the knowledge concerning HCV entry inside the cells has improved, however the relevance of cell polarity in HCV entry has been addressed in distant models and remains unclear. The aim of this project was therefore to establish a simple polarized cells model to study the mechanisms of HCV entry. Whereas polarized epithelial cells have a rather simple morphology with one basolateral domain and one apical domain, hepatocytes are complex cells. They have several apical domains that form bile canaliculi, a continuous network between adjacent cells in which the bile is secreted. Basolateral domains of hepatocytes face the Disse space and the blood stream. To study the entry of Hepatitis A virus, a picornavirus targeting also hepatocytes, Snooks et al. isolated a clone of HepG2 cells that maintains characteristics of polarized hepatocytes but displays the morphological advantages of simple columnar epithelial cells. Our objective was therefore to produce similar clones infectable by HCV. To obtain similar clones, we used HepG2 cells expressing CD81 (to confer susceptibility to HCV to HepG2 cells), available in the laboratory and we isolated clones by limit dilution.
A first screen of the clones was performed based on the recruitment of the tight junction protein ZO-1 to cell-cell contact region when the clones were grown at high density. To analyze the polarization capacities of the clones, cells were grown on semi-permeable supports that provide independent access to apical and basolateral domains. A lot of tests were required to determine the optimal cell culture conditions to induce cell polarization. The best results were achieved using a support harboring a polyester membrane with 3µm pore size. In order to induce polarization, cells were grown in Williams medium supplemented with 10% FBS and 1% DMSO. To determine the polarization properties of the clones, we first analyzed the localization of apical, basolateral and tight junctions markers. To assess the functionality of the polarization we quantified the basolateral secretion of the human serum albumin. Based on this analysis, we identified two clones with a strong capacity to polarize with a phenotype similar to epithelial cells, with one apical domain and one basolateral domain. We determined that optimal polarization of the cells was achieved between day 6 and 9 after adding the DMSO. Some of the other clones have a strong capacity to form bile canaliculi. Using our polarized clones with an epithelial-like morphology, we were able to confirm that HCV binding and entry occurs at the basolateral domain. We analyzed the membrane partition of HCV co-receptors in polarized cells in confocal microscopy. We found that both tight junctions protein, CLDN1 and OCDN are recruited at the cell-cell contact. SR-BI and CD81 are localized at the basolateral domain of the cells. Recently it has been shown that the proteoglycan, syndecan-1, might be responsible for HCV binding at the cell surface, therefore we analyzed its localization at the cell surface. Syndecan-1 is mainly detected at the basolateral domain, but we could also detect some expression at the apical domain.
The low expression of syndecan-1 at the apical domain is likely responsible for the low binding of HCV at this domain. The viral and cellular determinant responsible for the initial binding remains unclear. In the continuity of this work, we investigated this early step of the viral life cycle in non-polarized Huh-7 cells. Early studies have found that the hypervariable region (HVR1) of E2 was responsible for heparan sulfate binding, however due to its lipoviroparticle structure it has been recently suggested that the apolipoprotein E associated with the virion was responsible for this interaction. We have shown that deletion of HVR1 does not alter HCV binding to HS. In contrast, results from kinetic studies, heparin pull-down and inhibition experiments with anti-apolipoprotein E antibodies indicate that this apolipoprotein plays a major role in HCV-HS interaction. Finally, characterization of HS structural determinants required for HCV infection by silencing enzymes involved in the HS biosynthesis pathway and by competition with modified heparin indicated that N- and 6-O-sulfation but not 2-O-sulfation are required for HCV infection, and that the minimum HS oligosaccharide length required for HCV infection is a decasaccharide.
In polarized cells, in order to deliver the proteins to the appropriate compartment, the intracellular trafficking is also polarized. We monitored HCV secretion to determine if the export of the virus is also polarized. We found that around 90% of the virus produced is secreted in the basolateral chamber. Because of its lipoviroparticle structure, the density of the virus is heterogenous and densities of cell culture derived viruses are higher than those isolated from patient. This discrepancy may be explained by the unability of Huh7 cells to produce very low-density lipoproteins. We compared the densities of the virions produced by polarized cells with Huh7 cells derived virions and we did not find any change in the density profiles of the virus or any difference between the virus secreted in the basolateral or apical chamber.
In conclusion, we developed a simple model of polarized hepatocytes that will allow for studying the different aspects of HCV life cycle at the molecular level taking the cell polarity into considerations. For the first time, we were able to show that HCV entry and release occurs at the basolateral domain of the cells.