Gut. 2014 Sep;63(9):1490-500.

In vitro infection of primary human hepatocytes by HCV-positive sera: insights on a highly relevant model.

Gondeau C1, Briolotti P1, Razafy F1, Duret C1, Rubbo PA2, Helle F3, Rème T1, Ripault MP4, Ducos J5, Fabre JM6, Ramos J7, Pécheur EI8, Larrey D9, Maurel P1, Daujat-Chavanieu M10.

  • 1INSERM U1040, Montpellier, France Université Montpellier 1, Montpellier, France.
  • 2Université Montpellier 1, Montpellier, France INSERM U1058, Montpellier, France.
  • 3EA4294, Laboratoire de Virologie, Centre Hospitalier Universitaire et Université de Picardie Jules Verne, Amiens, France.
  • 4Department of Hepato-gastroenterology A, Hospital Saint Eloi, CHU Montpellier, Montpellier, France.
  • 5INSERM U1058, Montpellier, France Département de Bactériologie-Virologie, CHU de Montpellier, Montpellier, France.
  • 6Department of Digestive Surgery, Hospital Saint Eloi, CHU Montpellier, France.
  • 7Pathological anatomy department, CHU Gui de Chauliac, Montpellier, France.
  • 8UMR INSERM 1052/CNRS 5286, Centre de Recherche en Cancérologie de Lyon, Université de Lyon, Lyon, France.
  • 9INSERM U1040, Montpellier, France Department of Hepato-gastroenterology A, Hospital Saint Eloi, CHU Montpellier, Montpellier, France.
  • 10INSERM U1040, Montpellier, France Université Montpellier 1, Montpellier, France CHU Saint Eloi, Institute of Research in Biotherapy, Montpellier, France.



OBJECTIVE: Adult primary human hepatocytes (PHHs) support the complete infection cycle of natural HCV from patients’ sera. The molecular details underlying sera infectivity towards these cells remain largely unknown. Therefore, we sought to gain a deeper comprehension of these features in the most physiologically relevant culture system.

DESIGN: Using kinetic experiments, we defined the optimal conditions to infect PHH and explored the link between cell organisation and permissivity. Based on their infectivity, about 120 sera were classified in three groups. Concentration of 52 analytes was measured in 79 selected sera using multiplexed immunobead-based analyte profiling.

RESULTS: PHH permissivity towards HCV infection negatively correlated with cell polarisation and formation of functional bile canaliculi. PHH supported HCV replication for at least 2 weeks with de novo virus production. Depending on their reactivity, sera could be classified in three groups of high, intermediate or low infectivity toward PHH. Infectivity could not be predicted based on the donors’ clinical characteristics, viral load or genotype. Interestingly, highly infectious sera displayed a specific cytokine profile with low levels of most of the 52 tested analytes. Among them, 24 cytokines/growth factors could impact hepatocyte biology and infection efficiency.

CONCLUSIONS: We identified critical factors leading to efficient PHH infection by HCV sera in vitro. Overall, we showed that this cellular model provides a useful tool for studying the mechanism of HCV infection in its natural host cell, selecting highly infectious isolates, and determining the potency of drugs towards various HCV strains.


PMID: 24153249



Hepatitis C virus (HCV) infection is a major cause of chronic liver disease. Many different in vitro models that mimic the entire virus cycle have been described but are largely based on the molecular clone JFH1 (HCVcc, genotype 2a), subgenomic or full length replicons and hepatoma cell lines, which mimic partly the natural infection. We and others have previously reported that mature human hepatocytes in primary culture (PHH), isolated from non-infected patients, are sensitive to infection by HCV-positive sera from infected patients (HCVser, different HCV genotypes) or HCVcc (JFH1) and permissive to viral genome replication. This system remains so far the most physiologically relevant cell-culture model for studying hepatitis C virus infection but few sera give measurable level of intracellular HCV RNA and most of laboratories failed in using it. These considerations prompted us to study the relationship between host cell phenotype in culture, characteristics of viral isolates and serum infectivity.

Our findings clearly demonstrated that there are critical time windows post-seeding for inoculation and infectivity analysis. PHH susceptibility to HCV infection is strongly linked to cell polarization and membrane/cytoskeleton organization (Figure 1). Importantly, in our culture conditions, viral strains from HCVser can complete their viral cycle in PHH, as well as the cell culture-produced particle HCVcc/JFH1 (Figure 2). Infectious cycles could be followed up to 15 days and this is of particular interest for studying cell disorders involved in the switch from acute to chronic infection.
In our well-defined culture conditions, a standard test allowed classifying ~120 sera in 3 groups, based on their infectivity toward PHH. We found that 12 % of the sera tested are highly infectious (HI; more than 5×103 HCV RNA copies/µg of total RNA), 68 % are poorly infectious (PI; less than 1.3×103 copies/µg) and 20 % are intermediate (Inter).

Parameters including patient clinical profile, viral load, genotype, and cytokines content were then carefully examined as potential markers allowing the prediction of sera infectivity in vitro. We showed that sera infectivity cannot be predicted based on the viral genotype or viral load but has to be determined using a standard test. The donor patient’s clinical profile could be important: 8 of the 14 samples collected after liver transplantation were highly infectious.

Interestingly, we showed that the HI sera have a cytokine profile that clearly distinguishes them from other groups, with low levels of the majority of the 52 analytes tested, including cytokines involved in the regulation of immune responses and inflammatory reactions (Figure 3). Thus, it could be hypothesized that a combination of some of the analytes highlighted in our study impacts on sera infectivity in vitro.

Study importance: The use of sera for HCV-infection studies is considered challenging due their low infectivity and replication. Here, we demonstrate that well-defined culture conditions can lead to the identification of highly infectious strains, such as S310. Indeed, a full-length consensus genome was cloned from the S310 viral RNA and was used to establish the first infectious genotype 3a HCV system, providing a new tool for genotype 3a HCV studies in human hepatoma cells [1,2]. More than 30 % of the serum samples tested are infectious enough to be used in experimental protocols; 15 % can be used to evaluate the potency of antiviral drugs and to determine their efficiency against viral strains with different genotypes and quasi-species. Finally, based on the cytokine profile of HI sera, our results suggest that suppression of inflammation by immunosuppressive treatments could impact on hepatocyte biology and favor HCV recurrence after liver transplantation.

1- Saeed M, et al. Gastroenterology 2013 Jan;144(1):56-58.
2- Kim S, et al. Hepatology 2014 Dec;60(6):1838-50.


Gondeau_Figure-1Fig 1. PHH reorganization in polarized cell layers after seeding and sensitivity to HCV infection. A. The localization of the polarization markers ZO-1 and Occludin was assessed in PHH at days 1, 4 and 7 post-seeding by immunofluorescence staining; nuclei were stained with Hoechst (blue). B. PHH isolated from liver FT316 and FT351 were exposed to HCVser S317 (genotype 1b) for 16h at different days post-seeding (D). The intracellular HCV RNA concentration was evaluated 72h post-inoculation.


Gondeau_Figure-2Fig 2. HCV replication, persistence of infection in PHH and potency of NS3-4 inhibitors. A. PHH FT383 were transduced with lentiviruses expressing RFP-NLS-IPS at D1 post-seeding and then were inoculated with serum S310 (left panel), S317 or HCVcc (right panels). HCV infected cells were identified by translocation of the cleavage product RFP-NLS to the nucleus (pink nuclei indicated by arrows) 72 h post-infection. B. PHH isolated from liver FT316, FT351 and FT324 were infected at D3 post-seeding with HCVser S317 (genotype 1b), S310 (genotype 3a) and HCVcc (JFH1), respectively. Cells were washed 8h post-inoculation and cellular HCV RNA was evaluated at the indicated time points post-infection (I1=D4 post-seeding). HCV RNA is expressed as the mean copy number/µg of total RNA on a log10 scale (± SD, n= 2 replicates/condition). C. (Left) PHH FT385 cells were infected with S397 (genotype 1b) and treated with increasing concentrations of Telaprevir or BILN 2061 after washing off the inoculum. (Right) PHH FT383 cells were infected with serum S294 (genotype 3a) and treated with increasing concentrations of Telaprevir after washing off the inoculum. Curves represent the non-linear regression analysis of the data.


Gondeau_Figure-3Fig 3. Classification of serum samples based on the analyte content. A. Twenty four analytes differentially concentrated (p < 0.05) in highly infectious (HI) and poorly infectious (PI) serum samples were identified using a Mann-Whitney unpaired rank test, and visualized in a heat map by unsupervised hierarchical clustering. Concentrations in red: high; green: low; black: unchanged. Each row represents an analyte; each column represents a serum. Dendrogram: at the top of the matrix indicates the distance between HCVser; on the left, the distance between analytes.


Multiselect Ultimate Query Plugin by InoPlugs Web Design Vienna | Webdesign Wien and Juwelier SchönmannMultiselect Ultimate Query Plugin by InoPlugs Web Design Vienna | Webdesign Wien and Juwelier Schönmann