Immortalized stem cell-derived hepatocyte-like cells: An alternative model for studying dengue pathogenesis and therapy

Abstract

Suitable cell models are essential to advance our understanding of the pathogenesis of liver diseases and the development of therapeutic strategies. Primary human hepatocytes (PHHs), the most ideal hepatic model, are commercially available, but they are expensive and vary from lot-to-lot which confounds their utility. We have recently developed an immortalized hepatocyte-like cell line (imHC) from human mesenchymal stem cells, and tested it for use as a substitute model for hepatotropic infectious diseases. With a special interest in liver pathogenesis of viral infection, herein we determined the suitability of imHC as a host cell target for dengue virus (DENV) and as a model for anti-viral drug testing. We characterized the kinetics of DENV production, cellular responses to DENV infection (apoptosis, cytokine production and lipid droplet metabolism), and examined anti-viral drug effects in imHC cells with comparisons to the commonly used hepatoma cell lines (HepG2 and Huh-7) and PHHs. Our results showed that imHC cells had higher efficiencies in DENV replication and NS1 secretion as compared to HepG2 and Huh-7 cells. The kinetics of DENV infection in imHC cells showed a slower rate of apoptosis than the hepatoma cell lines and a certain similarity of cytokine profiles to PHHs. In imHC, DENV-induced alterations in levels of lipid droplets and triacylglycerols, a major component of lipid droplets, were more apparent than in hepatoma cell lines, suggesting active lipid metabolism in imHC. Significantly, responses to drugs with DENV inhibitory effects were greater in imHC cells than in HepG2 and Huh-7 cells. In conclusion, our findings suggest superior suitability of imHC as a new hepatocyte model for studying mechanisms underlying viral pathogenesis, liver diseases and drug effects.

Fig 1. Characterization of DENV binding ability and potential DENV receptors of imHC in comparison with commonly used hepatoma cell lines (Huh-7 and HepG2).

(A) The three hepatic cell types (imHC, Huh-7, HepG2) were incubated with different MOIs of DENV-2 for 1 h at 4°C, and the levels of DENV-2 bound to the cell surface were determined by immunofluorescent staining with anti-E antibody followed by flow cytometric analysis. In the absence of DENV, each cell type has unique background immunofluorescent staining. Therefore, the mean fluorescence intensity (MFI) values from the control (no DENV) condition of each cell type were subtracted from MFI values obtained from the corresponding cells in the presence of DENV. Data shown are mean ± S.D. of the normalized MFI values from three independent experiments. (B) The three hepatic cell types were incubated with DENV-2 at MOI of 5 for 1 h at 4°C in the presence of heparin at different concentrations, and the bound DENV levels were determined as described in A. The percentages of DENV binding to cells were calculated by comparing the subtracted MFI values of heparin-containing conditions to those from the control (no heparin) conditions of the same cell type. Data shown are mean ± S.D. of values from three independent experiments. An asterisk (*), “#” symbol and “+” symbol indicate significant differences between imHC vs HepG2, imHC vs Huh-7, and Huh-7 vs HepG2, respectively (* P < 0.05; ** P < 0.01; **** P < 0.0001). (C) Representative flow cytometry histograms showing signals on the surface of three cell types stained with specific antibodies for indicated DENV receptors (gray) compared with the isotype control.

Fig 2. Efficiency of imHC in DENV replication and production in comparison with hepatoma cell lines.

The three hepatic cell lines were cultured in 24-well plates and infected with different MOIs of DENV-2 (0.1, 1 and 5). The cells and the culture supernatants were collected at 24 and 48 hpi. Intracellular expression levels of NS3, a DENV antigen representing DENV replication, were determined by immunofluorescence and flow cytometry (A). NS1 secretion levels were determined by ELISA (B). The infectious virion production levels in supernatants (C) and cell-associated viruses in cell extracts (D) were assessed by FFU assay. The data are presented as mean ± S.D. of values from three independent experiments. Asterisks indicate significant differences among cell types at the specified time point (* P < 0.05; ** P < 0.01; *** P < 0.001; **** P < 0.0001). Levels of intracellular NS3 (E), secreted NS1 (F), extracellular virion E and C proteins (G) and cell-associated DENV E/C antigens (H) were evaluated by Western blotting analysis. Cell lysates containing equal amounts of proteins (15 μg/sample) or equal volumes of supernatants (10 μl/sample) were used for Western blotting. GAPDH was used as endogenous protein control for cell lysates.

Fig 3. Productive infection of all four DENV serotypes, both laboratory and clinically isolated strains, in imHC.

imHC were infected with four different DENV serotypes (DENV-1, DENV-2, DENV-3 and DENV-4) of either laboratory (A, B and C) or recently isolated clinical strains (D, E and F) at MOI of 0.1 for 48 h. The cell culture supernatants were subjected to the FFU assay to determine the levels of viral production (A and D). The percentages of DENV-infected cells were analyzed by immunofluorescent staining of cytoplasmic NS1 (B and E) and envelope (C and F) proteins, followed by flow cytometry. Data are presented as mean ± S.D. of values from three independent experiments.

Fig 4. Viability of imHC and the hepatoma cell lines over the time course of DENV-2 infection.

The cells were cultured in 24-well plates and mock-infected or infected with different MOIs of DENV-2 (0.1, 1 and 5), and then assessed for their viability at the indicated time points post-infection. (A) Graphs showing percentages of cells that had positive staining for annexin V but negative staining for propidium iodide (AnV+/PI-) (early apoptotic cells). (B) Graphs showing percentages of cells that were positively stained for both AnV and PI (AnV+/PI+) (late apoptotic cells). (C-D) Graphs showing percentages of apoptotic cells with positive staining for caspase-3/7 activation. All data are presented as mean ± S.D. of values from three independent experiments. An asterisk (*), “#” symbol and “+” symbol indicate significant differences between imHC vs HepG2, imHC vs Huh-7, and Huh-7 vs HepG2, respectively (* P < 0.05; ** P < 0.01; **** P < 0.0001).

Fig 5. Cytokine expression of hepatocytes upon DENV-2 infection.

The three hepatic cell lines and PHHs from 2 different donors (D1 and D2) were mock-infected (M) or infected with DENV-2 (DV) at MOI of 5. At the specified time points, cells and culture supernatants were harvested for assessment of cytokine expression. (A-I) A heatmap of differential cytokine protein production by hepatic cell lines and PHHs was quantified by Bio-Plex Pro Human Cytokine 27-plex assay. The numbers shown in the heatmap are average amounts (pg/ml) of secreted cytokines from two experiments of hepatic cell lines or two donors for PHHs. An equal number of each cell type was used for RNA extraction and an equal amount of RNA from each source was further subjected to RT-PCR and gel electrophoresis. The three hepatic cell lines and PHHs from 2 different donors (D1 and D2) were mock-infected (M) or infected with different MOIs (0.1, 1 and 5) of DENV-2. The cells were harvested for RT-PCR analysis of TRAIL, IFN-α, and IFN-β and a representative gel, showing RT-PCR products corresponding to these cytokines, is shown in J.

Fig 6. Changes in lipid droplet areas in imHC and hepatoma cell lines upon DENV infection.

(A) Representative images from the confocal microscopic analysis of lipid droplets stained with BODIPY 493/503 in the three cell lines cultured for 48 h under mock-infected (top panel) or DENV-infected conditions (MOI of 0.1) (bottom panel). Nuclei were stained with Hoechst 33342. Scale bar = 10 μm. (B) Area per lipid droplet of each cell type pre- and post-DENV infection was further analyzed using Image J software. The data shown are the mean ± S.D. values of the areas of lipid droplets analyzed from a total of 300–400 cells for each cell type. Asterisks indicate statistically significant differences (P < 0.05) of area per lipid droplet among the three cell types pre-infection as well as between the mock and DENV-infected conditions of each cell type.

Fig 7. LC-ESI-MS/MS analyses for quantification of triacylglycerols (TAGs), major components of lipid droplets, in three hepatic cell lines before and after DENV infection.

(A) Multiple neutral loss scans of ammoniated TAGs in hepatocyte lipid extracts. Representative MS/MS spectra from neutral loss scanning of 273, 271, 299, and 301, corresponding to the four common fatty acyl chains with C16:0, C16:1, C18:1, and C18:0, respectively, of the imHC extract. The major m/z signals from each scan (indicated by the larger m/z labeling) were selected for TAG species assignment, and quantification of the relative abundances of TAGs with different fatty acyl compositions (see S1 Table for details). (B-C) Changes of TAG relative abundances in the three hepatic cell lines upon DENV infection. Relative abundances of TAGs were quantified based on the peak areas of individual m/z signals from each neutral loss scan and normalized by the peak area of m/z 734 (DPPC–a major PC) in the same sample. The ratios to DPPC of selected TAG signals from the same neutral loss scan were combined and the sum of these ratios was used for comparison of TAGs containing different types of fatty acids (C16:0, C16:1, C18:0 and C18:1) between mock- and DENV-infected conditions of each cell type (B). The relative abundances of individual molecular species of TAGs with C16:0 in the three hepatic cell lines upon DENV infection are also shown (C). Data shown in B and C panels are the mean ± S.D. of values obtained from three sets of biological samples. Asterisks indicate statistically significant changes (P < 0.05) of TAG levels in imHC after DENV infection.

Fig 8. Differential inhibitory effects of anti-viral drugs on DENV replication in three hepatic cell lines.

The three cell lines (HepG2, Huh-7, imHC) were infected with DENV-2 at MOI of 0.1 and cultured in media with or without drugs [5 μM of ivermectin (IVER) or 10 μM of medium-soluble ribavirin (RBV)]. Plain medium or medium containing 0.5% DMSO (used for ivermectin solubilization) was also used as control vehicles for ribavirin and ivermectin treatment, respectively. At 48 h post infection, the cells and culture supernatants were collected for assessment of drug effects on viral production and DENV antigen expression as follows: (A-C) Levels of indicated DENV antigens were determined by Western blotting analysis and GAPDH was used as an internal control protein, (D and G) Percentages of cells with cytoplasmic NS3 expression determined by flow cytometry, (E and H) NS1 secretion levels in culture supernatants quantified by ELISA, (F and I) Virion levels in culture supernatants determined by FFU assay. The data from drug treatment conditions of different cell types were calculated as percentages of their corresponding control vehicles. All data are presented as mean ± S.D. of values from three independent experiments. Asterisks indicate significance differences between conditions or cell types (* P < 0.05; ** P < 0.01; *** P < 0.001; **** P < 0.0001).