Fig 1.
Induction of hepcidin gene expression and intracellular iron accumulation by HCV infection.
Iron concentrations in HCV-infected Huh7.5.1 cells 3 days post-infection (dpi) at MOIs of 0, 0.5, and 3 were evaluated through (A) measurement of Fe2+ and Fe3+ in cell lysates using the Nitroso-PSAP method and (B) measurement of relative intensity corresponding to Fe2+ using the fluorescent probe FeRhoNox-1. (B) For each group, cytoplasmic luminescence intensities of cells stained with the fluorescence probe were measured in Huh7.5.1 cells and the relative intensities were graphed with the average intensity of the uninfected-cell group as 1. Representative images of the cells used to measure luminescence intensities are shown at the bottom of the graph. (C) mRNA expressions of hepcidin, FPN1, DMT1, TfR1, TfR2 and ferritin in infected cells 3 dpi (MOI = 0.5, HCV+) and cells without infection (HCV-) were analyzed using qRT-PCR. (D) Time-course changes in hepcidin gene expression induced by HCV infection at MOIs of 1 and 3 were determined. At 12, 24, and 48 hpi, hepcidin mRNA was quantified as above. (E) In long-term cultures of infected cells, hepcidin mRNA was measured at 7, 14, and 21 dpi at MOI of 0.5. Values obtained from mock-infections at each point were set as 1. Intracellular iron concentrations were determined by the Nitroso-PSAP. (A)-(E) Results represent the means with SD from three independent measurements. Student’s t test; *P<0.05, **P<0.01, ***P <0.001.
Fig 2.
Role of CREBH in hepcidin and BMP6 expression induced by HCV infection or protein expression.
(A) Nuclear fractions from infected cells 3 dpi (MOI = 0.5) and cells without infection were used for ChIP assays. Protein-DNA complexes were immunoprecipitated with antibodies against either SMAD1 or CREBH or with rabbit IgG, analyzed by qPCR using three primer sets: Hepcidin pr ChIP set #1-#3. The sites potentially recognized by CREBH or BMP6 are indicated. (B) BMP6 mRNA in cells with and without HCV infection was analyzed using qRT-PCR. (C) BMP6 promoter reporter (BMP6pr) or basal reporter (EV; empty vector) was transfected into cells with HCV infection at 1 dpi and without infection. Firefly luciferase activities normalized by Renilla luciferase in cells at 2 days post-transfection (dpt) were determined. (D) ChIP assays and immunoprecipitations were performed as described in (A) with anti-CREBH antibody against or rabbit IgG. Precipitates were analyzed for qPCR using two primer sets: BMP6 pr ChIP set #1 and #2. The CREBH-binding sites are indicated. (E) CREBH-KO and parental Huh7.5.1 cells were cultured with or without HCV infection for 3 days, after which the mRNA of hepcidin (left) and BMP6 (middle) were quantified and the iron concentration (right) was measured by the Nitroso-PSAP. (F) Cells were transfected with a plasmid expressing Core-NS2 or NS3-5B or an EV. At 2 dpt, mRNAs of hepcidin, BMP6, and GAPDH were quantified. (G) Eight-week-old male WT and Creb3l3−/− mice were intravenously injected with AdV expressing Cre recombinase (AdEFNCre) together with AdV expressing either GFP (AdEFLNLGFP) or Core-NS2 (AdEFLNLHCVCore/NS2). At 5 dpi, total RNAs from liver tissues were extracted, followed by determining hepcidin and BMP6 mRNAs. Expression of HCV RNA was detected by RT-PCR. (B, E-G) Results represent the means with SD from three independent measurements. Student’s t test; *P<0.05, **P<0.01, ***P <0.001.
Fig 3.
Hepcidin-25 measurement in culture supernatants with HCV infection and in sera from hepatitis C patients.
(A) Hepcidin-25 concentrations in the supernatants of HCV-infected CREBH-KO cells and parental cells at 3 dpi at MOIs of 0, 0.2, and 1 were measured by the LC-MS/MS. As a positive control (PC), culture supernatants of cells transfected with the hepcidin-expressing plasmid were collected at 2 dpt and used. Expression of HCV Core and GAPDH in cells was assessed. (B) Changes in serum hepcidin-25 concentrations with decreasing viral loads in hepatitis C patients were analyzed. Paired serum samples from 26 hepatitis C patients were collected before (HCV positive) and after DAA therapy (HCV negative), and serum hepcidin-25 concentrations before vs. after viral elimination were compared. Significant differences between two groups before and after therapy were analyzed using Wilcoxon rank sum tests (***P<0.001).
Fig 4.
Proteolytic cleavage of FPN1 mediated by HCV NS3-4A protease.
(A) Huh7.5.1 cells with or without HCV infection 1 day after transfection of the FPN1-expression plasmid were harvested at 1 dpi and reacted with anti-FPN1 antibody. The percentage of FPN1-positive cells was counted by flow cytometry. Cells transfected with an EV (nega cont) were used as a negative control. The figure is representative of three experiments with similar results. (B, D) At 1 dpi, cells with or without HCV infection were transfected (B) with the FPN1-expression plasmid together with the NS3-4A expression plasmid or the empty vector, (D) with plasmids expressing FPN1 or FPN1 C326S. At 2 dpt, FPN1, HCV Core, and GAPDH were analyzed by western blotting. (C) FPN1 and GAPDH were analyzed by western blotting with cell lysates transfected with both plasmids expressing FPN1 or FPN1 C326S and Core-NS2, NS3-5B, NS3-4A, or NS3-4A S139A at 2 dpt. (E) HCV-infected cells 3 dpi were immunoprecipitated with anti-FPN1 antibody (BMP033) or rabbit IgG, followed by western blotting using anti-FPN1 antibody (NBP1-21502). HCV Core and GAPDH in input samples were assessed by western blotting without immunoprecipitation. A cell lysate transfected with plasmids expressing FPN1 and NS3-4A was used as a positive control (PC). Endogenous full-length- and processed FPN1 are indicated as an arrowhead and arrow, respectively. (F) MBP-FPN1 (aa 278–292)-sfGFP-FLAG fusion protein as a substrate was reacted with recombinant NS3-4A. After SDS-PAGE and transfer to a PVDF membrane, the processed band stained with CBB (left) was excised, followed by N-terminal analysis. Processing of the substrate fusion protein was confirmed by western blotting using anti-FLAG antibody (right). (G-I) FPN1, HA-tagged NS3-4A, and GAPDH were analyzed by western blotting. Cells were transfected with (F) the plasmid expressing WT- or mutated FPN1 together with the expression plasmid for HA-tagged NS3-4A or EV, (H) the FLAG-FPN1-expressing plasmid together with the HA-NS3-4A expressing plasmid from genotypes 1b, 2a, or 3a, and (I) the plasmids expressing FLAG-FPN1 and HA-NS3-4A in the presence or absence of 1 μM asunaprevir. (B, D, H, I) MG132 and NH4Cl were added at final concentrations of 10 μM and 10 mM, respectively, 12 h before cell harvesting.
Fig 5.
Iron store levels in mouse liver tissues and in cultured cells expressing HCV proteins.
(A) Eight-week-old male WT mice were intravenously injected with AdV expressing either GFP or NS3-4A genotype 1b. At 5 dpi, liver tissues were subject to western blotting to detect FPN1 and its cleaved form (red arrow) and HCV NS3 (left). Hepalclc7 cells transfected with the NS3-4A-expressing plasmid or the EV (right). (B) Iron concentrations in cells transfected with the plasmid expressing FPN1 together with the expression plasmid for NS3-4A WT or S139A mutant or EV were measured using the PSAP-method. Mean values obtained from transfectants with the FPN1-plasmid and EV were set as 100%. Results represent the means with SD from three independent measurements. Student’s t test; *P<0.05, **P<0.01. (C) Four groups of mice (N = 4 each) were created in which recombinant adenoviruses were injected as follows: GFP group as a negative control; AdEFLNLGFP, AdEFNCre and AdEFGFP, Core-NS2 group; AdEFLNLHCVCore/NS2 and AdEFNCre, NS3-4A group; AdEFHCVNS3/4A, Core-NS2 and NS3-4A group; AdEFLNLHCVCore/NS2, AdEFNCre and AdEFHCVNS3/4A. Thin section slices were prepared from livers at 5 dpi and stained with Berlin blue. Representative images of two mouse livers from each group are shown. (D) Elemental analysis was performed with the NanoSuit method at the six spots positive for staining (Blue staining positive) and unstained parts (Background) in slices from the Core-NS2 group described in S13 Fig. (Left) Elemental composition graphs of energy (keV) vs. cps per eV are shown. Areas including the Fe peaks in the sample positive for blue staining and the corresponding area in the background sample are expanded. (Right) The weight concentrations of Fe measured were quantified and are expressed as weight percentages (wt%). Student’s t test; **P<0.01. (E) Total iron scores of 10 fields of view in the livers from the four groups described in (C) were individually evaluated by means of histological hepatic iron index. Results represent the plots separately for each group. Welch’s t test; *P<0.05, **P<0.01, ***P<0.001, ns P>0.05.
Fig 6.
Working model of iron accumulation induced in HCV-infected hepatocytes.
Much of the replication process of HCV is dependent on the ER of the host cell; in HCV-infected cells, a viral precursor polyprotein is cleaved into 10 proteins by ER-resident peptidases and viral proteases associated with the ER membrane. ER-derived membranes also play an important role in viral replication complex formation. These characteristics of the HCV lifecycle induce ER stress in infected cells. This activates the ER resident transcription factor CREBH, which in turn upregulates the transcription both of BMP6 and hepcidin through its recognition of the promoters. The induction of BMP6 expression enhances BMP signaling, and the Smad complex, which is activated as a result, also plays a role in activation of the hepcidin promoter. The induction of hepcidin gene expression leads to an increase in extracellular hepcidin peptide levels, which in turn increases hepcidin-FPN1 association, thus promoting retrograde transport of FPN1 into the cell and ubiquitin-dependent degradation. The NS3-NS4A complex, whose serine protease activity is essential for processing the viral precursor, can cleave the intracytoplasmic portion of the central region of FPN1, which is presumably suppressed in its iron efflux function upon cleavage.