Hepatitis E virus persists in the presence of a type III interferon response

Abstract

The RIG-I-like RNA helicase (RLR)-mediated interferon (IFN) response plays a pivotal role in the hepatic antiviral immunity. The hepatitis A virus (HAV) and the hepatitis C virus (HCV) counter this response by encoding a viral protease that cleaves the mitochondria antiviral signaling protein (MAVS), a common signaling adaptor for RLRs. However, a third hepatotropic RNA virus, the hepatitis E virus (HEV), does not appear to encode a functional protease yet persists in infected cells. We investigated HEV-induced IFN responses in human hepatoma cells and primary human hepatocytes. HEV infection resulted in persistent virus replication despite poor spread. This was companied by a type III IFN response that upregulated multiple IFN-stimulated genes (ISGs), but type I IFNs were barely detected. Blocking type III IFN production or signaling resulted in reduced ISG expression and enhanced HEV replication. Unlike HAV and HCV, HEV did not cleave MAVS; MAVS protein size, mitochondrial localization, and function remained unaltered in HEV-replicating cells. Depletion of MAVS or MDA5, and to a less extent RIG-I, also diminished IFN production and increased HEV replication. Furthermore, persistent activation of the JAK/STAT signaling rendered infected cells refractory to exogenous IFN treatment, and depletion of MAVS or the receptor for type III IFNs restored the IFN responsiveness. Collectively, these results indicate that unlike other hepatotropic RNA viruses, HEV does not target MAVS and its persistence is associated with continuous production of type III IFNs.

Fig 1. HEV induces a type III-predominant IFN response in human hepatocytes.

(A) HepG2 cells were inoculated with HEV (1,000 HEV genome equivalents (GE)/cell) and stained with anti-ORF2 antibody at different times after inoculation. Cells were counterstained with DAPI for DNA. Scale bar: 100 μm. (B) Kinetics of HEV replication and IFN secretion responses in HepG2 cells. HEV RNA was measured by quantitative RT-PCR using primers targeting either ORF1 (representing the full-length HEV genome) or ORF2/3 (representing the sum of the full-length and the subgenomic HEV RNA). Concentrations of different IFNs (IFN-α, IFN-β, and IFN-λ) in the culture supernatants were measured by specific ELISA. Only IFN-λ protein was detected and, therefore, shown. Dotted line denotes the detection of limit of IFN-λ (15.6 pg/ml). The results show the mean ± SEM of the average of the duplicates in each of 2 independent experiments. (C) Kinetics of different IFN and ISG mRNA expression in HEV-infected HepG2 cells. Data are expressed as log₁₀ fold change relative to mock-infected cells. (D-E) Kinetics of HEV replication and IFN production in HEV-infected primary human hepatocytes (PHHs). PHHs from two different donors (HH1086 and HH1076) were inoculated with HEV (10⁴ GE/cell), and at different days post inoculation stained with chimpanzee immune serum (ch1313) and DAPI (top panels). Cells inoculated with irradiated HEV did not produce positive signal. Scale bar, 100 μm. Intracellular HEV RNA and supernatant IFNs were quantified by qRT-PCR and ELISA, respectively (bottom panels). IFN-α and IFN-β proteins were undetectable. The error bars indicate SEM of results from duplicate wells. (F) HEV (ORF1) RNA levels in two independently created clones of HepG2 replicon cells and corresponding replicon-cured cells. The results show the mean ± SEM of 2 independent experiments. (G-H) mRNA levels of IFNs and ISGs in HepG2 cells, HepG2 replicon cells and replicon-cured cells. Data are expressed as fold changes relative to the parental cells. The results show the mean ± SEM of 2 independent experiments performed in duplicate each.

Fig 2. Type III IFN response regulates HEV replication.

(A) Immunoblots of IFNAR1, IFNLR1 and β-actin in HepG2 cells transduced with lentiviruses expressing gene-specific shRNA or GFP (Ctrl). (B) ISRE promoter activity in different cells after IFN-α or IFN-λ treatment. HepG2 cells were transfected with an interferon-sensitive response element (ISRE) promoter-driven firefly luciferase reporter plasmid ISRE-Luc and a herpes simplex virus thymidine kinase promoter-driven Renilla luciferase (TK-RLuc) plasmid for 24 h, then treated with IFN-α (100ng/ml) or IFN-λ1 (220ng/ml) for 24 h. Data are expressed as fold changes relative to non-treated cells based on the relative luciferase activities (firefly luciferase vs. Renilla luciferase). Shown are representative results (mean ± SEM) from two independent experiments each performed in triplicate. (C-E) Effects of IFN receptor knockdown on ISG expression and HEV replication. HepG2 cells transduced with lentiviruses expressing gene-specific shRNA or GFP (Ctrl) were infected with HEV for 5 days. (C) Intracellular IFIT1 mRNA expression determined by qRT-PCR. Results are represented as fold changes relative to HEV-infected control cells. (D) HEV RNA abundance determined by qRT-PCR (fold changes relative to HEV-infected control cells). (E) Cells expressing different shRNA or GFP (Ctrl) were infected with HEV for 5 days, harvested and subjected to three rounds of freeze-thaws prior to inoculation to naïve HepG2 cells. Infected cells were detected by IFA and HEV foci were counted after 5 days. Each data point represents the mean ± SEM of at least 2 independent experiments in duplicate each. * P<0.05; ** P<0.01.

Fig 3. Signaling pathways involved in HEV-induced IFN response.

Immunoblots of RIG-I, MDA5, MAVS and β-actin in HepG2 cells expressing gene-specific shRNA, or GFP (Ctrl). (B-C) IFN-β promoter activity in HepG2 cells with different gene knockdown following Sendai virus (SeV) infection (B) or poly IC transfection (C). Cells were transfected with IFN-β-Luc and TK-RLuc (for normalization of transfection efficiency) 20 h prior to SeV infection or poly IC transfection. Cells were lysed and luciferase activity was determined 20 h after SeV infection or 12 h after poly IC transfection. Data are presented as fold changes relative to non-treated cells. Shown are representative results from two independent experiments each performed in triplicate. (D-F) Effect of RIG-I, MDA5 or MAVS knockdown on HEV replication and host IFN responses. Control and shRNA-expressing HepG2 cells were inoculated with HEV (1000 GE/cell). IFN-λ mRNA expression (D), HEV RNA abundance (E), and HEV-positive foci (F) were determined at 5 days after infection. The results show the mean ± SEM of 2 independent experiments performed in duplicate each. * P<0.05; ** P<0.01. (G) Immunoblots of IRF-3, IRF-7 and β-actin in HepG2 cells transduced with lentiviruses expressing GFP (Ctrl) or gene-specific shRNA. (H-J) Effect of IRF-3 or IRF-7 knockdown on HEV replication and IFN responses. Control and shRNA-expressing cells were inoculated with HEV (1,000 GE/cell). IFN-λ mRNA expression (H), HEV RNA abundance (I), and HEV-positive foci (J) in different cells were determined after 5 days. The results show the mean ± SEM of 2 independent experiments each performed in duplicate wells. * P<0.05; ** P<0.01. Scale bar (upper panel in F), 100 μm.

Fig 4. HEV does not target MAVS.

(A) Confocal images showing MAVS and viral antigens in HepG2 cells infected with either HEV (top) or HAV (bottom). Cells were stained 5 days after infection with a rabbit anti-MAVS, chimpanzee 1313 serum (HEV), or a murine monoclonal antibody K24F2 (HAV). DAPI was used to stain the nuclei. Scale bar: 10 μm. (B) Confocal images showing the mitochondrial localization of MAVS in HepG2 cells with or without HEV replicon. MAVS was stained with a rabbit antibody against MAVS (green). Mitochondria was visualized with MitoTracker (red). Nuclei were stained with DAPI. Scale bar: 10 μm. (C) HepG2 cells with or without the HEV replicon were transfected with a MAVS-expressing plasmid along with a HAV 3ABC-expressing plasmid or an empty vector. The endogenous (closed arrowheads) and overexpressed MAVS (open arrowheads) were detected with a rabbit anti-MAVS antibody. Note that co-expression of HAV 3ABC led to the degradation of MAVS. (D) HepG2 cells with or without the HEV replicon were transfected with poly IC (6 μg/ml). After 6 h, cells were lysed and subjected to Western blot analysis using antibodies against MAVS and β-actin. Crude mitochondria were isolated and subjected to SDD-AGE for detection of MAVS polymer. (E-F) HepG2 cells with or without the HEV replicon were transfected with poly IC (6 μg/ml). After 6 h, intracellular IFN mRNA levels were measured by qRT-PCR (E) and supernatant IFN-λ concentration was measured by ELISA (F). In panel E, data are expressed as fold changes relative to mock transfected cells containing no replicon, and the results show the mean ± SEM of 2 independent experiments. (G) Immunoblots of endogenous RIG-I, MDA5, ISG56, MAVS, and β-actin in HepG2 cells, replicon cells or replicon-cured cells following poly IC transfection (1.5 μg/ml, 12 h). (H-I) HepG2 cells with or without the HEV replicon were transfected with the hepatitis C virus (HCV) 3’UTR RNA. After 6 h, intracellular IFN mRNA levels were measured by qRT-PCR (H), and concentrations of supernatant IFN-λ were measured by ELISA (I). The results show the mean ± SEM of 2 independent experiments. (J) Immunoblots of endogenous RIG-I, MDA5, ISG56, MAVS, and β-actin in HepG2 cells with or without the replicon before or after HCV 3’UTR RNA transfection (3.6 μg/ml, 14 h).

Fig 5. HEV-infected cells are refractory to exogenous IFNs due to basal activation of the JAK/STAT signaling.

(A) Comparison of antiviral effect of IFN-α and IFN-λ on HCV and HEV replication. Top, a schematic diagram of HCV and HEV constructs used in the experiment. Bottom, Huh-7 cells were transfected with HCV (H77S3/Gluc2A) or HEV (Kernow-C1 p6/luc) RNA. After 24 h, cells were treated with IFN-α or IFN-λ at indicated concentrations. Supernatant Gluc activity was measured 4 days after transfection. Data were presented as percent inhibition relative to the non-treated controls. Shown are the representative results from 2 independent experiments each performed in triplicates. Error bars denote S.E.M. (B) Huh-7 cells were transfected with HEV (p6/luc) RNA and treated with IFN-α (100 ng/ml) at indicated times. Data were presented as percent inhibition relative to the non-treated controls. (C) Different HepG2 cell lines were transfected with ISRE-luc and TK-RLuc (an internal control for transfection efficiency). 24 h post transfection, cells were treated with IFN-α (100 ng/ml) or IFN-λ1 (220 ng/ml). Cells were harvested 24 h later for luciferase assays. Data are expressed as fold changes relative to non-treated cells based on the relative luciferase activities (firefly luciferase vs. renilla luciferase). The results show the mean ± SEM of 2 independent experiments. (D) Antiviral effects of IFN-α and IFN-λ in the HepG2 replicon cells with or without MAVS. Cells were treated with indicated doses of recombinant IFN-α or IFN-λ proteins. After 3 days, intracellular RNA levels were measured by qRT-PCR using primers targeting HEV ORF1. The results (mean ± SEM of 2 independent experiments) show the percentage of inhibition in HEV replication relative to non-treated cells. * P<0.05; ** P<0.01. (E) Immunoblots of total and phosphorylated STAT1, ISG56, and β-actin protein levels in different HepG2 cells treated with recombinant IFN-α (100 ng/ml) or IFN-λ (220 ng/ml) for various times. Lower panels show fold changes in total STAT1 proteins and pSTAT1/total STAT1 ratios relative to untreated parental cells. (F) Immunoblots of total and phosphorylated STAT1, MAVS, and β-actin protein levels in different HepG2 cells treated with IFN-α (100 ng/ml) or IFN-λ (220 ng/ml) for 1 h. Lower panel shows fold changes in total STAT1 protein levels and pSTAT1/total STAT1 ratios relative to untreated parental cells.