Abstract
Dengue virus (DENV), a common and widely spread arbovirus, causes life-threatening diseases, such as dengue hemorrhagic fever or dengue shock syndrome. There is currently no effective therapeutic or preventive treatment for DENV infection.
Next-generation sequencing analysis revealed that prostasin expression was decreased upon DENV infection. Prostasin expression levels were confirmed by real-time quantitative polymerase chain reaction in patients with dengue fever and a DENV-infected mice model. Short hairpin RNA against EGFR and LY294002 were used to investigate the molecular mechanism.
Based on clinical studies, we first found relatively low expression of prostasin, a glycosylphosphatidyl inositol-anchored membrane protease, in blood samples from patients with dengue fever compared with healthy individuals and a high correlation of prostasin expression and DENV-2 RNA copy number. DENV infection significantly decreased prostasin RNA levels of in vivo and in vitro models. By contrast, exogenous expression of prostasin could protect ICR suckling mice from life-threatening DENV-2 infection. Mechanistic studies showed that inhibition of DENV propagation by prostasin was due to reducing expression of epithelial growth factor receptor, leading to suppression of the Akt/NF-κB–mediated cyclooxygenase-2 signaling pathway.
Our results demonstrate that prostasin expression is a noteworthy clinical feature and a potential therapeutic target against DENV infection.
Dengue virus (DENV) is an arthropod-transmitted virus and a serious public health problem in tropical and subtropical regions of the world. DENV is divided into 4 serotypes on the basis of antigenetic distinction [1]. Primary infection by DENV is often asymptomatic or else results in self-limiting dengue fever (DF) [2]. Secondary infection with different serotypes may increase the risk of lethal DENV diseases [3]. Vaccines against dengue have a limited efficacy of only 64.7% [4, 5]. No effective antiviral treatment for the prevention and control of DENV infection is available.
Prostasin is a glycosylphosphatidyl inositol (GPI)-anchored membrane protease in most epithelial tissues [6, 7]. Prostasin has multiple functions in epithelial physiology, including the suppression of tumor invasiveness in vitro and attenuation of inflammation-induced gene expression [8, 9]. In addition, prostasin is implicated in the modulation of epithelial growth factor receptor (EGFR) signaling through the formation of the active prostasin-matriptase complex in the proteolytic cleavage of the EGFR extracellular domain [10]. EGFR is a cell-surface receptor that belongs to the erythroblastic leukemia viral oncogene family of receptor tyrosine kinases [11]. It is also a critical factor in the upstream of a complex signal transduction pathway for modulation of cell proliferation, adhesion, migration, survival, and differentiation [12]. EGFR signaling is crucial to the viral life cycle upon infection. For example, hepatitis C virus (HCV)-induced EGFR activation is required for viral internalization and entry [13]. By contrast, inhibition of EGFR activation by the kinase inhibitor erlotinib suppresses HCV infection [14]. In addition, enterovirus 71 modulates an EGFR/cyclooxygenase-2 (COX-2)/prostaglandin (PGE2)-dependent signaling pathway to facilitate viral replication [15]. The COX-2/PGE2 signaling cascade is required for the propagation of other viruses and is a potential target in the treatment of viral diseases, such as HCV and Japanese encephalitis virus [16, 17]. We recently found that COX-2 expression also facilitates DENV replication and that COX-2 inhibition can block DENV replication [18]. To date, many studies have revealed that prostasin downregulates EGFR expression to decrease the expression of COX-2 inflammatory response genes [8, 19]. Therefore, the correlation between the prostasin-mediated EGFR/COX-2 signaling pathway and DENV replication deserves the attention of researchers.
In this study, we first analyzed the expression levels of prostasin in clinical samples and in a DENV-infected mouse model. Then, we overexpressed prostasin in vitro and in vivo to evaluate the effect of prostasin on DENV infection. We used short hairpin RNA (shRNA) against EGFR and the specific inhibitor LY294002 to investigate the molecular mechanism that underlies the involvement of prostasin-mediated EGFR in DENV replication/propagation through the downregulation of the COX-2 signaling pathway.
MATERIALS AND METHODS
Ethics Statement
All patients with DF were followed the diagnosed criteria defined by the 2009 WHO guidelines and Centers for Disease Control, Department of Health (Taiwan). Patients’ personal information, clinical data, and medical records were collected, and the study was approved by the Institutional Review Board of Kaohsiung Medical University Hospital (IRB number: KMUH-IRB-20110451). All informed consent were obtained from patients diagnosed with DF. Breeder mice of the Institute of Cancer Research (ICR) strain were obtained from BioLasco Taiwan Co. Ltd, and 6-day-old suckling mice were used in the present study. The experimental protocol was approved by the Animal Research committee of Kaohsiung Medical University of Taiwan (IACUC number 104227) following the guideline of the Public Health Service policy on Human Care and Use of Laboratory Animals.
Prostasin Expression in Healthy Donors and Patients With DF
A laboratory-confirmed dengue patient was defined as a patient who had at least 2 positive results from 4 experimental examinations, including a real-time quantitative polymerase chain reaction (RT-qPCR) assay, dengue NS1 Ag STRIP test, capture IgM and IgG enzyme-linked immunosorbent assay, and dengue viral culture assay. Whole-blood cell samples were obtained from the patients at the febrile phase and were used for the analysis of prostasin expression. All samples were maintained at −80°C. Serum samples from healthy donors were used in the comparison of prostasin expression.
Cell Culture, Virus, and Reagent
Human hepatic Huh-7 cells and human monocytic THP-1 cells were cultured as previously described [18]. DENV-2 (strains 16681 and PL046) was propagated in C6/36 cells in accordance with the protocol of a previous study [20]. Ly294002 [21] and GW5074 [22] were purchased from Sigma-Aldrich (St. Louis, MO). The chemical agents were prepared at 100 mM as a stock solution in 100% DMSO. The final concentration of DMSO in all experiments was maintained at 0.1%.
DENV-2 Infection in ICR Suckling Mice
ICR suckling mice were infected with DENV-2 as previously described [18, 23]. Briefly, 6-day-old ICR suckling mice were intracerebrally injected with 2.5 × 105 plaque-forming units (PFU) of 60°C heat-inactivated DENV-2 PL046 (iDENV) or DENV-2 (DENV). At 6 days postinfection (dpi), brain tissue was collected for the quantification of prostasin expression through RT-qPCR. Six-day-old ICR suckling mice were randomly divided into 3 groups and infected with 2.5 × 105 PFU of iDENV or 2.5 × 105 PFU of DENV-2: group 1, injected with iDENV; group 2, injected with 105 PFU of adenovirus-green fluorescent protein (Ad-GFP) for 2 days and then infected with DENV-2; group 3, injected with 105 PFU of Ad-prostasin for 2 days and then infected with DENV-2. Each group comprised 10 suckling mice. The body weight and survival rate of the mice were measured daily after DENV-2 infection.
DENV-Infection in AG129 Mice
AG129 (129/Sv mice with interferon alpha/beta and gamma receptor deficiency) were obtained from B&K Universal (Hull, UK). They were maintained under specific pathogen free condition in ventilated cages. Eight to 10-week-old mice were injected with 104 plaque forming units (PFU) of heat-inactivated DENV-2 or DENV-2 via intravenous route. At 4 dpi, we collected blood cells to analyze prostasin levels.
Western Blotting
Cell lysates were collected with RIPA lysis buffer, and western blotting was performed as previously described [24]. Protein-transferred membranes were probed with specific antibodies against DENV NS2B (GeneTex, Irvine, CA), glyceraldehyde-3-phosphate dehydrogenase (GAPDH; GeneTex, Irvine, CA), prostasin (Abcam, Cambridge, MA), COX-2 (Cayman, ML), EGFR (Abcam, Cambridge, MA), Myc (Abcam, Cambridge, MA), and hemagglutinin (HA) tag (GeneTex, Irvine, CA).
RT-qPCR
Total cellular RNA was extracted using a total RNA miniprep purification kit (GeneMark Biolab, Taiwan) in accordance with the manufacturer’s instructions. Prostasin, COX-2 mRNA, and DENV-2 RNA levels were analyzed by RT-qPCR using specific primers. The DENV RNA copies were determined by absolute quantification for qPCR. The data were calculated using the standard curve method with the cloned target sequences as in the previous study [25].
Transfection and Luciferase Activity Assay
Huh-7 cells were cotransfected with 1 μg of pCOX-2–luciferase (Luc) and pCMV-prostasin-Myc at the indicated concentrations for 6 hours using T-pro reagent (Ji-Feng Biotechnology Co., Ltd., Taipei, Taiwan) in accordance with the manufacturer’s instructions. Huh-7 cells were transfected with 0.2 μg pCMV-Renilla-Luc served as the transfection efficiency control. Cell lysates were subjected to a luciferase activity assay using a Dual-Glo Luciferase Assay System (Promega). The shRNA expression vectors, shprostasin (pLKO.1-prss8, TRCN 0000047065), shGFP (pLKO.1-GFP, TRCN 0000072179), and shEGFR (pLKO.1-EGFR, TRCN 0000039634) were obtain from the National RNAi Core Facility, Academia Sinica, Taiwan.
Plaque Assay
BHK cells were seeded in a 12-well plate and infected with serially diluted virus for 2 hours of incubation. The viral inoculum was replaced with DMEM containing 0.8% methyl-cellulose (Sigma-Aldrich, St Louis, MO). The cells were then incubated for 5 days. The cells were fixed and stained with crystal violet solution for 1.5 hours and the virus titer was calculated.
Statistical Analysis
All data were presented as the mean ± SD. Quantification analysis was performed for 3 independent experiments, with at least triplicate samples for each experiment. Statistically significant differences between 2 data groups were determined using 1-way ANOVA or Student t test. Spearman correlation was used to analyze the correlation between prostasin and DENV RNA copies, and analysis was performed with Graphpad Prism (Graphpad Software, Inc. Version 6.0). A P value of < .05 was considered to indicate a statistically significant result.
RESULTS
Prostasin Expression Is Decreased in DENV-Infected Patients, Mice, Hepatoma Cells, and Monocytes
Next-generation sequencing (NGS) analysis revealed that prostasin expression was time-dependently decreased upon DENV infection relative to that in the DENV-uninfected hepatoma cell line (Supplementary Figure 1A). NGS results were further validated by RT-qPCR under identical experimental conditions (Supplementary Figure 1B). To determine the expression level of prostasin in clinical DENV infection, blood cell samples from 14 DENV-infected patients with DF and 8 healthy donors were analyzed by RT-qPCR. As shown in Figure 1A, prostasin expression was lower in DF patients than in the healthy donors. We further found that the relationship between prostasin and DENV RNA copy number showed a significant negative correlation (Figure 1B), indicating that this high correlation between prostasin expression and DENV RNA expression is a noteworthy clinical feature. Consistently, prostasin RNA levels in brain tissue and blood cells were decreased in ICR suckling mice (Figure 1C) and DENV-infected AG129 mice (lacking interferon-α/β and -γ receptors) (Supplementary Figure 1C), compared with those of iDENV-infected mice. The levels of gapdh mRNA (used as a control) were not significantly affected by DENV infection (Supplementary Figure 2). Initiation of DENV binding in the human hepatoma Huh-7 cells is heparan sulfate (HS)-dependent, and the cells are capable of DENV replication [26, 27]. Therefore, we infected Huh-7 cells with DENV to evaluate the inhibitory effect of DENV-2 infection on prostasin. Western blotting and RT-qPCR results for the cell-based experiments revealed that mRNA levels of prostasin decreased in DENV-2–infected Huh-7 cells (Figure 1D). In addition, we used THP-1 monocytes to confirm these important findings (Supplementary Figure 1D).
Dengue virus (DENV) infection decreases prostasin expression in dengue fever (DF) patients, DENV-infected mice, and human hepatoma cells. A, Downregulated prostasin expression in blood cell samples from DF patients. Prostasin mRNA levels in blood cell samples of 14 clinical DF patients and 8 healthy donors were determined through real-time quantitative polymerase chain reaction (RT-qPCR). B, Correlation between prostasin and DENV RNA copy number in DF patients. The levels of prostasin were correlated with DENV RNA copy number using Spearman rank correlation coefficient test. C, Downregulated prostasin RNA level in DENV-2–infected ICR suckling mice. DENV-2 or heat-inactivated DENV (iDENV) was intracerebrally injected into 6-day-old suckling mice at 2.5 × 105 plaque-forming units. Prostasin mRNA levels in mouse brain tissue at 6 days postinfection (dpi) were analyzed by RT-qPCR. Each group comprised 5 suckling mice (n = 5). D, Decreased prostasin mRNA levels in DENV-2–infected human hepatoma cells. Huh-7 cells were infected with DENV-2 at a multiplicity of infection of 1, and cell lysate and cellular RNA were extracted at 2 dpi. The RNA copies of prostasin were determined through absolute RT-qPCR. The relative RNA level of prostasin following normalization to the cellular gapdh mRNA level. The cell lysates were subjected to western blotting. All data from cell-based experiments are indicative of at least 3 independent experiments. Error bars represent mean ± SD of 3 independent experiments; *P < .05.
Prostasin Overexpression Decreases the Mortality Rate of DENV-Infected ICR Suckling Mice
To evaluate the biological value of prostasin upon DENV-2 infection in vivo, 6-day-old ICR suckling mice were intracerebrally injected with 108 PFU of recombinant Ad-GFP as a control or Ad-prostasin. The Ad-injected suckling mice were then intracerebrally injected with iDENV-2 or DENV-2 at 2 days after inoculation. We recorded the body weight and survival of all experimental mice daily for 6 days. As shown in Figure 2A, at 6 dpi, prostasin overexpression maintained the survival rate of DENV-2–infected mice at 60%. Furthermore, mice infected with Ad-prostasin had consistently higher body weights than those infected with Ad-GFP (Figure 2B). To evaluate the interference of prostasin in DENV-2 propagation, we extracted viral particles and protein from the brain tissue of suckling mice at 6 dpi. The viral titer in Ad-prostasin–infected mice decreased by 1.3 log relative to that in Ad-GFP–infected mice in the presence of DENV-2 infection (Figure 2C). Western blotting results also revealed that the protein levels of DENV-2 significantly decreased with prostasin overexpression (Figure 2D).
Prostasin overexpression protects mice from life-threatening Dengue virus (DENV) infection. Recombinant adenovirus expressing green fluorescent protein (Ad-GFP) or prostasin (Ad-prostasin) was intracerebrally injected into 6-day-old ICR suckling mice at a viral load of 108 plaque-forming units. The suckling mice were then intracerebrally injected with DENV-2 or heat-inactivated DENV (iDENV) at 2 dpi. The (A) survival rate and (B) body weights of the mice were recorded daily. Each group comprised 10 suckling mice (n = 10). C, Viral titer were determined through plaque assay and real-time quantitative polymerase chain reaction (RT-qPCR) at 6 dpi. D, The expression of prostasin and NS2B were identified through western blotting. Error bars represent the mean ± SD of 3 independent experiments; *P < .05. Abbreviation: PFU, plaque-forming units.
Prostasin Overexpression Attenuates DENV Propagation in Huh-7 and THP-1 Cells
Clinical observations and mouse model experiments have shown that monocyte and liver cells are the targets of DENV replication and pathogenesis [28]. Therefore, we first used Huh-7 cells to investigate the mechanism of antiviral action of prostasin against DENV replication. We transfected DENV-2–infected Huh-7 cells with various concentrations of a pCMV-prostasin-Myc expression vector. Western blotting and plaque assay results indicated that prostasin reduced DENV-2 protein synthesis and amount of DENV particles in a dose-dependent manner (Figure 3A and 3B). In addition, the inhibitory effect of prostasin on DENV replication was also observed in DENV-2–infected THP-1 cells (Supplementary Figure 3A). To confirm the inhibitory effect of prostasin overexpression on DENV replication, we cotransfected short hairpin GFP RNA (shGFP) or short hairpin prostasin RNA expression vector into prostasin-Myc–expressing Huh-7 cells with increasing plasmid concentrations of 0–1.5 μg/well. Then, all transfected cells were infected with DENV-2. As shown in Figure 3C, prostasin silencing retarded the reduction of DENV-2 protein levels (lane 4) relative to that in cells with exogenous prostasin-Myc overexpression (lanes 1–3). Additionally, downregulation of endogenous prostasin also could facilitate DENV-2 protein synthesis, revealing that prostasin is a potential suppressor of DENV-2 replication (Supplementary Figure 3B). Furthermore, prostasin also reduced viral replication of 4 serotypes of DENV (Figure 3D).
Prostasin overexpression decreases Dengue virus (DENV) replication and propagation. A and B, Exogenous prostasin expression decreased DENV-2 protein synthesis and viral propagation. Huh-7 cells were transfected with pCMV-prostasin-Myc at the indicated concentrations and then infected with DENV-2 at a multiplicity of infection (MOI) of 1. Cell lysate and supernatants were collected 2 days after transfection. C, Prostasin silencing restored the reduction of DENV-2 protein synthesis by exogenous prostasin expression. Huh-7 cells were cotransfected with pCMV-prostasin-Myc and green fluorescent protein short hairpin RNA (shGFP) or prostasin shRNA (shprostasin), and the transfected cells were infected with DENV-2 at a MOI of 1 for 2 days. D, Replication of 4 serotypes of DENV was inhibited by prostasin overexpression in a concentration-dependent manner. All data from cell-based experiments are indicative of at least 3 independent experiments. Error bars represent the mean ± SD of 3 independent experiments; *P < .05. Abbreviation: PFU, plaque-forming units.
Prostasin Overexpression Suppresses DENV Replication by Reducing COX-2 Expression
Prostasin overexpression has been reported to reduce COX-2 expression in PC-3, a human prostate carcinoma cell line [10]. We previously reported that COX-2 suppression can block DENV replication and propagation in vitro and in vivo [18]. To investigate the effect of prostasin on DENV-2–induced COX-2 expression, we transfected DENV-2–infected Huh-7 cells with various concentrations of pCMV-prostasin-Myc or the pcDNA4 plasmid vehicle. Western blotting results revealed that prostasin decreased DENV-2–induced COX-2 protein levels in a dose-dependent manner (Figure 4A). To investigate the effect of prostasin on DENV-2–induced COX-2 transcriptional activity, we cotransfected the pCOX-2-Luc plasmid, containing a COX-2 promoter-linked firefly luciferase reporter gene, and pCMV-prostasin-Myc into DENV-2–infected Huh-7 cells. The results of a reporter assay showed that prostasin decreased DENV-2–elevated COX-2 promoter activity relative to that in DENV-2–infected Huh-7 cells without prostasin overexpression (Figure 4B). To evaluate the contribution of prostasin-reduced COX-2 expression to the decrease in DENV-2 replication, we performed a restoration experiment on DENV-2 replication by increasing exogenous COX-2 expression in DENV-2–infected Huh-7 cells overexpressing prostasin. DENV-2–infected cells were cotransfected with 1.0 μg of pCMV-prostasin-His and increasing concentrations of pCMV-COX-2-Myc. As shown in Figure 4C, increased exogenous COX-2-Myc expression can gradually restore prostasin-reduced DENV-2 protein levels (lanes 3 and 4) relative to that in pcDNA4-transfected cells (lane 1) and prostasin-transfected cells (lane 2) in the presence of DENV-2 infection. Similar results were observed by quantifying the viral titer of the supernatant using a plaque assay (Figure 4D). The results collectively indicated that the suppression of COX-2 expression is an important mechanism by which prostasin blocks DENV-2 replication.
Prostasin reduces Dengue virus (DENV) replication by decreasing cyclooxygenase-2 (COX-2) expression. A, Prostasin overexpression reduced DENV-2–induced COX-2 expression in a dose-dependent manner. Huh-7 cells were transfected with pCMV-prostasin-Myc at the indicated concentrations and then infected with DENV-2 at a multiplicity of infection (MOI) of 1 for 2 days. Western blotting was performed with the indicated antibodies. B, Prostasin overexpression reduced DENV-2–elevated COX-2 promoter activity in a dose-dependent manner. Huh-7 cells were cotransfected with pCOX-2– luciferase (Luc) and pCMV-prostasin-Myc at the indicated concentrations and then infected with DENV-2 at a MOI of 1 for 2 days. The cell lysate was subjected to luciferase activity assay. C and D, Exogenous expression of COX-2 recovered the inhibitory effect of prostasin on DENV-2 protein expression and viral titer. Huh-7 cells were cotransfected with 1 μg pCMV-prostasin-His and increasing pCMV-COX-2-Myc concentrations and then infected with DENV-2 at a MOI of 1 for 2 days. All data from cell-based experiments are indicative of at least 3 independent experiments. Error bars represent the mean ± SD of 3 independent experiments; *P < .05. Abbreviation: PFU, plaque-forming units.
Prostasin Attenuates the Inhibitory Action of the EGFR-Mediated COX-2 Signaling Pathway on DENV Replication
Recently published studies revealed that EGRF/ErbB-2/ErbB-4 inhibitors reduced DENV replication, which indicated that EGFR activation may be a critical signaling pathway for DENV propagation [29]. Prostasin is involved in the extracellular proteolytic modulation of EGFR, an upstream regulator of COX-2 expression [30, 31]. Based on an increase in total EGFR and phospho-EGFR in response to DENV-2 infection (Supplementary Figure 4A), we suggested that the active EGFP may have an important role in DENV-2 replication. Therefore, we examined whether EGFR expression is required for DENV-2 replication by using shRNA against EGFR. We transfected DENV-2–infected Huh-7 cells with GFP or EGFR shRNA expression vector. The results showed that EGFR shRNA reduced DENV-2 protein synthesis in a dose-dependent manner (Figure 5A). The amount of DENV harvested from the supernatant of EGFR shRNA-transfected Huh-7 cells was measured by plaque assay, and the results showed a decrease in viral titer in the presence of EGFR shRNA expression (Figure 5B). To analyze the effect of EGFR activation on DENV-2 replication, DENV-2–infected Huh-7 cells were treated with EGF to stimulate the activation of EGFR. The results showed that elevated DENV-2 protein levels and viral titers were accompanied by increased concentrations of EGF (Supplementary Figure 4B and 4C). These results indicate that EGFR is critical for DENV-2 propagation. Interestingly, prostasin overexpression decreased DENV-2–elevated EGFR protein levels in a concentration-dependent manner (Figure 5C). Notably, the reduction of EGFR by prostasin suppressed DENV-2 protein synthesis at 24 and 48 dpi (Figure 5D top and middle panels, lines 4 and 6). Therefore, we hypothesized that prostasin-mediated proteolysis of EGFR is a mode of action against DENV-2 replication that acts through COX-2 suppression. PI3k/Akt and Raf/ERK activation is regulated by EGFR phosphorylation to induce COX-2 expression. We used specific inhibitors against PI3k/Akt or Raf/ERK signaling to confirm the signal molecules involved in DENV-2–mediated EGFR/COX-2 activation. As shown in Figure 6A and 6B, PI3K/Akt inhibitor LY294002 reduced the activity and protein expression of the DENV-2–induced COX-2 promoter-linked luciferase in a dose-dependent manner. In addition, pretreatment with PI3K/Akt (LY294002) inhibitor attenuated DENV-2 protein synthesis and amount of DENV-2 particles (Figure 6C and 6D). By contrast, the Raf inhibitor GW5074 did not decrease the activity and protein expression of DENV-2–induced COX-2 promoter-linked luciferase (Supplementary Figure 5A and 5B). As expected, the Raf inhibitor GW5074 did not reduce DENV-2 protein synthesis and RNA replication (Supplementary Figure 6A and 6B).
Prostasin knocks down epithelial growth factor receptor (EGFR) to suppress Dengue virus (DENV) replication. A and B, EGFR silencing decreased DENV-2 protein synthesis and viral titer. Huh-7 cells were transfected with green fluorescent protein short hairpin RNA (shGFP) and shEGFR at the indicated concentrations and infected with DENV-2 at a multiplicity of infection (MOI) of 1 for 2 days. The viral titer in the supernatant was measured by a plaque assay. C, Prostasin-reduced DENV-2–stimulated EGFR expression in a dose-dependent manner. Huh-7 cells were transfected with pCMV-prostasin-Myc at the indicated concentrations and infected with DENV-2 at a MOI of 1 for 2 days. D, Prostasin decreased DENV-2–stimulated EGFR expression against DENV protein synthesis in a time-dependent manner. Huh-7 cells were transfected with 1 μg pcDNA4 or pCMV-prostasin-Myc and infected with DENV-2 at a MOI of 1 for 2 days. Western blotting was performed with the indicated antibodies. All data from cell-based experiments are indicative of at least 3 independent experiments. Error bars represent the mean ± SD of 3 independent experiments; *P < .05. Abbreviation: GAPDH, glyceraldehyde-3-phosphate dehydrogenase; PFU, plaque-forming units.
Dengue virus (DENV) replication requires the activation of Akt/cyclooxygenase-2 (COX-2) signaling. A and B, The Akt inhibitor LY294002 reduced DENV-induced COX-2 promoter activity and protein synthesis. Huh-7 cells were transfected with pCOX-2-luciferase (Luc) and then infected with DENV-2 at a multiplicity of infection (MOI) of 1. The DENV-2–infected cells were treated with LY294002 at the indicated concentrations for 2 days. The cell lysate was subjected to luciferase activity assay and western blotting. The corresponding COX-2 protein levels shown as band intensity were quantified with a densitometer following normalization of total glyceraldehyde-3-phosphate dehydrogenase (GAPDH). C and D, The Akt inhibitor LY294002 reduced DENV-2 protein synthesis and viral titer. Huh-7 cells were infected with DENV-2 at a MOI of 1 and then treated with LY294002 at the indicated concentrations for 2 days. Minus symbol indicates treatment with 0.1% dimethyl sulfoxide. Western blotting was performed with indicated antibodies. Real-time quantitative polymerase chain reaction (RT-qPCR) was performed to analyze the relative RNA level of DENV-2 after normalization to the cellular gapdh mRNA level. The supernatant was subjected to a plaque assay to determine the viral titer. All data from cell-based experiments are indicative of at least 3 independent experiments. Error bars represent the mean ± SD of 3 independent experiments; *P < .05. Abbreviation: PFU, plaque-forming units.
DISCUSSION
In the present study, we observed low prostasin levels in blood cell samples from DF patients. This result was subsequently confirmed in DENV-infected ICR suckling mice and DENV-2–infected Huh-7 and THP-1 cells (Figure 1). We coinfected ICR suckling mice with recombinant adenoviruses carrying the prostasin gene and DENV to evaluate the biological role of prostasin in the increased survival rate of life-threatening DENV-2 infection in vivo (Figure 2). Through cell-based investigations, we further demonstrated that prostasin could decrease DENV-2 RNA replication and propagation via suppression of DENV-2–induced COX-2 expression (Figures 3 and 4). In this mechanism, prostasin mediated proteolysis of EGFR to interfere with the Akt/nuclear factor kappa B (NF-κB) signaling cascade and suppressed COX-2 expression, thus inhibiting DENV-2 replication (Figures 5 and 6). Several studies have shown that Akt/NF-κB and ERK were activated following stimulation of EGFR by viral infection, and the signaling pathways were critical for replication of viruses such as enterovirus 71 [15], herpes simplex virus [32], and HCV [33]. Our results showed that EGFR-mediated downstream molecules, including Akt and NF-κB, are actively phosphorylated upon DENV-2 infection (Supplementary Figure 7A). In contrast, the overexpression of prostasin reduced the DENV-2–induced phosphorylation of Akt and NF-κB (Supplementary Figure 7B). Furthermore, our results revealed that specific PI3K/Akt inhibitor LY294002, but not Raf inhibitor GW5074, reduced DENV-2–induced COX-2 expression and DENV-2 replication. These results confirmed the critical role of the Akt/NF-κB but not the Raf/ERK signaling pathway in viral propagation (Figure 6; Supplementary Figures 4 and 5). A previous study also reported that NF-κB–mediated but not ERK-mediated COX-2 expression was important for DENV replication [18]. Taken together, our results indicate that DENV-induced COX-2 expression is dependent on the EGFR/Akt/NF-κB signal cascade, but not the EGFR/Raf/ERK signal cascade.
EGFR has been reported to be involved in the entry of various viruses. For example, EGFR and EphA2 promote the association of HCV with CD81-claudin1 for membrane fusion during entry [14]. The activity of tyrosine kinase is required for the entry and budding of Ebola virions [34, 35]. Therefore, targeting EGFR may be a broad-spectrum antiviral strategy against multiple viral infections [36]. In the case of DENV, a previous study only showed that the assembly of DENV depends on the c-Src kinase-mediated EGFR signaling and that viral replication and propagation required the activation of EGFR defined by tyrosine kinase inhibitors in HepG2 cells [29]. Of the viruses belonging to the Flavivirus genus, only Japanese encephalitis virus entry was reported to be associated with EGFR-PI3K–mediated RhoA activation, which led to caveolin phosphorylation for viral endocytosis [37]. To date, there are still no studies indicating the association between EGFR and DENV entry. The detailed mechanism of EGFR involved in the entry process of DENV may be an interesting topic for further investigation. Previous studies revealed that the DENV-induced PI3K/Akt pathway can inhibit apoptosis for maintenance of cell survival and produce interleukin-10 (IL-10) for creation of a suitable environment to facilitate viral replication [38, 39]. Furthermore, a previous study indicated that PI3K/Akt is involved in the dynamic organization of actin cytoskeleton [40]. The active role of actin has been reported to be associated with DENV entry, production, and release of virions [41, 42]. Therefore, PI3K/Akt may serve as a potential anti-DENV therapeutic target. The detail relationship between prostasin and the PI3K/Akt pathway upon DENV infection is worth further investigation.
Type I interferon (IFN)-mediated antiviral responses constitute a critical cellular defense system against pathogen infection [43, 44]. In this defense system, dimerization of the phosphorylated signal transducer and the activator of transcription factors 1 and 2 (STAT1 and STAT2) are required to trigger the expression of several antiviral IFN-stimulated genes, such as PKR, 2′-5′-oligoadenylate synthetase 1 (OAS1), OAS2, and OAS3 [45–47]. Several DENV proteins, including NS2A, NS4A, and NS4B, inhibit STAT1 phosphorylation, causing formation of the STAT1-STAT2 heterodimer to fail and providing the appropriate environment for viral replication [48]. EGFR signaling activation impairs IFN-mediated antiviral activity by inhibiting STAT1 dimerization [49]. Based on the results of the present study, we hypothesize that DENV-activated EGFR signaling may interfere with antiviral IFN-mediated responses to facilitate viral replication given that DENV infection induces EGFR expression and the overexpression of its downstream genes. By contrast, prostasin-mediated EGFR suppression may be involved in the inhibition of DENV replication through the recovery of DENV-suppressed antiviral IFN-mediated responses. To confirm this hypothesis, we analyzed the expression of IFN and IFN-mediated antiviral genes in prostasin-overexpressing Huh-7 cells upon DENV-2 infection. As expected, the overexpression of prostasin dose-dependently induced the expression of IFN and IFN-stimulated antiviral genes, including OAS1, OAS2, OAS3, and PKR (Supplementary Figure 8). The detailed mechanism through which prostasin-mediated EGFR regulates type I IFN-mediated antiviral responses is an important issue requiring further investigation.
In summary, the model for the prostasin-mediated inhibition of DENV replication shown in Figure 7 revealed that DENV-elevated EGFR-Akt-NF-κB signaling facilitated DENV replication and propagation through aberrant COX-2 overexpression. Practically applied, prostasin overexpression can increase the survival rate following life-threatening DENV-2 infection in vivo.
Model for the mechanism of prostasin-mediated inhibition of Dengue virus-2 (DENV-2) replication via modulation of epithelial growth factor receptor (EGFR) proteolysis. Exogenous expression of prostasin protected ICR suckling mice from DENV-induced mortality and suppressed DENV-2 propagation. The antiviral mechanism of prostasin was the reduction effect on EGFR, leading to suppression of the Akt/nuclear factor kappa B (NF-κB)-mediated cyclooxygenase 2 (COX-2) signaling pathway. Abbreviations: COX-2, cyclooxygenase-2; GFP, green fluorescent protein; NF-κB, nuclear factor kappa B.
Supplementary Data
Supplementary materials are available at The Journal of Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.
Notes
Acknowledgments. We gratefully acknowledge Dr Charles Rice (Rockefeller University) and Aapth LLC for kindly supporting human hepatoma cell (Huh-7) and Dr Huey-Nan Wu (Institute of Molecular Biology, Academia Sinica, Taipei, Taiwan) for kindly supporting DENV replicon cell line (Huh-7-D2-FLuc-SGR-Neo). We also thank for Centers for Disease Control, Department of Health (Taiwan) for kindly supporting dengue virus (PL046 and 16681).
Financial support. This work was supported by the Ministry Science and Technology of Taiwan (grant numbers MOST 106-2911-I-037-502, MOST 106-2314-B-037-087, MOST 107-2314-B-037-079, and MOST 107-2311-B-037-005-MY3); Kaohsiung Medical University-Kaohsiung Medical University Hospital Co-Project of Key Research (grant numbers KMU-DK107011 and KMU-DK108010); and the National Sun Yat-Sen University-Kaohsiung Medical University Joint Research Project (grant number NSYSU-KMU 107-I003).
Potential conflicts of interest. All authors: No reported conflicts of interest. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.
