The Epidermal Growth Factor Receptor Is a Relevant Host Factor in the Early Stages of The Zika Virus Life Cycle In Vitro

Abstract

Zika virus (ZIKV) is a flavivirus that is well known for the epidemic in the Americas in 2015 and 2016 in which microcephaly in newborns and other neurological complications were connected to ZIKV infection. Many aspects of the ZIKV viral life cycle, including binding and entry into the host cell, are still enigmatic. Based on the observation that CHO cells lack expression of the epidermal growth factor receptor (EGFR) and are not permissive for various ZIKV strains, the relevance of EGFR for the viral life cycle was analyzed. Infection of A549 cells by ZIKV leads to a rapid internalization of EGFR that colocalizes with the endosomal marker EEA1. Moreover, infection by different ZIKV strains is associated with an activation of EGFR and the subsequent activation of the mitogen-activated protein kinase (MAPK)/extracellular signal-regulated kinase (ERK) signaling cascade. However, treatment of the cells with methyl-β-cyclodextrin (MβCD), which on the one hand leads to an activation of EGFR but on the other hand prevents EGFR internalization, impairs ZIKV infection. Specific inhibition of EGFR or of the Ras-Raf-MEK-ERK signal transduction cascade hinders ZIKV infection by inhibition of ZIKV entry. In accordance with this, knockout of EGFR expression impedes ZIKV entry. In the case of an already established infection, inhibition of EGFR or of downstream signaling does not affect viral replication. Taken together, these data demonstrate the relevance of EGFR in the early stages of ZIKV infection and identify EGFR as a target for antiviral strategies. IMPORTANCE These data deepen the knowledge about the ZIKV infection process and demonstrate the relevance of EGFR for ZIKV entry. In light of the fact that a variety of specific and efficient inhibitors of EGFR and of EGFR-dependent signaling have been developed and licensed, repurposing of these substances could be a helpful tool to prevent the spreading of ZIKV infection in an epidemic outbreak.

Keywords: ZIKV; virus entry.

FIG 1

Zika virus (ZIKV) entry is affected in CHO cells. (A) Western blot analysis of epidermal growth factor receptor (EGFR) and ZIKV envelope (E) derived from cell lysates of A549 and CHO cells infected with either the French Polynesia (FP) or the Uganda (U) strain at a multiplicity of infection (MOI) of 1 and analyzed at 48 h postinfection (hpi). Uninfected (Ø) cells were used as a negative control for ZIKV infection. Detection of β-actin served as a loading control. (B) Confocal laser scanning microscopy analysis of A549 and CHO cells infected with either the French Polynesia or the Uganda strain at an MOI of 1 and analyzed at 48 hpi. Nuclei were stained with 4′,6-diamidino-2-phenylindole (DAPI) (blue), and EGFR and the ZIKV envelope were visualized with specific antibodies (red and green, respectively). Bar, 10 μm. (C) Western blot analysis of EGFR derived from cell lysates of A549, CHO, and liver tissue of Syrian golden hamster. A549 and CHO cells were used as positive and negative controls for EGFR expression, respectively. Detection of β-actin served as a loading control. (D) Relative numbers of attached viral genomes of A549 and CHO cells infected with the French Polynesia or the Uganda strain at an MOI of 1. The number of genomes was quantified by quantitative real-time PCR (RT-qPCR). Uninfected cells of the respective cell lines served as a negative control and for normalization. (E, F) Relative number of intracellular viral genomes of A549 and CHO cells infected with the French Polynesia or the Uganda strain at an MOI of 1 and analyzed at 30 (E) and 180 (F) min postinfection (mpi). The number of genomes was quantified by RT-qPCR after a short trypsinization step to remove bound, but not completely internalized, virus. Uninfected cells of the respective cell lines served as a negative control and for normalization. (G) Relative number of ZIKV genomes of CHO-EGFR #22.2 and CHO-EGFR #45 cells infected with the French Polynesia or the Uganda strain at an MOI of 10 at 37°C for 1 h (entry). The number of intracellular ZIKV genomes plotted was quantified by RT-qPCR and corrected to the corresponding number of attached genomes of CHO-EGFR #22.2 and CHO-EGFR #45 cells infected with the French Polynesia or the Uganda strain at an MOI of 10 at 4°C for 1 h (attachment). CHO cells served as control group and for normalization. ns, nonsignificant; *, P ≤ 0.05; **, P ≤ 0.01.

FIG 2

CHO cells overexpressing EGFR are not permissive to ZIKV infection, but EGFR overexpression in CHO cells increases viral entry. (A) Confocal laser scanning microscopy analysis of A549, CHO, CHO-EGFR #22.2, and CHO-EGFR #45 cells infected with either the French Polynesia or the Uganda strain at an MOI of 50 and analyzed at 24 hpi. Uninfected cells served as negative and positive controls, respectively. Nuclei were stained with DAPI (blue), and ZIKV E and EGFR were visualized with specific antibodies (green and red, respectively). Bar, 50 μm. (B) Western blot analysis of phosphorylated EGFR (p-EGFR), total EGFR, p-ERK, and total ERK of cell lysates of A549, CHO, CHO-EGFR #22.2, and CHO-EGFR #45 stimulated with either 25 or 50 ng/ml of EGF for 30 min. Unstimulated cells were used as a negative control. Detection of β-actin served as a loading control. (C) Confocal laser scanning microscopy analysis of A549, CHO, CHO-EGFR #22.2, and CHO-EGFR #45 cells stimulated with EGF (50 ng/ml) for 30 min or not stimulated. Nuclei were stained with DAPI (blue), and EGFR was visualized with a specific antibody (red). Bar, 20 μm. (D) Confocal laser scanning microscopy analysis of A549, CHO, CHO-EGFR #22.2, and CHO-EGFR #45 cells electroporated with a Renilla luciferase reporter virus and analyzed at 96 h postelectroporation (hpe). Nuclei were stained with DAPI (blue), and ZIKV E was visualized with a specific antibody (green). (E) Luciferase activity expressed in relative light units (RLU) per microgram of protein of A549, CHO, CHO-EGFR #22.2, and CHO-EGFR #45 cells electroporated with a Renilla luciferase reporter virus and analyzed at 24, 48, 72, and 96 hpe. A549 cells were used as a positive control for viral replication. Luciferase activity is presented on a log₁₀ scale.

FIG 3

Internalization of EGFR in ZIKV-infected cells. (A) Confocal laser scanning microscopy analysis of A549 cells infected with either the French Polynesia or the Uganda strain at an MOI of 1 and analyzed at 16 hpi. Uninfected cells and EGF-stimulated cells (50 ng/ml EGF for 30 min) served as negative and positive controls, respectively. Nuclei were stained with DAPI (blue), and EGFR and EEA1 were visualized with specific antibodies (green and red, respectively). Bar, 5 μm. (B) The colocalization of EGFR (green signal) with EEA1 (red signal) was quantified by the threshold of Mander’s overlap coefficient (tMOC), in which 1 represents a total overlap and 0 the opposite. Quantification is based on at least 6 cells. (C) Confocal laser scanning microscopy analysis of A549 cells infected with either the French Polynesia or the Uganda strain at an MOI of 1 and analyzed at 16 hpi. Uninfected cells were used as a negative control for ZIKV infection. Nuclei were stained with DAPI (blue), and ZIKV E was visualized with a specific antibody (green). Bar, 10 μm. (D) Confocal laser scanning microscopy analysis of A549 cells infected with either the French Polynesia or the Uganda strain at an MOI of 20 and analyzed at 5, 10, 30, 60, and 120 min postinfection (mpi). Infection was synchronized at 4°C. Uninfected cells were used as a negative control. Nuclei were stained with DAPI (blue), and EGFR was visualized with a specific antibody (green). Bar, 5 μm. (E) z-stacking of A549 cells infected with the French Polynesia strain at an MOI of 20 and analyzed at 30 min postinfection. Yellow arrows highlight the intracellular EGFR. EGFR was visualized with a specific antibody (red), and nuclei and F-actin were stained with DAPI (blue) and phalloidin-Atto 633 (cyan). Bar, 5 μm. ****, P ≤ 0.0001.

FIG 4

Activation of EGFR and downstream ERK in ZIKV-infected cells. (A to E) Representative Western blot (A) and respective densitometric quantification of relative p-EGFR/total EGFR (B, C) and p-ERK/total ERK (D, E) ratios of cell lysates of A549 cells infected with either the French Polynesia (FP) (B, D) or the Uganda (U) strain (C, E) at an MOI of 10 and analyzed at 5, 10, 15, 30, 60, 90, and 120 min postinfection (mpi). Detection of β-actin served as a loading control and to normalize the respective values. (F) Relative kinase activity of EGFR of A549 cells infected with either the French Polynesia or the Uganda strain at an MOI of 10 or 30 and analyzed at 15 mpi. Relative kinase activity was determined using the PamChip peptide microarray system, and the data were log₂ transformed by the analysis software. The plotted data are relative to n = 6 and n = 2 for MOI of 10 and 30, respectively. ns, nonsignificant; *, P ≤ 0.05; **, P ≤ 0.01; ****, P ≤ 0.0001.

FIG 5

EGFR and MAPK/ERK inhibitors decrease ZIKV infection, but not replication. (A) Relative metabolic activity of erlotinib-, sorafenib-, and PD98059-treated cells determined using the PrestoBlue cell viability reagent. The treatment lasted 24 h. Untreated and 2% TX-100-treated cells served as negative and positive controls, respectively. (B, D, F) Relative number of ZIKV genomes of A549 cells pretreated for 2 h with 25 μM erlotinib (B), 2 μM sorafenib (D), or 50 μM PD98059 (F) and infected with either the French Polynesia or the Uganda strain at an MOI of 1 for 4 h. After a short trypsinization step to remove attached but not fully internalized infectious viral particles, the number of ZIKV genomes was quantified by RT-qPCR at 24 h postinfection (hpi). The compounds were present during the whole experiment. Dimethyl sulfoxide (DMSO) was used as a vehicle control. (C, E, G) Relative number of intracellular infectious viral particles of A549 cells pretreated for 2 h with 25 μM erlotinib (C), 2 μM sorafenib (E), and 50 μM PD98059 (G) and infected with either the French Polynesia or the Uganda strain at an MOI of 1 for 4 h. After a short trypsinization step to remove attached but not fully internalized infectious viral particles, the number of intracellular infectious viral particles were quantified by plaque assay at 24 h postinfection (hpi). The compounds were present during the whole experiment. DMSO was used as a vehicle control. (H) Confocal laser scanning microscopy analysis of ZIKV-luciferase-replicating A549 cells. Nuclei were stained with DAPI (blue), and ZIKV E was visualized with a specific antibody (green). (I to K) Relative luciferase activity of ZIKV-luciferase-replicating A549 cells treated with 25 μM erlotinib (I), 2 μM sorafenib (J), or 50 μM PD98059 (K) and quantified by a luciferase assay at 24 h posttreatment. DMSO was used as a vehicle control. (L to N) Relative number of ZIKV genomes of A549 cells pretreated for 2 h with 25 μM erlotinib (L), 2 μM sorafenib (M), and 50 μM PD98059 (N) and infected with either the French Polynesia or the Uganda strain at an MOI of 10 for 1 h at 37°C (entry). The number of ZIKV genomes was quantified by RT-qPCR. (O) Confocal laser scanning microscopy analysis of A549 cells pretreated for 2 h with 25 μM erlotinib, 2 μM sorafenib (E), or 50 μM PD98059 and stimulated with EGF (50 ng/ml) for 30 min or left unstimulated. Nuclei were stained with DAPI (blue), and EGFR was visualized with a specific antibody (red). Bar, 10 μm. ns, nonsignificant; *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001; ****, P ≤ 0.0001.

FIG 6

Lipid raft disruption by MβCD treatment affects ZIKV binding and entry. (A, B) Relative number of ZIKV genomes of serum-starved A549 cells pretreated for 1 h with 1, 5, and 10 mM MβCD and infected with either the French Polynesia (A) or the Uganda strain (B) at an MOI of 10 for 1 h at 4°C (binding) and incubated at 37°C for 1 h (entry). The number of ZIKV genomes was quantified by RT-qPCR. Untreated cells were used as a control. (C) Relative metabolic activity of MβCD-treated cells determined using the PrestoBlue cell viability reagent. Untreated and 2% TX-100-treated cells served as negative and positive controls, respectively. (D to F) Representative Western blot (D) and respective densitometric quantification of relative p-EGFR/total EGFR (E) and p-ERK/total ERK (F) ratios of cell lysates of serum-starved A549 cells treated for 1 h with 1, 5, and 10 MβCD and stimulated with EGF (50 ng/ml) for 30 min or not stimulated. Detection of β-actin served as a loading control and to normalize the respective values. All of the values were further normalized to untreated unstimulated cells. (G) Confocal laser scanning microscopy analysis of serum-starved A549 cells treated for 1 h with 1, 5, and 10 MβCD and stimulated with EGF (50 ng/ml) for 30 min or not stimulated. Nuclei were stained with DAPI (blue), and EGFR was visualized with a specific antibody (red). Bar, 10 μm. ns, nonsignificant; **, P ≤ 0.01; ****, P ≤ 0.0001.

FIG 7

ZIKV infection is decreased in A549-EGFR knockout (KO) cells. (A) Confocal laser scanning microscopy analysis of A549-EGFR KO cells generated using the CRISPR/Cas9 system. A549 cells were transfected with plasmids containing either one of the single guide RNA (sgRNA) sequences. A549 cells transfected with the plasmid containing the off-target sequence were used as control. Nuclei were stained with DAPI (blue), and EGFR was visualized with a specific antibody (red). Bar, 10 μm. (B) Representative Western blot of cell lysates of A549-EGFR KO cells stimulated with EGF (50 ng/ml) for 30 min. Activation of EGFR and downstream ERK was evaluated by the level of phosphorylated protein in comparison with the total amount of protein. EGF-stimulated off-target cells were used as a positive control, while unstimulated cells were used as a negative control. Detection of β-actin served as a loading control. (C, D) Relative number of ZIKV genomes of A549-EGFR KO cells infected with either the French Polynesia (C) or the Uganda strain (D) at an MOI of 1. The number of ZIKV genomes was quantified by RT-qPCR at 24 h postinfection (hpi). Off-target cells were used as a control. (E, F) Relative amount of intracellular infectious viral particles of A549-EGFR KO cells infected with either the French Polynesia (E) or the Uganda strain (F) at an MOI of 1. The number of intracellular infectious viral particles was quantified by plaque assay at 24 h postinfection (hpi). Off-target cells were used as a control. (G) Relative number of ZIKV genomes of A549-EGFR KO cells infected with either the French Polynesia or the Uganda strain at an MOI of 10 for 1 h at 37°C for 1 h (entry). The number of ZIKV genomes was quantified by RT-qPCR. ns, nonsignificant; *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001; ****, P ≤ 0.0001.