Zika virus modulates mitochondrial dynamics, mitophagy, and mitochondria-derived vesicles to facilitate viral replication in trophoblast cells

Abstract

Zika virus (ZIKV) remains a global public health threat with the potential risk of a future outbreak. Since viral infections are known to exploit mitochondria-mediated cellular processes, we investigated the effects of ZIKV infection in trophoblast cells in terms of the different mitochondrial quality control pathways that govern mitochondrial integrity and function. Here we demonstrate that ZIKV (PRVABC59) infection of JEG-3 trophoblast cells manipulates mitochondrial dynamics, mitophagy, and formation of mitochondria-derived vesicles (MDVs). Specifically, ZIKV nonstructural protein 4A (NS4A) translocates to the mitochondria, triggers mitochondrial fission and mitophagy, and suppresses mitochondrial associated antiviral protein (MAVS)-mediated type I interferon (IFN) response. Furthermore, proteomics profiling of small extracellular vesicles (sEVs) revealed an enrichment of mitochondrial proteins in sEVs secreted by ZIKV-infected JEG-3 cells, suggesting that MDV formation may also be another mitochondrial quality control mechanism manipulated during placental ZIKV infection. Altogether, our findings highlight the different mitochondrial quality control mechanisms manipulated by ZIKV during infection of placental cells as host immune evasion mechanisms utilized by ZIKV at the placenta to suppress the host antiviral response and facilitate viral infection.

Keywords: mitochondria-derived vesicles (MDVs); mitochondrial quality control; mitophagy; nonstructural protein 4A (NS4A); zika virus (ZIKV).

Figure 1

ZIKV NS4A disrupts mitochondrial dynamics and respiration profile, and induces mtROS production. (A) JEG-3 cells were transfected with mtDsRED-encoding plasmid. Next day, cells were mock-infected or infected with PRVABC59 (PRV) (MOI 10) and fixed at the indicated time points to visualize mitochondrial morphology using confocal microscopy. Scale bar = 10 μm. (B) ZIKV NS4A-induced mitochondrial fragmentation was visualized using JEG-3 cells transfected with mtDsRED and empty vector (EV) or FLAG-tagged ZIKV NS4A plasmids. Scale bar = 10 μm. CCCP-treated cells were included as a control for mitochondrial fragmentation. (C) Western blot analysis of proteins regulating mitochondrial dynamics in JEG-3 cells overexpressing EV, ZIKV NS4A or NS4B. Band intensities were quantified using ImageJ software and normalized against β-actin. (D) Real-time measurement of mitochondrial stress assay for EV or ZIKV NS4A-overexpressing cells was carried out to determine ZIKV NS4A-induced changes in oxygen consumption rate (OCR) following treatment with oligomycin (1.5 μM), FCCP (2μM) or rotenone + Antimycin A (0.5μM each) (E) Tetramethylrhodamine methyl ester (TMRM) assay of JEG-3 cells overexpressing increasing concentrations of ZIKV NS4A-encoding plasmid. NS4A-induced mitochondrial membrane depolarization was determined using fluorescence microplate reader. *p<0.05, **p<0.01, when compared to the control. (F) JEG-3 cells were transfected with different amount of ZIKV NS4A-encoding plasmid. Cells were treated with DMSO control (ctl), CCCP, or mito-TEMPO, an antioxidant targeted to mitochondria. Mitochondrial ROS (mtROS) was measured by MitoSox Red staining and mtROS levels were quantitatively determined using fluorescence analysis software (mean ± SD; n = 3).*p<0.05, ***p<0.001, when compared to the control.

Figure 2

ZIKV NS4A induces mitophagy in human trophoblasts and neural progenitor cells. (A) JEG-3 cells were infected with mock- (m) or PRV (MOI 10) and stained with MitoTracker (green) (500 nM) and LysoTracker (red) (75 nM). Scale bar = 10 μm. Mitolysosomes were quantified by measuring percentage of cells with MitoTracker and LysoTracker colocalization (graph). The graph shows mean ± SD (n = 3). Statistical analysis; ***p<0.001, when compared to the mock. (B-D) HeLa cells expressing mtKeima (pH-dependent fluorescent protein) were infected with mock (m) or PRV (B), mock (m), Brazil or MR766 (C), or transfected with empty vector (EV), ZIKV NS4A or NS4B (D) were treated with CCCP (25 μM) for 2 (h) Green fluorescence of mtKeima reflects mitochondria in the cytosol (green), while red fluorescence of mtKeima reflects mitochondria in lysosomes (red). pH-dependent shift in excitation peaks were quantified to calculate % mitophagy (graph) using Zeiss ZEN software. *p<0.05, **p<0.01, and ***p<0.001, when compared to the mock or EV-transfected control. (E) MitoTracker Green and LysoTracker Red staining of mock- vs. PRV-infected human neural progenitor cells (hNPCs). Percentage of cells with MitoTracker and LysoTracker colocalization was calculated (graph). (F) TMRM staining of hNPCs transfected with varying concentrations of ZIKV NS4A-encoding plasmid. *p<0.05, ***p<0.001, when compared to the EV-transfected control.

Figure 3

Drp1-mediated mitochondrial fission & PINK1 are important for ZIKV replication. (A) Western blot analysis of JEG-3 transfected with ctl (siCtl) vs. PINK1-specific siRNA (siPINK1) to confirm successful knockdown. Band intensities were quantified using ImageJ. (B) ZIKV vRNA levels were quantified using qRT-PCR in siCtl- vs. siPINK1-transfected cells infected with PRV. **p<0.01, when compared to the siCtl-treated cells (C, D) JEG-3 cells transfected with control siRNA (siCtl), Drp1-specific siRNA (siDrp1) (C) or MFN2-specific siRNA (siMFN2) (D) were infected with PRV (MOI 10). ZIKV vRNA expression levels were quantified with qRT-PCR. (E, F) Mdivi-1 was added to JEG-3 cells at various concentrations and cells were infected with PRV (MOI 10). Mdivi-1-induced downregulation of mitochondrial fission was confirmed via suppression of FIS1 mRNA levels (E). Mdivi-1 treatment at various concentrations suppressed ZIKV vRNA gene expression levels, as quantified by qRT-PCR (F) (mean ± SD; n = 3). *p<0.05, **p<0.01, and ***p<0.001, when compared to the control. Relative expression levels were normalized against GAPDH.

Figure 4

ZIKV NS4A upregulates mitophagy receptor expression, and BNIP3 and NIX modulate ZIKV replication. (A) Western blot analysis of mitophagy receptor expression levels in cytosolic (cyto) and mitochondrial (mito) fractions obtained from mock- vs. PRV-infected JEG-3 cells. Band intensities were quantified using ImageJ and normalized against TOM70. (B, C) JEG-3 cells were transfected with siCtl, siBNIP3 (B) or siNIX (C) and infected with PRV (MOI 10). BNIP3 and NIX mRNA levels were determined by qRT-PCR. Data represented as mean ± SD. **p<0.01, when compared to the siCtl-transfected cells. Relative expression levels were normalized against GAPDH. (D, F) qRT-PCR analysis of BNIP3 or NIX siRNA-transfected JEG-3 cells infected with PRV (MOI 10). Relative viral transcription levels (D) and IFN-β, ISG15, and OAS1 mRNA levels (F) were determined. Relative expression levels were normalized against GAPDH. (E) TCID₅₀ assay of Vero cells infected with supernatants collected at 48 hpi from siBNIP3- or siNIX-transfected JEG-3 cells infected with PRV (MOI 10). Data represented as mean ± SD. *p<0.05, **p<0.01, and ***p<0.001, when compared to the siCtl-treated PRV-infected cells.

Figure 5

ZIKV NS4A reduces mitochondrial mass & blocks MAVS oligomerization to suppress IFN response. (A, B) Genomic DNA isolated to determine relative mitochondrial DNA (mtDNA) present in JEG-3 cells infected with mock (m), PRV (MOI 10) or treated with CCCP (25 μM, 2 h) (A) or transfected with empty vector (EV), ZIKV NS4A, NS4B or treated with CCCP (25 μM, 2 h) (B). Relative mtDNA amount was normalized against 18S rRNA. *p<0.05, **p<0.01, and ***p<0.001, when compared to the m or EV control. (C) Western blot analysis of MAVS oligomerization and expression levels in JEG-3 cells transfected with MAVS in combination with increasing concentrations of ZIKV NS4A. (D) JEG-3 cells overexpressing IFN-β or ISRE reporter plasmids were transfected with EV, ZIKV NS4A or NS4B and treated with poly(I:C) (20 μg/mL, 2h). Luciferase reporter assay is shown (mean ± SD). (E, F) Luciferase assay of HEK293T cells transfected with IFN-β or ISRE reporter plasmids in combination with ZIKV NS4A and one of the following RLR signaling effectors (MDA5, RIG-I, MAVS, TBK1, IRF3). Data represented as mean ± SD. *p<0.05, **p<0.01, and ***p<0.001, when compared to the EV control.

Figure 6

Small extracellular vesicles (sEV) proteomics profiling revealed presence of mitochondrial components. (A) Schematic representation of the experimental procedures taken to isolate, purify, and characterize sEVs released from mock- vs. PRV-infected JEG-3 cells. Liquid chromatography with tandem mass spectrometry (LC–MS/MS) analysis was performed to characterize the contents of sEVs from PRV-infected JEG-3 cells (B) Nanoparticle tracking analysis of sEVs secreted by mock- or PRV-infected JEG3 cells is shown. (C) Western blot analysis of sEVs confirm the presence of ZIKV NS1 and sEV markers (CD9, CD63, TSG101, and ALIX). Endoplasmic reticulum localized calnexin was used as a negative marker. (D) Transmission electron microscopy of sEVs. Scale bar = 100 nm. (E) Pie chart of functional Gene Ontology (GO) terms associated with differentially expressed proteins detected within sEVs (PRV- vs. mock-infected; p < 0.05, fold change > 2). (F) Heatmap of top expressed proteins that were upregulated in PRV-infected sEVs. Two biological replicates per samples were included for further analysis. (G) LC–MS/MS-based proteomics data were validated by Western blot analysis. Snx6, Snx9, LAMP1, TOM70, IRF3, p62, and G3BP1 were upregulated in PRV-infected sEVs compared with mock sEVs. (H) Number of protein-encoding genes of the detected proteins were quantified according to their subcellular localization. Proteins that can localize to the mitochondria are shown in gray.

Figure 7

MDV components (Snx6 and Snx9) regulate ZIKV replication & infectivity. (A) Heatmap of upregulated sEV proteins that can localize to the endosome. Snx6 and 9 were chosen for further analysis. (B) Western blot analysis of JEG-3 cells transfected with Snx6- or Snx9-specific siRNAs and infected with mock or PRV (MOI 10). Snx6 and Snx9 protein levels were determined and quantified below the blot. (C) qRT-PCR quantification of ZIKV vRNA transcription levels in siSnx6- and/or siSnx9-transfected JEG-3 cells infected with PRVABC59. Expression levels were quantified relative to siCtl-transfected JEG-3 cells infected with PRV. (D) TCID₅₀ assay of Vero cells infected with supernatants collected at 48 hpi from siSnx6- and/or siSnx9-transfected JEG-3 cells infected with PRV. (E) siSnx6- and/or siSnx9-transfected JEG-3 cells were infected with PRV for 24h. Relative ISG15 mRNA levels were quantified relative to siCtl-transfected JEG-3 cells infected with PRV. (F) Cytosolic mtDNA amount was quantified with qRT-PCR. (+) = JEG-3 cells transfected with siSnx6 and siSnx9 in combination. Data represented as mean ± SD (n=3). *p<0.05, **p<0.01, and ***p<0.001, when compared to the siCtl-transfected cells infected with PRV.

Figure 8

ZIKV NS4A dysregulates mitochondrial dynamics and quality control mechanisms to suppress host antiviral response in trophoblast cells. ZIKV NS4A triggers changes in mitophagy, and MAVS-mediated antiviral immunity to evade the host innate immune response and facilitate viral infection, while ZIKV modulation of MDV components, such as Snx6 and Snx9, may affect ZIKV replication and infectivity.