Zika Virus NS4A and NS4B Proteins Deregulate Akt-mTOR Signaling in Human Fetal Neural Stem Cells to Inhibit Neurogenesis and Induce Autophagy

Abstract

The current widespread outbreak of Zika virus (ZIKV) infection has been linked to severe clinical birth defects, particularly microcephaly, warranting urgent study of the molecular mechanisms underlying ZIKV pathogenesis. Akt-mTOR signaling is one of the key cellular pathways essential for brain development and autophagy regulation. Here, we show that ZIKV infection of human fetal neural stem cells (fNSCs) causes inhibition of the Akt-mTOR pathway, leading to defective neurogenesis and aberrant activation of autophagy. By screening the three structural proteins and seven nonstructural proteins present in ZIKV, we found that two, NS4A and NS4B, cooperatively suppress the Akt-mTOR pathway and lead to cellular dysregulation. Corresponding proteins from the closely related dengue virus do not have the same effect on neurogenesis. Thus, our study highlights ZIKV NS4A and NS4B as candidate determinants of viral pathogenesis and identifies a mechanism of action for their effects, suggesting potential targets for anti-ZIKV therapeutic intervention.

Figure 1. ZIKV Infection Impairs Neurosphere Formation and Elevates Autophagy in fNSCs

(A) Representative images from Live/Dead cell viability assay in cultured fNSCs at 5 dpi with three strains of ZIKV (MR766, IbH30656, or H/PF/2013) or mock treatment. Images were taken using live cell imaging. Scale bar, 20 μm. (B) Quantification of percentage of cell death as described in (A). Mean ± SEM; p < 0.05 by one-way ANOVA. (C) Representative images showing neurosphere formation at 3 dpi with three strains of ZIKV or mock treatment. Scale bar, 100 μm. (D and E) Quantification of number of neurospheres formed per 1 × 10⁵ fNSCs (D) and neurosphere size by diameter measurement (E) in conditions as in (C). Mean ± SEM; p < 0.05 by one-way ANOVA. (F and G) Representative confocal images showing 3D reconstruction of neurospheres at 7 dpi with MR766 or mock treatment. Nestin and SOX2 were used as fNSC-specific markers; ZIKV was immunostained against its E protein in (F); apoptotic cell death was marked by TUNEL staining in (G). Scale bar, 50 μm. (H) ZIKV infection induces autophagy in fNSCs. fNSCs were infected with ZIKV and LC3 processing was examined by immunoblot at indicated time points. (I) fNSCs infected with ZIKV at MOI 0.1 were fixed and stained with indicated antibodies and LC3 puncta were counted. Mean ± SEM; p < 0.05 by one-way ANOVA. (J) Autophagy is required for the efficient replication of ZIKV. fNSCs were infected with ZIKV MR766 at MOI 0.5, and the medium was changed with indicated drugs (rapamycin 50 nM, 3-MA 2 μM, chloroquine 5 μM) at 1 hpi. The mRNA levels of ZIKV were measured by RT-qPCR at 10 hpi. See also Figures S1 and S2.

Figure 2. ZIKV NS4A and NS4B Impair Neurogenesis of fNSCs

(A) Representative images of neurospheres formed at 7 dpi from fNSCs transduced with lentiviruses expressing ZIKV NS4A, NS4B, NS4A-NS4B, or vector alone. Neurospheres were stained with fNSC markers Nestin and SOX2; upper panels represent primary neurospheres (P0), lower panels represent secondary neurospheres after passage (P1). Scale bar, 50 μm. (B) Quantification of number of neurospheres formed per 1 × 10⁵ fNSCs transduced with lentiviruses as indicated. Mean ± SEM; p < 0.05 by one-way ANOVA. (C) Percentage bar graph showing the distribution of neurospheres at different size ranges as described in (A). (D) BrdU incorporation into fNSC-based flow cytometry analysis at 5 days after transduction with different lentiviruses as indicated. Mean ± SEM; p < 0.05 by one-way ANOVA. (E and F) Representative confocal images of Nestin and Ki-67 double staining (E) and the quantification of proliferating Nestin+ and Ki-67+ double positive fNSCs (F) at 5 days after transduction with different lentiviruses as indicated. Mean ± SEM; p < 0.05 by one-way ANOVA in (F). (G and H) Representative confocal images of β3-tubulin immunostaining on fNSCs transduced with different lentiviruses as indicated followed by 10 days of differentiation (G) and quantification (percentage) of β3-tubulin-positive neurons differentiated from fNSCs (H). Dapi: nucleus staining. Mean ± SEM; p < 0.05 by one-way ANOVA in (H). (I and J) Representative confocal images of GFAP immunostaining and nucleus staining with Dapi (I) and the quantification (percentage) of GFAP-positive astrocytes differentiated from fNSCs (J), 10 days after transduction with different lentiviruses as indicated. Mean ± SEM; p < 0.05 by one-way ANOVA in (J). See also Figure S3.

Figure 3. ZIKV NS4A and NS4B Induce Autophagy

(A and B) Screening of ZIKV proteins for autophagy induction. HeLa-GFP-LC3 cells transiently expressed each ZIKV protein as indicated by lentivirus infection. The levels of GFP-LC3 puncta were measured and quantified at 2 dpi. Mean ± SEM; p < 0.05 by one-way ANOVA in (B). (C–F) LC3 processing from fNSCs or HeLa cells stably expressing vector, NS4A, NS4B, or NS4A-NS4B was measured by immunoblot with indicated antibodies. The levels of LC3-II/LC3-I were quantified by band intensity with Image Lab software (BioRad). See also Figure S4.

Figure 4. ZIKV NS4A and NS4B Inhibit the Akt-mTOR Signaling Pathway

(A) ZIKV replication inhibits Akt-mTOR signaling. fNSCs were infected with ZIKV strain MR766 at MOI 0.5. Cell lysates were harvested at various time points and subjected to immunoblot with indicated antibodies. (B) The levels of Akt and mTOR activities of HeLa cells expressing each ZIKV gene were measured with indicated antibodies. NS3H, NS3 helicase domain; NS3S, NS3 serine protease domain. (C–E) fNSCs or HeLa cells stably expressing vector, NS4A, NS4B, or NS4A-NS4B were stimulated with serum (20%) or insulin (2 μg/mL) after 8 hr starvation. Cell lysates were harvested and subjected to immunoblot with indicated antibodies. See also Figure S4.