Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
, 21 (3), 349-358.e6

Zika-Virus-Encoded NS2A Disrupts Mammalian Cortical Neurogenesis by Degrading Adherens Junction Proteins

Affiliations

Zika-Virus-Encoded NS2A Disrupts Mammalian Cortical Neurogenesis by Degrading Adherens Junction Proteins

Ki-Jun Yoon et al. Cell Stem Cell.

Abstract

Zika virus (ZIKV) directly infects neural progenitors and impairs their proliferation. How ZIKV interacts with the host molecular machinery to impact neurogenesis in vivo is not well understood. Here, by systematically introducing individual proteins encoded by ZIKV into the embryonic mouse cortex, we show that expression of ZIKV-NS2A, but not Dengue virus (DENV)-NS2A, leads to reduced proliferation and premature differentiation of radial glial cells and aberrant positioning of newborn neurons. Mechanistically, in vitro mapping of protein-interactomes and biochemical analysis suggest interactions between ZIKA-NS2A and multiple adherens junction complex (AJ) components. Functionally, ZIKV-NS2A, but not DENV-NS2A, destabilizes the AJ complex, resulting in impaired AJ formation and aberrant radial glial fiber scaffolding in the embryonic mouse cortex. Similarly, ZIKA-NS2A, but not DENV-NS2A, reduces radial glial cell proliferation and causes AJ deficits in human forebrain organoids. Together, our results reveal pathogenic mechanisms underlying ZIKV infection in the developing mammalian brain.

Keywords: Zika virus; adherens junction; cortical neurogenesis; flavivirus; human organoid; human protein microarray; microcephaly; neural stem cell; neuronal migration; radial glial cell.

Conflict of interest statement

COMPETING FINANCIAL INTERESTS: The authors declare no competing financial interests.

Figures

Figure 1
Figure 1. ZIKV-NS2A, but not DENV-NS2A, dysregulates radial glial cells in the embryonic mouse cortex
Embryonic mouse cortex was electroporated at E14.5 to express GFP, GFP and ZIKV-NS2A, or GFP and DENV-NS2A, followed by EdU labeling 2 hr before analysis at E17.5. Sample confocal images of immunostaining for Pax6 (A), Tbr2 (B), Tuj1 (C) and GFP, and staining for EdU and DAPI are shown. Scale bars: 100 μm (A, B, C). The regions in white boxes (top panels) are shown at a higher magnification (bottom panels). Quantifications are also shown (D–I). Values represent mean + SEM (n = 5–7 sections from 3–4 animals; ***: P < 0.001; **: P < 0.01; One-way ANOVA). Also see Figure S1 and S2.
Figure 2
Figure 2. ZIKV-NS2A, but not DENV-NS2A, expression leads to aberrant localization of neurons in the developing mouse cortex
(A) Embryonic mouse cortex was electroporated at E14.5 to express GFP, or GFP and ZIKV-NS2A, followed by analysis at E19.5. Sample confocal images of immunostaining for SATB2, CTIP2 and GFP, and staining for DAPI are shown. Scale bar: 50 μm. (B) Scatter plots and summary of cell body position of GFP+SATB2+ neurons and GFPCTIP2+ neurons in the mouse cortex at E19.5. The distance of each cell to the apical surface was normalized to the total thickness of the neocortex. Each dot represents one neuron. The box plots show the medians (line), means (square), interquartile ranges (box; 25–75%), and extremes of the distribution (whisker; 99%: upper crosshatch; 1%: lower crosshatch) (n = 8 sections from 4 animals for each condition; ** P < 0.01; n.s. P > 0.1; Student’s t-test).
Figure 3
Figure 3. Protein-protein interactomes of ZIKV-NS2A and DENV-NS2A across the human proteome
(A) 143 and 47 direct ZIKV-NS2A and DENV-NS2A interacting proteins were identified in vitro using a protein array containing 20,240 full-length human proteins. Among 143 ZIKV-NS2A interacting proteins, 83 proteins can be visualized in a connected network based on existing literatures (P < 4.91 × 10−13; STRING analysis), whereas the remaining 60 proteins are singletons. ZIKV-NS2A-specific interacting proteins are coded in gray and adhesion-related proteins are highlighted with circles. (B) Sample western blot images of co-IP analysis of HEK293 cells expressing GFP, ZIKV-NS2A and GFP, or DENV and GFP, and analyzed for adherens junction complex components. Histone H3 served as a negative control for Co-IP. Also see Figure S3.
Figure 4
Figure 4. ZIKV-NS2A, but not DENV-NS2A, expression leads to adherens junction complex component degradation
(A–B) Sample western blot images of expression levels of adherens junction complex components in mouse neural progenitors infected with ZIKV, expressing ZIKV-NS2A, or expressing DENV-NS2A (A), and quantifications (B). Data were normalized to that of mock infection for the ZIKV infection condition, or to that of GFP expression alone for the ZIKV-NS2A or DENV-NS2A conditions. Values represent mean + SEM (n = 3 cultures; ***: P < 0.001; *: P < 0.05; Student’s t-test). (C–D) Sample western blot images of expression levels of ZO-1 upon 24 hr treatment of DMSO, MG-132 (20 μM), Delanzomib (60 nM) and BFA (100 nM) in HEK293 cells expressing GFP and ZIKV-NS2A, or GFP alone (C), and quantifications (D). Data were first normalized to actin expression levels and then to the data from expression of GFP alone. Values represent mean + SEM (n = 3 cultures; **: P < 0.01; Student’s t-test). (E–F) Sample western blot images of expression levels of ZO-1 and NUMBL upon 24 hr treatment of DMSO, or BFA (100 nM) of mouse cortical neural progenitors expressing GFP and ZIKV-NS2A, or GFP alone (E), and quantifications (F). Data were normalized to that of actin. Values represent mean + SEM (n = 3 cultures; ***: P < 0.001; Student’s t-test). Also see Figure S4.
Figure 5
Figure 5. ZIKV-NS2A expression and direct ZIKV infection disrupt the formation of adherens junction complex in the embryonic mouse cortex
(A–B) Embryonic mouse brains were electroporated at E14.5 to express GFP, GFP and ZIKV-NS2A, or GFP and DENV-NS2A, and analyzed at E17.5. Sample confocal images of immunostaining for GFP, β-Catenin (A), or PKCλ and ZO-1 (B), and staining for DAPI under different conditions are shown (left panels). Scale bars: 50 μm. Arrows point to regions with discontinuous AJ formation. Regions in white boxes in (A) are shown at a higher magnification (bottom panels). Quantifications of continuous AJ formation and number of protrusions are also shown (right panels). Values represent mean + SEM (n = 5 sections from 3 animals; ***: P < 0.001; **: P < 0.01; Student’s t-test). (C–D) ZIKV-SZ strain was injected into lateral ventricles of E13.5/E14.5 mice. Similar to (A–B), sample confocal images of immunostaining for ZIKV, β-Catenin and ZO-1, and staining for DAPI at E18.5 (left panels) and quantifications of continuous AJ formation at E18.5 and ventricular protrusions at P3 (right panels) are shown. Values represent mean + SEM (n = 6 sections from 4 animals; ***: P < 0.001; *: P < 0.05; Student’s t-test). Also see Figure S5.
Figure 6
Figure 6. Expression of ZIKV-NS2A, but not DENV-NS2A, reduces proliferation and disrupts adherens junction formation of ventricular radial glial cells in human forebrain organoids
(A) Day 45 forebrain organoids were electroporated to express GFP, GFP and ZIKV-NS2A, or GFP and DENV-NS2A, and analyzed 3 days later (45+3) after pulsing with EdU (10 μM) for 1 hr. Sample confocal images for immunostaining for GFP and PAX6, and staining for EdU and DAPI are shown (top panels). Scale bar: 100 μm. Quantifications of percentages of EdU+GFP+ cells among all GFP+ cells, or GFP+PAX6+EdU+ cells among GFP+PAX6+ cells are also shown (bottom panels). Values represent mean + SEM (n = 10 organoids; ***: P < 0.001; Student’s t-test). (B) Sample confocal images of immunostaining for GFP, PKCλ, and staining for DAPI are shown (left panels). Scale bar: 100 μm. Arrows point to regions with discontinuous AJ formation. Quantifications of AJ continuity based on PKCλ expression are also shown (right panel). Values represent mean + SEM (n = 9 organoids; **: P < 0.01; Student’s t-test). (C) Expression of ZIKV-NS2A, but not DENV-NS2A, or GFP alone, led to a loss of typical radial glia morphology of PAX6+GFP+ cells at 7 days after electroporation (45+7). Sample confocal images of immunostaining for GFP and PAX6, and staining for DAPI (left panels; scale bar: 100 μm) and quantification of percentages of GFP+PAX6+ cells with multipolar morphologies are shown. Values represent mean + SEM (n = 7 organoids; **: P < 0.01; Student’s t-test). Also see Figure S6.

Similar articles

See all similar articles

Cited by 28 PubMed Central articles

See all "Cited by" articles

Publication types

Substances

LinkOut - more resources

Feedback