Zika virus E protein modulates functions of human brain microvascular endothelial cells and astrocytes: implications on blood-brain barrier properties

Abstract

Neurotropic viruses can cross the otherwise dynamically regulated blood-brain barrier (BBB) and affect the brain cells. Zika virus (ZIKV) is an enveloped neurotropic Flavivirus known to cause severe neurological complications, such as encephalitis and fetal microcephaly. In the present study, we employed human brain microvascular endothelial cells (hBMECs) and astrocytes derived from human progenitors to establish a physiologically relevant BBB model. We used this model to investigate the effects of ZIKV envelope (E) protein on properties of cells comprising the BBB. E protein is the principal viral protein involved in interaction with host cell surface receptors, facilitating the viral entry. Our findings show that the presence of ZIKV E protein leads to activation of both hBMECs and astrocytes. In hBMECs, we observed a decrease in the expression of crucial endothelial junction proteins such as ZO-1, Occludin and VE-Cadherin, which are vital in establishment and maintenance of the BBB. Consequently, the ZIKV E protein induced changes in BBB integrity and permeability. We also found upregulation of genes involved in leukocyte recruitment along with increased proinflammatory chemokines and cytokines upon exposure to E protein. Additionally, the E protein also led to astrogliosis, evident from the elevated expression of GFAP and Vimentin. Both cell types comprising the BBB exhibited inflammatory response upon exposure to E protein which may influence viral access into the central nervous system (CNS) and subsequent infection of other CNS cells. Overall, our study provides valuable insights into the transient changes that occur at the site of BBB upon ZIKV infection.

Keywords: E protein; Zika virus; astrocytes; blood-brain barrier; brain microvascular endothelial cells; tight junction.

FIGURE 1

Effect of ZIKV E protein on BBB integrity. Schematic representation of monoculture BBB model establishment using hBMECs (A). hBMECs were transfected with empty vector or ZIKV E-protein. TEER was measured after 24 h. Bar graph shows TEER values given in ohm.cm² (B). Dextran-FITC transendothelial permeability was assessed after 24 h of transfection. Bar graph shows relative fluorescence change (C). Schematic representation of coculture BBB model establishment using hBMECs and astrocytes (D). hBMECs transfected with empty vector or ZIKV E protein were grown in contact with astrocytes for 24 h followed by measurement of TEER values (E) and dextran-FITC transendothelial permeability (F). Data represents mean ± SEM for at least three independent experiments. **p < 0.01, ***p < 0.001 with respect to control.

FIGURE 2

Effect of ZIKV E protein on endothelial junction protein expression. hBMECs were transfected with ZIKV E protein and empty vector as control for 24 h. (A) Total protein was isolated and subjected to Western blotting against ZO-1 (B), VE-Cadherin (C), Occludin (D) antibodies. β-actin was used as loading control. One representative blot is shown. Bar graphs shows relative protein expression in indicated groups. To assess localization of junction proteins, transfected hBMECs were immunostained with ZO-1 (E,F), VE-Cadherin (G,H) and Occludin (I,J). Scale bar 20 μm. Bar graph shows relative protein expression in indicated groups (K–M). Data represents mean ± SEM for at least three independent experiments. *p < 0.05, **p < 0.01 with respect to control.

FIGURE 3

ZIKV E protein induces inflammation in hBMECs. Brain microvascular endothelial cells were transfected with empty vector and E protein for 24 h. Total RNA was isolated and transcript levels of inflammatory cytokines and chemokines were measured using quantitative RT-PCR for IL-6 (A), IL-8 (B), IL-1β (C), CCL2 (D), CCL5 (E) and CXCL10 (F). Transcript levels of genes involved in modulation of adhesion molecules and angiogenesis- ICAM-1 (G), VCAM-1 (H), and PTGS2 (I) were also measured. Bar graph shows relative transcript expression in indicated groups. Data represents mean ± SEM for at least three independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001 with respect to control.

FIGURE 4

ZIKV E protein induces release of proinflammatory cytokines in hBMECs. Schematic representation of experimental setup (A). Brain microvascular endothelial cells were transfected with empty vector or E protein for 24 h. Supernatants collected from transfected cells were subjected to cytokine bead array for measurement of inflammatory cytokines IL-6 (B) and IL-8 (C). Data represents mean ± SEM for at least three independent experiments. **p < 0.01 with respect to control.

FIGURE 5

ZIKV E protein triggers astrogliosis. Progenitor derived astrocytes were transfected with empty vector or E protein for 24 h. Total protein was isolated and subjected to Western blotting against markers for astrocyte reactivity, GFAP (A) and Vimentin (B). β-actin was used as loading control. One representative blot is shown. Bar graph shows relative protein expression in indicated groups (C,D). To assess the transcript levels, total RNA was isolated and subjected to quantitative RT- PCR for GFAP (E), Vimentin (F), IL-6 (G), and IL-8 (H). Data represents mean ± SEM for at least three independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001 with respect to control.

FIGURE 6

ZIKV E protein induced reactive astrocytes release proinflammatory cytokines, extracellular glutamate and MCP-1. Progenitor derived astrocytes were transfected with empty vector or E protein for 24 h. Schematic representation of experimental paradigm followed (A). Supernatant collected from transfected cells were subjected to cytokine bead array for measurement of inflammatory cytokines IL-6 (B) and IL-8 (C). Also, supernatant collected from transfected astrocytes was subjected to measurement of extracellular glutamate release (D) and also subjected to ELISA for measurement of MCP-1 (E). Data represents mean ± SEM for at least three independent experiments. *p < 0.05, **p < 0.01 with respect to control.

FIGURE 7

Reactive astrocytes affect endothelial junction proteins in hBMECs. Astrocytes were transfected with empty vector and E protein for 24 h. Schematic representation of experimental paradigm followed (A). Supernatants were collected and mixed with equal amount of fresh endothelial growth media. Brain microvascular endothelial cells were exposed to the astrocyte-conditioned media for 24 h. (B) Total protein was isolated and subjected to Western blotting against endothelial junction proteins- ZO-1 (C), VE-Cadherin (D), and Occludin (E). β-actin was used as loading control. One representative blot is shown. Bar graphs shows relative protein expression in indicated groups. Data represents mean ± SEM for at least three independent experiments. *p < 0.05, **p < 0.01, ns = not significant, with respect to control.

FIGURE 8

ZIKV E protein expression in hBMECs affects endothelial proteins in BBB of astrocytes and hBMECs. Transfected hBMECs were grown in contact with astrocytes for 24 h on transwell apparatus to form co-culture of BBB. Schematic representation of experimental paradigm followed (A). (B) Total protein was isolated from the luminal side of transwell and subjected to Western blotting against endothelial junction proteins- ZO-1 (C), VE-Cadherin (D), and Occludin (E). β-actin was used as loading control. One representative blot is shown. Bar graphs shows relative protein expression in indicated groups. Data represents mean ± SEM for at least three independent experiments. *p < 0.05, **p < 0.01 with respect to control.