Astrocyte-derived VEGF increases cerebral microvascular permeability under high salt conditions

Abstract

Excess salt (NaCl) intake is closely related to a variety of central nervous system (CNS) diseases characterized by increased cerebral microvascular permeability. However, the link between a high salt diet (HSD) and the breakdown of tight junctions (TJs) remains unclear. In the present study, we found that high salt does not directly influence the barrier between endothelial cells, but it suppresses expression of TJ proteins when endothelial cells are co-cultured with astrocytes. This effect is independent of blood pressure, but depends on the astrocyte activation via the NFκB/MMP-9 signaling pathway, resulting in a marked increase in VEGF expression. VEGF, in turn, induces disruption of TJs by inducing phosphorylation and activation of ERK and eNOS. Correspondingly, the HSD-induced disruption of TJ proteins is attenuated by blocking VEGF using the specific monoclonal antibody Bevacizumab. These results reveal a new axis linking a HSD to increased cerebral microvascular permeability through a VEGF-initiated inflammatory response, which may be a potential target for preventing the deleterious effects of HSD on the CNS.

Keywords: VEGF; astrocyte; cerebral microvascular permeability; high salt.

Figure 1

HSD increases cerebral microvascular permeability. (A) Representative images of the mice cerebral cortical microvessels and dye leakage at 10 min after injection of Rhodamine B isothiocyanate-dextran (n=3 per group). (B) Evans blue leakage analyzed in rats fed with selected diets for 180 days (n=3 per group). (C) ZO-1 (green) co-stained with CD31 (red), a microvascular endothelia marker, and DAPI (blue; n=6 per group) in the rats brain slices. (D, E) Expression of Occludin, Claudin-5 and ZO-1 in the brain tissues from rats analyzed by immunoblotting. (F) Systolic blood pressure (SBP) of rats fed with NSD or HSD for 180 days. *P<0.05, **P<0.01, compared with NSD; at least three separate experiments were conducted, means ± SD.

Figure 2

High salt activates astrocytes and up-regulates VEGF. (A) Representative images of double immunofluorescence staining for GFAP(red) and VEGF(green) in the brain specimens of rats; DAPI is stained in blue(n=5 per group). (B, C) Expression of GFAP and VEGF in the brain tissues from rats analyzed by western blotting and densitometry (n=5 per group). (D) Immunofluorescence staining of GFAP and AQP4 in red. (E, F) Expression of AQP4 and GFAP analyzed by western blotting in NaCl-treated astrocytes; *P<0.05, **P<0.01, compared with NS. At least three separate experiments were conducted; means ± SD.

Figure 3

Increased expression of VEGF is mediated by NFκB/MMP-9 pathway. (A, B) HS increases protein levels of VEGF, MMP9 and p-p65. (C) Immunofluorescence staining of NaCl (40mM, 24 h)-induced p65 phosphorylation and nuclear translocation; (D) HS upregulates mRNA expression levels of VEGF. (E) SN50 antagonized NaCl-induced up-regulation of VEGF and MMP-9 protein expression. *P<0.05, **P<0.01, compared with NS, n=3, means ± SD.

Figure 4

Astrocyte-derived VEGF mediates HS-induced BBB breakdown. (A) Representative double immunofluorescence staining of ZO-1⁺ bEnd.3 endothelium. Endothelial cells were cultured alone, or co-cultured with primary rats′ astrocytes, and treated with NS, HS, conditioned medium, or VEGF neutralizing antibody (NA). (B, C) Western blotting analysis of Occludin, Claudin-5, and ZO-1 in endothelial cells. (D) Permeability of tight junctions measured using NaF; *P<0.05, **P<0.01, compared with EC+NS. At least three separate experiments were conducted; means ± SD.

Figure 5

Effect of VEGF on tight junction is mediated via ERK/eNOS. (A, B) Protein levels of ERK/pERK, eNOS/pSer¹¹⁷⁷, eNOS, and p38/p-p38 in endothelial cells; *P<0.05, **P<0.01, compared with vehicle. At least three separate experiments were conducted; means ± SD.

Figure 6

Increased cerebral microvascular permeability in VEGF^hi/+ mice. (A) Schematic diagram of VEGF^hi/+ mice generation. (B) Immunoblotting showing VEGF, Occludin, Claudin-5 and ZO-1expression in VEGF^hi/+ and wildtype mice. (C) Cerebral cortical micro-vessel and dye leakage at 5 and 10 min after injection of Rhodamine B isothiocyanate-dextran; n=5, *P<0.05, **P<0.01, compared with NSD. At least three separate experiments were conducted; means ± SD.

Figure 7

Blocking VEGF attenuates disruption of tight junctions induced by HSD. (A) Representative images showing double immunofluorescence staining of ZO-1 in green and DAPI in the brain specimens of rats. (B, C) Expression of Occludin, Claudin-5 and ZO-1 determined by western blotting; n=5, *P<0.05, **P<0.01, compared with NSD. At least three separate experiments were conducted; means ± SD.