Development of ascorbate transporters in brain cortical capillary endothelial cells in culture

Abstract

Ascorbic acid in its reduced form is not transported across the capillary endothelial cell blood-brain barrier. This is thought to be due to absence of the SVCT2, a specific transporter for ascorbate. To assess this directly we prepared primary cultures of mouse cortical microvascular endothelial cells. When still in the capillaries, these cells did not express the SVCT2 protein as assessed by immunocytochemistry and by immunoblotting. However, during several days in culture, they developed SVCT2 expression and showed ascorbate transport rates comparable to those in immortalized endothelial cell lines. SVCT2 expression was inversely proportional to cell density, was enhanced by culture at low physiologic plasma ascorbate concentrations, was inhibited by ascorbate concentrations expected in the brain interstitium, and was stimulated by cobalt ions. Expression of the SVCT2 was associated with ascorbate-dependent maturation and release of type IV collagen by the cells in culture. Although the SVCT2 is induced by culture of cortical capillary endothelial cells, its absence in vivo remains perplexing, given the need for intracellular ascorbate to facilitate type IV collagen maturation and release by endothelial cells.

Fig. 1. Absence of the SVCT2 protein in brain microcapillaries

Panel A: Immunostaining. Freshly isolated cortical microcapillaries were immunostained for von Willebrand factor in the left image, and for the SVCT2 in the right image. Arrowheads point to the endothelial cell-lined capillaries. Panel B: Immunoblotting. Rinsed, concentrated cortical capillary vessels were blotted for the SVCT2 at day zero (0d) and day 7 (7d) in culture. The location of the SVCT2 as a 75 kDa band on electrophoresis is noted.

Fig. 2. Time-dependent appearance of the SVCT2 in CCECs in culture

Immunostaining of the SVCT2 in cells cultured for various times (upper panel). Actin staining of the capillary endothelial cell cultures shown in Panel A to confirm that similar protein amounts were loaded in each lane (lower panel).

Fig. 3. Appearance of specific ascorbate transport in CCECs in culture

Panel A. Time-dependent ascorbate transport measurements in CCECs (PC) and in two immortalized endothelial cell lines, EA.hy926 and bEND.3. Panel B: Specificity of ascorbate transport in brain primary culture endothelial cells not treated with an agent (-), treated with the concentration of dimethylsulfoxide required to dissolve the two inhibitors (DMSO), and treated with either 1 mM sulfinpyrazone (SPZ) or 100 µM phloretin just before the transport assay. Results are corrected for cell number and are representative of 3 such experiments performed.

Fig. 4. Inhibition of SVCT2 protein expression at with culture of CCECs at high cell densities

CCECs were cultured at the same plating density in different sized wells for 7 days, followed by protein extraction for immunoblotting of the SVCT2. Panel A shows a representative immunoblot, below which is β-actin expression in the same gel shown to confirm equal loading of protein in the gel lanes. Panel B is a densitometric analysis of 4 experiments performed, with an asterisk (*) indicating p < 0.05.

Fig. 5. Dependence of SVCT2 expression on medium ascorbate concentrations in CCECs

Cells were cultured for 7 days supplemented with the ascorbate concentrations noted before they were taken for gel electrophoresis and immunoblotting. Panel A shows a representative immunoblot and Panel B shows the densitometric readings of 3 immunoblots with fractional densities normalized to actin noted above each bar. An asterisk (*) indicates p < 0.05 compared to the sample treated with 25 mM ascorbate (N = 4 experiments).

Fig. 6. Induction of SVCT2 by cobalt chloride is independent of HIF

Panel A: Effect of cobalt exposure on SVCT2 expression. Primary endothelial cells (PC) were cultured for 5 days, and exposed to 250 µM cobalt alone or in combination with 400 µM ascorbate for additional 2 days. Confluent EA.hy926 cells (EA) were used as positive controls for HIF expression. The protein extracts were subjected to immunoblot analysis using SVCT2-specific antibody. Panel B: Expression level of HIF-α in cobalt treated CCECs. The cells were cultured as described above, and the whole-cell extracts were analyzed by immunoblot analysis using HIF-1α- or HIF-2α-specifc antibodies. Panel C: Early time course of cobalt (250 µM) on HIF-1α protein expression in EA.hy926 cells and in primary culture cells. The bottom panel in all the gels shows β-actin expression to confirm equal loading of protein in the gel lanes.

Fig. 7. Generation of type IV collagen in cultured endothelial cells

EA.hy926 endothelial cells (EA, lane 1) in culture were used as a control for generation of type IV collagen sub-types. CCECs (PC) at 7 days in culture were treated overnight without or with ascorbate as indicated. Both cells and medium were taken for immunoblotting of type IV collagen. Collagen sub-types are identified based on their relative migration in the gels.