Ascorbate protects endothelial barrier function during septic insult: Role of protein phosphatase type 2A

Abstract

Endothelial barrier dysfunction contributes to morbidity in sepsis. We tested the hypothesis that raising the intracellular ascorbate concentration protects the endothelial barrier from septic insult by inhibiting protein phosphatase type 2A. Monolayer cultures of microvascular endothelial cells were incubated with ascorbate, dehydroascorbic acid (DHAA), the NADPH oxidase inhibitors apocynin and diphenyliodonium, or the PP2A inhibitor okadaic acid and then were exposed to septic insult (lipopolysaccharide and interferon-gamma). Under standard culture conditions that depleted intracellular ascorbate, septic insult stimulated oxidant production and PP2A activity, dephosphorylated phosphoserine and phosphothreonine residues in the tight junction-associated protein occludin, decreased the abundance of occludin at cell borders, and increased monolayer permeability to albumin. NADPH oxidase inhibitors prevented PP2A activation and monolayer leak, showing that these changes required reactive oxygen species. Okadaic acid, at a concentration that inhibited PP2A activity and monolayer leak, prevented occludin dephosphorylation and redistribution, implicating PP2A in the response of occludin to septic insult. Incubation with ascorbate or DHAA raised intracellular ascorbate concentrations and mitigated the effects of septic insult. In conclusion, ascorbate acts within microvascular endothelial cells to inhibit septic stimulation of oxidant production by NADPH oxidase and thereby prevents PP2A activation, PP2A-dependent dephosphorylation and redistribution of occludin, and disruption of the endothelial barrier.

Copyright 2009 Elsevier Inc. All rights reserved.

Figures

Fig. 1

Incubations with ascorbate and DHAA increase intracellular ascorbate concentration and prevent LPS+IFNγ-induced oxidant production in microvascular endothelial cells. The cells were incubated with ascorbate (500 μM), DHAA (500 μM) or DMEM vehicle (Vehicle) for 12 h. Subsequently they were incubated with LPS (25 ng/ml) + IFNγ (100 U/ml) or vehicle (BSA, Control) for 24 h. (A) Intracellular ascorbate concentration was determined by HPLC-based electrochemical assay and expressed as nmol ascorbate/mg cell protein. (B) Oxidant production was assessed by dichlorodihydrofluorescein oxidation assay and expressed as a percentage of the value for vehicle control. Plotted are mean ± SD values for 3 experiments. *P<0.05 compared to vehicle control. #P<0.05 compared to the combination of vehicle (DMEM) and LPS+IFNγ.

Fig. 2

Ascorbate and DHAA prevent LPS+IFNγ-induced barrier dysfunction. Microvascular endothelial cells were treated as described in Fig. 1. (A) Permeability of endothelial monolayers was measured with Evans blue-coupled albumin and expressed as a percentage of the value for vehicle control. (B) Cell viability was measured by CellTiter-Fluor™ cell viability assay and expressed as a percentage of the value for vehicle control. Plotted are mean ± SD values for 4 experiments. *P<0.05 compared to vehicle control. #P<0.05 compared to the combination of vehicle (DMEM) and LPS+IFNγ.

Fig. 3

Incubations with ascorbate and DHAA increase type IV collagen production in the presence or absence of delayed LPS+IFNγ. Microvascular endothelial cells were treated as described in Fig. 1 and then aliquots of media were collected for competitive ELISA assay of type IV collagen. *P

Fig. 4

Ascorbate and DHAA prevent LPS+IFNγ-induced PP2A activation. Microvascular endothelial cells were treated as described in Fig. 1. (A) PP2Ac and β-actin (loading control) protein expression was measured by western blot analysis and the cell treatments (LPS+IFNγ, ascorbate and DHAA) were found to have no detectable effect on the PP2Ac:β-actin ratio. A representative western blot is shown. (B) PP2A activity was measured as okadaic acid-inhibitable phosphatase activity and expressed as a percentage of the value for vehicle control. Plotted are mean ± SD values for 9 experiments. *P<0.05 compared to vehicle control. #P<0.05 compared to the combination of vehicle (DMEM) and LPS+IFNγ.

Fig. 5

NADPH oxidase inhibitors prevent LPS+IFNγ-induced barrier dysfunction in microvascular endothelial cell monolayers. The cells were incubated with apocynin (250 μM), DPI (5 μM) or DMEM vehicle (Vehicle) for 1 h. Subsequently they were incubated with LPS (25 ng/ml) + IFNγ (100 U/ml) or vehicle (BSA, Control) for 24 h and then monolayer permeability was measured. Plotted are mean ± SD values for 3 experiments. *P#P<0.05 compared to the combination of vehicle (DMEM) and LPS+IFNγ.

Fig. 6

NADPH oxidase inhibitors prevent LPS+IFNγ-induced PP2A activation. Microvascular endothelial cells were treated as described in Fig. 4. (A) PP2Ac and β-actin protein expression was measured by western blot analysis and the cell treatments (LPS+IFNγ, apocynin and DPI) were found to have no detectable effect on the PP2Ac:β-actin ratio. A representative western blot is shown. (B) PP2A activity. Plotted are mean ± SD values for 6 experiments. *P<0.05 compared to vehicle control. #P<0.05 compared to the combination of vehicle (DMEM) and LPS+IFNγ.

Fig. 7

Ascorbate and DHAA prevent LPS+IFNγ-induced dephosphorylation and redistribution of occludin. Microvascular endothelial cells were treated as described in Fig. 1. (A) p-Serine and p-threonine levels were measured in occludin immunoprecipitates by western blot analysis. A representative western blot is shown. (B) Ratios of phosphoserine:occludin and phosphothreonine:occludin in occludin immunoprecipitates. Plotted are mean ± SD values for 3 experiments. *P<0.05 compared to vehicle control. #P<0.05 compared to the combination of vehicle (DMEM) and LPS+IFNγ. (C) Cellular distribution of immunoreactive occludin was visualized by immunofluorescence microscopy. Representative micrographs are shown.

Fig. 8

LPS+IFNγ-induced dephosphorylation and redistribution of occludin are prevented by the PP2A inhibitor, okadaic acid. Microvascular endothelial cells were incubated with okadaic acid (OA, 0.5 nM) or DMEM vehicle (Vehicle) for 1 h and then incubated with LPS+IFNγ or vehicle (BSA, Control) for 24 h. (A) p-Serine and p-threonine levels were measured in occludin immunoprecipitates by western blot analysis. A representative western blot is shown. (B) Ratios of phosphoserine:occludin and phosphothreonine:occludin in occludin immunoprecipitates. Plotted are mean ± SD values for 3 experiments. *P<0.05 compared to vehicle control. #P<0.05 compared to the combination of vehicle (DMEM) and LPS+IFNγ. (C) Immunofluorescence micrographs of occludin in microvascular endothelial cells.

Fig. 9

Diagram of the mechanism by which intracellular ascorbate and okadaic acid (OA) prevent the induction of endothelial barrier dysfunction by LPS+IFNγ. The solid arrows indicate stimulation and the dotted arrows indicate inhibition.