A, influence of TNFα treatment on the extracellular matrix component collagen IV could be shown by Western blot. For this, cells were seeded on plastic cell culture flasks without further coating, treated as indicated and subjected to SDS-gel electrophoresis and Western blotting. An 8 h treatment with TNFα reduced the extracellular matrix component collagen IV to 66.31 ± 2.33% of control cells as estimated by densitometric quantification. This effect could be reduced by simultaneous administration of TNFα and 100 nm dexamethasone (dex): collagen levels amounted to 83.18 ± 2.9% of control cells pointing to an role for GCs in reducing TNFα-induced gelatinase activity. Data are representative of 3 independent experiments. B, densitometric evaluation of A. Measurements were performed with Scion Image Beta 4.02 and data are shown as mean ± s.d. of 3 independent experiments.

A, quantitative real-time RT-PCR analysis of MMP-9 gene expression after TNFα stimulation and dexamethasone administration in cEND cells. PCR analysis for MMP-9 showed a 1.62 ± 0.2-fold overexpression of MMP-9 in response to TNFα treatment, which could be reversed by steroid treatment (0.41 ± 0.18-fold expression for TNFα–dexamethasone-treated cells and 0.76 ± 0.13-fold expression for TNFα–hydrocortisone-treated cells). B, quantitative real-time RT-PCR analysis of TIMP-1 gene expression after TNFα stimulation and dexamethasone administration in cEND cells. Real-time RT-PCR analysis for TIMP-1 showed a 1.91 ± 0.08-fold increase in TIMP-1 expression for cells treated with TNFα–dexamethasone. Dexamethasone treatment only led to a 2.2 ± 0.09-fold increase of TIMP-1 expression, while co-treatment with dexamethasone and the antagonist RU486 abolished transcription of the TIMP-1 mRNA. C, quantitative real-time RT-PCR analysis of TIMP-2 gene expression after TNFα stimulation and dexamethasone administration in cEND cells. Analysis for TIMP-2 showed no effect on TIMP-2 gene expression for treatments with TNFα or dexamethasone or a combination thereof. Co-treatment with dexamethasone and the antagonist RU486 had no effect on TIMP-2 transcript levels. Data are shown as mean ± s.d. of 3 independent experiments.

The modulation of expression levels of the integrin α1 and αv subunits were assessed by quantitative real-time RT-PCR. An 8 h administration of TNFα led to a down-regulation of integrin α1 expression (0.78 ± 0.13-fold of control cells) while elevated levels of the αv integrin subunit of the vitronectin receptor αvβ3 were observed (1.45 ± 0.13-fold of control cells). In the presence of dexamethasone, integrin α1 was not reduced by TNFα but instead up-regulated (1.78 ± 0.25-fold of control cells). The expression levels of αv integrin remained elevated (1.6 ± 0.23-fold of control cells). Data are shown as mean ± s.d. of 3 independent experiments.

A, MMP-9 Western blot analysis showed increased secretion of MMP-9 into cell culture supernatants from cEND cells 8 h after treatment with TNFα; while MMP-9 protein could not be detected in cell culture supernatants from untreated cells, treatment with TNFα caused a significant secretion of MMP-9 into supernatants. Addition of 100 nm dexamethasone reduced MMP-9 secretion to 15.03 ± 2.16% of the value determined for TNFα-treated cells. Addition of hydrocortisone reduced MMP-9 secretion to 56.4 ± 5.26% of TNFα-treated cells. B, densitometric evaluation of A. Measurements were performed with Scion Image Beta 4.02 and data are shown as mean ± s.d. of 3 independent experiments.

A, the −1718/+95 fragment of the human TIMP-1 promoter was transfected into MCF-7 cells to test GC responsivity. Treatment of the full-length TIMP-1 promoter region −1718/+95 transfected in MCF-7 cells with 100 nm dexamethasone stimulated reporter gene expression about 2.4 + 0.18-fold. In contrast, treatment with 100 nm dexamethasone did not stimulate reporter gene expression in MCF-7 cells transfected with the TIMP-1 minimal promoter −102/+95, not containing a putative degenerate GRE (0.7 + 0.28-fold). Data are shown as mean ± s.d. of 3 independent experiments. B, TIMP-1 promoter transactivation was verified to be mediated specifically via the glucocorticoid receptor (GR): treatment of COS-1 cells cotransfected with hGRα with 100 nm dexamethasone stimulated reporter gene expression about 3.2 + 0.26-fold, while a negative control, where cotransfection of COS-1 cells with hGRα was omitted, did not show glucocorticoid responsivity. These data indicate that the observed transcriptional activation by dexamethasone is mediated via the GR and is detectable at physiological glucocorticoid concentrations. Data are shown as mean ± s.d. of three independent experiments.

A, glucocorticoid treatment prevents a compromise of BBB function in response to TNFα administration: 8 h TNFα treatment led to a reduction in TER to 65 ± 8% of control values, while treatment with hydrocortisone or dexamethasone increased TER values to 160 ± 7% and 163 ± 6% of control values, respectively. Simultaneous administration of TNFα with hydrocortisone effectively prevented barrier breakdown, TER values amounted to 96 ± 7% of control values, while simultaneous administration of TNFα with dexamethasone even led to increased tightness of the barrier (TER = 145 ± 9% of control values). B, confluent monolayers of cEND cells were grown in gelatine-coated cell culture flasks to confluence for 5 days and maintained in differentiation medium containing 2% FCS for an additional 3 days. On day 3, cells were treated with 10 nm TNFα for 8 h and cell lysates prepared. Cell lysates were analysed by Western blot for occludin, claudin-5 and ZO-1. Eight hours of TNFα treatment caused a decrease in occludin protein of 21 ± 3% as compared to the untreated control. Simultaneous administration of TNFα and hydrocortisone or dexamethasone prevented occludin loss; we were even able to detect an increase in 17 ± 5% and 49 ± 7%, respectively, as compared to untreated cells. When treated with hydrocortisone and dexamethasone alone, occludin contents were increased by 22 ± 2% and 57 ± 4%, respectively. Eight hours of TNFα treatment caused a decrease in claudin-5 protein of 24 ± 4% as compared to the untreated control. Simultaneous administration of TNFα and hydrocortisone or dexamethasone, however, did not significantly prevent claudin-5 loss, claudin-5 protein levels were reduced by 26 ± 5% and 13 ± 3%, respectively, as compared to untreated cells. When treated with hydrocortisone and dexamethasone alone, claudin-5 contents did not change significantly (88 ± 7% and 93 ± 4%, respectively). Protein levels in ZO-1 were not changed significantly by either treatment.