Katharina Mertsch

Nitric oxide protects blood-brain barrier in vitro from hypoxia/reoxygenation-mediated injury.

A cell culture model of blood‐brain barrier (BBB, coculture of rat brain endothelial cells with rat astrocytes) was used to investigate the effect of nitric oxide (⋅NO) on the damage of the BBB induced by hypoxia/reoxygenation (H/R). Permeability coefficient of fluorescein across the endothelium was used as a marker of BBB tightness. The permeability coefficient increased 5.2 times after H/R indicating strong disruption of the BBB. The presence of the ⋅NO donor S‐nitroso‐N‐acetylpenicillamine (SNAP, 30 μM), authentic ⋅NO (6 μM) or superoxide dismutase (50 units/ml) during H/R attenuated H/R‐induced increase in permeability. 30 μM SNAP or 6 μM ⋅NO did not influence the function of BBB during normoxia, however, severe disruption was observed using 150 μM of SNAP and more than 24 μM of ⋅NO. After H/R of endothelial cells, the content of malondialdehyde (MDA) increased 2.3 times indicating radical‐induced peroxidation of membrane lipids. 30 μM SNAP or 6 μM authentic ⋅NO completely prevented MDA formation. The results show that ⋅NO may effectively scavenge reactive oxygen species formed during H/R of brain capillary endothelial cells, affording protection of BBB at the molecular and functional level.

Astrocytes enhance radical defence in capillary endothelial cells constituting the blood-brain barrier.

Astrocytes (AC) induce blood‐brain barrier (BBB) properties in brain endothelial cells (EC). As antioxidative activity (AOA) is assumed to be a BBB characteristic, we tested whether AC improve AOA of EC. Monocultivated AC showed higher AOA [manganese superoxide dismutase (SOD), catalase (Cat), glutathione peroxidase (GPx)] than EC. Cocultivation elevated AOA in EC (MnSOD, CuZnSOD, Cat, GPx), and AC (MnSOD, CuZnSOD, GPx). Hypoxia increased radical‐induced membrane lipid peroxidation in monocultivated, but not in cocultivated EC. Thus, EC/AC cocultivation intensifies AOA in both cell types, protects the EC, and therefore, the BBB against oxidative stress. The high AOA is regarded as an essential property of the BBB, which is induced by AC.

Phosphorylation of vasodilator-stimulated phosphoprotein: a consequence of nitric oxide- and cGMP-mediated signal transduction in brain capillary endothelial cells and astrocytes

There is contradictory information on the relevance of nitric oxide (NO) and cGMP for the function of brain capillary endothelial cells (BCEC) forming the blood-brain barrier (BBB). Therefore, NO/cGMP-mediated signal transduction was investigated in cell cultures of BCEC and of astrocytes (AC) inducing BBB properties in BCEC. Constitutive, Ca2+-activated isoforms of NO synthase (NOS) were found in BCEC (endothelial NOS: eNOS) and in AC (neuronal NOS: nNOS), leading to increased NO release after incubation with the Ca2+-ionophore A23187. Both cell types expressed inducible NOS (iNOS) after incubation with cytokines. Soluble guanylate cyclase (sGC) was detected in both cell types. NO-dependent cGMP formation were observed in BCEC and, less pronounced, in AC. Furthermore, both cell types formed cGMP independently of NO via stimulation of particulate guanylate cyclase (pGC). cGMP-dependent protein kinase (PKG) type Ibeta, but not type II, was expressed in BCEC and AC. In BCEC, vasodilator-stimulated phosphoprotein (VASP) was detected, an established substrate of PKG and associated with microfilaments and cell-cell contacts. Phosphorylation of VASP was intensified by increased intracellular cGMP concentrations. The results indicate that BCEC and, to a smaller degree, AC can form NO and cGMP in response to different stimuli. In BCEC, NO/cGMP-dependent phosphorylation of VASP is demonstrated, thus providing a possibility of influencing cell-cell contacts.

Effect of MK-801 and U83836E on a porcine brain capillary endothelial cell barrier during hypoxia.

The present study investigated the influence of MK-801 (N-methyl-D-aspartate receptor antagonist) and U83836E (antioxidative aminosteroid) on the permeability of sodium fluorescein through a cell barrier during hypoxia (2 h 95% N2/5% CO2). The barrier consisted of porcine brain capillary endothelial cells and of cerebral rat astrocytes cultivated on two sides of a filter. After hypoxia, the permeation of fluorescein was significantly increased (10.2 +/- 1.5 x 10(-3) cm/min, P < 0.001) compared to the normoxic control (2 h 95% O2/5% CO2, 1.8 +/- 0.6 x 10(-3) cm/min). The hypoxia-enhanced permeation was significantly (P < 0.05) reduced by 10 microM MK-801 (2.0 +/- 0.5 x 10(-3) cm/min) and 10 microM U83836E (3.1 +/- 1.3 x 10(-3) cm/min). The results demonstrate, for the first time in a cell culture system, that hypoxia impairs brain endothelial barrier function, and that this enhanced permeability can be influenced pharmacologically. It is concluded that two distinct pathogenic mechanisms are involved in hypoxic cerebral endothelial cell injury, and that cerebroprotection afforded by these agents may result, in part, from reductions in edema secondary to improved blood-brain barrier function.

4-hydroxynonenal impairs the permeability of an in vitro rat blood-brain barrier

The function of the blood-brain barrier (BBB) can be impaired by free radicals. Since free radicals have only a limited diffusion capacity, we tested the possibility whether one of the major secondary lipid peroxidation product - 4-hydroxynonenal, is able to influence the permeability of the BBB. Therefore, we established an in vitro BBB model and tested its capacity to degrade 4-hydroxynonenal. Although, endothelial cells and astrocytes, possess the ability to degrade 4-hydroxynonenal the aldehyde is able to increase the permeability of the BBB. Since aldehydic lipid peroxidation products are metabolized via conjugation with glutathione we proofed that a decrease in glutathione is also able to increase the permeability of the BBB. We concluded that 4-hydroxynonenal is able to impair the BBB function via the decrease of reduced glutathione.

Cytotoxicity of spin trapping compounds

Spin trapping compounds are used frequently to detect free radicals released by cells. Their cytotoxicity has to be considered in order to prevent perturbations of normal cell growth and viability. Eleven spin traps (eight nitrones and three nitroso traps) have been tested for their effects on bovine aortic endothelial cells (toxicity range, 50% survival rate). The lowest cytotoxicity was found for 5,5‐dimethylpyrroline‐1‐oxide and 2,2,4‐trimethyl‐2H‐imidazole‐1‐oxide whereas nitrosobenzene and 2‐methyl‐2‐nitrosopropane exerted the strongest cytotoxic effects. In addition, three nitronyl nitroxides were tested. Their cytotoxicity was found to be dependent on substitution, and the toxic concentration of a lipophilic derivative was found to be more than two orders lower as compared to a hydrophilic derivative. The results of this study indicate that most spin traps can be used in cell cultures at customary (i.e. millimolar) concentrations; caution is recommended when nitroso spin traps are applied to cells.

Superoxide-Mediated Reduction of the Nitroxide Group Can Prevent Detection of Nitric Oxide by Nitronyl Nitroxides

Nitronyl nitroxides (NN), a class of compounds which react with nitric oxide forming imino nitroxides, were applied in different systems for the detection of nitric oxide. Addition of a NN to planar monolayers of bovine aortic endothelial cells (BAEC) activated by Ca2+ ionophore A23187 immediately resulted in a strong decrease of the ozone-mediated .NO chemiluminescence. Simultaneously, a rapid diminution of the electron spin resonance (ESR) signal intensity of the NN (without detectable formation of the corresponding imino nitroxide) was observed; superoxide dismutase partially inhibited this decrease in the NN concentration. Model experiments using hypoxanthine/xanthine oxidase in aqueous solution and KO2 in dimethylsulfoxide as sources of O2.- revealed that there is a rapid reduction of nitronyl nitroxides by superoxide. The second order rate constant for the reaction of the water soluble NN with O2.- was determined to be 8.8 x 10(5) M-1s-1, which is more than two orders of magnitude higher than the value reported previously for reaction with .NO (Woldman et al., BBRC 202, 195-203, 1994). Reduction of the nitronyl nitroxide was also observed in the presence of glutathione, ascorbic acid or rabbit liver microsomes. Incorporation of both nitronyl and imino nitroxides into liposomes strongly decreased reduction by superoxide and other reductants, however, in the presence of microsomes, there was no protective effect by liposomal encapsulation of NN. The results indicate that in biological systems (in addition to other reducing agents) the presence of superoxide can prevent the detection of nitric oxide using nitronyl nitroxides.

Protective effects of the thiophosphate amifostine (WR 2721) and a lazaroid (U83836E) on lipid peroxidation in endothelial cells during hypoxia/reoxygenation

Little is known about pharmacological interventions with thiophosphates or lazaroids in endothelial cells injured by hypoxia/reoxygenation with respect to membrane lipid peroxidation (LPO) caused by reactive oxygen species. Therefore, a cell line of bovine aortic endothelial cells was studied after 120-min hypoxia followed by 30-min reoxygenation, resulting in moderate and predominantly reversible injury (energy depression/cytosolic Ca2+-accumulation during hypoxia, which almost normalized during reoxygenation; membrane blebs, an increasing amount of lysosomes, vacuolization, lipofuscin formation, alterations in mitochondria size, some lyzed cells). 18.9 +/- 4.3% of the cells died. Radical-induced LPO measured as malondialdehyde continuously increased to 2.18 +/- 0.17 nmol/mg of protein after reoxygenation vs control (0.41 +/- 0.13, P < 0.05). Simultaneously, the content of 4-hydroxynonenal, a novel indicator of LPO, increased from 0.02 +/- 0.01 to 0.11 +/- 0.02 nmol/mg of protein (P < 0.01). The results support the assumption that reoxygenation injury is accompanied by an increase in membrane LPO, causing structural and functional disturbances in the monolayer. The thiophosphate WR 2721 [S-2-(3-aminopropylamino) ethylphosphorothioic acid] and the lazaroid U83836E [(-)-2-[[4-(2,6-di-1-pyrrolidinyl-4-pyrimidinyl)-1-piperazinyl] methyl]-3,4-dihydro-2,5,7,8-tetramethyl-2H-1-benzopyran-6-ol (dihydrochloride)] were effective scavengers of .OH, being more efficient than trolox C (6-hydroxy-2,5,7,8-tetramethylchroman-2-carbon acid) used as standard (EC50: 12, 5 and 15 microM, respectively, measured by electron spin resonance spectroscopy). One mM WR 2721, 10 microM U83836E, and 5 microM trolox C reduced formation of malondialdehyde during hypoxia/reoxygenation to 53 +/- 7, 51 +/- 10 and 48 +/- 6%, respectively (P < 0.05 each, versus control). In general, WR 2721 and U83836E prevent radical-induced membrane LPO in a model of endothelial cells injured by hypoxia/reoxygenation. The use of these two agents is a new approach to protect the endothelium against oxidative stress.

High metabolic rates of 4-hydroxynonenal in brain capillary endothelial cells during hypoxia/reoxygenation.

We measured the accumulation of 4-hydroxynonenal (HNE), a major lipid peroxidation product during hypoxia/reoxygenation of brain capillary endothelial cells (BCEC). The concentration of HNE after 2 h of hypoxia was 0.23 nmol/mg protein and rose up to 0.28 nmol/mg protein after 30 min of reoxygenation. That reflects a 1.5-fold increase, whereas aortic endothelial cells (AEC) increased the HNE level 5-fold, compared to the control. Therefore, the ability of BCEC to degrade exogenously added HNE was tested. The HNE consumption in BCEC achieved a rate of about 600 nmol.min-1.mg protein-1, about two times higher than in AEC. The higher ability of BCEC to degrade HNE is probably the reason of the 2-fold higher IC50 value against the aldehyde. Therefore, we concluded that the high ability of BCEC to degrade HNE is a substantial part of the secondary antioxidative defense of the brain.