Alexander Mülsch

INDUCTION OF NITRIC OXIDE SYNTHASE BY CYTOKINES IN VASCULAR SMOOTH MUSCLE CELLS

We investigated the mechanisms by which cytokines lead to a diminished responsiveness of vascular smooth muscle to vasoconstrictors. The attenuation of noradrenaline‐induced contraction by 6 to 24 h incubations with the cytokines, tumor necrosis factor and interleukin‐1, in endotheliumdenuded rabbi, aorta was associated with an increase in intracellular cyclic GMP level. This increase was abolished by the stereoselective inhibitor of nitric oxide‐synthase. N o‐nitro‐L‐arginine and by cycloheximide. Formation of nitric oxide was detected in the cytosol of cytokine‐treated native and cultured smooth muscle cells by activation of purified soluble guanylate cyclase, and depended on tetrahydrobiopterin, but not on ? ‐calmodulin. The results indicate that cytokines induce a nitric oxide‐synthase of the macrophage‐type in vascular smooth muscle.

Calcium‐dependent nitric oxide synthesis in endothelial cytosol is mediated by calmodulin

We investigated whether calmodulin mediates the stimulating effect of Ca2+ on nitric oxide synthase in the cytosol of porcine aortic endothelial cells. Nitric oxide was quantified by activation of a purified soluble guanylate cyclase. The Ca2+‐sensitivity of nitric oxide synthase was lost after anion exchange chromatography of the endothelial cytosol and could only be reconstituted by addition of calmodulin or heat‐denatured endothelial cytosol. The Ca2+‐dependent activation of nitric oxide synthase in the cytosol was inhibited by the calmodulin‐binding peptides/proteins melittin, mastoparan, and calcineurin (IC50 450, 350 and 60 nM, respectively), but not by the calmodulin antagonist, calmidazolium. In contrast, Ca2+‐calmodulin‐reconstituted nitric oxide synthase was inhibited with similar potency by melittin and calmidazolium. The results suggest that the Ca2+‐dependent activation of nitric oxide synthase in endothelial cells is mediated by calmodulin.

Differential role of extra- and intracellular calcium in the release of EDRF and prostacyclin from cultured endothelial cells

1 The effects of extracellular Ca2+ on the release of endothelium‐derived relaxing factor (EDRF) and prostacyclin (PGI2), and on the intracellular free calcium concentration ([Ca2+]i), were studied in cultured bovine aortic endothelial cells. 2 Receptor‐mediated stimulation of endothelial cells with bradykinin (10 nM) elicited a transient release of EDRF (assayed by its stimulant effect on purified soluble guanylate cyclase) and of PGI2 (measured by radioimmunoassay for 6‐keto prostaglandin F1α). 3 Bradykinin (10 nM) also increased [Ca2+]i (measured with the fluorescent probe indo‐1) from 125 ± 11 nM to 631 ±59 nM, with the same time course as for autacoid release. 4 In Ca2+‐free medium, [Ca2+]i was still increased by bradykinin but declined faster (within 1 min) to resting levels than in the presence of extracellular Ca2+. 5 PGI2 release was almost completely abolished in Ca2+‐free medium. The intracellular calcium antagonist TMB‐8 evoked a similar inhibition of PGI2 release. 6 In contrast, bradykinin‐induced EDRF release was not significantly affected by TMB‐8 but was completely abolished in Ca2+‐free medium. 7 When endothelial cells were stimulated with the receptor‐independent drug thimerosal (an inhibitor of the enzyme acyl‐CoA‐lysolecithin‐acyl‐transferase; 5 μM), a long‐lasting release of EDRF (>90 min) and PGI2 (>20 min) was observed. 8 In contrast to bradykinin stimulation, thimerosal‐induced autacoid release was associated with only a slight increase of [Ca2+]i to 201 ± 13 nM after 40min. 9 After removal of extracellular Ca2+ from thimerosal‐stimulated endothelial cells, [Ca2+]i was little affected during the observation time of 90 s. EDRF release was completely abolished within 90 s whereas PGI2 release was unchanged. 10 We conclude that EDRF production is directly controlled by extracellular Ca2+ during both receptor‐dependent and independent stimulation. This effect of extracellular Ca2+ is not mediated by changes in [Ca2+]i. In contrast, PGI2 release is closely correlated to [Ca2+]i in bradykinin‐stimulated endothelial cells. However, the results obtained during thimerosal stimulation indicate that there is not necessarily a tight coupling between the absolute level of [Ca2+]i and the amount of PGI2 released.

Effects of native and oxidized low density lipoproteins on formation and inactivation of endothelium-derived relaxing factor

The influence of native (N-) and oxidized (Ox-) low density lipoproteins (LDLs) on endothelium-dependent vasomotion is still controversial. We investigated the short-term effects of N-LDL and Ox-LDL on the formation of endothelium-derived relaxing factor (EDRF) in native and cultured endothelial cells and on its inactivation after release from the cells. N-LDL was isolated from fresh human plasma via sequential ultracentrifugation and oxidized by incubation with Cu2+. EDRF released from cultured endothelial cells was inactivated by both N-LDL and Ox-LDL (1 mg/ml) as detected in a bioassay system. N-LDL reduced the EDRF-mediated vasodilations of the detector segments by 38.5 +/- 5.3%, and Ox-LDL, by 55.5 +/- 4.6%. The effects of lipoproteins on EDRF formation were studied in cultured endothelial cells preincubated with either N-LDL or Ox-LDL (1 mg/ml for 1 hour) and stimulated for EDRF release with bradykinin after washout of the lipoproteins. EDRF was assessed by measuring its stimulatory effect on the activity of a purified, soluble guanylate cyclase. Both N-LDL and Ox-LDL did not reduce the bradykinin-induced EDRF formation. Consistent with this finding, acetylcholine-induced, EDRF-mediated dilations of intact rabbit femoral artery segments were not impaired by luminal exposure to N-LDL or Ox-LDL (1 mg/ml for 1 hour). However, these relaxations were significantly reduced by preincubation of aortic ring preparations with the same concentrations of the same charges of N-LDL and Ox-LDL. In conclusion, neither N-LDL nor Ox-LDL acutely impairs the formation of EDRF but does inactivate EDRF after its release from endothelial cells.(ABSTRACT TRUNCATED AT 250 WORDS)

On-line detection of nitric oxide formation in liquid aqueous phase by electron paramagnetic resonance spectroscopy.

A method for the detection of the nitric oxide radical (NO) in oxygen-containing aqueous solution by means of electron paramagnetic resonance spectroscopy (EPR) is described. NO evolving from the spontaneous decomposition of 3-morpholinosydnonimine (SIN-1) was trapped by Fe(2+)-diethyldithiocarbamate (DETC) complex dissolved in yeast cell membranes. The resulting mononitrosyl-Fe(2+)-(DETC)2 complex was stable and exhibited a characteristic EPR signal at g perpendicular = 2.04 and g parallel = 2.02 with an unresolved triplet hyperfine structure at g perpendicular in frozen solution and an isotropic triplet signal at gav = 2.03 at 37 degrees C. The amount of NO trapped was calculated from the amplitude of one of the triplet lines calibrated by means of a dinitrosyl-Fe(2+)-thiosulfate standard. The lower detection limit of NO was 0.5 nmol/(ml x h) due to a low background NO signal. The upper detection limit was about 10 nmol NO/40 mg traps (DETC-loaded yeast cells), because of saturation of traps. The trapping efficiency approached 60% under anaerobic conditions and with low concentrations of SIN-1, but decreased progressively with higher concentrations and in the presence of oxygen. Nitrite (up to 0.1 mM) did not increase the background NO level. The sensitivity was sufficient to follow the rate of NO release from SIN-1 on-line at 37 degrees C in a flat quartz cuvette. The time course of NO release detected by EPR spectrometry correlated with the time course of nitrite accumulation measured by diazotation. In conclusion, this method will permit the on-line detection of NO formation from endogenous and pharmacological sources in oxygen-containing aqueous media.

The potent vasodilating and guanylyl cyclase activating dinitrosyl‐iron(II) complex is stored in a protein‐bound form in vascular tissue and is released by thiols

We studied the biological activity, stability and interaction of dinitrosyl‐iron(II)‐L‐cysteine with vascular tissue. Dinitrosyl‐iron((II)‐L‐cysteine was a potent activator of purified soluble guanylyl cyclase (EC50 (nM with and 100 nM without superoxide dismutase) and relaxed noradrenaline‐precontracted segments of endothelium‐denuded rabbit femoral artery (EC50 10 nM superoxide dismutase). Pre‐incubation (5 min; 310 K) of endothelium‐denuded rabbit aortic segments with dinitrosyl‐iron(II)‐L‐cysteine (0.036–3.6 mM) resulted in a concentration‐dependent formation of a dinitrosyl‐iron(II complex with protein thiol groups, as detected by ESR spectroscopy. While the complex with proteins was stable for 2 h at 310 K, dinitrosyl‐iron(II)‐L‐cysteine in aqueous solution (30–360 μM) decomposed completely within 15 min, as indicated by disappearance of its isotropic ESR signal at g av = 2.03 (293 K). Aortic segments pre‐incubated with dinitrosyl‐iron(II)‐L‐cysteine released a labile vasodilating and guanylyl cyclase activating factor. Perfusion of these segments with N‐acetyl‐L‐cysteine resulted in the generation of a low molecular weight dinitrosyl‐iron(II)‐dithiolate from the dinitrosyl‐iron(II) complex with proteins, as revealed by the shape change of the ESR signal at 293 K. The low molecular weight dinitrosyl‐iron(II)‐dithiolate accounted to an enhanced guanylyl cyclase activation and vasodilation induced by the aortic effluent. We conclude that nitric oxide (NO) produced by, or acting on vascular cells can be stabilized and stored as a dinitrosyl‐iron(II) complex with protein thiols, and can be released from cells in the form of a low molecular weight dinitrosyl‐iron(II)‐dithiolate by intra‐ and extracellular thiols.

Diethyldithiocarbamate inhibits induction of macrophage NO synthase

We investigated whether sodium diethyldithiocarbamate (DETC), an inhibitor of the nuclear transcription factor kappa B (NFkappaB), modulates induction of NO synthase (NOS) in murine bone marrow‐derived macrophages. A short exposure (between 1 and 16 h) of L929‐cell medium‐preconditioned macrophages to E. coli lipopolysaccharide (LPS) significantly increased the level of NOS mRNA, and elicited NO formation as detected by electron spin resonance spectroscopy and by the release of nitrite. DETC (0.1–1 mM) present during stimulation with LPS prevented the increase in NOS mRNA and the expression of NOS activity. These findings suggest that NFkappaB is involved in the signal transduction pathway linking stimulation of macrophages by LPS with transcription of, the gene encoding inducible NOS.

Nitric oxide synthesis in endothelial cytosol: Evidence for a calcium-dependent and a calcium-independent mechanism

SummaryRelease of nitric oxide (NO) from endothelial cells critically depends on a sustained increase in intracellular free calcium maintained by a transmembrane calcium influx into the cells. Therefore, we studied whether the free cytosolic calcium concentration directly affects the activity of the NO-forming enzyme(s) present in the cytosol from freshly harvested porcine aortic endothelial cells. NO was quantified by activation of a purified soluble guanylate cyclase coincubated with the cytosol. In the presence of 1 mM L-arginine, 0.1 mM NADPH and 0.1 mM EGTA, endothelial cytosol (0.2 mg of cytosolic protein per ml) stimulated the activity of guanylate cyclase 5.0 + 0.5-fold (from 31 + 9 to 153 + 15 nmol cyclic GMP formed per min per mg guanylate cyclase). Calcium chloride increased this stimulation further in a concentration-dependent fashion by up to 136 + 15% (with 2 μM free calcium; EC50 0.3 μM). The calcium-dependent and -independent activation of guanylate cyclase was enhanced by superoxide dismutase (0.3 μM) and was inhibited by the stereospecifically acting inhibitor of L-arginine-dependent NO formation NG-nitro-L-arginine (1 mM) and by LY 83583 (1 μM), a generator of superoxide anions. Our findings suggest a calcium-dependent and -independent synthesis of NO from L-arginine by native porcine aortic endothelial cells.

The relationship between L-arginine-dependent nitric oxide synthesis, nitrite release and dinitrosyl-iron complex formation by activated macrophages

We identified the source of the nitrogen included into nitric oxide (NO) and studied the relationship between formation of NO, intracellular dinitrosyl ferrous iron complex (DNIC) and release of nitrite by murine bone-marrow-derived macrophages stimulated with E. coli lipopolysaccharide (LPS). NO was trapped in the cell membrane by iron-diethyldithiocarbamate complex (FeDETC) and was detected as a paramagnetic NOFe(DETC)2 complex by electron paramagnetic resonance (EPR) spectroscopy. Macrophages stimulated for 7 h up to 48 h with LPS and then incubated for 2 h with DETC exhibited an anisotropic EPR signal of axial symmetry with g-factor values g perpendicular = 2.035, g parallel = 2.02 and a triplet hyperfine structure (hfs) at g perpendicular characteristic for NOFe(DETC)2. In cells incubated with [15NG]L-arginine instead of [14NG]L-arginine the EPR signal of [15N]OFe(DETC)2 was detected with a doublet hfs at g perpendicular, indicating that NO was generated exclusively from the terminal guanidino-nitrogen of extracellular L-arginine. The ratio of NO formation and of nitrite release changed with time of exposure to LPS, nitrite exceeding NO at early stages of macrophage activation, and NO exceeding nitrite at later stages. DNIC with thiolate ligands (0.5 nmol/10(7) cells) was observed in stimulated macrophages not loaded with DETC. Furthermore, DNIC released from macrophages was trapped in the extracellular medium by bovine serum albumin (BSA) (1 nmol/10(7) cells per 2 h) by formation of a paramagnetic DNIC with BSA. DNIC release not only provides a route for iron loss from activated macrophages, but may also play a role in the cytotoxic and microbiostatic activity of macrophages.

LY 83583 (6-anilino-5,8-quinolinedione) blocks nitrovasodilator-induced cyclic GMP increases and inhibition of platelet activation.

SummaryWe studied the effects and the mechanism of action of the cyclic GMP-lowering substance 6-anilino-5,8-quinolinedione (LY 83583) on cyclic GMP-mediated inhibition of platelet function. The activation of washed human platelets by thrombin was counteracted by 8-bromo-cyclic GMP and the direct activators of soluble guanylate cyclase, sodium nitroprusside and endothelium-derived relaxant factor (EDRF = nitric oxide). LY 83583 significantly antagonized the inhibitory effect of sodium nitroprusside and EDRF, but not that of 8-bromo-cyclic GMP, on thrombin-induced aggregation, ATP-release, adhesion to native endothelial cells and increase in concentration of free intracellular calcium ions. In accordance, increases in intracellular cyclic GMP by sodium nitroprusside and EDRF were attenuated by LY 83583. The inhibition of cyclic GMP-mediated effects on platelets by LY 83583 could be related to inhibition of platelet soluble guanylate cyclase, as the activation of the purified enzyme from platelets by sodium nitroprusside was directly inhibited by LY 83583. This effect of LY 83583 was attenuated in the presence of superoxide dismutase. Our findings support the hypothesis that sodium nitroprusside and EDRF inhibit platelet activation by stimulation of soluble guanylate cyclase via nitric oxide. Consequently, inhibition of nitric oxide-induced cyclic GMP formation by LY 83583, which may act by intracellular generation of superoxide anions, facilitates platelet activation.