Cytokine and intracellular signaling regulation of tissue factor expression in astrocytes

There is evidence that inflammatory cytokines such as IL-1beta, TNFalpha, and IL-6 are involved in the pathogenesis of cerebrovascular disorders including stroke. One action of cytokines that contributes to diseases in peripheral tissues is upregulation of the procoagulant receptor tissue factor (TF). In the CNS, astrocytes are the primary cells that express TF; although little is known about how TF is regulated in these cells. Experiments were performed to evaluate the effect of cytokine treatment on TF activity in primary cultures of murine cortical astrocytes and in the human astrocytoma cell line (CCF). IL-1beta treatment induced a 2.5-fold increase in TF activity in the primary astrocytes and a 3-fold induction in the astrocytoma cells. TNFalpha treatment induced a 2.5-fold increase in TF activity in both the primary astrocytes and astrocytoma cells. IL-6 upregulated TF activity 2-fold in primary astrocytes, however, it had no effect on TF activity in the astrocytoma cells. The signaling pathways regulating TF expression in these cells were examined by using staurosporine, a broad spectrum inhibitor of serine-threonine protein kinases, and by examining the effects of intermediates in the sphingomyelin signaling pathway. Staurosporine inhibited IL-1beta-induced TF activity in the primary astrocytes but did not effect IL-1beta- or TNFalpha-induced TF activity in the astrocytoma cells. TF activity in the astrocytoma cells was upregulated 1.5-fold over constitutive levels by a ceramide analogue or the enzyme sphingomyelinase, however the ceramide analogue had no effect on TF activity in the primary astrocytes. These results suggest inflammatory cytokines can upregulate TF activity in astrocytes and the astrocytoma CCF cell line although the two cell types appear to utilize different signaling pathways to mediate TF expression. Further studies will be important to more completely define the signaling regulation of TF in astrocytes since alterations in brain TF levels may play a key role in CNS pathophysiology.

TNF-alpha and IL-1alpha induce heme oxygenase-1 via protein kinase C, Ca2+, and phospholipase A2 in endothelial cells

Heme oxygenase-1 (HO-1), an enzyme important in protection against oxidant stress, is induced in human vascular endothelial cells by the cytokines tumor necrosis factor-alpha (TNF-alpha) and interleukin-1alpha (IL-1alpha). However, the signaling mediators that regulate the induction are not known. This study examined the involvement of protein kinase C (PKC), phospholipase A2 (PLA2), calcium, and oxidants in cytokine induction of HO-1. Acute exposure to the PKC activator phorbol 12-myristate 13-acetate (PMA) stimulated HO-1 mRNA. However, prolonged exposure, which downregulates most PKC isoforms, blocked induction of HO-1 mRNA by IL-1alpha and TNF-alpha. Additionally, the phosphatase inhibitors okadaic acid and calyculin enhanced cytokine induction of HO-1. Mepacrine, a PLA2 inhibitor, prevented HO-1 induction by cytokine, suggesting a role for arachidonate, the product of PLA2 hydrolysis of phospholipids, in HO-1 expression. The intracellular calcium chelator 1,2-bis(2-aminophenoxy)ethane-N,N,N',N'-tetraacetic acid acetoxymethyl ester (BAPTA-AM) blocked cytokine induction of HO-1. Paradoxically, the calcium ionophore A-23187 prevented HO-1 induction by cytokine but not by PMA. Finally, the oxidant scavenger N-acetylcysteine inhibited HO-1 induction by cytokines. These results demonstrate that TNF-alpha and IL-1alpha induction of HO-1 requires PKC-mediated phosphorylation and PLA2 activation as well as oxidant generation.

Effect of tumor necrosis factor-alpha and interleukin-1 alpha on heme oxygenase-1 expression in human endothelial cells

Heme iron exacerbates oxidant damage by catalyzing the production of free radicals. Heme oxygenase is the rate-limiting enzyme involved in heme catabolism. An inducible form of heme oxygenase, heme oxygenase-1 (HO-1), is upregulated in oxidant and inflammatory settings, and recent work suggests that HO-1 induction may serve a protective function against oxidant injury. The ability of the endogenous inflammatory mediators, interleukin (IL)-1 alpha, tumor necrosis factor-alpha (TNF-alpha), and IL-6, to enhance HO-1 expression in cultured human endothelial cells was examined in this study. HO-1 mRNA and protein expression were upregulated by IL-1 alpha and TNF-alpha exposure but not by IL-6. Induction of HO-1 mRNA by IL-1 alpha and TNF-alpha occurred in a concentration- and time-dependent fashion, with maximal expression occurring by 4 h for both cytokines. Induction depended on protein synthesis and occurred at the transcriptional level. Inhibition of the AP-1 transcription factor with curcumin decreased the cytokine induction of HO-1 mRNA, suggesting the involvement of this transcription factor in cytokine signaling of HO-1. The results of this study indicate that the endogenous inflammatory cytokines IL-1 alpha and TNF-alpha induce HO-1 in endothelial cells, providing further evidence that HO-1 may be an important cellular response to inflammatory stress.

ACE inhibitor effects on platelet function in stages I-II hypertension

Angiotensin II enhances platelet aggregation through activation of the G protein-linked pathway present in platelets. Studies of several angiotensin-converting enzyme (ACE) inhibitors have demonstrated marked differences on platelets. Therefore this prospective, randomized, double-blind, crossover study compared the ex vivo effects of equivalent antihypertensive doses of captopril, enalapril, and fosinopril on platelet aggregation and thromboxane B2 (TxB2) formation in subjects with stage I-II essential hypertension. Nineteen male subjects with a baseline mean seated blood pressure of 141 +/- 3/100 +/- 1 mm Hg were enrolled. The decline in mean arterial pressure after 4 weeks of stable dosing was 10 +/- 1, 12 +/- 1, and 11 +/- 1 mm Hg for captopril, enalapril, and fosinopril, respectively (p = NS). There was no significant change in adenosine diphosphate (ADP)-, epinephrine-, or thrombin-stimulated platelet aggregation from baseline or between ACE inhibitors. Compared with baseline, fosinopril decreased TxB2 concentrations 27.5-67.6% with all stimuli after 1 and 5 min. Captopril also decreased TxB2 formation, but this effect was stimulus and time dependent. Enalapril consistently increased TxB2 concentrations, independent of stimuli or time. We conclude that different ACE inhibitors have distinct effects on platelet TxB2 formation without significant effects on platelet aggregation. Fosinopril may be a direct antagonist ofTxA2 synthase, suggesting benefit in syndromes of platelet activation or vascular occlusion.

Mechanisms of ionomycin-induced endothelial cell barrier dysfunction

Myosin light chain (MLC) phosphorylation catalyzed by the Ca(2+)- calmodulin-dependent MLC kinase (MLCK) is critical to thrombin-mediated endothelial cell gap formation and barrier dysfunction. We have tested the hypothesis that the Ca2+ ionophore ionomycin stimulates MLCK-dependent endothelial cell contraction and permeability. Ionomycin significantly increased albumin clearance and decreased electrical resistance across confluent bovine pulmonary microvascular and macrovascular endothelial cell monolayers in a concentration-dependent manner that was temporally similar to that produced by thrombin. In contrast, however, ionomycin produced a significant Ca(2+)-dependent reduction in the levels of phosphorylated MLC with evidence of serine/threonine phosphatase activation. Potential MLCK-independent mechanisms of endothelial cell permeability were examined with little evidence to support a role for stimulated nitric oxide synthase or phospholipase A2 activities. Importantly, ionomycin produced 1) reductions in the activities of the barrier protective adenylate cyclase and the adenosine 3',5'-cyclic monophosphate-dependent protein kinase A, 2) dramatic dose- and time-dependent inhibition of endothelial cell tyrosine kinase activities, and 3) marked decreases in the phosphotyrosine content of the p125 focal adhesion kinase. These data indicate that ionomycin produces endothelial cell barrier dysfunction by mechanisms that are independent of MLCK activation and may involve reductions in endothelial cell tethering forces via inhibition of protein kinase A and tyrosine kinase activities, especially the p125 focal adhesion kinase.

Circulating concentrations of proinflammatory cytokines in mild or moderate heart failure secondary to ischemic or idiopathic dilated cardiomyopathy

Proinflammatory cytokines are capable of modulating cardiovascular function by a variety of mechanisms. These cytokines are elevated in patients with severe heart failure, but changes in mild or moderate heart failure have not been reported. Therefore, simultaneous arterial and coronary sinus concentrations of interleukin-1alpha, soluble interleukin-2 receptor, interleukin-6, and tumor necrosis factor-alpha were measured in 78 patients with New York Heart Association functional class II to IV heart failure and compared with 17 healthy volunteers. Concentrations of interleukin-1alpha, soluble interleukin-2 receptor, and interleukin-6 were determined by a "sandwich" enzyme-linked immunosorbent assay and tumor necrosis factor-alpha by tissue culture technique. There were no statistical differences in interleukin-1alpha, soluble interleukin-2 receptor, or tumor necrosis factor-alpha concentrations in mild to moderate heart failure versus control subjects. Interleukin-6 was significantly elevated, 75 +/- 16 versus 0.4 +/- 0.4pg/ml (p = 0.002). Cytokine concentrations did not differ by heart failure etiology. Paired arterial and coronary sinus concentrations were not significantly different. Soluble interleukin-2 receptor concentrations were significantly correlated with New York Heart Association functional class (r = 0.59, p = 0.04) and negatively associated with exercise tolerance time (r = -0.59, p = 0.007). Thus, interleukin-6 is significantly elevated in mild or moderate heart failure.

Protein kinase C regulates cytokine-induced tissue factor transcription and procoagulant activity in human endothelial cells

Interleukin-1 alpha (IL-1 alpha) and tumor necrosis factor-alpha (TNF-alpha) induce tissue factor in endothelium, which results in activation of the coagulation cascade. Despite extensive investigation, in which various stimuli that induce tissue factor have been defined, the intracellular processes that control tissue factor expression are not well understood. It has been proposed that protein kinase C regulates tissue factor expression primarily because phorbol myristate acetate, the protein kinase C activator, induces tissue factor expression. In this study we examined whether IL-1 alpha- or TNF-alpha-stimulated tissue factor production is regulated through a protein kinase C-dependent mechanism. Northern blot analysis showed that cytokine-induced tissue factor mRNA was significantly reduced in human umbilical vein endothelial cells treated with calphostin C, a specific protein kinase C inhibitor. Tissue factor functional activity was decreased in the presence of calphostin C as well. Calphostin C also inhibited phorbol myristate acetate-induced tissue factor expression. In contrast, calphostin C did not alter cytokine induction of E-selectin or prostacyclin release. Because calcium stimulates protein kinase C binding to the membrane and its resulting catalytic activity, human umbilical vein endothelial cells were exposed to IL-1 alpha or TNF-alpha in the presence of calcium ionophore A23187. A23187 had little effect alone but significantly augmented cytokine stimulation of tissue factor mRNA. Okadaic acid, a phosphatase inhibitor, increased cytokine-induced tissue factor mRNA compared with cytokine alone, which suggests that a phosphorylation event is important in tissue factor expression. These results indicate that protein kinase C is involved in cytokine activation of endothelial cell tissue factor expression.

Oxidant exposure stimulates cultured coronary artery endothelial cells to release 15-HETE: differential effects on PGI2 and 15-HETE synthesis

Oxidant stress to the endothelium is an important component of inflammatory processes involved in the pathogenesis of ischemic/reperfusion injury. The effects of acute oxidant exposure on cultured bovine coronary artery endothelial cell (BCA) functions including arachidonic acid metabolism, permeability, tissue factor expression, and viability were assessed after exposure of cells to the hydrogen peroxide-generating system of glucose-glucose oxidase (GO). GO markedly stimulated the synthesis of the arachidonic acid metabolites 15-hydroxyeicosatetraenoic acid (15-HETE) and prostacyclin (PGI2). Both sublethal and lethal concentrations of GO increased 15-HETE release from BCAs by as much as 15-fold. In contrast to 15-HETE, enhanced PGI2 synthesis occurred at concentrations of GO that did not injure the BCA monolayers, whereas lethal doses of GO had no stimulatory effect on PGI2 production. Moreover, the sublytic oxidant-induced stimulation of PGI2 synthesis in BCAs (50-fold) was significantly greater than that induced by other mediators or that observed in parallel studies with human umbilical vein endothelial cells. In vitro endothelial cell barrier function was determined by measuring iodine 125-labeled albumin clearance across confluent cell monolayers. GO increased cellular permeability in a concentration-dependent manner, although statistically significant increases were only observed at the highest (i.e., lethal) concentrations (C(alb) = 0.840 +/- 0.16 with 1.0 U/ml GO vs C(alb) = 0.24 +/- 0.02 in control cells). Finally, oxidant exposure did not induce BCA tissue factor activity at any concentration examined. These results suggest that oxidant exposure, as might occur during ischemic reperfusion, could affect subsequent coronary vascular responses by releasing the arachidonate metabolite 15-HETE, which can cause vasoconstriction as well as attract and activate leukocytes. In addition, oxidants may also modulate vascular reactivity by altering the release of the potent vasodilator and neutrophil modulator PGI2 as lower levels of oxidant generation stimulate its synthesis, whereas higher levels suppress PGI2 release. Thus the degree of oxidant stress may profoundly affect the endothelial synthesis and release of 15-HETE and PGI2, compounds with antagonist effects on vascular tone and neutrophil activation. Consequently the balance between oxidant-induced production of these mediators by the coronary endothelium may significantly affect the pathophysiology of myocardial ischemia and reperfusion injury.

Hyperoxic exposure in humans. Effects of 50 percent oxygen on alveolar macrophage leukotriene B4 synthesis

The pathogenesis of oxygen toxicity remains unknown but may involve leukocyte mediated injury. The effects of hyperoxia on several lower respiratory tract parameters were examined in bronchoalveolar lavage fluid of normal nonsmoking subjects who inhaled a fractional inspired oxygen concentration of 50 percent (mean exposure: 44 h). Evidence that 50 percent O2 produced oxidative stress in the lung included recovery of fluorescent products of lipid peroxidation and partial oxidation of alpha 1-antitrypsin in BAL fluid obtained after O2 exposure. To examine whether alveolar macrophage-derived leukotriene B4 may be generated in response to 50 percent O2, AM were isolated from O2-exposed subjects and compared with AM recovered from subjects breathing room air. Leukotriene B4 levels were elevated in supernatants from both unstimulated and arachidonic acid-stimulated AM obtained from hyperoxia-exposed subjects. In hyperoxia-exposed individuals, LTB4 levels were also elevated in extracted BAL fluid. The percentage of BAL neutrophils was also significantly increased after O2 exposure (2.8 +/- 0.6 vs 1.2 +/- 0.4 percent, p = 0.05). We conclude that an FIO2 of 50 percent inhaled for 44 h is associated with enhanced oxidative stress, stimulation of AM to release LTB4, and a small but significantly increased percentage of neutrophils recovered in BAL fluid.

Asbestos exposure results in increased lung procoagulant activity in vivo and in vitro

Enhanced fibrin deposition is a common histologic finding in fibrotic lung disorders including asbestosis and may be an important mechanism by which fibroblast proliferation is modulated. Asbestos-induced activation of lung interstitial cells may result in enhanced expression of procoagulant activity which contributes to the inflammatory response resulting in subsequent fibrin deposition. The current study examines procoagulant activity in bronchoalveolar lavage fluid from patients with clinically diagnosed asbestosis, patients with asbestos exposure without asbestosis, and normal, control subjects. Results indicated that asbestos exposure resulted in increased lung procoagulant expression in vivo, and furthermore, suggested that both endothelial cells and alveolar macrophages represented lung parenchymal cells which may contribute to this activity. This imbalance in coagulation homeostasis may be important in the regulation of fibrotic responses observed in asbestosis.

Thrombin-induced prostacyclin biosynthesis in human endothelium: role of guanine nucleotide regulatory proteins in stimulus/coupling responses

The regulation of prostacyclin (PGI2) synthesis by cultured human umbilical vein endothelium (HUVEC) was investigated. HUVEC monolayer generation of PGI2 was monitored by RIA of 6-keto PGF1 alpha and dose-dependent increases observed with human alpha- and gamma-thrombins, histamine, or arachidonate. Alpha thrombin (10 nM) produced levels of 6-keto PGF1 alpha approximating responses with 1 microM gamma-thrombin, 5 microM arachidonate, or 10 microM histamine. Diisopropyl phosphorofluoridate-inactivated alpha-thrombin did not stimulate PGI2 release, demonstrating that catalytic activity was required for thrombin-stimulated PGI2 release. Sodium fluoride (NaF), at concentrations known to activate guanine nucleotide regulatory proteins (G proteins), directly stimulated HUVEC PGI2 synthesis in a dose-dependent and time-dependent manner (20 mM NaF, 4.4 +/- 0.5-fold increase at 10 min, 11.9 +/- 1.5-fold increase at 30 min). Neither alpha-thrombin nor NaF-stimulated PGI2 release was dependent upon the availability of extracellular Ca++). The hypothesis that G proteins are involved in agonist-stimulated PGI2 synthesis was further supported by studies using digitonin-permeabilized HUVEC monolayers challenged with another G protein activator, guanosine 5'-0-3-thiotrisphosphate (GTP gamma S), which effected significant dose-dependent increases in PGI2 synthesis compared with control levels of 6-keto PGF1 alpha. In contrast, the G-protein inhibitor GDP beta S, (guanosine 5'-0-2-thiodiphosphate), attenuated alpha-thrombin-mediated prostaglandin generation. Treatment of HUVEC monolayers with pertussis toxin (1 microgram/ml) did not inhibit the PGI2 synthesis stimulated by either alpha-thrombin, NaF, or histamine but catalyzed the ADP ribosylation of a 40 kDa membrane protein which cross-reacted with antisera against a synthetic peptide corresponding to an amino acid sequence common to the alpha-subunit of other G-proteins. Preincubation of HUVEC microsomal membranes with alpha-thrombin diminished pertussis toxin-catalyzed ADP ribosylation in a time-dependent manner. These data suggest that thrombin stimulation of PGI2 synthesis by HUVEC monolayers requires the catalytically functional enzyme and further suggests that the thrombin-occupied receptor is coupled to phospholipase activities by a pertussis toxin-insensitive guanine nucleotide regulatory protein in human endothelial cell membranes.

Effect of environmental particulates on cultured human and bovine endothelium. Cellular injury via an oxidant-dependent pathway

The effects of respirable environmental fibers on cultures of human umbilical vein and bovine pulmonary artery endothelial cell monolayers were studied. Interaction among endothelial cell monolayers and amosite and chrysotile asbestos, attapulgite, fiberglass, or latex beads resulted in rapid phagocytosis of the particulates. A gradient of time-dependent and concentration-dependent endothelial cell injury (measured by specific 51Cr release) was observed with amosite and attapulgite being markedly toxic. Chrysotile and fiberglass were much less toxic, and latex beads were not significantly injurious at any time or dose examined. Responses of bovine pulmonary artery and human endothelial vein endothelial cells to fiber phagocytosis and fiber-induced injury were similar. In human umbilical cell monolayers, fiber-mediated stimulation of the arachidonate metabolite prostacyclin paralleled endothelial cell injury; i.e. amosite and attapulgite were stimulatory, whereas fiberglass (0-500 micrograms/ml) and latex beads (10(9) beads/ml) did not significantly increase prostacyclin generation. Although chrysotile was only weakly cytotoxic, significant stimulation of prostacyclin was observed at the highest dose tested (500 micrograms/ml). To investigate whether toxic oxygen species may be involved in fiber-induced cytotoxicity, oxidant scavengers or inhibitors were used in injury studies. Both superoxide dismutase (a scavenger of O2-) and catalase (an inhibitor of H2O2) produced significant protection against fiber-mediated endothelial cell injury. In addition, chelation by deferoxamine of elemental Fe present in the fiber preparations was also protective, suggesting Fe, via the modified Haber-Weiss reaction, may promote hydroxyl radical formation and contribute to endothelial cell injury induced by these particulates. These results suggest that the interaction within the interstitial environment between endothelial cells and occupationally relevant dusts may be important in fiber-mediated inflammatory processes in the lung.

Alveolar macrophages from patients with asbestos exposure release increased levels of leukotriene B4

The alveolar influx and subsequent activation of inflammatory cells such as neutrophils and eosinophils are believed to be important in the pathogenesis of many interstitial lung disorders, including asbestosis. Indices of lower respiratory tract abnormalities detected by bronchoalveolar lavage (BAL) were investigated in 93 asbestos-exposed workers as well as in smoking (n = 12) and nonsmoking (n = 10) control subjects. Patients with clinical asbestosis (n = 12) exhibited increases in both BAL neutrophils and BAL eosinophils, expressed as both percentage of total cells and total numbers, when compared to asbestos-exposed workers without asbestosis (n = 81) and control subjects. Significantly greater numbers of BAL neutrophils were also found in asbestos-exposed workers without asbestosis than in either smoking or nonsmoking control subjects. These abnormalities correlated significantly with in vitro BAL alveolar macrophage production of the potent leukocyte chemotaxin, leukotriene B4 (LTB4). For example, basal, unstimulated LTB4 production was 3.1 +/- 0.8 ng/10(6) alveolar macrophages for patients with asbestosis, 1.05 +/- 0.2 ng/10(6) cells in asbestos workers without asbestosis, 0.9 +/- 0.2 ng/10(6) cells in control nonsmokers, and 0.2 +/- 0.05 ng/10(6) cells in control smokers. Stimulated LTB4 release from BAL alveolar macrophages (A23187 or arachidonate) was even more pronounced in asbestos workers with or without asbestosis, suggesting an in vivo priming effect on alveolar macrophage synthesis of LTB4. Cell-free BAL supernatants from asbestos-exposed patients with or without asbestosis also contained significantly greater amounts of LTB4 than did those from control subjects, indicating enhanced in vivo production of this inflammatory mediator.(ABSTRACT TRUNCATED AT 250 WORDS)

Asbestos-induced endothelial cell activation and injury. Demonstration of fiber phagocytosis and oxidant-dependent toxicity

Vascular endothelial cell injury is important in the development of a variety of chronic interstitial lung disorders. However, the involvement of such injury in the inflammatory response associated with the inhalation of asbestos fibers is unclear and the mechanism of asbestos fiber cytotoxicity remains unknown. In the present study, human umbilical vein endothelial cells were challenged with amosite asbestos and several parameters of cellular function were examined. Electron microscopic examination revealed that endothelial cell exposure to asbestos resulted in active phagocytosis of these particulates. Biochemical evidence of dose-dependent asbestos-mediated endothelial cell activation was indicated by increased metabolism of arachidonic acid. For example, amosite asbestos (500 micrograms/ml) produced a ninefold increase in prostacyclin (PGI2) levels over those levels in non-exposed cells. Incubation of human endothelial cells with asbestos fibers induced specific 51Cr release in both a dose- and time-dependent fashion indicative of cellular injury. Injury induced by amosite asbestos was not significantly attenuated by treatment of the endothelial cell monolayer with either the iron chelator deferoxamine, which prevents hydroxyl radical (.OH) formation, or by the superoxide anion (O2-) scavenger, superoxide dismutase. However, significant dose-dependent protection was observed with the hydrogen peroxide (H2O2) scavenger, catalase. Chelation of elemental iron present within amosite asbestos fibers by deferoxamine produced a 33% reduction in asbestos cytotoxicity, suggesting a potential role for hydroxyl radical-mediated injury via the iron-catalyzed Haber-Weiss reaction.(ABSTRACT TRUNCATED AT 250 WORDS)

A glycolipid precursor of bacterial lipopolysaccharide (lipid X) lacks activity against endothelial cells in vitro and is not toxic in vivo

Lipid X (2,3-diacylglucosamine-1-phosphate) accumulates in mutants of Escherichia coli incapable of assembling the disaccharide backbone of lipid A, the principle endotoxic moiety of bacterial lipopolysaccharide (LPS). We compared the effects of lipid X on cultured bovine aortic endothelial cell (BEC) viability and prostacyclin (PGI2) release with those of lipid A and LPS. At 10(-5) M, both LPS and lipid A produced significant BEC cytotoxicity (percentage cytotoxicity 69 +/- 4 for LPS and 51 +/- 11 for lipid A) and induced a variable but consistent increase in the release of PGI2 (11- to 73-fold increase for LPS and 4- to 6-fold increase for lipid A). Lipid X, in contrast, was not toxic and did not induce PGI2 release at 10(-4) M. Pretreatment and coincubation of BEC with lipid X, at a concentration 100 times greater than LPS, failed to prevent LPS-mediated cytotoxicity. Intravenous infusion of lipid X in goats had no effect except for a modest elevation in the pulmonary artery pressure during the period of infusion. Moreover, pretreatment of goats with lipid X (70 micrograms/kg) did not block the effects of a subsequent infusion of LPS (5 micrograms/kg). These data suggest that a fatty acid-substituted disaccharide is the minimal molecular requirement for the numerous effects in vivo and activity in vitro induced by LPS. Furthermore, these effects are not prevented by pretreatment with a monosaccharide precursor of lipopolysaccharide, lipid X, at a dose 10- to 100-fold greater than that of LPS.

Effect of platelet activating factor on leukocyte-endothelial cell interactions

The proinflammatory effects of platelet activating factor (PAF) that impact upon tissue inflammation were studied in vitro using the adherence of human neutrophils to endothelium and the increase in macromolecule permeability of endothelial monolayers. PAF produced both a time- and dose-dependent increase in neutrophil-endothelial cell adhesion. The adhesion promoting properties observed were primarily due to an effect of PAF on endothelium and not on neutrophils and was independent of endothelial cell cyclooxygenase products. The PAF receptor antagonist kadsurenone inhibited the adhesion response suggesting endothelial surface PAF receptors are involved in these responses. Whereas PAF alone did not alter endothelial cell barrier properties, leukocyte activation (neutrophil and platelets) with PAF produced significant increases in 125I-albumin clearance across endothelial monolayers. These studies suggest that PAF has potent proinflammatory effects and that it can play a significant role in the endothelial response to injury.

Platelet-derived growth factor does not stimulate prostacyclin synthesis by cultured endothelial cells

We examined the effect of highly purified platelet-derived growth factor (PDGF) on prostacyclin (PGI2) release by cultured human umbilical vein and bovine aortic endothelial cells. PDGF tested at concentrations equal to or exceeding those observed in serum did not increase endothelial cell PGI2 synthesis as measured by radioimmunoassay of its metabolite, 6-keto-PGF1 alpha. In contrast, cells incubated with 20% human whole blood serum (WBS) demonstrated significantly increased PGI2 production (fivefold stimulation). Addition of anti-PDGF antibody to the 20% WBS did not attenuate the increased synthesis of PGI2. Incubation with 20% plasma-derived serum (PDS) that was deficient in PDGF produced stimulation of PGI2 release similar to 20% WBS. These results demonstrate that PDGF does not cause increased PGI2 synthesis in cultured human endothelial cells of human or bovine origin, and further suggest that the stimulation observed with serum is not due to a platelet-release product.

Synthesis of dihomoprostaglandins from adrenic acid (7,10,13,16-docosatetraenoic acid) by human endothelial cells

Human umbilical vein endothelial cells were found to contain adrenic acid (22:4) in their cellular lipids. Since this fatty acid may be metabolized by cyclooxygenase in the kidney, the metabolism of adrenic acid was studied in endothelial cell cultures. [14C]Adrenic acid was metabolized to several more polar metabolites. Two of these metabolites co-migrated on HPLC with 1 alpha,1 beta-dihomo-8-ketoprostaglandin F1 alpha (the metabolite of 1 alpha, 1 beta-dihomoprostaglandin I2) and 1 alpha,1 beta-dihomoprostaglandin E2. Indomethacin (10(-5) M) inhibited the synthesis of these metabolites. When cells were treated with adrenic acid (3 X 10(-5) M), a peak that co-migrated with dihomo-8-ketoprostaglandin F1 alpha was detected by radioimmunoassay using an antiserum directed against 6-ketoprostaglandin F1 alpha. The presence of dihomo-8-ketoprostaglandin F1 alpha was confirmed by gas chromatography-mass spectrometry. Immunoreactive peaks that co-migrated with dihomoprostaglandins E2 and F2 alpha were identified with antisera against prostaglandin E2 and prostaglandin F2 alpha, respectively. [14C]Arachidonic acid was metabolized to [14C]prostaglandin F2 alpha, 6-keto[14C]prostaglandin F1 alpha, and [14C]prostaglandin E2. Similar results were found with unlabelled arachidonic acid using specific antisera. When the two fatty acids were combined, adrenic acid reduced the metabolism of arachidonic acid. The culture media from endothelial cells inhibited thrombin-induced platelet aggregation, an effect blocked by aspirin. The inhibitory activity of the media was enhanced when arachidonic acid was added to the cells, but it was reduced by adrenic acid. Both prostaglandin I2 and dihomoprostaglandin I2 inhibited platelet aggregation, but prostaglandin I2 was 100-times more potent. We conclude that adrenic acid is metabolized in human endothelial cells to 1 alpha, 1 beta-dihomoprostaglandins and can compete with endogenous arachidonic acid for conversion by cyclooxygenase. These findings suggest that adrenic acid may reduce the formation of prostaglandin I2 by the blood vessel.

Enhancement of the antiaggregatory activity of prostacyclin by propranolol in human platelets

The beta-adrenergic antagonist propranolol has been found to inhibit platelet aggregation. We investigated the possibility that propranolol exerts this action by stimulating the synthesis or enhancing the antiaggregatory activity of prostaglandin (PG) I2. The media from cultures of human endothelial cells inhibited thrombin-induced platelet aggregation, an effect attributed to PGI2 production by the cells. When endothelial cells were incubated with dl- or d-propranolol, the media had two to three times the inhibitory activity of control media. However, this increased activity was not due to increased synthesis of PGI2 because control and propranolol-treated cultures synthesized similar amounts of the PGI2 metabolite, 6-keto-PGF1 alpha. Instead, propranolol enhanced the antiaggregatory activity of PGI2. Propranolol (1 microM) and PGI2 (0.05 nM), when tested separately, inhibited aggregation by 19% and 13%, respectively, whereas the combination inhibited aggregation by 51%. PGI2 inhibited platelet aggregation and thromboxane (Tx) B2 production but stimulated cyclic AMP formation. The adenyl cyclase inhibitor 2',5'-dideoxyadenosine (DDA) had no effect of its own on these parameters, but blocked the actions of PGI2. Propranolol inhibited aggregation and TxB2 synthesis without changing cyclic AMP levels. Unlike PGI2, propranolol's effects were not altered by DDA. While the combination of propranolol and PGI2 inhibited aggregation to a greater extent than either agent alone, this enhanced effect with the combination did not extend to TxB2 or cyclic AMP production. Propranolol, PGI2, and the combination inhibited TxB2 synthesis to a similar extent, and PGI2 produced a similar increase in cyclic AMP in the presence and absence of propranolol. These findings indicate that propranolol and PGI2 inhibit platelet aggregation through cyclic AMP-independent and dependent mechanisms, respectively. While propranolol does not alter the synthesis of PGI2, it enhances the inhibition of aggregation by PGI2, and this may contribute to its antiplatelet effect.

Role of hydrogen peroxide in the neutrophil-mediated release of prostacyclin from cultured endothelial cells

We have examined the effect of activated neutrophils on the release of prostacyclin (PGI2) from cultured endothelial cells by radioimmunoassay and thin layer chromatography of its stable metabolite, 6-keto-prostaglandin F1 alpha (6-keto-PGF1 alpha). Phorbol myristate acetate-activated neutrophils induced a time- and dose-dependent release of 6-keto-PGF1 alpha from human and bovine endothelial cell monolayers, whereas phorbol myristate acetate alone and neutrophils alone did not. Pretreatment of the endothelial cells with aspirin prevented neutrophil-mediated 6-keto-PGF1 alpha release, indicating that it did not depend upon neutrophil-generated endoperoxides. Phorbol myristate acetate-activated neutrophils from a patient with chronic granulomatous disease failed to induce endothelial 6-keto-PGF1 alpha release. Addition of catalase but not of superoxide dismutase significantly reduced human and bovine endothelial 6-keto-PGF1 alpha release by phorbol myristate acetate-activated neutrophils. Catalase-inhibitable endothelial 6-keto-PGF1 alpha release was also observed after the addition of the hydrogen peroxide-generating system, glucose-glucose oxidase, to bovine and human endothelial cell monolayers. Bovine endothelial 6-keto-PGF1 alpha release induced by exogenously generated hydrogen peroxide was attenuated by the phospholipase inhibitor mepacrine, suggesting that hydrogen peroxide may act by triggering endothelial membrane phospholipase activation. The release of 6-keto-PGF1 alpha by enzymatically or neutrophil-generated hydrogen peroxide was not associated with endothelial cell lysis as assessed by 51Cr release. We conclude that exogenously generated hydrogen peroxide or a hydrogen peroxide-derived product mediates rapid nonlytic release of PGI2 from cultured endothelial cells.

Glutathione redox cycle protects cultured endothelial cells against lysis by extracellularly generated hydrogen peroxide

We have examined the role of the glutathione redox cycle as an antioxidant defense mechanism in cultured bovine and human endothelial cells by disrupting the glutathione redox cycle at several points. Endothelial glutathione reductase was selectively inhibited with 1,3-bis(chloroethyl)-1-nitrosourea (BCNU). Cellular stores of reduced glutathione were depleted by reaction with diethylmaleate (DEM) or 1-chloro-2,4-dinitrobenzene (CDNB) or by inhibition of glutathione synthesis with buthionine sulfoximine (BSO). Whereas several strains of untreated bovine and human endothelial cells were resistant to lysis by enzymatically generated hydrogen peroxide, BCNU-treated cells were readily lysed in a time- and dose-dependent manner. Glucose-glucose oxidase-mediated lysis of BCNU-treated bovine endothelial cells was catalase-inhibitable and directly related to BCNU concentration and endogenous glutathione reductase activity. Pretreatment of bovine endothelial cells with BCNU did not potentiate lysis by distilled water, calcium ionophore, lipopolysaccharide, or hypochlorous acid. Depletion of cellular reduced glutathione by reaction with DEM or CDNB or by inhibition of glutathione synthesis by BSO also potentiated endothelial lysis by enzymatically generated hydrogen peroxide. Inhibition of endothelial glutathione reductase by BCNU or depletion of reduced glutathione by BSO increased endothelial susceptibility to lysis by hydrogen peroxide generated by phorbol myristate acetate-activated neutrophils. We conclude that the glutathione redox cycle plays an important role as an endogenous antioxidant defense mechanism in cultured endothelial cells.

Stimulation of prostaglandin formation by vasoactive mediators in cultured human endothelial cells

Human endothelial cells in culture synthesize prostaglandins and release these products into the culture medium. The major products of arachidonic acid metabolism were identified by high pressure liquid chromatography or thin layer chromatography, and release of prostaglandins was measured by radioimmunoassays. Addition of histamine or bradykinin enhanced release of prostaglandins in both arterial and venous endothelial cells. Other vasoactive compounds including angiotensin II, vasopressin, substance P, epinephrine, norepinephrine, or isoproterenol were ineffective. Release of prostaglandins by histamine was concentration-related, and involved H1 receptors, as determined by addition of histamine antagonists. Incubation of endothelial cells with 14C-arachidonic acid resulted in a time-dependent uptake into cell lipids, where most of the radioactivity was incorporated into phosphatidyl choline and neutral lipids. Endothelial cells released 14C-arachidonic acid as well as 14C-prostaglandins in response to either histamine or bradykinin. The enhanced release of 14C-prostaglandins was inhibited by either indomethacin or mepacrine, but 14C-arachidonic acid release was inhibited only by mepacrine. We conclude that the vasoactive compounds, histamine and bradykinin, stimulate formation of prostaglandins in endothelial cells by the release of arachidonic acid from phospholipids of the cell membrane.

Anti-platelet activity of beta-adrenergic antagonists: inhibition of thromboxane synthesis and platelet aggregation in patients receiving long-term propranolol treatment

Treatment of hypertensive patients with dl-propranolol (640 mg/day) significantly inhibited thromboxane synthesis by their platelets and platelet aggregation induced by thrombin or arachidonic acid. The effects were dose-related and were also caused by the stereoisomer, d-propranolol (640 mg/day), which has very little beta-blocking activity. These findings suggest that the cardioprotective effects of propranolol may be due partly to this anti-platelet activity, to a reduction in thromboxane-induced coronary-artery vasoconstriction, or to both. d-Propranolol treatment may be particularly useful, since this isomer provides similar benefits without causing pronounced beta-adrenergic blockade.