Mice lacking E-selectin show normal numbers of rolling leukocytes but reduced leukocyte stable arrest on cytokine-activated microvascular endothelium.

OBJECTIVE: Previous work indicated that E-selectin mediates transient interactions between leukocytes and cytokine-activated endothelium in vitro. Here we examine the role of E-selectin in blood leukocyte interactions with microvascular endothelium in vivo. METHODS: E-selectin-deficient (E-/-) mice were produced by gene targeting. The effect of this null mutation on leukocyte-endothelial interactions was determined by intravital microscopy before and 4 to 5 hours after local administration of the proinflammatory cytokine tumor necrosis factor alpha (TNF alpha) in dermal microvessels with low blood flow (dorsal skin-fold chambers, intact ear skin), and after endotoxin activation in exteriorized mesenteric microvessels with higher blood flow. RESULTS: E-/- mice were viable, fertile with normal circulating leukocyte and platelet profiles. Approximately 60% of circulating leukocytes rolled in dermal microvessels of both normal (E+/+) and E-/- mice without inflammatory stimulation. After local administration of TNF alpha, rolling increased modestly and equivalently in both genotypes. The main effect of TNF alpha was a dramatic increase in leukocyte stable adhesion and, unlike rolling, this manifestation of endothelial activation was significantly reduced in E-/- animals. This reflected fewer dermal microvessels supporting higher adhesion densities in E-/- mice, and a similar trend was observed in mesenteric microvessels. CONCLUSIONS: E-selectin plays a previously unappreciated role in facilitating and/or mediating stable adhesion of leukocytes to inflamed microvascular endothelium.

Microvascular endothelial cells from E-selectin-deficient mice form tubes in vitro.

Studies of capillary morphogenesis and angiogenesis in vitro have suggested a role for E-selectin (CD62E) in the process of differentiation into tube-like structures. Recent studies by our group and others have demonstrated that mice lacking E-selectin because of germline inactivation of the E-selectin gene by gene targeting are viable and fertile, without apparent deficiencies in vascular development. Murine lung endothelial cells from wild-type and E-selectin-deficient animals were isolated using an activation-dependent sterile sorting method, and differentiation into tube-like structures on sparse fibronectin, Matrigel, and collagen gels was compared. Both types of murine lung endothelial cells spontaneously organized to form multicellular tubes and extensive anastomotic networks. There were no major differences in either the time course of development or the general appearance of the multicellular cords or tube-like structures formed by murine lung endothelial cells from wild-type or E-selectin-deficient mice, although different patterns were observed on different extracellular matrices. These studies, thus, demonstrate that E-selectin is not required for morphogenesis of 3-dimensional vascular structures in vitro.

Evidence that angiotensin II is present in human monocytes.

BACKGROUND: The purpose of the present study was to determine whether human monocytes and polymorphonuclear leukocytes contain angiotensins I and II. METHODS AND RESULTS: Human mononuclear and polymorphonuclear leukocytes were isolated from blood. To identify angiotensins in human leukocytes, we performed immunocytochemistry using both alkaline phosphatase and fluorescence methods. With light microscopy immunocytochemistry with alkaline phosphatase, prominent staining of angiotensin II was observed in mononuclear leukocytes. Angiotensin I was also demonstrated in mononuclear leukocytes, but the signal was less pronounced than for angiotensin II. Polymorphonuclear leukocytes showed very little staining for angiotensin II. Fluorescence immunocytochemistry also demonstrated angiotensin II in mononuclear leukocytes. Angiotensins I and II in homogenate of leukocytes were quantified by radioimmunoassay. The concentration of angiotensins I and II in mononuclear leukocytes was 355 +/- 216 (mean +/- SEM) and 2331 +/- 106 fmol/mg protein, respectively, and the concentration in polymorphonuclear leukocytes was 36 +/- 10 and 336 +/- 120 fmol/mg protein. CONCLUSIONS: These findings suggest that human mononuclear leukocytes contain large amounts of angiotensin II and lesser amounts of angiotensin I. Human polymorphonuclear leukocytes contain small amounts of angiotensin I and II.

Interaction of human platelets and leukocytes in modulation of vascular tone.

We tested the hypothesis that the vasodilator response to human platelets is modulated by polymorphonuclear leukocytes (PMNs). Responses to platelets activated with thrombin, as well as PMNs activated with N-formylmethionyl-leucyl-phenylalanine (FMLP), were examined in perfused rabbit carotid arteries in vitro. Activation of platelets produced marked dilatation, and activation of PMNs produced modest constriction in arteries preconstricted with phenylephrine. Vasodilator responses to platelets were greatly impaired during infusion of activated PMNs. Pretreatment of PMNs with superoxide dismutase (SOD) partially restored dilator responses to platelets. Because SOD only partially restored vasodilator responses to platelets, we tested the possibility that adenosine-diphosphatase (ADPase) activity of PMNs may degrade ADP released by platelets and thus reduce vasodilator responses. After incubation with PMNs, dilator responses to ADP, but not acetylcholine, were significantly impaired. These findings indicate that vasodilatation produced by activated human platelets is profoundly impaired by activated leukocytes. We conclude that two mechanisms may account for this effect: 1) endothelium-derived relaxing factor, released in response to platelet-derived ADP, is inactivated by superoxide anion generated by activated PMNs and 2) ADP is degraded by ADPase activity of PMNs. We speculate that platelet-leukocyte interaction may have important effects on vasomotor tone.

Role of platelets and leukocytes in modulation of vascular tone.

Activated platelets and leukocytes release potent vasoactive factors that may modulate vascular tone. Activation of normal platelets produces dilatation of normal arteries. Vasodilatation is mediated by adenosine diphosphate (ADP), which releases endothelium-derived relaxing factor (EDRF) from endothelium. In atherosclerotic arteries, activation of platelets produces constriction and perhaps spasm. The constrictor response of atherosclerotic arteries is related in part to profound changes in vascular function: endothelial dysfunction with impaired endothelium-dependent vasodilator responses to ADP, and augmented vasoconstrictor responses to serotonin. In addition, hypercholesterolemia has a direct effect on platelets, resulting in impaired vasodilator responses. Thus, abnormal platelets and altered vascular function may both predispose to spasm of atherosclerotic arteries. Activation of leukocytes has little effect on resistance of large arteries in normal monkeys. In contrast, activation of leukocytes produces pronounced vasoconstriction in atherosclerotic monkeys. Possible mediators of vasoconstriction include prostaglandin E2, oxygen radicals, thromboxane, and a peptide (perhaps angiotensin II). In addition, leukocytes also alter vascular responses to platelets, in part by ADPase on the leukocyte membrane, which degrades ADP released by activated platelets. Leukocytes also release oxygen radicals, which may inactivate EDRF, thereby impairing ADP-mediated, endothelium-dependent vasodilator responses. Abnormal responses to platelets and leukocytes are largely reversed when atherosclerotic lesions regress in monkeys.

Functional improvement precedes structural regression of atherosclerosis.

BACKGROUND: Vasoconstrictor responses to serotonin are augmented in monkeys with diet-induced atherosclerosis and improve after 18 months of normal diet. We tested the hypothesis that functional improvement may occur early during regression, before evidence of structural improvement. METHODS AND RESULTS: Responses of the iliac artery to serotonin were measured by quantitative angiography and a Doppler flow probe in several groups of monkeys: (1) normal monkeys, (2) monkeys fed an atherogenic diet for 2 years (atherosclerotic), and (3) monkeys fed an atherogenic diet for 2 years (preregression) followed by a normal diet for 4, 8, or 12 months (regression). In normal monkeys, serotonin produced minimal constriction of the iliac artery, and blood flow to the legs increased. In atherosclerotic monkeys, there was pronounced constriction of the iliac artery, and blood flow to the legs decreased markedly. After 4 months of regression diet, four of eight monkeys demonstrated marked reduction in hyperresponsiveness to serotonin angiographically, and by 8 months, six of eight monkeys had significant improvement. After regression, serotonin produced minimal changes in flow. There was no reduction in intimal area (ie, atherosclerotic lesion) in iliac arteries from regression monkeys compared with atherosclerotic monkeys, but there was a marked reduction in cholesteryl ester in arteries from regression monkeys. CONCLUSIONS: Abnormal vasoconstrictor responses to serotonin usually return to or toward normal within a few months during regression of atherosclerosis. Functional improvement occurs in conjunction with early resorption of lipid from the arterial wall and occurs before detectable changes in mass of the atherosclerotic lesion.

Altered vascular responses to platelets from hypercholesterolemic humans.

Activated platelets release potent vasoactive factors. Previous studies have focused on mechanisms by which vascular abnormalities lead to altered responses of atherosclerotic arteries. We tested the hypothesis that the activation of platelets from hypercholesterolemic humans produces abnormal vascular responses. Responses to intraluminal and abluminal activation of platelets from normal subjects and type II hypercholesterolemic patients (total cholesterol, 274 +/- 16 [mean +/- SEM] mg/dl) were examined in carotid arteries from normal rabbits perfused in vitro. Intraluminal activation of normal platelets produced pronounced dilatation of arteries preconstricted with phenylephrine. Vasodilator responses to intraluminal activation of platelets from hypercholesterolemic patients were greatly impaired. Vasodilator responses to platelets from hypercholesterolemic patients were not restored to normal by LY53,857 (10(-5) M), a 5-hydroxytryptamine2-serotonergic antagonist, by SQ29,548 (10(-5) M), a thromboxane A2/prostaglandin H2 receptor antagonist, or by apyrase (1.5 units/ml), an enzyme with ADPase activity. Abluminal activation of normal platelets produced modest constriction in quiescent arteries, and abluminal activation of platelets from hypercholesterolemic patients produced augmented vasoconstrictor responses. The major finding is that vasodilator responses to platelets from hypercholesterolemic patients are profoundly impaired, and vasoconstrictor responses to platelets from hypercholesterolemic patients are augmented. Mechanisms in addition to increased release of serotonin, thromboxane, and ADP appear to contribute to impaired vasodilator responses to hypercholesterolemic platelets. Thus, alteration of platelets by hypercholesterolemia, as well as altered vascular reactivity, may contribute to abnormal vascular responses in atherosclerosis.

Effect of atherosclerosis on responses of the perfused rabbit carotid artery to human platelets.

Vascular responses to intraluminal and abluminal activation of human platelets were examined in carotid arteries from normal and atherosclerotic rabbits. The carotid artery was perfused in vitro, platelets were activated with thrombin (0.1 unit/ml), and changes in diameter were measured. In vessels from normal animals, intraluminal activation of platelets produced dilatation of preconstricted arteries. The dilator response was attenuated by N omega-nitro-L-arginine (10(-5) M), an inhibitor of synthesis of endothelium-derived relaxing factor-nitric oxide (EDRF-NO), and augmented by LY53,857 (10(-5) M), a 5-HT2-serotonergic antagonist. Abluminal activation of platelets produced modest constriction in quiescent arteries, which was inhibited by LY53,857. Intraluminal but not abluminal ADP produced pronounced dilatation of preconstricted arteries. In vessels from atherosclerotic animals, endothelium-dependent dilatation to intraluminal activation of platelets and to ADP was impaired and dilator responses to sodium nitroprusside were normal. These experiments indicate that 1) intraluminal activation of human platelets produces endothelium-dependent dilatation in perfused carotid arteries, whereas abluminal activation of human platelets produces vasoconstriction, which is mediated primarily by serotonin, and 2) atherosclerosis markedly impairs vasodilator responses to activation of human platelets, probably because vasodilatation to ADP released from platelets is impaired.

Activation of leukocytes with complement C5a is associated with prostanoid-dependent constriction of large arteries in atherosclerotic monkeys in vivo.

Activated leukocytes release a variety of substances which have been shown in vitro to modulate vascular tone. The chemotactic peptide complement C5a is a physiological activator of leukocytes. We injected human recombinant complement C5a (10 and 100 micrograms) into the blood-perfused hind limb of normal and atherosclerotic cynomolgus monkeys and examined vascular responses. In both normal and atherosclerotic monkeys, the high dose of C5a produced about 65% decrease in leukocyte cell count in venous blood drainage from the hind limb. Injection of C5a produced a pronounced increase in resistance of large arteries (segment from iliac artery to dorsal pedal artery) in atherosclerotic, but not in normal monkeys. The constrictor effect of C5a in atherosclerotic monkeys was abolished by the thromboxane A2 receptor antagonist SQ 29,548 (2 mg/kg i.v.). The platelet-activating factor antagonist WEB 2086 (5 mg/kg, i.v.) did not alter vascular responses to C5a. We conclude that activation of leukocytes produces constriction of large arteries in atherosclerotic, but not normal, monkeys in vivo. This response may be mediated in part by release of thromboxane A2.

Vascular responses to activated leukocytes after regression of atherosclerosis.

Activation of leukocytes in vivo produces marked constriction of large arteries in atherosclerotic, but not in normal, monkeys. We tested the hypotheses that vasoconstrictor responses to activated leukocytes in vivo may be abnormal during hypercholesterolemia before the development of atherosclerotic lesions and that responses may return to normal after the regression of atherosclerosis. Leukocytes were activated by injection of the chemotactic peptide formyl-methionine-leucine-phenylalanine (fMLP) into the blood-perfused hind limb of four groups of cynomolgus monkeys: monkeys fed a normal diet (normal group, n = 18), monkeys fed an atherogenic diet for 3-4 months (hypercholesterolemic group, n = 6), monkeys fed an atherogenic diet for 20 months (atherosclerotic group, n = 19), and monkeys fed an atherogenic diet for 18 months, followed by a normal diet for 20 months (regression group, n = 14). Baseline resistance of large arteries was 1.5 +/- 0.2 (mean +/- SEM), 2.0 +/- 0.6, 3.5 +/- 0.4 (p less than 0.05 versus normal), and 1.7 +/- 0.2 mm Hg/ml/min per 100 g tissue for the normal, hypercholesterolemic, atherosclerotic, and regression groups, respectively. Injection of fMLP did not change resistance of large arteries in normal or hypercholesterolemic monkeys. Injection of fMLP increased resistance of large arteries by 3.0 +/- 0.7 mm Hg/ml/min per 100 g tissue in atherosclerotic monkeys and by 1.3 +/- 0.4 mm Hg/ml/min per 100 g tissue in regression monkeys (p less than 0.05 versus atherosclerotic and normal). Thus, abnormal vasoconstriction in response to activation of leukocytes persists, but to a lesser extent, after regression. In contrast, vasoconstrictor responses to serotonin, which were potentiated in atherosclerotic monkeys, were normal after regression.(ABSTRACT TRUNCATED AT 250 WORDS)

Mechanisms of contraction induced by human leukocytes in normal and atherosclerotic arteries.

Activation of leukocytes results in the release of a variety of vasoactive substances that may modulate vascular tone. We studied the effect of human polymorphonuclear (PMN) and mononuclear (MONO) leukocytes on quiescent femoral arteries in vitro. Arteries were obtained from normal and atherosclerotic cynomolgus monkeys. In normal arteries, stimulation of PMNs (3 and 5 x 10(6) cells/ml) with either thrombin (5 units/ml) or complement C5a (0.5 micrograms/ml) resulted in endothelium-independent contraction (approximately 25% of maximum contraction with 80 mM KCl). Vasocontraction was augmented in the presence of superoxide dismutase (150 units/ml) and was significantly impaired in the presence of the hydroxyl radical scavengers mannitol (20 mM) and deferoxamine (1 mM). Catalase (1,200 units/ml) or L-alanine (20 mM) did not modify this effect of PMNs. In contrast to PMNs, vasocontraction in response to MONOs was not altered by the addition of radical scavengers. Pretreatment of PMNs and MONOs with indomethacin (10 microM) or nordihydroguaiaretic acid (20 microM) did not influence vascular responses. Supernatant of thrombin-stimulated PMNs and MONOs also produced vasocontraction (approximately two thirds of the effect of intact cells). This vasocontractor factor (or factors) was heat stable (30 minutes, 95 degrees C) and had a molecular weight less than 1,000 as determined by ultrafiltration. Stimulation of MONOs or PMNs (3 and 5 x 10(6) cells/ml) produced a similar response in normal arteries. In contrast, the constrictor response in atherosclerotic arteries to MONOs (5 x 10(6) cells/ml) was significantly greater than to PMNs. We conclude that stimulated human PMNs and MONOs contract arteries in vitro by release of at least two factors. One factor appears to be heat stable, with a molecular weight less than 1,000. The vascular response to PMNs, but not to MONOs, appears to involve the generation of hydroxyl radicals. The response to MONOs is greater than the response to PMNs in atherosclerotic, but not in normal, arteries.

Spontaneous production of interleukin-2 in the supernatants of cultured cutaneous sarcoidal granulomas.

Biopsies of cutaneous sarcoidal lesions were cultured for 24 hr in vitro, and the cell-free supernatants were examined for the presence of T cell growth factor (IL-2). Low levels of IL-2 were detected in these supernatants based on their ability to support the growth of cultured T cells. The apparent molecular mass of the active component in these supernatants, as assessed by exclusion chromatography, was 14,900 daltons. These findings suggest that IL-2 production within these sarcoidal lesions may be partly responsible for their evolution and maintenance.