Karen L. Hynes

Endothelial permeability and IL-6 production during hypoxia: role of ROS in signal transduction

Prolonged hypoxia produces reversible changes in endothelial permeability, but the mechanisms involved are not fully known. Previous studies have implicated reactive oxygen species (ROS) and cytokines in the regulation of permeability. We tested whether prolonged hypoxia alters permeability to increasing ROS generation, which amplifies cytokine production. Human umbilical vein endothelial cell (HUVEC) monolayers were exposed to hypoxia while secretion of tumor necrosis factor-α (TNF-α), interleukin (IL)-1α, IL-6, and IL-8 was measured. IL-6 and IL-8 secretion increased fourfold over 24 h in a pattern corresponding to changes in HUVEC permeability measured by transendothelial electrical resistance (TEER). Addition of exogenous IL-6 to normoxic HUVEC monolayers caused time-dependent changes in TEER that mimicked the hypoxic response. An antibody to IL-6 significantly attenuated the hypoxia-induced changes in TEER (86 ± 4 vs. 63 ± 3% with hypoxia alone at 18 h), whereas treatment with anti-IL-8 had no effect. To determine the role of hypoxia-induced ROS on this response, HUVEC monolayers were incubated with the antioxidants ebselen (50 μM) and N-acetyl-l-cysteine (NAC, 1 mM) before hypoxia. Antioxidants attenuated hypoxia-induced IL-6 secretion (13 ± 2 pg/ml with ebselen and 19 ± 3 pg/ml with NAC vs. 140 ± 15 pg/ml with hypoxia). Ebselen and NAC prevented changes in TEER during hypoxia (94 ± 2% with ebselen and 90 ± 6% with NAC vs. 63 ± 3% with hypoxia at 18 h). N-nitro-l-arginine (500 μM) did not decrease hypoxia-induced changes in dichlorofluorescin fluorescence, IL-6 secretion, or TEER. Thus ROS generated during hypoxia act as signaling elements, regulating secretion of the proinflammatory cytokines that lead to alterations of endothelial permeability.Prolonged hypoxia produces reversible changes in endothelial permeability, but the mechanisms involved are not fully known. Previous studies have implicated reactive oxygen species (ROS) and cytokines in the regulation of permeability. We tested whether prolonged hypoxia alters permeability to increasing ROS generation, which amplifies cytokine production. Human umbilical vein endothelial cell (HUVEC) monolayers were exposed to hypoxia while secretion of tumor necrosis factor-alpha (TNF-alpha), interleukin (IL)-1alpha, IL-6, and IL-8 was measured. IL-6 and IL-8 secretion increased fourfold over 24 h in a pattern corresponding to changes in HUVEC permeability measured by transendothelial electrical resistance (TEER). Addition of exogenous IL-6 to normoxic HUVEC monolayers caused time-dependent changes in TEER that mimicked the hypoxic response. An antibody to IL-6 significantly attenuated the hypoxia-induced changes in TEER (86 +/- 4 vs. 63 +/- 3% with hypoxia alone at 18 h), whereas treatment with anti-IL-8 had no effect. To determine the role of hypoxia-induced ROS on this response, HUVEC monolayers were incubated with the antioxidants ebselen (50 microM) and N-acetyl-L-cysteine (NAC, 1 mM) before hypoxia. Antioxidants attenuated hypoxia-induced IL-6 secretion (13 +/- 2 pg/ml with ebselen and 19 +/- 3 pg/ml with NAC vs. 140 +/- 15 pg/ml with hypoxia). Ebselen and NAC prevented changes in TEER during hypoxia (94 +/- 2% with ebselen and 90 +/- 6% with NAC vs. 63 +/- 3% with hypoxia at 18 h). N-nitro-L-arginine (500 microM) did not decrease hypoxia-induced changes in dichlorofluorescin fluorescence, IL-6 secretion, or TEER. Thus ROS generated during hypoxia act as signaling elements, regulating secretion of the proinflammatory cytokines that lead to alterations of endothelial permeability.

Role of Mitochondrial Oxidant Generation in Endothelial Cell Responses to Hypoxia

Endothelial cells increase their secretion of the cytokine interleukin-6 (IL-6) during hypoxia, which then acts in an autocrine fashion to increase the permeability of cell monolayers. These responses are attenuated by antioxidants, suggesting that reactive oxygen species (ROS) participate in signaling in hypoxic endothelium. We tested whether mitochondria are responsible for these ROS in human umbilical vein endothelial cells exposed to hypoxia. Oxidation of the probe 2′, 7′-dichlorodihydrofluorescein to fluorescent dichlorofluorescein or the probe dihydroethidium was used to assess oxidant signaling, whereas permeability was assessed by using transendothelial electrical resistance. Hypoxia elicited increases in dichlorofluorescein and dihydroethidium fluorescence that were abrogated by the mitochondrial electron transport (ET) inhibitors rotenone (2 &mgr;mol/L) and diphenyleneiodonium (5 &mgr;mol/L). The same ET inhibitors also attenuated hypoxia-induced increases in nuclear factor-&kgr;B (NF-&kgr;B) activation, although they did not abrogate NF-&kgr;B activation in response to endotoxin (lipopolysaccharide). ET inhibition also abolished the hypoxia-induced increases in IL-6 mRNA expression, hypoxia-stimulated IL-6 secretion into the media, and the hypoxia-induced increases in transendothelial electrical resistance of human umbilical vein endothelial cell monolayers. By contrast, the above responses to hypoxia were not significantly affected by treatment with the NAD(P)H oxidase inhibitor apocynin (30 &mgr;mol/L), the xanthine oxidase inhibitor allopurinol (100 &mgr;mol/L), or the NO synthase inhibitor N-nitro-l-arginine (100 &mgr;mol/L). We conclude that ROS signals originating from the mitochondrial ET chain trigger the increase in NF-&kgr;B activation, the transcriptional activation of IL-6, the secretion of IL-6 into the cell culture media, and the increases in endothelial permeability observed during hypoxia.

Cytokine-induced increases in endothelial permeability occur after adhesion molecule expression

BACKGROUND Transmigration of neutrophils (PMNs) through endothelial cell tight junctions is a critical stage in the tissue injury of ischemia-reperfusion (I/R). Although cytokines are released in I/R, it is unclear whether cytokines directly increase permeability or this phenomenon requires both expression of cell adhesion molecules and PMN adhesion-activation. METHODS We exposed confluent monolayers of human umbilical vein endothelial cells to physiologic concentrations of interleukin-1 (10 pg/ml) and tumor necrosis factor-alpha (10 pg/ml) in the absence of PMNs. Tight junction permeability was quantified with both transendothelial electrical resistance and albumin flux, whereas expression of endothelial-leukocyte adhesion molecule-1 was measured by flow cytometry (t test p < 0.05). RESULTS Stimulation with tumor necrosis factor-alpha or interleukin-1 produced maximal transendothelial electrical resistance decreases at 12 hours with return to baseline at 24 hours. Increases in albumin flux began at 6 hours, with maximum effects at 24 hours. These changes occurred soon after maximal expression of endothelial-leukocyte adhesion molecule-1 at 4 hours. CONCLUSIONS Cytokines induced increases in both cell adhesion molecule expression and endothelial permeability. This sequence of events is consistent with direct cytokine effects on cytoarchitecture, because it occurred without the adhesion-activation of PMNs.

Prolonged hypoxia alters endothelial barrier function

BACKGROUND It is well recognized that hypoxia/reoxygenation and exposure to inflammatory mediators such as cytokines and neutrophils alter the barrier function of the vascular endothelium. The experiments we conducted tested whether hypoxia alone could produce changes in permeability and whether a prolonged period of hypoxia alters the surface expression of cell adhesion molecules. METHODS Endothelial cells were cultured from human umbilical vein endothelial cells (HUVECs). Hypoxia was created by isolating the cells in a chamber through which 1% 02, 5% CO2, and 94% N2 were insufflated (30 min at 1/min). Oxygen tension was measured through oxygen-quenching phosphorescence. Hypoxia was maintained for 24 hours. Changes in endothelial permeability were measured by transendothelial electrical resistance (TEER). Endothelial leukocyte adhesion molecule 1 (ELAM-1) and intercellular adhesion molecule 1 (ICAM-1) expression were assessed by flow cytometry (mean +/ standard error of the mean [SEM]. RESULTS Exposure of endothelial cells to hypoxia resulted in increased permeability between 6 and 24 hours, with the greatest decrease in TEER at 18 hours (63% +/ 3%; P < .05). Prolonged hypoxia produced no change in the surface expression of ELAM-1 or ICAM-1. CONCLUSIONS Hypoxia alone produced a significant reversible alteration in endothelial permeability. However, this change was observed only under severe hypoxic conditions (eg, below 20 mm Hg); higher oxygen tensions (25 and 35 mm Hg) had no significant effect. Unlike observations made after cytokine exposure, hypoxic breakdown of endothelial barrier function was unassociated with up-regulation of either ELAM-1 or ICAM-1.

Loss of endothelial barrier function requires neutrophil adhesion

BACKGROUND The inflammatory response is characterized by cytokine-induced up-regulation of endothelial adhesion molecules followed by polymorphonuclear neutrophil (PMN) adhesion and breakdown of tight junctions between cells. The purpose of this investigation was to determine whether PMN adhesion is an essential element in the alteration of endothelial permeability or whether cytokines alone can produce this change. METHODS Human umbilical vein endothelial cells (HUVECs) were exposed to formylated met-leu-phe-activated PMNs. In a second series of experiments, PMNs were contained in a microporous membrane that allowed passage of secreted cytokines but not cells. Permeability was quantified by using transendothelial electrical resistance (TEER, ohm.cm2,) whereas expressions of two cell adhesion molecules (endothelial leukocyte adhesion molecule-1 [ELAM-1] and intercellular adhesion molecule-1 [ICAM-1]) were measured by flow cytometry (% shift). Cytokine production was monitored with enzyme-linked immunosorbant assays (picograms per milliliter). RESULTS Stimulated PMNs secreted comparable amounts of cytokines whether allowed access to HUVECs or trapped in a microporous membrane (interleukin-1 alpha, 5.88 +/- 2.38 versus 3.65 +/- 1.84 pg/ml; tumor necrosis factor-alpha, 10.27 +/- 3.21 versus 6.61 +/- 1.82 pg/ml). Up-regulation of ELAM-1 and ICAM-1 was observed whether PMNs were free or restricted (52.97% +/- 2.14% versus 75.32% +/- 4.19% and 71.66% +/- 7.37% versus 73.66% +/- 4.32%, respectively). TEER was unchanged in controls and when PMNs were membrane restricted. In contrast, TEER decreased precipitously (51% +/- 5.9% of control, p < 0.05) if PMNs were allowed access to HUVECs. CONCLUSIONS Cytokine secretion by PMNs is independent of endothelial contact and is sufficient to upregulate adhesion molecules. However, PMN adhesion is essential for the loss of endothelial barrier function, which leads to diapedesis of activated PMNs and eventual tissue injury.

Decrease in mucosal alkaline phosphatase: A potential marker of intestinal reperfusion injury ☆ ☆☆ ★

Intestinal ischemia necessitates rapid re-establishment of blood flow to prevent irreversible anoxic tissue damage. However, reperfusion results in additional injury as a consequence of the generation of oxygen free radicals. To date, no clear-cut marker to differentiate between ischemia versus reperfusion injury is available. In this regard, previous studies from our laboratory utilizing a rat in vitro lipid peroxidation model demonstrated that the generation of free radicals resulted in the inactivation of only the intestinal brush border alkaline phosphatase enzyme, with no effect on other membrane-bound digestive enzymes. Current studies were designed to assess the possibility of alkaline phosphatase being a specific marker of the reperfusion injury in canine and human ex vivo ischemia/reperfusion models. Small bowels harvested from canines and organ donors were subjected to ischemia followed by reperfusion. Brush border membrane enzymes, alkaline phosphatase, sucrase, maltase, and gamma-glutamyl transpeptidase were assayed in mucosal extracts from intestines with ischemia versus reperfusion. In both experimental models, there was no change in any enzyme activity with warm ischemia alone. In contrast, alkaline phosphatase activity was significantly decreased in both the canine and human reperfusion models, with no change in specific activities of sucrase, maltase, and gamma-glutamyl transpeptidase. Our data indicate that the alkaline phosphatase enzyme activity may represent a potential marker of intestinal reperfusion injury and may permit quantitative assessments of therapeutic interventions in human intestinal reperfusion injury.

Serum proteins facilitate neutrophil induction of endothelial leukocyte adhesion molecule 1.

BACKGROUND Although the individual actions of neutrophils and serum proteins such as complement in acute inflammation are well characterized, less is known about their effects in combination. We investigated the combined effects of neutrophil contact and active serum proteins on the expression of endothelial leukocyte adhesion molecule 1 (ELAM-1). METHODS Confluent monolayers of human umbilical vein endothelial cells were incubated with neutrophils in the presence and absence of fresh human serum. Flow cytometry was used to assess expression of endothelial intercellular adhesion molecule 1 (ICAM-1) and ELAM-1. In addition, neutrophils were retained in a semipermeable insert, which allowed their secretions to contact the endothelium but restricted neutrophil-endothelial contact. RESULTS ELAM-1 expression was significantly increased on the cells coincubated with neutrophils and fresh human serum (25.8%; p < 0.001). There was no significant change in ELAM-1 expression on endothelial cells incubated with fresh human serum alone (3.9%; p > 0.01) or in those incubated with neutrophils and heat-inactivated serum (9.3%; p > 0.01). In the absence of neutrophil contact, ELAM-1 expression was increased only in the presence of fresh human serum (9.6%; p < 0.05). CONCLUSIONS These findings suggest that serum proteins may potentiate the volume or potency of neutrophil-derived diffusable mediators of ELAM-1 expression. These effects are eliminated with the heat inactivation of serum proteins, implicating a heat sensitive mediator such as the complement cascade.