Inflammatory responses of endothelial cells experiencing reduction in flow after conditioning by shear stress.

Exposure of endothelial cells (EC) to shear stress reduces their response to tumour necrosis factor-alpha (TNF). We tested how shear-conditioned EC responded to reduction in flow, either by spontaneously binding leukocytes, or by increasing sensitivity to TNF. Human umbilical vein EC were exposed to shear stress of 2.0 Pa (20 dyn/cm(2)) for 24 h. Shear was then reduced to stasis (30 sec perfusion each hour to exchange medium) or 0.003 Pa (creeping flow). At chosen times, neutrophils were perfused over the EC at 0.1 Pa (effective reperfusion). EC developed an ability to capture flowing neutrophils that lasted from 1 to 3 h after flow reduction, which was reduced by antibody against P-selectin or pre-treatment of EC with an inhibitor of NADPH-oxidase. Adhesion of neutrophils to TNF-treated EC was greatly suppressed by shear-conditioning, remained suppressed immediately after cessation of flow and then took 48 h to approach the level in static cultures. Interestingly, the response to TNF remained suppressed in cultures switched to creeping flow. Gene array analysis confirmed that differently recovered cells had separate phenotypes. Thus, an acute response of EC to reduction in shear may contribute to leukocyte recruitment, along with hypoxia, in ischaemia and reperfusion. Prolonged cessation of flow may increase the sensitivity of EC to inflammatory stimuli, but this effect may be suppressed by residual flow.

Mechanisms of the anti-inflammatory effects of hydroxyethyl starch demonstrated in a flow-based model of neutrophil recruitment by endothelial cells.

OBJECTIVE: To determine whether plasma volume expander hydroxyethyl starch (HES) may protect against reperfusion injury through an ability to reduce neutrophil recruitment. DESIGN: An in vitro study using paired comparisons of adhesion of flowing neutrophils. SETTING: A collaboration between clinical and basic science departments in a university hospital. SUBJECTS: Neutrophils and cultured human umbilical vein endothelial cells (HUVEC). INTERVENTIONS: Treatment with HES (average molecular weight of 200 kd and substitution of 0.62) at clinically relevant concentrations or with gelatin solution (average molecular weight of 30 kDa) of comparable viscosity. MEASUREMENTS AND MAIN RESULTS: Glass capillaries were coated internally with either purified adhesion molecules or HUVEC, which were treated with tumor necrosis factor-alpha in the presence or absence of HES. Neutrophils were perfused over these surfaces (with or without HES) and their recruitment quantified by video microscopy. Expression of adhesion molecules and of the chemokine interleukin-8 by HUVEC were analyzed by enzyme-linked immunosorbent assay and quantitation of messenger RNA. HES over a wide range of concentrations had no effect on selectin- or integrin-mediated adhesion of neutrophils. However, when HUVEC were cultured with 1.5% wt/vol HES, neutrophil capture induced by low-dose (1 unit/mL) tumor necrosis factor-alpha and transendothelial migration induced by high-dose (100 units/mL) tumor necrosis factor-alpha were significantly inhibited (p < .05, in each case). The effects were linked with reductions in expression of E-selectin and interleukin-8 by HUVEC at these respective tumor necrosis factor-alpha concentrations (p < .05, in each case). Gelatin (2% wt/vol) had no significant effect in assays with HUVEC. CONCLUSIONS: Application of HES to HUVEC exerts an inhibitory effect on different stages of neutrophil recruitment, depending on the level of the inflammatory stimulus. These effects are associated with reduced adhesion molecule expression and chemokine production. In vivo, comparable effects might protect against complications associated with reperfusion injury.

Identification of a phenotypically and functionally distinct population of long-lived neutrophils in a model of reverse endothelial migration.

Recent studies have demonstrated that neutrophils are not a homogenous population of cells. Here, we have identified a subset of human neutrophils with a distinct profile of cell-surface receptors [CD54(high), CXC chemokine receptor 1(low) (CXCR1(low))], which represent cells that have migrated through an endothelial monolayer and then re-emerged by reverse transmigration (RT). RT neutrophils, when in contact with endothelium, were rescued from apoptosis, demonstrate functional priming, and were rheologically distinct from neutrophils that had not undergone transendothelial migration. In vivo, 1-2% of peripheral blood neutrophils in patients with systemic inflammation exhibit a RT phenotype. A smaller population existed in healthy donors ( approximately 0.25%). RT neutrophils were distinct from naïve circulatory neutrophils (CD54(low), CXCR1(high)) and naïve cells after activation with formyl-Met-Leu-Phe (CD54(low), CXCR1(low)). It is important that the RT phenotype (CD54(high), CXCR1(low)) is also distinct from tissue-resident neutrophils (CD54(low), CXCR1(low)). Our results demonstrate that neutrophils can migrate in a retrograde direction across endothelial cells and suggest that a population of tissue-experienced neutrophils with a distinct phenotype and function are present in the peripheral circulation in humans in vivo.

Effects of disturbed flow on endothelial cell function: Pathogenic implications of modified leukocyte recruitment.

Numerous studies have shown that intracellular signalling, transcription factor activation and gene expression in endothelial cells are modulated by the magnitude and patterns of shear stress to which they are exposed. Although these responses suggest that the haemodynamic environment will consequently modulate the ability of the endothelial cells to support leukocyte recruitment as part of an inflammatory response, direct evidence is quite sparse. It seems that disturbances of flow (such as local spatial or temporal variation or sudden cessation) are likely to be pathogenic co-factors, combined with mediators such as cytokines, oxidised lipids or hypoxia, in conditions such as atherosclerosis, post-surgical intimal hyperplasia and ischaemia/reperfusion injury. In fact there have been few experimental investigations of these scenarios that include measurement of leukocyte adhesion and migration. We recently demonstrated that the level of steady shear to which EC are exposed has a powerful effect on their ability to support cytokine-induced leukocyte adhesion and migration. However, more combined studies of flow and agonist-mediated responses, with functional readouts, appear necessary if we are to develop a better understanding of the mechanisms pre-disposing to vascular inflammatory responses and pathology.

Differing mechanisms of leukocyte recruitment and sensitivity to conditioning by shear stress for endothelial cells treated with tumour necrosis factor-alpha or interleukin-1beta.

The cytokines tumour necrosis factor-alpha (TNFalpha) and interleukin-1beta (IL-1B) induce endothelial cells to recruit leukocytes. However, the exact adhesion and activation mechanisms induced by each cytokine, and their relative sensitivities to modulation by endothelial exposure to shear stress remain unclear. We cultured human umbilical vein endothelial cells (HUVEC) in glass capillaries at various shear stresses, with TNFalpha or IL-1B added for the last 4 h. Subsequently, human neutrophils were perfused over the HUVEC, and adhesion and migration were recorded. Both cytokines induced dose-dependent capture of neutrophils. However, while conditioning of HUVEC by increasing shear stress for 24 h diminished their response to TNFalpha, the response of HUVEC to IL-1B was similar at all shear stresses. The differing sensitivities were evident at levels of adhesive function and mRNA for adhesion molecules and chemokines. Analysis of nuclear factor kappaB (NF-kappaB)/Rel family of transcription factors showed that their expression and activation were modified by exposure to shear stress, but did not obviously explain differential responses to TNFalpha and IL-1B. Antibodies against selectins were effective against capture of neutrophils on TNFalpha-treated but not IL-1B-treated HUVEC. Stable adhesion was supported by beta2-integrins in each case. Activation of neutrophils occurred dominantly through CXC-chemokine receptor 2 (CXCR2) for TNFalpha-treated HUVEC, while blockade of CXCR1, CXCR2 and of platelet-activating factor receptors caused additive inhibition of migration on IL-1B-treated HUVEC. The mechanisms which underlie neutrophil recruitment, and their modulation by the haemodynamic environment, differ between cytokines. Interventions aimed against leukocyte recruitment may not operate equally in different inflammatory milieu.