Importance of L-selectin-dependent leukocyte-leukocyte interactions in human whole blood.

The objective of this study was to investigate whether leukocytes could be recruited by rolling leukocytes in a human whole blood model system. In all experiments, either neutrophils, whole blood, or diluted blood was perfused over immobilized E-selectin. With isolated neutrophils (2 x 10(5)/mL), the free-flowing neutrophils were captured by attached neutrophils to form secondary interactions that resulted in lines of rolling leukocytes. These secondary tethers accounted for 50% to 60% of all interactions and were eliminated by an L-selectin antibody, which also eliminated the lines of rolling leukocytes. Perfusion of whole blood or diluted blood revealed no lines of rolling leukocytes. The addition of red blood cells to isolated neutrophils either in a 1000:1 or a 10:1 ratio also inhibited lines of rolling leukocytes. Leukocytes were fluorescently labeled with rhodamine-6G so that leukocyte-leukocyte interactions could be studied in whole blood. A small number of secondary tethers (less than 20%) occurred and could be reduced by more than 80% with an L-selectin antibody. However, the overall impact on leukocyte recruitment was negligible. Similar experiments were performed using murine whole blood or isolated murine leukocytes. In the absence of red blood cells, murine leukocytes also formed lines of rolling leukocytes on E-selectin, and secondary tethers accounted for 50% of total interactions. However, when murine blood (diluted 1:5 with buffer) was perfused over E-selectin, secondary tethers accounted for only 13% of total interactions. These interactions were completely absent when blood was used from L-selectin-deficient mice. These data demonstrate for the first time that the importance of L-selectin-dependent leukocyte-leukocyte interactions is greatly reduced in whole blood and does not enhance overall recruitment of leukocytes in this physiologic milieu. (Blood. 2000;95:2954-2959)

Endogenous interleukin-10 regulates hemodynamic parameters, leukocyte-endothelial cell interactions, and microvascular permeability during endotoxemia.

The objective of this study was to determine whether endogenous IL-10 is capable of regulating hemodynamic parameters, leukocyte recruitment, and microvascular permeability in response to endotoxin. Intravital microscopy was used to examine hemodynamic parameters, leukocyte rolling and adhesion, and microvascular permeability in cremasteric postcapillary venules in wild-type mice and in IL-10-deficient (IL-10(-/-)) mice exposed to lipopolysaccharide (LPS). Doses of LPS (3 or 30 microg/kg, IV), which did not reduce blood pressure and minimally altered microvascular hemodynamic factors in wild-type mice, caused significant reductions in these parameters in IL-10(-/-) mice, demonstrating at least a 10-fold increased sensitivity in IL-10(-/-) mice to LPS-induced hemodynamic alterations. Furthermore, in response to LPS (30 microg/kg, IV), leukocyte rolling, adhesion, and fluorescein isothiocyanate-albumin extravasation were increased in the IL-10(-/-) mice. Antibody blockade experiments showed that in both types of mice, leukocyte rolling was mediated by E-selectin and P-selectin. Leukocyte accumulation into other tissues, such as lung, also was enhanced greatly in IL-10(-/-) mice. This was specific to endotoxin, because acute chemotactic stimuli including N-formyl-methionyl-leucyl-phenylalanine elicited similar responses in IL-10(-/-) and wild-type mice. These results suggest that endogenous IL-10 may be a homeostatic regulator of hemodynamic parameters, leukocyte-endothelial cell interactions, and microvascular dysfunction in response to endotoxin and provide potential mechanisms to explain the protective effect of IL-10 against LPS-induced mortality.

Differential leukocyte recruitment from whole blood via endothelial adhesion molecules under shear conditions.

The objective of this study was to determine if vascular cell adhesion molecule (VCAM-1), E-selectin, and P-selectin could selectively recruit leukocyte subpopulations, and whether this was affected by shear force or adhesion molecule concentration. Cover slips coated with purified adhesion molecules were incorporated into laminar flow chambers. Whole human blood was perfused for 5 minutes over these cover slips at relative shear forces of 2 to 40 dynes/cm2. Chasing the whole blood with buffer permitted visualization of leukocyte-substratum interactions. Leukocytes were observed to roll on and adhere to VCAM-1 at shears between 2 and 15 dynes/cm2. As assessed by cover slip staining, the majority of these cells were lymphocytes, but eosinophils, monocytes, and, surprisingly, neutrophils were also recruited, events inhibitable by anti-4-integrin antibody (HP1/2). Neutrophils were effectively recruited onto the selectins, with interactions occurring at shears as high as 30 and 40 dynes/cm2 for E- and P-selectin respectively. Eosinophils had high affinity for P- but not E-selectin. Mononuclear cells did not have high affinity for either selectin, but interacted avidly with VCAM-1. Antibodies against P-selectin (G1) and E-selectin (ES-1) completely blocked interactions on these substrates. Reducing the concentration of adhesion molecules did not appreciably change recruitment patterns except for VCAM-1, where neutrophils were no longer recruited. The novel use of whole blood in flow chambers shows a partial selectivity of selectins and VCAM-1 for certain subpopulations of leukocytes under varying physiologic shear conditions.

Inducible nitric oxide synthase-deficient mice have enhanced leukocyte-endothelium interactions in endotoxemia.

Nitric oxide (NO) from constitutive NO synthase (NOS) has been postulated to be a homeostatic regulator of leukocyte-endothelial cell interactions. By contrast, the inducible NO synthase (iNOS) isoform has been invoked as a potential pathogenic enzyme in numerous inflammatory diseases. The objective of this study was to determine whether the iNOS isoform is also capable of functioning as a regulator of leukocyte recruitment. Mice received endotoxin (LPS, 30 microg/kg, i.v.); 2-4 h later, intravital microscopy was used to examine leukocyte rolling and adhesion in postcapillary venules of the cremaster muscle and the sinusoids and postsinusoidal venules of the hepatic microcirculation. Leukocyte recruitment into the lung was also examined. RT-PCR confirmed that this treatment induced iNOS mRNA expression in wild-type mice as early as 2 h after LPS treatment. Between 2 and 4 h after LPS administration, the number of rolling and adherent leukocytes in cremasteric postcapillary venules and of adherent cells in liver postsinusoidal venules of iNOS-deficient mice were significantly higher than in wild-type mice. Leukocyte accumulation in the lung (measured by myeloperoxidase assay) was also significantly elevated in iNOS-deficient animals. These effects could not be attributed to differences in systemic blood pressure, shear rates, circulating leukocyte numbers, or baseline levels of rolling and adhesion because these parameters were not different between the two groups. To establish whether the differences in leukocyte recruitment were related to the leukocytes per se, perfusion of iNOS+/+ or iNOS-/- septic blood over purified E-selectin (using parallel plate flow chambers) revealed much larger recruitment of iNOS-/- leukocytes. These results suggest that iNOS induced in response to LPS releases NO that is capable of reducing leukocyte accumulation by affecting leukocytes directly and raises the possibility that induction of iNOS is a homeostatic regulator for leukocyte recruitment.

Emigrated rat neutrophils adhere to cardiac myocytes via alpha 4 integrin.

Previous work has shown that neutrophils isolated from whole blood adhere to cardiac-myocytes via CD18 (beta 2 integrin) to cause injury to the heart cells. In vitro, we have found that upon endothelial transmigration, neutrophils can also express alpha 4 beta 1; however, whether this contributes to neutrophil adhesion to parenchymal cells remains entirely unknown. Unstimulated and tumor necrosis factor-alpha-stimulated rat cardiac myocytes adherent to gelatin-coated coverslips supported N-formyl-Met-Leu-Phe (fMLP)-induced neutrophil (isolated from whole blood) adhesion entirely via CD18 (blocked with monoclonal antibody [mAb] WT-3). Emigrated neutrophils spontaneously adhered to cardiac myocytes also entirely via CD18. However, if fMLP was used to restimulate emigrated neutrophils, the adhesion to cardiac myocytes was entirely independent of CD18. Although an anti-alpha 4 integrin antibody (mAb TA-2) alone did not reduce the emigrated neutrophil-myocyte interaction, dual administration of TA-2 and WT-3 reduced adhesion by 81%. alpha 4 integrin was expressed in small amounts on the surface of circulating neutrophils, increased following transmigration, and then increased > 5-fold after restimulation of these emigrated neutrophils. In the presence of the anti-CD18 antibody, a fibronectin fragment (FN-40) but not a vascular cell adhesion molecule-1 antibody (mAb 5F10) inhibitied neutrophil-myocyte interactions by 80%. Similar results were seen when the rat chemokine CINC-gro was used instead of fMLP, suggesting that the alpha 4-dependent adhesion was not specific to fMLP. These data demonstrate that alpha 4 integrin can be physiologically induced to increase in number and avidity after neutrophil emigration and that this adhesion molecule can cause firm adhesion to fibronectin on parenchymal cells, including rat cardiac myocytes.

Neutrophils can adhere via alpha4beta1-integrin under flow conditions.

In this study we investigated the possibility that an alternative pathway exists for neutrophil recruitment, namely an alpha4beta1-dependent pathway. A parallel plate chamber was used to investigate whether neutrophils could tether, roll, and adhere to tumor necrosis factor alpha (TNF alpha)-stimulated endothelium via alpha4beta1. alpha4beta1-integrin was induced on neutrophils using dihydrocytochalasin B and either an endogenous (endothelial-derived) chemotactic agent or an exogenous chemotactic molecule. alpha4beta1-expressing neutrophils could stably adhere under shear force (2 dyne/cm2) to TNF alpha-stimulated endothelium independent of the beta2-integrin. The firm adhesion was entirely abolished by antibodies directed against either the alpha4 or beta1-integrin subunits. However, the rolling interaction was not dependent on alpha4beta1 but was abolished by antiselectin therapy. Neutrophils expressing alpha4beta1 could also tether to the endothelium in the presence of antiselectin therapy, but at shear stresses less than 2 dyne/cm2. alpha4beta1-expressing neutrophils also tethered to and stably adhered (no rolling) to VCAM-1- but not to ICAM-1-transfected L cells. The interaction only occurred at shear stress less than 2 dyne/cm2. A cell line (Ramos) known to express high quantities of alpha4beta1-integrin interacted with VCAM-1-transfected L cells at very similar shear conditions. alpha4beta1-expressing neutrophils were also able to adhere to a second alpha4-integrin ligand, fibronectin; however, this interaction only occurred under static conditions. These data suggest that, under certain conditions, neutrophils can adhere independently of the beta2-integrin pathway and adhere via the alpha4beta1-integrin. This study refutes the concept that alpha4beta1-integrin adhesion is restricted to mononuclear leukocytes and is not functional on human neutrophils.

Promiscuity of clinical Plasmodium falciparum isolates for multiple adhesion molecules under flow conditions.

The central pathologic process in severe Plasmodium falciparum malaria is the cytoadherence of parasitized erythrocytes to capillary and postcapillary venular endothelium, with resultant tissue hypoxia, metabolic disturbances, and multiorgan dysfunction. The molecular basis of this process has been studied extensively using static adhesion assays. In the present study, we determined whether infected red blood cells (IRBC) from clinical parasite isolates would roll and adhere on CD36, ICAM-1, E-selectin, P-selectin, and VCAM-1 using a laminar flow system that allowed for the direct visualization of IRBC-substratum interactions. The results indicate that IRBC could tether and roll on CD36, ICAM-1, P-selectin, and VCAM-1 in a shear-dependent fashion, but significant adhesion was restricted to CD36. There was no interaction with E-selectin. When both CD36 and ICAM-1 were expressed on the same cellular substratum such as C32 melanoma cells, adhesion was significantly greater than when CD36 was present alone. The adhesive interactions were different from those between leukocytes and the same adhesion molecules. Furthermore, IRBC rolling on P-selectin and VCAM-1 was not inhibitable by Abs that entirely prevented leukocyte-receptor interactions. These findings suggest that cytoadherence under physiologic conditions may be a multistep process similar to that involved in the recruitment of a number of different cell types. Further elucidation of the molecular basis of these novel interactions is crucial for the development of therapeutic interventions aimed at inhibiting or reversing the process.

Tumor necrosis factor-alpha induces leukocyte recruitment by different mechanisms in vivo and in vitro.

It is well established that E-selectin is the endothelial adhesion molecule that is primarily responsible for mediating leukocyte rolling on TNF-alpha-stimulated cultured endothelial cells. Despite this, few studies in in vivo inflammatory models have observed reduced leukocyte accumulation using mAbs against E-selectin. The objective of this study was to compare the function of E-selectin on endothelial cells in vitro with its role in TNF-alpha-induced leukocyte recruitment in vivo using EL246, a mAb that blocks the function of E-selectin on activated feline endothelial cells. In vitro experiments using feline endothelial cells showed that EL246 functionally inhibits E-selectin-dependent leukocyte recruitment induced by TNF-alpha, without affecting the function of other rolling mechanisms. Intravital microscopy of single 25- to 40-microm venules in the feline mesentery was then used to examine leukocyte rolling and adhesion in response to superfusion with TNF-alpha. TNF-alpha treatment significantly increased the number of both rolling and adherent leukocytes and significantly decreased leukocyte rolling velocity. Treatment with EL246 (1 mg/kg), either i.v. at the start of the TNF-alpha protocol or directly into the superior mesenteric artery after 3 h of TNF-alpha treatment, had no effect on leukocyte rolling, adhesion, or rolling velocity. However, treatment with the selectin-binding carbohydrate, fucoidan, reduced leukocyte rolling to below baseline levels. These results suggest that in contrast to its prominent role on cultured endothelial cells, E-selectin does not contribute to leukocyte recruitment in TNF-alpha-stimulated feline mesenteric venules in vivo.

Monosodium urate crystals promote neutrophil adhesion via a CD18-independent and selectin-independent mechanism.

The primary objective of this study was to determine whether monosodium urate (MSU) crystals induced neutrophil adhesion to cellular substrata and, if so, then to elucidate the molecular mechanisms involved. Human umbilical vein endothelial cells (HUVEC), as well as various other cellular substrata, were treated with various sized MSU crystals, washed, and then coincubated in the presence of neutrophils for 60 min. HUVEC exposed to MSU crystals but not to silica crystals or uric acid promoted neutrophil adhesion in a dose- and size-dependent manner, an event also observed with monolayers of rabbit synovial cells and rat intestinal epithelial cells. The increased neutrophil adhesion could not be attenuated by anti-CD18, anti-intracellular adhesion molecule-1, or various anti-selectin antibodies, despite the fact that scanning electron microscopy revealed that neutrophils were adhering primarily to the endothelial cells rather than to exposed crystals. CD18-deficient neutrophils adhered to MSU crystal-treated HUVEC as effectively as their CD18-positive counterparts. The neutrophil adhesion was temperature dependent but did not require protein synthesis. Additionally, HUVEC phagocytosis of crystals was necessary for subsequent neutrophil-endothelial cell interactions to transpire. Pretreatment of endothelial cells and neutrophils with colchicine significantly reduced the adhesive interaction. Our data demonstrate that exposure of endothelial and other cells to MSU crystals promotes neutrophil adhesion that occurs by a firm CD18-independent and selectin-independent adhesive mechanism.

A novel beta 1-dependent adhesion pathway on neutrophils: a mechanism invoked by dihydrocytochalasin B or endothelial transmigration.

It is generally accepted that the beta 2-integrin is restricted to mononuclear leukocytes. The objective of this study was to determine whether neutrophils can also express beta 1-integrin (specifically alpha 4 beta 1) and whether this can support neutrophil adhesion to endothelial cells and to extracellular matrix. We stimulated neutrophils with dihydrocytochalasin B (DHCB) and various chemotactic stimuli and observed that chemotactic stimuli induced neutrophil adhesion via beta 2-integrin (CD18), whereas DHCB and either fMLP, PAF, or IL-8 induced adhesion to endothelium or protein-coated plastic that was not inhibitable by anti-CD18 antibody. beta 2-integrin-deficient cells, which did not respond to chemotactic stimuli alone, also adhered avidly in the presence of chemotactic stimuli and DHCB. The induced neutrophil adhesion was inhibited by antibody to beta 1- or alpha 4-integrin chains, but only if an anti-beta 2-integrin antibody was also present. Flow cytometry revealed increased expression of both beta 1 and alpha 4 in the presence of fMLP plus DHCB. Transendothelial migration of neutrophils induced by chemotactic stimuli alone also increased expression of beta 1 and alpha 4. Transmigration across deendothelialized membranes induced a similar beta 1 expression on neutrophils suggesting that events other than an endothelial signal elicited beta 1-integrin expression. Transmigration-induced beta 1-dependent expression translated into only modest adhesion to protein-coated plastic. These data suggest that both a pharmacological (DHCB) and a physiological (transmigration) stimulus can invoke expression of alpha 4 and beta 1 on human neutrophils to mediate adhesion.

Excess nitric oxide does not cause cellular, vascular, or mucosal dysfunction in the cat small intestine.

The overproduction of nitric oxide in the small bowel has been invoked as a cytotoxic event in the vascular, mucosal, and whole organ dysfunction associated with inflammation. We assessed whether exogenous administration of nitric oxide in the form of nitric oxide donors (CAS 754, SIN-1) could cause microvascular and mucosal barrier dysfunction in vivo or epithelial and endothelial cell permeability alterations and cell injury in vitro. Increasing concentrations of CAS 754 or SIN-1 were infused locally into autoperfused segments of cat ileum at 30-min intervals. Baseline epithelial permeability (blood-to-lumen clearance of 51Cr-EDTA) was not affected by CAS 754, whereas vascular protein clearance was reduced. The latter effect could almost entirely be explained by a decrease in intestinal capillary hydrostatic pressure. Therefore, in some experiments venous pressure was elevated and the microvascular reflection coefficient for total proteins was estimated at filtration-independent rates. This direct measurement of microvascular permeability was unaffected by exogenous nitric oxide. CAS 754 did not increase permeability across monolayers of endothelial or epithelial cells and did not cause cell injury. Next, we assessed the possibility that excess nitric oxide may be detrimental, but only in inflamed intestine, by infusing CAS 754 with platelet-activating factor; the latter directly increases microvascular and mucosal permeability. CAS 754 did not exacerbate but rather reduced platelet-activating factor-induced rise in microvascular and mucosal permeability. These results suggest that high concentrations of nitric oxide do not cause breakdown of mucosal or microvascular barrier integrity under normal or inflammatory conditions.

Effects of human neutrophil elastase (HNE) on neutrophil function in vitro and in inflamed microvessels.

The primary objective of this study was to test the hypothesis that human neutrophil elastase (HNE) affects neutrophil infiltration (adhesion and emigration) into inflamed vessels. To determine whether HNE contributes to neutrophil adhesion in vivo, intravital microscopy was used to study neutrophil-endothelial cell interactions in single inflamed postcapillary venules. Superfusion of platelet-activating factor (PAF) (100 nmol/L) onto the mesentery caused an increase in neutrophil-neutrophil interactions, neutrophil adhesion to postcapillary venules, and cellular emigration out of the vasculature. Both L658 758 (an elastase-specific inhibitor), and Eglin C (an elastase and cathepsin G inhibitor) significantly attenuated all of these parameters in vivo. To further characterize the mechanism(s) involved, various in vitro parameters were assessed. HNE, but not trypsin, caused a dose-dependent (0.01 to 1.0 microgram/mL) increase in the expression of the beta subunit (CD18) of the CD11/CD18 adhesive glycoprotein complex on neutrophils. An HNE-dependent increase in CD11b expression was also observed; however, HNE did not affect the expression of other neutrophil adhesion molecules (L-selectin), superoxide production, or degranulation. PAF-enhanced CD18 expression on neutrophils and neutrophil migration were both abolished by L658 758 but PAF-induced neutrophil adhesion to endothelial monolayers was not affected by the antiproteinase. The in vitro data suggest that the antiproteinases do not directly prevent neutrophil adhesion in vivo but may be important in other CD18-dependent events such as neutrophil-neutrophil interaction or neutrophil infiltration (chemotaxis). These results translate into an important, rate-limiting role for elastase in the process of leukocyte infiltration and accumulation in inflamed microvessels.