Down-regulation of human basophil IgE and FC epsilon RI alpha surface densities and mediator release by anti-IgE-infusions is reversible in vitro and in vivo

Previously, infusions of an anti-IgE mAb (rhumAb-E25) in subjects decreased serum IgE levels, basophil IgE and FcepsilonRIalpha surface density, and polyclonal anti-IgE and Ag-induced basophil histamine release responses. We hypothesized that these effects would be reversed in vivo by discontinuation of infusions and in vitro by exposing basophils to IgE. Subjects received rhumAb-E25 biweekly for 46 wk. Blood samples taken 0-52 wk after rhumAb-E25 were analyzed for serum IgE and basophil expression of IgE, FcepsilonRIalpha, and CD32. Basophil numbers were unaffected by infusions. Eight weeks after infusions, free IgE levels rose in vivo but did not reach baseline. Basophil IgE and FcepsilonRIalpha rose in parallel with free IgE while CD32 was stable. FcepsilonRI densities, measured by acid elution, returned to 80% of baseline, whereas histamine release responses returned to baseline. Basophils cultured with or without IgE or IgG were analyzed for expression of IgE, FcepsilonRIalpha, and CD32. By 7 days with IgE, expression of IgE and FcepsilonRIalpha rose significantly, whereas cultures without IgE declined. IgE culture did not effect CD32. IgG culture did not effect expression of any marker. The present results strongly suggest that free IgE levels regulate FcepsilonRIalpha expression on basophils.

alphadbeta2 integrin is expressed on human eosinophils and functions as an alternative ligand for vascular cell adhesion molecule 1 (VCAM-1)

The beta2 family of integrins, CD11a, CD11b, CD11c, and alphad, are expressed on most leukocytes. We show that the newest member of this family, alphad, is expressed on human eosinophils in peripheral blood, and surface expression can be upregulated within minutes by phorbol ester or calcium ionophore A23187. Culture of eosinophils with interleukin 5 (IL-5) leads to a two- to fourfold increase in alphad levels by 3-7 d without a change in alpha4 integrin expression. Eosinophils isolated from late phase bronchoalveolar lavage fluids express alphad at levels similar to that seen after 3 d of IL-5 culture. Regarding alphadbeta2 ligands, in both freshly isolated and IL-5-cultured eosinophils, as well as alphadbeta2-transfected Chinese hamster ovary cells, alphadbeta2 can function as a ligand for vascular cell adhesion molecule 1 (VCAM-1). This conclusion is based on the ability of monoclonal antibodies to alphad, beta2, or VCAM-1 to block cell attachment in static adhesion assays. In experiments with eosinophils, the relative contribution of alphadbeta2 integrin- mediated adhesion is enhanced after IL-5 culture. These experiments demonstrate that alphadbeta2 is an alternative ligand for VCAM-1, and this integrin may play a role in eosinophil adhesion to VCAM-1 in states of chronic inflammation.

Beta 1 integrin-dependent binding of Jurkat cells to fibronectin is regulated by a serine-threonine phosphatase

We investigated the effects of signaling molecule inhibitors on the expression and function of beta1 integrins in Jurkat cells. Jurkat cells expressed alpha4beta1 and alpha5beta1, with significant levels of constitutively activated beta1 integrins as assessed by labeling with mAb 15/7 that distinguishes between activation states. Adhesion to fibronectin (Fn) was mediated equally through alpha4 and alpha5 subunits, and was potentiated by the beta1 integrin activating mAb 8A2. Fn adhesion was decreased by okadaic acid through effects on both alpha4beta1, and alpha5beta1. Tyrphostin A23 also decreased adhesion but was less potent. Neither inhibitor had any effect on the surface expression of total or activated beta1 integrins. The effect of tyrphostin was completely reversed by 8A2; the effect of okadaic acid was only partially reversed. Using Calyculin A, we determined that Jurkat adhesion to Fn was regulated via protein phosphatase 1, independent of the levels of integrins or integrin activation epitopes. Activation of Jurkat cells with a CD3-stimulating mAb enhanced adhesion to Fn and was partially blocked by okadaic acid. These data demonstrate different regulatory pathways for constitutive versus activation-dependent adhesion via beta1 integrins, and implicate both tyrosine kinases and serine-threonine phosphatases in integrin function.

Rat neutrophils express alpha4 and beta1 integrins and bind to vascular cell adhesion molecule-1 (VCAM-1) and mucosal addressin cell adhesion molecule-1 (MAdCAM-1)

The alpha4 integrins, which are constitutively expressed on all human leukocyte subtypes except neutrophils, interact with vascular cell adhesion molecule-1 (VCAM-1) and mucosal addressin cell adhesion molecule (MAdCAM-1) on endothelium to mediate selective recruitment of leukocyte subpopulations, other than neutrophils, to sites of inflammation. However, here we report that a different paradigm of leukocyte recruitment may exist in the rat. Flow cytometric analysis of rat neutrophils using a panel of monoclonal antibodies which recognize rat alpha4 and beta1 integrins showed consistent, low levels of expression. Although alpha4 was expressed at lower levels on neutrophils than all other rat leukocytes, this level of expression was sufficient to mediate significant levels of alpha4- and beta1-dependent neutrophil adhesion to rat and human VCAM-1, and alpha4-dependent, but beta1-independent, adhesion to human MAdCAM-1. These data suggest that rat neutrophils, unlike other species, may use alpha4 integrins to traffic to sites of inflammation in vivo.

Phenotyping and cytokine regulation of the BEAS-2B human bronchial epithelial cell: demonstration of inducible expression of the adhesion molecules VCAM-1 and ICAM-1

Airway epithelium may actively participate in inflammatory responses, such as occur in asthma. The presence and regulation of surface molecules on the airway epithelium, however, is incompletely understood. We have determined the phenotype of the human bronchial epithelial cell line BEAS-2B by flow cytometry. We confirmed previous observations that human bronchial epithelial cells constitutively express CD29, CD44, CD49a, CD49b, CD49c, CD49d, CD49e, CD49f, CD51, CD54 (ICAM-1), CD61, and HLA class 1. BEAS-2B cells were also found to constitutively express CD9, CD13, CD15, CD15s, CD23, CD33, CD36, CD40, CD41b, CD42b, CD48, CD50, CD71, and CD102 (ICAM-2). Culture of BEAS-2B cells with tumor necrosis factor (TNF)-alpha or interleukin (IL)-1beta (1 ng/ml) was found to enhance intercellular adhesion molecule-1 (ICAM-1) expression (several fold) and induce de novo CD106 [vascular cell adhesion molecule-1 (VCAM-1)] expression. TNF-alpha or IL-1beta did not change the expression of CD9, CD13, CD16, CD23, CD29, CD31, CD32, CD35, CD45, CD61, or CD64 in BEAS-2B cells. IL-4 (1 ng/ml) also induced expression of VCAM-1 (1.5-fold) but not ICAM- expression while interferon-gamma (1 ng/ml) enhanced only ICAM-1 expression (2-fold). Maximal VCAM-1 expression was obtained with the combination of TNF-alpha and IL-4 (8-fold). Using Northern blot hybridization analysis, ICAM-1 and VCAM-1 mRNA was detected in BEAS-2B cells stimulated with cytokines. VCAM-1 on stimulated BEAS-2B was functionally active as determined by adhesion of purified eosinophils and blockade with specific antibodies. Primary isolates of bronchial epithelial cells produced detectable levels of VCAM-1 protein and mRNA as detected by enzyme-linked immunosorbent assay and reverse transcription-polymerase chain reaction, respectively. These results suggest that cytokine activation induces expression of ICAM-1 and VCAM-1 on airway epithelium, an event which may influence leukocyte infiltration and activation.

Studies of cell adhesion and flow cytometric analyses of degranulation, surface phenotype, and viability using human eosinophils, basophils, and mast cells

Products derived from eosinophils, basophils, and mast cells are considered critical to the development of allergic diseases. Studies of the selective recruitment, accumulation, and/or activation of these cells during human allergic inflammatory reactions in vitro and in vivo have been facilitated by a wide variety of methods. Some have been developed to identify and isolate these cells from a variety of sites, including blood, airway secretions, and surgical or autopsy tissues. Once enriched in purity, assays of cell adhesion to endothelium, epithelium, matrix proteins, and purified, immobilized counterligands for integrins, selectins, or immunoglobulin gene superfamily structures can be performed in vitro under both static and flow conditions. Techniques involving flow cytometry, utilizing characteristics of cellular light scatter and immunofluorescence, have permitted the elucidation of cell surface phenotype and have aided in quantification of cellular degranulation and viability. These approaches have yielded new information on the function of human eosinophils, basophils, and mast cells and have suggested unique cell-specific pathways of cell recruitment, activation, and survival that may contribute to the pathogenesis of allergic diseases.

Regulation of alpha 4 integrin-mediated adhesion of human eosinophils to fibronectin and vascular cell adhesion molecule-1

BACKGROUND

Eosinophils selectively accumulate at sites of allergic inflammation. Their recruitment is dependent on both the expression and functional activity of cell adhesion molecules. How the functional activity of cell adhesion molecules on eosinophils is regulated is poorly understood.

OBJECTIVE

Our objective was to examine the functional activity of alpha 4 integrins on human eosinophils and its regulation by various agents.

METHODS

Function of alpha 4 integrins on human eosinophils was examined by testing adhesion to immobilized fibronection and vascular cell adhesion molecule-1 (VCAM-1) in the presence or absence of a monoclonal antibody (mAb) (8A2) that activates beta 1 integrin function.

RESULTS

Spontaneous eosinophil adhesion to VCAM-1 was enhanced by 8A2, but adhesion to fibronectin could only be detected in the presence of 8A2. Concentrations of 8A2 that were approximately 100-fold less than saturating induced maximal eosinophil adhesion. Adhesion to VCAM-1 in the presence of 8A2 was effectively inhibited by alpha 4 and beta 1 integrin mAbs: beta 7 mAb had partial inhibitory activity. Connecting segment-1 peptide and alpha 4 mAb blocked 8A2-dependent fibronectin binding: beta 1, beta 2, and beta 7 integrin mAbs had partial inhibitory activity. Eosinophils obtained from bronchoalveolar lavage fluids and blood eosinophils stimulated with IL-5, platelet-activating factor, or RANTES displayed increased beta 2 integrin-dependent, not alpha 4 integrin-dependent, attachment. Spontaneous adhesion of eosinophils to VCAM-1 was significantly reduced by the tyrosine kinase inhibitor tyrphostin B46 (inhibitory concentration of 50% approximately equal to 20 mumol/L); this effect was reversed by 8A2.

CONCLUSIONS

The functional activity of integrins on eosinophils can be positively and negatively regulated. Altered integrin avidity may influence eosinophil recruitment in vivo.

Down-regulation of Fc(epsilon)RI expression on human basophils during in vivo treatment of atopic patients with anti-IgE antibody

Treatment of allergic disease by decreasing circulating IgE with anti-IgE Abs is currently under clinical study. Based on previous unrelated studies, it appeared likely that Fc(epsilon)RI expression on basophils and mast cells might also be regulated by levels of circulating IgE Ab. Therefore, the expression of IgE and Fc(epsilon)RI on human basophils was examined in 15 subjects receiving humanized anti-IgE mAb intravenously. Treatment with the anti-IgE mAb decreased free IgE levels to 1% of pretreatment levels and also resulted in a marked down-regulation of Fc(epsilon)RI on basophils. Median pretreatment densities of Fc(epsilon)RI were approximately 220,000 receptors per basophil and after 3 mo of treatment, the densities had decreased to a median of 8,300 receptors per basophil. Flow cytometric studies, conducted in parallel, showed similar results and also showed in a subset of 3 donors that receptors decreased with a t1/2 of approximately 3 days. The responsiveness of the cells to IgE-mediated stimulation using anti-IgE Ab was marginally decreased (approximately 40%) while the response of the same cells to stimulation with dust mite Ag, Dermatophagoides farinae, was reduced by approximately 90%. One possible explanation for these results is that Fc(epsilon)RI density is directly or indirectly regulated by plasma-free IgE levels.

Down-regulation of Fc(epsilon)RI expression on human basophils during in vivo treatment of atopic patients with anti-IgE antibody

Treatment of allergic disease by decreasing circulating IgE with anti-IgE Abs is currently under clinical study. Based on previous unrelated studies, it appeared likely that Fc(epsilon)RI expression on basophils and mast cells might also be regulated by levels of circulating IgE Ab. Therefore, the expression of IgE and Fc(epsilon)RI on human basophils was examined in 15 subjects receiving humanized anti-IgE mAb intravenously. Treatment with the anti-IgE mAb decreased free IgE levels to 1% of pretreatment levels and also resulted in a marked down-regulation of Fc(epsilon)RI on basophils. Median pretreatment densities of Fc(epsilon)RI were approximately 220,000 receptors per basophil and after 3 mo of treatment, the densities had decreased to a median of 8,300 receptors per basophil. Flow cytometric studies, conducted in parallel, showed similar results and also showed in a subset of 3 donors that receptors decreased with a t1/2 of approximately 3 days. The responsiveness of the cells to IgE-mediated stimulation using anti-IgE Ab was marginally decreased (approximately 40%) while the response of the same cells to stimulation with dust mite Ag, Dermatophagoides farinae, was reduced by approximately 90%. One possible explanation for these results is that Fc(epsilon)RI density is directly or indirectly regulated by plasma-free IgE levels.

Counter-receptors on human basophils for endothelial cell adhesion molecules

Ligands on human basophils for the endothelial adhesion molecules intercellular adhesion molecule-1 (ICAM-1), vascular cell adhesion molecule-1 (VCAM-1), mucosal addressin cell adhesion molecule-1 (MAdCAM-1), and E-selectin were investigated. Adhesion of basophils to endothelial cells was inhibited by mAb recognizing CD18, CD11a, and/or CD11b, with the pattern and magnitude of inhibition dependent upon the activation state of the basophils and endothelium. Adhesion to recombinant VCAM-1 was completely inhibited by mAb recognizing alpha 4 integrin and partially by mAb to the beta 1 or beta 7 subunit; surface expression of these integrins was also detected. Adhesion to recombinant MAdCAM-1 expressed on Chinese hamster ovary cells was completely inhibited by mAb recognizing alpha 4 and/or beta 7 integrins. Adhesion to recombinant E-selectin was completely inhibited by basophil pretreatment with neuraminidase and partially inhibited by endo-beta-galactosidase. By flow cytometry, bimodal patterns of expression of sialyl-Lewis X- and sialyl-dimeric-Lewis X were observed, and adherent cells tended to be sialyl-dimeric-Lewis X positive. Thus, basophils express beta 1, beta 2, and beta 7 integrins along with sialylated surface ligands that may interact with the endothelium during basophil recruitment responses.

Counter-receptors on human basophils for endothelial cell adhesion molecules

Ligands on human basophils for the endothelial adhesion molecules intercellular adhesion molecule-1 (ICAM-1), vascular cell adhesion molecule-1 (VCAM-1), mucosal addressin cell adhesion molecule-1 (MAdCAM-1), and E-selectin were investigated. Adhesion of basophils to endothelial cells was inhibited by mAb recognizing CD18, CD11a, and/or CD11b, with the pattern and magnitude of inhibition dependent upon the activation state of the basophils and endothelium. Adhesion to recombinant VCAM-1 was completely inhibited by mAb recognizing alpha 4 integrin and partially by mAb to the beta 1 or beta 7 subunit; surface expression of these integrins was also detected. Adhesion to recombinant MAdCAM-1 expressed on Chinese hamster ovary cells was completely inhibited by mAb recognizing alpha 4 and/or beta 7 integrins. Adhesion to recombinant E-selectin was completely inhibited by basophil pretreatment with neuraminidase and partially inhibited by endo-beta-galactosidase. By flow cytometry, bimodal patterns of expression of sialyl-Lewis X- and sialyl-dimeric-Lewis X were observed, and adherent cells tended to be sialyl-dimeric-Lewis X positive. Thus, basophils express beta 1, beta 2, and beta 7 integrins along with sialylated surface ligands that may interact with the endothelium during basophil recruitment responses.

Functional regulation of beta 1 integrins on human eosinophils by divalent cations and cytokines

Divalent cations and various soluble stimuli can alter cell adherence by affecting the avidity of adhesion molecules. We hypothesized that beta 1 integrin function of human eosinophils may be altered by divalent cations and eosinophil-activating cytokines such as interleukin-5 (IL-5). Expression of the beta 1 integrin activation epitope recognized by monoclonal antibody (mAb) 15/7 was evaluated by flow cytometry using purified eosinophils from allergic subjects, normal subjects, and late-phase bronchoalveolar lavage (BAL) fluids. Rapid and reversible 15/7 binding on eosinophils from each source was induced in Mn2+ (0.01-1 mM) but not in buffers containing other divalent cations and occurred without affecting the total level of beta 1 integrin expression (quantified using mAb 33B6). Augmentation of eosinophil adhesion to immobilized vascular cell adhesion molecule (VCAM-1) in Mn2+ followed a similar concentration dependence as mAb 15/7 binding. Net binding to VCAM-1 in Mn2+ was completely inhibited with a mixture of alpha 4 and beta 1 integrin mAb while beta 2 integrin mAb had no effect. Exposure of eosinophils from allergic subjects to as little as 1 pg/ml IL-5 completely inhibited mAb 15/7 binding induced by Mn2+. In contrast, increased binding of mAb 15/7 in Mn2+ was not blocked by IL-5 in eosinophils from normal subjects. For eosinophils from allergic subjects, IL-5 also inhibited Mn(2+)-induced adhesion to VCAM-1. Thus, beta 1 integrins on eosinophils from allergic and nonallergic subjects are modulated differently by Mn2+ and IL-5. Altered beta 1 integrin avidity may be one mechanism involved in preferential eosinophil recruitment in vivo.

Carboxyfluorescein diacetate labeling does not affect adhesion molecule expression or function in human neutrophils or eosinophils

Fluorescently labeled leukocytes are commonly used in in vitro and in vivo experimental systems. However, the effects of fluorescent labeling on the expression and function of leukocyte adhesion molecules has not been examined in part because the extreme intensity of fluorescence tends to obscure signals from other fluorochromes used for dual color analysis. We have utilized a novel technique involving a 7-amino-4-methylcoumarin-3-acetic acid (AMCA) fluorophore-conjugated F(ab')2 fragment excitable in the ultraviolet wavelength range (350-450 nm) and dual-laser flow cytometry to determine if labeling of human neutrophils and eosinophils with the fluorescent dye 5-(6)-carboxyfluorescein diacetate (CFDA) alters surface expression of the primary leukocyte adhesion molecules involved in leukocyte-endothelial interactions. Simultaneously, adhesion molecule function was assessed by comparing the ability of CFDA-labeled vs. control cells to adhere to cultured human umbilical vein endothelial cells (HUVEC) and purified immobilized adhesion molecules. Isolated human eosinophils and neutrophils were fluorescently labeled by incubation with CFDA. Flow cytometric comparisons of labeled and unlabeled cells demonstrated that fluorescence labeling of neutrophils and eosinophils with CFDA did not alter basal surface expression of the beta 2 integrins (i.e., CD11a CD11b or, CD18). Stimulation of neutrophils with fMLP and eosinophils with PMA resulted in increased surface expression of CD11b and CD18 which was not altered by CFDA labeling. Likewise, CFDA labeling of neutrophils and eosinophils did not significantly alter their integrin-dependent adhesion to activated HUVEC under static or rotational conditions. Similarly, adhesion to immobilized recombinant E- and P-selectin was unaltered. These data demonstrate that fluorescent labeling of human neutrophils and eosinophils with CFDA does not alter surface expression or function of several adhesion molecules necessary for leukocyte-endothelial interactions. The use of CFDA-labeled cells in experiments employing intravital microscopy should therefore provide valid information on adhesion molecule function in vivo.

Comparison of human eosinophil and neutrophil ligands for P-selectin: ligands for P-selectin differ from those for E-selectin

Eosinophils (EOS) and neutrophils (PMN) display different patterns of accumulation during various inflammatory reactions. We hypothesized that EOS and PMN may differ in their ligands for P-selectin, and that these ligands may differ from those previously identified for E-selectin. Recombinant human P-selectin was immobilized on plastic surfaces and adhesion of 51Cr-labeled human EOS or PMN was compared. EOS and PMN adhered in a concentration-dependent fashion, with similar maximal net adhesion. Preincubation with a blocking P-selectin antibody inhibited adhesion of both cell types, whereas a non-blocking antibody did not. To determine if the counterligands were sialylated proteins, cells were treated with various glycosidases and proteases before testing adhesion. Neuraminidase treatment markedly inhibited binding of both cell types, while endo-beta-galactosidase had no significant effect. Pretreatment with several proteases reduced adhesion of both cell types, although they consistently caused a greater inhibition of PMN binding than EOS binding. To determine whether the P-selectin ligands were surface structures whose expression or function may be altered by cell activation, leukocytes were pretreated with various stimuli; only platelet-activating factor (PAF) treatment reduced the capacity of leukocytes to adhere to P-selectin. Thus, the counterligands for P-selectin on EOS and PMN are similar sialylated, protease-sensitive, endo-beta-galactosidase-resistant structures, whose function and/or expression is reduced following treatment with PAF. These characteristics are clearly different than those reported for EOS and PMN ligands for E-selectin, and suggest disparate roles for P-selectin and E-selectin during EOS and PMN recruitment during inflammatory responses in vivo.

IL-13 selectively induces vascular cell adhesion molecule-1 expression in human endothelial cells

Previous studies with human umbilical vein endothelial cells (HUVEC) have shown that the cytokine IL-4 induces adherence of human eosinophils, but not neutrophils, because of its ability to selectively induce surface expression of vascular cell adhesion molecule-1 (VCAM-1). Because the cytokine IL-13 shares a number of biologic properties with IL-4, we examined the effect of IL-13 on the expression and function of adhesion molecules on HUVEC. Incubation of HUVEC for 4 to 48 h with IL-13 (0.1 to 15 U/ml) induced surface expression of VCAM-1, as detected by indirect immunofluorescence and flow cytometry, without significantly affecting expression of E-selectin or intercellular adhesion molecule-1. The kinetics and maximal IL-13-induced expression of VCAM-1 were similar to those seen with IL-4. Treatment of HUVEC with an optimal concentration of IL-13 (15 U/ml for 24 h) induced adhesiveness for eosinophils, but not for neutrophils, and adhesion was completely inhibited by mAb recognizing VCAM-1 or alpha 4 integrin (CD49d). These results demonstrate that IL-13, like IL-4, selectively stimulates HUVEC to express functional cell surface VCAM-1 and suggest a possible role for IL-13 in promoting VCAM-1/alpha 4 integrin-dependent accumulation of eosinophils during allergic and other inflammatory reactions in vivo.

IL-13 selectively induces vascular cell adhesion molecule-1 expression in human endothelial cells

Previous studies with human umbilical vein endothelial cells (HUVEC) have shown that the cytokine IL-4 induces adherence of human eosinophils, but not neutrophils, because of its ability to selectively induce surface expression of vascular cell adhesion molecule-1 (VCAM-1). Because the cytokine IL-13 shares a number of biologic properties with IL-4, we examined the effect of IL-13 on the expression and function of adhesion molecules on HUVEC. Incubation of HUVEC for 4 to 48 h with IL-13 (0.1 to 15 U/ml) induced surface expression of VCAM-1, as detected by indirect immunofluorescence and flow cytometry, without significantly affecting expression of E-selectin or intercellular adhesion molecule-1. The kinetics and maximal IL-13-induced expression of VCAM-1 were similar to those seen with IL-4. Treatment of HUVEC with an optimal concentration of IL-13 (15 U/ml for 24 h) induced adhesiveness for eosinophils, but not for neutrophils, and adhesion was completely inhibited by mAb recognizing VCAM-1 or alpha 4 integrin (CD49d). These results demonstrate that IL-13, like IL-4, selectively stimulates HUVEC to express functional cell surface VCAM-1 and suggest a possible role for IL-13 in promoting VCAM-1/alpha 4 integrin-dependent accumulation of eosinophils during allergic and other inflammatory reactions in vivo.

Comparison of human eosinophil and neutrophil adhesion to endothelial cells under nonstatic conditions. Role of L-selectin

To simulate adhesion that occurs under conditions of flow, we investigated the attachment of eosinophils to endothelium under rotational conditions. Tissue-culture plates containing monolayers of HUVEC were placed on a horizontal rotator (80 revolutions per minute (rpm)), and equal numbers of purified human eosinophils or neutrophils were added to separate wells at 4 degrees C. Binding of eosinophils and neutrophils to unstimulated endothelial cells was 15 +/- 3 and 31 +/- 11 cells/four high power fields (HPF), respectively. After preincubation of HUVEC with IL-1 beta (1 ng/ml, 4 h, 37 degrees C), adhesion increased to 56 +/- 4 and 290 +/- 26 cells/four HPF, respectively (p < 0.0002 for both, n = 8-14). Eosinophils with reduced levels of L-selectin (blood eosinophils activated in vitro or eosinophils obtained from bronchoalveolar lavage (BAL) performed after segmental lung allergen challenge of allergic subjects) demonstrated reduced binding under rotating conditions. Several L-selectin Abs inhibited adhesion of eosinophils and neutrophils (e.g., LAM1-3: 43 +/- 14% vs 63 +/- 3% inhibition; LAM1-6: 73 +/- 5% vs 36 +/- 6% inhibition, respectively, n > or = 6). Interestingly, one additional L-selectin Ab, LAM1-11, inhibited eosinophil but not neutrophil adhesion (51 +/- 2% vs 1 +/- 7% inhibition, respectively, n > or = 5). We conclude that eosinophils, like neutrophils, use L-selectin to bind to activated endothelial cells under conditions of flow, although mAb LAM1-11 can selectively inhibit eosinophil attachment to stimulated endothelial cells in vitro, suggesting different functional epitopes on L-selectin among eosinophils and neutrophils.

Comparison of human eosinophil and neutrophil adhesion to endothelial cells under nonstatic conditions. Role of L-selectin

To simulate adhesion that occurs under conditions of flow, we investigated the attachment of eosinophils to endothelium under rotational conditions. Tissue-culture plates containing monolayers of HUVEC were placed on a horizontal rotator (80 revolutions per minute (rpm)), and equal numbers of purified human eosinophils or neutrophils were added to separate wells at 4 degrees C. Binding of eosinophils and neutrophils to unstimulated endothelial cells was 15 +/- 3 and 31 +/- 11 cells/four high power fields (HPF), respectively. After preincubation of HUVEC with IL-1 beta (1 ng/ml, 4 h, 37 degrees C), adhesion increased to 56 +/- 4 and 290 +/- 26 cells/four HPF, respectively (p &lt; 0.0002 for both, n = 8-14). Eosinophils with reduced levels of L-selectin (blood eosinophils activated in vitro or eosinophils obtained from bronchoalveolar lavage (BAL) performed after segmental lung allergen challenge of allergic subjects) demonstrated reduced binding under rotating conditions. Several L-selectin Abs inhibited adhesion of eosinophils and neutrophils (e.g., LAM1-3: 43 +/- 14% vs 63 +/- 3% inhibition; LAM1-6: 73 +/- 5% vs 36 +/- 6% inhibition, respectively, n &gt; or = 6). Interestingly, one additional L-selectin Ab, LAM1-11, inhibited eosinophil but not neutrophil adhesion (51 +/- 2% vs 1 +/- 7% inhibition, respectively, n &gt; or = 5). We conclude that eosinophils, like neutrophils, use L-selectin to bind to activated endothelial cells under conditions of flow, although mAb LAM1-11 can selectively inhibit eosinophil attachment to stimulated endothelial cells in vitro, suggesting different functional epitopes on L-selectin among eosinophils and neutrophils.

Differences between human eosinophils and neutrophils in the function and expression of sialic acid-containing counterligands for E-selectin

Both neutrophils and eosinophils have been shown to bind to the inducible endothelial cell adhesion molecule E-selectin. For neutrophils, one of the reported ligands for E-selectin is the sialylated Lewis X Ag (sLe(x)). To analyze the counterligands on eosinophils for E-selectin, adhesion assays were performed in which purified leukocytes were allowed to adhere to a soluble recombinant form of the molecule immobilized on plastic plates. Eosinophils, like neutrophils, bound to immobilized E-selectin, but significantly more neutrophils than eosinophils adhered in this assay. Consistent with the greater ability of neutrophils to bind E-selectin was the observation by flow cytometry that neutrophils expressed significant levels of sLe(x) and a sialylated dimeric form of the Le(x) Ag (sialyl-dimeric Le(x), or sialyl-stage-specific embryonic Ag-1, recognized by mAb FH6), whereas the expression of these epitopes on eosinophils was extremely low or undetectable. Expression was similar on eosinophils from allergic and nonallergic donors, and was not altered on eosinophils after induction of L-selectin shedding in vitro by treatment with platelet-activating factor. For both eosinophils and neutrophils, treatment with sialidase was associated with the complete elimination of sLe(x) and sialyl-dimeric Le(x) surface expression, and abolished leukocyte adhesion to E-selectin. Another glycosidase, endo-beta-galactosidase, which specifically cleaves the beta 1-4 galactose linkage to N-acetyl-glucosamine when it exists in an extended chain form such as that found in sialyl-dimeric Le(x), significantly inhibited eosinophil and neutrophil adhesion and expression of sialyl-dimeric Le(x). Such treatment also reduced sLe(x) expression on eosinophils, while having little effect on total neutrophil sLe(x) expression. For both eosinophils and neutrophils the sialylated ligand did not appear to be a glycoprotein because pretreatment of leukocytes with several proteases had no effect on adhesion to E-selectin or on expression of sLe(x) and sialyl-dimeric Le(x). These data suggest that eosinophils, like neutrophils, use sialylated, protease-resistant structures to bind to E-selectin, although the eosinophil expresses much lower levels of these structures on its surface. A major proportion of the sLe(x)-containing E-selectin ligand on the surface of eosinophils appears to be in the form of sialyl-dimeric Le(x), whereas this represents a minor proportion on the surface of neutrophils. Based on results using endo-beta-galactosidase, it appears that these cells may rely disproportionately upon the cell surface sialyl-dimeric Le(x) to bind to E-selectin.(ABSTRACT TRUNCATED AT 400 WORDS)

Differences between human eosinophils and neutrophils in the function and expression of sialic acid-containing counterligands for E-selectin

Both neutrophils and eosinophils have been shown to bind to the inducible endothelial cell adhesion molecule E-selectin. For neutrophils, one of the reported ligands for E-selectin is the sialylated Lewis X Ag (sLe(x)). To analyze the counterligands on eosinophils for E-selectin, adhesion assays were performed in which purified leukocytes were allowed to adhere to a soluble recombinant form of the molecule immobilized on plastic plates. Eosinophils, like neutrophils, bound to immobilized E-selectin, but significantly more neutrophils than eosinophils adhered in this assay. Consistent with the greater ability of neutrophils to bind E-selectin was the observation by flow cytometry that neutrophils expressed significant levels of sLe(x) and a sialylated dimeric form of the Le(x) Ag (sialyl-dimeric Le(x), or sialyl-stage-specific embryonic Ag-1, recognized by mAb FH6), whereas the expression of these epitopes on eosinophils was extremely low or undetectable. Expression was similar on eosinophils from allergic and nonallergic donors, and was not altered on eosinophils after induction of L-selectin shedding in vitro by treatment with platelet-activating factor. For both eosinophils and neutrophils, treatment with sialidase was associated with the complete elimination of sLe(x) and sialyl-dimeric Le(x) surface expression, and abolished leukocyte adhesion to E-selectin. Another glycosidase, endo-beta-galactosidase, which specifically cleaves the beta 1-4 galactose linkage to N-acetyl-glucosamine when it exists in an extended chain form such as that found in sialyl-dimeric Le(x), significantly inhibited eosinophil and neutrophil adhesion and expression of sialyl-dimeric Le(x). Such treatment also reduced sLe(x) expression on eosinophils, while having little effect on total neutrophil sLe(x) expression. For both eosinophils and neutrophils the sialylated ligand did not appear to be a glycoprotein because pretreatment of leukocytes with several proteases had no effect on adhesion to E-selectin or on expression of sLe(x) and sialyl-dimeric Le(x). These data suggest that eosinophils, like neutrophils, use sialylated, protease-resistant structures to bind to E-selectin, although the eosinophil expresses much lower levels of these structures on its surface. A major proportion of the sLe(x)-containing E-selectin ligand on the surface of eosinophils appears to be in the form of sialyl-dimeric Le(x), whereas this represents a minor proportion on the surface of neutrophils. Based on results using endo-beta-galactosidase, it appears that these cells may rely disproportionately upon the cell surface sialyl-dimeric Le(x) to bind to E-selectin.(ABSTRACT TRUNCATED AT 400 WORDS)

Expression of a functional laminin receptor (alpha 6 beta 1, very late activation antigen-6) on human eosinophils

The mechanisms responsible for the accumulation of eosinophils at sites of allergic and other inflammatory reactions are unknown, but recent studies have implicated both eosinophil and endothelial adhesion molecules in this process. However, less well studied have been the adhesive interactions between eosinophils and the subendothelial basement membrane and interstitial connective tissues. To test the hypothesis that eosinophils might interact with extracellular matrix proteins, we analyzed purified human eosinophils for the expression and function of various beta 1 integrins. Using indirect immunofluorescence and flow cytometry, purified eosinophils from mildly allergic donors were found to consistently express the integrin subunits beta 1 (CD29), alpha 4 (CD49d, very late activation antigen [VLA]-4 alpha), and alpha 6 (CD49f, VLA-6 alpha). No significant expression of the alpha 1, alpha 2, alpha 3, alpha 5, or beta 4 subunits was detected. Platelet contamination of the eosinophil preparations was excluded by light microscopy and by the inability to detect expression of platelet glycoproteins alpha v, CD41b, and CD42b. Immunoprecipitation and sodium dodecyl sulfate-polyacrylamide gel electrophoresis analysis of eosinophils confirmed the expression of cell-surface beta 1, alpha 4, and alpha 6. Furthermore, eosinophils purified from allergic donors were shown to adhere to plate-bound laminin, but not to type 1 or type 4 collagen. Adhesion to laminin was concentration-dependent, required divalent cations, and was completely and specifically inhibited by the anti-alpha 6 monoclonal antibody (MoAb) GoH3 and by the anti-beta 1 MoAb 33B6. Interestingly, the anti-beta 1 MoAb 18D3 (which like 33B6 blocks eosinophil binding to VCAM-1) did not inhibit eosinophil adhesion to laminin, suggesting that there are functionally distinct epitopes on the beta 1 subunit. Eosinophils purified from 4 healthy, nonallergic donors also showed alpha 6-dependent adhesion to laminin, although these cells adhered less well. These studies establish the expression of alpha 6 beta 1 on human eosinophils and document its function as a laminin receptor. Interaction of eosinophil alpha 6 beta 1 with laminin, eg, in basement membranes, may contribute to the localization of these cells at inflammatory sites in vivo.

IL-4 induces adherence of human eosinophils and basophils but not neutrophils to endothelium. Association with expression of VCAM-1

The present studies were performed to explore potentially selective mechanisms of leukocyte adhesion in an attempt to understand how preferential recruitment of eosinophils and basophils might occur during allergic and other inflammatory reactions. Stimulation of human vascular endothelial cells for 24 h with IL-4 (30 to 1,000 U/ml) induced adhesion for eosinophils (up to approximately four-fold of control) and basophils (up to approximately twofold of control) but not neutrophils (less than 125% of control). Analysis of endothelial expression of adhesion molecules by flow cytometry revealed that IL-4 treatment induced vascular cell adhesion molecule-1 (VCAM-1) expression without significantly affecting the expression of other adhesion molecules, namely endothelial-leukocyte adhesion molecule-1 (ELAM-1) or intercellular adhesion molecule-1 (ICAM-1). The concentration-response curve for IL-4-induced VCAM-1 expression paralleled that for adhesion. Endothelial cells stimulated with IL-4 expressed adhesive properties for eosinophils by 3 h; the response increased steadily during a 24-h time course study. Eosinophils and basophils adhered to plates coated with a recombinant form of VCAM-1. This adhesion was blocked with antibodies to VCAM-1 but not ELAM-1. mAb directed against either VCAM-1 or VLA-4 inhibited (by approximately 75%) the binding of eosinophils and basophils to IL-4-stimulated endothelial cells. Because VLA-4 and VCAM-1 have been demonstrated to bind to each other in other adhesion systems, these results suggest that IL-4 stimulates eosinophil and basophil adhesion by inducing endothelial cell expression of VCAM-1 which binds to eosinophil and basophil VLA-4. The lack of expression of VLA-4 on neutrophils and the failure of IL-4 to stimulate neutrophil adherence support this conclusion. It is proposed that local release of IL-4 in vivo in allergic diseases or after experimental allergen challenge may partly explain the enrichment of eosinophils and basophils (vs neutrophils) observed in these situations.

IL-4 induces adherence of human eosinophils and basophils but not neutrophils to endothelium. Association with expression of VCAM-1

The present studies were performed to explore potentially selective mechanisms of leukocyte adhesion in an attempt to understand how preferential recruitment of eosinophils and basophils might occur during allergic and other inflammatory reactions. Stimulation of human vascular endothelial cells for 24 h with IL-4 (30 to 1,000 U/ml) induced adhesion for eosinophils (up to approximately four-fold of control) and basophils (up to approximately twofold of control) but not neutrophils (less than 125% of control). Analysis of endothelial expression of adhesion molecules by flow cytometry revealed that IL-4 treatment induced vascular cell adhesion molecule-1 (VCAM-1) expression without significantly affecting the expression of other adhesion molecules, namely endothelial-leukocyte adhesion molecule-1 (ELAM-1) or intercellular adhesion molecule-1 (ICAM-1). The concentration-response curve for IL-4-induced VCAM-1 expression paralleled that for adhesion. Endothelial cells stimulated with IL-4 expressed adhesive properties for eosinophils by 3 h; the response increased steadily during a 24-h time course study. Eosinophils and basophils adhered to plates coated with a recombinant form of VCAM-1. This adhesion was blocked with antibodies to VCAM-1 but not ELAM-1. mAb directed against either VCAM-1 or VLA-4 inhibited (by approximately 75%) the binding of eosinophils and basophils to IL-4-stimulated endothelial cells. Because VLA-4 and VCAM-1 have been demonstrated to bind to each other in other adhesion systems, these results suggest that IL-4 stimulates eosinophil and basophil adhesion by inducing endothelial cell expression of VCAM-1 which binds to eosinophil and basophil VLA-4. The lack of expression of VLA-4 on neutrophils and the failure of IL-4 to stimulate neutrophil adherence support this conclusion. It is proposed that local release of IL-4 in vivo in allergic diseases or after experimental allergen challenge may partly explain the enrichment of eosinophils and basophils (vs neutrophils) observed in these situations.

Adhesion of human basophils, eosinophils, and neutrophils to interleukin 1-activated human vascular endothelial cells: contributions of endothelial cell adhesion molecules

Cytokines such as interleukin 1 (IL-1) promote adhesiveness in human umbilical vein endothelial cells for leukocytes including basophils, eosinophils, and neutrophils, and induce expression of adherence molecules including ICAM-1 (intercellular adhesion molecule-1), ELAM-1 (endothelial-leukocyte adhesion molecule-1), and VCAM-1 (vascular cell adhesion molecule-1). In the present study, blocking monoclonal antibodies (mAb) recognizing ICAM-1, ELAM-1, and VCAM-1 have been used to compare their roles in IL-1-induced adhesion of human basophils, eosinophils, and neutrophils. IL-1 treatment of endothelial cell monolayers for 4 hours induced a four- to eight-fold increase in adhesion for each cell type. Treatment of endothelial cells with either anti-ICAM-1 or anti-ELAM-1 mAb inhibited IL-1-induced adherence of each cell type. In contrast, treatment with anti-VCAM-1 mAb inhibited basophil and eosinophil (but not neutrophil) adhesion, and was especially effective in blocking eosinophil adhesion. The effects of these mAb were at least additive. Indirect immunofluorescence and flow cytometry demonstrated expression of VLA-4 alpha (very late activation antigen-4 alpha, a counter-receptor for VCAM-1) on eosinophils and basophils but not on neutrophils. These data document distinct roles for ICAM-1, ELAM-1, and VCAM-1 during basophil, eosinophil, and neutrophil adhesion in vitro, and suggest a novel mechanism for the recruitment of eosinophils and basophils to sites of inflammation in vivo.

Altered surface expression of CD11 and Leu 8 during human basophil degranulation

Immunofluorescence and flow cytometric techniques have been used to study changes in surface Ag expression and viability that occur during human basophil degranulation. Treatment with polyclonal anti-IgE, FMLP, or the calcium ionophore A23187 induced histamine release, along with rapid and sustained unimodal increases in basophil CD11b mean fluorescence intensity. In contrast, treatment with anti-IgE or FMLP resulted in a decrease in Leu 8 expression. Degranulation did not significantly affect basophil viability (as determined by exclusion of propidium iodide), scatter characteristics, or percentage of identifiable IgE-bearing cells, and an inconsistent association was seen between percent histamine release and reduction in the percent of cells identified by light microscopy after staining with alcian blue. For anti-IgE, dose-dependent changes in CD11b, CD11c, and Leu 8 expression were seen (optimal at 0.1, 0.1, and 1 microgram/ml, respectively), although CD11a expression remained unchanged. Histamine release was optimal at 0.3 microgram/ml anti-IgE, and at superoptimal concentrations, reduced CD11b expression was observed which paralleled decreases in histamine release; reduction of the expression of Leu 8, however, occurred equally at optimal and superoptimal concentrations of anti-IgE. Kinetic analyses of these responses revealed that CD11b up-regulation proceeded more rapidly than histamine release, whereas Leu 8 down-regulation was much slower and did not plateau until 120 min of stimulation. Although changes in CD11b mean fluorescence intensity correlated with the magnitude of histamine release, exposure to stimuli in the absence of calcium (which blocked degranulation) resulted in similar alterations in CD11b and Leu 8, suggesting that degranulation was not required for changes in the surface expression of these adhesion molecules. Interestingly, pretreatment of basophils with drugs that either inhibited or enhanced histamine release (isobutylmethylxanthine and cyclosporin A vs cytochalasin B, respectively) significantly decreased the magnitude of anti-IgE-induced CD11b up-regulation; down-regulation of Leu 8 expression was also partially inhibited by treatment with isobutylmethylaxanthine. These studies demonstrate that activation of human basophils by secretagogues in vitro results in a variety of phenotypic changes including alterations in surface expression of adhesion molecules, and suggest that degranulation in vivo may be accompanied or preceded by changes in adhesion-related functions.

Altered surface expression of CD11 and Leu 8 during human basophil degranulation

Immunofluorescence and flow cytometric techniques have been used to study changes in surface Ag expression and viability that occur during human basophil degranulation. Treatment with polyclonal anti-IgE, FMLP, or the calcium ionophore A23187 induced histamine release, along with rapid and sustained unimodal increases in basophil CD11b mean fluorescence intensity. In contrast, treatment with anti-IgE or FMLP resulted in a decrease in Leu 8 expression. Degranulation did not significantly affect basophil viability (as determined by exclusion of propidium iodide), scatter characteristics, or percentage of identifiable IgE-bearing cells, and an inconsistent association was seen between percent histamine release and reduction in the percent of cells identified by light microscopy after staining with alcian blue. For anti-IgE, dose-dependent changes in CD11b, CD11c, and Leu 8 expression were seen (optimal at 0.1, 0.1, and 1 microgram/ml, respectively), although CD11a expression remained unchanged. Histamine release was optimal at 0.3 microgram/ml anti-IgE, and at superoptimal concentrations, reduced CD11b expression was observed which paralleled decreases in histamine release; reduction of the expression of Leu 8, however, occurred equally at optimal and superoptimal concentrations of anti-IgE. Kinetic analyses of these responses revealed that CD11b up-regulation proceeded more rapidly than histamine release, whereas Leu 8 down-regulation was much slower and did not plateau until 120 min of stimulation. Although changes in CD11b mean fluorescence intensity correlated with the magnitude of histamine release, exposure to stimuli in the absence of calcium (which blocked degranulation) resulted in similar alterations in CD11b and Leu 8, suggesting that degranulation was not required for changes in the surface expression of these adhesion molecules. Interestingly, pretreatment of basophils with drugs that either inhibited or enhanced histamine release (isobutylmethylxanthine and cyclosporin A vs cytochalasin B, respectively) significantly decreased the magnitude of anti-IgE-induced CD11b up-regulation; down-regulation of Leu 8 expression was also partially inhibited by treatment with isobutylmethylaxanthine. These studies demonstrate that activation of human basophils by secretagogues in vitro results in a variety of phenotypic changes including alterations in surface expression of adhesion molecules, and suggest that degranulation in vivo may be accompanied or preceded by changes in adhesion-related functions.

IL-3 augments adhesiveness for endothelium and CD11b expression in human basophils but not neutrophils

A number of natural and recombinant human cytokines have been tested for their ability to activate basophil and neutrophil adhesiveness for human umbilical vein endothelial cells in vitro. Coincubation of basophils and endothelial cell monolayers for 10 min with biologically relevant concentrations of rIL-1, natural IL-2, rIL-4, rIL-5, rIL-6, rIL-8, rGM-CSF, and rIFN-gamma had no effect on basophil adhesiveness. In contrast, rIL-3 induced basophil adhesiveness for endothelial cells (optimal at 1 ng/ml: 144 +/- 18% of control adherence (mean +/- SEM); control basophil binding, 13 +/- 3%, n = 9, p less than or equal to 0.05). This increase in adhesiveness was similar in magnitude to that induced by an optimal concentration of a known potent inducer of basophil adhesiveness (1 microM FMLP, 164 +/- 15% of control adherence, n = 9). Under these experimental conditions, the effects of rIL-3 occurred at concentrations of 0.1 to 30 ng/ml, were partially dependent on calcium, and were not accompanied by histamine release. Fixation experiments demonstrated that the effect of rIL-3 was directed against the basophil rather than the endothelial cell. Neither rIL-3 nor the other cytokines tested had any effect on the adherence of 51Cr-labeled neutrophils, even when tested simultaneously on cells from the same donors. Under experimental conditions that permitted histamine release, no correlation was seen between the ability of rIL-3 (0.3 to 300 ng/ml) to induce histamine release or enhance adhesiveness (n = 8). mAb blocking experiments demonstrated a role for both CD11 and CD18 adherence glycoproteins in basophil adherence induced by rIL-3, and indirect immunofluorescence and flow cytometric analysis revealed that rIL-3 treatment led to rapid and sustained increases in cell surface expression of CD11b antigens on basophils but not neutrophils (e.g., after 10 min: 217 +/- 29 vs 91 +/- 11% of control mean fluorescence intensity, p less than 0.05). However, no correlation was seen between the magnitude of changes in CD11b expression and changes in adhesion when tested simultaneously. These results suggest that local production of IL-3 during allergic reactions in vivo may selectively promote basophil activation, adhesion to endothelium, and recruitment to extravascular sites of inflammation.

IL-3 augments adhesiveness for endothelium and CD11b expression in human basophils but not neutrophils

A number of natural and recombinant human cytokines have been tested for their ability to activate basophil and neutrophil adhesiveness for human umbilical vein endothelial cells in vitro. Coincubation of basophils and endothelial cell monolayers for 10 min with biologically relevant concentrations of rIL-1, natural IL-2, rIL-4, rIL-5, rIL-6, rIL-8, rGM-CSF, and rIFN-gamma had no effect on basophil adhesiveness. In contrast, rIL-3 induced basophil adhesiveness for endothelial cells (optimal at 1 ng/ml: 144 +/- 18% of control adherence (mean +/- SEM); control basophil binding, 13 +/- 3%, n = 9, p less than or equal to 0.05). This increase in adhesiveness was similar in magnitude to that induced by an optimal concentration of a known potent inducer of basophil adhesiveness (1 microM FMLP, 164 +/- 15% of control adherence, n = 9). Under these experimental conditions, the effects of rIL-3 occurred at concentrations of 0.1 to 30 ng/ml, were partially dependent on calcium, and were not accompanied by histamine release. Fixation experiments demonstrated that the effect of rIL-3 was directed against the basophil rather than the endothelial cell. Neither rIL-3 nor the other cytokines tested had any effect on the adherence of 51Cr-labeled neutrophils, even when tested simultaneously on cells from the same donors. Under experimental conditions that permitted histamine release, no correlation was seen between the ability of rIL-3 (0.3 to 300 ng/ml) to induce histamine release or enhance adhesiveness (n = 8). mAb blocking experiments demonstrated a role for both CD11 and CD18 adherence glycoproteins in basophil adherence induced by rIL-3, and indirect immunofluorescence and flow cytometric analysis revealed that rIL-3 treatment led to rapid and sustained increases in cell surface expression of CD11b antigens on basophils but not neutrophils (e.g., after 10 min: 217 +/- 29 vs 91 +/- 11% of control mean fluorescence intensity, p less than 0.05). However, no correlation was seen between the magnitude of changes in CD11b expression and changes in adhesion when tested simultaneously. These results suggest that local production of IL-3 during allergic reactions in vivo may selectively promote basophil activation, adhesion to endothelium, and recruitment to extravascular sites of inflammation.