Interleukin-15 induces adhesion receptor redistribution in T lymphocytes

Chemotactic factors such as cytokines and chemokines direct the migration of leukocytes into inflammatory sites. Chemokines play a role regulating both the expression and adhesive properties of leukocyte integrins. We have recently described an additional function of chemokines in the induction of cell polarization and adhesion receptor redistribution during the initial step of leukocyte locomotion. We herein report that interleukin (IL)-15, a newly described cytokine with chemotactic properties, is able to induce uropod formation on T lymphoblasts to which intercellular adhesion molecule (ICAM)-3, a leukocyte-restricted counter-receptor for the lymphocyte function-associated antigen (LFA)-1 integrin, is redistributed. Other adhesion molecules, such as ICAM-1, ICAM-2, CD43 and CD44, also redistributed to the uropod, although in a lower proportion of the cells. The induction of uropod formation by IL-15 was observed on T lymphoblasts adhering to the integrin ligands fibronectin, vascular cell adhesion molecule (VCAM)-1 and ICAM-1, but not to bovine serum albumin or poly-L-lysine. The effect of IL-15 was dose dependent and specifically inhibited by a monoclonal antibody (mAb) against this cytokine. Blocking experiments with anti-IL-2 receptor beta chain mAb showed an inhibitory effect on IL-15-mediated redistribution of ICAM-3, whereas no effect was observed in the presence of anti-IL-2 receptor alpha chain mAb. The uropod induced by IL-15 is enriched in many different adhesion receptors and, being well exposed to the external milieu, is likely to modulate the adhesive properties of lymphocytes.     

Ionizing radiation mediates expression of cell adhesion molecules in distinct histological patterns within the lung

Inflammatory cell infiltration of the lung is a predominant histopathological change that occurs during radiation pneumonitis. Emigration of inflammatory cells from the circulation requires the interaction between cell adhesion molecules on the vascular endothelium and molecules on the surface of leukocytes. We studied the immunohistochemical pattern of expression of cell adhesion molecules in lungs from mice treated with thoracic irradiation. After X-irradiation, the endothelial leukocyte adhesion molecule 1 (ELAM-1; E-selectin) was primarily expressed in the pulmonary endothelium of larger vessels and minimally in the microvascular endothelium. Conversely, the intercellular adhesion molecule 1 (ICAM-1; CD54) was expressed in the pulmonary capillary endothelium and minimally in the endothelium of larger vessels. Radiation-mediated E-selectin expression was first observed at 6 h, whereas ICAM-1 expression initially increased at 24 h after irradiation. ICAM-1 and E-selectin expression persisted for several days. P-selectin is constitutively expressed in Weibel-Palade bodies in the endothelium, which moved to the vascular lumen within 30 min after irradiation. P-selectin was not detected in the pulmonary endothelium at 6 h after irradiation. The radiation dose required for increased cell adhesion molecule expression within the pulmonary vascular endothelium was 2 Gy, and expression increased in a dose-dependent manner. These data demonstrate that ICAM-1 and E-selectin expression is increased in the pulmonary endothelium following thoracic irradiation. The pattern of expression of E-selectin, P-selectin, and ICAM-1 is distinct from one another.     

Effects of aminoguanidine on adhesion molecule expression of human endothelial cells

The effect of aminoguanidine (AG) on the expression of adhesion molecules on nonactivated human umbilical vein endothelial cells (HUVEC) was investigated in vitro. Nonactivated HUVEC cultivated on long-term glycated fibronectin (FN) as compared to native FN showed a significant upregulation of intercellular adhesion molecule 1 (ICAM-1), vascular cell adhesion molecule 1 (VCAM-1) and CD31 which could be further promoted by long-term glycated bovine serum albumin. AG, at a concentration of 0.01 mol/l, caused an upregulation of ICAM-1 of 48 +/- 17.4% in HUVEC cultivated on gelatin. In contrast, VCAM-1 and E-selectin remained unaffected. At this concentration, formation of advanced glycation end products (AGE) was inhibited by 57%, as determined immunologically, and by 50%, as verified by AGE-specific fluorescence. A hypothesis concerning the upregulation of ICAM-1 by AG as compared to VCAM-1 is proposed relating to its relative redox insensitivity. Our results demonstrate that the beneficial effect of AG in reducing the risk of accelerated development of atherosclerosis in diabetic patients by inhibiting formation of AGE on matrix proteins such as FN might be hampered by its tendency to upregulate ICAM-1 on endothelial cells.     

Influence of adhesion molecule expression by human brain microvessel endothelium on cancer cell adhesion

Cultures of endothelial (En) cells derived from human brain microvessels were established in order to characterize adhesion molecule expression and to assay the adhesion properties of neoplastic cell lines to monolayers of En cells. Low constitutive expression of beta1 integrin (CD29), and ICAM-2 (CD102) was detected on human brain microvessel En cells. The beta1 chain of the VLA integrin family, ICAM-1, E-selectin (CD62E) and VCAM-1 (CD106) but not ICAM-2 and PECAM-1 (CD31) expression was upregulated by IL1-alpha, and TNF-alpha proinflammatory cytokines. High expression of PECAM-1 was found on non-activated human brain EN cells. In order to study the potential role of adhesion molecules in neoplastic cell adhesion two tumor cell lines were chosen. Adhesion of a cell line (DU145) derived from a cerebral metastasis of prostate carcinoma to human brain microvessel En cell monolayers was less pronounced compared to adhesion of a primary prostate carcinoma cell line (ND1). Adhesion of cerebral metastatic neoplastic cell line (DU145) was not significantly influenced by incubation of endothelial cells with different proinflammatory cytokines. The adhesion capability of primary prostate carcinoma line (NDI) was significantly upregulated by TNF-alpha proinflammatory cytokine. Furthermore, the adhesion of ND1 was partly inhibited using anti-E-selectin and VCAM-1 monoclonal antibodies. There was no significant effect of anti-adhesion antibodies on the adhesion characteristics of the cerebral metastatic (DU145) cell line. Our data demonstrate that different mechanisms are involved in the adhesion of neoplastic cells to cerebral En cells and turn our attention to the importance of adhesion molecule expression in the formation of metastases.     

Circulating intercellular adhesion molecule 1 (ICAM-1), E-selectin and vascular cell adhesion molecule 1 (VCAM-1) in head and neck cancer

The sera from patients with nasopharyngeal carcinoma (n = 30), oral carcinoma (n = 22) and laryngeal carcinoma (n = 22) was extracted before treatment. The concentration of circulating intercellular adhesion molecule 1 (ICAM-1), E-selectin and vascular cell adhesion molecule 1 (VCAM-1) was measured by enzyme-linked immunoassay and compared with those from normal subjects (n = 20). The concentration of circulating ICAM-1, E-selectin and VCAM-1 was significantly increased in nasopharyngeal carcinoma. Correspondingly, VCAM-1 and E-selectin were significantly increased in laryngeal carcinoma, whereas only E-selectin was elevated in oral carcinoma. The concentrations of these adhesion molecules did not significantly differ with respect to the early and late stages of these carcinomas. Elevated levels of soluble adhesion molecules in the sera of cancer patients at three different head and neck regions, although appearing to be implicated in these tumour formations, may be unrelated to tumour progression.     

Reduced susceptibility to collagen-induced arthritis in mice deficient in intercellular adhesion molecule-1

Intercellular adhesion molecule-1 (ICAM-1) plays an important role in the firm adhesion of leukocytes to venular endothelium and facilitates leukocyte extravasation from the vasculature into inflamed tissue. In addition, ICAM-1 is an important costimulatory molecule during Ag presentation to lymphocytes. Using mice deficient in ICAM-1, we have investigated the role of this molecule in the development of collagen-induced arthritis. After immunization with type II collagen, 71% of wild-type mice developed arthritis compared with 50% of ICAM-1 heterozygote mutants and 18% of ICAM-1 homozygous mutants. In those ICAM-1 mutants that developed arthritis, the mean day of onset, the mean number of involved paws, and the severity of paw inflammation were not significantly different from those in wild-type mice. The reduced incidence of arthritis in the ICAM-1 homozygous mutant mice was not due to lack of immunity to type II collagen, since these mice developed similar levels of anti-type II collagen IgG compared with wild-type mice and had a positive delayed-type hypersensitivity reaction to type II collagen. The reduction of arthritis in heterozygous as well as homozygous deficient mice indicates that expression of ICAM-1 can be a pivotal variable in the pathogenesis of collagen-induced arthritis in mice. The results suggest that naturally occurring genetic variation in the expression of ICAM-1 or related inflammatory cell adhesion molecules might influence susceptibility to the complex disease of rheumatoid arthritis in humans and support the concept that pharmacologic approaches to chronic reduction in the expression or the function of ICAM-1 may be of therapeutic value.     

Release of soluble intercellular adhesion molecule 1 into bile and serum in murine endotoxin shock

Neutrophil-induced liver injury during endotoxemia is dependent on the adhesion molecules Mac-1 (CD11b/CD18) on neutrophils and its counterreceptor on endothelial cells and hepatocytes, intercellular adhesion molecule 1 (ICAM-1). To investigate a potential release of a soluble form of ICAM-1 (sICAM-1), animals received 100 micrograms/kg Salmonella abortus equi endotoxin alone or in combination with 700 mg/kg galactosamine. In endotoxin-sensitive mice (C3Heb/FeJ), injection of endotoxin did not cause liver injury but induced a time-dependent increase of sICAM-1 in serum (300%) and in bile (615%) without affecting bile flow. In galactosamine/endotoxin-treated animals, which developed liver injury, the increase in both compartments was only 97% and 104%, respectively. In either case, the increase in sICAM-1 concentrations paralleled the enhanced ICAM-1 expression in the liver. The endotoxin-resistant strain (C3H/HeJ) did not show elevated sICAM-1 levels in serum or bile after endotoxin administration. In contrast, the intravenous injection of murine tumor necrosis factor alpha (TNF-alpha), interleukin-1 alpha (IL-1 alpha) or IL-1 beta (13-23 micrograms/kg) into endotoxin-resistant mice induced a 225% to 364% increase in serum sICAM-1 and a 370% elevation of the biliary efflux of sICAM-1, again independent of changes in bile flow. These data indicate that cytokines are major inducers of sICAM-1 formation during endotoxemia in vivo. The described experimental model can be used to investigate the role of sICAM-1 in the pathophysiology of inflammatory liver disease.     

Anti-adhesion molecule therapy in Theiler's murine encephalomyelitis virus-induced demyelinating disease

We examined the role of leukocyte function-associated antigen (LFA)-1 and its counter-receptor intercellular adhesion molecule (ICAM)-1, one of the most important pairs of adhesion molecules, in the development of Theiler's murine encephalomyelitis virus-induced demyelinating disease (TMEV-IDD). Immunohistochemical study showed hyper-expression of ICAM-1 on vascular endothelial cells and expression of LFA-1 on mononuclear infiltrating cells in the spinal cords of TMEV-infected mice. Treatment with mAb to ICAM-1 and/or LFA-1 molecules resulted in significant suppression of the development of demyelinating disease, both clinically and histologically, with down-regulation in the CNS of the respective adhesion molecules after treatment. In mice treated with these mAb, the specific delayed-type hypersensitivity and T cell proliferative responses for TMEV were decreased. The production of tumor necrosis factor-alpha and IFN-gamma in spleen cells was also decreased, but IL-4 production remained unchanged. These data suggest that ICAM-1/LFA-1 interaction is critically involved in the pathogenesis of TMEV-IDD and that antibodies to these adhesion molecules could be a novel therapeutic approach to the treatment of demyelinating diseases such as human multiple sclerosis.     

Cloning and sequencing of a cDNA encoding bovine intercellular adhesion molecule 3 (ICAM-3)

A cDNA encoding a putative bovine intercellular adhesion molecule (ICAM)-3, a ligand of the leukocyte integrin LFA-1 (CD11a/CD18), was sequenced and compared with human ICAM sequences. The 1635-bp bovine sequence codes for a protein of 544 amino acids (aa). This putative bovine ICAM-3 has five immunoglobulin (Ig)-like domains similar to human ICAM-1 and ICAM-3, and belongs to the Ig gene superfamily. The overall identities of the deduced aa sequence with those of human ICAM-3 and ICAM-1 are 61% and 58%, respectively. The predicted number and positions of Cys residues are all conserved between the bovine and human ICAM 3 aa sequences.     

Differential contribution of various adhesion molecules to leukocyte kinetics in pulmonary microvessels of hyperoxia-exposed rat lungs

To elucidate the differential role of various adhesion molecules in distorting leukocyte behavior in the microvasculature of hyperoxia-exposed rat lungs, we investigated fluorescein-labeled leukocyte and erythrocyte kinetics in isolated lungs taken from the animals exposed to 90% O2 for 48 h under conditions in which endothelial intercellular adhesion molecule-1 (ICAM-1) and P-selectin were inhibited by appropriate monoclonal antibodies (1A29 for ICAM-1 and ARP2-4 for P-selectin), while leukocyte L-selectin was restrained with fucoidin. Measurements of blood cell kinetics were made by a confocal laser luminescence microscope coupled with a high-speed video camera. In addition, we histologically examined leukocyte accumulation within the alveolar septa and ICAM-1 as well as P-selectin expressions in the lung. We found that P-selectin expression was sparsely enhanced only in arterioles, whereas ICAM-1 was significantly induced in both venules and capillaries. Firm adhesion of leukocytes was not identified in arterioles and venules, whereas leukocyte rolling was evident in both the vessels. Arteriolar rolling was regulated via a P-selectin- and ICAM-1-independent but L-selectin-dependent mechanism, whereas venular rolling was mediated via a P-selectin-independent but ICAM-1- and L-selectin-dependent pathway. Leukocyte sequestration within capillaries was augmented by an ICAM-1-related mechanism. These findings may suggest that, in hyperoxia-exposed lungs, induction of adhesion molecules and their obstacles to leukocyte behavior are qualitatively different among arterioles, venules, and capillaries.     

The cell-mediated response to schistosomal antigens at the clonal level. III. Identification of soluble egg antigens recognized by cloned specific granulomagenic murine CD4+ Th1-type lymphocytes

Granulomatous inflammation in schistosomiasis is a consequence of T cell-mediated hypersensitivity to parasite egg Ag. In the present study we used three consecutive independent chromatographic procedures to fractionate and identify the soluble egg Ag recognized by schistosome-specific, cloned, murine, CD4+ Th1-type lymphocytes, which had been shown previously to be capable of mediating granuloma formation in vivo when adoptively transferred to normal syngeneic hosts challenged with an i.v. injection of eggs. The stimulatory activity resided in two acidic egg molecules, with apparent molecular masses of 64 to 68 kDa and 38 to 42 kDa, each of which ran as a single band on SDS-PAGE after purification. Fast performance liquid chromatography and SDS-PAGE performed under reducing conditions suggested that the two molecules are related and that the 38- to 42-kDa molecule is a subunit of the 64- to 68-kDa molecule. Polyclonal lymphoid cells from schistosome-infected mice were similarly stimulated by the purified 64- to 68-kDa and 38- to 42-kDa molecules, implying that these are sensitizing Ag in the natural disease.     

Deficiency of intercellular adhesion molecule 1 attenuates microcirculatory disturbance and infarction size in focal cerebral ischemia

Recent evidence has shown crucial roles for cell-adhesion molecules in inflammation-induced rolling, adhesion, and accumulation of neutrophils in tissue. Intercellular adhesion molecule-1 (ICAM-1) is one of these adhesion molecules. Previous studies have shown marked reduction in the size of infarction after focal cerebral ischemia by depletion of granulocytes and administration of the antibody against ICAM-1. In the present study we investigated the role of ICAM-1 in the size of ischemic lesions, accumulation of granulocytes, and microcirculatory compromise in focal cerebral ischemia by using ICAM-1-knockout mice. Ischemic lesions were significantly mitigated in knockout mice after permanent and transient focal ischemia, even though the number of granulocytes in the infarcted tissue was almost the same between knockout and wild-type mice. Depletion of granulocytes further decreased the size of ischemic lesions after transient focal ischemia in ICAM-1-knockout mice. Microcirculation was reduced after focal ischemia, but it was better preserved in the cerebral cortex of knockout mice than that of wild-type mice. The present study demonstrated that ICAM-1 played a role in microcirculatory failure and subsequent development and expansion of infarction after focal cerebral ischemia. However, it is highly unlikely that ICAM-1 played a key role in accumulation of granulocytes after focal cerebral ischemia.     

Cytokine secretion and adhesion molecule expression by granuloma T lymphocytes in Mycobacterium avium infection

Mice experimentally infected with Mycobacterium avium develop a chronic disease characterized by widespread noncaseating granulomas. In this report, we describe the phenotype and cytokine secretion profile of these granuloma-infiltrating effector T lymphocytes. In response to specific antigen, granuloma T cells and, to a lesser extent, spleen cells secrete interferon-gamma, but no interleukin-4 or -5. The importance of this Th1-like response to the host was demonstrated by the massively increased bacterial load and lethal disease in interferon-gamma knockout mice. One function of localized cytokine secretion is to recruit inflammatory T cells bearing surface adhesion molecules complementary to counter-receptors on vascular endothelial cells. Granuloma T cells express high levels of these pro-inflammatory adhesion molecules but have down-regulated their expression of L-selectin (CD62L). The expression of these adhesion molecules on granuloma-infiltrating T lymphocytes would alter the migration pathway of these cells and is likely to be important in facilitating the traffic of effector T cells to the granulomatous inflammatory site. In addition, T cells from Schistosoma mansoni granulomas express the same set of adhesion molecules, showing that this phenotype is not specifically dependent upon the Th1 pattern of cytokine secretion.     

Blocking of classical complement pathway inhibits endothelial adhesion molecule expression and preserves ischemic myocardium from reperfusion injury

Myocardial injury after ischemia (I) and reperfusion (R) is related to leukocyte activation with subsequent release of cytokines and oxygen-derived free radicals as well as complement activation. In our study, the cardioprotective effects of exogenous C1 esterase inhibitor (C1 INH) were examined in a rat model of myocardial I + R (i.e., 20 min + 24 hr or 48 hr). The C1 INH (10, 50 and 100 U/kg) administered 2 min before reperfusion significantly attenuated myocardial injury after 24 hr of R compared to vehicle treated rats (P < .001). Further, cardiac myeloperoxidase activity (i.e., a marker of PMN [polymorphonuclear leukocyte] accumulation) in the ischemic area was significantly reduced after C1 INH treatment compared to vehicle treated animals (0.81 +/- 0.1, 0.34 +/- 0.13, 0.13 +/- 0.1 vs. 1.44 +/- 0.3 U/100 mg tissue, P < .001). In addition, C1 INH (100 U/kg) significantly attenuated myocardial injury and neutrophil infiltration even after 48 hr of reperfusion compared to vehicle treatment. Immunohistochemical analysis of ischemic-reperfused myocardial tissue demonstrated activation of classical complement pathway by deposition of C1q on cardiac myocytes and cardiac vessels. In addition, expression of the endothelial adhesion molecules P-selectin and intercellular adhesion molecule 1 (ICAM-1) was observed after reperfusion of the ischemic myocardium. In this regard, C1 INH administration abolished expression of P-selectin and ICAM-1 on the cardiac vasculature after myocardial ischemia and reperfusion. Blocking the classical complement pathway by exogenous C1 INH appears to be an effective means to preserve ischemic myocardium from injury after 24 and 48 hr of reperfusion. The mechanisms of this cardioprotective effect appears to be due to blocking of complement activation and reduced endothelial adhesion molecule expression with subsequent reduced PMN-endothelium interaction, resulting in diminished cardiac necrosis.     

Spatio-temporal diversity in the microenvironments for neural cell adhesion molecule, neural cell adhesion molecule-polysialic acid, and L1-cell adhesion molecule expression by sensory neurons and their targets during cochleo-vestibular innervation

Sixteen phases in the microenvironments were defined for the structural development and innervation of the cochleo-vestibular ganglion and its targets. In each phase the cell adhesion molecules, neural cell adhesion molecule, neural cell adhesion molecule-polysialic acid, and L1-cell adhesion molecule, were expressed differentially by cochleo-vestibular ganglion cells, their precursors, and the target cells on which they synapse. Detected by immunocytochemistry in staged chicken embryos, in the otocyst, neural cell adhesion molecule, but not L1-cell adhesion molecule, was localized to the ganglion and hair cell precursors. Ganglionic precursors, migrating from the otocyst, only weakly expressed neural cell adhesion molecule. Epithelial hair cell precursors, remaining in the otocyst, expressed neural cell adhesion molecule, but not L1-cell adhesion molecule. Post-migratory ganglion cell processes expressed both molecules in all stages. The cell adhesion molecules were most heavily expressed by axons penetrating the otic epithelium and accumulated in large amounts in the basal lamina. In the basilar papilla (cochlea), cell adhesion molecule expression followed the innervation gradient. Neural cell adhesion molecule and L1 were heavily concentrated on axonal endings peripherally and centrally. In the rhombencephalon, primitive epithelial cells expressed neural cell adhesion molecule, but not L1-cell adhesion molecule, except in the floorplate. The neuroblasts and their axons expressed L1-cell adhesion molecule, but not neural cell adhesion molecule, when they began to migrate and form the dorsal commissure. There was a stage-dependent, differential distribution of the cell adhesion molecules in the floorplate. Commissural axons expressed both cell adhesion molecules, but their polysialic acid disappeared within the floorplate at later stages. In conclusion, the cell adhesion molecules are expressed by the same cells at different times and places during their development. They are positioned to play different roles in migration, target penetration, and synapse formation by sensory neurons. A multiphasic model provides a morphological basis for experimental analyses of the molecules critical for the changing roles of the microenvironment in neuronal specification.     

Levels of circulating intercellular adhesion molecule 1 and E-selectin in serum and synovial fluid of juvenile chronic arthritis patients

OBJECTIVE: To measure the levels of two adhesion molecules (AM), soluble intercellular adhesion molecule 1 (sICAM-1) and soluble E-selectin (sE-selectin), in serum and synovial fluid (SF) of patients with juvenile chronic arthritis (JCA). METHODS: Both soluble AM levels were tested, in serum and synovial fluid (SF) samples, with an enzyme-linked immunosorbent assay (ELISA) method. RESULTS: Serum levels of sICAM-1 and sE-selectin in JCA patients were not significantly different from those of a control group. Synovial fluid levels of sICAM-1, but not of sE-selectin, assayed significantly higher (p < 0.05) in JCA patients than in controls. Moreover SF levels of both molecules correlated negatively with disease duration, being higher in the earliest phases. No significant correlations were found between JCA sICAM-1 and sE-selectin levels and leukocyte count or ESR. CONCLUSIONS: These observations may signify a more important role of ICAM-1 than E-selectin in the migration of inflammatory cells into JCA SF. The negative correlation of both AMSF levels in JCA patients with disease duration could reflect a higher expression of ICAM-1 and E-selectin during the earliest phases of the disease.     

Interference of nonsteroidal antiinflammatory drugs with very late activation antigen 4/vascular cells adhesion molecule 1-mediated lymphocyte-endothelial cell adhesion

OBJECTIVE: To study the effect of nonsteroidal antiinflammatory drugs (NSAIDs) on the adhesion of peripheral blood lymphocytes (PBL) to activated human umbilical vein endothelial cells (HUVEC) under conditions that resemble blood flow. METHODS: Assays of adhesion of PBL to HUVEC or recombinant vascular cell adhesion molecule 1 (rVCAM-1), intercellular adhesion molecule 1 (ICAM-1), and E-selectin were performed under continuous rotation at 37 degrees C. The phenotype of PBL subpopulations attached was characterized by flow cytometry. Lymphocytes were pretreated with different doses (5-100 microg/ml) of aceclofenac, diclofenac, indomethacin, or piroxicam or with inhibitory monoclonal antibodies (MAb) prior to the adhesion assays. The effect of NSAIDs on lymphocyte adhesion molecules was assessed by flow cytometry. To determine whether NSAIDs interfere with the affinity state of very late activation antigen 4 (VLA-4) integrin, we studied the effect of these drugs on the appearance of a beta1 activation-dependent epitope recognized by the HUTS21 MAb both on human T lymphoblasts and on synovial fluid lymphocytes (SFL). RESULTS: In the flow-resembling model, PBL-HUVEC adhesion was mainly mediated by the VLA-4/ VCAM-1 adhesion pathway. The major PBL subset attached was the CD3+, CD45RO+ memory T cell, with CD49d(high) expression. Aceclofenac, diclofenac, and indomethacin, but not piroxicam, were able to inhibit PBL adhesion to HUVEC or rVCAM-1. However, the quantitative expression of VLA-4 was not affected by treatment of PBL with any of the NSAIDs studied. On T lymphoblasts and SFL, mostly CD45RO+ cells, the expression of the beta1 activation-dependent epitope detected by HUTS21 MAb was significantly decreased by aceclofenac, diclofenac, and indomethacin. CONCLUSION: Some NSAIDs are able to inhibit the adhesion of PBL to HUVEC under conditions that resemble blood flow by interfering with the conformational change in VLA-4 that increases its affinity for VCAM-1.