Mechanisms of eosinophil adherence to cultured vascular endothelial cells Eosinophils bind to the cytokine induced ligand vascular cell adhesion molecule 1 via the very late activation antigen 4 integrin receptor

Mechanisms of eosinophil adherence to

cultured vascular endothelial cells. Eosinophils

bind to the cytokine-induced ligand vascular

cell adhesion molecule-1 via the very late

activation antigen-4 integrin receptor.

A Dobrina, … , J M Harlan, P Patriarca

J Clin Invest.

1991;

88(1)

:20-26.

https://doi.org/10.1172/JCI115278

.

We have examined the mechanisms involved in the adherence of normal peripheral blood

eosinophils to cultured human umbilical vein endothelial cells (HEC) under three

conditions: (a) adherence in the absence of treatment of HEC or eosinophils with activating

agents (basal adherence); (b) adherence induced by stimulation of eosinophils with phorbol

ester (eosinophil-dependent adherence); and (c) adherence induced by pretreatment of

HEC with LPS, tumor necrosis factor (TNF), or IL-1 (endothelial-dependent adherence). A

mechanism was identified that was equally active in basal, eosinophil-dependent, and

endothelial-dependent adherence. This mechanism was optimally active in the presence of

both Ca++ and Mg++, and reduced in the presence of Ca++ only or Mg++ only. Furthermore,

like the other mechanisms of eosinophil adherence, it was active at 37 degrees C but not at

4 degrees C. A second mechanism of adherence was involved in eosinophil- and in

endothelial-dependent adherence. This mechanism was dependent on the CD11/CD18

adhesion complex of eosinophils (i.e., inhibited by anti-CD18 MAb) and it was active in the

presence of Ca++ and Mg++ or Mg++ only, but not Ca++ only. The third mechanism of

adherence was specific for endothelial-dependent adherence. It involved the endothelial

ligand vascular cell adhesion molecule-1 (VCAM-1) and the eosinophil receptor very late

activation antigen-4 (VLA-4, CD49d/CD29, i.e., inhibited by anti-VCAM-1 MAb or anti-VLA-4

MAb). This mechanism was active in […]

Research Article

Mechanisms of Eosinophil Adherence to Cultured Vascular Endothelial Cells

Eosinophils Bind to the Cytokine-induced Endothelial Ligand Vascular Cell Adhesion

Molecule-1

via the

Very

Late Activation

_{Antigen-4 Integrin}

Receptor

A.Dobrina,**R.Menegazzi,* T. M.Carlos,$E. Nardon,* R. Cramer,* T. Zacchi,* J. M.Harlan,$and P.Patriarca*

*InstituteofGeneral Pathology, UniversityofTrieste, Trieste, Italy 34127;*Carloand Dirce CallerioFoundation, Trieste, Italy

34127;

and DepartmentofMedicine, Universityof Washington,Seattle, Washington98195

Abstract Introduction

We haveexaminedthe mechanisms involved in theadherence ofnormalperipheralbloodeosinophils to culturedhuman um-bilical veinendothelial cells(HEC)under threeconditions: (a) adherencein the absence of treatment of HEC or eosinophils with activatingagents (basal adherence); (b) adherence induced bystimulation ofeosinophils with phorbol ester (eosinophil-de-pendentadherence);and(c)adherenceinduced by pretreatment ofHEC with LPS,tumornecrosis factor (TNF), orIL-I (endo-thelial-dependentadherence). A mechanism wasidentifiedthat was equallyactivein basal, eosinophil-dependent, and endothe-lial-dependentadherence. This mechanism was optimally ac-tiveinthepresenceofboth

Ca"+

and

Mg",

and reduced in the presence of

Ca"+

only or Mg"+ only. Furthermore, like the other mechanisms of eosinophil adherence, it was active at 370Cbut not at4VC.Asecond mechanismofadherence was involved in eosinophil- and in endothelial-dependent adher-ence.This mechanismwas dependent on theCD11/CD18 ad-hesion complex of eosinophils (i.e., inhibited by anti-CD18 MAb)and itwasactivein the presenceof

Ca"+

and

Mg"+

or Mg++only, but notCa++only. Thethirdmechanism of adher-ence was specific for endothelial-dependent adherence. It in-volved theendothelial ligandvascular celladhesion

molecule-i

(VCAM-1)and theeosinophilreceptor very late activation an-tigen4 (VLA4,

CD49d/CD29,

i.e., inhibited by anti-VCAM-i MAboranti-VLA4MAb). This mechanismwasactiveinthe presenceofCa++andMg++but notofCa++only or Mg++only, andwas notup-ordownregulatedwheneosinophilswere stimu-latedwith phorbolester. Incontrast,theendothelial leukocyte adhesion molecule-i

(ELAM-I),

that binds neutrophils and monocytes, was notinvolvedineosinophiladherence toLPS-, TNF-, or

IL-i-stimulated

HEC (i.e., not inhibited by anti-ELAM-1MAb).Weconcludethateosinophils, likemonocytes and

lymphocytes,

bindtothe

cytokine-induced

endothelial li-gand VCAM-1 via the

integrin

receptorVLA4. (J. Clin. In-vest. 1991.88:20-26.) Keywords: vascular celladhesion

mole-cule-i

* verylate activation

antigen4

*

eosinophil

-endothelium

* adherence

AddresscorrespondencetoDr.JohnM.Harlan,Division of Hematol-ogy, Room 10,University of Washington, Seattle,WA 98195.

Receivedfor publication 9May 1990 and in revised form17 De-cember 1990.

Matureeosinophilsare locatedpredominantly in the extravas-cular space, even inphysiologicconditions(1-3), theskin, gas-trointestinal tract, and mucosa of the bronchi being the most heavily infiltrated tissues(3, 4). Increased levelsofcirculating eosinophils and localaccumulationofeosinophilsatsites of acute orchronic inflammation have long been associated with allergic reactions, parasitic infestations, and otheracuteand chronicinflammatory diseases suchasthyroiditis,somestages oftuberculosis, mycotic infections, recurrent staphylococcal infection, Hodgkin's disease, and other neoplastic processes (recently reviewedby Nutman et al. [5,6] andbySpry [7]).A series of studieshasemphasizedtheroleplayedin eosinophil-mediated inflammatory reactions by powerful toxic mecha-nisms ofeosinophils, such astheeosinophilic peroxidase-hy-drogen peroxide-halidesystem(8-10)andthe release of eosin-ophil majorbasicprotein (11, 12).In contrast, little is known about themechanisms involvedin thelocalization of eosino-philsintissues,both inphysiologicandpathologic conditions. Eosinophiladherence toendothelial cells,akeyevent in leuko-cyte emigration into tissues (13), was recently examined by Lamas etal.(14) andbyKimaniet al.(15). The authors have documented at least three mechanisms of

eosinophil

adher-ence to cultured human umbilical vein endothelial cells

(HEC)':

(a)adherence of unstimulatedeosinophilstoresting HEC (basal adherence); (b) adherence of

eosinophils

stimu-lated bychemotacticfactors, such asplatelet

activating

factor (PAF) (16,17),orbyphorbolester(PMA)to

resting

HEC

(eo-sinophil-dependent

adherence); and

(c)

adherence ofresting

eosinophils

to HEC stimulated by

cytokines,

such as tumor necrosis factor (TNF)and IL- 1, orLPS(endothelial-dependent adherence). The

eosinophil-dependent

adherence

mechanism,

similartothatof stimulated

neutrophils (18, 19),

involves acti-vation ofthe leukocyte adhesion complex CD 1/CD 18, as judged by inhibition by CD18 MAb

(14, 15).

Furthermore,

as already reported forneutrophils (18, 19), the endothelial-de-pendent adherence of

eosinophils

is

only

partially

inhibited

by

CD 18MAb,suggesting involvement ofa

second,

CD1 8-inde-pendentadherencemechanism

(15).

Finally,

basaladherence of

eosinophils

is

independent

of

CD1 1/CD

18

complex,

since it isnotaffected by CDl1/CD18 MAb

(14).

In thispaperwe presentevidence that several featuresof

1.Abbreviations usedin this paper: ELAM-1, endothelial leukocyte adhesion molecule-i; HEC, human umbilical vein endothelialcells; ICAM-1,intercellular adhesion molecule-1;LAD,leukocyteadhesion deficiency;PAF,plateletactivatingfactor (l-alkyl-2-acetyl-sn-glycero-3-phosphorylcholine); TNF,tumornecrosisfactor;VCAM- 1, vascular celladhesion molecule- 1; VLA-4,verylateactivationantigen-4.

20 Dobrina, Menegazzi,Carlos,Nardon, Cramer, Zacchi, Harlan,andPatriarca J.Clin. Invest.

©TheAmericanSocietyforClinicalInvestigation,Inc.

eosinophil

adherence are clearly distinct from those of

neu-trophils,

most notably: (a) basal adherence ofeosinophils is

higher

than that of

neutrophils,

is temperature-dependent, and

requires

Ca"+

and

Mg"+;

(b)

the CDlI/CDI8-independent component of

eosinophil

adherence toLPS-, TNF-, or IL-1-treated HEC involves the very late activationantigen-4

(VLA-4) (CD49d/CD29) integrin

receptor(20)ontheeosinophil and vascular cell adhesion molecule-1 _{(VCAM-1 [211;} inducible celladhesionmolecule-l 10

(INCAM)-

110

[22]),

anendothelial cell molecule that promotes adherence of peripheral blood

lymphocytes (22, 23)

andmonocytes(22, 24), butnot

neutro-phils;

and

(c)

the CDl

1/CD

18-independent eosinophil

adher-ence to

cytokine-

orLPS-stimulated HEC isnotdownregulated after direct activation of

eosinophils,

as occurs with

neutro-phils (25).

Methods

Cell culture. HEC were prepared by collagenase treatment of the vesselsasdescribed elsewhere(26) andmaintained inendotoxin-free

RPMI 1640 medium(Gibco-BRL,Uxbridge,UK) supplemented with 10% newborn bovineserum(FlowLaboratories Ltd., Irvine, Scotland, UK)and 10% fetal calfserum (Biochrom KG, Berlin, FRG) (NBS-FCS). Passaged HECweremaintained in RPMI 1640 supplemented with 20% NBS-FCS containing heparin (90

jg/ml;

Sigma Chemical Co., St. Louis, MO)andendothelial cellgrowthfactor (50

jug/ml)

as described_byThornton etal.(27). Endothelial cell growth factorwas

preparedfrom bovinehypothalamus accordingtothemethod of

Ma-ciag

etal.(28).

Neutrophilisolation. Peripheralbloodwasobtained by

venipunc-turefrom_healthydonors. Thebloodwascollected insyringes contain-ing15%(vol:vol)sterile ACD solution(acid-citrate-dextrose; 100 mM disodiumcitrate, 128 mMglucose; pH 5.0),and theneutrophilswere

isolatedby Ficoll-Hypaque (PharmaciaFineChemicals,Uppsala, Swe-den)

gradient

centrifugation, 3% dextran sedimentation, and hypo-tonic saline

lysis

ofcontaminatingredcells(29).This procedure

re-sulted inapreparation>95%neutrophils,which exceeded 95%

viabil-ity

bytrypan bluedyeexclusion. Isolatedperipheralbloodneutrophils

werewashedwith PBS(Gibco-BRL)and suspendedat afinal concen-tration of7 X

10'

cells/mlin PBScontaining5 mMglucose, 1 mM

CaC12,

and 1 mMMgCl2,unlessotherwise stated.

Eosinophil

isolation. Eosinophilswere isolated accordingto the methodofR. Cramer(manuscriptinpreparation).Peripheral blood fromhealthydonorscontaining<2XI0O eosinophils/mlwascollected in ACD solution. After erythrocyte sedimentation in 4.5% dextran

(Pharmacia),

thewhite cell-rich plasmawaswashedoncewith PBS

containing

13 mM sodium citrate and 0.5% BSA(MilesLaboratories Inc., Goodwood,SouthAfrica).The cellpelletwasthen resuspended in isotonic Percoll(Pharmacia) containing 13 mM sodiumcitrate and 0.5% BSA (pH 7.4). The density of the Percoll suspension was

1.0853±0.0002g/ml,asmeasuredat20°CbyaDMA 45density meter (A. Paar, Graz, Austria),and Percollosmotic valuewas290±2 mosM, asmeasuredby cryoscopicosmometer(Osmomat030; Gonotec, Ber-lin, FRG).Thecellsuspensionwaslayeredon aPercoll cushion with a

density higher

than1.1g/mlandwascentrifugedat1,000 g for 20min

at

20°C.

The cellringformedattheinterfacewascollectedandthe red cells presentwereremovedbyhypotoniclysisat4°C.Isolated

periph-eralbloodeosinophilswerewashedwith PBSandsuspended at a final concentration of 3X I05cells/mlin PBScontaining5mMglucose, 1 mMCaC12, and 1 mM

_MgCi2,

unlessotherwise stated. The resulting cellsuspensioncontained between 85% and 98% eosinophils and the

yield

was >50%of theeosinophilspresent in the starting bloodsamiple. Cellviabilitywas >98%,asdetermined bythetrypan blue dye exclu-siontest.

Adherenceassay. Firsttothird passage HEC were harvested with 0.05 trypsin and 0.02% EDTA in balanced salt solution(Gibco-BRL).

Thecells were then platedin6.4-mm diameter wells(CostarCluster, Cambridge, MA)at 1.5 X I04 cells/ml in RPMI 1640 supplemented with 20% NBS-FCS. Visually confluent monolayers were formed after overnight incubation. Cellswerepretreated with reagents for 4h,and the monolayerswerethen washed withathree-well volume ofPBS. For the adherence assayat40C, after the incubationat370C, HECwere

incubated for 30 min inacoldroom(2-40C) and washed witha three-wellvolume exchange of ice-cold PBS. Eosinophilsorneutrophils were then added(70 jl/well). Leukocytes for the 40C adherence assaywere

incubated for 30 minonice before additiontothe wells. Leukocytes and HECwerethen incubated for 30 min at 370Corin the coldroom.

Afterincubation the monolayerswerewashedwithatwo-well volume exchange of PBSorice-cold PBSto removenonadherentleukocytes.A

colorimetric assaywasthenappliedtodetect theeosinophils or

neutro-philsadhering to the monolayers, using tetramethylbenzidine (TMB)

asperoxidase substrate (30). The substrate solution consisted of 2 mM

TMB(Sigma) in 0.1Msodiumacetatebuffer, (pH 4.2) containing 0.1 % (wt:vol) cetitrimethylammonium bromide (Sigma)asperoxidase solu-bilizing agent.Inthis assay eosinophils showa > 10 timeshigher perox-idase activity than neutrophilson aper cell basis, thus rulingoutany significantinterference by the few contaminating neutrophils. When neutrophil adherencewasassayed, the selective eosinophil peroxidase inhibitor 3 amino- 1,2,4 triazol (I mM; Schuchardt, Munich, FRG) (31) was also addedtothe substrate solutiontoabolish the interference byeosinophils.The substrate solutionwasthen addedtothe mono-layers(75

Al/well)

followed,after 2min,by 0.7 mM hydrogen peroxide (75

Al/well).

After 2 min of incubationat roomtemperature, the perox-idasereaction was stopped bytheaddition of 50

_Al

of4Naceticacid, containing 10mMsodium azide. The absorbancewasthen determined

at 620 nm using a Titertek Multiskan (Flow Laboratories, Inc., McLean, VA). Percent leukocyte adherencewascalculated usinga cali-brationcurve.Thiswasobtained byperforming the peroxidase

reac-tion in wellscontainingknown amountsofeosinophilsorneutrophils. Immunofluorescenceflowcytometry. Immunofluorescence flow cy-tometrywasperformedaspreviouslydescribed(25). Eosinophil and neutrophilpreparationsused inthese assayswere>98% pure. Periph-eral blood mononuclear cellswereobtained by Ficoll-Hypaque separa-tion. The cellpopulationscomprisedanaverage of 70% lymphocytes and30% monocytes (29). Leukocytes (5XI05)weresuspended in 50

Al

PBS-0. 1% BSAcontaining MAb P4C2orMAb HP2/1 (1:50 dilution ofhybridomasupernatant media), MAb 60.3orMAb4B9 (20 jg/ml). The cells were incubated for 20 min at 4°C, washed free of unbound antibody, and incubated againat4°C for 20 min with a 1:50 dilution of FITC goat anti-mouse IgG (Sigma). After washing, the cells were sus-pendedin 500 jl of 1%paraformaldehydein PBS and stored at 4°C. Themeanfluorescenceof each cell population was quantified by a flow cytometer (EPICS-C; Coulter Corp.; Hialeah, FL) with quantitative determination of peak fluorescence intensity.

Monoclonal antibodies. MAb 60.3 is ofthe IgG2a subclass and rec-ognizes the CD1 8 subunit (common beta-chain) of the CDI l/CD18 antigencomplex (32). MAb 4B9 isamurineIgGl and recognizes cells transfectedwith VCAM-Ibutnotcells that aretransfected with inter-cellularadhesion molecule-I(ICAM- 1)orendothelial leukocyte adhe-sionmolecule-l (ELAM-1) cDNA (23). MAb BBl l is a murine IgG2b thatrecognizesafunctional epitope on ELAM-l (33), and was a gift of

Drs.Christopher Benjamin and Roy Lobb, Biogen Inc., Cambridge,

MA.MAbP4C2is a murine IgG3 that recognizes an epitope on CD49d

and(34) andwas agift ofDr.Elizabeth Wayner, Cytel Corp., La Jolla, CA(34).Monoclonalantibody HP2/1 is a murine IgG1 antibody that bindstoanepitopeonCD49d and was a gift of Dr. F. Sanchez-Madrid (35). MAbP4COisamurine IgGIthatrecognizesafunctional epitope

onCD29(34), andwasagift ofDr.Elizabeth Wayner.

Reagents. Escherichia coli 055:B5 LPS, extracted by phenol/water, wasobtained from Sigma Chemical Company. The LPS preparation wassuspended in PBS at a concentration of 1.5 mg/ml, dispersed by sonicationfor 5 minat4°C, and stored in aliquots at -35°C until used. Phorbol-12-myristate-13 acetate (PMA; Sigma) was dissolved at 1 mg/

solution at -350C. Purified recombinant human IL-1 alpha (1,000 U/ml) (IL-1) was purchased from Boehringer Mannheim GmbH, Mannheim, FRG. Human recombinant TNF-alpha was a generous gift of Prof. R. Lax, Bissendorf Biochemicals GmbH, Hannover, FRG.

Statistics. Significance was determined by paired, two-tailed t test.

Results

Influence ofanti-CD18, anti-VCAM-J, and anti-ELAM-J MAb oneosinophil adherence to HEC. Adherence assays were per-formed in the presence or absence of blocking MAbs to the inducedendothelial ligands ELAM- 1, VCAM- 1, or the leuko-cyte receptors CDl 1/CD18 and VLA-4. As reported previously (15), eosinophils adhered spontaneously to untreated HEC (Ta-ble Ia). Basal adherence of eosinophils to HEC was not signifi-cantly affected by the anti-VCAM-1 MAb 4B9, the anti-ELAM-1MAb BB11, orthe CD 18 MAb 60.3. Pretreatment of HECwithLPS caused a significant increase in eosinophil adher-enceabove control values, which was significantly reduced by MAb 4B9 or MAb 60.3. Moreover, adherence to LPS-stimu-lated HEC was reduced to thatofunstimulated HEC when MAb 60.3 was used in combination with MAb 4B9. In

con-TableI.

Effect

ofanti-VCAM-J,anti-CD18, andanti-ELAM-J MAbonEosinophiland Neutrophil Adherenceto HEC

Percentadherence

PMA-stimulated Unstimulated LPS-treated HEC eosinophils

(a) Eosinophils (6)

Control 17.0±2.2 29.4±2.0 58.7±3.8

MAb 4B9 15.6±2.6 17.1±2.3* 53.7±2.3

MAb60.3 13.6±1.6 19.7±0.4* 20.8±5.2*

MAb 4B9+MAb60.3 13.8±2.3 11.0±1.5*$ 20.3±4.8*

MAbBB11 17.3±2.2 33.1±4.4 64.8±3.1

MAb BBl 1 +MAb60.3 14.7±3.8 24.3±1.4 25.5±3.0* (b)Neutrophils(4)

Control 6.6±0.4 24.0±4.7 49.1±6.3

MAb 4B9 5.8±0.4 24.0±5.1 45.0±8.0

MAb60.3 4.0±0.3 14.1±1.8* 3.5±0.8*

MAb 4B9+MAb 60.3 ND 14.3±1.0* ND

MAbBBl1 4.5±0.2 13.2±3.4* 48.5±3.5

MAb BBl 1 +MAb60.3 ND 3.3±0.8* ND

HECmonolayers were pretreated for4hwith control medium or mediumcontainingE. coli LPS (100 ng/ml). Before the assay, eosin-ophilsorneutrophilssuspended in PBScontaining5mMglucose, 1

mMCaCI,and 1mMMgC1wereincubated for 15minatroom tem-perature in the presenceorabsence of MAb 60.3(20,ug/ml)andHEC monolayers were incubated in the presenceorabsenceof the anti-VCAM-I MAb 4B9(20

tg/ml)

or theanti-ELAM-1 MAb BBI 1 (20

gg/ml).

Leukocytes were then added to HEC monolayers followed by PBS(medium)or PMA(100 ng/ml,final). Percentage leukocyte adherencewasdetermined aftera30-min incubationat37°C. Values represent the means+SEof (n) experiments with four replicate wells in eachexperiment. ND, not done.*P <0.005(paired t test) com-paredtoadherence in the absence of MAb (controls). The other re-sultswerenotsignificantlydifferent from control values.*_{P <0.005}

for adherence of MAb 60.3- and MAb 4B9-treated eosinophilsvs.

MAb60.3-treatedeosinophils to LPS-pretreated HEC and for

adher-enceofMAb60.3- andMAb BB11-treatedneutrophils vs. MAb 60.3-treatedneutrophilstoLPS-pretreated HEC.

trast, MAb BB11 had no effect on eosinophil adherence to unstimulatedorLPS-stimulated HEC. Similar resultswere ob-tainedwhen HECwerepretreatedwith IL- I(10 U/ml)orwith TNF(500U/ml). Treatmentof HEC with IL-I increased eo-sinophil adherence from basal values of 18.5%±2.5 to 37.5%±2.8. MAb4B9 and MAb 60.3 reduced IL-l-stimulated adherence to 27.5%±1.8 and 25%±2.0, respectively, andto 15.5%±1.5 when the two MAbs were used in combination. Eosinophil adherence to TNF-treated HEC was 41.0%±3.8, butitwasreducedto31.0%±2.6and31.1%±2.8byMAb4B9 and MAb 60.3, respectively, andto13.0%±0.9 by the combina-tionofMAb4B9and MAb60.3.In contrast,noinhibitionof eosinophil adherence to IL-1- or TNF-treated HECwas ob-tained withMAb BB11 (means±SDof fourreplicatewells in oneexperiment). InhibitioncausedbyMAb 4B9wasdueto an effect on the endothelial cell rather than on the

eosinophil,

since preincubation of

eosinophils

with MAb4B9followed by washing did not inhibit subsequent eosinophil adherence to LPS-pretreatedHEC (notshown). Finally,asshownin TableI a, PMA-stimulated adherence ofeosinophilswas completely inhibitedbyMAb 60.3 butwas not

significantly

inhibitedby MAb 4B9 or by MAb BBI1.

The behaviorso far described for eosinophils differed in severalrespectsfrom that of

neutrophils.

Unstimulated adher-ence of

neutrophils

(Table I

b)

was low(6.6%±0.4) as com-paredtounstimulated adherence of

eosinophils

(17.6%±2.6).

Asreportedpreviously (18, 19), neutrophiladherencewas

sig-nificantly increasedbystimulation withPMAorby pretreat-ment of HEC with LPS. No influence ofMAb 4B9 was ob-servedon neutrophil adherencetoLPS-treated HEC.In

con-trast,

neutrophil

adherencetoLPS-treatedHECwasinhibited by45%byMAb BB11 andby42%

by

MAb60.3,

respectively,

andwas inhibited by 87% when the two MAb were used in combination. PMA-stimulated adherence ofneutrophilswas

completely inhibited by MAb 60.3,butwasnotinfluenced

by

MAb 4B9orMAb BB1 1.

Influence of

anti-VLA-4 MAbon

eosinophil

adherence to HEC. Previous studieshaveshown thatantibodiestothe VLA-4integrinreceptorblocklymphocyte adherenceto VCAM-1 onactivated endothelium (34). Since

eosinophil

adherence in ourassayswasinhibited by anti-VCAM-I MAb

4B9,

the

possi-bilityarosethatthe VLA-4moleculewasalsoinvolvedin

eo-sinophil

adherence to activated HEC.

Hence,

a search for VLA-4 oneosinophilswas

carried

out.TableII compares the

binding

oftwoVLA-4

alpha-chain-specific

MAbs

(CD49d)

to

Table II. ExpressionofAdhesionProteinson

Peripheral

BloodLeukocytes

Eosinophils Neutrophils Mononuclear cells

MAb

P4C2(CD49d) 26.7±1.4(3) 0(2) 15.3±2.2(3) HP2/1 (CD49d) 25.3±1.8(3) 0(2) 15.3±3.3(3) 60.3(CD18) 122.3±2.7(3) 155(2) 100.7±3.2(3) 4B9(VCAM-1) 0(3) 0₍₂₎ 0(3)

Binding of MAbswasassayed by flow cytometryasdescribed in Methods.Valuesfornet meanpeakfluorescencewerecalculatedby subtractingvaluesobtained with

FHTC-conjugated

secondantibody alone and represent the mean±SE of(n)experiments.

eosinophils, neutrophils, orPBMCs. The CD49dMAbs P4C2 and HP2/1 bound to eosinophilsaswell as to mononuclear cells, but nottoneutrophils. As expected, all cell types bound theanti-CD18 MAb60.3 but none bound theanti-VCAM-1 MAb4B9.Thereafter,studies on the effect of anti-VLA-4 anti-bodiesoneosinophiladherence were carried out. As shownin Table III, the CD49d MAb P4C2 significantly inhibited eosino-philadherence to LPS-pretreated HEC, but not to untreated HEC. Theanti-VCAM-1 MAb 4B9 also inhibitedeosinophil adherence to LPS-treated HEC. However, no additive effect was obtained when MAb P4C2 and MAb 4B9 were used in combination, thussuggestingthat the two antibodies exerted theireffecton thesameadherence mechanism.Aninhibitory effectoneosinophil adherencewasalsoobtainedbyusingthe CD49dMAb HP2/1 and theCD29MAbP4C10. Monoclonal antibody HP2/1 inhibited eosinophiladherence to LPS-treated HECby68.2%inoneexperimentand MAb

P4ClO

by 45.9% and34.8% intwoexperiments(results not shown).Finally,it is importantto note that adherence of PMA-stimulated eosino-philstountreated HEC was notinhibitedatallby the CD49d MAb P4C2 (Table III).

Influence oftemperature on eosinophil adherence to HEC. Adherence assays were performed at 37°C or at 4°C (Fig. 1). Eosinophiladherence to HEC was stimulated either by PMA orby preincubating HECmonolayerswithLPS. Valuesof eo-sinophil adherenceat37°C were 16.5%±1.3, 30.8%±2.5, and 64.6%±4.2 for unstimulated and for LPS-and PMA-stimu-lated adherence,respectively.However, both unstimulated and stimulated adherence of eosinophilswas almost abolished at 4°C. Again neutrophilsbehavedquite differentlyfrom eosino-phils. Theiradherence tounstimulatedHEC was low at 37°C andat4°C.PMAmarkedlyincreased neutrophiladherence to unstimulated HEC at 37°C, but not at 4°C. As described

TableIII.

Effect

ofanti- VLA-4andanti-VCAM-IMAbs onEosinophilAdherencetoHEC

Percent adherence

PMA-LPS-stimulated stimulated Unstimulated HEC eosinophils

MAb

control 18.6±1.2 29.6±1.4 63.3±6.7

MAbP4C2 (CD49d) 18.8±1.6 15.7±2.3* 63.5±6.3 MAb 4B9

(anti-VCAM-l) 16.6±2.0

19.1±1.8*

ND MAbP4C2+MAb 4B9 17.8±2.0 19.5±2.4* ND

HEC monolayers were pretreated for 4 h with control medium or medium containing E. coli LPS (100ng/ml).Before the assay, eosin-ophilssuspended in PBS containing 5mMglucose, 1 mM CaCl2,and

1mMMgCl2,wereincubated for 15 min at room temperature in the presence or absence of MAb P4C2 (1:50 dilution of hybridoma su-pernatant medium)and HEC monolayers were incubated in the presence or absence of MAb 4B9(10ug/ml): leukocytes were then addedtoHECmonolayers followed by PBS (medium) or PMA(100 ng/ml, final). Percentage leukocyte adherence was determined after a 30-minincubationat37°C. Values represent the means±SE of six experiments with three replicate wells in each experiment. ND, not done.*P <0.005; *P <0.025(paired t test) compared to adherence in the absence of MAb (controls). The other results were not signifi-cantly different from control values.

80 eosinophils neutrophils Figure1.Effect of

tem-peratureoneosinophil

0 - andneutrophil

adher-ence toHEC. HEC

.40 monolayers were

pre-treatedat370C for4h

20 h with medium alone

(control)

orwith

me-PA dium

containing

E.coli

- LPS PMA LPS PMA

LPS(100ng/ml). Eo-sinophilsorneutrophils suspended in PBScontaining5mMglucose, 1 mMCaCl2,and 1 mMMgCl2wereaddedtoHECmonolayerswithmedium

(control)

orwith PMA(100 ng/ml).For the adherence assayat4°C, leukocytes wereincubatedonice and HECwereincubated inacoldroom

(2-4°C) for 30 min before the assay.Percentageleukocyteadherencewas

determinedaftera30-min incubationat37°Corat40C. Valuesare

means±SE of sixexperiments,withfourreplicatewells in each ex-periment. *P<0.001 forleukocyteadherenceat370Cvs.adherence

at4°C. **P<0.002 forneutrophiladherencetoLPS-pretreatedHEC vs.untreated HECat4°C. o, 37°C; , 40C.

previously (19), stimulation ofHECby LPSpretreatment

re-sulted in an increased adherence both at 370C andat 4°C. Neutrophil adherence to LPS-treated HEC at4°Cwas

com-pletelyinhibited byMAb BB 11(not

shown),

suggesting

involve-mentofELAM- 1 inthisadherence.

Divalent

cation requirements

for eosinophil

adherenceto HEC. The

experiments

whose resultsareshown inTableIand Fig. 1 wereperformed in PBS medium

containing

both

Ca" (1

mM) and Mg++ (1 mM). Divalent cation

requirements

forthe variousmechanisms ofadherencewerefurther

investigated

in greaterdetailat370C

(Fig.

2). Detachment ofHEC occurred in theabsence ofbothcalciumand

magnesium,

orin thepresence ofEDTA (1 mM),

thereby

preventing

us from

determining

eosinophil

adherenceto HEC under these conditions. When the assaywas

performed

in the presenceof

Ca++ only (1 mM),

a

80

-

60-c W 40

2o

20

-Ca

I

I

Lar

Mg

[L4r

A

_I

- LPS PMA - LPS PMA

Caa and Ms

11

- LPS PMA

Figure 2. Effect of Ca++ and Mg++oneosinophiladherencetoHEC. HEC monolayerswerepretreatedat37°C for4hwith medium alone

ormediumcontainingE.coliLPS(100 ng/ml).Eosinophilswere

suspended in PBScontaining5mM glucose with 1 mMCaCl2only,

orwith 1 mMMgCl2only,orwith both cations.Eosinophilswere

thenaddedtoHECmonolayers followed by PBS(control)or PMA

(100ng/ml). Percentage eosinophil adherencewasdetermined after

a30-min incubationat37°C.Valuesaremeans±SE of five

experi-mentswith fourreplicate wells in each experiment. Statistical

signifi-cance(bypairedttest):P<0.05for adherence of MAb 60.3-treated eosinophilsvs.untreatedeosinophilstoLPS-treatedHECin the

pres-enceof

CaCI2

andMgCl2.P<0.02 for unstimulated adherence in the presence ofCaCl2onlyorMgCl2only vs. adherence in the presence of bothcations,andfor adherenceof MAb 60.3-treatedeosinophils

toLPS-treated HECvs.untreatedHEC in the presence ofCaCl2and

MgCI2.

o,medium;

.,

MAb60.3.

definite proportion (8.9%±1.9) of eosinophils adhered to un-treated HEC. Ofnote, unstimulated eosinophil adherence at 370C inthepresenceofCa"+only (as well as in the presence of Mg++only) wasstillstatisticallygreater than adherence at

4VC

inthe presenceofCa"+only, orMg"+ only, or both cations (P <0.05, fourexperiments). At 370C and in the presence ofCa"+ only, however, therewas noincrease of eosinophil adherence abovecontrollevels with LPS-pretreated HEC or in the pres-ence of PMA. Moreover, adherpres-ence in the prespres-ence of Ca"+ onlywasunaffected by MAb60.3.

Inthepresenceof

Mg"+

only(1mM),unstimulated eosino-phil adherencewassimilartothat observed in the presence of

Ca`+

only.Eosinophiladherence to LPS-pretreated HEC, how-ever, wassignificantly increasedas comparedto adherence to untreated HEC and was completely inhibited by MAb 60.3. Furthermore, in the presence ofMg"+ only,PMA stimulated eosinophil adherence and this was again inhibited by MAb 60.3. These results indicatethat,inthe presence ofMg`+only, eosinophil adherencedue toLPS pretreatment of HEC or to PMA is accounted for, almost completely, by

CDl

1/CD18. Raising

Mg`+

concentrationto 2 mM did not lead to apprecia-ble changes in the results obtained with 1 mM Mg`+ (not shown).

Inthe presenceofbothCa++ and

Mg++,

unstimulated eo-sinophil adherencewasstatisticallygreater thanin the presence of Ca++only or Mg++ only(Fig.2). Adherence toLPS-treated HECwasfurther increasedabove levels of unstimulated adher-encein thepresenceof

Ca"

and

Mg",

orlevels of adherence to LPS-treated HEC in thepresence ofMg++ only, and was onlypartially inhibited byMAb60.3(see alsoTableI).In con-trast, PMA-stimulated adherence of eosinophils in the pres-enceof Ca++ and Mg++wassimilartothatobserved withMg++ only, andwascompletely inhibited by the CD18mAb 60.3.

Influence of

PMAon CDI

I/CDJ8-dependent

and -indepen-dent adherence mechanisms.Ina

previous

study,wefound that neutrophil

CDl

1/CD18-independent adherence is downregu-lated whenneutrophilsareactivated withPMA(25).Todefine theeffect of phorbolester onCDl 1/CD

18-independent

adher-ence of

eosinophils, eosinophil

adherence to HEC or LPS-treated HECwasstimulated withPMAinthepresenceofthe CD 18(MAb) 60.3.AsshowninFig. 3, eosinophil adherenceto

80 eosinophils neutrophils Figure3. Effect of PMA

oneosinophiland

neu-60L

trophiladherenceto

' 40 . LPS-treated HEC in the

* 40

presenceof

_CDI

8 MAb.

20-

HEC

_monolayers

were

pretreated

for4hwith

o. s [ L s | L L medium aloneor

me-- LPS PMALPS/PMA - LPS PHALPS/PHA dium containingE. coli LPS(100ng/ml). Eo-sinophilsorneutrophils suspendedin PBScontaining5 mMglucose,

1mMCaCl2,and 1 mM

MgCl2

wereincubatedfor 15 minatroom

temperature in the presenceorabsenceof the CD18 MAb 60.3(20 ug/ml). Leukocyteswerethen addedtoHECmonolayers followed byPBS(control)or PMA(100ng/ml). Percentage leukocyte

adher-ence wasdetermined aftera30-minincubationat37°C.Valuesare

means±SE of sixexperimentswith fourreplicatewells in each exper-iment.*P<0.01 (pairedttest)comparedtoadherenceof unstimu-latedneutrophilstoLPS-pretreatedHEC in the presence of MAb 60.3 aloneorPMA alone. o,medium;.,MAb60.3.

untreatedHECwaspotentlystimulatedbyPMA in theabsence of MAb60.3,but itremained unaffected in its presence. Simi-larly, adherence ofeosinophils to LPS-treated HEC was in-creasedbyPMA in the absence of MAb60.3,butno

change

(i.e., no up- or downregulation) ofadherence between PMA-treated and untreated eosinophils was observed when MAb 60.3waspresent in the assay. Controlexperimentswere per-formedwith neutrophils. These cells adhered minimally to un-treatedHEC, but bound avidly to HEC when stimulated with PMA. PretreatmentofHEC with LPS also markedly increased neutrophil adherence. Adherence of neutrophils to LPS-treated HEC was only partially (by 50%) inhibited by MAb 60.3, whereas PMA-stimulated neutrophil adherence to un-treated HEC was completely inhibitedbythis MAb.However, MAb 60.3completelyabolishedneutrophiladherence to LPS-treatedHEC when PMA was added with the neutrophils, thus indicating that stimulation with PMA downregulated the CDl1/CD 1 8-independent mechanism of neutrophil adher-ence.

Discussion

Our results indicate that at least three binding mechanisms are involved in eosinophil adherence to endothelial cells: (a) a mechanism(s) that accounts for a small but definite (15-17%) proportionofeosinophiladherence to unstimulated HEC (ba-saladherence); (b)amechanisminvolvingtheleukocyte adhe-sion complex CDl

1/CD

18. This mechanism of adherence accounts, almostcompletely,for the increase of adherence fol-lowingeosinophil activation byagents such as PMA (eosino-phil-dependent adherence), and in part for the adherence of unstimulated eosinophils to endothelial cells that have been pretreatedwith LPS,TNF, or IL- 1 (endothelial-dependent ad-herence);and (c)amechanismspecificfor endothelial-depen-dent adherence that involves the interaction ofthe VLA-4 inte-grinreceptor(20, 36)oneosinophils with the

cytokine-

or LPS-inducible endothelial adhesion molecule VCAM-l (21) (also known asINCAM-110, 22).

Kimani et al. (15) have

previously reported

values from 25% to35% for basaleosinophil adherencetoendothelialcells, whereas Lamas et al. (14) havereported much lower values (< 5%). The higher values of basal eosinophil adherence ob-tained by us, in comparison to Lamas et al., were not ac-counted forby a failure toremovethe nonadherent cells,as indicated by parallel adhesionassayswith

eosinophils

and neu-trophils, usingthe sametechnique, and,inparticular,the same washingprocedure, in which the unstimulated adherence of eosinophilswas significantly higherthan that ofneutrophils, i.e., 15-17% vs.5-6%. The basal adherence ofeosinophilswas temperature-dependent, since itoccurredat

37°C

but not at 4°C,

suggesting

that an active

binding

mechanism(s)

is in-volved. Inaddition, basal

eosinophil

adherence

required

Ca++ orMg++and was greater inthepresence of both cations. These results suggest that twodistinctadherencemechanismsmay be involved inunstimulatedadherenceof

eosinophils,

one

requir-ingbothCa++and

Mg+',

and theother

requiring

either

Ca++

or

Mg++.

Furtherstudiesmay

identify

the

molecule(s)

involved in thisbasal adherence. Thehigherspontaneous adherence of eo-sinophils, as

compared

to

neutrophils,

may account for the propensity ofblood

eosinophils

to

emigrate

in the extravascu-lar space in

physiologic

conditions, i.e.,

in the absence of in-flammatorystimuli(1-4).

In a recentstudy,Lamasetal.(14)

reported

that eosinophil adherencewasincreased by soluble agents suchasthe tumor promoterPMA, thechemotactic bacterial

peptide FMLP,

and thechemotacticfactor foreosinophils PAF,aswellasby pre-treatment of HEC withLPS, TNF, orIL-1. Theproadhesive effect of PMA, FMLP,orPAFwasexertedthrougheosinophil activation,since the agents stimulated

eosinophil

adherenceon gelatincoated dishesaswellas onHEC. Similar resultswere obtained by Kimanietal.(15) by

stimulating eosinophil

adher-encewith PAF.PMA-, FMLP-,andPAF-stimulatedadherence involved the activation oftheglycoprotein adhesion complex CD1la,b,c/CD18 (LFA-1,

Mac-i, p150/95),

since adherence wascompletely inhibited by MAb directed

against

the com-monCD18subunit(14,15) ofCDl 1/CD18.In contrast, eosin-ophil adherence induced by LPS, TNF,orIL-I pretreatmentof HEC, wasonly inpartinhibited by the anti-CD 18 MAb, sug-gestingthataCD11/CD

18-independent

mechanismwasalso involvedinendothelial-mediated adherence (14).

Similarly,

we foundthatthe CD1 8MAb60.3

completely

inhibited PMA-in-duced adherence of

eosinophils

tountreated HEC, but only

partially

inhibited

eosinophil

adherenceto

LPS-, TNF-,

or

IL-1-treated

HEC. Three adhesion molecules havebeenidentified thatcanbeupregulated by endothelialcells uponstimulation with cytokinesorLPS:(a) ICAM-1,which

functions

asligand fortheleukocyte adhesionreceptorCD1

la/CD

18

(37,

38);

(b) ELAM-1 (39), which is involved in the adherenceto endothe-lium of neutrophils and monocytes (39a);and(c) VCAM-1, recentlyidentifiedas anendothelialsurface molecule involved in adherence of peripheral blood lymphocytes (22, 23) and monocytes(22)(39a),and somelymphocytic cell lines (21, 34). In ouradhesionassays,the

anti-VCAM-l

MAb4B9 caused a

significant

inhibition of

eosinophil

adherencetoLPS-or cyto-kine-treated HEC.'Moreover,when MAb4B9wasusedin com-bination with the CD18MAb60.3, the increase of adherence causedby the LPSorTNForIL-I treatmentof HECwas re-duced to control values

(i.e.,

values of unstimulated adher-ence). ThisindicatesthatVCAM-1 and theendothelial ligand for CD1 1/CD 18,most

likely

ICAM-1 (37, 38),arethe endothe-lial adherence molecules specifically involved in eosinophil

binding

toLPS-or

cytokine-stimulated

HEC. Recently, Elices et al. (36)reported that specific adherence of VLA-4-trans-fected cellstoVCAM-1 expressedoneither HEC monolayers, or on COS cells transfected with VCAM- 1, was completely inhibited by anti-VLA-4 MAb. UsingCD1

1/CD

18-deficient lymphocytes, Schwartz etal.(34)demonstratedthat lympho-cyte

binding

toTNF-stimulated HECwas

inhibited

byMAbs to VLA-4

(CD49d/CD29)

orVCAM-1. Similarly, we found thattwoanti-VLA-4MAbsinhibited eosinophiladherence to LPS-treated HEC, with no additive effectwhen anti-VLA-4 andanti-VCAM-1 MAbwereusedin combination.Hence, the

eosinophil

receptorrecognizing VCAM-1appears to be the in-tegrinreceptorVLA-4.In contrast to theanti-VCAM-l MAb, the anti-ELAM-1 MAb BB 1 1, that completely inhibited CD11/CD1

8-independent

neutrophil adherence to LPS-treated HEC, did not affect eosinophil adherence to LPS-, TNF- or

IL-I-treated

HEC, thus excluding involvement of ELAM-1 in endothelial-dependent eosinophil adherence. In-volvement of

ELAM-1

ineosinophil adherence to stimulated HEC couldbe excluded in ourexperimentsalso on the basis of the

following

observations: (a) eosinophil adherence to LPS-treatedHECwas

completely

inhibited at

40C, whereas

neutro-phil

adherence was

partially

_maintained.

_{Themechanism of}

neutrophil

adherence at

4VC

involves

ELAM-1,

since the

CD1

1/CD

1

8-dependent

adherence mechanismwasinactiveat

4VC,

andsince

_neutrophil

adherenceat

4VC

wasabolished

_by

the

anti-ELAM-l

MAb

_BB1

1.

(b) Neutrophil

CDI

l/CDI8-in-dependent

adherencetoLPS-treatedHECwas

reported

tobe

activein the_{presence of}

Ca"

only (19).

In contrast,

CDl

1/

CD1

8-independent

adherence of

eosinophils

to LPS-treated

HEC

(i.e.,

adherencein the presence ofanti-CD18

MAb)

re-quired

both

Ca"

and

Mg".

(c)

Asdescribed

previously

(25),

CD1 1/CD

1

8-independent

adherence of

neutrophils

to

LPS-treatedHECwas

downregulated

in the presence of

PMA,

an

effectof PMAthatmay be related with the

reported

downregu-lationof

neutrophil

MEL- 14

_antigen

_(40).

Incontrast, CDl

1/

CDl 8-independent

adherence of

eosinophils

to LPS-treated HEC was unaffected

by PMA,

thus

indicating

that

VLA-4/

VCAM-

1-dependent eosinophil

adherence is not

downregu-latedupon

eosinophil

activation.The last result mayprovidea

possible explanation

for theobserved

eosinophilic

infiltration intissues of

patients

with

partial

orcomplete deficiencyof the

leukocyte

membrane

CD1 1/CD

18 adhesion

complex

(leuko-cyte adhesion

deficiency, LAD) (41).

LAD

neutrophils

adhere in vitro toLPS- or

cytokine-stimulated

HEC

_by

the CD1

1/

CD1

8-independent

mechanism

_{(18, 19). However,}

_neutrophils

fail to accumulate in infected tissues in LAD

patients

(41).

SincetheCDl1/CD 18-independent adherencemechanism of LAD

neutrophils

isinhibitedin vitro when theneutrophilsare

activated

_by

agentssuchasPMAor

FMLP,

and in this condi-tion

they

do not adhere tostimulated HEC

(25),

a

possible

explanation

for thebehaviorof

neutrophils

inLAD

patients

is

that,

in

vivo,

theCDl

l/CD18-independent

adherence mecha-nism is

downregulated by inflammatory

stimuli produced at sites ofinflammation (25, 40).Theabsenceof

downregulation

of

eosinophil

CDl 1/CD

1

8-independent

adherence mecha-nismmay thenaccountforeosinophil adherencetoand

migra-tionacrosstheendothelium inLAD patients, giventhatLAD

eosinophils

areendowedwith thesameCD 1/CD1

8-indepen-dentmechanismsofadherenceasnormal

eosinophils.

Acknowledgments

We thank R.Gagliardi,F.Bassan,A. Parenzan, and A. Knowles for technical assistance.

Thiswork was supported in part by grants from the U. S. Public Health Service (HL 18645), the MPI and the CNR ofItaly,Target Project on Biotechnology and Bioinstrumentation, and grant No. 89.02733.04. Dr.Menegazziis arecipientofa Research Doctor Fellow-shipfrom the Anna Villa Rusconi Foundation. Dr. Carlos is arecipient ofaClinician-Scientist Award from the American Heart Association. Dr.Harlan is an Established Investigator ofthe American Heart Associ-ation. Dr. Nardon is arecipientof a grant from the Italian Association for Cancer Research(AIRC).

References

1.Hudson, G. 1968. Quantitativestudy of theeosinophil granulocytes.

Se-min. Hematol.5:166-186.

2. Stryckmans,P. A., E. P. Cronkite, M. L.Greenberg,and L. M. Schiffer.

1968. Kinetics ofeosinophilleukocyteproliferation in man. In Plenary Session

Papersofthe Twelfth Congress ofthe International Society ofHaematology.E. R.

Jaffe, editor. The International Society ofHaematology, New York. p. F19. (Abstr.)

3. Rytomaa, T. 1960. Organ distributionandhistochemical properties of

eosinophil granulocytesin therat.ActaPathol. Microbiol. Scand. 50(Suppl.

140):1-118.

Hand-bookofHaematology. H.Downey,editor. Hamish HamiltonMedical Books, London. 181-208.

5.Nutman,T.B.,E. A.Ottesen, and S. G. Cohen. 1989. The eosinophil,

eosinophilia,andeosinophil-relateddisorders. III. Clinical assessments. Allergy Proc.10:33-46.

6.Nutman,T. B.,E. A.Ottesen,andS. G. Cohen. 1989. The eosinophil, eosinophilia, and eosinophil-related disorders. IV. Eosinophil-related disorders

(continued). AllergyProc. 10:47-62.

7.Spry,C. J. F. 1988.Eosinophilsindisease. InEosinophils.A Comprehen-siveReview,andGuidetotheScientificandMedical Literature. C.J. F.Spry, editor.OxfordUniversityPress, Oxford. 131-316.

8.Migler,R., L. R. DeChatelet,and D.A.Bass.1978. Humaneosinophilic peroxidase:role in bactericidalactivity.Blood. 51:445-456.

9.Weiss, S. J., S. T. Test, C. M. Eckmann, D. Ross, and S. Regiani. 1986.

Brominatingoxidants generated by human eosinophils. Science(Wash. DC).

234:200-203.

10. Jong, E.C., A.F. Mahmond, and S. J. Klebanoff. 1981. Bactericidal

activityofeosinophilic peroxidase.J.Immunol. 124:1378-1382.

11.Gleich,G. J.,E.Frigas,D. A.Loegering, D. L. Wassom, and D. Stein-muller. 1979.Cytotoxic propertiesof the eosinophil major basic protein. J. Im-munol. 123:2925-2927.

12.Kroegel, C.,U.Costabel,and H. Matthys. 1987. Mechanism of mem-brane damage mediated by eosinophil major basic protein (letter). Lancet. i:1380-1381.

13.Butcher,E.C. 1990. Cellular and molecular mechanisms that direct

leu-kocyte traffic.Am.J.Pathol. 136:3-11.

14. Lamas,A.M., C.M.Mulroney, and R. P. Schleimer. 1988. Studies on the adhesive interaction betweenpurifiedhumaneosinophilsand cultured vascular endothelial cells. J.Immunol. 140:1500-1505.

15.Kimani,G., M. G. Tonnesen, and P. M. Henson. 1988. Stimulation of

eosinophiladherencetohumanvascularendothelial cells in vitro by platelet-acti-vating factor.J.Immunol. 140:3161-3166.

16. Wardlaw,A.J.,R.Moqbel, 0. Cromwell, and A.B.Kay. 1986.

Platelet-activatingfactor.Apotent chemotactic and chemokinetic factor for human

eo-sinophils.J.Clin.Invest.78:1701-1706.

17.Sigal,C. E., F. H. Valone,M. J.Holtzman, andE. J.Goetzl. 1987. Prefer-ential human eosinophil chemotactic activityof theplatelet-activating factor (PAF) 1-0-hexadecyl-2-acetyl-sn-glyceryl-3-phosphocholine (AGEPC). J. Clin.

Immunol. 7:179-184.

18. Pohlman, T. H., K.A.Stanness,P. G.Beatty, H. D. Ochs, and J.M.

Harlan.1986.Anendothelial cell surfacefactor(s)induced in vitro by lipopolysac-charide,interleukin- 1, andtumornecrosisfactor-alphaincreasesneutrophil ad-herencebyaCDw-dependentmechanism. J. Immunol. 136:4548-4553.

19.Dobrina,A.,B. R.Schwartz,T. M.Carlos,H.D.Ochs,P.G. Beatty, and J.M. Harlan.1989. CD1l/CDl8-independent neutrophiladherencetoinducible

endothelial-leukocyte adhesion molecules (E-LAM) in vitro. Immunology.

67:502-508.

20. Hemler,M.E., C. Huang,Y.Takada,L.Schwarz,J.L.Strominger,and

M. L.Clabby.1987. Characterizationof the cell surface heterodimerVLA-4and

relatedpeptides.J.Biol. Chem. 262:11478-11485.

21.Osborn, L., C.Hession,R.Tizard,C.Vassallo,S.Luhowsky,G.

Chi-Rosso,and R. Lobb. 1989. Directexpression cloningofvascular celladhesion molecule 1,acytokine-induced endothelialproteinthatbindstolymphocytes. Cell.59:1203-1211.

22. Rice,G.E.,J. M. Munro,and M. P.Bevilacqua. 1990.Inducible cell adhesion molecule 1 10(INCAM- 110)isanendothelial receptor forlymphocytes.

ACDIl/CDl8-independentadhesionmechanism. J.Exp.Med. 171:1369-1374. 23.Carlos,T.M.,B.R.Schwartz,N. L.Kovach,E.Yee,M.Rosso,L.Osborn,

G.Chi-Rosso,R.Lobb,andJ. M.Harlan.1990. Vascular celladhesion molecule-1mediateslymphocyteadherencetocytokine-activatedcultured human endothe-lial cells.Blood.76:965-970.

24.Carlos,T.M.,andJ.M. Harlan. 1990. Membraneproteinsinvolved in phagocyte adherencetoendothelium. Immunol.Rev. 114:5-28.

25. Dobrina,A.,T.M.Carlos,B. R. Schwartz, P.G. Beatty, H. D. Ochs, and

J. M.Harlan. 1990.Phorbol ester causesdown-regulation ofCDl l/CDl8-inde-pendentneutrophil adherence to endothelium. Immunology. 69:429-434.

26. Jaffe, E.A.,R.L. Nachman,C.G. Becker, andR.C.Minick. 1973. Culture of human endothelial cells derived from umbilical veins.Identification by morphologic and immunologiccriteria. J. Clin. Invest. 52:2745-2756.

27.Thornton, S. C., S. N. Mueller, and E. M. Levine. 1983. Human endothe-lial cells: use of heparin in cloning and long-term serial cultivation.Science

(Wash.DC). 222:623-625.

28. Maciag, T., J. Cerundolo, S. Ilsley, and P. R. and R. Forand. 1979. An endothelial cell growth factor from bovine hypothalamus: identificationand

par-tial characterization. Proc.NatL. Acad. Sci. USA. 76:5674-5678.

29. Boyum, A. 1968. Isolation of mononuclear cells and granulocytes from human blood.Scand. J. Clin. Lab. Invest. 21(Suppl. 97):77-81.

30. Andrews, P. C., and N. I. Krinsky. 1981. The reductive cleavage of myelo-peroxidase in half, producing enzymically active hemi-myelomyelo-peroxidase. J. Biol. Chem. 256:4211-4218.

31.Cramer, R., M. R. Soranzo, P. Dri, R. Menegazzi, A. Pitotti, G. Zabucchi, and P.Patriarca. 1984. A simple reliable assay for myeloperoxidase activity in mixed neutrophil-eosinophil cell suspensions: application to detection of myelo-peroxidase deficiency. J. Immunol. Methods. 70:119-125.

32. Beatty, P. G., J. A. Ledbetter, P. G. Martin, T. H. Price, and J. A. Hansen. 1983. Definition of a common leukocyte cell-surface antigen (Gp95-150) asso-ciated withdiverse cell-mediated immune functions. J. Immunol. 131:2913-2918.

33. Benjamin, C., I. Douas, G. Chi-Rosso, S. Luhowskyj, M. Rosa, B. New-man, L. Osborn, C. Vassallo, C. Hession, S.Goelz, K. McCarthy, and R. Lobb. 1990. Ablocking monoclonal antibody to endothelial-leukocyte adhesion mole-cule- 1(ELAM I). Biochem.Biophys.Res.Commun. 171:348-353.

34.Schwartz, B. R., E. A. Wayner, T. M. Carlos, H. D. Ochs, and J. M. Harlan. 1990. Identification of surface proteins mediating adherence ofCDl 1/ CD118-deficient lymphoblastoid cellstoculturedhumanendothelium. J. Clin. Invest. 85:2019-2022.

35.Sanchez Madrid, F., M. 0. De Londazuri, G. Morago, M. Cebrian, A. Acevedo, andC. Bernabeu. 1986. VLA-3: a novel polypeptide association with theVLAmolecular complex: celldistribution and biochemical characterization. Eur. J.Immunol. 16:1343-1349.

36.Elices,M. J., L.Osborn, Y. Takada, C. Crouse, S. Luhowskyj, M. E.

Hemler, and R. R. Lobb. 1990. VCAM-I on activatedendotheliuminteractswith theleukocyteintegrinVLA-4 at asitedistinct fromtheVLA-4/fibronectin bind-ing site. Cell. 60:577-584.

37.Marlin, S. D., and T. A. Springer. 1987. Purified intercellular adhesion molecule-1 (ICAM-1) is a ligandforlymphocyte function-associated antigen (LFA-1).Cell. 51:813-819.

38.Smith,C. W., R.Rothlein,B. J.Hughes,M. M.Mariscalco,H.E.Rudloff,

F. C. Schmalsteig, and D. C. Anderson. 1988.Recognition of an endothelial determinant for CDl 8-dependent human neutrophil adherence and

transen-dothelialmigration. J. Clin. Invest. 82:1746-1756.

39.Bevilacqua, M. P., J. S. Pober,D. L.Mendrick,R.S.Cotran, and M.A.

Gimbrone,Jr. 1987.Identificationofaninducibleendothelial-leukocyte adhe-sion molecule. Proc. Natl.Acad.Sci. USA. 84:9238-9242.

39a.Carlos,T. M., N.Kovach, B. R. Schwartz, M. Rosa,B.Newman,E. Wayner, C.Benjamin,L.Osborn, R. Lobb, and J.M. Harlan. 1991. Human monocytes bindto twocytokine-inducedadhesiveligandsonculturedhuman

endothelialcells:ELAM- IandVCAM-1. Blood.Inpress.

40.Kishimoto,T.K.,M.A.Jutila,E. L.Berg,and E. C. Butcher.Neutrophil

Mac-l andMEL-14adhesionproteins inversely regulated bychemotactic factors. Science(Wash. DC).245:1238-1241.

41.Anderson, D.C., F.C.Schmalsteig,M. J.Finegold, B. J.Hughes, R.

Rothlein,L. J.Miller,S.Kohl,M.F.Tosi,R. L.Jacobs,T.C.Waldrop,etal.

1985. Thesevereand moderatephenotypes ofheritableMac- 1,LFA- Ideficiency:

theirquantitative deficiencyand relationtoleukocyte dysfunctionand clinical

features. J.Infect.Dis. 152:668-689.