• No results found

Leukocyte adhesion deficient neutrophils fail to amplify phagocytic function in response to stimulation Evidence for CD11b/CD18 dependent and independent mechanisms of phagocytosis

N/A
N/A
Protected

Academic year: 2020

Share "Leukocyte adhesion deficient neutrophils fail to amplify phagocytic function in response to stimulation Evidence for CD11b/CD18 dependent and independent mechanisms of phagocytosis"

Copied!
11
0
0

Loading.... (view fulltext now)

Full text

(1)

Leukocyte adhesion-deficient neutrophils fail to

amplify phagocytic function in response to

stimulation. Evidence for

CD11b/CD18-dependent and -inCD11b/CD18-dependent mechanisms of

phagocytosis.

H D Gresham, … , D C Anderson, E J Brown

J Clin Invest.

1991;

88(2)

:588-597.

https://doi.org/10.1172/JCI115343

.

Stimulation of PMN with inflammatory mediators markedly augments Fc and CR1

receptor-mediated ingestion. However, CD11/CD18-deficient PMN from three patients with complete

leukocyte adhesion deficiency (LAD) failed to recruit phagocytic function in response to

phorbol esters, cytokine, or Arg-Gly-Asp-containing ligand stimulation. Because stimulated

ingestion is protein kinase C (PKC)-dependent, our data indicate that LAD PMN exhibit only

PKC-independent phagocytosis. The defect in PKC-dependent ingestion is specific for

CD11b/CD18 and not secondary to the chronic or recurrent infections which occur in this

disease. The LAD phenotype for phagocytic function can be reproduced in normal PMN by

the anti-CD11b MAbs OKM1 and OKM10. In contrast, MAb Mo1 (anti-CD11b) and MAb IB4

(anti-CD18) inhibit both CD11b/CD18-dependent and -independent mechanisms of

ingestion by normal PMN. Their ability to inhibit CD11b/CD18-independent ingestion may

be mediated by cAMP, as shown by experiments with a protein kinase A inhibitor HA1004

and by direct measurement of cAMP levels in immune complex- and FMLP-stimulated

PMN. These data indicate that CD11b/CD18-independent and -dependent mechanisms of

phagocytosis exist and that some effects of anti-CD11b/CD18 MAbs may be mediated by

alterations in cAMP levels.

Research Article

(2)

Leukocyte

Adhesion-deficient Neutrophils Fail

to

Amplify

Phagocytic Function

in

Response

to

Stimulation

Evidence for

CD1Ib/CD18-dependent

and -independent Mechanisms of

Phagocytosis

Hattie D.Gresham,*Irene L.

Graham,*

DonaldC.

Anderson,*

and

Eric

J.

Brownt

*Department

of

Pharmacology, UniversityofMissouri-Columbia, Columbia,Missouri

6S212;

DepartmentofPediatrics,

BaylorCollege ofMedicine, Houston, Texas77030;and

*Departments

ofMedicine, CellBiology,andPhysiology, andMicrobiology and Immunology, Washington University, St.Louis,Missouri 63110

Abstract

Stimulation of PMN withinflammatory mediators markedly augmentsFc andCR1 receptor-mediated ingestion.However,

CDM1/CD18-deficientPMNfromthreepatients withcomplete leukocyte adhesiondeficiency(LAD)failed torecruit

phago-cyticfunction in responsetophorbolesters,cytokine,or

Arg-Gly-Asp-containing ligand stimulation.Becausestimulated

in-gestionisprotein kinase C (PKC)-dependent,ourdataindicate that LAD PMNexhibit onlyPKC-independentphagocytosis. ThedefectinPKC-dependent ingestion is specific forCD1lb/

CD18andnotsecondarytothe chronicorrecurrentinfections

whichoccurinthisdisease.The LAD phenotypefor phagocytic

functioncanbereproducedin normal PMN by theanti-CD11b MAbs OKM1 and OKM10. In contrast, MAb Mol (anti-CD11b) and MAb IB4 (anti-CD18) inhibit both CDllb/ CD18-dependentand-independent mechanisms of ingestionby normal PMN.Their ability toinhibit

CDllb/CD18-indepen-dentingestionmay bemediated by cAMP,asshownby

experi-mentswith a protein kinase A inhibitor HA1004 and by direct measurementof cAMP levels in immune complex- and

FMLP-stimulatedPMN. These dataindicatethat

CD11b/CD18-inde-pendent and-dependent mechanismsof phagocytosis existand that some effects of anti-CDllb/CD18 MAbs may be me-diated byalterations in cAMP levels. (J. Clin. Invest. 1991. 88:588-597.) Key words: CD116/CD18 * cyclic adenosine monophosphate- phagocytosis

Introduction

Phagocytosis isacomplexbiologicalprocessofspecialized cells

whichis essential fornormal hostdefense. Itis necessary not

only fortheuptakeanddestruction of pathogensbutalsofor theclearanceof immune complexes, removal of damaged

tis-sue, andinitiation ofwoundrepair. Despite itscentral role in

hostdefense, knowledge oftheregulation of phagocytic func-tionat

inflammatory

siteshaslagged behind understanding of activation oftherespiratoryburst andexocytosis,other

func-This workwaspresented in part attheannualmeeting oftheAmerican

Federation for Clinical Research, Washington, DC,May 1989, ard

also appeared in abstract form (1989. Clin.Res.37:564A).

Addressreprintrequests to Dr.Gresham,Room517B, 1 Hospital

Drive,University ofMissouri-Columbia,Columbia, MO 65212.

Receivedforpublication 30 July 1990 and in revisedform 28 Febru-ary1991.

tions of professional phagocytes required for the maintenance

of homeostasis.Recent investigationsfrom ourgroupintothe regulation of human neutrophil (PMN) phagocytosishaveled

us to hypothesize that phagocytosis by PMN is primarily a

recruited functionatinflammatory sites (1-5). Stimulation of PMN with various inflammatory mediators markedly aug-ments both Fc- and CR1-mediated ingestion (1-9). This

re-cruitment of phagocytic function is reflected notonly in the

total number ofingested targetsbut also in thepercentageof

PMN participating in the phagocytic process(1, 2). Presum-ably, this regulation of ingestion acts tolimit the damaging products of PMN activation, such as toxic oxygen metabolites and secreted lysosomal products, to the site of inflammation.

We and othershave foundthatPMN can recruit several distinct molecular mechanisms for phagocytosisinresponseto

inflammatory signals, which may be regulated further by the specific opsonin (IgG, C4b, C3b,andC3bi)andopsonin recep-tor(FcRII, FcRIII, CR1,andCR3)engaged (1-9).Inaddition, monocytesandmacrophages employ quite different molecular mechanisms for phagocytosis from those employed by PMN,

evenwhenthesameopsoninoropsonin receptor is involved (1, 10). In this manner, at inflammatory sites where PMN and macrophages encounterpro-inflammatorycytokines,

interfer-ons, interleukins, chemotactic peptides, and components of

the extracellularmatrix, ingestion isnotonlyaugmented,but the molecular mechanism involved in phagocytosis may be

altered. As anexample,wehave shownrecentlythatingestion

ofIgG-opsonized targets by nonstimulatedPMNisnot depen-dent uponprotein kinase C(PKC)'activation. However,when

PMN are stimulated with tumornecrosis factor-a (TNF-a),

not only is ingestion of the IgG-opsonizedtargets markedly increased, but PKC translocation is also greatly augmented

overthe minimalamountobserved with either IgGorTNF-a

alone(11). This synergisminactivation of PKC is essentialfor the increasedingestion byPMNexposedto TNF-a(11).The

importance of these datatohost defense isdemonstrated bythe fact that PMN frompatientswith chronicgranulomatous dis-ease(CGD), which lack the abilitytogenerate a respiratory burst,aswellasPMNfrom patients withparoxysmalnocturnal

hemoglobinuria, whichlackexpression ofFcRIII,failto

am-plifyingestion ofIgG-opsonizedtargets in responsetocytokine (TNF-a) stimulation (4, 12). Therefore, defectsinthe different

1.Abbreviations used in this paper: CGD, chronic granulomatous dis-ease;DiBcAMP,dibutyrlcAMP; E, sheeperythrocyte(s); EC4b and EIgG, sheeperythrocyte(s) opsonizedwith C4b orIgG;respectively;

Fn,fibronectin; IBMX,isobutylmethylxanthine; LAD,leukocyte ad-hesion deficiency; PDBu, phorbol dibutyrate; PI, phagocytic index;

PKAandPKC, proteinkinaseAandC;RGD,Arg-Gly-Asp;TNF-a,

tumornecrosisfactor-a;Vn, vitronectin. J.Clin.Invest.

© The AmericanSocietyfor ClinicalInvestigation,Inc. 0021-9738/91/08/0588/10 $2.00

(3)

molecular mechanisms for phagocytosis employed by PMNat

inflammatory sites may be observed in both genetic and

ac-quired disorders which exhibitapredispositiontoserious and recurrentinfection.

Inthisregard, the purpose of the presentwork isto deter-mine whether CDl 1/CD18-deficient PMN from three patients with complete leukocyte adhesion deficiency (LAD) (13) would becapable of recruiting the ingestion mechanisms ex-pected to predominate atinflammatory foci. These patients lack the expression of a family of adhesion molecules

(CDI la,b,c/CD1 8) owing to deficient production or matura-tionofthe common ,3 chain of these heterodimers (13).Invivo, this defect resultsinapredispositionto severe recurrent

bacte-rial infections (14). The PMN from patients with LAD are thought to exhibit normal phagocytic functioninthe absence ofstimulation, aslongasthe target isopsonizedwithaligand for a receptor other than CDl lb/CD 18, (e.g., IgG) (14). How-ever, we and others have shown previously thatsome

monoclo-nal antibodies (MAbs) to either CDl lb (Mo1) or to CD1 8

(I1B4) inhibit both stimulated and nonstimulatedingestionof targetsopsonizedwith non-CD1lb/CDl8ligands( 15-17).We haveinterpreted these datato meanthatCDl lb/CD18playsa

role in the regulation of both stimulated and nonstimulated

phagocytosis independent of itsligand-binding function. The

purpose of the present work is to resolve this apparent

discrep-ancy.

Onepossibilitytoreconcile thesedatais thattheantibodies toCDllb/CD18 that inhibitingestiondo sobygeneratingan

intracellular signal which is inhibitory for ingestion.

Alterna-tively, the patients' PMN may have compensatory mecha-nisms ofingestion that are in factdifferent from the

mecha-nismsused by normal PMN.Our extensiveinvestigations into

the phagocytic mechanisms employed by normal PMN af-forded us theopportunitytoassess ifthesamemechanisms are usedbyLADPMN. Inthepresentwork,weshow thatPMN

frompatients withLADfailto

amplify

ingestion for either

IgG-orC4b-opsonizedtargets inresponse tostimulation by phorbol

esters, cytokines, or

Arg-Gly-Asp

(RGD)-containing

adhesive

proteins. WhereasLAD PMNaredefective inthemechanism for stimulated

ingestion,

ourdataindicate that theydo possess and use themechanism ofunstimulated IgG-dependent

inges-tion used bynormal PMN. In

addition,

ourdataindicatethat themonoclonalantibodiestoCDl lb, OKM 1, and OKM10 do

not inhibit unstimulated ingestion but do inhibit stimulated ingestion. In this manner, treatment ofnormal PMN with theseantibodiestoCD lIb resultsin reproduction of theLAD

phenotype for phagocytic function. In orderto explain why

antibodies MoI and 1B4 inhibit aCDI l/CDl8-independent mechanism ofphagocytosis,weinvestigated their effectson the

intracellular levels of cyclic adenosine monophosphate

(cAMP).

We had shown

previously

thatagents thatincrease intracellular cAMP inhibit unstimulated but not stimulated

ingestion(3). We present data which indicate that Mol and 1 B4, but not control antibodies, augment the accumulation of cAMP in response to either immune complex or

N-formyl-methionyl-leucyl-phenylalanine (FMLP) stimulation.In

addi-tion,

theability oftheseantibodiestoinhibitunstimulated

in-gestionbut notstimulated ingestion isabrogated by the

inclu-sionofHA1004, aputativeprotein kinaseA(PKA) inhibitor.

Thus, in thepresence ofa PKAinhibitor, Mol and 1B4can alsoreproduce theLADphenotype forphagocytic function.

These dataindicate thatat

inflammatory sites,

augmenta-tion ofPMN

phagocytic

function is absolutely dependenton

the CDl lb/CD 1 8 complex. Becausewebelieve thatthe mecha-nismfor stimulated ingestion is essential for normal host de-fense, this suggeststhat the susceptibility toinfection in

pa-tients with LADmayberelated inpart to afailuretoenhance

ingestion in response to inflammatory stimuli. Thismay be

particularly importantatsites, suchasthelung, where CDl 1/ CD18-deficient PMN ingress in response to inflammation is relatively normal. Finally, our data demonstrate that both CDllb/CD18-independent and -dependent mechanisms of phagocytosis exist in PMN. The ability of antibodies Mol and IB4 toinhibit the CDl1 /CD 1 8-independent mechanism of in-gestion may beexplainedby theirabilitytoaugment increases inintracellularcAMPinresponse tostimulation. This suggests thatother effects ofsomeantibodies and potential ligands for the CDI lb/CD 1 8 complex may result from alterations in cel-lular cAMP metabolism.

Methods

Special reagents. The following reagentswerepurchasedfromSigma

Chemical Co., St. Louis, MO: phorbol dibutyrate (PDBu), phorbol

myristate acetate (PMA), FMLP, isobutylmethylxanthine (IBMX),

catalase(bovine liver, 52,000 U/mg), superoxidedismutase (bovine

erythrocytes, 3,000U/mg),cytochromec(type VI,horseheart),and dibutyrl cAMP(DiBcAMP). PDBu (1 mg/ml),phorbolmyristate

ace-tate(PMA, 1 mg/ml), FMLP(22.9 mM), and IBMX(80mM)were prepared as stocksolutions in (DMSO, Aldrich ChemicalCo., Milwau-kee, WI), werestored at -70°C,andwerediluted into aqueous

me-dium immediately beforeuse. Equivalentconcentrations of vehicle were usedto controlfor solventeffects. Chicken egg albumin

(ovalbu-min) andhumanvitronectin (Vn)werepurchased from Calbiochem-Behring Corp., La Jolla,CA. Theisoquinolinesulfonamidederivatives H7and HA 1004 werepurchased fromSeikagaku America,St. Peters-burg, FL. Theywerereconstituted(10 mM)withphosphate-buffered

saline and stored protected fromlightat4°C.A 10-fold concentrated stockof Hanks' balanced salt solution (HBSS)waspurchased from GibcoLaboratories, Grand Island, NY. Human fibronectin(Fn)was

purifiedasdescribed(18). ThesyntheticpeptideGRGDSCwas pre-pared by the Protein Chemistry Facility, Washington University

School of Medicine. GRGDSC and Cas acontrolwerelinkedtopigeon

heart cytochrome c asdescribed (5).

MAbs. Purified MAbtoCDl lb/CDI8 wereobtainedasfollows: OKM1 and OKM10(anti-CD1Ib)werethe generousgiftsofDr. Patri-cia Rao, the R. W. Johnson Pharmaceutical ResearchInstitute, Rari-tan, NJ; Mo-I (IgG2a anti-CDlIb) and IB4 (IgG2a anti-CD18)as described (17), Leu-15 (IgG2a anti-CDlIb)from Becton, Dickinson & Co., SanJose, CA; andisotypecontrols 4B2(IgG2b)and l B5(IgG2a)

asdescribed (2). 4B2 and lB5 recognize unknownantigensexpressed onPMN. Purified murine monoclonal anti-FcRI (32.2), anti-FcRII (IV.3), andanti-FcRIII (3G8) were purchased from Medarex, Inc., WestLebanon, NH.Anti-CR 1 (3D9) wasobtained asdescribed (7).

IsolationofPMN.PMNwereisolated as described (3) from normal volunteers,three patientswith LAD followed at the Baylor Collegeof Medicine (14), and two patients with CGD followedattheUniversity

ofMissouri-Columbia School of Medicine. Patient sampleswere ob-tained whenpatientswerefree from infection. LADpatientblood sam-ples werealwaysaccompanied by blood drawnat the sametime froma normal control. ThePMN from these normal volunteers served to controlfor any effects oftransportation on PMNfunction.Multiple

studies of the normal transportedPMNshowed thattheirphagocytic,

respiratory burst, and excretory responses were not different from PMNisolatedfrom normal volunteers from the laboratory. TheLAD

patientshave beenstudiedpreviously(14) and theirPMN wereshown then andsubsequently byus to lackantigenicexpressionof CD18as

assessedby MAbIB4positivitymeasuredby fluorescent flow

(4)

Phagocytictargets. Sheep erythrocytes (E) were purchased from Whittaker M.A. Bioproducts, Walkersville, MD. IgG (EIgG and

EIgGhi)- and C4b (EC4b)- opsonized Ewere prepared asdescribed (3, 19).

Phagocytosisassays. PMN phagocytosis was assessed by a fluid-phase assay asdescribed (1-5). The reaction mixtures are described in thefigure legends.Phagocytosiswasassessed bylight microscopy and quantitated as a phagocytic index (PI), the numberof E ingested/ 100 PMN.

Superoxideanion assay. Superoxide anion was quantitated by cy-tochromecreductionasdescribed(4). The data are reported asthe

superoxide dismutase-inhibitablenanomolesofcytochromecreduced per1.0 x 106PMN/20min.

Proteinkinase C assay. PMN (1.0-1.5 X 106)in 1 ml ofHBSS`+

wereincubatedwith and without 10 or 100 ng of PMA for 15 min at

370C.Thesampleswerediluted intoice-coldHBSS,pelleted by

centrif-ugation, andresuspended in 1 mlof homogenization buffer, 5 mM Tris, pH 7.5,1mMMgCI2,2mMDTT,10mM EGTA, and 10Mg/ml

of eachleupeptin, aprotinin, pepstatin, and p-nitrophenyl

pi-guani-dino-benzoate. The suspensions weresonicated for 15 s onice before

centrifugationat 100,000 gfor 1 h. PKC activity in the supernatant

(cytosolic fraction)wasassessedasfollows:incubationmixtures (0.25 ml)containing20 mM Hepes, pH 7.5, 10 mMMgC12,5mM DTT, 500

jeg/ml

histone III-S(SigmaChemical Co.), 0.5 mMCaCI2, 12

4g/ml

1-oleoyl-2-acetylglycerol, 60 ug/mlphosphatidyl serine (both lipids

from Avanti PolarLipids,Birmingham, AL),25Mdof supernatant,

5,gl

of[y-32P]ATP(40-50MM;NewEngland Nuclear, Boston,MA),and 50MMATP wereincubated for 20 minat30°C.Additional unlabeled ATP wasadded toafinal concentration of 1 mM beforeapplicationof 10

gl

of each assay mixture toP81 phosphocellulose paper(Fisher

ScientificCo., Pittsburgh, PA).The paperwasdried inamicrowave oven for 2 minand, aftercoolingtoroomtemperature,waswashed twicefor 20 min in 10% TCAcontaining 10 mMpyrophosphate,and oncefor 20 min in 5% TCA. The paperwasrinsed in ethanol(ETOH)1 min, a 1:1 mixture of ETOH andacetone for 1 min, andfinallyin acetonefor 1 minbeforeairdrying. Theradioactivityboundtothe paperwasassessedbyscintillationcountinginScintiVerse(Fisher Sci-entificCo.).PKC-specificphosphorylationwascalculatedasthe differ-ence between TCA-precipitable radioactivity in complete reaction mixtures andphosphorylationin mixtureslacking Ca++and phospha-tidyl serine.PKCactivityisreportedasthepicomoles/minuteper 107 PMN.Under thesereactionconditions,assessmentof loss ofcytosolic activitywas moresensitive andreproduciblethanwastranslocationof

activitytoparticulatefractions,presumablybecause of therapid cleav-age of translocated PKCto aCa++andphospholipid-independent ki-naseactivity (20).

cAMPassay. PMN (2.0 X

107/ml)

were incubated in 400 uM IBMXfor 5 minat37°C. Without washing, 2X 106PMN were incu-batedwith

20Mgg

ofpurified antibodyinavolumeof 150

M1

for 30 min at0°C.Immunecomplexes(rabbit IgG anti-BSA/BSA)equivalent to 360Mgofantibody/mlwereadded and themixtureswereincubated at 1.0X106 PMN/ml for 30 min at 37°C.Insomeexperiments, FMLP at afinal concentration of 1 MMwasadded andthe mixtureswere incu-bated for 15 minat37°C.cAMPlevelswereassessedasdescribed using

the acetylationprocedure withthe radioimmunoassaykitfrom New England Nuclear (3). Cell supernatants were acetylated and assayed immediately upon generation. Supernatants which were frozen before

acetylationhadsignificantlyreduced andvariable levels of detectable cAMP.The dataarerepresentedas netpicomoles cAMP per 1.0X 107

PMN. The averaged valueforbuffer-treated PMN that was subtracted from the experimental values was 10.1±0.23

pmol/l.0

X 107PMN.

Results

LAD PMN

fail

to ingest IgG-

and

C4b-opsonized

targets

(CDJlb/CD18-independent

opsonins)

inresponse to

stimula-tion. NormalPMNin

suspension

bind EIgG but are not very efficient at

mediating

phagocytosis; however,

when they are

stimulated withanoptimal dose ofPDBu(15 ng/ml), ingestion

ismarkedly augmented (references 3 and 4 and Fig. 1 A). This augmentation is reflectednotonly inthe total number of in-gested targets(PI)but also in the percentage of phagocytosing PMN(- 30% for unstimulated PMN vs. - 80% for

PDBu-stimulated PMN). Unlike ingestion by buffer-treated PMN, ingestion stimulated by PDBu is dependentuponactivation of therespiratoryburst and is inhibitedbythe inclusion of

cata-lase in the reaction mixture (Fig. 1 A, and reference 4). Adhe-siveproteins containing the sequence RGD also stimulate the ingestion of EIgG; however, they dosobyadifferent mecha-nism because the inclusion of catalasetoinhibit the myeloper-oxidase-hydrogen peroxide-halide oxidant system is required

to demonstrate RGD-stimulated ingestion (5). As shown in Fig. 1 A, when normal PMN are stimulated with an optimal dose(40,Ag/ml) ofapolyvalent RGD ligand (cc-RGDS), inges-tion is augmented but only in the presence of catalase. In

con-trast, PMN fromtwopatients with LAD (Fig. 1, B and C) failed

torespond to an optimal dose of either PDBu or cc-RGDSto

augmentingestion either in the presenceorabsence ofcatalase. However, ingestion by buffer-treated PMN either in the pres-ence orabsence of catalasewasthesame(Fig. 1 C)orslightly elevated(Fig. 1 B)ascomparedtonormal PMN(Fig. 1 A). In addition, the failureto respondwas notsimply because ofa

change in the dose-response curveof thestimulant forLAD PMN, because stimulation ofLADPMN with either PDBu from 15to 300 ng/ml (optimal dose 15 ng/ml)or cc-RGDS from 20to 160 Mg/ml(optimal dose 40

Ag/ml)

still did not

result in augmentation of ingestion (data not shown). In addi-tion, the failureof LAD PMN torespondtocc-RGDSwas not

due tofailuretobind the stimulant because wehaveshown

previously that LAD PMN bind equivalent amounts of

cc-RGDSascompared to normal PMN (5). Two other

RGD-con-tainingadhesive proteins, Vn and Fn, also failed to augment ingestion of EIgG by LAD PMN (data not shown). Not only did PDBu stimulation fail to increaseingestionby LAD PMN butitactuallydecreased the level ofingestionascompared to

ingestion by buffer-treatedLAD PMN(Fig.1, B andC). Simi-lardatawereobtained for ingestion stimulatedwith a low mo-lecularweight cytokine YM-IOE (2) and ingestion stimulated withTNF-a (I 1) (data not shown). A similarobservationhas beenmade for the adherence ofLAD PMN to vascular

endothe-A.NORMAL PMN

400r

300F

PI 200F

1Q00

0~~~~~

Bufe PDu RO

B. LADPMN #1

_Buffer

_i Catalase

C. LAD PMN #2

Buff A ':

Buffer PDBu GCCRODS Buffer PDBu CC-RODS

Figure 1.LAD PMN fail toamplify ingestionof EIgG in response to stimulation.PMN(1.0x i05)from normal controls(A)or two pa-tients withLAD(B, patient 1 and C, patient 2) were incubated with EIgG in the presence(hatchedbars)orabsence (openbars)of5,000

Uof catalase and eitherbuffer, 15 ng/ml PDBu, or 40Mg/mlof cc-RGDS in afinal volume of 115 Mi for 30 minat370C.Ingestionwas

(5)

lium: adherence of FMLP-stimulated LAD PMN isd sed, ascompared toadherencebybuffer-treated LAD PMN (21).

Identicaldata as depicted in Fig. 1 were obtained with PMN from a thirdpatient with LAD(datanotshown).For our addi-tional studies we have used PDBu stimulation as a model of

cytokine-stimulated ingestionand cc-RGDS as a model of ex-tracellular matrix stimulated ingestion to avoid the issue of the

failure ofthestimulant to bind as anexplanationfor the failure of LAD PMN to demonstrate stimulated ingestion. However, the failure to demonstrate stimulatedingestionby LAD PMN could be observed withYM-IOE,TNF-a, Fn, and Vn.

Inges-tion via another CDllb/CD I8-independent opsonin, C4b, also was

investigated.

AsshowninFig. 2, normalPMN stimu-latedwithanoptimaldoseof PDBuavidly ingested EC4b.In contrast, PMN from the two LAD patients failed to respond to PDBu to ingest the EC4b.As we haveshownpreviously, the

anti-CD1lb/CD18monoclonalMol iscapableof completely

abrogating PDBu-stimulated ingestion of EC4b by normal PMN (17). Forboth EIgG and EC4b,the failure to demon-stratestimulated ingestionwas notdue to a failure to bind the

opsonized particles.PMNfrom the threepatientsbound equal numbersof both EIgG and EAC4basthe normalPMN(data

not shown). In addition, LAD PMN expressed normal

anti-genic levels of FcRII, FcRIII, and CR1 asassessed by fluores-centflow microcytometry (datanotshown). These data indi-catethat CD1l/CDl 8-deficientneutrophils failtorespondto

pro-inflammatoryagents toamplify phagocytic functioneven

for CDIlb/CD1

8-independent

ligands.

Ingestion

by

LAD PMN is

PKC-independent.

Ourprevious studieson the molecularmechanisms ofIgG-mediated inges-tion by non-stimulated and stimulated PMN have ledus to

hypothesize that ingestion by buffer-treatedPMNis indepen-dent of PKC activation. In contrast, PMN stimulated with TNF-a, granulocyte/macrophage colony-stimulating factor (GM-CSF),orPDBuexhibitingestion ofEIgG which is

depen-dent upon PKC activation (11). Because CDl

1/CD18-defi-140

120-O Buffer * Mol 100

80

Pi 60

40

20

0~

Normal LAD#1 LAD#2

Figure 2.LAD PMNfailtoingest EAC4b inresponse to stimulation. PMN(1.0X IO')fromnormal controls and two patients with LAD

(patient 1 and patient 2)werepretreated withbufferor0.83

,g

of

purified anti-CD 1Ib/18(Mol)for10min at room temperature. Without washing, the mixtureswerefurther incubated withEAC4b in thepresenceof15ng/mlofPDBuinafinalvolumeof1 5,ulfor 30 minat

370C.

PI=numberof EAC4b

ingested by

100PMN. Data

are

represented

asthe

mean±SEM,

n=3-6determinations.

cieut PMN-failedto respondtoactivation to enhance inges-tion, we sought to investigate the mechanism of ingestion uti-lized by these PMN as compared to normal PMN by examining the effects of PKC inhibitors and of PKA activators and inhibitors oningestion by LAD PMN. Possibly, the failure of LAD PMN to respond to stimulation was due to in vivo activation byinflammatory mediators as a result of subclinical infection. As shown in Fig. 3 A, ingestion of EIgG by buffer-treatednormal PMN was not affectedby the isoquinoline sul-fonamide derivative H7, a PKC inhibitor, or by a fourfold greater concentration ofthecontrolderivative HA1004,which

is more specific atthis concentration for PKA. In contrast, PDBu- and cc-RGDS-stimulated ingestion were completely

abrogated by H7 whereas the control HA 1004 had no effect

(Fig.3A).Inaddition, ingestionof EC4bbyPDBu-stimulated normal PMN wasinhibited 78.3% byH7andwasunaffected

byHA 1004(datanotshown).The inclusion of DiBcAMP

in-hibitedIgG-dependent ingestion by normalPMN, but had no effectonEIgG ingestionstimulatedbyPDBu orcytokines(Fig. 3A).This inhibition occurspresumably by activation ofPKA andcanbe overcomeby PDBuorcytokine stimulation of

nor-mal PMN (Fig. 3A and references 3 and 11). Importantly,

ingestion of EIgG by buffer-treatedLAD PMN was not inhib-itedbytheinclusionofeither H7 or HA1004(Fig.3 B). Further-more, DiBcAMP inhibited EIgG ingestion by LAD PMN.

Thus,

ingestion

by buffer-treatednormal PMN or LAD PMN

is independent of PKC andcanbeinhibitedby PKA, whereas stimulatedingestion, observable only with normal PMN,

re-quires

PKC activation. Thesedatademonstrate that the

molec-ular mechanisms used by CD1 1/CD18-deficient PMN for

phagocytosis

aresimilartothoseusedbyunstimulatednormal PMN.

Therefore,

it is unlikelythat thefailure ofLAD PMNto

respondtostimulation by PDBu, cytokines,or

RGD-contain-ing

adhesive proteins reflectsan in vivo preactivation to the

stimulatedstate.

The

defect

in LAD

PMNphagocyticfunction

issubsequent

toPKC activation.Because theCD1 8-deficientPMNfailedto demonstrate

PKC-dependent ingestion

in response tophorbol

esterstimulation, weinvestigated the ability ofLADPMNto

translocate PKCfrom the cytosol inresponse toactivation with phorbolesters.Asshown inFig.4, LAD PMN lost PKC

activ-ity

from

cytosolic

fractions similarlyto normal PMN in re-sponsetostimulation with

increasing

concentrations ofPMA. BecausetheLAD PMN could translocate PKC normally, we

investigated

superoxide anion generation, another cell

func-tion

thought

to bedependentuponPKCactivation,to

deter-mine whetherLAD PMN weredefectivein all PKC-dependent processes.Several reportsexaminingthegeneration of

super-oxide anion byLAD PMNhaveshown that these cells generate

superoxide

anionafter stimulation with high concentrations of either

phorbol

esters orFMLP (14). These studieswere

per-formedoncellsin

suspension.

However,there has been atleast

onereport ofa

patient

whose PMN in suspensionproduced elevated levels of

superoxide

anion ascompared to normal PMNinresponse to FMLP and areducedresponse tophorbol

esterstimulation

(22).

In

addition,

someforms of

adherence-dependent

activation ofthe

respiratory

burst appear to be di-minished inthesepatients (23, 24).Ourstudieswereperformed

oncellsin

suspension

under theconditions ofthephagocytosis

assay.Unlike their complete

inability

toenhancephagocytosis, both

patients'

PMN could produce superoxide anion in re-sponsetoPDBustimulation

(data

notshown).PMNfromone

(6)

A.

NORMAL

500

400

300

200

100

-0

-B. LAD

* Buffer

[ 0.625iMH7

2.54MHAl 004

E 500gMDIBcAMP

~~~~~~~~~~~~~~~~~~~~~

I

Buffer

PDBu

CC-RGDS

Buffer

PDBu

CC-RGDS

Figure 3. Effect of pharmacologic inhibitors of PKConingestionofEIgG by normal PMN and LAD PMN.PMN(1.0x 10)fromnormal

controls(A) and LAD patients1 and 3 (B)wereincubated with EIgG inthepresenceoftheindicatedpharmacologicagentsandeither buffer, 15ng/ml PDBu,or40

Ag/ml

of cc-RGDS inafinalvolumeof1155d for30minat370C. Tubes containing cc-RGDS also contained 5,000U

of catalase. PI=numberof EIgG ingested by 100 PMN. Dataarerepresentedasthemeanof duplicate determinations.

the otheraslightly diminishedresponse ascomparedtonormal PMN(LAD patient

1,

6.41+0.95 nmol superoxide anion/106 PMN; LAD patient 2, 14.2±1.6 nmol superoxideanion; and normal PMN, 8.86±0.95 nmol superoxide anion, mean± SEM,n = 3-4). The inclusion ofH7inthe reactionmixtures

inhibited superoxide anion generation by LAD and normal

PMNequivalently: 47.6% inhibition ofsuperoxide anion

gener-ationby PDBu-stimulated LAD PMNascomparedto 56.6% inhibition for PDBu-stimulated normal PMN (data not

shown). Thus, both patients' PMN were able to generate a PKC-dependent respiratory burst. Therefore, these data donot

indicatearole for CD11/CD 1 8 in superoxide anion generation subsequenttoPKCactivationasstudiedinourassay.The vari-ationintheresponse of theLADPMNindicatesthat CD11/ CDl8-independent factors (i.e., inflammatory cytokines to

which the PMN had been exposed in vivo)mayplayarole in regulating the magnitude oftheresponse.Weconclude that the failure of LAD PMNtorecruitPKC-dependent mechanisms forphagocytosis inresponsetoseveral stimuli doesnotreflecta

general failuretoactivate PKCortoexhibitPKC-dependent membrane functions.

Thedefectin LADPMNphagocyticfunctionis

specific

for thedeficiency of CDJ b/CDJ8. Becauseone ofthe patients' PMN expressed FcRI even in the absence of evidence of

currentinfection(datanotshown),weinvestigated if the fail-ureof LAD PMNtodemonstrate PKC-dependent ingestion wasspecific for the deficiency of CDI lb/CD1 8orwasduetoa

nonspecific effect of inflammation. We performed several

ex-periments to address this. First, we examined the ability of PMN fromtwo patients with CGDtodemonstrate PKC-de-pendent ingestion in responsetoRGDstimulation. We have reported previously that CGD PMN do not demonstrate

PDBu-orcytokine-stimulated ingestion because that stimula-tion isdependentuponactivation of therespiratoryburst(4). However, PMN fromaCGDpatient withahistory ofrecurrent

infection similar to the LAD patients increased ingestion of EIgGsignificantly when stimulated with the RGD-containing adhesiveprotein,Fn(Fig. 5 B). ThePMNfrom thispatient did express a smallamount of FcRI(data not shown); this may explain the high levelofunstimulatedingestionascomparedto

unstimulated normalPMN(Fig. 5 A).Eventhoughthese PMN

exhibited evidenceof in vivoinflammatorycytokineactivation

(25), Fn-stimulated ingestionwas stilldemonstrable. Similar

datawereobtainedwith PMNfromasecondpatientwith CGD (datanotshown). Pretreatment ofPMNwith the anti-CD18

monoclonal IB4completely abrogated ingestion byeither nor-malorCGD PMN(Fig. 5). As expected, Fn-stimulated inges-tionbyCGDPMN, unlikenormalPMN,wasobservable with-outtheinclusion of catalaseinthereaction mixture. In

addi-tion, when normal PMN were incubated with 100 U/mlof interferon-y for 6htoinducetheexpressionof FcRI(8), they stillretainedtheabilitytorespondto Fntoenhanceingestion (data not shown). EIgG ingestion by unstimulated PMN

in-creasedfromaPI =88±8with control PMNtoaPI= 157±5

with

interferon-y-treated

PMN. Fn stimulation increased EIgG ingestion for both control (PI=313±31)and

interferon-'y-treatedPMN(PI=332±49, mean±SEM,n=3).These data

indicate that the defect in phagocytic function observedwith

LAD PMN isunlikelytobeexplained byin vivoexposureto

inflammatory cytokines which inhibit PKC-dependent inges-tion.

Next, weinvestigated the ability ofseveral MAbsto the

CDl lb/CD18complextoinhibitonly PKC-dependentbut not

PKC-independent ingestion by normalPMN. We have shown 0.

200

(7)

-100-

A.

NORMAL PMN

B. CGD

PMN

90

0

0

80

70

60

50'

40

30

20-10'

0*

I

Control

5

LAD

PI

T

D

Control

I1B4

Buffer Fn

-F

10 100

PMA

(nM)

Figure 4.EffectofPMAoncytosolic PKCactivity from normal and LAD PMN. PMN (1.0-1.5X106)from normal controls (solid bars) andLADpatient I(hatchedbars)werestimulated with 10 or100ng of PMA inafinal volume of 1 ml for 15 min at370C.The suspen-sionsweresonicated and the PKCactivityof thecytosolic fractions

wasdetermined as described in Methods. Data represent the mean ofduplicatedeterminations of the percentage of cytosolic PKC

activ-ityofbuffercontrols and are represented as the mean±SEM, n=3. In arepresentative experiment, the100%cytosolic PKCactivity of the buffer controlwas4.2pmol/minper 107normalPMNand 3.5 pmol/min per107LAD PMN.

previously that both MAbIB4 and Mo 1 inhibit both unstimu-lated and stimulated ingestion (17). Therefore, they inhibit bothPKC-dependentand

-independent ingestion.

Asshownin

Fig. 6,PMNtreatedwithOKM 1, aMAbto

CDl

lb, exhibited normallevelsof PKC-independent ingestion but failedto

re-spondtoeither PDBuor Fn tostimulate

ingestion,

either in the

presence orabsenceofcatalase (Fig. 6 B). In contrast, PMN

treated withan

isotype

controlantibody orwith Leu-l 5,

an-other MAbto

CDl

lb, exhibited both

PKC-independent

and

PKC-dependent ingestion. Increasing the concentration of

OKM1 to 10

Aug/1.0

XI0 PMN

(10-fold increase)

stillhadno

effecton

unstimulated, PKC-independent ingestion

by buffer-treated normalPMN(datanotshown).In

addition,

asecond anti-CD1lb MAb, OKM10, gave similar results as

OKMl

(data not

shown).

Thus some MAb

against

CDl lb/CD18 couldconvertnormal PMNtothe

phenotype

ofLAD PMN:

capable

of

unstimulated,

PKC-independent ingestion,

but inca-pable ofanyform of

PKC-dependent

phagocytosis.

Theanti-CD18

MAb,

IB4, also inhibited

PKC-dependent

ingestion.

However,

unlike

LAD PMN or OKM

1-treated

PMN, IB4-treated

PMN

werealso

incapable

of

PKC-indepen-dent

ingestion

of EIgG. We sought toresolve this apparent

discrepancy

by

testing

the

ability

of the PKA

inhibitor,

control

IMI34

-Buffer Fn

Figure 5. Effect of fibronectin on ingestion of EIgG by CGD and normal PMN. PMN (1.0X 10)from normal controls (A) or from a patient with CGD (B) were incubated with 1.0

jtg

of purified anti-CD18 MAb IB4(hatched bars)orisotype control (openbars) for 15 min at room temperature. Without washing, the suspensions were further incubated with ESgG in either buffer or 5 ug/ml of Fn for 30 min at370C.Reactionmixtures with normal PMN (A) also contained 5,000 U of catalase. PI = number of EIgG ingested by 100 PMN. Dataarerepresentedasthemean±SEM,n = 3.

HA1004, to reverse IB4-mediated inhibition of EIgG ingestion by buffer-treated PMN. Aswehave shown previously and in thispaper(Fig. 3), activation of PKA by DiBcAMP inhibits PKC-independent ingestion of EIgG. Therefore, we reasoned that MAbIB4might inhibit ingestion byasimilar mechanism. Asshown in Fig. 7 A, PMN treated with MAb IB4, as

com-paredtotreatmentwithaPMN-binding isotype controlMAb,

wereinhibited in their ability to ingest EIgG even when the E were opsonized with a fivefold greater concentration of IgG

(EIgGhi).

However,when HA 1004 was included in the reaction

A.

CONTROL

ANTIBODY B.

OKMI ANTIBODY

EIgG PDBu Fn EIgG PDBu Fn

Figure6. Effect ofanti-CDI lb/l8 monoclonalOKM1oningestion

ofEIgGbynormal PMN. PMN(1.0x I0O)wereincubated with 1.0

ggofpurifiedisotypecontrolantibody (A)orOKM1 (B)for 1 hon

ice. Withoutwashing,thesuspensionswerefurther incubated with

EIgGinbuffer, 15ng/ml ofPDBu,or5

gg/ml

ofFnfor 30 minat

37°C.Reactionmixtureswereincubated in the absence(open

bars)

andin the presence of5,000Uof catalase(hatchedbars).PI= num-berofEIgGingested by1(00PMN.Dataarerepresentedasthe mean±SEM,n=3.

:

(8)

B. HA1004

EIgG EIgGhi

Figure7. Effect of HA1004on monoclonal antibody IB4-mediated inhibition of EIgG ingestion.PMN(1.0X 10)were incubated with 1.0 ,ug of isotype control antibody (open bars) orpurified IB4(hatched

bars)for 15minat roomtemperature. Without washing, the suspen-sionswerefurther incubated with either EIgG orEIgGhjin the absence (A) or presence (B) of 5MuMHAl004 inafinal volume of115Mlfor 30 minat37°C. PI=numberof EIgG ingested by 100PMN. Data arerepresentedasthemean±SEM,n=3.

mixture,treatmentofPMNwithMAb IB4 nolonger inhibited ingestion of either EIgGor

EIgGh

(Fig. 7B). This

concentra-tion ofHA1004 (5

gM)

also prevented DiBcAMP-mediated inhibition of ingestion (datanotshown).Next, weinvestigated ifHA1004 would alsopreventtheability ofMAbIB4 toinhibit stimulated orPKC-dependent

ingestion.

Asshown in Fig. 8, MAb IB4 treatment

of

PMN

completely abrogated

PDBu-stimulated ingestioneven in -the presence of 5 uM HA1004. Therefore, inclusion ofHA1004 revealed thatMAb IB4could alsoconvertnormalPMNtothe LADphenotype, capableof

baseline

IgG-mediated ingestion

but

incapable

ofstimulated ingestion. Identical data were obtained when the effect of

HA1004on Mo

1-inhibited

ingestionwasassessed.The

ability

ofHA1004 toreverse MAb IB4-mediated inhibition of

only

unstimulated

ingestion

couldnotbe

explained

by the absolute level of

ingestion

involved, becauseIB4 inhibition of

EIgGh

ingestion withPI=240-280was

prevented

byHA1004

(Fig.

7

B)yet IB4inhibitionof PDBu-stimulated

ingestion

with PI in

the same range was unaffected by PKA inhibition

(Fig. 8).

ThesedataindicatethatMAbsMo 1 and IB4 inhibit PKC-de-pendent

ingestion

bya

completely

different mechanismthan

400

-A

Figure 8. Effect of

mono-A

\ clonalantibody IB4on PDBu-stimulated ingestion 300 A by HA1004-treatedPMN.

PMN(1.0x 10)were in-* cubatedwith

1.0

,ug of

iso-PI 200 type control antibody

(A)

or

purifiedIB4(o) for15min

/00

atroomtemperature. 100

*

k k Without

washing,

the reac-tion mixtureswerefurther

__ ___, incubatedwith 15 Mlof

20 40 60

EIgG

inthepresenceof5 [PDBu]nM MMHA1004and the indi-catedconcentrations of PDBu. PI =number of EIgG ingested by 100PMN. Data are repre-sentativeoftwoexperiments.

themechanism by whichtheyinhibit PKC-independent

inges-tionof EIgG. Therefore,in the presenceofHA1004, IB4 and Mo1,aswellasOKM1 and OKM 10 convert normal PMN into thephenotype ofLADPMNforphagocytosis.

These data strongly imply that some antibodies against

CDl

lb/CD18canactivatePKAinPMN. Tofurthertestthis hypothesis we examined the effect of MAbs'Mo1

(anti-CDl

lb) andIB4(anti-CD18)onintracellularcAMP levelsin

normal PMN. Leu 15 served as the optimal control antibody

for the effects ofboth Mo 1 and 1B4 because it also binds

CDl

lb, is the same isotype as Mo 1 and 1B4, and does not

inhibit either stimulatedorunstimulated ingestion ofEIgG. As

shown inFig. 9, MAb Mol stimulation of PMN,atthe amount

of 1B4 usedper PMN inFigs. 7 and 8 (20 ,ug/2.0x 106 = I

,ug/

1.0 x

105

PMN), did not cause a statistically significant

increaseinintracellularcAMP levels as compared to stimula-tion withLeu 15.Inallexperiments, cAMP levels of untreated

PMN were not significantly different from Leul 5-treated

PMN. In contrast to incubation in buffer, stimulation with

IgG-containing immune complexes (which mimics stimula-tion by EIgG) resultedin astatistically

significant

increase in

netpicomolescAMPfor theMo 1-ascomparedtotheLeu 1

5-treatedPMN

(Fig.

9).In

addition,

weassessed the effect ofMol

onFMLP-stimulatedcAMP

accumulation.

FMLP was

exam-ined because it isknown toincrease surfaceMac- 1by

up-regu-lationof intracellular pools of CD1 lb/CD18. In

addition,

the effect of FMLP onintracellular levels ofcAMPin PMNhas

been

extensively

studied (26). FMLP doesnotstimulate

adenyl-atecyclase directly,butdoescause amodest increase incAMP (- net 10

pmol/107

PMN) presumably by inhibition ofthe cAMP

phosphodiesterase (26).

Asshownin

Fig. 9,

treatmentof

PMNwith Mol butnotLeu 15resulted inamarked

enhance-mentofcAMPaccumulationinFMLP-treatedPMN. Infact,

z

CL

,-0

x

0

Q

E

C-z

20

-10

-* Leu 15

* Mol T

+

CONTROL IMMUNE

COMPLEXES

FMLP

Figure9.Effect of antibodiesLeul5and Mol onnetintracellular cAMP levels of unstimulated and stimulatedPMN.IBMX-treated PMN(2.0X 106)wereincubated with 20

Mg

ofpurifiedLeul5orMol for 30 minat0°C.Either vehiclecontrol,immunecomplexesat360 Mgofantibody/ml,or I MM FMLP was added in a final volume of 2 mland incubatedat37°C. cAMP levelswereassessedasdescribed inMethods. Data arerepresentedasthe mean±SEMnetpicomoles

cAMP/1.0

X 107PMN,n=3. t*P<0.001.Theaveragedvalue for buffer-treatedPMNwhichwassubtracted from theexperimental val-ueswas10.1±0.23pmol cAMP/1.0X 107PMN,n=3.

300r-A.

BUFFER

250H 200H PI

1501-50- h

EIgG

EIgGhi

1fnl

(9)

Mol treatmentdoubled theincrement in-cAMP

accumulatioo

ascomparedto Leul 5treatmentof FMLP-stimulated PMN. OKM10 treatmentof PMN, which hasnoeffecton

unstimu-latedingestion,wasidenticaltoLeul 5 inits effectoncAMP

levels (data notshown). Inadditionto

Mol,

1B4 treatmentof

PMN as compared to isotype control antibody significantly augmented cAMP levels in response to FMLP stimulation (25.8±4.2netpmol

cAMP/10'

PMNwith1B4 ascomparedto

12.1±5.5 netpmol cAMPwith control

antibody,

P <0.01). These experiments were performed on PMN treated with

IBMX to inhibit the phosphodiesterase which breaks down

cAMP. Inprevious studies with FMLP, theuseofIBMX

re-sulted in an increasedsensitivity for the detection of cAMP

when examined over time (26). However, in the absence of IBMX,

significant increases

in cAMP could be detectedat

ear-lier time points (26). To assessthis,weexaminedLeul 5-and Mol-treated PMNwhich hadnotbeen treated withIBMX 5 minafter stimulation withFMLP.Thesametwofold increase innet cAMPlevels forMo

1-treated

PMN ascomparedto

con-trolLeu I5-treatedPMN wasobserved(datanotshown).

There-fore,

theuseofIBMXdidnotalterthequalitativeassessment

of the effect of these antibodiesonintracellular cAMP levels. These dataindicate that the ability

of

Mol and 1B4toinhibit the

CDl

lb/CD 1

8-independent

mechanism of

ingestion

may

be duetotheir effectontheaccumulation of cAMP levels in stimulated PMN. In conclusion, the MAb

experiments,

the studies with CGD PMN, and the studies with

IFN-y-treated

normalPMNdemonstratethat theinability of LADPMN to

increaseingestion inresponse toactivation isadirect result of the genetic

defect,

rather than aconsequence of the chronic inflammatory environmenttowhich thesePMN areexposed in

vivo.

Discussion

Careful study ofphagocytosisoverthepastfewyearshas

sug-gestedthat

optimal

phagocytic

function inPMNisrecruitedat

inflammatory sites. Whereas unstimulated ingestion byPMN

isaratherinefficientprocess,ingestionof equivalently

opson-ized

particles

canbestimulated bya

variety

of cytokines and

RGD-containing

adhesive proteins. Stimulated ingestion differs from unstimulated

ingestion

notonly in theextentand

efficiency

of the

phagocytic

process,butalso in the molecular pathways involved (4, 11).Becausecytokines and extracellular matrixproteinsareplentifulatinflammatory sites,the

phago-cytic

mechanisms involved in stimulated

ingestion

are

likely

to

predominateatsitesof inflammation and infection.The data

presented in this paperdemonstrate that, whereas unstimu-lated phagocytosis is

CDl

lb/CD

18 independent, a

CDl lb/

CD

18-dependent

mechanism of phagocytosis is recruited in

PMNstimulated with multiplepro-inflammatory signals. This isverysimilartothe recentstudies of the mechanisms of adhe-siontoendothelium used byPMNundervarious environmen-talconditions. Whereas adhesion by

unactivated

PMNis inde-pendent ofCDI

lb/CD

18,

FMLP

stimulationengages an

adhe-sion mechanism which is dependentonthis glycoprotein (21). Thus, activation ofPMN recruits a majorrole for

CDl

lb/

CD18 inbothadhesionandphagocytosis.Ourdata on

phagocy-tosis show that the rolefor

CD1

lb/CDl

8isat a step beyond target

binding

and may berelatedto anactivity for this integrin in the cytoskeletal rearrangements that occurduring

stimu-lated

ingestion.

It is

intriguing

to

speculate

that the

parallel

recruitment of

CDl lb/CD

18in both adhesion and

ingestion

byactivated PMNsuggests asimilarnon-ligand-binding role for CDl lb/CD1 8 in thespreadingof activated PMN on

endo-thelium.

The purposeoftheexperimentsreportedherewas to deter-mine whether CD11/CD 1 8-deficientPMNfrompatients with

LADwould becapable ofrecruiting theingestion mechanisms expected topredominateatinflammatoryfoci. These studies were performed with targets opsonized with IgG and C4b.

Theseopsonins bindto Fc receptorsandCR1 and donot de-pend on CDllb/CD18 for

ligand

binding.

As reported by others(14), LAD PMN bind EIgG normally and ingest equiva-lentamounts as normal PMNin the absence of stimulation. However,LAD PMN werecompletely unresponsiveto stimula-tion with phorbolesters,cytokines,orRGD-containing

adhe-sive

proteins (Figs.

1 and2).

This defect

wasdemonstrable for bothIgG and C4bopsonins. ThusLAD PMNlacked theability

to respond to inflammatory stimuli to amplify phagocytic function andwould be expected to exhibita marked

phago-cytic

defectat

inflammatory

sites in vivo.

Weexamined several alternative hypothesestoexplain this defect inLAD PMNphagocytic function.An

important

con-cern wasthat thefailuretostimulate

ingestion

was aresultnot

ofthe basic genetic defect but secondarytothechronicor recur-rentinfections which occurinthis disease. If thePMNwere

activated byinflammatory stimuli in vivo, the signals for stimu-lating ingestion mightalready have been encountered and the

resultantingestionbyunstimulatedLAD PMNwould use

mo-lecularmechanisms similartoinvitro activated normalPMN.

Alternatively,thecirculatingPMNmightrepresent a

popula-tion which, having been subjectedtopro-inflammatory

cyto-kines invivo,wererenderedunresponsivetootherstimulatory

signals.

Neither of these isalikely explanation forourresults

for severalreasons.

First,

thepatternof

pharmacologic

inhibi-tion of EIgG ingestionwas identicalforLADPMNand

un-stimulated normal PMN

(Fig.

3).

Ingestion

by unstimulated

normal PMNisunaffected byH7, aPKCinhibitor,butis inhib-ited byDiBcAMP, whereas phorbol ester-stimulated ingestion

exhibits

the

opposite

patternof inhibition. IfLADPMN were

upregulated in

vivo, IgG-mediated

phagocytosis by thesePMN

would beinhibited by H7 butresistanttotheeffects of

DiB-cAMP. In

fact,

theoppositewasthecase,demonstratingthat LAD PMN use amolecular mechanism for ingestion equiva-lenttothat used byresting, unstimulated,normal PMN.

Inaddition, other patients withrecurrentinfections exhib-itednormalactivation forenhancedingestioninvitro.We

ex-amined

ingestion

byPMNfrom patients with CGD, another genetic disease withapredispositionto recurrentbacterial

in-fections. These patients representedapopulationwhich, like theLAD

patients,

weresubjecttochronicor recurrent

infec-tions,

and theconsequentgeneration of pro-inflammatory

cy-tokines in vivo. Evidence forexposureof bothLADand CGD

PMN to

inflammatory

signals in vivowasgivenby thevariable

presence ofFcRI on these patients' PMN. Normal resting PMN expressminimalamountsofFcRI, butits expression can

beinduced in vitro byexposureofPMN to

interferon-y

(8)and

itcanbedetectedonthesurface ofPMNisolated frompatients with infectious diseases (25). CGD PMN (Fig. 6) aswell as

interferon-y-stimulated

normal PMN(datanotshown)were

abletoincrease

ingestion

of

EIgG

inresponseto RGD-contain-ing adhesive proteins,eventhough theyexpressedFcRI.These

datashow thatexposureto

pro-inflammatory cytokines

invivo did not affect the

ability

of

circulating

PMN to

respond

to

(10)

phagocytic function is unlikelyto be theconsequence of expo-sure toinflammatory signalsinvivo.

If the failure to amplify ingestion in response to stimulation was a consequenceof the absence of CD 1 b/CD 18 antigens, then treatment of normal PMNwith monoclonal antibodies

against CDI lb or CD1 8 might reproducethe LAD phenotype. Indeed, our dataindicatethat two MAbrecognizing CD1 lb,

OKM1 andOKM 10, inhibited theability ofnormal PMN to augmentingestion in response tostimulation. Neither MAb

affected unstimulated ingestion of EIgG (Fig. 6). ThusOKM

1-and OKM10-treated normal PMNbehavedin vitro like un-treated LAD PMN. We asked if this was a property of the

particular epitope recognized by these MAbs, becauseweand

othershavepreviously reportedthat someantibodies (suchas IB4and Mol)inhibitbothunstimulatedandstimulated inges-tion (PKC independent anddependent) (15-17). We postu-lated thatthe antibodies that inhibited both mechanisms of phagocytosis might actually inducethegeneration ofasignal inhibitory to ingestion. The one signal we had determined previously to negatively affect unstimulated ingestion was

cAMP, presumably working through PKA (3). When PMN wereincubated withHA1004toinhibitPKA, MAb-IB4 treat-mentof thosePMNnolongerinhibited unstimulatedingestion of EIgG, whereasit retained theabilitytoinhibit PDBu-stimu-latedingestion (Figs.7and8).BecauseMAbswhichrecognize both CDl lb andCD1 8 could reproduce the LAD defect in vitro, and the defectwasnot adirectconsequenceofthe in vivo

exposure to inflammatory signals, our data indicate that the

phagocyticdefect inLADisadirectconsequenceofthegenetic abnormality.

Inaddition,we assessed theability ofMoIand IB4 to aug-ment intracellular levelsofcAMP. Both Mo1 and 1B4 aug-mented intracellular cAMPlevels in response to stimulation witheither immune complexesor FMLP(Fig. 9).Eventhough

mostofourexperimentswere

performed

with IBMX-treated

PMN,identical fold increases innetcAMP could be observed withuntreated PMN atearlier time points.Therefore, the ef-fects ofMo 1 and 1 B4were notdueto anartifact of theIBMX treatment. Our data donotprove

directly

thatMol and 1B4

inhibit unstimulatedingestion by activation ofPKA. Conclu-sive

proof

will

require

elucidation ofthemechanism

by

which cAMPinhibits unstimulated

ingestion (i.e.,

phosphorylation

of a

specific protein)

and

showing

identical effects of 1B4 and Mo

1.

Recently published workindicatesthat 1B4 treatment canincreaseintracellular cAMP levels

by

preventing

CDl

lb/

CDl 8-dependent decreases in cAMP

(27).

It is

possible

that Mo 1 and 1B4 areworking inourstudies byasimilar

passive

mechanism. Our datamostclosely parallel studies

performed

with anti-CD29 treatmentof lymphocytes (28). Increasesin intracellularcAMP wereobservedupon MAb engagement of CD29, arelated

integrin

,

chain, expressed

by

lymphocytes,

butonlywhen the cellswereco-stimulatedwith anti-CD3

(28).

Thus, stimulation ofPMNwithEIgGinthepresenceofeither

Mo 1 or1B4

during

the

phagocytosis

assay could resultin local accumulations ofcAMP withactivation ofPKAand

genera-tionofa

signal

inhibitory for

PKC-independent ingestion.

The

significance

of this observationtoothereffects of Mol and 1B4

is unknown; however, it is

interesting

tospeculate that

anti-CDl lb/CD

18 activation ofPKA may

explain

some ofthe

anti-inflammatory

effects of the

anti-,B

chain antibodies in

various pathological conditions(29, 30). Itis equallyintriguing

to

speculate

that

ligand

engagementof

CDl lb/CD

1

8-bearing

receptors could also stimulate cAMP

accumulation,

which

may mediate via PKA some ofthe pleiotropic biological effects of thisintegrin family.

Because CDl lb/CD18-deficient PMN failed to demon-strate PKC-dependent phagocytic mechanisms, we investi-gated the ability of LAD PMN to activate PKC and to demon-strate otherPKC-dependent functions. UsingPKC transloca-tion as an assay forPKC activation, we foundthat phorbol esterstranslocatedPKC normally in LAD PMN (Fig. 4). Thus,

failure of PKC translocation could not explain the lack of phagocyticresponseofLAD PMN tophorbol esters. We also examined theability ofLAD PMN to generate arespiratory

burst in response to phorbol esters, another cell function thoughttobedependentonPKCactivation.Wefound,as have

others, that the generation of superoxide anion inresponse to

phorbol ester stimulation as well as FMLP stimulation oc-curred but wasquantitatively variableandwithoutaconsistent

patternofresponse. Someofthevariabilitymay represent

dif-ferences in prior in vivo exposure to inflammatory signals. Nonetheless, thesedata showed that there was no generaldefect in PKC-dependentresponsesinLAD PMN.Therefore,the in-ability ofLAD PMN torecruit phagocytic mechanisms

depen-dentonPKCactivation didnotarise fromafailuretoactivate

PKC orfromageneralfailure ofPKC-dependentsignal

trans-duction.

In summary, our data demonstrate a specific defect in phagocytosis byLADPMN. LAD PMNfailedtoamplify

in-gestion

in response tostimulation by phorbolesters,cytokines including TNF-a, and RGD-containing adhesive proteins. Theseexperiments indicate thata rolefor

CDl

lb/CD1 8is

es-sential for the

phagocytic

mechanism that

predominates

at

in-flammatory

foci:

i.e., ingestion

stimulated by

inflammatory

signals.

Our data indicatethat the

CDl

lb/CD18-dependent mechanism of ingestion isalso dependent on theactivationof PKC. Thissuggeststhat PKCmayplayanimportant role inthe

regulation of CDl lb/CD18 function, either by direct phos-phorylation ofthe

,3

chain (31),orindirectly by phosphoryla-tion ofa

cytoplasmic protein

which alters the conformationor

cytoskeletal association ofthe plasma membrane heterodimer. Theimportance of PKC-dependentand

CDl

lb/CD

18-depen-dent

phagocytic

mechanismstohost

defense

is demonstrated bythefactthat PMNfrom

patients

with diseases witha

predis-position

toseriousrecurrentinfection

(CGD, paroxysmal

noc-turnalhemoglobinuria,andLAD) failtoexhibitsome orallof

thesemechanismsfor the

amplification

of

ingestion.

The inabil-ity of phagocytesto normally upregulate their

ingestion

most

likely contributesvery

significantly

totheincreased

propensity

toinfection in

patients

withLAD. In supportofthis

possibility

is the clinical observation that LAD

patients

get

significant

bacterial

pneumonias, despite apparently

normal

ingress

of

PMNtothis siteof infection (Dr. Anderson,

personal

observa-tion).

Acknowledgments

This work wassupported bygrants AI-23790 (Dr. Gresham), GM-38330 (Dr. Brown), and AI-23521, AI-19031, DE-07875,and HL-11422 (Dr. Anderson) from theNational Institutes of Health, bya

postdoctoralfellowshipfrom theArthritis Foundation(Dr.Graham),

andbytheMedical ResearchService, Departmentof VeteransAffairs

(Dr.Gresham).

References

1.Pommier,C.G.,J.O'Shea,T.Chused,K.Yancey,M.Frank,and E. J.

(11)

leukocytes:morphologicandfunctional characterization. J.Exp.Med.

-159:137-151.

2.Gresham, H. D.,L.T.Clement,J. E.Lehmeyer,F. M.Griffin,and J. E. Volanakis. 1986.Stimulation of humanneutrophil Fcreceptor-mediated phago-cytosis by alowmolecularweight cytokine.J.Immunol. 137:868-875.

3.Gresham, H. D., L. T. Clement, J.E.Volanakis,and E.J. Brown. 1987. Cholera toxin andpertussistoxinregulatethe Fcreceptor-mediatedphagocytic

response of humanneutrophilsina manneranalogoustoregulationby monoclo-nalantibody1C2. J.Immunol.139:4159-4166.

4.Gresham,H.D.,J. A.McGarr,P.G.Shackelford,andE. J. Brown.1988. Studies on the molecular mechanisms ofreceptor-mediatedphagocytosis: ampli-fication of ingestion is dependent on thegeneration ofreactive oxygen metabo-lites and isdeficient in PMN from patients with chronic granulomatous disease. J.Clin.Invest.82:1192-1201.

5.Gresham, H. D., J. L.Goodwin,P.M.Allen,D.C.Anderson, and E. J. Brown.1989.Anovel memberofthe integrin receptorfamilymediates Arg-Gly-Asp-stimulatedneutrophil phagocytosis.J.CellBiol. 108:1935-1943.

6.Wright, S. D., and B. C. Meyer. 1986. Phorbol esters causesequential

activation and deactivation of complement receptorsonpolymorphonuclear leu-kocytes. J. Immunol. 136:1759-1764.

7.O'Shea, J. J., E. J. Brown,T. A.Gaither,T.Takahashi,and M. M. Frank. 1985.Tumor-promoting phorbol esters induce rapidinternalization of the C3b receptorvia a cytoskeleton-dependentmechanism. J.Immunol.135:1325-1330. 8.Perussia, B., M.Kobayashi,M. E.Rossi, I. Anegon, and G. Trichieri. 1987. Immuneinterferon enhancesfunctionalproperties of human granulocytes: role of Fc receptors and effect oflymphotoxin,tumornecrosisfactor,and granulo-cyte-macrophage colonystimulating factor.J.Immunol.138:765-774.

9.Klebanoff,S. J., M.A.Vadas, J. M.Harlan,L. H.Sparks, J. R. Gamble, J. M.Agosti, and A. M. Waltersdorph. 1986.Stimulation ofneutrophils by tumor necrosis factor. J.Immunol.136:4220-4225.

10.Brown, E. J., A. M. Newell, and H. D. Gresham. 1987. Molecular regula-tion ofphagocytefunction:evidenceforinvolvementofaGTPbinding proteinin

opsonin-mediatedphagocytosis by monocytes. J.Immunol.139:3777-3782. 11.Gresham,H.D., A.Zheleznyak, J. S. Mormol, and E. J. Brown. 1990. Studiesonthe molecularmechanisms ofhumanneutrophilFcreceptor-mediated

phagocytosis: evidence that a distinct pathway foractivation of the respiratory burst results inreactive oxygen metabolite-dependent amplification of ingestion.

J. Biol.Chem. 265:7819-7826.

12.Gresham, H. D., and E. J. Brown. 1989. Molecularmechanism offibronec-tin-stimulatedneutrophil phagocytosis.J.LeukocyteBiol. 46:3 10.(Abstr.)

13. Anderson, D.C., andT. A. Springer. 1987. Leukocyte adhesion defi-ciency:aninherited defect in theMac-i,LFA-1,p150,95glycoproteins. Annu. Rev.Med. 38:175-194.

14.Anderson, D.C., F. C. Schmalstieg, M. J. Finegold, B. J. Hughes, R. Rothlein, L. D. Miller, S. Kohl, M. F. Tosi, R. L. Jacobs, T. C. Waldrop, et al. 1985. The severe and moderate phenotypesof heritableMac-l,LFA-1,p150,95

deficiency: theirquantitativedefinition and relation to leukocyte dysfunction and clinical features. J.Infect.Dis. 152:668-689.

15. Arnaout, M. A., R. F. ToddIII,N. Dana, J.Melamed,S. F. Schlossman, andH. R.Colten.1983. Inhibition ofphagocytosisofcomplement C3- or immu-noglobulin G-coatedparticles and C3bi binding by monoclonal antibodies to a monocyte-granulocyte membraneglycoprotein(Mol).J.Clin. Invest. 72:171-179.

16. Brown,E.J., J.F.Bohnsack, and H. D.Gresham. 1988. Mechanism of

inhibition-ofimmunoglobulinG-mediatedphagocytosis bymonoclonal

antibod-iesthatrecognizetheMac-lantigen.J.Clin. Invest. 81:365-375.

17.Graham, I. L., H. D.Gresham,and E. J. Brown. 1989. An immobile subsetof plasmamembrane CDIlb/18(Mac-l)is involved inphagocytosisof targetsrecognized by multiplereceptors. J. Immunol. 142:2352-2358.

18.Bohnsack, J.F., T.Takahashi, andE.J.Brown. 1986.Interaction of culture-derivedmacrophageswiththefibroblast-bindingdomain of fibronectin is

anecessary but inefficientsignalfor fibronectin enhancement of CR1-mediated

phagocytosis.J.Immunol.136:3793-3798.

19.Bohnsack, J. F., H.K.Kleinman,T.Takahashi,J.J.O'Shea,and E. J. Brown. 1985. Connective tissueproteinsandphagocyticcellfunction. J.Exp.

Med. 161:912-923.

20.Melloni,E., S.Pontremoli,M.Michetti,0.Sacco,B.Sparatore,and B.L. Horecker. 1986. Theinvolvementof calpainin the activation ofproteinkinasec

inneutrophilsstimulated by phorbolmyristicacid. J.Biol. Chem. 261:4101-4105.

21.Smith,C. W., S. D.Marlin,R.Rothlein,C.Toman,and D.C. Anderson. 1989.Cooperativeinteractions ofLFA- Iand Mac-l with intercellular adhesion molecule-l infacilitatingadherence andtransendothelialmigrationof human

neutrophilsinvitro. J. Clin.Invest.83:2008-2017.

22.Nauseef,W.M.,P.deAlarcon, J.F.Bale, and R.A.Clark. 1986. Aberrant activation andregulation oftheoxidative burst inneutrophilswithMo1

glyco-proteindeficiency.J.Immunol. 137:636-642.

23.Nathan, C., S.Srimal,C.Farber,E.Sanschez,L.Kabbash,A.Asch,J.

Gailit,and S. D. Wright. 1989.Cytokine-inducedrespiratoryburstof human neutrophils: dependenceonextracellular matrixproteinsandCDIl/CDI8

inte-grins.J.CellBiol. 109:1341-1349.

24.Shappell, S. B., C. Toman,D. C.Anderson, A. A.Taylor,and C. W. Smith. 1989.Adherence-dependentneutrophil hydrogenperoxidesecretion is mediated by Mac-l(CD1lb/18).J. Immunol. 144:2702-2711.

25.Simms,H.H.,M.M.Frank, T. C.Quinn,S.Holland,andT. A.Gaither. 1989. Studiesonphagocytosisinpatientswithacutebacterialinfections.J.Clin.

Invest.83:252-260.

26.Verghese,M.W.,K.Fox,L.C.McPhail,and R.Snyderman.1985. Che-moattractant-elicited alterations in cAMP levels in humanpolymorphonuclear

leukocytesrequireaCa2"-dependentmechanism which isindependentof

trans-membraneactivationof adenylate cyclase. J.Biol.Chem. 260:6769-6775. 27. Nathan,C.,and E.Sanchez. 1990. Tumor NecrosisFactorandCD 1/

CD18integrinsactsynergisticallytolowercAMPin humanneutrophils.J.Cell

Biol. 111:2171-2181.

28.Groux,H., S. Ituet, H.Valentin,D.Pham,andA.Bernard. 1989.

Sup-pressor effectsandcyclicAMPaccumulationbytheCD29 moleculeof CD4+

lymphocytes.Nature(Lond.).339:152-154.

29.Vedder,N.B.,R. K.Winn,C.L.Rice,E.Y.Chi,K.-E.Arfors,and J.M.

Harlan. 1988.Amonoclonalantibodytotheadherence-promoting leukocyte glycoprotein, CD18,reduces organinjuryandimproves survival from

hemor-rhagicshock andresuscitation in rabbits. J.Clin.Invest.81:939-944. 30.Tuomanen,E.I.,K.Saukkonen,S.Sande,C.Cioffe,andS.D.Wright.

1989.Reduction ofinflammation,tissue damage, andmortalityinbacterial

men-ingitisinrabbitstreated with monoclonalantibodiesagainst adhesion-promoting

receptors ofleukocytes. J.Exp.Med.170:959-968.

31.Chatila,T. A., R.L.Geha,andM. A. Arnout. 1989. Constitutive and stimulant-induced phosphorylationofCD1 1/CD18 leukocyteadhesion

References

Related documents

She also served as the President of the Latinx Law Student Association, the Diversity Representative of the Student Bar Association, as a Staff Editor on the Seattle Journal

Surgical Site Infection Reduction Through Nasal Decolonization Prior to Surgery..

Institute Warranties (1 July 1976), American Institute Trade Warranties (1 July 1972) and the International Navigating Conditions (1 November 2003) are largely (but

trepreneurs with the previous start-up experience and first time company founders to highlight positive effect of education, experience and knowledge on start-ups. To better

Figure 2 indicates the representative chromatographic peaks of FFAs in the sediment samples in both seasons while the free fatty acid levels of the river sediments

Typical motor characteristic curve with nominal voltage and optimal motor controller. Observe the maximum permissible rotational speeds of add-on and installation components (such

The aim of this study was to evaluate the genotypes and allele frequencies of the rs749292 and rs700518 polymorphisms of the CYP19A1 gene in Polish postmenopausal women

The infoffi1ation obtained in the interviews that were undertaken as part of the case studies for three of the four sectors (the Irish mechanical engineering sector analysis