Allelic polymorphisms of human Fc gamma receptor IIA and Fc gamma receptor IIIB Independent mechanisms for differences in human phagocyte function

Allelic polymorphisms of human Fc gamma

receptor IIA and Fc gamma receptor IIIB.

Independent mechanisms for differences in

human phagocyte function.

J E Salmon, … , N L Brogle, R P Kimberly

J Clin Invest. 1992;89(4):1274-1281. https://doi.org/10.1172/JCI115712.

Two different allelic polymorphisms among the isoforms of human Fc gamma receptors have been defined: the low-responder (LR)-high-responder (HR) polymorphism of huFc gamma RIIA expressed on both PMN and monocytes and the NA1-NA2 polymorphism of the neutrophil Fc gamma RIII (huFc gamma RIIIB). To address the issues of whether the LR-HR polymorphism has a significant impact on Fc gamma R-mediated functions in human blood cells and whether any differences in LHR might be related to higher Fc gamma R-mediated phagocytosis in NA1 donors, we examined Fc gamma R-specific binding and internalization by donors homozygous for the two huFc gamma RIIA alleles. PMN from LR homozygotes showed consistently higher levels of internalization of erythrocytes opsonized with pooled human IgG (E-hIgG). The absence of an LR-HR phagocytic difference with erythrocytes opsonized with either anti-Fc gamma RIIA MAb IV.3 or rabbit IgG, as opposed to E-hIgG, suggested that the Fc piece of the opsonin might be important for this LR-HR difference. Accordingly, we studied HR and LR homozygotes with human IgG subclass-specific probes. Both PMN (independent of huFc gamma RIIIB phenotype) and monocytes from LR donors bound and internalized erythrocytes coated with human IgG2 (E-hIgG2) efficiently, whereas phagocytes from HR donors did so poorly. E-hIgG2 internalization was completely abrogated by blockade of the ligand binding site of huFc gamma RIIA with IV.3 Fab, […]

Research Article

Find the latest version:

http://jci.me/115712/pdf

Allelic Polymorphisms

of Human

Fcy

Receptor

11A and

Fcy

Receptor

111B

Independent Mechanismsfor Differences in Human Phagocyte Function

Jane E.Salmon,Jeffrey C. Edberg, Nina L. Brogle, and Robert P. Kimberly

DepartmentofMedicine, The Hospitalfor Special Surgery and The New York Hospital, TheCornell UniversityMedical College, New York 10021

Abstract

Two different allelicpolymorphisms among the isoforms of hu-manFcyreceptors have been defined: the low-responder (LR)-high-responder (HR) polymorphism ofhuFcyRIIAexpressed on both PMN and monocytes and the NA1-NA2 polymor-phism of the neutrophilFcyRIII (huFcyRIIIB).Toaddress the issues of whether the LR-HR polymorphism has a significant

impacton

Fc'yR-mediated

functionsin human blood cells and whether anydifferencesin LR-HR might be related to higher

Fc'yR-mediated

phagocytosis in NAI donors, we examined

Fc'yR-specificbinding and internalization by donors homozy-gous for the twohuFc-yRIIAalleles. PMNfromLR homozy-gotes showed consistently higher levels of internalization of erythrocytesopsonizedwith pooledhumanIgG (E-hIgG).The absenceof an LR-HRphagocytic differencewith erythrocytes

opsonizedwitheitheranti-FcyRIIAMAbIV3orrabbitIgG,

asopposed toE-hIgG,suggested that the Fcpieceof the op-soninmightbeimportant forthisLR-HRdifference. Accord-ingly, we studied HR and LR homozygotes with human IgG

subclass-specific probes. Both PMN (independent of

huFctyRIIIB phenotype)and monocytes from LR donors bound and internalized erythrocytes coated with human IgG2 (E-hIgG2)

efficiently,

whereasphagocytesfrom HR donors didso

poorly.E-hIgG2 internalizationwascompletely abrogated by blockade of the ligand bindingsite of

huFc'yRIIA

with IV3 Fab,indicating that

huFc'yRIIA

is essential for thehandling of hIgG2 and that the mechanism of the LR-HR phagocytic dif-ference isatthelevel ofligandbindingtohuFcyRIIA.In

con-trast, thedifferenceininternalization ofE-hIgGbetweenNA1 and NA2 homozygous donors was independent of the

huFc'yRIIA

phenotype and did not manifest differencesin

li-gand binding. Thus,the two knownallelic polymorphisms of human

Fc'yR

have distinct and independent mechanisms for

altering receptor function, which may influence host defense and immune complex handling. (J. Clin. Invest. 1992.

89:1274-1281.) Key words: IgG subclasses - phagocytosis .

monocyte *neutrophil

Introduction

Two different allelic polymorphisms among the isoforms of

humanFcyreceptors have beendefined. TheNA1-NA2

poly-Addressreprintrequests to Dr. Salmon, The Hospitalfor Special Sur-gery, 535 East 70thStreet,NewYork, NY 10021.

Receivedfor publication 14June1991and inrevisedform 6

De-cember 1991.

morphism of the neutrophil FcyRIII

(huFcyRIIIB)'

reflects

several differences in amino acid sequence and two potential

glycosylation sites (1-5). The low-responder

(LR)-high-re-sponder (HR) polymorphism of huFcyRIIA reflects two

amino acidsubstitutions atpositions27and131, with an

argi-nine residueataminoacidposition 131 being critical for

effi-cientbinding of mIgG1 (6-9). For each ofthese allelic systems, there has been substantial interest in identifying their func-tional consequences and the molecular mechanisms

underly-inganydifferences.Wehave shown that individuals homozy-gous for the NA1 and NA2 alleles ofhuFcyRIIIBhave distinct

phagocytic capacities(10).

The recentobservationby Warmerdametal.(8) that

posi-tion 131 ofFcyRIIA expressedintransfected fibroblasts

influ-encesthecapacity forbindinghuman IgG2 (hIgG2)suggests thatthispolymorphism may also beimportantinhuman sys-tems.Usingalarge donor populationwith definedhuFcyRIIA alleles,weexamined the effect of thispolymorphismon quan-titative phagocytosis by human phagocytes. Our results

re-vealedthat homozygousLRdonors haveahigher capacityto

internalize erythrocytes (E) opsonized with human IgG (E-hIgG)thanhomozygousHRdonors.Furtheranalysisof these homozygous donorsindicated that,whenstudiedwith human IgG (hIgG) subclass-specific probes, PMN and monocytes from bothLRand HRsubjectsareabletobind andinternalize E-IgGl comparably.WithE-IgG2,both cell types from homo-zygous LR subjects bind and internalize efficiently whereas cells fromHRdonorsdo sopoorly. Thus,with intact human

phagocytes, theLR-HRpolymorphisminfluences bothligand bindingtohuFcyRIIAandsubsequentFcyR-mediated phago-cytosis.

This difference in FcyR-mediated phagocytosis, coupled withrecent data localizingthe genesforboth huFcyRIIand

huFc'yRIII

tohumanchromosome 1(1

1-13),

raised the

possi-bilitythatanassociationbetween the NA alleles ofhuFcyRIIIB

and the LR-HR alleles ofhuFcyRIIAmight explainthe differ-encein

Fc'yR-mediated phagocytosis

betweenNA1

homozy-gous and NA2 homozygousdonors.

Thus,

we also examined

1.Abbreviations usedin this paper:Al, adherence index; ANOVA, analysisofvariance;E,erythrocytes;EA,IgG-sensitizedbovine erythro-cytes;EB, biotinylated erythrocytes; EBA,streptavidin-coated EB; E-hIgG, E coated withpooled hIgG; E-hIgGl, E coated with human

IgGI;E-hIgG2,Ecoated with humanIgG2; E-hIgG3,Ecoatedwith humanIgG3; E-hIgG4,Ecoatedwith human IgG4;E-IV.3,Ecoated with IV.3F(ab')2;hIgG, humanIgG containing mixed subclasses; HR,

high-responderalleleof huFcyRIIA;huFcyR,receptorsforFcportion

ofIgGin human cells; huFcyRIIA,40-kDreceptoronhuman

neutro-philsand monocytes for Fcportionof IgG; huFcyRIIIB, 50-78-kD

receptoron human neutrophilsfor Fc portion of IgG; LR,

low-re-sponderalleleofhuFc'yRIIA; MCF, meanchannel logfluorescence

units;mIgGl,murineIgG1 myelomaprotein; PE,phycoerythrin;PI, phagocyticindex.

1274 J.E.Salmon,J. C.Edberg,N. L.Brogle,andR. P.Kimberly

J.Clin. Invest.

© The American Society for Clinical Investigation, Inc.

0021-9738/92/04/1274/08 $2.00

the differences inphagocytosisforNAl and NA2homozygotes

matched for expression of the LR-HRpolymorphism.

Phago-cytosisof E coated withhIgGwashigherfor NAl donors

inde-pendent of theLR-HRphenotype. Furthermore,therewere

nodifferences in E-hIgGbinding by cellsexpressingeitherNAl or NA2 alleles. The mechanism of thephagocytic difference, therefore, appears to be distal to ligand binding. These data

demonstratethat the tworecognizedallelicpolymorphismsof humanFcy receptors haveindependentmechanisms for

me-diating functionally apparent differences in FcyR-mediated

phagocytosis. These differencesmayhaveimportant

implica-tionsforhumandisease.

Methods

Subjects. Peripheralbloodwascollected from 32 disease-free

volun-teerswhorangedin age from 20to56 yr(34±8yr[mean±SD]). Proto-colsfor these studieswereapproved bythe InstitutionalCommitteeon

HumanRightsin Research.

Determination ofNA antigensbyleukoagglutination. Typing for neutrophil-specific antigens, NAI andNA2, wasperformed by

leu-koagglutinationasdescribedbyLalezari (2, 14),withapanelof

anti-NA1andanti-NA2alloserakindly provided byDr.DavidStroncekof

theAmericanRedCrossNeutrophil Serology Reference Laboratory, St. Paul,MN.TheassignmentofNA typewasconfirmed by immuno-precipitationandflowcytometrywithMAbsCLB-FcRgran 1,

CLB-gran 11, and GRMI(15, 16).

DeterminationofLR-HRallelesbyflowcytometry.Phenotyping of

donors for the LR-HRallelesof huFcyRIIAwasperformed by quanti-tative flowcytometryusingMAbs41.H16 and IV.3 asdescribed by Gosselinetal.(17). Phenotypic assignmentwascorroboratedby

anti-CD3mitogenesisassaysasdescribedbelow.

Reagents.HBSS, RPMI1640, and FCS werefromGibco

Laborato-ries,Grand Island, NY. FCS was heatinactivatedat56°Cfor60 min

before use. FITC-conjugated rabbit anti-human IgG F(ab')2

FITC-conjugatedgoatanti-rabbit IgG F(ab')2,andphycoerythrin

(PE)-con-jugatedgoat anti-mouseIgGF(ab')2waspurchased fromTago, Inc., Burlingame, CA. Allconjugatedreagents wereabsorbedagainst

hu-man mononuclear cells before use. Sulfosuccinimidobiotin (Sulfo-NHS-biotin),sulfosuccinimidyl-6-(biotinamido)hexanoate (NHS-LC-biotin), and streptavidin were obtained from Pierce Chemical Co., Rockford, IL. FMLP(SigmaChemicalCo.,St.Louis, MO)was

dis-solved inAmerican ChemicalSociety gradeDMSOforastock

concen-trationofl1-0 M and storedat-20°C.

MAb3G8(mIgGl) recognizing huFcyRIII (CD16)waskindly pro-vided by Dr. J. Unkeless, Mt. Sinai Medical Center, New York(18).

The MAb IV.3 (mIgG2b) and IV.3 Fab fragments recognizing

huFcyRII (CD32)and 3G8 F(ab')2 werepurchased from Medarex,

Inc., (WestLebanon,NH) (19).Silver stain analysis ofthe IV.3Fab

fragment preparationand the3G8 F(ab')2preparation indicatedthat there was no intact IgG. The MAb 41H. 16 (mIgG2a) recognizing

huFc'yRII wasgenerouslyprovided Dr.ThomasZipf, Universityof TexasCancer Center, Houston, TX (20). Leu4 and OKT3(IgGl and

IgG2a anti-CD3, respectively) wereprovided by Dr. Robert Evans, Memorial Sloan-Kettering Cancer Center, New York. Isotype controls

(mIgGl [MOPC21] and mIgG2a [UPCI0]) were purchased from

SigmaChemicalCo.

PurifiedhIgGl and hIgG2 myeloma proteins were obtained from Dr.Elliot Osserman of Columbia University, New York. hIgG3 and hIgG4 myeloma proteinswereobtainedfrom The Binding Site, Inc.

(SanDiego, CA).Asecondsetofpreparations of

hlgGl,

hIgG2, hIgG3, and hIgG4 were obtained from Calbiochemic-Behring Corp., San

Diego, CA. Normal pooled hIgGwasobtained from Sigma Chemi-calCo.

Preparationofcells.Freshanticoagulatedhumanperipheralblood wasseparated bycentrifugation throughadiscontinuous two-step

Fi-coll-Hypaque gradient(10).PMNswereisolatedfrom the lower

inter-face and washed with HBSS.ContaminatingEwerelysedwith hypo-tonic saline(0.02%NaCl)for 20sfollowedby0.16% NaCl andafinal wash with HBSS. Mononuclear cellswere isolated from the upper

Ficoll-Hypaque interface and washed with HBSS. Thepercent

mono-cytes within the mononuclear celllayerwasdeterminedby peroxidase staining.Afterfinal washes,all cellswere_resuspendedto5X106 PMNs

ormonocytes/ml.

PreparationofE.Antibody-coatederythrocytes (EA)wereprepared byincubatingbovine E with rabbitIgGanti-bovine Eantibody

(Cap-pel Laboratories, Cochranville, PA)for 1 hat 370C. The cellswere

washed andresuspendedatI0O cells/mlin RPMI and 20% FCS(21).

Thestandardamountofantibodyusedwas afourfold dilution of the

minimumagglutinating titerandresultedin maximumphagocytosisof EAbymonocytes.

Ewerecoupledtopooledorsubclass-specific hIgG bya biotin-avi-dintechnique (22).To preparebiotinylatedE(EB),0.5 ml ofE(1 X109

cells/ml)in0.1M carbonate buffer(pH 8.6)wasincubated with

sulfo-NHS-biotin(500 g/ml) for20 minat4VCwithmixing,followedby

three washes withbuffercontainingdivalentcationsasfollows(mM):

(2.5 veronalbuffer, pH 7.4; 146NaCl,0.05%gelatin,0.15CaC12,and 0.05MgCl2).EBat1X109/mlwereincubated withanequalvolume of

streptavidin(250 g/ml)for 30 minat4VCwithmixing.The

streptavi-din-cbatedEB(EBA)werethen washed three times with the buffer and

resuspendedto1 X 109 E/mlfor immediateuse.PooledhIgGor puri-fiedhumanmyelomaproteinsofspecific IgGsubclasseswere

biotiny-latedwith NHS-LC-biotin(0.01 mgbiotin/mgprotein)for 60 minat roomtemperaturewith occasionalmixing. Tobind thebiotinylated

hIgGtotheEBA, EBA (12.5 Al at 1 X 109/ml) werecombined with

varyingamountsofbiotinylatedprotein(200 g/mlto1mg/ml)for 30

minat4°C withmixing. Afterthreewashes,thehIgG-coated EBA

(E-hIgG)werethenresuspendedin 125

Al

(1XI0O E/ml)andused

immedi-ately. Forcertainexperiments,theEBAwerecoupledtobiotinylated

IV.3 Fab. Theresults ofexperiments using E-IV.3 prepared bythe

biotin-avidinsandwichtechniquewerecomparabletothoseusing E-IV.3 Fabheteroantibodies(23-25).

Assayofphagocytosis.Quantitation ofphagocytosisbyPMNs and monocyteswasperformedaspreviouslydescribed(10,21, 26).Briefly,

cellswere resuspendedin RPMI with20% heat-inactivated_IgG-free FCS(GibcoLaboratories),at5X 106 cells/ml.Incertainexperiments,

the PMNswerepreincubated with eitherIV.3 Fabor3G8 F(ab')2(30 ,g/ml)for 5min;the MAbswerepresentthroughoutthe assay of phago-cytosis.Forexperiments withFMLP-treated PMN(i0-7M),FMLP

wasaddedsimultaneously withthe phagocyticparticle.

Toassess internalization ofE targetparticles,cells (100Ml)were combined with EA(250Ml),E-IV.3(250Ml),orE-hIgG (125Ml).The

leukocyte-erythrocyte mixtureswerecentrifugedat44 gfor3 min and thenincubatedat37°C for15mintoallowformaximum internaliza-tion. After hypotonic lysis of noninternalized E, phagocytosis was

quantitated by lightmicroscopy.Atleast 400cells/slidewerecounted in duplicate without knowledge of the donor huFc~yRIIA or

huFc~yRIIIB

allotype.Thedataareexpressed asphagocytic index (PI, numberofingestedE/100 phagocytes).For assayswithmonocytes,PI wascorrected for the percentperoxidase positivity of mononuclear cells.

Assayofadherence. Toquantitateadherence, PMN and E target

particleswere prepared and combinedas described in the assay of phagocytosisabove.After centrifugationat 44 gfor3 min, the PMN-E mixturesweremaintainedat roomtemperature for10 min and then gently resuspended. Adherence ofE to PMN wasquantitated by light

microscopy. Data are expressed asadherence index (AI, number of adherentorinternalizederythrocytes/100PMN). Under these

condi-tions,<1%ofphagocyteshadinternalizedE.

Immunofluorescentflow cytometry. Fresh leukocytes (5 X

₁₀₅

in

PBSwith 0.1%FCS)wereincubated with saturatingamountsof spe-cific MAborisotypecontrolsfor 30 minat4°C.Aftertwowasheswith

coldPBScontaining 1% FCS,cellswereincubated withsaturating

con-centrations of phycoerythrin (PE)-conjugated goat anti-mouse IgG

FCS. E-hIgG werestained with FITC-conjugated rabbitanti-human IgG F(ab')2, E coated with IV.3 Fab were stained withPE-conjugated

goat anti-mouseIgGF(ab')2,and Eopsonizedwith rabbit anti-ox E antibodies were stained with FITC-conjugatedgoat anti-rabbit IgG F(ab')2,followed bywashes with coldPBS/l%FCS.

After staining, cell-associated immunofluorescence was

quanti-tated on aCytofluorograf IIS(Becton DickinsonImmunocytometry Systems, Mountain View,CA)with a 2151 computer as previously

described(10, 15, 16). For each experiment, the instrument was

cali-brated withFITC-conjugated calfthymusnuclei (Fluorotrol-GF, Bec-ton,Dickinson& Co.,MountainView, CA) and quantitative PE

mi-crobeadstandards (FlowCytometryStandards Corp., Research

Trian-gle Park, NC)to allowforassessmentofboth absolute and relative levelsof immunofluorescence.

Anti-CD3proliferationassay.Peripheralblood mononuclear cells,

separated byFicoll-Hypaqueasdescribed above,wereresuspended at 1

X 106cells/mlinRPMI1640 with10%FCS(supplemented with gluta-mine,penicillin,andstreptomycin) and aliquoted into 96-well plates at

1 x I0O cells/well. OKT3(IgG2aanti-CD3, 5 _gg/mlfinal

concentra-tion),Leu4(IgGI anti-CD3,5ug/mlfinal concentration), nonspecific

control isotype antibodies, or mediaalone were added totriplicate

wells and the cells wereincubatedfor4d at370C.For thefinal8hr,

[3H]thymidine

(AmershamCorp.,ArlingtonHeights, IL),2

_,gCi/well

was added, after whichthe cells were harvested, washed, and

[3H]-thymidineincorporation measured inabeta counter(27-29).

Dataanalysis. For assessmentofphagocyticcapacity, all

experi-ments wereperformed inamatched-pairsexperimental design.

Accord-ingly,eachsubject, homozygous foragiven FcyR isoform(huFc'yRIIA

orhuFcyRIIIB),wasstudied in comparisonto asecondsubject,

homo-zygousforthe other alleleofthe sameisoformandmatchedforthe samephenotype ofthe otherFcyR (e.g.,HR-NA1 vs.LR-NA1).

Data aredisplayedasmean±SEM. PIfor specificprobes were com-paredusingapairedt test(twotailed). Differencesinphagocytic

capac-itybetween the groupsforEopsonizedoverarangeof densitieswere

compared by repeated measures analysis of variance (ANOVA). A

probability of0.05 was used torejectthe nullhypothesisthat thereisno

difference betweenthe groups.

Results

Characterization

ofhuFcyRIIA

and

huFcyRIIIB

alleles. The

allotypic characteristics of our normal population for

huFcyRIIIBweredetermined by serological typingand quanti-tative flowcytometry of PMN using NA1- and NA2-specific MAbs(10, 15, 16).The LR-HR allelesof huFcyRIIAinour

populationwereidentifiedbyquantitativeflow cytometry with

MAb 41H.16,whichrecognizestheHRalleleofhuFcyRIIA,

and MAbIV.3,whichrecognizesbothHRandLRalleles (17). Using the ratio of fluorescence intensity of 41H.16/IV.3 in

both monocytes andPMN,we

assigned

thehomozygousHR

phenotypeto donors havingaratio of0.88-1.1 (n = 8), the heterozygous HR/LRphenotype to donorshavingaratioof 0.42-0.59 (n = 11), and the homozygous LRphenotype to donorshavingaratio of<0.13 (n= 13).Thephenotypic

as-signmentofhomozygousindividualswascorroborated by

pro-liferation assays with anti-CD3 MAbs of both mIgGI and mIgG2a isotypes. In all cases, the results of the mitogenesis

assays wereinagreementwith the flowcytometryassignment

forhuFcyRIIA.

Quantitative expressionofhuFcyRIIasdeterminedbyflow

cytometryusingMAb IV.3 showedsmalldifferences between thephenotypes (PMN:HR 129±2, HR/LR

128+3,

LR121±2

meanchannel logfluorescence units[MCF];monocytes: HR

130±4, HR/LR 122±3,LR 118±4 MCF; forPMN, HR vs. LR P < 0.02, Mann-Whitney U test; all other comparisons, P

>0.1). The expression of huFcyRIII as determined by flow

cytometry using MAb 3G8, which recognizesthe NAl and

NA2 alleles, was the same for all phenotypes (NA1

154+4;

NA1/NA2 150±2;NA2152±6; P>0.1 forall comparisons).

Phagocytosis by PMNfromLRhomozygotesishigher than

HR homozygotes. Given the difference inligand bindingby Fc'yRIIAalleles transfected into fibroblasts orCOS cells (7-9),

weconsidered thepossibilitythattheFcyR-mediated phago-cytic capacity of blood cells expressing these alleles might

differ.Accordingly,weexaminedquantitative phagocytosis by

PMNfrom normal donors who arehomozygous for the LRor HR alleles. To control for differences due tohuFcyRIIIB,each

pairofhomozygousdonors wasmatched for NA alleles. Al-though not different when probed with anti-FcyRII

Fab-coated E (E-IV.3, a ligand-independent

huFcyRII-specific

probe),LRhomozygotesshowedsignificantly higher

phagocy-tosis than HRhomozygoteswhenprobedwithE-hIgG (PI [E-hIgGinternalized/100PMN]:LR vs.HR,50±7vs.27±10,P <0.003; Fig. 1).

Over a broadrangeofopsonization densities,LRsubjects

had agreaterlevel ofphagocytosis ofE-hIgG (P <0.02,

re-peatedmeasuresANOVA).Incontrast,phagocytosisEA and E-IV.3preparedat abroadrange ofopsonization densitieswas indistinguishable between the twogroups(P> 0.1, repeated

measures_ANOVA).Theabsenceofthedifference in

phagocy-tosisbetween LR and HR donors with EA and E-IV.3

sug-gestedthat the nature of theligand mightbeimportantin

de-finingthe LR-HRfunctional difference.

Higher

phagocytosis by

FcyRIIA LR

homozygotes:

rela-tionship tohIgG subclasses. Since the allelesof

_huFcyRIIA

differentially bind murine IgGi (9, 27-32) and hIgG2 in a transfected fibroblastsystem(8),weexamined thecapacityof the LR-HR alleles expressedonblood cells torecognizeand bind subclasses ofhIgG.In each ofsevenpairsofhomozygous

LR and HRsubjectsmatched forhuFcyRIIIB alleles,therewas

adramatic difference between the HR and LRindividualsin

x a)

-C)

>9

0 0) no

60

50

40

30

20

10

0

E-hIgG

E-IV.3

Figure1.PhagocytosisofE-hIgG byPMNsfromLRhomozygotesis

higherthatbyHRhomozygotes.PMNsfromnormaldonors, homo-zygous for the LRorHRalleles and matched foridenticalNAalleles

ofhuFcyRIIIB,werestudiedsimultaneouslyforinternalizationof E-IV.3 andE-hIgGinamatched-pairs design.PIwasquantitated by lightmicroscopy.Ineachof sixpairs,internalization of E-IgGwas

greater for thehomozygousLRsubjects(P<0.003,pairedttest). In contrast,phagocytosisof E-IV.3was_{indistinguishable}between the

twogroups(P=NS). (o) LR; (m)HR.

Figure

2. PMNsfromLR

homo-80 _{B zygotesinternalize E-hIgG2.}

1 20 _~ PMNs from pairs of homozygous

HR and LR donorswith

identi-X ₆₀ cal NA alleleswerestudied

si-c) 6 x

multaneously

with all_probes. CT1C8080 T _ (A) For the assay of phagocytosis,

* T 1PMNs wereincubatedwith

E-40 hIgGl, E-hIgG2, E-hIgG3,or

E-u ll _{hIgG4 opsonized}with similar

m

_{CP~~~~~~~~~~~~~~~~~~-}

rsCz*

intermediate densities ofIgG

2l 40 (MCF150-160). In each of

-- H20 2 sevenpairs,therewasminimal

<

|1^ | internalization ofE-hIgG2 by HRdonors(LRvs.HR,P 0

T

a

* * O_- <0.0004). No significant

differ-E-hE-h~g~lg G 1 E-E-hlgG2h g G 2 EE-h~gC3 E-h~gC4-h g G 3 E-h _gG 4 EE-hlg~l E-hlgC2-h

lgG

1 E-h

lgG

2 ence_probesin_wasphagocytosis_noted.(B)of the other_{For the} assayof attachment,PMNswereincubatedwithE-hIgGIorE-hIgG2.In each of sixpairs studied,therewasminimalbindingofE-hIgG2toHR PMNs(LRvs.HR, P<0.003),whereasbinding of E-hIgGIwassimilar for LR and HR.(n)LR; (n) HR.

phagocytosis of E-hIgG2 (PI [E internalized/l00 PMN]:LR vs.

HR, 47±7 vs. 3±1, P < 0.0004), whereas internalization of E-hIgGl, E-hIgG3, and E-hIgG4 wassimilar for both donor

types(Fig. 2A). The lack of internalization ofE-hIgG2

re-flectedaninabilityof theE-hIgG2tobindtoPMNsfrom HR

subjects (AI forEwith intermediatedensityofopsonization:

0-3E

attached/

100PMN; Fig.2B).Identical resultswere

ob-tained with a second and different myeloma preparation of each hIgG subclass. A series ofblocking experiments

con-firmedthathuFc'yRIIwasresponsiblefor thephagocytosis of

E-hIgG2.PretreatmentofPMNwith MAb IV.3 Fab(30

,g/ml)

toblockhuFcyRII completely abrogatedinternalization of

E-hIgG2(Fig.3). Indeed,pretreatmentwith3G8F(ab')2toblock

huFcyRIIIBdidnotblock, but rather enhanced,phagocytosis

ofE-hIgG2 consistent with ourprevious findingsthat

cross-linking huFcyRIIIB primes phagocytosis mediated by

huFcyRIIin PMN (PI: controlvs.3G8 F(ab')2, 47±16E inter-nalized/100PMN vs. 72±16,P>0.05) (25).

ThecapacityoftheLRalleleof huFcyRIIAtobindmIgGl andformsubclass-specific rosettes orinduceTcell prolifera-tion (32, 33) is dependentinpart ontheconcentrationofMAb presented to receptor-bearing cells. A similarrelationship is

apparentfor thecapacity of theHR alleleofhuFcyRIIAon PMN tobind hIgG2.Asshown inFigure4 A,higher densities

ofhIgG2 on E resulted in a higher PI in both LRand HR

donors. At alllevels ofhIgG2, however,amarked difference

between theLRandHR groups wasevident (P< 0.001,

re-peated measures ANOVA). In contrast, the phagocytosis of

E-hIgGl showednodifferencebetween thetwogroups at any

level ofE surface hIgGl (Fig. 4 B). Similarly, therewas no

differenceininternalization ofE-hIgG4opsonizedover a range

ofdensities betweenbothgroups. For example, thePIfor

E-hIgG4 with MCF 170-179 was 29±10 E internalized/100

PMNforLRdonors and 27±6 for HR donors.

Interestingly, internalization ofE-hIgG and E-hIgG1 by HR PMN wasidenticalover mostofthe range ofopsonization,

whereas internalization ofE-hIgG by LR PMN washigher than

thatof E-hIgGl with similar densities of opsonization. This difference is presumably due torecognition ofhIgG2 on the

E-hIgG by the LRallele, and this observation suggests that

functionaldifferencesbetweenthe

_huFcyRIIA

alleles may be apparent evenwith polyclonal antibodyresponses.

Detectionof the differentialcapacityofhuFcyRIIAalleles tobindhIgG2is not restricted to PMNs. Human monocytes,

whichexpressboth FcTRIIAandFcyRI,show efficient

inter-nalizationofE-hIgG2when derived from LR subjects but little internalization when derived from HR subjects(PI[E

internal-ized/100 monocytes]: LR vs. HR, 145+14 vs. 21±3, P <0.0005). In contrast, phagocytosis of both hIgG and

E-hIgGlwascomparablefor both groups(PI [E internalized/100 monocytes]:LR vs.HR, 152±10 vs. 146±25 and 149±13 vs.

153+37, respectively, P> 0.1). Blockade of huFcyRII with

IV.3 Fab blocked internalization ofE-hIgG2 bymonocytes in LRsubjects(PI [E internalized/100 monocytes]: 169±43 vs.

10±6,P<0.03), similartothefindingsin PMN(Fig. 3).

100

x

0D

-D

U

>4

0

.D_

0-n~

80

60

40

20

0

IV.3 Fab

blockade - + - N.D. - + - +

Low density IgG2 High density IgG2

Figure 3. BlockadeofhuFcyRIIAwith IV.3 Fab completely inhibits internalization of E-hIgG2 in bothLRandHRPMNs.PMNsfrom

HRand LR donorswerepretreatedwith IV.3 Fab (30,gg/ml)or con-trolmediumfor 10 min and then combined with E-hIgG2 prepared

atlower(intermediate density;MCF150-160)orhigher (MCF 170-180)density opsonization. Ineachexperiment both anHRandan LRdonorwerestudiedsimultaneously (n= 3-4pairs). Pretreatment of PMN with IV.3 Fab blockedphagocytosisof bothE-hIgG2 probes

inLRdonors(controlvs.IV.3Fab,P <0.03).In HRdonors, IV.3 Fab blockedphagocytosisof thehigh densityE-hIgG2(controlvs.

100 _r

80 F

A

E-hIgG2

_{200 r}

0

0

-'IO

_~

₀

60

F

40 F

20 F

0

100

80

60

40

20

0

140- 145- lS0o- 155- 160-144 149 154 159 164

Mean Channel Fluorescence

B

x

a) -0

C)

0

C-EL

A

160

F

120 F

80

1

40 F

0

165-169

E-hIgGl

200

X

_0

-o

C)

(9

U)

0

C-nL

FL

I

I

I-control FMLP 3G8(Fab')2

Low density G2

B

160

120

80

40

0

145- 150- 155- 160-

165-149 154 159 164 169

Mean Channel Fluorescence Figure4.The lack of internalization of E-hIgG2 by PMNs from

ho-mozygousHR subjects is relative,notabsolute. PMNs from HR and LRwerecombined withEopsonized with varyingamountsofhIgG2 (A)orhIgGl(B),asdetermined by flowcytometry(n=3-5

pairs/le-vel). Phagocytosis of all probeswasdirectly relatedtodegree of

op-sonization.(A) InternalizationofE-hIgG2washigher forLRdonors

atall levels of opsonization (P<0.001, repeatedmeasuresANOVA), (B) whereas that for E-hIgGIremained similarforbothgroups.(o) LR; (C) HR.

The

difference

inE-hIgG2 internalization betweenLRand

HR donorspersists inprimedPMNs. Simultaneous

engage-mentof severalreceptorsmaybe themostrelevant

physiologi-cal stimulus for phagocytes. Forexample, it isunlikely that microbes opsonized in vivoorendogenously formed immune

complexes would be composed exclusively of hIgG2 and thus ligate onlyhuFcyRII.The observation thattreatmentofPMNs with 3G8F(ab')2tocrosslinkhuFcyRIIIBaugmented internal-ization ofE-hIgG2 raised the possibility that primed or

acti-vated PMNs found in sites of infectionorinflammation might notshowadifference between LR and HR donors. To examine

this possibility, PMNs from LR and HR donors were

pre-treated with 3G8F(ab')2, FMLP,orcontrol medium.Inspite

ofanincreasein E-hIgG2 internalization, phagocytosis of

E-hIgG2 byLRPMNs remained significantly higher than that by

HRPMNs(Fig. 5). The increaseinE-hIgG2 phagocytosiswas

associated withanincrease inE-hIgG2 bindinginboth LR and

control FMLP 3G8(Fab')2

High density G2

Figure 5. The differenceinE-hIgG2 internalization between LR and HRdonors persists in primed PMN. PMN from HR and LR donors

weretreated with FMLP(l0-7M), 3G8 F(ab')2(30_,g/ml),orcontrol

medium before quantitation of phagocytosis of E-hIgG2 preparedat

lower(intermediate density; MCF 150-160) and higher (MCF 170-180) density opsonization. Phagocytosis of the (A) lower density E-hIgG2 (A) and the (B) higher density E-E-hIgG2wasgreaterfor LR

do-nors(n=3-9). Low density E-hIgG2, LRvs.HR: control P<0.01,

FMLP P<_{0.03, 3G8 F(ab')2 P}<0.01; High density E-hIgG2, LR

vs.HR: control P<0.001, FMLP P<_{0.03, 3G8 F(ab')2 P}<0.03.

(o)LR;(m)HR.

HRdonors, although thismaynotbe only mechanism for the increasein internalization(Al forLR[E-hIgG2adherent/100 PMN]: control 134+34,FMLP219±36, 3G8 F(ab')2150+42, n=5-6, FMLPvs.controlP<0.02;Alfor HR: control 8±3, FMLP72±15, 3G8 F(ab')2 40±10n=3-4,FMLPvs.control P

<0.04).Phagocytosis of E-hIgG2 by FMLP-stimulated PMNs fromhomozygousHRdonorswasalso completely inhibited by

blockade ofhuFcyRIIwith IV.3 Fab(PI [Einternalized/100

PMN]: FMLP-treatedPMNs=37±9, FMLP-treated PMNsin thepresence of IV.3 Fab = 5+2;n = 4;P<0.05). Thus, in

primedPMNs, hIgG2 binding and internalizationis predomi-nantlymediatedbyhuFc-yRII.

Higher phagocytic capacity ofPMNsfrom NAI subjectsis

notrelatedtoallelicpolymorphisms of huFcyRIIA.

Homozy-gousNA Idonors havehigher phagocytosisof EA than

homo-1278 J. E.Salmon,J. C.Edberg,N. L.Brogle, and R. P.Kimberly

x

->

I-C)

0

0)

C-

02-x

a)

>1

-o

0

0)

C-zygousNA2donors,which isFcyR specificrather than dueto

differences ingeneralizedphagocytic capacity (10). Given the

differencesin

FcyRIIA-mediated

phagocytosisand therecent

datalocalizingthegenesforbothhuFcyRIIandhuFcyRIIIto

humanchromosome 1(11-13),wedetermined whetheran

as-sociationbetween the allelesof huFcyRIIAandtheNAI-NA2 allelesof huFcyRIIIB might explainthehigher phagocytic

ca-pacityofNAI homozygotes. Ineach of sixpairsof NAI and NA2homozygousindividuals, eachwith identicalhuFcyRIIA alleles, internalizationofEAandofE-hIgGwashigherin the NAI donor,independentof thehuFcyRIIA phenotypeof the donorpair (Fig. 6).The difference inphagocytic capacitywas

notrelatedtoalterationsinbindingofE-hIgG (Alfor standard level of opsonization [E-hIgG

adherent/

100 PMN]: NAI vs.

NA2 = 265±69vs.

280+66,

n =5). Simultaneousanalysisof

huFc'yRII-specific

phagocytosisin the NA 1 and NA2

homozy-gousdonorswith E-IV.3 showednosignificant differences (PI

[Einternalized/100 PMN]:NAl vs.NA2 =32±12vs.29±10,

n= 5;Fig. 6). Thus,an

huFcyRIIIB-specific

mechanism,

unre-lated tothe quantitative binding of ligand and unrelated to

huFcyRIIA

alleles,appears toberesponsible forthedifference

inphagocytosisbetween NA 1 and NA2homozygousdonors.

Discussion

Allelicpolymorphisms ofhumanFcyRhavefunctionally

dis-tinct capacities thatare important in mediatinginteractions between humanphagocytesandIgGligands. Using E-hIgG,we

showconsistently higherPMNphagocytosis by individuals

ho-mozygous forthe LR alleleof huFcyRIIAandbyindividuals

homozygous for the NAI alleleof huFcyRIIIB.Thisdifference

100

80

ci<_t)\60 QO <

~t

n - 40

z

20

0

EA

E-hIgG

E-IV.3 Figure6.FcyR-mediatedphagocytosisofEAandE-hIgG,butnot

E-IV.3,ishigherin_homozygousNAIsubjectscomparedwith that

ofNA2_subjects.PMNfromNAl and NA2homozygousdonorswith

identicalHR-LRalleleswere_assayedsimultaneouslywithEA,

E-hIgG,andE-IV.3,and the percent difference between the donorswas

calculated.Ineachof six_{pairs studied,}the NA2 donor hadalower

phagocytic capacityforEAandE-hIgG (PIfor EA: 45±9%lower,P

<_0.03;PIforE-hIgG:33±8%_lower,P<_{0.003). Phagocytosis}of E-IV.3wassimilar for both groups(P=_{NS).Aswith EA(9),the} dif-ference inphagocytosisofE-hIgGbecamemorepronouncedwith lessopsonizationof theE,suchthatafourfolddilution of the

stan-dardconcentration ofhIgGshowedan80±10% difference in PI

(in-sert).Thedilutions of_{biotinylated hIgG}usedtoopsonizeEBA

corre-spondedtoMCFsof 164(std), 158(1:2),and144(1:4).

is even more pronounced in LR homozygous subjects for hIgG2 subclass-specific phagocytosis, since, incontrast toHR, the LR allele ofhuFcyRIIAbindshIgG2 efficiently.

Thedifferences in functionalcapacity for thetwoalleles of huFc-yRIIA appear to be mediated bya mechanism distinct fromthat involved in the differences between the two alleles of

huFcoyRIIIB. The biallelic HR-LR polymorphism ofhuFcy

RIIAalters theaffinity of the ligand binding site forsome sub-classes ofIgG ligand. Originally distinguished bydifferences in

bindingofmIgGIto humanmonocytesasassessedbyT cell

proliferation induced by anti-CD3 MAb (27-32), the func-tional HR-LRphenotypeis determinedbyasingleamino acid

present atposition 131, whichplaysacriticalrole in the bind-ing capacity for mIgG1 (6, 7, 9). Warmerdam and colleagues

(8)also showedthat fibroblast transfectantsexpressingthe LR allele (histidine residue atposition 131)lack the capacityto

bind mIgGI but can bind hIgG2, whereas transfectants

ex-pressingthe HR allele(arginine residueatposition 131) bind

mIgGI but do not recognize hIgG2. Importantly, the func-tionalimpact of this difference is notrestrictedtotransfected fibroblastsexpressingasinglereceptorfamilybutis evident in

resting and activated human phagocytes. Our data indicate

that for human PMN, which express both huFcyRII and

huFcyRIIIB, submaximally opsonizedE-hIgGareinternalized

moreefficientlybyLRdonors.This differencereflectsthe

abil-ity

of LR donors tobind

hIgG2,

which

comprises

20%of pooled hIgG (34, 35). E-hIgG2areinternalized efficientlyby individuals homozygous forLRalleleof huFcyRIIAbutare

internalizedpoorly byindividualshomozygousforthe HR al-lele(Fig.2). Even in"primed"PMNs, which demonstrate

en-hancedbinding and internalization of E-hIgG2, the difference betweenHRandLRdonorspersists (Fig. 5).Thisdifferencein

E-hIgG2phagocytosis isalsoevidentin humanmonocytesthat

express both huFcyRI and huFcyRII. The inability ofHR

PMN tobind hIgG2 is relative rather than absolute, andathigh levelsofopsonizationthereisamodest degreeof

internaliza-tionof E-hIgG2 (Fig.4). A similardiscriminative preference for the other human subclasses of IgG was not identifiable.

Although absolute comparisons of efficiency of binding and

internalization foreachofthe hIgG subclassesare notpossible, in partbecause determinations of"comparabledensities" of opsonizationmaybeinfluencedbyunique properties of indi-vidualmyelomaproteins and by properties of the second

anti-body,ourresultssuggestthat inLRPMNtherelative efficiency is hIgGl =hIgG3 =hIgG2>hIgG4and, in HR PMN,hIgGI =hIgG3 >hIgG4>hIgG2.

ThebiallelicNA l-NA2polymorphism ofhuFcyRIIIB,

ex-pressed onlyonhuman PMNs, is relatedtoseveralamino acid

differences,twoof which alterpotential N-linkedglycosylation sites(1-5).SincehuFcyRIIIBmayfunction inpart as abinding moleculetopresentligandtohuFcyRIIA(24), weconsidered thecontribution of the LR-HR alleles of huFcyRIIAto the

functional consequences of the huFcyRIIIB polymorphism.

The current results demonstrate higher phagocytosis of E-hIgG

byNA Idonors compared withNA2donors,despite matching ofdonorsfor huFcyRIIA alleles (Fig. 6)and nodifference in

phagocytosis of E-IV.3(anhuFcyRII-specific probe) byNAl

andNA2donors(Fig.6). These data, along withourprevious observation that blockade ofhuFcyRII withIV.3 Fabamplifies

the difference in phagocytosis between NA 1 and NA2 (10), supporttheconclusionthatthedifference in EAandE-hIgG

independent of

_huFcyRIIA

phenotype. The mechanism un-derlying the difference between NAl and NA2 homozygotes remains undefined, although IgG-independent, allele-specific interactions with othersurfacemembrane molecules provide provocative possibilities.

The implications of these two allotypic systems are intrigu-ing. The NAl-NA2 functional polymorphism appears to be IgG-ligand independent. No differences in opsonized target binding have been demonstrated, and no subclass preference has beenidentified since,among HR homozygous donors, the inability to bind hIgG2 is equally apparent for both theNAI

homozygous and NA2 homozygous phenotypes. The differ-ence ininternalization is evident with concanavalin-treated E, which engagehuFcyRIIIBthroughnonclassical (IgG-ligand in-dependent)carbohydrate-mediated interactions(10). The fact that the NA I-NA2 alleles are functionally distinct indicates, however, thathuFcyRIIIBserves as an integral participant in

phagocytosis by PMNs. In contrast, the HR-LR functional

polymorphismis directly ligand dependent, and the ability of LR PMNsand monocytes to bind hIgG2 efficiently suggests that a reexaminationof the role of hIgG2 humoral response may beappropriate.For example, since the antibody response to bacterial polysaccharide antigens ispredominantlyhIgG2 isotype (36), therelationshipbetween hIgG2 levels and

suscepti-bilitytoinfection withencapsulatedbacterialpathogens (34, 35, 37) may be influenced by the HR-LR phenotype of the host.Thus, theincreased risk of invasive Hemophilus

influen-zaeinfection incertain populationsthat haveimpaired anti-bodyresponsestoH.influenzaeand lowhIgG2levels(38, 39)

may begreaterforHRhomozygotesthai forLRhomozygotes.

Similarly,incystic fibrosis,wherebacterial colonization of the lungs by Pseudomonas aeruginosa is associated with high

serumlevelsof hIgG2 anti-Ps.aeruginosa antibodies and im-mune complexes containing hIgG2 anti-Ps. aeruginosa anti-bodies, thepotentialfor the hIgG2 antibodiesto actaspoor

opsoninsandinhibitclearance of theorganisms (40-42)may vary in accordwiththeHR-LRphenotype ofthe host.

Impor-tantly, activation ofPMNsdoesnotabrogatethis allelic

differ-enceinfunction. Indeed,one

might

speculatethat the evolu-tionarypressureforthe presenceoftheLRallele of

huFcyRIIA

(29) is related to more effective defense against microbial

agentselicitinganhIgG2humoral response.

Theprecedent for allelic differencesin receptor structure

havingasignificant impactonphagocytefunction is of

poten-tialimportance giventhestructural

diversity

of huFcyR.

Defi-nitionoffurther suchstructure-functionrelationshipsamong

huFcyRmayprovide insights into bothdiseasesusceptibility

and

pathogenesis.

Acknowledgments

Wegreatly appreciatetheparticipationof the normal volunteers

with-outwhoseassistancethesestudiescouldnothave beencompleted.We thank Dr.MaryCrow forperformingtheanti-CD3proliferation

as-says; Carl Triscari forassistancewithflowcytometry; Drs. Jan G. J.van

deWinkel,PetraA. M.Warmerdam,and AllanGibofskyforuseful discussions;Dr. Margaret Peterson for assistancein statisticalanalysis;

andDr.Charles L.Christian for his continued support.

This workwassupported in partby grants ROI-AR38889 (Dr. Salmon)andROI-AR33062 (Dr.Kimberly)awardedbythe National Institutes of Health. Dr. Edberg isaFellowoftheIrvington Institute for Medical Research. The Flow Cytometry CoreFacilityatTheHospital forSpecial Surgeryand the ResearchMethodologyCore(Biometry)

were supported in part by theCornellMultipurposeArthritis and

Mus-culoskeletalDiseasesCenter(P60-AR38520).

References

1. Werner, G., A. E. G. Kr. von dem Borne, M. J. E. Bos, J. F. Tromp, C. M. van der Plas-vanDalen, F. J. Visser, C. P.Engelfriet,and P. A.Tetteroo. 1985. Localization of the human NA 1 alloantigen on neutrophils toFc'yreceptors. In LeucocyteTypingII. Vol. 3. E. L. Reinherz, B. F. Haynes, L. M. Nadler, andI.D. Bernstein, editors. Springer-Verlag, New York, 109-121.

2. Tetteroo, P. A. T.,C. E. Van der Schoot, F. J. Visser, M. J. E. Bos, and A. G. E. Kr. von dem Borne. 1987. Three different types ofFcyreceptors on human leukocytes defined by workshop antibodies: FceyR,. of neutrophils,

FceyRk, of NK/K lymphocytes and Fc'yRII. In Leucocyte Typing III. A. J. McMichael, editor. Oxford University Press, Oxford,UK.702-706.

3. Ory, P.,I. M. Goldstein, E. E. Kwoh, and S. Clarkson. 1989. Characteriza-tion of polymorphic forms of Fc'yRIII on human neutrophils. J. Clin. Invest. 83:1676-1681.

4. Ory, P., M. R.Clark, E. E. Kwoh, S. B. Clarkson,andI.M. Goldstein. 1989. Sequences of complementary DNAs that encode theNAIand NA2 forms ofFc'y

receptor III on neutrophils. J. Clin. Invest. 84:1688-1691.

5. Ravetch, J. V., and B. Perussia. 1989. Alternative membrane forms of

FcyRIHI(CD16) onhuman natural killer cells and neutrophils. Cell type-specific expression of two genes that differ in single nucleotide substitutions. J. Exp. Med.

170:481-497.

6. Clark, M. R., S. B. Clarkson, P. A.Ory, N.Stollman, and I. M. Goldstein. 1989. Molecular basis for a polymorphism involving Fc receptor II on human monocytes. J. Immunol. 143:1731-1734.

7. Warmerdam, P. A. M., J. G. J. van de Winkel, E. J. Gosselin, and P. J. A. Capel. 1990. Molecular basis for a polymorphism of humanFcyreceptor II (CD32). J. Exp. Med. 172:19-25.

8. Warmerdam, P. A. M., J. G. J. van de Winkel, A. Vlug, N. A. C. Westerdal, and P. J. A. Capel. 1991. A single amino acid in the second Ig-like domain ofthe human Fcy receptor II is critical for human IgG2 binding. J. Immunol.

147:1338-1343.

9. Clark, M. R., S. G. Stuart, R. P. Kimberly, P. A. Ory, and I. M. Goldstein. 1991. A single amino acid distinguishes the high-responder from low-responder

form ofFc receptor II on human monocytes. Eur. J. Immunol. 21:1911-1916. 10. Salmon, J. E., J. C. Edberg, and R. P. Kimberly. 1990.FcyreceptorIII on human neutrophils. Allelic variants have functionally distinct capacities. J. Clin. Invest. 85:1287-1295.

11. Peltz, G. A., H.0. Grundy, R. V. Lebo, H. Yssel, G. S. Barsh, andK.W. Moore. 1989. Human Fc-yRlII:cloning, expression, and identification of the chromosomal locus of two Fc receptors for IgG. Proc. Natl. Acad. Sci. USA. 86:1013-1017.

12. Grundy, H.O., G. Peltz, K. W. Moore, M. S. Golbus, M. L. G. Jackson, and R. V. Lebo. 1989. The polymorphicFcyreceptor II gene maps to the human chromosomelq.Immunogenetics. 29:331-339.

13. Brooks, D. G., W. Q. Qiu, A. D. Luster, andJ. V. Ravetch. 1989. Structure and expression ofhuman IgG FcRII (CD32). Functional heterogeneity is en-coded by the alternatively spliced productsofmultiple genes.J. Exp. Med.

170:1369-1385.

14.Lalezari, P. 1977. Neutrophil antigens: immunology and clinical

implica-tions.In TheGranulocyte: FunctionandClinical Utilization.T.Greenwalt and G. S. Jamieson, editors. Alan R. Liss, New York, 209-225.

15. Edberg, J.C., P. B. Redecha, J. E. Salmon, and R. P. Kimberly. 1989. HumanFc-yRIII(CD16). Isoforms with distinct allelic expression, epitopes dis-plays and membrane anchors on PMN and NK cells. J.Immunol. 143:1642-1649.

16. Edberg, J. C., M. Barinsky, P. B. Redecha, J. E. Salmon, and R. P.

Kim-berly.1990.Fc'yRIIIexpressedonculturedmonocytesisanN-glycosylated trans-membraneprotein. SimilaritytoFcyRIIIexpressed on NK cells.J.Immunol.

144:4729-4734.

17. Gosselin, E. J., M. F. Brown, C.L. Anderson, T. F. Zipf, and P. M. Guyre. 1990. The monoclonal antibody 41H16 detects the Leu 4 responderformof humanFcyRII.J.Immunol.144:1817-1822.

18.Fleit, H. B., S. D.Wright,and J. C. Unkeless. 1982. Humanneutrophil

Fc-yreceptordistributionand structure. Proc.Nati. Acad.Sci. USA. 79:3275-3279.

19. Looney, R. J., D. H. Ryan, K.Takahashi,H. B.Fleit, H. J. Cohen, G. N. Abraham, and C. L. Anderson. 1986.Identificationof a secondclass of IgGFc receptors on humanneutrophils.A40kilodalton molecule alsofoundon eosino-phils. J. Exp. Med. 163:826-836.

20. Antoun, G. R., B. M. Longenecker, and T. F. Zipf. 1989.Comparison of

the 40 kDahematopoieticcellantigensboundby monoclonal antibodiesIV.3,

41H.16 and KB61. Mol. Immunol.26:333-338.

21.Salmon,J.E.,S.Kapur,and R. P.Kimberly.1987.Opsonin-independent

ligationof Fcy receptors: the3G8-bearingreceptorsonneutrophilsmediate the

phagocytosisof concanavalinA-treated erythrocytes andnonopsonizedE.coli.

J. Exp.Med. 166:1798-1813.

22. Edberg, J. C., and R. P. Kimberly. 1992. Receptor specific probes for the study ofFcyreceptorspecificfunction. J. Immunol. Methods. In press.

23. Shen, L., P. M. Guyre, C. L. Anderson, and M. W. Fanger. 1986.

Hetero-antibody-mediated cytotoxicity:antibody to thehighaffinity Fc receptor for IgG mediatescytotoxicity by human monocytes that is enhanced by interferon-gamma andis not blocked by human IgG. J. Immunol. 137:3378-3382.

24.Anderson, C. L., L. Shen, D. M. Eicher, M. D. Wewers, and J. K. Gill. 1990.Phagocytosis mediatedby threedistinct Fcyreceptor classes on human leukocytes. J. Exp.Med. 171:1333-1345.

25. Salmon, J. E., N. L.Brogle,J. C. Edberg, and R. P. Kimberly. 1991.Fcy

receptor III induces actin polymerization in human neutrophils and primes

phagocytosis mediatedbyFcyreceptor II. J.Immunol.146:997-1004. 26.Salmon, J. E., R. P.Kimberly,A.Gibofsky,and M.Fotino.1986. Altered

phagocytosisby monocytesfromHLA-DR2 andDR3-positivehealthy adults is

Fc-yreceptor-specific.J.Immunol. 136:3625-3630.

27. Tax, W.J.M., H. W. Willems, P. P. M. Reekers, P. J. A. Capel, and R. A. P.Koene. 1983.Polymorphisminmitogeniceffect ofIgGl monoclonal

antibodies againstT3antigenonhuman T cells. Nature(Lond.).304:445-447.

28.Tax, W. J. M., F. F. M. Hermes, R. W. Willems,P. J. A. Capel, and R. A. P.Koene. 1984. Fc Receptors for mouse IgGi on human monocytes:

polymorphismand roleinantibody-inducedTcellproliferation.J.Immunol.

133:1185-1189.

29. Abo, T., A. B. Tilden, C. M. Balch, K.Kumagai,G. M. Troup, and M. D.

Cooper. 1984. Ethnic differences inthelymphocyteproliferativeresponse in-ducedby murine IgGI antibody, Leu 4, to theT3molecule. J. Exp. Med.

160:303-309.

30.Cueppens,J.L.,andF. vanVaeck. 1987. Directdemonstration ofbinding ofanti-Leu 4antibodytothe 40 kDa Fc receptor on monocytes as aprerequisite for anti-Leu4inducedTcellmitogenesis.J.Immunol. 139:4067-4071.

31. van de Winkel, J. G. J., G. J. J. C. Boonen, P. L. W. Janssen, A. Vlug, N. Hoog,and W. J. M. Tax. 1989. Activity of two typesofFc receptors,Fc-yRIand FcyRII, in human monocyte cytotoxicity to sensitized erythrocytes. Scand. J.

Immunol.29:23-31.

32.Clement, L. T., A. B. Tilden, and N. E. Dunlap. 1985. Analysis of the monocyte Fcreceptor and antibody-mediated cellularinteractionsrequiredfor

the induction ofT cellproliferationby anti-T3 antibodies. J. Immunol. 135:165-171.

33. Boot, J. H. A., M. E. J. Geerts, and L. A. Aarden. 1989. Functional

polymorphisms ofFc receptors in humanmonocyte-mediated cytotoxicity

to-wardserythrocytes induced by murine isotype switch variants. J. Immunol. 142:1217-1222.

34.Schur, P. H. 1987. IgG subclasses-areview.Ann.Allergy. 58:89-99. 35.Jefferis,R., and D. S. Kumararatne. 1990. SelectiveIgGsubclass defi-ciency: quantificationandclinicalrelevance. Clin.Exp. Immunol.81:357-367.

36.Yount, W. J., M. M.Corner, H. G.Kunkel, and E. A. Kabat. 1968.

Studiesonhuman antibodies.VI.Selective variationsinsubgroupcomposition

andgeneticmarker. J. Exp.Med. 127:633-646.

37.Scott,M.G.,D. E.Briles,and M. H. Nahm. 1990.SelectiveIgGsubclass

expression: biologic, clinicalandfunctionalaspects. In The HumanIgG

Sub-classes: MolecularAnalysis ofStructure, Function andRegulation.F.Shakib, editor.Pergamon Press,Oxford,UK. 161-183.

38.Siber,G. R., M.Santosham,G.R.Reid,C.Thompson,J.Almeido-Hill,

A.Morell,G. deLange, J. K. Ketchum, and E. H. Callahan. 1990. Impaired antibodyresponse toHaemphilusinfluenzatypebpolysaccharideand lowIgG2

andIgG4concentrations inApachechildren.N.Engl.J.Med. 323:1387-1392.

39. Ward, J. G. Brenneman, G. W. Letson, W. L. Heyward, and the Alaska H.

influenza VaccineStudy Group.1990.Limited efficacy ofaHaemophilus

influ-enzatype bconjugate vaccine inAlaskainfants.N.Engl.J.Med. 323:1393-1401. 40.Fick,R.B., Jr., J.Olchowski, S. U.Squier,W. W.Merrill, andH. Y. Reynolds.1986. Immunoglobulin-Gsubclasses incystic fibrosis. IgG2response toPseudomonasaeruginosalipopolysaccharide.Am.Rev.Respir. Dis.

133:418-422.

41.Fick,R.B., Jr., G. P.Naegel, R. A. Matthay, and H. Y. Reynolds. 1981.

Cystic fibrosis Pseudomonas opsonin. Inhibitorynatureinanin vitrophagocytic

assay. J.Clin.Invest.68:899-914.

42.Hornick,D.B., and R. B.Fick,Jr.1990. Theimmunoglobulin G subclass composition ofimmunecomplexes incysticfibrosis. Implications for the