Interaction of human monocytes, macrophages, and polymorphonuclear leukocytes with zymosan in vitro Role of type 3 complement receptors and macrophage derived complement

Interaction of human monocytes, macrophages,

and polymorphonuclear leukocytes with

zymosan in vitro. Role of type 3 complement

receptors and macrophage-derived

complement.

R A Ezekowitz, … , G G MacPherson, S Gordon

J Clin Invest.

1985;

76(6)

:2368-2376.

https://doi.org/10.1172/JCI112249

.

Macrophages take up zymosan in the absence of exogenous complement via receptors for

iC3b (type 3 complement receptors) acting with or without lectin-like receptors for

mannosyl-fucosyl-terminated glycoconjugates. We previously provided evidence that macrophages

themselves secrete complement-alternative pathway components able to opsonize

zymosan locally (Ezekowitz et al., J. Exp. Med. 1984. 159:244-260). We show here that

covalently bound C3 cleavage products C3b and iC3b can be eluted from zymosan

particles cultivated with 36-h adherent human monocytes in the absence of serum. The

ligand binding site of type 3 complement receptors is involved in macrophage-zymosan

interactions as shown by inhibition studies of zymosan binding and uptake with Fab

fragments of anti-C3 antibodies and monoclonal antireceptor antibodies M01 and OKM10.

In contrast, antibody IB4, which binds to a receptor epitope distinct from the binding site,

does not inhibit zymosan uptake. Selective modulation of macrophage receptors onto

anticomplement receptor antibody and mannose-rich yeast mannan, respectively, confirms

that the complement and lectin-like receptors are distinct. Human polymorphonuclear

leukocytes, which express receptors for complement, but are not known to secrete

complement proteins, bind and ingest only exogenously opsonized zymosan. Unopsonized

zymosan is a poor trigger of respiratory burst activity in neutrophils or 7-d adherent human

macrophages, but induces cell aggregation and secretion of large amounts of superoxide

anion when these cells are co-cultivated in serum-free medium […]

Research Article

Find the latest version:

Interaction

of Human

Monocytes, Macrophages,

and

Polymorphonuclear

Leukocytes with Zymosan

In

Vitro

Role

of

Type

3

Complement Receptors

and

Macrophage-derived Complement

R. AlanB.Ezekowitz, R. B.Sim, G. G. MacPherson,and S. Gordon

Sir WilliamDunn SchoolofPathology,andMedical ResearchCouncilImmunochemistry Unit,Department ofBiochemistry, University of Oxford, Oxford OX] 3RE, England

Abstract

Macrophages take up zymosan in the absence of exogenous complement via receptors for iC3b

(type

3

complement

receptors) acting withorwithout lectin-like receptors for

mannosyl-fucosyl-terminated

glycoconjugates.

We

previously

provided

evidence that macrophages themselves secrete complement-alternative pathway components ableto

opsonize

zymosan

locally

(Ezekowitz et al., J.

Exp.

Med. 1984.

159:244-260).

We show here that covalently bound C3

cleavage

products

C3b and

iC3b

can be elutedfrom zymosan

particles

cultivated with 36-h adherent hu-man monocytes in the absence ofserum.The

ligand binding

site of type 3 complement receptors is

involved

in

macrophage-zy-mosan

interactions

as shown by inhibition studies of zymosan

binding

and uptake with Fab

fragments

of anti-C3 antibodies and monoclonal

antireceptor

antibodies

M01

and

OKM10.

In contrast,antibody IB4, which bindsto areceptor

epitope

distinct from the binding site, doesnotinhibit zymosan uptake. Selective modulation of macrophage receptors onto

anticomplement

re-ceptor

antibody

and mannose-rich yeast mannan,

respectively,

confirms

that the

complement

andlectin-like receptorsare dis-tinct. Human

polymorphonuclear leukocytes,

which express re-ceptors forcomplement, butare notknownto secretecomplement proteins, bind and ingest only exogenously opsonized zymosan.

Unopsonized

zymosan isapoor

trigger

of

respiratory

burst

ac-tivity

in

neutrophils

or7-d adherent human

macrophages,

but induces cell aggregation and secretion of large amounts of

su-peroxide

anion when these cellsare

co-cultivated

in serum-free medium and challenged with zymosan. Our studies

indicate

that complement

and/or

otherproducts

synthesized

by macrophages atextravascularsites could

play

an

important

rolein opsonization and

lysis

of pathogens abletoactivate the alternative pathway and mediate

macrophage-neutrophil

collaboration in

first-line

hostdefence.

Introduction

Various macrophage

_(MO)'

populations bind and ingest zymosan in the absence of an

exogenous

opsonin (1). Monoclonal

anti-Dr. Ezekowitz's present addressisDivisionofHematology/Oncology, Children's Hospital, Harvard Medical School, Boston,MA02115.

Addressreprintrequeststo Dr. Gordon,SirWilliamDunn School

of Pathology, University of Oxford, South Parks Road, Oxford OXI 3RE, England.

Receivedfor publication 7 February 1985 and in revised form 27 June1985.

bodies directed

_against

the functional site of the

type

3

comple-ment receptor

(CR3)

for iC3b-coated

particles (2, 3)

partially

inhibit

uptake

of

unopsonized

zymosan.Directsugar

recognition

by

the lectin-like

mannosyl-fucosyl receptor

on

_macrophages

(MFR[4])

has also been

implicated

inzymosan

uptake by

use

ofmannan as

inhibitor,

and additive effects of MFR and CR3 inhibitorswerenotedinsome

M) populations.

We and others

(5-7)

have

presented

evidence that

MO

themselves

produce

complement

components of the alternative and classical

path-ways.

Biosynthetic labeling

indicated that

MO

synthesize

and secrete active C3 under assay

conditions,

and C3 convertase components, factor H and I

activities,

were detected in

M4-conditioned medium. We therefore

postulated

that

locally

se-creted

complement

components could

opsonize

targets suchas

zymosanfor

uptake

viaCR3. Studies

by

Blackwelletal.

(8)

have

provided

evidence thatasimilar mechanism contributestothe initial

penetration

of resident

peritoneal

mouse

MO

by

promas-tigotes

of Leishmania

donovani,

which also activate the

alter-native

pathway

of

complement.

In this article we

provide

further evidence that the

ligand

iC3b is

deposited

onzymosan underassayconditions and

ex-amine the role of CR3 inzymosan

uptake by using

monoclonal

antireceptor

antibodies. Because

polymorphonuclear

leukocytes

(PMN)

express

CR3,

butare notknowntosecrete

complement

proteins,

we have studied their interaction with

unopsonized

and

iC3b-opsonized

zymosan.

We,

like others

(9, 10),

find that human PMN

only

take up zymosan after exogenous

opsoni-zation. Co-cultivation with

MO,

however, enables PMNto

re-spond

to zymosan

by

a

vigorous respiratory

burst

indicating

that

MO

complement and/or

other products enable the PMN toreact withzymosan particles in the absence of serum. This in vitrosystemindicatesanovel role for the

MO

in interaction with PMN and initiation of theacuteinflammatoryresponse.

Methods

Mice

CBAT6T6 werebred at the Sir William Dunn School of Pathology, Oxford,and both sexes were used at20-30 g.

Media and reagents

Iscove'smodification of Dulbecco's medium (IM) wasobtainedfrom Gibco-Biocult Ltd., Paisley, Scotland. Fetal bovine serum(FBS)from

the same source was_{routinely heat-inactivated (560C for 30 min) before}

1.Abbreviations used in this paper: CR3, type 3 complement receptor; E, sheeperythrocyte; EA, sheep erythrocyte coated with rabbit IgM anti-erythrocyte antibody; FBS, fetal bovine serum; IM, Iscove'smodified medium; KRPG, Krebs-Ringer phosphatebuffer with glucose; Man-BSA, mannosylated bovineserumalbumin; MFR, mannosyl-fucosyl re-ceptor;

M0,

macrophage; NBT, nitroblue tetrazolium; O°, superoxide ion; PAGE, polyacrylamidegelelectrophoresis;PMA, phorbolmyristate acetate;RPM, resident peritoneal macrophage.

2368 Ezekowitzetal.

J.Clin. Invest.

© The American Society for Clinical Investigation,Inc.

0021-9738/85/12/2368/09 $1.00

use. 100

_jsg/ml

kanamycin, 50

_og/ml

streptomycin, and 50

ug/ml

pen-icillin were added to media. Phosphate-buffered saline (PBS) A and B were obtained from Oxoid Ltd., Basingstoke, England. Ficoll-Paque was obtained from Pharmacia Fine Chemicals, Uppsala, Sweden. Mannan, ribonuclease B, zymosan type A, cytochrome c, superoxide dismutase, phorbol myristate acetate (PMA), and proteins used as molecular weight standards were all obtained from Sigma Chemical Co., St. Louis, MO.

Antibodies

The rat anti-mouse hybridomaMl/70(anti-MAC-l), used as a tissue culture supernatant at saturation or as an Fab fragment, was a kind gift from Dr. T. Springer, Harvard Medical School, Boston, MA. This an-tibody defines the CR3 on mouse M+, PMN, and natural killer cells (2). MO1 (3) and OKM 10(11)are mouseanti-human antibodies and were gifts of Dr. R. Todd, Harvard Medical School, and Dr. G. Goldstein, OrthoPharmaceuticals, Rahway, NJ,respectively.Bothantibodies inhibit bindinganduptake of iC3bligands and were used as ascites at saturating concentrations. The monoclonal antibody, IB4, which was a kind gift ofDr.S.D.Wright, TheRockefeller University, New York, is directed against a different epitope of the human iC3b receptor(11). Rabbit anti-human C3 was prepared and converted to the Fab fragment as de-scribed (12).

Cells

Mouse macrophages. Resident peritoneal macrophages (RPM) were

harvestedfrom unstimulated mice (1). The cellswerewashed,suspended in IM,andplatedonglass,mannan-coatedcoverslips,orMl /70-coated coverslipsat 1 X

101 MO

percoverslip.

Isolationofhuman monocytes andPMN. DextranT250 3% wt/vol wasadded tofresh whole blood from the John Radcliffe Hospital Blood Transfusion Service, Headington, Oxford,andmononuclear cells and

PMN wereisolatedbycentrifugationonFicoll-Paque.The mononuclear fractionwaswashedfive times in IMat200 gto removeplatelets. Cells wereresuspended inserumfree IM and adheredfor 1 hin24-well tissue culture trays and washedto removenonadherentcells,andIM+ 10% heat-inactivated human serum pooled from 20 different donorswere

added. The cellswerecultivated for 24hor7d.Cellswerecultivated

onglasscoverslipsforphase-contrastmicroscopyor asdescribed below for electron microscopy. Duringcultivation for7 dcells became well spread and resembledmature

MO

morphologically.Toremove serum

proteins prior to assay, all cultureswerewashedfive times in PBS and

thencultivatedforafurther2 hinserum-freeIM,whichwas

replaced

immediately before assay.PMN wereobtainedbyexposingthepellets from theFicoll-Paquegradientsto two treatmentswithhypotonic

lysis

buffer(0.2% wt/volNaCI)tolyse erythrocytes. Cellswere

resuspended

inIM orKrebs-Ringerphosphatebuffer solution withglucose

(KRPG)

and eitherassayed in

suspension

oradheredfor 15 min asdescribed below. 99% of cellswere PMNand>98% excluded trypan blue.

Ligands

Aglycoconjugateofmannose-bovineserumalbumin(Man-BSA)with 33-37 mol ofsugar/mol ofproteinwas agiftfromDr. P.Stahl,

Wash-ingtonUniversity,St.Louis, MO.This materialwastrace-labeled with

125I

andusedat8X 103 cpm/ng.Zymosanwasboiled for15min in 10

mlof PBSand waspelletedbycentrifugation. Zymosanwasiodinated

asdescribed (13).

Complement

HumanC3waspreparedaspreviouslydescribed(14).Factors H andI, respectively,wereisolatedaspreviouslydescribed( 15, 16).

Sensitized erythrocytes(E) bearingC3fragments (C3b and

iC3b)

werepreparedaspreviouslydescribed(12).TheIgMfraction from rabbit hemolysinwasusedtosensitizetheerythrocytes (EA).

Zymosan coated withhumanC3b waspreparedasdescribed(17)

and convertedtozymosan-iC3b byincubationfor4hat

37°C

inPBS

containing10

,g/ml

factorHand 2

gg/ml

factor I.Factors HandIwere removedbywashinginPBS. The zymosan had

14,000

C3boriC3b

sites/

particle.

Preparation of

mannan-coated

and

MJ/70-coated

coverslips

Glass coverslips were boiled in decon for 5 min and washed extensively in running tap water. Proteins were covalently coupled as described (18). Briefly, surfaces were treated with poly-L-lysine, with glutaraldehyde, then with the Fabfragment of M1/70 at a concentration of 50

_Ag/ml.

After 90 min at 20'C the unreacted glutaraldehyde was neutralized with 0.1 M glycine (pH 7.5) containing human serum albumin at 1 mg/ml. Mannan-coated coverslips were prepared as described (19). 50 mg/ml mannan and 10 mg/ml carbodiimide hydrochloride were layered on poly-L-lysine-coated coverslips overnight at4VC.

Deposition ofcleaved C3 on zymosan

Zymosan particles were incubated under serum-free conditions with hu-man monocytes in the presence ofinhibitors of uptake, recovered, washed free of noncovalently bound material and then stripped to remove ester-bound C3 fragments. Before zymosan challenge the monocytes (-5 x 10') were cultivated for 24-36 h in medium containing 10% heat-inactivated pooled human serum, washed five times, incubated for 2 h inserum-free medium to remove adsorbed proteins and incubated further for 6hin freshserum-free medium. Monolayers were then incubated for 30 min at37°Cwith well-suspended zymosan particles, 100 per cell, in the presence of anti-CR3 monoclonal antibody and mannan. The zymosan particles were carefully collected and washed successively with 2 MNaCI, water, 1% SDS, 1% Triton X-100 in 150 mM NaCI, andfinally in 6 Mguanidine-HCI (pH 7.0)to remove noncovalently boundmaterial. The material was then suspended inphysiologic saline, frozen, and thawed, rewashed in 0.2 M Tris-2% SDS-8 M urea, pH 8.0, andfinally incubatedfor 3hat37°C in 200Al50mM diethanolamine-HCl(pH 11.5)-100mMNaCI-0.1% SDS. The last incubation conditions remove ester-boundC3fragments from carbohydratesurfaces (14). The high-pH eluate from the zymosan was radioiodinated with 125I using chloramineT tofacilitate detection ofproteinin the eluate. The radio-iodinated sample was then reducedwithdithiothreitol,runon SDS poly-acrylamide gels, and compared with C3fragmentstandards.

Secretion

of

C3

and

factor

H

l-d monocyte cultures werepreparedasabove, washed, and incubated 2 hinserum-free medium. Cellswerethenincubatedfor 12 hin

me-thionine-free medium containing 1% bovine serum albumin and

[35S]methionine,

250

ACi/ml.

Conditioned mediumwascollected,cleared bycentrifugation, andprocessedasfollows.2.5 mlof supernatant

con-taining 3.5X 106 dpm of

I'S

incorporated intonondialyzablematerial was mixed with 1 mlof human plasma and fractionatedon a 27X

1.8-cmdiam column of DEAE-celluloseequilibratedin 25mMpotassium phosphate-S mM EDTA-50 mMe-aminocaproate, pH 7.0. -40% of the radioactivity wasretained onthe column andthiswaseluted by applyingalineargradient (300 ml) ofNaCl,from 0to0.35 M. This chromatographic procedure isastandard preparative step for C3 and factorHandthe elution positionsof C3 andHarewellcharacterized (20). Cold carrier C3andfactorHin the column eluateweredetected byOuchterlony immunodiffusionusingmonospecificantisera. The pro-teins eluted in the expected positions. Portions of successive column fractionswerethensubjectedtoSDS-polyacrylamide gel electrophoresis (PAGE). Thegels weretreated with Autofluor(National

Diagnostics,

Inc.,Somerville,NJ) and the"S-labeled proteinsweredetectedby fluo-rography.

Assays

Uptakeof 25I-zymosan. Human 1-d cultured monocyteswere

prein-cubated withorwithoutsaturatingamountsof anti-CR3

antibodies,

the

Fab fragmentofanti-C3

globulin

or 1 mg/mlyeastmannan in IM for

15minat37°Cin 5%CO2.Afterafurther incubation with

1251I-zymosan

(- 100 zymosanparticles/cell),the cellswerewashedfive timestoremove free zymosan. Attached zymosan wasremoved

by

a 10-mintreatment

with 2.5% wt/voltrypsin solution (Gibco Biocult Ltd.)at 370Cas de-scribed(1).Cells werethen solubilized in 1 M NaOH and cell-associated

radioactivity

wasmeasured inaPackard gamma spectrometer (Packard

InstrumentCo., Inc., DownersGrove, IL).

Mannose-specifc

endocytosis. Uptake

wasmeasuredusing saturating

amountsof trace-labeled Man-BSAasdescribed

(21)

andwasinhibited

by

yeastmannan

(1 mg/ml).

Resultswereexpressedasnanogramsligand/

microgram

of cell

protein.

Binding ofEAiC3b byRPMon

mannan-

orMJ/70-coated

coverslips.

M40

on

coverslips

were placedin 0.5 ml ofIMin wells. 50AL of EA coatedwith iC3b(26,500 sites/cell)wasadded.After45 min at_370C,

nonattached Ewereremoved,the

M40

werefixed in 0.25% glutaraldehyde inPBS,andattachedEwerecountedby phase-contrast microscopy.

Uptake of

zymosan

by

humanM4 andPMN.Human blood

mono-cyteswerecultivated for7d in IM+10% heat-inactivated humanserum

onsterile

glass

chips,

washed

extensively,

and cultivated forafurther2

hinIM,whichwas

replaced

with fresh medium

prior

toaddition of 10

til

of 1 mg/mlboiledunopsonizedzymosan.After 10minat

370C,

un-bound zymosan

particles

wereremoved

by

extensive

washing.

Freshly isolated human

peripheral

blood PMNwereadheredfor 15 min inIM

totissueculture

plastic

dishes which had been

preincubated

with 10% heat-inactivatedhumanserumasabove andcoatedwith 25gg/ml

poly-L-lysine.

Unopsonized

oriC3b-coated zymosan

particles

wereaddedfor

15minat

_37°C.

Electron microscopy

Scanning. Glasschipswithadherent cellswerefixed for 30 min in2%

glutaraldehydein0.1 Mcacodylatebuffercontaining1%sucrose.They

werewashedinbuffer, postfixed in 1%osmiumtetroxide for45 min, dehydratedingraded alcohols,andcriticalpointdried(Polaron

Equip-mentLtd.,Watford, England).After sputtercoatingwithgold (Polaron E5

100), they

were examined in aJEOL 100 CX electron microscope (JEOL, London, England)withASIDscanningattachmentoperatingat

40kV.

Transmission. AdherentPMN werefixed insitu inglutaraldehyde asabove. Themonolayerwasrecoveredwithpropylene oxide,washed in propylene

oxide,

pelleted, dehydrated, and embedded in Araldite. Sectionswere cut withadiamond

knife,

stained with leadcitrate,and examined ina JEOL 100 CX

microscope

operatingat80kV.

Respiratory burst assays-superoxide

anion

(O2)

release

PMNand

freshly

isolated monocyteswereassayedinsuspensionas de-scribed

_(22).

_Superoxide

dismutase-inhibitable reduction of ferricyto-chromec wasmeasuredspectrophotometrically afterincubation of cells withunopsonizedorserum-opsonizedzymosan -100particles per cell, orPMA.AssayswithPMNwereterminatedafter<10 min ofincubation thatwasfoundto

give optimal

reduction,beforesignificantreoxidation of

cytochrome

c;monocyte assayswere continued for upto 60 min. Cultivated M4

(7 d)

and PMN(15 min)werealso assayed with various stimuli after adherenceto _{glass coverslips.} With PMN thecoverslips

werepreincubatedfor7dwith IM+ 10% heat-inactivated human serum andwashed welltoreduceadherence-induced release of

O2.

Nitroblue tetrazolium (NBT) reduction

7-d human

_M40

and 15-min adherent PMN onglass coverslips were

incubated in KRPG aloneortogether, with or without unopsonized zymosan. Pretreatedglass coverslipswereusedforPMNalone. 200 mg ofNBTwasresuspendedin 1mlof KRPGandcentrifuged in a microfuge for 3 min. The supernatantwas filteredthrough a0.22-ym Millipore filterand30glwasaddedtoeachpreparation, with or without 5

g1

of

1

mg/mi

zymosan. After15minat37°C,thecells were washed extensively

withPBS,fixed in 0.25% glutaraldehyde in PBS and viewed by phase-contrastand

bright-field

microscopy.

Results

Biochemical evidence that iC3b

is

deposited on zymosan

and

confirmation of

C3

andfactor

H

secretion

by

human monocytes

We

_sought

direct

evidence

that

_iC3b,

aknown

ligand

ofthe CR3

(2),

was

_present

on_zymosan

_particles

that had been

_exposed

to

monocytes/MO

without serum and with inhibitors presentto preventuptake.Wetook

advantage

of the covalentinteraction between activated C3 andzymosantoelute bound C3

cleavage

products selectively. Fig. 1 A shows

that

the zymosan eluate containedpolypeptides that

co-migrate

with the three chains of iC3b:43,000, 68,000, and

70,000

molwt.In

addition,

it contains material (108,000 molwt) co-running with the C3ba'-chain and

acharacteristic38,000-molwt

degradation

fragment

of the

iC3b

43,000-mol

wtchain. Theresultis consistent with

deposition

of C3bonthe zymosan

surface,

followed

by degradation

ofover 50% of the C3btoiC3b.

Inprevious studieswedemonstrated thepresenceof native C3 inhumanmonocyteculturesupernatantbybiosynthetic la-beling and found low levels of factor I

plus

cofactor

activity

(?

factorH) by measuring iC3b formation from C3b(1). To estab-lish thatmonocytessynthesizeandsecretefactor Hunderassay conditions, we partially purified 35S-labeled factor H and C3 from 1-d cultured

monocyte-conditioned

medium

by

ion-ex-change chromatography. Analysis by SDS-PAGE in both

re-ducing and nonrere-ducing(not shown) conditions

(Fig.

1

B)

in-dicated that both proteins were--detectable by

fluorography,

compared with standards. Thesewereminor labeledcomponents in that C3 contained <3% of the protein-bound 35S, whereas factorH contained <0.25%.

Role of C3 in uptake of zymosan by

_monocytes/Mo

Previous studies (1) showedthat MO 1,amouseanti-human CR3 monoclonal antibody, inhibited uptakeof'25I-zymosan by hu-man monocytes and 7-d cultured

Mo.

To extend these

obser-

BH-A

116k-

70k--achain

-108k

-:-chain

-68k

-43k -38k

a b c a b c

Figure1.(A)SDS-PAGE of material eluted fromzymosanand radio-iodinatedasdescribed.Samplesreducedbefore electrophoresis. (Tracksaandb)Coomassie blue-stained standards; (track c)

autora-diograph.

(Track a)C3a-and ,8-chains.(Trackb)C3bdigested with factorH(15% wt/vol)andfactorI(1% wt/vol)toform iC3b.The up-permostbandis factorH.Trackcshows thechains of iC3b (43,000

and68,000-70,000-molwtdoubletwhich is unresolved)andC3ba',a

108,000-molwt_cleavageproduct of C3b. The 38,000-molwt

degrada-tion

fragment

ofthe43,000-molwtchain of iC3b is alsopresentinthe

standard.K,thousands in molecularweight.(B) SDS-PAGE(reduced)

of 35S-labeled factorHandC3 in the eluatefromaDEAE-cellulose column.Theelutionpositionof thetwoproteins overlappedand a

fluorograph

ofadjacentcolumnfractions isshownin bothtracks

la-beledb. TracksaandcshowCoomassie blue-stainedpurifiedfactorH andC3, respectively.Inadditiontofactor HandC3a-andj3-chains, the tracks labeledb also contain otherprominentcomponents

(95,000, 58,000, 43,000,and38,000molwt). The 95,000-molwt bandresemblesonthebasisof sizeandchargeagelatin-binding

monocytesecretion product described by Vartioetal.(23);the other

productshavenotbeenidentified.

Table I. Effects ofAnti-CR3Antibodies and Anti-C3 Fab on Uptake of Unopsonized Zymosan by Human Monocytes

Percentinhibitionof uptakeaftertreatment with

Ingestionof

Cells 1251-zymosan MOI OKMIO IB4 Anti-C3Fab Mannan

cpm 1-dcultivated

human monocytes 10,942±690 70±10 64±6 10±6 38±6 12±4

Humanmonocytes were cultivated for 24 h in 24-well tissue culture trays in IM +

10%o

heat-inactivated humanserum. 2 h prior to assay the cells were washed free of serum and reincubated with IM alone. Monolayers were incubated in the presence or absence of inhibitors for 15minprior to incubation with

1251-zymosan

for 10minat370Cin the continuous presence of inhibitor (1). Attached zymosan was removed by 15min of trypsintreatment(1).Results showmean±standarddeviation of triplicate wells from one experiment, representative of five independent

experi-ments.

vations we studied inhibition of zymosan uptake by two other

anti-CR3 antibodies

(1

1),

OKM 10 and

IB4,

and

by

the

Fab

frag-ment

of

arabbit anti-human C3

antibodies.

OKM10,

like MOI,

recognizes the ligand binding site of CR3, whereas IB4 binds to an

epitope

remote

from

the

binding site

(11). The results show

(Table

I) that MO1 and OKM10antibodies

inhibited

zymosan

ingestion

byadherent monocytes

by 70% and 64%,

respectively,

whereasinhibition by IB4 was 10%

of

untreatedcontrols. The

anti-C3

Fab

inhibited

38% of

uptake.

Mannaninhibition in these

experiments

was low in

keeping

with the low levels of MFR

expressed

by 4-d cultivated

monocytes

compared with cells

cul-tivated for

7d

(1,

24). Indirect

bindings

assays

confirmed

that all monoclonal

antibodies

boundtothecells

(results

not

shown).

These

inhibition studies indicate that

theuptake

of

zymosanby recently isolated monocytes

is

in large part

mediated

by the CR3

ligand binding site and together with the results

shown

in

Fig.

1 are

strongly

suggestive

that the

ligand is indeed iC3b deposited

onthe zymosan.

CR3

and

MFR

are

distinct receptors

Earlier studies

showed that zymosan that hadnotbeen

opsonized

with fresh

serum

retained

much

of its

biological

activity (22)

and that the MFR

contributed

to

uptake

of

these

particles (19).

Uptake of

"unopsonized"

zymosan

by

mononuclear

phagocytes

can be

mediated by CR3

as well as MFR

depending

on

the

expression of

each receptor

by

particular

MO

populations

(1).

Recent

studies

show that both receptorsare

required for uptake

of Leishmania donovani

promastigotes (8).

To

determine

whether these receptorscan

function

as

distinct

molecules in

phagocytic

recognition,

we

performed experiments

in

which each

receptorwas

selectively depleted from

the

_MO

plasma membrane

by

adherenceto

ligand-coated substrata

as

described by Michl

et

al.

(18). Receptor activities

weremeasured

by

the

ability

to rosette

defined EAiC3b

andas

specific,

mannan-inhibitable uptake of '25I-Man-BSA. Resident

peritoneal

MO

from

the mouse wereselected because these cells express both MFRand

CR3

andare

suitable

nonadherent cells

for modulation

experiments.

RPM were

cultivated

on

glass

coverslips

which

had been coated

with

the Fab

fiagment

of

ratanti-mouse CR3

(MAC-

1

antigen)

or mannan.

RPM

adheredon M

1/70

Fab

sur-faces

selectively

lost the

ability

to rosette

EAiC3b,

whereas MFR

activity

was

similar

tothat

of

cellsadheredtocontrol substrata

(Fig.

2 A and

legend).

The

reciprocal

experiment

(Fig.

2

B)

showed

that

MO-

on

mannan-coated

coverslips expressed

re-duced MFR

activity,

but

EAiC3b

rosetting

was

similar

to

con-trols. Therewas noloss

of

MAC-I

antigen

from

the

surface of

Mk

on a mannan

surface

as

determined by indirect

binding

studies using antireceptor antibodies (not shown). Neither the CR3 nor MFR was modulated on

_MO

cultivated on coverslips coated with the Fab fragment of 2.4G2, a rat anti-mouse Fc

Figure 2.

MO

receptors for

mannosyl-glycoconjugates

andiC3bare

distinct. RPMwerecultivated for4honcoverslipscoatedwith

man-nan,M1/70 Fab(anti-CR3),or oncontrol

coverslips.

Theywerethen incubated with

'25I-Man-BSA

orEAiC3b for 30 minat37°C. (A)

M4

adheringtocoverslipscoatedwithMl/70Fab and incubated with EAiC3b show loss ofrosetting. (B)

MO

adheringtocoverslipscoated withmannanandincubatedwith EAiC3b shows strongrosette forma-tion.Specific uptakeof'25I-Man-BSA by

MO

cultivatedon mannan

was0.20

ng/gg

cellproteincompared with 0.82 ng/ug cellprotein by

M4,

on M1/70orcontrolcoverslips.Results shownare

representative

of sixcoverslipsfromoneexperimentandare

representative

of three independentexperiments.

receptor(trypsin resistant) monoclonal antibody (25),

although

the Fc receptorwasselectively modulated (not shown). Therefore MFR and

CR3

can be

selectively modulated

andare

distinct

from each other and the

trypsin-resistant

Fc receptor.

Human PMN bind and ingest opsonized, but

not

unopsonized zymosan

PMN express receptors

for

the

cleaved

components of

active

C3, C3b (CR 1)

(26),

and

iC3b

(CR3) ( 11)

and monoclonal

anti-CR3

antibodies immunoprecipitate antigens

whichare

identical

to

MO

CR3

(11).

These

cells unlike

_MO

havenotbeenshown to secrete complement

proteins

andcanbeusedtotest

whether

interaction of zymosan

with

CR3

is direct

or

dependent

upon

an opsonin. We

compared uptake of

unopsonized

and

iC3b-coated zymosan by

freshly isolated

PMNand 7-d

cultivated

hu-man

_MO

in the absence

of

serum. At least

50-100

cellson

trip-licate coded

preparations

were

viewed

by electron

microscopy

to

determine binding and/or ingestion.

As

anticipated,

MO

were able to bind

(Fig.

3A andB) and

ingest unopsonized and

op-sonized

(not

shown)

zymosan

particles.

In contrast, PMNwere able to bind and ingest only

iC3b

zymosan

(Fig.

3 C

and

D).

Examination of electron

micrographs showed

that 22%

of

PMN in these sections bound or

ingested iC3b

zymosan

after

incu-bation for

15 min at

370C

compared with

2%

uptake of

unop-sonized

zymosan (n =

_{200). This method of analysis}

underes-timates

the

actual

percentage

of

PMN that

take

up

particles.

D

.L..

Figure 3. Phagocytosis of unopsonized andiC3b-opsonizedzymosan byhuman

MO

and PMN,respectively. (A and B) Scanning electron micrographs. 7-d cultured human

_MO-bound

unopsonized zymosan (arrows). (A)X4,500. (B) An enlargement of central region of cell,

X 18,000. (C and D) Transmission electron micrographs. Freshly

iso-latedPMNingested only

iC3b-zymosan.

(C)X9,900. (D)X 15,000.

PMNexposedtounopsonized zymosan bound or ingested very few particles. Note looseapposition betweenPMNplasma membrane and opsonizedzymosan surface inD.

2372 Ezekowitzetal.

.4&-..t,'.

Pr-,,.'_... ,.# ;*-:

1.

Phase-contrast microscopy confirmed that virtually all PMN could bind or ingest iC3b-zymosan, whereas uptake of unop-sonized zymosan was <10%.

Interaction of

_MO

and PMN in response to

unopsonized zymosan

Inview of the foregoing observations we asked whether

_MO

are able to opsonize zymosan particles for recognition by PMN. We

designed

an

experiment

in which

Mq

were exposed briefly to

unopsonized

zymosan under

serum-free

conditions, before

ad-dition of

PMN

which

by themselves would remain unreactive to

unopsonized

zymosan. We assayed respiratory burst induction as ameasureof the PMN response because this was suitable for

analysis

atthesingle cell level (NBT reduction) and preliminary

experiments

with PMN in suspension confirmed that PMN pro-ducedcopious amounts

of

O2

when triggered with PMA or fresh serum-treated zymosan, but not in response to unopsonized zymosan (legend to Fig. 4). We chose 7-d cultured blood mono-cytesasthe source of

MO

in these cell collaboration experiments because

prolonged adhesion

results in loss

of

respiratory burst

activity

tobackground levels ([27, 28] Fig. 4 legend), while the

ability

to bind and ingest unopsonized zymosan via the CR3 and to secrete complement

(5)

is substantially retained.

Fur-thermore, experiments

with

freshly

isolated monocytes con-firmed that these cells themselves release

O2

upondirect chal-lenge with unopsonized zymosan (not shown) and were unsuit-able

for

co-cultivation experiments. Finally, we

found

that

isolated

PMN weretriggered to release

0°

upon initial contact with fresh glass coverslips (27). This could be prevented by using glass

coverslips

which had been coated by prior incubation in

heat-inactivated

human serum, as used

for

MO

cultivation (Fig.

4legend).

We

therefore incubated

7-d

MO

with unopsonized

zymosan

for

10 minat

370C,

in the absence

of

serum,

prior

to

addition

of

freshly isolated

PMN. The PMNclustered around Mc

con-taining

zymosan and released

large

amounts

of

O2

as

determined

by

NBT

reduction and

cytochrome

c

reduction

(Fig.

4 EandF and

legend).

When the

interaction continued for

longer

than 15 min, the

MO

monolayer

was

destroyed

andtherewasalsoPMN death

(not shown).

Fig.

4C and D show that in the absence

of

zymosan therewasno

specific

MO-PMN

aggregation

and very

little

°2 release. These

experiments indicated

that

_MO

wereable to

provide opsonins and/or chemotactic products

that enable PMNto respond tothe presence

of

zymosan

with

a

vigorous

respiratory

burst.

Discussion

In

this study

weshow that

(a)

zymosan

particles

incubated with

MO

without

serum

accumulate iC3b

ontheir

surface;

(b)

mono-clonal antibodies directed

against

the

ligand binding

site of CR3 inhibit zymosan

uptake by

MO;

(c)

MO

receptors for

mannosyl

residues

and for iC3b that

mediate

zymosan

uptake

aredistinct and can be

independently modulated; (d)

human PMN that express CR3

bind

and

ingest

zymosan

only

after exogenous

op-sonization; (e)

7-d cultivated human Mq enable PMNto react

with

unopsonized

zymosan in the absence of serum, with

re-sultant release of °2

-Wehave

provided

further evidence

indicating

arolefor

MO-derived alternative

pathway

components and the CR3 in local

opsonization

of

zymosan.In

addition

to

_C3,

wedemonstrated factor H in

monocyte-conditioned

medium

by

biosynthetic

la-beling

and eluted polypeptide chains of iC3b and C3b from zymosanparticles incubated with monocytes in the absence of serum. The use

of

zymosan as a trap for covalently bound C3 provides a powerful direct probe to detect C3 and related mol-ecules. The

alternative

pathway components found in

monocyte/

MO-conditioned

media represent only a small fraction ofsecreted protein and have been detected by sensitive functional assays or

partial purification using affinity (1)

orion-exchange chro-matography. Other proteins found in relatively large amounts

after biosynthetic

labeling and not thought to be complement components includea95,000-mol wt polypeptide similar to a

major gelatin-binding

protein

described

by Vartio et al. (23).

This

molecule is distinct from fibronectin, which can also be secretedby human

MO

(29), and there is no evidence that it is

opsonic

orable to modulate complement receptor function (30), like

fibronectin.

Selective

inhibition

of

zymosanuptake by Fab

fragments of

anti-C3

and bytwo antibodies that

interfere

with

binding of

iC3b ligands (M01, OKM10), but not by IB4 antibody, which bindsto a

different

CR3 epitope,

provides additional

evidence that zymosan

binds,

atleast in part,tothe

ligand binding site

of

MO

CR3

although

other receptorsarealso

likely

tobe involved (1, 8). The exactnature

of

the

ligand-receptor interaction

is not known, but

inasmuch

asC3

is

a

glycoprotein it

has been pos-tulated that

CR3

might

bea

lectin-like

receptor

(31).

Zymosan-CR3

interaction

would

therefore

be

explained

bya

direct

inter-action of

asugar

shared with iC3b

hence

accounting for

inhi-bition by

the CR3

antibody.

Our

study

makes

this

explanation

unlikely

in thecase

of

monocyte/M4

interaction with

zymosan. Weshowthat CR3 and the

lectin-like MFR,

which

contributes

tozymosan

uptake

by

MO

populations

with this

receptor, can be modulated

independently. Modulation of complement

re-ceptors in RPM demonstrated here

with anti-CR3

antibody

re-sembles that observed with human monocytes

(30),

but

differs

from a

previous

report where substratum-bound C3

failed

to modulate the CR3 in RPM

(18). Successful modulation

may be

the

result

of

agreater

affinity

of

themonoclonal

antibody

than the

ligand for

the receptor. It

is

possible

that the

failure

of

RPM to

ingest

iC3b

ligands

isnot a

function

of lack

of mobility of

the receptors, but

reflects

theirstate

of nonactivation

(32).

In

addition,

solublemannandoesnot

inhibit

EAiC3b

rosetting

nor

does anti-CR3

antibody

inhibit

MFR

activity (1). Together

these results make it

unlikely

thatmannose

plays

arole in the

MO-iC3b/CR3

interaction

studied

here. Wedo notknow whether thetworeceptors

recognize

distinct

ligands

onthe zymosan

sur-face

or a

complex

ligand

suchasthe

mannose-rich

peptidoglycan

found

in cell walls

of

yeast,

parasites,

andsome

bacteria which

mayact asbothanacceptor

surface

for

initial

complement

de-position

and express

exposed

mannoseresidues. Further

depo-sition of

complement

could mask

fre

mannosyl-residues

sothat

uptake

of

serum-opsonized

targets

proceeds

almost

exclusively

via CR3 (1,

8). Finally,

exogenous

(30, 33)

or

Mkderived

fi-bronectin

(23)

caninfluence

phagocytic recognition by

director

indirect involvement of CR3.Itis

possible

thatiC3b isa

major,

butnotexclusive

ligand

for the active site of CR3 in these

com-plex interactions.

Although

PMNand

MO

share CR3 receptors,

PMN,

unlike

M4,

did not

ingest

zymosan

particles

in the absence

of

an ex-ogenous

opsonin (cf.

9,

10),

nordid

unopsonized

zymosan elicit a

respiratory

burst

by

PMN.It

is

possible

that the

CR3

andother lymphocyte function

antigen (LFA)

molecules inPMN

_(34,

₃₅₎

alsocontribute

_directly

to PMN

adhesion

to substrataor

.ot

4i

..L

Figure 4.Unopsonizedzymosantriggersarespiratory burst by human

PMNcocultivated withhuman M+. NBTreduction shownin phase-contrastmicrographs,except DandF, whicharebright-field

micro-graphs. O2releaseinparallelexperiments performed inabsence of NBTisgivenin brackets. (A) PMN cultivated 15 min on pretreated glasscoverslips(seeMethods)withNBT and zymosan(58 nmol

O2/

10 min).

Noteoccasional

clumps

ofextracellular zymosan;the

vast majority of PMN are free. (B) 7-d cultured

M4)

withNBT and zymosan(20nmol

05/10

min). Arrows show ingested zymosan

parti-cleswhichare notreadily discernedatthis magnification. (C-F) FreshlyisolatedPMNaddedto7-d

My0

inpresenceofNBTwithout

ticulate substances, in that patients genetically deficient in surface

expression of this family of

molecules

display

complex

defects

in PMN

adhesion,

even inthe absenceof

complement.

However,

(CandD, 32nmol05/10 min)orwith (EandF, 218nmol05/10 min)zymosan. Note the extensiveaggregation ofPMNandmarked

NBTreductionseen only in thepresenceofzymosanparticles, which

areobscured. Controlexperiments withPMAand serum-treated zy-mosan(STZ) showed0° release/1Omin asfollows:PMNalone +PMA230, +STZ 182; 7-d

MO

alone + PMA 44; PMNwith

MO

+PMA268, +STZ 204. Background release onsurface without

trig-ger(20-60)subtracted. Results show averageofduplicate valuesin a

singleassay,representative ofmorethanthreeindependent

experi-ments.

the ligands for other lymphocyte function antigens have not been identifiednor canthe defectsbe ascribed toCR3 dysfunc-tion alone becauseat least three surface molecules share a

mon ,8-chain which is lacking in leukocytes of these patients (36). Further studiesarealso needed withmonocytes from such patients.

We obtained evidence that

_MO

can interact withzymosan to induce a respiratory burst in PMN. In co-cultivation exper-iments we noted that PMN are attracted to zymosan-bearing M+, aggregate, and degranulate with resultant destruction of macrophagesaswellas PMN,presumablyaresult of the vigorous respiratory burst. These reactions dependedonthe simultaneous presence of both cell types andzymosan and could be due to reaction of PMN with "opsonized" zymosan orwithproducts ofthe M4-zymosan interaction. We havenotidentifiedthe pos-sible

MO

mediators involved, e.g.,complement products C3a, C5a (37), and/or leukotrienes (38), nor have we determined whether the CR3 is able to trigger a respiratory burst in PMN (39, 40). However, these observations indicate thateven non-activatedMkcancollaborate withPMNin host defense. Initial migration of PMN to an extravascular site maynotnecessarily beaccompanied by increased effilux of highlevelsofallplasma complement components. Local production of complement and other

potential

chemotactic products and opsonins bytissue

_MO

could serve not only to mediate destruction of pathogens by

MO

themselves, but alsoto recruit andactivate PMN,and thus to initiate and amplify a local inflammatory response.

Acknowledgments

We thank Mr. Entwistle and staff from the Oxford Regional Blood

Transfusion Service for their cooperation and Dr. Sam Wright, The RockefellerUniversity, forhelpfuldiscussions. Mr.Stan Buckingham and Mrs.ElwenaGregory helpedtopreparethemanuscript.

This work was supported bya grantfrom the Medical Research Council, United Kingdom.

References

1. Ezekowitz, R.A. B., R. B. Sim,M. Hill, and S. Gordon. 1984. Localopsonizationby secreted macrophage complementcomponents.

Role ofreceptorsforcomplement inuptakeof zymosan. J. Exp. Med. 159:244-260.

2. Beller, D. I., T. A. Springer,andR. D. Schreiber. 1982. Anti

Mac- Iselectively inhibits themouseandhuman typethreecomplement receptor.J.Exp.Med. 156:1000-1018.

3.Arnaout, M. A., R. F.Todd,I.N.Dana, J.Melamed,S. F.

Schloss-man,andH. R.Colten. 1983. Inhibition ofphagocytosisofcomplement C3 or immunoglobulin G-coated particles and of C3bi binding by monoclonal antibodiestoamonocyte-granulocytemembrane glycopro-tein(MOI).J.Clin.Invest.72:171-179.

4. Stahl, P. D., and S. Gordon. 1982.

Expression

ofa mannosyl-fucosylreceptorfor

endocytosis

onculturedprimary macrophagesand theirhybrids.J.CellBiol. 93:49-56.

5.Colten,H. R.1982.Biosynthesisof the MHC-linkedcomplement proteins (C2,C4 and factorB) bymononuclear

phagocytes.

Mol.

Im-munol. 19:1279-1285.

6.Brade, V.,and H. U.Beuscher. 1984. Formation of EAC 142 and EAC 1423 withmacrophage culture supernatantcontainingthesecreted complementcomponents CItoC3.

Immunobiology.

166:177-189.

7. Whaley, K. 1980. Biosynthesisof thecomplementcomponents and theregulatoryproteinsofthealternativecomplement pathway by

humanperipheral bloodmonocytes. J.Exp. Med. 151:501-516. 8.Blackwell,J.M.,R. A.Ezekowitz,M. B.Roberts,J. Y.Channon,

R.B.Sim,and S. Gordon. 1985.Macrophage complementand lectin-likereceptorsbind Leishmania in the absence ofserum.J.Exp. Med.

162:324-331.

9. Henson, P. M. 1971. Theimmunologic release of constituents

from neutrophilic leukocytes. 1. The role of antibody and complement on nonphagocytosable surfaces or phagocytosable particles. J. Immunol. 107:1535-1545.

10. Roos, D., A. A. M. Bot, M. L. J. Van Schaik, M. de Boer, and M. R. Daha. 1981. Interaction between human neutrophils and zymosan particles: the role of opsonins and divalent cations. J. Immunol. 126: 433-440.

11. Wright, S. D., P. E. Rao, W. C. van Voorhis, L. S. Craigmyle, K. lida, A.Talle, E. F. Westberg, G. Goldstein, and S. C. Silverstein. 1983. Identification of theC3bireceptor of human monocytes and mac-rophages with monoclonal antibodies. Proc. Natl. Acad. Sci. USA. 80: 5699-5703.

12. Micklem, K. J., R. B. Sim, and E. Sim. 1984. Analysis of C3 receptor activity on human B lymphocytes and isolation of the comple-ment receptor type 2 (CR2).

Biochlem.

J.224:75-86.

13. Badwey, J. A., J. T. Curnutte, J. M. Robinson, J. K. Lazdins, R. T. Briggs, M. J. Karnovsky, and M. L. Karnovksy. 1980. Comparative aspects of oxidative metabolism of neutrophils from human blood and guinea pig macrophages.J.Cell. Physiol. 105:541-552.

14.Sim, R. B., T. M. Twose, D. S. Paterson, and E. Sim. 1981. The covalent binding reaction of complement component C3. Biochem.J. 193:115-127.

15.Sim, R. B., and R. G. DiScipio. 1982. Purification and structural studies on the complement system control protein BH (factor H). Biochem.J.205:285-293.

16. Hsiung, L. M., A. N. Barclay, M. R. Brandon, E. Sim, and R. R. Porter. 1982. Affinity purification of human C3binactivator by monoclonal antibody affinity chromatography. Biochem. J. 203:293-298.

17. Malhotra, V., and R. B. Sim. 1984. Role ofcomplement receptor CRI inthe breakdown of soluble and zymosan-boundC3b.Biochem. Soc. Trans. 12:781-782.

18.Michl, J., M. M. Pieczonka, J. C. Unkeless, and J. C. Silverstein. 1980. Effects ofimmobilized immune complexes on Fc- and complement-receptorfunction in resident and thioglycollate-elicited mouse peritoneal macrophages.

J.

Exp. Med. 150:607-623.

19.Sung, S. J., R. S. Nelson, and S. C. Silverstein. 1983. Yeast

man-nansinhibit binding and phagocytosis of zymosan by mouse peritoneal macrophages. J. Cell Biol. 96:160-166.

20.Tack, B. F., and J. W. Prahl. 1976. Third component of human complement: Purification from plasma andphysiochemical character-isation. Biochemistry. 15:4513-4521.

21.Ezekowitz,R. A.B., J. Austyn,P. D.Stahl, and S. Gordon. 1981. Surface properties of bacillus Calmette-Guerin activated mouse

mac-rophages. Reduced expression of mannose-specific endocytosis,Fc

re-ceptors and antigen F4/80accompanies induction of Ia.J. Exp. Med. 154:60-76.

22. Berton, G., and S. Gordon. 1983. Modulation of macrophage mannosyl-specific receptors by cultivationonimmobilised zymosan. Ef-fectsonsuperoxide-anion release andphagocytosis. Immunology. 49: 705-715.

23. Vartio, T.,T.Hovi, andA.Vaheri. 1982. Humanmacrophages synthesize andsecrete amajor 95000 daltongelatin binding protein dis-tinct fromfibronectin. J.Biol.Chem.252:8862-8866.

24.Mokoena, T., and S. Gordon. 1985.Humanmacrophage acti-vation. Modulation of mannosyl, fucosylreceptoractivityin vitroby

lymphokines,

gamma and alpha interferons and dexamethasone. J. Clin.

Invest.75:624-631.

25. Unkeless,J.C. 1979. Characterisation ofamonoclonalantibody directedagainst mouse macrophageand lymphocyte Fcreceptors. J.

Exp. Med. 150:580-596.

26. Fearon,D. T. 1980. Identification of the membrane

glycoprotein

thatis theC3breceptor ofthe humanerythrocyte,

polymorphonuclear

leukocyte,B

lymphocyte

and monocyte. J.Exp.Med. 152:20-35.

27.Berton, G.,andS.Gordon.1983.Superoxidereleaseby

peritoneal

and bonemarrow-derivedmacrophages. Modulation

by

adherence and cell activation.Immunology.49:693-704.

28.

Nakagawara,

A.,

N. M.de

Santis,

N.

Noguiera,

andC. F. Nathan.

1982.Lymphokines enhance the capacity of human monocytes to secrete reactive oxygenintermediates. J.Clin. Invest. 70:1042-1048.

29.Alitalo, K., T. Hovi, and A.Vaheri. 1980. Fibronectin is produced by humanmacrophages. J. Exp.Med. 151:602-613.

30. Wright, S. D.,M. R.Licht,L.S.Craigmyle, and S. C. Silverstein. 1984.Communication between receptorsfor different ligands on a single cell: ligation of fibronectin receptors inducesa reversible alteration in thefunction of complement receptors on cultured human monocytes.

J.Cell Biol. 99:336-339.

31. Ross,G. D., J. A. Cain, and P. J. Lachmann. 1985. Membrane complement receptor type three (CR3) haslectin-like properties analogous

tobovine conglutinin and functions asareceptorfor zymosan and rabbit erythrocytes as well as a receptorforiC3b. J. Immunol. 134:3307-3315. 32. Griffin, T. A., and F. M. Griffin. 1979. Augmentation of mac-rophagecomplement receptorfunction in vitro. I. Characterisation of thecellularinteractions required for the generation of a T-lymphocyte product thatenhances macrophage complement receptor function. J. Exp.Med. 150:653-675.

33.Pommier, C. G., J. O'Shea, T. Chused, T. Takahashi, M. Ochoa, T.B.Nutman,C. Bianco, and E. J. Brown.1984. Differentiation stimuli induce receptors for plasma fibronectin on the human myelomonocytic celllineHL-60.Blood. 64:858-866.

34. Sanchez-Madrid, F., J. A. Nagy, E. Robbins, P. Simon, and T. A. Springer. 1983. Characterization of a human leukocyte differen-tiation antigen family with distinct alpha subunits and a common beta

subunit: the lymphocyte function-associated antigen (LFA-l), the C3bi complement receptor(OKMI1/Maci) and the P150, 95 molecule. J. Exp. Med. 158:1785-1803.

35. Arnaout,M.A., H.Spits, C.Terhorst,J.Pitt, and R. F. Todd III. 1984.Deficiency of leukocyte surface glycoproteins (LFA-1) in two patients withMOI deficiency. Effects of cell activationonMOI/LFA-I surface expression in normal and deficient leukocytes. J. Clin. Invest. 74:1291-1300.

36. Springer, T. A., W. S. Thompson, L. J. Miller, F. C. Schmalstieg, andD. C. Anderson. 1984. Inherited deficiency of the Mac-l,LFA-I, p150, 95 glycoprotein family and its molecular basis. J. Exp. Med. 160:

1901-1918.

37. Fernandez, H. N., P. M. Henson, A. Otani, and T. E.Hugh. 1978.Chemotactic responsetohuman C3a and CSa anaphylatoxins. I. Evaluation of C3a and CSa leukotaxis in vitro and under simulated in vivo conditions. J. Immunol. 120:109-114.

38. Rouzer, C. A., W. A. Scott, A. C. Hamill, and Z. A. Cohn. 1982. Synthesisof leukotriene C and other arachidonic acid metabolitesby mousepulmonary macrophages. J. Exp. Med. 155:720-734.

39. Yamamoto, K., and R. B. Johnston, Jr. 1984. Dissociation of phagocytosis from stimulation of the oxidative metabolic burst in

mac-rophages. J. Exp. Med. 159:405-416.

40.Wright,S. D., and S. C. Silverstein. 1983.Receptorsfor C3b and C3bi promotephagocytosis butnotthe release of toxic oxygen from humanphagocytes. J. Exp. Med. 158:2016-2023.