Binding of Cryptococcus neoformans by human cultured macrophages Requirements for multiple complement receptors and actin

Binding of Cryptococcus neoformans by human

cultured macrophages. Requirements for

multiple complement receptors and actin.

S M Levitz, A Tabuni

J Clin Invest.

1991;

87(2)

:528-535.

https://doi.org/10.1172/JCI115027

.

We studied the receptors on human cultured macrophages (MO-M phi) responsible for

binding encapsulated and isogenic mutant acapsular strains of Cryptococcus neoformans,

and whether such binding leads to a phagocytic event. Both strains required opsonization

with complement components in normal human serum in order for binding to occur. Binding

of the acapsular, but not the encapsulated, strain led to phagocytosis. MAb directed against

any of the three defined complement receptors (CR) on MO-M phi (CR1, CR3, and CR4)

profoundly inhibited binding of serum-opsonized encapsulated (and to a lesser extent

acapsular) organisms to MO-M phi. Immunofluorescence studies demonstrated migration of

CR to the area of the cryptococcal binding site. Trypsin and elastase inhibited binding of

encapsulated and, to a lesser extent, acapsular yeasts to MO-M phi. Binding of

encapsulated C. neoformans was profoundly inhibited by incubation in the cold or by

inhibitors of receptor capping and actin microfilaments. Thus, multiple CR appear to

contribute to binding of serum-opsonized encapsulated C. neoformans by MO-M phi.

Binding is an energy-dependent process that requires conformational changes in actin yet

does not lead to phagocytosis of the organism. In contrast, energy is not required for binding

of acapsular yeasts by MO-M phi and binding triggers phagocytosis.

Research Article

Binding of

Cryptococcus

neoformans by Human Cultured

Macrophages

Requirements for Multiple Complement Receptors and Actin

Stuart M. Levitz and Abdulmoneim Tabuni

EvansMemorialDepartmentof ClinicalResearch and the Department ofMedicine, The University Hospital, Boston UniversityMedicalCenter, Boston, Massachusetts 02118

Abstract

We studied the receptors on human cultured macrophages

(MO-MO)

responsible for binding encapsulated and isogenic mutant acapsular strains of Cryptococcus neoformans, and whether such binding leads to a

phagocytic

event. Both strains required opsonization

with

complement components in normal humanserum

in

order

for

binding to occur. Binding of the

acap-sular, but

not

the encapsulated,

strain led

to

phagocytosis.

MAb directed against

any

of

the three

defined

complement re-ceptors

(CR)

on

MO-MO

(CR1, CR3,

and

CR4) profoundly

inhibited binding

of serum-opsonized encapsulated (and to a

lesser

extent

acapsular) organisms

to

MO-MO.

Immnunofluores-cence

studies demonstrated migration of CR

to

the

area

of

the cryptococcal

binding site. Trypsin

and elastase inhibited

bind-ing of

encapsulated

and,

to a lesser

extent,

acapsular yeasts to

MO-Mo.

Binding

of

encapsulated

C.

neoformans

was

pro-foundly inhibited

by

incubation

in the cold or by

inhibitors

of receptor capping and actin

microfilaments.

Thus, multiple CR appear to

contribute

to

binding of serum-opsonized

encapsu-lated

C.

neoformans

by

MO-MO.

Binding

is an

energy-depen-dent

process

that

requires conformational changes

in

actin

yet

does

notlead to

phagocytosis

of

the

organism.

In contrast, en-ergy

is

not

required for binding of acapsular

yeasts

by

MO-Mo

and

binding triggers

phagocytosis. (J. Clin. Invest. 1991.

87:528-535.)

Key words:

cryptococcosis

-

divalent cations

*

fun-gus *

phagocytosis

-receptor

capping

Introduction

Infections

due to

the

encapsulated

yeastlike fungus

Cryptococ-cus

neoformans

are

significant

causes

of

morbidity

and

mortal-ity

in

patients

with

impaired

cell-mediated

immunity,

espe-cially

those with AIDS

(1). Macrophages

are

thought

to

play

a

major

role in host

defenses

against

this

opportunistic pathogen

(2-6).

In

the absence of

specific antibody,

a

functional

comple-mentsystem

has been shown

to

be

required

for

binding

of

C.

neoformans

to

human

macrophages

in vitro

(7,

8).

C.

neofor-mans

is

apotent

activator of the

complement

cascade,

with

large

amounts

of C3,

primarily

in the form of

iC3b, deposited

onthe

surface ofthe capsule after incubation

in

normal

human serum

(8-1 1).

Binding

by phagocytes of C.

neoformans

opson-ized with

human serum occursat a

relatively

lowrate com-AddresscorrespondencetoStuartM.Levitz, M.D.,RoomE540,The

University Hospital,88 E. NewtonSt., Boston,MA02118.

Receivedfor publication 21 February1990and inrevisedform 10 August 1990.

pared with mutant yeasts lacking capsule, despite being cov-ered with seemingly

sufficient

amounts

of iC3b

(5, 6, 12). Whereas capsuleconfers upon

C. neoformans

a negative sur-face charge and a hydrophilic sursur-face, the

antibinding

effects of capsulecanbedissociated from thesephysicochemical proper-ties(13, 14).

Human monocyte-derived macrophages

(MO-MO)'

and

other

phagocytes

have been shown to possess at least three

dis-tinct

complement receptors (CR) thatrecognize C3

degrada-tion

fragments (reviewed in reference

15).

CR1 (CD35)

and

CR3

(CDl

lb/CD

18) have greatest

affinity

for particles coated

with

C3b and iC3b, respectively, whereasCR4

(CDl

Ic/CD

18, p

150,95) preferentially

recognizes both

iC3b-

and C3dg-coated

stimuli. Although CR3 and CR4 share

a common

_P

chain

(CD1 8), their complement-binding activities

have

been

local-ized

to regions on

their distinct

a

chains

(15,

16). These recep-tors can

be

differentiated

on the

basis

of their reactivities

to MAb,

their sensitivities

to proteases and

their

requirements for temperature

and

divalent cations

(17-20).

In

addition

to

sites

for

binding

complement, the a

chain

of CR3

also appears to have

sites for binding

f-glucans

(21),whereas the common ,

chain ofCR3, CR4,

and

LFA-l

(the

third

member

of

theCD1 8

family)

has

been shown

to

play

a

role in cellular

adhesion and

in

binding unopsonized Histoplasma

capsulatum and bacterial

endotoxin

(18, 22, 23).

Whether a

microorganism is

able to

establish itself

as a

pathogen depends

to a

large

extentupon

which

receptors on

host

cells bind

to

it,

and the consequences

of

that

binding.

Therefore, in this

paper, we

investigated

the

requirements for

binding (attachment) of serum-opsonized encapsulated and

mutant

acapsular strains of C.

neoformans

to

MO-MO,

and whetherbinding leads to actual

phagocytosis (internalization).

Taken

together,

several

lines of evidence suggested

that

multi-ple CR

are

involved in

binding

of

serum-opsonized

encapsu-lated

C.

neoformans

by

MO-Mo. Moreover,

binding

was an

energy-dependent

process

that

required conformational

changes

in

actin

yet

did

not

lead

to

phagocytosis

of the

organ-ism.

In contrast,

binding

of

serum-opsonized

acapsular C.

neo-formans

by

MO-Mo

did

not

require

energy

and led

to

phagocy-tosis of

the yeast.

Methods

Materials. Allreagentsused,except where noted

otherwise,

were ob-tained in thehighest qualityavailablefrom

Sigma

Chemical Co.

(St.

Louis, MO).Medium used in allexperiments,unless stated

otherwise,

was RPMI1640(Gibco, Grand Island, NY)supplementedwith L-glu-tamine,penicillin,and

streptomycin.

Human

neutrophil

elastasewas

obtained from Elastin ProductsCorp. (Pacific,MO). Sheep

erythro-1.Abbreviationsusedinthispaper: CR, complement

receptor(s);

EC3b andEiC3b, erythrocytescoated with C3b and

iC3b; MO-MO,

mono-cyte-derived macrophages;PHS, pooledhumanserum.

528 S. M.Levitzand A. Tabuni

J.Clin. Invest.

© The AmericanSocietyforClinical

Investigation,

Inc.

0021-9738/91/02/0528/08

$2.00

cytesselectively opsonizedwith C3b(EC3b) and iC3b(EiC3b)were

prepared asdescribedpreviously usingpurified components of the al-ternative complement pathway (24) andwere agenerousgift of Dr. Simon Newman (Cincinnati, OH).

Opsonins. Pooled human serum(PHS) wasobtained by pooling serumfrom 10 healthy volunteers andwaskept inaliquotsat-70'C until use.PHSwasheat-inactivated by incubationat560C for 30 min. Twoseparatetechniqueswereusedtodeplete PHS of antibodies

reac-tive with C.neoformans.First,antibodieswereadsorbedtothe yeastsat

40C in the presence of EDTA, aspreviously described(25).Second, PHS wasincubated three times with ProteinASepharose4B (Pharma-cia FineChemicals;Piscataway,NJ)at00C in the presence of EDTA. PHS wasdialyzedagainstPBS, concentratedtoitsoriginalvolumewith

anAmicon filter and recalcified. This secondtechnique depletedover

97%of the IgG in the PHS(from 12.1 to0.3mg/ml)asmeasured by nephelometry.Anticapsularantibody,agenerousgiftof Dr. John Ben-nett(Bethesda, MD),wasprepared byimmunizingrabbits with

sero-typeDcapsularpolysaccharide,and usedat asubagglutinating

concen-tration (1:8,600 finaldilution).

Monoclonalantibodies.Mouse MAbdirectedagainsthuman mac-rophagesurface antigenswere asfollows: anti-CRI

_(IgGj,

bindsto

CR1, Becton-Dickinson, Mountain View,CA);OKM 10 and OKMI

(IgG2,

and

IgG2b,

respectively,bothbind to theachain ofCR3,kindly provided by Dr. Patricia Rao, Ortho Pharmaceutical Corp. Raritan, NJ) (26); anti-leu-M5

(IgG2b,

bindsto theachain ofCR4, Becton-Dickinson); 7C3 (IgG3, bindstositesonneutrophils,butnot

MO-MO,

kindlyprovidedby Dr. DavidMelnick, Bethesda, MD) (27);and IV.3

(IgG2b,

bindstoFcRII,kindlyprovided byDr.Michael Fanger, Han-over, NH)(28).Commerciallyobtained MAbweredialyzedtwice against500 vol of PBSto remove azide. MAbwere usedat afinal concentrationof 25

Ag/ml,

except for anti-CRIandanti-leu-M5 which wereused at 12.5 yg/ml and4

_,g/ml,

respectively. F(ab)fragmentsof OKM 10 wereprepared usingimmobilized papain, accordingtothe manufacturer'sdirections (Pierce Chemical Co.,Rockford,IL). Resid-ual Fcfragmentsand anyundigestedMAbwereremovedwith Protein A-Sepharose.Purity of thepreparationwasascertained by SDS-PAGE. Human

MO-Mo.

Heparinized bloodwasobtained byvenipuncture from normal volunteers and the PBMCpurifiedbycentrifugation over

acushion ofFicoll-Hypaque.Cellsweresuspendedat5 X 105/mlin mediumcontaining10% humanABmaleserumand100

_Al

added per wellof 96-wellflat-bottom polystyrene tissue culture plates (No. 25860, CorningGlassWorks,Corning, NY). Forsomeexperimentswith monoclonalantibodies, to conserve reagents,"halfarea" 96-well plates wereused (Costar,Cambridge, MA). Plates wereincubated in a 37°C,

5%CO2 humidifiedenvironment,washedafter 24 htoremove nonad-herent cells and culturedfor 6-21 d before challenge withC.

neofor-mans. Variation of the culture period over this range did not affect results. After the washing step, the remaining cells in the well were

>98% monocytesasassessed by morphology and nonspecific esterase staining. Forimmunofluorescentstudiesdetermining the localization of receptors,aswellasthe assaytodistinguish attached from phagocy-tosedorganisms,

MO-Mo

werecultured asabove, except in eight-chamber glass slides (Lab-Tek, Miles Scientific, Naperville, IL).

C. neoformans. Encapsulated,serotype D strain B3501 (alsoknown

asstrain MP415) and acapsular, isogenic mutant strain CAP67 were maintained and harvested as previously described (6). To facilitate rapid,accurateidentificationof yeasts, most experiments used

FITC-orRITC-labeled, heat-killed organisms. Selected experiments yielded similar results with live, unlabeled organisms. Yeasts were heat-killed by immersion ina60°C water bathfor 30 min and then labeled with either FITCorRITCaspreviously described (5, 6).C.

neoformans

were preopsonized by incubation in either PHS or anticapsular antibody for 30 minat37°C, followed bythreewashes inPBS.

Bindingassays. Wellscontaining

_MO-MO

werewashed with me-dium,except forexperiments examining divalent cation requirements where cells were washed with PBS containing glucose and the indicated cations. Monolayers were then treated with MAb, proteases or inhibi-tors, and

preopsonized

C.

neoformans

(2.5

x

I05

and2.5x

I04

per well

for theencapsulated and acapsularstrains, respectively)wasadded. The 10-fold higher inoculum of the encapsulated strain was necessary

tocorrectfor the lessefficient

MO-MO

binding of encapsulated,

com-pared with acapsular, organisms. For the protease experiments, pro-teaseswerewashedoff before addition of yeastssothat complement andother componentsonthe surface of the organisms wouldnotbe cleaved.

MO-MO

and yeastswereincubatedat370C for 5to30min following which unbound organisms were washed off and the mono-layerfixed with 1% formaldehyde in PBS. Using an inverted micro-scope at200x, 100-200 MO-MI were randomly selected under bright field and then examined under epifluorescence to determine the

ber ofyeasts associated per MO-M+. Binding index represents the num-berof cell-associated yeasts per 100 MO-M4. Due tovariability in baselinebinding indices between experiments, resultsareexpressedas

percent inhibition ofbindingwhichwascalculatedas:[1 -(binding indexofexperimental group/binding index of controlgroup)]X 100.

Assay todistinguishedattachedfromfully internalized organisms. Theabove binding assay does not distinguish between organisms bound to the surface of MO-M4 versus those that have been fully phagocytosed(internalized). To make thisdistinction,ourpreviously described Uvitex assay, with slight modifications, was used (5, 25).

MO-MO

were cultured as above, except in wellsofeight-chamber tissue culture slides (Lab-Tek, Miles Scientific)at4 x I0 per well. After three washeswith RPMI1640,

_MO-MO

werechallenged for 30 min at 370C withRITC-labeled C.neoformans in the presence of 10% PHS. Wells werewashed with PBS, and the cells incubated with 0.1% Uvitex (diaethanolorFungiqual,Ciba-Giegy, Basel, Switzerland) for 1 min at 230C. Wells wereagain washed in PBS and the cellsfixed with 1% formaldehyde.Usinganepifluorescentmicroscope,200 cell-associated yeasts per well were identified afterexcitation at546 nm, and then examined with UV excitation to determine which yeasts stained with Uvitex. Organisms that stain with Uvitex under these conditions are bound but not internalized, whereas those that do not stain are fully internalized (5, 25). Bound only organisms were also distinguished from those internalized by examination under phase-contrast micros-copy,asdescribed (29).

Immunofluorescence localization of complement receptors. Wells containingMO-M( were sequentiallyincubated with (a) unlabeled encapsulated or acapsular strains of C. neoformans for 10 min; (b)

z

z

U-

0

z

co

0

m

z

F-100

80

60

40

20

0

-20

NONE HEATED CN-ADS PrA-ADS

POOLED

HUMAN

SERUM

Figure

1. Role of

opsonins

in

_binding

ofC.

neoformans

to

_MO-MO.

MO-MO

wereincubatedat

370C

for 30 min withencapsulated (m)

or

_{acapsular (X)}

strains of C.

neoformans

in the presence of 10% PHS,

noPHS

_(NONE),

heat-inactivated PHS

(HEA

TED),

orPHSdepleted of reactive antibodies

by adsorption

with C.

neoformans

(CN-ADS)or

protein

A

_(PrA-ADS).

Resultsare

_expressed

aspercent inhibition of

TableI. InternalizationofSelectively Opsonized Encapsulated and Acapsular C. neoformans by

_MO-Mo

Strain Opsonins % Internalized

Encapsulated PHS 23±1

Encapsulated Ig-depleted PHS 19±2 Encapsulated Anti-cap Ig 64±6 Encapsulated PHS + Anti-cap Ig 75±5

Acapsular PHS 88±4

Acapsular Ig-depleted PHS 87±5

MO-M0

wereincuabated with a selectively opsonized encapsulated or acapsularstrainof C.neoformansfor 30 min at370C.Ofthe organ-ismsbound to

_MO-MO,

thepercentage that was completely internal-ized (phagocytosed) was determined by the Uvitex assay. Rabbit anticapsular immunoglobulin (anti-cap Ig) was used at a subagglu-tinating concentration. Results represent the means±SEM of two to four separate duplicate experiments.

~~A

MAb (5 gg/ml)for 10 min; (c) 1:2,000 dilution of Biotin-SP F(ab)'2 fragmentgoat anti-mouseIgG (JacksonImmunoresearch Laborato-ries,WestGrove, PA)for30min; (d) fluorescein-conjugated strepavi-din (50 ueg/ml;Jackson Immunoresearch Laboratories,WestGrove, PA)for 30min;ard(e)1%formaldehyde.Wellswerewashedsix times with medium between the above incubations. MO-M) containing C. neoformanswereidentified under light microscopy, and then observed under epifluorescence where the distribution of fluorescentstaining

wasnoted. For all experiments, controlsweredone omitting incuba-tions b, c,ord. Under suchconditions,fluorescencewas not seen.

Results

Opsonic factors

necessary

for binding

and internalization

of

Cryptococci

by

MO-Mo.

Initial experiments examined

the op-sonic

requirements

for

binding

of

encapsulated

and

acapsular

C.

neoformans

to

human

MO-M46

(Fig. 1). Omission of

opson-ins,

or heatinactivation of PHS, resulted in nearly complete

inhibition

of

binding

compared

with

wells containing

PHS.

Figure 2. InternalizationofencapsulatedC. neoformans by MO-M4,

asdetermined by the Uvitex assay.MO-M40wereincubated with RITC-labeled C.neoformansin the presence of 10% PHS for 30 min and then stained with Uvitex.Onlyextracellular yeasts stain with Uvitex. Brightfieldphotomicrograph(A) demonstratesMO-M4,some

of which contain cell-associated C.neoformans.Note that the yeast cells tendtoformrosettesaround the MO-M4. Fluorescent photomicrographof thesamefield taken with peak excitationat490

nm(B)and 545nm(C)demonstrates C.neoformanswith RITC and Uvitexstaining, respectively.Most of theRITC-stained organisms also stain withUvitex,indicatingthattheyareextracellular. Arrows pointto rareintracellularorganismsthat stain withRITC,but not Uvitex.

AX

Depleting PHS of

immunoglobulins

reactive with C.

neofor-mans

by either of

two

methods

hada

minimal effect

on

bind-ing. Thus, consistent with

previous

studies

(7,

8),

a

heat-labile

factor in

human serum,

presumably

complement, appeared

to be

required

for

binding.

Using

the

Uvitex

assay, wenext

examined whether

the

bound

organisms

were

internalized

(Table

I,

Figs.

2 and

3).

Of

the

cell-associated

serum-opsonized

encapsulated organisms,

only

23%

were

internalized

by

the

_MO-Mo.

Depleting

PHS

of

immunoglobulins reactive with C.

neoformans

did

not

signifi-cantly

alter

the percentage of

organisms

internalized.

However,

when the

yeasts were

opsonized with

specific

anticapsular

anti-body,

most

of the bound

organisms

were

internalized.

In con-trast, 88%

of acapsular

yeastswere

completely

internalized

after

opsonization

in PHS

alone,

even

if the PHS

was

depleted

of

immunoglobulins

reactive with C.

neoformans.

Identical

trends

were

observed

using phase-contrast

microscopy

instead

of

the

Uvitex

assay

(data

not

shown).

Figure3.Internalization of acapsularC.neofornansby

MO-M4,

as determined by the Uvitex assay. Fig. 3, A-C, correspond to Fig. 2, A-C,exceptacapsularC. neoformanswereused. Note that rosettes oforganismsare notseen onthebrightfieldphotomicrograph. The majority ofyeastcells do notstain with Uvitex, indicating that they

areintracellular.

Inhibition

ofbindingbyanti-receptorMAb.

A

panel

ofMAb

known

tobe

reactive with specific

CR and FcR on phagocytes wasexamined for its ability to inhibit binding of

PHS-opson-ized encapsulated and acapsular C. neoformans

to

_MO-MO

(Table II). To control for the specificity, EC3b and EiC3b

were

included

as targets

in

some

of the experiments.

MAb

directed

against

any

of the three defined CR

on

_MO-MO

(CR1,

CR3, and

CR4)

profoundly inhibited binding of

serum-opsonized

encapsulated (and

to a

lesser

extent

acapsular) C. neoformans

to

MO-Mo.

F(ab)

fragments

of OKM

10

also

inhibited,

al-though

to a

lesser

extent

than whole MAb.

In contrast, MAb

IV.3

directed against

FcRII, or

MAb 7C3

which

reacts

with

PMN(but not

_MO-MO),

did

not inhibit. If encapsulated yeasts were

opsonized

with IgG in lieu

of

PHS, then MAb directed

against

CR no

longer inhibited (data

not shown).

Binding of

EC3b

to

MO-M)

was not

inhibited

by any

of

the MAbtested.

Even

anti-CR

1

did

not

inhibit binding,

suggesting that this MAb

does

not

block

the

C3b

binding site

on

CR

1.

Both

of the

.:.%

A

'o

I

f,Tr

L

A

Figure4. Brightfield (A) andcorrespondingfluorescent (B) photomicrographsof

MO-MO

incubated withC.neoformansfor 10 min followed byanti-leu-M5 (anti-CR4). Immunofluorescent localization of anti-leu-M5 wasthenaccomplished by sequential incubations with

biotin-conjugated F(ab)'2

fragmentgoat anti-mouse IgGand

fluorescein-conjugated

strepavidin.Arrowsonthe

brightfield

photomicrograph pointtocell-associated yeasts.Notethatareasof

Table II. Effect ofAnti-receptor MonoclonalAntibodies

on

_{Binding of}

C.

neoformans, EC3b,

andEiC3btoMO-M)

%Inhibition of_binding

Encapsulated Acapsular

MAb Specificity strain strain EC3b EiC3b

Anti-CRI CRI 69±5 67±7 1±9 7±3 OKMI CR3 83±5 42±4 -5±12 81±2 OKM 10 CR3 76±3 36±5 -5±6 77±3 OKM 10F(ab) CR3 34±4 19±7 -4±6 15±8 Anti-Leu-M5 CR4 83±5 79±5 5±9 -7±11

IV.3 FcRII 2±7 0±6 ND ND

7C3 PMN 5±9 3±5 ND ND

MO-Mo

weretreatedwith the indicated MAb for 30 min and then challengedfor 30min withencapsulatedoracapsularC.

neoformans

that had been preopsonized in PHSorwith EC3b or EiC3b. Con-centrations of MAb used for theseexperimentswereatleast five times greater than the maximum inhibitory dose. Results are expressedas

percent inhibition of binding compared with

MO-Mo

not pretreated withantibody and represent the means±SEM of 3 to 10 separate triplicate experiments.

anti-CR3

MAb

(OKM

1

and

OKM

10)

profoundly

inhibited

binding

ofEiC3bto

_MO-MO.

Immunofluorescence

studies were nextperformed to see

whether CR

could be localizedtothe regionofthe cryptococcal

binding

site. Under

_{epifluorescent microscopy, intense staining}

of

_MO-MO

incubated with

either

anti-leu-M5 (Fig.

4) or

OKM

1 was seenin the

immediate vicinity

of

bound

encapsu-lated and acapsular C.

neoformans,

butwasnot seenin other areasof the cell. No

fluorescence

was

visualized

when

similar

experiments

were

performed

with

anti-CR

1,

perhaps because

by FACS scanning

_MO-MO

stained with anti-CR 1 had

approxi-mately

10-fold

less fluorescent intensity

than

MO-MO

stained

with either

anti-leu-M5

or

OKM

1 (datanotshown).

Divalent

cation requirement

for

binding.

Binding of

com-plement-coated particles

to

CR3

and

CR4,

butnotCR1,

has

been shown to

require

the divalent cations Ca++ and

Mg++

(16-18, 30, 31).

We

therefore

next examined the Ca++ and

Mg++

requirements for binding serum-opsonized C.

neofor-mans to

_MO-MO

(Table

III).Omission of Ca++ and Mg++ from the culture medium resulted in>80% inhibition of

binding

of both cryptococcal strains.

Inhibition of binding by proteolyticenzymes. Trypsinand human neutrophil elastase both have been showntoaffect ex-pression and function of CR (19, 20, 32). The effect of these proteases on

_MO-MO

binding

of C. neoformanswastherefore investigated (Fig. 5, aand b). MO-M4 were incubated with

trypsin

or

elastase, washed,

andthen incubated with preopson-izedyeastsfor either 5, 15, or 30min. Both proteases

signifi-cantly inhibited

binding

of encapsulated and,to alesserextent, acapsular C. neoformansto

_MO-Mo.

Inhibitionof

binding

was greatest at5

min,

and diminished progressively (althoughnot entirely)overthe time period studied. Because these proteases should

only'cleave

the susceptible

pool

of external receptors

fluorescence in Bcorrespondtothe location of yeasts. No fluorescence is visible on the twoMO-MOthat do not contain cell-associated yeasts.

TableIII. Influence ofDivalent Cations onBinding of C. neoformanstoMO-Mk

%Inhibition ofbinding

Encapsulated Acapsular strain strain

Nodivalent cations 83±2 81±2

MO-MO

werechallengedfor 5 min withC.neoformans preopsonized inPHS. Medium used for the control groupwasPBScontaining5.5 mMglucose, 1.3 mM

Ca",

and 0.8 mM

Mg".

Medium for the experimentalgroupwasPBS withglucose only.Dataareexpressedas

percent inhibition ofbinding comparedwith the control group and represent the means±SEM of fivetosix separate

triplicate experiments.

Table IV.

Effect

of

Temperatureon

Binding

of

C.

neoformans

toMO-M,

%Inhibition of binding

Encapsulated Acapsular

Temperature strain strain

40C 99±1 26±2

MO-MO

wereeither chilledto 4VCorkeptat370Cand then chal-lenged for 5 min withC.neoformans preopsonizedinPHS.Dataare

expressed as percent inhibition ofbinding comparedwith the370C group and represent the means±SEM of three separatetriplicate experiments.

(33, 34),

the loss of inhibition over time is likely secondary to replacement with

intact

receptors from the internal pool. Treatment

of

MO-MO

with elastase resulted in > 70% inhibi-tion of binding of EC3b and EiC3b, whereas treatment with trypsin inhibited EC3b binding by 26% and EiC3b

binding

by 94%.

Effect

of

temperature

and inhibitors of

cappingandactin on

cryptococcal

binding. Binding

(as

opposed

to

internalization) of

1

.00

z

- 80

a

o 60

z

0 401

m

I 20

n_{D 10 20 30}

TIME

(minutes)

O 100_

z z

s

-Z 80

IL 0 60

z

0 40

m

I 20

z

0

-Aoo

10 20 30

TIME (minutes)

Figure 5. Effect ofproteasesonbinding ofC.neoformansto

_MO-Mo.

(A)

MO-Mo

weretrypsin-treated by incubationfor30 min with 100

Ag/ml

trypsinat4VC,followedbytreatmentwith100

,g/ml

soybean

trypsininhibitorand threewashes withwarmRPMI1640.

MO-MO

werethenimmediatelychallengedfor5, 15, or30 minwith

encapsulated (-A-)oracapsular(---o---)strains of C. neoformans preopsonizedinPHS. Resultsareexpressedaspercentinhibition of

bindingcomparedwith

MO-Mo

nottreated withproteaseand

representthemeans±SEMofthreetriplicate experiments.(B) Asin Fig.3A,except

_MO-Mo

weretreatedwith10

Ag/ml

human leukocyte

elastasefor 60minat370C.

opsonized

erythrocytes

tophagocytes

does

not

require

energy

and

occurs at

40C (18, 35).

We

therefore studied whether the

situation

was

similar

for C.

neoformans (Table

IV).

For

these

experiments, a

5-min incubation

was

used.

Preliminary

experi-ments

established

that even at

4VC,

MO-MO

remained

adher-ent to

the cell

wells

during this incubation period. Nearly

no

binding of serum-opsonized encapsulated C. neoformans

to

MO-MO

occurred at

40C.

Incontrast,

_MO-MO

bound

serum-opsonized acapsular

yeasts at

4VC, although binding

was some-whatdiminished compared with at

370C.

The

loss of binding of PHS-opsonized encapsulated

yeasts at

40C

suggested

that

this

binding

was an

energy-dependent

process. Because a

population of

CR has been shown to be

linked

to the cytoskelaton

of

the cell (36, 37), we next

exam-ined whether inhibitors of capping

and

actin

filaments

would

inhibit

cryptococcal

binding.

Both azide and

cytochalasin

D

inhibited binding of PHS-opsonized encapsulated,

and to a lesserextent, acapsular yeasts to

_MO-MO

(TableV).

Discussion

The data presented here

provide

several lines of evidence that, taken

together, strongly

suggest

multiple

CR are

involved in

binding of serum-opsonized C. neoformans

by

MO-M4.

First,

monoclonal

antibodies directed against

CR1,

CR3, or CR4

each

profoundly inhibited binding of serum-opsonized

yeasts

to

_MO-MO

(Table

II).

The

inhibition

was not secondary to

nonspecific depression of

cellular

functions,

because these

TableV.

Effect

ofInhibitors

of

Capping andActinPolymerization on

Binding of

C.

neoformans

to

MO-Mq,

%Inhibition ofbinding

Encapsulated

Inhibitor strain Acapsular strain

Azide (1.5 X

10-2

M) 53±5 29±4 CytochalasinD(8X

IO-'

M) 91±2 72±5 CytochalasinD(2X

_10-'

M) 43±5 30± 5

MO-MO

were treatedwith the indicated inhibitorfor 30 min and then challenged for 30 min with encapsulatedoracapsular strains ofC. neoformans preopsonized in PHS. Resultsareexpressedas percent inhibition of binding compared with

MO-MO

notpretreated with inhibitors and represent themeans±SEM ofthree to four separate triplicate experiments.

T

B

0.L

0

1

MAb did not

inhibit binding

of organisms opsonizedwith anti-capsular Ig rather than PHS. Moreover, the MAb were

specific

in their effectson binding of EC3b and EiC3b. Second, immu-nofluorescentstudies localized CR3 and CR4 preferentially to the area of the

cryptococcal

binding site (Fig. 4). Finally,

pro-found inhibition of binding

occurred underconditions where divalent cations were

omitted

from the system (TableIII). Di-valentcations have been shown to be necessary for the binding of complement-coatedparticles by CR3 and CR4, but not CR 1 (17, 18, 30, 31).

Cleavage of

receptors

with

the proteases

trypsin

and human neutrophil elastase

profoundly inhibited

binding of

encapsu-lated,

and, to a lesser extent,

acapsular C.

neoformans

to

MO-MO

(Fig. 5).

Treatment

with either

trypsin

orhuman

leukocyte

elastase

has been shown to

cleave

CR1,

butnot

CR3,

on

acti-vated human PMN as measured by FACS

scanning (20, 32).

In

preliminary studies

in

our

laboratory, we

have found

that elas-tase treatment

of

_MO-MO

reduced

anti-CR 1

binding

by> 50% as

measured

by FACS

scanning,

while

having

noeffecton

bind-ing

of

OKM 10 or

anti-leu-M5.

In contrast,

trypsin

treatment

failed

to

significantly inhibit

binding

of

any of the three mono-clonals

(Levitz,

S.

M.,

and R. T.

Maziarz,

unpublished

data).

However, results

obtained by

FACS

analysis

maynot

correlate

with

functional

data. For

instance,

if a

protease cleaves the

complement

binding

siteor asite necessary for

signal

transduc-tion,

but does not

cleave the

epitope recognized by

the

MAb,

then the

receptor may be

functionally

inactivated withouta

decrement

in FACS

staining. Indeed,

inour

experiments,

bind-ing of EC3b

and

EiC3b

were both

inhibited

when

_MO-MO

were

protease-treated. Previous

investigators

also

found

inhibi-tion of

binding

of

EiC3bto

trypsin-treated

MO-M4

(1

9j.

Involvement

of

multiple

CR in

binding

ofa

complement-coated

particle

appearsnot tobe limitedto

C.

neoformans,

as

Payne and

Horwitz,

using

conditions similar

toours,

demon-strated

>

60% inhibition of

_MO-M,0

binding

of

serum-opson-ized

Legionella

pneumophilia by

MAb

directed

against

either

CR1

or CR3

(38).

In our

studies,

anti-CR1 and anti-leu-M5

profoundly inhibited

binding

of C.

neoformans

to

MO-MO

under conditions

where these MAb failed to

appreciably

in-hibit

binding

of EC3b and

EiC3b.

Thus,

these MAb appearto be

inhibiting

cryptococcal binding

by

mechanisms distinct

from

simple

physical

blockade

of

complement-binding

sites.

Generally,

binding

is

thought

tobea

passive

process in that

it

does not

require

energy and willoccurin the

cold,

whereas

actual

phagocytosis

is

energy-dependent, requires

aminimum temperature

of

18°C,

and is

accompanied

by assembly

and

cross-linking

of actin filaments

inthe cell's

cytoskeleton

(35).

However,

binding

of

serum-opsonized encapsulated

C.

neofor-mans to

MO-MO

was

profoundly

inhibited

by

cold and

_by

inhibitors of actin

assembly

and receptor

capping (Tables

IV and

V).

The

mechanism(s)

responsible

forthe

temperature

and

cytoskeleton

requirements

of

binding

remain

speculative.

One

possibility is

that the CRon

MO-M4 bind

complement

compo-nents on

encapsulated

yeasts very

weakly.

Active

migration

of

multiple

CR to the

organism,

perhaps

accompanied by

some

degree

of cytoskeletal

reorganization,

would then be

required

to reach the

critical

strength

necessarytohold

the

organism

in

place.

Interestingly,

after

receptor-ligand

binding,

several

re-ceptorson

phagocytes,

including

CR1

and

CR3,

are

recruited

to the

cytoskeleton

and

specifically

associate with actin

fila-ments

(36, 37). Moreover,

a

subset of

plasma

membrane CR3

has

recently

been

described

that appears tobe

involved

in

phagocytosis

of

particles

by

several different receptors

includ-ing

CR1

and FcR.This subset is attached to the cytoskeleton of thephagocyte and

its

function is inhibited

by both cytochalasin and MAb

directed

at sites on

CR3

distinct from

the

comple-ment-binding site

(39). For FMLP receptors, receptor occu-pancy has been shown to lead to

actin

polymerization, and,

concomitantly, conversion

of

the receptor complex

from

a fast

dissociating

to aslow

dissociating form (40,

41). Thus, cytoskel-etal

changes

could

contribute

tocryptococcal binding by

in-creasing

the

affinity

of CR

for

ligands on the yeast's surface. Recently, both attachment and phagocytosis of unopsonized

Histoplasma

capsulatum (but not complement-coated

erythro-cytes)

tohuman

_MO-MO

was shown to be

inhibited

by

cyto-chalasin

D

(23).

The

requirements

formultiple CR, energy, and actin may

help

explain why binding of

serum-opsonized encapsulated

C.

neoformans

is so

inefficient, despite the dense accumulation of

iC3b

on the surface of theorganism (8, 9, 1 1). However, it still remains unclear what factors on the organism contribute to

this inefficient binding.

For

complement-coated

erythrocytes,

clustering of

C3

fragments

has been shown to lead to more

efficient binding

by

phagocytes

than when the

C3

molecules are

randomly

distributed

on the

surface

(42). It is thus possible

that

the

distribution of C3 fragments

onthe capsular

surface

thwarts

efficient

phagocytic

binding.

Alternatively,

the

site

on the

C3

molecule that

interacts with

the

ligand-binding

site on

phagocytic

CR may not be

optimally

exposed so that

high-af-finity binding

can occur.Further

studies

on the molecular biol-ogy

of

complement

activation

and

deposition

on

cryptococcal

capsule

areneeded to answer these

questions.

After

binding of

a

particle

to

phagocytic

receptor(s),

inter-nalization of

theparticle may or may not take place,

depending

upon the

particular

receptor(s)

involved

and

its

state

of

activa-tion (24, 43).

Whereas FcR and mannose receptors

constituti-vely phagocytose bound

particles,

CR do not, although under

certain conditions they

canbe

upregulated

todo so(44). More-over,

CR

may collaborate with

internalization-promoting

re-ceptors to

phagocytose

a

particle.

For

example,

mannose re-ceptors can

synergize with

CR to

efficiently

internalize

comple-ment-coated

zymosan or Leishmania donovani

promastigotes

and

amastigotes (35, 44, 45). Consistent with

these

observa-tions, of

the

serum-opsonized encapsulated

C.

neoformans

bound to

_MO-MO,

only

23%

actually

was

internalized (Table

I).

Incontrast, 64 and

88% of

bound

Ig-opsonized encapsulated

and

serum-opsonized acapsular

C.

neoformans, respectively,

were

internalized.

The

ability of

MO-MO

to

phagocytose

serum-opsonized acapsular

yeasts

likely

relates to

exposed

mannans,

,B-glucans

orother

ligands

exposed

onthe cell wall

of

the

organism.

Such cell wall

ligands

are

masked

from

the

sur-face of encapsulated

organisms

by capsule

and thusare not

available

to

bind

to

their

appropriate

rec~eptor.

These

phagocy-tosis

dataare

consistent

with

results

previously reported using

mouse

peritoneal macrophages (5, 29).

Thus, our data

provide evidence

that

binding

of

serum-op-sonized

encapsulated

C.

neoformans by

MO-MO

occurs

by

a novel

mechanism.

First,

multiple

CRare

involved.

Second,

binding

isan

energy-dependent

event

that

doesnot

trigger

phagocytosis.

The

disparities

between our results

with

comple-ment-opsonized

C.

neoformans,

and those

reported

in the

liter-ature

using

other

complement-opsonized

particles,

emphasize

thevastrepertoire

of

responses the

macrophage

hasat

its

dis-posal.

Further

characterization of

the receptors involved in

rec-ognition

of C.

neoformans,

and the

biochemical

events

trig-gered

after

receptor

binding,

should

provide important

insights

into thepathogenesis of cryptococcosis aswell asthe

mecha-nisms of phagocytic

function.

Acknowledgments

The authors thankDrs.Simon Newman, John Bennett, PatriciaRoa, DavidMelnick,andMichaelFanger forgenerousgifts ofreagents,and Dr. Richard Diamond forhelpful discussions and critically reading this

manuscript.

This workwassupportedin part by grants to Dr. Levitz from the

National Institutesof Health(AI-00658 and AI-25780). Dr. Levitz is a

recipientof aPhysician-ScientistAward from the National Institutes of Health.

References

1.Diamond,R. D.1989. Cryptococcus neoformans.InPrinciples and Practice ofInfectiousDiseases. 3rd ed. G.L.Mandell, R. G. Douglas,and J. E.Bennett, editors. ChurchillLivingstone,New York.1980-1989.

2.Gentry, L. O., and J. S. Remington. 1971. Resistance against Cryptococcus conferred by intracellular bacteria and protozoa. J. Infect.Dis. 123:22-31.

3. Monga,D. P.1981. Role ofmacrophages in resistance ofmice toexperimen-talcryptococcosis. Infect.Immun.32:975-978.

4.Granger,D.L., J.R.Perfect, andD. T.Durack. 1986.

Macrophage-me-diatedfungistasisinvitro:requirements forintracellularand

extracellularcytotox-icity.J. Immunol.136:672-680.

5. Levitz, S. M.,and D. J.DiBenedetto. 1988. Differential stimulationof

murine residentperitoneal cells by selectively opsonized encapsulatedand acap-sularCryptococcus neoformans. Infect.Immun.56:2544-2551.

6.Levitz,S. M., and D. J. DiBenedetto. 1989. Paradoxicalroleofcapsule in

murine bronchoalveolar macrophage-mediated killing of Cryptococcus

neofor-mans.J.Immunol. 142:659-665.

7.Diamond,R.D.,J.E.May,M. A.Kane,M. M.Frank, and J.E.Bennett. 1974. The roleof the classical and alternate complement pathwaysin host

de-fensesagainst Cryptococcus neoformans'infection.J. Immunol. 112:2260-2270. 8. Kozel, T. R., G. S. T. Pfrommer,A.S. Guerlain,B. A.Highison,and G. J.

Highison.1988. Roleofthe capsule in phagocytosis of Cryptococcus neoformans.

Rev.Infect.Dis. 10:S436-S439.

9. Kozel,T.R.,B.Highison,andC. J. Stratton. 1984. Localizationon encap-sulatedCryptococcus neoformansofserumcomponents opsonic for phagocytosis

by macrophagesandneutrophils.Infect.Immun. 43:574-579.

10. Kozel,T.R., and G. S. T. Pfrommer. 1986. Activation ofthecomplement systemby Cryptococcus neoformansleads to binding of iC3b to the yeast. Infect. Immun.52:1-5.

11.Kozel,T.R.,M. A.Wilson, G. S.T. Pfrommer,and A. M.Schlageter.

1989.Activationandbinding of opsonic fragments ofC3 onencapsulated

Cryp-tococcusneoformansby usinganalternative complement pathway reconstituted from six isolated proteins.Infect.Immun. 57:1922-1927.

12.Kozel,T.R., and C.A.Hermerath. 1984. Binding ofcryptococcal polysac-charidetoCryptococcus neoformans. Infect.Immun. 43:879-886.

13. Kozel,T.R., andE.C. Gotschlich. 1982.ThecapsuleofCryptococcus

neoformanspassivelyinhibits phagocytosis of the yeast by macrophages. J.

Im-munol. 129:1675-1680.

14.Kozel, T. R. 1983. Dissociation ofahydrophobic surfacefrom phagocyto-sis ofencapsulatedand non-encapsulated Cryptococcus neoformans. Infect.

Im-mun.39:1214-1219.

15.Law, S.K. A.1988. C3 receptorsonmacrophages.J. CellSci. 9:S67-S97. 16.Myones,B. L., J. G. Daizell, N. Hogg,andG.D. Ross. 1988. Neutrophil andmonocytecellsurfacep150,95hasiC3b-receptor(CR-4) activityresembling CR-3. J. Clin. Invest.82:640-651.

17.Ross, G. D., S. L. Newman,J. D.Lambris,J. E.Devery-Pocius,J. A. Cain, and P. J.Lachmann. 1983. Generation of three different fragmentsof bound C3

withpurifiedfactorI orserum.II.Locationofbinding sitesintheC3fragments

for factorsBandH,complement receptors, and bovine conglutinin. J.Exp.Med. 158:334-352.

18.Wright,S. D., andM. T. C.Jong.1986.Adhesion-promoting receptorson

human macrophages recognize Escherichia coli by binding to lipopolysaccharide. J.Exp.Med. 164:1876-1888.

19.Speert,D.P., S. D. Wright, S.C.Silverstein, andB. Mah.1988. Functional characterizationof macrophage receptors for in vitro phagocytosisof unopson-izedPseudomonasaeruginosa.J. Clin. Invest. 82:872-879.

20.Berger,M.,R.U.Sorensen,M. F.Tosi,D. G.Dearborn,and G.Doring. 1989.Complementreceptor expressiononneutrophilsat aninflammatory site, thePseudomonas-infectedlung in cystic fibrosis.J. Clin.Invest. 84:1302-1313.

21.Ross,G. D., J.A.Cain, B.L.Myones,S. L.Newman,and P. J.Lachmann.

1987.Specificityof membranecomplementreceptor type three(CR-3)for beta-glucans.Complement.4:61-74.

22. Bullock, W. E., and S. D. Wright. 1987. Role ofthe adherence-promoting receptors, CR3, LFA-1, andp150,95,inbindingofHistoplasma capsulatum by

humanmacrophages.J.Exp. Med. 165:195-2 10.

23. Newman, S. L., C.Bucher,J.Rhodes,and W. E. Bullock. 1990.

Phagocyto-sis of Histoplasma capsulatumyeasts and microconidia by human cultured mac-rophages and alveolarmacrophages.J. Clin. Invest. 85:223-230.

24. Newman, S. L., J. E.Devery-Pocius,G. D.Ross,and P. M. Henson. 1984.

Phagocytosisof humanmonocyte-derived macrophages.Independent function

ofreceptors for C3b(CR,)and iC3b(CR3). Complement. 1:213-227. 25. Levitz,S. M.,D.J. DiBenedetto, and R. D. Diamond. 1987. A rapid fluorescent assay todistinguishattached fromphagocytizedyeastparticles.J.

Immunol. Methods.101:37-42.

26.Wright, S. D., P. E. Rao,W.C. VanVoorhis, L. S. Craigmyle, K.fida,

M.A.Talle,E. F.Westberg, G.Goldstein,and S. C. Silverstein. 1983. Identifica-tionofthe C3bi receptor of human monocytes and macrophages by using mono-clonalantibodies.Proc.Natl.Acad.Sci. USA. 80:5699-5703.

27.Melnick, D. A., W. M. Nauseef, S. D. Markowitz, J. P. Gardner, and H. L. Malech. 1985. Biochemical analysis and subcellularlocalizationofa neutrophil-specific antigen, PMN-7, involved in the respiratory burst. J. Immunol. 134:3346-3355.

28. Graziano, R. F., and M. W. Fanger. 1987.FcrRlandFcrRII on mono-cytes and granulomono-cytes arecytotoxictrigger molecules fortumorcells. J.

Im-munol. 139:3536-3541.

29. Griffin,F. M.,Jr. 1981. Roles of macrophage Fc and C3b receptors in

phagocytosis ofimmunologically-coated Cryptococcus neoformans.Proc.Natl.

Acad.Sci.USA.78:3853-3857.

30.Micklem,K.J., andR. B.Sim. 1985. Isolation of

complement-fragment-iC3b-bindingproteinsbyaffinity chromatography. Biochem. J. 231:233-236. 31. Malhotra, V., N. Hogg, andR. B. Sim. 1986.Ligand binding by the

p150,95 antigen of U937monocyticcells:propertiesincommonwith comple-mentreceptor type 3 (CR3). Eur.J. Immunol. 16:1117-1123.

32. Tosi, M. F., andM.Berger. 1988. Functional differences between the 40 kDaand 50 to 70 kDa IgG Fc receptors on human neutrophils revealed by elastase treatment andantireceptorantibodies. J. Immunol. 141:2097-2103.

33.O'Shea, J.J.,E. J.Brown,B. E.Seligmann,J.A.Metcalf,M. M.Frank,

andJ.I.Gallin. 1985. Evidence for distinct intracellular pools of receptors for C3b and C3bi in humanneutrophils.J.Immunol. 134:2580-2587.

34. Lacal, P., F. Pulido, F. Sanchez-Madrid, and F. Mollinedo. 1988. Intracel-lularlocation of T200 and Mol glycoproteins in human neutrophils. J.Biol. Chem. 263:9946-9951.

35.Silverstein,S.C., S. Greenberg, F.DiVirgilio,andT.H.Steinberg. 1989.

Phagocytosis.In Fundamental Immunology. 2nd ed. W. E. Paul, ed.RavenPress Ltd., New York. 703-720.

36. Jack,R.M., R. M. Ezzel, J. Hartwig, and D. T. Fearon. 1986. Differential interaction of the C3b/C4b receptor and MHC class I with the cytoskeleton of humanneutrophils. J. Immunol. 137:3996-4003.

37.Suchard, S.J.,and L.A.Boxer. 1989. Characterizationandcytoskeletal association ofamajorcellsurface glycoprotein,GP 140, in human neutrophils. J. Clin.Invest.84:484-492.

38. Payne, N. R., and M.A.Horwitz. 1987.Phagocytosis ofLegionella

pneu-mophilais mediated by human monocyte complement receptors. J. Exp. Med.

166:1377-1389.

39. Graham,I. L.,H. D.Gresham, and E. J. Brown. 1989. Animmobile

subsetofplasma membrane CDI Ib/CD I 8 (Mac- I) is involved in phagocytosis of targets recognized by multiple receptors. J. Immunol. 142:2352-2358.

40. Jesaitis,A.J., J. R. Naemura, R. G. Painter, L. A. Sklar, and C. G. Cochrane.1983. The fate of anN-formylatedchemotacticpeptide in stimulated humangranulocytes. J.Biol.Chem. 258:1968-1977.

41.Jesaitis, A. J., J.0.Tolley, G. M. Bokoch, and R. A. Allen. 1989. Regula-tion ofchemoattractant receptor interacRegula-tion with transducing proteins by organi-zational control in the plasma membrane of human neutrophils. J. Cell Biol. 109:2783-2790.

42.Hermanowski-Vosatka, A.,P.A.Detmers,0.Gotze, S. C. Silverstein, and S. D.Wright. 1988.Clustering ofligand on the surface of a particle enhances adhesiontoreceptor-bearingcells. J. Biol. Chem. 263:17822-17827.

43.Bianco,C.,F.M.Griffin,Jr., and S. C. Silverstein.1975.Studies of the macrophage complement receptor. Alteration of receptor function upon macro-phage activation.J.Exp.Med. 141:1278-1290.

44.Ezekowitz, R.A.B., R. B. Sim, G. G. MacPherson, and S. Gordon. 1985. Interaction of human monocytes, macrophages, and polymorphonuclear leuko-cytes with zymosan in vitro. Role of type 3 complement receptor and

macro-phage-derivedcomplement. J. Clin. Invest. 76:2368-2376.

45.Blackwell,J.M., R. A. B. Ezekowitz, M. B. Roberts, J. Y. Channon, R. B.

Sim,andS. Gordon. 1985. Macrophage complement and lectin-like receptors bindLeishmania in the absence of serum.J.Exp. Med. 162:324-331.