Subcellular localization and dynamics of Mac 1 (alpha m beta 2) in human neutrophils

Subcellular localization and dynamics of Mac-1

(alpha m beta 2) in human neutrophils.

H Sengeløv, … , T A Springer, N Borregaard

J Clin Invest.

1993;

92(3)

:1467-1476.

https://doi.org/10.1172/JCI116724

.

The subcellular localization of Mac-1 was determined in resting and stimulated human

neutrophils after disruption by nitrogen cavitation and fractionation on two-layer Percoll

density gradients. Light membranes were further separated by high voltage free flow

electrophoresis. Mac-1 was determined by an ELISA with monoclonal antibodies that were

specific for the alpha-chain (CD11b). In unstimulated neutrophils, 75% of Mac-1 colocalized

with specific granules including gelatinase granules, 20% with secretory vesicles and the

rest with plasma membranes. Stimulation with nanomolar concentrations of FMLP resulted

in the translocation of Mac-1 from secretory vesicles to the plasma membrane, and only

minimal translocation from specific granules and gelatinase granules. Stimulation with PMA

or Ionomycin resulted in full translocation of Mac-1 from secretory vesicles and gelatinase

granules to the plasma membrane, and partial translocation of Mac-1 from specific granules.

These findings were corroborated by flow cytometry, which demonstrated a 6-10-fold

increase in the surface membrane content of Mac-1 in response to stimulation with FMLP,

granulocyte-macrophage colony stimulating factor, IL-8, leukotriene B4, platelet-activating

factor, TNF-alpha, and zymosan-activated serum, and a 25-fold increase in response to

Ionomycin. Thus, secretory vesicles constitute the most important reservoir of Mac-1 that is

incorporated into the plasma membrane during stimulation with inflammatory mediators.

Research Article

Find the latest version:

Subcellular Localization and

Dynamics

of Mac-1

(amfi2)

in Human Neutrophils

Hennk

Sengelov,

* LarsKjeldsen,* MichaelS.

Diamond,t Timothy

A.

Springer,t

and Niels

Borregaard*

*Granulocyte Research

Laboratory,

Department

ofHematology

L,

University

Hospital,

Rigshospitalet,

DK-2100

Copenhagen,

Denmark;

and tThe CenterforBloodResearch, Harvard MedicalSchool, Boston,Massachusetts 02115

Abstract

The

subcellular localization of Mac-l

was

determined in

rest-ing and stimulated human

neutrophils

after

disruption by

nitro-gen

cavitation and fractionation

on

two-layer Percoll density

gradients.

Light membranes

were

further separated by

high

voltage

free

flow electrophoresis. Mac-l

was

determined by

an

ELISA with monoclonal antibodies that

were

specific

for the

a-chain

(CDilb).

In

unstimulated

neutrophils,

75%

of Mac-l

colocalized with

specific

granules

including

gelatinase

gran-ules, 20% with secretory vesicles and the

rest

with plasma

membranes.

Stimulation with nanomolar concentrations

of

FMLP resulted in the

translocation of Mac-l from secretory

vesicles

tothe

plasma

membrane, and only

minimal

transloca-tion from

specific

granules and

gelatinase granules. Stimulation

with PMA or

Ionomycin resulted

in

full translocation of

Mac-l

from

secretory

vesicles and

gelatinase granules

tothe

plasma

membrane, and

partial translocation

of

Mac-l from

specific

granules.

These

findings

were

corroborated

by

flow

cytometry,

which demonstrated

a

6-10-fold increase in the surface

mem-brane content

of Mac-l

in response to

stimulation

with FMLP,

granulocyte-macrophage colony

stimulating

factor,

IL-8,

leu-kotriene

B4, platelet-activating

factor,

TNF-a,

and

zymosan-activated

serum,

and

a

25-fold increase in

response to

Ionomy-cin.

Thus,

secretory vesicles constitute the

most

important

res-ervoir of

Mac-i

that

is

incorporated

into the plasma membrane

during stimulation with

inflammatory

mediators.

(J. Clin.

In-vest.

1993.

92:1467-1476.)

Key words:

secretory

vesicles

-

gela-tinase granules

*

specific

granules

*

CD lb

*

free-flow

electro-phoresis

Introduction

Human

neutrophils constitute the

primary

mobile cellular

de-fense

against

intruding microorganisms. Essential

components

of their

defense

mechanisms

are

stored in

membranes of

intra-cellular

mobilizable

granules and

recruited

to

the

surface

dur-ing

degranulation (

1

).

Mac-

1,

the

most

abundant

of the

2

integrins in neutrophils, is essential for neutrophil adherence

to

endothelium, chemotaxis,

and

diapedesis,

as

well

as

phagocyto-sis of complement opsonized objects (2-5). Previous

reports

have

demonstrated that

the

majority of

Mac- 1

in resting

cells

is

Address correspondence and reprint requests to Henrik

Sengel0v,

M.D.,Granulocyte Research Laboratory,Department of Hematology

L-4041, Rigshospitalet,9 Blegdamsvej,DK-2100 Copenhagen,

Den-mark.

Receivedforpublication 22 December 1992 and in revisedform 30

April1993.

localized in membranes of mobilizable intracellular

granules

or

vesicles (6, 7).

FACSO analysis (Becton Dickinson

Immunocy-tometry

Systems,

Mountain

View, CA)

has

demonstrated

that the Mac-

1

content in

plasma membranes increases

consider-ably when

neutrophils

are

manipulated during isolation

from

blood,

and

by stimulation of neutrophils with inflammatory

mediators

such as

FMLP,

leukotriene B4 (LTB4)', TNF-a, and

C5a (8). Paradoxically, subcellular fractionation data have

in-dicated that Mac- 1 is localized in specific granules (6),

al-though

these

granules

are notmobilized to

any significant

de-gree

by the

manipulations that apparently result in full

upregu-lation of Mac- 1 in the plasma membrane (9). Thus, it has been

debated whether

more

easily mobilizable intracellular

organ-elles exist that

may

furnish the surface membrane with

adhe-sion

proteins.

We have

discovered

such a

compartment,

the

secretory

vesicle, which is identified by latent alkaline

phos-phatase and plasma

proteins (10, 11). Secretory vesicles have

recently been separated from plasma membranes ( 12),

and

their role in

neutrophil activities

can now

be investigated.

Gelatinase granules have been claimed

to

exist

as a

separate

entity distinct from specific granules ( 13). On the other hand,

a

later study showed colocalization between the specific granule

marker

lactoferrin and

gelatinase

by

immunoelectron

micro-scopy

( 14).

The

reason

for these

contradictory

observations

may

be that

neutrophil

gelatinase,

previously

considered

pure,

is in

part a

complex

of

92 kD

gelatinase and

a

novel

25-kD

protein termed neutrophil gelatinase-associated lipocalin

(NGAL),

giving rise

to a

135-kD

form gelatinase in addition

to

the 92-kD form ( 15). Antibodies raised against "pure"

gelatin-ase

recognized both

gelatinase

and

NGAL, but

affinity

purifica-tion

resulted in monospecific antibodies against gelatinase

and

NGAL,

respectively.

Using affinity-purified antigelatinase

anti-bodies,

we

have demonstrated that gelatinase is mainly

local-ized

in granules with

a

slightly lower density

than

lactoferrin-containing specific

granules, and that gelatinase is mobilized

more

readily than lactoferrin (Kjeldsen,

L., D. F.

Bainton,

H.

Sengelov,

and

N.

Borregaard, manuscript submitted for

publi-cation,

and

reference

16). On the

contrary,NGAL was

found

to

colocalize with lactoferrin in specific

granules, and only a

small

fraction of NGAL is

complexed

with

gelatinase

in

gran-ules with

a

density

and

mobility intermediary

between

specific

granules and

gelatinase

granules(Kjeldsen, L., D. F. Bainton,

H.

Sengelov,

and N. Borregaard, manuscript in

preparation).

Having

introduced

secretory vesicles

andgelatinase granules as

highly mobilizable

granules, we

found

itrelevant to readdress the

question

of

the

intracellular reservoir of

Mac- 1 inhuman

neutrophils

anddeveloped an ELISA for

CDl

lb.

1.Abbreviationsused in this paper: CTAB,cetyltrimethylammonium bromide;KRP,Krebs-Ringer phosphate;LTB4, leukotrieneB4; MFI,

meanfluorescence intensity; NGAL, neutrophil gelatinase-associated

lipocalin; PMA, phorbolmyristate acetate; ZAS,zymosan-activated

serum. J.Clin.Invest.

C) The AmericanSocietyforClinicalInvestigation,Inc.

0021-9738/93/09/1467/10 $2.00

Methods

Human neutrophilswereisolated fromfreshly drawn bloodas

previ-ouslydescribed (1).Sedimentation of redcellswasinducedby Dextran

T-500(Pharmacia, Uppsala, Sweden),and the leukocyte-rich

superna-tantwascentrifuged(Lymphoprep; Nygaard, Oslo, Norway). Residual erythrocyteswerelysed by hypotonic shock, and the neutrophilswere

resuspended in bufferasindicated. Allprocedureswerecarriedout at

40C,exceptfor thesedimentation of red cells.

Stimulation of neutrophilswasperformedat370Catacell

concen-tration of3 X 107cells/ml in Krebs-Ringer phosphate (KRP): 130

mM NaCl, 5 mM KCI, 1.27 mM MgSO4, 0.95 mM CaCI2, 5mM

glucose, 10 mM NaH2PO4/Na2HPO4,pH 7.4.Stimulationwas

initi-ated aftera5-min preincubation period and terminated after15min by

dilutionwith 2 vol of ice-cold buffer and centrifugationat200gfor10

min. The pelleted cellswereresuspended in ice-cold buffer, and

sam-plesweretaken for FACS® analysis, determination ofgranule markers,

and subcellular fractionation. Release of granulecontentwas

deter-minedasmarkerproteinpresentinsupernatantasapercentof marker proteinpresentinsupernatantplus pellet.

Subcellular fractionation wasperformed as previously described

(1). In short, neutrophilswereresuspended in 15 ml disruption buffer

(100 mM KCl, 3 mM NaCl,1mMNa2ATP, 3.5 mM MgCl2, 0.5mM

PMSF, 10 mM Pipes, pH 7.2)at3 X 107

cells/ml

and disrupted by

nitrogen cavitation. Nuclei and unbroken cells weresedimented by

centrifugationat400gfor 15min (P.), and 10 ml ofthepostnuclear

supernatant(SI)wasloadedontopofa28-mltwo-layer Percoll density gradient ( 1.05/ 1.12 g/ml) and centrifugedasdescribed (1). This

re-sultedingeneration of fourseparateregions thatcould bevisually

iden-tified in thegradient: the bottom band (a band) containing azurophil granules, the intermediate band (,B band) containing specific granules and gelatinase granules, thetop band (y band) containing plasma membranes andsecretoryvesicles, and the clearsupernatant(S2)

con-taining cytosol.Gradientswereaspirated from the bottomthrougha

peristalticpumpattachedtoafractioncollectorset todeliver 1.4 ml in each fraction.

Separation of plasma membranesfromsecretoryvesicleswas

per-formedasdescribed(12). In short, Percollwasremovedfrom the z bandby ultracentrifugation. The'ybandwasresuspended in chamber

buffer(270mMsucrose,5 mMtriethanolamine, 5 mM acetic acid, pH

7.4, conductivity: 0.42mg'), and treated with neuraminidaseto mod-ify the charge of plasma membrane vesicles. The materialwasthen appliedtofree-flowelectrophoresisat5°Con anElphor Vap22from

Bender &Hobein(Munich, Germany). Itwasascertained that

neur-aminidase treatmentdid not interfere with detection of Mac-l by ELISA.

Markerproteins. Azurophil granuleswereidentifiedby

myeloper-oxidase measuredby ELISA using rabbit antimyeloperoxidase anti-body (A 398; Dakopatts; Glostrup, Denmark)ascatching antibody diluted1:10,000. Thesameantibodywasbiotinylatedat1mg/ml ( 17)

andwasusedasdetectingantibody diluted 1:10,000. Myeloperoxidase,

purifiedfromisolatedazurophil granules,wasusedasstandard.

Spe-cificgranuleswereidentifiedbylactoferrin measuredby ELISA ( 16)

and by vitamin B12-binding protein (1). Gelatinase granules were

identifiedbygelatinase and measured by ELISA ( 18). Secretory vesi-cleswereidentifiedby alkaline phosphatase (19) that could be

mea-sured only inthe presenceof 0.2% (vol/vol) Triton X-100 (latent

alkalinephosphatase)(10, 20) and by albumin measured by ELISA (11). Plasma membraneswereidentified byHLAclassIassayed bya

mixed ELISA(21 ) and by alkaline phosphatase activity measured in theabsence of detergent (nonlatent alkaline phosphatase) (10).

Quantitation of Mac-Iwasdoneby ELISA: CBRM1/23,anIgG2a mousemonoclonalantibody against the COOH-terminal extracellular

region of theachain of Mac- 1 (22)waspurified from ascites fluid by

proteinASepharoseaffinity chromatography, diluted 1:1,000 froma

2.3 mg/ml stock, and usedfor coating immunoplates. LM2/ 1, an

IgGImousemonoclonal antibody againstthe I domainof theachain of Mac-1(22, 23)waspurifiedfrom ascitesfluidbyproteinA

Sepha-roseaffinitychromatography and biotinylated at a concentration of 1 mg/ml (17). This was used as detecting antibody at a 1: 1,000-fold dilution. Serial dilutions ofybandisolatedfrom neutrophils stimu-latedwith 2

_Ag/ml

phorbol myristate acetate (PMA) at370Cfor 15 minwereused as standards. A stockofyband from 5 X 1O'cells in

disruption buffer (0.9mg/ml protein) was prepared and used forall experiments. Thisstock wasdiluted 16-foldin buffer B(seebelow) and used as thehigheststandard,definedas 64 arbitrary U/ml. Samples for CDl lb ELISA were solubilized by addition of 25

,l

250 mM N-octyl

glucoside (SigmaChemical Co., St. Louis, MO), and 25

_tl

2%

cetyltri-methylammonium bromide (CTAB) (BDH Chemicals LTC, Poole,

United

Kingdom)

to200

,l

sample followedbyincubation on ice for 1 h. The samples were then diluted with 750

Al

500 mMNaCl, 3 mM

A

CD11b,

U/ml

600

400-

200-6 11 16 21

Fraction No

|

I

I

I

1

11

af

0

l

_S2

B

E

z

-I

NORMfAL-C LAD-CD11B IORMAL-CD11B

LAD-C

FLI

Io jY3

Figure1. Mac-l in normal andLAD

neutrophils.

CDl lbin normal

(3x 108 cells)and LAD(2.25 x 108 cells)

neutrophils

was deter-minedbyELISA insubcellular fractions(A)andbyFACSO

analysis

onintact cells(B).InA,thelocalization of the

gradient

bandsare

depicted

belowthe

figure.

Theabandcontains

azurophil granules,

the d band contains

specific

and

gelatinase granules,

andthe yband

contains

secretory

vesicles and

plasma membrane, S2

is cytosol. Nor-malcontrol,-+-;LAD

patient,

-s-.Abscissa in B islog.

fluores-cence

intensity.

The cellswerestimulatedwith10-8MFMLPbefore

FACS®

analysis

to

optimize

detection of surface bound Mac-1. C

denotes

isotype-specific

nonimmune control

antibody.

Exocytosis

of

albumin, gelatinase

andlactoferrin fromLAD

neutrophils

in response

toFMLPwasnormal

(not shown).

1468 H.Sengelov, L.Kjeldsen,M.S.Diamond, T. A.Springer,andN.Borregaard

(5D I AFe....I..I§I. . - . ... I.' .... ....

B

CD1 1b,U/

Lactoferrin,pg/ml

8 11 18 21

FractionNo.

(Ml Vitamin B12, ng bound/ml

11 1

Fraction No. C CD 1lb,U/mi

E _{CD11b, U/ml} 300

Latent Alk.Phos., mU/ml

1 6 11 18 21

F

Fraction

No.

G

CD11b, U/ml Albumin, ng/ml

8 11 18 21

FractionNo.

CD11b,U/mi HLA, U/ml

6 11 16 21

Fraction No.

D

_{CD11b, U/ml}

150-

100-

50-Myeioperoxidase,_pjg/ml

6 11 1

Fraction No.

ie 21

Figure2.Subcellularprofile ofMac-i. Subcellular distributionof CDI lb and markers forspecific granules (AandB), gelatinase granules (C), azurophil granules (D),secretoryvesicles(EandF),andplasmamembranes(G). Assaysarefromasingle, representativefractionation(out

offive)ofunstimulatedneutrophilsheldat4°C. CD1lbis shownforcomparisonineachprofile (grayline)toothermarkers(blackline).

A

_CD1I

b,

U/mI

-15

-10

l6 21

11 1

A

CD11bU/ml

1400

r

1200

1000

600

600 400 200

1

B

Non-latentALP, mU/ml

1!

1001

501

6

11

16

21

Fraction

No.

-Control + FMLP * PMA

C

Gelatinase,

pg/ml

30

20

1 6 11 16

Fraction

No.

-Control +FMLP *

_PMA

D Laten

_ALP

_mU/ml

120 r

100

80

60

40 20

1 6 11 16 21 1 6 11 16 21

Fraction No.

Fraction

No.

-Control

+

FMLP

*

PMA

_

--Control

+

FMLP

*

_PMA

E

I-Lactoforrin, pg/ml

1200

r

S0

HLA,

U/ml

D01

500

400

300

200

100

1 6 11 16 21 1 6 11 16 21

Fraction No.

Fraction No.

-Control

+ FMLP * PMA

-Control + FMLP *

PMA

1470 H.Sengel0v, L.Kjeldsen,M.S.Diamond, T.A._Springer,and N.Borregaard

21

1U

B

CD11b;HLA,U/ml LatentALP,mU/ml CD11b; HLA, U/ml LatentALP, mU/ml,

12 13 14 15 16 17 18 1 20 21 22 23 24 25

Fraction No.

-CD11b + Latent ALP HLA

150

100

150

12 13 14 15 16 17 18 19 20 21 22 23 24 25 Fraction No.

O _CDIlb + Latent_ALP

-HLA

Figure4. Mac-I inlightmembranes. Subcellular localizationof CDl lband markers forsecretoryvesicles(latentalkalinephosphatase)and

plasmamembranes(HLA)wasdeterminedin control cells(A)and FMLP-stimulated cells(B).Samedata as inFig. 3.

KCl, 1%(vol/vol)TritonX-100,1%(wt/vol)BSA(Sigma),25mM

N-octyl glucoside, 0.2% CTAB, 8 mM Na2HPO4/KH2PO4, pH7.2, incubated overnight in the cold and then appliedto96-well immuno-plates (Nunc, Roskilde, Denmark) that had been coated with catching antibodyandblockedasdescribed below.

ELISAwasperformed usingthesamegeneral procedure

indepen-dent of the antibodies exceptasspecifiedinprevious publications

re-gardingmixed ELISA(21)and lactoferrin (16)whereavidin-biotin interactionwasnotused.Immunoplateswerecoated with 100,l

catch-ing antibody, diluted in 50 mM Na2HCO3/NaH2CO3, pH 9.6,and incubatedovernightatroomtemperature.The wellswerewashed in buffer A(500mMNaCi, 3 mMKCI,8mMNa2HPO4/KH2PO4, pH 7.2,1%Triton X-100). Additional binding siteswereblockedby

incu-bating with 200 buffer B (500 mM NaC1, 3 mM KCI, 8 mM

Na2HPO4/KH2PO4, pH 7.2,1% TritonX-100, 1%BSA)for1h. After

washing three timesinbufferA,100,lsamplewasapplied alongwith serial dilutions of standard and incubated for 1 h. Allsamplesand

standardswerediluted in buffer B(exceptfor quantitation ofCD Ibas

describedabove). After three additional washesinbufferA, 100 Ml

biotinylated catching antibodywasapplied, appropriately dilutedin

bufferB,and incubated for 1 h followedby washingthree times in

buffer A. 100 Mlperoxidase-conjugated avidin(P347; Dakopatts)in

buffer Bwasthen addedandincubatedforI hfollowed by washing

three times in buffer A andoncein100 mMNa2HPO4/NaH2PO4, 100

mM sodiumcitrate, pH 5.0. Color developed duringa30-min

incuba-tioninsodiumphosphatecitric acid buffercontaining 0.04% (wt/vol) o-phenylenediamineand0.03% H202,andwasstopped by addition of

100 Mi 1 MH2SO4. Absorbancewasreadat492nminanautomatic

ELISAreader(Multiscan Plus; Labsystems, Helsinki, Finland). FACS" analysis. Isolated neutrophils, either keptonice,at370Cor

stimulatedat370Casindicated,werefixedat3 x I07cells/mlin 2%

paraformaldehyde, 0.05% glutaraldehydeinPBS for 15 minat40C,

washed twice in PBS containing 0.5% (wt/vol) BSA. Labeling of cells

wasperformed by incubating 25 Mlcellsat3X 10'/mlinPBS

contain-ing 0.5% BSA with 50,lmonoclonal antibody (LM2/ 1oranti-HLA class1(Dakopatts M736))ormurine preimmuneIgGI

(Becton-Dick-inson Immunocytometry Systems) for1hat4VC.After washing twice,

50 Mlfluorescein-conjugatedrabbit anti-mouseantibody(Dakopatts)

wasadded,and thecellswereincubated for 30 minat4VC. Aftertwo

additionalwashes,the cellswereresuspendedin PBScontaining 1% formaldehyde and analyzed in aFACScan"(Becton-Dickinson

Im-munocytometrySystems). Mean fluorescence intensityforspecific

an-tibodies was corrected for nonspecific fluorescence by subtracting

meanfluorescenceforpreimmune IgG1.

Source ofstimuli. FMLP (Sigma)1 mMinethanol.Granulocyte

macrophage-colony stimulating factor (Sandoz Pharmaceuticals, East Hanover, NJ) 104U/ml in H20. rIL-8 (agenerousgift fromDr.K. Thestrup-Pedersen, Department of Dermatology, Marselisborg Hospi-tal, Aarhus, Denmark) 10 Mg/ml in PBS. Ionomycin (Calbiochem Corp.,LaJolla, CA)0.1 mM inDMSO. LTB4 (Sigma) 1mM in

eth-anol.Platelet-activatingfactor(Sigma) 1 mM inPBS 0.5% (wt/vol)

HSA (Sigma). TNF(Amersham International, Amersham, United Kingdom) 104 U/mlinH20.

Results

A sandwich ELISAwasdeveloped for Mac-I usingtwo

mono-clonal antibodies that localizedtoepitopesondistinct regions of Mac-l a chain(CDl Ib): CBRM1 /23, which binds_tothe

COOH-terminalsegment, andLM2/ 1, which bindstothe I domain (22, 23). Using thisassay, weinvestigated the

subcel-lular localization of Mac-1 in human neutrophils.Tovalidate thespecificityof the assay, CDl lbwasmeasuredinfractions of

Figure3.Dynamics ofMac-i.Effect of stimulationwith FMLP(10-8M)orPMA(2,g/ml)on:the subcellular localization ofCDl lb (A),

nonlatentalkaline phosphatase (membrane incorporated intotheplasma membrane fromsecretoryvesicles) (B), gelatinasegranules (C), latent

alkalinephosphatase(secretoryvesicles)(D),lactoferrin (specificgranules) (E), and HLA (plasma membrane)(F).Neutrophilsfromtwo

normaldonorswerepooled, resuspended inKRPcontaining1mMsodium azide_and2,000U/ml_{catalase(bovine liver; Sigma), and divided}

inthree.Onepartwaskeptonice (control)onepartwasstimulated with10-8 M FMLPat370C, andonepartwasstimulated with 2 Mg/ml

PMAat370C before subcellular fractionation.The resultsofasingleexperiment, representativeof three, is given.

A

B

35 37 39 41 43 45 47 49 51 53 55 Fraction No.

+Latent ALP -CD11 b *HLA

3001

-1

I 250k-

.-C.,

35 37 39 41

I , W.M

43 45 47 49 51 53 *55 FractionNo.

+LatentALP-CD11b *HLA

Figure 5. Free-flow electrophoresis. Membranes from the yband of4VCcontrolneutrophils(A) andneutrophilsstimulated with FMLP ( 10-8

M)at370C (B)weresubjectedtofree flowelectrophoresis. Material from 3x 108 cellswasused for each condition.

neutrophils fromapatient with the lesssevereform of

leuko-cyteadhesion deficiency

(24-26).

Fig. 1 A shows that in this

patient, only 5% of the normalamountof CDl lbwasfound,

whichagreeswell with

immunofluorescence

studies (Fig. 1B).

Thus, the ELISA is specific for CDl lb.

The major peak of Mac-l is

located

in the : band, which contains both gelatinase and lactoferrin (Fig. 2). We confirm herethe separation of peak

lactoferrin

and peak gelatinase that hasbeen

established

before

(16,

27). As seen in Fig. 2, we

found that Mac-l had an intermediate

distribution

among

granules that contained

lactoferrin

and gelatinase, but mainly colocalized with lactoferrin in specific granules of the ,3 band. Inaddition, Mac-1wasfound inasecond broad peak in the-y

bandcolocalizing with latent alkaline phosphatase and albu-min, both markers ofthemostrapidly and extensively mobiliz-ablecompartment, secretoryvesicles( 11), but withadifferent peak profile than the plasma membrane marker, HLA classI.

Theresolutionbetween

organelles

by subcellular fraction-ation isimproved by examining the

dynamics

oftheorganelles inresponsetostimulation. Therefore,westudied the

transloca-tion ofMac- from intracellular membranes to the plasma membrane under conditions that discriminate between mobili-zation ofsecretoryvesicles,gelatinase granules, and lactoferrin

granules (Fig.

3). Mac- is

translocated

from intracellular membranesto theplasma membrane in response to FMLP, whichmobilizessecretoryvesicles almost completely, but only results in 15-25% release ofgelatinase and <3% release of

lactoferrin (16), and in response to PMA, which results in

complete mobilization of secretory vesicles and gelatinase granules, and almost complete mobilizationofspecific

gran-ules(

1, 16, 20, 28).AfterstimulationwithFMLP,Mac-l

redis-tributes within theregionofthegradientthat contains thelight membranes(secretory vesicles and plasma membranes).

Thisismoreeasily observedin Fig. 4,which displaysthe

fractions containing light membranes (fractions 12-25). In

unstimulated cells, the Mac-1 profile colocalizes with latent

alkaline

phosphatase,

themarker for secretoryvesicles, inthe

light membrane region (Fig. 4A),but notentirely with the

profile of plasma membrane marker,HLA.FMLPstimulation

promotestranslocation of alkaline

phosphatase

fromsecretory vesicles tothe plasma membrane with

corresponding

lossof latency (Fig. 3, B and D). Mac-l also translocates ina

corre-sponding manner and now colocalizes

completely

with the

plasma membrane marker (Fig. 3 A and Fig. 4 B). The contri-bution of membrane from secretory vesicles to the plasma membrane is illustrated in Fig. 3 B, which gives the increase in nonlatent alkaline phosphatase present in the plasma

mem-braneas aresult ofstimulation by FMLP and PMA. The alka-line

phosphatase

level in the plasma

membrane

increases four-fold afterFMLP stimulation, and no further increase is

ob-served after PMA stimulation, since secretory vesicles are

almostfully mobilized by FMLP. Tailing of gelatinase (Fig. 3 C)wasstillpresentinfractions13-15,eventhough Mac-1 had

disappeared after stimulation by FMLP, excluding these

gran-ulesas amajorstoreoflight membrane Mac-1.Thus,

mobiliza-tion ofsecretoryvesiclescanaccountformostofthe increase in plasma membrane CDl lbcontentthat isinduced by stimula-tion with FMLP. The separastimula-tion ofsecretory vesicles and plasma membranes can be greatly improved by combining

density gradient centrifugationwithhigh voltagefree-flow elec-trophoresis. This modality takes advantage of differences in

charge thatcan be selectively induced in plasma membrane vesicles bytreatmentwith neuraminidase ( 12). Using this

pro-cedure,weobtainedanalmost total

separation

ofplasma

mem-branesand secretory vesicles(Fig. 5A).It isclearlyseenthat verylittle Mac- is associated with the plasma membrane in

unstimulated cells, andthatthemajorpartofMac-

coloca-lizeswith latent alkaline phosphatasein secretoryvesicles. In contrast, complete colocalization of the plasma membrane

markerHLA with Mac-l isseeninthey band from

FMLP-sti-mulated cells(Fig. 5B).

To discriminate between colocalization ofMac- I with gela-tinase granules and lactoferrin granules, thetranslocation of

Mac- to the plasma membrane in response tostimulation

with PMA was determined (Fig. 3). The disappearance of

Mac-l from the ,B band (85%, median of threeexperiments,

range82-89%)wasintermediate betweenthedisappearanceof

lactoferrin(63%,median of threeexperiments,range47-70%) 1472 H._Sengelov, L._Kjeldsen,M. S.Diamond, T. A.Springer,and N.Borregaard

c1

~~~~~~~~~~~~~1

_

5

I

25 L <-*. _E 20D0E

15

10 200 -4'0

is

la

and the

disappearance of gelatinase (93%, median of three

ex-periments,

range

89-95%). Not all Mac-i that disappeared

from the

_#

band

was

recovered in

the

plasma membranes of

the

'y

band.

This

may be caused

by

a

change in the

epitopes of

the

CD

1 b

molecule

by

exposure to

reactive

oxygen

species

and

proteases. However,

inclusion

of azide

and

catalase increased

the

recovery

of CD

1

lb in the plasma membrane considerably

but

was

unable

to

completely

protect

the

molecule. Ionomycin,

another

potent

secretagogue, also induced translocation of

Mac-I from the

#

band

to

the

y

band

(data

not

shown).

Al-though

ionomycin

does

not

activate the

plasma

membrane

bound NADPH oxidase

(29,

30),

only 70% of Mac-I that

dis-appeared from the ( band

was

recovered in the

y

band. Other

yet

unclarified mechanisms

must

be

responsible

for the

de-crease

in

Mac-1

immunoreactivity

after translocation

to

the

neutrophil surface induced by

potent

stimuli. These

observa-tions

on

the

subcellular localization and mobilization of Mac-

I were

corroborated by

flow

cytometry as

shown in

Fig. 6,

which

gives the

upregulation

of Mac-I in the

surface

membrane of

intact cells in

response to

stimulation

with

a

variety

of

inflam-matory

mediators. FMLP,

GM-CSF, LTB4,

IL-8, PAF,

TNF-a,

and

zymosan-activated

serum

(C5a)

result in

a

6-10-fold

increase in the

surface

membrane

content

of Mac-i and

a

corresponding

almost

complete

mobilization of

secretory

vesi-cles

(release of

albumin),

but

little

release

of

gelatinase

and

no

release

of lactoferrin.

Ionomycin

results in

a

25-fold

increase in

the

plasma membrane

content

of Mac-

I

consistent with

incor-poration

of the

specific

granule membrane pool of Mac-

1

into

the

plasma membrane.

In contrast, PMA

stimulation did

not

significantly

increase the

amount

of Mac-I detectable

on

the

surface

of the cells beyond what is

seen

after

FMLP

stimula-tion,

despite

significant

mobilization of

specific granules.

This

is probably

aconsequence

ofboth inactivation

of

Mac-I

as

also

observed in

subcellular

fractionation,

and

folding

and

invagi-nation

of

the

plasma

membrane, which makes the

epitopes

inaccessible

to

the

antibody.

The latter

is

supported by

the

sig-nificant reduction in detectable

HLA

observed

by

FACS®,

but not

by

ELISA after stimulation with

PMA.

Discussion

Defense against microorganisms

depends on the ability of

neu-trophils

to

stick

to

endothelium, migrate

through it, phagocy-tose,

and kill

microorganisms

by

mechanisms

that are both

toxic and largely unspecific. On

theother hand, integrity of the

vascular

bed depends on the ability of neutrophils to slide

through

the

microcirculation

in

intimate

contact with

endothe-lium,

without

causing

anydamage to

it.

Storing proteins, active

in mediating

both

adherence, phagocytosis, and generation

of

toxic

oxygen

species, in

themembrane of mobilizable

intracel-lular organelles,

seems

teleologically wise.

Theexistence offour separate

mobilizable intracellular

organelles ( 10,

11,

16, 27), each

with

adistinct

membrane

composition, and with a strict

hierarchy of mobilization

(31), ties neutrophil activation

tightly

tocontrol

of exocytosis.

Thus,it is important to eluci-date

both

the

mechanisms

that mediate the selective

mobiliza-tion of

these granules and vesicles, and the

contribution

that

each granule

subset makes to the functional capacity of the

plasma

membrane. By combining subcellular

fractionation

of

unstimulated

cells

with

high voltage free-flow

electrophoresis

and

with

fractionation

of

cells that have been stimulated to

selectively

mobilize individual

granule subsets,

a

high degree

of resolution between

granules

with differences in

density

or

mo-bilization canbe achieved.

Using

this

approach,

we have identified the intracellular

location

ofMac- I as the membranes of

specific granules,

gela-tinase

granules,

and

secretory

vesicles.

Stimulation

of

neutro-phils

with

inflammatory

mediators results in

incorporation

of Mac- I from

secretory

vesiclesinto the

plasma membrane,

and this accounts for most of the increase in

plasma

membrane Mac-1.

Clearly,

not all

intracellular

Mac-I is mobilized to the

plasma

membrane

by

FMLP stimulation of the cells. This is in contrast to data

previously published by

us

based

on

FACS®

analysis

of

fixed permeabilized

versus

unpermeabilized

cells

(8).

The reason for this

discrepancy

most

likely

is the failure of the

permeabilization procedure used

to

make the intragranular

Mac- I

fully

available

to

antibody. This

was

also

an

initial

prob-lem in

this study, until the solubilization protocol combining

N-octyl glucoside and CTAB

was

introduced. Failure

to

com-pletely solubilize

the

membrane of specific granules

may also

have

been a

problem in

other

studies in

which the subcellular

localization

of Mac- I was

determined by immunoprecipitation

(32-36). Since secretory vesicles

aremore

readily solubilized

than specific granules (Sengel0v, H., unpublished

observa-tion),

incomplete solubilization will result in relative

overesti-mation of the Mac-I

present

in the

y

and pre-y regions

that

contain secretory vesicles (32, 33).

The

intracellular localization of Mac-

I

has been the subject

of much debate. The issue has been whether Mac-

I

is

localized in

gelatinase, so called "tertiary,"

granules or

in

specific

gran-ules.

Previous ultrastructural examination has demonstrated

80%

colocalization of gelatinase and lactoferrin ( 14).

How-ever,

the

discovery ofNGAL

as a component

ofgelatinase

casts some

doubt

on

the

specificity of the antibodies used

to

detect

gelatinase. Subcellular fractionation studies reporting

colocali-zation of Mac-

I

and

gelatinase have made

use

of

a

gelatinolytic

assay

for

gelatinase (32-37). This

assay,

in

which the

enzyme

has

to

be

activated in

the presence

of

other proteases,

is

prob-lematic since it is

not

ascertained that all gelatinase is activated

and

none

is

inactivated. Using

an

ELISA

for quantitation of

gelatinase, these problems

are

avoided ( 16,

18).

Furthermore,

the

above mentioned studies (32-37) have

not

taken into

ac-count

the

existence of secretory vesicles,

and have

therefore

overlooked the

most

readily available

pool

of intracellular

Mac-l.

It

is

apparent

from

our

studies

(Kjeldsen,

L.,D. F. Bainton, H.

Sengelov, and

N.

Borregaard,

manuscript submitted

for

publication,

and

reference

16.) that gelatinase

does not

coloca-lize with lactoferrin. Our subcellular fractionation

data indi-cate

that Mac-I is

present

in the

membranes

of

both

gelatinase

and

lactoferrin granules.

We conclude that secretory vesicles

are the most important reservoir

of

Mac- I that is mobilized to the plasma membrane

during

stimulation of neutrophils

with

inflammatory

media-tors. In

the

interplay between endothelium

and neutrophils,

which

is

most pronounced in venules (4), it is critical that

mobilization of

Mac- I takes place rapidly, and before the

neu-trophil leaves this

vascular bed, otherwise, the

neutrophil

will not

adhere

firmly

to

inflamed endothelium,

butwill arrive acti-vated to the capillary bed of the lungs. Secretory vesicles, by far the most

rapidly

and

extensively

mobilized organelle of

Relae(%) 4C MFI

Figure 6.Upregulationof Mac-Ion intactneutrophils.Normalneutrophilswereisolated,resuspended in KRP, and divided. Somewerekept on ice

(40C)

othersincubated at370Cwith orwithout stimuli(10-8MFMLP, 100 U/mlgranulocytemacrophage-colony stimulating factor, 0.2

,gg/ml

IL-8, I

gM

lonomycin, 10-8 MLTB4, 10-8 MPAF,2Mg/ml PMA,500U/ml TNF,and 5%

zymosan-activated

serum

[ZAS])

as indi-cated.Upregulationofsurface boundCDlIbdetected by

LM2/

1wasdeterminedby FACSO

analysis,

andmeanfluorescence intensity(MFI) is shown incombination with simultaneouslydeterminedreleaseofgranule markers. HLA classI wasquantitatedby FACSs on the same cells andgivenasMR.Cellskept on icedisplayedasmall and variable subset of cells with apparentupregulation,observed only uponfixation.

These weregated awaybeforedetermination of MFI. Double hatched bars, lactoferrin release; lined bars, gelatinase release; single-hatched bars, albuminrelease; solid bars, MFI for CDlIb; open bars, MFI for HLA. Albumin release could not be determined after ZAS stimulation because

ofthe presence of albumin in ZAS. Note the different axis scale in thelonomycinandPMAexperiments.Dataaregivenas mean(bar)+SD

(error bar)of threetosix

experiments.

The

exocytosis of gelatinase-rich

granules

doesnot

contribute

significantly

to the enrichment of the plasma membrane in

Mac-

I that takes place

during

stimulation with

inflammatory

mediators

suchasC5a,

FMLP, IL-8, LTB4, PAF,

and

TNF,

but

probably plays

amajor role

during migration

and

diapedesis

by

virtue

of the

exocytosis ofgelatinase. Two

(3

integrins (laminin

receptor and

fibronectin receptor),

a

(2

integrin (Mac- 1),

and

a 13

integrin

(vitronectin

receptor)

have been found

only

in specific granules by immunoelectron

microscopy

(38).

How-ever,

Mac-

1 isalso present in

secretory

vesicles as shown in

this

study, and careful

analysis

of the surface

upregulation

of

lam-inin

receptors is

suggestive

oflocalization of laminin

receptors

in a more mobilizable

compartment

than

specific

gran-ules

(39).

In

platelets,

the

f3

integrin GPIIb/IIIa

has been

shown

to

be located

intracellularly

both in the a

granules

(40)

and in

vesicles which

may be

distinct from the open canalicular

sys-tem

(41). Translocation

of

GPIIb/IIIa

to

the platelet plasma

membrane from

a

granules may

facilitate

the

observed

cluster-ing

of

GPIIb/IIIa in areas of cell-cell

interaction

(42).

In the

platelet vesicles,

GPIIb/IIIa

is located together with GPIb (a

f3

integrin)

(41

).

Interestingly,

the

residence

and

translocation

of

PGIIb/IIIa

in

platelet

vesicles

closely resembles the location

of

Mac-

1in

secretory

vesicles

in neutrophils.

This similarity

prob-ably reflects that

both

platelets

and

neutrophils

are

circulating

cells with inherent

extensive adhesive

properties,

which

must be

highly

regulated.

The

localization

of

Mac-

1 in

membranes of intracellular

organelles with marked

differences

in

mobilization

may reflect

the two

different functions laid down in the

same structure: the

function

as an

integrin,

important

during chemotaxis and

dia-pedesis,

is best served

by

storing Mac-

1 in

the membrane of

a

highly

mobilizable

organelle

as

the

secretory vesicles that

re-sponds to

inflammatory mediators,

whereas the

function

as

receptor for C3bi is best served

by

storing Mac-

1

in granules

that are

mobilized during

phagocytosis,

such as

specific

gran-ules.

Acknowledgments

The expert technical assistance of Charlotte Horn and Pia L. Olsen is

gratefully appreciated.

Thiswork was_{supported by the Danish Cancer Society, the Danish}

Medical Research Council, National Institutes of Health grant CA

31799,The Novo Fund, The Lundbeck Fund, Emil C. Hertz's Fund, Amalie J0rgensens Mindelegat, Brochner Mortensen's Fund,Anders

Hasselbalch's Fund, and Ane Katrine PlesnersMindelegat. N.

Borre-gaardistherecipient ofaNeye-Research Professorship.

References

1.Borregaard,N., J. M. Heiple, E. R. Simons, and R.A.Clark. 1983. Subcel-lularlocalizationof theb-cytochromecomponentof the human neutrophil mi-crobicidal oxidase. Translocation during activation. J.CellBio.97:53-61.

2. Beller,D.I.,T.A. Springer, and R. D. Schreiber. 1982.Anti-Mac-l

selec-tively_{inhibits the mouse and human type three complement receptor. J.Exp.}

Med. 156:1000-1009.

3.Hickstein,D.D., J. Ozols,S. A.Williams, J. U. Baenziger, R. M. Locksley, andG.J. Roth. 1987. Isolation and characterization of the receptor on human neutrophils that mediates cellular adherence.J.Biol. Chem. 262:5576-5580.

4.Andrian,U. H.,J. D.Chambers,L.M. McEvoy, R. F. Bargatze, K.-E. Arfors, andE.C. Butcher. 1991. Two-step modelofleukocyte-endothelial cell

interactionininflammation:distinct roles for LECAM-Iand the leukocyte_f2 integrinsinvivo. Proc.Natl. Acad.Sci. USA. 88:7538-7542.

5.Anderson,D.C.,L. J.Miller,F.C.

_Schmalstieg,

R._{Rothlein, and}T. A.

Springer.1986.Contributionof the Mac-I_{glycoprotein family}to

adherence-de-pendent granulocytefunctions:structure-functionassessmentsemploying

sub-unit-specificmonoclonal antibodies. J. Immunol. 137:15-27.

6.Todd,R. F. III, M. A.Arnaout,R. E.Rosin,C. A.Crowley,W. A.Peters,

and B. M. Babior. 1984. Subcellular localization ofthelargesubunit MoI(MoIa,

formerlyGP 110),asurfaceglycoproteinassociated withneutrophiladhesion.J.

Clin. Invest.74:1280-1290.

7._Bainton,D.F.,L. J.Miller,T. K.Kishimoto,andT. A.Springer. 1987.

Leukocyteadhesionreceptorsarestoredinperoxidase negative granulesof

hu-manneutrophils.J.Exp.Med. 166:1641-1653.

8.Miller, L. J.,D. F. Bainton,N. Borregaard,and T. A.Springer. 1987. Stimulated mobilization of monocyteMac-1 andp150,95adhesion proteins

fromanintracellular vesicularcompartmenttothe cell surface.J.Clin. Invest.

80:535-544.

9.Berger,M.,E.Wetzler,andR. S. Wallis. 1988. Tumor necrosis factor is the

majormonocyteproductthat increasescomplementreceptorexpressionon

ma-turehumanneutrophils.Blood. 71:151-158.

10._{Borregaard,N.,}L. J.Miller,and T. A.Springer.1987.

Chemoattractant-regulatedmobilization ofanovel intracellularcompartmentin human

neutro-phils.Science(Wash. DC).237:1204-1206.

11._Borregaard,N.,L._Kjeldsen,K.Rygaard,L.Bastholm,M. H.Nielsen,H.

Sengelov,0. W.Bjerrum,and A. H. Johnsen. 1992. Stimulusdependent secre-tion ofplasma proteinsfrom humanneutrophils.J. Clin. Invest.90:86-96.

12._Sengelov, H.,M. H.Nielsen, and N.Borregaard. 1992.Separationof humanneutrophilplasmamembrane fromintracellularvesiclescontaining alka-line_phosphataseand NADPH oxidase_activitybyfree flowelectrophoresis. J. Biol. Chem.267:14912-14917.

13.Dewald, B.,U.Bretz,andM._Baggiolini.1982. Release ofgelatinasefrom

anovelsecretory compartmentof humanneutrophils. J. Clin. Invest. 70:518-525.

14.Hibbs,M.S.,andD. F. Bainton. 1989. Humanneutrophilgelatinaseisa

component ofspecificgranules.J. Clin. Invest.84:1395-1402.

15._Kjeldsen,L.,A. H.Johnsen,H._Sengelev,and N._Borregaard.1993. Isola-tion and_primarystructureofNGAL,anovelproteinassociated with human

neutrophilgelatinase.J.Biol. Chem. 268:10425-10432.

16.Kjeldsen,L.,0. W.Bjerrum,J.Askaa,and N.Borregaard.1992. Subcellu-lar localization and release of humanneutrophil gelatinase confirmingthe exis-tenceof separategelatinase containinggranules.Biochem.J.287:603-610.

17.Wilchek, M.,andE. A.Bayer.1990. Avidin-biotin_technology.In Meth-odsEnzymol.M. Wilchek and E. A.Bayer,editors. AcademicPress, Inc.,New York. 184:123-166.

18._Kjeldsen,L.,0. W.Bjerrum,D._Hovgaard,A.H.Johnsen,M._Sehested, and N.Borregaard.1992. Human_{neutrophil gelatinase:}amarker for

_circulating

bloodneutrophils.Purification andquantitationbyenzyme-linked immunosor-bentassay.Eur. J.Haematol.49:180-191.

19._DeChatelet,L.R.,andM. R.Cooper.1970.Amodifiedprocedureforthe determination ofleukocytealkalinephosphatase.Biochem. Med.4:61-68.

20._Borregaard, N.,L. _{Christensen,0. W.Bjerrum,}H. S.Birgens,and I. Clemmensen. 1990. Identification ofa_highlymobilitablesubset of human

neu-trophilintracellular vesicles that containstetranectinandlatentalkaline

phos-phatase. J.Clin. Invest.85:408-416.

21.Bjerrum,0.W.,andN.Borregaard.1990.Mixedenzyme-linked

immun-osorbent assay(MELISA)for HLA class Iantigen:a_plasmamembrane marker. Scand.J.Immunol.31:305-313.

22.Diamond,M.S.,J._{Garcia-Aguilar,}J. K.Bickford,A.L.Corbi,and T.A.

Springer.1993. The I domain isa_majorrecognitionsiteon theleukocyte

integrin

Mac-I (CDIlb/CD18) for four distinct adhesionligands. J. Cell Biol. 120: 1031-1043.

23.Miller,L.M.,R._Schwarting,and T. A.Springer.1986.Regulated expres-sion oftheMac-1,LFA- 1,p150,95 glycoproteinfamily duringleukocyte differen-tiation.J.Immunol. 137:2891-2900.

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

Rothlein,L. J.Miller,S.Kohl,M.Tosi,R. L.Jacobs,T. C.Waldrop,etal. 1985. Thesevereand moderatephenotypesof heritable Mac-1,LFA- 1deficiency:their

quantitativedefinition and relationto_{leukocyte dysfunction}and clinical fea-tures.J.Infect.Dis. 152:668-689.

25.Springer,T.A.,W. S.Thompson,L.J.Miller,F.C.Schmalstieg,and D. C. Anderson. 1984. InheriteddeficiencyoftheMac-i,LFA-1,p150,95_glycoprotein

familyandits molecular basis. J.Exp.Med. 160:1901-1918.

26._Wang,P.,andT.A._Springer.1987. Steroidtreatmentofapatient with

moderateMac-i,LFA- 1deficiency.Eur.J.Clin. Invest. 17:376a._(Abstr.) 27.Nitsch, M.,D.Gabrijelcic,andH. Tschesche. 1990.Separationofgranule subpopulationsin humanpolymorphonuclearleukocytes.Biol. Chem. Hoppe-Seyler371:611-615.

28._{Borregaard, N.,}andA.I. Tauber. 1984. Subcellularlocalizationof the humanneutrophilNADPH oxidase.b-Cytochromeandassociated

flavoprotein.

J.Biol. Chem. 259:47-52.

29.Dahlgren,C. 1987. Difference inextracellularradical release after chemo-tactic factorandcalcium ionophore activationofthe oxygenradical-generating

systemin humanneutrophils.Biochim.Biophys.Acta. 930:33-38.

Borre-gaard. 1992. Activation of the oxygen-radical-generating system in granules of intact human neutrophils by a calcium ionophore (ionomycin). Biochim. Biophys. Acta. 1 137:182-188.

31.Sengel0v, H., L. Kjeldsen, and N. Borregaard. 1993. Control ofexocytosis in early neutrophil activation. J. Immunol. 150:1535-1543.

32. Jones, D. H., D. C. Anderson, B. L. Burr, H. E. Rudloff, C. W. Smith, S. S. Krater, and F. C. Schmalsteig. 1988. Quantitation of intracellular Mac-l

(CD1 lb/CD18)poolsinhuman neutrophils.J.Leukocyte Biol.44:545-544. 33. Jones, D. H., F. C. Schmalsteig, K. Dempsey, S. S. Krater, D. D. Nannen, C. W.Smith,and D. C. Anderson. 1990. Subcellular distribution and

mobiliza-tionofMac-l(CDllb/CD18)inneonatalneutrophils.Blood.75:488-498.

34.Lacal, P., R.Pulido,F.Sanchez-Madrid,and F.Mollinedo.1988. Intracel-lular localizationof T200 andMolglycoproteins in human neutrophils. J.Biol. Chem.263:9946-9951.

35.English,D., and V.Graves. 1992. Simultaneous mobilization of Mac-l (CDI Ib /CD 1 8) andformylpeptide chemoattractant receptors in human neutro-phils.Blood.80:776-787.

36.Mollinedo,F., R. Pulido,P. M.Lacal, and F.Sanchez-Madrid. 1991. Mobilizationofgelatinase-richgranules as a regulatorymechanism ofearly

func-tionalresponses in humanneutrophils. Scand.J.Immunol. 34:3343. 37.Petrequin, P.R.,R. F.Todd III, L. J. Devall, L. A. Boxer, and J. T.

Curnutte III. 1987.Association between gelatinase release and increased plasma membraneexpression of the Mol glycoprotein. Blood. 69:605-6 10.

38.Singer,I.S., S. Scott, D. W. Kawka, and D. M.Kazazis.1989. Adhesomes:

specific granules containingreceptorsforlaminin, C3bi/fibrinogen, fibronectin,

and vitronectin in humanpolymorphonuclearleukocytesand monocytes. J. Cell

Biol.109:3169-3182.

39. Yoon, P.S., L. A. Boxer, L. A. Mayo, Y. Y. Angela, and M. S. Wicha. 1987. Humanneutrophil lamininreceptors:activation-dependentreceptor ex-pression. J. Immunol. 138:259-265.

40.Gogstad,G.O., I. Hagen, R. Korsmo, and N. 0. Solum. 1981. Character-izationoftheproteins of isolated human platelet a-granules. Evidence for a separatea-granulepooloftheglycoproteinsI1b andIll.Biochim. Biophys. Acta.

670:150-162.

41.Wencel-Drake,J.D., E. P.Flow,T.J.Kunicki,V. L.Woods,D. M.Keller,

and M. H.Ginsberg.1986. Localization of internalpoolsof membrane

glycopro-teins involved inplateletadhesive responses. Am. J. Pathol. 124:324-334. 42.Lewis,J.C.,R. R.Hantgan,S. C.Stevenson,T.Thornburg,N.Kieffer,J.

Guichard,and J.Breton-Gorius. 1990.FibrinogenandglycoproteinIlb/IIIa

la-calizationduring plateletadhesion. Localizationtothegranulomereand sites of plateletadhesion.Am.J.Pathol. 136:239-252.