The Role of Protected Extracellular Compartments in Interactions between Leukocytes, and Platelets, and Fibrin/Fibrinogen Matrices" J. D. LOIKE!jc R. SILVERSTEIN," S. D. WRIGHT: J. I. WEITZf A. J. HUANGYC AND S. C. SILVERSTEINC CRLmrLabmatmy Deprtment of Phyhlogy and cellrclav Bthpbysics Columk U n d t y CO&e of Physicians and Surgeons 630 W 168th Sheet New Ywk, New Ywk 10032 CmneU UniveWity Medical Col& 1300 Ywk Avenue New Ywk, New Ywk 10021 eThe Rock&lhr U n k i t y 1200 Ywk Avenue New Ywk, New Ywk 10021 fMcMa&r Unha'ty 711 Concession Sheet Hamiltun, Onturi~,Canada L.8V 1C3 INTRODUCTION As inflammatory cells emigrate h m the blood to sites of inflammation they must pass through three barriers: the layer of endothelial cells that lines the blood vessels, the basement membrane, and the underlying extracellular matrix. We have been examining whether polymorphonuclear leukocytes (PMNs) degrade extracellular matrices as they migrate through them. Since both the plasma and the interstitial spaces are replete with protease inhibitors, proteases released by migrating PMNs will be rapidly inactivated. Thus, if these proteases are to affect extracellular matrix proteins, This investigation was supported by grants from the National Institutes of Health (DK39110, AI20516, AI22003, AI30556), Medical Research Council of Canada, Cystic Fibrosis Foundation Research Development Program at Columbia University, Heart and Stroke Foundation of Ontario, and a generous gift from Samuel Rover. J. I. W.is a Caner investigator of the Hean and Stroke Foundation of Ontario. S. D. W.is an Established Investigator of the American Heart Association. Author to whom correspondence should be addressed; Tel.: (212) 305-1510.
163
164
ANNALS NEW YORK ACADEMY OF SCIJ3NCE.S
they must be resistant to plasma anti-proteases, or be released into a compartment that is inaccessible to anti-proteases. In 1982, Campbell et ul. showed that when PMNs are stimulated to adhere to fibronectin-coated surfaces they degrade the substrate-adherent fibronectin even when the surrounding medium contains high levels of protease inhibitors. Subsequently, Wright and Siverstein2 showed that adherence of monocyte-derived macrophages to IgG or complement-coated surfaces leads to the formation of “sealed” extracellular compartments in the zones of contact between the cells and the surface. They showed that these compartments excluded Fab fragments (50,000 MW). Electron microscopic2 and reflection interference contrast microscopic3 studies indicate that this seal surrounds a space(s) between the macrophages and the underlying ligand-coated surhce(s). Baron etul.4 showed that osteoclasts secrete protons and acid hydrolases into a protected compartment formed between themselves and the underlying bone and that bone resorption by osteoclasts depends upon the ability of these cells to form such protected compartments. We have focused on the membrane receptors that mediate formation of extracellular compartments by neutrophils and platelets, the impermeability of these compartments to macromolecules, and the physiological consequences of their impermeability.
PROTECTED EXTRACELLULAR COMPARTMENTS FORMED BY PMNs Fibrinopeptide A a 1-21 is cleaved from the A-ar chain of fibrinogen by PMN elastase, a serine proteinase. No other known PMN enzyme cleaves fibrinogen in this way. By use of a radioimmunoassay that is specific for fibrinopeptide Aa 1-21, Weitz et aL5T6found that smokers, and patients with a1 protease inhibitor deficiency, had much hlgher plasma levels of Aa 1-21 than the plasma of nonsmokers and of individuals who express normal levels of a 1 antiprotease. However, the plasma of individuals with normal levels of plasma antiproteases also contained detectable levels of Aa 1-21. This suggested to us that during their emigration from the blood into the tissues, PMNs might form extracellular compartments that exclude plasma antiproteases and that fibrinogen trapped within these compartments might be degraded by PMN elastase, unopposed by the antiproteases in the surrounding plasma. To explore this hypothesis, Weitz et u1.’ used release of Aa 1-21 from fibrinogencoated nitrocellulose filters to measure elastase activity of PMNs chemotaxing through these filters. Control experiments showed that low concentrations of plasma (1%), completely inhibited A a 1-21 release by purified PMN elastase or by lysed PMNs WLE 1 and ref. 7). In contrast, PMNs stimulated to migrate across fibrinogen-coated filters by the chemoattractant fMLP (N-formyl-methionyl-leucyl-phenyalanine), released significant amounts of Aa 1-21 into the medium (TABLE1 and ref. 7 ) , even when the PMNs were bathed in 100% plasma. Furthermore, concentrations of a 1 proteinase inhibitor or soybean trypsin inhibitor that totally blocked PMN elastase activity (TABLE1 and ref. 7), incompletely inhibited the fibrinogenolytic activity of chemotaxing PMNs WLE 1 and ref. 7 ) . Addition of a low molecular weight elastase inhibitor (Me-Suc-Alaz-ProValCH2CI) to chemotaxing PMNs completely inhibited their fibrinogenolytic activity (TULE 1 and ref. 7). Control experiments showed that this inhibitor did not inactivate intracellular elastase, and did not block PMN adhesion to or chemotaxis through
LOIRE ct al.: PROTECTED EXTIUCELLULAR COMPARTh4ENl-S
TABLE1.
165
Fibrinogenolysis by PMN Lysates and Intact PMNs Aa 1-21 Peptide Released (pmol)
Inhibitor
PMN Lysates
Intact PMNs
None 550 161 Plasma (100%) 0.9 24 a1 Proteinase-inhibitor(0.03 mM) 0.1 19 Soybean trypsin-inhibitor (0.03 mM) 2.1 17 MeO-Suc-Alaz-Pro-ValCHzCl (0.01 mM) 0.1 0.5 Inhibitors at the concentrations indicated in the table were incubated with intact 106 fMLPstimulated PMNs or the lysates of 106 fMLP-stimulated PMNs in chemomxis chambers containing fibrinogen-coatednitrocellulose filters. The preparations were then incubated for 60 min at 37OC and the amount of Aa 1-21 released into the supernatant was assayed as described.’
fibrinogen-coated filters.’ These results indicate that chemoattractant-stimulated PMNs secrete elastase into a site that is inaccessible to high molecular weight antiproteases, but is accessible to low molecular weight antiproteases. We suggest that this site is a protected extracellular compartment formed between PMNs and a fibrinogencoated surface. Fluorescence microscopy was then utilized to visualize whether chemoattractantstimulated PMNs form protected compartments on fibrinogen-coated surfaces. MLPor phorbol ester-stimulated PMNs were allowed to adhere to fibrinogenioated surhces for 4 5 min. The preparations then were incubated with fluorescein-conjugated F(ab’)z goat anti-human fibrinogen. We observed unifbrm staining of the fibrinogencoated surfaces with the exception of the areas underlying the PMNs where no staining was seen (FIG. 1 and data not shown). Control experiments, using PMNs permeabilized with Triton X-100,confirmed that immunoreactive fibrinogen was present beneath these cells (data not shown). Thus the F(ab’)2 did not stain the fibrinogen beneath the PMNs because it did not have access to this compartment. Further experiments showed that when stimulated PMNs adhered to fibrinogencoated surfaces for >60 min at 37OC, they degraded the fibrinogen on the substrate directly underneath them (data not shown). Similar observations were made by Campbell et af. who reported degradation of fibronectin by chemoattractant-stimulated PMNs.
’
82
INTEGRINS MEDIATE ADHESION OF PMNs TO PIBRINOGEN-COATED SURFACES
To identify the plasma membrane receptors that mediate the fbrmation of protected compartments we tested the effects of monoclonal antibodies directed against several PMN surface proteins on the formation of protected compartments. Adherence of phorbol ester-stimulated PMNs to fibrinogen-coated sudces was inhibited by 375% by the addition of monoclonal antibodies directed against the 95 kD subunit of the /32 leukocyte integrins.* Using monoclonal antibodies directed against the several a-chains of these integtvls we h u n d that anti-CDllb (OKM10, directed against the a-chain of complement receptor 3), blocked the adherence of phorbol ester-
166
ANNALS NEW YORK ACADEMY OF SCIENCES
FIGURE 1. F(ab')2 is excluded from mnes of contact between stimulated PMNs and fibrinogn-coatedsu&ces: Phorbol ester-stimulatedPMNs were prepared and allowed to adhere to fibrinogen-coatedglass wells for 5 min as described.8The fibrinogen surface was visualized by incubating the wells with fluorescein conjugated F(ab')z fragments of anti-human fibrinogen and observed under phase (A) or fluorrscence microscopy (B) at 4OOx magnification.
LOME ct al.: PROTECI'ED EXTRACELLULAR COMPARTMENTS
167
300
PDB 250
cu E
E !x W a v)
TNF 200
NONE
150
z E
a
100
50
0
SUBSTRATE ADDITIONS
FIGURE 2. E&ct of peptides or of anti-CDIKD18 antibodies on adhesion of TNF- or phorbol ester-stimulated PMN to Fg-coated surfices. Unstimulated PMNs (5000 ceUs/well) or PMNs treated with TNF (2 ng/ml) or phorbol dibutyrate (300 ng/ml) were allowed to settle onto fibrinogen-coated surfices tbr 30 min at 4OC in the absence or presence of the indicated antibodies or peptides. CeUs were then warmed to 37O for 15 min, washed, fixed and counted as described.12 Adhesion of untreated PMN to surfices coated with human serum albumin (HSA) served as a control. Monoclonal antibodies OKM10, and LeuM5 were used at 20 and 5 &ml, respectively. GPRP was used at a concentration of 1 mg/ml.
stimulated PMNs to fibrinogen-coatedsudces (FIG.2 and ref. 8). Recent experiments have shown that soluble fibrinogen blocks the interaction of C3bi with purified receptor (VanStrijp, Russell, Tuomanen, Brown, and Wright, manuscript in preparation). Furthermore, the ability of CDllbCD18 to recognize fibrinogen has been confirmed by Altieri et aL9 Binding of phorbol ester-stimulated PMN to fibrinogen depends on the presence of the carboxy terminus of fibrinogen. Fragment D, a proteolytic fiagment of fibrinogen that lacks the amino termini of the a,0 and y chains, supported adhesion of stimulated PMNs. Soluble hgment D also blocked PMN adhesion to intact fibrinogens and blocked binding of radiolabeled fibrinogen to CDllb/CD18 on leukocytes.9 We also observed that adhesion of PMN to fibrinogen-coated surfaces and the fbrmation of a tight seal with these sudces was blocked with peptides containing the sequence Lys-Gln-Ala-Gly-AspVal (KQAGDV) which corresponds to amino acids 406411 of the carboxy terminus of fibrinogen's y-thain.8 Closely related peptides were without effect in these experiments.8 Our initial interpretation of these findings was that
168
ANNALS NEW YORK ACADEMY OF SCIENCES
CDllblCD18 directly recognized KQAGDV. However, this interpretation must be modified because KQAGDV Wed to block binding of C3bi to purified CDllb/CD18 (VanStrijp, Russell, Tuomanen, Brown, and Wright, manuscript in preparation) and attempts to purify CDllb/CDl8 on matrices coated with such peptides were not successful (S. D. Wright, unpublished observations). In addition, Altieri9 has reported that a merit D analogue which lacks KQAGVD was still capable of binding to this receptor. Thus, the precise amino acid sequence recognized by CDllblCD18 on the carboxy terminus remains to be identified. The profound effects of peptides containing KQAGDV sequences on PMN may involve other integrins on the leukocyte membrane that induce effects on CDllb/CDl8. This hypothesis is consistent with previous studies1O showing that one class of integrins affects the biological activity of different integrins on the same cells. Tumor necrosis factor (TNF), like phorbol esters, activates 8 2 integrins on leukoc y t e ~ We . ~ ~have shown that TNF also stimulates PMN adhesion to fibrinogencoated sudces.12 To our surprise, however, adherence of TNF-stimulated PMNs to fibrinogen-coatedsudces was not blocked by monoclonal antibodies directed against CDllb (FIG. 2 and ref. 12). Monoclonal antibodies against CDlla also were without effect.12 In contrast, monoclonal antibodies LeuM513 and 3.9,14 which are directed against CDllc (the a-chain of p150/95), inhibited adherence of TNF-stimulated PMNs to fibrinogen-coated sudces by X O % (FIG. 2 and ref. 12). Several lines of evidence indicated that CDlldCD18 recognizes a different ligand on fibrinogen than CDllb/CD18, and that this ligand resides in the amino terminal half of the fibrinogen molecule12; 1) Peptides containing the sequence KQAGDV, which blocked CDllb/CDM-mediated adherence of phorbol ester stimulated PMN to fibrinogen did not inhibit adherence ofTNF-stimulated PMNs to fibrinogen-coated surfaces. 2) TNF markedly stimulated adhesion of PMNs to su&ces coated with a fragment of fibrinogen [the N-terminal disuffide knot (NDSK)] that lacks the carboxy terminus of fibrinogen. Adhesion of TNF-stimulated PMNs to NDSK-coated sudces was blocked by >50% by monoclonal antibody LeuM5 (directed against CDllc), but not by monoclonal antibody OKMlO (directed against the C3bi-binding domain of CDllb). 3) Approximately the same number of unstimulated as TNF-stimulated PMNs adhered to sudces coated with the D fragment (the carboxy terminus), offibrinogen, and this adhesion was unaffected by monoclonal antibody LeuM5. To identify the amino acid sequence of fibrinogen that is a ligand b r CDllc/CD18 we tested the capacity of peptides to block adhesion of TNF-stimulated PMNs to fibrinogen-coatedsu&ces. We found that at a concentration of 1mg/ml, peptides containing the sequence Gly-Pro-4 (GPR) (k.,Gly-Pro-4-Pro, and fibrinogenfragment Aa 1-21) reduced adherence of TNF-stimulated PMNs to fibrinogen-coated sudces by about 50% (FIG.2 and ref. 12). Gly-Pro-Arg-Pro blocked the adherence of TNF-stimulated PMNs to fibrinogen-coatedsurfaces half maximally at about 0.2 mg/ml (0.38 mM). G-H-R-P (corresponding to amino acids 15-18 of fibrinogen’s &chain), peptide G-15 (corresponding to amino acids 397-411 at the carboxy terminus of fibrinogen’s gamma chain), G-P, G-P-G-G, all at concentrations of 1 mg/ml, had no significant efkct on the adhesion of TNF-stimulated PMNs to fibrinogen-coated surfaces.’* In contrast, GPRP did not affect the adherence of phorbol ester-stimulated PMNs to Fg-coated sudces (FIG.2). The requirement b r high concentrations of GPR containing peptides to inhibit PMN adhesion is consistent with our previous observations that high concentrations of peptides are required to block binding of Cfbi-coated erythrocytes to CDllb/
LOME ct al.: P R m D EXTRACELLULAR COMPARTMENTS
169
CDWreceptor bearing cells.l5 This may reflect the fict that peptides act as monomeric ligands, while cell bound C3bi and surfice bound fibrinogen fbrm multimeric ligands. 15 These results show that TNF activates CDllc/CD18 on PMNs, and that when activated, this receptor preferentially binds to a region at the amino terminus of fibrinogen's Aar chain and may recognize a sequence Gly-Pro-A~g(corresponding to residues 17-19) in this region. This is the first indication that the ligand-binding activities of CDllb/CD18 and CDllc/CD18 can be regulated independently of one another on the same cell. The capacity of CDllcKD18 on other cells to mediate cell attachment to fibrinogen-coatedsurfices was recently described by Postigo etal.I6They showed that B lymphocyte activation with a variety of stimuli including phorbol esters and TNF induced dc mCDllc/CD18 cell surfice expression on most B cells without affecting CDllb expression. In addition, the newly expressed CDllc/CD18 mediates B cell adhesion to fibrinogen-coated surfices and enhances cell proliferation.
ANTIBODIES DIRECTED AGAINST CDllb/CD18 BLOCK THE FORMATION OF PROTECTED EXTRACELLULAR COMPARTMENTS BY fMLP-STullULATED PMNs Since antibodies directed against CDllbKD18, and peptides that block the C3bi binding site of CDllb/CDU, block adhesion of phorbol ester- or fMLP-stimulated PMNs to fibrinogen-coated surfices, we expected that these antibodies also would block the formation of protected extracellular compartments. In the presence of these antibodies or peptides, PMN elastase-mediated fibrinolysis was less than that observed in the absence of these antibodies or peptides (TABLE 2 and ref. 8 ). Thus, 0-10 and L10 (a peptide containing KQAGDV) blocked the fbrmation of protected compartments, allowing antiproteinases in the medium to gain access to any elastase secreted by the PMNs, even in regions between the cell and the substrate. Acetylation of the lysine in L10 rendered the peptide incapable of blocking the formation of protected compartments between PMNs and fibrinogen-coated surfices (TABLE 2 and ref. 8). Thus, CDllb/CD18 is an essential PMN receptor that mediates phorbol esteror fMLP-stimulated protected extracellular compartment fbrmation by PMNs. Preliminary work indicates that protected compartment fbrmation between TNF-
TABLE 2. Fibrinogenolysis by PMNs in the Presence of Soybean Trypsin Inhibitor Inhibitor
A a 1-21 Released (pmollml) ~~
None Monoclonal antibody: OKMlO Peptide: L10 Peptide: €-amino lysine acetylated L-10
8.6 0.3 0.15 7.0
fMLP-stimulated PMNs were incubated at O°C for 30 min with various antibodies or peptides. Monoclonal antibody OKMlW was u x d at a concentration of 20 &mI and the peptides8 were used at 1 mg/ml. The PMNs were then incubated at 37OC with fibrinogen-coated filters in the presence of medium containing 60 &ml soybean trypsin inhibitor. After 60 min, supernatants were taken and assayed for Aa 1-21 as described.8
170
ANNALS NEW YORK ACADEMY OF SCIENCES
stimulated PMNs and fibrinogen-coated sudces is altered by the presence of Gly-ProArg-Pro containing peptides and/or monoclonal antibodies directed against CDllc/ CD18 on PMN. These results are consistent with the hypothesis that CDllc/CD18 is essential in the formation of protected compartments by TNF-stimulated PMNs.
PLATELETS FORM PROTECTED EXTRACELLULAR COMPARTMENTS ON FIBRINOGEN-COATED SURFACES In our studies with PMNs, we observed that platelets in the PMN preparation also had the capacity to form protected extracellular compartments when they adhered to fibrinogen-coated surfices. These observations led us to examine the formation of protected extracellular compartments between platelets and fibrinogen-coated sudces. Human platelets were first incubated with thrombin (1U/ml) or ADP (1 pM) and allowed to interact with fibrinogen-coated sudces. Platelets stimulated in this way adhered avidly to fibrinogen-coated sudces where they formed protected extracellular compartments, as measured by the capacity of these platelets to exclude fluorescein conjugated F(ab')2 fragments of goat-anti human fibrinogen antibodies from their undersurfices (manuscript in preparation). Control experiments revealed that the absence of staining underneath the thrombin- or ADP-stimulated platelets was not due to degradation of the underlying fibrinogen by the platelets. However, the protected compartments formed on fibrinogen-coated sudces by stimulated platelets were significantly more permeable to macromolecules than those formed by stimulated PMNs on the same sudces (manuscript in preparation). Fluorescein conjugated Fab antifibrinogen stained the fibrinogen on the substrate underlying stimulated platelets, but was unable to stain the fibrinogen on the substrate underlying PDB or TNF-stimulated PMNs (manuscript in preparation). Thus, the protected compartments formed by stimulated platelets on fibrinogen-coated surfices are permeable to molecules of 50,000-C100,OOO daltons, while those formed by stimulated PMNs on the same surfaces exclude proteins of 50,000 daltons.
DISCUSSION Monocytes, PMNs and platelets form protected extracellular compartments between their plasma membranes and protein ( e . j . , fibrinogen, fibronectin, or elastin) -coated surfices.1 . 2 ~ 1 The ~ , ~ protected ~ compartments formed by PMNs and monocytes exclude most of the macromolecular components of plasma. Therefore, digestive enzymes secreted into these compartments by PMNs and monocytes are free to function in an environment free of most of the protease inhibitors in plasma and in glandular secretions. One possible exception to this generalization is the secretory leukoprotease inhibitor (SLPI) in the lung. SLPI is an 11.7 kD protease inhibitor that is reported to penetrate the protected compartments formed when stimulated PMNs adhere to elastin fibers,l8 thereby inhibiting PMN-mediated elastin degradation in circumstances where SLPI has access to PMNs interacting with elastin fibers. Protected extracellular compartments play an essential role in the reabsorption of bone by osteoclasts.4 The roles of protected extracellular compartments in the movement of leukocytes across basement membranes, the remodeling and reabsorption of
LOIKE
ct al.: PROTWXED EXTklCELLULAR COMPARThUiIUTS
171
extracellular matrices, and in the cytotoxic effector functions of these cells remain to be explored. Our finding that platelets form protected extracellular compartments that are significantly more permeable to proteins than those formed by PMNs was unexpected. One consequence of this difference is that proteases such as plasminogen activator and plasmin may be able to penetrate the compartments formed between platelets and fibrinogedfibrin in clots. The penetration of these enzymes into the compartments formed between platelets and fibrinogen may Eacilitate thrombolysis. Conversely, Eactors that slow or reduce penetration of plasminogen activators and of plasminogen into these compartments may inhibit thrombolysis. Studies are now in progress to evaluate these possibilities.
SUMMARY Polymorphonuclear leukocytes express multiple surfice receptors that mediate their adhesion to extracellular matrices and to other cells. These receptors also play roles in cell migration and phagoqtosis. We have studied the role of one class of polymorphonuclear leukocytes surfice receptors, the /32 integrins, in the interactions of these cells with fibrinogen. We have found that the /32 integrins, CD11b/CD18 (complement receptor three) and CDllc/CD18 mediate attachment of polymorphonuclear leukocytes to fibrinogen-coated surfices. Polymorphonuclear leukocytes must be s t i m ulated with chemoattractants, such as fMLP, or with cytokines, such as tumor necrosis Eactor to bind to fibrinogen via these integrins. Moreover, each of these integrins interacts with a dfferent segment of the fibrinogen molecule. PMN adherence to fibrinogen via CDllblCD18 depends on the carboy terminus of fibrinogen, whereas adherence via CDllc/CD18 depends on the amino terminus of fibrinogen's ar-chain. One of the physiological consequences of these interactions is that polymorphonuclear leukocytes stimulated with either chemoattracmts or TNF form protected compartments at their interfice with fibrinogen-coated surfices and that elastase released into these compartments is inaccessible to protease inhibitors present in the plasma. These protected compartments exclude plasma proteins of >40,000daltons such as crl anti-proteinase, thereby allowing polymorphonuclear leukocyte proteases to degrade matrix proteins within this compartment without interference by plasma anti-proteinases. REPERENCES 1. CAMPBELL, E. J., R M. SENIOR,J. A. MCDONALD & D. L. Cox. 1982. Proteolysis by neutrophils. Relative importance of cell-substrate contact and oxidative inactivation of proteinase inhibitors in vitro. J. Clin. Invest. 70: 845-852. 2. WRIGHT, S. D. & S. C. SILVERSTEIN. 1984. Phagocytosing macrophages exclude proteins from the mnes of contact with opsonizcd targets. Nature 309: 359-361. 3. HEIPLE, J. M., S. D. WRIGHT, N. S. ALLEN& S.C. SXLVERSTJDI. 1990. Macrophages fbm circular mnes of very close apposition to IgG-coated surfdces. Cell Mod. Cytoskeleton 15: 260-270. 4. BARON, R , L. NESS,B. RDWARD & P.J. C~URTOY. 1985. Cell mediated exnacellularacidification and bone-resorption:Evidence of a low pH in resorbing lacunae and localization of a 100 kd lysosomal membrane protein at the osteoclast d e d border. J. Cell Biol. 101: 2210-2222.
172
ANNALS NEW YORK ACADEMY OF SCIENCES
S. BIRKHN & F. J. MORGAN.1986. De5. WEITZ,J. I., S. L. LANDMAN,K. A. CROWLBY, 6. 7.
8.
9. 10.
11.
12. 13. 14.
15. 16.
17. 18.
velopment of an assay for in vivo neumphd dastase activity. Increased elastase activity in patients with alpha-1-proteinase inhibitor deficiency. J. Clin. Invest. 78: 155-162. WEITZ,J. I., K. A. CROWLEY, S. L. LANDMAN, B. I. LIPMAN& J. Yu. 1987. Increased neutrophil elastase activity in cigamtte smokers. Ann. Internal Med. 107: 680-682. WEITZ,J. I., A. J. HUANG,S. L. LANDMAN, S. C. NICHOLSON & s. C. SILVERSTBIN. 1987. Elastase-mediated fibrinanolysis by chemoattractant-stimulated neutrophils occurs in the presence of physiolc&al.con~enuations of antiproteinases. J. Exb. Med. 166: 1836-1850. WRIGHT, S. D., J. I. WEITZ,A. J. HUANG,S. M. LBVIN,S.C. SILVERSTEIN & J. D. L~IKE. 1988. Complement receptor type three (CDllbICD18) of human polymorphonuclear leukocytes recognize fibrinogen. Proc. Natl. Acad. Sci. USA 85: 7734-7738. ALTIERI, D. C., F. R AGBANYO, J. PLBSCIA,M. H. GINSBFXG,T. S. EDGINGKIN & E. F. PLOW.1990. A unique recognition site mediates the interaction of fibrinogen with the leukocyte integrin Mac-1 (CDllb/CD18). J. Biol. Chem. 265: 12119-12122. WRIGHT, S. D., M. R LICHT,L. S. CMGMYL & S. C. SILVBRSTBIN. 1984. Communication between receptors b r difkrent ligands on a single cell: Ligtion of fibronectin recep tors induces a wersible alteration in the function of complement receptors on cultured human monocytes. J. Cell Biol. 99: 336339. MILLER, L. J., D. F. B ~ NN. ,BORREGAARD &T. A. SPRINGER. 1987. Stimulated mobilization of monocytc Mac-1 and p150/95 adhesion proteins from an intracellular vesicular compartment to the cell suhce. J. Clin. Invest. 80: 535-544. L~IKE, J. D., B. SODEIK, L. CAO,S. LEWCONA, J. I. WE^, P.A. DBTMBRS, S.D. WRIGHT & S.C. SILVERSTEIN. 1991. CDc/CD18 on neutrophils is a receptor for fibrinogen.Proc. Natl. Acad. Sci. USA 88: 11044-11048. LAN- L. L., M. A. ARNAOUT, R SCHWARTING,N. L. WARNER & G . D. Ross. 1985. ~150195,third member of the LFA-lCR3 polypeptide t i d y identified by anti Lcu M5 monoclonal antibody. Eur. J. Immunol. 15: 7l3-7l8. N. H o c c & G. D. Ross. 1988. Neumphd and monocyte MYONES,B. L., J. G. DALZELL, cell sulface p150/95 has iC3b-receptor (CR4) activity resembling CR3. J. Clin. Invest. 82: 640-651. HERMANOWSKI-VOSATKA,A., P. A. DFITMERS, 0.W E , S. C. SILVEP.STEIN & S. D. WRIGHT.1988. Clustering of hpnd on the surface of a particle enhances adhesion to receptor-bearing cells. J. Biol. Chem. 263: 17822-17827. POsncO, A. A., A. L. &RBI, F. SANCHEZ-MADBID,M. 0.DE LANDAZURI. 1991. Regulated expression and h c t i o n of CDllc/CD18 integcin on human B lymphocytes. Relation between attachment to fibrinogen and triggering of profitation through CDllc/ CD18. J. Exp. Med. 174 1313-1322. SANDHAUS, R A. 1987. Elastase may play a central role in neutrophil migration through connective tissue. In Pulmonary Emphysema and Proteolysis. J. C. liylor & C. Mittman, Eds. Academic Press. Ncw York. RICE, W. G. &S. J. WEISS.1990. Regulation ofproteolysis at the neutrophil-substrate interi c e by secretory leukoproteax inhibitor. Science 249 178-181.
163
164
ANNALS NEW YORK ACADEMY OF SCIJ3NCE.S
they must be resistant to plasma anti-proteases, or be released into a compartment that is inaccessible to anti-proteases. In 1982, Campbell et ul. showed that when PMNs are stimulated to adhere to fibronectin-coated surfaces they degrade the substrate-adherent fibronectin even when the surrounding medium contains high levels of protease inhibitors. Subsequently, Wright and Siverstein2 showed that adherence of monocyte-derived macrophages to IgG or complement-coated surfaces leads to the formation of “sealed” extracellular compartments in the zones of contact between the cells and the surface. They showed that these compartments excluded Fab fragments (50,000 MW). Electron microscopic2 and reflection interference contrast microscopic3 studies indicate that this seal surrounds a space(s) between the macrophages and the underlying ligand-coated surhce(s). Baron etul.4 showed that osteoclasts secrete protons and acid hydrolases into a protected compartment formed between themselves and the underlying bone and that bone resorption by osteoclasts depends upon the ability of these cells to form such protected compartments. We have focused on the membrane receptors that mediate formation of extracellular compartments by neutrophils and platelets, the impermeability of these compartments to macromolecules, and the physiological consequences of their impermeability.
PROTECTED EXTRACELLULAR COMPARTMENTS FORMED BY PMNs Fibrinopeptide A a 1-21 is cleaved from the A-ar chain of fibrinogen by PMN elastase, a serine proteinase. No other known PMN enzyme cleaves fibrinogen in this way. By use of a radioimmunoassay that is specific for fibrinopeptide Aa 1-21, Weitz et aL5T6found that smokers, and patients with a1 protease inhibitor deficiency, had much hlgher plasma levels of Aa 1-21 than the plasma of nonsmokers and of individuals who express normal levels of a 1 antiprotease. However, the plasma of individuals with normal levels of plasma antiproteases also contained detectable levels of Aa 1-21. This suggested to us that during their emigration from the blood into the tissues, PMNs might form extracellular compartments that exclude plasma antiproteases and that fibrinogen trapped within these compartments might be degraded by PMN elastase, unopposed by the antiproteases in the surrounding plasma. To explore this hypothesis, Weitz et u1.’ used release of Aa 1-21 from fibrinogencoated nitrocellulose filters to measure elastase activity of PMNs chemotaxing through these filters. Control experiments showed that low concentrations of plasma (1%), completely inhibited A a 1-21 release by purified PMN elastase or by lysed PMNs WLE 1 and ref. 7). In contrast, PMNs stimulated to migrate across fibrinogen-coated filters by the chemoattractant fMLP (N-formyl-methionyl-leucyl-phenyalanine), released significant amounts of Aa 1-21 into the medium (TABLE1 and ref. 7 ) , even when the PMNs were bathed in 100% plasma. Furthermore, concentrations of a 1 proteinase inhibitor or soybean trypsin inhibitor that totally blocked PMN elastase activity (TABLE1 and ref. 7), incompletely inhibited the fibrinogenolytic activity of chemotaxing PMNs WLE 1 and ref. 7 ) . Addition of a low molecular weight elastase inhibitor (Me-Suc-Alaz-ProValCH2CI) to chemotaxing PMNs completely inhibited their fibrinogenolytic activity (TULE 1 and ref. 7). Control experiments showed that this inhibitor did not inactivate intracellular elastase, and did not block PMN adhesion to or chemotaxis through
LOIRE ct al.: PROTECTED EXTIUCELLULAR COMPARTh4ENl-S
TABLE1.
165
Fibrinogenolysis by PMN Lysates and Intact PMNs Aa 1-21 Peptide Released (pmol)
Inhibitor
PMN Lysates
Intact PMNs
None 550 161 Plasma (100%) 0.9 24 a1 Proteinase-inhibitor(0.03 mM) 0.1 19 Soybean trypsin-inhibitor (0.03 mM) 2.1 17 MeO-Suc-Alaz-Pro-ValCHzCl (0.01 mM) 0.1 0.5 Inhibitors at the concentrations indicated in the table were incubated with intact 106 fMLPstimulated PMNs or the lysates of 106 fMLP-stimulated PMNs in chemomxis chambers containing fibrinogen-coatednitrocellulose filters. The preparations were then incubated for 60 min at 37OC and the amount of Aa 1-21 released into the supernatant was assayed as described.’
fibrinogen-coated filters.’ These results indicate that chemoattractant-stimulated PMNs secrete elastase into a site that is inaccessible to high molecular weight antiproteases, but is accessible to low molecular weight antiproteases. We suggest that this site is a protected extracellular compartment formed between PMNs and a fibrinogencoated surface. Fluorescence microscopy was then utilized to visualize whether chemoattractantstimulated PMNs form protected compartments on fibrinogen-coated surfaces. MLPor phorbol ester-stimulated PMNs were allowed to adhere to fibrinogenioated surhces for 4 5 min. The preparations then were incubated with fluorescein-conjugated F(ab’)z goat anti-human fibrinogen. We observed unifbrm staining of the fibrinogencoated surfaces with the exception of the areas underlying the PMNs where no staining was seen (FIG. 1 and data not shown). Control experiments, using PMNs permeabilized with Triton X-100,confirmed that immunoreactive fibrinogen was present beneath these cells (data not shown). Thus the F(ab’)2 did not stain the fibrinogen beneath the PMNs because it did not have access to this compartment. Further experiments showed that when stimulated PMNs adhered to fibrinogencoated surfaces for >60 min at 37OC, they degraded the fibrinogen on the substrate directly underneath them (data not shown). Similar observations were made by Campbell et af. who reported degradation of fibronectin by chemoattractant-stimulated PMNs.
’
82
INTEGRINS MEDIATE ADHESION OF PMNs TO PIBRINOGEN-COATED SURFACES
To identify the plasma membrane receptors that mediate the fbrmation of protected compartments we tested the effects of monoclonal antibodies directed against several PMN surface proteins on the formation of protected compartments. Adherence of phorbol ester-stimulated PMNs to fibrinogen-coated sudces was inhibited by 375% by the addition of monoclonal antibodies directed against the 95 kD subunit of the /32 leukocyte integrins.* Using monoclonal antibodies directed against the several a-chains of these integtvls we h u n d that anti-CDllb (OKM10, directed against the a-chain of complement receptor 3), blocked the adherence of phorbol ester-
166
ANNALS NEW YORK ACADEMY OF SCIENCES
FIGURE 1. F(ab')2 is excluded from mnes of contact between stimulated PMNs and fibrinogn-coatedsu&ces: Phorbol ester-stimulatedPMNs were prepared and allowed to adhere to fibrinogen-coatedglass wells for 5 min as described.8The fibrinogen surface was visualized by incubating the wells with fluorescein conjugated F(ab')z fragments of anti-human fibrinogen and observed under phase (A) or fluorrscence microscopy (B) at 4OOx magnification.
LOME ct al.: PROTECI'ED EXTRACELLULAR COMPARTMENTS
167
300
PDB 250
cu E
E !x W a v)
TNF 200
NONE
150
z E
a
100
50
0
SUBSTRATE ADDITIONS
FIGURE 2. E&ct of peptides or of anti-CDIKD18 antibodies on adhesion of TNF- or phorbol ester-stimulated PMN to Fg-coated surfices. Unstimulated PMNs (5000 ceUs/well) or PMNs treated with TNF (2 ng/ml) or phorbol dibutyrate (300 ng/ml) were allowed to settle onto fibrinogen-coated surfices tbr 30 min at 4OC in the absence or presence of the indicated antibodies or peptides. CeUs were then warmed to 37O for 15 min, washed, fixed and counted as described.12 Adhesion of untreated PMN to surfices coated with human serum albumin (HSA) served as a control. Monoclonal antibodies OKM10, and LeuM5 were used at 20 and 5 &ml, respectively. GPRP was used at a concentration of 1 mg/ml.
stimulated PMNs to fibrinogen-coatedsudces (FIG.2 and ref. 8). Recent experiments have shown that soluble fibrinogen blocks the interaction of C3bi with purified receptor (VanStrijp, Russell, Tuomanen, Brown, and Wright, manuscript in preparation). Furthermore, the ability of CDllbCD18 to recognize fibrinogen has been confirmed by Altieri et aL9 Binding of phorbol ester-stimulated PMN to fibrinogen depends on the presence of the carboxy terminus of fibrinogen. Fragment D, a proteolytic fiagment of fibrinogen that lacks the amino termini of the a,0 and y chains, supported adhesion of stimulated PMNs. Soluble hgment D also blocked PMN adhesion to intact fibrinogens and blocked binding of radiolabeled fibrinogen to CDllb/CD18 on leukocytes.9 We also observed that adhesion of PMN to fibrinogen-coated surfaces and the fbrmation of a tight seal with these sudces was blocked with peptides containing the sequence Lys-Gln-Ala-Gly-AspVal (KQAGDV) which corresponds to amino acids 406411 of the carboxy terminus of fibrinogen's y-thain.8 Closely related peptides were without effect in these experiments.8 Our initial interpretation of these findings was that
168
ANNALS NEW YORK ACADEMY OF SCIENCES
CDllblCD18 directly recognized KQAGDV. However, this interpretation must be modified because KQAGDV Wed to block binding of C3bi to purified CDllb/CD18 (VanStrijp, Russell, Tuomanen, Brown, and Wright, manuscript in preparation) and attempts to purify CDllb/CDl8 on matrices coated with such peptides were not successful (S. D. Wright, unpublished observations). In addition, Altieri9 has reported that a merit D analogue which lacks KQAGVD was still capable of binding to this receptor. Thus, the precise amino acid sequence recognized by CDllblCD18 on the carboxy terminus remains to be identified. The profound effects of peptides containing KQAGDV sequences on PMN may involve other integrins on the leukocyte membrane that induce effects on CDllb/CDl8. This hypothesis is consistent with previous studies1O showing that one class of integrins affects the biological activity of different integrins on the same cells. Tumor necrosis factor (TNF), like phorbol esters, activates 8 2 integrins on leukoc y t e ~ We . ~ ~have shown that TNF also stimulates PMN adhesion to fibrinogencoated sudces.12 To our surprise, however, adherence of TNF-stimulated PMNs to fibrinogen-coatedsudces was not blocked by monoclonal antibodies directed against CDllb (FIG. 2 and ref. 12). Monoclonal antibodies against CDlla also were without effect.12 In contrast, monoclonal antibodies LeuM513 and 3.9,14 which are directed against CDllc (the a-chain of p150/95), inhibited adherence of TNF-stimulated PMNs to fibrinogen-coated sudces by X O % (FIG. 2 and ref. 12). Several lines of evidence indicated that CDlldCD18 recognizes a different ligand on fibrinogen than CDllb/CD18, and that this ligand resides in the amino terminal half of the fibrinogen molecule12; 1) Peptides containing the sequence KQAGDV, which blocked CDllb/CDM-mediated adherence of phorbol ester stimulated PMN to fibrinogen did not inhibit adherence ofTNF-stimulated PMNs to fibrinogen-coated surfaces. 2) TNF markedly stimulated adhesion of PMNs to su&ces coated with a fragment of fibrinogen [the N-terminal disuffide knot (NDSK)] that lacks the carboxy terminus of fibrinogen. Adhesion of TNF-stimulated PMNs to NDSK-coated sudces was blocked by >50% by monoclonal antibody LeuM5 (directed against CDllc), but not by monoclonal antibody OKMlO (directed against the C3bi-binding domain of CDllb). 3) Approximately the same number of unstimulated as TNF-stimulated PMNs adhered to sudces coated with the D fragment (the carboxy terminus), offibrinogen, and this adhesion was unaffected by monoclonal antibody LeuM5. To identify the amino acid sequence of fibrinogen that is a ligand b r CDllc/CD18 we tested the capacity of peptides to block adhesion of TNF-stimulated PMNs to fibrinogen-coatedsu&ces. We found that at a concentration of 1mg/ml, peptides containing the sequence Gly-Pro-4 (GPR) (k.,Gly-Pro-4-Pro, and fibrinogenfragment Aa 1-21) reduced adherence of TNF-stimulated PMNs to fibrinogen-coated sudces by about 50% (FIG.2 and ref. 12). Gly-Pro-Arg-Pro blocked the adherence of TNF-stimulated PMNs to fibrinogen-coatedsurfaces half maximally at about 0.2 mg/ml (0.38 mM). G-H-R-P (corresponding to amino acids 15-18 of fibrinogen’s &chain), peptide G-15 (corresponding to amino acids 397-411 at the carboxy terminus of fibrinogen’s gamma chain), G-P, G-P-G-G, all at concentrations of 1 mg/ml, had no significant efkct on the adhesion of TNF-stimulated PMNs to fibrinogen-coated surfaces.’* In contrast, GPRP did not affect the adherence of phorbol ester-stimulated PMNs to Fg-coated sudces (FIG.2). The requirement b r high concentrations of GPR containing peptides to inhibit PMN adhesion is consistent with our previous observations that high concentrations of peptides are required to block binding of Cfbi-coated erythrocytes to CDllb/
LOME ct al.: P R m D EXTRACELLULAR COMPARTMENTS
169
CDWreceptor bearing cells.l5 This may reflect the fict that peptides act as monomeric ligands, while cell bound C3bi and surfice bound fibrinogen fbrm multimeric ligands. 15 These results show that TNF activates CDllc/CD18 on PMNs, and that when activated, this receptor preferentially binds to a region at the amino terminus of fibrinogen's Aar chain and may recognize a sequence Gly-Pro-A~g(corresponding to residues 17-19) in this region. This is the first indication that the ligand-binding activities of CDllb/CD18 and CDllc/CD18 can be regulated independently of one another on the same cell. The capacity of CDllcKD18 on other cells to mediate cell attachment to fibrinogen-coatedsurfices was recently described by Postigo etal.I6They showed that B lymphocyte activation with a variety of stimuli including phorbol esters and TNF induced dc mCDllc/CD18 cell surfice expression on most B cells without affecting CDllb expression. In addition, the newly expressed CDllc/CD18 mediates B cell adhesion to fibrinogen-coated surfices and enhances cell proliferation.
ANTIBODIES DIRECTED AGAINST CDllb/CD18 BLOCK THE FORMATION OF PROTECTED EXTRACELLULAR COMPARTMENTS BY fMLP-STullULATED PMNs Since antibodies directed against CDllbKD18, and peptides that block the C3bi binding site of CDllb/CDU, block adhesion of phorbol ester- or fMLP-stimulated PMNs to fibrinogen-coated surfices, we expected that these antibodies also would block the formation of protected extracellular compartments. In the presence of these antibodies or peptides, PMN elastase-mediated fibrinolysis was less than that observed in the absence of these antibodies or peptides (TABLE 2 and ref. 8 ). Thus, 0-10 and L10 (a peptide containing KQAGDV) blocked the fbrmation of protected compartments, allowing antiproteinases in the medium to gain access to any elastase secreted by the PMNs, even in regions between the cell and the substrate. Acetylation of the lysine in L10 rendered the peptide incapable of blocking the formation of protected compartments between PMNs and fibrinogen-coated surfices (TABLE 2 and ref. 8). Thus, CDllb/CD18 is an essential PMN receptor that mediates phorbol esteror fMLP-stimulated protected extracellular compartment fbrmation by PMNs. Preliminary work indicates that protected compartment fbrmation between TNF-
TABLE 2. Fibrinogenolysis by PMNs in the Presence of Soybean Trypsin Inhibitor Inhibitor
A a 1-21 Released (pmollml) ~~
None Monoclonal antibody: OKMlO Peptide: L10 Peptide: €-amino lysine acetylated L-10
8.6 0.3 0.15 7.0
fMLP-stimulated PMNs were incubated at O°C for 30 min with various antibodies or peptides. Monoclonal antibody OKMlW was u x d at a concentration of 20 &mI and the peptides8 were used at 1 mg/ml. The PMNs were then incubated at 37OC with fibrinogen-coated filters in the presence of medium containing 60 &ml soybean trypsin inhibitor. After 60 min, supernatants were taken and assayed for Aa 1-21 as described.8
170
ANNALS NEW YORK ACADEMY OF SCIENCES
stimulated PMNs and fibrinogen-coated sudces is altered by the presence of Gly-ProArg-Pro containing peptides and/or monoclonal antibodies directed against CDllc/ CD18 on PMN. These results are consistent with the hypothesis that CDllc/CD18 is essential in the formation of protected compartments by TNF-stimulated PMNs.
PLATELETS FORM PROTECTED EXTRACELLULAR COMPARTMENTS ON FIBRINOGEN-COATED SURFACES In our studies with PMNs, we observed that platelets in the PMN preparation also had the capacity to form protected extracellular compartments when they adhered to fibrinogen-coated surfices. These observations led us to examine the formation of protected extracellular compartments between platelets and fibrinogen-coated sudces. Human platelets were first incubated with thrombin (1U/ml) or ADP (1 pM) and allowed to interact with fibrinogen-coated sudces. Platelets stimulated in this way adhered avidly to fibrinogen-coated sudces where they formed protected extracellular compartments, as measured by the capacity of these platelets to exclude fluorescein conjugated F(ab')2 fragments of goat-anti human fibrinogen antibodies from their undersurfices (manuscript in preparation). Control experiments revealed that the absence of staining underneath the thrombin- or ADP-stimulated platelets was not due to degradation of the underlying fibrinogen by the platelets. However, the protected compartments formed on fibrinogen-coated sudces by stimulated platelets were significantly more permeable to macromolecules than those formed by stimulated PMNs on the same sudces (manuscript in preparation). Fluorescein conjugated Fab antifibrinogen stained the fibrinogen on the substrate underlying stimulated platelets, but was unable to stain the fibrinogen on the substrate underlying PDB or TNF-stimulated PMNs (manuscript in preparation). Thus, the protected compartments formed by stimulated platelets on fibrinogen-coated surfices are permeable to molecules of 50,000-C100,OOO daltons, while those formed by stimulated PMNs on the same surfaces exclude proteins of 50,000 daltons.
DISCUSSION Monocytes, PMNs and platelets form protected extracellular compartments between their plasma membranes and protein ( e . j . , fibrinogen, fibronectin, or elastin) -coated surfices.1 . 2 ~ 1 The ~ , ~ protected ~ compartments formed by PMNs and monocytes exclude most of the macromolecular components of plasma. Therefore, digestive enzymes secreted into these compartments by PMNs and monocytes are free to function in an environment free of most of the protease inhibitors in plasma and in glandular secretions. One possible exception to this generalization is the secretory leukoprotease inhibitor (SLPI) in the lung. SLPI is an 11.7 kD protease inhibitor that is reported to penetrate the protected compartments formed when stimulated PMNs adhere to elastin fibers,l8 thereby inhibiting PMN-mediated elastin degradation in circumstances where SLPI has access to PMNs interacting with elastin fibers. Protected extracellular compartments play an essential role in the reabsorption of bone by osteoclasts.4 The roles of protected extracellular compartments in the movement of leukocytes across basement membranes, the remodeling and reabsorption of
LOIKE
ct al.: PROTWXED EXTklCELLULAR COMPARThUiIUTS
171
extracellular matrices, and in the cytotoxic effector functions of these cells remain to be explored. Our finding that platelets form protected extracellular compartments that are significantly more permeable to proteins than those formed by PMNs was unexpected. One consequence of this difference is that proteases such as plasminogen activator and plasmin may be able to penetrate the compartments formed between platelets and fibrinogedfibrin in clots. The penetration of these enzymes into the compartments formed between platelets and fibrinogen may Eacilitate thrombolysis. Conversely, Eactors that slow or reduce penetration of plasminogen activators and of plasminogen into these compartments may inhibit thrombolysis. Studies are now in progress to evaluate these possibilities.
SUMMARY Polymorphonuclear leukocytes express multiple surfice receptors that mediate their adhesion to extracellular matrices and to other cells. These receptors also play roles in cell migration and phagoqtosis. We have studied the role of one class of polymorphonuclear leukocytes surfice receptors, the /32 integrins, in the interactions of these cells with fibrinogen. We have found that the /32 integrins, CD11b/CD18 (complement receptor three) and CDllc/CD18 mediate attachment of polymorphonuclear leukocytes to fibrinogen-coated surfices. Polymorphonuclear leukocytes must be s t i m ulated with chemoattractants, such as fMLP, or with cytokines, such as tumor necrosis Eactor to bind to fibrinogen via these integrins. Moreover, each of these integrins interacts with a dfferent segment of the fibrinogen molecule. PMN adherence to fibrinogen via CDllblCD18 depends on the carboy terminus of fibrinogen, whereas adherence via CDllc/CD18 depends on the amino terminus of fibrinogen's ar-chain. One of the physiological consequences of these interactions is that polymorphonuclear leukocytes stimulated with either chemoattracmts or TNF form protected compartments at their interfice with fibrinogen-coated surfices and that elastase released into these compartments is inaccessible to protease inhibitors present in the plasma. These protected compartments exclude plasma proteins of >40,000daltons such as crl anti-proteinase, thereby allowing polymorphonuclear leukocyte proteases to degrade matrix proteins within this compartment without interference by plasma anti-proteinases. REPERENCES 1. CAMPBELL, E. J., R M. SENIOR,J. A. MCDONALD & D. L. Cox. 1982. Proteolysis by neutrophils. Relative importance of cell-substrate contact and oxidative inactivation of proteinase inhibitors in vitro. J. Clin. Invest. 70: 845-852. 2. WRIGHT, S. D. & S. C. SILVERSTEIN. 1984. Phagocytosing macrophages exclude proteins from the mnes of contact with opsonizcd targets. Nature 309: 359-361. 3. HEIPLE, J. M., S. D. WRIGHT, N. S. ALLEN& S.C. SXLVERSTJDI. 1990. Macrophages fbm circular mnes of very close apposition to IgG-coated surfdces. Cell Mod. Cytoskeleton 15: 260-270. 4. BARON, R , L. NESS,B. RDWARD & P.J. C~URTOY. 1985. Cell mediated exnacellularacidification and bone-resorption:Evidence of a low pH in resorbing lacunae and localization of a 100 kd lysosomal membrane protein at the osteoclast d e d border. J. Cell Biol. 101: 2210-2222.
172
ANNALS NEW YORK ACADEMY OF SCIENCES
S. BIRKHN & F. J. MORGAN.1986. De5. WEITZ,J. I., S. L. LANDMAN,K. A. CROWLBY, 6. 7.
8.
9. 10.
11.
12. 13. 14.
15. 16.
17. 18.
velopment of an assay for in vivo neumphd dastase activity. Increased elastase activity in patients with alpha-1-proteinase inhibitor deficiency. J. Clin. Invest. 78: 155-162. WEITZ,J. I., K. A. CROWLEY, S. L. LANDMAN, B. I. LIPMAN& J. Yu. 1987. Increased neutrophil elastase activity in cigamtte smokers. Ann. Internal Med. 107: 680-682. WEITZ,J. I., A. J. HUANG,S. L. LANDMAN, S. C. NICHOLSON & s. C. SILVERSTBIN. 1987. Elastase-mediated fibrinanolysis by chemoattractant-stimulated neutrophils occurs in the presence of physiolc&al.con~enuations of antiproteinases. J. Exb. Med. 166: 1836-1850. WRIGHT, S. D., J. I. WEITZ,A. J. HUANG,S. M. LBVIN,S.C. SILVERSTEIN & J. D. L~IKE. 1988. Complement receptor type three (CDllbICD18) of human polymorphonuclear leukocytes recognize fibrinogen. Proc. Natl. Acad. Sci. USA 85: 7734-7738. ALTIERI, D. C., F. R AGBANYO, J. PLBSCIA,M. H. GINSBFXG,T. S. EDGINGKIN & E. F. PLOW.1990. A unique recognition site mediates the interaction of fibrinogen with the leukocyte integrin Mac-1 (CDllb/CD18). J. Biol. Chem. 265: 12119-12122. WRIGHT, S. D., M. R LICHT,L. S. CMGMYL & S. C. SILVBRSTBIN. 1984. Communication between receptors b r difkrent ligands on a single cell: Ligtion of fibronectin recep tors induces a wersible alteration in the function of complement receptors on cultured human monocytes. J. Cell Biol. 99: 336339. MILLER, L. J., D. F. B ~ NN. ,BORREGAARD &T. A. SPRINGER. 1987. Stimulated mobilization of monocytc Mac-1 and p150/95 adhesion proteins from an intracellular vesicular compartment to the cell suhce. J. Clin. Invest. 80: 535-544. L~IKE, J. D., B. SODEIK, L. CAO,S. LEWCONA, J. I. WE^, P.A. DBTMBRS, S.D. WRIGHT & S.C. SILVERSTEIN. 1991. CDc/CD18 on neutrophils is a receptor for fibrinogen.Proc. Natl. Acad. Sci. USA 88: 11044-11048. LAN- L. L., M. A. ARNAOUT, R SCHWARTING,N. L. WARNER & G . D. Ross. 1985. ~150195,third member of the LFA-lCR3 polypeptide t i d y identified by anti Lcu M5 monoclonal antibody. Eur. J. Immunol. 15: 7l3-7l8. N. H o c c & G. D. Ross. 1988. Neumphd and monocyte MYONES,B. L., J. G. DALZELL, cell sulface p150/95 has iC3b-receptor (CR4) activity resembling CR3. J. Clin. Invest. 82: 640-651. HERMANOWSKI-VOSATKA,A., P. A. DFITMERS, 0.W E , S. C. SILVEP.STEIN & S. D. WRIGHT.1988. Clustering of hpnd on the surface of a particle enhances adhesion to receptor-bearing cells. J. Biol. Chem. 263: 17822-17827. POsncO, A. A., A. L. &RBI, F. SANCHEZ-MADBID,M. 0.DE LANDAZURI. 1991. Regulated expression and h c t i o n of CDllc/CD18 integcin on human B lymphocytes. Relation between attachment to fibrinogen and triggering of profitation through CDllc/ CD18. J. Exp. Med. 174 1313-1322. SANDHAUS, R A. 1987. Elastase may play a central role in neutrophil migration through connective tissue. In Pulmonary Emphysema and Proteolysis. J. C. liylor & C. Mittman, Eds. Academic Press. Ncw York. RICE, W. G. &S. J. WEISS.1990. Regulation ofproteolysis at the neutrophil-substrate interi c e by secretory leukoproteax inhibitor. Science 249 178-181.
