Vol.
172,
November
No.
3, 1990
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
Pages
15, 1990
1349-1356
REQUIREMENT FOR SIALIC ACID ON NEUTROPHILS IN A GMP- 140(PADGEM) MEDIATED ADHESIVE INTERACTION WITH ACTIVATED PLATELETS Laura Corral, Mark S. Singer, *Bruce A. Macher, and Steven D. Rosen Departmentof Anatomy andProgramin Immunology, University of California, SanFrancisco,CA 94143-0452 *Departmentsof Chemistry and Biochemistry, SanFranciscoState University, SanFrancisco,CA 94132 Received
October
4,
1990
Summarv. Platelet GMP- 140,alongwith ELAM- 1 andgp9OmL, comprise the l.,EC-C&l family of cell-cell adhesionproteins. The three proteinsdemonstratea highly relateddomainorganization, which includesan extracellular calcium-typelectin motif. gp90MEL,a lymphocyte homing receptor, mediateslymphocyte attachmentto high endothelialvenulesof lymph nodesthrough recognition of a sialylatedligand on the endothelialcells. The rosettingof neutrophilsor promyelocytic HL60 cellsby activated plateletsis mediatedby GMP- 140on the platelets. We showherethat treatmentof neutrophilsor HL60 cellswith 3 broad spectrumsialidasescompletely prevents rosetting. However, the Newcastlediseasevirus sialidase,an enzyme specific for a2,3 and a28 linkagesof sialic acid doesnot affect rosettingof HL60 cells. Theseresultsindicate that the ligand for GMP- 140requiressialicacid and suggestthat an a2,6 linkage may be critical, ‘31990Academic Press.IIIC. Granular-membraneprotein (GMP- 140), alsoknown asPADGEM or CD62, is a glycoprotein of 140,000daltonsfound in the alphagranulesof plateletsand the Weibel-Palade bodiesof endothelialcells(l-5). Upon stimulationof thesecellsby agonistssuchasthrombin, degmnulationleadsto the rapid externalization of this membrane-bound protein to the plasma membrane.Recentstudieshave establishedthat GMP-140 mediatesthe calcium-dependent adhesionof neutrophilsor monocytesto activated platelets(6, 7) aswell asthe early binding of neutrophilsto histamine-or thrombin-stimulatedendothelium(8). Possiblefunctional consequences of theseadhesiveeventsincludethe removal of activated plateletsfrom the blood by neutrophils(or monocytes)and the rapid adherenceof neutrophilsto endotheliumat sitesof acute inflammation. The molecularcloning of GMP- 140hasrevealeda modularprotein consistingof an aminoterminal lectin-like domain(calcium-dependent),an EGF motif, 9 tandemconsensus repeats relatedto thosefound in complement-bindingproteins,a transmembranedomain,anda short cytosolic tail (9). Two other proteins, ELAM- 1andgp90MEL, sharea high degreeof sequence homology with GMP-140 and the samebasicdomainstructure,but are distinctive in the numberof consensus repeats(lo- 12). ELAM- 1 is a cytokine-inducible moleculeon endothelialcellswhich Abbreviations: Au, Arthrobacer ureafaciens ; C-type, calcium-type; Cp, Cfostridiumperfringens HEV, high endothelialvenule; NeuAc, 5 N-acetylneumminicacid; NDV, Newcastlediseasevirus; Vc, Vibrio choler .
Vol.
172,
No.
3, 1990
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
mediates neutrophil attachment (10, 13). 1~ viva, this molecule may promote neutrophil adherence to endothelium during the later stages of certain acute inflammatory reactions. gp90MEL, a constitutive cell surface protein on murine lymphocytes identified by the MEL- 14 mAb, is responsible for the organ-selective adherence (“homing”) of lymphocytes to the endothelial cells of high endothelial venules (HEV) in lymph nodes, an adhesive event that initiates extravasation during the process of lymphocyte recirculation (14- 16). These three proteins, each of which is involved in a highly selective cellular interaction within the blood vascular compartment, comprise a new family of cell-cell adhesion molecules designated as the LEC-CAMS (17) or selectins (8). In the case of gp9OMEL, the lectin domain has been directly implicated in the adhesive function of the receptor. Specific carbohydrates have been identified that interact with this domain in a calcium-dependent manner and competitively inhibit lymphocyte attachment to HEV ( 18,22). Furthermore, the HEV-ligand for gp’)@ELdemonstrates
a functional requirement for sialic acid as
follows: 1) lymphocyte attachment to HEV is blocked by pretreatment of the HEV with sialidases or with the sialic acid specific lectin, Limaxflavus agglutinin (23-25); 2) the binding of a recombinant form of gp90MELto HEV in tissue sections is abrogated by exposure of the HEV to either sialidases or Limax agglutinin (25); and 3) desialylation of the isolated HEV-ligand for gp90MELprevents
its binding to gp90 MEL (Imai, Singer, Fennie, Lasky, and Rosen, submitted).
In view of the strong structural similarities between gp90 MELand the other members of the LECCAM family, we investigated whether a sialic acid requirement also exists for the binding of GMPI40 to its ligand. MATERIALS
AND METHODS
Cell Culture and Cell Preparation. The HL60 human promyelocytic cell line was cultured in RPMI- 1640, supplemented with 10% fetal bovine serum, 100 U/ml penicillin and 100 ug/ml streptomycin. All culture reagents were obtained from the UCSF Cell Culture Facility. The cells were washed 3 times by centrifugation (150 x g) and resuspended in Dulbecco’s PBS at a concentration of 3 x 107 cells/ml. The cells were fixed by the addition of an equal volume of 1% parafotmaldehyde (EM Sciences, Cherry Hill, NJ) in 0.1 M sodium cacodylate, pH 7.3. After 1hr at 40 C, the cells were washed 3 times and suspended in PB buffer (0.0 1M Na,H;P04,0.1 SM NaCl, 5mM CaC12, pH 6.5) or PBS. Polymorphonuclear neutrophils were isolated from fresh heparinized human blood by density gradient centrifugation (Mono-Poly Resolving Media, Flow Laboratories, McLean, VA) following the manufacturer’s instructions. The neutrophils were fixed according to the procedure of Hamburger and McEver (7) and washed 3 times. Enzyme Treatment. Fixed HL60 cells (150 ul at 3 x 107 cells/ml in PB) were treated with either Arthrobacter ureafaciens neuraminidase (Calbiochem, Inc., San Diego, CA) or Clostridium perftingens neuraminidase (Type X, affinity purified, Sigma Chemical Co., St. Louis, MO) at a final concentration of 0.1 U/ml (international units) or with Vibrio cholera neuraminidase (Gibco, Grand Island, NY) at a final concentration of 250 U/ml (according to the manufacturer’s definition of a unit) and incubated at 37O C for 1 hr. Fixed neutrophils were treated with C. perfringens neuraminidase under the same conditions used for HL60 cells. In some experiments, HL60 cells were exposed to Arthrobacterureafaciens neumminidase (0.1 U/ml final ) in PBS instead of PB. Newcastle disease virus (NDV) sialidase was prepared and assayed as described by Paulson et al. (26). Fixed HL60 cells at 4 x 107/ml in PBS were mixed with NDV sialidase (0.2 U/ml final) and incubated at 370 C for 1 hr. After exposure to an enzyme, the cells were washed 3 times with PBS before use in the platelet binding assay. Control samples were treated in parallel without exposure to enzymes. Platelet isolation. Blood from normal human donors was anticoagulated with ACD (85 mM sodium citrate, 7 1 mM citric acid, 111 mM D-glucose) at a 6: 1 (v/v) mtio. After centrifugation at 1350
Vol.
172, No. 3, 1990
BIOCHEMICAL
AND BIOPHYSICAL
RESEARCH COMMUNICATIONS
200 x g for 20 min at 22OC, the platelet-rich plasmawaspooledand fractionated by gel filtration at 220C with SepharoseCL-2B (Sigma ChemicalCo.), equilibratedwith HEPES buffer (5.5 mM D-glucose, 137mM NaHC03,2.68 mM KCl, 20.0 mM HEPES, 0.35% humanalbumin), pH 7.4. Plateletswere stimulatedwith humanthrombin (T-6759, SigmaChemicalCo.) at a final concentrationof 0.15 U/ml for a minimum of 10 min and were usedin the adhesionassayswithin 30 min of activation. In somecases,the isolatedplateletswere already activated asdetermined morphologically and were usedwithout thrombin treatment. Similar resultswere obtainedwith thrombin-activatedplateletsand thoseactivated during isolation. Platelet Rosetting Assay. The procedurefor platelet rosettingwasslightly modified from the protocol of Larsenet al. (6). For eachsample,50 ul of cells (3 x 107cells/ml in PBS) were mixed with an equalvolume of activatedplatelets(2 x IO* cells/ml) in a microfugetube and allowedto react for 20 min at 220 C without agitation. Each treatmentwascarried out in 3-4 replicates. The cell suspensions were fixed by the addition of glutaraldehydeto a final concentrationof 1.25%. Platelet rosettingof cellswas determinedby Hoffman modulationcontrastmicroscopy. Cellswith two or more adherentplateletswere scoredasrosetted. Samples(l-2) of cells (SO-loo) were evaluatedfrom eachreplicatetube. Resultsare indicatedaspercent rosettedcellsbasedon the meansof the 3-4 replicates. SEMs denotethe statisticalvariability. The effect of calcium chelation was determinedby exposingactivated plateletsto 10mM EGTA for 10min prior to mixing with the HL60 cellsor neutrophils Flow Cytometry Analysis. Fixed HL60 cells preparedasabove (3.3 x 107/ml) were exposedto NDV sialidase(0.2 U/ml) in PBS, Arthrobacter ureafaciensneuraminidase(0.1 U/ml final) in PBS, or to PBS aloneand incubatedfor 1 hr at 37OC. The cellswere washed3 timesin PBS containing 1% fetal bovine serum(FBS). The washedcellswere exposedto CSLEX- 1 mAb (27) or to VIM-2 mAb (28), eachdiluted asan ascitesfluid l/50 in PBS-l% FBS. Incubation was for 30 min at 4OC. The cellswere sequentiallyexposedto to biotinylated-horseanti-mouseIgG (Vector Laboratories,Burlingame,CA) and FITC-extrAvidin (SigmaChemicalCo.) accordingto the manufacturers’recommendations.Fluorescencehistogramprofiles were determinedwith a Becton-DickinsonFACScan II flow cytometer. The backgroundlevel of fluorescencewas defined with no primary antibody added, RESULTS A well-describedphenomenonis the binding of activatedplateletsto neutrophilsor the HL60 promyelocytic cell line (6, 7,29). Two groups(6, 7) have shown that the rosettingis calcium-dependentandappearsto be mediatedalmostentirely by GMP- 140. We found that fixed HL60 cellsor neutrophilswere rosettedby thrombin-activatedplateletsat levelscomparableto thosepreviously reported. In further correspondence,the interactionwascompletely inhibited by the inclusion of EGTA in the rosetting step(Figs. l-3), and rosettingby restingplateletswas greatly diminished(lo- 15%rosetting,not shown). To determinethe role of sialic acid on HL60 cellsin rosettingby activated platelets,we treatedfixed HL60 cellswith broad-spectrumbacterial sialidasesprior to the rosettingstep. Each of the threesialidasestested,including an affinitypurified sialidasefrom Clostridiumperfringens,abolishedthe capacity of HL60 cells to be rosetted (Fig. 1). In the caseof the Arthrobacterureahciens sialidase,inclusionof the substratesialyllactoseduring the enzyme incubation, substantiallypreservedthe rosettingwith HL60 cells, thus demonstratingthat the sialidaseitself, rather than a contaminatingactivity, wasresponsiblefor the diminishedrosetting(data not shown). Sialidasetreatmentof viable HL60 cells resultedin a 6070% reduction in platelet resettingin contrastto the completeinhibition observedwith fixed cells (not shown). This incompleteeffect may reflect the ability of metabolicallyactive cellsto restore critical sialylatedresiduesduring the 20 min incubationstepwith the platelets. To determine whether neutrophilswere alsosusceptibleto sialidasetreatment, fared neutropbilswere treated 1351
Vol.
172, No. 3, 1990
BIOCHEMICAL
AND BIOPHYSICAL RESEARCH COMMUNICATIONS
with the Closttidium perfringens enzyme. As wasthe casefor HL60 cells, sialidasetreatment of neutrophilseliminatedrosetting(Fig. 2). The bacterial sialidasesemployedwere all broad spectrumenzymeswhich can cleavea2,3, a2,6, anda2,8 linked NeuAc from glycoconjugates(30). In an attempt to define a particular linkage of sialic acid that might be important in platelet rosetting,we employedthe Newcastle diseasevirus (NDV) sialidase,which cleavesa2,3 and a2,8 but not a2,6 linked NeuAc (26, 31). As shownin Figure 3, treatmentof HL60 cellswith a high concentrationof NDV sialidasehad no effect on their rosettingby platelets,whereasthe cellswere still susceptibleto the broad spectrum Atihrobacter
ureafaciens
sialidase.To confirm that the NDV sialidasewas,in fact, active upon
cell surfacesialyloligosaccharides of HL60, we testedthe effect of this enzyme on two independent epitopesdefined by the monoclonalantibodies,CSLEX-l(27) and VIM-2 (32). CSLEX- 1 recognizesthe sialylatedLewis X structure: NeuAca2,3Galp1,4[Fucal,3]GlcNAcp1,3GalpwhereasVIM-2 bindsto the heptasaccharide: NeuAca2,3Gal~l,4GlcNAc~l,3Ga1[31,4[Fucal,3]GlcNAc~l,3GalFor both antibodies,binding showsan absolutedependenceon the presenceof a2,3 linked NeuAc (27,28). Neither of thesecarbohydratesepitopesis itself the ligand for GMP- 140,sincethese antibodieshaveno effect on rosettingof HL60 or neutrophilsby activated platelets(data not shown). As shownin Fig. 4, treatment of HL60 cellswith either NDV sialidaseor Arthrobacter ureafaciens
sialidaseunder the sameconditionsemployedfor the rosettingexperiment (Fig. 3)
completelyeliminatedthe epitopesdefinedby both CSLEX- 1 andVIM-2. 80
1
,
0
1
Siatidnses
Treatment
02
ControlSialidax (Cp)
EGl-A
Treatment
Figure 1. Effects of sialidases or EGTA ontheresettingof HL60cellsby platelets.Fixed HL60 cells were treatedwith Arthrohxter ureafaciens (Au) sialidase. Closttidium verfrinens (CD) sialidase, Vibrio cholera (Vc) sialidase or control‘buffer and thek tested for r&.etting%y ac&ted platelets. HL60 cells were also tested for rosetting in the presence of EGTA (10 mM). The means and SEMs were computed on the basis of three independent replicates. Fimre 2. Effects of Clostridium pcrfhgem sialidase or EGTA on the resetting of neutrophils by platelets. Fixed neutrophils were treated with Cbstkiium perftingens (Cp) sialidase or control buffer and then tested for rosetting with activated platelets. Neutrophils were also tested for rosetting in the presence of EGTA (10 mM). The means and SEMs were computed on the basis of three independent replicates. 1352
Vol.
172,
No.
3, 1990
BIOCHEMICAL
COlltd
AND
NDV Sialidase
BIOPHYSICAL
Au Sialidase
RESEARCH
COMMUNICATIONS
EGTA
Treatment
Figure 3. Effects of NDV sialidase and Arthrobacter ureafkiens sialidase on the rosetting of HL60 cells by platelets. Fixed HL60 cells were treated with NDV sialidase or Arthrobacterureaf (Au) sialidase in PBS as described in MATERIALS AND METHODS. A control sample was sham treated with PBS alone. Rosetting of control cells was also determined in the presence of EGTA (10 mM). The means and SEMs were computed on the basis of four independent replicates. The incomplete effect of the Arthrobacter is probably attributable to the higher pH of PBS vs. the PB buffer used for the data in Figure 1.
DISCUSSION Our previouswork hasshowna functional requirementfor sialic acid on the HEV-ligand for gp9WEL, a lymphocyte homing receptor. It is not known in this systemwhether sialic acid contributesdirectly to a recognitiondeterminantof the ligand or modulatesthe conformationor activity of a distal recognition element. However, the known capacity of different forms and linkagesof sialic acid to encodehighly specific recognitionelements(33-35) arguesfor the former possibility. A notablefeatureof the lectin domainof gp9wEL (11) and that of its closely related humanhomologue(22,36,37) is a very large numberof lysine residues,i.e., 16 in the caseof the mousemoleculeand 12in the caseof the humanmolecule. This high concentrationof positive chargeswithin the lectin domainis compatiblewith the sialicacid requirementfor the HEV-ligand aswell aswith the finding that all of the sugarsknown to mimic the ligand are anionic (2 1). In contrast,other receptorscontainingthe C-type lectin domain(e.g., chicken, rat, and humanhepatic lectins, mannose-bindingreceptor, etc.) contain an averageof 5 lysine residuesand, insofar asis known, recognize only neutral sugars(38). It is, thus, notablethat the other two membersof the LEC-CAM family alsoexhibit a large numberof lysine residuesin their lectin domains(i.e., 14 for GMP- 140and 10 for ELAM- I), suggestingthat their cognateligandsmight be anionic, a prediction which is supportedfor GMP- 140by the finding that it canbind to severalanionic polysaccharides(39). The high degreeof sequencehomology amongthe lectin domainsof the LEC-CAMS (60-700/o)lendsfurther supportto the possibility of relatedsugar-bindingspecificities. In the presentstudy, we provide direct evidencefor a similarity by showingthat sialidasetreatment of HL60 cellsand neutrophilspreventsrosettingby activated platelets,an intelaction attributed to 1353
Vol.
172, No. 3, 1990
BIOCHEMICAL
AND BIOPHYSICAL RESEARCH COMMUNICATIONS
VIM-2
CSLEX-1
D
RelativeFluorescence Figure 4. Effects of NDV sialidase and Arthrobacter ureafkiens sialidase on the expression of the epitopes for CSLEX-1 and VIM-Z. Fixed HL60 cells were treated with NDV sialidase, Admbacter urediciens sialidase or PBS alone under the identical conditions used in Figure 3. The binding of CSLEX- 1 and VIM-2 mAbs was determined by indirect bnmunofluorescence using flow cytometry. The y-axis indicates the relative cell frequencies and the x-axis indicates relative fluorescence intensity. A) no primary antibody; B) control-treated cells; C) cells treated with Arthrohcter ureafaciens (Au) sialidase; and D) cells treated with NDV sialidase. The modal fluorescence values were: no primary antibody (3); CSLEX-control(328); CSLEX-Au (5); CSLEX-NDV (7); VIM-2 control (215); VIM-2 Au (3); VIM-2 NDV (6).
GMP- 140. As three different broadspectrumsialidasesproducethis effect (on HL.60 cells) and competitive inhibition by sialyl-lactoseblocksthe inactivation, a sialicacid requirementis clearly indicated. The useof the linkage-specificsialidaseof NDV hasprovided information about the 13.54
Vol.
172,
No.
3, 1990
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
possible nature of the linkage of the essential sialic acid on HL60 cells. Under conditions of enzyme treatment that completely eliminated two epitopes dependent on a2,3 linked NeuAc (CSLEX- 1 and VIM-2), there was no detectable effect on the ligand activity for GMP- 140. NeuAca2,3Gal on isolated oligosaccharides or on cell surface associated glycoconjugates is completely susceptible to hydrolyis by NDV sialidase, whereas the NeuAca2,6Gal and NeuAca2,6GalNAc
linkages are completely resistant (26). From the activity of NDV sialidase on
the NeuAca28NeuAc linkage in a milk oligosaccharide, it is strongly suspected that the enzyme would also cleave a2,8 linked NeuAc at the cell surface (26). If it is assumed that the ligand for GMP- 140 does not involve an unusual form or linkage of sialic acid with resistance to NDV sialidase (e.g., 4-0-acetyl NeuAc [30]), our results argue for the existence of a critical a2,6 linked NeuAc. Further biochemical chamcterization of the ligand will allow evaluation of this prediction. A question of considerable additional interest is whether the sialic acid requirement extends to ELAM- 1 and thus, is general for all members of the LEC-CAM
family. If so, the ligands for
ELAM- 1 and GMP- 140 would be predicted to involve distinct sialyloligosaccharides,
since
functional evidence suggests that the ligands are independent and nonoverlapping (40). The results of this study may shed light on the interaction between certain tumor cells and platelets. It has long been known that some tumor cells induce platelets to aggregate (4 1). Interest in this phenomenon has been considerable, since the formation of tumor-platelet emboli in the blood may be an important determinant of metastatic arrest and growth of the tumor cells at secondary sites (41,42).
For certain tumor cells, a correlation exists among the degree of cell
surface sialylation, their ability (or that of shed factors derived from them) to aggregate platelets, and their metastatic potential in animals (42,43).
Recently, it has been shown that sialic acid
moieties on the surface of a metastatic colon adenocarcinoma cell line (mouse) are directly involved in platelet co-aggregation (44). Since, as reported here, GMP- 140 is likely to mediate plateletneutrophil binding through recognition of a sialylated structure, the participation of GMP- 140 in tumor cell-induced platelet aggregation warrants investigation. AcknowledgmentsWe thank Drs. David Brown and Israel Charo of COR Therapeutics for their advice on the isolation of platelets. We thank Dr. Walter Knapp (Institute of Immunology, University of Vienna, Vienna) and Dr. Paul Terasaki (Dept. of Surgery, UCLA School of Medicine, Los Angeles) for their gifts of VIM-2 and CSLEX- 1, respectively. We are grateful to Dr. David True and Supama Bhattacharya for culturing of NDV. The research was supported by NIH Grants: MAC P60 AR20684 and GM23547 to SDR and CA32826 to BAM.
REFERENCES 1. 2. 3. 4. 5. 6. 7.
McEver, R. P., and Martin, M. N. (1984) J. Biol. Chem. 259, 9799-9804. Stenberg, P. E., McEver, R. P., Shuman, M. A., Jacques, Y. V., and Bainton, D. F. (1985) J. Cell Biol. 101, 880-886. Berman, C. L., Yeo, E. L., Wencel-Drake, J. D., Furie, B. C., Ginsberg, M. H., and Furie, B. (1986) J. Clin. Invest. 78, 130-137. McEver, R. P., Be&stead, J. H., Moore, K. L., Marshall-Carlson, L., and Bainton, D. F. (1989) J. Clin. Invest. 84, 92-99. Bonfanti, R., Furie, B. C., Furie, B., and Wagner, D. D. (1989) Blood 73, 1109-l 112. Larsen, E., Celi, A., Gilbert, G. E., Furie, B. C., Erban, J. K., Bonfanti, R., Wagner, D., and Furie, B. (1989) Cell 59, 305-312. Hamburger, S. A., and McEver, R. P. (1990) Blood 75, 550-554. 1355
Vol.
8. 9. 10. 11. 12. 13. 14. 1.5. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. :;: 33. 34. 2: 37. 38. 39. 40. 41. 1:: 44.
172,
No.
3, 1990
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
Geng, J. G., Bevilacqua, M. P., Moore, K. L., McIntyre, T. M., Prescott, S. M., Kim, J. M., Bliss, G. A., Zimmerman, G. A., and McEver, R. P. (1990) Nature 343, 757-760. Johnston,G. I., Cook, R. G., and McEver, R. P. (1989) Cell 56, 1033-1044. Bevilacqua, M. P., Spengeling, S., Gimbrone, M. A., Jr., and Seed, B. (1989) Science 243, 1160-l 165. Lanky, L. A., Singer, M. S., Yednock, T. A,, Dowbenko, D., Fennie, C., Rodriguez, H., Nguyen, T., Stachel, S., and Rosen,S. D. (1989) Cell 56, 1045-1055. Siegelman,M. H., Van, de, Rijn, M, and Weissman,I. L. (1989) Science243, 1165 1172. Bevilacqua, M. P., Pober, J. S., Mendrick, D. L., Cotran, R. S., and Gimbrone, M. A. J. (1987) Proc. Natl. Acad. Sci. USA 84, 9238-9242. Gallatin, W. M., Weissman,I. L., and Butcher, E. C. (1983) Nature 303, 30-34. Geoffroy, J. S., and Rosen, S. D. (1989) J. Cell Biol. 109, 2463-2469. Watson, S. R., Imai, Y., Fennie, C., Geoffroy, J. S., Rosen, S. D., and Lasky, L. A. (1990) J. Cell Biol. 110, 2221-2229. Stoolman, L. M. (1989) Cell 56, 907-910. Stoolman, L. M., Tenforde, T. S., and Rosen,S. D. (1984) J. Cell Biol. 99, 15351540. Yednock, T. A., Stoolman, L. M., and Rosen,S. D. (1987) J. Cell Biol. 104, 713-723. Yednock, T. A., Butcher, E. C., Stoolman, L. M., and Rosen,S. D. (1987) J. Cell Biol. 104, 725-731. Imai, Y., True, D. D., Singer, M. S., and Rosen,S. D. (1990) J. Cell Biol. 111, 12251232. Bowen, B. R., Nguyen, T., and Lasky, L. A. (1989) J. Cell Biol. 109, 421-427. Rosen, S. D., Singer, M. S., Yednock, T. A., and Stoolman, L. M. (1985) Science228, 1005-1007. Rosen,S. D., Chi, S.-I., True, D. D., Singer, M. S., and Yednock, T. A. (1989) J. Immunol. 142, 1895-1902. True, D. D., Singer, M. S., Lasky, L. A., and Rosen,S. D. J. Cell Biol., in press Paulson,J., Weinstein, J., Dorland, L., van Halbeek, H., and Vliegenthart, J. F. G.( 1982) J. Biol. Chem. 257, 12734-12738. Fukushima, K., Hirota, M., Terasaki, P. I., Wakisaka, A., Togashi, H., Chia, D., Suyama, N., Fukushi, Y., Nudelman,E., and Hakomori, S.-I. (1984) Can. Res. 44, 5279-5285. Majdic, O., Bettelheim, P., Stockinger, H., Aberer, W., Liszka, K., Lutz, D., and Knapp, W. (1984) Int. J. Can. 33, 617-623. Jungi, T. W., Spycher, M. O., Nydegger, U. E., and Barandun, S. (1986) Blood 67, 629636. Schauer,R. (1982) Adv. Carb. Chem. Biochem. 40, 131-233. Dtzeniek, R. (1973) Histochem. J. 5, 27 l-290. Macher, B. A., Buehler, J., Scudder, P., Knapp, W., and Feizi, T. (1988) J. Biol. Chem. 263, 10186-10191. Schauer,R. (1985) TrendsBiochem. Sci. 10,357-360. Sharon, N., and Lis, H. (1989) Science246, 227-234. Wiley, D. C., and Skehel, J. J. (1987) Ann. Rev. Biochem. 56, 365-394. Siegelman,M. H., and Weissman,I. L. (1989) Proc. Natl. Acad. Sci. USA 86,55625566. Tedder, T. F., Isaacs,C. M., Ernst, T. J., Demetri, G. D., Adler, D. A., and Disteche, C. M. (1989) J. Exp. Med. 170, 123-133. Drickamer, K. (1988) J. Biol. Chem. 263, 9557-9560. Skinner, M. P., Foumier, D. J., Andrews, R. K., Gorman, J. J., Chesterman,C. N., and Bemdt, M. C. (1989) Biochem. Biophys. Res.Commun. 164, 1373-1379. Gamble, J. R., Skinner, M. P., Bemdt, M. C., and Vadas, M. A. (1990) Science249, 414-417. Gasic, G. J., Gasic, T. B., Galanti, N., Johnson,T., and Murphy, S. (1973) Int. J. Cancer 11, 704-718. Karpatkin, S., and Pearlstein,E. (1981) Ann. Intern. Med. 95,636-641. Pearlstein,E., Salk, P. L., Yogeeswaran,G., and Karpatkin, S. (1980) Proc. Natl. Acad. Sci. USA 77, 4336-4339. Kijima-Suda, I., Miyazawa, T., Itoh, M., Toyoshima, S., and Osawa, T. (1988) Can.Res. 48, 3728-3732.
1356
172,
November
No.
3, 1990
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
Pages
15, 1990
1349-1356
REQUIREMENT FOR SIALIC ACID ON NEUTROPHILS IN A GMP- 140(PADGEM) MEDIATED ADHESIVE INTERACTION WITH ACTIVATED PLATELETS Laura Corral, Mark S. Singer, *Bruce A. Macher, and Steven D. Rosen Departmentof Anatomy andProgramin Immunology, University of California, SanFrancisco,CA 94143-0452 *Departmentsof Chemistry and Biochemistry, SanFranciscoState University, SanFrancisco,CA 94132 Received
October
4,
1990
Summarv. Platelet GMP- 140,alongwith ELAM- 1 andgp9OmL, comprise the l.,EC-C&l family of cell-cell adhesionproteins. The three proteinsdemonstratea highly relateddomainorganization, which includesan extracellular calcium-typelectin motif. gp90MEL,a lymphocyte homing receptor, mediateslymphocyte attachmentto high endothelialvenulesof lymph nodesthrough recognition of a sialylatedligand on the endothelialcells. The rosettingof neutrophilsor promyelocytic HL60 cellsby activated plateletsis mediatedby GMP- 140on the platelets. We showherethat treatmentof neutrophilsor HL60 cellswith 3 broad spectrumsialidasescompletely prevents rosetting. However, the Newcastlediseasevirus sialidase,an enzyme specific for a2,3 and a28 linkagesof sialic acid doesnot affect rosettingof HL60 cells. Theseresultsindicate that the ligand for GMP- 140requiressialicacid and suggestthat an a2,6 linkage may be critical, ‘31990Academic Press.IIIC. Granular-membraneprotein (GMP- 140), alsoknown asPADGEM or CD62, is a glycoprotein of 140,000daltonsfound in the alphagranulesof plateletsand the Weibel-Palade bodiesof endothelialcells(l-5). Upon stimulationof thesecellsby agonistssuchasthrombin, degmnulationleadsto the rapid externalization of this membrane-bound protein to the plasma membrane.Recentstudieshave establishedthat GMP-140 mediatesthe calcium-dependent adhesionof neutrophilsor monocytesto activated platelets(6, 7) aswell asthe early binding of neutrophilsto histamine-or thrombin-stimulatedendothelium(8). Possiblefunctional consequences of theseadhesiveeventsincludethe removal of activated plateletsfrom the blood by neutrophils(or monocytes)and the rapid adherenceof neutrophilsto endotheliumat sitesof acute inflammation. The molecularcloning of GMP- 140hasrevealeda modularprotein consistingof an aminoterminal lectin-like domain(calcium-dependent),an EGF motif, 9 tandemconsensus repeats relatedto thosefound in complement-bindingproteins,a transmembranedomain,anda short cytosolic tail (9). Two other proteins, ELAM- 1andgp90MEL, sharea high degreeof sequence homology with GMP-140 and the samebasicdomainstructure,but are distinctive in the numberof consensus repeats(lo- 12). ELAM- 1 is a cytokine-inducible moleculeon endothelialcellswhich Abbreviations: Au, Arthrobacer ureafaciens ; C-type, calcium-type; Cp, Cfostridiumperfringens HEV, high endothelialvenule; NeuAc, 5 N-acetylneumminicacid; NDV, Newcastlediseasevirus; Vc, Vibrio choler .
Vol.
172,
No.
3, 1990
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
mediates neutrophil attachment (10, 13). 1~ viva, this molecule may promote neutrophil adherence to endothelium during the later stages of certain acute inflammatory reactions. gp90MEL, a constitutive cell surface protein on murine lymphocytes identified by the MEL- 14 mAb, is responsible for the organ-selective adherence (“homing”) of lymphocytes to the endothelial cells of high endothelial venules (HEV) in lymph nodes, an adhesive event that initiates extravasation during the process of lymphocyte recirculation (14- 16). These three proteins, each of which is involved in a highly selective cellular interaction within the blood vascular compartment, comprise a new family of cell-cell adhesion molecules designated as the LEC-CAMS (17) or selectins (8). In the case of gp9OMEL, the lectin domain has been directly implicated in the adhesive function of the receptor. Specific carbohydrates have been identified that interact with this domain in a calcium-dependent manner and competitively inhibit lymphocyte attachment to HEV ( 18,22). Furthermore, the HEV-ligand for gp’)@ELdemonstrates
a functional requirement for sialic acid as
follows: 1) lymphocyte attachment to HEV is blocked by pretreatment of the HEV with sialidases or with the sialic acid specific lectin, Limaxflavus agglutinin (23-25); 2) the binding of a recombinant form of gp90MELto HEV in tissue sections is abrogated by exposure of the HEV to either sialidases or Limax agglutinin (25); and 3) desialylation of the isolated HEV-ligand for gp90MELprevents
its binding to gp90 MEL (Imai, Singer, Fennie, Lasky, and Rosen, submitted).
In view of the strong structural similarities between gp90 MELand the other members of the LECCAM family, we investigated whether a sialic acid requirement also exists for the binding of GMPI40 to its ligand. MATERIALS
AND METHODS
Cell Culture and Cell Preparation. The HL60 human promyelocytic cell line was cultured in RPMI- 1640, supplemented with 10% fetal bovine serum, 100 U/ml penicillin and 100 ug/ml streptomycin. All culture reagents were obtained from the UCSF Cell Culture Facility. The cells were washed 3 times by centrifugation (150 x g) and resuspended in Dulbecco’s PBS at a concentration of 3 x 107 cells/ml. The cells were fixed by the addition of an equal volume of 1% parafotmaldehyde (EM Sciences, Cherry Hill, NJ) in 0.1 M sodium cacodylate, pH 7.3. After 1hr at 40 C, the cells were washed 3 times and suspended in PB buffer (0.0 1M Na,H;P04,0.1 SM NaCl, 5mM CaC12, pH 6.5) or PBS. Polymorphonuclear neutrophils were isolated from fresh heparinized human blood by density gradient centrifugation (Mono-Poly Resolving Media, Flow Laboratories, McLean, VA) following the manufacturer’s instructions. The neutrophils were fixed according to the procedure of Hamburger and McEver (7) and washed 3 times. Enzyme Treatment. Fixed HL60 cells (150 ul at 3 x 107 cells/ml in PB) were treated with either Arthrobacter ureafaciens neuraminidase (Calbiochem, Inc., San Diego, CA) or Clostridium perftingens neuraminidase (Type X, affinity purified, Sigma Chemical Co., St. Louis, MO) at a final concentration of 0.1 U/ml (international units) or with Vibrio cholera neuraminidase (Gibco, Grand Island, NY) at a final concentration of 250 U/ml (according to the manufacturer’s definition of a unit) and incubated at 37O C for 1 hr. Fixed neutrophils were treated with C. perfringens neuraminidase under the same conditions used for HL60 cells. In some experiments, HL60 cells were exposed to Arthrobacterureafaciens neumminidase (0.1 U/ml final ) in PBS instead of PB. Newcastle disease virus (NDV) sialidase was prepared and assayed as described by Paulson et al. (26). Fixed HL60 cells at 4 x 107/ml in PBS were mixed with NDV sialidase (0.2 U/ml final) and incubated at 370 C for 1 hr. After exposure to an enzyme, the cells were washed 3 times with PBS before use in the platelet binding assay. Control samples were treated in parallel without exposure to enzymes. Platelet isolation. Blood from normal human donors was anticoagulated with ACD (85 mM sodium citrate, 7 1 mM citric acid, 111 mM D-glucose) at a 6: 1 (v/v) mtio. After centrifugation at 1350
Vol.
172, No. 3, 1990
BIOCHEMICAL
AND BIOPHYSICAL
RESEARCH COMMUNICATIONS
200 x g for 20 min at 22OC, the platelet-rich plasmawaspooledand fractionated by gel filtration at 220C with SepharoseCL-2B (Sigma ChemicalCo.), equilibratedwith HEPES buffer (5.5 mM D-glucose, 137mM NaHC03,2.68 mM KCl, 20.0 mM HEPES, 0.35% humanalbumin), pH 7.4. Plateletswere stimulatedwith humanthrombin (T-6759, SigmaChemicalCo.) at a final concentrationof 0.15 U/ml for a minimum of 10 min and were usedin the adhesionassayswithin 30 min of activation. In somecases,the isolatedplateletswere already activated asdetermined morphologically and were usedwithout thrombin treatment. Similar resultswere obtainedwith thrombin-activatedplateletsand thoseactivated during isolation. Platelet Rosetting Assay. The procedurefor platelet rosettingwasslightly modified from the protocol of Larsenet al. (6). For eachsample,50 ul of cells (3 x 107cells/ml in PBS) were mixed with an equalvolume of activatedplatelets(2 x IO* cells/ml) in a microfugetube and allowedto react for 20 min at 220 C without agitation. Each treatmentwascarried out in 3-4 replicates. The cell suspensions were fixed by the addition of glutaraldehydeto a final concentrationof 1.25%. Platelet rosettingof cellswas determinedby Hoffman modulationcontrastmicroscopy. Cellswith two or more adherentplateletswere scoredasrosetted. Samples(l-2) of cells (SO-loo) were evaluatedfrom eachreplicatetube. Resultsare indicatedaspercent rosettedcellsbasedon the meansof the 3-4 replicates. SEMs denotethe statisticalvariability. The effect of calcium chelation was determinedby exposingactivated plateletsto 10mM EGTA for 10min prior to mixing with the HL60 cellsor neutrophils Flow Cytometry Analysis. Fixed HL60 cells preparedasabove (3.3 x 107/ml) were exposedto NDV sialidase(0.2 U/ml) in PBS, Arthrobacter ureafaciensneuraminidase(0.1 U/ml final) in PBS, or to PBS aloneand incubatedfor 1 hr at 37OC. The cellswere washed3 timesin PBS containing 1% fetal bovine serum(FBS). The washedcellswere exposedto CSLEX- 1 mAb (27) or to VIM-2 mAb (28), eachdiluted asan ascitesfluid l/50 in PBS-l% FBS. Incubation was for 30 min at 4OC. The cellswere sequentiallyexposedto to biotinylated-horseanti-mouseIgG (Vector Laboratories,Burlingame,CA) and FITC-extrAvidin (SigmaChemicalCo.) accordingto the manufacturers’recommendations.Fluorescencehistogramprofiles were determinedwith a Becton-DickinsonFACScan II flow cytometer. The backgroundlevel of fluorescencewas defined with no primary antibody added, RESULTS A well-describedphenomenonis the binding of activatedplateletsto neutrophilsor the HL60 promyelocytic cell line (6, 7,29). Two groups(6, 7) have shown that the rosettingis calcium-dependentandappearsto be mediatedalmostentirely by GMP- 140. We found that fixed HL60 cellsor neutrophilswere rosettedby thrombin-activatedplateletsat levelscomparableto thosepreviously reported. In further correspondence,the interactionwascompletely inhibited by the inclusion of EGTA in the rosetting step(Figs. l-3), and rosettingby restingplateletswas greatly diminished(lo- 15%rosetting,not shown). To determinethe role of sialic acid on HL60 cellsin rosettingby activated platelets,we treatedfixed HL60 cellswith broad-spectrumbacterial sialidasesprior to the rosettingstep. Each of the threesialidasestested,including an affinitypurified sialidasefrom Clostridiumperfringens,abolishedthe capacity of HL60 cells to be rosetted (Fig. 1). In the caseof the Arthrobacterureahciens sialidase,inclusionof the substratesialyllactoseduring the enzyme incubation, substantiallypreservedthe rosettingwith HL60 cells, thus demonstratingthat the sialidaseitself, rather than a contaminatingactivity, wasresponsiblefor the diminishedrosetting(data not shown). Sialidasetreatmentof viable HL60 cells resultedin a 6070% reduction in platelet resettingin contrastto the completeinhibition observedwith fixed cells (not shown). This incompleteeffect may reflect the ability of metabolicallyactive cellsto restore critical sialylatedresiduesduring the 20 min incubationstepwith the platelets. To determine whether neutrophilswere alsosusceptibleto sialidasetreatment, fared neutropbilswere treated 1351
Vol.
172, No. 3, 1990
BIOCHEMICAL
AND BIOPHYSICAL RESEARCH COMMUNICATIONS
with the Closttidium perfringens enzyme. As wasthe casefor HL60 cells, sialidasetreatment of neutrophilseliminatedrosetting(Fig. 2). The bacterial sialidasesemployedwere all broad spectrumenzymeswhich can cleavea2,3, a2,6, anda2,8 linked NeuAc from glycoconjugates(30). In an attempt to define a particular linkage of sialic acid that might be important in platelet rosetting,we employedthe Newcastle diseasevirus (NDV) sialidase,which cleavesa2,3 and a2,8 but not a2,6 linked NeuAc (26, 31). As shownin Figure 3, treatmentof HL60 cellswith a high concentrationof NDV sialidasehad no effect on their rosettingby platelets,whereasthe cellswere still susceptibleto the broad spectrum Atihrobacter
ureafaciens
sialidase.To confirm that the NDV sialidasewas,in fact, active upon
cell surfacesialyloligosaccharides of HL60, we testedthe effect of this enzyme on two independent epitopesdefined by the monoclonalantibodies,CSLEX-l(27) and VIM-2 (32). CSLEX- 1 recognizesthe sialylatedLewis X structure: NeuAca2,3Galp1,4[Fucal,3]GlcNAcp1,3GalpwhereasVIM-2 bindsto the heptasaccharide: NeuAca2,3Gal~l,4GlcNAc~l,3Ga1[31,4[Fucal,3]GlcNAc~l,3GalFor both antibodies,binding showsan absolutedependenceon the presenceof a2,3 linked NeuAc (27,28). Neither of thesecarbohydratesepitopesis itself the ligand for GMP- 140,sincethese antibodieshaveno effect on rosettingof HL60 or neutrophilsby activated platelets(data not shown). As shownin Fig. 4, treatment of HL60 cellswith either NDV sialidaseor Arthrobacter ureafaciens
sialidaseunder the sameconditionsemployedfor the rosettingexperiment (Fig. 3)
completelyeliminatedthe epitopesdefinedby both CSLEX- 1 andVIM-2. 80
1
,
0
1
Siatidnses
Treatment
02
ControlSialidax (Cp)
EGl-A
Treatment
Figure 1. Effects of sialidases or EGTA ontheresettingof HL60cellsby platelets.Fixed HL60 cells were treatedwith Arthrohxter ureafaciens (Au) sialidase. Closttidium verfrinens (CD) sialidase, Vibrio cholera (Vc) sialidase or control‘buffer and thek tested for r&.etting%y ac&ted platelets. HL60 cells were also tested for rosetting in the presence of EGTA (10 mM). The means and SEMs were computed on the basis of three independent replicates. Fimre 2. Effects of Clostridium pcrfhgem sialidase or EGTA on the resetting of neutrophils by platelets. Fixed neutrophils were treated with Cbstkiium perftingens (Cp) sialidase or control buffer and then tested for rosetting with activated platelets. Neutrophils were also tested for rosetting in the presence of EGTA (10 mM). The means and SEMs were computed on the basis of three independent replicates. 1352
Vol.
172,
No.
3, 1990
BIOCHEMICAL
COlltd
AND
NDV Sialidase
BIOPHYSICAL
Au Sialidase
RESEARCH
COMMUNICATIONS
EGTA
Treatment
Figure 3. Effects of NDV sialidase and Arthrobacter ureafkiens sialidase on the rosetting of HL60 cells by platelets. Fixed HL60 cells were treated with NDV sialidase or Arthrobacterureaf (Au) sialidase in PBS as described in MATERIALS AND METHODS. A control sample was sham treated with PBS alone. Rosetting of control cells was also determined in the presence of EGTA (10 mM). The means and SEMs were computed on the basis of four independent replicates. The incomplete effect of the Arthrobacter is probably attributable to the higher pH of PBS vs. the PB buffer used for the data in Figure 1.
DISCUSSION Our previouswork hasshowna functional requirementfor sialic acid on the HEV-ligand for gp9WEL, a lymphocyte homing receptor. It is not known in this systemwhether sialic acid contributesdirectly to a recognitiondeterminantof the ligand or modulatesthe conformationor activity of a distal recognition element. However, the known capacity of different forms and linkagesof sialic acid to encodehighly specific recognitionelements(33-35) arguesfor the former possibility. A notablefeatureof the lectin domainof gp9wEL (11) and that of its closely related humanhomologue(22,36,37) is a very large numberof lysine residues,i.e., 16 in the caseof the mousemoleculeand 12in the caseof the humanmolecule. This high concentrationof positive chargeswithin the lectin domainis compatiblewith the sialicacid requirementfor the HEV-ligand aswell aswith the finding that all of the sugarsknown to mimic the ligand are anionic (2 1). In contrast,other receptorscontainingthe C-type lectin domain(e.g., chicken, rat, and humanhepatic lectins, mannose-bindingreceptor, etc.) contain an averageof 5 lysine residuesand, insofar asis known, recognize only neutral sugars(38). It is, thus, notablethat the other two membersof the LEC-CAM family alsoexhibit a large numberof lysine residuesin their lectin domains(i.e., 14 for GMP- 140and 10 for ELAM- I), suggestingthat their cognateligandsmight be anionic, a prediction which is supportedfor GMP- 140by the finding that it canbind to severalanionic polysaccharides(39). The high degreeof sequencehomology amongthe lectin domainsof the LEC-CAMS (60-700/o)lendsfurther supportto the possibility of relatedsugar-bindingspecificities. In the presentstudy, we provide direct evidencefor a similarity by showingthat sialidasetreatment of HL60 cellsand neutrophilspreventsrosettingby activated platelets,an intelaction attributed to 1353
Vol.
172, No. 3, 1990
BIOCHEMICAL
AND BIOPHYSICAL RESEARCH COMMUNICATIONS
VIM-2
CSLEX-1
D
RelativeFluorescence Figure 4. Effects of NDV sialidase and Arthrobacter ureafkiens sialidase on the expression of the epitopes for CSLEX-1 and VIM-Z. Fixed HL60 cells were treated with NDV sialidase, Admbacter urediciens sialidase or PBS alone under the identical conditions used in Figure 3. The binding of CSLEX- 1 and VIM-2 mAbs was determined by indirect bnmunofluorescence using flow cytometry. The y-axis indicates the relative cell frequencies and the x-axis indicates relative fluorescence intensity. A) no primary antibody; B) control-treated cells; C) cells treated with Arthrohcter ureafaciens (Au) sialidase; and D) cells treated with NDV sialidase. The modal fluorescence values were: no primary antibody (3); CSLEX-control(328); CSLEX-Au (5); CSLEX-NDV (7); VIM-2 control (215); VIM-2 Au (3); VIM-2 NDV (6).
GMP- 140. As three different broadspectrumsialidasesproducethis effect (on HL.60 cells) and competitive inhibition by sialyl-lactoseblocksthe inactivation, a sialicacid requirementis clearly indicated. The useof the linkage-specificsialidaseof NDV hasprovided information about the 13.54
Vol.
172,
No.
3, 1990
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
possible nature of the linkage of the essential sialic acid on HL60 cells. Under conditions of enzyme treatment that completely eliminated two epitopes dependent on a2,3 linked NeuAc (CSLEX- 1 and VIM-2), there was no detectable effect on the ligand activity for GMP- 140. NeuAca2,3Gal on isolated oligosaccharides or on cell surface associated glycoconjugates is completely susceptible to hydrolyis by NDV sialidase, whereas the NeuAca2,6Gal and NeuAca2,6GalNAc
linkages are completely resistant (26). From the activity of NDV sialidase on
the NeuAca28NeuAc linkage in a milk oligosaccharide, it is strongly suspected that the enzyme would also cleave a2,8 linked NeuAc at the cell surface (26). If it is assumed that the ligand for GMP- 140 does not involve an unusual form or linkage of sialic acid with resistance to NDV sialidase (e.g., 4-0-acetyl NeuAc [30]), our results argue for the existence of a critical a2,6 linked NeuAc. Further biochemical chamcterization of the ligand will allow evaluation of this prediction. A question of considerable additional interest is whether the sialic acid requirement extends to ELAM- 1 and thus, is general for all members of the LEC-CAM
family. If so, the ligands for
ELAM- 1 and GMP- 140 would be predicted to involve distinct sialyloligosaccharides,
since
functional evidence suggests that the ligands are independent and nonoverlapping (40). The results of this study may shed light on the interaction between certain tumor cells and platelets. It has long been known that some tumor cells induce platelets to aggregate (4 1). Interest in this phenomenon has been considerable, since the formation of tumor-platelet emboli in the blood may be an important determinant of metastatic arrest and growth of the tumor cells at secondary sites (41,42).
For certain tumor cells, a correlation exists among the degree of cell
surface sialylation, their ability (or that of shed factors derived from them) to aggregate platelets, and their metastatic potential in animals (42,43).
Recently, it has been shown that sialic acid
moieties on the surface of a metastatic colon adenocarcinoma cell line (mouse) are directly involved in platelet co-aggregation (44). Since, as reported here, GMP- 140 is likely to mediate plateletneutrophil binding through recognition of a sialylated structure, the participation of GMP- 140 in tumor cell-induced platelet aggregation warrants investigation. AcknowledgmentsWe thank Drs. David Brown and Israel Charo of COR Therapeutics for their advice on the isolation of platelets. We thank Dr. Walter Knapp (Institute of Immunology, University of Vienna, Vienna) and Dr. Paul Terasaki (Dept. of Surgery, UCLA School of Medicine, Los Angeles) for their gifts of VIM-2 and CSLEX- 1, respectively. We are grateful to Dr. David True and Supama Bhattacharya for culturing of NDV. The research was supported by NIH Grants: MAC P60 AR20684 and GM23547 to SDR and CA32826 to BAM.
REFERENCES 1. 2. 3. 4. 5. 6. 7.
McEver, R. P., and Martin, M. N. (1984) J. Biol. Chem. 259, 9799-9804. Stenberg, P. E., McEver, R. P., Shuman, M. A., Jacques, Y. V., and Bainton, D. F. (1985) J. Cell Biol. 101, 880-886. Berman, C. L., Yeo, E. L., Wencel-Drake, J. D., Furie, B. C., Ginsberg, M. H., and Furie, B. (1986) J. Clin. Invest. 78, 130-137. McEver, R. P., Be&stead, J. H., Moore, K. L., Marshall-Carlson, L., and Bainton, D. F. (1989) J. Clin. Invest. 84, 92-99. Bonfanti, R., Furie, B. C., Furie, B., and Wagner, D. D. (1989) Blood 73, 1109-l 112. Larsen, E., Celi, A., Gilbert, G. E., Furie, B. C., Erban, J. K., Bonfanti, R., Wagner, D., and Furie, B. (1989) Cell 59, 305-312. Hamburger, S. A., and McEver, R. P. (1990) Blood 75, 550-554. 1355
Vol.
8. 9. 10. 11. 12. 13. 14. 1.5. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. :;: 33. 34. 2: 37. 38. 39. 40. 41. 1:: 44.
172,
No.
3, 1990
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
Geng, J. G., Bevilacqua, M. P., Moore, K. L., McIntyre, T. M., Prescott, S. M., Kim, J. M., Bliss, G. A., Zimmerman, G. A., and McEver, R. P. (1990) Nature 343, 757-760. Johnston,G. I., Cook, R. G., and McEver, R. P. (1989) Cell 56, 1033-1044. Bevilacqua, M. P., Spengeling, S., Gimbrone, M. A., Jr., and Seed, B. (1989) Science 243, 1160-l 165. Lanky, L. A., Singer, M. S., Yednock, T. A,, Dowbenko, D., Fennie, C., Rodriguez, H., Nguyen, T., Stachel, S., and Rosen,S. D. (1989) Cell 56, 1045-1055. Siegelman,M. H., Van, de, Rijn, M, and Weissman,I. L. (1989) Science243, 1165 1172. Bevilacqua, M. P., Pober, J. S., Mendrick, D. L., Cotran, R. S., and Gimbrone, M. A. J. (1987) Proc. Natl. Acad. Sci. USA 84, 9238-9242. Gallatin, W. M., Weissman,I. L., and Butcher, E. C. (1983) Nature 303, 30-34. Geoffroy, J. S., and Rosen, S. D. (1989) J. Cell Biol. 109, 2463-2469. Watson, S. R., Imai, Y., Fennie, C., Geoffroy, J. S., Rosen, S. D., and Lasky, L. A. (1990) J. Cell Biol. 110, 2221-2229. Stoolman, L. M. (1989) Cell 56, 907-910. Stoolman, L. M., Tenforde, T. S., and Rosen,S. D. (1984) J. Cell Biol. 99, 15351540. Yednock, T. A., Stoolman, L. M., and Rosen,S. D. (1987) J. Cell Biol. 104, 713-723. Yednock, T. A., Butcher, E. C., Stoolman, L. M., and Rosen,S. D. (1987) J. Cell Biol. 104, 725-731. Imai, Y., True, D. D., Singer, M. S., and Rosen,S. D. (1990) J. Cell Biol. 111, 12251232. Bowen, B. R., Nguyen, T., and Lasky, L. A. (1989) J. Cell Biol. 109, 421-427. Rosen, S. D., Singer, M. S., Yednock, T. A., and Stoolman, L. M. (1985) Science228, 1005-1007. Rosen,S. D., Chi, S.-I., True, D. D., Singer, M. S., and Yednock, T. A. (1989) J. Immunol. 142, 1895-1902. True, D. D., Singer, M. S., Lasky, L. A., and Rosen,S. D. J. Cell Biol., in press Paulson,J., Weinstein, J., Dorland, L., van Halbeek, H., and Vliegenthart, J. F. G.( 1982) J. Biol. Chem. 257, 12734-12738. Fukushima, K., Hirota, M., Terasaki, P. I., Wakisaka, A., Togashi, H., Chia, D., Suyama, N., Fukushi, Y., Nudelman,E., and Hakomori, S.-I. (1984) Can. Res. 44, 5279-5285. Majdic, O., Bettelheim, P., Stockinger, H., Aberer, W., Liszka, K., Lutz, D., and Knapp, W. (1984) Int. J. Can. 33, 617-623. Jungi, T. W., Spycher, M. O., Nydegger, U. E., and Barandun, S. (1986) Blood 67, 629636. Schauer,R. (1982) Adv. Carb. Chem. Biochem. 40, 131-233. Dtzeniek, R. (1973) Histochem. J. 5, 27 l-290. Macher, B. A., Buehler, J., Scudder, P., Knapp, W., and Feizi, T. (1988) J. Biol. Chem. 263, 10186-10191. Schauer,R. (1985) TrendsBiochem. Sci. 10,357-360. Sharon, N., and Lis, H. (1989) Science246, 227-234. Wiley, D. C., and Skehel, J. J. (1987) Ann. Rev. Biochem. 56, 365-394. Siegelman,M. H., and Weissman,I. L. (1989) Proc. Natl. Acad. Sci. USA 86,55625566. Tedder, T. F., Isaacs,C. M., Ernst, T. J., Demetri, G. D., Adler, D. A., and Disteche, C. M. (1989) J. Exp. Med. 170, 123-133. Drickamer, K. (1988) J. Biol. Chem. 263, 9557-9560. Skinner, M. P., Foumier, D. J., Andrews, R. K., Gorman, J. J., Chesterman,C. N., and Bemdt, M. C. (1989) Biochem. Biophys. Res.Commun. 164, 1373-1379. Gamble, J. R., Skinner, M. P., Bemdt, M. C., and Vadas, M. A. (1990) Science249, 414-417. Gasic, G. J., Gasic, T. B., Galanti, N., Johnson,T., and Murphy, S. (1973) Int. J. Cancer 11, 704-718. Karpatkin, S., and Pearlstein,E. (1981) Ann. Intern. Med. 95,636-641. Pearlstein,E., Salk, P. L., Yogeeswaran,G., and Karpatkin, S. (1980) Proc. Natl. Acad. Sci. USA 77, 4336-4339. Kijima-Suda, I., Miyazawa, T., Itoh, M., Toyoshima, S., and Osawa, T. (1988) Can.Res. 48, 3728-3732.
1356
