CD15, PHAGOCYTES THROUGH CARBOHYDRATE SIALYL-CD15 ...

0022-1767/92/14810-3221$02.00/0 THEJOURNAL OF 1MMUNOLOCY Copyrlght 0 1992 by The Arnerlcan Assoclatlon of lmmunoioglsts

Vol. 148. 3221-3229. No. 10. May 15. 1992 Printed in U . S . A .

ACTIVATION OF HUMAN PHAGOCYTES THROUGH CARBOHYDRATE ANTIGENS (CD15, SIALYL-CD15, CDwl7,ANDCDw65)l FRIDTJOF LUND-JOHANSEN,'+JOHANNAOLWEUS.* JOHN S. THOMPSON,'RAMONVILELLA,"

VACLAV HOREJSI,+ KEITH M. SKUBITZ,' AND FRANK W.SYMINGTONq

From the 'Department of Pathology, The Gade Institute, Universityof Bergen, Haukeland Hospital,N-5021Bergen. Norway; 'Institute of Molecular Genetics, Czechoslovak Academyof Sciences, Videnska 1083, 142 20 Praha, Czechoslovakia; 'Department of Medicine, Universityof Minnesota Medical School, Minneapolis, M N 55455;$Department of Medicine, Chandler Medical Center, Universityof Kentucky, Lexington, KY 405360084;flservei d'Immunologia, Hospital Clinic i Provincial, Barcelona 08036.Spain; and 'SeattleBiomedical Research Institute, Seattle,W A 98109

The leukocyte carbohydrate (CHO) Ag CD15, sia- the ELAM-1 molecule on activated endothelium (8, 12). lyl-CD15, and CDw65 have recently been found to Several cell adhesion molecules serve a dual role, acting function as ligands for CD62 andELAM-1 cell adhe- both as binding sites formolecules on extracellular masion molecules on platelets and endothelium, retrix or other cells, a n d as receptors capableof transmemspectively. Cell adhesion ligands also may act as branesignaltransduction. In experimentswherethe receptors capableof signal transduction.We there- physiologic multivalent engagement of adhesion molefore investigated the possibility thatthese CHO Ag cules has been mimicked by antibody cross-linking, it andCDwl7,aglycolipid Ag whose expression is has been shown that molecules such as CD2 (13).CD1 l a regulated by leukocyte activation, may have recep(14),CD1 l b , CD35 (15), andCD44 (16) can actas receptor-likecharacteristics. The effects ofantibody tors and mediate cell activation. Although several reports cross-linking of CHO Ag on phagocyte activation have shown that mAb against CHO Ag may influence were measured by using flow cytometry and fluorescent indicators for cytoplasmic calcium ions, phagocyte functions(17-24), there islimited information about potential involvement ofCHO Ag in transmemoxidative burst, and the granule-associated proteins brane signaling in human monocytes and granulocytes. CDllb andCD67.Cross-linkingofCD15, sialylThepresentstudyinvestigatedthe possibility that CD15. CDw65, or CDwl7 induced a moderate release phagocyte CHO Ag may have receptor-like characterisof calcium ions into the cytoplasm of granulocytes, a strong activation of oxidative burst, and a low up-tics. Flow cytometry and fluorescent indicators were used regulation of CDllb and CD67 compared to the ef- to study the effects of antibody cross-linking of these burst, fects of treatmentwith 4 pM FMLP. The results molecules on cytoplasmic calcium levels, oxidative suggest a role for CHO Ag in leukocyte signal trans- and degranulation of leukocyte subsets. The antibodies duction and support the view thatthese molecules used were specific for the carbohydrate epitopes CD15, are involved in phagocyte activation. s-CD15, CD43 (leukosialin) (25), CDwl7 (lactosylceramide) (3), and CDw65, adhesion molecules including CDllb/CD18 (CR3) (26), CDlla (LFA1) (26). and CD44 (hyaluronic acid receptor) (27) and CD45 (protein tyrosine phosphatase) (28). Our results demonstrate that in the absence of other stimuli, cross-linking of cell-bound mAb to several CHO Ag induces cytoplasmic calcium fluxes, activation of the respiratory burst, and degranulation.

Humanmonocytesandgranulocytesexpresslarge amounts of complex CH03Ag on their cell surface (1-5). Recently, it has been shown that some of these CHO mediate adhesion of granulocytes to activated platelets andendothelium(6-12).The CD15epitope(Galbl~ ( F u c " ~ - ~ ) G ~ c N A(1) c ~and ~ - RCD15 ) (2) have been characterized as ligands for CD62 (GMP-140) present on acMATERIALS AND METHODS tivated platelets (6. 1 l ) , whereas s-CD15 and CDw65 (a Reagents. Dulbecco's PBS with or without calcium and magneceramide-dodecasaccharide a 1-3 fucosylated on the pe- sium (PBS-Ca or PBS, respectively) and FCS were purchased from FMLP, PMSF. and pepsin nultimate glCnAC) (5)have been shown to interact with Gibco (Paisley, Scotland, United Kingdom). Received for publication December 4, 1991. Accepted for publication February 20. 1992. The costs of publlcation of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked aduerttsernent in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. This study was supported by grants from Inger-Margrethe and Per Jzger, Frank Mohn A/S. The Norwegian Council for Research and the Humanities, and The Blix Foundation. 'Address for correspondence: Fridtjof Lund-Johansen. Department of Pathology, The Gade Institute, University of Bergen, Haukeland Hospital, N-5021 Bergen. Norway. Abbreviations used In this paper: CHO. carbohydrate: s-CD15. sialylated form of CD15: PBS-FCS. PBSwith 2% FCS and 5 mM glucose: GAMM. F(ab'), fragments of goat antibodies to mouse IgM p-chain speciflc: GAM-HL.F(ab')' fragments of goat antibodies to mouse IgG heavy and light chains; PBS-FCS-Ca. PBSwith calcium, magnesium, 2% FCS, and 5 mM glucose.

were from Sigma Chemicals Co. (St. Louis. MO). Dihydrorhodamine 123 and Fluo-3 were from Molecular Probes (Eugene,OR]. FMLP (10 mM), Dihydrorhodamine 123 (10 mgml), Fluo-3( 1 mM), and PMSF (100 mM) were dissolved in dry DMSO (Merck, Darmstadt. Germany) and stored in aliquots at -70°C. The solvent concentrations were less than 0.5% in final dilutions. Paraformaldehyde, and all salts used in laboratory-made solutions, were analytical grade. Microwell plates (polystyrene, V-bottomed) were from Nunc (Copenhagen, Denmark). Polyclonal secondary antibodies. Fluorochrome-conjugated and unlabeled GAM". GAM-HL, and goat anti-mouse IgG (Fc-specific) were generouslyprovided by Jackson lmmunoresearch (West Grove, PA]. Unconjugated secondary antibodies were used as 1 . 2 m g m l i n PBS. whereas conjugated antibodies were usedas 20 pdml in PBSFCS and 0.1% sodium azide. The unconjugated secondary antibodies were found to contain>99%pure F(ab')z fragmentsby HPLC analysis using a TSK 3000W column (Toyo Soda Manufacturing, Japan) (data not shown). mAb. The mAb used in the study are listed in TablesI and 11. All

322 1

3222

PHAGOCYTE ACTIVATION THROUGH CARBOHYDRATE Ag

mAb were diluted in PBS-FCS with 0.1% sodium azide and used at Measurement of oxidatiue burst. Cellular oxidative burst was concentrations giving the maximal fluorescent staining and minimal quantified by using the fluorescent indicator, Dihydrorhodamine cell aggregation a s determined by flow cytometry. The concentra123 (32). Dihydrorhodamine 123 is a nonfluorescent, cell memtions of the commercial mAb varied from 2 to 50 pglml, whereas brane-permeable compound. During the oxidative burst, the intraascites fluidswere used a t 1:100 to1:200 dilutions. cellular Dihydrorhodamine 123 is irreversibly converted to the fluPreparation ofF(ab’Jz fragments ofZgG mAb. IgG was purified on orescentcompound,Rhodamine 123 (32, 33). This rhodamine is a protein A Sepharose 4 fast flow column (Pharmacia, Stockholm, membrane-impermeable, and so accumulates in thecells. The cellSweden),dialyzed against citrate buffer(pH 3.5 for IgCl, pH 4.3 for ular oxidative burst is measured as a function of cellular Rhodamine IgGZa), and incubated with pepsin (25pglml) for 18 h a t 37°C. The 123 fluorescence intensity (32). Thespecificity of this indicator for digestion was stopped by the addition of 1 mM PMSF. and the oxidative burst has been demonstrated in experiments with cells solution dialyzed against PBS before intact IgG was removed by from patients with chronic granulomatous disease (33). After labeling with mAb a t 20°C (see above), cells were washed once in PBSprotein A chromatography. The purity of F(ab‘)z fragments was demonstrated by measuring a >99%reduction in cell staining with FCS (20°C). Ten microliters of GAM”. GAM-HL. FMLP (40 pM). or phycoerythrin-conjugated antibodies to theFc part of mouse IgG. as PBS were added to each pellet, and the cellsresuspended on a compared to the intact IgG mAb. All F(ab’)z fragmentsprepared in whirlmixer before 100 pl PBS-FCS-Ca prewarmed to 37°C. containthis way stained cells with the same patternsas the corresponding ing 10 pglml Dihydrorhodamine 123 wereadded, and the cells intact IgG antibodies (datanot shown). incubated a t 37°C. After 20 min of incubation, 100 pl of ice-cold Removal of contaminating IgC f r o m IgM ascitesfluids. Ascites PBS-FCS containing 0.02% paraformaldehyde were added, and the fluid preparations were dialyzed against 20 mM phosphate buffer samples kept on ice until flow-cytometric measurement. The com(pH 7.2) andpassed over a protein G Superose HPLC column (Phar- bination of low paraformaldehyde concentration and low temperamacia). The IgG-depleted preparations were then dialyzed against ture was used to prevent leakage of intracellular dye that occurred PBS without calcium and magnesium. over time with higher temperatures and higher concentrations of Isolation of leukocytes. Venous blood from healthy laboratory fixative. Increases of twofold or greater in cellular Rhodamine 123 personnel between 20 and45 yearsold was drawn into Vacutainers fluorescence relative to control were considereda s significant stimcontaining ACD solution A (Becton Dickinson, Oxnard, CA) and ulus-induced respiratory bursts. diluted 1:lO in a solution containing 0.8% NH4C1, 0.08% NaHC03, Up-regulation of surface expressionof granule-associated proand 0.08%EDTA (pH 6.8, 20°C) to lyse RBC. The leukocytes were teins. Activation-inducedup-regulation of the granule-associated then pelleted by centrifugation at 180 x g for 5 min and washed proteins CD1 l b and CD67 was determined by flow-cytometric measonce in PBS-FCS. urement of immunofluorescence (34, 35). CDllb and CD67 are Labeling of cells with mAb. For functional experiments, as well associated with thespecific granules of granulocytes and are transas for immunofluorescence measurements, leukocytes were labeled located to the plasma membrane during degranulation (34. 35).In with mAb and washed before the addition of secondary antibodies. addition, CD67 has been suggested to be associated with the secre(35.36). This labeling procedure was chosen to prevent the formation of tory granules or latent alkaline phosphatase compartment immune complexes when the secondaryantibodies were added, and For measurements of Ag up-regulation, the cells were activated in to allow better control of the ratio of F(ab’)z fragments of goat the same manner asfor oxidative burst measurements (see above) antibodies to mAb when the concentrations of the mAb were un- except that Dihydrorhodamine 123 wasexcluded from the solutions. known. Leukocytes were transferred to microwell plates (2x lo5 to After 20 min of incubation at 37°C thecells were cooled, washed, 8 x 105/well) and pelleted by centrifugation a t 160 x g for 3 min. and incubated on ice for 20 min with PBS containing 10% normal The supernatant was then discarded, and the cells resuspended by mouse serum to block nonspecific binding sites for IgG antibodies. whirlmixing. Ten microliters of the mAb solutions were added to The cells werethen washed, stainedon ice with the FITC-conjugated each pellet. and the plates kept a t 20°C for 25 min with whirlmixing mAb Bear1 (anti-CD1 lb) or8 0 H 3 (anti-CD67) for20 rnin, washed in ice-cold PBS-FCS, fixed in 1 % paraformaldehyde, and measured by every 7 min. The cells were then washed once in PBS-FCS. Immunofluorescence. Leukocytes were labeled with mAb as de- flow cytometry. The specificity of 80H3 is based on it recognizing a 100-kDa phosphatidyl-inositol linked protein, and on the ability of scribed above, washed once with PBS-FCS, and stained with 10 pl CD67 mAb to block 8 0 H 3 binding to granulocytes. of fluorochrome-conjugated antibodies for 30 min on ice. The anti- the G10F5 anti Priming of leukocytes with FMLP. Venous blood was collected body-labeled cells were washed, fixed in PBS containing 1%parainto 3-ml heparin Vacutainers (Becton Dickinson). FMLP (2 x formaldehyde, and analyzed by flow cytometry. Measurement offluctuationsin intracellular calcium concentra- M) or PES was then added, and the tube incubatedin a 37°C water tion. Leukocytes (107/ml)were incubated 25 rnin a t 37°C in PBS- bath for 20 rnin with constant agitation. The leukocytes were then FCS containing 2 pM Fluo-3-acetoxymethylester (29). washed, and isolated and labeled with mAb as descrived above. Flow cytometry and data analysis. Calcium measurements were transferred to microwell plates. The cells were then labeled with mAb as described above, washed, whirlmixed. and resuspended in performed with a Coulter Epics V flow cytometer (Coulter Electronics, Luton, United Kingdom) interfaced to a CICERO PC-based data 200 pl of cold (4-6°C) PBS-FCS-Ca. The cells were then kept a t 46°C and prewarmed to 37°C for 90 s before the flow-cytometric aquisition and analysis system (Cytomation). The excitation source was a 488-nm argon laser, and standard fluorescein filters were measurement. Samples could be kept a t 4-6°C for a t least 4h without detectable changes in the calcium fluxes induced by any used for Fluo-3 detection. The kinetics and amplitudes of calcium mAb, whereas storage a t 20°C led to decreased responses to several responses wereanalyzed using specially designed software from Cytomation. The software draws a curve of the mean fluorescence mAb. including Mol, VIM2, and DAKO-15 (data not shown). During the flow-cytometric analyses, the sample temperature wasintensity of the cells ina population during thetime of measurement. kept a t 37°C with the help of a laboratory-made flow cytometry Measurements of immunofluorescence and Rhodamine 123 fluorescence were performed with Coulter Profile I1 or Becton Dickinson sample chamber. Monocytes and granulocytes were distinguished by measurements of cellular light scatter, and their Fluo-3fluores- FACSscan flow cytometers with standard filter setup. Data obtained cence gatedvs time to separate cytograms (30,31).Basal cytoplasmic from the Profile cytometer were processed with Verity Isocontour calcium levels were recorded for 10 to 20 s before 10 p1 F(ab’)z software from Verity Softwarehouse(Topsham, ME] forgraphic fragments of goat antibodies were added to cross-link cell-bound presentation. Statistical methods. Significance of difference was evaluated by mAb. Fluctuations in cytoplasmic free calcium concentration were recognized as alterations in Fluo-3 fluorescence intensity over time paired Student’s t-tests. All experiments were performed a t least three times and with (30,31). The mean fluorescence intensity of the cells at anygiven time point was calculated by software from Cytomation, Inc. (Engle- leukocytes from a t least three different persons. wood, CO). Inthis mean curve, the maximal cellular Fluo-3 fluorescence was compared to mean prestimulated levels, and a doubling RESULTS or more was considered indicativeof a stimulus-induced cytoplasmic calcium flux. Fluo-3 is regarded as a good indicator for observing Expression of CHO Ag on leukocyte subsets. Granurelative changes in cytoplasmic calcium levels because it has high sensitivity a t both high and low calcium concentrations (29).Fluo-3 locytes expressed high levels ofCHO Ag a s determined is less reliable as a n indicator forabsolute calciumlevels (29). by flow-cytometric measurement of staining with antiHowever, this study only measured relative changes in cytoplasmic CHO mAb (Table I). The mAb to CD15 and s-CD15 were calcium, and no attempts were made to determine absoluteconcenfound to be the strongest reacting, followed by those to trations. To determine the source of mobilized calcium ions, antibody- CDw65 andCDwl7 [Table I). Anti-CHOmAb stained labeled cells were resuspended in either PBS-FCS-Ca or PBS-FCS. granulocytes more brightly than mAb to adhesion moleTo chelate anycalcium ions present in thePBS-FCS solution. 100 pl of the sample were mixed with 100 p1 of PBS-FCS containing 1.0 cules such as CD1 l b , CD18, CD35, and CD44 (Table I). Monocytes were negative for the anti-CD15 mAb and had mM EGTA less than 1 min before flow-cytometric analysis.

3223

PHAGOCYTE ACTIVATION THROUGH CARBOHYDRATEAg TABLE I Binding

2 f 0

CD15

PR.S ~

MFF MFI

Ig Class

mAb "

Lymphocyte Form

Ag

MFI Monocyte

2 f0

2 f0

CD15

1 f0 1f0 1fO

3 f 0 5 f0 3f 1

333f43 235 f 21 260 f 63

Dako-CD15 My- 1 AHNl. 1

IgM

T5A7 GO35 MEM-74

IgM CDwl7 Supernatant 17 IgM Ascites CDwl7 IgM Ascites CDwl7

Supernatant Ascites

Ascites CD15

B. CD43 CD43

HElO CF4 VIM-2

IgM CDw65 Ascites IgM Ascites CDw65 IgM CDw65

CSLEXl FH6

IgM IgM

s-CD15 s-CD15

Reference

Sigma DAKO C. 1. Civin K. Skubitz

(481 1191

165 f 2 6 Symington F. J.Thompson 10 V. Horejsi

f9 20 f 3 1 1880f 9f2 2 324 11 f 21 5 2 51 3 3 f

ND ND ND 281 f 10282 Ascites Supernatant163 f 18140

Source

Granulocyte

I f 0 f0

IgM

B1 B6 Purified CD43 IgCl MEM-59 IgCl 84-3C1 IgC

leukocyte subsets

2 f 0 Negative control Purified 1.0

MOPC 104 E

a

of anti-CHO rnAb to human

f 20370 f 12230

f 26

f 10

Axelsson V. Horejsi R. Vilella

Ascites 1

32 f 8b 1f0 f O

43 f 201 12 f 12 2f1 37 f 1 93f30 2 3 2 f 9

M. Dockhelar M. Dockhelar BehringWerke

Purified Supernatant31

34 f 7" f 7"

77 f 12255 f 2 3 P. Terasaki 54 f 9 354 f 100 F. Symington

121 (371

MFI. mean fluorescence intensity givena s relative channel numbers (meanf SEM; n = 31. Data given for subpopulationof positive cells (5-30%).

lower staining than granulocytes with all other anti-CHO mAb tested (TableI). In contrast, monocytes were brightly stained withmAb to CD43, CD44 and with the anti-CD45 mAb MEM-28 (Table 11). Cross-linking of carbohydrate Ag induces a cytoplasmic calciumflux independentlyof IgG FcR. Following the addition ofGAM" to cells labeled with ascites fluid of My1 (anti-CD15). there was a rapid and transient increase in cytoplasmic calcium concentration in granulocytes, but not in monocytes (Fig. 1: Table 111). Monocytes did, however, respond with a calcium flux when GAMHL was added to the same sample(Fig. 1). The monocyte response was abolished when IgG-depleted ascites fluid was used, whereas the granulocyte response was unchanged (Fig. 1). This suggests that the response of monocytes toGAM-HL was due to cross-linking of contaminating IgG antibodies, whereas the response of granulocytes to GAM" was due to specific cross-linkingof IgM mAb. A s similar results were obtained with other mAb including AHN 1.1 (anti-CD15) and HElO (anti-CDw65) (data notshown).and as GAM" failedtoinducecalcium

responses in cellslabeled with several differentIgG mAb that caused calcium fluxes when cross-linked by GAMHL (data not shown), GAM" was used throughout the study toavoid nonspecific IgG-FcR interactions. When cells werelabeled with a nonspecific IgM control antibody (MOPC-104E). no response was seen after the addition of any of thesecondaryantibodies (Fig. 1). whereas both monocytes and granulocytes gave positive control responses to 4PM FMLP (Fig. 2). Treatment with FMLP yielded a 4.7 k 0.3- and 5.0f 0.5-fold increase in meanFluo-3fluorescence of monocytes and granulocytes, respectively (mean SEM; n = 5). This increase was significantly greater than the response seen after cross-linking of CD15 ( p < 0.01). Cytoplasmic calcium fluxes in granulocytes were also seen after specific cross-linking of IgM mAb to s-CD15. CDwl7, andCDw65 withGAM" (Fig. 3, Table 111). Monocytes responded only to cross-linking of one s-CD15mAb, CSLEX-1 (Table 111). This difference between the functional effects of two mAb to the sameAg could be due to low expression of the difucosyl-s-CD15 epitope preferred

*

TABLE II Binding of rnAb to glycoprotein Ag expressed by human leukocytes

Ig Class

mAb

PBS 2f0 TEPC 183 MOPC 104 E MEM-25

IgM IgM

Ag

MFI

MFI Form

MFI" Lymphocyte

Negative control Ascites Negative control Purified

CDlla lgCl

Ascites

Bear 1 Mo 1

IgCl IgM

CDllb CDllb

Ascites 21 Supernatant 12

Ascites MEM48 CD18 68-5A5

lgCl IgC

CD18

Supernatant 47

MEM-85 BRIC-222 33-383

IgCl IgGl

IgC

CD44 CD44 CD44

MEM-28

IgCl

CD45 Ascites 485

a

Purified CD67 IgCl

913327 f1 4 f lb f 2"

33 f 4 f4 59fr5b

J3D3 Ascites CD35 IgCl

80H3

1*0 I f 0 I f 0

Ascites 250 Supernatant287 Supernatant169

f 50 f 60

k8 f 485 20

21 97 55 4 "

Monocyte

Granulocyte

4fO 2 f 0

2 f 0 3 f0 2 f0

7

49f6

Source

Reference

Sigma Sigma V. Horejsi

(531

136 f 20 115 90 f 7

f8 Immunotech 70Coulter f2

80f 5 118 f 5

57 f 6 75 f 6

V. Horejsi R. Vilella

(531

40f8

36f 5

lmmunotech

(551

f9 V. Horejsi f8 D. J. Anstee 10 R . Vilella

(561

f7

(501

490 f157 20 440 f 20207 314 f 16107 f129 122

5

(541

*

42f 3

V. Horejsi

Immunotech

MFI, mean fluorescence intensity givena s relative channel numbers (meanf SEM: n = 31. Data given for subpopulationof positive cells (5-30WJ.

3224

PHAGOCYTEACTIVATIONTHROUGHCARBOHYDRATE

Ag

responses at least partially are due to release of calcium ions from intracellular stores, and suggests that the phospholipase C/inositol trisphosphate pathway(38)could be involved in anti-CHO-induced activation. Cross-linking of other Ag does not lead to cytoplasmic calciumfluxes. When nonspecific IgG FcR interactions were avoided by using F(ab')z fragmentsof IgG mAb, no significant calcium fluxes were seen after cross-linking I I of CD1la, CD18, CD35,CD43, or CD45(data not shown). 90LS In the granulocyte population, there were no calcium fluxes induced by any mAb to CD1 la, CD43, CD44, or MON GRAN CD45 even when whole IgG mAb were used (data not shown). Theonly mAb to other molecules that triggered FcR-independentcalciumfluxes was the purified IgM anti-CD1 l b (Mol), which upon cross-linking with GAMM induced a 3.1 f 0.3- and 2.3 0.3-fold increase in the Fluo-3 fluorescence of monocytes and granulocytes, respectively. In contrast, no response was triggered when cell-bound F(ab')z fragments of another anti-CD11b mAb F (Bearl)were cross-linked withGAM-HL (data not shown). L Cross-linking of CHO Ag induces strong activation of U thegranulocyterespiratoryburst. Afteraddition of 0 GAM" to cells prelabeled with anti-CD15 (IgG-depleted 3 ascites of Myl) or anti-s-CD15 (FH6 supernatant), there was an increase in granulocyte Rhodamine 123 fluoresF as compared to granulocytes that had been incucence L bated with PBS only (Fig. 4.; Table 111). No response was U seen when GAM" was added to unlabeled cells or to 0 cells that had been incubated with nonspecificIgM antiR body (TEPC-183) (Fig. 4). Incubation of Myl- or FH6E labeled cells in the absence ofGAM" did not result in S increased Rhodamine 123 fluorescence above PBS the or C irrelevant IgM controls (Fig. 4), demonstrating that crossE linking of the cell-bound IgM mAb was necessary for N activation. Addition of FMLP to unlabeled or mAb-labeled C cells also induceda n increase in granulocyte Rhodamine E 123 fluorescence (mean relative granulocyte Rhodamine 123 fluorescence, 5.5 k 0.5; mean f SEM, n = 5). This increase was, however, significantly smaller than the response to cross-linking of CD15 and s-CD15 with GAMM in the same experiments ( p < 0.01). No significant H increase in the FMLP-induced respiratory burst was seen I min IIME in cells prelabeled with mAb, suggesting that cells were Figure 1 . Flow-cytometric determination of fluctuations of leukocyte not primed by potential contaminants in the antibody cytoplasmic calcium concentration after cross-linkingof CD15 with mAb preparation (Fig. 4). No difference in the results was seen Myl. Upper diagram: Leukocyte subsets identified by light scatter (L. lymphocytes: M. monocytes; G. granulocytes; 90LS. 90" light scatter; when IgG-containing ascites fluid of anti-CD15 mAb My1 FALS. 4" light scatter]. Lower diagrams: Linear Fluo-3 fluorescence of monocytes (MON) and granulocytes (GRAN) gated vs time. Fluo-3-loaded was used (data not shown). Similar results for granulocytes were obtained with cells were labeled with indlcated mAb preparation, washed, and resuspended in 200 pl PBS-FCS-Ca. Tospecifically cross-link the IgM mAb. 10 other anti-CD15 and anti-s-CD15 mAb as well as with p1 of GAM" were added at indicated time point. To cross-link possible mAb to CDwl7 and CDw65 (Table 111) (data not shown). cell-bound IgG antibodies, 10 @I of GAM-HL were added at indicated time point. The diagrams are taken from an experiment representative of three Monocytesresponded with a n increase in Rhodamine performed. 123 fluorescence when incubated with CSLEX-1 (antisCD15) or VIM2 (anti-CDw65) followed by GAM" but by FH6 (37).When PBS was added instead ofGAM", not with othermAb to these Ag (Table 111). these rapid increases in cellular Fluo-3 fluorescence wereCross-linking of other Ag does not lead to activation never observed (data not shown). of the respiratory burst. When FcR-dependent antibody Calciumflux induced by CHO cross-linking involves binding was avoided by using F(ab')z fragments ofIgG release of calcium ions from intracellular stores. To mAb, no significant respiratory bursts were seen after investigate the source of the calcium ions mobilized by cross-linking of C D l l a , CD18,CD35,CD43,orCD45 CHO-cross-linking. granulocyte calcium fluxes were (data not shown). In the granulocyte population, there measured in the presence of the extracellular calcium was no oxidative burst induced by cross-linking of any chelator EGTA. These experiments showed that chelat- mAb to CD1 la, CD35, CD43, CD44, or CD45, even when ing of extracellular calcium ions did not abolish the cal- whole IgG mAb were used (data not shown). The only cium fluxes induced by cross-linking of granulocyte CHO- mAb to other molecules that triggered FcR-independent respiratory burst was the purified IgM anti-CD11 b (Mol). Ag (data not shown). This demonstrates that the observed

*

3225

PHAGOCYTE ACTIVATION THROUGH CARBOHYDRATEAg TABLE I11 Monocyte and granulocyte activation after cross-linkingof carbohydrate Ag (CDI 5. s-CD15. CDW17. CD43. CDw65J Degranulation Burst

Oxidative

Flux

Antibody

4

Calcium

Fluo-9 relative to control^ Rhodamine 123 relative controlCtocontrolbto Monocyte

Granulocyte

Monocyte

CDI Ib-FITC relative

CD67-FITC relative to controld

Granulocyte Granulocyte Granulocyte

CD15

4 1 . 21f.632.f0.08.k210. .31f‘0 . 3 1 . 6 20 . 2 3.3 f 0.31.6 f 0 . 32 4 . 4 f 3.3 DAKO 15 l.lfO.l 1 6 . 2 223. .02 1.1f2 0.7 0f0.3

2.2 f 0.2 2.0 f 0.1 f 0.2

2.1 k O . 1 1.7 k 0.2 1.8 f 0.1

S-CD15

FH6 CSLEXl

2.1 f1.9 0.1 0.2 2.1 f 2.1

f 0.1 f 0.1

1.41.4 f 0.1 1.8 f 0.1 1.9fO.l

f 0.1 1.5 f 0.1 2.0 f 0.1

kO.1 1.1 1.0 1.0 f 0.1

1.0 f 0.1 f 0.1

13.6 l4 ..0 2lf f.O 4O 0.2..l37 l .2 4.2 3.2 f 0.4

f0.5

2.0 f 0.6 45.4

”-

CDwl7

T6A7 ” . ~ GO35 MEM74

1 . 32f .05.lf1.0l3f. 3 .O6.fl0 . 5 1 . 3 f3O . 5. lf10. 5 . 4f80. 0 . 12 3 . 0 1 .22.1f6.3O f290.f.l.035.f15 . 3

CD43

MEM-59F(ab)* 1.1 f Ol .. ll f O . l B l B 6 F 1( a. 1b l .f)l0,.f0l.O f2. O 9 . lf. O l .l

CDw65

VIM2 HE- 10 CF4 -~

~

~~

~

~~

f 6.0

~

1.1 “ 0 . 1

1.1 f O . l

1.5 f 0.2 1.8 f 0.1 _ 23 . 82 .1 f.106 f 50 .21..033.f34 . 0 ~ _ _ - _ 1.4f0.1 1 . 63f 0 . 1.11 f0 . 7.13 f0 6.2f3.0 1. O f O . 1 1 . 1 f 0.0 ~

~~

1.1 f 0.0 ~~~

1.1 f 0.2 ~

1l0. .8 O3ff O . l ~~

1 . 1 f 0.1

The maximal Fluo-3 fluorescence after Ag cross-linking wasdivided by the mean fluorescence intensity before addition of Flab’), fragments of goat antibodies. Resultsare given as mean f SEM; n > 4. bRhodamine-123 fluorescence after Ag cross-linking was divided by the fluorescence intensity of unlabeled cells incubated with Flab’), fragments of goat antibodies. Results aregiven as mean f SEM: n = 5. Bearl (anti-CD1 l b ) FITC fluorescence after Ag cross-linking was divided by the fluorescence intensity of unlabeled cells incubated with F(ab’), fragments of goat antibodies. Results are given a s mean f SEM: n = 5. 8 0 H 3 (anti-CD67)FITC fluorescence afterAg cross-linking wasdivided by the fluorescence intensityof unlabeled cells incubated with F(ab’), fragments of goat antibodies. Results are given a s mean f SEM: n = 5. e Underlined values, p < 0.01. a

ANTI-CHO mAb

F L U 0 3

H

TIME

I min Figure 2. Flow-cytometric determination of FMLP-induced fluctuations of leukocyte cytoplasmic calcium concentration. Monocytes (MON) and granulocytes (GRAN]were recognized by light scatter and theirFluo3 fluorescence (linear] was gatedvs time. FMLP (4 r M ) was added at the indicated time point. The diagramsare taken from an experiment representative of five performed.

F L U 0

FH6 S -CD15

GO35 CDwl7

R

E S C E

Upon cross-linking of cell-bound Mol with GAM”, there was a 2.5 ? 0.3- and 3.4 f 0.5-fold increaseinthe HE10 Rhodamine 123 fluorescenceof monocytes and granuloN CDw65 cytes, respectively (mean f SEM, n = 5). whereas no C response was seen in the absence ofGAM” (data not E shown). When F(ab’), fragmentsof another anti-CD1 l b H mAb (Bearl) wereapplied in combination withGAM-HL, I min there was no increase in granulocyte or monocyte RhoTIME damine 123 fluorescence [data not shown), suggesting Figure 3. Flow-cytometric measurement of changesin monocyte that extensive cross-linking is necessary for activation (MON)and granulocyte (GRAN)calcium concentration after cross-linking of the carbohydrate Ag s-CD15, CDwl7, and CDw65. The cells were through CDl 1 on b unprimed cells. prelabeled with the indicated mAb. Fluo-3 fluorescence(linear) was gated Staining and activation with anti-CDwl7mAb is de- vs time for cell subsets recognized by light scatter. To specifically crosscreased in FMLP-primed granulocytes. Stimulation of link the IgM mAb. 10 pl of GAM“ were added at indicated time points. The diagrams are taken from an experiment representative of five pergranulocytes withFMLP leads toa decrease in the expres-formed. sion of CDw 17 and an increase the expression of CD 1 1 b (39. 40). To investigate whether the oxidative responses induced by mAb to these two Ag paralleled the cell sur- led to a 50 ? 9% decrease in staining with anti-CDw17 face Ag levels, we examined the effects of 2 X 10” M mAb MEM-74 and a corresponding decrease[53f 8%)in FMLP on staining andoxidative burst inducedby MEM74 the oxidative burst after cross-linkingof cell-bound anti(anti-CDw17) and Mol (anti-CD1l b ) . Priming with FMLP CDwl7 with GAM” (mean ? SEM; n = 3) (Fig. 5).Typi-

3226

PHAGOCYTE ACTIVATION THROUGH CARBOHYDRATEAg

+ PBS

+ 4pM FMLP + GAM”

MEM=74(CDwl7) NOT PRIMED

FMLP-PRIMED

.

.

- . .- .-...-... .........

1

C E L L

N U M B E

R

LOG RHODAMINE 123 FL Figure 4. Flow-cytometric measurements of granulocyte oxidative burst after activation with anti-CHO mAb and FMLP. Unlabeled cells and Figure 5. Flow-cytometric measurements of monocyte (MI and grancells labeled with themAb indicated inside each panel were incubated 20ulocyte (GI cell surface staining [top] and activation (bottom] with antimin at 37OC with PBS-FCS-Ca containing 10 pg/ml Dihydrorhodamine CDwl7. Cells were eitherprimed (left] or notprimed (right]with 2 X lo” 123, no stimulus, GAM”, or 4 pM FMLP. Granulocytes were recognized M FMLP and labeled with anti-CDwl7 (MEM-74).The cells were then by light scatter measurements and their Rhodamine 123 fluorescence either stained on ice with FITC-conjugated GAM” (top] to measure cell gated to one-parameter histograms. Histograms are taken from an expersurface CDwl7 (top] or incubated at 37°C for 20 min with unconjugated iment representative of five performed. GAM” and 10 pglml Dihydrorhodamine 123 in PBS-FCS-Ca to assess oxidative burst (bottom]. Fluorescence and 90’ light scatter (90LS)(both log mode] were then measured for unprimed and primed cells. Theblack cally,granulocyteswereheterogeneouslyaffected by histogram on the right border of each diagramrepresents fluorescence from an experiment FMLP with regard to staining and oxidative burst with intensity vs numberof cells. The diagrams are taken representative of three performed.

anti-CDw17 mAb (Fig. 5).In contrast to reduced staining and burst seen with anti-CDw17-primed granulocytes, oxidative burst than was observed after stimulation with priming with FMLP increased the staining and oxidative 4 pM FMLP ( p < 0.01) (Figs. 4 and 7). Stimulation with 4 burst seenwith the anti-CD11 bmAb Mol. Priming with pM FMLP led to a 3.1 f 0.1-fold increase of granulocyte FMLP led to a 205 f 40% increase in Mol fluorescence CD 1 1 expression, b whereas cross-linkingof CHO Ag ink SEM; intensity as compared to nonprimed cells (mean duced a doubling of granulocyte CD 1 1b expression (Fig. n = 3)(Fig. 6).In addition, there was a 650 f 80%increase 7, Table 111). The difference between the responses inin the granulocyte oxidative burst induced by the addition duced by FMLP and anti-CHO-mAb was even larger when of GAM” to Mol-labeled cells (mean5 SEM; n = 3) (Fig. increases in granulocyteCD67 were measured. InFMLP6). The background levels of Rhodamine 123 fluorestreated granulocytes, the mean CD67 fluorescence intencence were not altered by priming with FMLP, and in the sity was 5.5 k 0.6 times higher than in cells incubated absence of GAM”, there was no increase in Rhodaminewith PBS. In contrast, CD67 fluorescence typically only 123 fluorescence in any sample (data not shown). Theredoubled after cross-linking ofCHO Ag (Table 111) ( p < fore, oxidative responses induced by either anti-CDwl7 0.01).Preincubationwith mAb did notincreasethe or anti-CD 11b required cross-linkingof cell-bound mAb. FMLP-induced up-regulation of either CD1 l b or CD67 In addition, these data demonstrate that the strengthof (Fig. 7) (data not shown). the oxidative burst induced by cross-linking of CDw 17 or C D l l b parallels changes in cell surface expression of DISCUSSION these Ag. We have demonstrated that cross-linking ofCHO Ag Respiratory burst induced b y CHO cross-linking is cytoassociated with low up-regulationof C D l 1 b and CD67. on humangranulocytesandmonocytesinduces of therespiratory When changes in granulocyte CD1 l b expression and plasmiccalciumfluxes,activation burst, and increased cell surface expression of granuleoxidative bursts were measured in the same experiments, of evidence suggest that cross-linking CD15, s-CD15, CDwl7, and CDw65 led to associated proteins. Several lines a lesser increase in C D l l b expression and a stronger the observed responses reflect functional characteristics


3227

Mol K D l l b )

M B E

R

L o ( ; 901,s Figure 6. Flow-cytometric measurements of monocyte (M) and granulocyte (G) staining ( t o p ) and activation (bottom] with anti-CDllb mAb Mol. Cells were either (right)or not (left) primed with 2 x M FMLP and labeled with anti-CD1 l b . The cells were then either stained on ice with FITC-conjugated GAM" (top)to measure cell surface CD1 l b (top) or incubated at 37OC for 20 min with unconjugatedGAM" and 10 pg/ml Dihydrorhodamine 123 inPBS-FCS-Cato assess oxidative burst (bottom). Fluorescence and 90" light scatter (9OLS) (both log mode) were then measured for unprimed and primed cells. The black histogram on the right border of each diagramrepresents fluorescence intensity vs number of cells. The diagrams are from the same experiment as those in Figure 5 and arerepresentative of three performed.

that arespecific for CD15, s-CD15, CDwl7, andCDw65. 1)Cell activation induced by CHO cross-linking was not due to FcR interactions, inasmuch as IgM mAb and IgMspecific secondary antibodieswere used. 2) Specific binding of the mAb to the cells was necessary for activation. 3)No activation was seen unless the cell-bound mAb was cross-linked by a secondary antibody.4) Activation could be reproduced with more than one mAb to each of these CHO Ag, but not withmAb specific for a numberof other Ag including the high density CHO-epitope of CD43 (leukosialin). 5) FMLP-induced decrease in stainingwith anti-CDw17 was followed by a similar decrease in the ability of anti-CDw17 mAb to induce cell activation. 6) Preincubation withmAb didnot alter theoxidative burst or increase in cell surface expression of CD 1 1 induced b by 4 pM FMLP, demonstrating that the cells were not nonspecifically primed by potential contaminants in the mAb preparations. The results, therefore, suggest that CD15, s-CD15, CDwl7, and CDw65 can act as receptor molecules capable of mediating calcium fluxes, degranulation, and strong activation of the oxidative burst in granulocytes. Our results support and extend earlier reportssuggest-

LOG CD11B-FITC Figure 7. Flow-cytometric measurements of cellular CDll b expression after activation with anti-CHO mAb and FMLP. Unlabeled cells and cells labeled with indicated mAb (Ag in parentheses) were incubated 20 min at 37°C with PBS-FCS-Caalone or with PBS-FCS-Cacontaining GAMM or 4 pM FMLP. The cells were then labeled with FITC-anti-CDllb (Bearl).Granulocytes were recognized by light scatter measurements and their CDllb-FITC fluorescence gated to one-parameter histograms. The dotted curue at the extremeleft in the upper panel shows the binding of a FITC conjugated nonspecific control mAb. Histograms are takenfrom the sameexperiment as those shown inFigure 4, and are representative of five experiments.

ing involvement of the CHO Ag in initiation of granulocyte activation. mAb to the CD15 molecule have been shown to inhibit the oxidative burst induced by zymosan particles (17, 19, 22) andto inhibit phagocytosis of IgG- and C-opsonized particles and bacteria (18, 19). Similarly, mAb to CDwl7 and CDw65 have been found to inhibit the zymosan-induced respiratory burst (221, and antiCDw 17-coated bacteria are specifically phagocytosed and halogenated by human neutrophils (41). In addition to these effects, mAb to CD15 have also been found to cause activation of granulocytes, inducing responsessuch as a cytoplasmic calcium flux (20).degranulation (211, cell movement (21), and increased adhesion to endothelium (24).Furthermore,Sullivanetal.(23) reported that F(ab')z fragments of the IgG3 anti-CD15 mAb PMN7C3 (IgG3) could trigger an oxidative burst in GM-CSF-activated granulocytes. Attempts to trigger the oxidative burst by CD15 mAb in unprimed cells have, however, failed (21). The present study demonstrates an oxidative burst inunstimulated cells mediated by CD 15, as well as s-CD15, CDwl7, andCDw65, and supports the view that these Ag are involved in initiating cell activation. The reason for the discrepancy in the results from this and earlier studiescould be that thecombination of IgM mAb

3228


and IgM-specific cross-linking antibodies used in this ture of the ligands for the CHO molecules investigated in study may have caused more extensive aggregation of this study favors the hypothesis that cross-linking is a CHO molecules than has been obtained in earlier studies. physiologic phenomenon that triggerscell activation. This hypothesisis supported by the results showing that In conclusion, the present study demonstrates that mAb alone was without effects, and that the IgM mAb transmembrane signaling in human granulocytes and Mol to CD11b causedcalciumfluxesand oxidative monocytes can be triggeredby cross-linking CHO Ag. The bursts, whereas F(ab')z fragmentsof a n IgG mAb to the results strengthen the view that these molecules a n d same molecule failed to do so. these CHO and the molecules that bear them are imporThe mechanism by which anti-CHO mAb induces cell tant for phagocyte activation. Additional studies using activation remains to be determined. The rapid release the newly characterizedligandsfor CD15,s-CD15, of calcium ions after cross-linking suggests that antiCDwl7, andCDw65 may revealwhether or not the stimCHO mAb may activate phospholipase C as does FMLP ulatory effects of mAb reflect the biologic roles of these (38).The increase in cytoplasmic calcium concentration 4 . is, however, not likely to be responsible for all of the respiratory burst activity, inasmuch as FMLP induced Acknowledgments. The authors wishto thank Dr. calcium fluxes of higher amplitude thananti-CHO mAb, Robert Bjerknes, Prof.Ole Didrik Laerum, andProf. Holm butinducedlessoxidativeburstactivity. In addition, Holmsen for reviewing the manuscript andTor ChristenFMLP caused more extensive degranulation than antisen for help with the preparation of figures. Thanks also CHO mAb. These differences in the response patterns of go to Dr. Lage Aksnes for help with HPLC. We also thank FMLP and anti-CHO mAb indicate that these stimuli may Dr. D. J . Anstee, Dr. B. Axelsson, Dr. C. I. Civin. Dr. M. activate qualitatively different signaling pathways. This C. Dockhelar, and Dr. M. Foreman at Coulter Immunolsubject is presently being investigated. ogy, Mrs. M. Kjaervig Broe at Dako Patts. and Dr. Rosey The mAb to CD15 bind to carbohydrates present on Mushens for their generous gifts of mAb. We are greatly glycolipids and glycoproteins including CD1 l b (42, 43) indebted to Jackson Immunoresearch for making the and CD35 (43, 44). Similarly, mAb to s-CD15bind to work possible through their generous supply of the Affiepitopespresent glycolipids and on the cell adhesion nipure and Affinisorbed secondary antibodies used for as the CDl l b mol- ali immunofluorescence and cross-linking experiments. molecule LECAM-1 (4, 45). Inasmuch ecule was found to be capable of mediating cell activation whencross-linked by Mol a n d GAM", some of the REFERENCES effects seen with anti-anti-CD15 mAb might be ascribed 1 . Huang,L. C., C. I. Civin, J. L. Magnani, J. H. Shaper,and V. to cross-linkingof this glycoprotein. Several lines of eviGinsburg. 1983. My-1 , the human myeloid-specificantigen detected by mouse monoclonal antibodies is a sugar sequence found In lactodence suggest, however, involvementof other structures N-fucopentatose 111. Blood 61:1020. than C receptors in mediating activation by anti-CHO 2. Fukushima. K., M. Hirota, P. Terasaki, A. Wakisaka, H. Togasaki, mAb. 1) No intracellular responses were seen after cross- H. Chia, N. Suyama. Y. Fukushi, E. Nudelman, and S. Hakomori. 1984. Characterization of sialosylated Lewis" as a new tumor-assolinking of CD18 (which is associated with CD1 lb) or ciated antigen. Cancer Res. 44:5279. led to significantly CD35. 2) Cross-linking ofCHOAg 3. Symington, F. W., I. D. Bernstein. and S. Hakomori. 1984. Monoclonal antibody specific for lactosylceramide. J . Biol. Chem. stronger activation of the respiratory burst than cross259:6008. linking of CD1 l b . 3) Priming of cells with FMLP, which 4. Symington, F. W., D. L. Hedges, andS . Hakomori. 1985. Glycolipid down-modulates granulocyte expression of CDw 17 (39). antigens of human polymorphonuclear neutrophils and the inducible HL-60 myeloid leukemia line. J . Immunol. 134:2498. led to a specific decrease in the oxidative burst induced 5. Macher, B. A., J. Buehler, S . W.. W. Knapp. and T. Feizi. 1988. A by anti-CDw 17, thereby strengthening the view that the novel carbohydrate differentiation antigen on fucogangliosides of human myeloid cells recognized by monoclonal antibody VIM-2. J . glycolipid, lactosylceramide, is involved in activation by Biol. Chem. 263:10186. anti-CDw17. Theintracellulareventsobservedafter 6. Larsen, E.. T. Palabrica. S. Sajer. G. E. Gilbert, D. D. Wagner, B. C. cross-linking of CD 15, s-CD 15, CDw 17, andCDw65 may Furie, andB. Furie. 1990. PADGEM-dependentadhesion of platelets to monocytes and neutrophils is mediated by a lineage-specific cartherefore reflect engagement of CD11b as well as glycobohydrate, LNF 111 (CD15).Cell 63:467. lipids and glycoproteins with as yet undescribed func7 Lowe, J. B.. L. M. Stoolman, R. Nair, R. D. Larsen, T. L. Berhend. and R. M. Marks. 1990. ELAM-1-dependentcell adhesion to vascular tions. endothelium determined by a transfected human fucosyl transferase Whereas activating effectsof mAb may indicatea link cDNA. Cell 63:475. between a n Ag and a specific function, the recognition of 8 Phillips. M. L., E. Nudelman, C. A. Gaeta, M. Perez, A. K. Singhal, S . Hakomori, and J. C . Paulson. 1990. ELAM-1 mediates cell adhephysiologic ligands capable of triggering similar cellular sion by recognition of a carbohydrate ligand. sialyl-LeX.Science responses is necessary to classify the molecule as a re250: I i30. ceptor. I t is therefore interesting that the interactionof 9. Brandley, B. K., S. J. Swiedler. and P. W. Robbins. 1990. Carbohydrate ligands of the LEC cell adhesion molecules. Cell 63:861. granulocytes with a physiologic ligand for CHO Ag. the 10. Springer, T. A., and L. A. Lasky. 1991. Sticky sugars for selectins. ELAM- 1 molecule,leads to enhanced adhesive properties Science 349:196. 1 1 . Polley, M. J.. M. L. Phillips, E. Wayner. E. Nudelman, A. K. Singhal, of their membrane integrins (46) similar to what has S . Hakomori, and J. C. Paulson. 1991. CD62 and endothelial cellbeendescribedforanti-CD15 mAb (24). A s CHO-meleukocyte adhesion molecule 1 (ELAM-1) recognizethe same carbodiatedadhesionto ELAM-1 on activatedendothelium hydrate ligand. sialyl-Lewis".Proc. N a t l . Acad. Sci. USA 88:6224. most likelyleads to multivalent engagement of s-CD15 or 12. Tiemeyer, M., S . J. Swiedler, M. Ishihara, M. Moreland.H. Schweingruber, P. Hirtzer. andB. K. Brandley. 1991. Carbohydrate ligands CDw65 on the granulocyte surface, membrane alterafor endothelial leukocyte adhesion molecule-1 . Proc. NQtl.Acad. Sci. USA 88:1 138. tions like those occurring during antibody cross-linking 13. Meuer, S. C., R. E. Hussey, M. Fabbi. D. Fox. 0. Acuto, K. A. maybeinduced.Aggregates ofCHOAg mayalsobe Fitzgerald, J. C. Hodgon, J. P. Protentis, S . F. Schlossmann. and formed when bacteria with binding sites for lactosylcerE.L. Reinherz. 1984. An alternative pathway of T-cell activation: a functional role forthe 50 kd sheep erythrocyte receptorprotein. Cell amide(CDw17) (47)interactwithgranulocytemem36:897. branes or when GMP-140 (CD62) on activated platelets 14. Pardi. R.. J. R. Bender, C. Dettori. E. Gianazza, and E. G. Englemann. 1989. Heterogeneous distribution and transmembrane sigbind to granulocyteCD15 or s-CD15. Therefore, the na-


3229

and A. E. G. K. von dem Borne, eds. Oxford University Press,Oxford, naling properties of lymphocyte function-associated antigen (LFA-1) p. 887. in human lymphocyte subsets. J . Immunol. 143:3157. 36. Borregaard, N., L. Christensen, 0. W. Bjerrum, H. S. Birgens, and 15. Couturier, C.. N. Haeffner-Cavaillon. L. Weiss, E. Fischer, and M. I. G. Clemmensen. 1990. ldentification of a highly mobilizable subset D. Kazatchkine.1990.Induction of cell-associatedinterleukin1 of human neutrophil intracellular vesicles that contains tetranectin throughstimulation of the adhesion-promotingproteins LFA-1 and latent alkaline phosphatase. J . Clin. Inuest. 85~408. (CD1la/CD18)and CR3(CDl lb/CDlS]of human monocytes.Eur. J . 37. Fukushi. Y.,E. Nudelman, S. B. Levery. H. Rauvala. and S. Hakolmmunol. 20:999. mori. 1984. Novel fucolipids accumulating in human adenocarci16. Shimizu, Y., G . A. van Seventer. R. Siraganian, L. Whal, and S. noma. 111 A hybridoma antibody (FH6) defining a human cancerShaw. 1989. Dual role of the CD44 molecule in T cell adhesion and associated difucoganglioside (V13NeuAcV31113Fuc2nLc,). J . Biol. activation. J. Immunol. 143:2457. Chem. 259:10511. 17. Nauseef. W. M., R. K. Root, S. L. Newmann, and H. L. Malech. 1983. Inhibition of zymosan activationof human neutrophil oxidative burst 38. Berridge, M. J..and R. F. Irvine. 1989. Inositol phosphates and cell signalling. Nature. 34 1: 197. by a mouse monoclonal antibody.Blood. 62:635. glycolipid antigen 39. Symington, F. W. 1989.CDwl7:aneutrophil 18. Bjerknes. R.. 0. D. Laerum, and W. Knapp. 1984. Inhibition of regulated by activation. J . Irnmunol. 142:2784. phagocytosis by monoclonal antibodies to human myeloid differen40. Fischer, G. F.. 0. Majdic, and W. Knapp. 1987. Altered expression tiation antigens. Exp. Hematol. 12:856. of surface antigens after stimulation of human granulocytes. In 19. Skubitz. K. M., D. J. Weisdorf, and P. K. Peterson. 1985.Monoclonal Leukocyte Typing Ill: White Cell Dwerentlation Antigens. A. J. antibody AHN-1 inhibits phagocytosis by human neutrophils.Blood McMichael. eds. Oxford University Press, Oxford, p. 698. 65:333. 20. Apfeldorf, W. J., D. A. Melnick, T. Meshulam. H. Rasmussen. and 41. Symington, F.W., and S. Hakomori. 1985. Glycosphingolipids of immune cells. Lymphokines 12:201. H. Malech. 1985. A transient rise in intracellular free calcium is not a sufficient stimulus for respiratory burst activation in human pol- 42. Melnick, D. A., W. M. Nauseef, S. D. Markowitz, J. P. Gardner, and H. Malech. 1985. Biochemical analysis and subcellular localization ymorphonuclearleukocytes. Blochem.Elophys. Res. Commun. of a neutrophil-specific antigen, PMN-7, involved in the respiratory 132:674. burst. J . ImmunoL 134:3346. 21. Melnick, D. A,, T. Mesulham, A. Manto, and H. Malech. 1986. Activation of human neutrophils by monoclonal antibody PMN7C3: 43. Skubitz. K.M., and R. Snook. 1987. Monoclonal antibodies that recognize lacto-n-fucopentaose I11 (CD15) react with the adhesloncell movement and adhesion can be triggered independently from promoting glycoprotein family (LFA-l/HMACl/GP150.95) and CR1 the respiratory burst.Blood. 67: 1388. on human neutrophils.J. Immunol. 139: 1631. 22. Sunder-Plassmann. G., L. Wagner, A. Apperl, K. Hruby, and P. Balcke. 1989.Modulation of neutrophil respiratory burst and chemo-44. Stocks, S. C., M. Albrechtsen, andM. A. Kerr. 1990. Expression of the CD15 differentiationantigen (3-fucosyl-N-acetyl-lactosamine. taxis by the myeloid panel mAb. In Leukocyte Typing IV: White Cell Le") on putative neutrophil adhesion molecules CR3 and NCA-160. Dwerentiatlon Antigens. W. Knapp, B. Dorken, W. R. Gilks, E. P. Eiochem. J . Vol. 268, p . 275. Rieber. R. E. Schmidt, H. Stein, andA. E. G. K. von dem Borne, eds. 45. Picker, L. J., R. A. Warnock. A. R. Burns, C. M. Doerschuck. E. L. Oxford University Press, Oxford. p. 896. Berge, and E. C. Butcher. 199 1. The neutrophil selectinLECAM- 1 23. Sullivan, R., J. P.Fredette, J. D. Griffin, J. L. Leavitt, E. R. Simons, presents carbohydrate ligands to the vascular selectings ELAM-1 and D. A. Melnick. 1989. An elevation in the concentrationof free and GMP-140. Cell. 66:921. cytosolic calcium is sufficient to activate the oxidative burst of gran46 Lo, S. K., S. Lee, R. A. Ramos, R. Lobb, M. Rosa, G. Chi-Rosso. and ulocytes primed with recombinant human granulocyte-macrophage S. D. Wright. 1991. Endothelial-leukocyte adhesionmolecule 1 stimcolony-stimulating factor. J . Biol. Chem. 26433302. ulates the adhesive activity of leukocyte integrin CR3 (CD11b/CDl8, 24. Forsyth, K. D.. A. C. Simpson, andR. Levinsky. 1989. CD15 antimac-l, am82)on human neutrophils.J . Exp. Med. 173: 1493. bodies increase neutrophil adhesion to endothelium by a n LFA-147. Karlsson. K. 1989. Animalglycosphingolipids a s membrane attachdependent mechanism. Eur. J . lmmunol. 19:1331. ment sites for bacteria.Annu. Rev. Eiochem. 58:309. 25. Borche. L.,F. Lozano. R. Y. Vilella. a n d J. Vives. 1987. CD43 4 8 Civin, C. I., J. Mirro, and M. L. Banquerigo. 1981. My-1, a new monoclonal antibodies recognize the large sialoglocoprotein of humyeloid-specific antigen identified by a mouse monoclonal antibody. man leukocytes. Eur. J . lmmunol. 17:1523. Blood. 57:842. 26. Springer, T. A., and D.C. Anderson. 1986. The importance of the 49. Thompson, J. S., S.A. Brown, J. L. Rhoades, J. Burch, E. M. Oberle. Mac-I, LFA-1 glycoprotein family in monocyte and granulocyte adherence. chemotaxis, and migration into inflammatory sites: insights D. A. Jezek, C.D. Jennings, H. Sinn, and C. Spain. 1987. G10F5 118, (Workshop no. 310) reacts with a Pronase resistant epitope whose from a n experiment of nature. In Ciba Foundation Symposium tissue distribution differs from CD15 monoclonal antibodies. In LeuBiochemistry of Macrophages. Ciba Foundation, Raven Press,Lonkocyte Typing 111: White Cell Dwerentiation Antigens. A. J. Mcdon. p. 102. Michael. ed. Oxford University Press, Oxford, p. 713. 27. Aruffo. A.. I. Stamenkovic. M. Melnick, C. B. Underhill. and B. Seed. 1990. CD44 is the principal cell surface receptor for hyaluron- 50 Horejsi, V.. P. Angelisova. V. Bazil. H. Kristofova, S. Stoyanov, I. Stefanova. P. Hausner, M. Vosecky, and I. Hilgert. 1988. Monocloate. Cell. 61: 1303. nalantibodiesagainsthuman leukocyte antigens 11. Antibodies 28. Tonks, N. K., C. D. Diltz. and E. H. Fischer. 1990. CD45. a n integral membrane protein tyrosine phosphatase.J . Eiol. Chem. 26510674. against CD45 (T200). CD3 (T3). CD43. CD10. (CALLA),transferrin receptor (T9), a novel broadly expressed 18 kDa antigen (MEM-43) 29. Minta. A.. J. P. Y. Kao, andR. Y. Tsien. 1989.Fluorescent indicators for cytosolic calcium based on rhodamine and fluorescein chromoand a novel antigen of restricted expression (MEM-74).Folia Blol. phores. J . Biol. Chem. 264:8171. (Prague)34:23. 51. Stockinger, H. 1989. Cluster report: CDw65. In Leukocyte Typing 30. Lund-Johansen, F., J. Olweus, A. Aarli, and R. Bjerknes. 1990. Signal transduction in human monocytes and granulocytes through IV: White Cell Dwerentiation Antigens. W. Knapp. B. Dorken. W. the PI-linked CD14 antigen. FEBSoLett.273:55. R. Gilks. E. P. Rieber. R. E. Schmidt, H. Stein, and A. E. G. K. von dem Borne, eds.Oxford University Press, Oxford, p. 836. 31. Lund-Johansen. F., J. Olweus, A. Aarli. and R. Bjerknes. 1991. The IgG FcRIIand thePI-linked IgG FcRIIItrigger cytoplasmic calcium 5 2 Axelsson, B., A. Kimura. S. Hammarstrom. H. Wigzell, K. Nilsson, fluxes independently in human granulocytes. Scand. J . Immunol. and H. Mellstedt. 1988. Induction of aggregation and enhancement 33:261. of proliferation andIL-2 secretion in human T-cells by antibodies to CD43. J . Immunol. 141:2912. 32. Rothe. G., A. Oser. and G. Valet. 1988. Dihydrorhodamine 123: a 53. Bazil, V., I. Stefanova, I. Hilgert, H. Kristofova. and V. Horejsi. new flow cytometric indicator for respiratory burst activity in neutrophil granulocytes. Naturwissenschaften 75354. 1990. Monoclonal antibodies against human leukocyte antigens. IV. 33. Emmendorfer. A., M. Hecht. M. Lohmann-Matthes, andJ. Roesler. Antibodies against subunits of the LFA-I (CD1la/CD18) leukocyte1990. A fast and easymethod todetermine the production adhesion glycoprotein.Folia Eiol. (Prague)36:41. of reactive oxygen intermediates by human and murine phagocytes using dihy- 54. Todd, R. F.. L. M. Nadler, and s. F. Schlossman. 1981. Antigenson drorhodamine 123.J . Immunol. Methods 131:269. human monocytesidentified by monoclonal antibodies. J . lmmunol. 34. Tetteroo. P. A. T., M. J. E. Bos, F. J. Visser, and A. E. G . K. von 126: 1435. demBorne.1986. Neutrophil activationdetected by monoclonal 55. Knapp, W. 1989. Myeloid section report. In Leukocyte Typing IV: antibodies. J . lmrnunol. 136:3427. White Cell Dwerentiation Antigens. W. Knapp. B. Dorken. W. R. 35. van der Schoot, E., T. W. J. Huizinga. S. Gadd, 0. Majdic, R. Gilks, E. P. Rieber. R. E. Schmidt, H. Stein, and A. E. G. K. von dem Wijmanns, W. Knapp, and A. E. G. K. vondemBorne.1989. Borne. eds. Oxford University Press, Oxford, p. 747. Identification of three novel PI-linked proteins on granulocytes. In 56. Stefanova, I., I. Hilgert, V. Bazil, H. Kristofova, and V. Horejsi. LeukocyteTyping IV: WhiteCellDflerentiationAntigens. W. 1989. Humanleukocyte surfaceglycoprotein CDw44 and lymphocyte Knapp. B. Dorken. W. R. Gilks, E. P. Rieber. R. E. Schmidt, H. Stein, homing receptorare identical molecules. Immunogenetics 29:402. ~