The Anti-ganglioside Monoclonal Antibody AA4 Induces Protein ...

3 downloads 38 Views 6MB Size Report
William D. SwaimS, Kenji Minoguchi, Constance Oliver, Majed M. Hamawy, Hidetoshi Kihara,. Volker Stephan, Elsa H. Berenstein, and Reuben P. Siraganian.
THE JOURNAL OF BIOLWICAL Cm~lsmv

Vol. 269, No. 30,Issue of July 29, pp. 1946C19473, 1994 Printed in U.S.A.

The Anti-ganglioside Monoclonal Antibody AA4 Induces Protein Tyrosine Phosphorylations,but Not Degranulation, in Rat Basophilic Leukemia Cells* (Received for publication, January 27, 1994, and in revised form, April 25, 1994)

William D. SwaimS, Kenji Minoguchi, Constance Oliver, Majed M. Hamawy, Hidetoshi Kihara, Volker Stephan, Elsa H. Berenstein, and Reuben P. Siraganian From the Laboratory of Immunology, NIDR, National Institutes of Health, Bethesda, Maryland 20892

The monoclonal antibody AA4 (mAb AA4) recognizes novel a-galactosylderivatives of G,,, on rat basophilic leukemia (RBL-2H3) cells. The binding of mAb AA4 induced protein tyrosine phosphorylations without histamine release. Several of the same proteins including FccRIP, FccRIy, p72wh,and phospholipase C-yl were tyrosine-phosphorylatedby mAb AA4 binding and by the activation of the high affinityIgE receptor, FccRI. There and p72wh proteinwas also activation of the ~53156'~" tyrosine kinases, but compared to direct FceRI activation, mAb AA4 did not result in increased tyrosine phos~ ~ the , receptor phorylation of pp105-115 or p ~ 1 2 5 and subunits (FccRIP and FceRIy) were more heavily phosphorylated. Furthermore, the time course of the phosphorylations with mAb AA4 was slower than that induced by FccRI aggregation. By immunofluorescence, the tyrosine-phosphorylated proteins after mAb AA4 stimulation were localized in patches at the cell membrane and in areas of cell-cell contact, whereas after FccRI activation, there was a reticular cytoplasmic pattern. There were no protein tyrosine phosphorylations either whenFccRIwas saturated with IgE or when F(ab'), fragments of mAb AA4 were used, although the F(ab'), fragments still induced morphological changes. There was alsocoprecipitation of the P and y subunits of FccRI with the anti-ganglioside antibody. These data strongly suggest the involvement of FceRI in the antiganglioside-induced protein tyrosine phosphorylations. Moreover, phosphorylations of these proteins including the P and y chains of FccRI and activation of p53156'Y" and ~ 7 2 " did ~ ' not result inhistamine release.

basophils or mast cells (12-20). One group of proteins is tyrosine-phosphorylated rapidly after FccRI aggregation independently of the rise in intracellular calcium or the activation of protein kinase C. These include the p and y subunits of FceRI, phospholipase C-yl, p53/56'Y", p95""", p72Wk, and other 72-kDa proteins (pp72) (13, 16, 18, 19, 21, 22). Another group of 105115-kDa proteins, including the focal adhesion kinase p ~ 1 2 5is~ tyrosine-phosphorylated ~ , at later stages of cell activation (23-25). These phosphorylations are the resultof the increase in intracellular calcium andlor the activation of protein kinase C (23-25). Tyrosine phosphorylation of these proteins depends on adherence of the cells to extracellular matrix proteins (24, 25). Experiments suggest that the phosphorylation of the 105-115-kDa proteins is required for optimal degranulation of the cells. mAbl AA4 recognizes two unique a-galactosyl derivatives of GDlbthat are cell-type specific and present only on RBL-2H3 and rat mast cells (26, 27). These gangliosides are found close to FceRI, in 10-15-fold higher molar concentrations than FceRI (28, 29). The bindingof either intact mAb AA4 or its Fab fragment inhibits the binding of IgE to FceRI (28). Recently, we observed that these ganglioside derivatives of GDlbstrongly associate with the Src family tyrosine kinase p53156'Y" and a serine kinase. This association of Lyn with the gangliosides is not mediated by FceRI (30). Although the binding of anti-ganglioside mAb AA4 to RBL2H3 cells results in morphological changes similar to those induced by FccRI aggregation, it does not induce histamine release (27). mAb AA4 also causesa small increase in intracellular calcium, some phosphoinositide hydrolysis, and redistriThe aggregation of the high affinity IgE receptor, FccRI, on bution of protein kinase C (27). However, the magnitude of mast cells or basophils results in complex biochemical and mor- these changes is less than after FccRI aggregation. Here, we phological events. These include the activation of phospho- report that the binding of mAb AA4 also results in protein calcium; stimulation tyrosine phosphorylations. These mAb AA4-induced phospholipases C, 4,and D; a rise in intracellular of protein kinase C; and proteinphosphorylations (1-5). There rylations involve the participation of FccRI. mAb AA4 copreare also striking morphological and cytoskeletal changes (6-8). cipitated FccRI proteins, further supporting observations on These biochemical events result in the secretion of granular the close association of the gangliosides recognized by this anphosphorylations contents, in the generation of newly synthesized lipid media- tibody with the receptor. The protein tyrosine differences comtors, and in the synthesis and release of several lymphokines induced by mAb AA4 had both similarities and (9). Protein tyrosinephosphorylations play an important role in pared with those produced by direct FceRI aggregation. These phosphorylation could indicate a receptor-mediated signaling (10, 11).Multiple proteins are ty- differences in protein tyrosine is rosine-phosphorylated in stimulated rat basophilic leukemia point of divergence in the signal transduction pathway that RBL-2H3 cells, which are an in vitro model for the study of critical for secretion.

* The costs of publication of this article were defrayed in part by the be hereby marked payment of page charges. This article must therefore "uduertisement"in accordance with 18 U.S.C.Section 1734 solely to indicate this fact. $ To whom correspondence should be addressed: Lab. of Immunology, Bldg. 10, Rm. 1N106, NIDR, NIH, Bethesda, MD 20892. Tel.: 301-4965105; Fax: 301-496-2443.

'The abbreviations used are: mAb,monoclonalantibody; GDlb, Galp1-3GalNAc~1~4[NeuAca2~8NeuAca2-,3lGalp~4Glc~1-1Cer; GD3, NeuAca2-8NeuAca2-3Galp1-4Glcpl1Cer; BSA, bovine serum albumin; PBS, phosphate-buffered saline; PAGE, polyacrylamide gel electrophoresis;Tricine, N-[2-hydroxy-l,l-bis(hydroxymethyl)ethyllglycine.

19466

mAb AA4-induced Protein Pyrosine Phosphorylation

19467

the membranes were blocked fora minimum of 4 hin TTBS (10 mM Tris, 0.9% NaCl, and 0.05%Tween 20, pH 7.2) containing 4% BSA. Blots were Antibodies-The IgGl anti-ganglioside antibody (mAbAA4) has been probed with horseradish peroxidase-conjugated anti-phosphotyrosine described previously (28). For these experiments, it was purified from antibody (1:25,000 PY20-HRP; ICN) in TTBS + 0.5% BSAfor 2 h. They ascites fluid by ammonium sulfate precipitation followed by ion-ex- were then washed twice with TTBS + 0.5% BSA followed by extensive change chromatography on DE52. In some experiments, it was further washing with TTBS. Detection was by the enhanced chemiluminespurified by high pressure liquid chromatography on a gel filtration cence method (ECL kit, Amersham Corp.). The membranes were column (TSK 3000 SWG, 21.5 mm, inner diameter, x 60 cm). Ffab'), stripped as recommended by the manufacturer and reprobed with the fragments of mAb AA4were generated by pepsin digestion and purified immunoprecipitation antibody to verify the equal precipitation of proby high pressure liquid chromatography. teins. Anti-Lyn, anti-Syk, and anti-FccRIy were visualized with horseMonoclonal anti-dinitrophenyl IgE (DNP48 BC2), anti-trinitroradish peroxidase-conjugated donkey anti-rabbit antibody. Donkey phenyl IgE (TNP-142), mAb BGD6 (IgGl), mAb AD1 (IgGl), and the anti-mouse antibodies were used for the FceRIP and phospholipase C-yl anti-FceRI antibodies (mAb BC4to the a subunit and mAb BE to the P immunoblots. chain) havebeen described previously (28,31-34). Anti-G,, monoclonal Zbo-dimensional Gel Electrophoresis-After stimulation, cells were antibody R24 (35) was kindly provided by Dr. Lloyd J. Old (Memorial solubilized in urea lysis buffer (9.5 M urea, 2% Nonidet P-40, 5% preSloan-Kettering Center, New York, NY). Polyclonal rabbit anti-Lyn and blended pH 3-10 ampholytes, 1 mM Na,VO,, 5% 2-mercaptoethanol). anti-phosphotyrosine antibodies were prepared and purified as previ- Postnuclear lysates were separated by two-dimensional gel electrophoresis according to the method of O'Farrell(40). For the first dimenously described (30, 36, 37). Anti-FceRIy and a n t i - ~ 7 2have ~ ~ been described previously (34). a n t i - p ~ l 2 and 5 ~ ~anti-phospholipase C-yl sion, 40 p1 of the lysates (-400,000 cell eq) were separated in 2.4-mm (inner diameter) tubesby isoelectric focusing at 17 "C for 12,000V-h (16 were obtained from Upstate Biotechnology, Inc. (Lake Placid, NY). All other antibodies were purchased from Jackson ImmunoResearch Labo- h at 750 V). Following equilibration in SDS buffer (124 mM Tris, 5% glycerol, 2.1% SDS, 2% 2-mercaptoethanol, pH 6.8) for 30 min a t room ratories, Inc. (West Grove, PA). Where indicated, antibodies were coupled to CNBr-activated Sepharose 4B as recommended by the manutemperature, the cylindrical gels were annealed to the top of a 6 or 9% polyacrylamide slab gel (10.5 x 14.0 cm) and electrophoresed for -3 h. facturer (Pharmacia Biotech Inc.). Buffers and Solutions-The following buffers were used: SDS/Triton A laneof whole lysate was run next tothe standardlane for comparison. lysis buffer (10 mM Tris-HC1, pH 7.4, 100 mM NaCl, 50 mM NaF, 1% After separation, proteins were electrotransferredto nitrocellulose Triton X-100, 0.5% sodium deoxycholate, 0.1% SDS), 0.5% Triton lysis membranes, blocked in 4% BSA, and immunoblotted with horseradish buffer (50 ~lu Tris, l pH 7.5, 150 mM NaC1, 0.5% Triton X-loo), and 3% peroxidase-conjugated anti-phosphotyrosine antibody (PY20-HRP),and the proteins were visualized as described above. Brij 96 lysis buffer (0.2 M borate, pH 8.0, 0.15 M NaCl, 3% Brij 96). The ImmunofluorescenceMicroscopy-Cells were plated a t 2.5 x lo5 cells/ following protease and phosphatase inhibitors were added to each of well in 6-well plates on sterile glass coverslips and cultured for 48 h, these buffers: 1mM Na,VO,, 1mM phenylmethylsulfonyl fluoride, 3.0 p~ aprotinin, 23 w leupeptin, and 1.5 w pepstatin A. and the medium was replaced after 24 h. The monolayers were rinsed Cell Culture and Stimulation-RBL-2H3 and B6A4C1 cells were twice with warm 10 mM Hepes-buffered Eagle's minimal essential memaintained in culture as monolayers (38). B6A4C1 is a variant of the dium containing 1%BSA, pH 7.4, and incubated for the appropriate RBL-2H3 cell line that has decreased expression of the a-galactosyl interval with this medium alone or with different antibodies. Following stimulation, cells were rinsed twice in PBS containing 1 mM Na,VO,, derivatives ofG,,,, but unchanged numbers of FceRI (29, 30). For stimulation, 10' cells in 7 ml of culture medium were plated in 10-cm fured for 1h with 2% formaldehyde (electron microscopy-grade),rinsed tissue culture plates. After overnight culture, the cell monolayers were again in PBS/Na,VO,, and incubated in PBS/Na,VO, with 0.1 M glycine washed twice with Eagle's minimal essential medium with Earle's salts for 10 min, all at room temperature. Cells were then incubated in cold containing 1% BSA and 10 mM Hepes, pH 7.4, at 37 "C. The cells were methanol for 6 min at -20"C, returned to room temperature, and then stimulated in thesame medium at 37 "C with mAb AA4 (10pg/ml) washed twice in PBS/Na,VO, followed by incubation for 2 hwith rabbit or mAb BC4 (0.1 pg/ml). For IgE-mediated cell activation, the cells were anti-phosphotyrosine antibody (10 pg/ml) in PBS containing 1%BSA. cultured overnight with a 1:5000 dilution of the anti-trinitrophenyl IgE After two washes with PBS + 1%BSA, the cells were incubated with ascites fluid and stimulated with 0.1 pg/ml dinitrophenyl coupled to fluorescein isothiocyanate-conjugated F(ab'), donkey anti-rabbit IgG at human serum albumin. Following stimulation, the supernatants were a 1:50 dilution in PBS + 1%BSAfor 1h andthen washed five times with aspirated, centrifuged, and analyzed for histamine content. Cells were PBS followed bythree rinses in distilled water. The slides were mounted washed twice with warm (37 "C) PBS containing 1 mM Na,VO,; then, in Fluoromount (Fisher), and cells were photographed using a Nikon ice-cold lysis buffer was added, and theplates were incubated on ice for Microphot-FX epifluorescence microscope under 40 x or 60 x oil immer10 min. After centrifugation, the protein concentrations of the supersion objectives. natants were determined using the bicinchoninic acid method (Pierce Presentation ofResults-Immunoblots show typical results of experiChemical Co.). In experiments to deplete extracellular Ca", the cell ments repeated at least three times unless indicated otherwise. In all monolayers were washed with calcium-free Eagle's minimal essential cases, similar results were obtained in all experiments. medium (Life Technologies,Inc.) and stimulatedin thismedium supplemented with 0.5-10 mM EGTA. RESULTS Immunoprecipitation-Cell lysates (750 pg of protein) were incuAnti-ganglioside mAb AA4-induced Protein Tyrosine Phosbated with the different antibodies prebound on beads then washed extensively (six times), and the proteins were eluted by boiling in phorylations-Because the binding of mAb AA4 to RBL-2H3 Laemmli sample buffer (39). Cell lysates were in the buffer conditions cells results in biochemical and morphological changes similar that were found to be optimal for each immunoprecipitation. The anal- t o thoseinduced by FceRI aggregation, we investigated ysis of ~ ~ 1 was2 as5described ~ ~ previously (25). whether it also induces proteintyrosine phosphorylations. In Vitro Kinase Reactions-Cell lysates (250 pg of protein) were precleared with protein A beads then incubated with either anti-Lyn or RBL-2H3 cells were incubated with mAb AA4, and the lysates were then analyzed by anti-phosphotyrosine immunoblotting anti-Syk antibody for 1 h, and the complex was immunoprecipitated with protein A beads for 1h at 4 "C. SDS/Tritonlysis buffer was used for (Fig. lA). There was strong tyrosine phosphorylation of prothe anti-Lyn experiments, and 0.5% Triton lysis buffer for the anti-Syk teins of 140,78, 72,61, and 52 kDa. In contrast, normalmouse experiments. Immunoprecipitates were washed three times with lysis IgG and two antibodies (AD1 and BGD6) that bind to surface buffer; once with 50 mM Tris, 150 mM NaCl, pH7.4; and once with glycoproteinson these cells (31, 32) did not induce protein kinase buffer (30 mM Hepes, pH 7.5, 10 II~MMgCI,, and 2 f l l ~MnCl,); tyrosine phosphorylations. Moreover, anti-G,, mAb R24 bound and were resuspended in 20 pl of kinase buffer. The reaction was then started by the addition of 5 pCi of [Y-~'P]ATPfor 10 min at room t o RBL-2H3 cells, but did not induce protein tyrosinephosphotemperature, stopped by the addition of 20 pl of 2 x Laemmli sample rylations(datanotshown).In dose-response experiments, buffer, and boiled for 5 min (39). The eluted proteins were separated by phosphorylation wasdetectable at 1 pg/ml mAb AA4 and 10% SDS-PAGE under reducing conditions, electrotransferred to Im- reached maximum levels at 10 pg/ml. With as little as 1pg/ml mobilon P membranes, and visualized by autoradiography. SDS-PAGE and Immunoblotting-Cell lysates or immunoprecipi- mAb AA4, there was tyrosine phosphorylation of only a 140tated samples were separated electrophoretically on 10% gels (39). The kDa protein, followed by the appearance of 52-, 61-, 72-, and 78-80-kDa bands with 5 pg/ml mAb AA4. In time course exgels were equilibrated in carbonate buffer, and theproteins were transferred to nitrocellulose membranes overnight at -5 "C. After transfer, periments, mAb AA4-induced tyrosine phosphorylations were MATERIALSANDMETHODS

mAb AA4-induced Protein

19468 A.

B.

0 kc3

Sr6?LJ3

49.5-

- l-AA4-+

(D

o z q m d a m

10680-

Tyrosine Phosphorylation

7

-"

bzs

Min

30

p g *

2 5 10 15 30 15 15

106 80 49.5

-

rrd rw.e

32.5-

L.

"

27.5-

32.5 .".

Probe: Anti-PY

Probe: Anti-PY

FIG.1.Anti-ganglioside mAbAA4-induced protein tyrosine phosphorylations. A, RBL-2H3 cells were incubated for 30 min a t 37 "C with medium alone (Control),normal mouse IgG (NMG), mAb A D 1 , mAb BGDG, mAb AA4, antigen ( A g ) , or anti-FceRI mAb BC4. B , shown is thetime course of protein tyrosine phosphorylations induced by anti-ganglioside mAb AA4. Cells were incubated with mAb AA4 (10pg/ml) for the indicated times a t 37 "C. Cell lysates (40 pg of protein) were separated by 10% SDS-PAGE and analyzed by immunoblotting with anti-phosphotyrosine antibodies (Anti-PY).Control antibodies were tested twice.

detectable as early as 2 min, slowly increased through the 30-min time point (Fig. lB), and were unchanged a t 1h (data not shown). Moreover, the tyrosine phosphorylation pattern was the samewhether or not Ca2+was present in the extracellular medium (mAb AA4 stimulation in medium containing 0.5-10.0 mM EGTA). Although mAbAA4 binding induced prominent protein tyrosine phosphorylations, it did not result in histamine release. The pattern of phosphorylated proteins with mAb AA4 was remarkably similar tothat induced by the aggregation ofFceRI (Fig. 1). It was therefore important to demonstrate that themAb AA4-induced tyrosine phosphorylations were not due to contamination with IgE aggregates or anti-receptor antibodies. Five different preparations ofmAb AA4, four of which had been purified from separate ascites, still induced these phosphorylations. Further purification ofmAb AA4by high pressure liquid chromatography to remove any trace contamination of IgE aggregates or monomeric IgE did not change the protein tyrosine phosphorylations. None of the mAb AA4 preparations induced histamine release, again demonstrating that therewas no direct FceRI aggregation. Therefore, these phosphorylations were not due toIgE or anti-receptor antibody contamination of mAb AA4. By two-dimensional gel electrophoresis, the proteins tyrosine-phosphorylated dueto FceRI aggregation, whetherinduced by IgE-antigen or by anti-receptor antibody BC4, were very similar to thoseresulting from mAb AA4 binding (Fig. 2). There were some minor differences. The signal withmAb AA4 was of a lesser intensity, although the dissimilarity in the5080-kDa range was lessdramatic upon longer exposures. There were also several proteins that appeared to be tyrosine-phosphorylated only after IgE-antigen or anti-receptor aggregation. Thus, although mAb AA4 did not result in degranulation, the pattern of tyrosine-phosphorylated proteins induced by mAb AA4 was similar to thoseresulting from FceRI aggregation. Involvement of Fc&I in mAb AA4-induced Protein Tyrosine Phosphorylations-mAb AA4 does not directly bind to FceRI on cells or to any membrane proteins on immunoblots (291, and it did not bind to a Chinese hamster ovary cell line that stably expressed FceRIa, FceRIP, and FceRIy on its cell surface (data not shown). Nevertheless, previous experiments had demonstrated a close spatial and physical relationship between the ganglioside-binding sites of mAb AA4 and FceRI (29). For example, both intact mAb AA4 and its Fab fragments inhibit IgE binding, although IgE does not inhibit mAb AA4 binding (28, 29). Therefore, we investigated if FceRI proteins were coprecipitated with the anti-ganglioside antibodies. In Brij 96 cell

lysates, a small amount of the /3 and y chains of FceRI was immunoprecipitated by mAb AA4 (data not shown). To eliminate possible Fc interactions with FceRI, F(ab'), fragments of mAb AA4were then used for immunoprecipitations from Triton X-100-solubilized cells. Both the /3 and y subunits of FceRI were in immunoprecipitates from RBL-2H3 cells, but not in those from B6A4C1 variant cells that haddecreased mAb AA4-binding sites (Fig. 3). The fraction of FceRI coprecipitated with F(ab'), was -2% in Triton X-100, but slightly less in Brij 96 (data not shown). In immunoprecipitations with anti-Lyn antibodies, as has been reported previously (16), there was a small amount of FceRIP and FceRIy associated with p53/56'Yn. These results suggest that the a-galactosyl derivatives of G,,, recognized by mAb AA4 and p53/56'Y" associate weakly with FceRI. We then investigated the possible role of FceRI in the mAb AA4-induced protein tyrosine phosphorylations. The mAb AA4induced protein tyrosine phosphorylations were completely blocked by the saturation of FceRI with IgE, although these cells responded normally to antigen-induced receptor aggregation (Fig. 4A).Furthermore, by immunofluorescence, mAb AA4 bound to the IgE-saturated cells. These results demonstrate that protein tyrosine phosphorylations induced bymAbAA4 require an unoccupied FceRI. The inhibition of protein tyrosine phosphorylation by saturation ofFceRI with IgE raised the possibility that the Fc portion of mAb AA4 could be implicated in these phosphorylations. Although F(ab'), fragments ofmAbAA4 bound to the RBL-2H3 cells and inhibited lZ5I-IgEbinding (data not shown), they did not induce any protein tyrosine phosphorylation (Fig. 4B 1. There was also no phosphorylation even after theaddition of secondary rabbit anti-mouse Igto cross-link the F(ab'), fragments. However, the F(ab'), fragments still induced morphological changes similar tothose induced by the intact antibodies, i.e. there was spreading and ruffling of the cells (data not shown). These results demonstrate that theFc portion of mAb AA4 is critical for the induction of protein tyrosine phosphorylation and that thereis an involvement of FceRI. Furthermore, the aggregation of the gangliosides alone is capable of inducing cytoskeletal rearrangements and morphological changes. The aggregation of FceRI results in rapid tyrosine phosphorylation of the p and y subunits that areassociated with crosslinked receptors (13). The binding of mAb AA4induced tyrosine phosphorylation of both the p and y subunits that was much greater than with direct FceRI stimulation (Fig. 5). m o s i n e phosphorylation of the p subunit was detectable at 2 min and

mAb AA4-induced Protein Qrosine Phosphorylation

8.2

-

IEF-

4.7

- 205

T

19469

8.2

4.7

BC4 8.2

4.7

-

- 116 - 106 -

w (3

Qa v)

-

n

cn

80

-

- 49.5 Control 4.7

8.2

- 205 -

.

-

- 116 - 106 - 80 -

- 49.5 AA4

Ag

Probe: Anti-PY FIG.2. Two-dimensional analysis of tyrosine-phosphorylatedproteins. Cell lysates from control cells (Control)or from cells stimulated for 30 min at 37 "C with anti-FccRI mAb BC4, mAb AA4, or antigen ( A g ) were separatedby isoelectric focusing (ZEF)followed by SDS-PAGE and analyzed by immunoblotting with anti-phosphotyrosine antibodies.

AA4-stimulated cells, tyrosine phosphorylation of the y chain persisted even at 30 min, whereas therewas a decrease in the signal with antigen. mAb AA4-induced tyrosine phosphorylation of the p and y subunits further supports the concept of the involvement and activation of FceRI. Phosphorylation and Activation of Protein-tyrosine Kinases by mAb AA4 and Comparison to Fc&I Aggregation-Several protein-tyrosine kinases in RBL-2H3 cells are tyrosine-phosphorylated and/or activated after FceRI aggregation (16,19,25, is with FceRI and is FIG.3. Association of gangliosides with FceRI. RBL-2H3 or 34). The Src-related ~53156'~~"associated Recently, we also B6A4C1 cells were solubilized in 0.5% Triton lysis buffer; immunopre- activated after receptoraggregation(16). cipitated with 5 pg of F(ab'), fragments of mAb AA4, anti-Lyn antibod- found that p53156'Y" associates with thegangliosides recognized ies, or rabbit IgG (rlgG);separated by SDS-PAGE; and analyzed by by mAbAA4 (30). The binding of mAbAA4 resulted in an immunoblotting with anti-FceRIB(upper panel) or anti-FccRIy (lower panel) antibodies. The immunoprecipitation was from 5 x lo6 cells by increase in the i n vitro kinase activity of p53156'Y" (Fig. 6A). the F(ab'), fragments of mAb AA4 and from 20 x IOficells by anti-Lyn This activation was much slower than that induced by FceRI antibodies. Total lysates from 2 x IO5 cells were also loaded in the aggregation: with antigen,Lyn was activated a t 1min and had indicated lane. IP, immunoprecipitate. returned to resting levels by 5 min, whereas with mAbAA4, the activation was present a t 10min and had not returned to progressively increased up to30 min. At15 min, thesignal was resting levels even after 20 min. Although there wasactivation greater withmAb AA4than in the receptor-activated cells. The of ~53/56'~", there wasno detectable change in its tyrosine phosphorylation of the y subunit was also much stronger with phosphorylation with either FceRI aggregation or mAbAA4 mAb AA4 than with receptor activation (Fig. 5). In the mAb binding (data not shown). Nevertheless, both FceRI aggrega-

mAb AA4-induced Protein

19470

Qrosine Phosphorylation B.

A.

AA4 L

0)

6

0

106-

..

."

80 -

80-

-

32.5-

32.5

-

27.5-

27.5

-

Probe:

Anti-PY

.-

"

nIIImaq

49.5

49.5-

Anti-PY

"

106 -

Probe:

FIG.4. Role of FcsRI in mAb AA4-inducedprotein tyrosine phosphorylations. A, tyrosine phosphorylations wereblocked by saturation of FccRI with IgE. RBL-2H3 cells were cultured for 16 h with or without 10 pg/ml monoclonal anti-dinitrophenyl IgE. The cells were then washed twice and stimulatedfor 30 min. NMG,normal mouseIgG;AA4, mAbAA4 stimulation in nonsensitized cells; ZgEIAA4, cells cultured with10 pglml IgE and stimulated with mAbAA4; ZgE/AA4/Ag,mAbAA4 added to the IgE-sensitized cells for 10 min followed by antigen for 20 min; Ag, antigen; BC4, stimulation with anti-FccRImAb BC4. B, F(ab'), fragments of mAb AA4 did not induce protein tyrosine phosphorylations. Whole cell lysates were analyzed by immunoblotting with anti-phosphotyrosine antibodies (Anti-PY).Similar results were obtained with two different preparations of F(ab'), fragments of mAb AA4 tested intwo different experiments.

IP: Anti FceRI-p

- Ag BC4

1A - A4{-

Time(min)

2

5 10 15 2030

30 15 15

1068049.5-

32.527.518.5-

tion and mAb AA4 induced equal activation of ~53156'"". A protein prominently phosphorylated after FceRI aggregation is pp72, a component of which is p72Vk(12, 18-20,34,41). Intheseexperiments,therewas also a 72-kDa tyrosine-phosphorylated protein. By in vitro kinase assays, there was stronger activation of p72"ykafter FceRI aggregation than with mAb AA4 (Fig. 6B). With FceRI stimulation, peak activation was within 5-10 min, whereas mAb AA4 binding induced a slow increase inactivity that reached a maximum by 20 min. There was much lessactivation of p72wk by mAb AA4 than by direct FceRI aggregation. By immunoprecipitation followed by phosimmunoblotting, photyrosine p 7 P k was tyrosine-phosphorylated, although the time course for this phosphorylation was with slower that than activation receptor direct (Fig. 6C).

Probe: Anti-PY

IP: Anti-FceRI-y

+AA4 +

Time (min)

5 10 20 30

14.3

-+Ag"--I 30 5 10 20 30

-

5Probe: Anti-PY

msine

5' phosphoQ'lation Of p and 7' subunits Of FcsR1 by either anti-gangliosidemAb AA4 or FcsRI aggregation. Cell lysate (750-1000 pg of protein) was immunoprecipitated with antiFccRIp or anti-FccRIy, separated by SDS-PAGE (14% gels for p and 16% %cine for y), electrotransferred, and immunoblotted with antiphosphotyrosineantibodies (Anti-PY). The FceRIp lysateswereprepared inSDSPPriton lysis buffer, whereas FccRIy preparation was in 3% Brij 96 lysis buffer. Therewerecomparableamounts of theproteins immunoprecipitated from mAbAA4- and mAb BC4-stimulatedcells a s assessed by reprobingthemembraneswithanti-FceRIPand anti-FccRIy antibodies (data not shown). IP,immunoprecipitate, A g , antigen.

Thus, the time course of the changes in both Lyn and Syk is slower than that with directFceRI aggregation. Anti-ganglioside mAb AA4-induced Tyrosine Phosphorylation of Phospholipase C-yl-The aggregation of FceRI results in tyrosine phosphorylationof the phospholipase C-yl thatmay play a role in phosphoinositide hydrolysis (21, 42). The morphological changes induced mAb byAA4 are accompanied by a slight increase in intracellular calcium and a small increase in phosphatidylinositol hydrolysis (27). With mAb AA4, tyrosine phosphorylation of phospholipase C-yl wasbarely detectableat 10 min, but steadily increased from 20 to 30 min (Fig. 7). In contrast with direct FceRI aggregation, this reached a maximum by 10 min and was followed by a steady decrease. Lack of Tyrosine Phosphorylation ofppIO5-115 and ~ ~ 1 by A b AAC-Changes in tyrosine phosphorylation of a group of 105-115-kDa proteins including the focal adhesion tyrosine kinase ~ ~ 1occur 2 a5t later ~ stages ~ of cell activation and may be required for optimal degranulation (12, 24, 25). Cell adherence and the aggregation of Fc&I synergistically regulate tyrosine phosphorylation of these proteins. Unlike of the

2

other substrates, the phosphorylation of these proteins requires extracellular calcium and is induced by calcium ionophores andlor the activation of protein kinase C. There was no increase in tyrosine phosphorylation ofpp105-115 or p p 1 2 5 ~ ~ after mAb AA4 stimulation of RBL-2H3 cells (Fig. 8). Thus, changes intyrosine phosphorylationof ~ ~ 1may 2 represent 5 ~ ~ ~

5

~

mAb ~ ~ 4 - t n u u c Yrotesn eu I yrossne rnosphorylatson A.

IP: Anti-PLC-y,

Lyn Kinase Assay - t " g - i

Time (min.)

20 1

- I- Ag -I- AA4 "I 30 10 20 10 20 30

-t"4+

5 10 20

1

20

warn--

I Y 4I I

5

10 20

97

mmw"

&A

.J-

106-

80.

--

Probe: Anti - P Y FIG.7. Anti-ganglioside mAb AA4-induced tyrosine phosphorylation of phospholipase C-yl. Immunoprecipitation wasfrom 1000

Syk Kinase Assay

B.

- F A g +

Time (min.)

30

1

30

5 10 20 30

1

5

10 20 30

Syk Tyrosine Phosphorylation

C.

--1 Time (min.)

M

+ "1

30 5 10 20 30 5

A A ~

pg of protein of lysate prepared in SDS/Triton lysis buffer. The proteins were separated by 8% SDS-PAGE and analyzed by anti-phosphotyrosine immunoblotting. There were comparable levels of phospholipase C-yl immunoprecipitated as assessed by reprobing the membranes with anti-phospholipase C-yl antibody (data not shown). IP, immunoprecipitate; Ag, antigen; Anti-PY, anti-phosphotyrosine antibodies.

"- _+ 1

+ -

2 13 24 3

-

.-

- + + - - - +

Ag

+

AA4

4

.-.I

- 200 -

.,

=gay

-97.4-

10 20 30

" "

- 69 - 46 -

FIG.6. mAb AA4-induced activation and or protein tyrosine phosphorylation of protein kinases p53/56'Yn and ~'72". Cells

30

-

" "

Cell lysate

~ 1 2 precipitate 5 ~ ~ ~

were either stimulated by IgE-antigen ( A g ) or incubated withmAbAA4, Probe: Anti-PY and cell lysates were preparedat the times indicated and immunopreFIG.8. Anti-ganglioside mAb AA4 does not induce protein tycipitated with either anti-Lyn ( A )or anti-Syk (Band C) antibodies. Part 5 were ~ ~ . of each sample was used for in vitro kinase reaction( A and B 1, and the rosine phosphorylation of pp105-115 or ~ ~ 1 2 Cells stimulated with two different concentrations of antigen ( A g ) or with rest was used for immunoblotting with anti-phosphotyrosine antibodies mAb AA4 for 30 min. The whole cell lysates and the pp125F.AK immu(C). There was comparable amounts of the protein-tyrosine kinases immunoprecipitated from the mAb AA4- and mAb BC4-stimulated cells noprecipitates were analyzedby anti-phosphotyrosine immunoblotting. a s assessed by immunoblotting of the same samples with anti-Lyn and Lane 1, nonstimulated control cells; lane 2, IgE-sensitized cells plus antigen (0.1ng/ml); lane 3, IgE-sensitized cells plus antigen ( 1 ng/ml); anti-Syk antibodies (data not shown). lane 4 , mAb AA4 (10pg/ml)-stimulated cells. Anti-PY, anti-phosphotya divergence in the signaling pathwaybetween direct receptor rosine antibodies.

activation and mAb AA4 binding. Intracellular Distribution of Qrosine-phosphorylated Proteins Depends upon Stimulus-Immunofluorescence was used to study the distribution of tyrosine-phosphorylated proteins in cells stimulated with mAb AA4 compared with those stimulated by FceRI aggregation (Fig. 9). In nonstimulated cells, there was only diffuse intracellular staining of tyrosine-phosphorylated proteins that appeared not beto associatedwith the plasma membrane. After mAb AA4 stimulation, the tyrosinephosphorylated proteins were localized in long portions of cellcell contacts as well as in patches a t t h e cell membrane. In contrast, in IgE-antigen- oranti-FceRI-stimulated cells, there was staining of small intracellular vesicles with a reticular pattern and much less staining in ofareas cell-cell contact. This pattern progressively became stronger between 5 and 15 min and was stillvisible, although somewhatdecreased, at 30 min. The strong patchy localization on the plasma membrane was observed only with mAb AA4 and was evident by comparing different levels of focus and confirmed by scanning laser confocal microscopy. This patchy pattern was evident within 5 min after the addition of mAb AA4 and still apparent a t 30 min. These differences between mAb AA4- and IgE-antigen-induced (5-30 min for changes were evident at all time points examined both antigen andmAb AA4). Thus, although manyof the same proteins are tyrosine-phosphorylated by mAb AA4 or direct receptor aggregation, there was a distinctly different pattern of intracellular distribution.

DISCUSSION

mAb AA4-induced Intracellular Signals-The mAb AA4-induced protein tyrosine phosphorylations had both similarities and differencescompared withthoseresulting from direct FceRI aggregation.Both stimuliresultedinthe activation and/or tyrosine phosphorylation of Lyn and Syk, and manyof the same substrates were tyrosine-phosphorylated either by FceRI aggregation or by mAb AA4 including FceRIP, FceRIy, and phospholipase C-yl. However, there were several differences. First, the rateof kinase activation and substrate phosphorylation was slower with mAb AA4 than with FceRI aggregation. Second, both FceRIp and FceRIy were more heavily phosphorylated after mAb AA4 stimulationthan by direct FceRI aggregation. Third, mAb AA4 did not increase tyrosine phosphorylation of ~ ~ 1 or 2pp105-115 5 ~ proteins. ~ ~ These are late events that aresecondary to the rise in intracellularcalcium and/or protein kinaseC activation and areprobably critical for degranulation. Fourth, by immunofluorescence, the intracellular distribution of phosphoproteins was different after mAb AA4 stimulation compared with antigen stimulation. In general, mAb AA4 induced the early intracellularprotein tyrosine phosphorylation signals, but thesefailed to generate late signals and did not result in significant secretion. Nevertheless, even these early phosphorylation signals with mAb AA4 were different in time course and/or intensityfrom those with direct receptor activation. These anomalies in earlybiochemi-

19472

mAb AA4-induced Protein Tyrosine Phosphorylation

FIG. 9. Immunofluorescent studies of tyrosine-phosphorylated proteins in RBL-2H3 cells on activation with mAbAA4 or after receptor activation. A, nonstimulated cells;B , 5 min with mAb AA4 stimulation; C , 15 minwithmAb AA4 stimulation; D, 20 minwithmAb AA4 stimulation; E, 15 min with IgE-antigenstimulation; F, 15 minwithantiFccRI (mAb BC4) stimulation.

cal changes may contribute to the lack of the progression of the signaling cascade to later events suchas increases in intracellular calcium, activation of protein kinaseC, and Sermhrphosphorylation of substrates, allof which may be critical for secretion. For example, after mAb AA4 binding, there was little tyrosine phosphorylation of phospholipase C-yl, and this may account for the small changes in intracellular calcium and phosphatidylinositol hydrolysis (27). These resultsalso demonstrate that thephosphorylation a n d o r activation of Lyn, Syk, FceRIP, and FceRIy requires still other signals to result in degranulation. Association of Gangliosideswith FcdZI-mAbAA4 binds through its Fab fragment t o a-galactosyl derivatives of G,,, that associate with FceRI. The spatial proximity of these mAb AA4-binding sites is apparent from the capacity of mAb AA4 as Fab, F(ab'),, or intactmolecules to inhibit the binding of IgE to the receptor (28) and from fluorescent energy transfer experiments (43). Here, we observed coprecipitation of FceRI proteins by mAb AA4,although thepresence of -2% of cellular FceRI in the immunoprecipitates suggests that there is only a weak association of these gangliosides with FceRI. There is alsospe-

cific association of p53/56"" with thea-galactosyl derivatives of G,,,,, recognized by mAb AA4 (30). The association of Lyn with the gangliosides is much stronger than the association of p531 56'''" with FceRI. Thus, there appears to be a complex of molecules that includes these gangliosides, FceRI, and p53/56'.'". Role of FcdZI in mAb AA4-induced ProteinQrosine Phosphorylations-The association of the a-galactosyl derivatives of GDlbwith FceRI and alsowith Lyn implicates FceRI in the mAb AA4-induced protein tyrosine phosphorylations. The mAb AA4-induced protein tyrosine phosphorylations required an unoccupied FceRI and the Fc portion of mAb AA4. This suggests that phosphorylations result from mAb AA4 binding to gangliosides associated with FceRI and theFc portion of the antibody then interacting with the receptor. The fact thatmAb AA4 induced the tyrosinephosphorylation of P and y subunits suggests that FceRI was activated. In this model, mAbAA4 cross-links the gangliosides (including its associated Lyn and other proteins) with FceRI without true receptor aggregation. In contrast, inIgE-mediated cell activation, there is direct aggregation of FceRI either by antigen binding with receptorbound IgE or by anti-receptor antibodies (44-51). However,

mAb AA4-induced Phosphorylation Protein Tyrosine F(ab'),-mediated aggregation of the gangliosides and presumably the ensuing aggregation of Lyn and associated molecules did not result in protein tyrosine phosphorylations. Similarly, Src family tyrosine kinases fail t o initiate activation signals when expressed in T cells fused through their NH, terminus to transmembrane proteins and then aggregated by antibodies to the extracellular domain of these proteins (52). Thus, protein tyrosine phosphorylation may not result from simple aggregation of Src family kinases and may require signals generated uniquely from receptors including FceRI. The close association between these a-galactosyl derivatives of GDlband FceRI may indicate a functional role for these gangliosides that are uniquely expressed on rat mast cells or basophils. Gangliosides present in theouter leaflet of the plasma membrane may associate with cell-surface receptors (53, 54) and have multiple cellular functions (55, 56). They influence membrane fluidity and cell proliferation, act as receptors for other cells or bacterial toxins, and by modulating the function of membrane proteins, regulate transmembrane signaling (55, 56). The function of growth factor receptor tyrosine kinases such asthe platelet-derived growth factor, the epidermal growth factor, and the insulin receptor is modulated by gangliosides (54, 57-62). Gangliosides prevent the ligand-induced dimerization that is critical for signaling from these receptors (63,64).Different gangliosides on mousemast cells at different stages of differentiation couldbe related to their functional heterogeneity (65). Further experiments are needed to define the role of these gangliosides in FceRI function. Acknowledgments-We are indebted to Cynthia Fischler, Judy Waters, andLynda Weedon for skillful technical assistance. We also thank Drs. N. Tbmlinson and M. Swieter for critical reading of this manuscript. REFERENCES 1. Ishizaka. T., and Ishizaka, K. (1984)Prog. Allergy 34, 188-235 2. Metzger, H., Alcaraz,G., Hohman, R., Kinet, J. P., Pribluda, V., and Quarto,R. (1986) Annu. Reu. Immunol. 4, 419470 3. Beaven, M. A., and Cunha-Melo, J. R. (1988) Prog. Allergy 42, 123-184 4. Garcia-Gil, M., and Siraganian, R. P. (1986) J . Immunol. 136,259-263 5. Metzger, H. (1992) Immunol. Rev. 1 2 5 , 3 7 4 8 6. Pfeiffer, J. R., Seagrave, J. C., Davis, B. H., Deanin, G. G., and Oliver, J. M. (1985)J. Cell Biol. 101,2145-2155 7. Oliver, J. M., Seagrave, J., Stump, R.F., Pfeiffer, J. R., and Deanin, G. G. (1988) Proe. Allerev 42. 185-245 8. Sahara, N., Siraganian, R. P., and Oliver, C. (1990)J. Histochem. Cytochem. 38,975-983 9. Siraganian, R. P. (1993) in Allergy: Principles andPractice (Middleton, E., Jr., Reed, C. E., Ellis,E. F., Adkinson, N. F., Jr., Yunginger, J. W., and Busse,W. W., eds) pp. 105-134, Mosby-Year Book, Inc., St. Louis, MO 10. Ullrich, A., and Schlessinger, J. (1990) Cell 61, 203-212 11. Bolen, J. B., Rowley, R. B., Spana, C., and Tsygankov, A. Y.(1992) FASEB J. 6, 3403-3409 12. Benhamou, M., Gutkind, J. S., Robbins, K. C., and Siraganian, R.P. (1990) Proc. Natl. Acad. Sci. U. S. A. 87,5327-5330 13. Paolini, R., Jouvin, M. H., and Kinet, J. P. (1991) Nature 363, 855-858 14. Adamczewski, M., Paolini, R., and Kinet, J. P. (1992) J. Biol.Chem. 267, 18126-18132 15. Yu, K. T.,Lyall, R., Jariwala, N., Zilberstein, A., and Haimovich, J. (1991)J. Biol. Chem. 266,22564-22568 16. Eiseman, E., and Bolen, J. B. (1992)Nature 366, 78-80 17. Stephan, V., Benhamou, M., Gutkind, J. S., Robbins, K. C., and Siraganian,R. P. (1992) J . B i d . Chem. 267, 54345441 18. Li, W., Deanin, G. G., Margolis, B., Schlessinger, J., and Oliver, J. M. (1992) Mol. Cell. Bid. 12, 3176-3182 19. Hutchcroft, J. E., Geahlen, R. L., Deanin, G. G., and Oliver, J. M. (1992)Proc. Natl. Acad. Sci. U.S. A. 89, 9107-9111

-

Y"

I

19473

20. Benhamou, M., and Siraganian, R. P. (1992) Immunol. Today 13,195-197 21. Park, D. J., Min,H. K., andRhee, S. G. (1991)J. Biol. Chem. 266,24237-24240 22. Margolis, B., Hu, P., Katzav, S., Li, W., Oliver, J. M., Ullrich, A,,Weiss, A,, and Schlessinger, J. (1992) Nature 366, 71-74 23. Benhamou, M., Stephan, V., Robbins, K. C., and Siraganian, R.P. (1992) J. Biol. Chem. 267,7310-7314 24. Hamawy, M.M., Mergenhagen, S. E., and Siraganian, R. P. (1993) J. Biol. Chem. 268,5227-5233 25. Hamawy, M. M., Mergenhagen, S., and Siraganian,R. P. (1993)J. Biol. Chem. 268,6851-6854 26. Guo, N. H., Her, G. R., Reinhold, V. N., Brennan, M. J., Siraganian, R. P., and Ginsburg, V. (1989) J. Bid. Chem. 264, 13267-13272 27. Oliver, C., Sahara, N., Kitani,S., Robbins, A. R., Mertz, L. M., and Siraganian, R. P. (1992) J. Cell Biol. 116, 635-646 28. Basciano, L. K, Berenstein, E. H., Kmak, L., and Siraganian, R.P. (1986) J. B i d . Chem. 261, 11823-11831 29. Stephan, V., Guo, N., Ginsburg, V., and Siraganian, R. P. (1991)J . Immunol. 146,42714277 30. Minoguchi, K., Swaim, W. D., Berenstein, E. H., and Siraganian, R. P. (1994) J. Biol. Chem. 269,524%5254 31. Kitani, S., Berenstein, E., Mergenhagen, S., Tempst, P., and Siraganian, R. P. (1991) J. Biol. Chem. 266,1903-1909 32. Hamawy, M.M., Oliver, C., and Siraganian, R.P. (1992) J . Immunol. 148, 524-531 33. Stracke, M. L., Basciano,L. K., Fischler, C., Berenstein, E. H., and Siraganian, R. P. (1987)Mol. Immunol. 24,347356 34. Benhamou, M., Ryba, N. J. P., Kihara, H., Nishikata, H., and Siraganian,R. P. (1993) J . Biol. Chem. 268,23318-23324 35. Dippold, W. G.. Lloyd, K. O., Li, L. T., Ikeda, H., Oettgen, H. F., and Old, L. J. (1980)Proc. Natl. Acad. Sci. U. S. A. 77, 6114-6118 36. Kamps, M. P., and Sefion, B. M. (1988) Oncogene Res. 3, 105-115 37. Minoguchi, K , Kihara, H., Nishikata, H., Hamawy, M. M., and Siraganian,R. P. (1994) Mol. Immunol. 31,519-529 38. Barsumian, E. L., Isersky, C., Petrino,M. G., and Siraganian,R. P. (1981) Eur. J . Immunol. 11,317423 39. Laemmli, U. K. (1970) Nature 227, 680-685 40. O'Farrell, P. H. (1975) J. Bwl. Chem. 260,4007-4021 41. Minoguchi, K., Benhamou, M., Swaim, W. D., Kawakami, Y., Kawakami, T., and Siraganian, R. P. (1994)J. Biol. Chem. 269, 16902-16908 42. Sehneider, H., Cohen-Dayag,A., and Pecht, I. (1992) Int. Immunol. 4 , 4 4 7 4 5 3 43. Estes, K. A. S. (1988) Studiesof the Interaction between the Receptor for Immunoglobulin E and Other Cellular Components during Signal 'Dunsduction. Ph.D. dissertation, Cornell University 44. Siraganian, R. P., Hook, W. A,, and Levine, B. B. (1975) Immunochemistry 12, 149-157 45. Segal, D.M., Taurog, J. D., and Metzger, H. (1977) Proc. Natl. Acad. Sei. U. S. A. 74, 2993-2997 46. Ortega, E., Schweitzer-Stenner, R., and Pecht, I. (1988)EMBOJ. 7,41014109 47. Balakrishnan, K., Hsu, F. J., Cooper, A. D., and McConnell, H. M. (1982) J. B i d . Chem. 267,6427-6433 48. Kane, P., Erickson, J., Fewtrell, C., Baird, B., and Holowka, D. (1986) Mol. Immunol. 23,783-790 49. Chabay, R., DeLisi, C., Hook, W. A,, and Siraganian,R. P. (1980)J. B i d . Chem. 266,4628-4635 50. DeLisi, C., and Siraganian, R.P. (1979) J. Immunol. 122, 2286-2292 51. Kagey-Sobotka, A., Dembo, M., Goldstein, B., Metzger, H., and Lichtenstein, L. M. (1981) J. Immunol. 127,2285-2291 52. Kolanus, W., Romeo, C., and Seed, B. (1993) Cell 74, 171-183 53. Kielczynski, W., Hamson, L. C., and Leedman, P. J. (1991) Proc. Natl. Acad. Sci. U. S. A. 88, 1991-1995 54. Hanai, N., Nores,G. A,, MacLeod, C., Torres-Mendez, C. R., and Hakomori, S. (1988) J. Biol. Chem. 263,10915-10921 55. Hakomori, S. (1990) J. Bid. Chem. 265, 18713-18716 56. Zeller, C. B., and Marchase, R. B. (1992)Am. J. Physiol. 262, C134141355 57. Weis, F. M., and Davis, R.J. (19901 J. Bid. Chem. 266,12059-12066 58. Nojiri, H., Stroud, M., and Hakomori. S. (1991) J. Biol. Chem. 266,4531-4537 59. Song, W. X.,Vacca, M. F., Welti, R., and Rintoul, D.A. (1991) J. B i d . Chem. 266, 1017P10181 60. Bremer, E.G., Hakomori, S., Bowen-Pope, D. F., Raines, E., and Ross, R. (1984) J . Biol. Chem. 269,6818-6825 61. Bremer, E. G., Schlessinger, J., and Hakomori, S. (1986) J. Biol. Chem. 261, 2434-2440 62. Davis, R. J., Girones, N., and Faucher, M. (1988) J . Biol. Chem. 263, 53735379 63. Yarden, Y., and Schlessinger, J. (1987) Biochemistry 26, 1434-1442 64. Van Brocklyn. J., Bremer, E. G., and Yates, A. J. (1993) J. Neurochem. 61, 371-374 65. Katz, H. R., Dayton, E. T., Levi-SchaRer, F., Benson, A. C., Austen, K. F., and Stevens, R. L. (1988) J. Immunol. 140, 30904097