(Received for publication, May 21, 1984). Robin J. Andy$, Connie L. Finstads, Lloyd J. Oldp, Kenneth 0. Lloyd#, and Rosalind KornfeldS. From the $Washington ...
THEJOURNAL OF BloLOclcAL CHEMISTRY 0 1984 by The American Society of Biological Chemists, Inc.
Vol. 259, No. 20, Issue of October 25, pp. 12844-12849 1984 Printed in &%A.
The Antigen Identified by a Mouse Monoclonal Antibody Raised against Human Renal Cancer CellsIs the Adenosine Deaminase Binding Protein* (Received for publication, May 21, 1984)
Robin J. Andy$, Connie L. Finstads, Lloyd J. Oldp, Kenneth 0. Lloyd#, and Rosalind KornfeldS From the $Washington University School of Medicine, Departments of Internal Medicine and Biological Chemisty, Division of Hemntology-Oncology, St. Louis, Missouri 63110 and the $Memorial Sloan-Kettering Cancer Center, New York, New York 10021
The antigen recognized by a mouse monoclonal antibody (mAb527) raised againsta human renal cancer cell line has been identified as the adenosine deaminase binding protein. mAb S27 immunoprecipitates binding protein purified from a soluble fraction of human kidney. It also recognizes themature120,000-dalton membrane form of binding protein from [3"S]methionine-labeled human fibroblasts, HepG2 cells, and the renal cancer cell line against which the antibody was raised. A rabbit polyclonal antibody raised against purified kidney binding protein completely precipitates mAb S27-reactive material from labeled membrane extracts. mAb 527 does not precipitate the initially synthesized 110,000molecular weight precursor of binding protein in fibroblasts and recognizes only a small portion of binding protein precursor in labeled HepG2 cells suggesting that theantigenic determinant recognized by mAb 527 may be a post-translational modification present on the mature form of binding protein or that mAb 527 recognizes molecules in a certain conformation. Glycopeptides derived from purified soluble kidney binding protein or exogenously added adenosine deaminase do not inhibit the immunoprecipitation of binding protein by mAb 527, indicating that the mature oligosaccharide chains of binding protein are not the determinantrecognized bymAb S27 and that bound adenosine deaminase doesnot mask the antigenic sites on binding protein. The fact that monoclonal antibody 527, previously shown (Ueda, R., Ogata, S., Morissey, D. M., Finstad, C. L., Szkudlavek, J., Whitmore, W. F., Oettgen, H. F., Lloyd, K. O., and Old, L. J. (1981)Proc. Nutl. Acad. Sci. U. S . A. 78, 5122-5126) to detect a cell surface antigen on culturedrenal cancer cells, is directed against the adenosine deaminase binding protein confirms and extends the earlier observation (Andy, R. J., and Kornfeld, R. (1982) J. Biol. Chem. 257, 79227925) that binding protein is located on the cell surface.
A variety of novel cell surface antigens of human cancers have been detected with monoclonal antibodies obtainedafter immunizing mice with tumor cells (1).Among these are a
* This research was supported in part by Grants R01 CA08759, 5 T32 GM07067, CA08748,and CA21445 from the United States Public Health Service. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
series of 17 antibodies to human renal cancer cells described by Ueda et al. ( 2 ) . The set of five monoclonal antibodies that immunoprecipitated a 120,000-dalton polypeptide from [3HJglucosaminelabeled kidney tumor cells (2) was of particular interest to us. The serological specificity of these antibodes was tested on a large panel of established cell lines using the prototype antibody mAb' S6. It was found to react with epithelial cancers of renal, bladder, colon, and lung origins as well as with normal kidney epithelial cells, fetal kidney cells, adult skin fibroblasts, fetal lung fibroblasts, and VERO monkey kidney cells ( 2 ) .The molecular weight and distribution of the antigen recognized by this set of monoclonal antibodies was similar to thatof the adenosine deaminase binding protein. Binding protein was originally characterized as a soluble glycoprotein localized in thecytoplasm of tissues suchas kidney, liver, and lung (3-5). It has recently been shown to be present on the cell surface of human fibroblasts (6) and has been purified from a crude membrane fraction of human kidney (7) and human placenta (8).The membrane form of binding protein in human fibroblasts has a molecular weight of 120,000 (9). In biosynthetic studies, binding protein was shown to be initially synthesized as a precursor with a molecular weight of 110,000. Within 30-60 min of chase, this 110-kDaprecursor was converted to a 120-kDa mature form. This increase in molecular weight is most likely due to processing of the oligosaccharide chains of binding protein (9). In this study, we examine the reactivity of one of the monoclonal antibodies (mAb S27) with purified soluble kidney adenosine deaminase binding protein and with radiolabeled binding protein from human fibroblasts, a human hepatoma cell line, and a renal cancer cell line. EXPERIMENTALPROCEDURES
Materials-Na'2SI and [35S]methionine(1120 Ci/mmol) were purchased from Amersham Corp. [3H]Glucosamine(30-60 mCi/mmol) was from New England Nuclear. Protein A-Sepharose was from Pharmacia. Bovine serum albumin and aprotinin were from Sigma. EN3HANCE was from New England Nuclear. Purification and Radioiodinatwn of Soluble Kidney AdenosineDeaminase Binding Protein-Binding protein was purified by a modification (6) of the method described by Daddona and Kelley (11). Binding protein was iodinated by incubating 10 pg of protein with 1 pg of iodogen (Pierce Chemical Co.) and 20 pCi of Nal"I for 10 min on ice in a final volume of 25 rl. The sample was removed from the tube, and 2 mg of bovine serum albumin and 5 mg of KI were added. The sample was then applied to a Bio-Rad P-2 column equilibrated in phosphate-buffered saline, 0.02% sodium azide, and radioiodinated The abbreviations used are: mAb, monoclonal antibody; SDS, sodium dodecylsulfate; MEM, minimal essential medium; Da, dalton; BP, binding protein.
12844
Monoclonal Antibody Adenosine to binding protein appearing in the void volume was used for immunoprecipitation. Cells-Human fibroblasts (AG1518) obtained from the Human Genetic Cell Repository and a human hepatoma cell line (HepG2) from the Wistar Institute were maintained in monolayer culture in 95% air/5% COz.They were routinely cultured in a-minimal essential medium (a-MEM)containing 10%fetal calf serum (Gibco), 100units/ ml of penicillin, and 100 pg/ml of streptomycin. The renal cancer cell line (SK-RC-7) has been described (10). Cultures were maintained in Eagle's MEM, 100 units/ml penicillin, 1 pg/ml streptomycin, and 10% fetal calf serum. Radioactive Labeling-Fibroblasts and HepG2 cells were pulse labeled with [%]methionine after washing the cells 2times with methionine-free MEM and preincubating for 0-30 min a t 37 "C in methionine-free MEM containing 10% dialyzed serum as indicated in the figure legends. The preincubation medium was removed and replaced with medium containing 75-100 pCi/ml [35S]methionineand incubated for various times. The labeling medium was removed and cells were chased in a-MEM containing 500 pM methionine and 10% fetal calf serum. Renal cancer cell line (SK-RC-7) was labeled with [3H]glucosamine(20-50 pCi/ml) for 72 h. Preparation of Membrune Extracts-Membranes from labeled fibroblasts and HepG2 cells were prepared and detergent extracted as previously described (9). Antibodies-Mouse mAb S27 has been described (2); it has the same serological specificity as the prototype mAb S6 (2). An I& fraction of rabbit polyclonal antiserum raised against purified soluble kidney binding protein (6) was prepared as described previously (9). Immunoprecipitation of 12SI-lnbeled SolubleKidney Binding Protein-To a sample (25 pl) containing approximately 0.125 pg of '%Ilabeled binding protein were added 245pl of phosphate-buffered saline containing 0.5 mg/ml bovine serum albumin. Binding protein was immunoprecipitated by the addition of either 2 pl of S27 serum or 5 p1 of rabbit anti-binding protein IgG. Samples were incubated on ice, followedby the addition of 20 pl of rabbit anti-mouse Ig serum (Cappel Laboratories, Cochranville, PA) to thesample containing the monoclonal antibody, and the incubation was continued for 2 h on ice. Immune complexes were precipitated by the addition of Protein A-Sepharose and incubated on ice for 30 min. The immunoprecipitates were washed 3 timeswith PBS containing 0.2 unit/ml aprotinin, 1 mM methionine, 1% TritonX-100,0.5% sodium deoxycholate, and 0.005% SDS (Buffer A) and once with phosphate-buffered saline containing 0.2 unit/ml aprotinin. Each immunoprecipitate was solubilized in 125 mM Tris-HC1, pH 6.8, 2% SDS, 20% glycerol, 2% @mercaptoethanol, and 0.002% bromphenol blue by boiling 5 min, and subjected to SDS-polyacrylamide gel electrophoresis. A control sample was treated as above with 5 pl of a nonspecific ascites fluid. Immunoprecipitation of Binding Protein from Labeled Cell Extracts-As outlined in Fig. 1, [35S]methionine-labeledfibroblast membrane extracts in Buffer A were supplemented with bovine serum albumin to aconcentration of 3 mg/ml and were precleared by incubating with 50 pl of a 1:l suspension of Protein A-Sepharose for 20 min on ice. Suspensions were centrifuged for 2 min at 4 "C in an Eppendorf centrifuge. The supernatant was divided into2 equal aliquots and either 200 p1 of preimmune IgG (sample 1) or 10 pl of nonspecific ascites fluid (sample 2) was added, and the samples were incubated on ice for 60-120 min. To the ascites fluid sample 100 pl of rabbit anti-mouse Ig serum were added, and theincubation continued for 60 min on ice. IgG wasprecipitated by the addition of Protein A-Sepharose (1 p1/10 pg of IgG) as described above. Following centrifugation to remove the Protein A-Sepharose, 50 p1 of anti-binding protein IgG or 5 pl of mAb S27 were added to the supernatant and left on ice overnight. After incubating the mAb S27-treated sample with 40 pl of rabbit anti-mouse Ig serum for 60 min on ice, immune complexes were precipitated with Protein A-Sepharose as described. Samples 1 and 2 were then retreated with anti-binding protein IgG or mAb S27, respectively, to ensure complete immunoprecipitation. Then, to sample 1 were added 5 pl of mAb S27 and to sample 2 were added 75 p1 of anti-binding protein IgG. The incubations and precipitation of immune complexes were performed as described above. Each immunoprecipitate was washed and solubilized as described above and analyzed by SDS-polyacrylamide gel electrophoresis. HepG2 membrane extracts in Buffer A were supplemented with bovine serum albumin to a concentration of 3 mg/ml and preincubated with 25-40 pl of a 1:1 suspension of Protein A-Sepharose for 20 min on ice. Suspensions were centrifuged in an Eppendorf centrifuge for 2 min at 4 "C, and the supernatantwas divided into 2 equal
Deaminase Binding Protein
12845
knbraneExtract
+ZOO "I preirmvlne IgG 120 min, on i c e
+lo 111 non-specificascites f l u i d 60 min, on i c e
+Protein ASepharose
t
supernatant=sample 1
+ l o 0 111 r a b b i t anti-mouse I g 60 min. on i c e +Protein ASepharose
t
supernatant=sample2
+5 "1 527
+50 ul anti-BP IgG overnight, on i c e
A
a n t i BP-1A s u p e r n a t a n t
+50 min. BP "1 IgG on anti- 120 ice
a n t i BP-2
supernatant +5X 527 overnight, on i c e
anti BPS27
A
supernatant
/
527-1
60 min on i c e
supernatant supernatant
+50 "1 a n t i BP IgG120 min. on i c e
a n t i BP-3
overnight, on i c e
527-2 +Protein ASepharose'
supernatant
S27-3
+75 111 a n t i BP lgG overnight, on i c e
FIG. 1. Scheme for sequential immunoprecipitationof binding protein from labeled fibroblasts. The asterisk indicates that additional Protein A-Sepharose was added to ensure complete precipitation of rabbit anti-mouse IgG. aliquots. One aliquot (sample 1) was incubated with 100 pl of preimmune IgG for 120 min followed by precipitation with Protein ASepharose as described above. To sample 1, 200 p1 of anti-binding protein IgG were added and to the other aliquot (sample 2), 50 pl of a mAb S27-Protein A-Sepharose suspension were added. For this experiment 100pl ofmAb 527 were preadsorbed onto 100pl of Protein A-Sepharose. After washing the beads with Buffer A they were resuspended in an equal volume of Buffer A and used in all subsequent immunoprecipitations. The samples were incubated at 4 'C overnight. Immune complexes were precipitated with Protein ASepahrose, and samples 1 and 2 were retreated with anti-binding protein IgG or mAb S27-Protein A-Sepharose, respectively, for 8 h to ensure complete immunoprecipitation. Immune complexeswere precipitated as described and then, to sample 1 were added 25 pl of S27-Protein A-Sepharose and tosample 2 were added 150 p1 of antibinding protein IgG. Incubations were continued overnight at 4 "C followedby precipitation of immune complexes. Each immunoprecipitate was washed and solubilized as described. Renal cancer cell (SK-RC-7) lysates were treated as above except that both mAb S27 and theanti-binding protein IgG werepreadsorbed to Protein A-Sepharose. Aliquots of protein A-coupled mAb S27 or Protein A-coupled anti-binding protein IgG were incubated with [3H] glucosamine-labeled cell extracts in 100 p1of 0.01 M Tris-HC1, 0.15 M NaCl, pH 8.0. The nonbound material from mAb S27 treatment was incubated with anti-binding protein IgG. Similarly, 2 aliquots of Protein A-coupled anti-binding protein IgG were incubated successively with labeled cell extracts and the supernatant finally incubated with mAb S27-Sepharose. The first and second incubations of cell extracts with antibody were for 4 h at room temperature, and the third incubation was for 8 h at room temperature. Immunoprecipitates were analyzed by SDS-polyacrylamide gel electrophoresis as described below. Gel Electrophoresis-SDS-polyacrylamide gel electrophoresis was performed according to Laemmli (21) using 7.5% polyacrylamide slab gels. Electrophoresis was carried out at 120 V/slab. The gels were stained with 0.25% Coomassie Blue R-250 in 50% methanol, 10% acetic acid and destained in 30% ethanol, 10% acetic acid. Following
Monoclonal Antibody to Adenosine Deaminase Binding Protein
12846
destaining, the gelsweresoaked in EN3HANCEfor 1 h at room temperature and then 1 h in water. The gels were dried and exposed to preflashed x-ray film (Kodak XAR-5) at -70 “C. Preparation of Binding Protein Glycopeptides-Purifiedsoluble kidney binding protein(500 pg) was incubated with 25 pg of Pronase in 0.1 M Tris, pH 8.0.2 mM CaC12at 56 “C overnight undera toluene atmosphere. Another25 pg of Pronase were added,and the incubation was continued for 24 h. A final 25 pg were added, and the incubation was continued for the 48 h at 37 “C.The reaction was stopped by boiling, and the glycopeptides were desalted on a Sephadex G25-80 column equilibrated in 7% 1-propanol. RESULTS
Immunoprecipitation of ’251-labeledSoluble Kidney Adeno-120K sine Deaminase Binding Protein-To test the reactivity of -1 10K monoclonal antibody S27 with soluble kidney binding protein, immunoprecipitations were performed as described under “Experimental Procedures.” As shownin Fig. 2, mAb S27 (lane 2) and anti-binding protein IgG ( l a n e I ) specifically immunoprecipitated binding protein whereas a control monoclonal antibody did not precipitate binding protein (data not shown). The immunoprecipitationby S27 was complete since no ”9-labeled binding protein remained in the supernatant (data not shown). This result supports the conclusion that monoclonal antibody S27 specifically recognizes the adenosine deaminase binding protein. Immunoprecipitation of Labeled Cell Extracts-To test the reactivity of mAb S27 with the membrane form of binding 1 2 3 4 5 6 7 8 9 protein, the experiment outlined in Fig. 1 wasperformed. Fibroblasts were pulse labeled for 60 min with [35S]methioFIG. 3. Immunoprecipitation of labeled fibroblast memwas brane extracts. Confluent cultures of fibroblasts (in 100-mmdishes) nine, anda membrane extract was prepared. The extract to sequential were pulse labeled with 100 pCi/ml [?S]methionine in methione-free divided into two equalaliquotsandsubject MEM-Earle’s containing10% dialyzed calf serum for60 min at 37 “C. immunoprecipitation. By sequentiallyimmunoprecipitating with anti-binding protein IgG followed by mAb S27 or mAb The plates were washed three times with PBS containing 0.17 unit/ S27 followed by anti-binding protein IgG, it could be deter- ml aprotinin and 1 mM methionine. Membranes were prepared and as described under “Experimental Procedures”and subject mined if both antibodies were recognizing the same antigen. extracted to sequential immunoprecipitationas outlined in Fig. 1. Lanes 1 and As shown in Fig. 3, lane 5, anti-binding protein IgG immu- 9 are controls, treated with nonspecific ascites and preimmune I$, respectively. Lane 2, mAb S27 3; lane 3, mAb S27 2; lane 4, mAb S27 1; lane 5,anti-BP 1; lane 6,mAb S27 + anti-BP; lane 7, anti-BP + mAb S27; lane 8,anti-BP 3. Numbers on the right denote the mature (120,000) and precursor (110,000) forms of binding protein.
1
2
3
FIG.2. Immunoprecipitation of soluble kidney binding protein. Purified soluble kidney binding protein was iodinated as describedunder“ExperimentalProcedures.”Immunoprecipitations were performed with anti-binding protein IgG (lane I ) or monoclonal antibody S27 (lane2) as described under“ExperimentalProcedures.” ”‘I-Binding protein is shown in lane 3. Arrow indicates soluble kidney binding protein (112 kDa).
fibronoprecipitated 110- and 120-kDa components from the blast extract. These correspond to the precursor and mature form of binding protein, respectively. When this samplewas subsequently treated with mAbS27, no additional 120- or 110-kDa bands were obtained (Fig. 3, lane 7) suggesting that anti-bindingprotein IgG haddepletedtheextract of the antigen recognized by mAb S27. The bandsvisualized in this lane are contaminants not seenevery in immunoprecipitation with mAb S27. When the membrane extract was treated first with mAb S27, the 120-kDa band was immunoprecipitated, butnoprecursor was obtained (Fig. 3, lane 4 ) . A second addition of mAb S27 precipitated slightly more of the 120kDa band(Fig. 3, lane 3).When this sample was subsequently treated with anti-binding proteinIgG, the 110-kDa precursor was immunoprecipitated alongwith some additional 120-kDa mature form (Fig. 3, lane 6 ) .These results suggest that mAb S27 cannot immunoprecipitate the precursorof binding protein in fibroblasts and can only recognize a certain population of the mature (120 kDa) binding protein molecules which are recognized by anti-binding proteinIgG. When cells werepulse labeled for 15 min and chasedfor 16 h,a t which time labeled cell bindingprotein is fully maturedandpresentonthe surface,onceagainonly a portion of the binding protein polypeptides that could be immunoprecipitated by anti-bindingprotein IgG were precipitated by mAb S27(datanot shown). These results suggest that there is a determinant
Monoclonal Antibody to Adenosine Deaminase Binding Protein
12847
-200K -116K 125K -92K
llOK-
-66K
-45K
1
2
3 4 5 6 7 8 9 1 0 1 1 1 2
FIG. 4. Immunoprecipitation of labeled HepGt membrane extracts. Cultures of HepC2 cells (in 100-mm dishes) were pulse labeled with 75 pCi/ml [?3]methioninein methionine-free a-MEM containing 10% dialyzed fetal calf serum for 15min at 37 “C after a 30-min preincubationin methionine-free a-MEM. Cells were chasedin a-MEMcontaining 10% fetal calfserumand 500 p~ methionine for 0 (lanes 1 - 6 ) or 120 min ( l a n e s 7-12). Following the chase,plateswerewashed three times withPBScontaining 0.2 unit/mlaprotininand 1 mM methionine. Membranes were prepared, extracted, and subjected to sequential immunoprecipitation as described under “Experimental Procedures.”Lane 1, mAb S27 1; lane 2, mAbS27 2; lane 3, mAb S27 + anti-BP; lane 4, anti-BP 1; lane 5,anti-BP 2; lane 6, anti-BP mAb S27; lane 7, mAbS27 1; lane 8, mAb S27 2; lnne 9, mAb S27 + anti-BP; lane 10, anti-BP 1; lane 11; anti-BP 2; lane 12, anti-BP + mAb S27. The mature form (125,000) and precursor (110,OOO) are indicated with lines. The numbers to the right show the migration of molecular weight standards.
recognized by mAb S27 that is present on only some of the binding protein molecules in fibroblasts. Apparently this determinant is expressed only on mature binding protein since the fibroblast precursor is not immunoprecipitated by mAb S27. The reactivity ofmAb S27 with binding protein from HepG2 cells was also tested. HepG2 cells were pulse labeled with [35S]methioninefor 15 min and chased for 0 or 120 min. With a 15-min pulse, no chase, only the 110-kDa precursor, was labeledin HepG2. As shown in Fig. 4 (lanes 1 and 2), two additions of mAb S27 precipitated a small amount of material migrating with a molecular weight of 110,000. When antibinding protein IgG was subsequently added to this sample, the majority of binding protein precursor was immunoprecipitated (Fig. 4, lane 3). If the extract was treated first with anti-binding protein IgG, the precursor was completely precipitated (Fig. 4, lane 4 ) , and no additional mAb S27-reactive material remained (Fig. 4, lane 6). When cells were pulse labeled for 15 min and chased for 120 min, only the mature form of binding protein waslabeled. In HepG2 cells this mature form has a molecular weightof 125,000 (9). When the membrane extract was treated with mAb S27, this 125-kDa band was immunoprecipitated (Fig. 4, lane 7). A second addition of mAb S27 precipitated slightly more binding protein (Fig. 4, lane 8). If the extract was subsequently treated with anti-binding protein IgG, some additional 125-kDa material was precipitated (Fig. 4, lane 9). If the extract was treated first with anti-binding protein I&, the 125-kDa mature form was completely precipitated (Fig. 4, lane lo), and no mature form wasprecipitated by the subsequent addition of mAb S27 (Fig. 4, lane 12). These results suggest that mAb S27 recog-
nized the mature form of binding protein in HepG2cells similar to the results obtained in fibroblasts. In contrast to the results in fibroblasts, however, is the observations that a small amount of HepG2 precursor was precipitated by mAb S27. This result suggests that in HepG2 some of the binding protein precursor molecules carry the antigenic determinant recognized by the monoclonal antibody S27 whereas the majority of mature binding protein carries this determinant. The human renal cancer cell line (SK-RC-7) against which mAb S27 was raised waslabeled with [3H]glucosamineas described under “Experimental Procedures.” From a cell lysate, anti-binding protein IgG immunoprecipitated a band of approximately 120 kDa (Fig.5, lane 4). When this sample was subsequently treated wtih mAb S27, no additional binding protein was immunoprecipitated (Fig. 5, lane 3). If the lysate was treatedfirst with mAbS27, a band of 120kDawas obtained (Fig. 5, lane 1). A second addition ofmAb S27 precipitated slightly more of the 120-kDa band (Fig. 5, lane 2). When this sample was subsequently treated with antibinding protein IgG, more of the 120-kDa material was immunoprecipitated (Fig. 5, lane 6). These results confirm those obtained with labeled HepG2 cells and fibroblasts in that mAb S27 immunoprecipitated some, but not all, of the mature form of binding protein. Inhibition of Innunoprecipitation-One possible explanation for the observation that binding protein precursor in fibroblasts is not recognized bymAb S27 is that a posttranslational modification present on mature binding protein is the antigenic determinant recognized by the monoclonal antibody. A potential location for this modification would be on the oligosaccharide chains. It is known that theprecursor
MonoclonalAntibody to Adenosine Deaminase Binding Protein
12848
,120K
116K94K -
68K 45K-
1
2
3
4
5
6
7
FIG. 5. Immunoprecipitation of labeled SK-RC-7cell extracts. Cultures of SK-RC-7 cells were grown in the presence of 40 pCi/ml [3H]glucosaminein Eagle's MEM containing 10% fetal calf serum for 72 h at 37 "C. Cell lysates, prepared as previously described (2), were used for immunoprecipitations as outlined under "Experimental Procedures." Lane I , mAb S27 1; lane 2, mAb S27 2; lane 3, anti-BP + mAB S27; lane 4, anti-BP 1; lane 5, anti-BP 2; lane 6, mAb S27 4anti-BP; lone 7,normal rabbit serum used as negative control. The mature form (120,000) is indicated on the right. The numbers to the left indicate the migration of molecular weight standards.
contains high mannose-type chains whereas the mature form contains complex-type chains (9). It is unlikely that the typical complex chains are recognized since these would be present onmany glycoproteins. However, it may bethat there is a specific structure on these chains, such as sialic acid or fucose, that is recognized. To test this possibility, glycopeptides were prepared from soluble kidney binding protein and these glycopeptides were then added in approximately 50- to 100-fold excess to a labeled HepG2 membrane extract. Binding protein was then immunoprecipitated with mAb S27 or anti-binding protein IgG. Binding protein glycopeptides did not inhibit the immunoprecipitation of binding protein by mAb S27 or by the polyclonal IgG (data notshown) suggesting that theantigenic determinant recognized bythe monoclonal antibody is not the oligosaccharide chains of binding protein. In the threecell lines tested, mAb S27 did not completely precipitate mature binding protein since anti-binding protein IgG added after mAb S27 can immunoprecipitate additional binding protein. One possible explanation is that a small per cent of the binding protein molecules contain bound adenosine deaminase and that the bound enzyme inhibits precipitation by mAb S27. It is known that only a small per cent of total binding protein is occupied by adenosine deaminase in many cells (6). To test this possibility, labeled membrane extracts from HepG2 cells were incubated with a saturating concentration of adenosine deaminase followed by immunoprecipitation of binding protein with mAb S27. There was no inhibition of the immunoprecipitation inthe sample containing excess adenosine deaminase (data notshown). DISCUSSION
The results presented here show that a monoclonal antibody (S27) raised against a humantumor cell line specifically recognizes the mature form of the adenosine deaminase binding protein. This was demonstrated using purified soluble
kidney adenosine deaminasebinding protein and labeled cell extracts. Interestingly, infibroblasts only the mature form of binding protein was recognized by the monoclonal antibody since the precursor was not immunoprecipitated by mAb S27. In contrast,in HepG2 cells, a small amount of precursor was precipitated bymAbS27. Theseresults suggest that the determinant recognized by mAb S27 may be a post-translational modification of binding protein that is present only on the mature form of binding protein infibroblasts and on some precursor molecules and most mature binding protein molecules in HepG2 cells. One possible localization for this modification would be the oligosaccharide chains. There aremany examples of monoclonal antibodies that recognize carbohydrate determinants,particularly those of glycolipids (12-17). For example, a monoclonal antibody raised against a human gastric cancer cell linereacts with glycolipids having the structure Fuccrl-2GalS1-4(Fucal-3)GlcNAc(18). A monoclonal antibody that specifically reacts with the epidermal growth factor receptor of A431 cells has recently been characterized (19, 20). This monoclonal antibody precipitates receptor from A431 cells but not from human KB cells, normal rat kidney cells, or Swiss 3T3 cells (19). The reason for this specificity is that themonoclonal antibody is directed against the human blood group H type 1 sugar sequence Fuccrl2GalB1-3GlcNAc- (20) which is presenton the epidermal growth factor receptor in A431 cells but not in other cells. However, since dycopeptides derived from soluble kidney binding protein did not inhibit the immunoprecipitation of binding protein from labeled HepG2 cells, it seems unlikely that this monoclonal antibody recognizes a carbohydrate determinant. Also,. the determinant recognizedbymAb S27 must be present on only a portion of the mature binding protein molecules since additional binding protein is precipitated by anti-bindingprotein IgG added after membrane extracts have been treated with mAb S27. It was possible that mAb S27 was precipitating only unoccupied binding protein and, therefore, that the portion of binding protein molecules not immunoprecipitated bymAb S27 was complexed with adenosine deaminase. Since exogenously added enzyme did not inhibit the immunoprecipitation of binding protein, this seems unlikely. It appears then that in the various cell lines studied not all of the mature binding protein molecules carry the antigenic determinant although the majority of binding protein in HepG2 cells and the renal cancer cells is precipitated by mAb S27.Of course, it is possible that theadditional 120-kDa material immunoprecipitated by the polyclonal antibody is a contaminant. This seems unlikely, however, since mAb S27 completely immunoprecipitated purified soluble kidney binding protein which was used for the raising of antibinding protein antiserum. It may be that the determinant recognized bymAb S27 is formed by phosphorylation or sulfation of binding protein. It is also possible that a conformational change occurs as the precursor is converted to the mature form and thatbinding protein must be in this conformation to be recognized by mAb S27. Also, the soluble form of binding protein is known to exist as a dimer, and itmay be that mAb S27 recognizes the dimeric form and not the monomeric form. It is not known, however, when dimerization occurs. These possibilities have not been investigated further. In addition to identifying the antigen, the reactivity of monoclonal antibody S27 with binding protein demonstrates in another way the presence of binding protein on the cell surface. The assay used in previous studies (2) to analyze mAb S27, i.e. anti-mouse Ig-mixed hemagglutination assay, detects only cell-surface antigens. These studies showed that the antigen is present on the surface of renal, colon, bladder,
Monoclonal Antibody to Adenosine Deaminase Binding Protein and lung tumor cell lines, normal kidney, and monkey kidney cell lines. This distributionparallels that of the soluble form of binding protein(3). In earlier studies(6,9), it was suggested that the soluble form of binding protein arose by proteolysis of the membrane form. The fact that binding protein is presenton the surface of cells known to contain the soluble form supports this conclusion and demonstrates that adenosine deaminase binding Protein is Present on the surface of many cell types. REFERENCES 1. Lloyd,K.
2.
3.
4. 5. 6.
7. 8.
0. (1983) in Basic and Clinical Tumor Immunology (Herberman, R.B., ed) pp. 159-214, Nijhoff, Boston Ueda, R., Ogata, S., Morissey, D. M., Finstad, C. L., Szkudlarek, J., Whitmore, w. F., oettgen, H. F.,Lloyd, K. O., and Old, L. J. (1981) Proc. Natl. Acud. Sci. U. 5'. A. 7 8 , 5122-5126 Van der WeydenpM. B.*and Ke11ey9w' N' (1976) " Bioi' 25 1,5448-5456 Nishihara, H., Ishikawa, S., Shinkai, K., and Akedo, H. (1973) Biochim. Biophys. Acta 302,429-442 Swallow, D. M., Evans, L., and Hopkinson, D. A. (1977) Nature (Lond.) 269,261-262 Andy, R. J., and Kornfeld, R. (1982) J. Biol. Chern. 2 5 7 , 79227925 Schrader, W. P., and Bryer, P. J. (1982) Arch. Biochem. Biophys. 215, 107-115 Trotta, P. (1982) Biochemistry 21,4014-4023
12849
9. Andy, R. J., and Kornfeld, R. (1984) J . B i d Chem. 2599 98329839 Ueda, R., Shiku, H., Pfeundschuh, Takahashi, T., Li, L. T. C., Whitmore, W. F., Oettgen, H. F., and Old, L. J. (1979) J. Exp. Med. 150,564-579 11. Daddona, P. E., and Kelley, W.N. (1978) J. Biol. Chem. 2 5 3 , 4617-4623 . 12. Young, w. w.9 Jr., Portoukalian, Jv and Hakomori, s. (1981) J . Biol. Chem. 256, 10967-10972 13. Nudelman, E., Hakomori, S.,Kannagi, R., Levery, S., Yeh, M.Y., Hellstrom, E., K. and Hellstrom, I. (1982) J. Biol. Chem. 257,12752-12756 14. Hakomori, S.,Nudelman, E., Levery, S., Solter, D., and Knowles,
B. B. (1981) Biochem. Biophys. Res. Commun. 100, 1578-1586 15. Gooi,H. C., Feizi, T., Kapadia, A., Knowles,B.B., Solter, D., and Evans, J. M. (1981) Nature (Lond.) 2 9 2 , 156-158 16. Brockhaus7M., Ma@ani, J. L.,Blaszczyk7 M.9 SWewski, z-, Koprowski, H., Karlsson, K.-A., Larson, G., and Ginsburg, V. (1981) J. Biol. Chem. 2 5 6 , 13223-13225 17. Pukel, C. S.,Lloyd, K. O., Travassos, L. R.,Dippold,W. G., Oettgen, H. F., and Old, L. J. (1982) J. Exp. Med. 155, 11331147 18. Abe, K., McKibbin, J. M., and Hakomori, S. (1983) J. Biol. Chem. 258, 11793-11797 19. Richert, N. D., Willingham, M. C., and Pastan, I. (1983) J. Biol. Chem. 258,8902-8907 20. Fredman, P., Richert, N. D., Magnani, J. L., Willingham, M. C., Pastan, I., and Ginsburg, V. (1983) J . Biol. Chem. 258,1120611210 21. Laemmli, U.K. (1970) Nature (Lond.) 2 2 7 , 680-685
