Monoclonal Antibodies Directed against the Human Leb Blood Group ...

THEJOURNALOF BIOLOGICAL CHEMl6TRY

Vol. 256, No. 24. Issue of December 25, pp. 13223-13225, 1981 Primed in U.S.A .

Monoclonal Antibodies Directed against the Human Leb Blood Group Antigen* (Received for publication, June 16, 1981)

Manfred BrockhausSfj,John L. MagnaniS, Magdalena Blaszczykfl,Zenon Steplewskil, Hilary Koprowskin, Karl-Anders Karlssonll, Goran Larsonll, and Victor Ginsburg$ From the +National Institute of Arthritis, Metabolism and Digestive Diseases, National Institutesof Health, Bethesda, Maryland 20205, the 1Wistar Institute of Anatomy and Biology, Philadelphia, Pennsylvania 19104, and the 1) Department of Medical Biochemistry, University of Goteborg, S-400 33 Goteborg, Sweden

stable hybridoma clones, 19 of which secreted antibodies with apparFourmonoclonalantibodiesproducedbyhybridomas obtained from a mouse immunizedwith a human ent specificity for colon adenocarcinoma cells (1).Antibodies 1116NS10, 1116NS-33a,and 1116NS-38aare of the immunoglobulin M type. adenocarcinoma cell line SW1116 (Koprowski,H., Antibody 1116NS-43ais of the immunoglobulin Gyl type. Serum-free Steplewski, Z., Mitchell, K., Herlyn,M., Herlyn, D., and supernatant fluids from hybridoma cell cultures (7) containing apFuhrer, P. (1979) Somatic Cell Genet. 5, 957-972) are proximately 10 pg/ml of antibody were used for the experiments directed against the Leb antigen of the human Lewis described here. Oligosaccharides were isolated from human milk (8); bloodgroupsystem.Their specificities were estab- Leb-active ceramide hexwaccharide, lished by bindingstudies using purified Leb-activecerFuca-l-2Galpl-3GlcNAc/l1-3Galp1-4CIc~l-1Cer, amide hexasaccharide and by hapten inhibition studies 4 involving oligosaccharides obtained from human milk.

I

Fucc~l

Nineteen monoclonal antibodies produced by hybridomas obtained from a mouse immunized with a human colon adenocarcinoma cell line have an apparentspecificity for cell lines derived from human intestinal tumors (1).The antigen for one of these antibodies (1116NS-52a)has recently been characterized as a monosialoganglioside (2). Four other antibodies (1116NS-10,1116NS-33a,1116NS-38a, and 1116NS-43a) are identified by the data in the present paper as being direct,ed against the Leh antigen of the human Lewisblood group system (3).The Leb antigen contains the terminal sugar sequence F u c c x ~ - ~ G ~ ~ ~ ~ -. ~ . GIcNA~. 4

I

Fucal

and occurs in the glycolipids and glycoproteins of approximately 75% of the population who belong to the Le(a-b+) blood group (4). The remaining 25%of the population are unable to synthesize the Leb sequence of sugars because they lack either the fucosyltransferase responsible for the formation of Fucal-2Gal linkages (5) and belong to the Le(a+b-) blood group or the fucosyltransferase responsible for the formation ofFuccul-4GlcNAc linkages (6) and belong to the Le(a-b-) blood group. EXPERIMENTALPROCEDURES

Materials-The monoclonal antibodies were produced by hybridomas obtained from a mouse immunized with the human colon adenocarcinoma cell line SW1116. The cellfusion resulted in 76

* This work was supported in part by research grants CA-10815 and CA-21124 from the National Cancer Institute, grant RR-05540 from the Division of Research Resources, and funds from the W. W. Smith Foundation. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 5 Recipient of a Fellowship from the Deutsche Forschungsgemeinschaft.

and anALebceramide heptasaccharide,

GalNAcal-3Gal/ll-3GlcNAc~l-3Gal/ll-4Clc~l-1Cer, 2

4

I

I

FucalFucal

were isolated from adult human small intestine (9). Their structures were c o n f i i e d with methods described elsewhere (10).The ALeh glycolipid does not react with commercial human Leh or A antisera but reacts with the Siedler antibody which is an AILehantibody (3, 11).’ Other glycolipids were commercial products (Supelco, Bellafonte, PA). Lipid extracts of colon carcinoma line SW1116 and melanoma cell line WM 9 were prepared by the method of Folch et al. (12). The neutral glycolipids of human meconium were isolated as previously described (2, 10). Anti-mouse Fab from rabbit ?-labeled F(ab‘)n fragment (about 40 gCi/pg) was obtained from The Radiochemical Centre, Amersham, England. Solid Phase Radioimmunoassay-Binding of antibody to glycolipid is assayed by the procedure of Young et at. (13) with modifications as follows: glycolipid in 20 p1 of methanol is added to wells in a polyvinylchloride microtiter plate (Dynatech, Alexandria, VA) and the solutions dried by evaporation. After 30 min, the wells are filed with phosphate-buffered saline (0.15 M NaC1, 0.01 M sodium phosphate, pH 7.2) containing 1% bovine serum albumin (buffer A). After 2 h, the wells are emptied and to each is added 20 gl of buffer A and 5 pl of the monoclonal antibody solution. After covering the wells with p a r a f i , the microtiter plate is slowly rotated at a 45’ angle to facilitate diffusion for 6 h at 22 “C. The wells are emptied, washed once with buffer A, and then to each well is added about 13,000 cpm of anti-mouse Fab from rabbit, “‘I-labeled F(ab)n in 20 pl of buffer A. After rotation of the microtiter plate for an additional 12 h, the wells are washed six times with cold phosphate-buffered saline, cut from the plate, and assayed individually for ‘‘$1in a Auto-Gamma spectrometer. Thin Layer Chromatography and Autoradiography-Thin layer chromatography is carried out on HPTLC plates precoated with Silica Gel 60 (E. Merck, West Germany) developed with chloroformmethanol-water (60:40:9 byvolume). Glycolipid antigens are detected by autoradiography using a previously published procedure (2) with minor modifications as follows: after chromatography, the chromat-

13223

C. A. Tilley, unpublished data.

Monoclonal Antibodies against Human Antigen the Leh

13224

Cell Extract lpg w e 1 welght 01 cells extracted1

Glycohpnd lngl

FIG. 1. Binding of antibody 1116NS-10to totallipid extracts of cultured cells and to isolated glycolipids. Solid phase radioimmunoassays were carried out as described under “Experimental Procedures.” A , binding of antibody to extracts of colorectal carcinoma cells SW1116 ( 0 ” Q andtoextracts of melanoma cells WM9 1-;( B, binding of antibody to Leh-active ceramide hexasaccharide (M and ) GalNAc~l-3Galnl-4-Gal~l-4Glc~l-lCeramide (globoside) (-).

L,

0.1

1

10 IN0 I

Oligosaccharide I n ”

FIG. 3. Inhibition of binding of antibody 1116NS-10to

activeceramidehexasaccharide

- (:XI

- C’fH

-c

?I I

- c D 1;1

Leb-

by oligosaccharides. Solid

phase radioimmunoassays were carried out as described under “Experimental Procedures” using 0.5 pg of Le1’-activeceramide hexasaccharide perwell. LNT, lacto-N-tetraose;L N F Z,lacto-N-fucopentaose I; L N F ZI, lacto-N-fucopentaose 11; LND Z, lacto-N-difucohexaose I. For structures, see Footnote3. drying, the chromatogram is covered with monoclonal antibody solutiondiluted 1:4 withbufferA andincubated 6 h a t 4 “C. The chromatogram is washedby dipping in four successive changes of cold phosphate-buffered saline for 1 min each time and thencovered with buffer A containing about IO” cpm/ml of anti-mouse Fab from rabbit, ”‘1-labeled F(ab)?. After 12 h a t 4 “C, the plates are washed by dipping in six successive changes of cold phosphate-buffered saline for 1 min each time, dried, and exposed to XR-5 X-ray film (Eastman Kodak, Rochester, NY). RESULTS AND DISCUSSION

Total lipid extracts of adenocarcinoma cell line SW1116 but not melanoma cells WM 9 contain antigen for monoclonal antibody 1116NS-10 as evidenced by solid phase radioimmunoassay (Fig. lA). 1 3 The glycolipids in the extractof adenocarcinoma cells were FIG. 2. Binding of antibody 1116NS-10to thin layer chromatograms of the neutral glycolipids of meconium and of fractionated by DEAE-Sepharose chromatography and thin purified Leb-active ceramide hexasaccharide as detected by layer chromatography aspreviously described ( 2 ) followed by autoradiography as described under *‘ExperimentalProceautoradiography. The major antigen of antibody 1116NS-10 dures. Lane 1, 1.0 pg of neutral glycolipids from the meconiumof an in the extract was revealed by this technique to be a neutral individual belonging to the Le(a+b-) blood group; lane 2, 1.0 pg of glycolipid with the mobility of GM,under the conditions of neutral glycolipids from pooled meconia; and lane 3, 0.5 pgof Le”thin layer chromatography described under “Experimental activeceramide hexasaccharide. The positions of somestandard gangliosides and neutral glycolipids after chromatography are shown Procedures.” An antigen with the same mobility is present in on the right. The abbreviations used are: CMH, Galpl-ICeramide; the neutral glycolipid fraction obtained from pooled meconia CDH. Gal~l-4Glc~l-1Ceramide;CTH, Galal-4Gal~1-4Glc~l-1Cerincluding samples from individualsbelonging to the Le(a-b+) amide;paragloboside, Gal~l-4-GlcNAc~l-3Gal~l-4Glc~l-l-Ceramblood group (Fig. 2 , lane 2) but not in the neutral glycolipid ide; G.w, Gal~l-3GalNAc~l-4[NeuNAcn2-3]Gal~l-4Glc~l-lCeramfraction obtained from an individual belonging to the ide; and G,,,o, NeuNAca2-3Gal/31-3GalNAc~1-4[NeuNAca2-3]. Le(a+b-) blood group (Fig. 2, lane I). As this finding sugGal~1-4Glc~l-1Ceramide. gested that antibody 1116NS-10 recognizes the Leh antigen, purified Leh-active ceramide hexasaccharide was tested for ograms are dried and dipped in petroleum ether containing 0.05% polyisobutylmethacrylate (Polysciences, Warrington, PA).’ After binding by solid phase radioimmunoassay. As shown in Fig. 1B, the glycolipid effectively binds antibody 1116NS-10 with ’This treatment reducesunspecific binding of antibodies and helps maximal binding attained by 5 ng of glycolipid. No binding prevent the detachment of silica gel from the chromatograms during was observed with other glycolipids such as globoside (Fig. subsequent steps in the procedure. The silica gel of some HPTLC 1B). Similar results were obtained with antibodies 1116NSplates detach even after treatment with polymer. Glass plates coated 33a, 1116NS-38a, and 1116NS-43a. None of the four bound to with Silica Gel 60 F254 (E. Merck, West Germany) or Si-HPF (J. the ceramide heptasaccharidethatcontainsthesugar seBaker, Phillipsburg, NJ)can also be used. -Origin

Monoclonal Antibodies Human against the quences of both the A antigen and theLeb antigen. Autoradiography of the Le"-active ceramide hexasaccharide after thin layer chromatography using antibody 1116NS-10is shown in Fig. 2, lane 3. Its mobility is the same as the major antigen in meconium and in colorectal carcinoma cells. The slower running, minor antigen in meconium (Fig. 2, lane 2) is probably Leb-active ceramide octasaccharide reported previously (10, 14). Autoradiography using antibody 1116NS-43 revealed the same two antigens inmeconium as those detected by antibody 1116NS-10.Antibodies 1116NS-33aand 1116NS38a detected several additionalantigens. The specificity of the antibodies for the Leb antigen are confirmed by inhibition studies (Fig. 3). Binding of antibody 1116NS-10to Leb-active ceramide hexasaccharide is inhibited by lacto-N-difucohexaose I, which contains the Leb sequence of sugars, but not by the structurally related oligosaccharides lacto-N-tetraose, lacto-N-fucopentaose I, and lacto-N-fucopentaose I1 at theconcentrations t e ~ t e dBinding .~ of antibody 1116NS-10 isinhibited 50% by 0.8mM lacto-N-difucohexaose I. Binding of the other antibodiesto Leb-active ceramide hexasaccharide is also specifically inhibited by lacto-N-difucohexaose I. Inhibition of 50% is achieved by the following concentrations of oligosaccharide: with antibody 1116NS-33a, 0.2 mM; with antibody 1116NS-38a,0.3 mM; and with antibody 1116NS-43a,0.6 mM. The apparent specificity of the antibodies for colorectal carcinoma cell lines (1)may be explained by the high levels of Le"- and Leb-active glycolipids that occur in some adenocarcinomas of colon and pancreas regardless of the Lewis blood group of the donor (14). Monoclonal anti-A and anti-Bantibodies are useful reagents for routine blood grouping (15, 16). Preliminary experiments with antibodies 1116NS-10, 1116NS-33a,and 1116NS-38a indicate that they agglutinate Le(a-b+) erythrocytes but not Le(a+b-) or Le(a-b-) erythrocytes. More extensive studies,

Antigen Le

13225

however, are required to evaluate the usefulness of these antibodies for routine blood grouping. Monoclonal anti-Leb antibodies should be useful for histochemical studies and may offer an advantage over affinitypurified polyclonal antibodies with Lehspecificity which have recently been used for this purpose (17). Acknowledgments-We are grateful to Mary H. McGinniss for help in blood typing. We would also like to thank Urner Chase for technical assistance and JulieReddick for preparing the manuscript.

REFERENCES 1. Koprowski, H., Steplewski, Z., Mitchell, K., Herlyn, M., Herlyn, D., and Fuhrer, P. (1979)Somatic Cell Genet. 5,957-972 2. Magnani, J . L., Brockhaus, M., Smith, D. F., Ginsburg, V., Blaszczyk, M., Mitchell, K.F., Steplewski, Z., and Koprowski, H. (1981)Science 212,55-56 3. Race, R. R., and Sanger, R. (1975)Blood Groups in Man, 6th Ed, pp. 323-349,Blackwell, Oxford 4. Watkins, W. M. (1980)Adu. Hum. Genet. 10, 1-135 5. Shen, L.,Grollman, E. F., and Ginsburg, V. (1968)Proc. Natl. Acad. Sci. U. S. A . 59,224-230 6. Grollman, E. F., Kobata, A., and Ginsburg, V. (1969)J . Clin. Znuest. 48, 1489-1494 7. Chang, T. H., Steplewski, Z., and Koprowski, H. (1980)J. Immunol. Methods 39,369-375 8. Kobata, A. (1972)Methods Enzymol. 28,262-271 9. McKibbin, J . M. (1978)J . Lipid Res. 19, 131-147 10. Karlsson, K.-A., and Larson, G. (1981)J. Biol. Chem. 256, 35123524 11. Seaman, M. J., Chalmers, D. G., and Franks, D. (1968)Vox Sang. 15,25-30 12. Folch, J., Lees, M., and SloaneStanley, G. H. (1957)J. Biol. Chem. 226,497-509 13. Young, W. W., Jr., MacDonald, E. M. S., Nowinski, R. C., and Hakomori, S. (1979)J. Exp. Med. 150,1008-1019 14. Hakomori, S., and Andrews, H. D. (1970)Biochim. Biophys.Acta 202, 225-228 15. Voak, D., Sacks, S., Alderson, T., Takei, F., Lennox, E., Jarvis, J., The structures of these oligosaccharides are as follows: lacto-AIMilstein, C., and Darnborough, J. (1980)Vox Sang. 39, 134difucohexaose I, Fuccvl-2Galpl-3[Fucal-4]GIcNAc~l-3Gal~1-4Glc; 140 lacto-N-tetraose, G@1-3GlcNAc~l-3Gal~l-4Glc;lacto-N-fucopen16. Sacks, S. H., and Lennox, E. (1981)Vox. Sang. 40,99-104 and lacto-N-fucotaose I, Fucal-2Galpl-3GlcNAc~l-3Gal~l-4Glc; 17. Lemieux, R. U., Baker, D. A., Weinstein, W. M., and Switzer, C. pentaose 11, G@1-3[Fucal-4]GlcNAc~1-3Gal~1-4Glc. M.(1981)Biochemistry 20, 199-205.