HYBRIDOMA Volume 11, Number 2, 1992 Mary Ann Liebert, Inc., Publishers

Identification of an Epitope Recognized by the Monoclonal Antibody PEP80 in the C-Terminal Cytoplasmic Fragment of

Glycophorin A

MARIA DUK, MARCIN

CZERWIÑSKI, and ELWIRA LISOWSKA

Department of Immunochemistry, Ludwik Hirszfeld Institute of Immunology and Experimental Therapy, Polish Academy of Sciences, Wroclaw, Poland

ABSTRACT The monoclonal antibody PEP80 (IgG1) was raised by immunization of BALB/c mice with asialo-agalacto-glycophorin from human erythrocytes. The antibody is specific for glycophorin A (GPA) and reacts strongly with the GPA-derived tryptic peptide which is the C-terminal cytoplasmic portion of GPA, containing amino acid residues 102-131. Using the smaller chymotryptic fragments of this peptide and a set of solid phase-synthesized peptides allowed to establish that the MAb PEP80 is directed against an epitope comprising amino acid residues 112-121 of GPA. The peptides terminated with 120th or 119th amino acid residue were slightly less active, and the minimal structure which still gave a weak reaction with the anti¬ body was the sequence of amino acid residues 112-118. The MAb PEP80 did not bind to live human erythroleukemic K562 cells, but showed a strong binding to the cells permeabilized with methanol. INTRODUCTION A (GPA), a major sialoglycoprotein of human erythrocyte membra¬ consists of 3 domains: the N-terminal portion located outside the cell (a.a. 1-72), which is densely glycosylated up to the 50th amino acid residue, the hydrophobic transmembrane fragment (a.a. 73-95), and the cytoplasmic tail (a.a. 93-131) [1], Due to a general interest in surface antigens of erythrocytes, most anti-GPA monoclonal antibodies described so far are directed against various outer fragments of the molecule [2,3]. Epitopes in the glycosylated region of GPA usually show a direct or indirect dependence on the oligosaccharide chains, and for some studies it is convenient to have an antibody reacting with GPA polypeptide chain with sensitivity independent of its glycosylation level. This require¬ ment could be fulfilled by the antibodies directed against nonglycosylated portions of GPA. However, the antibodies recognizing peptidic epitopes in the nonglycosylated outer segment of GPA (a.a. 51-72) have a limited applicability, since usually they do not react or react weakly with the isolated GPA [4,5]. It happens most probably due to sensitivity to denaturation of this region of the

Glycophorin

nes,

181

molecule, or(and) due to the contribution of membrane lipids to these epitopes. This disadvantage does not seem to concern the antibodies against epitopes in the nonglycosylated cytoplasmic tail of GPA. One such antibody (233) was already applied by Morrow and Rubin [6] for studies on biogenesis of GPA in K562 human erythroleukemic cells, and another one (BRIC163) was used by King et al. [7] for characterization of variant glycophorins. We describe here a new monoclonal antibody PEP80 which reacts with a high affinity with an epitope in the C-terminal cytoplasmic portion of GPA. The exact location of this epitope has been established with the use of a set of synthetic peptides. MATERIALS AND METHODS

Glycophorin

and Glycophorin-Derived Products The crude glycophorin (mixture of glycophorins A, and C) was isolated by the phenol extraction of membranes of outdated human erythrocytes [8] which were supplied by the District Blood Transfusion Center in Wroclaw. To remove sialic acid, the glycophorin was hydrolyzed in 0.025 M sulfuric acid for 4 h at 60°C, and the sample was dialyzed and lyophilized. The subsequent release of galactose from asialoglycophorin was performed by Smith degradation, which included per¬ iodate oxidation, reduction with sodium borohydride, and hydrolysis of the sample in 0.025 M sulfuric acid for 1 h at 80°C [9]. Glycophorin A (GPA) was purified from crude glycophorin by gel filtration in the presence of SDS [10]. Tryptic fragments of GPA were obtained by digestion of GPA with TPCK-trypsin (Siqma) and fractionation of the products on the Sephadex G-150 column [11]. Some peptides were additionally purified in the high-pres¬ sure liquid Chromatograph (HPLC, LKB, Bromma, Sweden) on the Aquapore RP-300 col¬ umn (4x100 mm, 10 µ , Brownlee Labs., Santa Clara, CA, USA), using a gradient of (A)0.05% trifluoroacetic acid (TEA) in water and (B)0.1?ó TEA in 70% acetonitrile.

from K562 Cells The human erythroleukemic cell line K562 was obtained in 1978 from Dr. L.C. Anderson (University of Helsinki). Glycophorin from K562 cells was solubilized and partially purified by dissolving the cell pellet in 1% SDS and by treatment of this solution with trichloroacetic acid (TCA, 5%), as described by Silver et al. [12]. The SDS-TCA supernatant was exhaustively extracted with ether (to re¬ move TCA) and used as a crude K562 glycophorin preparation.

Glycophorin

Monoclonal

Antibody

PEP80 immunized with asialo-agalacto-glycophorin and antibody producing hybridoma clones were obtained as described elsewhere [9]. The super¬ natants from the growing clones were screened by agglutination of untreated and desialylated erythrocytes (RBC) and by binding to ELISA plates coated with un¬ treated, asialo-, and asialo-agalacto-glycophorin. The selected antibody PEP80 did not agglutinate RBC, but was bound in ELISA to the three glycophorin prepara¬ tions tested. The cells of the hybridoma clone secreting the MAb PEP80 were grown intraperitoneally in BALB/c mice and the collected ascites fluids were used for experiments. The MAb PEP80 was typed as IgG1 by double immunodiffusion in agarose gel, using monospecific anti-mouse Ig chain-type sera purchased from Bionetics

BALB/c

mice

were

(Kensington, MD, USA).

Microtiter Plate Enzyme-Linked Immunosorbent Assay (ELISA) The plates (Nunc, MaxiSorp, Vienna, Austria) were coated with untreated or modified glycophorin (1 µg/50 µ /well), and binding of the MAb PEP80 was deter¬ mined with alkaline phosphatase-conjugated rabbit antibodies against mouse Ig

(Dakopatts, Copenhagen, Denmark) and Sigma 104 Phosphatase Substrate Tablets, as described earlier [10]. To test the inhibition of the mAb binding, the samples of serially diluted inhibitors were mixed with an equal volume of the MAb PEP80 solution (ascites fluid diluted 1/5000) and incubated for 1 h at 20°C; binding of the mAb in these samples to glycophorin-coated ELISA plates was determined as de¬ scribed above. All tests were done in duplicate.

182

Immunoblotting

Membranes prepared from untreated RBC, or from RBC treated (under routine conditions) with proteases were submitted to SDS-polyacrylamide gel electrophoresis (SDS-PAGE) in 10?ó gel [13]. The fractionated material was electrophoretically transferred to nitrocellulose BA85 (SchleicheraSchuell, Dassel, Germany) [14]. The blots were quenched in 5% human serum albumin for 1 h at 37°C, and then were incubated in the MAb PEP80 solution (ascites fluid diluted 1/1000) overnight at 4°C. After washings, the blots were overlayed with horseradish peroxidaseconjugated goat antibodies against mouse Ig (Heintel, Vienna, Austria) for 1 h at 20°C, and finally with 4-chloro-1-naphtol (Sigma) solution as a substrate. The antibodies were diluted TBS (0.05 M Tris-HCl/0.15 M NaCl, pH 7.0) which was also used for washing the blots between the incubations. Amino Acid Analysis Peptides were hydrolyzed in 6 M HC1 for 22 h at 115°C in ampoules sealed under a decreased pressure. The samples were dried in a vacuum, and amino acids were analyzed as fluorescent-derivatives according to the method of Rajendra [15]. Briefly, the amino acids were modified with o -phthaldialdehyde(Sigma) and ana¬ lyzed by HPLC (Waters, Bedford, MA, USA), using the Spherisorb 0DS-II column (4x30 mm, 3µ , Phase separations, Queensferry, Great Britain). The fluorometric detection was done using excitation at 338 nm and emission at 425 nm. The stan¬ dard mixture of amino acids (Pierce, Rockford, IL, USA), which was treated iden¬ tically as the peptides, was used as a reference sample. Epitope Analysis with Synthetic Peptides The epitope scanning kit was purchased from Cambridge Research Biochemicals (Northwich, England). Peptides were synthesized on chemically derivatized plastic pins by the stepwise elongation of the peptide chain from C- to N-terminus, using Emoc-amino acid active esters, as recommended by the manufacturer. The amino groups of N-terminal amino acid residues of the synthesized peptides were acetylated. Binding of the MAb PEP80 to the pins with immobilized peptides was deter¬ mined by the microtiter plate ELISA, according to the method of Geysen et al.[161 using the horseradish peroxidase-conjugated antibodies against mouse Ig (Heintel, Vienna, Austria) and Or-pherryidiamine (Sigma) as a substrate. The results are mean values of at least two determinations, each on duplicate pins. of the MAb PEP80 to K562 Cells The K562 cells were treated with cold methanol for 30 min at 4°C. The 50 µ samples of cell suspension (10 cells/ml) were mixed with an equal volume of the MAb PEP80 solution and incubated for 30 min on ice. After washings the cells were incubated with FITC-conjugated goat antibodies to mouse Ig (Sigma, diluted 1/50), and fluorescence was measured in a fluorescence-activated cell sorter equipped with an argon ion laser operating at 488 nm and 500 mW power (FACstar, Becton Dickinson, Mountain View, CA, USA). A minimum 5000 events were analyzed for each sampleiya FACstar Plus Data Management System.

Binding

RESULTS AND DISCUSSION The purpose of immunization of mice with asialo-agalacto-glycophorin was to obtain anti-Tn monoclonal antibodies [9]. This immunogen also raised antibodies against peptidic epitopes of glycophorin A (GPA) and the MAb PEP80 was one of them. The MAb PEP80 did not agglutinate erythrocytes. To ascertain that the epi¬ tope recognized by the antibody is not present at erythrocyte surface, a direct and indirect (with the use of anti-mouse Ig antibodies) agglutination assay was performed at 20°C and 37°C with untreated RBC, and with RBC treated with neuraminidase, trypsin, chymotrypsin, or papain. The agglutination was not observed in any of these tests. However, the MAb PEP80 showed a distinct binding to ELISA plates coated with untreated glycophorin, asialo-glycophorin, or asialo-agalactoglycophorin. The activity of the MAb PEP80 was found to be pH-dependent: the bin¬ ding to untreated glycophorin was determined at pH range 6-9 and was highest at

183

405

3

81 9 27 ANTIBODY DILUTION

x10-

Figure 1. The pH-dependence of the MAb PEP80. Binding of the serially diluted MAb PEP80 to ELISA plate coated with glycophorin was determined at pH 6.0, 7.0, 8.0, and 9.0. For details see Materials and Methods.

U

7R

CH

PA

U

K562

Binding of the MAb PEP80 to glycophorin A and its degradation shown by immunoblotting with erythrocyte membrane components products separated by SDS-PAGE. The membranes were obtained from untreated eryth¬ rocytes (U), or from erythrocytes treated with trypsin (TR), chymotrypsin (CH), or papain (PA.). The last lane: immunoblotting with a crude glycophorin A glycophorin preparation from K562 cells. Positions ofheterodimer (AB) and A and monomer and dimer (2A) (A), glycophorin are indicated on the left side of the blot. Figure

2.

184

6 (Fig.1). The immunoblotting assay with erythrocyte membranes (Fig.2) showed 0 is specific for GPA and detects the fragments of GPA molecule that the MAb which are left in the membrane after treating erythrocytes with trypsin, chymotrypsin, or papain. The MAb PEP8Q also reacted in the blot with GPA from K562 cells (Fig.2). All these results indicated that the mAb PEP80 reacts with a peptidic epitope which is characteristic for GPA and is not available at erythro¬ cyte surface. A further mapping of the PEP80 epitope was done with tryptic fragments of GPA. The soluble products of digestion of GPA with trypsin were fractionated on the Sephadex G-150 column and the fractions were monitored for reactivity with the MAb PEP80 in ELISA-inhibition assay. A highly active material was found in the second protein peak, containing mainly the mixture of the glycopeptide T3 (a.a. 40-61) and C-terminal peptide T4 (a.a. 102-131). [1Ï ]. The active fraction T3/T4 was further factionated by HPLC and the activity was found in the last peak. The eluate corresponding to the active peak was pooled and analyzed for amino acid composition. The results (Table 1) indicated that it contained the peptide T4 which was found earlier [11] to have the following amino acid sequence (presented by one-letter amino acid code):

pH

(102)SPSDVKPLPSPDTDVPLSSVEIENPETSDQ(131)

This peptide was digested with chymotrypsin, which cleaves the bond between Leu118 and Ser119 [11], and the products were fractionated by HPLC. Two well sep¬ arated peptide fractions (T4-1 and T4-2) were obtained and analyzed for amino acid composition (Table 1). Distinct differences in the content of some amino acids allowed to establish that the fraction T4-1 contained the peptide 119-131, TABLE 1 Amino acid composition of the peptide T4 containing amino acid residues 102-131 of GPA and its two chymotryptic fragments T4-1 and T4-2. The results are presented as a number of residues per 3, .1, or 2 Val resi¬ due, respectively. The number of residues calculated from the sequence of the peptides is given in parantheses. Pro was not determined (n.d.). The amino acids which are not listed in the Table are not present in the analyzed peptides and were not detected. .

Peptides r

.

Amino acid .

.

T4-1

T4

(a.a.102-131)

(a.a. 119-131 )

T4-2

(a.a.102-118)

Number of residues

Asx Glx lie Leu

Lys Pro Ser Thr

Val

4.40 4.00 1.00 1.80 1.10 n.d. 5.20

1.50 3.00

(5) (4) (1) (2) (1) (6) (6) (2) (3)

2.33 3.71 0.91 0 0 n.d. 2.97 0.96 1.00

(2) (4) (1) (0) (0) (1) (3) (1) (1)

3.27 0 0

2.20 1.10 n.d. 3.00 0.95 2.00

(3) (0) (0) (2) (1) (5) (3) (1) (2)

the peptide 102-118. The inhibition of the MAb PEP80 by GPA, asialoGPA peptides T4, T4-1 and T4-2 is shown in Fig.3. Evaluation of the inhibi¬ tory activity was based on the comparison of molar concentrations which were re¬ quired for 50?i inhibition of the MAb PEP80 binding to GPA-coated ELISA plate. The antibody was strongly inhibited by GPA, asialoGPA and the peptide T4. The peptide was 2 to 3 times more inhibitory than the whole glycoproteins. Digestion of T4 with chymotrypsin significantly damaged the epitope, since the concentration of the peptide 102-118 required to inhibit the antibody was over 600-fold greater

and T4-2 and the

-

185

100-

80

Concentrations giving 50% inhibition of binding

o t-

¡íg/ñii

03

20

3.55

0.12

asialoGPA

1.58

0.07

T4

0.11

T4-1

»150

T4-2

0.1

1

µ

GPA

0.035 »100 22.3

39.8

x/xiyy\7i7 100

10

CONCENTRATION OF INHIBITOR

(µ /ml)

3. Inhibition of binding of the MAb PEP80 to glycophorin-coated ELISA plate by serially diluted GPA (·), asialoGPA (o), and the pepti¬ des T4 (A), T4-1 (V) and T4-2 ( ). Molar concentrations of inhibitors were calculated taking into account the molecular weights of the pep¬ tides and approximate molecular weights of GPA (30,000) and asialoGPA (22,000) monomers, deduced from their primary structures.

Figure

109 LPSPDTDV

PSPDTDVP SPDTDVPL PDTDVPLS DTDVPLSS TDVPLSSV DVPLSSVE PSPDTDVPLS SPDTDVPLSS PDTDVPLSSV PDTDVPLSSVE PDTDVPLSSVEI 123

0.2

0,4

0,6

0,5

1.0

1.2

490 Figure 4. Binding of the MAb PEP80 to synthetic peptides attached to plastic pins. The binding was measured by ELISA, as described in Mate¬ rials and Methods. The peptides, which correspond to various fragments of GPA polypeptide chain within amino acid residues 109-123, are pre¬ sented by one-letter .amino acid code in front of each bar.

186

than that of the peptide 102-131, and the second fragment 119-131 was inactive. These results suggested that the epitope PEP80 is located in the peptide 102-131 on both sides of the chymotryptic cleavage site at Leu118. However, a larger part of the epitope, giving a detectable reaction with the antibody, must be present in the peptide 102-118. This conclusion was fully confirmed by the epitope analysis with a set of peptides obtained by the multipin peptide synthesis (Fig.4). Among 7 overlapping octapeptides covering the sequence of amino acids 109-122 of GPA, only the pep¬ tide 112-119 showed a high binding of the MAb PEP80 and the peptide 111-118 showed an approx. 3-fold lower binding, whereas the other peptides were inactive. A strong binding of the antibody to one octapeptide only indicated that the epi¬ tope may consist of 8 or more amino acid residues. To check whether additional amino acid residues on both sides of the sequence 112-119 contribute to the epi¬ tope, several longer peptides were synthesized and tested (Fig.4). The results showed that addition of one or two amino acid residues at the N-terminal end, or one residue at the C-terminal end of the peptide 112-119 did not change its bin¬ ding ability. However, elongation of the peptide 112-119 by two residues at the C-terminal end improved the binding of the MAb PEP80 by approx. 30%, and this maximal binding remained unchanged after a further elongation of the peptide in~ this direction. Altogether, the results obtained allow to conclude that the epi¬ tope PEP80 is formed by the decapeptide sequence of amino acid residues 112-121 of GPA, and that the heptapeptide sequence 112-118 is a minimal structure re¬ quired for a detectable reaction with the antibody. The latter contusion results from a weak binding of the MAb PEP80 to the synthetic peptide 111-118 (Fig.4) and from a weak inhibitory activity of the proteolytic fragment 102-118 of GPA (Fig.3). The epitope PEP80 has the following sequence:

Pro( 112)-Asp-Thr-Asp-Val-Pro-Leu-Ser-Ser-Val(121)

The presence of two Asp residues in the epitope and pH-dependence of the MAb PEP80 (Fig.1) suggest that ionic interactions between these Asp residues and positively charged residues in the antibody binding site contribute to the Ag-Ab reaction. It is not known whether the other monoclonal antibodies against the cytoplasmic tail of GPA [6,7] have specificity similar to that of PEP80, since the epitopes for the MAbs 233 and BRIC163 have not been characterized.

Fi

PP?

t

h*P?

FB9i

PP?

.

IVa'

FLURESCENCE INTENSITY

Figure 5. Flow cytometry analysis of binding of the MAb PEP80 to K562 cells permeabilized with methanol. The MAb PEP80 (ascites fluid) was used diluted 1/50 (rare dots), 1/200 (dense dots) and 1/1000 (solid line). C control cells treated with normal mouse serum diluted 1/20. -

187

Raising and characterization of the MAb PEP80 were done with the use of gly¬ cophorin isolated under denaturing conditions, or treated with SDS (immuno¬ blotting), and the results obtained did not supply an evidence whether the anti¬ body reacts with native GPA inserted in cell membranes. Therefore, the binding of the MAb PEP80 to human erythroleukemic K562 cells was measured in the fluores¬ cence-activated cell sorter. Accordingly to the intracellular location of the epitope, the antibody did not bind to untreated K562 cells (the result not shown) but was strongly bound to the cells permeabilized with methanol (Fig.5). In conclusion, the MAb PEP80 is specific for the defined fragment of the cytoplasmic C-terminal portion of GPA and reacts strongly with GPA in cell mem¬ branes, isolated GPA, its C-terminal proteolytic fragment and synthetic peptideSj in all assays used. It shows that the epitope is not sensitive to denaturation. Low concentrations of antigens required for the antibody inhibition indicate its high affinity. All these properties make the MAb PEP80 a specific and highly sen¬ sitive reagent for GPA polypeptide chain, reacting independently of the antigen glycosylation level. ACKNOWLEDGMENT The authors thank dr. Danuta Dus (Department of Tumor for K562 cells and for performing the flow cytometry

Institute)

Immunology of experiments.

our

REFERENCES

1. Furthmayr H. & Marchesi V.T. (1983) Glycophorins

isolation, orientation and specific domains. Meth. Enzymol. 96: 268-280. Lisowska E. (1988) Antigenic properties of human erythrocyte glycophorins. In: The Molecular Biology of Complex Carbohydrates, Wu A.M. ed.,Plenum Press, New York & London, pp. 265-315. Anstee D.J. & Lisowska E. (1990) Monoclonal antibodies against glycophorins and other glycoproteins. J. Immunogenet. 17: 55-68. Dahr W., Wilkinson S., Issitt P.D., Beyreuther K., Hummel M. & Morel P.(1986) High frequency antigens of human erythrocyte membrane sialoglycoproteins. and Wr antigens. Biol. Chem. Hoppe Seyler, III. Studies on the En ER, Wr localization of

2.

3. 4.

-

,

367: 1033-1045.

5. Wasniowska K., Schroer K.E., McGinniss M.H., Reichert CM.A Zopf D. Monoclonal antibodies against glycophorin A. Hybridoma 7: 49-54. 6. Morrow

7.

B.

&

Rubin

S.C. (1987) Biogenesis of glycophorin A J. Biol. Chem. 262: 13812-13820.

erythroleukemia cells. King M.J., Poole J. &

the classification of fusion 29: 106-112.

Anstee D.J.

(1989)

Miltenberger

An application of series of blood group

in

K562

(1988). human

immunoblotting in antigens. Trans¬

8. Lisowska E., Messeter L., Duk ., Czerwinski M. & Lundblad A. (1987) A mono¬ clonal anti-glycophorin A antibody recognizing the blood group M determi¬ nant: studies on the subspecificity. Molec. Immunol. 24: 605-613.

M., Steuden I., Dus D., Radzikowski C & Lisowska E. An application of chemically desialylated and degalactosylated glycophorin for induction and characterization of anti-Tn monoclonal antibodies. Glycoconjugate J., sub¬

9. Duk

mitted.

10. Wasniowska K., Reichert CM., McGinniss M.H., Schroer K.R., Zopf D., Lisowska E., Messeter L. & Lundblad A. (1985) Two monoclonal antibodies highly speci¬ fic for blood group

determinant.

Glycoconjugate

188

J. 2: 163-176.

Tomita M., Furthmayr H. & Marchesi V.T. (1978) Primary structure of human erythrocyte glycophorin A. Isolation and characterization of peptides and complete amino acid sequence. Biochemistry 17: 4756-4769. 12. Silver R.E., Adamany A.M. & Blumenfeld 0.0. (1987) Glycophorins of human erythroleukemic K562 cells. Arch. Biochem. Biophys. 256: 285-294. 13. Laemmli U.K. (1970) Cleavage of structural proteins during the assembly of bacteriophage T4. Nature (London) 227: 680-685. 14. Towbin H., Staehelin T. & Gordon J. (1979) Electrophoretic transfer of pro¬ teins from polyacrylamide gel to nitrocellulose sheets: procedure and some applications. Proc. Nati. Acad. Sci. U.S.A. 76: 4350-4354. 15. Rajendra W. (1987) High performance Chromatographie determination of amino acids in biological samples by precolumn derivatization with o-phthaldialde~ hyde. J. Liquid Chromât. 10: 941-955. 16. Geysen H.M., Rodda S.J., Mason T.J., Tribbick G. & Schoofs P.G. (1987) Stra¬ tegies for epitope analysis using peptide synthesis. J. Immunol. Meth. 102: 11.

159-174.

Address

reprint requests

to:

Elwira Lisowska Institute of Immunology and

Experimental Therapy

Czerska 12 53-114 Wroclaw, Poland

189