Biochimica et Biophysica Acta, 1080(1991) 1-10

© 1991 ElsevierScience PublishersB.V. All rights reserved 0167-4838/91/$03.50 ADONIS 016748389100296D

BBAFRO 34005

Purification and partial characterization of a shed 66 kDa melanoma-associated antigen identified by autologous antibody * D a n i e l R . V l o c k t, D o n a l d J. A u l 2, A v i s h a g T o p o r o w i c z ~, J. P h i l i p M c C o y and William E. Brown 2 i Division of Medical Oncology, Department of Pattmlogy, Pittsburgh Cancer Institute, Uni,.~ersityof Pittsburgh School of Medicine and VA Medical Center Pittsburgh, PA (U.S.A.) and 2 Department of Biological Sciences, Carnegie Mellon Unirersity. Pittsburgh, PA (U.S.A.)

(Received22 Februaw 1991)

Key words: Melanoma:Tumor-associatedantigen;Autologousantibody;Immunology;Serology We have previously reported the isolation of a 66 kDa melanoma.associated antigen, identified by autologons antibody, in serum and unfractionated spent tissue culture media by Western blot analysis. The antigen, (letected by autologous serum Sl$O, was found to be broadly represented on melanoma, glioma, renal cell carcinoma, neuroblastoma and head and neck carcinoma cell lines. $150 did not react with bladder or colon carcinoma, fetal fibroblasts, pooled platelets, lymphocytes and red blood cells, autologous cultured lymphocytes or fetal calf serum. To further characterize the antigen, spent tissue culture media, obtained from autologans melanoma cell fine, Y-Mel 84:420, was separated by an isoclectric focusing column. Unabsorbed control serum ~;1$0 was noted to have a maximum titer of 1: 2040 against autologous melanoma cells as measured by protein A hemadsorption. Following isoclectric focusing the greatest decrease in autologous antibody titer (30-fold) occurred with fractions having a p l between 2 and 3. Further resolution of the antigen was accomplished with high-pressure ion-exchange chromatography. One of these fractions showed a significantly higher concentration of antigen and was distinctly resolved from bulk serum albumin. Subsequent Western blot analysis, with autologons antibody, of the isolated antigen-containing fraction, confirmed the presence of a single 66 kDa hand. Exposure of the antigen, purified by high-pressure ion-exchange chromatography, to neuraminidase ablated recognition by autologons antibody and suggests that sialic acid is present on the protein and may be part of the antigenic epitope. Binding of antigen, obtained following DEAE anion exchange chromatography, was noted to lectins derived from Tr/t/cum vulgaris, Dolichos biflorus a n d Lycopersicon esculenmm. Preparative purification of the antigen was accomplished by anion exchange followed by iectin affinity chromatography with a Do//chos biflorus column. Antigen obtained following lectin afl'mity chromatography subjected to SDS-PAGE and silver stain revealed a single band at 66 kDa. We conclude that a melanoma-associated antigen detected by autologous antibody in spent tissue culture media is an unusually acidic glycoprotein

(pl 2-3).

Introduction * Supported in part by National Institutes of Health Grant ROI CA42922. Abbreviations:FCS, fetal calf serum; PA, protein A hemadsorption; AD&U, acid dissociation and ultrafiltration; SDS-PAGE, sodium dodecyl sulfate-polyacrylamidegel electrophoresis;IEF, isoelectric focusing; HPLC, high-performanceliquid chromatography;HP-IEC, HPLC ion-exchange chromatugraphy; NC, nitrocellulose: BSA, bovine serum albumin; PBS, phosphate-bufferedsaline; TBS, trisbuffered saline. Correspondence: D.R. Vlock, Hematology/Oncology,VA Medical Center, UniversityDrive C, Pittsburgh,PA 15240, U.S.A.

The def'mition and characterization of tumor-associated antigens has been a major focus of tumor immunology for over two decades. By identifying antigens that are restricted to tumor cells it may be possible to better understand the nature of the host-tumor interaction and tailor more effective agents for immunodiagnostic and therapeutic use, A variety of antibodies have been utilized to detect tumor-associated antigens. These antibodies have included polyclonal xenogeneic

and allogeneic antisera [1-3] and more recently, monoclonal murine and human antibodies [4-10]. As an alternative to utilizing heteroantisera, we have focused our attention on the autologous immune response. The advantage of studying autologous antibody reactivity is that it may recognize antigens not apparent to heteroantisera. Because these antibodies are produced by the host, they may detect physiologically relevant antigens that could shed light on the host-tumor interaction. Antibodies against autologous tumor cell surface antigens were first reported by Carey et al. [11,12] and Shiku et al. [13,14], who used sensitive microserological assays and cultured tumor cells. These authors demonstrated antibodies against autologous cell surface antigens in a third of melanoma patients studied. Antigens described in these studies were divided into three classes: those which are restricted to the tumor of a single individual (class I); those which are shared among tumors of a similar histotype or ontogeny and rarely, if ever, found in non-neoplastic cells (class II); those which are found in non-neoplastic cells as well as tumors (class Ill). Studies in this laboratory have confirmed the presence of autologous antibody reactive against melanoma in 6 of 22 patients studied, including reaction patterns which fit classes I, I! and IlI [15]. However, due to the low incidence and weak titer of autologous antibody to melanoma questions have been raised regarding the relevance of humoral immunity in melanoma. The presence of circulating antigen in patients and the resulting formation of immune complexes may explain this. In previous work, we examined the incidence and specificity of autologous antibodies in patients with melanoma. As has been reported by others, in native serum these antibodies are usually of low incidence and titer. We demonstrated that dissociation of immune complexes by acidification and ultrafiltration of sera augments autologous antibody reactivity in the majority of cases studied [15]. Subsequent studies have shown correlations between autologous antibody titers, clinical course and prognosis [16-18]. We have previously reported the isolation of a 66 kDa antigen, Identified by autologous antibody, in serum arid unfractionated spent tissue culture media from autologous system 84:420, by Western blot analysis [18]. This antigen was sensitive to trypsin digestion but resistant to pepsin and heat inactivation. We now report the first successful purification and further characterization of a class II melanoma-associated antigen identified by the host from spent tissue culture media.

Materials a,.~ M-thods

Cell culture and cryopreservation. The acquisition, establishment and maintenance of melanoma and other

cell lines has been reported previously [15,18]. Protein A hemcdsorption (PA). Indicator cells for protein A mixed hemadsorption tests were prepared by conjugating staphylococcal protein A (Pharmacia Fine Chemicals) to the surface of selected human blood group O-Rh positive red blood cells with 0.01% CrC! 3 at pH 5.0 [19]. Target monolayers were seeded in micro-Terasaki assay plates and reacted with serial dilutions of autologous native serum, or derived fractions. After incubation of target cells with sera at 37 ° C, wells were washed x 3 with PBS containing 2% agammaglobulinemic FCS at 37 *C and 0.01 ml of an 0.15% suspension of indicator cells were added to each well. Plates were washed 2-4 × with PBS-2% agamma fetal calf serum after 45 min and positive cells read as follows: ( + ) cells were those which bear a > 50% erythrocyte rosette. The endpoint of the assay is read as the last well with 10% of target cells ( + ) .

Acid dissociation and ultrafiltration (AD&U) of serum. The method, initially described by Sjogren, et al. [20], was used to dissociate immune complexes in serum to augment autologous antibody reactivity. Briefly, a 60 ml ultrafiltration chamber and Amicon XM-100 membrane (Amicon Corporation) were used. 2 to 3 ml of prefiltered (0.45/~m, Millipore) serum was added to 50 ml of glycine-saline buffer, 0.1 M, pH 3.1, in the ultrafiltration chamber. Uitrafiltration was performed at 4°C under 10 psi N 2 until the original volume of the test solution is reached. This was repeated twice and then washed three times with 50 ml PBS and concentrated to the original volume of the test solution. The total time involved varied from 24 to 48 h depending on the serum used.

Detection of isolated antigen by competitive inhibition. To determine that the isolated fraction of spent media contained the antigen of interest, competitive binding of the isolated antigen for antibody directed against autologous melanoma cell lines was performed. Fractions tested for the presence of melanoma-associated antigen were mixed at a 1:1 dilution with autologous antibody, incubated for 30 min at 4 ° C then 30 min at 37 ° C and retested with PA against Y-Mel 84: 420 (see above). A 2-fold drop in autoiogous antibody titer was considered to be indicative of the presence of antigen in the fraction tested.

Isoelectric focusing (IEF) chromatography.

The

method described by Vesterberg and Svensson [21] was used to separate proteins in spent media. Briefly, 1 ml of concentrated, unfractionated spent media (protein concentration 300 m g / m l ) was applied to a 110 ml, water jacketed IEF column (LKB 8101 Electrofocusing Column) which was cooled to 2-3°C. The column was run at 260 v for 24 h followed by 660 v for 72 h. Following focusing the column was drained into 5 ml sequential fractions. These fractions were then dialyzed against PBS to remove ampholytes and sucrose.

The collected fractions were concentrated to 200 #1 in a Minicon Spinal Fluid Concentrator (Am~con, Danvers, MA) and the presence of antigen was determined by competitive inhibition. HPLC ion exchange chromatography (HP-IEC). HPIEC was performed after the method described by Henry [22]. 25 #1 of spent media was applied to a SynChropak AX300 column (SynChrom, Lafayette, IN) equilibrated with 0.02 N Tris-formate (pH 8.0). Following a wash with the equilibration buffer, a step gradient to 0.5 M sodium formate (pH 2.8) was applied. Collected fractions were dialyzed against PBS and tested for the presence of antigen by immunoblot and competitive inhibition.

SDS-PAGE and transfer to nitrocellulose (NC) of melanoma-associated antigens. One-dimensional SDSP A G E was performed after the method of Laemmeli [23] with 10 o n gradient gels of 7.5-15% acrylamide. Fractions were dissolved in 0.025 M Tris-HC! (pH 6,8), 2% SDS, 10% glycerol, 5% 2-mercaptoethanol with 0.001% bromphenol blue dye marker. Final sample vols. contained 5-10 /zg protein in 0.2-0.3 ml. Electrophoresis was carded out with a current of 20 mA until the bromphenol blue marker reached the margin of the gel. The methods of Towbin et al. [24] were used for transfer of proteins with a Hocfer transblot apparatus (Hoeffer Scientific Instruments, San Francisco, CA). Transfer was performed at 40 v for 2-3 h increased to 100 v during the last hour.-The NC (0.2 #.m, Scheicher and Schuell, Keene, NH) to which electrophoretic transfer was accomplished was stained immunologically and with biotinylated lectins. Immune detection of transferred proteins. Following transfer to NC, the sheet was air dried for 30 min at 20 °C and then quenched in tris-buffered saline (TBS) (pH 7.5) plus 10% BSA for 2 h or overnight on a rocker. Antibody diluted in TBS in 0.05% TWEEN-20 was overlaid on the NC paper and incubated on a rocker at 4 ° C for 16 h. The paper was washed with TBS +0.05% Tween 20 for 5 min × 5. The second peroxidase-conjugated rabbit anti-human ig antibody (Cappel), diluted with TBS plus 0.05% Tween 20 was overlaid on the NC paper and incubated at room temperature while rocking for 1 h. Following washing, the NC paper was developed in 60 mg of 4 chloro-1naphthol diluted in 20 ml methanol + 60 itl 30% H 2 0 2 diluted in 100 ml TBS for 15 rain. The reaction was stopped by removing the NC and placing it in a fresh container of d H 2 0 and washing 10 rain x 2. Blue agarose affinity chromatography. Blue agarosc (Sigma) was equilibrated in 0.01 M Tris-HCI (pH 7.8) and the 1 ml of concentrated spent media applied to the column. Serial fractions were collected and monitored at Az~0. When A~o returned to baseline, the bound material was eluted with 0.05 M Tris-HC1 (pH 7.5) plus 0.2 M NaSCN. Fractions were collected, dia-

lyzed and tested for the presence of antigen by compelitive inhibition. Lectin binding. Following transfer of the antigen to NC, the strips were quenched in TBS plus 10% BSA and lectin binding was performed. Biotinylated lectin (25 #./ml) (EY Laboratories, San Mateo, CA and Sigma, S t Louis, MO) was diluted in TBS, overlaid on the NC strips and incubated for 1 h at room temperature. The NC strips were then washed in TBS and incubated with an avidin-biotin peroxidase complex for 1 h using reagents from the Vectastain ABC Kit (Vector Laboratories, Budingame, CA) Lectin binding was visualized using diaminobenzidine as the chromagen. Neuraminidase digestion. 1 ml of antigen, purified by HP-IEC, (---10 #g) was dialyzed against dH20, iyophilized and resuspended in 0.5 ml of dilution buffer (50 mM Na Acetate, 154 mM NaCI, 4 mM CaCI2, pH 5.5) containing 0.5 m g / m i of lysozyme as a carrier protein. 5 . 1 0 -5 units of neuraminidase (~,/brio cholera isolate, Calbiochem, San Diego, CA) was added and incubated at 3 7 ° C frr 30 rain. The reaction was stopped by addition of 2.50 mM EDTA (pH 7.0) and dialyzed against P_nS. Following exposure to neuraminidase, the treated antigen was tes'ed by competitive inhibition.

Preparative purification DEAE anion.exchange chromatography. 3 ml of concentrated spent media (protein concentration 300 m g / m l ) was applied to a 1.8 x 15.5 cm Whatman DE-52 weak anion-exchange column that had been equilibrated with OC2 M tris formate (pH 8.0). The pH was changed in step-wise fashion with 0.5 M sodium formate (pH 2.8). Fractions were collected, dialyzed against PBS and tested for the presence of antigen by competitive inhibition. Lectin affinity chromatography. Follov, ing DEAE anion-exchange chromatography, 4 ml of antigen containing fractions (prgtein concentration 100 p.g/ml) were applied to a 1 cc Dolichos biflorus affinity column (E-Y Laboratories, San Mateo, CA), pre-equilibrated with PBS at 4 ° C. The antigen was incubated on the column for 15 rain and the column was washed with 5 vols. of PBS. Antigen bound to the column was eluted with 0.1 M acetic acid in 0.15 M NaCl. Fractions were collected, dialyzed and tested for the presence of antigen by competitive inhibition. Determination of yield and acticity. Measurement of protein concentration following each purification step was determined by absorption at A 2s0 with the exception of final purification step where the level was below the limits of detection and was estimated based on the appearance of a silver stained SDS-PAGE gel of the antigen [25]. Carbohydrate concentration was measured by the phenol sulfuric assay [26]. Assessment of

activity of antigen obtained at each purification step was expressed as the dilution of antigen required to give the same reduction in autologous antibody titer against melanoma cell line Y-Mcl 84:420 as unfractionated spent media.

12 2,000

t0

g_

8

E

e

2

Results

~

4

Acid dissociation and ultrafiltration. We previously reported that in native sera 5 of 12 patients demonstrated IgG reactivity against autologous cultured melanoma cells. Following ultrafiltration, 11 of 12 sera showed reactivity of IgG against autologous cultured melanoma cells [15]. We have now expanded our evaluation to a total of 21 autologous systems (Table I). In native serum, 11 of 21 patients were noted to have IgG reactivity against autologous melanoma cells. Following A D & U all but one patient was noted to have augmented autologous antibody reactivity. Normal human sera did not bind to melanoma cells before or after A D & U [151. Serologic analysis. Serologic analysis of the 66 kDa antigen detected by autologous antibody has been reported previously [18]. Briefly, following A D & U serum S150 was noted to have titer of 1 : 2048 against autologeus melanoma cell line Y-Mel 84:420. "l~e antigen, detected by autologous serum S150, was found to be broadly represented on melanoma, glioma, renal cell carcinoma, neuroblastoma, and head and neck carcinoma cell lines. S150 did not react with bladder or colon carcinoma, fetal fibroblasts, pooled platelets,

TABLE i Titers of melanoma patients' sera tested lry protein a hemadso~t;.on before and after AD&U against autologous cultured tumor Cell line

Native sera

AD&U sera

Y-Mel 81:370 Y-Me181:060 Y-Mel 81:090 Y-Mel 81 : 180 Y-Me181 : 170 Y-Me181 : 120 Y-Mel 78:010 Y-Me183:070 Y-Me182:550 Y-Mel 81:710 Y-Me184:420 P-Me186:150 P-Me187:210 P-Me187:180 P-Mel 87:183 P-Mel 88: 208 P-Me188:321 P-Mel 89:285 P-Mel 88:136 P-Me187:107 P-Me187:183

0 0 0 0 0 1:4 1:64 1:32 0 0 1:32 1:64 1:32 1:32 0 0 0 i :2 1:4 i :4 1: 1

0 1:8 1:64 1 : 128 1:256 1 : 16 1 : 128 1 : 128 1 : 128 1 : 128 1:2048 1 : 1024 1:512 1:2048 1 : 16 1:8 I : 16 1:512 I : 1024 1:512 l : 16

l 2

J

I

)

J

50 C

2

4

6

i

I

[

Q 10 12 Fro:Ion Numbw

I

[

I

J

14

16

18

20

0

Fig. 1. Absorptionof autologousantibodyreactivityfollowingisoeleetric focusingof spent tissue culture media. Control antibodyreactivity (C) was noted to have a maximum titer of 1:2048. Following incubation with IEF fractions three drops in autologous antibody titer can be seen. The most significantdrop in antibodytiter (30-fold) occurred between pH 2 and 3 (fractions14-17), a point quite distinct from BSA.

lymphocytes and red blood cells, autologous cultured lymphocytes or fetal calf serum, in addition, a 60-fold concentration of tissue culture media (F-10 plus 5% FCS), that had not been exposed to melanoma cells, failed to reduce autologous antibody titers (data not shown). lsoelectric focusing (IEF). Because of the similarity in molecular weights between the melanoma antigen and serum albumin, separation of the proteins was performed based on charge utilizing an IEF column (Fig. 1). Serum S150 (sample 'C'), had a maximum titer of 1 : 2048 against autologous melanoma cell line Y-Mel 84: 420. Three separate IEF fractions caused reductions in autologous antibody titer. Fraction 4 reduced the titer by one doubling dilution and was not felt to be significant. The remaining two, fractions 9-13 and 1417, caused significant decreases in autologous antibody titers. Fractions 9-13 correspond to a pH range of 4 - 5 and contained serum albumin ( p l 4.8-5.0) as confirmed by SDS-PAGE analysis (data not shown). However, the most significant decrease in autologous antibody titer (a 30-fold decrease) occurred br tween pH 2 and 3, a point quite distinct from serum albumin. H P L C ion-exchange chromatography (HP-IEC). Based upon the finding with 1EF of widely different p l values for BSA and the melanoma-associated antigen, HP-IEC was performed to attempt the complete resolution of these two molecules. As shown in the chromatograph (Fig. 2a), after injection of the sample a large peak is noted. The buffer was then changed (indicated by the vertical line) and a large peak was eluted which corresponds to BSA. Immediately following the BSA elution, a smaller peak can be seen (see arrow). When tested for the presence of antigen by competitive inhibition (Fig. 2b), only fraction 12, corre-

5 TABLE il

:1 a

Competltice inhibition o] autologou.s antibody re~:~lit iiy ]bllowing treatment of antigen n'ith neuraminidase +Spent media ~ unfractionated spent media, +antigen - antigen obtained following ttP-IEC, +neuraminidase control - HP-IEC antigen was exposed to the same conditions used in neuraminidasc digestion but without the addition of enzyme. Maximum titer Autologous antibody + spent media +antigen + antigen + neuraminidase + antigen + neuraminidase control

sponding to the second small peak on the chromatograph, reduced autologous antibody reactivity which is indicative of the presence of antigen.

Fraction Number

2,0OO ' b 1,000 5OO

~OO 100

C

8M

1

3

6

9 ~

12

15 18 Numbw

---200,000

---

92,000

--

70,000

O --- 46,000

--- 30,000

L

1 : 2048 1 : 256 1 : 16 I : 2048 1 : 128

21

24

2"

30

33

Western blot analysis o f isolated antigen. T o f u r t h e r a n a l y z e t h e a n t i g e n i s o l a t e d by H P - I E C , t h e a n t i g e n c o n t a i n i n g f r a c t i o n w a s s u b j e c t e d to W e s t e r n blot a n a l ysis (Fig. 2c). A single b a n d at 66 k D a d e t e c t e d by a u t o l o g o u s a n t i b o d y is n o t e d . T h e 66 k D a b a n d n o t e d h e r e c o r r e s p o n d s to w h a t we h a v e r e p o r t e d p r e v i o u s l y with W e s t e r n blot a n a l y s i s o f u n f r a c t i o n a t e d s p e n t m e d i a [18]. Blue agarose affinity chromatography. B e c a u s e b u l k s e r u m a l b u m i n r e p r e s e n t s t h e m a j o r c o n t a m i r ~ a n t in t i s s u e c u l t u r e m e d i a c o n c e r n ti~at t h e a n t i g e n w a s a n a l b u m i n c o n t a m i n a n t w a s raised. A u t o l o g o u s a n t i b o d y reactivity c o u l d n o t b e a b s o r b e d with i n c u b a t i o n with fetzl c a l f s e r u m o r a 60 x c o n c e n t r a t i o n o f t i s s u e c u l t u r e m e d i a t h a t h a d n o t b e t a inc, : e t e d with m e l a n o m a cells ( d a t a n o t s h o w n ) A c h r o m a t o g r a p h o f unfractionated spent media applied ~ a Blue agarose c o l u m n is s h o w n in Fig. 3a. f r a c t i o n A, r e p r e s e n t i n g t h e initial l a r g e p e c k w h i c h did n o t b i n d to t h e c o l u m n w a s n o t e d to c o n t a i n ~b a n t i g c p a,; d e t e r m i n e d by c o m p e t i t i v e i n h i b i t i o n (Fi-,. 3b) ,~O a n t i g e n w a s n o t e d in t h e f r a c t i o n t h a t b o u n u . ~ ' h e c o l u m n .

Fig. 2. (a) HPLC ion-exchange chromatography (HP-IEC) of spent media. Chromatograph of spent media following HP-IEC. After injection of the sample a large peak is noted. The buffer was then changed (indi:ated by vertical line) and a large peak was eluted which corresponds to BSA. Immediately following the BSA elution, a smaller peak can be seen (see arrow). Collected fractions were dialyzed against PBS and tested for the presence of antigen by competitive inhibition. (b) Competitive inhibition of autolo~ous antibody reactivity following HP-IEC. Control serum (C) was noted to have a maximum titer of 1:204& The addition of unfractionated spent media (SM) reduced autologous antibody titers to 1:256. Following HP-IEC only fraction 12 (see arrow Fig. 2a) reduced autologous antibody titers which is indicative of the presence of antigen. (c) Western blot analysis of isolated antigen-SDS/PAGE of antigen isolated by HP-IEC. A single band at 66 kDa is noted.

Neuraminidase digestion. Exposure of HP-IEC purified antigen to neuraminidase ablated recognition of the antigen by autologous antibody (Table II). Lectin binding. Binding of the antigen, obtained following DEAE anion exchange chromatography, to various biotinylated iectins was evaluated and are summarized in Table III. Binding was noted with lectins derived from Triticum vulgaris (wheat germ agglutinin), Dolichos biflorus (Horsegram) and Lycopersicon esculentum (tomato agglutinin).

Preparative purification Anion exchange chromatography. Based on our ability to consistently define the conditions under which the antigen may be isolzted by HP-IEC, we turned to anion exchange column chromatography to isolate larger quantities of the artigen. Autologous antibody reactivity was markedly reduced by samples obtained from Fractions 92-101 (Fig. 4b). This corresponds to a small peak noted on the chromatograph (see arrow, Fig. 4a). No other fraction contained antigen as determined by competitive inhibition. Lectin affinity chromatography. Through the use of biotinylated lectins we determined that the antigen binds to the lectin, Dolichos biflorus (Horse gram). As

seen in Fig. 5, antigen was detected in fractions 1-5, prior to elution with acetic acid and in fractions 24-36 following elution. The presence of the antigen in the unbound fraction suggests that the capacity of the column for the antigen had been exceeded and that a higher yield of purified antigen might be achieved with a larger lectin column. Adding antigen obtained from fraction 1 back onto the lectin column demonstrated additional binding, confirming our observations about the limited capacity of the column (data not shown). Assessment o f the purity and yield o f isolated antigen. SDS-PAGE analysis of each step of the purification process is illustrated in Fig. 6. Silver stain of the material obtained following lectin affini~, chromatography revealed a single protein band at 66 kDa (lane C). No other bands previously noted in unfractionated spent media or following anion-exchange chromatography were seen. Measurement of protein and carbohydrate concentration following each purification step is shown in Table IV. Measurement of the protein concentration following lectin affinity chromatography was difficult as the level was below the limits of detection by most assays. Based on the detection of a single band by silver stain, we estimate that the final yield of antigen was approx. 1 txg/ml. The final concentration of carbohydrate was determined to be 14/zg/ml. Be-

TABLE III Lectin binding to 66 kDa melanoma-associated antigen Positive Trilicum vulgaris GIcNAc(fl(1,4)GIcNAc)2> fl-GIcNAc Dolichos biflorus GaINAc(1,3)GaINAc Lycopersicon esculentum GIcNAc(fl(1,4)GIcNAc)I_3

Negative Canat'alia ensiformis a-man > a-Glc Arachis hypogaea Gal~q(l,3)GalNAc> Gal Phytolacca americana GIcNAc(fl(1,4)G{cNAc)1-5= (Gal(fl(l,4)GIcNAc)2-5 Glycine max a- and/3-GalNAc Erythina cristagalli [D-GaI-(I,4)GIcNAc]> -GalNAc Vicia cillosa a-GalNAc Griffonia simphcifolia I-B4 a-D-Gal Phytolacca americana GIcNAc(fl 1,4 GIcNAc)I -s = (Gal fl 1,4 GIcNAc)2 Lotus tetragonolobus L-Fuca(l,2)Gal fl(i,4)L-Fuca(l,3)GIcNAc Griffonia simplicifolia H a-GIcNAc,fl-GIcNAc Euonymus europaeus Gala(l,3)[L-Fuc-a-(1,2)]Gal> Gala(l,3)Gal= L-Fuca(l,2)Gal Maackia amurensis NANA a(2,3)Galfl(l,4)GIc Ulex europaeus ! L-Fuca( 1,2)Gal-fl(1,4)GIcNAc

a 0.200 - ,---A280

Bf

~

2400

0.180

2000 C)