ARCHIVES

Vol.

OF BIOCHEMISTRY

298, No. 2, November

AND

BIOPHYSICS

1, pp. 538-543,

1992

Purification of a Tumor-Specific PNA-Binding Glycoprotein, gp200, from a Human Em bryonal Carcinoma Cell Line W. Michael

Schopperle,*

D. Randall

Armant,?

and William

C. DeWolf*,l

*Division of Cellular Genetics, Charles A. Dana Research Institute, Beth Israel Hospital, Boston, Massachusetts 02215; and TThe Department of Obstetrics and Gynecology, Charles S. Mott Center for Human Growth and Development, Wayne State University School of Medicine, Detroit, Michigan 48201

Received

April

6, 1992, and in revised

form

July

6, 1992

A 200-kDa peanut agglutinin (PNA)-binding glycoprotein, gp200, has been purified and partially characterized from the human embryonal carcinoma cell line, HT-E (833k). Tissue distribution analysis of this molecule by lectin blotting with PNA of detergent-extracted proteins from human cell lines and tissues demonstrated expression limited to nonseminomatous germ cell tumors. The 200-kDa protein was purified with lectin affinity and gel filtration chromatography. Purification to apparent homogeneity was demonstrated by one- and twodimensional gel electrophoresis. Characterization of gp200 revealed it to be a surface integral membrane glycoprotein; however, gp200 could also be purified from the culture media of EC cells, suggesting gp200 has an extracellular role. The carbohydrate groups of gp200 are N-linked and partially sialylated and contain terminal galactose residues. These initial studies suggest that the PNA-defined glycoprotein, gp200, is a candidate for a nonseminomatous germ Cell tUmOr marker. 0 1992 Academic Press,

Inc.

Testis malignancy is the most common solid tumor in young adult males (1). Tumors of germ cell origin comprise 90% of all testicular tumors and are generally divided into two categories, seminomatous and nonseminomatous tumors. Identification of unique antigens on human germ cell tumors (GCT)2 provide a source for new tumor mark-

ers and information to aid in the delineation of the origin of these tumors. Peanut agglutinin, PNA, is a lectin that binds terminal D-galactose residues (2). Binding sites for PNA are uncommonly expressed on mature somatic cells but have been described in association with active differentiation and development such as in immature thymocytes (3) and preimplantation mouse blastocysts (4). PNA receptors have also been identified in mouse and human spermatogenie cells (5, 6). In mouse spermatogenic cells, PNA binding has been shown to be stage specific; spermatocytes are positive for PNA binding while spermatogonia do not bind PNA (7). Tumors of germ cell origin have also been reported to have PNA binding sites. PNA binding sites that disappear during in vitro differentiation have been described on murine embryonal carcinoma cells (8-10). Histochemical studies of human germ cell tumors have shown that PNA binds to the surface of nonseminomatous but not that of seminomatous GCT, indicating a potential role for PNA in classification of GCT (11). Lectin blotting was used to identify unique PNA-binding proteins expressed in human GCT. A 200-kDa PNAbinding protein, gp200, was identified in a nonseminomatous GCT embryonal carcinoma (EC) cell line, HT-E (833k), and is being evaluated as a potential tumor marker for nonseminomatous GCT. This report describes the purification, tissue distribution, and initial characterization of gp200. MATERIALS

i To whom correspondence should be addressed. ’ Abbreviations used: PNA, peanut agglutinin; PBS, phosphatebuffered saline; SDS-PAGE, sodium dodecyl sulfate-polyacrylamide gel electrophoresis; EC, embryonal carcinoma; GCT, germ cell tumor; NSGCT, nonseminomatous germ cell tumor; BSA, bovine serum albumin.

AND

METHODS

Tissue sources. Human EC cell lines HT-E (833k) and HT-H (2061H) are characterized cell lines and were maintained in RPM1 1640 with 15% fetal calf serum, 2 mM L-glutamine, and 1% gentamicin (1214). Renal cell carcinoma cell line HTB-45, Tera-I embryonal carcinoma cell line, ovarian carcinoma cell line A2780, and leukemic cell line, HL60, were obtained from American Type Culture Collection (Rockville, MD). Normal human organ tissues and tumors were obtained from surgical

538 All

Copyright 0 1992 rights of reproduction

0003.9861/92 $5.00 by Academic Press, Inc. in any form reserved.

PURIFICATION

AND

TISSUE

specimens. Protein extractions were done by homogenizing tissue or cells in 10 mM Tris-HCl, pH 7.2,0.1 mM phenylmethysulfonyl fluoride, 1 pg/ml leupeptin, 36 KI units/ml Aprotinin, 1 fig/ml soybean trypsin inhibitor, 1 mM EDTA, and 1% Triton X-100. Purification ojgp200. Protein extracts from EC cell line HT-E (833k) were mixed with PNA coupled to Sepharose 4B beads (Pharmacia) (PNA coupled to CNBr Sepharose 4B beads at 5 mg/ml beads by manufacturer recommendations) and incubated overnight at 4°C. The suspension was poured into a column and washed with 2.0 M NaCl/PBS, and PNAbinding proteins were eluted from the column with 0.2 M D-galactose at room temperature. The eluate was dialyzed against 10 mM Tris-HCl, pH 7.4, vacuum concentrated, and chromatographed at 4°C over a 100 X l.O-cm Sephacryl S-200 (Pharmacia) column (flow rate 9 ml/h). Protein concentrations were determined by the Bradford method (15). Gel electrophoresis and blotting. One- and two-dimensional gel electrophoresis (SDS-PAGE) was performed using standard methodology (16,17). Gels were silver stained (18) and autoradiograms were obtained by drying gels and exposing them to Kodak AR X-ray film for 1 to 24 h. Flurography was done by soaking gels in EnaHance (New England Nuclear) for 30 min, drying gels, and exposing them to Kodak X-ray film for 1 to 7 days. Blotting was done by transfering proteins to nitrocellulose (20). Following the transfer, the nitrocellulose was incubated for 30 min in blocking solution (3% BSA-PBS, pH 7.4), followed by overnight incubation at 4°C in blocking solution containing 0.2 fig/ml of peroxidase-conjugated PNA (Sigma). The nitrocellulose was washed for 2 h with five changes of PBS and developed in PBS containing 20% methanol, 0.5 mg/ml of 4-chloro-1-napthol, and 0.02% hydrogen peroxide. Carbohydrate analysis. To remove sialic acid, approximately 100 pg of protein extract from cells or tissue or 1 to 5 pg of purified gp200 was dialyzed, lyopholized, and resolubilized in 50 ~1 of 10 mM calcium acetate, 20 mM sodium cacodylate, pH 6.0, and 0.1% SDS. Samples were boiled for 5 min and NP-40 was added to a final concentration of 1% followed by 0.05 units of neuraminidase (Sigma). Samples were incubated for 1 to 8 h at 37°C. For enzymatic removal of O-linked carbohydrate chains from purified gp200, 0.5 milliunits of 0-glycanase enzyme (Genzyme) was added to 5 +g of neuraminidase-treated gp200 or to untreated gp200. Samples were incubated at 37’C for 24 to 48 h. To remove N-linked carbohydrate groups from purified gp200, approximately 5 pg of gp200 was lyophilized and resolubilized in 10 pl of 0.5% SDS and 0.1 M flmercaptoethanol. After boiling for 5 min, 15 ~1 of 0.5 M sodium phosphate, pH 8.6, 5 ~1 10% NP-40, and 1 unit of N-glycanase (Genzyme) were added to the sample. Chemical removal of O-linked carbohydrates was performed by mild alkaline hydrolysis. Approximately 5 pg of ‘%Ilabeled gp200 was lyophilized, resolubilized, in 5,10,20, or 50 mM NaOH, and incubated at 37°C for 16 h. Tunicamycin treatment of EC cells was performed by growing EC cells for 72 h in media containing either 10, 40, or 120 pg/ml of tunicamycin (Sigma). Radiolnbelinggp200. Iodination of purified gp200 was performed by adding 5 pg of gp200 to a reaction vial containing 10 pg of IODO-GEN (Pierce) reagent and sodium iodine-125 (NaiZ51, 100 PCi, New England Nuclear). Unlabeled Na iz51 was removed by gel filtration over a G-50 (Pharmacia) column that was equilibrated in 0.5% BSA, 10 mM Tris, pH 7.0 buffer. Gp200 was labeled with tritiated galactose by growing HT-E (833k) cells for 7 days in media containing 5 PCi D-[3H]galactose (ICN). Labeled gp200 was purified from the protein extract. The spent culture medium was filtered through 0.2-wrn filter and run over a 3-ml PNA-Sepharose column at a flow rate of 1 ml/min. The column was washed with 2 M NaCl buffer and PNA binding proteins isolated from the media were eluted from the column with 0.2 M D-galactose. Proteinase K digestion of EC membrane HT-E (833k) EC cells were harvested, aliquots, each suspended in 2 ml PBS, K (Sigma) was added to one aliquot of as a control without proteinase K. After 100 gl of 1 mM PMSF was added to the

proteins. Approximately lo7 washed, and divided into two pH 7.4. One unit of proteinase cells. The second aliquot acted 30 min of incubation at 37°C cells to inhibit proteinase K.

DISTRIBUTION

OF

539

gp200

RESULTS

Peroxidase-conjugated PNA was used as a probe to identify PNA-binding glycoproteins from human germ cell tumors and normal testis tissue. Analysis of detergentextracted GCT proteins by lectin blotting with PNA revealed a prominent 200-kDa PNA-binding protein, gp200, in nonseminomatous GCT (Fig. 1, Lane 1). Gp200 was not detectable in normal testis or seminomatous tumor protein extracts (Fig. 1, Lanes 2,3). However, a prominent PNA-binding protein was identified in the testis protein extract with an estimated molecular weight of 230 kDa; gp200 was also found to be expressed in an EC cell line HT-E (83313) (Fig. 1, Lane 4). The results of PNA blotting of human tissue, tumors, and cell lines showed that detection of gp200 was limited to nonseminomatous GCT and EC cell lines (Table I). Two additional EC cell lines, HT-H and Tera-1, also expressed gp200, confirming the presence of gp200 in different NSGCT-established cell lines. Many glycoproteins in mammalian cells have galactosecontaining carbohydrate structures that do not bind PNA because the galactose residues are capped by sialic acid. To determine whether somatic tissues contain gp200 with sialic acid masking the PNA binding site, the extracted proteins were treated with neuraminidase prior to lectin blotting with PNA. The results show the presence of nu-

.

ZOOkDa

FIG. 1. Identification of gp200 by PNA blotting. Detergent-extracted proteins from normal human testis (Lane 2), nonseminomatous germ cell tumor (Lane l), seminoma tumor (Lane 3), and human embryonal carcinoma cell line, HT-E (833k) (Lane 4) were separated on a 7.5% SDS gel and blotted onto nitrocellulose. PNA binding proteins were identified by peroxidase-conjugated PNA. Lane 5 contains prestained molecular weight markers. Myosin (206 kDa), B-galactosidase (116 kDa), and albumin (68 kDa). Approximately 50 pg of protein was loaded onto Lanes l-4.

540

SCHOPPERLE. TABLE

ARMANT,

AND

DEWOLF

I

Detection of gp200 in Human Tissue and Cell Lines before and after Neuraminidase Treatment Tissue

(-) Human

Neuraminidase malignant

+ ND ND ND ND ND ND

Human

EC HT-E (833k) EC HT-H (2061H) EC Tera-1 (105.HTB) Renal cell carcinoma (HTB 47) Ovarian (A2780) Leukemic (HL60) Colon carcinoma (HCT-15) Prostate carcinoma

cell lines + + +

(PC3)

Human

.SPZOO

Neuraminidase

tissue

+ -

Nonseminomatous Seminoma GCT Melanoma tumor Breast tumor Bladder tumor Prostate tumor Kidney tumor Liposacaroma

Testis Lymphocyte Kidney Lung Colon Prostate Liver Spleen Thymus Brain Stomach Muscle

(+)

4s

+ ND ND

-

ND ND ND

-

ND

-

ND

normal

-

1

i *

FIG. 2. Gel analysis of purified gp200. (A) Lectin blot with PNA after transfer from a 10% polyacrylamide gel of affinity-purified g-p200 from EC cell line HT-E (833k) (Lane 2). Lane 1, prestained molecular weight markers. (B) Silver-stained 12% polyacrylamide gel of total detergentextracted protein from HT-E (833k) (Lane 2), and affinity-purified gp200 (Lane 3). Lane 3 is overloaded with protein to enhance visualization. Lane 1, molecular weight markers. *The gp200 aggregate formed during the purification of gp200.

tissue

* ND * * * * ND * -

in gp200 affinity-purified fractions. The gp400 disappeared under reducing conditions (seeFig. 5A), suggesting that the band may represent an aggregate of gp200. Purified gp200 was labeled with 1251and subjected to twodimensional gel analysis. The autoradiogram (Fig. 3) shows gp200 with an approximate isoelectric point of 6.0.

A

-c*

B

~+gPtoo

Note. +, detection of gp200 by lectin blotting of detergent extracted proteins with peroxidase-conjugated PNA. -, no 200-kDa PNA-binding protein could be detected. ND, not done. *, Numerous PNA binding proteins present in the 150-250 kDa range and specific identification of gp200 was not possible.

merous high molecular weight PNA-binding proteins in many tissues such as testis, kidney, lung, liver, spleen, and brain (Table I) and specific identification of gp200 in these blots could not be done. Purification of gp200 from EC cell line HT-E (83313) was performed by PNA affinity and gel filtration chromatography (see Materials and Methods and Fig. 2). The purified fraction of gp200 also contained a major PNAbinding protein with an estimated molecular weight of 400 kDa. This larger PNA-binding protein was not found in cell protein extracts (Fig. l), but was consistently found

FIG. 3. Autoradiogram of two-dimensional gel of purified gp200. Affinity-purified gp200 was labeled with iz51 and analyzed by two-dimensional gel electrophoresis. Bovine serum albumin is present on the gel because of its addition during the iodination procedure of purified gp200. The pZ of gp200 is approximately 6.0. The acidic (A) and basic (B) ends of the gel are labeled. *Note the presence of the dimer form of the gp200 molecule with same pZ as gp200.

PURIFICATION

AND

TISSUE

DISTRIBUTION

1

FIG. 4. Cellular localization of gp200 in EC cells. Lectin blot with PNA of: Protein extracted from EC cells without detergent (Lane 1); insoluble pellet from non-detergent-extracted EC cells resuspended in extraction buffer with detergent (Lane 2); detergent-extracted protein (Lane 3); insoluble debris from detergent-extracted EC cells resuspended in 9 M urea (Lane 4); EC cells treated with proteinase IS prior to lysis and detergent extraction (Lane 5); and EC cells with no enzymatic treatment prior to lysis and extraction (Lane 6). Approximately 50 bg of protein is loaded onto Lanes 1-6.

The aggregate form of gp200 is also present and has a similar isoelectric point. Treatment of intact cells with proteinase K demonstrated that most gp200 was present on the cell surface and exposed to enzymatic degradation. Although greater than 50% of the proteinase K-treated cells were able to exclude trypan blue dye, PNA lectin blotting of protein from treated cells showed a complete loss of gp200 (Fig. 4, Lane 5) as compared to control cells (Fig. 4, Lane 6). To determine whether gp200 is an integral membrane component or a peripheral extracellular glycoprotein, EC cells were lysed in extraction buffer with or without nonionic detergent. The gp200 was solubilized when detergent was present during extraction (Fig. 4, Lane 3), but it was barely detectable when the detergent was omitted (Fig. 4, Lane l), suggesting that gp200 is an integral membrane protein. EC cells were grown in media containing [3H]galactose. The galactose-labeled gp200 protein is expressed in large amounts as compared to other galactose containing proteins in the EC cells (Fig. 5A, Lanes 1 and 2). The gp200 aggregate is not detectable in the whole cell galactoselabeled protein lysates but after affinity purification of gp200 the aggregate is present (Fig. 5A, Lanes 3 and 4). Most of the aggregate is lost when electrophoresed under reducing conditions. gp200 was also purified from the media of the galactose-labeled cells (Fig. 5B, Lane 2). The extracellular form of gp200 did not form the gp200 aggregates during purification.

OF

2

gp200

3

541

4

1

2

FIG. 5. Galactose labeling of gp200. (A) Autoradiogram of 4-20% SDS polyacrylamide gel loaded with detergent-extracted proteins from EC cells grown in the presence of [3H]galactose (Lanes 1,2) and affinitypurified gp200 from the same cells (Lanes 3, 4). Lanes 2 and 4 were nonreduced and Lanes 1 and 3 were reduced. The gp200 aggregate is not detected in the galactose-labeled whole cell homogenates but is present in the affinity-purified preps. Most of the aggregate is lost when electrophoresed in reducing buffer. (B) Autoradiogram of a nonreduced 7% polyacrylamide gel showing affinity-purified gp200 from EC cells labeled with [3H]galactose (Lane 1) and gp200 purified from spent culture media of EC cells labeled with [3H]galactose (Lane 2).

Purified, lz51-labeled gp200 was treated with different carbohydrate-removing enzymes to determine its carbohydrate moiety. Removal of O-linked oligosaccharides with 0-glycanase (data not shown) or by treatment with 5 to 20 mM NaOH (Fig. 6A) caused no observable electrophoretic shift of gp200. The electrophoretic mobility of gp200 was altered, however, by treatment with N-gly-

C 1

\

SP200

\

Y

2

3

u 1) 2DDkDa

SPZOO

12

3

FIG. 6. Carbohydrate analysis of gp200. (A) Autoradiogram of iodinated gp200 treated for 24 h with 5,10, or 20 mM NaOH. Untreated gp206 (c) is shown in Lane 1. (B) Autoradiogram of iodinated gp200 treated with N-glycanase enzyme for 24 h (Lane 2) and 48 h (Lane 3). Lane 1 is untreated gp200. (C) PNA blots of detergent-extracted proteins from EC cells treated with (Lane 1) and without (Lane 2) neuraminidase. Note the upward shift of gp200 after treatment with neuraminidase.

542

SCHOPPERLE,

ARMANT,

canase (Fig. 6B, Lanes 2 and 3) as compared with nontreated protein (Fig. 6B, Lane 1). The staining pattern of N-glycanase-treated gp200 is diffuse and probably indicates that the carbohydrate digestion is incomplete. Longer digestion with N-glycanase or the addition of more N-glycanase enzyme did not increase the removal of Nlinked sugars. To confirm the presence of N-linked sugars on gp200, EC cells were grown for 7 days in medium containing the N-glycosylation inhibitor tunicamycin (data not shown). No gp200 could be detected in the EC cells grown with tunicamycin, suggesting that the terminal galactose residues of gp200 are on N-linked oligosaccharide chains. Removal of sialic acid from gp200 by neuraminidase treatment of HT-E (83313) protein extracts caused an upward shift of gp200 on SDS-PAGE analysis (Fig. 6C, Lane 1) when compared with untreated gp200 (Fig. 6C, Lane 2). PNA staining of the neuraminidase-treated gp200 was increase due to the exposure of additional galactose residues. Similar treatment of purified gp200 produced an identical result (data not shown). DISCUSSION

This report describes the initial characterization of a PNA binding membrane protein, gp200, expressed on the surface of an EC cell line HT-E (83313). Using PNA lectin blotting, a variety of human tissues were screened for the presence of gp200 and its expression was found limited to nonseminomatous GCT; gp200 was not detectable on seminomatous GCT, normal human tissue, malignant tissue, or non-GCT cell lines. This finding is consistent with previously published histochemical studies of human GCT using fluorescence-conjugated PNA which showed positive PNA binding to nonseminomatous GCT and negative binding with seminomatous GCT (11). PNA-binding glycoproteins are uncommon on somatic tissues because terminal galactose residues are usually capped with sialic acid. Neuraminidase treatment of proteins extracted from normal tissues yielded diffuse PNAbinding bands in the range of 200 kDa in many tissues. Whether or not these proteins represent an altered form of gp200 awaits further structural identification and will be resolved with the generation of gp200-specific antibodies. Analysis of detergent-extracted proteins from human testis revealed a major PNA-binding protein with an apparent molecular weight of 230 kDa. This PNAbinding molecule is presumed to originate from spermatogenic cells within the testis because nongerm cells are reported to be PNA nonreactive (6, 7). gp200 was purified from human EC cell line HT-E (833k) by lectin affinity and gel filtration chromatography. PNA affinity chromatography was an effective method for isolating gp200 because of the limited presence of glycoproteins bearing terminal D-galactose residues. Homogeneity of gp200 was demonstrated by 2-dimensional gel electrophoresis of iodine-125labeled purified gp200.

AND

DEWOLF

The diffuse staining pattern of gp200 on one- and twodimensional gels as well as poor staining characteristics by conventional gel stains are typical of heavily glycosylated proteins. Preliminary carbohydrate analysis of purified gp200 reveals the presence of N-linked carbohydrate structures. Although only a partial digest of gp200 was obtained, the results indicate that all of the carbohydrate groups are N-linked. This finding was confirmed with experiments using a N-glycosylation inhibitor, tunicamycin. Gp200 could not be detected by PNA blotting in EC cells grown in the presence of tunicamycin. Gel analysis of purified gp200 revealed the presence of a large PNA-binding protein with a molecular weight of approximately 400 kDa which was present in nonreduced purified extracts and not in whole cell HT-E (833k) protein extracts. Treatment of purified gp200 with a reducing agent prior to electrophoresis resulted in an almost complete loss of the 400-kDa protein on gels and was accompanied by an increase in the staining density of gp200. These observations suggest that a gp200 aggregate is formed during the purification procedure of gp200. Whether this aggregate is an artifact of purification or has a biological role in HT-E (83313) cells is unknown. It is interesting to note that although gp200 can be isolated and purified from the media of EC cells, the aggregate form of gp200 cannot be identified in purified fractions from media. The absence of the gp200 dimer in purified fractions of gp200 from media may be due to loss of a “sticky” hydrophobic membrane-anchoring tail from gp200 when it is secreted into the media. It is also possible that purified gp200 from EC cells and purified gp200 from the media of EC cells are not the same molecule. Many PNA-binding proteins have been identified in human cells but usually these proteins require the enzymatic (neuraminidase) removal of sialic acid to expose galactose residues. Some cell types have been identified that have the de novo capacity to react with PNA such as bone marrow cells (21), spermatogenic cells (5,6), and mouse EC cells and morulae (4, 8, 9). A molecule similar to gp200 has been previously reported in association with human teratocarcinoma and identified by Western blotting with lz51-labeled PNA (22). However, in contrast to gp200, the molecule was not secreted and was nonsialylated. The findings reported here suggest that gp200 is a biochemical marker for some testis cancers. The tissue-specific expression of both gp200, in nonseminomatous GCT, and the 230-kDa PNA-binding protein in testis tissue suggests a possible specialized role for these molecules. Further comparison of the two PNA-binding molecules may establish a relationship providing insight into the exact cellular origin of nonseminomatous GCT which has yet to be determined. An interesting characteristic of gp200 after treatment with neuraminidase is its upward electrophoretic shift on a SDS acrylamide gel. The apparent molecular weight of the neuraminidase-treated

PURIFICATION

AND

TISSUE

DISTRIBUTION

OF

543

gp200

8. Reisner, Y., Gachelin, G., Dubois, P., Nicolas, J.-F., Sharon, N., and Jacob, F. (1977) Dev. Biol. 61,20-27. 9. Muramatsu, T., Gachelin, G., Damoneville, M., Delarbre, C., and Jacob, F. (1979) Cell 18, 183-191.

gp200 is similar to that of the testicular 230-kDa PNAbinding protein. The addition of sialic acid to proteins has been proposed to be part of an oncodevelopmental process (23). It is tempting to speculate that gp200 is a sialyated form of the testis-specific 230-kDa PNA-binding protein which is formed during the development of nonseminomatous GCT.

11. Teshima, tasumoto,

S., Hirohashi, S., Shimosuto, K., and Yamada, T. (1984)

ACKNOWLEDGMENTS

12. Andrews, Knowles,

P. W., Bronson, D. L., Benham, F., Strickland, B. B. (1980) Znt. J. Cancer 26, 269-280.

We thank Andrea Titelbaum for secretarial assistance and Dr. Abraham Morgentaler for helpful discussions and suggestions. DRA acknowledges grant support from NIH-HD25795.

10. Muramatsu, T., Gachelin, G., Nicholas, J. F., Condamine, H., Jakob, H., and Jacob, F. (1978) Proc. Natl. Acad. Sci. USA 75,2315-2322.

13. Schwarting, G. A., Carroll, Biophys. Rex Commun. 14. Carroll,

Y., Kishi, Lab. Invest.

P. G., and DeWolf,

K., Ino, Y., Ma50, 271-278. S., and

W. C. (1985)

B&hem.

112,940-945.

P. G., and DeWolf,

W. C. (1983)

131, 1007-

J. Zmmunol.

1012.

REFERENCES 1. Feldman, N. Engl.

A. R., Kessler, L., Myers, J. Med. 316,1894-1897.

M., and Naughton,

M. D. (1986)

2. Lotan, R., Skuttelsky, E., Danon, D., and Sharon, N. (1975) J. Biol. Chem. 250,8518-8523. 3. Reisner, Y., Linker-Israeli, M., and Sharon, N. (1976) J. Cell. Zmmunol. 26,129-135. 4. Chavez, D., and Enders, A. (1981) Dev. Biol. 8’7, 267-281. 5. Maekawa, M., and Nishimune, Y. (1986) Biol. Reprod. 32,419-425. 6. Schopperle, W. M., Kang, S., and DeWolf, W. C. (1986) J. Ural. 135,191A. 7. Watanabe, M., Muramatsu, T., Shirane, H., and Ugai, K. (1981) J. Histochem. Cytochem. 29, 779-787.

15. Bradford,

M. M.

16. Laemmli,

U. K. (1970)

17. O’Farrell,

P. (1975)

18. Morrissey,

(1976)

J. Biol.

J. H. (1981)

19. Towbin, H., Straehelin, Sci. USA 76,4350-4354.

Anal.

Biochem.

Nature Anal.

72, 248-253.

227,680-685. Chem.

250,

Biochem.

4007-4021.

117, 307-310.

T., and Gordon,

T. (1979)

20. Resisner, B. (1979)

Y., Biniaminou, M., Rosenthal, E., Sharon, Proc. Natl. Acad. Sci. USA 76, 447-453.

21. McIhinney, 151-152,

R. A. J., and Patel, A. R. Liss, New York.

S. (1981)

Germ

Proc. Natl.

Acad.

N., and Ramot, Cell Tumors,

22. Roth, J., Zuber, C., Wagner, P., Tuatjes, D., Waisgerber, P., Goridis, C., and Bitter-Sverman, D. (1988) Proc. Natl. USA 85,2799-3003.

pp.

C., Heitz, Acad. Sci.