Echinococcus granulosus: Antigen characterization by chemical treatment and enzymatic deglycosylation

Echinococcus granulosus: Antigen characterization by chemical treatment and enzymatic deglycosylation

EXPERIMENTAL PARASITOLOGY 73,433-439(1991) Echinococcus F. MARCH,* granulosus: Antigen Characterization Treatment and Enzymatic Deglycosylation C. ...

787KB Sizes 0 Downloads 101 Views

EXPERIMENTAL PARASITOLOGY 73,433-439(1991)

Echinococcus F. MARCH,*

granulosus: Antigen Characterization Treatment and Enzymatic Deglycosylation

C. EwucH,t M. MERCADER,* F. SANCHEZ,* AND G. PRATS*'$

by Chemical

C. MuAoz,*,$ P. COLL,*+

*Servei de Microbiologia, Hospital de la Santa Creu i Sant Pau, Barcelona, Spain; fDepartament de Biologia Cel. lular, Universitat de Barcelona, Barcelona, Spain; and .#Departament de Gen2tica i Microbiologia, Universitat Autbnoma de Barcelona, Barcelona, Spain MARCH, F., ENRICH C., MERCADER, M., SANCHEZ, F., MuRoz, C., COLL, P., AND PRATS, G. 1991. Echinococcus granulosus: Antigen characterization by chemical treatment and enzymatic deglycosylation. Experimental Parasitology, 73, 433439. Parasite antigenic

fractions obtained by biochemical purification of sheep hydatid fluid were subjected to enzymatic digestion. The relative mobilities of the 5 and B antigens, before and after treatment, were analyzed by polyacrylamide gel electrophoresis (SDS-PAGE) and Western blot. Antigenic fractions transferred to nitrocellulose were also treated with sodium metaperiodate and concanavalin A. The results indicate that antigen 5 contains a substantial amount of carbohydrates covalently linked to a polypeptide backbone, which strongly bind to concanavalin A and is removed by N-glycosidase F (PNGase F). Antigen 5 possesses complex N-linked oligosaccharides (PNGase F sensitive), without terminal N-acetyl-D-glucosamine residues (N-acetyl-o-glucosaminidase nonsensitive) and has no high-mannose oligosaccharides (endo+-N-acetylglucosaminidase H nonsensitive). In contrast, the antigen B of low molecular weight is not susceptible to either enzymatic digestions (PNGase F, Endo H, and N-acetyl-D-glucosaminidase) or sodium metaperiodate oxidation and it does not bind to concanavalin A. Polyclonal antibodies prepared against the two antigens reacted with the deglycosylated antigen 5 in Western blot. The dominant epitopes are, therefore, polypeptides, although the presence of carbohydrate epitopes in the native glycoproteins cannot be excluded. D 1991 Academic press, hc. INDEX DESCRIPTORSAND ABBREVIATIONS: Echinococcus granulosus; Sheep hydatid fluid (SHF); Anti-serum, rabbit, against SHF antigens; Glycoprotein; Concanavalin A (Con A); Sodium metaperiodate; N-glycosidase F (PNGase F); endo+-N-acetylglucosaminidase H (Endo H); Trypsin; Molecular weight estimate on SDS polyacrylamide gels (SDS-PAGE); Immunogenicity; Western blot.

poprotein antigen 5 (Capron et al. 1967; Pozzuoli et al. 1975) and the thermostable Unilocular hydatidosis is a disease lipoprotein antigen B (Oriol et al. 1971). Oriol et al. (1971) and Pozzuoli et al. caused by infection with the metacestode (1972, 1974, 1975, 1977) partially characterstage of the dog tapeworm, Echinococcus granulosus. This is recognized as one of the ized the two major antigens of hydatid cyst world’s major zoonoses, affecting both hu- fluid. Further studies showed that antigenic activity of “antigen 880” (similar to antigen mans and their domestic animals. B for the properties described by Njeruh et The material most frequently used in the serological diagnosis of human hydatidosis al. 1989)was not affected by trypsinization, is sheep hydatid fluid (SHF), which is a pepsinization, or delipidization. Both antimixture of sheep serum proteins and para- gens were periodate-sensitive (Al-Yaman et site antigens (Craig et al. 1981; Tassi et al. al. 1989) and concanavalin A (Con A) binding lipoproteins (Rickard et al. 1986). 1981). This study is concerned with the nature Parasite antigens comprise two major components, namely the thermolabile li- of the various antigenic components using 433 0014-4894/91$3.00 Copyright 8 1991 by Academic Press, Inc. All rights of reproduction in any form reserved.

434

MARCH ET AL.

enzymatic deglycosylation. For correlation with the study of the role of carbohydrates in antigenicity, it was of particular interest to determine the presence of covalent association of carbohydrates moieties (glycan) with proteins and the types of glycanprotein linkages. Primary emphasis is placed on two endoglycosidases which appear to have the most utility in clarifying the relationship between structure and function, namely IV-glycosidase F (PNGase F) (Trimble et al. 1984) and endo+-Nacetylglucosaminidase H (Endo H) (Tarentino et al. 1974). Together with these deglycosylations, chemical treatments with sodium metaperiodate and carbohydrateCon A interactions were performed. MATERIAL

AND METHODS

Antigens. Sheep hydatid fluid (SHF) was obtained from hepatic and pulmonary cysts in viscera. The protoscoleces were allowed to sediment, and remaining particulate material was removed by centrifugation at 15OOOgfor 30 min at 4°C. The solution was then filtered through a membrane (0.45pm pore size; Millipore Corp.) and stored at - 20°C. Concentration of pools of SHF was achieved with polyethylene glycol20.000 (Fluka) for 8 hr at 4°C. This was generally sufficient to concentrate the volume lofold. It was determined that the concentrated fluid used in these studies had a protein content of 563 mg (5.7 mg/ml). Purification of two major antigens of E. granulosus was performed by the method of Oriol et ul. (1971). Briefly, by means of dialysis against acetate buffer pH 5, 5 mM, the parasite components and host immunoglobulins were left in solution at a concentration of 104 mg of protein (5.2 mg/ml) without detectable host albumin. The SHF was shown to contain less than 20% of purified parasite antigens and about 70% of contaminating host serum albumin. The purified material was stored at -20°C until used for antigen characterization. Sera. A rabbit hyperimmune antiserum to antigens of E. granulosus was obtained by repeated intramuscular immunization every 8 days over a period of 2 months with about 1 mg of SHF purified with complete Freund’s adjuvant (Sanchez ef al. 1991). Quantitation of protein. Quantitative determinations of protein were performed by the Bradford method (Bradford 1976). Interaction of polysaccharide with Con A. The separated antigenic material in the preparative gels (500

pg) was transferred to nitrocellulose paper (0.45 pm, Millipore Corp.), which was rinsed with I nuI4 Ca’+ , 1 mM Mg’+, 1 mM Mn’+ Tris-buffered saline (pH 7.6) (TBS-C) for 30 min and was then incubated for 60 min at room temperature by adding 0.2 kg/ml or 0.5 &ml of biotin labeled Con A (from Canavalia ensiformis; Sigma) on a rocking platform (Goldstein et al. 1965; Becker et al. 1971). The strips were washed twice in TBS-C and once with TBS-C 0.1% (v/v) Tween 20. The strips were immersed in a solution of avidinbiotinylated peroxidase (Vectastain ABC Kits; Vector Laboratories) for 30 min at room temperature. Color was allowed to develop by the addition of substrate (0.1% (v/v) 30% HzOz in 0.1 M TBS; 0.05% (w/v) 3,3-diaminobenzidine tetrahydrocloride) and the reaction was stopped by rinsing with tap water. Periodate treatment. Antigenic fractions separated by SDS-PAGE (500 pg of protein) and transferred to nitrocellulose were subjected to severe periodate treatment, which was performed by the addition of IO mM and 100 mM sodium metaperiodate in 50 mM and 20 mM sodium acetate buffer, respectively. The reaction was allowed to proceed for 30 min at 4°C in the dark (Jamieson et al. 1971). Enzymatic treatments. Enzymes used in this study were N-glycosidase F from Flavobacterium meningosepticum (Boehringer-Mannheim), endo-B-Nacetylglucosaminidase H from Streptomyces plicatus (Boehringer-Mannheim), B-N-acetylglucosaminidase from Aspergillus niger (Sigma), and trypsin (Gibson). The conditions used to screen the activity of endoglycosidases on this purified antigenic material were as follows: N-glycosidase F (PNGase 0. Fifty micrograms of antigen were incubated with 10 U of PNGase F overnight at 37°C in the presence of 50 yl TBS (150 mM NaCl, 50 mM Tris-HCI, pH 7.3). 5 ul 10% SDS (w/v), and 25 ul 10% Triton X-100 (Stamatoglou et al. 1990). To protect against proteolytic degradation, aprotinin (Sigma) was added. The reaction mixture was quenched by boiling the incubate for 3 min and the protein was pelleted. The resultant pellet was resuspended in PAGE sample buffer and was stored until electrophoresis. A sham digest (identical conditions except no enzyme was added), as well as positive (fetuin) and negative controls were routinely analyzed to ascertain the intrinsic degradation of the proteolytic activity of the endoglycosidase. Endo-P-N-acetylglucosuminidase H (endo H). Fifty micrograms of protein sample in 2 ul of 10% SDS was mixed with 50 pl of 50 mM sodium acetate buffer (pH 6.5) and 10 pl of 10 mM EDTA and boiled for 2 min. The sample was then incubated with IO mU of endo H. The reaction was carried out overnight at 37°C (Chu 1986). P-N-ucetylglucosaminidase. Fifty micrograms of antigen were mixed with 50 )LI of 50 mM sodium-

E. @UnuhUs:

CHEMICAL

AND

ENZYMATIC

acetate buffer (pH 4.5), 10 mM CaCl,, and 0.1 M NaCI. The sample was incubated for 90 min at 37°C with 200 mU of B-IV-acetylglucosaminidase. The reactions of both endo H and B-N-acetylglucosaminidase were stopped as described above for PNGase F. Ovalbumin and asialofetuin (positive controls) were used to demonstrate the specificity of the endo H and B-N-acetylglucosaminidase reactions, respectively. Trypsin. Protease digestion of the purified antigenic material was performed by addition of 1% trypsin (w/ w) (Shimonkevitz et al. 1983). After 4 hr at 37”C, the reaction was allowed to proceed overnight at room temperature with constant agitation. Then it was stopped by adding trypsin inhibitor (from chicken egg white type II-O; Sigma) and the digest was subjected to SDS-PAGE. Western-blots. Samples were subjected to SDSPAGE (Laemmli 1970)and electrotransferred to nitrocellulose sheets by the method of Towbin et a/. (1979).

ANTIGEN

435

CHARACTERIZATION

demonstrated by binding to Con A (Fig. 2). Nevertheless, in the present paper, Antigen B exhibited insensitivity to sodium metaperiodate (Fig. 1) and neither of its subunits could be detected by binding to Con A (Fig. 2) as described in previously published works. Characterization moieties of Antigen 1234

of the carbohydrate

5. To our knowledge, 56

RESULTS AND DISCUSSION

Identification

of a glycoprotein

antigen.

Although the protein and antigen profile of E. granulosus has been defined, the existence and contribution of the carbohydrate moieties are largely unknown. Two major lipoprotein antigens have been identified in the cyst fluid and these have been given various names. The different nomenclatures have been reviewed and their characteristics summarized by Rickard et al. 24(1986). Briefly, Antigen B is an iodinatable, Con A binding lipoprotein of approximately 16150 kDa, comprising three subunits in the M,. range 10-20 kDa which are unaffected aby reduction. Antigen 5 is an iodinatable, Con A binding lipoprotein of varying high M, (approximately 400 kDa) comprising FIG. 1. Analysis of sodium metaperiodate oxidation subunits of 60-70 kDa which dissociate on of the antigen 5 (65 and 56 kDa) and antigen B (24, 16, disultide bond reduction to two subunits of and 8 kDa) by Western blot. Antigenic fractions iso39 and 20 kDa. In both antigens, isolated lated from SHF were run on 5-20% polyacrylamide and blotted onto nitrocellulose paper. Strips were from E. granulosus hydatid cyst fluid, some gels incubated with 10 mM (lane 1) and 100 mM (lane 3) of the epitopes were periodate-sensitive sodium metaperiodate. Sham treatments (identical (Al-Yaman et al. 1989). These properties conditions except no sodium metaperiodate was for Antigen 5 agree with the results ob- added) were also analyzed (lanes 2 and 4). A rabbit tained in the present work. Antigen 5 was a hyperimmune antiserum to antigens of Echinococcus was used for development. Lanes 5 and 6 glycoprotein which, when transferred to ni- granulosus show control reactions represented by blotting of SHF trocellulose, was sensitive to high concen- antigens recognized by positive and negative human trations of sodium metaperiodate (Fig. 1). It sera respectively. Calculated M, values, in kDa, of the contained mannose or glucose or both as revealed bands are shown on the left margin.

436

MARCH

ET AL.

123

treated antigens and some differences were found. Determination of i%4,without prior reduction of the sample indicates that although variation exists in the mobility of Antigen 5, the mobility of Antigen B was unchanged by the PNGase F treatment. Antigen 5 (65 and 56 kDa) disappears with this enzyme treatment, leaving 60- and 50kDa molecules (small discrepancies in molecular weights could be attributed to direct measurement of specific size on SDSPAGE gels, since the M, obtained are only estimates). No traces of Antigen 5 are detectable at native it4, values, suggesting that the PNGase F digestion was maximal overnight at 37°C as described above. As seen in Figs. 3 and 4, Antigen 5 showed no change in molecular weight following digestion with Endo H but demonstrated a reduction by ca. 5000 Da following digestion with PNGase F. Based upon these findings and the known specificity of FIG. 2. Interaction with Con A of the antigen 5 (65 Endo H (it only cleaves high-mannose and 56 kDa) and antigen B (24, 16, and 8 kDa) separated and transferred to nitrocellulose. Lanes 1 and 2 N-linked oligosaccharide) and PNGase F are antigenic fractions treated with Con A (0.5 ug/ml (cleaves both high-mannose and complex and 0.2 &ml, respectively). Lane 3 shows a Con A sham reaction. SDS-PAGE was carried out in 5-20% acrylamide.

characterization of a carbohydratecontaining antigen (Antigen 5) from E. granulosus by deglycosylation using enzymatic treatments with PNGase F and Endo H (Maley et al. 1989) has not been reported previously. The method proved to be effective (i) to study the nature of the glycanpeptide linkage (0- or N-glycosidic linkages) and (ii) to determine the relative amounts of polymannose and complex N-linked carbohydrate. The presence of N-linked carbohydrate in Antigen 5 was demonstrated by treating the antigenic material isolated from SHF with PNGase F. By SDS-PAGE and Western blot (Figs. 3 and 4), native 5 and B antigens were compared with the enzyme

2

3

4

5

6

7

6

FIG. 3. Analysis of analytical-scale enzymatic deglycosylation reactions of the antigen 5 (65 and 56 kDa) and antigen B (24, 16, and 8 kDa) by SDS-PAGE. Lanes 3, 5, and 7 are the reaction products from PNGase F, Endo H, and B-TV-acetylglucosaminidase reactions, respectively. Equal amounts of protein (50 ug) were loaded onto the gel. In the same way, lanes 2, 4, and 6 represent PNGase F, Endo H, and B-Nacetylglucosaminidase sham reactions. Lanes 1 and 8 show molecular weight markers whose value in kDa is depicted on the left margin. A 10% polyacrylamide gel was used, stained with Coomassie blue R-250.

E. @wU4bSus.’

CHEMICAL

AND

ENZYMATIC

ANTIGEN

CHARACTERIZATION

437

667

654

564

2416-

8FIG. 4. (A and B) Analysis of enzymatic deglycosylation reactions of the antigen 5 (65 and 56 kDa) and antigen B (24, 16, and 8 kDa) by Western blot. Lanes 1 and 5 represent the sham reactions of the antigenic material treated with PNGase F and Endo H, respectively. Lanes 2 and 6 represent the reaction products from enzymatic treatments (PNGase F and Endo H). For electrophoresis, equal amounts (50 pg) of protein were applied to 5-20% polyacrylamide gels, transferred to nitrocellulose, and probed with rabbit hyperimmune antiserum to antigens of Echinococcus grunulosus. Lanes 3,4, and 7 show negative controls represented by blotting of deglycosylated antigens recognized by normal human serum. Calculated M, values, in kDa, of the revealed bands are shown in the margins.

N-linked oligosaccharide), it was concluded that the observed decrease in molecular weight of the glycoprotein was likely the result of specific removal of complex oligosaccharide. It should be stressed that precautions were taken to ensure that the endoglycosidic reactions were performed in the presence of the appropriate inhibitors. Furthermore, proteolytic activity from the endoglycosidases (measured by incubation with nonglycosylated substrate) and from possible contamination of buffers or proteins (measured by sham digests) could not be detected. The activity of each endoglycosidase was also assessed by including glycoprotein, known to be susceptible, under conditions identical to those of the gly-

coprotein of interest as positive controls (data not shown). To extend the characterization of the carbohydrate moieties of Antigen 5, a reaction was performed by using P-N-acetylglucosaminidase. A parallel sham reaction was also performed with asialofetuin. The reaction products from both the P-Nacetylglucosaminidase digest (Fig. 3) and the sham reaction (not shown) were then evaluated by SDS-PAGE. No increased relative mobility was found. These observations are in agreement with our results using both Con A and sodium metaperiodate. It is beyond the scope of this study to determine the use of these enzymes in defining the contribution of carbohydrate to

438

MARCH

the structure and function of the protein and the precise alignment of sugars within oligosaccharide chains, areas in which they have been found to be valuable (Montreuil et al.). Insensitivity of Antigen 5 and Antigen B to trypsin treatment. Identical bands were observed by electrophoretic analysis of native and trypsin-treated antigens (not shown), indicating that both antigens are not susceptible to trypsin digestion. Based upon the known specifity of trypsin, our results are in agreement with a recent report about the relative contribution of different amino acids to the thermostable lipoprotein. In this study, Njeruh et al. (1989) demonstrated the absence of detectable amounts of lysine and arginine, which are the trypsin targets. Antigenicity studies. Two major antigens (Antigen 5 and Antigen B) isolated from hydatid cyst fluid are considered as the dominant immunogen in E. granulosus human infection (Pozzuoli et al. 1975). To investigate the role of the carbohydrate moieties in the immunogenicity of the Antigen 5, antigenicity of native and deglycosylated glycoprotein was revealed by Western blotting analysis. Transferred antigens were recognized by anti-E. granulosus antibodies present in rabbit hyperimmune serum. As seen in Fig. 4, no significant differences were observed after enzymatic digestions (PNGase F and Endo H). Although these results suggest the dominant epitopes are polypeptides, because of the lack of quantitative value with the Western blot we can not definitively exclude the contribution of complex N-linked oligosaccharides in the immunoreactivity of the native glycoprotein. In order to reach more precise conclusions regarding the role of carbohydrates in the immunogenicity of the glycoproteins, a further study to rule out the presence of O-linked oligosaccharides in these antigens is being undertaken.

ET AL.

ACKNOWLEDGMENTS

This work was supported by a grant from the “Fond0 de Investigacibn Sanitaria (FIS)” (8910021). REFERENCES

AL-YAMAN, F. M., AND KNOBLOCH, J. 1989. Isolation and partial characterization of species-specific and cross-reactive antigens of Echinococcus granulosus cyst fluid. Molecular and Biochemical Parasitology 37, 101-107. BECKER,J. W., REEKE, G. N., AND EDELMAN, G. M. 1971. Location of the saccharide binding site of concanavalin A. Journal of Biological Chemistry 246, 61234125. BRADFORD, M. M. 1976. A Rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Analytical Biochemistry 72, 248-254. CAPRON,A., VERNE% A., AND BIGUET, J. 1967. Le diagnostic immunoClectrophorCtique de l’hydatidose. In “Le Kyste Hydatique du Foie” (Simep, Ed.), pp. 27-40. Journees Lyonnaises d’Hydatidologie, Lyon. CHU, F. K. 1986. Requirements of cleavage of high mannose oligosaccharides in glycoproteins by N-glycosidase F. The Journal of Biological Chemistry 261, 172-177.

CRAIG, P. S., AND RICKARD, M. D. 1981. Studies on the specific immunodiagnosis of larval cestode infections of cattle and sheep using antigens purified by affinity chromatography in an enzyme-linked immunosorbent assay (ELISA). Znternational Journal of Parasilology

11, 441449.

GOLDSTEIN,I. J., HOLLERMAN,C. E., AND MERRICK, J. M. 1965.Protein-carbohydrate interaction. I. The interaction of polysaccharides with concanavalin A. Biochimica

and Biophysics

Acta 97, 68-76.

HARRISON, L. J. S., JOSHUA, G. W. P., WRIGHT, S. H., AND PARKHOUSE,R. M. E. 1989. Specific detection of circulating surface/secreted glycoproteins of viable cysticerci in Taenia saginata cysticercosis. Parasite Immunology 11, 351-370. JAMIESON,G. A., JETT, M., AND DEBERNARDO,S. L. 1971.The Carbohydrate Sequence of the Glycopeptide Chains of Human Transferrin. Journal of Biological Chemistry

246, 3686-3693.

LAEMMLI, U. K. 1970.Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680485. MALEY, F., TRIMBLE, R. B., TARENTINO, A. L., AND PLUMMER, T. H. 1989. Characterization of glycoprotein and their associated oligosaccharide through the use of endoglycosidases. Analytical Biochemistry 180, 195-204.

E. grurzulosus:~~~~~~~~

AND ENZYMATIC

MONTREUIL, J., BOLJQUELET, S., DEBRAY, H., FOURNET, B., SPIK, Cl., AND STRECKER, G. Glycoprotein. In “Carbohydrate Analysis. A Practical Approach” (M. F. Chaplin, and J. F. Kennedy, Eds.), pp. 143202. Oxford-Washington. NJERUH, F. M., GATHUMA, J. M., TUMBOH-OERI, A. G., AND OKELO, G. B. 1989. The amino acid and fatty acid composition of the thermostable lipoprotein (“antigen 880”) of hydatid cyst fluid (HCF). East African Medical Journal 66, 219-230. NJERUH, F. M., GATHUMA, J. M., TUMBOH-OERI, A. G., AND OKELO, G. B. 1989. Purification and partial characterization of a thermostable lipoprotein (“antigen 880”) of hydatid cyst fluid (HCF). East African Medical Journal 66, 507-5 15. ORIOL, R., WILLIAMS, J. F., PEREZ ESANDI, M. V., AND ORIOL, C. 1971. Purification of lipoprotein antigens of Echinococcus granulosus from sheep hydatid fluid. American Journal of Tropical Medicine and Hygiene 20, 569-574. PIANTELLI, M., POZZUOLI, R., ARRU, E., AND MusrANI, P. 1977. Echinococcus granulosus: Identitication of subunits of the major antigens. Journal of Immunology 119, 1382-1386. POZZUOLI, R., MUSIANI, P., ARRU, E., PIANTELLI, M., AND MAZZARELLA, R. 1972. Echinococcus granulosus: Isolation and characterization of sheep hydatid fluid antigens. Experimental Parasitology

32, 45-55. POZZUOLI, R., MUSIANI, P., ARRU, E., PATRONO, C., AND PIANTELLI, M. 1974. Echinococcus granulosus: Evaluation of purified antigens’ immunoreactivity. Experimental Parasitology 35, 52-60. POZZUOLI, R., PIANTELLI, M., PERUCCI, C., ARRU, E., AND MUSIANI, P. 1975. Isolation of the most immunoreactive antigens of Echinococcus granulosus from sheep hydatid fluid. Journal of Immunology 115, 1459-1463. RICKARD, M. D., AND LIGHTOWLERS, M. W. 1986. Immunodiagnosis of hydatid disease. In “The Biology of Echinococcus and Hydatid Disease”

ANTIGEN

CHARACTERIZATION

439

(R. C. A. Thompson, Ed.), pp. 217-249. George Allen and Unwin, London. SANCHEZ, F., MARCH F., MERCADER, M., COLL, P., MuRoz, C., AND PRATS, G. 1991. Immunochemical localization of major hydatid fluid antigens in protoscoleces and cysts of Echinococcus granulosus from human origin. Parasite Immunology, in press. SHIMONKEVITZ, R., KAPPLER, J., MARRACK, P., AND GREY, H. 1983. Antigen recognition by H-2 restricted T Cells. I. Cell-free antigen processing. Journal of Experimental Medicine 158, 303-316. STAMATOGLOU, S. C., RUI-CHANG, G., MILLS, G., BUTTERS, T. D., ZAIDI, F., AND HUGHES, R. C. 1990. Identification of a novel glycoprotein (AGpl 10) involved in interactions of rat liver parenchymal cells with tibronectin. The Journal of Cell Biology 111, 2117-2127. TARENTINO, A. L., AND MALEY, F. 1974. Purification and properties of an endo-B-N-acetylglucosaminidase from Streptomyces griseus. Journal of Biological Chemistry 249, 81 l-817. TASSI, C., DOTTORINI, S., SCALISE, G., AND GERANIO, N. 1981. Echinococcus granulosus: Diagnosis of human hydatid disease by the indirect haemagglutination reaction with antigens from hydatid haemagglutination reaction with antigens from hydatid fluid and scoleces. International Journal of Parasitoly 11, 85-88. TOWBIN, H., STAEHELIN, T., AND GORDON, J. 1979. Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets; Procedure and some applications. Proceedings of the National Academy of Sciences of the United States of America 76, 4350-4354. TRIMBLE, R. B., AND MALEY, F. 1984. Optimizing hydrolysis of N-linked high-mannose oligosaccharide by endo+-N-acetylglucosaminidase H. Analytical Biochemistry 141, 515-522. Received 11 January, July, 1991

1991; accepted with revision

2