- Email: [email protected]
Characterization and Distribution of O-Glycosylated Carbohydrates in the Cell Adhesion Molecule, Contact Site A, fromDictyostelium discoideum
EXPERIMENTAL CELL RESEARCH ARTICLE NO.
230, 393–398 (1997)
EX963434
Characterization and Distribution of O-Glycosylated Carbohydrates in the Cell Adhesion Molecule, Contact Site A, from Dictyostelium discoideum MOTONOBU YOSHIDA,*,1 SADAKI YOKOTA,†
AND
SEIJI OUCHI‡
*Research Institute of Food Science and ‡Department of Agriculture, Kinki University, Higashi-Osaka, Osaka 577, Japan; and †Department of Anatomy, Yamanashi Medical School, Yamanashi 409-38, Japan
This paper presents further investigation of the properties of carbohydrate II in the cell adhesion molecule, contact site A, from Dictyostelium discoideum. A purified contact site A was digested with Achromobacter protease I to produce a 31-kDa fragment to which carbohydrate II was mainly bound and a 21-kDa fragment containing the NH2 terminus of contact site A, which was identified as Ala-Pro-Thr-Ile-Thr-Ala. The NH2 terminus of the 31-kDa fragment was ThrGlu-Ala-Thr-Thr-Ser. It was estimated from the cDNA sequence data of contact site A that more than 20 Ser/ Thr residues exist as target sites for the O-linked oligosaccharides in the 31-kDa fragment, but not for the N-linked oligosaccharides. These results suggest that carbohydrate II exists as clustered O-linked oligosaccharides in the COOH terminus of contact site A. The results of two-dimensional electrophoresis confirm that oligosaccharides of contact site A contain sialic acids. Immunoelectron microscopy was carried out to define the organelle in which O-glycosylation by carbohydrate II occurs and how carbohydrate II antigens are distributed on the cell surface. The results show that O-glycosylation can occur in the Golgi apparatus in D. discoideum as observed in other cells, although this O-glycosylation was inhibited by tunicamycin. Furthermore, gold particles were densely concentrated in cell–cell contact regions but sparsely distributed in noncontact regions. q 1997 Academic Press
INTRODUCTION
Recent studies have indicated that cell surface glycoproteins play an important role in various biological processes and that their carbohydrates are involved in determining intracellular sorting, intracellular stability, protein clearance from plasma, cell adhesion activity, and other biological functions [1]. The cell adhesion molecule, contact site A, is involved in EDTA-resistant 1
To whom correspondence and reprint requests should be addressed. Fax: /81 742 431155.
cell contact and appears in the aggregation process of Dictyostelium discoideum. Contact site A is a glycoprotein with an apparent molecular weight of 80 kDa and is anchored in the cell membrane by glycophosphatidyl inositol (GPI) linkages. This glycoprotein consists of two types of carbohydrates: carbohydrate I, which is sulfatable [2], and carbohydrate II, which is mainly labeled with wheat germ agglutinin (WGA) [3]. Previous papers have suggested that carbohydrates might be involved in EDTA-resistant cell contact [4–11]. In particular, the mod B mutants not retaining carbohydrate II have been shown to be impaired in EDTAresistant cell contact [7, 8]. These mutants have been isolated using the monoclonal antibodies, which recognize carbohydrate II and partially block EDTA-resistant cell contact. Consequently, we isolated and determined the carbohydrate structures of contact site A [12] and showed the existence of sialic acids in D. discoideum [13]. We have previously reported that carbohydrate I is composed of N-linked complex type oligosaccharides and that carbohydrate II is composed of Olinked oligosaccharides with a core of Gal-GalNAc [12]. However, the antibiotic tunicamycin, an inhibitor of Nlinked glycosylation, inhibited the two-step glycosylation by carbohydrates I and II of contact site A, although the glycosylation by carbohydrate I was much more sensitive to tunicamycin than that by carbohydrate II [10]. The properties of carbohydrate II were thus investigated in more detail in order to confirm whether carbohydrate II is indeed composed of Olinked oligosaccharides. Here, we report on the characterization and distribution of carbohydrate II. MATERIALS AND METHODS Cell culture. Cells of D. discoideum AX2-214 were cultivated at 227C in a nutrient medium with 1.8% maltose as described by Watts and Ashworth [14]. Development was begun by washing the cells in 17 mM Soerensen’s phosphate buffer, pH 6.1 (standard buffer), and continued for 8 h with shaking at 150 rpm at 227C. Cells of the mod B mutant HL220 were provided by W. F. Loomis. HL220 mutant cells were grown and developed, and tunicamycin treatment was carried out as described in a previous paper [9]. Purification of contact site A. Purified contact site A was prepared as described in the previous work [9]. Purified contact site A was
393
AID
ECR 3434
/
6i19$$$181
01-20-97 17:24:30
0014-4827/97 $25.00 Copyright q 1997 by Academic Press All rights of reproduction in any form reserved.
ecal
394
YOSHIDA, YOKOTA, AND OUCHI
detected as a single band by staining sodium dodecyl sulfate (SDS)– polyacryamide gels with silver according to the procedure of Oakley et al. [15]. Gel electrophoresis and immunoblotting. SDS–polyacryamide electrophoresis in 10% gels was carried out using the standard method [16]. To obtain particulate fractions, cells were frozen, thawed, and centrifuged at 15,000 rpm for 20 min. The particulate fractions were dissolved in sample buffer for SDS–polyacryamide gel electrophoresis. For two-dimensional electrophoresis [17], 1 vol of ampholytes at pH 3–10 and 4 vol of ampholytes at pH 4–6 were used. For transfer to nitrocellulose (BA85; Schleicher & Schuell) or polyvinylidene difluoride (PVDF) membranes, the method of Towbin et al. [18] was used. Preparation of antibodies and labeling with antibodies and lectins. The antibodies against carbohydrate II (anti-carbohydrate II) were prepared by the following procedures (Fig. 1). The polyclonal antibodies against purified contact site A were prepared and absorbed with particulate fractions from growth phase cells as described in the previous paper [9]. To remove the antibodies against protein moieties of contact site A, the absorbed antibodies were further absorbed with particulate fractions from the cells, which were treated with tunicamycin during growth/starvation and expressed a 53-kDa contact site A devoid of carbohydrates I and II [9, 10]. Staining was performed with goat anti-rabbit IgG conjugated with horseradish peroxidase (E-Y Labs.). We used WGA (Sigma) for GlcNAc and NeuAc and Maackia amurensis leukoagglutinin (MAL; E-Y Labs.) for NeuAc, all of which were conjugated with horseradish peroxidase. In the case of MAL, the blotting amplification system (BLAST; DuPont) was used to amplify the sensitivity of lectin reactivity. Proteolysis of contact site A. Purified contact site A was suspended in 0.1% SDS and digested with Achromobacter protease I (Wako Chemicals), which is highly specific for lysine [19], in 10 mM phosphate buffer at pH 7.8 and 377C for 2 h at an enzyme to substrate ratio of 1:50 (wt:wt). The digestion was terminated by the addition of electrophoresis sample buffer to the reaction mixture. The proteolytic fragments were separated by SDS–polyacryamide of 10% gels, transferred onto PVDF membranes, and processed for immunoblotting as described above.
FIG. 1. Reactivity of contact site A with anti-carbohydrate II. Particulate fractions from growth-phase cells (lane 1), aggregationcompetent cells (lane 2), tunicamycin-treated cells (lane 3), and HL220 mutant cells (lane 4) were applied to SDS–polyacrylamide gel electrophoresis. Anti-carbohydrate II reacted with a 66-kDa contact site A retaining carbohydrate II in tunicamycin-treated cells, but it did not react with a 68-kDa contact site A retaining carbohydrate I in the mutant HL220. An 80-kDa contact site A retains carbohydrates I and II. The nonreactivity with the HL220 mutant indicates that anti-carbohydrate II does not contain antibodies against carbohydrate I and protein moieties of contact site A.
AID
ECR 3434
/
6i19$$$182
01-20-97 17:24:30
Amino-terminal sequence analysis. A part of the PVDF membrane was incubated with WGA or anti-carbohydrate II. In the case of a 21-kDa fragment, a part of the PVDF membrane was stained with Coomassie brilliant blue. The regions containing fragments of 31, 55, or 21 kDa were identified and cut away from the bulk of the PVDF membrane. These isolated regions were sequenced in a Shimadzu PSQ-1 protein sequencer equipped for on-line analysis of phenylthiohydantoin derivatives of amino acids; this was done according to the procedure of Matsudaira [20]. Preparation of protein A–gold probe. Colloidal gold was prepared in 15-nm particles as described in a previous paper [21]. The protein A–gold probe was mixed with glycerol (40%) and bovine serum albumin (0.1%) and stored at 0207C. Immunoelectron microscopy. Aggregation-competent cells were fixed with 2.5% glutaraldehyde in standard buffer on ice for 2 h. Excess fixative was removed by washing with standard buffer. The fixed cells were incubated with standard buffer containing 100 mM lysine and processed with postembedding immunoelectron microscopy as described previously [21]; briefly, the samples were dehydrated with graded dimethylformamide at 0207C and embedded in LR-White (Bio-Rad Labs.). The resin was polymerized under ultraviolet light at 0207C. Ultra-thin sections of LR-White-embedded materials were stained with anti-carbohydrate II and protein A–gold.
RESULTS
Determination of Amino Acid Sequences to Which Carbohydrate II Is Bound and Determination of the NH2 Terminus of Each Fragment In order to clarify the function of carbohydrate II, it is important to determine the amino acid sequences of contact site A to which carbohydrate II is bound. After digestion of purified contact site A with Achromobacter protease I, the products were electrophoresed, transferred onto PVDF membranes, and incubated with WGA and anti-carbohydrate II (Fig. 2). The reactivities of these two probes with fragments of contact site A were slightly different. WGA reacted with the 55- and 31-kDa fragments, whereas anti-carbohydrate II reacted with the 31- and 21-kDa fragments. However, both probes primarily reacted with the 31-kDa fragment, suggesting that the 31-kDa fragment retained carbohydrate II. The region corresponding to the 31kDa fragment was isolated and applied to an amino acid sequencer. We determined the NH2 terminus as Thr-Glu-Ala-Thr-Thr-Ser on the 31-kDa fragment. It was estimated from the cDNA sequence data [22] that the 31-kDa fragment was located in the COOH terminal region of contact site A. On the 31-kDa fragment, there were more than 20 Ser/Thr residues as target binding sites for O-linked oligosaccharides, while there was no motif for N-glycosylation of Asn-X-Ser/Thr. These findings provide conclusive evidence that carbohydrate II was composed of O-linked carbohydrates and suggest that WGA and anti-carbohydrate II reacted with the clustered O-linked oligosaccharides formed in the COOH terminal region. In order to clarify the relationship between the 31- and 55-kDa fragments, the NH2 terminus of the 55-kDa fragment was identified as Gln-Val-Asn-Asp-Ser. This identification
ecal
CLUSTERED O-LINKED OLIGOSACCHARIDES IN D. discoideum
395
carried out to verify whether the glycosylation by carbohydrate II occurs in the Golgi apparatus or in the endoplasmic reticulum. Figure 4 shows that the glycosylation by carbohydrate II clearly occurs in the Golgi apparatus, which was characterized by a few stacked cisternae but numerous vesicles and vacuoles. Next, the cell–cell contact regions were heavily labeled with gold particles. On the other hand, gold particles were distributed sparsely on the cell surface of noncontact regions (Figs. 4A–4C). No gold particles were observed when treated with preimmune IgG (data not shown). DISCUSSION
FIG. 2. Reactivities of protease-digested contact site A with WGA and with anti-carbohydrate II. Purified contact site A was digested with Achromobacter protease I and applied to SDS–polyacryamide gel electrophoresis. Proteins were revealed by silver staining (A). Alternatively, after transfer, PVDF membranes were incubated with WGA (B) and anti-carbohydrate II (C). (A and B) Lane 1, undigested contact site A; lane 2, digested contact site A. Arrowheads indicate bands of 55-, 31-, and 21-kDa fragments.
implies that 151 amino acids exist between the NH2 terminus of the 55-kDa fragment and that of the 31kDa fragment. Moreover, the results show that the 21kDa fragment contained the NH2 terminus of contact site A: Ala-Pro-Thr-Ile-Thr-Ala. Reactivity of Contact Site A Fragments with LectinRecognizing Sialic Acids We have previously reported that contact site A is a major glycoprotein containing sialic acids [13]. It is important to confirm whether sialic acids are indeed included in carbohydrates of contact site A in order to understand the function of contact site A. Here, supplementing our previous data, results of two-dimensional electrophoresis indicate that MAL for a2,3-linked sialic acid reacted with contact site A (Fig. 3). Localization of Carbohydrate II by Immunoelectron Microscopy Tunicamycin blocks the two-step glycosylation by carbohydrates I and II [9, 10]. However, carbohydrate II is composed of O-linked oligosaccharides. This is clearly inconsistent with the fact that tunicamycin inhibits only the glycosylation by N-linked oligosaccharides. It was then conjectured that O-glycosylation might occur in the endoplasmic reticulum in D. discoideum. Accordingly, immunoelectron microscopy was
AID
ECR 3434
/
6i19$$$182
01-20-97 17:24:30
It is known that the cell adhesion molecule, contact site A, retains two types of carbohydrates: carbohydrate I, which is sulfatable, and carbohydrate II, which is mainly labeled with WGA [3]. In this paper, we elucidated the properties of carbohydrate II in greater detail. The 31-kDa fragment of contact site A digested with Achromobacter protease I was recognized as the most prominent fragment by both WGA and anti-carbohydrate II; consequently, it is conceivable that carbohydrate II was included in this fragment. In addition to the 31-kDa fragment, the 55-kDa fragment was faintly labeled by WGA. The result of identifying the NH2 terminus of the 55-kDa fragment suggests that the 55kDa fragment was partially digested with Achromobacter protease I and that it contained clustered Olinked oligosaccharides. In a previous paper [11], we indicated that WGA reacted with an incomplete 68kDa contact site A, which does not retain carbohydrate II, of an HG794 mutant strain. Therefore, we cannot rule out the possibility that WGA reacts with oligosaccharides other than clustered O-linked oligosaccharides in the 55-kDa fragment. Anti-carbohydrate II, retaining the only reactivity with carbohydrate moieties, reacted with the 21-kDa fragment. However, it should be noted that this might be a different substance from the 21-kDa fragment containing the NH2 terminus of contact site A (21-kDaN fragment); rather than being produced by protease digestion, this substance with O-linked oligosaccharides (21 kDaC fragment) might be a degradation product from a 31-kDa fragment. This indicates that the 21kDaN fragment with NH2 terminus was contaminated with a small amount of 21-kDaC fragment with Olinked oligosaccharides. It is possible that the 21-kDaC fragment is a degradation product having insufficient antigenicity of carbohydrate II. Therefore, differences in 21-kDaC reactivities between WGA and anti-carbohydrate II might be due to the strength of affinity for carbohydrate II and/or the variant recognition for carbohydrate II. The glycosylation by carbohydrate II was inhibited by tunicamycin despite O-linked oligosaccharides, and
ecal
396
YOSHIDA, YOKOTA, AND OUCHI
FIG. 3. Two-dimensional electrophoresis of contact site A. Particulate fractions from aggregation-competent cells were applied to twodimensional electrophoresis. After transfer, the nitrocellulose membrane was incubated with MAL (A). The same membrane was subsequently incubated with anti-carbohydrate II (B). Arrowheads indicate contact site A.
this fact initially led us to a false conclusion concerning the structure of carbohydrate II. Since the part of the 53-kDa contact site A devoid of carbohydrates I and II was transported to the Golgi apparatus [23], we can discount the possibility that contact site A is unmodified by O-glycosylation because of the protein not reaching the Golgi apparatus. We alternatively conjectured that O-glycosylation might occur in the endoplasmic reticulum in D. discoideum. In fact, evidence in previ-
AID
ECR 3434
/
6i19$$$182
01-20-97 17:24:30
ous work shows that O-glycosylation occurs in the endoplasmic reticulum [24–26]. In yeast, for example, it is known that O-glycosylation occurs in the endoplasmic reticulum [27]. An immunoelectron microscopic staining study using anti-carbohydrate II suggested that O-glycosylation occurs in the Golgi apparatus in D. discoideum as observed in mammalian cells. In fungus, dolichol phosphate intermediates are commonly implicated in O-glycosylation [25]. It seems likely that O-
ecal
CLUSTERED O-LINKED OLIGOSACCHARIDES IN D. discoideum
397
glycosylation in D. discoideum was inhibited by tunicamycin because of the involvement of dolichol intermediates, although such an influence of dolichol intermediates has not yet been observed in O-glycosylation in the Golgi apparatus of D. discoideum. The current paper is the first to report that O-glycosylation in the Golgi apparatus was inhibited by tunicamycin, although it remains to be determined whether GalNAc was transferred into protein moieties at the endoplasmic reticulum or the Golgi apparatus of D. discoideum. Moreover, immunoelectron microscopy showed an irregular distribution of gold particles in cell–cell contact and noncontact regions. High concentrations of gold particles were observed on the contact regions of two cells, but gold particles were distributed sparsely on the noncontact regions. These results were consistent with the data of Choi and Siu [28], who observed the distribution of contact site A using monoclonal antibodies against protein moieties. On the other hand, Ochiai et al. [29] reported that contact site A antigens were uniformly distributed on the cell surface on the basis of observation with monoclonal antibodies against carbohydrate II. Though in both their study and the current work antibodies were raised against carbohydrate II, different results were obtained. The discrepancy may possibly reflect differences in the specificity of antibodies. Our immunoelectron microscopic studies suggested that a GPI anchor in contact site A might function in the rapid changes of cell contact sites caused by the reorientation in the cyclic AMP gradient. In this regard, chimeras of transmembrane and GPI-anchor proteins show that the GPI anchor plays a significant role in the higher mobility of the GPI-linked protein [30]. Another possibility that has been pointed out is that the GPI anchor of contact site A is involved in longevity on the cell surface [31]. We have previously reported that carbohydrate II might be functional in the stability of contact site A [32] and that WGA inhibited the cell adhesion by contact site A [3]. Taken together, the present results and the cell-adhesion inhibition by WGA suggest that the cell-adhesion inhibition is caused by the binding of WGA to clustered O-linked oligosaccharides. However, it remains to be determined how the three-dimensional structures of contact site A change through cell–cell interaction under natural conditions. We thank Dr. W. F. Loomis for providing the mod B mutant HL220 and Dr. A. Hayashi for critically reading the manuscript. This study was supported in part by Special Coordination Funds for promoting Science and Technology of the Science and Technology Agency of the Japanese Government. FIG. 4. Immunoelectron micrographs of aggregation-competent cells. Labeling was carried out by a protein–gold method, using anticarbohydrate II. The labels show the localization of carbohydrate II in the Golgi apparatus (G in A and B), in the cell–cell contact regions (solid arrow in C), and in the noncontact regions (open arrow in A– C). Bars, 0.5 mm.
1. Brandley, B. K., and Schnaar, R. L. (1986) J. Leukocyte Biol. 40, 97–111.
AID
ecal
ECR 3434
/
6i19$$$182
01-20-97 17:24:30
REFERENCES
398
YOSHIDA, YOKOTA, AND OUCHI
2. Stadler, J., Gerisch, G., Bauer, G., Suchanek, C., and Huttner, W. B. (1983) EMBO J. 2, 1137–1143. 3. Yoshida, M., Stadler, J., Bertholdt, G., and Gerisch, G. (1984) EMBO J. 3, 2663–2670. 4. Lam, T. Y., and Siu, C.-H. (1982) Dev. Biol. 92, 398–407. 5. Ochiai, H., Stadler, J., Westphal, M., Wagle, G., Merkl, R., and Gerisch, G. (1982) EMBO J. 1, 1011–1016. 6. Yamada, H., Hirano, T., Miyazaki, T., Takatsuki, A., and Tamura, G. (1982) J. Biochem. 92, 339–406. 7. Murray, B. A., Wheeler, S., Jongens, T., and Loomis, W. F. (1984) Mol. Cell. Biol. 4, 514–519. 8. Loomis, W. F., Wheeler, S. A., Springer, W. R., and Barondes, S. H. (1985) Dev. Biol. 109, 111–117. 9. Yoshida, M. (1987) J. Biochem. 101, 1233–1245. 10. Yoshida, M. (1991) Comp. Biochem. Physiol. 98B, 563–568. 11. Yoshida, M., Ishida, S., and Iizuka, Y. (1991) Cell Struct. Funct. 16, 383–390. 12. Yoshida, M., Matsui, T., Fuse, G., and Ouchi, S. (1993) FEBS Lett. 318, 305–309. 13. Yoshida, M., Fuse, G., Matsui, T., and Ouchi, S. (1992) Biochem. Biophys. Res. Commun. 188, 794–798. 14. Watts, D. J., and Ashworth, J. M. (1970) Biochem. J. 119, 171– 174. 15. Oakley, B. R., Kirsch, D. R., and Morris, N. R. (1980) Anal. Biochem. 105, 361–363. 16. Laemmli, U. K. (1970) Nature 227, 680–685. 17. O’Farrell, P. H. (1975) J. Biol. Chem. 250, 4007–4021.
18. Towbin, H., Staehelin, T., and Gordon, J. (1979) Proc. Natl. Acad. Sci. USA 76, 4350–4354. 19. Tsunasawa, S., Masaki, T., Hirose, M., Soejima, M., and Sakiyama, F. (1989) J. Biol. Chem. 264, 3832–3839. 20. Matsudaira, P. (1987) J. Biol. Chem. 262, 10035–10038. 21. Yokota, S., (1990) Eur. J. Cell Biol. 53, 112–121. 22. Kamboj, R. K., Wong, L. M., Lam, T. Y., and Siu, C.-H. (1988) J. Cell Biol. 107, 1835–1843. 23. Hohmann, H.-P., Bozzaro, S., Yoshida, M., Merkl, R., and Gerisch, G. (1987) J. Biol. Chem. 262, 16618–16624. 24. Strous, G. J. A. M. (1979) Proc. Natl. Acad. Sci. USA 76, 2694– 2698. 25. Tanner, W., and Lehle, L. (1987) Biochim. Biophys. Acta 906, 81–99. 26. Spielman, J., Hull, S. R., Sheng, Z., Kanterman, R., Bright, A., and Carraway, K. L. (1988) J. Biol. Chem. 263, 9621–9629. 27. Strahl-Bolsinger, S., Immervoll, T., Deutzmann, R., and Tanner, W. (1993) Proc. Natl. Acad. Sci. USA 90, 8164–8168. 28. Choi, A. H. C., and Siu, C.-H. (1987) J. Cell Biol. 104, 1375– 1387. 29. Ochiai, H., Schwarz, H., Merkl, R., Wagle, G., and Gerisch, G. (1982) Cell Differ. 11, 1–13. 30. Zhang, F., Crise, B., Su, B., Hou, Y., Rose, J. K., Bothwell, A., and Jacobson, K. (1991) J. Cell Biol. 115, 75–84. 31. Barth, A., Mueller-Taubenberger, A., Taranto, P., and Gerisch, G. (1994) J. Cell Biol. 124, 205–215. 32. Hohmann, H.-P., Bozzaro, S., Merkl, R., Wallraff, E., Yoshida, M., Weinhart, U., and Gerisch, G. (1987) EMBO J. 6, 3663– 3671.
Received November 5, 1996 Revised version received October 29, 1996
AID
ECR 3434
/
6i19$$$183
01-20-97 17:24:30
ecal
230, 393–398 (1997)
EX963434
Characterization and Distribution of O-Glycosylated Carbohydrates in the Cell Adhesion Molecule, Contact Site A, from Dictyostelium discoideum MOTONOBU YOSHIDA,*,1 SADAKI YOKOTA,†
AND
SEIJI OUCHI‡
*Research Institute of Food Science and ‡Department of Agriculture, Kinki University, Higashi-Osaka, Osaka 577, Japan; and †Department of Anatomy, Yamanashi Medical School, Yamanashi 409-38, Japan
This paper presents further investigation of the properties of carbohydrate II in the cell adhesion molecule, contact site A, from Dictyostelium discoideum. A purified contact site A was digested with Achromobacter protease I to produce a 31-kDa fragment to which carbohydrate II was mainly bound and a 21-kDa fragment containing the NH2 terminus of contact site A, which was identified as Ala-Pro-Thr-Ile-Thr-Ala. The NH2 terminus of the 31-kDa fragment was ThrGlu-Ala-Thr-Thr-Ser. It was estimated from the cDNA sequence data of contact site A that more than 20 Ser/ Thr residues exist as target sites for the O-linked oligosaccharides in the 31-kDa fragment, but not for the N-linked oligosaccharides. These results suggest that carbohydrate II exists as clustered O-linked oligosaccharides in the COOH terminus of contact site A. The results of two-dimensional electrophoresis confirm that oligosaccharides of contact site A contain sialic acids. Immunoelectron microscopy was carried out to define the organelle in which O-glycosylation by carbohydrate II occurs and how carbohydrate II antigens are distributed on the cell surface. The results show that O-glycosylation can occur in the Golgi apparatus in D. discoideum as observed in other cells, although this O-glycosylation was inhibited by tunicamycin. Furthermore, gold particles were densely concentrated in cell–cell contact regions but sparsely distributed in noncontact regions. q 1997 Academic Press
INTRODUCTION
Recent studies have indicated that cell surface glycoproteins play an important role in various biological processes and that their carbohydrates are involved in determining intracellular sorting, intracellular stability, protein clearance from plasma, cell adhesion activity, and other biological functions [1]. The cell adhesion molecule, contact site A, is involved in EDTA-resistant 1
To whom correspondence and reprint requests should be addressed. Fax: /81 742 431155.
cell contact and appears in the aggregation process of Dictyostelium discoideum. Contact site A is a glycoprotein with an apparent molecular weight of 80 kDa and is anchored in the cell membrane by glycophosphatidyl inositol (GPI) linkages. This glycoprotein consists of two types of carbohydrates: carbohydrate I, which is sulfatable [2], and carbohydrate II, which is mainly labeled with wheat germ agglutinin (WGA) [3]. Previous papers have suggested that carbohydrates might be involved in EDTA-resistant cell contact [4–11]. In particular, the mod B mutants not retaining carbohydrate II have been shown to be impaired in EDTAresistant cell contact [7, 8]. These mutants have been isolated using the monoclonal antibodies, which recognize carbohydrate II and partially block EDTA-resistant cell contact. Consequently, we isolated and determined the carbohydrate structures of contact site A [12] and showed the existence of sialic acids in D. discoideum [13]. We have previously reported that carbohydrate I is composed of N-linked complex type oligosaccharides and that carbohydrate II is composed of Olinked oligosaccharides with a core of Gal-GalNAc [12]. However, the antibiotic tunicamycin, an inhibitor of Nlinked glycosylation, inhibited the two-step glycosylation by carbohydrates I and II of contact site A, although the glycosylation by carbohydrate I was much more sensitive to tunicamycin than that by carbohydrate II [10]. The properties of carbohydrate II were thus investigated in more detail in order to confirm whether carbohydrate II is indeed composed of Olinked oligosaccharides. Here, we report on the characterization and distribution of carbohydrate II. MATERIALS AND METHODS Cell culture. Cells of D. discoideum AX2-214 were cultivated at 227C in a nutrient medium with 1.8% maltose as described by Watts and Ashworth [14]. Development was begun by washing the cells in 17 mM Soerensen’s phosphate buffer, pH 6.1 (standard buffer), and continued for 8 h with shaking at 150 rpm at 227C. Cells of the mod B mutant HL220 were provided by W. F. Loomis. HL220 mutant cells were grown and developed, and tunicamycin treatment was carried out as described in a previous paper [9]. Purification of contact site A. Purified contact site A was prepared as described in the previous work [9]. Purified contact site A was
393
AID
ECR 3434
/
6i19$$$181
01-20-97 17:24:30
0014-4827/97 $25.00 Copyright q 1997 by Academic Press All rights of reproduction in any form reserved.
ecal
394
YOSHIDA, YOKOTA, AND OUCHI
detected as a single band by staining sodium dodecyl sulfate (SDS)– polyacryamide gels with silver according to the procedure of Oakley et al. [15]. Gel electrophoresis and immunoblotting. SDS–polyacryamide electrophoresis in 10% gels was carried out using the standard method [16]. To obtain particulate fractions, cells were frozen, thawed, and centrifuged at 15,000 rpm for 20 min. The particulate fractions were dissolved in sample buffer for SDS–polyacryamide gel electrophoresis. For two-dimensional electrophoresis [17], 1 vol of ampholytes at pH 3–10 and 4 vol of ampholytes at pH 4–6 were used. For transfer to nitrocellulose (BA85; Schleicher & Schuell) or polyvinylidene difluoride (PVDF) membranes, the method of Towbin et al. [18] was used. Preparation of antibodies and labeling with antibodies and lectins. The antibodies against carbohydrate II (anti-carbohydrate II) were prepared by the following procedures (Fig. 1). The polyclonal antibodies against purified contact site A were prepared and absorbed with particulate fractions from growth phase cells as described in the previous paper [9]. To remove the antibodies against protein moieties of contact site A, the absorbed antibodies were further absorbed with particulate fractions from the cells, which were treated with tunicamycin during growth/starvation and expressed a 53-kDa contact site A devoid of carbohydrates I and II [9, 10]. Staining was performed with goat anti-rabbit IgG conjugated with horseradish peroxidase (E-Y Labs.). We used WGA (Sigma) for GlcNAc and NeuAc and Maackia amurensis leukoagglutinin (MAL; E-Y Labs.) for NeuAc, all of which were conjugated with horseradish peroxidase. In the case of MAL, the blotting amplification system (BLAST; DuPont) was used to amplify the sensitivity of lectin reactivity. Proteolysis of contact site A. Purified contact site A was suspended in 0.1% SDS and digested with Achromobacter protease I (Wako Chemicals), which is highly specific for lysine [19], in 10 mM phosphate buffer at pH 7.8 and 377C for 2 h at an enzyme to substrate ratio of 1:50 (wt:wt). The digestion was terminated by the addition of electrophoresis sample buffer to the reaction mixture. The proteolytic fragments were separated by SDS–polyacryamide of 10% gels, transferred onto PVDF membranes, and processed for immunoblotting as described above.
FIG. 1. Reactivity of contact site A with anti-carbohydrate II. Particulate fractions from growth-phase cells (lane 1), aggregationcompetent cells (lane 2), tunicamycin-treated cells (lane 3), and HL220 mutant cells (lane 4) were applied to SDS–polyacrylamide gel electrophoresis. Anti-carbohydrate II reacted with a 66-kDa contact site A retaining carbohydrate II in tunicamycin-treated cells, but it did not react with a 68-kDa contact site A retaining carbohydrate I in the mutant HL220. An 80-kDa contact site A retains carbohydrates I and II. The nonreactivity with the HL220 mutant indicates that anti-carbohydrate II does not contain antibodies against carbohydrate I and protein moieties of contact site A.
AID
ECR 3434
/
6i19$$$182
01-20-97 17:24:30
Amino-terminal sequence analysis. A part of the PVDF membrane was incubated with WGA or anti-carbohydrate II. In the case of a 21-kDa fragment, a part of the PVDF membrane was stained with Coomassie brilliant blue. The regions containing fragments of 31, 55, or 21 kDa were identified and cut away from the bulk of the PVDF membrane. These isolated regions were sequenced in a Shimadzu PSQ-1 protein sequencer equipped for on-line analysis of phenylthiohydantoin derivatives of amino acids; this was done according to the procedure of Matsudaira [20]. Preparation of protein A–gold probe. Colloidal gold was prepared in 15-nm particles as described in a previous paper [21]. The protein A–gold probe was mixed with glycerol (40%) and bovine serum albumin (0.1%) and stored at 0207C. Immunoelectron microscopy. Aggregation-competent cells were fixed with 2.5% glutaraldehyde in standard buffer on ice for 2 h. Excess fixative was removed by washing with standard buffer. The fixed cells were incubated with standard buffer containing 100 mM lysine and processed with postembedding immunoelectron microscopy as described previously [21]; briefly, the samples were dehydrated with graded dimethylformamide at 0207C and embedded in LR-White (Bio-Rad Labs.). The resin was polymerized under ultraviolet light at 0207C. Ultra-thin sections of LR-White-embedded materials were stained with anti-carbohydrate II and protein A–gold.
RESULTS
Determination of Amino Acid Sequences to Which Carbohydrate II Is Bound and Determination of the NH2 Terminus of Each Fragment In order to clarify the function of carbohydrate II, it is important to determine the amino acid sequences of contact site A to which carbohydrate II is bound. After digestion of purified contact site A with Achromobacter protease I, the products were electrophoresed, transferred onto PVDF membranes, and incubated with WGA and anti-carbohydrate II (Fig. 2). The reactivities of these two probes with fragments of contact site A were slightly different. WGA reacted with the 55- and 31-kDa fragments, whereas anti-carbohydrate II reacted with the 31- and 21-kDa fragments. However, both probes primarily reacted with the 31-kDa fragment, suggesting that the 31-kDa fragment retained carbohydrate II. The region corresponding to the 31kDa fragment was isolated and applied to an amino acid sequencer. We determined the NH2 terminus as Thr-Glu-Ala-Thr-Thr-Ser on the 31-kDa fragment. It was estimated from the cDNA sequence data [22] that the 31-kDa fragment was located in the COOH terminal region of contact site A. On the 31-kDa fragment, there were more than 20 Ser/Thr residues as target binding sites for O-linked oligosaccharides, while there was no motif for N-glycosylation of Asn-X-Ser/Thr. These findings provide conclusive evidence that carbohydrate II was composed of O-linked carbohydrates and suggest that WGA and anti-carbohydrate II reacted with the clustered O-linked oligosaccharides formed in the COOH terminal region. In order to clarify the relationship between the 31- and 55-kDa fragments, the NH2 terminus of the 55-kDa fragment was identified as Gln-Val-Asn-Asp-Ser. This identification
ecal
CLUSTERED O-LINKED OLIGOSACCHARIDES IN D. discoideum
395
carried out to verify whether the glycosylation by carbohydrate II occurs in the Golgi apparatus or in the endoplasmic reticulum. Figure 4 shows that the glycosylation by carbohydrate II clearly occurs in the Golgi apparatus, which was characterized by a few stacked cisternae but numerous vesicles and vacuoles. Next, the cell–cell contact regions were heavily labeled with gold particles. On the other hand, gold particles were distributed sparsely on the cell surface of noncontact regions (Figs. 4A–4C). No gold particles were observed when treated with preimmune IgG (data not shown). DISCUSSION
FIG. 2. Reactivities of protease-digested contact site A with WGA and with anti-carbohydrate II. Purified contact site A was digested with Achromobacter protease I and applied to SDS–polyacryamide gel electrophoresis. Proteins were revealed by silver staining (A). Alternatively, after transfer, PVDF membranes were incubated with WGA (B) and anti-carbohydrate II (C). (A and B) Lane 1, undigested contact site A; lane 2, digested contact site A. Arrowheads indicate bands of 55-, 31-, and 21-kDa fragments.
implies that 151 amino acids exist between the NH2 terminus of the 55-kDa fragment and that of the 31kDa fragment. Moreover, the results show that the 21kDa fragment contained the NH2 terminus of contact site A: Ala-Pro-Thr-Ile-Thr-Ala. Reactivity of Contact Site A Fragments with LectinRecognizing Sialic Acids We have previously reported that contact site A is a major glycoprotein containing sialic acids [13]. It is important to confirm whether sialic acids are indeed included in carbohydrates of contact site A in order to understand the function of contact site A. Here, supplementing our previous data, results of two-dimensional electrophoresis indicate that MAL for a2,3-linked sialic acid reacted with contact site A (Fig. 3). Localization of Carbohydrate II by Immunoelectron Microscopy Tunicamycin blocks the two-step glycosylation by carbohydrates I and II [9, 10]. However, carbohydrate II is composed of O-linked oligosaccharides. This is clearly inconsistent with the fact that tunicamycin inhibits only the glycosylation by N-linked oligosaccharides. It was then conjectured that O-glycosylation might occur in the endoplasmic reticulum in D. discoideum. Accordingly, immunoelectron microscopy was
AID
ECR 3434
/
6i19$$$182
01-20-97 17:24:30
It is known that the cell adhesion molecule, contact site A, retains two types of carbohydrates: carbohydrate I, which is sulfatable, and carbohydrate II, which is mainly labeled with WGA [3]. In this paper, we elucidated the properties of carbohydrate II in greater detail. The 31-kDa fragment of contact site A digested with Achromobacter protease I was recognized as the most prominent fragment by both WGA and anti-carbohydrate II; consequently, it is conceivable that carbohydrate II was included in this fragment. In addition to the 31-kDa fragment, the 55-kDa fragment was faintly labeled by WGA. The result of identifying the NH2 terminus of the 55-kDa fragment suggests that the 55kDa fragment was partially digested with Achromobacter protease I and that it contained clustered Olinked oligosaccharides. In a previous paper [11], we indicated that WGA reacted with an incomplete 68kDa contact site A, which does not retain carbohydrate II, of an HG794 mutant strain. Therefore, we cannot rule out the possibility that WGA reacts with oligosaccharides other than clustered O-linked oligosaccharides in the 55-kDa fragment. Anti-carbohydrate II, retaining the only reactivity with carbohydrate moieties, reacted with the 21-kDa fragment. However, it should be noted that this might be a different substance from the 21-kDa fragment containing the NH2 terminus of contact site A (21-kDaN fragment); rather than being produced by protease digestion, this substance with O-linked oligosaccharides (21 kDaC fragment) might be a degradation product from a 31-kDa fragment. This indicates that the 21kDaN fragment with NH2 terminus was contaminated with a small amount of 21-kDaC fragment with Olinked oligosaccharides. It is possible that the 21-kDaC fragment is a degradation product having insufficient antigenicity of carbohydrate II. Therefore, differences in 21-kDaC reactivities between WGA and anti-carbohydrate II might be due to the strength of affinity for carbohydrate II and/or the variant recognition for carbohydrate II. The glycosylation by carbohydrate II was inhibited by tunicamycin despite O-linked oligosaccharides, and
ecal
396
YOSHIDA, YOKOTA, AND OUCHI
FIG. 3. Two-dimensional electrophoresis of contact site A. Particulate fractions from aggregation-competent cells were applied to twodimensional electrophoresis. After transfer, the nitrocellulose membrane was incubated with MAL (A). The same membrane was subsequently incubated with anti-carbohydrate II (B). Arrowheads indicate contact site A.
this fact initially led us to a false conclusion concerning the structure of carbohydrate II. Since the part of the 53-kDa contact site A devoid of carbohydrates I and II was transported to the Golgi apparatus [23], we can discount the possibility that contact site A is unmodified by O-glycosylation because of the protein not reaching the Golgi apparatus. We alternatively conjectured that O-glycosylation might occur in the endoplasmic reticulum in D. discoideum. In fact, evidence in previ-
AID
ECR 3434
/
6i19$$$182
01-20-97 17:24:30
ous work shows that O-glycosylation occurs in the endoplasmic reticulum [24–26]. In yeast, for example, it is known that O-glycosylation occurs in the endoplasmic reticulum [27]. An immunoelectron microscopic staining study using anti-carbohydrate II suggested that O-glycosylation occurs in the Golgi apparatus in D. discoideum as observed in mammalian cells. In fungus, dolichol phosphate intermediates are commonly implicated in O-glycosylation [25]. It seems likely that O-
ecal
CLUSTERED O-LINKED OLIGOSACCHARIDES IN D. discoideum
397
glycosylation in D. discoideum was inhibited by tunicamycin because of the involvement of dolichol intermediates, although such an influence of dolichol intermediates has not yet been observed in O-glycosylation in the Golgi apparatus of D. discoideum. The current paper is the first to report that O-glycosylation in the Golgi apparatus was inhibited by tunicamycin, although it remains to be determined whether GalNAc was transferred into protein moieties at the endoplasmic reticulum or the Golgi apparatus of D. discoideum. Moreover, immunoelectron microscopy showed an irregular distribution of gold particles in cell–cell contact and noncontact regions. High concentrations of gold particles were observed on the contact regions of two cells, but gold particles were distributed sparsely on the noncontact regions. These results were consistent with the data of Choi and Siu [28], who observed the distribution of contact site A using monoclonal antibodies against protein moieties. On the other hand, Ochiai et al. [29] reported that contact site A antigens were uniformly distributed on the cell surface on the basis of observation with monoclonal antibodies against carbohydrate II. Though in both their study and the current work antibodies were raised against carbohydrate II, different results were obtained. The discrepancy may possibly reflect differences in the specificity of antibodies. Our immunoelectron microscopic studies suggested that a GPI anchor in contact site A might function in the rapid changes of cell contact sites caused by the reorientation in the cyclic AMP gradient. In this regard, chimeras of transmembrane and GPI-anchor proteins show that the GPI anchor plays a significant role in the higher mobility of the GPI-linked protein [30]. Another possibility that has been pointed out is that the GPI anchor of contact site A is involved in longevity on the cell surface [31]. We have previously reported that carbohydrate II might be functional in the stability of contact site A [32] and that WGA inhibited the cell adhesion by contact site A [3]. Taken together, the present results and the cell-adhesion inhibition by WGA suggest that the cell-adhesion inhibition is caused by the binding of WGA to clustered O-linked oligosaccharides. However, it remains to be determined how the three-dimensional structures of contact site A change through cell–cell interaction under natural conditions. We thank Dr. W. F. Loomis for providing the mod B mutant HL220 and Dr. A. Hayashi for critically reading the manuscript. This study was supported in part by Special Coordination Funds for promoting Science and Technology of the Science and Technology Agency of the Japanese Government. FIG. 4. Immunoelectron micrographs of aggregation-competent cells. Labeling was carried out by a protein–gold method, using anticarbohydrate II. The labels show the localization of carbohydrate II in the Golgi apparatus (G in A and B), in the cell–cell contact regions (solid arrow in C), and in the noncontact regions (open arrow in A– C). Bars, 0.5 mm.
1. Brandley, B. K., and Schnaar, R. L. (1986) J. Leukocyte Biol. 40, 97–111.
AID
ecal
ECR 3434
/
6i19$$$182
01-20-97 17:24:30
REFERENCES
398
YOSHIDA, YOKOTA, AND OUCHI
2. Stadler, J., Gerisch, G., Bauer, G., Suchanek, C., and Huttner, W. B. (1983) EMBO J. 2, 1137–1143. 3. Yoshida, M., Stadler, J., Bertholdt, G., and Gerisch, G. (1984) EMBO J. 3, 2663–2670. 4. Lam, T. Y., and Siu, C.-H. (1982) Dev. Biol. 92, 398–407. 5. Ochiai, H., Stadler, J., Westphal, M., Wagle, G., Merkl, R., and Gerisch, G. (1982) EMBO J. 1, 1011–1016. 6. Yamada, H., Hirano, T., Miyazaki, T., Takatsuki, A., and Tamura, G. (1982) J. Biochem. 92, 339–406. 7. Murray, B. A., Wheeler, S., Jongens, T., and Loomis, W. F. (1984) Mol. Cell. Biol. 4, 514–519. 8. Loomis, W. F., Wheeler, S. A., Springer, W. R., and Barondes, S. H. (1985) Dev. Biol. 109, 111–117. 9. Yoshida, M. (1987) J. Biochem. 101, 1233–1245. 10. Yoshida, M. (1991) Comp. Biochem. Physiol. 98B, 563–568. 11. Yoshida, M., Ishida, S., and Iizuka, Y. (1991) Cell Struct. Funct. 16, 383–390. 12. Yoshida, M., Matsui, T., Fuse, G., and Ouchi, S. (1993) FEBS Lett. 318, 305–309. 13. Yoshida, M., Fuse, G., Matsui, T., and Ouchi, S. (1992) Biochem. Biophys. Res. Commun. 188, 794–798. 14. Watts, D. J., and Ashworth, J. M. (1970) Biochem. J. 119, 171– 174. 15. Oakley, B. R., Kirsch, D. R., and Morris, N. R. (1980) Anal. Biochem. 105, 361–363. 16. Laemmli, U. K. (1970) Nature 227, 680–685. 17. O’Farrell, P. H. (1975) J. Biol. Chem. 250, 4007–4021.
18. Towbin, H., Staehelin, T., and Gordon, J. (1979) Proc. Natl. Acad. Sci. USA 76, 4350–4354. 19. Tsunasawa, S., Masaki, T., Hirose, M., Soejima, M., and Sakiyama, F. (1989) J. Biol. Chem. 264, 3832–3839. 20. Matsudaira, P. (1987) J. Biol. Chem. 262, 10035–10038. 21. Yokota, S., (1990) Eur. J. Cell Biol. 53, 112–121. 22. Kamboj, R. K., Wong, L. M., Lam, T. Y., and Siu, C.-H. (1988) J. Cell Biol. 107, 1835–1843. 23. Hohmann, H.-P., Bozzaro, S., Yoshida, M., Merkl, R., and Gerisch, G. (1987) J. Biol. Chem. 262, 16618–16624. 24. Strous, G. J. A. M. (1979) Proc. Natl. Acad. Sci. USA 76, 2694– 2698. 25. Tanner, W., and Lehle, L. (1987) Biochim. Biophys. Acta 906, 81–99. 26. Spielman, J., Hull, S. R., Sheng, Z., Kanterman, R., Bright, A., and Carraway, K. L. (1988) J. Biol. Chem. 263, 9621–9629. 27. Strahl-Bolsinger, S., Immervoll, T., Deutzmann, R., and Tanner, W. (1993) Proc. Natl. Acad. Sci. USA 90, 8164–8168. 28. Choi, A. H. C., and Siu, C.-H. (1987) J. Cell Biol. 104, 1375– 1387. 29. Ochiai, H., Schwarz, H., Merkl, R., Wagle, G., and Gerisch, G. (1982) Cell Differ. 11, 1–13. 30. Zhang, F., Crise, B., Su, B., Hou, Y., Rose, J. K., Bothwell, A., and Jacobson, K. (1991) J. Cell Biol. 115, 75–84. 31. Barth, A., Mueller-Taubenberger, A., Taranto, P., and Gerisch, G. (1994) J. Cell Biol. 124, 205–215. 32. Hohmann, H.-P., Bozzaro, S., Merkl, R., Wallraff, E., Yoshida, M., Weinhart, U., and Gerisch, G. (1987) EMBO J. 6, 3663– 3671.
Received November 5, 1996 Revised version received October 29, 1996
AID
ECR 3434
/
6i19$$$183
01-20-97 17:24:30
ecal