Immunocytochemical localization of the protein reactive to human ╬▓-1 4-galactosyltransferase antibodies during chick embryonic skin differentiaion.код для вставкиСкачать
THE ANATOMICAL RECORD 243:109-119 (1995) lmmunocytochemical Localization of the Protein Reactive to Human p-1,4-GalactosyIt ransferase Antibodies During Chick Embryonic Skin Differentiation YOSHIHIRO AKIMOTO, AKIKO OBINATA, HIROYOSHI ENDO, KIYOSHI FURUKAWA, DAISUKE AOKI, SHIRO NOZAWA, AND HIROSHI HIRANO Department of Anatomy, Kyorin University School of Medicine, Mitaka, Tokyo (Y.A., H.H.); Department of Physiological Chemistry, Faculty of Pharmaceutical Sciences, Teikyo University, Sagamiko, Kanagawa (A.O.,H.E.); Department of Obstetrics and Gynecology, Keio University School of Medicine, Shinanomachi, Shinjuku-ku, Tokyo ( D A . , S.N.); and Department of Biosignal Research, Tokyo Metropolitan Znstitute of Gerontology, Sakae-cho, Itabashi-ku, Tokyo (K.F.); Japan ABSTRACT Background P-1, 4-Galactosyltransferase (GalTase) transfers galactose from UDP-galactose to terminal N-acetylglucosamine in glycoconjugates and is located both in the Golgi apparatus and in the plasma membrane. The cell surface GalTase is thought to be involved in cell-to-cell recognition and cell-to-extracellular matrix interaction. Methods: By the use of specific monoclonal antibodies against human GalTase, changes in cell surface localization of the protein reactive to the antibodies in chick embryonic skin during its differentiation in vivo and in vitro were detected immunohistochemically at both light- and electron microscopic levels. The distribution of glycoconjugates having terminal N-acetylglucosamine residues was detected by staining with succinylated wheat germ agglutinin (s-WGA). Results: Under the light microscope, intense immunostaining was observed in the keratinized epidermis, particularly in the intermediate layer. Marked changes in the localization of the staining were observed in vitamin A-induced mucus-secreting skin, in which keratinization was suppressed. The localization of the immunostaining was in parallel with that of glycoconjugates having terminal N-acetylglucosamine residues. Immunoelectron microscopically the immunostaining was located on the cell surface and in the intercellular space of the desmosomes in the intermediate cells of the keratinized epidermis. However, the staining was not present on the cell surface but was detected on the limiting membrane of the mucous granules, in the mucous metaplastic epidermis. In contrast, the staining was always found in the Golgi apparatus in all of the cells. Conclusions: These results suggest that the protein reactive to human GalTase antibody may be involved in chick epidermal differentiation. 0 1995 Wiley-Liss, Inc. Key words: Antibody against human 0-1, 4-Galactosyltransferase, Immunocytochemistry, Chick embryonic skin, Differentiation, Keratinization, Mucous metaplasia In our investigations of the differentiation processes of chick embryonic tarsometatarsal skin in vivo and in vitro, we observed marked changes in the desmosome (Takata et al., 1981), basement membrane (Hirano et al., 1985; Akimoto et al., 1988), plasma membrane (Takata and Hirano, 1983), and intercellular matrix (Akimoto et al., 1992). By using plant lectins as a probe, changes in the glycoconjugates on the cell surface of epidermis have also been shown t o occur during differentiation of chick embryonic skin (Takata and Hirano, 1983). Moreover, we found that changes in 0 1995 WILEY-LISS, INC localization of endogenous (3-galactoside-binding lectins occur during the skin differentiation in association with those in distribution of glycoconjugateson the cell surface (Hirano et al., 1988; Oda et al., 1989; Akimoto et al., 1992,1993). During the period of the skin differ- Received March 24, 1994; accepted March 2, 1995. Address reprint requests to Professor Hiroshi Hirano, Department of Anatomy, Kyorin University School of Medicine, Mitaka, Tokyo 181, Japan. 110 Y. AKIMOTO ET AL. entiation, in which keratinization starts from day 13 in ovo (stage 39; Hamburger and Hamilton, 1951) and is completed by day 17 in ovo (stage 43), the changes in glycoconjugates on the cell surface are closely related to the skin differentiation processes. There have been reported several mechanisms by which cells interact with either neighboring cells or extracellular matrix or with both (Takeichi, 1988; Sharon and Lis, 1989; Lasky et al., 1992). One of these is mediated by the cell surface carbohydrates. Glycosyltransferases (Roseman, 1970) and lectins (Sharon and Lis, 1989) including selectins (Lasky et al., 19921, are known to be receptors for this type of the cell recognition. p-1, 4-Galactosyltransferase (GalTase, EC 2. 4. 1, 22), which is an enzyme that transfers galactose from UDP-galactose to terminal N-acetylglucosamine in glycoconjugates (Trayer and Hill, 1971; Aoki et al., 19901, is located both in Golgi apparatus, more precisely in its trans region, and in the plasma membrane (Pestalozzi et al., 1982; Roth and Berger, 1982; Roth et al., 1985; Hartel-Schnenke et al., 1991; Suganuma et al., 1991; Watzele et al., 1991; Aoki et al., 1992; Cooke and Shur, 1994; Kawano et al., 1994). Several lines of evidence suggest that the cell surface GalTase is involved in cell to cell recognition in the early embryo and cell to extracellular matrix interaction such as cell migration and fertilization (Romagnano et al., 1990; Miller et al., 1991; Shur, 1991; Eckstein and Shur, 1992). Little is known about the cell surface expression and distribution of GalTase during skin differentiation. In the present study, we detected the protein reactive to human GalTase antibodies on the cell surface during epidermal keratinization but not mucous metaplasia induced in chick embryonic skin in vivo andlor in vitro. The similar distribution of glycoconjugateshaving terminal N-acetylglucosamine residues was also observed using s-WGA (succinylated wheat germ agglutinin) as a probe during the differentiation. Monoclonal Antibody We used mouse monoclonal antibodies (MAb 8513 and MAb 8628) against human GalTase (Uejima et al., 1992; Uemura et al., 1992). Uejima et al. (1992) reported previously on the epitopes of these monoclonal antibodies by using the recombinant GalTase expressed in Escherichia coli. The epitopes of the monoclonal antibodies resides on the protein moiety of GalTase but not on the carbohydrate moiety (Uejima et al., 1992). lrnmunoblotting Procedures Seventeen-day-old chick embryonic tarsometatarsal skin was dissected, homogenized in PBS containing 4 mM 2-mercaptoethanol, mixed with Laemmli sampling solution, and boiled for 3 min as described previously (Akimoto et al., 1992). The homogenate was subjected t o SDS-polyacrylamide gel electrophoresis (PAGE, 12.5% gel) and transferred to nitrocellulose sheets (Towbin et al., 1979). The sheets after blocking with bovine serum albumin were incubated with a mouse monoclonal antibody against human GalTase (Uemura et al., 1992) or normal mouse IgG or IgM, followed by horseradish peroxidase (HRP)-conjugated goat antimouse IgG or IgM. The protein band was visualized by immersion in 3,3’-diaminobenzidine.4HCl(DAB, 0.2 mg/ml)-H,O, (0.005%)for 5 min at room temperature. lmrnunoprecipitation and GalTase Assay Since the monoclonal antibody used in the present study was raised against human GalTase (Uemura et al., 19921, whether or not the antibody binds to the chick embryo transferase was studied. Livers (3 g) from 17-day-oldchick embryos were homogenized in 3 ml of phosphate-buffered saline (PBS), pH 7.2, containing 1%Triton X-100. The homogenate was centrifuged at 1,500g for 15 min. An aliquot (100 p1) of the supernatant was incubated with mouse monoclonal antiMATERIALS AND METHODS body 8628 (100 pl containing 3.5 mg protein) at 4°C for Skin 60 min. The mixture was further incubated with proTarsometatarsal skin of 13- and 17-day-oldchick em- tein A-Sepharose suspension (v/v, 1:l) (200 p1) at 4°C bryos was used. In some cases, skin explants from the for 60 min. After centrifugation a t 600g for 1min, the tarsometatarsal region of 13-day-old chick embryos immunoprecipitates were washed with PBS. The final were cultured for 9 days in a chemically defined me- pellet was suspended in 50 p1 of 20 mM Mes buffer, pH dium, BGJb (Biggers et al., 1961), containing 5% deli- 6.5. GalTase activity was determined as described pidized fetal calf serum (dFCS) and 20 nM hydrocorti- previously (Furukawa et al., 1990). In brief, the sone hemisuccinate (Japan Upjohn, Ltd., Tokyo) by the transferase assays were conducted in 50 p1 of 20 mM Millipore filter-roller-tube method (Sugimoto et al., Mes buffer, pH 6.5, containing 3 mM 5’-AMP,5 mM 2, 1974). Under these conditions, keratinization of the 3-dimercapto-1-propanol,0.05% Nonidet P-40, 10 mM epidermis occurs in vitro similarly as in ovo (Takata et MnCl,, 100 pM UDP-L3H1Gal (323 pcilmmol), 250 pg al., 1981). Or the skin explants were first cultured for asialo-agalacto-transferrin as an acceptor, and the 1 day in BGJb medium containing 5% dFCS, 20 nM immunoprecipitated material as an enzyme source. hydrocortisone, and 20 pM vitamin A (Sigma, St. Mixtures were incubated at 37°C in a shaking water Louis, MO), and then in BGJb supplemented with 2 bath for 60 min, and the reaction was terminated by mM Bt,cAMP (Sigma, St. Louis, MO) for 8 days by the adding 70 p1 of bovine serum albumin (1mg) and 30 p1 same method. As reported previously, keratinization of 50% trichloroacetic acid (TCA). After standing at was suppressed and mucous metaplasia was induced by 0°C for 60 min, reaction mixtures were transferred to the addition of vitamin A in vitro (Hirano et al., 1985; Whatman micro glass filters, and the filters were Obinata et al., 1991).Moreover, vitamin A-induced epi- washed thoroughly with 5% TCA. Radioactivities on dermal mucous metaplasia was accelerated by the ad- the filters were determined by a liquid scintillation dition of Bt,cAMP (Obinata et al., 1991). The medium counter as described previously (Furukawa and Roth, 1985). was renewed every other day. LOCALIZATION OF THE PROTEIN REACTIVE TO GALTASE ANTIBODY lmmunostaining for Light Microscopic Observation Tarsometatarsal skin specimens of 13- and 17-dayold chick embryos were fixed in 4% formaldehyde in 0.1 M phosphate buffer (pH 7.3) for 1 h at 4°C. The specimens were subsequently immersed in 2.3 M sucrose-10 mM PBS, and then frozen in liquid nitrogen. Frozen sections of 4 pm thickness were cut with a coldtome CM-41 (Sakura, Ltd, Tokyo), washed with PBS, and treated for 10 min with 1%bovine serum albumin (BSA) in PBS. Sections were incubated with a mouse monoclonal antibody against GalTase or with normal mouse IgG or IgM for 1h at room temperature, washed with PBS, incubated with the HRP-conjugated goat anti-mouse IgG or M for 1h, washed again with PBS, and then immersed in DAB (0.2 mg/ml)-H,Oz (0.005%) for 5 min at room temperature. Finally sections were rinsed in distilled water, dehydrated, cleared, and mounted. Nonspecific staining was checked by the incubation with the HRP-conjugated goat anti-mouse IgG or IgM alone. To detect endogenous peroxidase activity, other sections were incubated with DAB-H,O, solution alone. To further confirm the specificity of the staining, we also carried out pre-absorption and competition experiments. The antibody (10 pg/ml) was preincubated with excess human milk GalTase (50 pg/ml, Sigma, St. Louis, MO) for 24 h a t 4°C and then applied to the sections. Lectin Staining for Light Microscopic Observation Specimens were fixed overnight with Bouin’s solution and embedded in paraffin. Four-micrometer sections were cut, deparaffinized with xylene, and dehydrated through a graded series of ethanols. Lectin binding sites were visualized with biotinylated lectin (s-WGA: succinylated wheat germ agglutinin, Vector Lab, Inc., Burlingame, CA) and avidin-biotin-peroxidase complex (ABC) reagent (Vector Lab Inc., Burlingame, CA). In this procedure the sections were first treated for 10 min with 1% BSA in PBS and then incubated with biotinylated lectin at a concentration of 25 pg/ml in 0.1% BSA-PBS for 30 min at room temperature. After a wash in PBS, the sections were incubated with ABC reagent for 30 rnin at room temperature, washed again with PBS, and then immersed in DAB-H202for 5 rnin at room temperature. Finally the sections were rinsed in water, dehydrated, cleared, and mounted. To confirm the specificity of lectin staining, some sections were preincubated with an appropriate hapten sugar (N-acetylglucosamine) at a concentration of 0.2 M for 30 min and then incubated with biotinylated lectin in the presence of 0.2 M hapten sugar. To detect endogenous peroxidase activity, other sections were incubated with DAB-H2O2 solution alone. Nonspecific binding of ABC reagent was also checked by incubation, first with ABC reagent, and then with DAB-H,02 solution. lmmunostaining for Electron Microscopy Specimens were fixed in 4% formaldehyde in 0.1 M phosphate buffer (pH 7.3) for 1h at 4°C. After washing, the specimens were frozen and cut into 40 pm slices with a cryomicrotome. Slices were incubated overnight at 4°C with the antibody, washed with PBS, and incubated for 1 h at room temperature with HRP-conju- 111 gated goat anti-mouse IgG or IgM. For a cytochemical control, normal mouse IgG or IgM was used. The specimens were washed with PBS again and fixed in 2.5% glutaraldehyde in 0.1 M phosphate buffer, pH 7.4, for 10 rnin to obtain the better preservation of the ultrastructure and to immobilize the HRP-conjugates. After another washing with PBS, the specimens were immersed in DAB-H,02 for 5 min at room temperature, and then washed with distilled water. Finally the specimens were osmicated for 30 min, dehydrated through a graded series of ethanols, and embedded in Epon 812. Ultrathin sections were cut, stained with lead citrate, and observed in a JEM-1200EX electron microscope. Nonspecific staining was checked by the incubation with the secondary antibody-HRP conjugate alone. Some skin explants were fixed in 2.5%glutaraldehyde and processed for conventional electron microscopic observation. nzwc rs Western Blot Analysis Using the Antibody In order to characterize the protein reactive to human GalTase antibody, 17-day-old chick embryonic skin homogenate solubilized with Triton-X 100 was subjected to SDS-PAGE and then immunoblotted (Fig. 1). The immunoblot result showed that the antiGalTase antibody detected a 68-kDa protein (Fig. la), which is in good agreement with the molecular weight of chick embryonic liver GalTase reported previously (Furukawa and Roth, 1985; Hathaway and Shur, 1992). No protein band was detected with preimmune mouse IgG or IgM instead of the specific antibody (Fig. lb). lmmunoprecipitationof the Protein With Monoclonal Antibody Since the monoclonal antibody used in the present study was raised against human GalTase (Uemura et al., 19921, whether or not the protein reactive to the antibody contains the GalTase activity was investigated. Solubilized liver homogenate prepared from chick embryos was incubated with the antibody, and GalTase activity was assayed using the immunoprecipitated material. Only a small amount of the radioactivity was incorporated into the asialo-agalactotransferrin acceptor (data not shown). Tritium galactosylated transferrin glycopeptides were isolated by ConA-Sepharose column chromatography after exhaustive digestion of reaction mixtures with pronase. Tritium-labeled galactose was released by digestion of the glycopeptides with diplococcal p-galactosidase, which cleaves the Galpl-4GlcNAc but not the Galpl3GlcNAc linkage (Glasgow e t al., 1977), indicating that galactose was transferred to N-acetylglucosamine residues via pl-4 linkage (data not shown). Since the antibody is specific to human GalTase and since the immunoprecipitated enzyme activity of chick embryo is low with this antibody, whether or not the protein reactive to the antibody is chicken GalTase has to be further determined. Light Microscopic Observation With the Antibody In skin specimens obtained from 13-day-old chick embryos, immunostaining was observed at the apical surface of the epidermis and, a t a very low level, in the Y. AKIMOTO ET AL. 112 not stained (Fig. 3b,c). The epidermal-dermal junctional zone was scarcely stained. In the mucous metaplastic epidermis, the cytoplasm of the superficial cells was intensely stained, while intermediate and basal cells were not stained (Fig. 3d). The lectin binding was completely inhibited by the addition of the haptenic sugar, N-acetylglucosamine (Fig. 3e). Neither endogenous peroxidase activity nor other non-specific binding were observed. 116K97K66K45K- Electron Microscopic Observation With the Antibody 29K- a b Fig. 1. Immunoblotting of skin lysates with the use of anti-GalTase antibody (MAb 8513). Seventeen-day-old chick embryonic skin explants were lysed, electrophoresed on 12.5%SDS polyacrylamide gel under reducing conditions, and transferred to nitrocellulose. Proteins were immunostained with the antibody. A single 68-kDa band is detected (a).No staining was observed with normal mouse IgG or IgM (b). cell surface of the epidermis (Fig. 2a). As the embryo developed, the staining became more intense; and the location of the protein was more clearly detected on the cell surface. In the epidermis keratinized both in vivo (17-day-old chick embryo) and in vitro, intense staining was observed along the cell surface of cells in the superficial and intermediate layers (Fig. 2b,c), while staining was weak in keratinized and the basal cell layers. The epidermal-dermal junctional zone also showed moderate staining. A positive staining was observed in the superficial cells of the epidermis in which keratinization had been suppressed and mucous metaplasia had been induced by vitamin A (Fig. 2d). When the antibodies were omitted or replaced with the preimmune mouse IgG or IgM, no positive staining was observed. All the positive staining was disappeared when 17-day-old chick embryonic skin was incubated with the antibody in the presence of human milk GalTase (Fig. 2e). The precise localization of the protein reactive to the antibody was studied by electron microscope. In the 13-day-old chick embryo, the epidermis consisted of 3-5 cell layers and was not keratinized yet (Fig. 4a). The cell surface of intermediate cells were stained weakly (Fig. 4b). In contrast, in the 17-day-old chick embryonic epidermis, in which several layers of the superficial cells were keratinized (Fig. 5a), staining intensity was increased significantly. In the superficial and intermediate cell layers, intense staining was observed along the cell surface and in the intercellular space of the desmosomes (Fig. 5b,c). The cell surface of the basal cell layer was stained at its basal side as well as the basement membrane (Fig. 5d). The staining pattern in the epidermis in which keratinization had been induced in vitro by hydrocortisone was the same as that of the 17-day-old embryonic epidermis in vivo (Fig. 6a,b). In the epidermis in which mucous metaplasia had been induced by vitamin A, many mucous granules were observed in the cytoplasm of cells in the superficial layer (Fig. 6c). Positive staining was detected on the limiting membrane of the mucous granules, in the Golgi apparatus, and on a part of the nuclear membrane (Fig. 6d). The limiting membrane of mucous granules was more intensely stained by MAb 8513 than MAb 8628. In the intermediate layer, staining was scarcely observed or the cell surface or in the intercellular space of the desmosomes. Positive staining was not observed when incubated with preimmune mouse IgG or IgM instead of the specific antibody (Fig. 7a,b). Neither endogenous peroxidase activity nor non-specific binding of the secondary antibody was observed. DISCUSSION GalTase is mainly localized in Golgi apparatus where it is involved in the synthesis of complex carbohydrates by concerted actions of many other glycosyltransferases. GalTase is also found on cell surfaces of a variety of cells, and implicated to be involved in cell to cell and cell to matrix interactions (Pierce et al., 1980). Light Microscopic Observation With s- WGA In fact, the cell surface GalTase is shown to promote In order t o elucidate the localization of glycoconju- cell adhesion as well as cell migration (Shur, 1989; gates which contain N-acetylglucosamine residues at Hathaway and Shur, 1992). GalTase in the plasma the non-reducing termini, lectin staining was con- membrane of mouse sperm binds to polylactosaminoducted using s-WGA which specifically recognizes a glycans of the zona pellucida glycoproteins (Lopez et terminal N-acetylglucosamine residue. Staining was al., 1989; Macek et al., 1991; Miller et al., 1991). In the observed at the apical surface of the epidermis and, at early stage of mouse embryonic development, cell sura very low level, along the cell surface of the epidermis face GalTase is involved in compaction and the adheof 13-day-oldchick embryo (Fig. 3a). In the skin kera- sion between blastomeres (Bayna et al., 1988; Hathatinized both in vivo and in vitro, the superficial and way et al., 1989). Binding between cell surface GalTase intermediate cell layers were positively stained (Fig. and polylactosaminoglycans is considered to be in3b,c). However, both keratinized and basal cells were volved in the implantation of embryos to the uterus LOCALIZATION OF THE PROTEIN REACTIVE TO GALTASE ANTIBODY Fig. 2. Immunohistochemical localization by light microscopy of the protein reactive to the antibody in chick embryonic skin from 13- (a) and 17-day-old (b)embryos, and in the skin of 13-day-old embryos cultured in the presence of hydrocortisone for 9 days ( c ) or hydrocortisone and vitamin A for 1 day and then Bt,cAMP for 8 days (d). Stained with mouse anti-GalTase antibody and with HRP-conjugated goat anti-mouse IgG or IgM. Arrowheads indicate the basal surface of the epidermis. P, periderm. (a) In the undifferentiated skin from a 13-day-old embryo, positive staining is observed a t a low level. In 17-day-old (b) and hydrocortisone-treated (c) chick embryonic epider- 113 mis, in which keratinization had taken place, intense staining is observed mainly in the intermediate layers. Positive staining is also seen in the epidermal-dermal junction. d: Positive staining is found in the superficial cells of the mucous metaplastic epidermis induced by vitamin A treatment in vitro. In the superficial cells of the epidermis, patchy staining (arrows) is observed in the cytoplasm. e: Seventeenday-old chick embryonic skin was preincubated with antigen, then incubated with anti-GalTase antibody in the presence of antigen. Positive staining is noticeably diminished. x 900. 114 Y. AKIMOTO ET AL. Fig. 3. Staining of 13- and 17-day-old chick embryonic skin (a,b) and cultured skin explants (c, d) with s-WGA. a: Thirteen-day-old embryonic skin. The epidermal cells are scarcely stained. b: Seventeen-day-old embryonic skin. The superficial and intermediate cells are intensely stained. c: Skin that was keratinized in vitro by treatment with hydrocortisone for 9 days. The cell membrane of intermediate cells is intensely positive for the s-WGA reaction. d Mucous metaplastic skin that was induced in vitro by treatment with hydrocortisone and vitamin A for 1 day and then with Bt,cAMP for 8 days. The cytoplasm of superficial cells is positively stained (arrows). e: A cytochemical control. Seventeen-day-old embryonic skin was incubated with s-WGA in the presence of 0.2 M N-acetylglucosamine. Positive reaction is completely inhibited. Arrowheads indicate the basal surface of the epidermis. P, periderm. x 800. (Dutt et al., 1987) and in cell adhesion of embryonal carcinoma (Shur, 1983). In the present study, the localization of the protein reactive to human GalTase antibody was found in the Golgi apparatus and on the cell surface of chick embryonic skin. Changes in the cell surface localization of the LOCALIZATION OF THE PROTEIN REACTIVE TO GALTASE ANTIBODY 115 Fig. 4.Electron micrographs of epidermis from the tarsometatarsal region of 13-day-old chick embryos. a: Osmium-fixed and Epon-embedded section of the epidermis. The epidermis consists of 3-5 cell layers and it not keratinized yet. S, superficial cells; I, intermediate cells; B, basal cells; P, peridermal cells, and D, dermis. b: Immunocytochemical localization by electron microscopy of the protein in the intermediate cells. Formaldehyde-fixed and frozen sections were reacted with mouse anti-GalTase antibody, then, with HRP-conjugated goat anti-mouse IgG or IgM, and processed for transmission electron microscopy. Plasma membrane of intermediate cells is stained weakly (arrows). a, X 1,600; b, X 24,000. protein were observed immunocytochemically in the chick embryonic tarsometatarsal skin in association with the differentiation and metaplasia both in vivo and in vitro. The localization of the protein in the keratinized epidermis was the same as that of the endogenous P-galactoside-binding lectins previously reported (Akimoto et al., 1992, 1993). The protein was also detected in the limiting membrane of mucous granules in chick embryonic mucous metaplastic epidermis. In general, changes in carbohydrate structures of glycoconjugates have important roles in embryonic development and differentiation (Kawai et al., 1979; Barnes, 1988). These changes have been also observed in developing chick embryonic epidermis (Takata and Hirano, 1983). In the undifferentiated epidermis of the 13-day-old embryo, glycoconjugates terminating with N-acetylglucosamine, galactose, and N-acetylgalactosamine residues are not detected by cytochemical studies. As the epidermis develops toward keratinization, however, these glycoconjugates accumulate on the cell surface of the intermediate cells. The expression of these glycoconjugateschronologically correspond to the appearance of the protein reactive to human GalTase and P-galactoside-binding lectin on the cell surface. The protein is also present in desmosomes which are well developed in the keratinized epidermis (Takata et al., 1981). In the desmosomes which are not well developed in the mucous metaplastic epidermis (Hirano et al., 1985; Obinata et al., 1991), glycoconjugates containing terminal N-acetylglucosamine residues do not appear on the cell surface, and the protein is hardly detected on the cell surface or in the intercellular space of desmosomes. These results indicate that in chick embryonic skin the antibody reactive protein and glycoconjugates terminating N-acetylglucosamine residues might play important roles in the skin differentiation probably by mediating specific cellular interactions. The protein was also located in the epidermal-dermal junctional region. However, s-WGA staining was scarcely observed in the epidermal-dermal junctional region. This might be due to the masking of terminal N-acetylglucosamine residues with other sugar residues at this region. Immunostaining of the protein in the superficial cells of the epidermis was increased in association with the mucous metaplastic changes in vitro. Immunoelectron microscopic observation revealed that the protein was located in the limiting membrane of the mucous granules in addition to the Golgi apparatus in the mucous metaplastic epidermis. Positive staining was also observed on the nuclear membranes of the metaplastic epidermis cultured in the presence of vitamin A. This was only detected in the cultured epidermis. The localization of GalTase in the nuclear membranes was already reported in cultured F9 embryonal cells (Suganuma et al., 1991). Al- 116 Y. AKIMOTO ET AL. Fig. 5. Electron micrographs of epidermis from the tarsometatarsal region of 17-day-old chick embryos. a: Osmium-fixed and Epon-embedded sections of the epidermis. Several layers of the superficial cells (S) are keratinized. Peridermal cells (P) are above the keratinized layer (K). I, intermediate cells; B, basal cells; and D, dermis. b-d. Immunocytochemical localization by electron microscopy of the protein in the epidermis. Formaldehyde-fixed and frozen sections were reacted with mouse anti-GalTase antibody, then, reacted with HRP- conjugated goat anti-mouse IgG or IgM, and processed for transmission electron microscopy. In the superficial (b) and intermediate (c) layers, the plasma membrane (arrows in b) and intercellular spaces of the desmosomes (arrows in c) are stained. d: At the boundary between the epidermis and the dermis, the basal plasma membrane of the basal cells, the basal lamina (arrows), and immediate environs (arrowheads) are stained. B, basal cell; CF, collagen fibers; N, nucleus. a , x 1,700; b, ~ 2 5 , 0 0 0c, ; ~ 2 0 , 0 0 0d, ; ~31,000. LOCALIZATION O F THE PROTEIN REACTIVE TO GALTASE ANTIBODY Fig. 6. Electron micrographs of epidermis from the tarsometatarsal region of chick embryonic epidermis cultured in the presence of hydrocortisone for 9 days (a and b), or hydrocortisone and vitamin A for 1 day and then Bt,cAMP for 8 days (c and d). Immunocytochemical localization by electron microscopy of the protein (b and d). Formaldehyde-fixed and frozen sections were reacted with mouse antiGalTase antibody and then reacted with HRP-conjugated goat antimouse IgG or IgM, and processed for transmission electron microscopy. a: Osmium-fixed and Epon-embedded section of the keratinized epidermis. Epidermis consists of superficial (S), intermediate (I), and basal (B) layers. Keratinized cells (K) are seen in the superficial layer of the epidermis. D, dermis. b: In the intermediate layer of 117 the epidermis, in which keratinization was induced by hydrocortisone, the plasma membrane (arrows) is intensely stained. c: Osmiumfixed and Epon-embeddedsection of the epidermis. Epidermis consists of superficial (S), intermediate (I), and basal (B) layers. No keratinized layer is seen in the epidermis. D, dermis. d: Positive staining is detected in the limiting membrane of the mucous granules (MI, in the Golgi complex (G), and in part of the nuclear membranes (arrowheads) of the superficial cells of the mucous metaplastic epidermis. Desmosomes (DE) and the cell surface are scarcely stained. a, x 1,800; b, x 9,300; c, x 1,700; d, x 14,000. Inset Staining is positive in the Golgi apparatus ( G )as indicated by arrows. x 27,500. Y. AKIMOTO ET AL. 118 Fig. 7. Electron micrographs of epidermis from the tarsometatarsal region of 17-day-old chick embryos. Cytochemical control. The specimens were incubated with normal mouse IgG instead of the specific antibody. No positive staining is observed in the desmosomes (arrows) and cell surface of the intermediate cells (a),and in the basal plasma membrane of epidermal basal cells and the basement membrane (b). B, basal cell; BL, basal lamina. a, x 31,000; b, x 30,000. though glycoproteins having an N-acetylglucosamine residue are shown to exist in nuclear membranes (Holt and Hart, 1986), the functions of the GalTase and glycoconjugates in the nuclear membrane remain to be elucidated. The present study suggests that the protein reactive to the human GalTase antibody is a useful differentiation marker for chick embryonic skin. Although the antibody is always reactive to the Golgi-apparatus, which is a primary location of the GalTase, it is important to determine whether the proteins reactive to the antibody a t the various locations in the chick skin are really a GalTase and whether the protein at the cell surface possesses a binding activity to a n N-acetylglucosamine residue of glycoconjugates. Currently, studies are going to elucidate the above points, and the results to be obtained will provide a functional aspect of the protein during the chick skin differentiation. ACKNOWLEDGMENTS The authors wish to express their appreciation to Dr. H. Kawakami (Department of Anatomy, Kyorin University School of Medicine) and Dr. N. Susumu (Department of Obstetrics and Gynecology, Keio University School of Medicine) for their invaluable discussion, to Dr. M. Uemura (Konica Corporation) for his kind gift of monoclonal antibody against GalTase and also to Mr. M. Fukuda, Ms. S. Matsubara, Ms. C . Kamata, Ms. M. Kanai, and Ms. T. Shibata (Kyorin University School of Medicine) for their technical assistance. This study was supported in part by grants-in-aid from the Ministry of Education, Science, and Culture of Japan. LITERATURE CITED Akimoto, Y., A. Obinata, H. Endo, and H. Hirano 1988 Epidermal growth factor (EGFI-induced morphological changes in the basement membrane of chick embryonic skin: An electron-microscopic study. Cell Tissue Res., 254:481-485. Akimoto, Y., H. Kawakami, Y. Oda, A. Obinata, H. Endo, K. Kasai, and H. Hirano 1992 Changes in expression of the endogenous p-galactoside-binding 14-kDa lectin of chick embryonic skin during epidermal differentiation. Exp. Cell Res., 199~297-304. Akimoto, Y., A. Obinata, J . Hirabayashi, Y. Sakakura, H. Endo, K. Kasai, and H. Hirano 1993 Secretion of endogenous 16-kDa p-galactoside-binding lectin from vitamin A-pretreated chick embryonic cultured skin. Exp. Cell Res., 205:251-261. Aoki, D., H.E. Appert, D. Johnson, S.S. Wong, and M.N. Fukuda 1990 Analysis of the substrate binding sites of human galactosyltransferase by protein engineering. EMBO J., 9:3171-3178. Aoki, D., N. Lee, N. Yamaguchi, C. Dubois, and M.N. Fukuda 1992 Golgi retention of trans-Golgi membrane protein, galactosyltransferase, requires cysteine and histidine residues within the membrane-anchoring domain. Proc. Natl. Acad. Sci. U.S.A., 89: 4319-4323. Barnes, D.M. 1988 Orchestrating the sperm-egg summit. Science, 239:1091-1092. Bayna, E.M., J.H. Shaper, and B.D. Shur 1988 Temporally specific involvement of cell surface pl,4-galactosyltransferaseduring mouse embryo morula compaction. Cell, 53t145-157. Biggers, J.D., R.B.L. Gwatkin, and S. Heyner 1961 Growth of embryonic avian and mammalian tibiae on a relatively simple chemically defined medium. Exp. Cell Res., 25:41-58. Cooke, S.V. and B.D. Shur 1994 Cell surface pl,4-galactosyltransferase: expression and function. Dev. Growth Diff., 36:125-132. LOCALIZATION OF THE PROTEIN REACTIVE TO GALTASE ANTIBODY Dutt, A., J.-P. Tang, and D.D. Carson 1987 Lactosaminoglycans are involved in uterine epithelial cell adhesion in vitro. Dev. Biol., 119:27-37. Eckstein, D.J. and B.D. Shur 1992 Cell surface p-1,4-galactosyltransferase is associated with the detergent-insoluble cytoskeleton on migrating mesenchymal cells. Exp. Cell Res., 201:83-90. Furukawa, K., K. Matsuta, F. Takeuchi, E. Kosuge, T. Miyamoto, and A. Kobata 1990 Kinetic study of a galactosyltransferase in the B cells of patients with rheumatoid arthritis. Internat. Immunol., 2:105-112. Furukawa, K. and S. Roth 1985 Co-purification of galactosyltransferases from chick-embryo liver. Biochem. J., 227r573-582. Glasgow, L.R., J.C. Paulson, and R.L. Hill 1977 Systematic purification of five glycosidases from Streptococcus (Diplococcuus) pneumoniue. J . Biol. Chem., 252:8615-8623. Hamburger, V. and H.L. Hamilton 1951 A series of normal stages in the development of the chick embryo. J . Morphol., 88:49-92. Hartel-Schnenk, S., N. Minnifield, W. Reutter, C. Hanski, C. Bauer, and D.J. Morre 1991 Distribution of glycosyltransferases among Golgi apparatus subfractions from liver and hepatomas of the rat. Biochim. Biophys. Acta, 1115:108-122. Hathaway, H.J., L.C. Romagnano, and B.S. Babiarz 1989 Analysis of cell surface galactosyltransferase activity during mouse trophectodermal differentiation. Dev. Biol., 134:351-361. Hathaway, H.J. and B.D. Shur 1992 Cell surface pl,4-galactosyltransferase functions during neural crest cell migration and neurulation in vivo. J . Cell Biol., 117:369-382. Hirano, H., Y. Akimoto, H. Kawakami, Y. Oda, and K. Kasai 1988 Localization of endogenous p-galactoside-binding lectin in embryonic chick epidermis. In: Glycoconjugates in Medicine. Ohvama M. Muramatsu T. eds. Professional Posteraduate Servicis, Tokyo, pp. 8-13. ’ Hirano, H., N. Oishi, A. Obinata, and H. Endo 1985 Metaplastic changes in chick embryonic skin induced by vitamin A. In: Basement Membranes. Shibata s,ed. Elsevier, Amsterdam, pp. 395406. Holt, G.D. and G.W. Hart 1986 The subcellular distribution of terminal N-acetylglucosamine moieties. J . Biol. Chem., 261:80498057. Kawai, N., F. Nishiyama, and H. Hirano 1979 Changes of lectinbinding sites on the embryonic muscle cell surface in the developing ascidian, Hulocynthiu uuruntium. Exp. Cell Res., 122:293304. Kawano, J., S. Ide, T. Oinuma, and T. Suganuma 1994 A proteinspecific monoclonal antibody to rat liver pl-4 galactosyltransferase and its application to immunocytochemistry. J . Histochem. Cytochem., 42~363 -369. Lasky, L.A. 1992 Selectins: Interpreters of cell-specific carbohydrate information during inflammation. Science, 258:964-969. Lopez, L.C., C.M. Maillet, K. Oleszkowicz, and B.D. Shur 1989 Cell surface and golgi pools of ~-1,4-galactosyltransferase are differentially regulated during embryonal carcinoma cell differentiation. Mol. Cell Biol., 9:2370-2377. Macek, M.B., L.C. Lopez, and B.D. Shur 1991 Aggregation of p-1,4galactosyltransferase on mouse sperm induces the acrosome reaction. Dev. Biol., 147:440-444. Miller, D.J., M.B. Macek, and B.D. Shur 1991 Complementarity between sperm surface p-1,4- galactosyltransferase and egg-coat ZP3 mediates sperm-egg binding. Nature, 357:589-593. Obinata, A,, Y. Akimoto, H. Hirano, and H. Endo 1991 Stimulation by Bt,cAMP of epidermal mucous metaplasia in retinol-pretreated chick embryonic cultured skin, and its inhibition by herbimycin A, an inhibitor for protein-tyrosine kinase. Exp. Cell Res., 193: 36-44. Oda, Y., Y. Ohyama, A. Obinata, H. Endo, and K. Kasai 1989 Endog- - 119 enous p-galactoside-binding lectin expression is suppressed in retinol-induced mucous metaplasia of chick embryonic epidermis. Exp. Cell Res., 182:33-43. Pestalozzi, D.M., M. Hess, and E.G. Berger 1982 Immunohistochemical evidence for cell surface and Golgi localization of galactosyltransferase in human stomach, jejunum, liver, and pancreas. J . Histochem. Cytochem., 30:1146-1152. Pierce, M., E.A. Turley, and S. Roth 1980 Cell surface glycosyltransferase activities. Int. Rev. Cytol., 65:l-47. Romagnano, L. and B. Babian 1990 The role of murine cell surface galactosyltransferase in trophoblast: Laminin interactions in uztro. Dev. Biol., 241.254-261. Roseman, S. 1970 The synthesis of complex carbohydrate by multiglycosyltransferase systems and their potential function in intercellular adhesion. Chem. Phys. Lipids, 5:270-297. Roth, J . and E.G. Berger 1982 Immunocytochemical localization of galactosyltransferase in Hela cells: codistribution with thiamine pyrophosphatase in trans-Golgi cisternae. J . Cell Biol., 93:223229. Roth, J., M.J. Lentze, and E.G. Berger 1985 Immunocytochemical demonstration of ecto-galactosyltransferase in absorptive intestinal cells. J . Cell Biol., 100:118-125. Sharon, N. and H. Lis 1989 Lectins as cell recognition molecules. Science, 246:227-234. Shur, B.D. 1989 Expression and function of cell surface galactosyltransferase. Biochem. Biophys. Acta 988:389-409. Shur, B.D. 1991 Mini Review: Cell surface p1,4 galactosyltransferase: Twenty years later. Glycobiology, 1:563-575. Shur, B.D. 1983 Embryonal carcinoma cell adhesion: The role of surface galactosyltransferase and its 90K lactosaminoglycan substrate. Dev. Biol., 99:360-372. Suganuma, T., H. Muramatsu, T. Muramatsu, K. Ihida, J. Kawano, and F. Murata 1991 Subcellular localization of N-acetylglucosaminide p1+4 galactosyltransferase revealed by immunoelectron microscopy. J . Histochem. Cytochem., 39:299-309. Sugimoto, M., K. Tajima, A. Kojima, and H. Endo 1974 Differential acceleration by hydrocortisone of the accumulation of epidermal structural proteins in the chick embryonic skin growing in a chemically defined medium. Dev. Biol., 39:295-307. Takata, K., A. Obinata, H. Endo, and H. Hirano 1981 Induction of the a-type keratinization by hydrocortisone in embryonic chick skins grown in a chemically defined medium: An electron microscopic study. Dev. Biol., 85:370-379. Takata, K. and H. Hirano 1983 Changes in soybean agglutinin (SBA) and peanut agglutinin (PNA) binding pattern in the epidermis of the developing chick embryo. Dev. Growth. Diff., 25r299-305. Takeichi, M. 1988 The cadherins: Cell-cell adhesion molecules controlling animal morphogenesis. Development, 102:639-655. Towbin, H., T. Staehelin, and J . Gordon 1979 Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: Procedure and some applications. Proc. Natl. Acad. Sci. U.S.A., 76: 4350-4354. Trayer, I.P. and R.L. Hill 1971 The purification and properties of the A protein of lactose synthetase. J . Biol. Chem., 246:6666-6675. Uejima, T., M. Uemura, S. Nozawa, and H. Narimatsu 1992 Complementary DNA cloning for galactosyltransferase associated with tumor and determination of antigenic epitopes recognized by specific monoclonal antibodies. Cancer Res., 52:6158-6163. Uemura, M., T. Sakaguchi, T. Uejima, S. Nozawa, and H. Narimatsu 1992 Mouse monoclonal antibodies which recognize a human (pl4)galactosyltransferase associated with tumor in body fluids. Cancer Res, 52:6153-6157. Watzele, G., R. Bachofner, and E.G. Berger 1991 Immunocytochemical localization of the Golgi apparatus using protein-specific antibodies to galactosyltransferase. Eur. J . Cell Biol., 56:451-458.