Stage-specific Changes in Lectin Binding to Mouse Cerebellar Cells in Vitro1

Eleven fluorescein isothiocyanate-conjugated (FITC) lectins, each with distinct carbohydratebinding properties, were used to assess cell surface glycoconjugates of embryonic and early postnatal cerebellar cells in vitro. Fluorescence staining of embryonic day 13 (E13) cerebellar cells with FITC Ricinus communis agglutinin diminished markedly between 24 and 72 hr in vitro. No staining of postnatal day 0 (PO) or postnatal day 7 (P7) cells was observed with FITC Ricinus communis agglutinin. A similar, but less pronounced decrease in FITC concanavalin A, FITC Lens culinaris, and FITC wheat germ agglutinin was observed between embryonic day 13 and birth. No specific staining of E13, PO, or P7 cultures was observed with FITC peanut agglutinin, FITC Dolichos bifloris agglutinin, FITC soybean agglutinin, FITC Wistaria floribundis agglutinin, FITC Phaseolus vulgaris agglutinin, FITC Limulus polyphemus agglutinin, or FITC Ulex europaeusl agglutinin. Similar results were obtained with 12”1-lectin binding assays. Ricinus communis ‘““I-agglutinin binding decreased dramatically between embryonic day 13 and birth. Less pronounced decreases were observed in 12”1-concanavalin A and wheat germ ‘251-agglutinin binding. Very low levels of soybean ‘251-agglutinin or Ulex europaeusl 1251-agglutinin were bound by either embryonic or early postnatal cerebellar cells in uitro. Changes in cell surface properties have been suggested to mediate specific cell contacts in the developing mammalian brain (Sidman, 1974; Sidman and Rakic, 1973). In the mouse cerebellum, lectins have been used to catalog alterations in carbohydrate-containing macromolecules (Hatten and Sidman, 1978; Hatten et al., 1979; Wood et al., 1974; Zanetta et al., 1978). Studies with anti-carbohydrate antibodies (Trenkner and Sarkar, 1978) and carbohydrate-specific toxins (Willinger and Schachner, 1980) also have revealed alterations in cerebellar glycoconjugates that are regulated developmentally. Histochemical studies of lectin binding to developing mouse cerebellum have demonstrated that some alterations detected by lectins are restricted to particular cell types (Hatten et al., 1979). Although light microscopic histochemical studies allow precise localization of lectin binding from the cellular geometry of the tissue, lectin binding is not restricted to the cell surface. One way to I We thank A. M. Francois for her expert technical assistance and Drs. R. K. H. Liem and F. R. Maxfield for their advice and criticism throughout this study. This work was supported by National Institutes of Health Grant NS 15429 (M. E. H.). S. Huck was supported by a Max Kade Postdoctoral Research Exchange grant. ’ Present address: Institut fur Neuropharmakologie, Universitat Wien, Wahringerstrasse 13a, A-1090 Wein, Austria. ’ To whom correspondence should be addressed at Department of Pharmacology, New York University School of Medicine, 550 First Avenue, New York, NY 10016. visualize lectin labeling of the cell surface with light microscopy is to label dissociated cells maintained as a monolayer in vitro with fluorescein isothiocyanate-conjugated (FITC)4 lectins. In the present study, 11 FITC lectins, each with a different carbohydrate-binding specificity, have been used to define the lectin binding properties of cerebellar cells dissociated at embryonic day 13 and postnatal days 0 and 7 and maintained in microcultures. The results with FITC lectins have been compared to studies of 1251lectin binding to dissociated cerebellar cells in vitro. Materials and Methods Cerebellar cell cultures. All experiments were carried out with tissue from C57B1/6J mice derived from a breeding colony in this department. Females were ’ The abbreviations used are: Con A, concanavalin A, DBA, Dolichos bifloris agglutinin; EO, embryonic day 0; E13, embryonic day 13; FITC, fluorescein isothiocyanate-conjugated; L-FUC, L-fucose; D-Gal, n-galactose; n-GalNAc, N-acetyl-n-galactosamine; D-Glc, n-glucose; DGlcNAc, N-acetyl-n-glucosamine; LCA, Lens culinaris A agglutinin; LPA, Limulus polyphemus agglutinin; D-Man, n-mannose; Me-oc-nMan, methyl-a-n-mannose; PO, postnatal day 0; P7, postnatal day 7; PBS, phosphate-buffered saline; PHA, Phaseolus vulgaris agglutinin; Plys, poly-n-lysine; PNA, peanut agglutinin; RCA,, Ricinus communis agglutinin; RITC, rhodamine isothiocyanate-conjugated, SBA, soybean agglutinin; UEAr, Ulex europaeusr agglutinin; WFA, Wistaria floribundis agglutinin; WGA, wheat germ agglutinin. 1076 Huck and Hatten Vol. 1, No. 10, Oct. 1981 checked daily for the presence of vaginal plugs, with the day of impregnation being designated EO. Cerebellar tissue was removed at embryonic day 13 (E13) or postnatal days 0 (PO, birth) or 7 (P7). Cells were dissociated from whole cerebellum as described (Hatten and Sidman, 1978; Hatten, 1981) and plated at a final cell density of 0.8 to 2 X lo6 cells/ml on glass coverslips treated with poly-n-lysine (Plys, 25 pg/ ml) in microcultures (Hatten, 1981). In brief, culture dishes were prepared by drilling a 4-mm hole in 50-mm Petri dishes (Falcon, No. 1006) and attaching a glass coverslip (20 mm diameter, Arthur H. Thomas Co., Philadelphia, PA) as a false bottom with a mixture of Vaseline and Tissue-Tek (Fisher Scientific, 3:l). Plates were sterilized by exposure to shortwave ultraviolet irradiation (30 to 60 min, 2O”C), washed three times with Hz0 (2O”C), air dried, and treated with Plys (25 pg/ml) as described (Hatten and Sidman, 1978). In total, cultures were prepared from cerebellar tissue from 14 litters of El3 embryos and more than 60 postnatal animals. More than 750 microcultures were assayed for FITC lectin binding. More than 5800 microwell cultures were assayed for ““I-lectin binding studies. Histochemical procedures: Binding of FITC lectins. The binding of FITC concanavalin A (FITC-Con A), FITC Dolichos bifloris agglutinin (FITC-DBA), FITC Lens culinaris A agglutinin (FITC-LCA), FITC Limulus polyphemus agglutinin (FITC-LPA), FITC peanut agglutinin (FITC-PNA), FITC Phaseolus vulgaris agglutinin (FITC-PHA), FITC Ricinus communis agglutinin (120,000 daltons, FITC-RCAi), FITC soybean agglutinin (FITC-SBA), FITC UZex europaeua agglutinin (FITCUEAi), FITC wheat germ agglutinin (FITC-WGA), and FITC Wistaria floribundis agglutinin (FITC-WFA) to dissociated E13, PO, and P7 cells was assayed in microcultures. Cultures were washed with phosphate-buffered saline (PBS, 0.2 M phosphate, pH 7.2) three times (5 min, 4°C or 20°C) and FITC lectin was added (10 to 500 pg/rnl in PBS, 15 min, 4°C or 20°C). FITC lectin was removed and the cultures were washed three times with PBS (5 min, 20°C). In some experiments, FITC lectin solutions were prepared with hapten carbohydrate. The haptens were as follows: for Con A, methyl-a-D-mannose (50 mM); for LCA, D-Glc (50 mM); for WGA, N,N’-diacetyl chitobiose (50 mM); for RCAi, lactose (50 mu); for PNA, DGal (50 mrvr); for SBA, DBA, WFA, and PHA, rr-GalNAc (50 mM); for UEAi, L-FUC (50 mM); and for LPA, sialic acid (50 mM). For FITC-LPA, all procedures were carried out in TrisHCl (pH 7.4). FITC lectins were purchased from E. Y. Laboratories (San Mateo, CA). We confirmed the carbohydrate specificity of each lectin by hemagglutination (Goldstein and Hayes, 1978). The ratio of OD495/OD280 for the lectins were as follows: Con A, 1.25; WGA, 1.24; RCAi, 1.6; PNA, 0.8; SBA, 0.93; UEAi, 1.35; DBA, 1.05; PHA, 1.21; and LPA, 1.08. As a positive control for FITC lectin staining of dissociated cerebellar cells in microcultures, erythrocytes from different species were labeled with FITC lectins by a modification of the procedure of Burger (1976). In brief, washed, packed, fresh cells were resuspended at 2 x 10” cells/ml in PBS (4°C). The suspension (90 ~1) was transferred to a porcelain plate and 0.01 ml of FITC lectin solution of varying concentrations (10 to 1000 pg/ml) was added. The suspension was aspirated gently, and after 1 min, 20 ~1 was transferred as a hanging drop onto a microscope slide. After 4 to 15 min, FITC lectin binding was visualized by fluorescence microscopy. As a control for carbohydrate specificity, FITC lectin solutions with hapten carbohydrate (0.1 M) were used. Rhodamine isothiocyanate-conjugated (RITC) lectins also were purchased from E. Y. Laboratories and used for some experiments. Methyl-a-n-mannose, D-ghCOSe, lactose, D-galactose, N-acetyl-D-glucosamine, and sialic acid were purchased from Sigma (St. Louis, MO). N,N’Diacetyl chitobiose and chitin hydrolysate were prepared by the method of Rupley (1964). ‘251-labeling of lectins. Con A, WGA, RCAi, SBA, and UEAi were iodinated by the chloramine T method (Freeman, 1967). Lectins were preincubated with hapten carbohydrate (100 mM) to protect the carbohydrate binding site. Briefly, 1 mCi (10 ~1 of carrier-free iodine in NaOH, New England Nuclear, Boston, MA) of ‘““I-sodium was added to lectin (50 ~1 of 1 mg/ml of solution containing 100 mM hapten carbohydrate, pH 7.2). Chloramine T (10 ~1 of 3.5 mM solution in PBS) was added, and after 20 set, the reaction was stopped by the addition of sodium metabisulfite (10 ~1 of 3.5 mM solution in PBS). Labeled lectins were applied to a Sephadex G-50 (Pharmacia, Piscataway, NJ) column (bed volume, 8 ml) and separated from free iodine and carbohydrate by elution with PBS. Labeled Con A was eluted with PBS containing methyl-cw-n-mannose (50 mM) and dialyzed against PBS (4°C). The specific activities of 12”1-labeled lectins (given as counts per min per gm x 10m4) were: Con A, 12.1; WGA, 6.38; RCAi, 11.7; SBA, 8.14; and UEAi, 9.09. We assayed the carbohydrate-binding specificity of lZ51-labeled lectins by hemagglutination (Goldstein and Hayes, 1978). All lectins were purchased from E. Y. Laboratories. Chloramine T was purchased from Eastman (Rochester, NY) and sodium metabisulfite was obtained from Sigma. Binding of ‘2”I-lectins to cerebellar cells. For ““I-labeling experiments, E13, PO, and P7 cerebellar cells were plated at 2.5 x lo6 cells/ml in flexible Microtest IT plates (Cooke Engineering, Alexandria, VA). Cell viability, as measured by exclusion of the dye trypan blue (O.l%, GIBCO), was greater than 85% for the cell cultures used for lz51-lectin binding assays. After 24 hr in vitro, cultures were rinsed three times with PBS (5 min, 4°C) and 1251lectin was added (50 ~1 of lectin, 10 to 100 pug/ml final concentration, 15 min, 4°C). The wells were rinsed three times with PBS (5 min, 4”C), after which, individual wells were cut out and transferred to a 12 x 75 mm tube (Falcon, No. 2054). Radioactivity was measured with a Beckman Gamma 4000 counter. All assays were carried out at 4°C. Binding experiments were carried out in triplicate or in groups of six wells for each of the four lectin concentrations (10,25,50, and 100 pg/ml) tested. For each lectin concentration assayed, a paired well was labeled with 1251-lectin with hapten carbohydrate (50 mM). Specific The Journal of Neuroscience Stage-specific Changes in Lectin Binding to Cerebellar Cells 1077 binding was calculated by subtracting the counts in the presence of hapten carbohydrate from counts in the absence of hapten carbohydrate. As a positive control, erythrocytes from different species were labeled with ‘““I-lectins. Washed, packed, fresh erythrocytes were resuspended at 2 x lo6 cells/ml in PBS (1.0 ml). lz51-Lectin was added (50 ~1 of lectin, 10 to 100 pg/ml final concentration, 15 min, 4°C). The cells were washed three times by centrifugation (1000 rpm) and the final pellet was assayed for radioactivity with a Beckman Gamma 4000 counter. For each lectin concentration assayed, a paired sample was labeled with lz51lectin with hapten carbohydrate (50 mM) and specific binding was calculated as described. Cell lysis was assayed by measuring the absorption of the supernatant of washed cells at 520 nm. The results reported were from samples with absorptions less than 0.1.