11EVEI,OPMENTAL

BIOLOGY

138,

400-409

Granule

(1 %)o)

Cell Induction of 9-0-Acetyl Gangliosides on Cerebellar Glia in Microcultures

R~SALIAMENDEZ-OTERO*

AND

Accepted

MARTHACONSTANTINE-PATON?

December

8, 1989

In previous studies we have shown that the expression of acetylated gangliosides recognized by the JONES monoclonal antibody is correlated with regions of cell migration in the developing rat nervous system. In this study we have investigated the expression of these gangliosides in two different types of cultures prepared from dissociated postnatal rat cerebella. In the first type, cells are plated after dissociation under conditions where most of the glial cells develop a stellate morphology that anchors neurons but does not support their migration. In the second type of culture, cells are plated in a ratio of four neurons to one glial ccl1 and under these conditions the predominant form of astroglia is an elongate form that supports the migration of granule neurons. Granule neurons express JONES antigens in dissociated cell suspensions and in cultures in which cells are plated either after dissociation or in a 4:l neuron:glia ratio. On the other hand, glial cells grown in the absence of neurons are JONES negative. In addit.ion, the expression of JONES gangliosides by glial cells is different in the two types of culture. In cultures where the astroglial cells display the stellate morphology only a small proportion show JONES staining. Cultures in which the glial cells assume the elongate morphology have a significantly higher number of JONES-positive astroplia. Ccl 1990 Academic Press, Inc.

INTRODUCTION

(Constantine-Paton et al., 1986; Mendez-Otero et al., 1988; Schlosshauer et al., 1988). In the postnatal cerebellum, it is possible to correlate the appearance of the JONES antigens with specific waves of granule cell migration in the various folia (Mendez-Otero et al., 1988). The radially oriented pattern of JONES binding in these folia suggested that the gangliosides might be expressed by the radial glia which were serving as substrates for the migrating granule cells. However, the plasma membrane location of the JONES antigens made it difficult to definitely localize them to a particular cell type. In addition, in the intact system, it was impossible to unambiguously link the expression of the antigens to the migratory event or to determine whether their appearance on the cell membranes was autonomous or controlled by cell-cell interactions (Mendez-Otero et nl., 1988). To address these issues, we have investigated the expression of Jones antigen in vitro using two different types of microwell cultures. The first one is prepared under conditions in which most of the astroglial cells express a stellate morphology, a form that anchors neurons rather than supports their migration (Hatten and Liem, 1981). In the second type of culture, granule cells are separated from glia using a Percoll gradient and seeded in a proportion of four neurons to one glia (Hatten, 1985). In these cultures, a large proportion of the glial cells assume an elongated form which supports

Results of numerous investigations of cell motility and axonal sprouting have suggested that gangliosides serve an important function in cell substratum interactions and neurite extension (e.g., Stallcup et al., 1989; Cheresh et nL, 1986, 1987; Spirman et al., 1982; Schwartz and Spirman, 1982). Studies of these cell surface, sialic acid-bearing lipids have, in general, lagged behind those on similarly distributed proteins because of technical difficulties in their isolation and identification. Nevcrthelcss, monoclonal antibodies directed against the carbohydrate moieties of gangliosides have become increasingly important aids in dissecting the biological function of these molecules. Such probes have, for example, implicated disialogangliosides as important factors in modulating the adhesion between integrins and fibronectin substrates (e.g., Cheresh et al, 1986, 1987; Burns et al., 1988; Stallcup et al., 1989) and a number of laboratories have recently described ganglioside antigens that are spatially and temporally regulated in developing neural tissue (e.g., Goldman et al., 1984, Grunwald et al., 1985; Levine et al., 1984; Rosner et al., 1988; Willinger and Schachner, 1980). In our own earlier studies we described the monoclonal antibody JONES which recognizes a subset of acetylated gangliosides in developing rat brain. The acetylated JONES antigens are associated with cell migration and axon elongation throughout the brain OOlZ-1606/90 $3.00 Copyright ~11 rights

Q 1990 by Academic Press. Inc. of reproduction in any form reserved.

400

MENIIEZ-OTERO

ANI)

CONSTANTINE-PATON

extensive neuronal migration (Hatten and Liem, 1981, Edmondson and Hatten, 1987, Gregory et al., 1988). Our results support our early proposal of in. viva association of JONES gangliosides and neuron migration (Mendez-Otero et al., 1988). Granule cells in isolation and in both types of cultures express these antigens, but, in the absence of neurons, glial cells are JONESnegative. Moreover, most glial cells with an elongated morphology are also JONES-positive and only a small proportion of the stellate astroglia express these gangliosides. Some of the results presented here have been described previously in abstract form (Mendez-Otero and Constantine-Paton, 1987). MATERIALS

AND

METHODS

Cell cultures. Cerebellar cultures were prepared according to published procedures (Hatten and Liem, 1981, Hatten, 1985). Briefly, whole cerebellum was dissected from Long-Evans rats on the second to sixth postnatal days, washed in Dulbecco’s calcium and magnesium-free phosphate-buffered saline solution (PBS), treated with 0.25% trypsin and 0.05% deoxyribonuclease I (DNase; Sigma, St. Louis, MO) for 15 min at 37”C, and washed three times in PBS solution to remove trypsin. Basal Eagle’s medium (GIBCO, Grand Island, NY) and 0.05% DNase were added and cells were dissociated by repeated passage through a fire-polished Pasteur pipet. The resulting cell suspension was passed through a monofilament polyester screen (33 pm in mesh size), counted, and plated immediately or applied to a two-step Percoll (Pharmacia, Piscataway, NJ) gradient to prepare enriched populations of astroglia and granule cells (Hatten, 1985). Cells were resuspended in the following culture medium: 90% basal Eagle’s medium with Earl’s salts, 10% heat-inactivated horse serum, 4 mM glutamine, 8 mM glucose, 20 U/ml penicillin-streptomycin (GIBCO). Cells were plated at a cell density of 2 X 10” cells/ml on either glass coverslips or tissue culture eight-chamber slides (Lab-Tek, Nunc, Inc., Naperville, IL) pretreated with 50 pug/ml poly-Llysine (Sigma, St. Louis, MO) and incubated at 35.5”C in a humidified atmosphere. Inl~nunocyfochemistr:y. Cultures were washed with culture medium without serum and fixed with 4% paraformaldehyde in 0.1 M phosphate buffer (pH 7.4) for 15 min at room temperature, washed three times in PBS, and stained in a two-step indirect immunofluorescence procedure. The cells were first incubated for 15 min with 5% normal goat serum (NGS) and then incubated with a 1:500 dilution of JONES monoclonal antibody purified ascites IgM at 4”C, overnight. For double-immunofluorescence staining, some cultures were then incubated with tetanus toxin (1 pg/ml, a gift from

9-0-Acrtyl

401

Gay~/liosidc~s

Dr. E. M. Steven, Wellcome Research Laboratories, England) for 1 hr at room temperature, washed three times in PBS, and incubated with a rabbit antiserum to tetanus toxin (1:50) for 2 hr at room temperature. The cultures to be stained with anti-glial fibrillary acidic protein (GFAP), anti-vimentin, or Rat-401 (generously donated by Dr. Susan Hockfield, Yale University) were postfixed with 4% paraformaldehyde after the incubation with JONES antibody and permeabilized by treatment with 0.3% (v/v) Triton X-100 (Sigma) in PBS for 15 min. After washing, the cells were incubated with a polyclonal rabbit antiserum to GFAP (Dako Corp., Santa Barbara, CA), diluted 1:200, for 2 hr at room temperature, or to vimentin (Boehringer-Mannheim, Indianapolis, IN) at 1:2 dilution, or with Rat-401 supernatant. After washings, cultures were then incubated for 2 hr at room temperature with a mixture of the two appropriate secondary antibodies (rhodamine-conjugated goat anti-mouse IgM, Organon Teknika Corp., West Chester, PA, and fluorescein-conjugated goat anti-rabbit, Jackson Immunoresearch, or fluoresceinconjugated goat anti-mouse IgG) diluted 1:50 in 5%) goat serum. JONES-positive cells were then identified by rhodamine immunofluorescence and glial cells or neurons by fluorescein immunofluorescence. After washing, the cells were coverslipped with n-propyl gallate (Sigma) in 80%) (v/v) glycerol in 0.1 M phosphate buffer, pH 8.5, sealed with nail polish, and visualized using rhodamine and fluorescein epifluorescence and phase-contrast optics (Leitz and Zeiss). Cell suspensions were stained following the same steps. To evaluate the percentage of neurons and glia expressing JONES we counted 100-200 cells per coverslip in double-labeled cultures. The specimens were photographed on Ektachrome or Tri-X film (Kodak). RESULTS

Identification of the cell types expressing the JONES antigens in dissociated cell cultures utilized both morphological criteria and immunological markers. In these cultures, neurons were small, phase-bright cells that bound tetanus toxin as evidenced by anti-tetanus toxin antisera (Fields, 1985; Mirsky et al., 1978). Astroglia were phase-dark cells with a flattened morphology in comparison with neurons. They were immunocytochemically identified with antibodies to vimentin or GFAP, or with Rat-401, which recognizes immature radial glia (Hockfield and McKay, 1985). Double-labeling with JONES and tetanus toxin of mixed cultures after 3 hr istl vitro revealed small, phase-bright JONES-positive neurons either isolated or in small clusters (Figs. lA-1C). Phase-dark cells showed an occasional elongated process and some of them expressed RAT-401 (Figs. lD-1F).

402

DEVELOPMENTAL

BIOLOGY

VOLIJME 138,199O

FIG. 1. Ccrebellar cultures fixed 3 hr after plating and double-stained with JONES antibody and tetanus toxin (A, B) or with JONES antibody and Rat-401 (D, E). (A, D) ImmunoAuorescence photomicrographs showing JONES-positive cells in two different cultures. (B) Same field as in A stained with tetanus toxin. Note that several cells were double-labeled. (E) Same field as in D showing Rat-401-positive cells. None of the Rat-4Ol-positive cells were JONES-positive. (C, F) Phase-cont,rast photomicrographs of the double immunofluorescence images shown in A, B and D, E, respectively. Calibration bar = 2.5 pm.

Within the first 24 hr after plating, neurons formed aggregates and started to develop neurites. After 2 days in vitro, neurons were clustered in small clumps connetted by thin bundles of neurites (Fig. 2). Most of the neurons were associated with astroglial cells and both

the neuronal cell soma and processeswere intensely and uniformly stained with JONES antibody (Figs. 2A-2C). The GFAP-positive cells showed predominantly a stellate morphology and only a small proportion exhibited the elongated radial glia-like form (Figs. 2D-2F).

MENIIEZ-OTERO

ANI)

CONSTANTINE-PATON

!/-0-Acet!yl

Gc~ngliosides

403

Frc:. 2. Cerebellar cultures after 48 hr in vifro. Immunofluorescence (RI, and GFAP (El. Note the intense JONES staining of the neuron; astroglial cells (E). Arrow in E points to a glial cell with an elongate shown above. Calibration bar = 50 em.

13hotomicrographs of cultures stained with JONES (A, D), tetanus toxin tl cell bodies and processes (A, D) and the stellate morphologyy of the d morphology. (C, F) Phase-contrast photomicrographs of the cultures

To quantify the number of neuronal and glial cells expressing JONES in these cultures we counted a total of 1200 cells in three different experiments. Only cells clearly recognizable as individuals were evaluated for antibody binding. The JONES antibody stained approximately 705% of the tetanus toxin-positive cells but only 10% of the GFAP-positive cells. The astroglial cells showing JONES antigenicity were, in general, less in-

tensely labeled than the neurons and the antigen appeared to be distributed in a punctiform pattern along the cell contour. The stellate form of astroglial cells in these mixed cultures has been associated with the clustering and anchoring of granule cells rather than supporting their migration (Hatten et ul., 1984). Only a small fraction of the GFAP-positive cells in such cultures expressed the

404

DEVELOPMENTALBIOLOGY

elongated form reminiscent of the Bergman radial glia, which has been associated with extensive granule cell motility by time-lapse video microscopy (Hatten et al., 1984, Edmondson and Hatten, 1987; Gregory et al., 1988). The proportion of Bergman glia-like cells and the amount of neuronal migration on these glia can be significantly increased by plating the cells in ratios of four

VOLIJME

138, 1990

or more neurons to one glial cell (Edmondson and Hatten, 1987). Figure 3 illustrates a typical field from a 4%hr microwell culture of rat cerebellar cells plated at these high neuron to glia ratios. Several morphological features of these cultures are characteristic of extensive granule cell migration. First, very few (tetanus toxin-

Frc:. 3. Cerchellar cultures plated at a 4:l ratio of newons to glia. Cells were fixed 48 hr after plating and double-stained with JONES antibody and tetanus toxin (A, B) or with JONES antihody and GFAP (D, E). (A and D) JONES-positive cells. (B) Same field as in A stained with tetanus toxin. Note that several cells were double-labeled with JONES and tetanus toxin. (E) GFAP-labeled astroglial cells showing a radial morphology. (C, F) Phase-contrast view of the cultures shown above. Calibration bar = 50 pm.

MENDEX-OTERO

AND CONSTANTINE-PATON

8-0-Aceful

Gangliosides

405

positive) granule cells are in large clusters that do not afford contact with a GFAP-positive process. Second, most tetanus toxin-positive cells show long bipolar protoplasmic processes stretched out along a GFAP-positive process (Fig. 4). Third, most GFAP-positive cells have the elongated Bergman glia-like form (Fig. 3E). Observations on 1600 cells in four different experiments with carefully controlled neuron to glia ratios indicated that most neurons were JONES-positive (Figs. 3A and 3B). However, unlike the earlier mixed cultures, in these cultures, most GFAP-positive processes showed pun&ate staining with the JONES antigen (Figs. 4 and 5). Even in regions where high-power morphological examination revealed no evidence of protoplasmic neuronal extensions, the glial cell membranes appeared to express the JONES gangliosides (arrows in Fig. 5). As is illustrated in Fig. 5, JONES staining was also present on the few GFAP-positive cells isolated from neurons in these cultures. By contrast, in cultures of purified glial cells without the addition of neurons, none of the GFAP-positive cells showed any reactivity with the JONES antibody (Fig. 6). The GFAP-positive cells showed a polygonal or flat morphology and only a small population of cells residing on top of the astrocytic monolayer was Jones-positive (Fig. 6A). These cells may represent neurons that remained as a small contaminant of the astroglial fraction. Dissociated granule cells immunostained immediately after the purification step expressed the antigen in high quantities (Fig. 7). These cells when plated in the absence of glia do not attach to the culture dish, form aggregates, and survive poorly (Hatten, 1985).

DISCIJSSION

The tissue culture observations presented above demonstrate that the JONES antigens are expressed by both the granule neurons and a subpopulation of astroglial cells of the postnatal rodent cerebellum. However, the control of this expression appears to be significantly different in the two cell types. Dissociated granule cells stained immediately after the purification step or after a few days in culture expressed the antigen in high quantities. Whether this is true cell autonomous expression, expression induced by interactions the cells experienced prior to dissociation, or expression maintained by some factor or factors in the tissue culture medium remains to be determined. Nevertheless, under identical conditions, cultures composed almost entirely of astroglia do not express 9-0-acetyl gangliosides. The glial cells appear to require some form of neuron-glia interaction to express the JONES antigens. The data available at present do not allow us to determine

FIG;. 4. ImmunoHuorcscence (A, BI anti phase-contrast (0 phc ,tomicrographs of cells double-stained with JONES (A) and GFAP (B). (A) JONES staining outlinrs the ncuronal cell body and cell proce! 3St?S and is associated with a radial glia, GFAP-positive (B) process. ( :&Iibration bar 10 j1n1.

whether the neuron-glia interaction is mediated by contact or by a.diffusible signal. Nor can we discril ninate a direct effect on glial ganglioside expression fr om an indirect effect that is associated with the neura ins’ obvious role in stimulating glial cell differentiation. In populations of glial cells maintained without neurons

406

DEVELOPMENTAL

FIG. 5. Bergman-like GFAP-positive cells double-stained views of the same cells. Note the punctate staining outlining

BIOLOGY

VOLUME

138,199O

with Jones. (A, D) JONES staining; (8, E) GFAP staining; (C, F) Phase-contrast the glial cell body and processes (arrows). Calibration bar = 20 Wm.

the cells fail both to assume a radial morphology and to express the acetylated gangliosides. The neuronal effect on glial differentiation has been well documented by Hatten (Hatten, 1985; Hatten, 1987). Astroglia cultured in the absence of neurons express a flattened morphology and proliferate rapidly, but, when they are grown with neurons, the same cells differentiate along different lines according to the relative proportions of neurons to glia. With relatively small proportions of neurons, the stellate form of glia

predominates and, as the proportion of neurons to glia is increased, the dominant glial type becomes the elongated Bergman-like form. Simultaneously the neurons undergo extensive motile activity along the glial processes.Our results clearly associate the JONES ganglioside with the elongated migration-supporting state of glial differentiation. A possibility that has to be considered is whether the neurons transfer the JONES gangliosides to the astroglial cells. Recent reports indicate that cultured as-

MENDEZ-OTF,RO

AND CONSTANTINE-PATON

9-0Acetyl

Gangliosides

407

possesslittle or no ability to synthesize gangliotetraose gangliosides (Byrne et al., 1988; Sbaschnig-Agler et al., 1988). The limited biosynthetic ability of the astroglial cells as compared to neurons and the fact that both cell types show similar patterns of gangliosides in vivo suggests that astroglial gangliosides may have derived, at least in part, from neurons. It is important to note that a ganglioside-transfer protein was recently isolated from brain which would be capable of catalyzing this process (Gammon et al., 1987). Recent experiments have shown that in the glioma line G26-24 the 9-0-acetyl-GD3 is synthesized from GD3 by acetylation of the terminal sialic acid (D. M. Bonafede, A. C. Missias, and M. Constantine-Paton, unpublished results). Whether this is also true for astrocytes in viva and in vitro remains to be determined. However, these results and the observations that GM3 and GD3 constitute 75-85s of the total gangliosides in cultured astrocytes (Sbaschnig-Agler et al., 1988) suggest that astrocytes in culture have large amounts of

FIN;. 6. Astroglial cells fixed after 48 hr in vitro. (A) Immunofluoresccnce photomicrograph of JONES staining. (B) Immunofluorescence photomicrograph of the GFAP staining. (C) Phase-contrast view. Note that all the flat polygonal GFAP-positive cells were JONESnegative. A few phase-bright JONES-positive cells (arrow in A) were seen growing on top of the astrocgtic layer. Calibration bar = 25 Km.

trocytes have an appreciable ganglioside content, comparable to that i7~ ~ivo, although the patterns differ markedly in both situations. Cultured astrocytes have virtually no gangliotetraose gangliosides, while this structure predominates in freshly isolated cells. It has also been shown that astrocytes both in vivo and in vitro

Frc:. ‘i. Isolated cerebcllar fluorescence photomicrograph ence-contrast photomicrograph = 25 pm.

neurons stained in suspension. Immuno(A) of JONES staining and interfer(B) of the same field. Calibration bar

408

DEVELOPMENTAL

BIOLOGY

GD3 and could synthesize the 9-0-acetyl-GD3. The absence of JONES staining in astroglial cells cultured in the absence of neurons and the different expression in the two different cultures suggests that the O-acetylation of GD3 could be associated with the functional state of the glia. The current in vitro studies are consistent with previous in vivo studies of localization of the 9-O-acetylated gangliosides in regions of neuron migration within the developing brain. Though gangliosides have previously been associated with the migratory activity of cells there is relatively little information on how they may facilitate motility. Gangliosides are known to function as cell surface receptors for a number of toxins (Van Heyningen, 1974) which interact on the cytosolic side of the cell with G proteins (e.g., Cassel and Pfeuffer, 1978; Lai, 1980). In addition, gangliosides have been reported to modulate kinase activities (Goldering et ah, 1985; Kim et al., 1986). Thus, the JONES gangliosides on the surface of migratory cells could potentially facilitate the migratory process directly through a second messenger-induced change in cytoskeletal organization (Forscher et ab, 1987) or a change in rates of membrane insertion (Bourne, 1988). Although gangliosides have been suggested as candidate receptors for fibronectin (Kleinman et ah, 1979; Perkins et al., 1982; Yamada et ah, 1981), it is becoming apparent that the role of gangliosides in integrin-based adhesion systems is that of accessory molecules to facilitate binding or to modulate receptor function rather than to function themselves as specific receptors (Stallcup, 1988; Stallcup et al., 1989, Burns et ah, 1988). The sialic acid residues of gangliosides are known to be powerful chelators of Ca2+ ions (Harding and Halliday, 1980). Consequently, Cheresh (1987) suggested that the disialogangliosides may serve the important function of modulating the strength of cell to substratum adhesions, which depend upon protein-based but calciumdependent adhesion systems. The positioning and amount of disialogangliosides could play a major role in controlling the amount of Ca”+ that is available to these adhesion systems. This would alter the strength of the adhesion they could effect. In addition, the terminal sialic acid residues of cell surface gangliosides could provide the initial electrostatic attraction between the cell and the substratum. The present observations would seem to support a role for JONES gangliosides in modulating cell surface adhesion via Ca2+ binding. The 9-0-acetylated gangliosides are quite clearly localized on the glial substrates which support migration, as well as on the migratory cells themselves. However, it should be pointed out that, if the JONES antigens are involved in modulating the amount of Ca2+available in the microenvironment of a

VOLUME

138.1990

migrating cell process, then they could interact at all of the numerous stages in the process of motility, where Ca2+fluxes have been implicated. Indeed, one intriguing possibility for the function of the JONES antigens was suggested by the fact that many regions of the developing brain express GD3 (Schlosshauer et al., 1988; Goldman et al., 1984; Sparrow and Barnstable, 1988) as well as the 9-0-acetylated GD3 recognized by the JONES mab, although the acetylated forms appear to have a more restricted distribution (Sparrow and Barnstable, 1988). The essential idea is that the 9-0-acetyl group, either by decreasing the negative charge of sialic acid or by protecting the disialoganglioside from sialidases (Schauer, 1987), may alter the amount of Ca2+ binding to GD3. Thus, by regulating the amount of the g-O-acetyl form of this ganglioside in the microenvironment, migratory cells or their substrates may be able to titrate a sensitive Ca”+ buffering system that would provide differential pathway selection as a result of the types of cells that are interacting. In short, gangliosides could be specific receptors for signals that modulate cell motility via second messenger systems. Alternatively, they may alter the charge environment conducive to ligand binding and modulate motility through changes in adhesion. The present observations of migratory neurons which appear to induce JONES ganglioside expression on their relatively nonmotile substrates tend to support the hypothesis for the 9-0-acetylated variant, but clearly the numerous other mechanistic possibilities cannot, at present, be ruled out. Thanks are due to Dr. Mary B. Hatten for invaluable advice regarding microcultures. We thank Dr. Leny A. Cavalcante and Daniela Bonafede for helpful suggestions concerning this work and Levi Amorim for excellent technical assistance. This work was supported by CNPq, FINEP, and FAPERJ grants to R.M-0. and NIH (HD-22498) and NSF(BSN-8616965) grants to M.C-P. REFERENCES BONAFEUK, D. M., MISSIAS, A. C., and CONSTANTINE-PATON, M. (1989). Expression and synthesis of ganglioside 9-0-acetyl-GD3 in mouse glioma subclones. Nwrosci. Atxtr. 15, 567. BOURNE, H. E. (1988). Do GTPases direct membrane traffic in secretion? Cell 53, 669-671. BURNS, G. F., LIJCAS, C. M., KRISSANSEN, G. W., WERKMEISTER, J. A., S(:ANLON, D. B., SIMPSON, R. J., and VADAS, M. A. (1988). Synergism between membrane gangliosides and Arg-Gly-Asp-directed glycoprotein receptors in attachment to matrix proteins by melanoma cells. J. Cdl Bid. 107, 1225~1230. BY’KNE, M. C., FAROOQ, M., SBAXIIING-AGI,ER, M., NORTON, W. T., and LEEIIEN, R. W. (1988). Ganglioside content of astroglial and neurons isolated from maturing rat brain: Consideration of the source of astroglial ~an~liosides. Bmiv Res. 461, 87-9’7. CASSEL, D., and PFIKIFFF.R, T. (1978). Mechanism of cholera toxin action: Covalent modification of the guanyl nucleotide-binding protein of adenylate cyclase system. Proc. N&L Aced Sci. 7JS’A 75, 2669-2673.

MENI)F,Z-OTEKO

!j-0-Awtyl

ANJ) CONSTANTINE-PATON

CHEKESH, Il. A. (1987). Ganglioside involvement in tumor cell-substratum interactions. In “Development and Recognition of the Transformed Cell” (M. I. Greene and T. Hamaoka, Eds.), pp. 407-428. Plenum, New York. CHEREXH, D. A., and KLLER, F. G. (1986). Disialoganglioside GD2 distributes preferentially into substrate-associated microprocesses on the surface of human melanoma cells. Prux. N&l. Acud. Sci. USA 81.5767-5771. CIIERESH, D. A., PIERSBACHER, M. D., HERZIG, M. A., and MIJJ~Q J. (1986). Disialogangliosides GD2 and GD3 are involved in the attachment of human melanoma and neuroblastoma cells to extracellular matrix proteins. J. Cf4/ Bid. 102, 688-696. CHER~~XII, D. A., PYTELA, R., PIERSCHHACIIER, M. D., KI,IER, F. G., RUOSI,AIITI, E., and REISFELD, R. A. (1987). An Arg-Gly-Asp-directcd receptor on the surface of human melanoma cells exist in a divalent cation-dependent functional complex with the disialogangliaside GD2. J. (%/I Biol. 105, 1163-11’72. CONSTANTINE-PATON, M., BLUM, A. S., MENUEZ-OTERO, R., and BARNSTABLE, C’. J. (1986). A cell surface molecule distributed in a dorsoventral gradient in the perinatal rat retina. Natzcw ILtmdm) 324, 459-462. EI>MONLX%)N, J. C., and HATTEN, M. E. (1987). Glial-guided granule neuron migration in/ oitro: A high-resolution time-lapse video microscopic study. J. N~rrosci. 7, 1928-1934. FIELL)S, K. (1!185). Neuronal and glial surface antigens on cells in culture. III “Cell Culture in Neuroscience” (J. E. Bottrnstein and G. Sato, Eds), pp. 45-93. Plenum, New York. FORSUIER, P., KACZMAREK, L. K., BIJCHANAN, J., and SMITH, S. J. (1987). Cyclic AMP induces changes in distribution and transport of organelles within growth cones of Aplysicr bag cell neurons. J. i\‘m roxi. 7, 3600-3611. GAMMON, C. M., VASWANI, K. K., and L~:EL)EN, R. W. (1987). Isolation of two glycolipid transfer proteins from bovine brain: Reactivity toward gangliasides and neutral ,&cosphingolipids. Biocd~mnistrg 26, 621’1-6243. t GOLI)ERJNC;, J. R., OTIS, L. C., YIJ, R. K., and DELORENZO, R. L. (1985). Calcium/ganglioside-dependent protein kinase activity in rat brain membrane. ?J. N~~~~ro~hem. 44, 1229-1234. GOLDMAN, J. E., HIRANO, M., Y~J, R. K., and SEYFRIED, T. N. (1984). GD3 ganglioside is a glycolipid characteristic of immature neuroectodermal cells. X Nm.woin~ w taoi. 7, 179- 192. GKECXJR’~, W. A., EDMONIISON, J. C., HATTEN, M. E., and MASON, C. A. (1988). Cytology and neuron-glial apposition of migrating cerebellar granule cells i?/ t,itro. J. Nwrowi. 8, 1728-1738. GRI:NwAI,n, G. B., FREDMAN, P., MA(:NANI, J. L., TRISLER, D., GINSB~JRG, V., and NIRENHEKC;, M. (1985). Monoclonal antibody 18B8 detects gangliosidcs associated wi-ith neuronal differentiation and synapse formation. Proc:. Nolf. rlctrci. Sci USA 82, 4008-4012. HAR~INC;, S. E., and HAI,I,JI)AY, J. (1980). Removal of sialic acid from cardiac sarcolemma does not affect contractile function in electrically stimulated guinea pig left atria. Nuttcr~ (Lontlorl) 286,819-8X HATTEN, M. E. (1985). Neuronal regulation of astroglial morphology and proliferation in &ro. J. Cell Riol. 100, 384-396. HATTEN, M. E. (1987). Neuronal inhibition of astroglial cell proliferation is membrane-mediated. J. Co11 Bid. 104, 1353-1360. HATTI’:N, M. E., and LIEM, R. K. H. (1981). Astroglia provide a template for the organization of cerebellar neurons ire. ,ritro. J. Cell Biol. 90, 622-630. HATTEN, M. E., LIEM, R. K. H., and MASON, C. A. (1984). Two forms of cerebellar glial cells interact differently with neurons in ~witro. J Cd Bid. 98, 193-204. HOCKFJEI,I), S., and MCKAY, R. D. G. (1985). Identification of major cell classes in the developing mammalian nervous system, J. Neurosci. 5 3 3310-3328. KJM, J. Y. H., GOI.DERING, J. R., DELORENZO, R. J., and Y~J, R. K.

Gcrngliosides

409

(1986). kinase

Gangliasides inhibit phospholipid-sensitive Ca”-dependent phosphorylation of rat myelin basic proteins. J. N~rrosc1. Rrs. 15, 159-166. KLEJNMAN, H. K., MAR’I’IN, G. R., and FISEIMAN, P. H. (1979). Ganglioside inhibition of Iibronectin-mediated cell adhesion to collagen. Proc. Nafl. Amrl. SC;. 16’A 76, 33673371. LAI, C-Y. (1980). The chemistry and biology of cholera toxin. Crif. Ret!. Biochem. 9, 171-206. LF:\‘JNE, J. M., BEASI,Y, L., and STAL,I.CI’~, W. B. (1984). The D.l.l antigen: A cell surface marker for germinal cells of the central nervous system. J. iVf,urtwi. 4, 820-831. MENDI%GTERO, R., and CUNST~ZNTINE-PATON, M. (1987). Expression of JONES gangliasides by ccrebcllar neurons and glia ill 11itro. Nvrcrosci. AO.dn 13, 1636. MENUEZ-OTERO, R., %HI.OSSIIAIJER, B., BAKNSTABLE, C. J., and GUNSTANTIXE-PATON, M. (1988). A developmentally regulated antigen associated with neural cell and process migration. J Neurosci. 8, 564-579. MIRSKY, R., WICNI)ON, I,. M. B., BUCK, P., STOI,KIN, C., and BKAY, I). (1978). Tetanus toxin: A cell surface marker for neurons in cell culture. Bruin Rw. 148, 251-259. PERKINS, R. M., KELLI~C:, S., PATEI,, B., and CRITCHLEY, D. R. (1982). Gangliasides as receptors for fibronectin. ET/J. Cd/ Rc,.s. 141, 231-243. ROSNEK, H., GREIS, C., and HF,NKI+FAHLE, S. (1988). Developmental expression in embryonic rat and chicken brain of a polysialoganglioside-antigen reacting with the monoclonal antibody ($211. UVC. Brain Rrs. 42, 161-171. SH~ZSCIINI(:-AC:LER, M., DRF:YF~JS, H., NORTON, W. T., SENSENBRENNF,R, M., FAKOOQ, M., BYRNE, M. C., and LEEDEN, R. W. (1988). Gangliosides of cultured astroglia. Brain Rm. 461, 98-106. SCHAUER, R. (1987). Sialic acids: Metabolism of 0-acetyl groups. Ire “Methods in Enzymology” (V. Ginsburg, Ed.), Vol. 138, pp. 611-626. Academic Press, San Diego. SCI~OSSHAIIER, B., BLUM, A. S., MENDEZ-OTERO, R., BAKNSTABLE, C. J., and CONSTANTINE-PATON, M. (1988). Developmental regulation of ganglioside antigens recognized by the Jones antibody. J. 1veu rosci. 8, 5x0-592. S~II~ARTZ, M., and SPIRMAN, N. (1982). Sprouting from chicken embryo dorsal root ganglia induced by nerve growth factor is specifically inhibited by affinity purified antiganglioside antibodies. PWK N/f 11.Acad. SYi. USA 79, 6080~6083. SFARRO~, J. R., and BAKNSTABLE, C. J. (1988). A gradient molecule in developing rat retina: Expression of 9-0-acetyl GD3 in relation to cell type, developmental age, and GD3 ganglioside. J. ZVeriro.sci. Rw. 21,398-409.

S~IKMAN, N., SELL, B.-A., and SCHWARTZ, M. (1982). Antiganglioside antibodies inhibit neuritic outgrowth from regenerating goldfish retinal explants. J. Neurochem. 39, 874-877. STALLCXV, W. B. (1988). Involvement of gangliasides and glgcoprotein Iibronectin receptors in cellular adhesion to fibroneetin. E.rp. Cd Res. 177 1 .‘X-102. STALLCUP, W. B., PYTELA, R., and RUOSJ.AHTI, E. (1989). A neuroectoderm-associated gangliaside participates in fibronectin receptormediated adhesion of germinal cells to fibronectin. Lhr Bid 132, 212-229. VAN HF,YNINC:EN, W. E. (1974). Gangliasides as membrane receptors for tetanus toxin, cholera toxin, and serotonin. Nufurf~ (Ltmdm) 249,415m417. WILI,INGF.R, M., and SCHACHNEK, M. (1980). GM1 as a marker for neuronal differentiation in mouse cerebellum. De?,. Bid. 74, lOlL117. YAMALL% K., KENNEIIY, D. W., GROTENDORST, G. R., and MOMOI, T. (1981). Glycolipids:Receptors for fibronectin? J. Cd. P’hysiol. 109, 343-8351.