Lipid rafts mediate the interaction between myelin-associated glycoprotein (MAG) on myelin and MAG-receptors on neurons

Molecular and Cellular Neuroscience 22 (2003) 344 –352

www.elsevier.com/locate/ymcne

Lipid rafts mediate the interaction between myelin-associated glycoprotein (MAG) on myelin and MAG-receptors on neurons Mary Vinson,a,* Oliver Rausch,a Peter R. Maycox,b Rab K. Prinjha,a Debra Chapman,a Rachel Morrow,a Alex J. Harper,a Colin Dingwall,a Frank S. Walsh,a Stephen A. Burbidge,a and David R. Riddella a b

Neurology Centre of Excellence for Drug Discovery, Glaxo SmithKline, New Frontiers Science Park North, Third Ave., Harlow CM19 5AW, UK Psychiatry Centre of Excellence for Drug Discovery, GlaxoSmithKline, New Frontiers Science Park North, Third Ave., Harlow CM19 5AW, UK Received 23 August 2002; revised 25 October 2002; accepted 25 October 2002

Abstract The interaction between myelin-associated glycoprotein (MAG), expressed at the periaxonal membrane of myelin, and receptors on neurons initiates a bidirectional signalling system that results in inhibition of neurite outgrowth and maintenance of myelin integrity. We show that this involves a lipid-raft to lipid-raft interaction on opposing cell membranes. MAG is exclusively located in low buoyancy Lubrol WX-insoluble membrane fractions isolated from whole brain, primary oligodendrocytes, or MAG-expressing CHO cells. Localisation within these domains is dependent on cellular cholesterol and occurs following terminal glycosylation in the trans-Golgi network, characteristics of association with lipid rafts. Furthermore, a recombinant form of MAG interacts specifically with lipid-raft fractions from whole brain and cultured cerebellar granule cells, containing functional MAG receptors GT1b and Nogo-66 receptor and molecules required for transduction of signal from MAG into neurons. The localisation of both MAG and MAG receptors within lipid rafts on the surface of opposing cells may create discrete areas of high avidity multivalent interaction, known to be critical for signalling into both cell types. Localisation within lipid rafts may provide a molecular environment that facilitates the interaction between MAG and multiple receptors and also between MAG ligands and molecules involved in signal transduction. © 2003 Elsevier Science (USA). All rights reserved.

Introduction Within the central nervous system, expression of myelinassociated glycoprotein (MAG)1 is restricted to paranodal loops and periaxonal membranes of myelin (Bartsch et al., 1989; Martini and Schachner, 1986; Trapp et al., 1989) and interacts with a receptor(s) on the surface of neurons. This MAG-mediated interaction between oligodendroctye and neuron is thought to initiate a bidirectional signalling system. Treatment of postnatal neurons with MAG in vitro results in inhibition of neurite outgrowth (Debellard et al., 1996; McKerracher et al., 1994; Mukhopadhyay et al., * Corresponding author. Fax: ⫹44-0-1279-622555. E-mail address: [email protected] (M. Vinson). 1 Abbreviations used: MAG, myelin-associated glycoprotein; p75, the neurotrophin receptor p75; CGC, cerebellar granule cells; MCD, methylcyclodextrin; TGN, trans-Golgi network; Endo H, endoglycosidase H.

1994; Tang et al., 1997) through a pathway involving Rho and Rho kinase activation (Lehmann et al., 1999; Vinson et al., 2001). Expression of MAG on the surface of oligodendroctyes plays a role in the maintenance of myelin integrity (Weiss et al., 2000), which may involve activation of Fyn kinase (Umemori et al., 1994), a molecule key to oligodendrocyte differentiation and myelination (Osterhout et al., 1999; Sperber 2001a,b). Analysis of the interaction between MAG and neurons is therefore important in understanding these events. MAG is a member of the siglec family and binds to sialic acid at the terminii of glycans via a binding site at the N-terminal domain. Recent reports show that sialylated gangliosides GT1b and GD1a are MAG receptors that mediate inhibition of regeneration (Vinson et al., 2001; Vyas et al., 2002). For GT1b, transduction of the signal from MAG to Rho may involve p75, with which it forms a close interac-

1044-7431/03/$ – see front matter © 2003 Elsevier Science (USA). All rights reserved. doi:10.1016/S1044-7431(02)00031-3

M. Vinson et al. / Molecular and Cellular Neuroscience 22 (2003) 344 –352

345

Fig. 1. Myelin-associated glycoprotein (MAG) is not localised within Triton X-100-insoluble lipid rafts. A, O2A cells differentiated for 6 days were fixed and the nucleus stained with Hoescht (blue). Cells were stained for cell surface MAG (green). Cells were then stained either for GM1, or permeabilised and stained for caveolin (red). Each image represents the deconvoluted and projected image from serial sections. Distinct red and green areas can be seen in all images, indicating no colocalisation. B, Whole mouse brain was lysed in the presence of 1% Triton X-100 at 4°C and lipid rafts were isolated by flotation through a sucrose gradient. Fractions are numbered from the top of the tube. Fractions containing lipid rafts (marked with an asterisk) were those at the 5% and 35% sucrose interface, contained ⬍5% of total protein, and were positive for caveolin-1. MAG was not found in the caveolin-containing fractions.

346

M. Vinson et al. / Molecular and Cellular Neuroscience 22 (2003) 344 –352

tion (Yamashita et al., 2002). Although GT1b and GD1a have been identified as functional MAG receptors, additional receptors are likely to exist. MAG is able to interact with specific proteins (De Bellard and Filbin, 1999; Debellard et al., 1996; Franzen et al., 2001; Strenge et al., 1999, 2001), and recently, the GPI-linked Nogo-66 receptor has been identified as a functional MAG receptor (Domeniconi et al., 2002; Liu et al., 2002). The molecular mechanism by which Nogo-66 receptor inhibits neuronal regeneration is not yet known (see note added in proof). In this study, we show that the MAG-mediated interaction between myelin and neurons involves lipid rafts on both cell types. We find that MAG is located within a novel type of Triton X-100-soluble but Lubrol WX-insoluble lipid raft on the surface of myelin and cultured cells. Furthermore, we find that MAG interacts with lipid rafts on the surface of neurons that contain GT1b, p75, Rho, and Nogo-66 receptor. We propose a model of raft–raft interaction between oligodendrocytes and neurons that is likely to have functional significance to the inhibition of central nervous system regeneration following injury and in the maintenance of myelin integrity.

Results and discussion MAG is located within a novel type of Lubrol WX-insoluble lipid raft Over the course of 6 days in culture, O2A progenitors differentiate to form promyelinating oligodendrocytes that form membranous sheets expressing myelin markers (Baron et al., 1998). The bipolar O2A progenitor expresses a low level of cell surface MAG that is visible on some cells (data not shown). This cell surface expression increases during differentiation, finally detectable as highly punctate immunoreactivity restricted to cellular processes (Fig. 1A). To investigate the nature of the punctate immunoreactivity observed, we considered that MAG could be located within discrete microdomains on the cell surface. “Lipid rafts” are a heterogeneous population of membrane microdomains, enriched in cholesterol and glycosphingolipids (Brown and London, 2000; Hakomori, 2002; Simons and Ikonen, 1997). As myelin is rich in cholesterol and sphingolipids (Stoffel and Bosio, 1997), we investigated the possibility that MAG is located within lipid rafts. First, we used two markers of lipid-raft structures, caveolin (located in caveolae, a specialised lipid-raft domain) and GM1 (present in caveolae and noncaveolar glycosphingolipid-rich lipid rafts). MAG showed no colocalisation with either caveolin or GM1 on the surface of oligodendroctyes (Fig. 1A). Caveolae and some glycosphingolipid-rich lipid rafts are commonly prepared based on their Triton X-100 insolubility. To confirm our immunocytochemical observations at a biochemical level, a Triton X-100 lipid-raft preparation

Fig. 2. Myelin-associated glycoprotein (MAG) localisation within Lubrol WX-insoluble lipid rafts. Lysates were prepared in the presence of Lubrol WX from: whole brain (A), cultured primary oligodendrocytes after 6 day differentiation (B) or chinese hamster ovary (CHO) cells expressing MAG at the surface (C). Raft-containing fractions were defined as for Fig. 1. Fractions were probed on Western blots for caveolin-1 and MAG. D, four dishes of CHO/MAG cells were vehicle treated (medium only), or treated with 5 mM methylcyclodextrin (MCD) for 1 h at 37°C. Cell lysates were then prepared in the presence of Lubrol WX and a sucrose gradient separation carried out. Lipid raft-containing fractions of each prep were pooled, 2 ␮g of protein was loaded in each lane. E, 50 ␮l of fraction 4 (raft) or fraction 9 (nonraft) from the preparation in B was incubated with 100 mU of endoglycosidase H (Endo H) (⫹) or the equivalent volume of phosphate-buffered saline (⫺) overnight at 37°C and blotted for MAG.

from whole mouse brain was examined for caveolin and MAG (Fig. 1B). Whereas caveolin was located in the fractions containing low buoyancy Triton X-100-insoluble lipid rafts, MAG was solubilised under these conditions. This suggests that MAG is not located within Triton X-100insoluble microdomains. A recent report describes a novel type of lipid raft that, although soluble in Triton X-100, is insoluble in the detergent Lubrol WX and is restricted to cellular processes (Roper et al., 2000). As we had observed process-restricted expression of MAG, we investigated whether MAG is enriched in this type of raft. Fig. 2A shows fractions from a sucrose gradient of whole mouse brain prepared in the presence of Lubrol WX. In this case, MAG is exclusively localised in the Lubrol WX-insoluble, low buoyancy membrane fractions. Similarly, we found MAG to be localised in

M. Vinson et al. / Molecular and Cellular Neuroscience 22 (2003) 344 –352

these fractions prepared from rat primary oligodendrocytes (Fig. 2B) and from chinese hamster ovary (CHO) cells expressing MAG at the surface (Fig. 2C). The integrity of lipid rafts is known to depend on cellular cholesterol (Scheiffele et al., 1997). To investigate whether the location of MAG within the Lubrol WX-insoluble fractions was dependent on cholesterol, CHO cells expressing MAG were treated with methylcyclodextrin, an agent that removes cholesterol from the cell. After methylcyclodextrin treatment, a 49% reduction in the quantity of MAG within the Lubrolinsoluble fractions was observed compared to vehicletreated cells (Fig. 2D). Similarly, a 32% reduction of MAG levels was seen in Lubrol-insoluble fractions prepared from primary oligodendrocytes treated with methylcyclodextrin compared to vehicle-treated cells (data not shown). This indicates that the localisation of MAG within Lubrol-insoluble domains is at least partly dependent on cellular cholesterol. In our oligodendrocyte preparations, we noticed that MAG exists as two immunoreactive bands of approximately 100 and 80 kDa (Fig. 2B). These correspond to the endoglycosidase H-resistant, mature MAG and the endoglycosidase H-sensitive, endoplasmic reticulum form of MAG, respectively (Fig. 2E). The smaller form was not observed in the lipid-raft fractions, suggesting that, in common with many other raft-associated proteins, MAG becomes associated with microdomains following terminal glycosylation within the trans-Golgi network (Riddell et al., 2001; Roper et al., 2000; Simons and Ikonen, 1997). Due to the fact that MAG is located within the low buoyancy Lubrol-insoluble fractions and that this localisation is partly dependent on cellular cholesterol and occurs after terminal glycosylation within the trans-Golgi network, we hypothesise that MAG is located within a lipid raft on myelin membranes and cultured cells. The enrichment of MAG in Lubrol-insoluble lipid rafts could be important for both the trafficking to and retention in specific domains within myelin (Corbeil et al., 2001).

347

exogenous sialyllactose, both of which significantly reduced binding (Fig. 3B). To characterise the MAG-binding lipid-raft population further, we used cultured cerebellar granule cells, which are known to respond to MAG (Vinson et al., 2001). A recent report has suggested that GT1b interacts with p75 on the surface of neurons and that this interaction results in Rho activation leading to inhibition of neurite outgrowth (Yamashita et al., 2002). In agreement with this hypothesis, we found GT1b and p75 to partially colocalise on the surface of cerebellar granule cells as did GT1b and RhoA (Fig. 4A). Fig. 4b shows a fractions from a sucrose gradient prepared from cerebellar granule cell lysate. In agreement with previous findings (Bilderbeck et al., 1997; Prinetti et al., 2000) lipid-raft fractions contained GT1b and p75. Lipid-raft fractions also contained Rho, previously demonstrated to be located in raft fractions in nonneuronal cells (Michaely et al., 1999). Furthermore, the GPI-linked Nogo-66 receptor was almost exclusively located within these domains. Furthermore, MAG-Fc bound exclusively to lipid raft-containing fractions. This suggests that MAG interacts with a lipid-raft population on the surface of neurons that contains known binding partners GT1b and Nogo-66 receptor as well as molecules required to transmit MAG-mediated intracellular signals into neurons. The significance of a raft–raft interaction Like other lectins, the affinity of a monomeric siglec molecule for sialic acid ligand is thought to be relatively low (Crocker et al., 1999). Increasing the avidity of the interaction by multivalent presentation of both siglec and carbohydrate ligand at their respective cell surfaces is thought to be important for cell– cell interaction (Crocker

MAG interacts with a lipid-raft population on the surface of neurons Gangliosides GT1b and GD1a have recently been identified as functional MAG receptors (Vinson et al., 2001; Vyas et al., 2002). As previously described (Prinetti et al., 2000), we found GT1b to be located within lipid-raft fractions prepared from whole mouse brain in the presence of either Triton X-100 or Lubrol WX (Fig. 3A). To examine the location of MAG-binding molecules, the lipid-raft preparations were probed with MAG-Fc, a soluble recombinant form containing the MAG extracellular domain fused to human IgG. MAG-Fc exclusively bound to the lipid raftcontaining fractions that contained GT1b (Fig. 3B). The specificity of this interaction was demonstrated by either treatment of the membrane with sialidase or addition of

Fig. 3. Myelin-associated glycoprotein (MAG) binds to raft fractions from whole brain that contain GT1b. Triton X-100 or Lubrol WX preparations shown in Figs. 1B and 2A were spotted onto nitrocellulose and blotted for GT1b by using biotinylated anti-GT1b IgM antibody followed by streptavidin-HRP (A) or for MAG binding partners using MAG-Fc followed by anti-human Fc HRP (B). MAG binding was abolished by preincubation of the membrane with 500 mU/ml sialidase for 1 h at 37°C or inclusion of 1 mM 3⬘-sialyllactose (3⬘sl).

Fig. 4. Myelin-associated glycoprotein (MAG) binds to lipid-raft fractions prepared from cultured primary cerebellar neurons that contain known binding partners and molecules required for signal transduction. A, cerebellar neurons were fixed and stained for cell-surface GT1b (red), then permeabilised and

M. Vinson et al. / Molecular and Cellular Neuroscience 22 (2003) 344 –352

349

Fig. 5. Schematic diagram of raft–raft interaction between myelin and neurons. Lipid rafts enriched in cholesterol are represented by gray shading. Myelin-associated glycoprotein (MAG) molecules are shown in black, neuronal MAG receptors in red, coreceptors in blue, and raft-associated signalling molecules in yellow. The localisation of MAG and MAG-binding partners within lipid rafts may result in discrete patches of high avidity interaction between the two cell types. Furthermore, rafts may also provide a platform in which multiple interactions may occur between MAG and MAG receptors, between MAG receptors and coreceptors required for signal transduction, and between receptors and signalling molecules.

stained for either p75 (green, top panel) or RhoA (green, bottom panel). Deconvoluted images are shown of serial sections through the cells. Areas of yellow in the merged images indicate colocalisation. GT1b partially colocalises with both p75 (top panel) and RhoA (bottom panel). B, lipid rafts were prepared from cerebellar granule cells. Triton X-100 lipid raft-containing fractions were defined as described in Fig. 1. Fractions were immunoblotted for GT1b, p75, RhoA, Nogo-66 receptor, and caveolin or ligand blotted using MAG-Fc (as described for Fig. 3).

350

M. Vinson et al. / Molecular and Cellular Neuroscience 22 (2003) 344 –352

and Feizi, 1996). A dramatic increase in binding potency generated by multivalent presentation of siglec protein (Crocker and Kelm, 1996) and sialic acid-bearing ligand (Hashimoto et al., 1998) has been clearly demonstrated. Furthermore, multivalent recognition of MAG receptors GT1b and GD1a is required to activate the pathway leading to inhibition of neurite outgrowth (Vyas et al., 2002). Similarly, on the oligodendrocyte cell surface, multivalent recognition of MAG is required for MAG phosphorylation and Fyn activation (Umemori et al., 1994). For the first time, we show that MAG and known MAG receptors are located within lipid rafts on the surface of opposing cells. Within rafts, lipids exist in a tightly packed, liquid-ordered state (London and Brown, 2000). Localisation within these ordered domains may therefore increase the avidity of cell interaction molecules. Thus, localisation of both MAG and binding partners within lipid rafts on opposing cell surfaces may create discrete areas of highly avid interaction, critical for activating signalling pathways in both cell types (Fig. 5). Since the identification of gangliosides as MAG receptors, the hypothesis that MAG interacts with ligands within lipid rafts has been suggested (McKerracher, 2002). Here we show that this is indeed the case. First, recombinant MAG bound exclusively to immobilised lipid-raft fractions. Second, these fractions were shown to contain known MAG receptors and molecules required to mediate inhibition of neurite outgrowth. Finally, partial colocalisation of GT1b with both p75 and Rho indicates that these molecules may be located within the same raft type. MAG-mediated inhibition of neurite outgrowth is attenuated in the absence of either complex gangliosides (including GT1b and GD1a; Vyas et al., 2002), p75 (Yamashita et al., 2002), or Nogo-66 receptor (Strittmater et al., 2002). One interpretation of these findings is that these molecules interact with each other. Their enrichment within lipid rafts may therefore provide a microenvironment that facilitates multiple interactions, both between MAG and multiple binding partners and between molecules required for response to MAG (Fig. 5). In conclusion, we suggest that a raft–raft interaction may therefore play a key role in MAG-mediated events in both neurons and glia. Furthermore, as Nogo-66 receptor and Rho are involved in inhibition of neurite outgrowth mediated by Nogo and OMgp, it is possible that they act as signalling platforms for multiple inhibitory signals.

Experimental methods Cell culture O2A progenitor cells were prepared as described in Baron et al. (1998). Briefly, primary mixed brain cell cultures were prepared from brains of 1–2-day-old SpragueDawley rats and plated onto poly-L-lysine-coated T75 flasks. After 12 days in culture, O2A progenitor cells were

removed by mechanical shaking (McCarthy and de Vellis, 1980) and incubated in a bacterial culture dish for 15 min to differentially remove microglia. The supernatant was then placed in a 50-ml tube for 2 min. Cell aggregates sediment at the bottom of the tube allowing the single-cell suspension to be removed and centrifuged at 1,000 rpm for 5 min. The cell pellet was resuspended in Sato’s medium and plated at the required cell density on poly-L-lysine-coated wells. After 24 h in culture this preparation contained predominantly O2A progenitor cells. Cells were used for the experiments described here after 6 days in culture. Cerebellar granule cells were prepared as described (Vinson et al., 2001). CHO cells expressing MAG at the surface were prepared as follows. Full-length MAG was cloned by reverse transcription–polymerase chain reaction from human spinal cord total RNA (Clontech). First strand cDNA was primed by using oligo(dT). Primer sequences for PCR amplification were as follows, and incorporated a 5⬘ HindIII site and 3⬘ XhoI site. Forward: TC AAGCTT CAGGTGGACCCAGAAGACGTCC; reverse: TA CTCGAG ACCTCACTGTCACCTGCC. The 2.3-kb product was cloned into pcDNA3 vector (Invitrogen). CHO cells were transfected by using Lipofectamine and mass-selected with G418. A high-expressing clone was selected using flow cytometry (FACS). Immunocytochemistry Immunocytochemistry was carried out by using standard techniques. Primary antibodies used were as follows: antiMAG mouse monoclonal antibody (Chemicon) at a 1:500 dilution; anti-caveolin rabbit polyclonal (Transduction Labs) at 1:500; biotinylated cholera toxin for GM1 staining (Sigma) at 1:500; biotinylated anti-GT1b (Seikagaku) at 1:500; p75 rabbit polyclonal (Promega) at 1:500. Secondaries were from Jackson Laboratories (preabsorbed against multiple species) and used at 0.2 ␮g/ml as follows: antimouse Cy2 (for detection of anti-MAG), anti-rabbit Cy5 (for caveolin), streptavidin Cy5 (for cholera toxin and GT1b), and anti-rabbit Cy2 (for p75). For analysis of cell surface MAG, GT1b, or GM1, cells were not permeabilised before staining. For costains of cell-surface MAG or GT1b with intracellular epitopes on caveolin or p75, MAG or GT1b staining (primary and secondary antibody steps) was carried out on nonpermeabilised cells. Cells were then permeabilised by using 0.1% Triton X-100 prior to incubating with anti-caveolin or anti-p75 followed by the appropriate secondaries. Staining was visualised by using an inverted Nikon microscope with appropriate excitation and emission barrier filters. Digital images were sequentially captured for each fluorophore on a CCD camera from sequential optical sections using a 0.1-␮m step size. Collected images were processed by using the DeltaVision deconvolution system (Applied Precision Inc., Washington).

M. Vinson et al. / Molecular and Cellular Neuroscience 22 (2003) 344 –352

351

Lipid-raft preparation

References

Whole mouse brain was homogenised in MBS buffer [25 mM 2-(N-morpholino)ethanesulfonic acid, 150 mM NaCl, pH 6.5] containing 1% (vol/vol) Triton X-100 or 1% (vol/ vol) Lubrol WX and fractionated in a 5– 45% discontinuous sucrose interface as described (Riddell et al., 2001). Fractions were collected from the top of tube. Low buoyancy detergent-resistant membrane fractions were those at the 5–35% sucrose interface, were caveolin 1-positive, and typically contained 4% (Triton X-100, whole brain), 5% (Lubrol WX, whole brain), and 7% (Lubrol WX, oligodendrocytes) total protein. For lipid-raft preparation from cultured oligodendrocytes, CHO cells and cerebellar neurons, cells were scraped into ice-cold MBS buffer containing 1% Lubrol WX and the method followed as above. For methylcyclodextrin treatment of cultured cells, cells were incubated with 5 mM methylcyclodextrin (Sigma, Poole, UK) or medium alone for 1 h at 37°C, washed, and scraped. Western blot analysis was carried out by using standard techniques using the caveolin and p75 antibodies described above. For detection of MAG, the anti-MAG rabbit polyclonal (Santa Cruz sc-9543, purchased from Autogen Bioclear, Wiltshire, UK) at 1:1,000 was used. For detection of RhoA, the anti-Rho rabbit polyclonal (Santa Cruz) at 1:100 was used.

Baron, W., de Jonge, J.C., de Vries, H., Hoekstra, D., 1998. Regulation of oligodendrocyte differentiation: protein kinase C activation prevents differentiation of O2A progenitor cells toward oligodendrocytes. Glia 22, 121–129. Bartsch, U., Kirchoff, F., Schachner, M., 1989. Immunohistological localization of the adhesion molecules L1, N-CAM, and MAG in the developing and adult optic nerve of mice. J. Comp. Neurol. 284, 451– 462. Bilderback, T.R., Grigsby, R.J., Dobrowsky, R.T., 1997. Association of p75NTR with caveolin and localisation of neurotrophin-induced sphingomyelin hydrolysis to caveolae. J. Biol. Chem. 272, 10922–10927. Brown, D., London, E., 2000. Structure and function of sphingolipid- and cholesterol-rich membrane rafts. J. Biol. Chem. 275, 17221–17224. Corbeil, D., Roper, K., Fargeas, C.A., Huttner, W.B., 2001. Prominin: a story of cholesterol, plasma membrane protrusions and human pathology. Traffic 2, 82–91. Crocker, P.R., Feizi, T., 1996. Carbohydrate recognition systems: functional triads in cell-cell interactions. Curr. Opin. Struct. Biol. 6, 679 – 691. Crocker, P.R., Kelm, S., 1996. Methods for studying the cellular binding properties of lectin-like receptors, in: Weir, D.M. (Ed.), Handbook of Experimental Immunology, Vol. II, Blackwell, Edinburgh, pp. 166.1– 166.11.. Crocker, P.R., Vinson, M., Kelm, S., Drickamer, K., 1999. Molecular analysis of sialoside binding to sialoadhesin by NMR and site-directed mutagenesis. Biochem. J. 341, 355–361. De Bellard, M., Filbin, M.T., 1999. Myelin-associated-glycoprotein, MAG, selectively binds several neuronal proteins. J. Neurol. Res. 56, 213– 218. De bellard, M.E., Tang, S., Mukhopadhyay, G., Shen, Y.J., Filbin, M.T., 1996. Myelin-associated glycoprotein inhibits axonal regeneration from a variety of neurons via interaction with a sialoglycoprotein. Mol. Cell. Neurosci. 7, 89 –101. Domeniconi, M., Cao, A., Spencer, T., Sivasankaran, R., Wang, K.C., Nikulina, E., Kimura, N., Cai, H., Deng, K., Gao, Y., He, Z., Filbin, M.T., 2002. Myelin-associated glycoprotein interacts with the Nogo66 receptor to inhibit neurite outgrowth. Neuron 35, 283–290. Franzen, R., Sandra, T.L., Dashiell, S.M., Rottkamp, C.A., Hammer, J.A., Quarles, R.H., 2001. Microtubule-associated protein 1B: a neuronal binding protein for myelin-associated glycoprotein. J. Cell Biol. 155, 893– 898. Hakomori, S., 2002. Inaugural article: the glycosynapse. Proc. Nat. Acad. Sci. USA 99, 225–232. Hashimoto, Y., Suzuki, M., Crocker, P.R., Suzuki, A., 1998. A streptavidin-based neoglycoprotein carrying more than 140 GT1b oligosaccharides: quantitative estimation of the binding specificity of murine sialoadhesin expressed on CHO cells. J. Biochem. (Tokyo) 123, 468 – 478. Lehmann, M., Fournier, A., Selles-Navarro, I., Dergham, P., Sebok, A., Leclerc, N., Tigyi, G., McKerracher, L., 1999. Inactivation of Rho signaling pathway promotes CNS axon regeneration. J. Neurosci. 19, 7537–7547. Liu, B.P., Fournier, A., GrandPre, T., Strittmatter, S.M., 2002. Myelinassociated glycoprotein as a functional ligand for the Nogo-66 receptor. Science 297, 1190 –1193. London, E., Brown, D., 2000. Insolubility of lipids in Triton X-100: physical origin and relationship to sphingolipid/cholesterol membrane domains (rafts). Biochim. Biophys. Acta 1508, 182–195. Martini, R., Schachner, M., 1986. Immunoelectron microscopic localization of neural cell adhesion molecules (L1, N-CAM, and MAG) and their shared carbohydrate epitope and myelin basic protein in developing sciatic nerve. J. Cell Biol. 103, 2439 –2448. McCarthy, K.D., de Vellis, J., 1980. Preparation of separate astroglial and oligodendrocyte cell cultures from rat cerebral tissue. J. Cell Biol. 85, 890 – 892.

Dot-blot analysis Two microliters of each lipid-raft fraction was spotted onto nitrocellulose and allowed to dry. For sialidase treatment of membrane, blots were incubated with 2 ml of 500 mU/ml V. cholerae sialidase (Glyko, Bkester, UK) for 1 h at 37°C, then washed. Membranes were blocked in 5% fat-free dried milk and then incubated overnight with either biotinylated anti-GT1b at 1:500 or 20 ␮g/ml MAG-Fc (Vinson et al 2001) in the presence or absence of 1 mM 3⬘-sialyllactose (Glyko). Blots were washed and incubated with either streptavidin-HRP (Amersham) at 1:1,000 (for GT1b) or goat anti-human IgG-Fc (Sigma) at 1:1,000 followed by anti-goat-HRP (Perbio, Cheshire, UK) at 1:4,000. Blots were developed by using ECL (Amersham, Buckinghamshire, UK).

Acknowledgments Thanks to Zinat E. Enayat and Laura Facci for help with cell culture. Note added in proof. p75 has recently been shown to act as a co-receptor for the GPI-linked Nogo-66 receptor. Wang et al. 2002 Nature 420, 74 –78.

352

M. Vinson et al. / Molecular and Cellular Neuroscience 22 (2003) 344 –352

McKerracher, L., 2002. Ganglioside rafts as MAG receptors that mediate blockade of axon growth. Proc. Natl. Acad. Sci. USA 99, 7811–7813. McKerracher, L., David, S., Jackson, D.L., Kottis, V., Dunn, R.J., Braun, P.E., 1994. Identification of myelin-associated glycoprotein as a major myelin-derived inhibitor of neurite growth. Neuron 13, 805– 811. Michaely, P.A., Mineo, C., Ying, Y., Anderson, R.G.W., 1999. Polarised distribution of endogenous Rac1 and RhoA at the cell surface. J. Cell Biol. 274, 21430 –21436. Mukhopadhyay, G., Doherty, P., Walsh, F.S., Crocker, P.R., Filbin, M.T., 1994. A novel role for myelin-associated glycoprotein as an inhibitor of axonal regeneration. Neuron 13, 757–767. Osterhout, D.J., Wolven, A., Wolf, R.M., Resh, M.D., Chao, M.V., 1999. Morphological differentiation of oligodendrocytes requires activation of Fyn tyrosine kinase. J. Cell Biol. 141, 1209 –1218. Prinetti, A., Chigorno, V., Tettamanti, G., Sonnino, S., 2000. Sphingolipidenriched membrane domains from rat cerebellar granule cells differentiated in culture. A compositional study. J. Biol. Chem. 275, 11658–11665. Riddell, D.R., Christie, G., Hussain, I., Dingwall, C., 2001. Compartmentalisation of ␤-secretase (ASP2) into low-buoyant density, noncaveolar lipid rafts. Current Biology 11, 1288 –1293. Roper, K., Corbeil, D., Huttner, W.B., 2000. Retention of prominin in microvilli reveals distinct cholesterol-based lipid micro-domains in the apical plasma membrane. Nat. Cell Biol. 2, 582–592. Scheiffele, P., Roth, M.G., Simons, K., 1997. Interaction of influenza virus haemagglutinin with sphingolipid-cholesterol membrane domains via its transmembrane domain. EMBO J. 16, 5501–5508. Simons, K., Ikonen, E., 1997. Functional rafts in cell membranes. Nature 387, 569 –572. Sperber, B.R., Boyle-Walsh, E.A., Engleka, M.J., Gadue, P., Peterson, A.C., Stein, P.L., Scherer, S.S., McMorris, F.T., 2001a. A unique role for Fyn in myelination. J. Neurosci. 21, 2039 –2047. Sperber, B.R., McMorris, F.A., 2001b. Fyn tyrosine kinase regulates oligodendroglial cell development but is not required for morpho-

logical differentiation of oligodendrocytes. J. Neurosci. Res. 63, 303–312. Stoffel, W., Bosio, A., 1997. Myelin glycolipids and their functions. Curr. Opin. Neurobiol. 5, 654 – 661. Strenge, K., Brossmer, R., Ihrig, P., Schauer, R., Kelm, S., 2001. Fibronectin is a binding partner for the myelin-associated glycoprotein (siglec4a). FEBS Lett. 499, 262–267. Strenge, K., Schauer, R., Kelm, S., 1999. Binding partners for the myelin-associated glycoprotein of N2A neuroblastoma cells. FEBS Lett. 444, 59–64. Tang, S., Woodhall, R.W., Shen, Y.J., Debellard, M.E., Saffell, J.L., Doherty, P., Walsh, F.S., Filbin, M.T., 1997. Soluble myelin-associated glycoprotein (MAG) found in vivo inhibits axonal regeneration. Mol. Cell. Neurosci. 9, 333–346. Trapp, B.D., Andrews, S.B., Cootauco, C., Quarles, R.H., 1989. The myelin-associated glycoprotein is enriched in multivesicular bodies and periaxonal membranes of actively myelinating oligodendrocytes. J. Cell Biol. 109, 2417–2426. Umemori, H., Sato, S., Yagi, T., Aizawa, S., Yamamoto, T., 1994. Initial events of myelination involve Fyn tyrosine kinase signalling. Nature 367, 572–576. Vinson, M., Strijbos, P.J.L.M., Rowles, A., Facci, L., Moore, S.E., Simmons, D.L., Walsh, F.S., 2001. Myelin-associated glycoprotein interacts with ganglioside GT1b: a mechanism for neurite outgrowth inhibition. J. Biol. Chem. 276, 20280 –20285. Vyas, A.A., Patel, H.V., Fromholt, S.E., Heffer-Lauc, M., Vyas, K.A., Dang, J., Schachner, M., Schnaar, R.L., 2002. Gangliosides are functional nerve cell ligands for myelin-associated glycprotein (MAG), an inhibitor of nerve regeneration. Proc. Natl. Acad. Sci. USA 99, 8412– 8417. Weiss, M.D., Hammer, J., Quarles, R.H., 2000. Oligodendrocytes in aging mice lacking myelin-associated glycoprotein are dystrophic but not apoptotic. J. Neurosci. Res. 62, 772–780. Yamashita, T., Higuchi, H., Tohyama, M., 2002. The p75 receptor transduces the signal from myelin-associated glycoprotein to Rho. J. Cell Biol. 157, 565–570.