Tubulin Anchoring to Glycolipid-enriched, Detergent-resistantDomains of the Neuronal Plasma Membrane*

(Received for publication, October 29, 1999, and in revised form, January 20, 2000)

Paola Palestini‡, Marina Pitto, Gabriella Tedeschi§, Anita Ferraretto¶, Marco Parenti,Joseph Brunneri, and Massimo Masserini

From the Department of Experimental, Environmental Medicine and Biotechnologies, Medical School,University of Milano-Bicocca, Hospital S. Gerardo, 20052 Monza, Italy, the ¶Department of Medical Chemistryand Biochemistry and the §Institute of Veterinary Physiology and CH-8092 Biochemistry,University of Milano, 20133 Milano, Italy, and the iEidgenossische Technische Hochschule, Zurich, Switzerland

After incubation of intact living cultured rat cerebel-lar granule cells at 37 °C with a new GM1 gangliosideanalog, carrying a diazirine group and labeled with 125Iin the ceramide moiety, followed by photoactivation, arelatively small number of radiolabeled proteins weredetected in a membrane-enriched fraction. A protein ofabout 55 kDa with a pI of about 5 carried a large portionof the radioactivity even if incubation and cross-linkingwere performed at 4 °C and in the presence of inhibitorsof endocytosis, suggesting that it is cross-linked at theplasma membrane. Immunoprecipitation and Westernblotting experiments showed the positivity of this pro-tein for tubulin. Trypsin treatment of intact cells ruledout the involvement of a plasma membrane surface tu-bulin. Release of radioactivity from cross-linked tubulinafter KOH treatment (but not hydroxylamine treatment)suggested that the photoactivated ganglioside reactswith an ester-linked fatty acid anchor of tubulin. Lowbuoyancy, detergent-resistant membrane fractions, iso-lated from cells after incubation with the GM1 analogueand photoactivation, proved their enrichment in endog-enous and radioactive GM1 ganglioside, sphingomyelin,cholesterol, signal transduction proteins, and tubulin. Itis noteworthy that radioactive tubulin was also detectedin this fraction, indicating the presence of tubulin mol-ecules carrying a fatty acid anchor in detergent-resis-tant, ganglioside-enriched domains of the plasma mem-brane. Parallel experiments carried out with aphosphatidylcholine analogue, also carrying a diazirinegroup and labeled with 125I in the fatty acid moiety,showed the specificity of tubulin interaction with GM1.Taken together, these results indicate that some tubulinmolecules are associated with a lipid anchor to deter-gent-resistant glycolipid-enriched domains of theplasma membrane. This novel feature of membrane do-mains can provide a key for a better understanding oftheir biological role.

Gangliosides are amphiphilic components of the plasmamembrane of vertebrate cells and are known to participate in anumber of cell surface-mediated events, such as cell-cell recog-

nition processes, modulation of cell growth and differentiation,receptor function, and membrane-mediated transfer of infor-mation (1–5). This wide range of functions is likely attributableto their membrane topology. In fact, they are asymmetricallylocated in the outer leaflet of the plasma membrane bilayer,with the oligosaccharide moiety exposed toward the externalmedium and the ceramide portion embedded in the hydropho-bic core of the bilayer. Therefore, they are available to interacteither with membrane-associated molecules or with externalligands. As a matter of fact, the occurrence of ganglioside-protein interaction has been reported (6–9).

Another peculiar feature of glycolipids is their enrichment—along with sphingomyelin, cholesterol, and a series of function-ally related proteins—within discrete plasma membrane do-mains, the involvement of which in the mechanisms of signaltransduction, cell adhesion, and lipid/protein sorting has beenpostulated (10–12). In particular, among glycolipids, domainsare typically enriched in GM1 ganglioside, which is often uti-lized as a lipid marker of these membrane structures (11, 13).

Owing to their peculiar membrane distribution, interactionof glycolipids with specific proteins of domains is expected.However, despite the reported co-segregation of glycolipids andspecific proteins inside domains (14–16), proof of their directinteraction has not been provided, with the exception of caveo-lin in caveolae (17).

Among the tools utilized to investigate the interaction withmembrane-associated proteins, photoactivable radioactive an-alogs of lipids, gangliosides included, which are able to co-valently cross-link and thus radiolabel neighboring moleculesupon illumination, have been repeatedly used (17–23). In thepresent investigation, we used this approach in order to iden-tify proteins interacting with gangliosides in neurons, in whichthese glycosphingolipids are particularly abundant (24), but inwhich the role of domains has been only partially investigated(13). For this purpose, cultured rat cerebellar granule cellswere utilized, coupled with the use of a new GM1 gangliosideanalog, TID-GM1,1 carrying a photoactivable diazirine groupand labeled with 125I in the ceramide moiety.

* This work was supported by Grant Cofin 1998 from Ministerodell’Universita e della Ricerca Scientifica e Tecnologica (Rome, Italy) (toM. M.) and Grant CT98.00488.CT04.115.33097 from the CNR (Rome,Italy) (to P. P.). The costs of publication of this article were defrayed inpart by the payment of page charges. This article must therefore behereby marked “advertisement” in accordance with 18 U.S.C. Section1734 solely to indicate this fact.

‡ To whom correspondence should be addressed: Medical School, Uni-versity of Milano-Bicocca, via Saldini 50, 20133 Milano, Italy. Tel.: 39-02-70645238; Fax: 39-02-70645254; E-mail: paola.palestini@ unimib.it.

1 The abbreviations used are: TID-GM1, 9-[[[2-[125I]iodo-4-((triflu-oromethyl)-3H-diazirin-3-yl)benzyl]oxy]carbonyl]nonanoyl-GM1 gan-glioside; CAPS, 3-(cyclohexylamino)-1-propanelsulfonic acid; DRF, de-tergent-resistant membrane fraction; HDF, high density membranefraction; MES, 2-(N-morpholino)ethanesulfonic acid; TID-PC, (1-O-do-decanoyl-2-O-(9-[[[2-(125I)iodo-4-(trifluoromethyl-3H-diazirin-3-yl)-ben-zyl]oxy] carbonyl] nonanoyl-sn-glycero-3-phosphocholine; 2D, two-dimensional; EF, gel electrophoresis; PAGE, polyacrylamide gelelectrophoresis.

THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 275, No. 14, Issue of April 7, pp. 9978–9985, 2000© 2000 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A.

This paper is available on line at http://www.jbc.org9978

at UN

IV D

EG

LI ST

UD

I DI M

ILAN

on January 5, 2007 w

ww

.jbc.orgD

ownloaded from

MATERIALS AND METHODS

Chemicals—The reagents used (analytical grade) and highperformance TLC plates (Kieselgel 60) were purchased fromMerck GmbH (Darmstadt, Germany). Modified Eagle’s basalmedium; fetal calf serum; trypsin; CAPS; MES; and antibod-ies against a-, b-, and acetyl-tubulin, against actin, andagainst GAP-43 were from Sigma. Antibodies against Fynwere from Transduction Laboratories (Lexington, KY); mono-clonal anti-Tau was a kind gift of Dr. A. Matus (FriedrichMiescher Institute, Basel, Switzerland). Rabbit anti-Go pro-tein a-subunit (Goa) and polyclonal antiserum raised againsta synthetic peptide corresponding to amino acids 2–17 of theGoa, was kindly provided by G. Milligan (University of Glas-gow, Scotland, United Kingdom). All of the material for theelectrophoresis was from Bio-Rad. 125I (IMS-300), 14C-labeledmethylated protein standard for electrophoresis, protein A-Sepharose, and autoradiographic films were from AmershamPharmacia Biotech. GM1 ganglioside was extracted and pu-rified from calf brain (25). Preparation and purification oftritiated GM1, labeled at the 3-position of the long chainbase, was accomplished as described (26, 27). Horseradishperoxidase-labeled cholera toxin B subunit was from ListBiological (Vandell Way, CA).

Cell Cultures—Granule cells, obtained from the cerebella of8-day-old Harlan Sprague-Dawley rats (Charles River, Milan,Italy), were prepared and cultured as described (28, 29). Mor-phological differentiation of granule cells in culture was fol-lowed by microscopic examination, and cell viability was mon-itored with fluorescein diacetate and propidium iodide (30).The experiments were performed with cells cultured for 8 daysin vitro. The protein content was determined by the method ofLowry et al. (31).

Chemical Synthesis of Photoactivable, 125I-Labeled GM1Ganglioside (TID-GM1) and of Photoactivable, 125I-LabeledPhosphatidylcholine (TID-PC)—For the preparation of TID-PC, the procedure previously described (32) has been fol-lowed. The final specific radioactivity was 2000 Ci/mmol, andthe radiochemical purity, assessed by TLC using differentsolvent systems followed by autoradiography, was over 98%.

The new photoactivable radioactive GM1 ganglioside analog,TID-GM1, carrying a 9-[[[2-[125I]iodo-4-((trifluoromethyl)-3H-diazirin-3-yl)benzyl]oxy]carbonyl]nonanoyl-fatty acyl residuein substitution of the native fatty acid present in the ceramidemoiety was synthesized relying on procedures previously de-scribed (32, 33), adapted for the synthesis of this particularglycolipid. The procedure is summarized here. First, the car-boxylic acid function of 9-[[[2-tributylstannyl-4-(trifluorom-ethyl-3H-diazirin-3-yl)-benzyl]oxy]carbonyl]nonanoic acid (32)was activated by formation of the N-hydroxysuccinimide ester,following the procedure described for the corresponding propa-noic acid N-hydroxysuccinimide ester (32). The activated esterwas purified by silica gel column chromatography using ether/hexane (3:1; v/v). The fractions containing the product (Rf ofhydroxysuccinimide ester 5 0.32) were subjected to TLC usingthe same solvent system. After elution from the gel and evap-oration of the solvent, the colorless oily hydroxysuccinimideester (90% yield) was obtained. Next, 6 mmol of lyso-GM1,prepared as described in Ref. 33, were N-acylated using anexcess (21 mmol) of the hydroxysuccinimide ester in 250 ml ofTHF, 6 ml of N-methylmorpholine, and 5 ml of H2O. The initialsuspension was stirred overnight at 40 °C, until a clear solu-tion was obtained. The reaction mixture was subjected to pre-parative TLC (silica gel) using chloroform/methanol/water, 60:35:4 (v/v/v), as the solvent system. The main band (having andRf value almost identical to that of standard GM1 ganglioside)was eluted from the gel using the same solvent system. The

material was further purified by gel filtration using a column ofSephadex LH20 (1.5 3 40 cm) equilibrated and eluted withCH2Cl2/methanol (1:3) (v/v). Fractions containing the stanny-lated photoactivable ganglioside were identified by analyticalTLC (detection by UV fluorescence quenching). Finally, stan-nylated photoactivable GM1 was subjected to radioiododestan-nylation following the protocol reported for various lipid ana-logues (32). Briefly, to a solution of photoactivable GM1 (20–50nmol) in 20 ml of glacial acetic acid, 5 mCi of Na125I (IMS-300;Amersham Pharmacia Biotech) and 5 ml of peracetic acid wereadded. After 2 min, the reaction was quenched by the additionof 20 ml of 100 mM sodium iodide and, after additional 2 min, of100 ml of 10% sodium bisulfite solution. The reaction mixturewas diluted with water (1 ml) and applied to a Sep-Pak PlusC18 Cartridge (Waters, Milford, MA). After elution of salts andfree radioactivity with water, TID-GM1 was eluted from thecolumn with methanol. The specific radioactivity of the mate-rial was estimated to be .100 Ci/mmol. Its radiochemical pu-rity was assessed by TLC, using different solvent systemsfollowed by autoradiography, and was .98%. Samples werephotolyzed using a UV lamp (300 W, Jelosyl, Milan, Italy). Inorder to define the time needed for photolysis of the TID re-agents, a 1025 M solution of the stannylated, unlabeled photo-activable ganglioside (see above) was irradiated for differentperiods of time. Photolysis of the diazirine group was monitoredspectroscopically, by measuring the disappearance of the char-acteristic diazirine band at 350 nm (molar extinction coefficientat 350 nm 5 300).

Treatment of Granule Cells with TID-GM1 and Cross-linkingExperiments—Cells plated in 10-cm dishes were washed withLocke’s solution and incubated for 2 h at 37° °C, or for 4 h at4 °C, with 2.5 ml of the same solution containing 1026 M TID-GM1. After incubation, the cells were washed three times withLocke’s solution at 4 °C, irradiated 5 min on ice with UV light(20), and then incubated with Eagle’s basal medium containing10% fetal calf serum for 30 min at 37 °C. In some experiments,cell incubation with TID-GM1 at 4 °C was carried out in thepresence of 20 mM nocodazole or 100 mM colchicine, added 1 hbefore the photoactivable ganglioside (34, 35).

In other experiments, after cell incubation with TID-GM1 at4 °C and irradiation, cells were treated for 5 min at 37 °C withtrypsin (0.1% in 2 ml of phosphate-buffered saline solution)(36).

At the end of the treatments, cells were washed twice withLocke’s solution; scraped in 2.5 ml of a solution containing 250mM sucrose, 0.1 mM EDTA, and 1 mM chymostatin, leupeptin,antipapain, and pepstatin protease inhibitor mixture (37) in 1mM potassium phosphate buffer, pH 7.4; and finally centri-fuged (8000 3 g for 10 min). The pellet was homogenized andcentrifuged (1000 3 g for 10 min) three times in the samebuffer, and the pooled supernatants were centrifuged at100000 3 g for 1 h. The pellet obtained, from now on called the“membrane-enriched fraction,” was used for further analysis(38).

Treatment of Granule Cells with TID-PC and Cross-linkingExperiments—Cells plated in 10-cm dishes were washed withLocke’s solution and incubated for 2 h at 4 °C with 2.5 ml ofthe same solution containing 2 3 1029 M TID-PC. After in-cubation, the cells were washed three times with Locke’ssolution at 4 °C and irradiated for 5 min on ice with UV light.Subsequent preparation of the membrane-enriched fractionwas performed as above described for cells treated withTID-GM1.

2D Gel Electrophoresis of the Membrane-enriched Frac-tions—The membrane-enriched fractions were subjected tolipid extraction (25). The delipidized pellet was solubilized in

Membrane Tubulin-GM1 Interaction 9979

at UN

IV D

EG

LI ST

UD

I DI M

ILAN

on January 5, 2007 w

ww

.jbc.orgD

ownloaded from

a proper buffer (39), and 2D gel analysis was performed usingthe Bio-Rad Mini-protean II 2D system according to the man-ufacturer’s recommendations, except for the first dimensionmixture, which was prepared as described (39). 10% gels wereused for the second dimension. Radioactive spots were de-tected using Phosphorus Imager (Bio-Rad) and then sub-jected to autoradiography. The total protein pattern wasassessed by silver staining. 30 mg (as protein) were used foreach sample.

Immunoprecipitation Experiments—Aliquots of the mem-brane-enriched fraction were suspended at 37 °C for 20 minin 500 ml of lysis buffer (25 mM Tris-HCl, pH 7.4; 150 mM

NaCl; 1% Triton X-100; 0.1% gelatin; and 0.1% chymostatin,leupeptin, antipapain, and pepstatin mixture). The lysatewas successively incubated at 4 °C for 2 h with mouse anti-a,anti-b, or anti-acetyl-tubulin (1:200) antibodies. Protein A-Sepharose, preincubated with rabbit anti-mouse IgG for 2 h,was added, and the incubation prolonged for additional 2 h at4 °C. The protein A-Sepharose beads, recovered by centrifu-gation (800 3 g for 5 min, three times) were heated at 100 °Cfor 5 min in Laemmli buffer containing 1% 2-mercaptoetha-nol and centrifuged, and the supernatant was utilized for10% SDS-polyacrylamide gel electrophoresis, followed byautoradiography.

Preparation and Characterization of Detergent-resistantMembrane Fraction (DRF)—Cells (60 3 106), some after incu-bation with TID-GM1 or with TID-PC and illumination, wereharvested in Locke’s solution and incubated in 2 ml of 1%TritonX-100 in 25 mM MES buffer, pH 6.5, containing 150 mM NaCl,1 mM phenylmethylsulfonyl fluoride, and 75 units/ml aprotinin,for 30 min on ice. The cell lysate was subjected to discontinuoussucrose density gradient centrifugation, for the separation oflow density DRF, as described (40). 1.2-ml fractions were col-lected from the gradient and subjected to radioactivity count-ing. Fraction 4 (DRF) and high density membrane fraction(HDF) (pool of fractions 8–10) were assayed for lipid and pro-tein content, as follows.

Lipids—DRF and HDF were dialyzed against distilled waterat 4 °C and then lyophilized. Lipids were extracted according toRef. 41, with slight modifications, as described in Ref. 42. Thelipids of the organic phases were separated by high perform-ance TLC (solvent system chloroform/methanol/acetic acid/H2O, 60:45:4:2, v/v/v/v) and revealed with I2. For cholesterolvisualization, the extracted lipid samples were separated byhigh performance TLC (solvent system hexane/diethylether/acetic acid, 20:35:1, v/v/v) and then sprayed with anisaldehydereagent. Cholesterol was detected by heating the plate at180 °C for 15 min. For detection and quantification of GM1,TLC separation, blotting with horseradish peroxidase-labeledcholera toxin B subunit, ECL detection, and quantification ofGM1 were performed as described (9).

Proteins—DRF and HDF were subjected to trichloroaceticacid precipitation. The pellet washed with acetone was sus-pended in water for protein assay and then subjected to gelelectrophoresis (EF). For detection of proteins by Western blot-ting, samples were suspended in Laemmli buffer containing 1%2-mercaptoethanol, heated at 100 °C and resolved by SDS-polyacrylamide gel electrophoresis (10% acrylamide, 7 mg/lane)followed by Western blotting that was carried out as follows.Proteins were transferred to nitrocellulose membranes. Blotswere incubated overnight at 4 °C in TBST (20 mM Tris-HCl, pH7.5, 150 mM NaCl, 0.1% Tween 20) containing 5% (w/v) dryskimmed milk. After washing with TBST, membranes wereincubated at room temperature for 3 h with the primary anti-body diluted in TBST/milk (anti-a-tubulin, 1:500; anti-b-tubu-lin, 1:500; anti-acetyl-tubulin, 1:2000; anti-actin, 1:500; anti-

GAP-43, 1:500; anti-Fyn, 1:250; anti-Tau, 1:200; anti-Go

protein a-subunit, 1:1000) and then for 2 h with horseradishperoxidase-conjugated goat anti-rabbit/mouse IgG (5000–10,000-fold diluted in TBST/milk; Pierce). Proteins were de-tected by chemiluminescence using the SuperSignal detectionkit (Pierce).

In separate experiments, an aliquot (8 mg as protein) of thepellet obtained from fraction 4 (DRF) was subjected to 2D-EF,as described for the membrane-enriched fractions, and sub-jected to silver staining. Another aliquot, after 2D-EF, wastransferred to polyvinylidene difluoride membrane and sub-jected to Western blotting and ECL detection, as describedabove, with the exception that a mixture of anti-a-tubulin(1:500), anti-b-tubulin (1:500), and anti-acetyl-tubulin (1:2000)antibodies was utilized.

Hydroxylamine and KOH Treatment—The immunoprecipi-tate obtained from the membrane-enriched fraction using amixture of anti-a-, anti-b-, and anti-acetyl-tubulin (1:200)antibodies was divided into three identical aliquots and sub-jected to 10% SDS-polyacrylamide gel electrophoresis. Thethree lanes were cut from fresh gels and incubated for 2 h atroom temperature in 1 M Tris-HCl (pH 7.0), 1 M hydroxyla-mine (pH 7.0), or 0.2 M KOH in methanol, respectively, asdescribed (43) and then dried and subjected altogether toautoradiography.

Assessment of the Specific Radioactivity of Tubulin in theMembrane-enriched and Detergent-resistant Fraction—Twoaliquots of the membrane-enriched and of the detergent-resis-tant fraction were subjected to immunoprecipitation using amixture of anti-a-, anti-b-, and anti-acetyl-tubulin (1:200) an-tibodies. The immunoprecipitates were subjected to EF on thesame gel, transferred to the same polyvinylidene difluoridemembrane, and subjected to radioactivity quantification. Then,the polyvinylidene difluoride membrane was subjected to West-ern blotting using anti-tubulin antibody, followed by ECL de-tection and densitometric quantification. Finally, the specificradioactivity was calculated as the ratio of the radioactivityassociated with the tubulin band (directly related to theamount of TID-GM1 bound to the protein) to the digitized valueof the same band in the ECL film (related to the amount oftubulin as protein).

Sequence Determination and Computer Analysis—Auto-mated sequence analysis was performed on a pulsed-liquidsequencer (model 477, Applied Biosystems, Foster City, CA)equipped with a 120A Applied Biosystems PTH analyzer. Ho-mologies with the entries in the nonredundant GenBankTM

were searched using the BLAST program (44).

RESULTS

After cell incubation with TID-GM1 for 2 h at 37 °C, 145 62.5 pmol of probe/mg of protein were bound to granule cells.A parallel experiment performed with isotopically labeled[3H]GM1 (45) gave very similar figures (138 6 1.9 pmol/mg ofprotein), indicating that the association of the two moleculeswas comparable. After illumination of cells with UV light,inducing cross-linking of TID-GM1 with neighboring mole-cules, a membrane-enriched fraction was prepared. Autora-diography of the 2D-EF of this fraction (Fig. 1A) showed thepresence of a relatively small number of radiolabeled pro-teins, compared with those present in a 2D-EF after silverstaining (Fig. 1C). The main radioactive spots correspondedto proteins with masses of about 55 and 30 kDa, with pIvalues of about 5.

The experiment was repeated carrying out the incubation at4 °C. After 4 h of incubation, 20 6 0.18 pmol of TID-GM1/mg ofprotein were bound to granule cells. After photoactivation, themembrane-enriched fraction was prepared, and 2D-EF was

Membrane Tubulin-GM1 Interaction9980

at UN

IV D

EG

LI ST

UD

I DI M

ILAN

on January 5, 2007 w

ww

.jbc.orgD

ownloaded from

carried out. A lesser number of radiolabeled proteins (Fig. 1B)were detected in comparison with the experiment carried out at37 °C. The major radioactive spots were again corresponding toproteins of about 55- (from now on called P55) and 30-kDamass. A nearly identical radiolabeling pattern was observedwhen cell incubation with TID-GM1 was carried out at 4 °C inthe presence of inhibitors of endocytosis, nocodazole, or colchi-cine (data not shown).

In the 2D-EF of the membrane-enriched fraction after silverstaining, a major spot, apparently due to at least two proteinspartially overlapping and having electrophoretic characteris-tics similar to P55, was present. This major spot, after transferto PDVF membrane, elution, and amino acid sequencing,showed its identity with a- and b-tubulin, suggesting that P55could correspond to the product of cross-linking between TID-GM1 and tubulin.

In order to verify this hypothesis, immunoprecipitation ex-periments were performed. For this purpose, the membrane-enriched fraction, prepared after cell incubation with TID-GM1and photoactivation at 4 °C, was lysed at 37 °C and immuno-precipitation carried out using monoclonal antibodies againsta-, b-, or acetyl-tubulin. As shown in Fig. 2, the radioactive

protein P55 was present in the immunoprecipitates obtainedwith any of the three antibodies. Using preimmune seruminstead of anti-tubulin antibody and submitting the cells to theimmunoprecipitation procedure, P55 was undetectable in theimmunoprecipitate under the same conditions.

In order to test the specificity of interaction between TID-GM1 and P55, parallel experiments were performed, in whichTID-PC was used instead of the photoactivable ganglioside.After incubation at 4 °C for 2 h, 1 6 0.2 pmol of phospholip-id/mg of protein were bound to cells. The pattern of radiola-beled proteins was different from that obtained with TID-GM1(the autoradiography of the 2D-EF of the membrane-enrichedfraction is reported in Fig. 3).

In further experiments, intact living cells were incubatedwith TID-GM1 and, after photoactivation, treated with trypsin.The membrane-enriched fraction was prepared, and 2D-EFwas performed. As reported in Fig. 4, P55 was still presentafter this treatment. Immunoprecipitation experiments usingmonoclonal antibodies (Fig. 4B) demonstrated that the maintubulin isoforms present were a and b.

Successively, the susceptibility of P55 to NH2-OH or to meth-anolic KOH treatment was assessed in order to establishwhether or not the cross-linking is occurring with a lipid an-chor of the protein (43). The results, reported in Fig. 5, showthat NH2-OH treatment did not cause substantial loss of ra-dioactivity from P55, whereas approximately 90% of radioac-tivity was released after treatment with KOH. No effect wasexerted by a mere treatment with methanol.

Further experiments were performed in order to investigatethe involvement of DRF in the interaction between GM1 andtubulin. Cells were treated with Triton X-100 at low tempera-ture and spun on a sucrose gradient, and 1.2-ml fractions of thegradient were assayed for radioactivity, protein content, andlipid content. The low density fraction (fraction 4), located atthe 5–30% sucrose interface and corresponding to the DRF (40),displayed the highest GM1 enrichment, almost mirroring thedistribution of TID-GM1 associated to the cells (Fig. 6A). Onthe contrary, TID-PC was mostly localized outside the DRF,within the HDF.

A comparison between DRF and HDF showed a higher pro-portion of sphingomyelin in the detergent resistant-fraction(Fig. 6B). Moreover, DRF was also enriched in cholesterol (Fig.6C). Western blotting experiments showed that the DRF, al-

FIG. 1. Detection of the proteins cross-linked by the photoac-tivable radioactive GM1 ganglioside analog TID-GM1. Culturedrat cerebellar granule cells were incubated with TID-GM1. After cross-linking induced by illumination with UV light, a membrane-enrichedfraction was prepared, and 2D electrophoresis carried out. A, autora-diography of the electrophoresis obtained after incubation of TID-GM1for 2 h at 37 °C; B, autoradiography of the electrophoresis obtainedafter incubation of TID-GM1 for 4 h at 4 °C; C, silver staining of theelectrophoresis shown in B. The arrow shows the direction of the iso-electric focusing (IEF) or the direction of PAGE. pI, isoelectric point. pHrange is 4–8.

FIG. 2. Assessment of the presence of tubulin among the pro-teins cross-linked by the photoactivable radioactive GM1 gan-glioside analog TID-GM1. Cells were incubated for 4 h at 4 °C withTID-GM1, and after illumination with UV light, the membrane-en-riched fraction was prepared. Three identical aliquots of the fractionwere subjected to immunoprecipitation using anti-a-, anti-b-, or anti-acetyl-tubulin antibodies. The autoradiography of the electrophoresis ofthe immunoprecipitates obtained with anti-a-, anti-b-, or anti-acetyl-tubulin antibodies are shown in lanes A, B, and C, respectively.

Membrane Tubulin-GM1 Interaction 9981

at UN

IV D

EG

LI ST

UD

I DI M

ILAN

on January 5, 2007 w

ww

.jbc.orgD

ownloaded from

ways compared with the HDF, was enriched in signal trans-duction-related molecules such as Fyn, GAP-43, and Ga (Fig.7). The bulk of cytoskeleton markers (tubulin, actin, and Tau)were present in the HDF, as expected. Only faint bands werevisible in the DRF, corresponding to the a- and acetyl isoformsof tubulin, and even fainter bands, corresponding to b-tubulinand to actin, whereas Tau was not detected in this fraction (Fig.7).

On the other hand, a comparison of the DRF (Fig. 8A) andthe membrane-enriched fraction (Fig. 1C) showed that the pro-tein pattern, detected by silver staining, was much simpler in

the detergent resistant fraction. The presence of tubulin in thisfraction was confirmed by Western blotting (Fig. 8B). Simplecalculations based upon densitometric quantification of theprotein bands visible in the two gels indicated that the propor-tion of tubulin over the total proteins in the DRF (about 18%)was higher than in the membrane-enriched fraction (about7%). Next, the specific radioactivity of tubulin in the mem-brane-enriched and in detergent-resistant fraction was as-sessed, relying on the amount of radioactivity associated andon the amount of tubulin (as protein) in each fraction. Controlexperiments showed that under the conditions required fortubulin detection by ECL, the film utilized was not sensitizedby the low amount of radioactivity associated to the protein.

FIG. 3. Detection of the proteins cross-linked by the photoac-tivable radioactive phosphatidylcholine analog TID-PC. Cul-tured rat cerebellar granule cells were incubated with TID-PC for 2 h at4 °C. After cross-linking induced by illumination with UV light, a mem-brane-enriched fraction was prepared, and 2D electrophoresis was car-ried out. The autoradiography of the electrophoresis is shown.

FIG. 4. Assessment of the resistance to trypsin treatment oftubulin cross-linked by the photoactivable radioactive GM1ganglioside analog TID-GM1. Cells were incubated for 4 h at 4 °Cwith TID-GM1, illuminated with UV light, and treated with trypsin,and the membrane-enriched fraction was prepared. Three identicalaliquots of the membrane-enriched fraction were subjected to immuno-precipitation using anti-a-, anti-b-, or anti-acetyl-tubulin antibodies. A,autoradiography of the electrophoresis obtained after carrying out theincubation with TID-GM1 for 4 h at 4 °C, followed by trypsin treatment.B, autoradiography of the electrophoresis of the immunoprecipitatesobtained with anti-a-tubulin (lane A), anti-b-tubulin (lane B), or anti-acetyl-tubulin (lane C) antibodies.

FIG. 5. Susceptibility to hydroxylamine or KOH treatment oftubulin cross-linked with the photoactivable radioactive GM1ganglioside analog, TID-GM1. Cells were incubated for 4 h at 4 °Cwith the photoactivable radioactive GM1 ganglioside analog TID-GM1and illuminated with UV light, and the membrane-enriched fractionwas prepared. The immunoprecipitate obtained from the membrane-enriched fraction using anti-tubulin antibodies was divided into threeidentical aliquots and subjected to electrophoresis on the same gel. Thethree lanes were cut from the gel and incubated in 1 M Tris-HCl, pH 7.0(control), 1 M hydroxylamine, or 0.2 M KOH in methanol, respectively.After incubation, the gel was reassembled and subjected to autoradiog-raphy. Lane A, control sample; lane B, sample treated with hydroxyla-mine; lane C, sample treated with KOH.

FIG. 6. Characterization of detergent-resistant fraction pre-pared from cerebellar granule cells: content of endogenous lip-ids and distribution of radioactivity in fractions withdrawnfrom the sucrose gradient. Cells, after incubation with either TID-GM1 or TID-PC and cross-linking induced by illumination with UVlight, were treated with 1% Triton X-100-containing buffer for 30 minon ice. The suspension was subjected to discontinuous sucrose densitygradient centrifugation. 1.2-ml fractions were withdrawn from the gra-dient and subjected to GM1 determination and radioactivity counting(A). M, GM1; E, TID-GM1; ‚, TID-PC. B, the phospholipid pattern ofDRF and HDF was assessed by submitting their lipid extracts to TLC.SM, sphingomyelin; PC, phosphatidylcholine; PS, phosphatidylserine;PE, phosphatidylethanolamine. C, cholesterol (CHL) content was as-sessed by submitting to TLC the lipids extracted from the same amountof DRF and HDF (as protein).

Membrane Tubulin-GM1 Interaction9982

at UN

IV D

EG

LI ST

UD

I DI M

ILAN

on January 5, 2007 w

ww

.jbc.orgD

ownloaded from

The specific radioactivity of tubulin was 2.1 arbitrary units inthe DRF and 0.28 arbitrary units in the membrane-enrichedfraction.

DISCUSSION

In the present investigation the GM1 ganglioside analog,TID-GM1, carrying a photoactivable diazirine group and la-beled with 125I in the ceramide moiety, has been utilized for thedetection of interacting proteins in cultured neurons. The mainfeatures of these analogs are the ability to form covalent bondswith neighboring proteins upon photoactivation, their ex-tremely high specific radioactivity, and, in the absence of UVlight, their remarkable stability under a range of differentchemical and physical conditions (32). After incubation of cer-ebellar granule cells with TID-GM1 at 37 °C and photoactiva-tion, a series of proteins become radiolabeled, indicating theirproximity to the probe. Clues about the localization of theseproteins were obtained performing incubation and successivephotoactivation at 4 °C. In fact, at this temperature, ganglio-side internalization is prevented (46–48), and therefore TID-GM1/protein cross-linking at the plasma membrane is privi-

leged. A protein (P55) with a mass of about 55 kDa and anotherof about 30 kDa and a pI of about 5 were radiolabeled under allof the experimental conditions adopted. A further confirmationof their cross-linking at the plasma membrane was obtained byexperiments carried out with TID-GM1 in the presence of in-hibitors of endocytosis (34, 35) and at low temperature, exclud-ing the possibility that part of the probe is endocytosed andinteracts with intracellular proteins.

Our attention was attracted by P55 because it displayedelectrophoretic features similar to those of tubulin, present asa major spot and recognized by amino acid sequencing in theendogenous protein pattern of the membrane-enrichedfraction.

Immunoprecipitation and Western blotting experimentsshowed that P55 is indeed the product of cross-linking betweenthe photoactivated ganglioside and a-, b-, and acetylated tubu-lin isoforms. The existence of a covalent bond between the twomolecules, typical of photoreagents of this class (32), is furtherstrengthened by the following considerations: (a) 2D electro-phoresis and autoradiography were carried out on membrane-enriched fractions after delipidization, removing lipids that arenot covalently bound. In addition, under the EF conditions,unbound lipids dissociate from proteins and run at the front (asdiscussed in Ref. 17). (b) Tubulin immunoprecipitation wascarried out on membrane-enriched fractions after lysis in de-tergent-containing buffer at 37 °C. At this temperature, mem-brane domains, detergent-resistant only at low temperature(10, 40, 56), are disrupted. Therefore, immunoprecipitation ofmembrane rafts, containing tubulin along with noncovalentlybound TID-GM1 or other cross-linked proteins, is prevented,reducing the possibility of artifacts.

Taken together, these results suggest that TID-GM1 cross-links tubulin at the plasma membrane of cerebellar granulecells. Even if microtubules, in which tubulin is the main com-ponent, are intracellular structures and gangliosides are typi-cal membrane components, their apparent interaction has anexplanation in the long known existence of membrane-associ-ated tubulin (49–53). However, the topology of membrane tu-bulin is not completely understood. For instance, it has beenpostulated that tubulin-like, trypsin-sensitive proteins can bepresent at the cell surface in neurons (54). It has also beensuggested that membrane-associated tubulin is an integralmembrane protein (50, 51). Other investigations claim that itshydrophobicity arises from the interaction with other mem-brane components (52). Recently, Caron (53) showed that tu-bulin is palmitoylated and ascribed its hydrophobic behavior atleast in part to this feature.

Experiments with trypsin, to which P55 was resistant, on theone hand ruled out that cross-linking was occurring with tu-bulin at the plasma membrane surface, and on the other handgave additional information. In fact, cell-associated exogenousgangliosides that are not removed by trypsin treatment (theso-called trypsin-stable form of ganglioside association withcells) are considered to be correctly inserted with the ceramidemoiety in the hydrophobic core of the bilayer (55). Because theganglioside photoactivable group is located at the end of theceramide fatty acid moiety, these results indicate that gangli-oside-tubulin interaction is taking place within the hydropho-bic core of the plasma membrane bilayer.

Therefore, we inspected the possibility that TID-GM1 wascross-linked with a lipid anchor of tubulin. In this case, arelease of radioactivity from P55 would occur after treatmentwith hydroxylamine or KOH, which are able to selectivelyremove different protein lipid anchors (43). On the contrary, ifcross-linking was occurring with an amino acid residue of theprotein, the radioactivity would be retained. The experiments

FIG. 7. Characterization of proteins in detergent-resistantfractions prepared from cerebellar granule cells. DRF and HDF(fraction 4 and a pool of fractions 8–10 of the gradient described in thelabel to Fig. 6, respectively), were separated on 10% SDS-polyacryl-amide gel electrophoresis (7 mg of protein/lane), transferred to nitrocel-lulose membranes, and immunoblotted with the indicated antibody.

FIG. 8. Characterization of proteins in detergent-resistantfractions prepared from cerebellar granule cells. DRF (fraction 4of the gradient described in the label to Fig. 6) was subjected to 2D-EFfollowed by silver staining (A) or by Western blotting using anti-tubulinantibodies (B).

Membrane Tubulin-GM1 Interaction 9983

at UN

IV D

EG

LI ST

UD

I DI M

ILAN

on January 5, 2007 w

ww

.jbc.orgD

ownloaded from

showed that the radioactivity was released only by KOH, con-sistent with the hypothesis that upon photoactivation, TID-GM1 is cross-linked with a fatty acid anchor that is linked withan ester-linkage to membrane-associated tubulin.

As a further step, we investigated the possible involvementof specialized domains of the neuronal plasma membrane. Infact, in recent years, substantial progress has been made inunderstanding the biological role of glycolipid-enriched, de-tergent-resistant membrane domains in eukaryotic cells (11,35, 56). However, the neuron, which perhaps represents oneof the greatest challenges to research on membrane trafficand function, has only been partially investigated (13, 42,57). Therefore, we prepared DRF from a lysate of cerebellargranule cells after incubation with TID-GM1 and cross-link-ing. This fraction was found to be enriched in endogenousGM1 (and TID-GM1), sphingomyelin, and cholesterol, inanalogy with other cellular systems (10, 11, 13). Also signaltransduction molecules were enriched in DRF, again in anal-ogy with other neuronal and nonneuronal cell types (10–12,42, 58). Although the bulk of cytoskeleton proteins werelocalized within the HDF, some of them were also detected intrace amounts in DRF. This finding is somewhat expected,because the presence of actin, in particular, in membranedomains has been already reported (11, 59). However, theneuronal microtubule-associated protein Tau (60) was notdetected in DRF, reducing the possibility of a mere contam-ination of DRF by the cytoskeleton. According to the abovereported results, tubulin in DRF is present in very smallamounts in comparison with HDF, which contains the bulk ofcell tubulin. On the contrary, upon comparison of DRF withmembrane-enriched fractions (which contain all of the tubu-lin associated to the membrane), it turns out that membranetubulin is enriched in detergent-resistant membranedomains.

Next, we assessed the presence of radioactive, GM1-cross-linked, tubulin within DRF. Radioactive tubulin was de-tected, indicating that ganglioside-enriched detergent-resis-tant domains do contain lipid-anchored tubulin. The presenceof lipid-anchored tubulin in DRF could explain why cell treat-ment with nocodazole does not affect the protein cross-link-ing with TID-GM1; in fact, it is conceivable that tubulinmolecules associated to the membrane with their lipid anchorare insensitive to this microtubule-depolimerizing drug.However, the evaluation of this hypothesis deserves furtherinvestigation.

The specificity of the interaction between tubulin and TID-GM1 could be debated. In this particular case, at least twotypes of specificity may be involved: (a) tubulin interaction withganglioside with respect to other lipids, and (b) TID-GM1 in-teraction with DRF tubulin in comparison with tubulin in thebulk membrane. Concerning the first type, experiments werecarried out with the phosphatidylcholine analogue TID-PC.TID-PC was localized in the bulk membrane, and cross-linkingexperiments suggested that tubulin interacts specifically withGM1. Concerning the second type of interaction (of GM1 withDRF tubulin in comparison with tubulin in the bulk mem-brane) the higher specific radioactivity in DRF suggests thatTID-GM1 is more specific toward tubulin associated with do-mains. However, this latter interaction is unlikely to be drivenby a mutual recognition, because it is occurring between lipids(ganglioside ceramide moiety and palmitoyl residues of theprotein). Instead, the apparent interaction likely depends onco-segregation and enrichment of both molecules—in particu-lar ganglioside—within the same domain. In this sense, wepropose to name it a “domain-specific interaction.” In this view,TID-GM1 is instrumental to ascertaining which proteins are

present within DRF; in this case, it detects tubulin anchoringto glycolipid-enriched detergent resistant domains of the neu-ronal plasma membrane.

The functional implications of the presence of lipid-an-chored tubulin within detergent-resistant domains could bedebated. First, it is likely that such localization is importantfor structural remodeling of the plasma membrane, as itoccurs during cell mitosis or axon extension in neurons. Gly-colipid-enriched domains could play a key role as sites wheremicrotubules, via lipid-anchored tubulin, come into contactwith the plasma membrane and contribute to physically driv-ing its changes. Second, the presence of lipid-anchored tubu-lin within DRF could play a role in signal transduction. Infact, domains are enriched in signal-transducing molecules,G-protein families included (10–12), and it is known thattubulin is involved in G-protein-mediated signal transduc-tion in a variety of systems (61). The interaction betweenthese two protein families inside domains can affect signaltransduction. GM1-tubulin interactions may participate inthe modulation of this process.

In conclusion, herein we have shown that some tubulinmolecules are associated with a lipid anchor to detergent-resistant glycolipid-enriched domains of the plasma mem-brane. This novel feature of glycolipid-enriched domains in-creases the evidence of their proven or postulatedparticipation in a series of important cell functions (10–12)and provides a key for a better understanding of their bio-logical role in the plasma membrane.

REFERENCES

1. Karlsson, K.-A. (1986) Chem. Phys. Lipids. 42, 153–1722. Hakomori, S.-I. (1981) Annu. Rev. Biochem. 50, 733–7643. Bremer, E. G., Hakomori, S., Bowen-Pope, D. G., Raines, E., and Ross, R.

(1984) J. Biol. Chem. 259, 6818–68254. Fishman, P. H. (1988) in New Trends in Ganglioside Research: Neurochemical

and Neurodegenerative Aspects (Ledeen, R. W., Hogan, E. L., Tettamanti,G.,. Yates, A. J., and Yu, R. K., eds) Fidia Research series, Vol. 14, pp.183–201, Liviana Press, Padova

5. Hakomori, S.-I. (1990) J. Cell Biol. 265, 18713–187166. Rebbaa, A., Hurh, J., Yamamoto, H., Kersey, D. S., and Bremer, E. G. (1996)

Glycobiology 6, 399–4067. Fueshko, S. M., and Schengrund, C.-L. (1992) J. Neurochem. 59, 527–5358. Tiemeyer, M., Yasuda, Y., and Schnaar, R. L. (1989) J. Biol. Chem. 264,

1671–16819. Pitto, M., Mutoh, T., Kuriyama, M., Ferraretto, A., Palestini, P., and

Masserini, M. (1998) FEBS Lett. 439, 93–9610. Simons, K., and Ikonen, E. (1997) Nature 387, 569–57211. Anderson, R. G. W. (1998) Annu. Rev. Biochem. 67, 199–22512. Lisanti, M. P., Scherer, P. E., Tang, Z., and Sargiacomo, M. (1994) Trends Cell

Biol. 4, 231–23513. M. Masserini, P. Palestini, and M. Pitto (1999) J. Neurochem. 73, 1–1114. Kasahara, K., Watanabe, Y., Yamamoto, T., and Sanai, Y. (1997) J. Biol.

Chem. 272, 29947–2995315. Iwabuchi, K., Handa, K., and Hakomori, S. (1998) J. Biol. Chem. 273,

33766–3377316. Iwabuchi, K., Yamamura, S., Prinetti, A., Handa, K., and Hakomori, S. (1998)

J. Biol. Chem. 273, 9130–913817. Fra, A., Masserini, M., Palestini, P., Sonnino, S., Simons, K. (1995) FEBS Lett.

375, 11–1418. Sonnino, S., Chigorno, V., Acquotti, D., Pitto, M., Kirschner, G., and

Tettamanti, G. (1989) Biochemistry 28, 77–8419. Sonnino, S., Chigorno, V., Valsecchi, M., Pitto, M., and Tettamanti, G. (1992)

Neurochem. Int. 20, 315–32120. Garner, J., Durrer, P., Kitchen, J., Brunner, J., and Crooke, E. (1998) J. Biol.

Chem. 273, 5167–517321. Eicher, J., Brunner, J., and Wickener, W. (1997) EMBO J. 16, 2188–219622. Shapiro, R. E., Specht, C. D., Collins, B. E., Woods, A. S., Cotter, R. J., and

Schnaar, R. L. (1997) J. Biol. Chem. 272, 30380–3038623. Kuhn, C. S., Lehmann, J., and Sandhoff, K. (1992) Bioconjugate Chem. 3,

230–23324. Tettamanti, G., and Riboni, L. (1994) in Progress in Brain Research

(Svennerholm, L., Asbury, A. K., Reisfeld, R. A., Sandhoff, K., Suzuki, K.,and Tettamanti, G., eds) Vol. 101, pp. 77–100, Elsevier Science BV,Amsterdam

25. Tettamanti, G., Bonali, F., Marchesini, S., and Zambotti, V. (1973) Biochim.Biophys. Acta 296, 160–170

26. Ghidoni, R., Sonnino, S., Masserini, M., Orlando, P., and Tettamanti, G. (1982)J. Lipid. Res. 22, 1286–1295

27. Sonnino, S., Ghidoni, R., Gazzotti, G., Kirschner, G., Galli, G., and Tettamanti,G. (1984) J. Lipid. Res. 25, 620–629

28. Gallo, V., Ciotti, M. T., Coletti, A., Aloisi, F., and Levi, G. (1982) Proc. Natl.

Membrane Tubulin-GM1 Interaction9984

at UN

IV D

EG

LI ST

UD

I DI M

ILAN

on January 5, 2007 w

ww

.jbc.orgD

ownloaded from

Acad. Sci. U. S. A. 79, 7919–792329. Vaccarino, F., Guidotti, A., and Costa, E. (1987) Proc. Natl. Acad. Sci. U. S. A.

84, 8707–871130. Favaron, M., Manev, H., Alho, H., Bertolini, M., Ferret, B., Guidotti, A., and

Costa, E. (1988) Proc. Natl. Acad. Sci. U. S. A. 85, 7351–735531. Lowry, O. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. (1951) J. Biol.

Chem. 193, 265–27532. Weber, T., and Brunner, J. (1995) J. Am. Chem. Soc. 117, 3084–309533. Sonnino, S. Acquotti, D., Kirschner, G., Uguaglianza, A., Zecca, l., Rubino, F.,

and Tettamanti, G. (1992) J. Lipid. Res. 33, 1221–122634. Gundersen, G. G., and Bulinski, J. C. (1988) Proc. Natl. Acad. Sci. U. S. A. 85,

5946–595035. Lu, S. C., Garcia-Ruiz, C., Kuhlenkamp, S., Ookhtens M., Salas-Prato, M., and

Kaplowitz, N. (1990) J. Biol. Chem. 265, 16088–1690536. Chigorno, V., Pitto, M., Cardace, G., Acquotti, D., Kischner, G., Sonnino, S.,

Ghidoni, G., and Tettamanti, G. (1985) Glycoconjugate J. 2, 279–29137. Harder, T., Scheiffele, P., Verkade, P., and Simons, K. (1998) J. Cell Biol. 141,

929–94238. Caimi, L., Marchesini, S., Aleo, M. F., Bresciani, R., Monti, E., Casella, A.,

Giudici, M. L., and Preti, A. (1989) J. Neurochem. 52, 1722–172839. Dessi, F., Ben-Ari, Y., and Charriaut-Marlangue, C. (1994) Brain Res. 654,

27–3340. Sargiacomo, M., Sudol, M., Tang, Z.-L., and Lisanti, M. P. (1993) J. Cell Biol.

122, 789–80741. Bligh, E. G., and Dyer W. J. (1959) Can. J. Biochem. Physiol. 37, 911–91742. Wu, C., Butz, S., Ying, Y., and Anderson, R. G. W. (1997) J. Biol. Chem. 272,

3554–355943. Bizzozero, O. A., Tetzloff, S. U., and Bharadwaj, M. (1994) Neurochem. Res. 19,

923–93344. Altschul, S. F., Gish, W., Miller, W., Myers, E. W., Lipman, D. J. (1990) J. Mol.

Biol. 215, 403–41045. Schwarzmann, G., and Sandhoff, K. (1990) Biochemistry 29, 10865–1087146. Parton, R. G. (1994) J. Histochem. Cytochem. 42, 155–16647. Sofer, A., Schwarzmann, G., and Futerman, A. H. (1996) J. Cell Sci. 109,

2111–211948. Riboni, L. Prinetti, A., Bassi, R., and Tettamanti, G. (1991) FEBS Lett. 287,

42–4649. Stephens, R. E. (1986) Biol. Cell 57, 95–11050. Zisapel, N., Levi, M., and Gozes, I. (1980) J. Neurochem. 43, 26–3251. Babich, J. A. (1981) J. Neurochem. 37, 1394–140052. Nunez-Fernandez M., Beltramo D. M., del C. Alonso, A., and Barra, H. S.

(1997) Mol. Cell. Biochem. 170, 91–9853. Caron, J. M. (1997) Mol. Biol. Cell 8, 621–63654. Estrige, M. (1977) Nature 268, 60–6355. Saqr, E. H., Pearl, D. K., and Yates, A. J. (1993) J. Neurochem. 61, 395–41056. Hakomori, S., Handa, K., Iwabuchi, K., Yamamura, S., and Prinetti, A. (1998)

Glycobiology. 8, XI–XIX57. Gorodinsky, A., and Harris, D. A. (1995) J. Cell Biol. 129, 619–62758. Oehrlein, S. A., Parker, P. J., and Herget, T. (1996) Biochem. J. 317, 219–22459. Hope, H. R., and Pike, L. J. (1996) Mol. Biol. Cell 7, 843–85160. Kaech, S., Ludin, B., and Matus, A. (1996) Neuron 17, 1189–119961. Ravindra, R. (1997) Endocrine 7, 123–143

Membrane Tubulin-GM1 Interaction 9985

at UN

IV D

EG

LI ST

UD

I DI M

ILAN

on January 5, 2007 w

ww

.jbc.orgD

ownloaded from