Journal of Neurochemistry Raven Press, Ltd., New York 0 199 1 International Society for Neurochemistry

Exogenous Gangliosides GD 1 b and GD 1 b-Lactone, Stably Associated to Rat Brain P2 Subcellular Fraction, Modulate

Differently the Process of Protein Phosphorylation

Rosaria Bassi, Vanna Chigorno, Amelia Fiorilli, Sandro Sonnino, and Guido Tettamanti

Study Center for the Functional Biochemistry of Brain Lipids, Department of Medical Chemistry and Biochemistry, The Medical School, University of Milan, Milan, Italy

Abstract: GDI b and GD 1 b-lactone (GD 1 b-L) gangliosides bind to the same extent to a Pp crude membrane preparation from rat brain. After 30 min of incubation with and M solutions of ganglioside, 1,800, 450, and 100 pmol of ganglioside/mg of protein, respectively, were found to be stably associated to the P2 fraction. This association modifies the phosphorylation process of the P2 membrane proteins in a dose-dependent manner, the maximal effect being reached at a ganglioside association of I .85 nmol/mg of protein and in large part at 450 pmol/mg of protein. The effects of GD 1 b and GD 1 b-L on the phosphorylation of five proteins, showing apparent molecular masses of 17, 20, 36, 41, and 44 kDa, were different after 0.5 min of phosphory- lation reaction as well as after 15 min. After 0.5 rnin of re-

action, in the presence of stably associated GDlb, the phos- phorylation of the 36-, 4 I-, and 44-kDa proteins was increased with reference to the control, whereas the phosphorylation of the 17- and 20-kDa proteins was decreased. GDlb-L ex- erted qualitatively similar effects only on the 44-, 41-, and 36-kDa proteins and to a strongly reduced degree. After 15 rnin of reaction, only the phosphorylation of the 36-kDa pro- tein was stimulated by GD I b; G D 1 b-L exerted a similar ef- fect, but to a low degree. Key Words: Gangliosides-Gan- glioside-lactones-Protein phosphorylation. Bassi R. et al. Exogenous gangliosides GD1 b and GD 1 b-lactone, stably as- sociated to rat brain P2 subcellular fraction, modulate differ- ently the process of protein phosphorylation. J. Neurochem. 57, 1207-1211 (1991).

Gangliosides, sialic acid-containing glycosphingoli- pids, are components of the external lipid layer of cell plasma membranes of vertebrates and are abundant in the nervous system (Wiegandt, 1985). Gangliosides occur in small amounts as lactones in the brains of rodents (Sonnino et al., 1983) and in more abundant quantities in some cerebral areas of aged humans (Ri- boni et al., 1986). One of them, the monolactone de- rivative of GD 1 b, GD 1 b-monolactone (GD 1 b-L), has been characterized completely (Riboni et al., 1986; Acquotti et al., 1987), and the process of its in vitro formation from GD 1 b has been studied in detail (Bassi et al., 1989).

It is assumed that gangliosides participate in the pro- cess of signal transduction through the membrane (Fishman, 1988) and that they play a crucial role in

the regulation of differentiation and proliferation pro- cesses (Hakomori, 198 1 ; Ledeen, 1984). In some cases their action is associated with the modulation of mem- brane-linked enzymes (Partington and Daly, 1979; Davis and Daly, 1980; Leon et al., 1981; Goldenring et al., 1985; Chan, 1987a,b; Kreutter et al., 1987).

The present article presents the effects of exogenous ganglioside GD 1 b and the lactone derivative GD 1 b-L on the phosphorylation of proteins from rat brain crude membranes.

MATERIALS AND METHODS

Reagents Commercial chemicals were the purest available, common

solvents were distilled, and water was doubly distilled in a

Received September 17, 1990; revised manuscript received Feb- ruary 18, 199 1 ; accepted March 1, 199 1.

Address correspondence and reprint requests to Dr. S. Sonnino at Dipartimento di Chimica e Biochimica Medica, Via Saldini 50, I- 20133 Milano, Italy.

Abbreviations used: Cer, ceramide; DTT, dithiothreitol; NeuSAc, N-acetylneuraminic acid nonulosamine, reduced Neu5Ac (-CHIOH instead of -COOH); PAGE, polyacrylamide gel electrophoresis; SDS,

sodium dodecyl sulfate. Ganglioside nomenclature is in accordance with Svennerholm ( 1980) and IUPAC-IUB Recommendations ( 1977, 1982): GDl b, I13(Neu5Ac)zGgOse4Cer,~-Gal-( 1 +3)-@-GalNAc- ( 1 +4)-[o1-Neu5Ac-(2+8)-a-NeuSAc-(2+3)]-p-Gal-( 1 +4)-P-Glc- ( 1 + 1 )-Cer; GD 1 bmonolactone, GD 1 b-L, B-Gal-( 1 +3)-@-GalNAc- (1 +4)-[o(-NeuSAc-(2+8,1+9)-a-Neu5Ac-(2+3)]-(3-Gal-( 1 +4)-p- Glc-( 1 + I )-Cer; GD 1 b-ol, P-Gal-( 1 -*3)-o-GalNAc-( 1 +4)-[a-non- ulosamine-(2+8)-a-Neu5Ac-(2+3)]-~-Gal-( 1 +4)-p-Glc-( 1 + 1 )-Cer.

1207

1208 R. BASSI ET AL.

glass apparatus. Silica gel precoated TLC plates (HPTLC, Kieselgel 60, aluminum sheets) were from Merck, F.R.G.; N-acetylneuraminic acid (Neu5Ac) and crystalline serum bovine albumin from Sigma, St. Louis, MO, U.S.A.; En3Hance spray from New England Nuclear, U.K.; [y- 32P]ATP (>5,000 Ci/mmol) was from Amersham, U.K. Ganglioside GDl b was extracted from pig brain (Tettamanti et al., 1973), purified, and characterized (Ghidoni et al., 1980). [3H]GD 1 b was prepared according to the galactose oxidase- tritiated sodium borohydride procedure (Gazzotti et al., 1984). GDlb-monolactone, GDlb-L, and [3H]GDlb-L were prepared from GD 1 b and [3H]GD 1 b, according to Acquotti et al. (1987). Standard GDIb-ol and [3H]GD1b-ol were pre- pared by reduction of GD 1 b-L and [ 3H]GD 1 b-L, respectively, and structurally characterized according to Fronza et al. (1988). Sprague-Dawley adult male rats, 100-125 g body weight, were from Charles River, Italy. Rats were killed by decapitation and the whole brains were immediately removed the forebrains were dissected and promptly submitted to pro- cedures for the preparation of the subcellular Pz fraction.

Preparation of rat brain crude mitochondrial fraction (P2 fraction)

The crude mitochondrial fraction, P2 fraction, was prepared from rat brains at 4°C according to Gray and Whittaker (1962), using 0.32 M sucrose, 20 mM Tris-HC1 (pH 7.4), 2 mM EDTA, 1 mMEGTA, and 0.5 Mdithiothreitol (DTT), as homogenizing buffer (buffer A).

Binding of GDlb and GDlb-L to the P2 fraction All the following operations were performed at 0-2°C.

Immediately after preparation, the P2 fraction (200 pg of pro- tein/ml) was suspended in buffer A (see above) without su- crose, containing GDlb or freshly prepared GDlb-L at a final concentration of lo-,, and M. Unlabeled gangliosides were used for the phosphorylation assay (see be- low), and tritiated gangliosides were used for the evaluation of the time course of ganglioside association to P2 membranes. After incubation, the suspension was centrifuged at 12,000 g for 10 min, to remove unbound ganglioside, and the pellet washed with 1 ml of the same buffer plus 0.6% albumin, to remove micelles of gangliosides weakly associated to the membranes (Radsak et al., 1982; Chigorno et al., 1985) and centrifuged as described above.

Quantification, extraction, and analyses of gangliosides associated to the P2 fraction

After incubation with [3H]GDlb or [3H]GDlb-L, one-half of the P2 fraction was dispersed in 1 M NaOH (4 mg of pro- tein/ml) and submitted to liquid scintillation counting; the remaining P2 was treated with NaBH, (10 mg/ml final con- centration), to reduce GDlb-L to GDlb-ol (Fronza et al., 1988), and submitted to ganglioside extraction according to Tettamanti et al. (1973). 3H-Gangliosides were then analyzed by TLC followed by fluorography and radiochromatoscan- ning.

Protein phosphorylation assay Endogenous protein phosphorylation in the P2 fraction was

determined after 30 min of incubation at 0-2°C in buffer A without sucrose in the presence or absence (controls) of lo-,,

Mgangliosides; then P2 proteins (200 pg) were suspended in the same buffer (50 pl final volume) containing 10 mM MgC12; the phosphorylation reaction was started by adding [T-~~PIATP to a final concentration of 10 pM and

and

carried out at 30°C. At 0.5-15 min of incubation, aliquots of the reaction mixture, containing 40 pg of protein, were quenched with 1% sodium dodecyl sulfate (SDS) and 5% 2- mercaptoethanol. Proteins were resolved on 10% SDS-poly- acrylamide gel electrophoresis (SDS-PAGE) according to Laemmli (1 970). The gels, after staining, destaining, and drying, were submitted to autoradiography. Quantification of 32P incorporation into proteins was achieved by excising the corresponding gel bands or gel portions and measuring the radioactivity by liquid scintillation counting.

TLC and fluorography Gangliosides and ganglioside den vatives were analyzed by

TLC using chloroform/methanol/C1.2% aqueous CaClz, 50: 42: 1 1 by vol. Radioactive gangliosides and ganglioside de- rivatives were located on the plates by fluorography. The plates were sprayed with a surface autoradiographic enhancer and exposed to films for different tiimes (24-48 h) at -80°C.

Colorimetric methods Gangliosides were assayed as bound sialic acid by the re-

sorcinol-HC1 method (Svennerholm, 1957; Miettinen and Takki-Luukkainen, 1959), pure Neu5Ac being used as the reference standard. Protein contenl was determined in 1 M NaOH-solubilized P2 pellet (4 mg of protein/ml) according to Lowry et al. (1 95 l), with bovine serum albumin used as the reference standard.

RESULTS AND DISCUSSION

Gangliosides, when exogenously administered to cell membranes, seem to regulate, with different effects, the activity of the cell kinases and phosphatases (Gold- enring et al., 1985; Kim et al., 1986; Chan, 1987a,b; Kreutter et al., 1987; Yates et al., 1989). On the basis of this knowledge, gangliosides that are widely ex- pressed in the outer lipid layer ofthe neural cell plasma membranes are considered as possible modulating molecules of phosphorylation-dephosphorylation of proteins, which is itself a regulatory system of many biological processes (Krebs and Beavo, 1979; Kennedy, 1983; Kuo et al., 1986). Moreover, gangliosides con- taining a disialosyl residue, such as GD1 b, can lactonize in the presence of protons (Bassi et al., 1989), the pro- cess modifying the ganglioside structure and confor- mation (Acquotti et al., 1987), and perhaps the mo- dalities of interaction occurring between gangliosides and membrane proteins.

Our specific interest was to verify the effect of gan- glioside GD 1 b and its monolactone derivative GD 1 b- L on the protein phosphorylation process after exog- enous administration to a P2 crude membrane prep- aration from rat brain. The phosphorylation reaction was carried out after incubation of the P2 fraction with ganglioside, followed by removal of all the ganglioside that was not stably associated with the membrane preparation. According to this procedure, a portion of the stably associated ganglioside was presumably in- serted into the membrane (Radsak et al., 1982; Chi- gorno et al., 1985). Therefore, ;any subsequent event observed thus could be correlated with the new mem- brane organization.

J. Neurnchem., Vol. 57, No. 4, 1991

EFFECTS OF GANGLIOSIDES ON PROTEIN PHOSPHOR YLA TION 1209

2000

n

10-514

10-6M

" 0 15' 30' 60

Incubation time (min)

FIG. 1. Time course of ganglioside stable association to P2 rat brain subcellular fraction. GD1 b (A); GDl b-L (0).

The time course of ganglioside association to rat brain PZ fraction is shown in Fig. 1. GD 1 b and GD1 b- L behaved similarly. This is not surprising if we recall that the two compounds contain the same ceramide portion, which is probably strongly involved in the ganglioside-membrane interactions. After 30 rnin of ganglioside-protein incubation, and at a ganglioside concentration in the incubation medium of lop5, and M, about 1,800, 450, and 100 pmol of gan- glioside/mg of protein were stably bound to the sub- cellular preparation.

As expected, GD 1 b associated to the P2 fraction did not undergo lactonization or any type of enzyme-as- sisted structural modification; therefore, under the ex- perimental conditions used, all the radioactivity ex- tracted from the NaBH4-treated subcellular fraction was carried by the GDlb ganglioside. Also, GDlb-L did not undergo metabolization and it was quite stable. After the above treatments, at up to 60 rnin of incu- bation, >96% of the total radioactivity was associated with GD 1 b-ol; the rest was associated with GD 1 b pro- duced from GDlb-L hydrolysis (Fig. 2).

The time course of the protein phosphorylation pro- cess was followed for up to 15 min, and we found that at this later time of reaction a large part of the radio- activity incorporated into the proteins at the beginning of process, about 60921, was removed (Fig. 3, lane 3 and 6). This, in good agreement with previous data (Chan,

FIG. 2. Autoradiography of gan- gliosides extracted from P2 rat brain subcellular fraction, after 30 rnin of incubation with lo-, M (lane 2). 1 0-5 M (lane 3), and 1 O-' M GDlb-L (lane 4) followed by reduction with NaBH,. Lane 1, standard GD1 b-ol and GD1 b.

1987a), is probably an indication that the process is regulated by two steps, first involving kinases and then phosphatases. We confirmed the capability of GD1 b to modify the extent of phosphorylation of some pro- teins, such as those showing on SDS-PAGE apparent molecular masses of 17, 20, 36, 41, and 44 kDa. After 0.5 min of reaction, the phosphorylation of the 17- and 20-kDa proteins was strongly inhibited by GDlb (Fig. 3, lane 4). In contrast, GD 1 b enhanced the phosphor- ylation of the 36-, 41-, and 44-kDa proteins (Fig. 3, lane 4). With prolongation of the reaction time for up to 15 min, the effects on the phosphorylation of the 17-, 20-, 4 1 -, and 44-kDa proteins disappeared, but a rather large incorporation of radioactivity into the 36- kDa protein was observed (Fig. 3, lane 7). This strongly changed the radioactive protein pattern, the 36-kDa protein becoming the major component of the total radioactive protein mixture. The quantitative data on the effects exerted by GDlb stably associated to P2 fraction on the phosphorylation of those proteins showing on SDS-PAGE apparent molecular masses of 17, 20, 36, 41, and 44 kDa are reported in Fig. 4. The extent of the effects exerted by GDlb on the phos-

FIG. 3. SDS-PAGE separation of protein from P2 rat brain sub- cellular fraction after association of 1,800 pmol of gangRosides/ mg of protein, and incubation with [y-32P]ATP. Coornassie Blue staining: lane 1, standard proteins; lane 2, protein pattern of P2 fraction. Autoradiographic detection: 0.5 min of phosphorylation: lane 3, control; lane 4, 30 rnin of preincubation with GD1 b 1 0-4 M; lane 5, 30 min of preincubation with GO1 b-L 1 O-, M; 15 rnin of phosphorylation: lane 6, control; lane 7, 30 rnin of preincubation with GD1 b lo-, M; lane 8, 30 min of preincubation with GD1 b-L 1 0 - ~ M.

f. Neurochem., Vol. 57, No. 4, 1991

1210

200

160 h - e 8 s 80-

120

- ul

G 40 c + o - a N 0

-40

-80

R. BASSI ET AL.

-

-

-

-

-

-

05 min - Reaction time

GDtb GDlb-L

1 A B C D E A B C D E

15 rnin- Reaction t ~ m e

GDlb GDlb-L which occurs when and if GDlb undergoes lactoni- zation producing GD 1 b-L, as postulated recently (Bassi et al., 1989). On the other hand, the small change in the natural ganglioside-lactone content, obtained when a 1 Op6 M solution of GD 1 b-L was incubated with pro- teins, was not able to modify the protein phosphory- lation process.

In conclusion, our results confirm the hypothesis that gangliosides participate in the modulation of membrane-linked enzymes. Our experimental model uses a membrane preparation and follows a total membrane process involving many kinases as well as phosphatases; therefore, it cannot suggest a specific in- teraction between gangliosides and one of these en- zymes, but it supports the idea that ganglioside lacton- ization could provide a further opportunity for a fine modulation of membrane processes.

11 A B C D E A B C D E

FIG. 4. Effects of gangliosides GD1 b and GD1 b-L stably associated to P2 rat brain subcellular fraction on the phosphorylation of the 44- (A), 41- (B), 36- (C), 20- (D), and 17- (E) kDa P2 proteins. Results are expressed as percent of increase (or decrease) as compared to control values taken for 100%. Ganglioside concentration in the incubation medium: closed column, 1 0-6 M; hatched column, 1 0-5 M; open column, 1 0-4 M.

phorylation process was dependent on the amount of ganglioside stably associated to the membrane, and was already near the maximum at a ganglioside membrane enrichment of about 450 pmol of gangliosidelmg of protein. This amount of ganglioside produces a small change in the original endogenous total ganglioside (800 pg of Neu5Ac/g wet tissue) and GDlb contents (20% on total lipid-bound sialic acid), which is about 0.9% and 4.35%, respectively. On the other hand, ex- periments performed with a photoreactive ganglioside (Sonnino et al., 1989) showed that gangliosides ad- ministered to cells and membrane preparations can become components of specific membrane microdo- mains containing proteins. In this case the increase in a specific part of the membrane of the number of gan- glioside molecules, also at a picomolar level, would correspond to a dramatic change of the total original ganglioside content.

The effects exerted by GDlb-L on the phosphory- lation of the P2 proteins were similar to those exerted by GDlb, but to a strongly reduced degree. The only difference is that after 0.5 min, associated GDIb-L did not modify the phosphorylation of the 17- and 20-kDa proteins (Fig. 3 , lane 5). The quantitative data of the effects exerted by GDlb-L on the phosphorylation of some P2 proteins are reported in Fig. 4.

GD 1 b-L ganglioside is a minor component of the rat brain ganglioside mixture, i.e., 0.7% of the total lipid bound sialic acid. Therefore, the exogenous in- sertion of hundreds of picomoles of the ganglioside- lactone into the rat brain crude membrane preparation produces a dramatic increase in the GDlb-L content. This new membrane composition would mimic that

Acknowledgment: This work was partially supported by a grant (89.00229.70) from the Consiglio Nazionale delle Ri- cerche (Progetto finalizzato Biotecnologie e Biostrumenta- zione), Rome, Italy.

REFERENCES

Acquotti D., Fronza G., Riboni L., Sonnino S., and Tettamanti G. (1987) Ganglioside lactones: 'H-NMR determination ofthe inner ester position of GD 1 b-ganglioside lactone naturally occumng in human brain or produced by chemical synthesis. Glycocon- jugute J. 4, 119-127.

Bassi R., Riboni L., Sonnino S . , and Tettamanti G. (1989) Lacton- ization of GD 1 b ganglioside under acidic conditions. Curbohydr. Res. 193, 141-146.

Chan K.-F. J. (1987~) Ganglioside-modulated protein phosphory- lation in myelin. J. Biol. Chem. 262, 2415-2422.

Chan K.-F. J. (19876) Ganglioside-modulated protein phosphory- lation. Partial purification and characterization of a ganglioside stimulated protein kinase in brain. J . Biol. Chem. 262, 5248- 5255.

Chigorno V., Pitto M., Cardace G., Acquotti D., Kirschner G., Son- nino s., Ghidoni R., and Tettamanti G. (1985) Association of gangliosides to fibroblasts in culture: a study performed with GM 1 ['4C]-labelled at the sialic acid acetyl group. Glycoconjugute J. 2, 279-29 1.

Davis C. W. and Daly J. W. (1980) Activation of rat cerebral cortical 3',5'-cyclic nucleotide phosphodiesterase activity by ganglioside. Mol. Pharmacol. 17, 206-2 1 I .

Fishman P. H. (1988) Gangliosides as cell surface receptors and transducers of biological signals. Firliu Res. Ser. 14, 183-202.

Fronza G. , Kirschner G., Acquotti D., Bassi R., Tagliavacca L., and Sonnino S. (1988) Synthesis and structural characterization of the dilactone derivative of GDl a ganglioside. Curbohydr. Rex

Gazzotti G., Sonnino S., Ghidoni R., Cprlando P., and Tettamanti G. (1984) Preparation of the tritiated molecular forms of gan- gliosides with homogeneous long chain base composition. Gly- coconjugutr J . 1, 1 1 1-121.

Ghidoni R., Sonnino S . , Tettamanti G., Baumann N., Reuter G., and Schauer R. (1980) Isolation and characterization of a tri- sialoganglioside from mouse brain, containing 9-O-acetyl-N- acetylneuraminic acid. J. Biol. Chem. 255, 6990-6995.

Goldenring J. R., Otis L. C., Yu R. K., and DeLorenzo R. J . (1985) Calcium/ganglioside-dependent protein kinase activity in rat brain membranes. J. Neurochem. 44, 1229- 1234.

Gray E. G. and Whittaker V. P. (1962) The isolation of nerve endings from brain-an electron microscopic study of cell fragments

182, 3 1-40,

J . Neiirocliem., Vol. 57, No. 4, 1991

EFFECTS OF GANGLIOSIDES ON PROTEIN PHOSPHORYLA TION 1211

divided by homogenization and centrifugation. J. Anal. 96, 79- 88.

Hakomori S. (I98 1) Glycosphingolipids in cellular interactions, dif- ferentiation, and oncogenesis. Annu. Rev. Biochem. 50, 733- 764.

IUPAC-IUB Commission on Biochemical Nomenclature (1977) The nomenclature of lipids. Lipids 12,455-468; (1982) J. Biol. Chem. 251,3347-3351.

Kennedy M. B. ( 1 983) Experimental approaches to understanding the role of protein phosphorylation in the regulation of neuronal function. Annu. Rev. Neurosci. 6, 493-526.

Kim J. Y. H., Goldenring J. R., DeLorenzo R. J., and Yu R. K. ( 1986) Gangliosides inhibit phospholipid-sensitive Ca2+-depen- dent kinase phosphorylation of rat myelin basic proteins. J. Neurosci. Res. 15, 159-166.

Krebs E. G. and Beavo J. A. (1979) Phosphorylation-dephosphor- ylation of enzymes. Annu. Rev. Biochem. 48, 923-959.

Kreutter D., Kim J. Y. H., Goldenring J. R., Rasmussen H., Uko- madu C., DeLorenzo R. J., and Yu R. K. ( I 987) Regulation of protein kinase C activity by gangliosides. J. Biol. Chem. 262, 1633-1637.

Kuo J. F., Shoji M., Girard P. R., Mazzei G. J., Turner R. S., and Su H.-D. (1986) Phospholipid/calciumdependent protein kinase (protein kinase C) system: a major site of bioregulation. Adv. Enzyme Regul. 25, 387-400.

Laemmli U. K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 221, 680- 685.

Ledeen R. W. ( 1 984) Biology of gangliosides: neuritogenic and neu- rotrophic properties. J. Neurosci. Res. 12, 147-1 59.

Leon A,, Dacci L., Toffano G., Sonnino S., and Tettamanti G. (1981) Activation of (Na+,K+)-ATPase by nanomolar concentrations of GMI ganglioside. J. Neurochem. 37, 350-357.

Lowry 0. H., Rosebrough N. J., Farr A. L., and Randall R. J. (1951) Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193, 265-275.

Miettinen J. and Takki-Luukkainen J. T. (1959) Use ofbutyl acetate in determination of sialic acid. Acta Chem. Scand. 13,856-858.

Partington C. R. and Daly J. W. (1979) Effect of gangliosides on adenylate cyclase activity in rat cerebral cortical membranes. Mol. Pharmacol. 15,484-491.

Radsak K., Schwarzmann G., and Wiegandt H. (1982) Studies on the cell association of exogenously added sialo-glycolipids. Hoppe Seylers Z. Physiol. Chem. 363, 263-272.

Riboni L., Sonnino S., Acquotti D., Malesci A,, Ghidoni R., Egge H., Mingrino S., and Tettamanti G. (1986) Natural occurrence of ganglioside lactones. Isolation and characterization of GD 1 b inner ester from adult human brain. J. Biol. Chem. 261, 85 14- 8519.

Sonnino S., Ghidoni R., Chigorno V., Masserini M., and Tettamanti G. ( 1983) Recognition by two-dimensional thin-layer chroma- tography and densitometric quantification of alkali-labile gan- gliosides from the brain of different animals. Anal. Biochenr.

Sonnino S., Chigorno V., Acquotti D., Pitto M., Kirschner G., and Tettamanti G. (1989) A photoreactive derivative of radiolabeled GM 1 ganglioside: preparation and use to establish the involve- ment of specific proteins in GM 1 uptake by human fibroblasts in culture. Biochemistry 28, 77-84.

Svennerholm L. (1957) Quantitative estimation of sialic acid. 11: A colorimetric resorcinol-hydrochloric acid method. Biochim. Biophys. Acta 24, 604-6 1 I .

Svennerholm L. (1 980) Ganglioside designation. Adv. Exp. Ned. Biol. 125, 11.

Tettamanti G., Bonali F., Marchesini S., and Zambotti V. (1973) A new procedure for the extraction and purification of brain gan- gliosides. Biochim. Biophys. Acta 296, 160- 170.

Wiegandt H. (1985) Gangliosides. N . Compr. Biochem. 10, 199-260. Yates A. J., Walters J. D., Wood C. L., and Johnson D. (1989) Gan-

glioside modulation of cyclic AMP-dependent protein kinase and cyclic nucleotide phosphodiesterase in vitro. J. Neurochem.

128, 104-1 14.

53, 162-167.

J. Neurochem., Vol. 57, No. 4, 1991