Stimulation of cyclic adenosine 3′,5′-monophosphate-dependent protein kinase with brain gangliosides

~ Pergamon

Neurochem. Int. Vol. 26, No. 2, pp. 187-193, 1995

0197-0186(94)00102-2

Copyright© 1995ElsevierScienceLtd Printed in Great Britain.All rights reserved 0197~)186/95 $9.50+0.00

STIMULATION OF CYCLIC ADENOSINE 3',5'MONOPHOSPHATE-DEPENDENT PROTEIN KINASE WITH BRAIN GANGLIOSIDES F U T O S H I A R A K A N E 1"2, K O H J I F U K U N A G A l, M E I S A T A K E 3, K O H J I M I Y A Z A K I z, H I T O S H I O K A M U R A 2 a n d E I S H I C H I M I Y A M O T O l* tDepartment of Pharmacology and ~Department of Obstetrics and Gynecology, Kumamoto University School of Medicine, Kumamoto 860, Japan 3Department of Neurochemistry, Brain Research Institute, Niigata University, Niigata, Japan (Received 6 June 1994 ; accepted 11 July 1994)

Abstract--The holoenzyme of cAMP-dependent protein kinase (cAMP-kinase) partially purified from the particulate fraction of rat brain was stimulated by gangliosides. Among various gangliosides tested, GM1 was most potent, giving K~ value of 19.5/~M. The maximal activation of the kinase was obtained with 100 /~M GM 1 using kemptide as substrate. Gangliosides inhibited the kinase activity of the catalytic subunit of cAMP-kinase. Of various substrates tested, the ganglioside-stimulated cAMP-kinase could phosphorylate microtubule-associated protein 2, synapsin I and myelin basic protein, but not histone H 1 and casein. The molecular mechanisms of the stimulatory effect of gangliosides were investigated. The kinase activated with GM 1 was inhibited by the addition of PKItide, a specific inhibitor for cAMP-kinase. However, GMI did not dissociate the holoenzyme into the catalytic and regulatory subunits and did not interfere with the binding ability of cAMP to the holoenzyme. These results suggest that the gangliosides can directly activate cAMP-kinase in a different manner from cAMP.

Gangliosides, which are sialic acid-containing glycosphingolipids, are found in high concentrations in neural cells (Leeden, 1985). They may be involved in the modulation of cell growth, differentiation and recognition of various bioeffectors such as bacterial toxins and hormones (Hakomori, 1990). It was reported that exogenous gangliosides induced the differentiation of several types of neural cells with concomitant sprouting and extension of neurites (Rosenberg et al., 1992). Although the exact mechanisms by which gangliosides may influence these diverse systems are not defined, accumulating evidence suggests the involvement of protein kinases,

including tyrosine kinases associated with epidermal growth factor and platelet derived growth factor receptors (Bremer, 1986; Hanai et al., 1988), PKC (Kreutter et al., 1987), calmodulin (CAM) kinase II (Fukunaga et al., 1990), phosphorylase b kinase (Chan, 1989) and undefined protein kinases (Chan, 1987b, 1988; Goldenring et al., 1985; Tsuji et al., 1985). We recently reported evidence for the activation of the holoenzyme of cAMP-kinase by glycolipids isolated from Aplysia kurodai (Arakane et al., 1994). Glycolipids could interact with the regulatory subunit of cAMP-kinase and resulted in stimulation of the activities in place of cAMP (Arakane et al., 1994). *Author to whom all correspondence should be addressed. Chan (1987a) reported that a mixture of bovine brain Abbreviations : cAMP, cyclic adenosine Y,5'-monophosphate; cAMP-kinase, cAMP-dependent protein gangliosides inhibited phosphorylation of myelin kinase ; CaM kinase II, Ca2+/calmodulin-dependent pro- basic protein by the catalytic subunit ofcAMP-kinase. tein kinase II; DEAE, diethylaminoethyl; DTT, dithi- Likewise, Yates et al. (1989) found the direct inhibiothreitol; MAP2, microtubule-associated protein 2; tion of autophosphorylation of the catalytic subunit PKC, protein kinase C ; PKItide, cAMP-dependent pro- as well as the inhibition of activity. tein kinase inhibitor peptide; PMSF, phenylmethylOn the other hand, the occurrence of gangliosidesulfonyl fluoride; PPEA, phosphatidylethanolamine; dependent protein kinases have been reported in synSDS-PAGE, sodium dodecyl sulfate-polyacrylamidegel eleetrophoresis. aptosomes of guinea pig brain (Chan, 1987b, 1988) 187

188

Futoshi Arakane et al.

and in the plasma m e m b r a n e fraction o f neuroblastoma cell line (Tsuji et al,, 1985). The enzyme was partially purified from the brain (Chan, 1987b). In the present study, we addressed the question o f whether gangliosides derived from m a m m a l i a n brain have similar effects on c A M P - k i n a s e as observed with glycolipids from Aplysia kurodai. Gangliosides have dual effects on the c A M P - k i n a s e activity. That is, gangliosides activated the holoenzyme o f c A M P - k i n ase but inhibited the activity o f the catalytic subunit.

EXPERIMENTAL

PROCEDURES

Materials DEAE cellulose (DE 52) and phosphocellulose P-81 paper were purchased from Whatman ; Sephacryl S-300 from Pharmacia Fine Chemicals; bovine serum albumin from Schwarz/Mann ; monosialoganglioside GMI, GM3, asialoGM1, disialoganglioside GDIa, trisialoganglioside GTlb and total ganglioside from Biosynth AG; cAMP, casein, arginine-rich histone, phosphorylase b, protamine, ceramide and glucocerebroside from Sigma; PKItide and kemptide from Bachem Inc. ; histone H 1 from Boehringer Mannheim ; [3H]cAMP and [~,-32p]ATPfrom New England Nuclear (Du pont). Synapsin I was purified from rat brain (Schiebler et al., 1986); MAP2 from bovine brain (Yamamoto et al., 1985) ; myosin light chain from chicken gizzard (Miyamoto et al., 1981) and myelin basic protein from bovine brain (Deibler et al., 1972). The catalytic subunit of cAMP-kinase was purified from rat brain to apparent homogeneity by the method of Reimann and Beham (1983). GMI oligosaccharide-PPEA was prepared by the methods previously reported (Ito and Yamagata, 1986; Stoll et al., 1988), and contained the structure in which dipalmitoyl L-~-phosphatidylethanolamine was conjugated with oligosaccharide derived from GM 1a. Oligosaccharide was prepared by enzymatic cleavage from GT1 b. Preparation of cA MP-kinase Rats were killed by decapitation, and 30 cerebri were immediately homogenized in 10 vol. (w/v) of 10 mM TrisHCI (pH 7.5), 2 mM EDTA, 0.5 mM EGTA, 0.5 mM DTT, 0.5 mM PMSF, 25 mg/l trypsin inhibitor and 2 mg/l pepstatin A (buffer A) containing 0.32 M sucrose in a Teflon-glass homogenizer with a clearance of 0.3 mm at 1000 rpm with 10 up-and-down strokes. All procedures were carried out at 0-4°C. The homogenate was centrifuged at 1000 g for I0 min. The pellet was rehomogenized in 5 vol. of buffer A containing 0.32 M sucrose and centrifuged at 1000 g for 10 min. The low-speed supernatant was subjected to ultracentrifugation at 100,000 g for 40 rain. The supernatant was discarded. The particulate fraction obtained was exposed to hypotonic conditions in 10 vol. of 10 mM Tris-HCI buffer (pH 7.5) and 0.05 mM PMSF for 30 min. After centrifugation at 100,000 g for 40 rain, the pellet was resuspended and rehomogenized in 20 mM Tris-HC 1 buffer (pH 7.5), 2 mM EDTA, 0.5 mM EGTA, 20 mM NaCI, 0.5 mM DTT, and 0.5 mM PMSF with 0.5% Nonidet P-40. After extraction for 15 min with stirring, the supernatant was obtained by centrifugation at 120,000 g for 60 min and applied to a DEAE-cellulose column (4× 10 cm). The

enzyme was eluted with 2 1of a linear gradient of NaC1 (20 400 mM), and fractions of 18 ml each were collected. The active fractions of type II cAMP-kinase were pooled. After ammonium sulfate fractionation (50% saturation), the precipitate was dissolved in a few milliliters of 20 mM Tris-HC1 buffer (pH 7.5), 10 mM 2-mercaptoethanol, 0.1 mM EGTA and 200 mM NaCI (buffer C) and the enzyme solution was applied to a gel filtration column (2 × 95 cm) of Sephacryl S300. In a separate experiment, skeletal muscle of hind legs (5 g) was minced and homogenized on 10 vol. of buffer A containing 0.32 M sucrose. Then, type I and type II cAMPkinase was partially purified from cytosol fraction of rat skeletal muscle by the same procedure as from the rat brain. The concentrate was dialyzed over night against 5 1 of 20 mM Tris-HC1 buffer (pH 7.5), 0.1 mM EGTA, 1 mM DTT and 50% glycerol. The dialyzed solution was stored at - 80°C until use. The specific activities of the partially purified type II cAMP-kinase from rat brain and type I and type II cAMP-kinase from skeletal muscle were 125, 39.6 and 44.3 nmol/mg/min, respectively. A ssav./or cA MP-kinase The standard assay for cAMP-kinase, contained 50 mM HEPES (pH 7.5), 10 mM magnesium acetate, 1 mM EGTA, 0.1 mM [7-32P] ATP (3,000 5,000 cpm/pmol), I mg/ml bovine serum albumin, 40/aM kemptide (standard condition) and the indicated concentrations of gangliosides in a total volume of 25/al. The reaction was initiated by addition of partially purified cAMP-kinase (0.1 /ag of protein). After a 10 min incubation at 30°C, 15 #l aliquots of assay mixture were spotted on phosphocellulose P-81 paper squares and processed as described (Roskoski, 1983). Phosphorylation o f substrate proteins Partially purified cAMP-kinase (0.2 pg of protein) was incubated with 25 mM HEPES buffer (pH 7.5), 10 mM magnesium acetate, 1 mM EGTA and 0.1 mM [~' 32p]ATP with 12 pg of MAP2 and myelin basic protein, 10 /zg of synapsin I, or 4.8 #g of histone H1 in the absence or presence of 10 6 M cAMP and 100/aM GMI in a total volume of 50 Id. After a 10 min incubation at 30°C, the reaction was stopped by addition of the SDS sample buffer (Laernmli, 1970). After boiling for 3 rain, the samples were subjected to SDS PAGE in 10% acrylamide. Each protein band was cut from the gel and the radioactivity counted by a liquid scintillation counter. Assay.lbr c A M P bindin 9 The binding of cAMP to cAMP-kinase was assayed in an incubation medium that contained, in a final volume of 0.1 ml, 50 mM potassium-phosphate buffer (pH 6.8), 0.5 mg/ml histone II and 1 pmol of [3H]cAMP (56,000 cpm), in the presence or absence of nonradioactive cAMP or GMI as described previously (Corbin et al., 1978). Incubation was initiated by addition of 80 #g of cAMP-kinase and carried out for 60 min at 0°C. After incubation, the samples were passed through a 24 mm diameter cellulose ester filter (pore size, 0.45 #m; Millipore) previously rinsed with the same buffer. After the filter was washed with 10 ml of the buffer, it was placed in a counting vial with 4 ml of a scintillation mixture of toluene-cellosolve (3:1 v/v) plus fluors, in which the filter is readily dissolved and counted for radioactivity.

Activation of cAMP-kinase with gangliosides

lZ~_

100~

RESULTS

Ganglioside-dependent protein kinase activities by D EAE-cellulose column chromatography The enzyme solution extracted from the membrane fraction of rat cerebrum was applied to a D E A E cellulose column. The cAMP-kinase activity corresponding to type II was observed as a single peak at 0.24-0.27 M NaC1 using kemptide as substrate. At the same position, a major peak of the ganglioside(GT 1b)-dependent protein kinase activity was detected with the same substrate. Next, the same enzyme solution was pretreated with 10 6 M cAMP for 60 min at 0°C to dissociate the holoenzyme of cAMP-kinase into the catalytic and regulatory subunits. Then, the solution was applied to a D E A E cellulose column equilibrated with the buffer containing 20 mM NaC1 and 10 - 6 M cAMP. After washing, the enzyme was eluted with a linear gradient of NaCI (20-400 mM) in the presence of 10 - 6 M cAMP, as described above. The cAMP-kinase activity disappeared at the position of 0.24-0.27 M NaCI, while the peak of the catalytic subunit activity was observed at the position of about 0.12 M NaC1. Likewise, the ganglioside-dependentprotein kinase activity also totally disappeared and the addition of ganglioside slightly inhibited the catalytic activity of cAMP-kinase rather than the stimulation.

Activation of cAMP-kinase by gangliosides The effects of gangliosides on the partially purified type II cAMP-kinase from the rat brain were investigated using kemptide as substrate. This preparation did not contain activities of PKC and CaM kinase II. GM 1 had dual effects on the cAMP-kinase activity at a concentration of less than 100 #M. In the absence of cAMP, cAMP-kinase was activated in a dosedependent manner with a Ka value of 19.5/~M and the V~ax of 40.0 pmol/min (Fig. 1). The maximal activation was obtained with 100 #M GM1 with about 30% of the activity obtained in the presence of 10 ~ M cAMP, while the activity in the presence of 10 -5 M cAMP was inhibited by the addition of GM 1 in a dose-dependent manner (ICs0 = 120/tM). The inhibitory effect of GM 1 was observed with the purified catalytic subunit of cAMP-kinase using kemptide as substrate (data not shown). Since the activity of type I cAMP-kinase in the rat particulate fraction was relatively low as reported previously (Arakane et al., 1994), we tested the effect of GM 1 on type I and type II cAMP-kinase from the skeletal muscle. Consistent with the activation of type II cAMP-kinase from the brain, GM 1 showed similar

189

r5 :

50~ ~

U 0.o5

Q.O 0.1 1/(OM1,~.~M)

i

O0

~

I

i

0.2

I

100 200 CONCENTRATION OF GM1 (Ix.M)

300

Fig. 1. Dose-response curve for kinase activity with GMI and additive effect with 10 6 M cAMP. Kinase activities were assayed with the indicated concentrations of GM 1 in the absence or presence of cAMP with 40/~M kemptide as substrate. An inset shows the double reciprocal plots. A line of best fit was computed from the experimental data of the kinetic analysis. Values represent means + SE of three determinations. stimulatory effects on both type I and type II cAMPkinase from the skeletal muscle. Furthermore, the activities of both cAMP-kinase from the skeletal muscle in the presence of 10 5M cAMP were inhibited by the addition of GM1, indicating that the activity of the catalytic subunit of both types of cAMP-kinase from the skeletal muscle was also inhibited. With type II cAMP-kinase from the rat brain, polysialogangliosides such as G D l a and G T I b were less potent activators than monosialogangliosides such as GM1 and GM3 (Table 1). Therefore, we tested the effect of GM1 oligosaccharide-PPEA on cAMPkinase. Although the compound could stimulate cAMP-kinase activity, the stimulation was much less than GM1. On the other hand, asialo-GM1, glucocerebroside, or the oligosaccharide and ceramide derived from GT 1b could not stimulate cAMP-kinase (Table 1). It suggests that the whole structure including a moiety of glycoside, ceramide and sialic acid may be important to activate cAMP-kinase. The effect of PKItide was tested on the gangliosideand cAMP-stimulated kinase activities (Fig. 2). Both kinase activities were maximally inhibited with 25 #M PKItide. The concentration of PKItide to give a half maximal inhibition was about 8 and 12 #M for ganglioside- and cAMP-stimulated kinase activities, respectively. The patterns of the inhibitory curve for both kinase activities were similar to each other (Fig. 2).

Futoshi Arakane et al.

190

Table 1. Effects of various compounds on kinase activity

Compound None GMI GM3 GDla GTIb Total ganglioside Asialo-GM 1 GM 1 oligosaccharide-PPEA Glucocerebroside Oligosaccharide Ceramide cAMP

Kinase activity (pmol/min) 1,5 + 0.1 12.1_+0.1 10.6_+0.1 6.8_+0.1 5.4+0.1 2.7 _+0. I 2.3 _+0.1 2.2 _+0.1 1.9 + 0.2 1.5 _+0.2 1.4 + 0.2 56.4+ 1.9

Kinase activity was assayed in the presence of 100 #M compounds and 10 6 M cAMP using 40 pM kemptide as substrate. Values represent means _+SE of three determinations. Total ganglioside contains the mixture of gangliosides from bovine brain.

The stimulation of cAMP-kinase by gangliosides may suggest that the stimulatory effect is due to the binding of the ganglioside to the substrate. Therefore, we tested whether the concentrations of the ganglioside had an effect on the kinase activity in the different concentrations of substrate. In other words, if the kinase activity increases by the interaction between ganglioside and substrate, the concentration of ganglioside to give the maximal activity should be changed by the concentrations of substrate. As shown

120/ 1

~_

•

cam,

7

in Fig. 3, the maximal activity of the kinase was obtained with the same concentration of the ganglioside using different concentrations of the substrate. These results indicate that the activation of cAMP-kinase with gangliosides is not due to the interaction with the substrate, but with the kinase.

Substrate specS'city q f ganglioside-stimulated cAMPkinase The ability of G M 1 to stimulate cAMP-kinase was assessed using several different protein substrates of cAMP-kinase. As shown in Table 2, GM1 stimulated the phosphorylation of kemptide, M A P 2 and synapsin I but had little effect on the phosphorylation of myelin basic protein. Phosphorylase b, protamine, arginine-rich histone and myosin light chain were also poor substrates for cAMP-kinase stimulated with gangliosides (data not shown). The phosphorylation of histone HI and casein was inhibited with gangliosides. Although histone HI is a good substrate for c A M P kinase, the protein was not phosphorylated in a G M 1dependent manner. It suggests that the activation manner of cAMP-kinase with GM1 is different from that elicited with cAMP.

Effects o f the gan,qlioside on cA MP-binding to cAMPkinase To analyze the mechanism of ganglioside stimulation, we tested the effect of ganglioside on c A M P binding to cAMP-kinase (Table 3). The addition of 10/~M nonradioactive c A M P completely inhibited the binding of [3H]cAMP to the kinase. In contrast, the ganglioside gave no significant inhibition of [3H]cAMP-binding even at 100 /~M (Table 3). The

~pM

Fig. 2. Inhibitory effects of PKltide on the activity. Effects of PKltide on the kinase activities were determined with the holoenzyme of cAMP-kinase in the presence of 100 pM GM1 (O) or 10 -6 M cAMP (O) using 40 pM kemptide as substrate, under standard assay conditions. Values are expressed as percentage. Kinase activities were 9.1 and 22.0 pmol/min with 100 pM GMI and 10 ~' M cAMP, respectively, in the absence of PKItide. Each value was taken as 100% and from this value, respective values were calculated as percentage. Data represent means +SE of three determinations.

~ ~

2 .# ~

~"~"~0

o ~ 50

-

,

,

1 O0

'I 50

200

/ -"~[ ";' [ 250

GM1 (~M)

Fig. 3. Effects of various concentrations of kemptide on kinase activity with GM1. Kinase activities were assayed with the indicated concentrations of GMI and kemptide. Values represent means _+SE of three determinations.

Activation of cAMP-kinase with gangliosides Table 2. Substrate specificity of kinase activities with cAMP and GMI Kinase activity (pmol/min) Substrate

None

GMI(100 #M)

cAMP (10 6 M)

1.17_+0.03 1.54 ±0.04 1.90±0.15 0.19_+0.01 11.44_+0.40 2.34±0.31

5.9±0.43 3.14_+0.23 2.47_+0.55 1.20_+0.03 7.86_+0.52 1.61 _+0.02

10.52±0.53 9.14+0.03 4.79±0.10 3.71 _+0.20 92,72± 1.86 3,75+0.08

191

fractions were still observed (Fig. 4). The results suggest that the ganglioside can activate cAMP-kinase without the dissociation of the enzyme. DISCUSSION

Kemptide Synapsin 1 MBP MAP2 Histone HI Casein

The holoenzyme of cAMP-kinase (0.2 #g) was incubated with 40 #M kemptide, 10 #g of synapsin 1, 12 /tg of myelin basic protein (MBP) and MAP2, 4.8 #g of histone H1 and 10/tg of casein in the absence or presence of 10 +6 M cAMP and 100/~M GM1, as indicated, in a total volume of 0.05 ml under the standard assay conditions. Values represent means ± SE of three determinations.

results suggest that the binding sites of the kinase differ between the ganglioside and cAMP.

Effect of the ganylioside on dissociation of cAMPkinase The different substrate specificity of the GM1stimulated kinase and the lack of inhibition of the [3H]cAMP binding by G M I may suggest that the stimulatory effect is due to the interaction of the ganglioside in a different manner from that of cAMP. To address this problem, the ability of G M 1 to dissociate the holoenzyme into the regulatory and catalytic subunits was tested (Fig. 4). The preincubation of the holoenzyme with 10 5 M cAMP completely dissociated and the peak of the kinase activity shifted to the position of the catalytic subunit, in which the activity was totally independent of cAMP (data not shown). In contrast, the pretreatment with 100/~M G M I could not dissociate the holoenzyme at all. The basal activity in the absence of cAMP increased because of the inclusion of G M I in the active fraction. The stimulatory effects of GM1 on the kinase activities of the Table 3. Effects of cAMP or GM1 on cAMP-binding activity of the kinase Concentration (~M) None cAMP GM 1

1 10 5 100

Bound [3H]cAMP (cpm) 9858.5 _ 1096.1 1875.8 +482.5 601.1 + 101.0 12369.5+737.4 11902.0 + 800.8

The assay for [3H]cAMP binding to cAMP-kinase was performed at 0'~C for 60 min in 50 mM potassium phosphate buffer (pH 6.8), 0.5 mg/ml histone I1, and 1 pmol [3H]cAMP (56,000 cpm) with 80 /~g of cAMP-kinase in the presence or absence of nonradioactive cAMP or G M I . Values represent means + S E of four determinations.

There is increasing evidence that the functions of membrane proteins are influenced by the surrounding lipids in the lipid bilayer. In the present study, we found that gangliosides could stimulate the activities of type I and type II cAMP-kinase partially purified not only from the particulate fraction of rat brain but also from the cytosol fraction of rat skeletal muscle. It is interesting that the compound with a sialic acid residue can modify to activate cAMP-kinase and that monosialoganglioside (GMI) is a more potent activator of cAMP-kinase than G D l a and G T I b (Table 1). However, since asialo-GM 1 had only a slight effect on the activation of the kinase, a portion of sialic acid may be needed for the effect of gangliosides. In contrast, glycolipids isolated from Aplysia kurodai which contain no sialic acid could stimulate cAMPkinase (Arakane et al., 1994). On the other hand, oligosaccharide or ceramide had little effect on the activation of the kinase, indicating that only a partial structure in whole ganglioside does not have an effect to stimulate the kinase. Furthermore, G M I oligosaccharide-PPEA in which a portion of sphingosine of G M 1 is replaced by phosphatidylethanolamine had only a slight effect, suggesting that a portion in sphingosine is needed for the activation of the enzyme. These results indicate that a specific moiety of ganglioside to activate the kinase does not occur, but the whole ganglioside structure is needed. Thus, we could not find any common rule about the structures for the activation of cAMP-kinase with gangliosides at present. The stimulation of cAMP-kinase by the gangliosides was dependent on the protein substrate utilized. With G M 1, significant stimulation was obtained using MAP2, synapsin I and myelin basic protein but was not with histone H1. Since these were good substrates for cAMP-kinase when stimulated by cAMP, this substrate selectivity could suggest that the stimulation of cAMP-kinase is due to binding of the ganglioside to the substrate. However, our observation would tend to argue against this mechanism. For example, the maximal activity of the kinase with different concentrations of the substrate was obtained with the same concentration of the ganglioside (Fig. 3). The ganglioside did not compete with [3H]cAMP binding on the enzyme and could activate cAMPkinase by interaction with the regulatory subunit with-

192

Futoshi Arakane etal.

25 Control 10.6 M cAMP 100 ~M GM1

15

W

00

Z

5

A V

0

10

ill

Jlh_l

~

A

~

~

20

30

FRACTION

40

50

60

NUMBER

Fig. 4. Dissociation of cAMP-kinase with GM1. cAMP-kinase (0.32 mg of protein) was preincubated at 0~C for 60 min with 100/~M GMI and applied to a sephacryl S-300 column (0.8 x 40 cm) equilibrated with buffer C containing 50 #M GM1. Then the enzyme was eluted with buffer C plus 50 #M GMI. Fractions of 300 #1 each were collected and aliquots (15 #1) of the fractions were assayed in the absence (@) and presence of 10 6 M cAMP (©) or 100 pM GM1 (A) using 40/~M kemptide as substrate.

out the dissociation. These results suggest that the mechanism of the kinase activation with ganglioside may differ from that with cAMP. The occurrence of ganglioside-dependent protein kinases has been reported (Chan, 1987b, 1988 ; Tsuji et al., 1985). The protein kinases were designated as ecto-protein kinase, since they were stimulated by the addition of exogenous gangliosides and, therefore, considered to be stimulated with gangliosides localized in the outer surface of the cellular membranes (Tsuji et al., 1988). The purification procedures for ganglioside-dependent protein kinase (Chan, 1987b) differed from those for cAMP-kinase reported. On the other hand, ecto-protein kinase activity in H e L a cells and Chinese hamster ovary cells was stimulated by the addition of c A M P (Kflbler et al., 1992). Since the cAMP-kinase inhibitor protein inhibited this kinase activity, they suggested that this ecto-protein kinase may be cAMP-kinase. Together with the above results, the present study indicates that the known protein kinase (cAMP-kinase) is activated with gangliosides and therefore suggests that at least one of

the ganglioside-dependent protein kinases is c A M P kinase. Although most gangliosides are bound to the membrane surface of the cell, 2 - 5 % of total cellular gangliosides is soluble in the cytosol (Miller-Podraza and Fishman, 1983 ; Sonnino et al., 1979, 1981). They are soluble as micelles in an aqueous solution at a similar concentration used in this study, van Genderen et al. 1991) reported that 66% of Forssman glycolipid, a glycosphingolipid, is localized in the intracellular space. It seems likely that they could reach significant levels in specific intracellular compartments, or could interact with a soluble component in the cytosol. The lack of information concerning local concentrations of gangliosides in subcellular compartments limits the specific interpretation of our in vitro studies with respect to effects of gangliosides on cAMP-kinase in vivo. Acknowledgements--This work was supported in part by a

Grant-in-aid for Scientific Research from the Ministry of Education, Science, and Culture of Japan. We would like to

Activation of cAMP-kinase with gangliosides thank Drs S. Ando and K. Kon of Tokyo Metropolitan Institute for the analysis of GM 1 oligosaccharide-PPEA. REFERENCES

Arakane F., Fukunaga K., Araki S., Abe S., Satake M., Miyazaki K., Okamura H. and Miyamoto E. (1994) Glycolipids isolated from Aplysia kuradai can activate cyclic adenosine 3',5'-monophosphate-dependent protein kinase from rat brain. J. Neurochem. 62, 86 93. Bremer E. G. (1986) Ganglioside-mediated modulation of cell growth: specific effects of GM3 on tyrosine phosphorylation of the epidermal growth factor receptor. J. biol. Chem. 261, 2434-2440. Chan K.-F. J. (1987a) Ganglioside-modulated protein phosphorylation in myelin. J. biol. Chem. 262, 2415-2422. Chan K.-F. J. (1987b) Ganglioside-modulated protein phosphorylation: partial purification and characterization of a ganglioside-stimulated protein kinase in brain. J. biol. Chem. 262, 5248 5255. Chan K.-F. J. (1988) Ganglioside-modulated protein phosphorylation : partial purification and characterization of a ganglioside-inhibited protein kinase in brain. J. biol. Chem. 263, 568-574. Chan K.-F. J. (1989) Ganglioside-modulated protein phosphorylation in muscle. Activation of phosphorylase b kinase by gangliosides. J. biol. Chem. 264, 18,632-18,637. Corbin J. D., Sugden P. H., West L,, Flockhart D. A., Lincoln T. M. and McCarthy D. (1978) Studies on the properties and mode of action of the purified regulatory subunit of bovine heart adenoside 3',5'-monophosphatedependent protein kinase. J. biol. Chem. 253, 3997-4003. Deibler G. E., Marteson R. E. and Kies M. W. (1972) Large scale preparation of myelin basic protein from central nervous tissue of several mammalian species. Prep. Biochem. 2, 139 165. Fukunaga K., Miyamoto E. and Soderling T. R. (1990) Regulation of Ca2+/calmodulin-dependent protein kinase II by brain gangliosides. J. Neurochem. 54, 102-109. van Genderen I. L., van Meet G., Slot J. W., Geuze H. J. and Voorhout W. F. (1991) Subcellular localization of Forssman glycolipid in epithelial MDCK cells by immunoelectronmicroscopy after freeze-substitution. J. Cell Biol. 115, 1001-1019. Goldenring J. R., Otis L. C., Yu R. K. and DeLorenzo R. J. (1985) Calcium/ganglioside-dependent protein kinase activity in rat brain membrane. J. Neurochem. 44, 1229 1234.

Hakomori S.-I. (1990) Bifunctional role of glycosphingolipids: modulators for transmembrane signaling and mediators for cellular interactions. J. biol. Chem. 265, 18,713-18,716. Hanai N., Nores G. A., MacLeod C., Torres-Menderz C.-R. and Hakomori S. (1988) Ganglioside-mediated modulation of cell growth : specific effects of GM3 and lyso-GM 3 in tyrosine phosphorylation of the epidermal growth factor receptor. J. biol. Chem. 263, 10,915-10,9121. lto M. and Yamagata T. (1986) A novel glycosphingolipiddegrading enzyme cleaves of the linkage between the oligosaccharide and ceramide of neutral and acidic glycosphingolipids. J. biol. Chem. 261, 14,278-14,282. Kreutter D., Kim J. Y. H., Goldenring J. R., Rassmussen H., Ukomadu C , DeLorenzo R. J. and Yu R. K. (1987)

193

Regulation of protein kinase C activity by gangliosides. J. biol. Chem. 262, 1633-1637.

Kt~bler D., Reinhardt D., Reed J., Pyerin W. and Kinzel V. (1992) Atrial natriuretic peptide is phosphorylated by intact cells through cAMP-dependent ecto-protein kinase. Eur. J. Biochem. 206, 179 86. Laemmli U. K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680 685. Leeden, R. W. (1985) Gangliosides of the neuron. Trends Neurosci. 8, 169-174. Miller-Podraza S. and Fishman P. H. (1983) Soluble gangliosides in cultured neurotumor cells. J. Neurochem. 41, 860-867. Miyamoto E., Fukunaga K., Matsui K. and Iwasa Y. (1981) Occurrence of two types of Ca2+-dependent protein kinases in the cytosol fraction of the brain. J. Neurochem. 37, 1324-1330. Reimann E. M. and Beham R. A. (1983) Catalytic subunit of cAMP-dependent protein kinase. Method. Enzymol. 99, 51 55. Rosenberg A., Sauer, A., Noble E. P., Gross H. J., Chang R. and Brossmer R. (1992) Developmental patterns of ganglioside sialosylation coincident with neuritogenesis in cultured embryonic chick brain neurons. J. biol. Chem. 267, 10,607-10,612. Roskoski R. J. (1983) Assays of protein kinase. Method. Enzymol. 99, 3 6~

Schiebler W., Jahn R., Doucet, J.-P., Rothlein J. and Greengard P. (1986) Characterization of synapsin I binding to small synaptic vesicles. J. biol. Chem. 261, 8383-8390. Sonnino, S., Ghidoni R., Marchsini S. and Tettamanti G. (1979) Cytosolic gangliosides : occurrence in calf brain as ganglioside-protein complexes. J. Neurochem. 33, 117121. Sonnino S., Ghidoni R., Masserni M., Aporti F. and Tettamanti G. (1981) Changes in rabbit brain cytosolic and membrane-bound gangliosides during prenatal life. J. Neurochem. 36, 227 232. Stoll M. S., Mizuochi T., Childs R. A. and Feizi T. (1988) Improved procedure for the construction of neoglycolipids having antigenic and lectin-binding activities from reducing oligosaccharides. Biochem. J. 256, 661-664. Tsuji S., Nakajima J., Sasaki T., and Nagai Y. (1985) Bioactive gangliosides. IV. Ganglioside GQIb/Ca :÷ dependent protein kinase activity exists in the plasma membrane fraction of neuroblastoma cell line, GOTO. J. Biochem. 97, 969-972. Tsuji S., Yamashita T. and Nagai Y. (1988) A novel carbohydrate signal-mediated cell surface protein phosphorylation: ganglioside G Q l b stimulates ecto-protein kinase activity on the cell surface of a human neuroblastoma cell line, GOTO. J. Biochem. (Tokyo) 104, 4098 5003. Yamamoto H., Fukunaga K., Goto S., Tanaka E. and Miyamoto E. (1985) Ca :÷, calmodulin-dependent regulation of microtubule formation via phosphorylation of microtubule-associated protein 2, ~ factor, and tubulin, and comparison with the cyclic AMP-dependent phosphorylation. J. Neuroehem. 44, 759-768. Yates A. J., Walter J. D., Wood C. L. and Johnson J. D. (1989) Ganglioside modulation of cyclic AMP-dependent protein kinase and cyclic nucleotide phosphodiesterase in vitro. J. Neurochem. 53, 162-167.