Synthesis and subcellular transport of sulfogalactosyl glycerolipids in the myelinating mouse brain

294

Biochimica et Bioph.ysicu Acta, 753 (3983) 294-299 Elsevier

BBA 51487

SYNTHESIS AND SUBCELLULAR TRANSPORT THE MYELINATING MOUSE BRAIN THOMAS

BURKART

a, LUIGI

a Department of Pediatrics, Milan (ItaJy) (Received

CAIMI

University

b and ULRICH

OF SULFOGALACTOSYL

N. WIESMANN

of Berne, Berne ~Switzerlffnd)

GLYCEROLIPIDS

IN

a

and b Department

of Biological

Chemistry,

University

of M&m,

July 4th. 1983)

Key words: Sulfogulactosyl

glycerolipid; Myelin synthesis; Membrane

turnover; Glycolipid; Lysosome;

(Mouse brain)

In the 17-day-old myelinating mouse brain the site of sulfogalactosyl glycerolipid synthesis and the kinetics of its su~elluIar dis~bution were studied by a 2 h else-ladling with p%]sulfate followed by a 4 h chase of ~3~S~s~fog~aetosyi ~ycerolipid. At several time intervals after the in~a~ritone~ [35S]suIfate injection, subceIlular fractions of brain were obtained by differential and di~ontin~us sucrose gradient centrifugation. Tbe crude microsomal membrane fraction (17 500 X g supernatant) was further subfractionated into light myelin, plasma membranes, Golgi vesicles, endoplasmic reticulum membranes and heavy vesicles associated with acid hydrolase activities. The results of the [35S]sulfogalactosyl glycerolipid labeling kinetics indicate that these lipids are synthesized in the Golgi-endoplasmic reticulum complex and transferred in vesicles associated with lysosomes to the myelin membranes. During this transfer part of the sulfogalactosyl glycerolipids appears to be degraded, similarly as described for brain sulfatides. This double function of lysosomes may be part of a general regulation mechanism of brain myelin glycoiipid content. In~~uction Brain sulfogalactosyl glycerolipids occurring in a diacyl and an alkylacyl form [l-4] have been shown to be associated with myelin [3]. Recently we have demonstrated [5] that the biosynthesis and the biodegradation as well as the accumulation of sulfogalactosyl glycerolipids in mouse brain follow developmental patterns comparable to the ones observed for brain sulfatides 161, except that sulfog~actosyl glycerolipids are synthesized with a peak around 14 days compared with 20 days for sulfatide. Their accumulation rate is highest during the period of active myelination. The newly synthesized sulfogalactosyl glycerolipids have a very large turnover rate, 40-8058 being degraded within 1 day. The rate of degradation closely parallels the rate of biosynthesis during myelination and appears to co-regulate the net synthesis of this myelin lipid [5]. In this paper we report the results of a study of OOOS-2760/83/$03.00

0 1983 Elsevier Science Publishers

B.V.

the sites of synthesis and degradation of sulfogalactosyi glycerolipids and their subcellular transport to the myelin membranes. The results are compared with the previous study on sulfatide [7]. The present data suggest that sulfogalactosyl glycerolipids are synthesized in the Golgi-endoplasmic reticular membrane complex. The newly synthesized sulfolipids are rapidly packed in vesicles, which are at least in part associated with brain lysosomes, before the lipids become incorporated into myelin. Materials and Methods Chemicals Carrier-free [35S]sulfate was obtained from the Radiochemical centre, Amersham. Bovine cerebrosides and crystallized bovine serum albumin were purchased from Sigma. As scintillation chemicals Aquasol from New England Nuclear and Permablend II@ from Packard were used. Sucrose

295

(density gradient grade) and all the other chemicals (analytical grade) used originated from Merck. Animals and experimental procedures

17-day-old mice (average weight 8 g) of the C57BL/J strain (Jackson Laboratories, Bar Harbor, ME, USA) were injected intraperitoneally with 20 PCi carrier-free [35S]sulfate per g body weight in 200 ~1 0.9% NaCl solution (25pl/g). After 2 h a second injection of 200 ~1 of 150 mM unlabeled Na,SO, solution (25 pi/g body weight) followed. At 1, 2, 3, 4 and 6 h after [35S]sulfate injection groups of eight animals were decapitated and their brains (without medulla oblongata) were isolated and pooled in an ice-cooled glass-Teflon homogenizer and weighed. Subcellular fractionation procedures

The brains were homogenized in 3 vol. ice-cold 0.32 M sucrose containing 1 mM EDTA and 25 mM Na,SO, at O’C in a 15 ml glass-Teflon homogenizer (0.2 mm clearance) by five slow down-up strokes with the rotating pestle (500 rpm). The 25% (w/v) homogenate was fractionated by differential centrifugation steps (1000 x g,17 500 x g, 100000 X g) and the resuspended 1000 X g and 17 500 X g pellets were further subfractionated on discontinuous sucrose gradients essentially according to the method of Eichberg et al. [8]. More details as well as data on the characteriza-

TABLE

tion of the fractions have been published previously [7]. The 17 500 x g supernatant material (crude microsomes) was further subfractionated on a continuous sucrose gradient ranging from 0.8 to 1.3 M sucrose with a 2 M sucrose cushion on the bottom, essentially according to the method of Siegrist et al. [9]. Determination of [-?3Jsulfogalactosyl glycerolipids

Radioactively labeled sulfolipids were determined as described previously [6,7]. After extraction with chloroform/methanol (2 : 1 v/v) and drying by nitrogen [lo], duplicate lipid samples were saponified in 0.2 M NaOH in methanol [l] and other duplicates were incubated in methanol alone. The products of mild alkaline hydrolysis were removed by a further partitioning according to the method of Folch et al. [lo]. The radioactivity of sulfogalactosyl glycerolipids was all recovered in the upper aqueous phase, as checked by thin-layer chromatography using solvent system I described by Ishizuka et al. [4] and system II described by Pieringer et al. [3], respectively. In aliquots of the upper phases and of the N,-dried lower phases radioactivity was measured by liquid scintillation counting (Intertechnique SL-4000 scintillation spectrometer) using toluene containing 0.4% Permablend II’ (98% 2,5-diphenyloxazole, 2% 1,6bis (2-(5-phenyloxazozyl)) ben-

I

ABSOLUTE AMOUNT OF [“S]SULFOGALACTOSYL FROM I7-DAY-OLD MOUSE BRAINS ACCORDING

GLYCEROLIPIDS TO THE METHOD

IN SUBCELLULAR FRACTIONS OF EICHBERG ET AL. [8]

DERIVED

The animals were intraperitoneally injected with [35S]sulfate (20 pCi/g) at time zero, followed by a second injection of unlabeled sulfate at 2 h (chase). The radioactivity data represent mean values of two independent experiments. The ranges were IS-22% of the mean. Values are dpm [ 35S]sulfogalactosyl glycerolipid/g brain (fresh weight) (. IO- ‘). n.d.. not detectable. Fraction

Time after 35S0, injection(h)

1

2

3

4

6

Homogenate (H) Microsomes (9) Soluble (SN,) Mitochondrial/lysosomal Synaptosomes (P2 B) Small myelin (P,A) Large myelin (P,A) Nuclear fraction (P,B) Debris (P,C)

9.09 1.91 n.d. 1.44 1.59 0.48 0.70 2.59 1.22

37.17 1.01 n.d. 3.72 4.65 1.81 4.00 5.07 16.20

36.34 0.99 n.d. 1.77

16.85 0.63 nd. 1.34 1.15 2.21 1.66 2.89 6.20

20.65 0.46 n.d. 1.84 0.75 2.71 1.54 5.17 7.31

(PzC)

1.80 3.37 4.96 5.82 14.05

296

zene) as the scintillation medium. The method is presented in full length in a previous paper of ours

Results

[51.

Kinetics of the [ ?‘T]sulfogaluctosyl glycerolipid labeling in subcellular fractions after injection of [ TS]suifate -_ In the homogenate (H) [ “Sjsulfogalactosyl glycerolipids increased during the first 2 h. reaching a plateau. and decreased after 3 h (Fig. la). In the microsomal fraction (I’,) radioactivity rose to a peak at 1 h and then decreased (Fig. la). In both the mitochondrial/lysosomal (P2C) and in the synaptosomal/ lysosomal ( Pz B) fractions, a peak was reached 2 h after labeling (Fig. lb). Between 2 and 4 h, the [ “5S]sulfogalactosyl glycerolipids fell to 36 and 25% of the peak values in the mitochondrial/ lysosomal fraction and synaptosomal/lysosomal fractions. respectively. [ “SS]Sulfogalactosyl glycerolipid radioactivity was maximal in the small ( P2A) and large (P, A) myelin fractions (Fig. Ic) 3 h after labeling, and thereafter decreased. In Table I the [ 35S]sulfogalactosyl glycerolipid content per g fresh brain in subcellular fractions is shown. 2 h after [“Slsulfate injection a high proportion of the j5S-labeled lipid was associated with the two fractions containing lysosomes. namely the mitochondrial (P2C) and the synaptosomal (P,B) fraction. At 3 h similar labeling was reached in the small (P,A) and large (PI A) myelin fractions while in P2C and P,B the radioactivity decrease to less than 50%. No radioactive lipid was detected in the soluble fraction (100000 x R supernatant. SN,). The remaining radioactivity was found to be associated with the nuclear fraction (P,B) and was pelleted with the debris fraction (P,C).

Protein determination Protein concentration was measured according to the method of Lowry et al. [ 1 l] using crystallized bovine serum albumin as the standard.

Fig. I. Kinetics of [ %]sulfogalactosyl glycerolipid ([ 3SS]SGG) labeling in subcellular fractions derived from mouse brain homogenates by differential centrifugation and by discontinuous sucrose density gradient centrifugations according to the method of Eichberg et al. [g]. 17-day-old mice were injected intraperitoneally with 20 PCi of carrier-free [35S]sulfate/g body weight (25 pi/g). After a labeling period of 2 h a chase was made by a second intraperitoneal injection of 150 mM unlabeled Na,SO, solution (25 pi/g body weight). At several time intervals after [3SS]sulfate injection, eight animals were killed by decapitation and from their brains subcellular fractions were prepared and characterized as previously described [7]. [ 35S]Suifogalactosyl glycerolipids were determined in the fractions as described in Material and Methods. All points are mean values from two independent labeling-fractionation experiments, individual determinations being within lo-1548 of the mean value (eight brains were pooled in each experiment). The [ 35S]sulfogalactosyl glycerolipid radioactivities are expressed per mg protein of the fractions. H, homogenate; P3, microsomal fraction; P,B synaptosomal/lysosomaI fraction: P,C, mitochond~al/lysosoma~ fraction: P, A. large myehn: P2 A. small myelin.

Kinetics of [“sS]su(fogalactosyI glycerolipids in microsomal subfractions prepared from I7-da.v-o/d mouse bruin A continuous sucrose gradient system (91 was used to separate microsomal membranes into light myelin membranes floating on the top of the gradient (fractions l-3), plasma membranes (fractions 4-6) Golgi vesicles in the middle of the gradient (fractions 7-9). endoplasmic reticulum membranes (fractions lo- 12) and vesicles associated with acid hydrolase activities in the densest fractions (Fractions 13-16). The labeling pattern of the [ 3’SJsuIfogalactosyl glycerolipids in the microsomal subfractions is

297

ity was highest 1 h after labeling. Between 1 and 3 h the radioactivity in these fractions decreased reaching about 50% of the peak value. Later, a further decrease to 10-20s was noted. A similar pattern was observed in the vesicles of the dense fractions (fractions 13-16). In the fractions enriched with plasma membranes (fractions 4-6) a high radioactivity was observed at 1 h, which continuously disappeared up to 6 h, whereas [ “S]sulfogalactosyl glycerolipids reached a peak at 3 h in the light myelin membranes (fractions l-3), followed by a decrease. In Fig. 2b the total amount (dpm/g brain) of [ “S]sulfogalactosyl glycerolipids and their relative distribution (%) throughout the gradient are given. For the calculation of the distribution in percent the total radioactivity in the crude microsomal fraction (17 500 X g supernatant) was taken as 100%. The data show that the association of newly formed sulfogalactosyl glycerolipid with lighter vesicles (plasma membranes and premyelin membranes) was proportionally much higher than with heavy vesicles (fractions 13-16).

shown in Fig. 2a: in the Golgi (fractions 7-9) and endoplasmic reticulum membranes (fractions lo- 12) [ j5 Slsulfogalactosyl glycerolipid radioactiv-

100".=1750dpm

Discussion

I

_

~~~~

6 11 16 fraction nr.

~_L____

1

6

11

fraction nr.

I6

,

Fig. 2. Kinetics of [ 35S]sulfogalactosyl glycerohpid ([35S]SGG) labeling in microsomal subfractions derived from the 17 500 x g supernatant by a linear sucrose gradient system according to the method of Siegrist et al. [9]. [35S]Sulfogalactosyl glycerolipid pulse-labeling-chase experiments were performed as described in the legend to Fig. 1. The microsomal subfractions were prepared and enzymatically characterized as published elsewhere [7,9]. [ 35S]Sulfogalactosyl glycerolipid was determined as described in Materials and Methods. All data plotted are mean values of the same two independent labelingfractionation experiments as shown in Fig. 1. The range of individual values was 12-188 of the mean value. In a. the kinetics of the [ 35S]sulfogalactosyl glycerolipid per mg protein of microsomal subfractions (fractions l-16) is shown. In b. the time-dependent change of the percent distribution of total [ 35S]sulfoglactosyl glycerolipid is given, calculated by taking the total radioactivity of the gradient. corresponding to 1 g of brain, as 100%. The following membrane populations have been found to be enriched throughout the gradient [9]: fractions 1-3. light myelin (dotted bars); fractions 4-6, plasma membranes (shaded bars); fraction 7-9. Golgi vesicles (open bars); fractions 10-12, endoplasmic reticulum membranes (cross-hatched bars); fractions 13-16. (pre)lysosomes (black bars).

All subcellular membrane fractions were monitored for protein recovery and the activities of marker enzymes, namely arylsulfatase A [12], pglucuronidase [ 131 and a-fucosidase [ 141 as markers for lysosomes, 2’,3’-cyclic nucleotide-3’-phosphohydrolase [ 151 for myelin, cerebroside sulfotransferase [ 161 as a Golgi marker, cytochrome c reductase [17] for mitochondria and acetylthiocholine esterase [ 181 for synaptosomes. Quite similar results have been obtained previously [7]. Brain subfractions comprised a microsomal and soluble fraction, a large and small myelin fraction, a a synaptosomal/ lysosomal and a nuclear, mitochondrial/lysosomal fraction [8]. The subfractionation of microsomes from the 17 500 x g supernatant yielded fractions enriched with light myelin, plasma membranes, Golgi vesicles, endoplasmic reticulum membranes and Golgi lysosomes [7,9]. The biosynthesis of brain sulfogalactosyl glycerolipids seems to occur exclusively in the oligodendrocytes, since it was shown in rat brain that monogalactosyl diacylglycerol synthesis is associ-

298

ated with this cell type [19]. Both the synthesis of monogalactosyl diacylglycerol, the precursor of sulfogalactosyl diacylglycerol [20] as well as the sulfotransferase enzyme which catalyzes the sulfate transfer from 3’-phosphoadenosine-5’-phosphosulfate to monogalactosyl diacylglycerol is located in the microsomal fraction [21]. In developing rat brain Inoue et al. [20] demonstrated a temporal relationship between the microsomal and myelin glactosyl diacylglycerol pool, suggesting a transfer of galactosyl diacylglycerol from their microsomal site of synthesis to their deposition in myelin. Our kinetic results also suggest that sulfogalactosyl glycerolipids are synthesized in the microsomes and, more specifically, in the Golgi and endoplasmic reticular membrane complex. During the first 2 h the newly synthesized lipid appears rapidly in microsomal plasma membranes and simultaneously in fractions containing brain lysosomes, whereas the proportion of labeled lipid decreased in the microsomal fraction between 1 and 2 h. After 2 h the radioactivity decreased in both lysosome-containing fractions and a simultaneous increase of labeled lipid was noted in all myelin fractions (premyelin, large and small myelin). In the soluble fraction no [35S]sulfogalactosyl glycerolipids were detectable. These results suggest that the lipid is synthesized in the endoplasmic reticulum and Golgi region and then is transferred by a vesicular transport to the premyelin and incorporated into the myelin membranes. The vesicles are associated with the lysosomes, and the lipid, being within lysosomes, could be partially degraded. This is supported by the observation that whereas more than 50% of the peak radioactivity disappeared from the lysosome-containing fractions between 2 and 3 h after labeling (Table I, P,B and P,C fractions), only about 25% reappeared in the myelin fractions (P, A and P, A). As suggested earlier for brain sulfatide [7], the present data indicates that the sulfogalactosyl glycerolipids may be transported in vesicles to the myelin membranes. The molecular form by which sulfogalactosyl glycerolipids occur in these vesicles has not been characterized up to date, but they may well be bound to a protein in analogy to the sulfatide lipoprotein complex found in brain [22,23] and in peripheral nerve [24]. The existence of a

free cytoplasmic lipoprotein complex appears to be an experimental artefact resulting from lysing the vesicles [7]. ‘As known from several studies in rat brain, sulfated glycerolipids occur in two forms, as a diacyl and alkylacyl derivative [l-4]. According to Ishizuka et al. [4] the molar ratio of the diacyl and alkylacyl forms changes with age. Up to 19 days more of the diacyl type is present. After 22 days the alkylacyl type predominates in whole rat brain. From these developmental data [4] and from turnover studies of radioactively labeled total brain sulfogalactosyl glycerolipids [3] it was deducted that the diacyl form has a higher turnover than the alkylacyl derivative and that the latter may be the more stable component of myelin. Unfortunately, we could not determine the two forms separately because of the scarcity of material in the subcellular fractions. The myelin sulfogalactosyl glycerolipids appeared to have a very rapid turnover with a half-life of few hours, in contrast to myelin sulfatide which has a very slow turnover with a half-life of several weeks [25-271. In spite of these differences, both sulfolipids follow the same subcellular route after their synthesis in the Golgi endoplasmic reticular membranes. From there, both sulfolipids are transferred by vesicles associated with lysosomes to their site of incorporation, the myelin. During this subcellular transport part of the lipids is degraded by lysosomal function. This double role of lysosomal function could be a general mechanism in the regulation of the net content of brain myelin glycolipids. Acknowledgements This work was supported by the Swiss National Foundation (Grants 3.666.80 and 3.143.81). We thank L. Ramseier and B. Salvisberg for their skillful technical assistance and Dr. R. Reynolds for his help in the revision of the manuscript. References I Flynn, T.J., R.A. (1975) 2 Levine. M., J. Biochem.

Deshmukh, D.S., Subba Rao, G. and Pieringer. Biochem. Biophys. Res. Commun. 65. 122- 128 Kornblatt. M.J. and Murray, R.K. (1975) Can. 53. 679-689

299

3 Pieringer,

J., SubbaRao,

G., Mandel,

P. and Pieringer,

R.A.

(1977) B&hem. J. 166, 421-428 4 Ishizuka, I., Inomata, M., Ueno, K. and Yamakawa, T. (1978) J. Biol. Chem. 253, 898-907 5 Burkart, T., Caimi, L., Herschkowitz, N.N. and Wiesmann, U.N. (1983) Dev. Biol. 98, 182-186 6 Burkart, T., Hofmann, K., Siegrist, H.P., Herschkowitz, N.N. and Wiesmann, U.N. (1981) Dev. Biol. 83, 42-48 7 Burkart, T., Caimi, L., Herschkowitz, N.N. and Wiesmann, U.N. (1982) J. Biol. Chem. 257, 3151-3156 8 Eichberg, J., Whittaker, V.P. and Dawson, R.M. (1964) Biochem. J. 92, 91-100 9 Siegrist, H.P., Burkart, T., Wiesmann, U.N., Herschkowitz, N.N. and Spycher, M.A. (1979) J. Neurochem. 33,497-504 10 Folch, J., Lees, M. and Sloane-Stanley, G.H. (1957) J. Biol. Chem. 266, 497-509 11 Lowry, O.H., Rosebrough, N.J., Farr, A.L. and Randall, R.J. (1951) J. Biol. Chem. 193, 265-275 12 Baum, H., Dodgson, K.S. and Spencer, B. (1959) Clin. Chim. Acta 4, 453-455 13 Woolen, J.W. and Walker, P.G. (1961) Proc. Assoc. Clin. B&hem. 2, 14- 15 14 Robinson, D. and Thorpe, R. (1974) Clin. Chim. Acta 55, 65-69

15 Weissbarth,

S., Maker,

H.S.,

Lehrer,

G.M.,

Schneider,

S.

and Bomstein, M.B. (1980) J. Neurochem. 35, 503-505 16 Burkart, T., Siegrist, H.P., Herschkowitz, N.N. and Wiesmann, U.N. (1977) B&him. Biophys. Acta 483, 303-311 17 Hodges, T.K. and Leonard, R.T. (1974) Methods Enzymol. XxX11, Part B, 397-398 18 Ellman, G.L., Courtney, K.D., Andres, V. and Featherstone, R.M. (1961) Biochem. Pharmacol. 7, 88-95 19 Deshmukh, D.S., Flynn, T.J. and Pieringer, R.A. (1974) J. Neurochem. 22, 479-485 20 Inoue, T., Deshmukh, D.S. and Pieringer, R.A. (1971) J. Biol. Chem. 246, 5688-5694 21 SubbaRao, G., Norcia, L.N., Pieringer, J. and Pieringer, R.A. (1977) Biochem. J. 166, 429-435 22 Herschkowitz, N., McKhann, G.M., Saxena, S. and Shooter, E.M. (1968) J. Neurochem. 15, 1181-1188 23 Yahara, S., Singh, I. and Kishimoto, Y. (1980) Biochim. Biophys. Acta 619, 177-185 24 Pleasure, D.E. and Prockop, D.J. (1972) J. Neurochem. 19, 283-295 25 Pritchard, E.T. (1966) J. Neurochem. 13, 13-21 26 Jungalwala, F.B. (1974) J. Lipid Res. 15, 114-123 27 Farooqui, A.A., Rebel, G. and Mandel, P. (1977) Life Sci. 20, 569-584