Gangliosides and neutral glycolipids of human adrenal medulla

480

Biochimica et Biophyaico Acta, 618 (1980) 480-486 @ Elsevier/North-Holland Biomedical Press

BBA 57572

GANGLIOSIDES AND NEUTRAL GLYCOLIPIDS OF HUMAN ADRENAL MEDULLA

TOSHIO ARIGAa, SUSUMU AND0 b, ATSUSHI TAKAHASHI TADASHI MIYATAKE *ld

c and

a Department of Biochemirtry and Metaboliom, Tokyo Metropolitan Institute of Medical Science, Honkomagome, Bunkyo-ku, Tokyo 113, u Department of Biochemistry, Tokyo Metropolitan Inetitute of Gerontology, Sakae-cho, Itabashi-ku, Tokyo 173, c Department of Pathology, Jichi Medical School, YakushQi, Tochigi 329-04 and d Department of Neurology, Jichi Medical School, Yakushiji, Tochigi 329-04 (Japan) (Received August 24th, 1979) Key words: Ganglioside;

Neutral glycolipid;

(Human adrenal medulla)

summary Glycolipids were isolated from human adrenal medulla by DEAE-Sephadex A-25 and Iatrobeads column chromatography. The lipid-bound sialic acid was about 234 a/g fresh tissue. The ganglioside fraction contained two major gangliosides which accounted for 93% of the total lipid-bound sialic acid. They were identified as GMJ, N-acetyhmuraminylgalacto8ylglucosylceramide and GD3, ZV-acetylneuraminyl N-acetylneuraminylgalactosylglucosylceramide on the basis of cochromatography with authentic standards, sugar composition aIialySi& and neuraminidase digestion. GMJ, ~-aCe~~euraminy~~aC~8y~UCO8ylglucoaylcenuide and GD3, ZV-acetylneuramin yl N-aCety~eUr~inylgahWtO8ylglUCO8ylCeramide occurred in a ratio of approximately 3 : 2, and the ratio seemed to be rather constant irrelevant of age and sex differences. The neutral glycolipid fraction consisted of GL la, glucosylceramide (Is%), GLlb, galactosylceramide (23%), GL2,, lactosylceramide (27%), GL3, digalaCto8ylglUCOBylCe~ide (2OW), and GL,, globoside (12%). The major fatty acid8 of all these glycolipids were 16 : 0,18 : 0,22 : 0,24 : 0 and 24 : 1. Introduction Adrenal medulla contains chromaffin granules, which store high concentrations of catecholamines. These are released into the blood stream in response to Abbreviationa: GLla, ~ucosylcemmide; GLlb. &ctorylceramide: GLh. kctosylcerrmide; GL3. ~~ctorylQlucosylcemmide; 04. doborlde; GM3, N-acetylneuaminylgalactosyWucorylceramide; GD3. N-ecetylneurPminyl N-ncetylneuraminylgalactosylglucosylceramide. The nomenclature for other eangliosides is also based on the system of Svennerholm [ 11.

Ledeen et al. [6,7] found that hmnatoside (GMs) was the major ganglioside of bovine adrenal medulla. Rice and Yu (81 reported considerable variations in hematoside concentrations among adrenal medulla from several mammals. The hematoside was reported to be located in the chromaffin granules ]S]. In this paper, we describe the characterixation of the major gangliosides of human adrenal medulla, and the variations of ganglioeide patterns among several human cases. We have also analysed the neutral glycolipid patterns and their fatty acid compositions m comparison with those of the gangliosides. To our knowledge, this is the first report on the neutral glycolipid composition of this tissue. Materials and Methods

Glycolipids were isolated from adrenal medullae which were taken from 9 patients without any known neurological or endocrinological disorders. Lipids were extracted from adrenal meduiIa (2 g) with 40 ml each of chloroform/ methanol (I : 2 and 1 : 1, v/v) at room temperature. The extracts were applied to a DRAE&phadex A-26 column (acetate form; bed volume, 6 ml) [IO]. The neutral lipids were eluted with 20 ml of methanol, and acidic lipids (including sulfatidea and gangliosides) were then recovered by an elution with 20 ml of 0.2 M sodium acetate in methanol. Roth lipid fractions were subjected to mild alkaline treatment with 0.1 N methanolic NaOH at 40% for 2 h. The liberated fatty acid methyl e&em were removed by the extraction with n-hexane. The neutral lipid fraction was evaporated to dryness, and the salt13were removed by partitioning according to Folch et al. [ll]. The neutral lipid fraction was further applied to an Iatroheads column (latron Lab. Inc., Tokyo, Japan) in order to obtain a purified neutral glycolipid iraction [ 123. The acidic lipid fraction was evaporated to dryness and dissolved in 0.6 ml of distilled water. The solution was desalted by Sephadex G-60 column chromatography (bed volume: 40 ml) to yieId a crude ganglioside &action [13 1. The lipid-bound sialic acid content was estimated by the method of Svennerhohn [ 141. In order to isolate pure GM3 and GDJ, the pooled ganghoside fraction was applied to an Iatrobeads column whioh was eluted with a linear gradient of chloroform/methanol/ water (from 66 : 36 : 4 to 36 : 66 : 4, v/v) [16]. The neutral glycolipid composition was examined by thin-layer chromatography (TLC). An upper half part of a silica gel TLC plate (E. Merck, Darmstadt, F.R.G.) was sprayed with aqueous 1% sodium borate solution, and the plate was then activated for 2 h at 110°C. After the aamplee had been applied to the plate, it was developed &et with chlorofo~/methanol/2.6 N ammonia (66 : 36 : 8, V/V). Thb step moved cerebro&es into the borate impregnated layer. The piate was dried in air, and then redeveloped with be/n-hexanelacetic acid (40 : 10 : 1, v/v) to remove less .polar contaminants, which were present near the cerebroside, in order to facilitate later densitometric scanning. Glyco-

482

lipids were visualized by heating at 125 * 2°C for 30 min after spraying with 0.25% sodium dichromate in 15% H#O, [ 161. The chromatogram was scanned at 440 nm with a Shimadzu dual wave-length TLC densitometer, model 910. The ganglioside composition was examined by TLC with the following solvent systems [lo]: (a) chloroform/methanol/water (55 : 45 : 10, v/v) containing 0.02% CaClz - 2 H,O; (b) chloroform/methanol/2.5 N ammonia (60 : 40 : 10, v/v). The gangliosides were visualized by heating at 95 t 2°C with resorcinol-HCl reagent and the chromatogram was scanned at 580 nm. Compositional analysis was carried out by gas-liquid chromatography (GLC) as follows: The lipids were methanolyzed for‘18 h at 75°C with 1 ml of 2.5% methanolic hydrochloride. Fatty acid methyl esters were extracted with n-hexane, and portions of the extracts were injected into a column (1.5 m X 3 mm) of 10% diethylene glycol succinate maintained at 19O’C. The lower methanolic layers were evaporated under a stream of nitrogen and dried in vacua. The dried residues were reated with 0.1 ml of acetic anhydride at room temperature for 1 h in order to re-l\r-acetylate the aminosugars [ 171. The solvent was evaporated under a stream’ of nitrogen. The residue was further dried in vacua, and then trimethylsilylated with 0.1 ml of hexamethyldisilazane/ trimethylchlorosilane/pyridine (2 : 1 : 5, v/v) at 60°C for 20 min to form N-acetyl0trimethylsilyl derivatives of saccharides [ 18 1. Aliquots of the reaction mixtures were injected into a column of 3% SE-30 programmed at 3”C/ min from 175°C to 230°C. For long-chain base analysis, the lipids were subjected to methanolysis with 2 N HCI in 82% methanol [19], and then trimethylsilylated with bis(trimethylsilyl)trifluoroacetamide/trimethylchlorosilane (9 : 1, v/v). Longchain bases were analysed by using a column of 1% OV-101 at 230°C. N-Acetyl and glycolyl neuraminic acids were determined by GLC as described by Yu and Ledeen [ 201. Neuraminidase treatment of gangliosides (20 erg) was carried out in 0.2 M sodium citrate buffer (pH 5.6) containing 0.5 units of neuraminidase (EC 3.2.1.18; Clostridium perfringens, Sigma Type IV) at 37°C for 2 h. The glycolipid products were examined by TLC. Results and Discussion The ganglioside patterns of human adrenal medulla are shown in Fig. 1, and the data obtained from 9 cases are summarized in Table I. The level of lipidbound sialic acid was about 234 pg (640 nmol)/g fresh tissue, which was approximately equivalent to 880 a of gangliosides. The ganglioside content is very close to that reported for bovine adrenal medulla (223 gg sialic acid/g fresh tissue) [6]. The ganglioside level of adrenal medulla, which is a part of autonomic nervous system, seems to be relatively high as compared to those for other visceral organs such as muscle, liver and lung [ 21-351. It might be interesting to notice that the ganglioside content of adrenal medulla is almost comparable to that of mammalian brain white matter [lo]. Human adrenal medulla contained two major ganghosides (approx. 93% of total gangliosides) as reported by Price and Yu 181. The two major gangliosides cochromatographed on TLC with authentic GM3 and GDB (Fig. 1). The fast migrating ganglioside contained glucose, galactose, and sialic acid in a molar ratio of 1.00 :

483

2

3

4

5

6

7

6

9 101112

Fig. 1. Thin-layer cbromatogram of ltumau adrenal medulla ganglioddes. Lone 1. normpl human Gay matter gnngltodde mixture; 2-10, adrenal medulln gang&sides from 9 different humaa subWte; 11. authentic GM3 which WM Lolated from a patient with Gaucher’s disease: 12, GD3 isolated from bovine brain. The pint8 WM developed with eblorof&Wmetbanol/wa~ (SS : 45 : 10. v/v) containing 0.02% Cdl2 * 2 W20. The ganpliosidsr were vieuaked with recorcinol-WC1 reagent.

1.03 : 0.97. Its structure was therefore consistent with that of GM3. The second ganglioside contained glucose, galactose, and sialic acid in a molar ratio of 1.00 : 0.96 : 1.97, suggesting that it had a structure of GD3. Both gangliosides yielded the same GLza backbone by neuraminidase treatment as shown TABLE I GANGLIOSIDES

OF HUMAN ADRENAL

MEDULLAE

Case No.

Age

Sex

Lipid-bound sialic acid *

GM3 **

GDJ **

GM3 + GDg **

GM3 lGDj

1 2 3 4 6 6 7 8 9

66 14 66 32 73 65 57 39 68

F M M F M M M M F

197.4 221.2 239.4 203.2 234.0 248.3 321.0 263.6 178.0

62.3 54.2 55.4 60.0 55.1 57.2 66.8 64.2 60.8

38.7 89.6 38.0 29.3 89.9 36.8 37.5 38.2 34.2

91.0 98.8 91.4 89.8 95.0 94.0 94.8 $2.4 95.0

1.86 1.37 1.54 2.06 1.38 1.65 1.61 1.42 1.78

284.0 f 42.2

66.2 f 2.8

38.7 f 8.3

92.9 t 2.1

1.65 i 0.23

Mean f 8.D. * #&g/efre& tissues. ** % of total panlliosides.

484

in Fig. 2. The rem& therefore lend support to our structural assignment. The

sialic acid of these ganglia was found to be Z’Aecdy~ acid by GLC. In contrast to human adrenal medulla, bovine adrenal medtrth gangliosides have been reported to contain a high concentzation of Nglycolylneuraminic acid, in addition to the N-acetyi type [6,8 J. In human adrenal medulla, GM3 and GDJ occurred in a ratio of approximately 3 : 2. The ratio seemed to be constant among various age and sex groups. The neutral glycolipid fractions contained CL,, (18%)‘ GLtb (23%), Gl& (2756), GL3 (29%) and GL, (12%). The sulfatide, which should be recovered in the same acidic lipid fraction as gangliosides, was not detected by TLC. Table II shows the fatty acid compositions of various glycolipids. The maor fatty acids were 16 : 0, 18 : 0, 22 : 0, 24 : 0 and 24 : 1. The fatty acid 24 : 0 was predominant in GL,,, GLlb, and GL+ On the other hand, the fatty acid 24 : 1 was much more abundant in Gl+,. Since the fatty acid 24 : 1 was also concentrated in GL3 and GM3, it suggested #at GL,, might serve as a precursor for GLJ and GM3. The presence of relatively high concentrations of long-chain fatty acids (22 : 0, 24 : 1, 24 : 0) in adrenal medulla gqgliosides is interesting because brain gangliosides tend to have a preponderance of 18 : 0. Both GM3 and GDs contained only Cl8 long-chain base, which was composed of 97% sph~gos~e and 3% sphinganine. No C20 long-chain base could be detected. The report by Dreyfus et al. 191 suggests a possible function for gangliosides. These workers describe the. ganglioside of chromaffm granule membranes isolated from bovine adrenal medulla as having exclusively GM3 containing N-acetylneuraminic acid. They further suggest that GM3 might be involved in the releasing of catechokmines from chromaffin granules by forming a complex between ganglioside sialic acid and Ca*‘. Further studies on the sub-

Fig. 2. Thtn-layor chmmt~gmu of glycolipfd producta of 6anJioddes after neuramlnfduc treatment. Lane 1. authsatic l3La from pi# rr~throeyta membranea: 2. human ulronal medulla mlodde GM3 after notreatment: 8. bumln M medulla 6aauHorlde (3D3 after neurarnhddus treatmmti 4. ~uthaatlc 0%. GL3 and 04 horn pig styttuocyta membruw. The plate was dewiotmd with ch&roform/methaaol/watcr (65 : 26 : 4, v/v). The chromatogram wu vhalized w&h m*Oae-H2204 regent.

TABLE II FATTY

ACID COMP08ITlONS

OF HUMAN

AWMU

YtwHw

QLYCOId?~

tr,tracrunountl6nthAnO.l%

Oatty acid

(W

14:o 18:O 18:l 18:O 18:l 20:o 2O:l 22:o 22:1 23 :0 23: 1 24:0 24:l 2ci:o

Gha

GLlb

GL2.

GL3

-4

(-3

GD3

2.1 17.0 0.4 14.4 2.1 4.4 1.1 18.8 tr 7.0 4.9 23.0 4.1 tr

0.5 9.9 0.2 5.8 0.6 3.1 0.1 28.3 tr 11.8 0.4 27.6 14.1 tr

tr 19.8 0.1 4.4 1.2 2.3 0.8 12.8 0.7 4.4 1.1 15.4 37.2 ti

0.1 17.8 0.4 8.1 2.0 4.8 1.0 14.6 0.9 3.3 0.7 18.0 29.2 t.r

1.8 10.8 0.2 5.6 1.0 4.6 0.2 24.4 tr 8.0 0.9 24.1 9.8 tr

t? 18.0 0.8 5.1 1.9 2.9 0.1 20.4 tr 8.5 0.1 19.4 24.0 0.8

tr 18.0 1.1 11.4 2.9 2.2 0.2 18.8 tr 7.8 0.1 20.2 17.1 tr

cellular distribution clarify ita function.

of gangliosides in human adrenal medulla are required to

Acknowledgements The authors would @e to thank Dr. Robert K. Yu (Department of Neurology, Yale University, School of Medicine, New Haven, CT, U.S.A.) for vaiuabie discussions in preparing the manuscript. The authors also ac%nowledge Dr. Mieko Oshima (Department of B&hem&y, KiWato UniversiQ, School of Medicine, Sagamihara) for the generous gift of authentic GM3 f&m a patient with Gaucher’s disease. Reference8 1 2 3 4 5 8 7 8 9 10 11 12 I3 14 16 18 17 18 19 20 21

Svaaazbolm.L. (1983)J.Neumchem. 10.813-823 Foldberg. W..Mlnz. B. ~dT~ddm~.H.(1934)J.PhydoL 81.288-304 Douglu.W.W.(l988)Br.J.PhumroL 34.461-474 ~puhi,K.md~~.A.(1974)Biochlm.BIo~h~~ Acta 337.107-117 Rahmaan.H.,Rorn~,H.~d~r.H.(l978)J.Th~or.loL 87.231-237 Ldeen.R..Salamau. K. mdCabr8a.M. (1988)Bioche~Mry7.2287-2296 Pricr.H.,Kundo.S.and Ind~.R.(1976)~~mk~14,1~12-1618 Pdcr.H.C.and Yu. R.K. (1978) Cow. Biochom. PhydoL MB. 401+M Drmrfua H.. Aunts, D..Hart& 8.and Aluukl, ?.(1977) moehIm. &DhYS. Acta

489.89-97

Aado. 8.. Chuu, N.-C. and Yu. B.K. (1979 ArmI. Biocbun. 89.437-460 Folch. J.. Lees,M. and SlasBMr. G.H. (1967) J. Biol. Cham. 22th 497-SO9 Ando. 8.. hobo, M. and NaSal. Y. (1976) RIoddm. BioDh9r Acta 424.98-105 Ueno,K.. Ando.S.rd Yu.RL(197S)J.~&w.l9,B+Sd71 svennerholm.L.(1967) Btochim. INo~hyr. AataU. 604-811

Momad.T..Ando.8.~dN~Y.(lS76)Bbahtm. ~Dh9S.hb 441.48(1--407 Ron, K..Aado. 8..T~.Y.~N~Y.(1979)Roo.~.cOnt.BLoabrm.Ilpi~il.~1 Hobnr, E.W.and O'B&n. J.S. (lS7S) Bioahaa. J. 171,94b9S3 Cmxm. ILL. and.Gw=. R.C. Clerf, 3. r@qh& 0.3Sl-q Swo&h. C.C. mtd MoaataDi. E. <19&I J. L@dRa. 1.46-47 Yu. R.t ti brdean.R.W.‘flWO) J. I&d Rw. 11.66641tI

J.-l!.. Ummak, B.-T. md Vmlu. M.-F. Laucs, A..Yw. 6264SS 22 Serkted.T.N.. Aado.S.aadYu.R.K.(1978)J.Id~Rea.1S.63S-U48 23 Iwu11ori.M.mdNyli.Y.(1978)J.BIochem.(Tok~o)84.1809-l816 v

Dh98.

-

h@t%

(lS72)

-.

Rio-