Formation of ether lipids and wax esters in mammalian cells Specificity of enzymes with regard to carbon chains of substrates

Biochimicu

et Biophysim

Elsevier Biomedical

Actu, 71

I

(1982)

197

197-207

Press

BBA 5 1082

FORMATION

OF ETHER LIPIDS AND WAX ESTERS IN MAMMALIAN

SPECIFICITY

OF ENZYMES WITH REGARD TO CARBON CHAINS OF SUBSTRATES

NIKOLAUS

WEBER

Bundesanstalt

fiir

(Received

Key words:

and INGOLF

Fettforschung,

November

CELLS

RICHTER

Piusuilee

68.7b,

D-4400

Miinster

(F. R.G.)

4th, 1981)

Enzyme specific+;

Homologous

ulcohol;

Octudecenol;

Ether lipid;

Wax ester; (Mummulian

cell)

The incorporation of radioactivity from individual constituents of an equimolar mixture of saturated straight-chain alcohols (14: 0, 16: 0, 18: 0, 20: 0) and of nearly uniform mixtures of isomeric cis- or trans-octadecenols (A* -At6) into alkyl, alk-1-enyl and acyl moieties of diradylglycerophosphocholines and diradylglycerophosphoethanolamines and into alkyl and acyl moieties of wax esters was studied in rat brain as well as in L 1210 and S 180 ascites cells. The pattern of incorporation of radioactivity from the substrates into alkyl and alk-1-enyl moieties of ether phospholipids and into alkyl moieties of wax esters reveals the following (1) The enzymes catalyzing the biosynthesis of alkylacylglycerols, the common intermediates of cholinephospholipids and ethanolaminephospholipids, have no substrate specificity with regard to position of the double bond of either cis- or &ans-octadecenols or of intermediate ether lipids derived therefrom. (2) CDPcholine:diradylglycerol cholinephosphotransferases exhibit a strong preference for alkylacylglycerols with cis-8, cis-9 and &lO-octadecenyl moieties, but no preference for the double bond position in the &a.ns-octadecenylacylglycerols. (3) CDPethanolamine:diradylglycerol ethanolaminephosphotransferases have no substrate specificity with regard to position of the double bond in cis- or transoctadecenyl moieties of alkylacylglycerols. (4) The enzyme systems catalyzing the biosynthesis of alkylacylglycerophosphocholines and alkylacylglycerophosphoethanohunines exhibit substrate specificity with regard to chain-length of saturated alcohols and intermediate ether lipids derived therefrom. (5) Alkylacylglycerophosphoethanohunine desaturase and (6) wax ester synthase are highly specific for alkylacylglycerophosphoethanohunines and long-chain alcohols, respectively, with regard to chain-length of saturated alkyl moieties, but not with regard to position of double bonds of cjs- or bans-octadecenyl moieties.

Introduction

the biosynthesis of ether lipids and wax esters with regard to chain-lengths of substrates have been investigated [g-lo], but little is known about the specificities with regard to position of double bonds in the alkyl chain of substrates [ 111. In the present study an equimolar mixture of saturated long-chain ( l-‘4C)-labeled alcohols as well as mixtures of positional isomers of cis- or truns-[ l-‘4C]octadecenols were applied to rat brain as well as to ascites cells of Leukemia 1210 (L 1210) and Sarcoma 180 (S 180). We have used mixed

Long-chain alcohols and their derivatives, such as ether lipids and wax esters, are found in normal and neoplastic mammalian cells [l-3]. It is known that saturated and unsaturated long-chain alcohols are utilized in the biosynthesis of ether lipids and wax esters [l-4]. Rat brain and murine ascites cells are suitable systems for studying the formation of ether lipids and wax esters from long-chain alcohols [5-71. Specificities of enzymes involved in OOOO-~oO/~2/ooOO-00/$02.75

‘CC1982 Elsevier

Biomedical

Press

19x

subtrates in order to study the metabolism of long-chain alcohols under competitive conditions. The aim of this study was to assess qualitatively the substrate specificity of enzymes involved in the biosynthesis (1) of alkyl and alk-1-enyl glycerophospholipids and (2) of wax esters, both with regard to chain length and position of the double bond in alcohols or intermediate substrates derived therefrom. A preliminary note of this work, in part, has appeared recently [ 121. Materials and Methods Chemicals. All reagents and adsorbents were purchased from E. Merck AG, Darmstadt, F.R.G. Distilled solvents were used throughout. Reference substances for chromatography were obtained from Applied Science Laboratories, Inc., State College, PA, U.S.A. [l-‘4C]Oleic acid, 2.08 GBq/mmol, [ 1 -‘4C]linolenic acid, 2.22 GBq/mmol, and K14CN, 2.19 GBq/mmol, were purchased from Amersham Buchler, Braunschweig, F.R.G. of substrates. cis - 9 - [ 1 -14C] Preparation Octadecenol, 1.04 GBq/mmol, was prepared from [ l-l4 Cloleic acid by reduction of the corresponding methyl ester with LiAlH, [13]. An equimolar mixture of ( l-‘4C)-labeled tetradecanol, hexadecanol, octadecanol and icosanol, 0.37 GBq/mmol, was prepared via chain elongation of odd-chain alkyl methanesulfonates using K14CN [ 141. A nearly uniform mixture of positional isomers (A’-A”) of methyl trans-[ l-‘4C]octadecenoates, 0.38 GBq/mmol, was prepared from [l-‘4C]linolenic acid by partial hydrogenation of the methyl ester and subsequent fractionation by argentation thin-layer chromatography [ 151. A nearly uniform mixture of positional isomers (A8 -A”) of methyl cis-[ l“C]octadecenoates, 0.38 GBq/mmol, was obtained from the isomeric methyl trans- [ l“C]octadecenoates by trans-cis equilibration and argentation chromatography [ 151. Mixtures of cis[ 1-‘4C]octadecenols, 0.38 GBq/mmol, and of trans-[ l-‘4C]octadecenols, 0.38 GBq/mmol, each of known composition with regard to positional isomers, were prepared from the corresponding methyl esters by reduction with LiAlH,. Substrates included cis-9-[ 1 -‘4C]octadecenol ( 1.04 GBq/mmol), an equimolar mixture of [l“C]tetradecanol, [ 1 -‘4C]hexadecanol, [1-

I4 Cloctadecanol and [ 1 -I4 Clicosanol (0.37 GBq/mmol), as well as mixtures of isomeric cis-[ l“C]octadeceno 1s ( 0.38 GBq/mmol) and of isomerit trans-[ l-‘4C]octadecenols (0.38 GBq/mmol). Application of substrates. Albino rats of the Sprague-Dawley strain, 15 days old at the day of the experiment, were purchased from Mus-Rattus, Brunnthal, F.R.G. Approximately 55 kBq of substrate ( l-‘4C)-labeled alcohols suspended in 20 ~1 of aqueous Tween 20 solution (35 mg/ml) was injected into each animal through the sutura sagittalis in the manner described by Bickerstaffe and Mead [16]. The animals were killed by cervical dislocation, 10 h after injection, and the brains were removed immediately. For each substrate the brains from three animals were pooled. Leukemia 12 10 and Sarcoma 180 ascites cells were grown in female NMRI mice (Mus-Rattus, Brunnthal, F.R.G.), average weight 20 g, and transplanted weekly. The cells were harvested from the peritoneal cavity 5 days after inoculation. Cells obtained from six mice were pooled and separated from the ascites fluid by centrifugation at 480 g for 5 min at room temperature. The cells were washed with tissue culture medium, TCM 199, supplemented with Hanks’ salts and NaHCO, (Seromed GmbH, Miinchen, F.R.G.) and resuspended for incubation. Suspensions of 6 . lo* L 12 10 ascites cells or 4 * lo8 S 180 ascites cells in 30 ml TCM 199 medium were used for incubations. The substrate ( l-‘4C)-labeled alcohols, 110 kBq, dissolved in 30 ~1 ethanol, were added to the cell suspensions and incubations were carried out by shaking at 37°C for 3 h under aerobic conditions. Lipid analyses. Radioactivity was determined in a Packard Tri-Carb C 2425 liquid scintillation counter (Packard Instruments Company, Downers Grove, IL, U.S.A.) using Aquasol(NEN Chemicals GmbH, Dreieichenhain, F.R.G.) as scintillation solution. Thin-layer radiochromatograms were assayed by means of a Berthold Scanner LB 2760 (BF-Vertriebsgesellschaft, Wildbad, F.R.G.). Autoradiograms were obtained by exposing AgfaGevaert Curix RP 1 X-ray films (Agfa-Gevaert N.V., Mortsel, Belgium) to thin-layer chromatograms for a period of 3 weeks. Radio-gas chromatography was conducted in a Perkin-Elmer F 22 gas chromatograph (Perkin-Elmer and Co GmbH, oberlingen, F.R.G.) equipped with a thermal con-

199

ductivity detector, in combination with a Packard Gas Proportional Counter, Model 894 (Packard Instruments Company). Peak areas were measured with an Autolab System I Spectra Physics Reporting Integrator (Autolab System GmbH, Darmstadt, F.R.G.). Separations were carried out on glass columns (1.8 m X 3 mm) packed with 3% OV-101 on Gas Chrom Q, loo-120 mesh (Applied Science Laboratories, Inc.) using helium (40 ml/min) as carrier gas. The temperature was programmed from 160 to 260°C (4”C/min) for the analysis of alkylacetates or from 220 to 260°C (4”C/min) and then kept at 260°C for 10 min for the analysis of alkyldiacetylglycerols. The temperature for the analysis of products of ozonolysis obtained from unsaturated alkylacetates and alkyldiacetylglycerols was programmed from 80 to 260°C (4”C/min) and from 160 to 260°C (2”C/min), respectively. The lipids of rat brain and of murine ascites cells were extracted according to an established procedure [17]. The total lipids were fractionated into diradylglycerophophocholines, diradylgly cerophosphoethanolamines and neutral lipids on layers of silica gel H with chloroform/ methanol/water (65:25:4, v/v) [ 181, and the distribution of radioactivity in the various fractions was determined with the thin-layer chromatogram scanner. The lipid fractions were then eluted from the adsorbent with chloroform/methanol/water (1:2:0.8, v/v). The purity of the phospholipid fractions was checked by two-dimensional thin-layer chromatography [ 191 and subsequent autoradiography. Diradylglycerophosphocholines and diradylglycerophosphoethanolamines were reduced with LiAlH, [ 131. The resulting mixtures were separated on layers of silica gel H using hexane/diethyl ether (2:8, v/v), the distribution of radioactivity was determined by scanning and fractions of long-chain alcohols and those of alkylglycerols plus alk- 1-enylglycerols were isolated. The alk- lenyl moieties of alk- 1-enylglycerols were converted to aldehydes by acid hydrolysis [20] and these were reduced with LiAlH, to the corresponding longchain alcohols. Mixtures of long-chain alcohols and alkylglycerols were separated and isolated after scanning as described above. The neutral lipids were fractionated on layers of

silica gel H using hexane/diethyl ether/acetic acid (85: 15: 1, v/v) [21] and the distribution of radioactivity in the various lipid classes was determined as described above. The fractions containing steryl esters and wax esters were eluted from the adsorbent with water-saturated diethyl ether, dried, and resolved on layers of magnesium oxide using hexane/diethyl ether/ ethyl acetate (75:25: 1, v/v) [22]. The distribution of radioactivity in the two fractions was determined by scanning. The fractions of wax esters were eluted from magnesium oxide with water-saturated diethyl ether, dried and transmethylated [23]. The resulting mixtures of long-chain alcohols and methyl esters were separated on layers of silica gel H with hexane/diethyl ether (2:8, v/v) and the distribution of radioactivity in the two fractions was determined by scanning. Long-chain alcohols as well as alkylglycerols were acetylated [24] and the resulting alkylacetates and alkyldiacetylglycerols, respectively, were purified by argentation thin-layer chromatography using hexane/diethylether (8:2, v/v) or (2:8, V/V) as developing solvents. Samples consisting of saturated alkylacetates or alkyldiacetylglycerols were analyzed by radio-gas chromatography directly. Samples consisting of monounsaturated alkylacetates or alkyldiacetylglycerols were added to 50 pg of cis-9-octadecenylacetate and rut - 1 - (cis - 9’ - octadecenyl)2,3 diacetylglycerol, respectively, and analyzed by radio-gas chromatography after reductive ozonolysis as decribed [12,15]. Results c&9-[ l- I4 C]Octadecenol, mixtures of long-chain saturated alcohols and mixtures of isomeric cis- as well as truns-octadecenols, which were separately applied to rat brains as well as to L 1210 and S 180 cell suspensions, were taken up readily by the cells. Approximately 30% of radioactivity of the long-chain alcohols applied to rat brains and neoplastic cells was incorporated into the total lipids excluding unchanged substrates. Table I gives the distribution of radioactivity in lipid fractions of rat brain and murine ascites cells after application of ( l-‘4C)-labeled long-chain alcohols. Radioactivity is incorporated to large ex-

200

TABLE I DISTRIBUTION APPLICATION

OF RADIOACTIVITY OF (I -14C)-LABELED

IN LIPID FRACTIONS OF RAT BRAIN LONG-CHAIN ALCOHOLS

AND

MURINE

ASCITES

CELLS

AFTER

Substrates used were cis-9-octadecenol (1.04 GBq/mmol) as well as mixtures of saturated straight-chain alcohols (0.37
Rat brain

substrates

1210

180

a Including

(%)

Diradylglycerophosphoethanolamines

Triacylglycerols

40

49

2

6

tr

3

40

47

4

5

tr

A

cis-octadecenols Mixture of isomeric rruns-octadecenols

34

60

2

2

tr

22

68

3

3

tr

4

cis-9-Octadecenol

32

58

2

2

tr

6

33

60

2

tr

4

21

12

2

tr

3

22

71

I

tr

5

19

40

IO

tr

26

29

30

II

tr

27

18

41

8

tr

27

22

43

6

tr

26

Mixture of saturated straight-chain alcohols Mixture of isomeric cis-octadecenols Mixture of isomeric nuns-octadecenols Sarcoma

of radioactivity

Diradylglycerophosphocholines

cis-9-Octadecenol Mixture of saturated straight-chain alcohols Mixture of isomeric

Leukemia

Distribution

cis-9-Octadecenol Mixture of saturated straight-chain alcohols Mixture of isomeric cis-octadecenols Mixture of isomeric trans-octadecenols

Alkyldiacylglycerols

Cholesteryl esters

Wax esters

a

alk- I -enyl-diacylglycerols.

tent into diradylglycerophosphocholines and diradylglycerophosphoethanolamines of all three cell types as well as into wax esters of S 180 cells. Minor proportions of label are incorporated into triacylglycerols and alkyldiacylglycerols whereas other lipid classes are practically devoid of radioactivity. Table II shows the distribution of radioactivity

in various radyl moieties of major lipid classes of rat brain and murine ascites cells after application of (l-‘4C)-labeled long-chain alcohols. In all three systems a high proportion of radioactivity occurs in alkyl moieties of glycerophospholipids and wax esters. Substantial amounts of labeled alk-1-enyl moieties are found in glycerophosphoethanolamines only. The proportion of labeled acyl moie-

201

TABLE

II

DISTRIBUTION AND MURINE

OF RADIOACTIVITY IN VARIOUS RADYL MOIETIES OF MAJOR LIPID CLASSES FROM ASCITES CELLS AFTER APPLICATION OF ( I-‘4C)-LABELED LONG-CHAIN ALCOHOLS

RAT BRAIN

Diradylglycerophosphochohnes, diradylglycerophosphoethanolamines and wax esters that had been resolved by thin-layer chromatography (Table I) were isolated. The diradylglycerophospholipids were reduced with LiAlH, and the resulting alkylglycerols, alk-lenylglycerols and long-chain alcohols were analyzed by radio thin-layer chromatography on silica gel H and isolated. The fractions consisting of alkylglycerols plus alk-I-enylglycerols were subjected to acid hydrolysis and reduction with LiAlH,. The ratio of radioactivity in mixtures of alkylglycerols and long-chain alcohols, derived from alk-I-enylglycerols, was determined by radio thin-layer chromatography on silica gel H. Wax esters were transmethylated and the resulting long-chain alcohols and methyl esters were analyzed by radio thin-layer chromatography on silica gel H. The analytical procedures are described in Materials and Methods. tr, trace (
Distribution

of radioactivity

Diradylglycerophosphocholines

Rat brain

Leukemia

Sarcoma

1210

180

(%) in: Wax esters

Diradylglycerophosphoethanolamines

Alkyl

Alk1-enyl

Acyl

Alkyl

AlkI -enyl

Acyl

Alkyl

Acyl

cis-9-Octadecenol Mixture of saturated straight-chain alcohols Mixture of isomeric cis-octadecenols Mixture of isomeric rruas-octadecenols

35

tr

65

25

48

27

>98

tr

24

tr

76

27

40

33

198

tr

14

tr

86

23

49

28

>98

tr

16

tr

84

29

41

30

198

tr

cis-9-Octadecenol Mixture of saturated straight-chain alcohols Mixture of isomeric cis-octadecenols Mixture of isomeric fruns-octadecenols

83

tr

17

74

16

IO

>98

tr

79

tr

21

71

22

7

>98

tr

13

tr

27

80

12

8

>98

tr

70

tr

30

71

15

14

>98

tr

cis-9-Octadecenol Mixture of saturated straight-chain alcohols Mixture of isomeric cis-octadecenols Mixture of isomeric truns-octadecenols

64

tr

36

76

8

16

>98

tr

52

tr

48

80

12

8

>98

tr

55

tr

45

69

18

13

>98

tr

41

tr

59

74

12

14

>98

tr

ties of glycerophospholipids is higher in rat brain than in L 1210 and S 180 ascites cells. The relative proportion of labeled acyl moieties is much higher in diradylglycerophosphocholines than in diradylglycerophosphoethanolamines. In the wax esters of each cell type the label is present almost exclusively in the alkyl moieties. Analyses of alkylacetates derived from alkyl moieties of wax esters or from alk-1-enyl and acyl

moieties of diradylglycerophospholipids as well as of alkyldiacetylglycerols derived from alkylacylglycerophospholipids show that neither elongation or desaturation of all added substrates nor geometrical isomerisation of the added cis- and transoctadecenols had occurred (data not shown). Table III shows the distribution of chain lengths in labeled radyl moieties of phospholipids and wax esters from rat brain and murine ascites cells after

202 TABLE

III

DISTRIBUTION OF CHAIN LENGTHS IN LABELED RADYL MOIETIES OF GLYCEROPHOSPHOLIPIDS ESTERS FROM RAT BRAIN AND MURINE ASCITES CELLS AFTER APPLICATION OF AN EQUIMOLAR SATURATED STRAIGHT-CHAIN (I -14C)-LABELED ALCOHOLS

AND WAX MIXTURE OF

Alkylglycerols and long-chain alcohols derived from diradylglycerophosphocholines and diradylglycerophosphoethanolamines via reduction with LiAIH,, hydrolysis of the vinyl ether bond and subsequent reduction of the liberated aldehydes as well as long-chain alcohols derived from wax esters via transmethylation were isolated by thin-layer chromatography (Table II) and acetylated. The distribution of chain-lengths in the a1ky1diacety1g1ycero1s and alkylacetates was determined by radio-gas chromatography. The analytical procedures were described in Materials and Methods. Values represent distribution of radioactivity (%). Saturated straight-chain alcohols were an equimolar mixture of [ I-‘4C]tetradecano1, [ I-‘4C]hexadecanol. [ I-‘4C]octadecano1 and [I -‘4C]icosanol. Chain length of the homologous

Substrate:

Mixture

of saturated

straight-chain

Rat brain Diradylglycerophosphocholines Diradylglycerophosphoethanolamines

Wax esters Leukemia 1210 Diradylglycerophosphocholines Diradylglycerophosphoethanolamines

Wax esters Sarcoma IX0 Diradylglycerophosphocholines Diradylglycerophosphoethanolamines

Wax esters

radyl moieties

I4

I6

IX

20

25

25

25

25

Alkyl Acyl Alkyl Alk- I -enyl Acyl Alkyl

3s 3 21 X 3 II

44 4X 40 49 21 25

17 43 29 35 6X 34

4 6 IO x 8 30

Alkyl Acyl Alkyl Alk- I -enyl Acyl Alkyl

IS 4x 45 26 26 IX

22 31 4x 67 45 34

3 20 I 7 29 34

Alkyl Acyl Alkyl Alk- I-enyl Acyl Alkyl

71 59 40 20 IX 25

26 32 54 73 39 3X

3 9 6 7 43 32

alcohols

application of an equimolar mixture of saturated straight-chain ( l-‘4C)-labeled alcohols. In all three systems the distribution of chain lengths in labeled alk - 1 -enyl and acyl moieties of alkyl, glycerophospholipids is quite different from the composition of chain lengths in substrates. Labeled radyl moieties with chain-length 20 are incorporated into the glycerophospholipids of all three cell types to a minor extent, if at all. In general, diradylglycerophosphocholines and diradylgly cerophosphoethanolamines contain preferentially labeled alkyl moieties with chain lengths of 14 and 16 carbon atoms. Diradylglycerophosphoethanol-

I _ _ I4 _ _ _ _ _ 5

amines of rat brain contain preferentially labeled alk-1-enyl moieties with chain lengths 16 and 18, whereas in L 1210 and S 180 ascites cells the chain length of 16 carbon atoms is preferred strongly. In contrast to neoplastic cells, labeled acyl moieties with chain length of 14 carbon atoms are practically excluded from glycerophospholipids of rat brain. In the wax esters of all three cell types the radioactivity is present almost exclusively in the alkyl moieties (Table II). Saturated labeled alcohols with chain lengths of 18 and 20 carbon atoms are incorporated preferentially into alkyl

203

esters from rat brain and murine ascites cells after application of mixtures of isomeric cis- and trans[ l-‘4C]octadecenols, respectively, is shown in Tables IV and V. All three cell types exhibit a similar distribution of isomers in labeled radyl moieties of the corresponding lipids. It is striking that alkylacylglycerophosphocholines exhibit a remarkably high enrichment of labeled cis-8, cis-9 and

moieties of wax esters of rat brain (Table III). In contrast, wax esters of neoplastic cells are found to contain preferentially labeled alkyl moieties with chain lengths of 16 and 18 carbon atoms, whereas alkyl moieties with chain lengths of 20 carbon atoms are found to a minor extent. The distribution of positional isomers in labeled radyl moieties of glycerophospholipids and wax

TABLE

IV

DISTRIBUTION OF POSITIONAL ISOMERS IN LABELED RADYL MOIETIES WAX ESTERS FROM RAT BRAIN AND MURINE ASCITES CELLS AFTER ISOMERIC cis-[ I-‘4C]OCTADECENOLS

OF GLYCEROPHOSPHOLIPIDS AND APPLICATION OF A MIXTURE OF

Alkylglycerols and long-chain alcohols derived from diradylglycerophosphocholines and diradylglycerophosphoethanolamines via reduction with LiAlH,, hydrolysis of the vinyl ether bond and subsequent reduction of the liberated aldehydes as well as long-chain alcohols derived from wax esters via transmethylation were isolated by thin-layer chromatography (Table II) and acetylated. The distribution of positional isomers in the alkyldiacetylglycerois and alkylacetates was determined by reductive ozonolyis and subsequent radio gas chromatography. The analytical procedures are described in Materials and Methods, Values represent distribution of radioactivity (W). The isomeric cis-octadecenols were a mixture of cis-[ I-‘4C]octadecenols with a fairly uniform distribution of positional isomers.

Substrate:

Wax esters Leukemia 1210 Diradylglycerophosphocholines Diradylglycerophosphoethanolamines

Wax esters Sarcoma 180 Diradylglycerophosphocholines Diradylglycerophosphoethanolamines

Wax esters

of double

8

9

3

bonds of isomeric

radyl moieties

IO

II

12

13

14

I5

16

15

9

II

22

IO

IO

14

6

6

30

14

IO

IO

9

7

8

6

Alkyl Alk- 1-enyl Alkyl

3 5 4

II 12 14

II 12 12

II

II 12

18 17 21

II 13 II

13 IO IO

13 13 II

9 7 5

Alkyl Acyl

6 3

21 12

17 7

IO II

IO 25

8 IO

6 8

IO 12

6 12

Alkyl Alk- 1-enyl Acyl Alkyl

4 3 5 5

18 14 14 13

IO II 9 8

14 I1 II 10

16 24 I8 23

II 12 II IO

9 8 IO 10

IO II 13 14

8 6 9 7

Alkyl Acyl

8 5

31 13

16 8

II 15

IO 22

I II

5 8

I II

5 I

Alkyl Alk- I-enyl Acyl Alkyl

6 2 3 4

15 16 14 16

12 IO 8 IO

16 II II II

17 21 18 25

9 14 II 8

8 9 IO IO

IO 10 16 IO

7 7 9 6

Mixture of isomeric cis-octadecenols

Rat brain Diradylglycerophosphocholines Diradylglycerophosphoethanolamines

Position

204

TABLE V DISTRIBUTION OF POSITIONAL WAX ESTERS FROM RAT BRAIN trcms-[ I-‘4C]OCTADECENOLS

ISOMERS IN LABELED RADYL MOIETIES OF GLYCEROPHOSPHOLIPIDS AND AND MURINE ASCITES CELLS AFTER APPLCATION OF A MIXTURE OF ISOMERIC

Alkylglycerols and long-chain alcohols derived from diradylglycerophosphocholines and diradylglycerophosphoethanolaminex via reduction with LiAlH,, hydrolysis of the vinyl ether bond and subsequent reduction of the liberated aldehydes as well as long-chain alcohols derived from wax esters via transmethylation were isolated by thin-layer chromatography (Table II) and acetylated. The distribution of positional isomers in the alkyldiacetylglycerols and alkylacetates was determined by reductive ozonolysis and subsequent radio gas chromatography. The analytical procedures are described in Materials and methods. Values represent distribution of radioactivity (%). The isomeric trues-octadecenols were a mixture of trams-[ I-‘4C]octadcccnols with a fairly uniform distribution of positional isomers.

Substrate:

Position

of double

8

9

IO

II

I2

I3

I4

15

16

3

17

IO

12

23

9

9

I2

5

Alkyl

2

II

I7

7

I8

II

II

I6

7

Alkyl Alk- I -enyl Alkyl

3 3 4

12 II I8

I2 II 9

IO IO I2

22 I8 I6

9 I3 II

I2 I2 I0

I3 15 I3

7 7 7

Alkyl Acyl

4 4

10 I3

I5 7

8 IO

I7 22

8 I2

II II

IX I3

9 8

Alkyl Alk- I -enyl Acyl Alkyl

3 2 3 3

16 II I2 20

I3 IO 9 IO

II 7 IO I4

I7 I8 21 20

8 I4 IO x

9 II I5 9

I4 I5 II 9

9 I2 9 7

Alkyl Acyl

4 3

II I2

I6 6

8 I4

I7 24

IO 9

9 I2

I7 I3

x 7

Alkyl Alk- 1-enyl Acyl Alkyl

3 3 2 4

9 IO I2 21

II 8 6 7

7 7 I2 9

21 I7 26 21

IO I7 II 9

II I2 II x

I7 I5 12 13

II II x 8

Mixture of isomeric rrrms-octadecenols

bonds of isomeric

radyl moieties

Rat brain Diradylglycerophosphocholines Diradylglycerophosphoethanolamines Wax esters Leukemia 1210 Diradylglycerophosphocholines Diradylglycerophosphoethanolamines

Wax esters Sarcoma

I80

Diradylglycerophosphocholines Diradylglycerophosphoethanolamines

Wax esters

cis-lo-octadecenyl moieties as compared to the substrates (Table IV). The isomer composition of labeled alkyl, alk-lin diradylglyenyl and acyl moieties cerophosphoethanolamines of rat brain and murine ascites cells fairly resembles the isomer composition of the added mixtures of cis- and transoctadecenols (Tables IV and V). Furthermore, the of cisand tram-[ lisomer composition

“C]octadecenyl moieties of wax esters in all three systems is quite similar to that of the added mixed substrates. Discussion

The results of this study show that mixtures of saturated straight-chain alcohols or cis- 9 octadecenol as well as mixtures of its positional

205

and geometrical isomers are utilized rapidly for the biosynthesis of membrane lipids in rat brain and murine ascites cells (TableI). The radioactivity from substrate alcohols, which are incorporated into the radyl moieties of glycerophospholipids of all three cell systems, is located only in diradylglycerophosphocholines and diradylglycerophosphoethanolamines, but not in other phospholipids. Therefore, it appears that in rat brain and murine ascites cells the biosynthesis of glycerophosphocholines and glycerophosphoethanolamines mainly occurs via the 1,Zdiradylglycerol pathway [25-271. Possibly also the deacylation-reacylation cycle via lysoglycerophospholipids is operative; other biosynthetic routes, such as the transmethylation pathway or base-exchange pathway, are of minor significance in the formation of glycerophosphocholines and glycerophosphoethanolamines [25-271. Major proportions of the radioactivity from substrate alcohols are incorporated into alkyl moieties of diradylglycerophosphocholines and into both alkyl and alk-1-enyl moieties of diradylglycerophosphoethanolamines of all three cell types investigated (Table II). The virtual absence of labeled alk- 1-enyl moieties in diradylglycerophosphocholines is consistent with earlier findings [28]. We have found that all three cell types, especially S 180 ascites cells, readily incorporate exogenous long-chain alcohols into wax esters (Table I). This could be expected as wax ester synthase activity has been found in normal and neoplastic mammalian cells [4,6,8,29], whether they normally contain wax esters or not [30]. The label from radioactive long-chain alcohols is found almost exclusively in the alkyl moieties of wax esters, although oxidation of long-chain alcohols to the corresponding fatty acids occurs (Table II). Longchain alcohols are known to be oxidized rapidly in mammalian cells, and the substrate specificity of oxidizing enzymes with regard to the chain-length of saturated alcohols has been studied [31,32]. Apparently the labeled fatty acids derived from the alcohols are channeled mainly to acyl lipids, as evident from extensive labeling of the acyl moieties of glycerophospholipids and neutral glycerolipids (Tables I and II). It is interesting to note that only small proportions of labeled acyl

moieties with 20 carbon atoms occur in glycerophospholipids, suggesting that icosanol is oxidized to a minor extent (Table III). Chain elongation and desaturation of acyl moieties does not occur in all three cell systems. Traces of radioactivity found in cholesterol are likely to have resulted from acetyl-CoA formed by P-oxidation of ( 1-‘4C)-labeled fatty acids (Table I). The various steps involved in the de novo synthesis of diradylglycerophosphocholines and diradylglycerophosphoethanolamines from longchain alcohols and acyldihydroxyacetonephosphate as well as from acyl-CoA and glycerol 3phosphate or dihydroxyacetonephosphate are well known [1,3]. In the final step the phospholipids are formed by microsomal cholinephosphotransferases and ethanolaminephosphotransferases from diradylglycerols, the common intermediates of cholinephospholipids and ethanolaminephospholipids [25-271. Assuming that the same pool of to both phosdiradylglycerols is available photransferases, the difference in the pattern of labeling of saturated alkyl and acyl moieties of both diradylglycerophosphocholines and diradylglycerophosphoethanolamines (Table III) reveals that in all three cell types investigated cholinephosphotransferases and ethanolaminephosphotransferases have distinctly different specificities with regard to the chain length of alkyl and acyl moieties of diradylglycerols. Although a number of enzymatic reactions is involved in the formation of alkyl and alk- I-enyl ethanolamineglycerophospholipids, none of the enzymes appears to exhibit pronounced specificity for any of the positional isomers of either cis- or truns-octadecenols or intermediate ether lipids derived therefrom. This is evident from the data given in Tables IV and V, which show that the isomer distribution of labeled alkyl and alk-1-enyl moieties in diradylglycerophosphoethanolamines of rat brain and murine ascites cells closely resembles the composition of the mixtures of cis- or trans-octadecenols added. These results imply that the isomer composition of labeled alkyl moieties in the intermediate alkylacylglycerols also resembles the composition of the added mixtures of cis- and trans-octadecenols. On the other hand, the results given in Table IV show that in alkylacylglycerophosphocholines of

206

all three systems

investigated

the level of labeled moieties is considerably increased, as compared to the substrate cis-octadecenols, and that a concomitant decrease in the level of labeled cis-12-octadecenyl moieties occurs. Assuming that the same pool of alkylacylglycerols is available to both phosphotransferases, it must be concluded from the foregoing findings that cholinephosphotrans ferases, in contrast to ethanolaminephosphotransferases, are highly specific for alkylacylglycerols containing cis-8, cis-9 and cis-looctadecenyl moieties. It has been reported that alk-l-enylacylglycerophosphoethanolamines are synthesized from alkylacylglycerophosphoethanolamines by a microsomal desaturase that inserts a A’ double bond in the alkyl chain [33,34]. The data in Table III show that in diradylglycerophosphoethanolamines of rat brain and murine ascites cells the distribution of labeled alkyl moieties with regard to chain-length is remarkably different from the distribution of labeled alk-1-enyl moieties. On the assumption that alkylacylglycerophosphoethanolamines are the direct precursors of alk- l-enylacylglycerophosphoethanolamines [28,33-351 it is evident that alkylacylglycerophosphoethanolamine desaturase is highly specific with regard to chain length of saturated alkyl moieties of alkylacylglycerophosphoethanolamines. Similar results have been reported for small intestine of the rat [36]. The composition of labeled acyl moieties in diradylglycerophosphocholines and diradylglycerophosphoethanolamines of L 12 10 and S 180 ascites cells shows that none of the isomeric octadecenols added is excluded from oxidation (Tables IV and V). Furthermore, the distribution of positional isomers of labeled octadecenoyl moieties in both glycerophospholipids of the neoplastic cells closely resembles the composition of the mixed substrates. These findings strongly suggest that neither the enzymes involved in the oxidation of long-chain alcohols nor the enzymes catalyzing the incorporation of acyl moieties into the precursors of diradylglycerophosphocholines and diradylglycerophosphoethanolamines exhibit specificity with regard to position of double bonds of the cis- or truns-octadecenols. The data given in Table III show that wax ester cis-8, cis-9 and cis-lo-octadecenyl

synthase in rat brain prefers as substrates saturated alcohols with chain lengths of 18 and 20 carbon atoms, whereas in L 1210 and S 180 ascites cells alcohols with chain-lengths 16 and 18 are the preferred substrates. Similar preferences for the incorporation of long-chain alcohols have been reported in rat intestinal mucosa [8] and mouse preputial gland tumor [29]. The results given in Tables IV and V show that the isomer composition of labeled cis- and truns-octadecenyl moieties in wax esters of rat brain as well as of murine ascites cells closely resembles the composition of the substrates. Thus, it is evident that wax ester synthase is specific to individual saturated alcohols with regard to chain-length but not to positional isomers of cis- and trans-octadecenols.

Acknowledgement This study was supported, in part, by Deutsche Forschungsgemeinschaft, Bonn - Bad Godesberg (We 927/1-l). References 1 Snyder, F. (1972) in Ether Lipids: Chemistry 2 3 4 5 6 7 8 9

and Biology (Snyder, F., ed.), pp. 273-295, Academic Press, New York Mahadevan, V. (1978) Prog. Chem. Fats Other Lipids IS, 255-299 Mangold, H.K. (1979) Angew. Chem. Int. Ed. Engl. 18. 493-503 Mukhejee, K.D., Weber, N., Mangold, H.K., Volm, M. and Richter, I. (1980) Eur. J. Biochem. 107, 289-294 Snyder, F. and Wood, R. (1968) Cancer Res. 28. 912-978 Snyder, F., Malone, B. and Blank, M.L. (1970) J. Biol. Chem. 245, 1790- I799 Su, K.L. and Schmid, H.H.O. (1972) J. Lipid Rcs. 13, 452-457 Bandi, Z.L. and Mangold. H.K. (1973) FEBS Lctt. 31. 97-100 Natarajan. V. and Schmid, H.H.O. (1977) Fed. Proc. 36.

851 IO Natarajan. V. and Schmid. H.H.O. (I 978) Arch. Biochem. Biophys. 187. 215-222 II Mukherjee, K.D. and Kiewitt, I. (1980) FEBS Lett. 122, 133-134 Z. Physiol. 12 Richter, I. and Weber, N. (198 I) Hoppe-Seyler’s Chem. 362. I l63- I 166 13 Brown. W.G. (I 95 I) Organic Reactions 6, 469-509 14 Mangold, H.K., Becker, H., Cramer, U. and Spener, F. ( 1976) Chem. Phys. Lipids 17, 176- I8 I I., Mukherjee, K.D. and Weber. N. (1978) Z. 15 Richter, Naturforsch. 33~. 629-633

20-l

16 Bickerstaffe, R. and Mead, J.F. (1967) Biochemistry 6, 655-662 17 Bligh, E.G. and Dyer, W.S. (1959) Can. J. Biochem. Physiol. 37, 911-917 18 Wagner, H., H&hammer, L. and Wolff, P. (I 96 I) Biochem. z. 34, 175-184 19 Parsons, J.G. and Patton, S. (1967) J. Lipid Res. 8, 696-698 20 Anderson, R.E., Garrett, R.D., Blank, M.L. and Snyder, F. (1969) Lipids 4, 327-330 21 Mangold, H.K. (1961) J. Am. Oil Chem. Sot. 38, 708-727 22 Kaufmann, H.P., Mangold, H.K. and Mukherjee, K.D. 23 24 25 26 27

(1971) J. Lipid Res. 12, 506-509 Chalvardjian, A. (1964) Biochem. J. 90, 518-521 Christie, W.W. (1976) Lipid Analysis, pp. 94-95. Pergamon Press Ltd., Oxford Van den Bosch, H. (I 974) Annu. Rev. Biochem. 43,243-277 Bell, R.M. and Coleman, R.A. (1980) Annu. Rev. Biochem. 49, 459-487 Holub, B.J. and Kuksis, A. (1978) Adv. Lipid Res. 16, l-125

28 Schmid, H.H.O., Muramatsu, T. and Su, K.L. (1972) Biochim. Biophys. Acta 270, 3 17-323 29 Wykle, R.L., Malone, B. and Snyder, F. (I 979) J. Lipid Res. 20, 890-896 30 Sargent, J.R. (1976) in Biochemical and Biophysical Perspectives in Marine Biology (Malins, D.C. and Sargent, J.R., eds.), Vol. 3, pp. 149-212, Academic Press, London 31 Lee, T-C. (1979) J. Biol. Chem. 254, 2892-2896 32 Natarajan, V. and Schmid, H.H.O. (1977) Lipids 12, 872875 33 Paltauf, F. (1972) FEBS Lett. 20, 79-82 34 Paltauf, F. and Holasek, A. (1973) J. Biol. Chem. 248, 1609-1615 35 Debuch, H., Mtiller, J. and Ftirniss, H. (1’971) HoppeSeyler’s Z. Physiol. Chem. 352, 984-990 36 Bandi, Z.L. and Mangold, H.K. (1978) Nutr. Metab. 22, 190-199