BY EUGENE P. KENNEDY and SAMUEL B. WEISS†

(From the Ben May Laboratory for Cancer Research and the Department of Biochemistry, University of Chicago, Chicago, Illinois)

(Received for publication, January 23,1966)

In 1952, Kornberg and Pricer (1) reported in a preliminary communication that P-choline1 labeled with P" and CM was converted by liver enzymes to a lipide which was not further characterized. The ratio of P" to CM in the product was closely similar to that of the labeled P-choline, suggesting incorporation of both phosphorus and choline as an intact unit into a phospholipide, presumably lecithin. The following reaction mechanism was suggested (2, 3) as a possible explanation for these results:

(1) Phoephatidic acid + phosphorylcholine-P22 →lecithin-P22 + P1

Rodbell and Hanahan (4) have studied the incorporation of P-choline into the lipides of isolated mitochondria from the liver of the guinea pig. These workers identified the product of the enzymatic reaction as lecithin, but their results cast little light on possible intermediates.

Although no definite information is available on this point, it appears that a reaction such as Equation 1 should be readily reversible and a phos-phorolytic cleavage of lecithin should occur, as a result of which inorganic phosphate labeled with P32 should be incorporated into phosphatide acids. Work in this laboratory on the enzymatic incorporation of P1^22 into phos-pholipides failed to reveal such a reaction. Accordingly, it was decided to investigate the conversion of P-choline to lecithin more closely in an effort to learn the true mechanism of the reaction.

*Aided by grants from the Nutrition Foundation, Inc., and from the American Cancer Society, recommended by the Committee on Growth of the National Research Council.

† Postdoctoral Fellow of the American Heart Association.

1 The following abbreviations will be used in this paper: P-choline - phosphoryl-choline; P-ethanolamine - phosphorylethanolamine; ATP - adenosine triphosphate; UTP - uridine-a'-triphosphate; GDP - guanosine-S'-diphosphate; ITP - inosine triphosphate; CTP - cytidine-5'-triphosphate; CDP-choline - cytidine diphosphate choline; CDP-ethanolamine - cytidine diphosphate ethanolamine; UDP-choline - uridine diphosphate choline; GDP-choIine - guanosine diphosphate choline; Tris - tris(hydroxymethyl)aminomethane; CoA - coenzyme A; CMP - cytidine-5'-monophosphate; AMP - adenosine-fi'-monophosphate; Vereene - ethyl-enediaminetetraacetate; Pi - inorganic orthophoephate; P-O-P - inorganic pyrophosphate.

In the course of this investigation, a new type of coenzyme, of which the novel compounds cytidine diphosphate choline and cytidine diphosphate ethanolamine are examples, has been found to play an essential rdle in the biosynthesis of phospholipides. This paper will describe the enzymatic synthesis of cytidine diphosphate choline and cytidine diphosphate ethanolamine, the isolation of these compounds from liver and yeast, and their function in the biosynthesis of lecithin and phosphatidylethanolamine.

In 1954, Kennedy (5, 6) reported the incorporation of free choline into the lipides of mitochondria by a pathway not involving P-choline, but requiring ATP and CoA. The enzymatically synthesized lipide was identified as lecithin principally by a procedure which involved the isolation by chromatographic methods of a highly purified lecithin fraction containing alkali-labile radioactive choline. Later work2 has shown that this identification was erroneous. The actual product is a lipide difficult to distinguish from lecithin by chromatography on alumina, but separable from lecithin on cation exchange resins. It does not contain phosphorus. The suggestion that a separate enzymatic pathway exists for the conversion of free choline into lecithin must therefore be withdrawn. The reactions to be described in the present paper represent the only route for the enzymatic synthesis of phospholipides known at present.

A preliminary account of some aspects of this work has been published (7).

Materials And Methods

A mitochondrial fraction was prepared from the livers of normal adult albino rats by the following procedure. The livers were quickly removed from rats killed by decapitation and homogenized in 5 volumes of ice-cold 0.25 h sucrose containing 0.001 m Versene, in a glass homogenizer of the Potter-Elvehjem type. Nuclei, debris, and whole cells were removed by centrifugation for 5 minutes at 500 X g at 0°. The mitochondrial fraction was then obtained by centrifugation at 18,000 X g for 20 minutes at 0°, the high speed head of an International refrigerated centrifuge being used, and was washed twice with the sucrose-Versene solution, and finally suspended in sucrose-Versene or water just prior to use. Since it is now recognized that" mitochondria" prepared by conventional procedures of differential centrifugation may represent a range of cytoplasmic granules of varying chemical and biochemical properties (26, 27), this preparation will be described simply as rat liver particles.

The acid phosphatase of human semen was prepared from semen which had been stored at -15° for many months. After thawing, the seminal plasma was clarified by centrifugation and the protein was precipitated by saturation with ammonium sulfate. The precipitate was removed by centrifugation and dissolved in water (about one-tenth of the volume of the original semen). The enzyme solution was dialyzed against distilled water and then stored at -15°.

1 E. P. Kennedy, unpublished experiments.

Egg lecithin was purified by chromatography on alumina by the method of Hanahan, Turner, and Jayko (8). A mixture of D-a,β-diglycerides was prepared from this egg lecithin by the action of the lecithinase D present in Clostridium perfringens type A toxin (9), which was the gift of Dr. H. D. Piersma of the Lederle Laboratories.

Details of the methods for separation of labeled lipides from the enzyme incubation mixture and for the quantitative removal of unchanged labeled substrate have been described by Kennedy (10). In the experiments reported in this paper, carbon tetrachloride rather than ethyl acetate was used when equilibrating the lipide extracts against aqueous buffer. Ali-quots of the carbon tetrachloride solution of phospholipides, completely freed of water-soluble radioactive contaminants, were dried in aluminum dishes and counted in a windowless gas flow counter under conditions of negligible self-absorption. The effectiveness of the equilibration procedures was tested by "zero time" controls, in which the trichloroacetic acid was added prior to the enzyme.

Choline-1,2-C14 bromide was purchased from Tracerlab, Inc. Ethanol-amine-1,2-C14 was the gift of Dr. I. Zabin. P-choline and P-ethanolamine were synthesized by a variation of the method of Plimmer and Burch (11). Dipalmitoyl-l-«-glycerophosphoric acid was the gift of Professor E. Baer.

The nucleotides used in this work were products of the Pabst Laboratories, except for the GDP, which was the gift of Dr. D. Sanadi. The amorphous ATP preparation (lot No. 116) was about 95 per cent pure. The crystallized ATP (lot No. 122) was a highly purified product in which no CTP could be detected. The authors wish to express their gratitude to Dr. S. Morell, Dr. A. Frieden, and Dr. S. Lipton of the Pabst Laboratories for generous gifts of several of these nucleotides.

Chromatographic Procedures-Chromatography on Dowex 1 formate (2 per cent cross-linked) ion exchange resin was carried out with an apparatus similar to that described by Busch, Hurlbert, and Potter (12). The mixing chamber initially contained 300 ml. of distilled water. The upper reservoir contained either 0.04 n formic acid (formic acid system) or 0.2 n formic acid adjusted to pH 9.4 by the addition of concentrated ammonium hydroxide (ammonium formate system). Paper chromatography was carried out with the solvents previously described (13).

Other materials and methods were described in previous publications (10,13).