Metabolism of Brain Glycolipid Fatty Acids '': Yasuo Kishimoto and Norman S. Radin, Mental Health Research Institute, University of Michigan, Ann Arbor, Michigan ABSTRACT and sulfatides contain NFA and tIFA, The metabolism of the fatty acid moieties saturated and unsaturated; the gangliosides, of brain cerebrosides, sulfatides, and however, contain only NFA in which there are gangliosides is reviewed and discussed. only traces of unsaturated acids. In the cere- The methodology involved in the isolation t)rosides and sulfatides there are two clusters of the fatty acids is described briefly. It of FA: those around 18 carbons long and those seems clear now that most of these acids around 24 carbons long. In the gangliosides are made by chain elongation of inter- there is only one cluster, centering around 18:0, mediate length fatty acids by addition of with negligible amounts of 22:0 and 24:0. acetate residues. The unsaturated acids Other points of contrast between gangliosides are made by desaturation of the inter- and the other two can be made: the former mediate length acids (palmitic, heptade- occurs primarily in brain gray matter, the canoic, stearic) followed by chain elonga- latter are primarily in white. The former tion. The hydroxy acids are made directly has glucose attached to the ceramide residue, from the corresponding nonhydroxy acids, the latter have galactose. The former has saturated, unsaturated, and odd-numbered. only traces of odd-numbered FA; the latter All the hydroxy acids undergo oxidative can contain considerable amounts of C~ and decarboxylation to yield fatty acids con- C2.~ FA. Further differences, particularly in taining one less carbon atom. The odd- the metabolism of the two groups of glyeo- numbered acids are also made from lipids, will be described below. propionate, which is elongated to inter- This paper describes mainly the studies mediate length acids and then to longer carried out in this laboratory, generally with acids. The major intermediate length live rats. It discusses the special problems of "primer" acid seems to be palmitate, but isolation that were encountered, as well as the there is evidence that the stearate used interpretation of the isotopic data that were for cerebroside synthesis is also made obtained. It is hoped that the approaches used will be of value in other types of studies. de novo from acetate. The ganglioside fatty acids were found to turn over some- what faster than the other fatty acids. Two SPECIAL PROBLEMS OF ISOLATION metabolic pools for the cerebroside acids Technical problems made this study partic- were found, one with a very high turnover ularly difficult. These arose from several rate, the other with a very low turnover factors : rate. (a) Judging by experiments with adult ani- INTRODUCTION mals the blood-brain barrier causes the brain to come out second-best when there is a com- I~E BtCAII~ GLYCOLIPIDS covered in this petition for intraperitoneally administered T paper are the cerebrosides, sulfatides, and isotopic precursors. The liver and other organs gangliosides. More specifically, these are pick up most of the injected material, so that ceramide galaetoside, cerebroside sulfate (sul- large amounts of radioactivity must be ad- furic acid ester of eerebrosides, attached to nfinistered to each animal in order to get useful the 3-position of the galactose), and the ce- levels of incorporation into the brain. While ramide polysaccharides containing neuraminic intraeranial injection yields much better utili- acid. These lipids occur as families, differing zation of labeled material, it does not seem to within each family as to the nature of the fatty be reproducible or physiological enough to acid in the ceramide residue. The cerebrosides warrant use in turnover and other quantitative 1 Presented at the Prof. :Ernst Klenk Symposium on studies. Because of the blood-brain barrier and Glycolipids and the Nervous System, AOCS meeting, the relatively leisurely rate of metabolism in Houston, April 1965. 2 Abbreviations: RCA, relative carboxyl activity the brain, the glycolipid FA obtained from (activity in COOH groupX 100/activity in total labeled acetate administration have low specific FA); FA, fatty acid, HFA, 2-hydroxy fatty acid; NFA, nonhydroxy fatty acid; 18:0, s.tearic acid; 18:1, oleic activities. This prevents use of the convenient acid; 16:19, palmitoleic acid as is 16:1s7. An h GLC-ionization chamber combinations, which symbolizes a 2-hydroxy l~A, thus h24:0 is cerebronie allow rapid determination of the radioactivity in acid (2-hydroxy]ignoceric acid). A k indicates a 2-keto FA. C is chloroform; E, ether; H, hexane; 1~, methanol. each FA. 47 48 YASUO KISHI~IOTO AND ~NT. S. RADIN (b) If the individual FA in each glyeolipid methanolysis step convenient and reliable by family is to be quantified and its relative radio- using a test tube with an O-ring closure (25). activity is to be measured, the glyeolipid fam- ily must be isolated from the brain in nearly Cerebrosides and Sulfatides ]00% yield. Methods which involve losses in- In our earlier work (36,18) we used Florisil evitably give unequal losses of the various to remove nonpolar lipids (mainly cholesterol), individual members. gangliosides, and nearly all the phosphatides (e) Because of the high amounts of odd FA from the total brain lipids. The potentialities in many glyeolipid samples, GLC separations of this interesting and economical adsorbent of the highest quality are needed. were thus brought to the attention of workers (d) Since GLC methods, on a preparative in the field of the complex lipids. Following scale at least, do not give complete separation the suggestion of Carroll (5) we have more recently been using Florisil deactivated by of saturated from unsaturated FA, and NFA from HFA, it is necessary to carry out pre- addition of water. We first dry the powder liminary class separations. thoroughly by heating at 600C for 60 min, then add 8 nfl water per 100 g powder. The (e) The methyl esters of the 2-hydroxy FA separating power of such Florisil undergoes a do not, in our hands, undergo GLC separation rather sudden deterioration after 3 months, without destruction. It is therefore necessary and we discard the outdated material. It is to protect the OH group prior to GLC. In likely that cold storage of the wetted material experiments where the isolated HFA is to be would prolong its life. Perhap~ the adsor- degraded chemically, it is necessary to use an bent, which is copreeipitated silica gel and easily removed protecting group. magnesia, slowly reacts in the presence of (f) Because of the great length of the water to form magnesium silicate. glyeolipid FA (up to C~), particularly strin- Sulfatides are not readily separated from gent requirements are placed on the techniques eerebrosides by Florisil, so we used synthetic for GLC. "Bleed" materials from packings ion exchange resins to isolate the sulfatides contaminate the GLC effluents. from the mixture (37). Impurities in the This list of difficulties is offered to enlist resins, as well as other exchangers we tried, not only the reader's sympathy but also his kept us from preparing pure sulfatides by this skepticism toward studies in which the im- approach. As Rouser and his co-workers have portance of the factors was inadequately shown (39), it is possible to make ion exchange appreciated. practical by purifying the ion exchanger and eluting reagent very extensively. SPECIt~IC METHODS OF ISOLATION Our previous methods called for the use of large Florisil eolunms since much o~ the total Ganglioside l~atty Acids brain lipids had to be retained by the adsor- As in most of the recent methods for ganglio- bent. We now carry out a preliminary cleavage side isolation or analysis, we ~se a version of of the ester-linked lipids, following removal of the later extraction-partitioning system of gangliosides by solvent partitioning (17). The Foleh, Lees, and Sloane Stanley (8,26,17). The lipids are stirred 1 hr at room temperature in upper layer of this solvent system contains 0.07 N NaOH in C-M 2:1, then partitioned nearly all the gangliosides, as well as an un- with aqueous acetic acid to remove glyeero- known lipid in which the I~A are ester-linked. phosphate esters. This treatment converts the Isolation procedures which rely on dialysis ester-linked FA and part of the free FA (7) as the primary purification step (following to methyl esters, which are then readily isolated partition) cannot eliminate this impurity. The by Florisil chromatography. The same column ester-linked FA account for nearly all the also yields brain, cholesterol quantitatively. ]8:1 in such extracts. Cerebrosides and sulfatides are eluted with For simple analytical work, we renmve most C-M 3:1. Because of the prior methanolysis of the esterified FA by a back-extraction with step, the Florisil column has to adsorb just "Foleh lower phase" (8), but for isotopic the cholesterol, cerebrosides, sulfatides, and work we saponify the ganglioside extract with lyso derivatives of the alkenyl and alkyl phos- mild aqueous alkali, extract the free FA, and phatides. If the cholesterol is not wanted, it then process the purified gangliosides. The FA can be eluted with the methyl esters, thereby are obtained by evaporating the solution to reducing the size of the column further. dryness, methanolyzing with HCI-IV[, and ex- This procedure is of additional interest as tracting with hexane. We have made the it yields the ester-linked FA in the form of METAROLISSI OF ]~I~AIN GLYCOLIPID 49 methyl esters, ready for GLC, by a very mild one another in chain length distribution, ex- and rapid process. There is no contamination cept for the relative absence of h18:0 (18,35).