A. N. Davison

The chemical composition of vertebrate myelin and microsomes.

ELECTRON microscopic studies have shown that the myelin sheath is composed of a compact ordered arrangement of unit membranes (FINEAN, 1957 ; ROBERTSON, 1959, 1964). Myelin therefore provides a particularly suitable model for the investigation of the molecular organisation of biological membranes. However, until quite recently it was not possible to isolate myelin from nervous tissue and reliable information about its chemical composition was not available. Improved methods for the separation of subcellular structures have now been applied to neural tissue and a number of laboratories have described techniques for the isolation of myelin in a form suitable for direct analysis (WHITTAKER, 1959; PATTERSON and FINEAN, 1961 ; AUGUST, DAVISON and WILLIAMS, 1961 ; LAATSCH, KIES, GORDON and ALVORD, 1962; DAVISON and GREGSON, 1962; HULCHER, 1963). Some reports of the lipid composition of myelin from individual species have already appeared (PATTERSON and FINEAN, 1961 ; NUSSBAUM, BIETH and MANDEL, 1963; AUTILIO, NORTON and TERRY, 1964; EICHBERG, WHITTAKER and DAWSON, 1964; SEMINARIO, HREN and GOMEZ, 1964; CUZNER, DAVISON and GREGSON, 1965). Although it has been suggested that all membranes may have a common ‘unit membrane’ structure (ROBERTSON, 1964) both electron microscopic and biochemical evidence indicate that the various membranes found in the cell are not necessarily similar in all properties (STOECKENIUS, 1964). Earlier work from this laboratory has indicated that mammalian myelin may represent a unique class of membrane for it has a characteristically slow rate of turnover with little associated enzyme activity (ADAMS, DAVISON and GREGSON, 1963). In addition indirect evidence led to the idea that the lipid composition of myelin also differs from that of other membranes (ROSSITER, 1962; ADAMS and DAVISON, 1965; GREGSON 1965). The purpose of the present work has been to investigate this possibility more thoroughly by comparing the chemical composition of myelin with that of another cell membrane fraction from the same tissue. Myelin and microsomal fractions have been prepared therefore from the brains of a number of mammalian and more primitive vertebrates ; samples have been examined by electron microscopy and submitted to chemical analysis. The results of these and earlier observations will be discussed with particular reference to the possible molecular organisation of membranes. EXPERIMENTAL Materiais. The following animals were used in the present work: white Wistar rats, rabbits, the feral pigeon, common frog (Rana temporaria) and dogfish (Scyllium cuniculu). These animals were killed, exsanguinated and the brain removed. Ox brain, collected in ice from the slaughter house, and human frontal cerebral cortex obtained at autopsy, were brought to the laboratory and white matter removed. Human brains were obtained from patients who had died from non-neurological causes.

ENZYME INACTIVITY OF MYELIN: HISTOCHEMICAL AND BIOCHEMICAL EVIDENCE

EARLIER work on the lipid and protein metabolism of grey and white matter of brain suggested that the myelin sheath may be in a metabolically dynamic state (SCHEINBERG and KOREY, 1962). Quantitative cytochemical studies of Ammon’s horn in the rabbit (LOWRY et af., 1954) first suggested that myelin may itself be enzyinically active. This view was based on finding more enzyme activity and more acid soluble phosphorus derivatives in the myelinated layer than would then be expected from the content of axons and neuroglia. More recently, it has been claimed that a wide range of oxidative and hydrolytic enzymes are present i n the myelin sheath of the peripheral nerve and in its neurokeratiri network (WOLFGRAM and ROSE, 1960; TEWARI and BOURNE, 1960a, b and c): some enzyme activity has also been reported in CNS myelin (TEWAIU and BOURNE, 1962n and h). TEWARI and BOURNE regard the tieurokeratin network as the locus of metabolic activity within the myelin sheath. If these histological and histochemical observations on neurokeratin and myelin enzymes can be substantiated, the myelin sheath would appear to be potentially metabolically active and could be regarded as a functional part of the cytoplasm of its formative cell-oligodendrocyte or Schwann cell~-and therefore, susceptible to injury by metabolic derangements. However, evidence has been accumulated from isotope studies to show that myelin lipids (DAVISON et al., 1959 a and 6; DAVISON and D o B B I N c ; , 1960) and proteins

THE OCCURRENCE OF ESTERIFIED CHOLESTEROL IN THE DEVELOPING NERVOUS SYSTEM

J ~ I E R I F I E D cholesterol, in relatively large amounts, has been found in the brain of developing chickens by MANDEL, BETH and STOLL (1949). The presence of these esters in the young animal and their apparent absence (SPFXRY, 1955) in the adult led us to postulate that esterified cholesterol may play an important role in the development of the central nervous system. This paper describes histochemical observations on the developing CNS of man and chickens, together with parallel chemical and chromatographic determinations of cholesterol and its esters. Evidence is presented to show that part of the ester cholesterol is located in the newly formed myelin sheath. The relevance of these histochemical and chemical observations to the process of myelinogenesis is also discussed.

THE DEPOSITION AND DISPOSAL OF (4‐14C) CHOLESTEROL IN THE BRAIN OF GROWING CHICKENS

MUCH has been learned in recent years through the use of electron microscopy and X-ray diffraction studies about the laminated structure of the myelin sheath (FERNANDEZ-MORAN and FINEAN, 1957). Such studies have disclosed its distinctively organized character, and the likelihood that the various types of protein and lipid molecules recoverable from it by chemical extraction are mutually linked in a highly orientated manner. Little biochemical work has yet been done, however, to determine whether this apparent stability of internal structure in the sheath has as its counterpart any tardiness in the metabolic ‘turnover rate’ of any or all of its known constituent materials. Of the various lipids typical of myelin (JOHNSON, MCNABB and ROSSITER, 1948), cholesterol possesses several notable advantages for such ‘turnover’ studies. In addition to being quantitatively one of the principal members of this group of lipids, it is well characterized chemically, readily extractable and capable of accurate estimation. It can furthermore be obtained labelled with radioactive carbon, so that the presence and duration of persistence of this identifiable form of the compound can be ascertained with considerable accuracy. In the following study, radioactive cholesterol in trace amounts of high specific activity has been introduced into newly-hatched chickens at a time when myelination is proceeding with great rapidity. By extracting their brains for lipids at various intervals afterwards, it has been possible to determine the extent to which this cholesterol had become incorporated in the central nervous system and to gain some indication of its ‘turnover rate’ from the rate at which it subsequently disappears.

CHEMICAL AND METABOLIC STUDIES OF RAT MYELIN OF THE CENTRAL NERVOUS SYSTEM.

The brain differs from other organs in that dynamic metabolism of total lipids is restricted in the former in comparison to the rest of the body. Previous investigators (Davison & Dobbing, 1960a, b; Davison, Morgan, Wajda & Payling Wright, 1959) have shown, by introduction of radioactive precursors into developing anini’als, that part of the brain lipids remained metabolically stable during growth and subsequently in the adult animal. It was further suggested that this metabolic stability was associated with the myelin sheath and possibly other anatomical structures in the brain. Evidence in support of this concept has recently been reviewed (Davison, 1964).

Cerebral lipids in multiple sclerosis.

IN an investigation into the cerebral lipids from a number of cases of multiple sclerosis CUMINGS (1953, 1955) found that apparently normal areas of the brain were deficient in phospholipids. This important observation suggests that patients with the disease might have a generaliscd deficiency of brain lipid and therefore a predisposition to multiple sclerosis (THOMPSON, 1961). The present study was undertaken in an attempt to confirm and extend the original observations of CUMINGS on the lipid composition of apparently normal areas of brain from patients with multiple sclerosis. In the first stage of this work modern techniques for the extraction and isolation of lipids were used for the complete analysis of normal human brain lipids (DAVISON and WAJDA, 1962). These same methods have now been used for determining the brain lipid composition in five cases of multiple sclerosis.

QUANTITATIVE ANALYSIS AND RECOVERY OF CEREBRAL LIPIDS BY MODIFIED THIN LAYER CHROMATOGRAPHY

IN EARLIER work from this laboratory (DAVISON and WAJDA, 1962) it was reported that analysis of cerebral lipids could be considerably simplified by prior fractionation of a lipid extract on an alumina column. Since the method is time consuming and since phosphatidylserine cannot be completely removed from the column without decomposition (LONG and STAPLES, 1959, 1960, 1961), it was thought desirable to find an alternative method for an initial separation of lipids prior to their estimation. Separation of lipids by thin layer chromatography is now well established as a reliable technique (MANGOLD, 1961 ; WAGNER, HORHAMMER and WOLFF, 1961) and this method seemed suitable for present purposes. Although quantitative analysis of phospholipids and sphingolipids from silicic acid thin layer chromatograms has been reported (JATZKEWITZ, 1961), the specialised micro-methods employed necessitate ashing or hydrolysis of lipids and, under these conditions, quantitative analysis is difficult. VACIKOVA, FELT and MALIKOVA (1 962) have recovered neutral lipids and phospholipids from alumina spread layers; however, in this investigation only serum lipids were used and no attempt was made to separate phospholipids. The present paper describes a simple method for the quantitative recovery of intact lipids from thin layer chromatograms. A preliminary account of this work has already been published (DAVISON and GRAHAM-WOLFAARD, 1963).