Myelin maturation: the evidence from organ cultures of cerebellum [proceedings].

Before the advent of critical isotope-labelling experiments (Eichberg & Dawson, 1965), myelin was believed to be metabolically inert. Evidence soon accumulated which indicated that not only did an active myelin metabolism persist in young adult animals, but was accompanied by a progressive change in chemical composition of the sheath. Typically the myelin isolated by centrifugation from developing brains of rat (Eng & Noble, 1968), mouse (Horrocks, 1968) and human (Eng et al., 1968) had a high phospholipid/cholesterol ratio and was relatively deficient in cerebrosides compared with mature myelin from adults. Myelin development in vitro has now been examined in myelinating organotypic cultures. Neonatal rat cerebellar explants were placed on collagen-coated coverslips and maintained for up to 60 days in vitro in roller tubes at 37°C. The culture method and the culture feed comprising Simm’s balanced salt solution, medium 199 (Wellcome Reagents), human non-icteric ascitic fluid, bovine serum ultrafiltrate and chick-embryo extract were described by Lumsden (1968). Groups of 200 explants and an equivalent weight of cerebella from littermates were made up to 0.5ml with water and homogenized in a Teflon-glass homogenizer (clearance 0.2mm) with an equal volume of 0.65 s sucrose (lo%, w/v) at 1800rev./min for 1 min. The homogenates were layered over 1 ml of 0 . 7 0 ~ sucrose and centrifuged at 15000g for 30min at 4°C (Christ Omega I1 swing-out rotor) yielding a pellet and a crude myelin layer. The myelin layer was mixed with 1Ovol. of water, kept for IOOmin at 4°C and centrifuged at 45000g for 10min. The pellet was suspended in ice-cold water and centrifuged at lOOO00g for lOmin; this step was repeated. The myelin pellet was suspended in 0.25 ml of 0.32~-sucrose and layered on a discontinuous gradient of 0.80~-, 0.63 Mand 0.5O~-sucrose in ascending order in a 2ml tube. The gradient was centrifuged at 75000g for 1 h. Three layers and a pellet were recovered, and each was suspended in ice-cold water and centrifuged at l00OOOg for 15min. This centrifugation was repeated. The pellet yielded insufficient material for analysis. Chloroform/methanol (2 : I , v/v) lipid extracts (Folch et al., 1957) were examined by t.1.c. (Nussbaum et al., 1969). Galactolipids were determined by g.1.c. (Carter & Gaver, 1967), cholesterol by the method of Rappaport & Eichhorn (1955) and phospholipids by the method of Bartlett (1959) or by neutron activation. The three myelin layers were termed ‘light myelin’, ‘heavy myelin’ and ‘membrane fraction’ by analogy with similar layers prepared from 15-day-old rat brains by Agrawal et al. (1974). From 10 days in vitro onwards, all three layers could be recovered from explant homogenates. Electron microscopy revealed heavy myelin to contain multilamellar fragments. The light-myelin fraction contained no large aggregates, and the membrane fraction comprised single-walled vesicles with a dense central line similar to those observed in membrane fractions of myelin in vivo (Agrawal et uf., 1974). No gross ultrastructural differences were noted between the myelin fractions derived from the cerebellar explants and the equivalent fractions prepared from the cerebella of I i ttermates. Table 1 presents the molar proportions of the major lipid groups in the culture myelin fractions. Between 14 and 21 days in vitro, when myelination of the explants is rapid, the phospholipid/cholesterol ratio of the three myelin fractions increased. The increase was greatest in the membrane fraction. The phospholipid/cholesterol ratios in fractions of myelin after 14 days in vifro are significantly higher than the ratios of the comparable fractions prepared from the cerebella of 15-day-old littermates. The most important discrepancy between the lipid composition of cultured myelin fractions and that of litter-