- Email: [email protected]
Some observations on early myelination in the human spinal cord. Light and electron microscope study
Brahl Research, 104 (1976) 21-32
21
© Elsevier Scientific Publishing Company, Amsterdam - Printed in The Netherlands
SOME OBSERVATIONS ON E A R L Y M Y E L I N A T I O N IN T H E H U M A N SPINAL CORD. L I G H T A N D E L E C T R O N MICROSCOPE STUDY
CLAUS MEIER
Department o[ Neurology, University of Berne, Berne (Switzerland) (Accepted August 10th, 1975)
SUMMARY
Segments of cervical spinal cord from a 23-week-old human foetus have been examined by light and electron microscopy. Myelinated fibres were found in the dorsal, ventral and peripheral lateral tracts, while the lateral corticospinal tract was completely unmyelinated. Myelin sheaths appeared to be formed by spiral wrapping of elongated mesaxons which originated from the apposition of the plasma membranes of oligodendrocytes. Preparations stained with Sudan red and Sudan black revealed the occurrence of lipid inclusions in the interfascicular glia. The topographical relation and the ultrastructural features of these inclusions are described. The possible significance of the inclusion bodies is discussed.
INTRODUCTION
Antenatal diagnosis by biochemical studies of amniotic fluid cell cultures have been proved successful in an increasing number of human hereditary metabolic disorders affecting the nervous systema2, is. After abortion morphological and biochemical investigations have been carried out to confirm diagnosis and for the purpose of studying the early pathological changesl,12,18, 26. However, very little is known about the ultrastructure of the normal developing human central nervous system (CNS), which obviously impairs comparison with the growing body of knowledge about the pathological materialS,13,14. This investigation was carried out on the CNS of a human foetus on which abortion was performed at risk of the mother after 23 weeks of gestation. The white matter of the cerebrum and cerebellum was not yet myelinated; the spinal cord, however, showed the beginning of myelination in some tracts, while others remained unmyelinated. The description of the topographical relation and the ultrastructural characterization of the lipid inclusions, known since Virchow's z4 early
22 report to occur in the developing human CNS, is regarded as the main purposc ol this study. MATERIAL AND METHODS
The observations were made on a foetus obtained by hysterectomy performed at risk of the mother after 23 weeks of gestation. The history of the mother did not suggest that foetal abnormality would be present and none was found. The foetus was 28 cm long (crown-to-heel) and weighted 920 g, compatible with a gestational age of 22-24 weeks. Small segments of the cervical spinal cord were dissected immediately after death and immersed in chilled 3 ~ glutaraldehyde in phosphate buffer at pH 7.3 for 4 h. Postfixation was performed with 1 ~ phosphate buffered osmium tetroxide for 4 h. After dehydration in acetone and embedding in Araldite, 1/zm semithin sections were stained with methylene blue for light microscopy. For electron microscopy thin sections from anterior, lateral and posterior fasciculi, representing the fasciculus gracilis et cuneatus, and the lateral and the ventral corticospinal tracts, were prepared and stained with uranyl acetate and lead citrate. The medulla oblongata adjacent to the segment of the cervical spinal cord embedded for electron microscopy, was fixed by immersion in 10 ~ formalin. Frozen sections were stained with Sudan red and Sudan black dyes*. RESULTS
Light microscopy In 1/~m semithin cross sections stained with methylene blue scattered myelinated fibres were readily identified in the ventral, dorsal and the peripheral lateral tracts, the latter representing the spinocerebellar, the spinoolivary and the spinotectal pathways (Fig. I a, c). Some of the commissural fibres were also myelinated. The region of the lateral corticospinal tract, however, was completely unmyelinated (Fig. 1b). The highest density of myelinated fibres was found in the depth of the ventral and dorsal tracts including the region of the anterior corticospinal pathway. Corresponding to the degree of myelination the incidence of glial cells showed the highest density in these regions. Active oligodendrocytes could easily be identified by their dark-staining cytoplasm with clumsy processes often found in contact with myelinated axons. Positive identification of astrocytes was possible by their large, light-staining nuclei. A third glial cell type, probably representing glioblasts, with small round or oval nuclei, was found in all tracts, but most often in unmyelinated regions. Mitotic figures were found with the same frequency in myelinating and unmyelinated regions with no predilection for the subependymal field around the central canal. About two to three dark-
* The histochemicalinvestigationwas carried out by Professor J. Ulrich, head of the Department of Neuropathotogy, Institute of Pathology, University of Basle, Switzerland. His help is gratefully acknowledged.
23
Fig. I. Spinal cord of 23-week-old human foetus. Cross-sectien of methylene blue-stained I t*m semitbin section, a: both ventral tracts show a fair amount of myelinated fibres. Asterisk on anterior median fissure, x 420. b: lateral corticospinal tract exhibiting no myelinated fibres. The density of glial cells is rather low. Arrow points to lipid accumulation, x 1050. c: myelinating ventral tract exhibiting a high density of glial cells, Most of them can be identified as 'active' oligodendrocytes by their dark staining clumsy processes sometimes found in touch to myelinating axons. Arrow points to lipid accumulation, x 1050.
staining n e c r o b i o t i c cells exhibiting p y k n o t i c nuclei were f o u n d on one entire crosssection o f the spinal cord. L i p i d a c c u m u l a t i o n s could be observed in myelinating a n d u n m y e l i n a t e d tracts (Fig. l b , c). In S u d a n b l a c k - s t a i n e d p r e p a r a t i o n s few m y e l i n a t e d fibres c o u l d be s u b s t a n t i a t e d in the long tracts. S o m e o f the interfascicular glia cells c o n t a i n e d deeply black-staining inclusions. In S u d a n red p r e p a r a t i o n s rare redstained material was f o u n d in the regions o f the l o n g t r a c t s ; however, a m o r e distinct localization was not possible. The S u d a n red-stained material exhibited a w e a k d o u b l e refraction in p o l a r i z e d light.
Electron microscopy In transverse section m o s t o f the axons in the m y e l i n a t i n g tracts exhibited calibres o f 0.2-0.5 fire. T h e y were sheathless a n d a r r a n g e d in fascicular bundles by radial-
Fig. 2. Spinal cord of 23-week-old human foetus. Small unmyelinated axon (ax) is completely enguticd by the perinuclear cytoplasm of an oligodendrocyte (OL). Arrow points to mesaxon formation. Two myelinated axons and one axon surrounded by loose myelin membranes seem to be aitached to che same oligodendrocyte. The oligodendroglial cytoplasm contains membraneous material ~ith a structure of myelin (asterisk). :: 30,000. Fig. 3. Transverse section of 23-week-old h u m a n foetus spinal cord. A stage of early myelmation m the lateral corticospinal tract. An axon (ax) is surrounded by a spiral of loose myelin. The oligodendroglial cytoplasm contains a lysosomal body with a 'tuffstone'-like appearance. The arrow points lo the beginning of mesaxon formation. : 30,000.
25
Fig. 4. Spinal cord of 23-week-old human foetus. Cross-section. Axon (ax) completely engulfed by the perinuclear cytoplasm of a myelinating oligodendrocyte (OL). Arrow points to external mesaxon linking the plasma membrane to the compact myelin sheath. × 15,000.
ly oriented astrocytic cell processes. A few axons of larger diameter were ensheathed by several loose layers or up to 18 layers of compact myelin. The calibres of myelinated fibres ranged between 0.5 and 2 #m. Processes of oligodendrocytes with a welldeveloped cytoplasm containing numerous microtubules were found in relation to myelinating axons (Fig. 2). Some of the oligodendroglial processes encircled the axon completely forming a mesaxon by conjunction of the plasma membranes of the opposed cytoplasmic lips. Elongation of the mesaxon led to the formation of a loose spiral around the central axon (Fig. 3). Twice, a myelinating axon was found which was surrounded completely by the perinuclear cytoplasm of an oligodendrocyte similar to the myelination of peripheral axons by Schwann cells (Figs. 2 and 4). Membranebound dense bodies, obviously of lysosomal origin, were found in the oligodendroglial cytoplasm adjacent to the myelinating axon outside or inside the myelin sheath, and
26
Fig. 5. Spinal cord of 23-week-old human foetus. Longitudinal section. Lysosomal bodies with a structured matrix in the inner cytoplasmic tongue (ol) ofa myelinating oligodendrocyte (a) and in the axoplasm (ax) ofa myelinated axon (b). More inclusion bodies are found in a glial cell process, probably of oligodendroglial origin (right). a, 19,000~b, ~ 13,000. also in the c y t o p l a s m i c loops near the nodes o f R a n v i e r (Figs. 3, 5a a n d 6). Some of these dense bodies exhibited circular m e m b r a n e profiles with a lamellar periodicity o f 5.8-6.0 nm a r o u n d a g r a n u l a r m a t r i x (Fig. 7). similar to the 'tuffstone bodies" as described by Bischoff a n d Ulrich 3 in m e t a c h r o m a t i c l e u c o d y s t r o p h y ( M L D ) material. C o n c e n t r i c l a m e l l a t e d bodies were found in u n m y e l i n a t e d a n d m y e l i n a t i n g axons (Fig. 5b). Indications o f a b u n d a n t myelination, as o b s e r v e d in the d e v e l o p i n g white
27
Fig. 6. Spinal cord of 23-week-old human foetus. Longitudinal section. Structured lysosomal inclusion (asterix) in a paranodal loop of an oligodendrocyte, × 76,0~)3. Fig. 7. High-power view of a lysosomal inclusion of a 'tuffstone' appearance, exhibiting a circular pattern of lamellated material with a periodicity of 5.8-6.0 nm (arrows). × 100,000.
m a t t e r o f kitten 9, r a t 19 a n d m o u s e 16 were very seldom e n c o u n t e r e d . O c c a s i o n a l l y small a m o u n t s o f vesicular dissolution o f myelin at the inner or o u t e r c y t o p l a s m i c t o n g u e were observed, b u t not myelin b r e a k d o w n . Infrequently o l i g o d e n d r o c y t e s were f o u n d which c o n t a i n e d large a m o u n t s o f h o m o g e n e o u s bodies o f weak o s m i o philia, p r o b a b l y representing a c c u m u l a t i o n s o f n e u t r a l lipids (Fig. 8a, b). Lipidc o n t a i n i n g glioblasts were f o u n d in the n o n - m y e l i n a t e d tracts. N e c r o b i o t i c cells a p p e a r very seldom. In the two instances in which they could be studied electron microscopically, it was not possible to discern their origin. Breakd o w n p r o d u c t s o f these cells were f o u n d in the intracellular space or in n e i g h b o u r i n g
28
Fig. 8. a and b: longitudinal section of 23-week-old foetus spinal cord. Glial cell closely attached to a myelinating axon (ax) is filled with m e m b r a n e - b o u n d homogeneous inclusions o f weak osmiophility. b: detail o f a . a, :~ 8000: b, × 30,000.
29
Fig. 9. Spinal cord of 23-week-old human foetus. Area of cell death in ventral tract. Parts of necrobiotic cells (asterisks) are phagocytozed by neighbouring astrocyte (AS). × 13,000.
astrocytic cell processes (Fig. 9). They could easily be distinguished from the inclusion bodies described above. DISCUSSION
The observations on myelination in human spinal cord as presented, confirm the current views about the formation of central myelin based upon studies of animal material as reviewed by BungO and Peters and Vaughn 19. Processes of glial cells with the cytological criteria of active oligodendrocvtes 23 were found in relation to myelin-
30 ating axons. Many times loose spirals of an elongated mesaxon, consisting of conjugated plasma membranes of these processes, could be observed encircling a central axon. In two instances myelinating axons were found completely engulfed in perinuclear oligodendroglial cytoplasm similar to the myelination process in peripheral nerves. The transient appearance of sudanophilic material at the time of initial myelination in the developing human central nervous system has been a matter of controversy since Virchow's first description of what he called 'Congenitale Encephalitis und Myelinitis' in 1867. While he considered it a pathological finding representing myelin breakdown products, others have suggested that this is a normal event during myelination2,17. In the view of the latter the lipid inclusions are thought to be myelin precursors which are gradually transformed into myelin2,17. In the early ultrastructural studies of myelination in the CNS in mouse 21 and rat ~9 lipid inclusions were not mentioned, but later lamellar lipid bodies have been recognized in the myelinating optic nerve of mice 22 and in the developing feline spinal cord white matter 9. While Uzman and Hedley-Whyte z2 regarded the lamellar bodies as transient storage depots of myelin lipid constituents, Hildebrand 9 interpreted them as breakdown products of myelin-glial units appearing during normal development. The observations made in this study confirm the view that lipid inclusion bodies occur in the normal myelinating human CNS. They are found either as homogeneous membrane-bound weakly osmiophilic droplets, or as structured bodies of varying appearance and strongly osmiophilic. Homogeneous lipid inclusions were found in glioblasts and oligodendrocytes. From their morphology these inclusions are probably neutral lipids, and according to chemical analysis of myelinating brains 2,z° they most likely are cholesterol esters. If these inclusions are cholesterol esters, they may be myelin precursors, as they were also found in unmyelinated tracts of the spinal cord where myelin breakdown could not have contributed to their occurrence. In this context it may be reasonable to mention that similar material, but in increased amounts, has been found in oligodendroglial cells of Jimpy mice 1,5 in which, because of a defect of the myelin synthesis, myelin precursor material may accumulate. Osmiophilic lipid bodies of lamellated or granular structure appeared in oligodendrocytes, and in the axoplasm of myelinated and unmyelinated fibres. In oligodendrocytes these bodies were particularly found near myelinating axons, i.e. in the external cytoplasmic tongue, or in the inner cytoplasmic loop between axolemma and myelin sheaths, or in the lateral cytoplasmic loops seen in longitudinal sections of the paranodes. They exhibited a granular ground substance and a circular lamellated arrangement with a lamellar periodicity of about 5.8 nm around a light centre. These bodies were similar to the sulphatide storage material in metachromatic leukodystrophy characterized by Bischoff and Ulrich 3 as 'tuffstone bodies'. The inclusions in axons mostly showed a circular lamellated pattern with a lamellar periodicity of 5.5-6.0 nm. Myelin formation is characterized by a very active synthesis of sulphatide 6,2'', part of which may be degraded normally in lysosomes before it ever can be incorporated into the myelin sheath. It has been shown 5 that the rate of sulphatide synthesis is
31 closely paralleled by the total activity of arylsulphatase A in the myelinating mouse brain, possibly to meet functionally the need of increased catabolism of sulphatide spilled over from myelin formation 26. Thus, if the morphological similarity of the inclusion bodies in normal human material and metachromatic leukodystrophy also reflects a similar chemical composition, their (probably transient) appearance could be the expression of an altered balance between sulphatide synthesis and degradation, the former being much larger than the latter. The close topographical relationship between these lipid bodies and myelinating axons is consistent with this hypothesis. From the results of the histochemical studies it seems to be likely that neutral lipid and complex lipid inclusions are present in the material investigated. They probably correspond to the inclusion bodies characterized ultrastructurally. Nevertheless, the possibility that some of the inclusion bodies are artefacts of fixation cannot be excluded completely, since fixation was done by immersion. This handicap however, is involved in any study of human material and should be kept in mind in the interpretations of similar findings in pathological material. ACKNOWLEDGEMENTS
I thank Dr. A. Bischoff for his support, and Dr. J. S. O'Brien for critical discussion of the biochemical questions involved. Miss. Th. Lauterburg kindly helped with the light microscopic preparations. The help of the stud. med. G. Nagel with the English translation is gratefully acknowledged. The study was supported by the Swiss National Fund, Grant No. 3.747.72.
REFERENCES 1 ADhfttI, M., SCHNECK,L., AND VOLK,B. W., Ultrastructural studies of eight cases of fetal TaySachs disease, Lab. Invest., 30 (1974) 102 112. 2 ADAMS,C. W. M., ANDDAVlSON,A. N., The occurrence of esterified cholesterol in the developing nervous system, Y. Neurochem., 10 (1959) 383 395. 3 BISCHOFF,A., UND ULRJCH, J., Amaurotische ldiotie in Verbindung mit metachromatischer Leukodystrophie: Uebergangsform oder Kombination?, Acta Neuropath. (Berl.), 8 (1967) 292-308. 4 BUNGE,R. P., Glial cells and the myelin sheath, Physiol. Rev., 48 (1968) 197-251. 5 BURKART,T., WJESMANN,U. N., ANDHERSCHKOWITZ,N. N., Developmental activity patterns of arylsulfatase A and fl-glucuronidase, Develop. Biol., in press. 6 DAVlSON,A. N., AND GREGSON,N. A., Metabolism of cellular membrane sulfolipids in the rat brain, Biochem. J., 98 (1972) 915-922. 7 FLECHSIG,P., Anatomic des menschlichen Gehirns undRiickemnarksaufmyelogenetischer Grundlaffe, Thieme, Leipzig, 1920. 8 GAMBLE,H. J., Electron microscope observations on the human foetal and embryonic spinal cord, d. Anat. (Lond.), 104 (1969) 435-453. 9 HILDEBRAND,C., Ultrastructural and light-microscopic studies of the feline spinal cord white matter, Acta physiol, scand., Suppl. 364 (1971) 81-134. 10 JACOB.H., Neurobiologie der Lebensalter, Fortschr. Neurol. Psychiat., in press. 11 LANGWORTHY,O. R., Development of behavior patterns and myelination of the nervous system in the human fetus and infant, Contr. Embryol. Carneg. Instn, 24 (1933) 1-57. 12 LEROY,J. G., VAN ELSEN,A. F., MARTIN,J. J., DUMON,J. E., HULET,A. E., OKODA,S., AND NAVARRO,C.. Infantile metachromatic leucodystrophy, N. EngL J. Med., 288 (1973) 1365-1369.
32 13 MALINSKY,J., Ontogenetic development of glial cells in the grey matter of the human spinal cord, Neuropath. pol., I0 (1972) 331 335. 14 MALINSKA,J., Developmental changes of glial elements in the white matter of the human spinal cord, Nenropath. pol., 10 (1972) 337 342. 15 MEIER, C., AND BISCHO~, A., Dysmyelination in "Jimpy" mouse, J. Neuropath. exp. Neuro/., 33 (1974) 343-353. 16 MEIER, C., AND BISCHOW, A., Oligodendroglial cell development in Jimpy mice and controls, J. neurol. &'i., 26 (1975) 517--528. 17 MICKtE, H. S., AND GJLLES, F. H., Changes in glial cells during human telencephalic myetinogenesis, Brain, 93 (1970) 337-346. 18 O'BR1EN, J. S., OKODA, S., FmLERUe, D. L., VEATH, M. L., ADORNATO,B., BRENNER, P. H., AND LERO'¢, J. G., Tay-Sachs disease: prenatal diagnosis, Science, 172 (1971) 61-64. 19 PETERS, A., AND VAUGHN, J. E., Morphology and development of the myelin sheath. In A. N. DAWSON AND A. PETERS(Eds.), Myelination, Springfield, 1970. 20 SVENNERHOLM,L., The distribution of lipids in the human nervous system, J. Neurochem.. I I (1964) 839 853. 21 UZMAN, B. G., The spiral configuration of myelin lamellae, J. Ultrastruct. Res., 11 (1964) 208-212. 22 UZMAN, B. G., AND HEDLEY-WHITE,E. T., Myelin: dynamic or stable?, J. gen. Physiol., 51 (1968) 8 18. 23 VAUGHN,J. E., An electron microscopic analysis of gliogenesis in rat optic nerves, Z. ZellJbrsch., 94 (1969) 293-324. 24 VmCHOW,R., Zur pathologischen Anatomie des Gehirns. I. Congenitale Encephalitis und Myelitis, Virchows Arch. path. Anat., 38 (1867) 129-138. 25 WELLS, M. A., AND Dn~TraEm J. C., A comprehensive study of the postnatal changes in the concentration of the lipids of the developing rat brain, Biochemistry, 6 (1967) 2169-3175. 26 WIESMANN, U. N., MEIER, C., SPYCHER, M. g., SCI-IMID, W., BISCHOIVF,A., GAUTIER, E., AND HERSCHKOWITZ,N., Prenatal metachromatic leucodystrophy, Heir. paediat. Acta, 30 (1975) 31 42.
21
© Elsevier Scientific Publishing Company, Amsterdam - Printed in The Netherlands
SOME OBSERVATIONS ON E A R L Y M Y E L I N A T I O N IN T H E H U M A N SPINAL CORD. L I G H T A N D E L E C T R O N MICROSCOPE STUDY
CLAUS MEIER
Department o[ Neurology, University of Berne, Berne (Switzerland) (Accepted August 10th, 1975)
SUMMARY
Segments of cervical spinal cord from a 23-week-old human foetus have been examined by light and electron microscopy. Myelinated fibres were found in the dorsal, ventral and peripheral lateral tracts, while the lateral corticospinal tract was completely unmyelinated. Myelin sheaths appeared to be formed by spiral wrapping of elongated mesaxons which originated from the apposition of the plasma membranes of oligodendrocytes. Preparations stained with Sudan red and Sudan black revealed the occurrence of lipid inclusions in the interfascicular glia. The topographical relation and the ultrastructural features of these inclusions are described. The possible significance of the inclusion bodies is discussed.
INTRODUCTION
Antenatal diagnosis by biochemical studies of amniotic fluid cell cultures have been proved successful in an increasing number of human hereditary metabolic disorders affecting the nervous systema2, is. After abortion morphological and biochemical investigations have been carried out to confirm diagnosis and for the purpose of studying the early pathological changesl,12,18, 26. However, very little is known about the ultrastructure of the normal developing human central nervous system (CNS), which obviously impairs comparison with the growing body of knowledge about the pathological materialS,13,14. This investigation was carried out on the CNS of a human foetus on which abortion was performed at risk of the mother after 23 weeks of gestation. The white matter of the cerebrum and cerebellum was not yet myelinated; the spinal cord, however, showed the beginning of myelination in some tracts, while others remained unmyelinated. The description of the topographical relation and the ultrastructural characterization of the lipid inclusions, known since Virchow's z4 early
22 report to occur in the developing human CNS, is regarded as the main purposc ol this study. MATERIAL AND METHODS
The observations were made on a foetus obtained by hysterectomy performed at risk of the mother after 23 weeks of gestation. The history of the mother did not suggest that foetal abnormality would be present and none was found. The foetus was 28 cm long (crown-to-heel) and weighted 920 g, compatible with a gestational age of 22-24 weeks. Small segments of the cervical spinal cord were dissected immediately after death and immersed in chilled 3 ~ glutaraldehyde in phosphate buffer at pH 7.3 for 4 h. Postfixation was performed with 1 ~ phosphate buffered osmium tetroxide for 4 h. After dehydration in acetone and embedding in Araldite, 1/zm semithin sections were stained with methylene blue for light microscopy. For electron microscopy thin sections from anterior, lateral and posterior fasciculi, representing the fasciculus gracilis et cuneatus, and the lateral and the ventral corticospinal tracts, were prepared and stained with uranyl acetate and lead citrate. The medulla oblongata adjacent to the segment of the cervical spinal cord embedded for electron microscopy, was fixed by immersion in 10 ~ formalin. Frozen sections were stained with Sudan red and Sudan black dyes*. RESULTS
Light microscopy In 1/~m semithin cross sections stained with methylene blue scattered myelinated fibres were readily identified in the ventral, dorsal and the peripheral lateral tracts, the latter representing the spinocerebellar, the spinoolivary and the spinotectal pathways (Fig. I a, c). Some of the commissural fibres were also myelinated. The region of the lateral corticospinal tract, however, was completely unmyelinated (Fig. 1b). The highest density of myelinated fibres was found in the depth of the ventral and dorsal tracts including the region of the anterior corticospinal pathway. Corresponding to the degree of myelination the incidence of glial cells showed the highest density in these regions. Active oligodendrocytes could easily be identified by their dark-staining cytoplasm with clumsy processes often found in contact with myelinated axons. Positive identification of astrocytes was possible by their large, light-staining nuclei. A third glial cell type, probably representing glioblasts, with small round or oval nuclei, was found in all tracts, but most often in unmyelinated regions. Mitotic figures were found with the same frequency in myelinating and unmyelinated regions with no predilection for the subependymal field around the central canal. About two to three dark-
* The histochemicalinvestigationwas carried out by Professor J. Ulrich, head of the Department of Neuropathotogy, Institute of Pathology, University of Basle, Switzerland. His help is gratefully acknowledged.
23
Fig. I. Spinal cord of 23-week-old human foetus. Cross-sectien of methylene blue-stained I t*m semitbin section, a: both ventral tracts show a fair amount of myelinated fibres. Asterisk on anterior median fissure, x 420. b: lateral corticospinal tract exhibiting no myelinated fibres. The density of glial cells is rather low. Arrow points to lipid accumulation, x 1050. c: myelinating ventral tract exhibiting a high density of glial cells, Most of them can be identified as 'active' oligodendrocytes by their dark staining clumsy processes sometimes found in touch to myelinating axons. Arrow points to lipid accumulation, x 1050.
staining n e c r o b i o t i c cells exhibiting p y k n o t i c nuclei were f o u n d on one entire crosssection o f the spinal cord. L i p i d a c c u m u l a t i o n s could be observed in myelinating a n d u n m y e l i n a t e d tracts (Fig. l b , c). In S u d a n b l a c k - s t a i n e d p r e p a r a t i o n s few m y e l i n a t e d fibres c o u l d be s u b s t a n t i a t e d in the long tracts. S o m e o f the interfascicular glia cells c o n t a i n e d deeply black-staining inclusions. In S u d a n red p r e p a r a t i o n s rare redstained material was f o u n d in the regions o f the l o n g t r a c t s ; however, a m o r e distinct localization was not possible. The S u d a n red-stained material exhibited a w e a k d o u b l e refraction in p o l a r i z e d light.
Electron microscopy In transverse section m o s t o f the axons in the m y e l i n a t i n g tracts exhibited calibres o f 0.2-0.5 fire. T h e y were sheathless a n d a r r a n g e d in fascicular bundles by radial-
Fig. 2. Spinal cord of 23-week-old human foetus. Small unmyelinated axon (ax) is completely enguticd by the perinuclear cytoplasm of an oligodendrocyte (OL). Arrow points to mesaxon formation. Two myelinated axons and one axon surrounded by loose myelin membranes seem to be aitached to che same oligodendrocyte. The oligodendroglial cytoplasm contains membraneous material ~ith a structure of myelin (asterisk). :: 30,000. Fig. 3. Transverse section of 23-week-old h u m a n foetus spinal cord. A stage of early myelmation m the lateral corticospinal tract. An axon (ax) is surrounded by a spiral of loose myelin. The oligodendroglial cytoplasm contains a lysosomal body with a 'tuffstone'-like appearance. The arrow points lo the beginning of mesaxon formation. : 30,000.
25
Fig. 4. Spinal cord of 23-week-old human foetus. Cross-section. Axon (ax) completely engulfed by the perinuclear cytoplasm of a myelinating oligodendrocyte (OL). Arrow points to external mesaxon linking the plasma membrane to the compact myelin sheath. × 15,000.
ly oriented astrocytic cell processes. A few axons of larger diameter were ensheathed by several loose layers or up to 18 layers of compact myelin. The calibres of myelinated fibres ranged between 0.5 and 2 #m. Processes of oligodendrocytes with a welldeveloped cytoplasm containing numerous microtubules were found in relation to myelinating axons (Fig. 2). Some of the oligodendroglial processes encircled the axon completely forming a mesaxon by conjunction of the plasma membranes of the opposed cytoplasmic lips. Elongation of the mesaxon led to the formation of a loose spiral around the central axon (Fig. 3). Twice, a myelinating axon was found which was surrounded completely by the perinuclear cytoplasm of an oligodendrocyte similar to the myelination of peripheral axons by Schwann cells (Figs. 2 and 4). Membranebound dense bodies, obviously of lysosomal origin, were found in the oligodendroglial cytoplasm adjacent to the myelinating axon outside or inside the myelin sheath, and
26
Fig. 5. Spinal cord of 23-week-old human foetus. Longitudinal section. Lysosomal bodies with a structured matrix in the inner cytoplasmic tongue (ol) ofa myelinating oligodendrocyte (a) and in the axoplasm (ax) ofa myelinated axon (b). More inclusion bodies are found in a glial cell process, probably of oligodendroglial origin (right). a, 19,000~b, ~ 13,000. also in the c y t o p l a s m i c loops near the nodes o f R a n v i e r (Figs. 3, 5a a n d 6). Some of these dense bodies exhibited circular m e m b r a n e profiles with a lamellar periodicity o f 5.8-6.0 nm a r o u n d a g r a n u l a r m a t r i x (Fig. 7). similar to the 'tuffstone bodies" as described by Bischoff a n d Ulrich 3 in m e t a c h r o m a t i c l e u c o d y s t r o p h y ( M L D ) material. C o n c e n t r i c l a m e l l a t e d bodies were found in u n m y e l i n a t e d a n d m y e l i n a t i n g axons (Fig. 5b). Indications o f a b u n d a n t myelination, as o b s e r v e d in the d e v e l o p i n g white
27
Fig. 6. Spinal cord of 23-week-old human foetus. Longitudinal section. Structured lysosomal inclusion (asterix) in a paranodal loop of an oligodendrocyte, × 76,0~)3. Fig. 7. High-power view of a lysosomal inclusion of a 'tuffstone' appearance, exhibiting a circular pattern of lamellated material with a periodicity of 5.8-6.0 nm (arrows). × 100,000.
m a t t e r o f kitten 9, r a t 19 a n d m o u s e 16 were very seldom e n c o u n t e r e d . O c c a s i o n a l l y small a m o u n t s o f vesicular dissolution o f myelin at the inner or o u t e r c y t o p l a s m i c t o n g u e were observed, b u t not myelin b r e a k d o w n . Infrequently o l i g o d e n d r o c y t e s were f o u n d which c o n t a i n e d large a m o u n t s o f h o m o g e n e o u s bodies o f weak o s m i o philia, p r o b a b l y representing a c c u m u l a t i o n s o f n e u t r a l lipids (Fig. 8a, b). Lipidc o n t a i n i n g glioblasts were f o u n d in the n o n - m y e l i n a t e d tracts. N e c r o b i o t i c cells a p p e a r very seldom. In the two instances in which they could be studied electron microscopically, it was not possible to discern their origin. Breakd o w n p r o d u c t s o f these cells were f o u n d in the intracellular space or in n e i g h b o u r i n g
28
Fig. 8. a and b: longitudinal section of 23-week-old foetus spinal cord. Glial cell closely attached to a myelinating axon (ax) is filled with m e m b r a n e - b o u n d homogeneous inclusions o f weak osmiophility. b: detail o f a . a, :~ 8000: b, × 30,000.
29
Fig. 9. Spinal cord of 23-week-old human foetus. Area of cell death in ventral tract. Parts of necrobiotic cells (asterisks) are phagocytozed by neighbouring astrocyte (AS). × 13,000.
astrocytic cell processes (Fig. 9). They could easily be distinguished from the inclusion bodies described above. DISCUSSION
The observations on myelination in human spinal cord as presented, confirm the current views about the formation of central myelin based upon studies of animal material as reviewed by BungO and Peters and Vaughn 19. Processes of glial cells with the cytological criteria of active oligodendrocvtes 23 were found in relation to myelin-
30 ating axons. Many times loose spirals of an elongated mesaxon, consisting of conjugated plasma membranes of these processes, could be observed encircling a central axon. In two instances myelinating axons were found completely engulfed in perinuclear oligodendroglial cytoplasm similar to the myelination process in peripheral nerves. The transient appearance of sudanophilic material at the time of initial myelination in the developing human central nervous system has been a matter of controversy since Virchow's first description of what he called 'Congenitale Encephalitis und Myelinitis' in 1867. While he considered it a pathological finding representing myelin breakdown products, others have suggested that this is a normal event during myelination2,17. In the view of the latter the lipid inclusions are thought to be myelin precursors which are gradually transformed into myelin2,17. In the early ultrastructural studies of myelination in the CNS in mouse 21 and rat ~9 lipid inclusions were not mentioned, but later lamellar lipid bodies have been recognized in the myelinating optic nerve of mice 22 and in the developing feline spinal cord white matter 9. While Uzman and Hedley-Whyte z2 regarded the lamellar bodies as transient storage depots of myelin lipid constituents, Hildebrand 9 interpreted them as breakdown products of myelin-glial units appearing during normal development. The observations made in this study confirm the view that lipid inclusion bodies occur in the normal myelinating human CNS. They are found either as homogeneous membrane-bound weakly osmiophilic droplets, or as structured bodies of varying appearance and strongly osmiophilic. Homogeneous lipid inclusions were found in glioblasts and oligodendrocytes. From their morphology these inclusions are probably neutral lipids, and according to chemical analysis of myelinating brains 2,z° they most likely are cholesterol esters. If these inclusions are cholesterol esters, they may be myelin precursors, as they were also found in unmyelinated tracts of the spinal cord where myelin breakdown could not have contributed to their occurrence. In this context it may be reasonable to mention that similar material, but in increased amounts, has been found in oligodendroglial cells of Jimpy mice 1,5 in which, because of a defect of the myelin synthesis, myelin precursor material may accumulate. Osmiophilic lipid bodies of lamellated or granular structure appeared in oligodendrocytes, and in the axoplasm of myelinated and unmyelinated fibres. In oligodendrocytes these bodies were particularly found near myelinating axons, i.e. in the external cytoplasmic tongue, or in the inner cytoplasmic loop between axolemma and myelin sheaths, or in the lateral cytoplasmic loops seen in longitudinal sections of the paranodes. They exhibited a granular ground substance and a circular lamellated arrangement with a lamellar periodicity of about 5.8 nm around a light centre. These bodies were similar to the sulphatide storage material in metachromatic leukodystrophy characterized by Bischoff and Ulrich 3 as 'tuffstone bodies'. The inclusions in axons mostly showed a circular lamellated pattern with a lamellar periodicity of 5.5-6.0 nm. Myelin formation is characterized by a very active synthesis of sulphatide 6,2'', part of which may be degraded normally in lysosomes before it ever can be incorporated into the myelin sheath. It has been shown 5 that the rate of sulphatide synthesis is
31 closely paralleled by the total activity of arylsulphatase A in the myelinating mouse brain, possibly to meet functionally the need of increased catabolism of sulphatide spilled over from myelin formation 26. Thus, if the morphological similarity of the inclusion bodies in normal human material and metachromatic leukodystrophy also reflects a similar chemical composition, their (probably transient) appearance could be the expression of an altered balance between sulphatide synthesis and degradation, the former being much larger than the latter. The close topographical relationship between these lipid bodies and myelinating axons is consistent with this hypothesis. From the results of the histochemical studies it seems to be likely that neutral lipid and complex lipid inclusions are present in the material investigated. They probably correspond to the inclusion bodies characterized ultrastructurally. Nevertheless, the possibility that some of the inclusion bodies are artefacts of fixation cannot be excluded completely, since fixation was done by immersion. This handicap however, is involved in any study of human material and should be kept in mind in the interpretations of similar findings in pathological material. ACKNOWLEDGEMENTS
I thank Dr. A. Bischoff for his support, and Dr. J. S. O'Brien for critical discussion of the biochemical questions involved. Miss. Th. Lauterburg kindly helped with the light microscopic preparations. The help of the stud. med. G. Nagel with the English translation is gratefully acknowledged. The study was supported by the Swiss National Fund, Grant No. 3.747.72.
REFERENCES 1 ADhfttI, M., SCHNECK,L., AND VOLK,B. W., Ultrastructural studies of eight cases of fetal TaySachs disease, Lab. Invest., 30 (1974) 102 112. 2 ADAMS,C. W. M., ANDDAVlSON,A. N., The occurrence of esterified cholesterol in the developing nervous system, Y. Neurochem., 10 (1959) 383 395. 3 BISCHOFF,A., UND ULRJCH, J., Amaurotische ldiotie in Verbindung mit metachromatischer Leukodystrophie: Uebergangsform oder Kombination?, Acta Neuropath. (Berl.), 8 (1967) 292-308. 4 BUNGE,R. P., Glial cells and the myelin sheath, Physiol. Rev., 48 (1968) 197-251. 5 BURKART,T., WJESMANN,U. N., ANDHERSCHKOWITZ,N. N., Developmental activity patterns of arylsulfatase A and fl-glucuronidase, Develop. Biol., in press. 6 DAVlSON,A. N., AND GREGSON,N. A., Metabolism of cellular membrane sulfolipids in the rat brain, Biochem. J., 98 (1972) 915-922. 7 FLECHSIG,P., Anatomic des menschlichen Gehirns undRiickemnarksaufmyelogenetischer Grundlaffe, Thieme, Leipzig, 1920. 8 GAMBLE,H. J., Electron microscope observations on the human foetal and embryonic spinal cord, d. Anat. (Lond.), 104 (1969) 435-453. 9 HILDEBRAND,C., Ultrastructural and light-microscopic studies of the feline spinal cord white matter, Acta physiol, scand., Suppl. 364 (1971) 81-134. 10 JACOB.H., Neurobiologie der Lebensalter, Fortschr. Neurol. Psychiat., in press. 11 LANGWORTHY,O. R., Development of behavior patterns and myelination of the nervous system in the human fetus and infant, Contr. Embryol. Carneg. Instn, 24 (1933) 1-57. 12 LEROY,J. G., VAN ELSEN,A. F., MARTIN,J. J., DUMON,J. E., HULET,A. E., OKODA,S., AND NAVARRO,C.. Infantile metachromatic leucodystrophy, N. EngL J. Med., 288 (1973) 1365-1369.
32 13 MALINSKY,J., Ontogenetic development of glial cells in the grey matter of the human spinal cord, Neuropath. pol., I0 (1972) 331 335. 14 MALINSKA,J., Developmental changes of glial elements in the white matter of the human spinal cord, Nenropath. pol., 10 (1972) 337 342. 15 MEIER, C., AND BISCHO~, A., Dysmyelination in "Jimpy" mouse, J. Neuropath. exp. Neuro/., 33 (1974) 343-353. 16 MEIER, C., AND BISCHOW, A., Oligodendroglial cell development in Jimpy mice and controls, J. neurol. &'i., 26 (1975) 517--528. 17 MICKtE, H. S., AND GJLLES, F. H., Changes in glial cells during human telencephalic myetinogenesis, Brain, 93 (1970) 337-346. 18 O'BR1EN, J. S., OKODA, S., FmLERUe, D. L., VEATH, M. L., ADORNATO,B., BRENNER, P. H., AND LERO'¢, J. G., Tay-Sachs disease: prenatal diagnosis, Science, 172 (1971) 61-64. 19 PETERS, A., AND VAUGHN, J. E., Morphology and development of the myelin sheath. In A. N. DAWSON AND A. PETERS(Eds.), Myelination, Springfield, 1970. 20 SVENNERHOLM,L., The distribution of lipids in the human nervous system, J. Neurochem.. I I (1964) 839 853. 21 UZMAN, B. G., The spiral configuration of myelin lamellae, J. Ultrastruct. Res., 11 (1964) 208-212. 22 UZMAN, B. G., AND HEDLEY-WHITE,E. T., Myelin: dynamic or stable?, J. gen. Physiol., 51 (1968) 8 18. 23 VAUGHN,J. E., An electron microscopic analysis of gliogenesis in rat optic nerves, Z. ZellJbrsch., 94 (1969) 293-324. 24 VmCHOW,R., Zur pathologischen Anatomie des Gehirns. I. Congenitale Encephalitis und Myelitis, Virchows Arch. path. Anat., 38 (1867) 129-138. 25 WELLS, M. A., AND Dn~TraEm J. C., A comprehensive study of the postnatal changes in the concentration of the lipids of the developing rat brain, Biochemistry, 6 (1967) 2169-3175. 26 WIESMANN, U. N., MEIER, C., SPYCHER, M. g., SCI-IMID, W., BISCHOIVF,A., GAUTIER, E., AND HERSCHKOWITZ,N., Prenatal metachromatic leucodystrophy, Heir. paediat. Acta, 30 (1975) 31 42.