The Ultrastructure of Normal and Abnormal Oligodendroglia’ SARAH A. LUSE Departments of Anatomy and Pathology, and the Beaumont-May Institute of Neurology, Washington University School of Medicine, Saint Louis, Missouri This review of the oligodendroglia is disposed principally in the white matter. presented as a tribute to Professor Nor- In Hortega’s opinion, oligodendrocytes mand Hoerr, under whose guidance the were analogous to the Schwann cells of the author received her basic training in the peripheral nerves and related in some way discipline and technics of neurohistology. to the formation and maintenance of the By means of metallic impregnations, myelin sheath, irrespective of what other Ramon y Cajal (’13) separated from the functions they might possess. astrocytic glia a group of small cells that Participation of the oligodendroglia in had not previously been stained selectively neoplasia was first suggested by Bailey and by any method then known. These aden- Hiller (’24) and two years later confirmed dritic cells he called the “third element” in by Bailey and Cushing (’26). Bailey and the central nervous system. Del Rio- Bucy (’29) were able to prove by the Hortega (’19, ’21) devised a silver carbo- silver impregnation methods of Hortega nate method that selectively stained the that these tumors actually were formed of “third element of Cajal” and demonstrated oligodendrocytes. Little can be added to that it was composed of two distinct cell their description. In hematoxylin and types : oligodendroglia and microglia. eosin or similar stains, they are characterHortega credited Robertson (1899, ’00) ized by numerous peculiarly uniform cells with the first description of the oligoden- with ovoid nuclei centrally placed in clear droglia and also equated them with the cytoplasmic compartments. The margins adendritic cells of Cajal. The historical of the clear halos about the nuclei stain developments in this field have been re- densely and have been considered fine viewed by both del Rio-Hortega (’28) and processes of either the oligodendroglia or of stromal astrocytes. Bailey and Bucy Glees (’55). Del Rio-Hortega (’21, ’28) described the (’29) and other observers have pointed out oligodendroglia as of variable size, and of the astrocytic stroma of these tumors as round, polyhedral or pyriform shape. Their well as the transitional neoplastic forms nuclei were round, and usually eccentri- that have both astrocytic and oligodendrogcally located with an accumulation of lial characteristics. The role of the oligodendroglia in cerecytoplasm at the pole from which the thickest process arose. Although Hortega bral edema has been more gradually stressed the paucity of both cytoplasm and recognized. Although Weed and McKibben processes, Penfield (’24, ’30) demonstrated (’19) described the accentuation of perithat this had been due to incomplete im- vascular spaces in experimental swelling pregnation. Both authors described the of the brain, they did not equate these oligodendrocytes as arranged in rows be- changes with an alteration in oligodendrogtween the nerve fibers of the white matter lial cells. Penfield and Cone (’26a, ’26b) (fig. 1); as satellites to neurons (fig. 2 ) ; and Cone (’28) studied the oligodendrogand as vascular satellites (fig. 3). Their lia in a variety of experimental and cliniubiquitous distribution, however, was in 1This work was supported in part by grants contrast to the protoplasmic astrocytes B-1539 and B-425 from the United States Public which were restricted to the gray matter Health Service, the National Institutes of Health, and to the fibrous astrocytes which were Bethesda, Maryland. 461 462 SARAH A. LUSE cal conditions and observed that the earliest manifestations of anoxia or toxicity in the central nervous system was an acute swelling of the oligodendrocytes. Unfortunately, these cells responded in a similar manner to autolysis. Courville ('42), Perret and Kernohan ('43) and Reid ('43), among others, have all ascribed at least the early changes in edema of the brain to swelling of the oligodendrocytes, although they could not exclude the possibility of an additional accumulation of extracellular fluid. Recently electron microscopy has shown that edema of the brain is associated with an increase in volume of the oligodendroglial cells without evidence of increased extracellular fluid (Luse and Harris, '59, '60; Torack, Terry and Zimmerman, '59, '60). Gerschenfeld, Wald, Zadunaisky and de Robertis ('59) also demonstrated an increase in volume of the pale cells in edema, although they identified these cells as astrocytes. In contrast to neoplasia and edema, the role of oligodendroglia in dehydration, metabolic alterations and demyelination has remained little understood. MATERIALS AND METHODS The tissues on which the present study is based were obtained from a variety of sources. Blocks were prepared from the cerebral cortex and cerebellum of adult cats; the spinal cord and cerebral cortex of adult rabbits; the spinal cord, medulla and cortex of young and adult mice and rats and from the spinal cord of chicks. Specimens from human cerebral cortex and from oligodendrogliomas were obtained at the time of surgery. Rabbit brains in which edema or dehydration had been produced experimentally were also examined, as were the spinal cords of rabbits with experimental allergic encephalomyelitis. Small blocks of tissue for electron microscopy were fixed one to two hours in a 1 % solution of osmium tetroxide in potassium dichromate, adjusted to a pH of 7.4 as recommended by Dalton ('55). Occasionally tissue was fixed in potassium permanganate as described by Luft ('56). Dehydration was in graded concentrations of ethanol (10% to 100% ) and infiltration was with a 7: 1 mixture of butyl and methyl methacrylates. Tissues were embedded in the partially prepolymerized methacrylate mixture to which benzoyl peroxide had been added as a catalyst, and then were hardened at 60°C overnight. Some tissues were embedded in epoxy resins (Araldite) according to the procedure of Richardson, Jarrett and Finke ('60). Thin sections were cut on a PorterBlum microtome ('53) using glass knives, and were mounted on collodion-covered copper grids. Sections were examined in RCA electron microscopes (models EMU2E, 3B, or 3C) without removing the plastic. Original micrographs were made at 1000 to 10,000 diameters and subsequently enlarged photographically. For light microscopy, tissue was fixed in formalin and embedded in paraffin for staining. Thick sections of osmium fixed plastic embedded material were treated with 5% periodic acid to remove the osmium tetroxide and then stained with a solution of 1% azure A and 1% methylene blue at an alkaline pH. OBSERVATIONS N o r m a l oligodendroglia A d u l t . Interfascicular oligodendrocytes in the white matter are arranged in short or long rows between the myelinated fibers, and tend to be smaller than either the vascular or neuronal satellite cells. Although usually round, their cytoplasm is plastic and may vary in shape depending on the pressures from adjacent cells. Their plasma membranes are smooth and usually delicate so that they may be broken during dehydration or embedding. Cytoplasm of the interfascicular oligodendrocytes is of two types. A majority have a pale cytoplasm (fig. 4 ) similar to that of satellite cells and have scant small mitochondria, scant Golgi membranes and only a small amount of ergastoplasm. Interspersed between these pale cells are others with similar rounded nuclei and outlines but with a more dense cytoplasm (figs. 5 and 6 ) . Mitochondria and the Golgi apparatus are similar in both types of oligodendroglia, the difference in the two being that the darker cells possess numerous small intracytoplasmic vesicles, some of which have attached RNA granules. Ocasionally, entirely different cells with dense cytoplasmic ULTRASTRUCTURE OF OLIGODENDROGLIA fibrils, the fibrous astrocytes, are intercalated in the rows of oligodendrocytes but are readily distinguished from the smoothy outlined oligodendroglia such as those demonstrated in figure 6. In both the white and gray matter, oligodendrocytes are arranged so that either their soma or processes are in intimate contact with vessels, forming vascular satellites. Hortega demonstrated their perivascular arrangement by the classic gold and silver impregnations (figs. 2 and 3). In electron micrographs pale oligodendroglial processes almost invariably contact the capillary wall (fig. 14) and occasionally as in figure 9 practically surround it. Between these processes that seemingly only touch the vascular wall are the firmly adherent astrocytic processes, the so-called astrocytic feet (fig. 14). In fortuitous sections one can actually demonstrate that the same cell is a satellite both to a vessel and to a neuron (fig. 9). A majority of the neuronal satellites are oligodendroglia, although protoplasmic astrocytes and microglia may also occur in this position. Satellite oligodendrocytes have uniformly round or oval nuclei with a single small nucleolus. Their most characteristic feature, however, is abundant pale cytoplasm with scant ergastoplasm and a paucity of Golgi membranes or mitochondria. Their plasma membranes are sharply outlined and most cells tend to be round in section with a single stout process that may extend from the cell body (fig. 9) or may branch dichotomously as in figure 8. In the neuropil the round or ovoid pale processes of oligodendrocytes are, for the most part, readily distinguished from other structures, but may be confused with unmyelinated axons if neurofilaments are not seen. Not all oligodendrocytes of the gray matter have their cell bodies adjacent to either vessels or neurons. Some may be isolated, or arranged in short rows (fig. 7). Whether or not their processes touch those of other oligodendroglial cells, vessels, or neurons, cannot be determined in thin sections and can only be surmised. The gliosomes of light microscopy apparently correspond to mitochondria since no other organelles of the proper size are present and since the gliosomes are stained 463 by specific mitochondrial stains. During myelin formation electron opaque cytoplasmic structures occur that are not mitochondria, but which may have constituted some of the gliosomes of light microscopy (fig. 11 ). In addition, in adult animals of all species examined, and especially in the human being, there are accumulations of dense irregularly outlined particles that we have equated with the lipochrome pigment granules associated with aging (fig. 21). Immature: during myelination. The brains and spinal cords of rats and mice one to 20 days of age were examined. No attempt was made to investigate the development of oligodendroglia prior to the time of birth or to identify their precursors. During the fust three weeks of life oligodendrocytes were more numerous than in the adult and somewhat larger with an irregular cytoplasmic contour. Their abundant cytoplasm, although pale, differed from that of the adult in their cytoplasmic components. Mitochondria were numerous and sometimes large. The ergastoplasm (endoplasmic reticulum) in some cells was arranged in distinct parallel arrays, whereas in others it was as inconspicuous as in adult cells. Occasionally large round to oval dense bodies were present as cytoplasmic inclusions, but their significance is unknown (fig. 11). Rare cytoplasmic lipid droplets occurred. Expansions arising from oligodendroglia were broad based, but distant from the cell body it was difficult if not impossible to distinguish them from unmyelinated axons in the one- to three-day-old animal. As myelination proceeded, this distinction was more readily made. In the 5- to 14-day-old animal, the voluminous pale expansions were often in close apposition to or surrounded the myelinating axon as has been described previously (Luse, '56b, '59). Vesicles were numerous in the terminal portions of oligodendroglial processes surrounding myelinating axons (fig. 12). Although the story of myelin formation is still incomplete, the presence of vesicles adjacent to myelinating fibers, both centrally and peripherally, suggests that the myelin lamellae are formed, at least in part, by fusion of cytoplasmic membranesurrounded vesicles which later become 464 SARAH A. LUSE organized to form uniform myelin membranes. The possibility of overlapping glial processes and their membranes contributing to the sheath has also been postulated (Luse, '56b). A b n m a l oligodendroglia Neoplasia. The tumors which we examined were all well differentiated oligodendrogliomas of the classical light microscopic appearance (fig. 18). At low magnifications, electron micrographs presented a picture strikingly remininscent of that seen with hematoxylin and eosin stains: rather uniform cells containing round to oval nuclei set in voluminous pale cytoplasm (figs. 18 and 19). The nuclear chromatin varied from finely granular to coarsely clumped and often formed a dense rim at the inner margin of the nuclear membrane. Although most nuclei had round or oval outlines, some were bizarre. Mitotic figures occurred. Nucleoli of tumor cells were large, sometimes multiple, and frequently had a unique punched out appearance (fig. 20). Cytoplasmic membranes of tumor cells were variable. Some were smooth and sharply delineated while others had numerous short, delicate expansions that interlocked with those of adjacent cells (fig. 24). Tumor cells often were clumped, the entire group being surrounded by stromalastrocytic processes or cells. Cytoplasm of the neoplastic oligodendroglia was abundant and pale due to the paucity of ergastoplasm, which was only slightly more prominent than in normal cells. The Golgi apparatus was rarely evident. Mitochondria of these cells often were vastly altered. Sometimes they were so numerous that they actually filled the cytoplasm (fig. 24), measuring as much as three CI in width. In addition there were cells with characteristic oligodendroglial nuclei that had cytoplasmic fibrils or even processes resembling those of astrocytes. Other cells with reniform or indented nuclei and astrocyte-like processes possessed the clear cytoplasm and numerous mitochondria of neoplastic oligodendroglial cells. The end result is that with the resolution and magnifications of electron microscopy, one still finds a few cells in oligodendrogliomas that defy attempts to classify them. Indeed, they possess characteristics of both the oligodendrocyte and of the astrocyte. This same dilemma was faced by Hortega ('21) and by Penfield ('32), who decided that there existed forms intermediate between oligodendrocytes and astrocytes. Dense irregular lipid droplets occurred in some tumor cells. An occasional cell was almost filled with fine or coarse extremely electron dense granules (fig. 22) which corresponded with the presence of calcium in contiguous thicker sections. Mitotic figures were not infrequent (fig. 23). Edema and dehydration Electron microscopy of the swollen brains of rabbits given intravenous distilled water revealed a marked increase in volume of oligodendroglial cytoplasm, both in the cell body and in its processes. This was particularly distinct about the vessels, where the distended perivascular space of light microscopy proved to be dilated oligodendroglial processes (fig. 15). With more severe swelling, cytoplasmic expansion of oligodendrocytes was sufficient to rupture cell membranes and to compress both astrocytic and neural processes. Spontaneous edema at the margins of metastatic tumors in the human being showed the same picture of oligodendroglial swelling without extracellular accumulation of fluid. With injection of hypertonic solutions ('50% sucrose, 25% glucose, 30% urea or 50% sorbitol), the cytoplasm of oligodendroglia was markedly shrunken (fig. 16). Nuclei of oligodendrocytes apparently were unaltered in dehydration, but their cytoplasm was reduced to a mere perinuclear rim. Cytoplasmic organelles were concentrated in the residual cytoplasm so that it was no longer pale. Similarly reduced in volume were the oligodendroglial processes about vessels and in the neuropil. With rehydration the cytoplasm of oligodendrocytes re-expanded. Injection of 5 % glucose intravenously produced little, if any, change in the appearance of the brain. In contrast excess intravenous 0.9% saline produced an unequivocal change in oligodendroglial cells. Numerous fine cytoplasmic tonofibrils ap- ULTRASTRUCTURE OF OLIGODENDROGLIA peared in the cytoplasm of almost all oligodendrocytes (fig. 17). The significance of this observation is not known, nor has the reaction of oligodendroglia to other salts yet been investigated. Demyelination: (experimental allergic encephalomyelitis) Allergic encephalomyelits is a disease in which demyelination occurs with sparing of the axis cylinders. Oligodendrocytes are distinctly altered in this disease. They are somewhat increased in volume. Their mitochondria are swollen, whereas mitochondria of adjacent microglial cells or axons are normal (fig. 26). Ergastoplasmic sacs of the oligodendroglia may also be distended. It appears that the primary insult in allergic encephalomyelitis may well be to the oligodendrocyte, and that the myelin disintegration is secondary to the oligodendroglial lesion (Luse and McDougal, '60). DISCUSSION The importance of oligodendroglia in the maintenance of the integrity of the nervous system has become increasingly apparent in recent years. Although Cajal postulated that they were in symbiosis with neurons, Hortega believed them to be the Schwann cells of the central nervous system and related to formation and maintenance of the myelin sheath. Both views have by now been amply verified. Penfield ('24, '30, '32) in particular has amplified the work of Hortega and substantiated his observations on the role of the oligodendrocyte in myelin formation; while at the same time Bailey extended the silver technics to neoplasia (Bailey and Hiller, '24; Bailey and Cushing, '26; and Bailey and Bucy, '29). Electron microscopy has opened a new field in the area of morphology. The identification of the various cellular components of the nervous system at the magnifications now available should be relatively simple. The reverse, however, appears to be the case, and a controversy has arisen over the identification of glia, cf. Farqhuar and Hartmann ('57), Gerschenfeld et al. ('59), de Robertis et al. ('60) and Schultz, Maynard and Pease ('57). 465 The cells with folded membranes and dense cytoplasm have been identified by Luse ('56a, '58), Palay ('57) and Dempsey and Luse ('57) as astrocytes. They bear a close resemblance to: ( 1 ) the cells in astrocytic tumors; and (2) cells proliferating at the margins of stab wounds, abscesses or injections of alumina gel, cells that may also be impregnated by the gold chloride-sublimate method of Cajal; and ( 3 ) many of the cells forming the pia-glial membrane of the mouse optic nerve and of the cat cortex. Furthermore, the cells equated by Luse ('56a, '56b, '59) and Palay ('57) with oligodendroglia (the satellite cells of the neurons and vessels and the rows of cells in the white matter) are strikingly similar to the cells of the 5 clear-cut oligodendroglial tumors that we have examined. The close investment of myelinating axons by pale cell cytoplasm (Luse '56b, '59); the degenerative changes in these same cells in allergic encephalomyelitis; and their increase in volume in cerebral edema further point to their being oligodendroglia. An exactly opposite evaluation of the glial cells has been arrived at by Farqhuar and Hartmann ('57), de Robertis, Gerschenfeld and Wald ('60), Gerschenfeld et al. ('59) and Schultz, Maynard and Pease ('57). Their opinions have been based in part on the abundance of pale processes that abut upon the blood vessels and that they have interpreted as astrocytic feet. Actually, oligodendroglia make as many, if not more, contacts with vessels than astrocytes, as Hortega ('21, '28) pointed out. Hortega did not consider these contacts of oligodendroglia and vessels to be as firmly adherent as the astrocytic "sucker feet" although Ferraro and Davidoff ('28) apparently did. Furthermore, these investigators have considered the pale processes at the pia-glial membrane as necessarily being astrocytic feet. However, their work has mostly been with the rat, rather than the cat, monkey or man. The piaglial membrane in the latter has a higher density of definite astrocytes than in the mouse, rat or rabbit, excepting the optic nerve. The astrocytes forming the pial surface of the brain are specialized in type and less complex than elsewhere, nor are all cells reaching the surface of the brain 466 SARAH A. LUSE necessarily astrocytes. Indeed, both Hortega ('21, '28) and Bailey and Schaltenbrandt ('27) have contended that the oligodendrocyte extends to and contributes to the pia-glial membrane. In edema of the brain, experimental or spontaneous, (Luse and Harris, ' 6 0 ) , the pale cells considered to be oligodendrocytes by us, have been massively increased in volume. Gerschenfeld et al. ('59) and de Robertis, Gerschenfeld and Wald ('60) have also seen swelling of pale cells in edema of the brain, but have identified them as astrocytes basing this in part on the recent work of Klatzo et al. ('57). Klatzo's identification of edema as being an astrocytic change is at variance with an extensive literature on this subject: Courville ('42), Hassin ('28), Scheinker ('41, '47), Perret and Kernochan ('43) Evans and Scheinker ('45), as well as the experimental work of Reid ('43), Penfield and Cone ('26a, '26b) and Cone ('28). All of these investigators demonstrated swelling of oligodendrocytes in cerebral edema and Reid ('43) and Cone ('28) also demonstrated the swollen oligodendrocytes in edema by the silver carbonate stains of Hortega. Electron microscopy has revealed the lack of an extracellular fluid compartment in edema of the brain (Luse and Harris, '59, '60; Gerschenfeld et al., '59; Torack, Terry and Zimmerman, '59, '60). Even more remarkable is the fact that the only cellular element that undergoes detectable change in edema or dehydration is the oligodendrocyte. The function of the neuroglial cells has tantalized both the morphologist and the chemist. The supportive functions of the fibrous astrocyte are clear enough; but little is known of the role of the protoplasmic astrocyte, nor of the oligodendroglial cells. Electron microscopy has confirmed the beliefs and observations of del Rio-Hortega and of Penfield as to the close association of oligodendrocytes with the formation of myelin (figs. 12 and 13). Electron microscopy of the spinal cord in allergic encephalomyelitis has further strengthened this relation by showing that the first lesion in myelin destruction is a degenerative alteration in the oligodendrocyte. Greenfield ('33) and Adams and Kubik ('52), among others, had suggested that some clinical types of demyelinating diseases might be related to changes in the interfascicular cells. Cajal's concept of symbiosis of oligodendrocytes and neurons also has been amplified by electron microscopy. In the central nervous system, cells and cellular processes are so closely wedged together that the distances between any two are of the order of 100 to 200 Angstrom units, with a resulting extracellular space calculated by Horstmann and Meves ('57) as not exceeding 3 to 5% ; whereas the chloride space had been calculated as approximately 30 to 35% (Woodbury, '57). Thus, no extracellular space is available in the central nervous system for transport of fluid or electrolytes, the most likely site of such transport being the oligodendroglial cell. The oligodendroglial cell expands in edema and contracts with dehydration; and since it makes both vascular and neuronal contacts (fig. 9 ) forms a possible channel for cellular exchange by neurons otherwise isolated from the blood vascular system. Although little has yet been done in the areas of enzyme or microchemistry of the glial cells, some interesting differences have been suggested. Kluver ('55) has described a high porphyrin content in the nervous system, apparently located in the oligodendrocytes. C avanaugh, Thompson and Webster ('54) have demonstrated large amounts of pseudocholinesterase in the white matter which they belive to be in oligodendrocytes; and Ashby and coworkers ('52) an elevated carbonic anhydrase. These changes could indicate a role in the carbon dioxide or oxygen transport systems of the cell or in membrane permeability. In addition, Pope, Hess and Allen ('55), Victor and Wolf ('37) and Heller and Elliott ('55) have shown that the oxygen consumption of oligodendroglial tumors is considerably higher than that of astrocytic tumors and that mixed tumors with an increased number of oligodendrocytes have an increased level of cytochrome oxidase. The findings in tissue culture of both normal and neoplastic oligodendroglial cells has also been informative. Canti, Bland and Russell ('37) first cultured the ULTRASTRUCTURE OF OLIGODENDROGLIA oligodendroglioma. Normal oligodendrocytes have shown a similar morphology and a similar pulsatile activity which distinguishes them from the astrocytes (Lumsden and Pomerat, '51; Pomerat, '57). Benitez, Murray and Wooley ('55) and Murray ('57) have added to this their demonstration, in vitro, of the tonic contraction response of oligodendrocytes to the addition of serotonin, which suggests that the metabolically active oligodendrocyte may have a pumping action which is stimulated by serotonin. SUMMARY Normal and pathologic oligodendroglia have been studied. They are usually pale round cells arranged in rows in the white matter or as satellites to neurons and vessels. During the time of myelin formation and in the immature animal, they are more voluminous than in the adult and are intimately associated with the rnyelinating axons. In experimental or spontaneous cerebral edema, oligodendroglial cytoplasm is dilated and conversely in dehydration the oligodendroglial cytoplasm shrinks. In allergic encephalomyelitis the earliest lesion apparently is a degenerative change in the oligodendrocyte with swelling of both mitochondria and ergastoplasm, and secondarily degeneration of myelin. Oligodendroglial tumors are composed of cells reminiscent of normal oligodendrocytes with voluminous pale cytoplasm and round or bizarre nuclei. It is suggested that the oligodendrocytes subserve at least two functions: (1) formation of the myelin sheath, and (2) a pathway for fluid and electrolyte transport within the nervous system in lieu of an extraceUuIar space. ACKNOWLEDGMENTS It is a pleasure to acknowledge the collaboration of Dr. Basil Harris in the studies on edema and dehydration of the brain. I am also indebted to Miss Ann Jones and Miss Helen Ferguson for their technical assistance. LITERATURE CITED Adams, R. D., and C. Kubik 1952 The morbid anatomy of the demyelinative diseases. Am. J. Med., 12: 510-546. 467 Ashby, W., R. F. Garzoli and E. M. Schuster 1952 Relative distribution patterns of three brain enzymes, carbonic anhydrase, choline esterase and acetyl phosphatase. Am. J. Physiol., 170: 116-120. Bailey, P., and P. C. Bucy 1929 Oligodendrogliomas of the brain. J. Path. Bact., 32: 735751. Bailey, P., and H. Cushing 1926 A Classification of the Tumors of the Glioma Group on a Histogenetic Basis with a Correlated Study of Prognosis. Lippincott, Philadelphia. Bailey, P., and G. Hiller 1924 The interstitial tissues of the central nervous system; a review. J. Nerv. Ment. Dis., 59: 337-361. Bailey, P., and G. Schaltenbrandt 1927 Die mukoese Degeneration de Oligodendroglia. Deutsche Ztschr. Nervenh., 97: 231-237. Benitez, H. H.,M. R. Murray and D. W. Wooley 1955 Effects of serotonin and certain of its antagonists upon oligodendroglial cells in vitro. Excerpta Med., Sect. 8, 8: 877. Canti, R. B., J. W. Bland and D. S. Russell 1937 Tissue culture of gliomata. Cinematographic demonstration. Res. Pub. Assoc. Res. Nerv. Ment. Dis., 16: 1-24. Cavanaugh, J. B., R. H. S. Thompson and G. R. Webster 1954 The localization of pseudocholinesterase activity in nervous tissue. Quart. Exp. Physiol., Lond., 39: 185-197. Cone, W. 1928 Acute pathologic changes in neuroglia and microglia. Arch. New. Psychiat., 20: 34. Courville, C. B. 1942 Structural changes in the brain consequent to traumatic disturbances of intracranial fluid balance. Bull. Los Angeles Neur. SOC., 7: 55-76. Dalton, A. J. 1955 Chrome-osmium fixative for electron microscopy. Anat. Rec., 121: 281. Dempsey, E. W.,and S. Luse 1957 Fine structure of the neuropil in relation to neuroglia cells. In: Biology of Neuroglia, ed. W. F. Windle, Charles C Thomas, Springfield, Illinois, chap. 7. Evans, J. A., and I. M. Scheinker 1945 Histologic studies of the brain following head trauma. I. Post-traumatic cerebral swelling and edema. J. Neurosurg., 2: 306-314. Farqhuar, M. G., and J. F. Hartmann 1957 Neuroglial structure and relationships as revealed by electron microscopy. J. Neuropath. Exp. Neur., 16: 18-39. Ferraro, A., and L. M. Davidoff 1928 The reaction of the oligodendroglia to injury of the brain. Arch, Path., 6: 1030-1053. Gerschenfeld, H. M., F. Wald, J. A. Zadunaisky and E. D. P. de Robertis 1959 Function of astroglia in the water-ion metabolism of the central nervous system. Neurology, 9: 41-25. Glees, P. 1955 Neuroglia, Morphology and Function. Charles C Thomas, Springfield, Illinois. Greenfield, J. G. 1933 A form of progressive cerebral sclerosis in infants associated with primary degeneration of the interfascicular glia. J. Neur. Psychopath., 13: 289-302. Hassin, G. G. 1928 Changes in the brain in increased intracranial pressure. Arch. Neur. Psychiat., 20: 1172-1189. 468 S A R A H A. LUSE Heller, I. H., and K. A. C. Elliott 1955 The intracranial tumors (so-called edema or swellmetabolism of normal brain and brain tumors ing of the brain). J. Neuropath. Exp. Neur., in relation to cell type and density. Canad. J. 2: 341-352. Biochem. Physiol., 33: 395-403. Pomerat, C. M. 1957 Functional concepts based Horstmann, E., and H. Meves 1957 Die Feinon tissue culture studies of neuroglia cells. struktur des molekularen Rindengraves und In: Biology of Neuroglia, ed. W. F. Windle. ihre physiologische Bedeutung. Ztschr. ZellCharles C Thomas, Springfield, Ill., pp. 167forsch., 49: 569-604. 175. Klatzo, Igor, A. Piraux and E. J. Laskowski 1958 Pope, A., H. H. Hess and J. N. Allen 1957 The relationship between edema, blood-brain Quantitative histochemistry of proteolytic and barrier and tissue elements in a local brain oxidative enzymes in human cerebral cortex injury. J. Neuropath. Exp. Neur., 17: 548-564. and brain tumors. Chapter 10 in: Progress in Kliiver, H. 1955 Porphyrins in relation to the Neurobiology, vol. 2, “Ultrastructure and Celludevelopment of the nervous system. In: Biolar Chemistry of Neural Tissue,” ed. H. chemistry of the Developing Nervous System, Waelsch. P. B. Hoeber, New York. H. Waelsch, ed. Academic Press, New York. Porter, K. R., and J. Blum 1953 A study in Luft, J. H. 1956 Permanganate-A new fixamicrotomy for electron microscopy. Anat. Rec., tive for electron microscopy. J. Biophys. Bio117: 685-710. chem. Cytol., 2: 799-801. Ramon y Cajal, S. 1913 Sobre u n nuevo proLumsden, C. E., and C. M. Pomerat 1951 ceder de impregnacion de la neuroglia y sus Normal oligodendrocytes in tissue culture. Exp. resultados en 10s centros nerviosos del hombre Cell Res., 2: 103-114. y animales. Trab. Lab. Inv. Biol. Univ. Madrid, Luse, S. A. 1956a Electron microscopic observa11, 219. tions of the central nervous system. J. Biophys. Reid, W. L. 1943 Cerebral oedema. Austral. Biochem. Cytol., 2: 531-541. N. Zealand J. Surg., 17: 439446. 19561, Formation of myelin in the central nervous system of mice and rats, as stud- Richardson, K. C., L. 0. Jarrett and E. H. Finke 1960 Notes on the use of Araldite epoxy resins ied with the electron microscope. Ibid., 2: for ultra-thin sectioning in electron microscopy. 777-784. Stain Tech., in press. 1958 Ultrastructure of reactive and neoplastic astrocytes. Lab. Invest., 7: 401-417. del Rio-Hortega, P. 1919 El “tercer elements” 1959 The fine structure of the morphode 10s centros nerviosos. I. La microglia e n genesis of myelin. Progress in Neurobiology, estado normal. 11. Intervencion de la microg4: 59-81. lia e n 10s procesos patologicos. 111. NaturLuse, S. A., and B. Harris 1959 Electron mialeza probable de l a microglia. Bol. SOC. Esp. croscopy of the brain in experimental edema. Biol., 9: 69. Anat. Rec., 133: 305. 1921 Estudios sobre la neurolgia. La 1960 Electron microscopy of the brain glia de escasas radiaciones (oligodendroglia). in experimental edema. J. Neurosurg., 17: Bol. Real SOC.Expan. Hist. Nat., 21: 63-92. 439-446. 1928 Tercera aportacion a1 conocimiLuse, S. A., and J. B. McDougal, Jr. 1960 ento morfologico e interpretacion funcional de Electron microscopic observations on allergic la oligodendroglia. Mem. Real SOC. Expan. encephalomyelitis in the rabbit. J. Exp. Med., Hist. Nat., 14: 1-122. in press. de Robertis, E. D. P., H. &I. Gerschenfeld and Murray, M. R. 1957 Response of oligodendroF. Wald 1960 Some aspects of glial function cytes to serotonin. I n : Biology of Neuroglia, as revealed by electron microscopy. In: ed. W. F. Windle. Charles C Thomas, SpringFourth International Conference on Electron field, Illinois, pp. 176-180. Microscopy, Springer-Verlag, Berlin, pp. 443Palay, S. L. 1957 An electron microscopical 447. study of neuroglia. Ibid., pp. 24-38. Robertson, W. F. 1899 On a new method of Penfield, W. 1924 Oligodendroglia and its relaobtaining a black reaction in certain tissuetion to classical neuroglia. Brain, 47: 430-452. elements of the central nervous system (plat1930 Further modification of del Rioinum method). Scott. Med. Surg. J., 4: 23Hortega’s method of staining oligodendroglia. 1900 A microscopic demonstration of Am. J. Path., 6: 445448. the normal and pathologic histology of mesog1932 Neuroglia and microglia. The inlia cells. J. Ment. Sci., 46: 724. terstitial tissue of the central nervous system. In: Special Cytology, ed. E. C. Cowdry. Paul Scheinker, I. 1941 Cerebral swelling and edema associated with cerebral tumor. A histogenetic B. Hoeber, N. Y., 2nd ed., pp. 1447-1482. and histopathologic study. Arch. Neur. PsyPenfield, W., and W. Cone 1926a Acute swellchiat., 45: 117-129. ing of oligodendroglia. A specific type of neuroglia change. Arch. Neur. Psychiat., 16: 131- Scheinker, I. M. 1947 Cerebral swelling. Histopathology, classification and clinical signifi153. cance of brain edema. J. Neurosurg., 4: 25519261, The acute regressive changes of 275. neuroglia (ameboid glia and acute swelling of oligodendroglia). J. f. Psych. Neur., 34: 204- Schultz, R. L., E. A. Maynard and D. C. Pease 224. 1957 Electron microscopy of neurons and Perret, G. E., and J. W. Kernohan 1943 Histoneuroglia of cerebral cortex and corpus callopathologic changes of the brain caused by sum. Am. J. Anat., 100: 368408. - - - - - ULTRASTRUCTURE OF OLIGODENDROGLIA Torack, R. M., R. D. Terry and H. M. Zimmerman 1959 The fine structure of cerebral fluid accumulation. I. Swelling secondary to cold injury. Am. J. Path., 35: 1135-1147. 1960 The fine structure of cerebral fluid accumulation. 11. Swelling produced by triethyl tin poisoning and its comparison with that in the human brain. Ibid., 36: 273-287. 469 Victor, J. V., and A. Wolf 1937 Metabolism of brain tumors. Res. Publ. Assoc. Res. New. Merit. Dis., 1 6 : 44-58. Weed, L. H., and P. S. McKibben 1919 Experimental alteration of brain bulk. Am. J. Physiol., 48: 531-558. Woodbury, D. M. 1957 I n : Biology of Neuroglia, ed. W. F. Windle. Charles C Thomas, Springfield, Ill., pp. 120-126. PLATE 1 EXPLANATION OF FIGURES 470 1 This is a copy of figure 2, Wortega (’21), “Neuroglia de la substancia blanca del cerebro humano, tenida por el metodo aurjco de Cajal. A, glia interfascicular de escasas radiaciones; B, gliocio fibroso de largas radiaciones.” It demonstrates oligodendroglial cells “A” arranged in rows in the white matter and their lack of processes in the gold chloride-sublimate stain, whereas radiations of astrocyes “B” are depicted. 2 This is a copy of figure 1 from Hortega (’21). “Neuroglia de la corteza cerebral humana, tenida por el metodo aurico de Capal: A oligodendroglia; B, gliocito protoplasmic0 o de cortas radiaciones; C, nucleo de microglia.” 3 This a copy of figure 66 from Hortega (’21). “Substancia blance de una laminilla cerebeloso del mono. Junto a 10s vasos existen abundantes gliocitos con escasas radiaciones, bastante largas y rainificadas.” This demonstrates the difference in staining with the silver carbonate method. ULTRASTRUCTURE OF OLIGODENDROGLIA Sarah A. Luse PLATE 1 471 PLATE 2 EXPLANATION O F FIGURES 4 Electron micrograph of a small oligodendroglial cell i n the white matter of the brain of a n adult mouse. Numerous myelinated axons are present. The cytoplasm is pale with scant mitochondria, ergastoplasm or Golgi membranes. 5 x 9,000. Numerous large and small myelinated fibers surround this oligodendrocyte in the white matter of the brain of a n adult albino rat. The oligodendrocyte is almost round with a single process evident at its upper margin (arrow). The round nucleus is centrally located. The cytoplasm of this cell is in contrast to figure 4, in that numerous vesicles, some associated with RNA granules, are present. X 4,000. 6 Electron micrograph from spinal cord of a newborn chick. Parts of 4 cells are lined up in a row between myelinated axons. A margin of myelin is evident at the arrows at the bottom of the figure. Two of the cells, A and C, have dense cytoplasm, and two others, B and D, have pale cytoplasm. The cytoplasm of C is of intermediate density and is readily recognizable as olgiodendroglial, whereas that of A is so dense that its possible oligodendroglial origin is suggested by position only. X 7,500. 4 72 ULTRASTRUCTURE OF OLIGODENDROGLIA Sarah A. Luse PLATE 2 473 PLATE 3 EXPLANATION OF FIGURE 7 474 Two oligodendroglial cells deep in the cortex of the brain of an adult rabbit. Part of a neuron is present at the left. The oligodendroglial nuclei are almost round. The cytoplasm is abundant, pale, and has few organelles. The upper cell has two blunt processes and the lower one a single process (arrows). X 10,000. ULTRASTRUCTURE OF OLIGODENDROGLIA Sarah A. Luse PLATE 3 475 PLATE 4 EXPLANATION O F FIGURES 8 This is a n electron micrograph of a perineuronal satellite. The oligodendrocyte is surrounded by two dendrites ( D ) and is not separated from them by any other cell or cell process. The oligodendroglial nucleus is round and its cytoplasm pale. The cytoplasmic organelles are, for the most part, concentrated near the zone where the two oligodendroglial processes (arrows) are arising. Rabbit cortex. x 6,000. 9 Electron micrograph of oligodendrocyte (oligo) that is a satellite to a neuron and also is i n contact with a blood vessel ( V ) . Rabbit cortex. 476 x 4,500. ULTRASTRUCTURE OF OLIGODENDROGLIA Sarah A. Luse PLATE 4 477 PLATE 5 EXPLANATION O F FIGURES 10 Electron micrograph of oligodendrocyte i n the medulla of a mouse 6 days after birth. The cytoplasm is more abundant than in similar cells from mature animals and contains more cytoplasmic organelles. The number of cytoplasmic vesicles with or without RNA granules is greater than in cells from adult animals. X 6,500. 11 478 Electron micrograph of oligodendroglial process from spinal cord of a one-day-old mouse. The prominent granular component of the cytoplasm is evident. In addition, three dense aggregates are present (arrows). Their significance is unknown, but they are seen almost exclusively i n immature animals. These aggregates are readily distinguished from the lipofuchsin granules seen i n aged animals or i n humans (fig. 21). >( 16,000. ULTRASTRUCTURE OF OLIGODENDROGLIA Sarah A. Luse PLATE 5 479 PLATE 6 EXPLANATION O F FIGURES 12 Electron micrograph of a n axon which is i n the process of myelination. Glial cytoplasm (G) is loculated between the axolemma and the outer layers of myelin. Surrounding the outermost myelin lamella are at least oligodendroglial processes, some of which contain numerous vesicles. Their plasma membranes are evident at the arrows. An oligodendroglial nucleus is present a t the left ( N ) . X 12,000. 13 Electron micrograph of medulla of an adult mouse. A myelinated axon is present in the center of the field and is in close apposition to a large oligodendroglial process (oligo). Numerous neuronal processes that are approaching their terminations are present ( N ) and are recognizable by their numerous vesicles. Embedded i n Araldite and stained with potassium pcrmanganate. X 12,000. ULTRASTRUCTURE OF OLIGODENDROGLIA Sarah A. Luse PLATE 6 481 PLATE 7 EXPLANATION 482 O F FIGURES 14 Capillary in the brain of a n adult rabbit. Much of the outer surface of the vessel is covered by astrocytic process (arrows). Three oligodendroglial process ( O L ) are present at the lower right. : :9,000. 15 Cerebral edema i n a n adult rabbit given intravenous distilled water. The capillary is surrounded by distended oligodendroglial processes. Mitochondria (M) are still present. Plasma membranes ( P ) are evident. The neuropil actually has been compressed by the distended oligodendroglial processes. X 6,000. ULTRASTRUCTURE OF OLIGODENDROGLIA Sarah A. Luse PLATE 7 483 PLATE 8 EXPLANATION O F FIGURES 16 Electron micrograph of a n oligodendrocyte in the brain of a rabbit that had been given intravenous 50% sucrose. The oligodendroglial cytoplasm is shrunk to only a narrow rim about the nucleus. A process is extending out a t the lower left (PR) and most of the organelles of the cell are concentrated in this region. x 12,000. 17 Oligodendrocyte from the brain of a rabbit that had received 300 cm3 of IV normal saline. The cytoplasm (arrows) contains numerous delicate fibrils. X 12,000. 484 ULTRASTRUCTURE OF OLIGODENDROGLIA Sarah A. Luse PLATE 8 485 PLATE 9 EXPLANATION O F FIGURES 18 Electron micrograph of a n oligodendroglioma (light micrograph is shown i n insert at upper left). The cells are irregular with delicate plasma membranes and ovoid to round nuclei. The cytoplasm is somewhat pale with scant organelles. X 5,000. 19 486 Electron micrograph of another region from the same tumor as that in figure 18. The pale cytoplasm is more voluminous in this group of cells. Part of a blood vessel is evident at the upper right (vessel). X 5,000. ULTRASTRUCTURE OF OLIGODENDROGLIA Sarah A. Luse PLATE 9 487 PLATE 10 EXPLANATION O F FIGURES 4aa 20 Electron micrograph of part of the nucleus of a tumor cell in another oligodendroglioma. The nucleolus is abnormally large with a number of punctate pale zones. x 10,000. 21 Lipochrome pigment in the cytoplasm of a normal oligodendrocyte in the brain of a n adult human being. X 10,000. 22 Cytoplasm of part of a neoplastic oligodendroglial cell from the same tumor as figures 18 and 19. Dense irregular aggregates i n the cytoplasm (arrows) are one of the manifestations of calcium deposition in these cells. X 10,000. 23 Mitotic figure occurring in an oligodendroglioma. The nuclear membrane has disappeared. The cytoplasm is more dense than usual. Mitochondria are not seen. The cell is in metaphase, and has been sectioned equatorially through the metaphase plate. x 10,000. ULTRASTRUCTURE OF OLIGODENDROG1,IA Sarah A. Luse PLATE 10 489 PLATE 11 EXPLANATION OF FIGURES 490 24 Electron micrograph of parts of 4 cells i n a n oligodendroglioma. This figure illustrates two points: in the center the fine process of adjacent cells are interdigitated with each other, and at the upper right the cytoplasm of one cell is almost filled with normal sized mitochondria. X 12,000. 25 Electron micrograph of another oligodendroglial tumor cell in which some of the mitochondria are of normal size and others are abnormally large ( M ) . In the lower central mitochondrion (arrow) there are a few cristae extending across the cell. In the two upper large ones there are lipid inclusions. X 12,000. ULTRASTRUCTURE OF OLIGODENDROGLIA Sarah A. Luse PLATE 11 491 ULTRASTRUCTURE OF OLIGODENDROGLIA Sarah A. Luse 26 492 PLATE 12 Electron micrograph of a n oligodendroglial cell in allergic encephalomyelitis. The nucleus is normal. The cytoplasm is increased in amount. The mitochondria ( M ) and ergastoplasmic sacs ( E ) are swollen. A few fragmented mitochondria1 cristae remain (arrows). The microglial phagocytic cell at the upper right contains myelin debris and mitochondria ( M ) that are normal. x 10,000.