* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download differentiation of neuronal types and synapses in myelinating
Nonsynaptic plasticity wikipedia , lookup
Biological neuron model wikipedia , lookup
Clinical neurochemistry wikipedia , lookup
Molecular neuroscience wikipedia , lookup
Premovement neuronal activity wikipedia , lookup
Node of Ranvier wikipedia , lookup
Apical dendrite wikipedia , lookup
Subventricular zone wikipedia , lookup
Neuroregeneration wikipedia , lookup
Stimulus (physiology) wikipedia , lookup
Circumventricular organs wikipedia , lookup
Eyeblink conditioning wikipedia , lookup
Neuropsychopharmacology wikipedia , lookup
Nervous system network models wikipedia , lookup
Chemical synapse wikipedia , lookup
Multielectrode array wikipedia , lookup
Synaptic gating wikipedia , lookup
Axon guidance wikipedia , lookup
Optogenetics wikipedia , lookup
Development of the nervous system wikipedia , lookup
Synaptogenesis wikipedia , lookup
Feature detection (nervous system) wikipedia , lookup
Published July 1, 1964 DIFFERENTIATION OF N E U R O N A L TYPES AND SYNAPSES IN MYELINATING C U L T U R E S OF M O U S E C E R E B E L L U M MERRILL K. WOLF, M.D. From the Laboratory of Cellular Neuropathology, Harvard Medical School~ Boston ABSTRACT When central nervous tissue from fetal or neonatal animals is cultivated in vitro, the starting material is a complex neuroepithelium which contains many types of neurons at various stages of maturation, as well as neuroglial, ependymal, and mesodermal cells. In the subsequent development of a culture, neurons might die, dedifferentiate, survive unaltered, or undergo some part of their normal maturation. The precision with which morphogenetic events in vitro can be analyzed and the architecture of individual cultures described will directly determine the kinds of information about central nervous tissue which can be obtained from tissue culture experiments. The cerebellum was the first part of the central nervous system from which myelin formation in culture was obtained (5). Several investigators have now prepared cultures of cerebellum containing myelinated axons (1, 2, 5, 7, 12-15, 18, 20). However, myelin formation is only one aspect of the differentiation of intact cerebellum. Developing cerebellar neurons undergo complex migrations, acquire characteristic shapes, and enter into elaborately ordered relationships with other neurons. Little is known about the extent to which these aspects of the differentiation of cerebellar neurons may proceed in vitro. Pomerat and Costero (16) described several types of neurons in cultured cerebellum, and laid special emphasis on the emigration of granule cell neurons into the outgrowth zone of cultures. Similar findings are reported by Nakai and Okamoto (13) and by Yonezawa et al. 259 Downloaded from on June 16, 2017 T h e Holmes silver impregnation method has made possible the recognition of multiple ncuronal types and synapses in myelinating cultures of mouse cerebellum. Well stained large and medium-sized neurons are always found in small numbers near epcndymal formations and arc considered to be roof nuclear neurons. Neurons with poorly stained somas, abruptly dcmarkcd from intcnsely stained axons, arc numerous and often are arranged in palisades. With prolonged maintcnance in vitro these neurons develop some but not all of the features of mature Purkinjc cells. A few small, densely stained, bipolar neurons, often with one process bifurcated, are found in dense regions of some cultures of newborn ccrebcllure. These neurons are commoncr in cultures from cerebella of older mice. They closely resemble the immature granule cell in vivo. All the neuron types recognized in cultures are present in thc initial explants; neurons differentiate further in vitro, but new neurons probably do not form. Synaptic boutons arc found on somas and dendrites of many Purkinjc cells. Two cultures contained structures resembling the basket cndings which surround Purkinje cell somas in vivo. Thc complexity of neuronal relationships in cultures of central nervous tissue is emphasized. Published July 1, 1964 (20). However, Piper (I 5) reported degeneration of all the granule cells in cultured cerebellum. Bornstein a n d his collaborators (1, 2, 18) h a v e e m p h a sized the uniformity of the neurons in their cultures of cerebellum, as studied b o t h by light a n d electron microscopy. T h e Holmes silver nitrate pyridine m e t h o d for MATERIALS AND METHODS Mouse cerebellum was chosen, so that observations on cultures might be related to the schedule of cell divisions and migrations worked out for the developing cerebellum of that species ( l l ) . The corpus cerebelli of newborn mice was removed, stripped of meninges, and divided by parasagittal section into 6 neurofibrils has, in the a u t h o r ' s hands, shown rem a r k a b l e advantages as a stain for whole m o u n t cultures of central nervous tissues. T h e favorable properties of this stain have m a d e it possible to examine the detailed a n a t o m y of cerebellar cultures. A greater degree of differentiation is d e m o n strated t h a n has previously been reported. Several types of neurons a n d occasional synaptic contacts between t h e m c a n be recognized. These findings are described below. 260 explants of about equal size. Cerebellar folla of 7- to ll-day-old mice were cut into explants 1 to 2 m m 3 in size. The study is based on examination of more than 200 cultures of newborn and about 60 cultures of older cerebellum. Explants were placed on collagen-coated coversllps, usually one explant per coverslip, and fed and maintained exactly according to the procedures of Bornstein and Murray (2) except for the trivial substitution of Tyrode's for Simms' balanced saline and the elevation of the final glucose concentration of the medium to 600 nag per cent (12). TH~ JOURNAL OF CEI~ BIOt~)Gv • VOLV~tE ~ , 1964 Downloaded from on June 16, 2017 FmURE 1 Transverse section of newborn mouse brain stem and cerebellum, stained with toluidinc blue. The cerebellum forms a simple bar of tissue overlying and extending along the sides of the brain stem. The newborn cerebellum has little of the complex architecture of the adult cerebellum. Note the relatively large ependymal surface (E). The roof nuclei (R) can be identified as clusters of darker staining cells embedded in the presumptive white matter. The cortex (C) has a conspicuous external granular layer. X 84. Published July 1, 1964 One group of 12 cultures was maintained with medium containing only 100 nag per cent of glucose, following Yonezawa's statement (20) that this low concentration, although detrimental to the general health of the cultures, favored the maintenance of granule ceils in vitro. Cultures were observed regularly by brightfield microscopy. Most cultures were examined in detail and then fixed after 2 to 3 weeks For silver impregnation, cultures were fixed in 10 per cent formalin in balanced saline at 4 ° for 1 week or more. Fixation for as long as 4 months still gave good results. Cultures were then immersed in 70 per cent ethanol at 4 ° for 1 week or more, a step which appeared to be absolutely essential. When this step was omitted or shortened, or when the 70 per cent ethanol was replaced by 50 per cent methanol: 50 Downloaded from on June 16, 2017 FmuRs 2 Higher magnification of the area labeled R in Fig. 1. The roof nuclear neuron can be distinguished from surrounding cells by its large oval nucleus, single large nucleolus, and cytoplasm containing dark staining Nissl substance. Some cells (arrows) have recognizable dendrites. X 870. FIGURE 3 Higher magnification of the area labeled C in Fig. 1, rotated about 90°. From the top downwards, note first the external granular layer with small, darkly stained, closely packed cells; then a relatively cell-free zone; then the layer of loosely arranged Purkinje cells (bracket). The Purkinje cells lack large nucleoli and Nissl substance and cannot really be distinguished from surrounding cells. Precursors of glial cells are undergoing mitosis in the Purkinje cell layer (arrow). X 370. in vitro, but one group was followed for 7 weeks in vitro. In some cases, before fixation, the small coverslip bearing the culture was removed froin the Maximow assembly, mounted in a sandwich chamber filled with balanced saline such as described by Hild and Tasaki (7), and the culture then examined by phase contrast microscopy with Zeiss optics. Myelin was readily examined in the living culture by brightfield microscopy with a long working distance oil immersion lens. A few cultures were fixed with formalin or acrolein and stained with sudan black B to confirm the identity of the myelin sheaths. per cent methyl cellosolve, the silver impregnation always failed. Cultures were washed in distilled water and stained as whole mounts by Holmes' method (8, also cited in 10) with the following minor adaptations. Solutions were placed in Columbia dishes and coverslips individually handled with plastic-coated forceps. Ten ml of impregnating solution, the full capacity of a Columbia dish, appeared adequate for two coverslip cultures. Gold-toning appeared to be improved by exposing the cultures to the gold chloride and oxalic acid solutions for approximately 5 minutes and 10 minutes, respectively, M. K. WoL~ Differentiation of Neuronal Types and Synapses 261 Published July 1, 1964 instead of the shorter times given by Holmes. Counterstaining was unnecessary and indeed detrimental: it increased the optical density of the already thick cultures and thereby concealed many of the most interesting structures. The Luxol Fast Blue stain for myelin, which has been successfully combined with the Holmes stain on paraffin and polyester wax sections (10, 19) failed after Holmes staining of whole mount cultures. Acrolein was tried in several vehicles as a fixative for whole mount cultures. Limited success with the Holmes stain was obtained after fixation in 10 per cent acrolein in M/10 phosphate buffer, p H 7.0, at 0 ° overnight, followed by the usual immersion in RESULTS Starting Material T h e cerebellum in the n e w b o r n mouse (Fig. 1) forms a simple b a r of tissue with only r u d i m e n t a r y fissures. Perforating blood vessels are few a n d u n developed. Presumably for this reason, the m e n i n ges can be peeled away from the living tissue with relative ease. No myelin was seen by light microscopy. T h e neurons of roof nuclei were readily identified deep in the presumptive white m a t t e r adjacent to the ethanol. Neurofibrils in acrolein-fixed specimens were more weakly stained, and glial cell cytoplasm more strongly stained, than after formalin fixation. The interpretation of this image was more difficult and ambiguous, and acrolein fixation was, therefore, little used. Limited, unsuccessful trials were also made of the silver methods of Bodian, Palmgren, R a m o n y Cajal, and of the Golgi-Cox and rapid Golgi procedures. These trials were suspended when Holmes' method yielded preparations worthy of intensive study. The histology of newborn mouse cerebellum was reviewed in brains which had been fixed in acrolein and embedded in polyester wax (19). Sections were stained with toluidine blue or by Holmes' or Palmgren's silver methods. Reference was also made to published accounts (9, 11). 262 e p e n d y m a l lining of the fourth ventricle (Fig. 2). These cells have oval, vesicular nuclei, 10 to 15 # in diameter, a n d large nucleoli. T h e i r Nissl substance is moderate in a m o u n t a n d dendrites are visible. T h e Purkinje neurons, w h i c h u n d e r w e n t their final mitosis together with the roof nuclear neurons on the 1 l t h to 13th embryonic days (11), h a v e migrated to the cortex b u t have not yet acquired their m a t u r e structure. T h e y form a loosely ordered layer several cells deep, have neither large nucleoli n o r Nissl substance, a n d are difficult to distinguish from other cells (Fig. 3). T h e small darkly staining cells of the external g r a n u l a r layer (Fig. 3) are the precursors of most other cerebellar neurons. These cells will multiply several times THE JOURNAL OF CELL BIOLOGy • VOLUME ~ , 1964 Downloaded from on June 16, 2017 lq~GURES 4 through 27 are photomicrographs of cultures of mouse cerebellum, shined as whole mounts by Holmes' method. FmuaE 4 Newborn mouse, 17 days in vitro. A relatively thin portion of the explant region, showing axons intertwining in all directions at numerous planes of focus. Note one axon making a U-turn (arrow). X 370. Published July 1, 1964 during the first 2 weeks after birth and their descendants will migrate inwards to become granule cell and probably basket cell neurons. Some granule cells have begun this migration even at birth, but are too few in number to form a conspicuous internal granular layer (11). In the cerebellum of the 1- to 2-week-old mouse, fissuration is more advanced. The meninges are firmly adherent and cannot be removed without cells, and the cortex will assume the adult threelayered architecture. Appearance of Living Cultures Cultures of newborn mouse cerebellum, in their general pattern of outgrowth, reorganization, and myelination, closely resembled the cultures of rat cerebellum described by Bornstein and M u r r a y (2) except for certain details. Myelination pro- Downloaded from on June 16, 2017 FIGURE 5 Newborn mouse, ~1 days in vitro. A portion of the outgrowth zone, showing an isolated axon which enters the field at bottom center, makes three successive U-turns (arrows), and leaves the field at bottom left. X 93. FIGURE 6 Newborn mouse, 14 days in vitro. Axons in the proximal outgrowth zone bearing clusters of pseudopod-like appendages. )< 460. macerating the tissue. Myelin is developing in the white matter, and axons are well stained by Holmes' or Palmgren's method. Purkinje cells now form a monolayer, and have large nucleoli and prominent Nissl substance. Multiplication of the cells of the external granular layer is at its height. M a n y granule cells have migrated inward, and form a distinct internal granular layer. The cerebellar cortex of the neonatal kitten, as described by Pomerat and Costero (16), resembles that of the 1to 2-week-old mouse. Later, the external granular layer will disappear by inward migration of all its ceeded more rapidly than in rat cultures. It began as early as the 9th day in vitro, and was underway by the 12th day in vitro in about 60 per cent of cultures. After 3 weeks in vitro, myelin was present in some amount in 80 to 90 per cent of cultures. No new myelin appeared subsequently, although individual axons and myelin sheaths might continue to increase in diameter. Emigration of macrophages and thinning of the explanted fragment were somewhat less thorough and less well synchronized than in rat cultures, and the macrophages themselves were more variable in size. M. K. Wota" Differentiation of Neuronal Types and Synapses 263 Published July 1, 1964 parasagittal plane produced scanty myelin or none at all. T h e opposite experiment, of culturing roof nuclei isolated from cortex, was impractical because of the small size of the roof nuclei, the absence of l a n d m a r k s in the living state, a n d the extreme softness a n d fragility of n e w b o r n mouse cerebellum. I n cultures from cerebellum of 7- to 11-day-old mice, the meninges produced a n epithelioid a n d fibrocytic outgrowth w h i c h tended to i n h i b i t or mask the outgrowth of glia a n d neurites. T h e explants persisted as dense clumps, seldom flattening out in the m a n n e r of n e w b o r n cerebellum. Myelin was obtained in only 5 to 10 per cent of cultures. T h e cultures m a i n t a i n e d with m e d i u m containing 100 m g per cent glucose instead of 600 m g per cent showed a large proportion of degenerating cells in b o t h explant a n d outgrowth. Several cultures degenerated completely by the 12th day in vitro. Only one culture in the group of 12 developed any myelin whatever. Appearance of Holmes Impregnations NEWBORN CULTURES: T h e nuclei of all cells were stained various gray or reddish hues, a n d a great n u m b e r of axons were stained j e t black. All the myelinated axons were well stained; in addition, large n u m b e r s of presumably u n m y e l i n a t e d axons were demonstrated. T h e y proceeded in all directions at r a n d o m t h r o u g h o u t the explant, forming so dense a network t h a t the individual axons could never be traced unambiguously for their full length (Fig. 4). Some axons extended FIGURE 7 Newborn mouse, 14 days in vitro. Survey view of entire explant. Note the ependymal formations in the lower right corner of the culture (arrow). X 87. FIGURE 8 Same culture, higher magnifcation of the region of the ependymal formation. Two large Type I neurons (I A) are seen below the ependyma. Above and to the left is a region of the explant zone (I C) which is somewhat less dense than the remainder. This region contains small Type I neurons which cannot be resolved at this magnification. X 93. FIGURE 9 Same culture. The lower of the two cells seen in Fig. 8 at I A. Note tim continuous network of neurofibrils through the soma and the four broad-based, radially arranged dendrites. The axon hillock is visible (Ax), but the axon is totally concealed by being out of focus. )< 1170. b~GUI~E 10 Same culture. A high-magnification view of cells in the region labeled I C in Fig. 8 Seven or more small Type I cells are seen at various focal planes. These ceils are about half the diameter of the large Type I cell shown at the same magnification in Fig. 9, but are similar in their staining characteristics. X 1170. 264 THE JOURNAL OF CELL BIOLOGY • VOLUME ~ , 1964 Downloaded from on June 16, 2017 Mouse e p e n d y m a often assumed the form of low c o l u m n a r sheets lining bizarre shaped cavities, within which debris m i g h t be whirled a b o u t u n ceasingly by the activity of the e p e n d y m a l cilia. T h e separate a n d distinct fluid-filled cystic formations associated with r a t e p e n d y m a (2) were not seen in mouse cultures. However, at the basal surface of sheets of e p e n d y m a l cells a layer of refractile vesicles of variable diameter, m a x i m u m 6 to 8 #, was regularly seen a n d was conspicuous e n o u g h to aid in locating the ependyma. T h e content of these vesicles was neither sudanophilic nor argyrophilic a n d the vesicles m i g h t be homologous w i t h the m u c h larger cysts associated with r a t ependyma. Bornstein a n d M u r r a y did not describe the fate of explants from different regions of r a t cerebellum. I n the present study, mouse cerebellum was divided into parasagittally oriented slices. T h e medial a n d intermediate slices, containing portions of roof nuclei as well as cortical anlagen, generally produced more n u m e r o u s a n d larger myelin sheaths t h a n the lateral slices, w h i c h m i g h t be deficient or lacking in roof nuclear material. A correlation seemed to exist between the presence of a b u n d a n t myelin a n d the presence of large ependymal formations, which were presumably derived from the roof of the fourth ventricle adjacent to the roof nuclei since the chorioid plexus h a d been carefully removed. Most cultures which failed to produce myelin also lacked ependyma. Fragments of n e w b o r n cerebellar cortex deliberately isolated from roof nuclei by cutting in planes other t h a n the Published July 1, 1964 Downloaded from on June 16, 2017 M. K. WOLF Differentiation of Neuronal Types and Synapses 265 Published July 1, 1964 several millimeters into the outgrowth zone, where their course was sometimes elaborately tortuous, with multiple changes of direction (Fig. 5). Occasionally, axons in the outgrowth zone bore clusters of pseudopod-like projections (Fig. 6). These projections were only found in a few of the healthiest well myelinated cultures and, therefore, it seems unlikely that they were degenerative in nature, although their significance is unknown. had been overlooked during the life of the culture. The staining of the somas and dendrites of neurons was highly variable. Analysis of more than 130 successfully impregnated cultures showed that neurons differing in size and density of stain were consistently located in different and characteristic sites in the culture. These properties permitted the definition of three principal types of neurons, with certain variants and intergrades (Table I). FIGURE 1~ Same culture. High magnification of the ependymal cells seen in Fig. 11, showing the well stained, individually separated ependymal cilia. )< 1170. The ependymal formations were readily recognized, both by their characteristic bizarre shapes (Figs. 7, 8, 11) and by the fact that the ependymal cilia themselves were lightly stained (Fig. 12). This property of the Holmes stain is of more than passing interest since, in a previous study, (6) the cilia did not stain with hematoxylin eosin, iron hematoxylin, azan, or by Jacobson's or Bodian's methods. The staining of ependymal cilia aided in the anatomical analysis of cultures, since the existence and location of ependymal formations could be unequivocally demonstrated even when they 266 Type I neurons (Figs. 8 to 11, 13) contained densely stained black neurofibrils in a continuous network throughout soma, dendrites, and axon. These cells were invariably located close to ependymal formations. Three varieties of Type I neurons were distinguished: (a) Large cells, with nucleus 10 to 12 # in diameter and soma ranging from 20 to 40 ~. The dendrites were broad-based, tapering, branched, somewhat radially arranged and extremely long--100 to 200 # or even longer. These cells were found in about 25 per cent of the cultures but were never more numerous than one to THE JOURNAL OF CELL BIOLOGY • VOLUME ~g, 1964 Downloaded from on June 16, 2017 FIGURE 11 Newborn mouse, 17 days in vitro. A group of large Type I neurons in a thicker region of a culture. Three neurons (arrows) are sufficiently in focus to distinguish soma and dendrites. Other neurons in this field are wholly or partly out of focus. At the right of the field an ependymal formation (E) is seen. X 460. Published July 1, 1964 TABLE I Types of Neurons in Cultures of Mouse Cerebellum Neuron Type I Type III Size A One to five per culture near e p e n d y m a in about 1~ of cultures Nucleus 10 to 12 /z. Soma 20 to 40/z in diameter Intensely stained neurofibrils continuous throughout radially arranged dendrites, soma, and axon Large neurons of roof nuclei B Same as I A in only a few cultures Same as I A Intensely stained oval adendritic soma continuous with axon Large neurons of roof nuclei in partially degenerate state C In small clusters near e p e n d y m a in some cultures About half the diameter of I A Like I A Small neurons of roof nuclei Singly, in small clusters or loosely arranged palisades round the perimeter of the explant. A b u n d a n t in all cultures Nucleus 8 to 12 /~. Soma 12 to 20 /z in diameter Soma flask shaped, dendrites apical, always less strongly stained than axon. Axon hillock and proximal 10 to 20 # of axon generally unstained. Often soma and dendrites unstained--appears as naked nucleus with neighboring axon Purkinje cells. Increasing somatodendritic stain in cultures with increasing age in vitro signifies continuing maturation. In dense p a r t of explant. Often close to palisades of Type II neurons. In about of cultures from newborn mice. Commoner in cultures from older mice. Soma 5 to 6 # in diameter Oval soma and two or three processes, jet black. One process often bifurcated 6 to 18 # from the soma I m m a t u r e granule cells (and perhaps basket cells) five per culture (Figs. 8, 9, 11, 13). (b) Adendritic cells with oval somas and eccentric nuclei also 10 to 12 # in diameter, found in a few cultures of both n e w b o r n and older cerebellum, but again only to the extent of one to five per culture (Fig. 13). (c) Cells about half the size of the first variety but similar in staining properties and conformation, found in certain explants which had initially been dissected so as to contain m u c h of the anlage of the dentate nucleus. These cells were distributed in a single c l u m p of one to two dozen (Fig. 10). O n e exceptional culture of this sort contained two of the large Type I cells and a clump of 40 to 50 small Type I cells. T y p e I cells were present in about half of all Interpretation cultures, in about two-thirds of the cultures in good health after 2 to 3 weeks in vitro, and in nearly all the a b u n d a n t l y myelinated cultures. In one culture, a few conspicuous myelinated axons observed during life were recognized after silver staining, and were traced back to somas of large Type I neurons. However, in the a b u n d a n t l y myelinated cultures the myelin sheaths seemed to be ten to twenty times more numerous t h a n the T y p e I neurons, although even an approximate count of myelin sheaths was impossible because they intertwined and b r a n c h e d so extensively. Type I I neurons (Figs. 14 to 19, 22 to 27) had somas and dendrites which were more weakly stained t h a n their axons, and very often r e m a i n e d M. K. WoL~ Differentiation of Neuronal Types and Synapses 267 Downloaded from on June 16, 2017 Type II Cell form and staining characteristics Numberand location Published July 1, 1964 Downloaded from on June 16, 2017 bXmvaE 18 Two-day-old mouse, 16 days in vitro. This field contains four large Type I A neurons and (at I B) a densely stained neuron with oval, adendritic soma. Short lengths of the axons of this neuron and of one other Type I neuron (Ax) are in focus. Note the great length of the tapering, branched dendrites. The dense part of the explant contains a narrow channel, not visible at this magnification, which is lined by ciliated ependymal cells. X 450. Published July 1, 1964 times in clusters of 6 or 7, a n d were well developed a n d readily e x a m i n e d in detail (Figs. 20, 21, 29). LOW GLUCOSE MEDIUM: Cultures of newb o r n cerebellum fed m e d i u m with glucose content reduced to 100 m g per cent were also lacking in myelin. T y p e I neurons were totally absent, a n d Type I I I neurons were present in only two of the cultures. Type I I neurons were numerous, h a d unstained somas a n d dendrites, b u t long, well developed, extensively b r a n c h e d axons w h i c h in some cases went far into the outgrowth zone. SYNAPSES : I n suitably flattened cultures, synaptic boutons terminaux a n d boutons en passant were occasionally found applied to the surfaces of Type II n e u r o n somas (Figs. 22 to 30). T h e largest boutons were 2 to 3/z in d i a m e t e r (Fig. 22) ; others were half this size or less (Fig. 24), a n d were visible only with the X100 oil immersion lens. Some boutons a p p e a r e d in the silver preparations as delicately outlined neurofibrillar rings with u n stained centers (Figs. 23, 26), while others showed i m p r e g n a t i o n of their g r o u n d cytoplasm (Figs. 22, 27). O n e bouton terminal was found on a dendrite of a Type I neuron. A t t e m p t s to trace a n axon from a b o u t o n to the p a r e n t soma were almost always unsuccessful, like all attempts to trace axons t h r o u g h the dense maze in the center of the culture. I n one culture from a 7-day-old mouse a terminal axon filament w h i c h coiled a r o u n d a T y p e I I n e u r o n a n d bore structures consistent w i t h boutons terminaux a n d boutons en passant was traced back to a Type I I I n e u r o n (Fig. 29). M a n y synapses, however, were found in cultures lacking T y p e I I I neurons. I n one culture of n e w b o r n cerebellum, a series of highly b r a n c h e d t e r m i n a l axon filaments surrounded the soma of a T y p e I I n e u r o n in a m a n n e r strongly resembling the basket ending which envelops the Purkinje cell in vivo. This structure was traced to a b r a n c h of the axon of the same n e u r o n - - a cell in a p p a r e n t synaptic contact with itself (Fig. 30). Phase Microscope Observations No n e u r o n soma of any type was ever found in the outgrowth zone of a culture. N u m e r o u s attempts were m a d e to correlate the a p p e a r a n c e of living cultures, m o u n t e d in sandwich c h a m b e r s a n d e x a m i n e d by phase microscopy, w i t h their a p p e a r a n c e after silver impregnation. T h e cells in the o u t g r o w t h w h i c h a p p e a r e d in phase contrast as possible " n e u r o n s " always r e m a i n e d totally u n stained, a n d could not even be located in the glial network. G e n u i n e neurons revealed by silver im- M. K. WOLF Differentiationof Neuronal Types and Synapses 269 Downloaded from on June 16, 2017 completely unstained. T h e nuclei of unstained Type I I neurons were recognizable by their diameter of 8 to 12 #, m u c h larger t h a n t h a t of a n y neuroglial nucleus, a n d by their single large nucleolus (Figs. 14, 15, 18, 23, 27). T h e soma, w h e n stained, a p p e a r e d more or less flask-shaped with apical dendrites (Figs. 19, 26). T h e dendrites tended to be directed toward the periphery of the explant. Even w h e n the soma was stained, the axon hillock often r e m a i n e d unstained, creating a n a p p a r e n t hiatus between soma a n d axon (Fig. 19). T h e proximal end of the stained axon m i g h t be a club-shaped mass located within 10 to 15 # of the nucleus (Fig. 14), or a fragile wisp first visible at a greater distance from the nucleus (Fig. 15). I n the latter case, it was often impossible to d e t e r m i n e w h e t h e r a given nucleus a n d axon belonged to the same n e u r o n unless t h a t n e u r o n was favorably isolated in a flat p a r t of the culture. A few T y p e II neurons were so isolated, singly or in small clusters, b u t most were a r r a n g e d in loose rows or palisades a r o u n d the perimeter of the explant (Fig. 18) or within the thick p a r t of the explant close to its m a r g i n (Figs. 16, 17). T h e staining of soma a n d dendrites was most often weak or absent in cultures fixed after 2 to 3 weeks in vitro; it was stronger in the one series of cultures m a i n t a i n e d for 7 weeks (Fig. 26). T h e Type I I n e u r o n was by far the most a b u n d a n t type, a n d was present in all cultures. Type III neurons (Figs. 16, 20, 21, 29) were the smallest a n d most densely stained. T h e i r oval somas, a b o u t 6 # in d i a m e t e r a n d stained solid black, gave rise to two or three processes of w h i c h the principal one generally bifurcated within 6 to 18 # of the soma. These cells were usually found in the dense p a r t of the explant somewhat closer to the center t h a n the palisades of Type I I neurons b u t often associated with t h e m (Fig. 16). Rarely, a T y p e I I I n e u r o n was found in the t h i n n e d - o u t m a r g i n of the explant, again associated w i t h one or more T y p e II cells (Fig. 29). Type I I I cells were few in n u m b e r a n d lacking altogether from more t h a n half the cultures of n e w b o r n cerebellum, including m a n y well myelinated cultures. OLDER (~ULTURES : Cultures from cerebell u m of 7- to 11-day-old mice were inferior to cultures from n e w b o r n mouse in most respects. Type I neurons w h e n present assumed the rounded, adendritic form with eccentric nuclei; T y p e I I neurons were few in n u m b e r a n d their axons were degenerate or absent. Myelin, as already noted, was rarely developed. Type I I I neurons, on the other h a n d , were present in every culture, some- Published July 1, 1964 Downloaded from on June 16, 2017 270 THE JotmNAL OF CELL BIOLOaY • VoLura~ ~o, 1964 Published July 1, 1964 pregnation were always in areas which may have appeared thin in the stained culture but were actually several cells thick and had been totally opaque to phase contrast. Phase contrast also gave no hint of the number, location, or course of axons in the outgrowth, since these structures during life were indistinguishable from the glial processes a m o n g which they meandered. T h e numerous, radially directed, gently undulating neuron processes which were so conspicuous a feature of the outgrowth zone of young cultures either stained diffusely and weakly or failed to stain at all. I n most cases they appeared to degenerate during the 3rd week in vitro. DISCUSSION Previous workers who have attempted to define neuronal types in cultures of cerebellum have con- centrated on study of the outgrowth zone; neurons remaining within the explant have either explicitly or tacitly been assumed to form a homogeneous population. T h e Holmes stain reveals no neurons in the outgrowth, but provides means for the recognition and classification of several types of neurons within the explant. These types may be interpreted as representing survival, with partial differentiation, of neurons already present in the tissue at the time of explantation. This interpretation is summarized in Table I. T h e Type I neurons are invariably located near ependymal formations, just as the roof nuclei in vivo are located near the ependymal lining of the fourth ventricle. Type I cells are, therefore, considered to be the large and small neurons of the roof nuclei. T h e Type II neurons are considered to be the Purkinie cells of the cortex. Their ar- Downloaded from on June 16, 2017 FmURE 14 Newborn cerebellum, 17 days in vitro. A single Type II neuron, well isolated and in focus at the margin of the explant. Although the soma is very weakly stained, there is a suggestion of a dendrite (at D). The proximal end of the axon appears as a club-shaped mass. At C, the axon gives rise to a recurrent collateral branch. X 460. FmURE 15 Same culture as in Fig. 14. A group of five Type II neurons, one out of focus. The single large nucleolus in each neuron can be better seen than in Fig. 14, in which the whole nucleus is darkened. The cytoplasm of soma and dendrites, however, is unstained. The axon of one of these cells (Ax) originates close to the nucleus as does that of the neuron in Fig. 14. Other axons originate at a greater distance from nuclei. As a result, there is even in this small group of neurons some ambiguity about the assignment of any given axon to the cell of origin. X 460. FIGURE 16 Same culture as in Fig. 14. A thicker region of the perimeter of the explant, showing a loose row or palisade (II, bracket) of Type II neurons. Their cytoplasm is poorly stained and only their solid black nuclei are evident. Below this palisade are four densely stained Type III neurons (III, arrows). X 460. I~GURE 17 The identical field shown in Fig. 16, photographed at a different focal plane to demonstrate the thickness of this area and the dense network of axons underlying the neuron somas. The palisade of Type II neuron nuclei is barely visible in positions corresponding to the black nuclei in proper focus in Fig. 16. X 460. FIGURE 18 Newborn cerebellum, 20 days in vitro. A palisade of Type II neurons which has come to lie in a flatter region of the explant than the one in Fig. 16. The cytoplasm of these cells is almost unstained. Note the many fine branches of axons running around and between somas in all directions. X 460. FIGURE 19 Newborn cerebellum, 17 days in vitro. This culture is stained more heavily than those illustrated in Figs. 14 to 18. Five Type II neurons with flask-shaped somas are seen, four in a row running diagonally from upper left to lower right, and one in the lower left corner. The dendrites are more strongly stained than the somas, and are directed toward the periphery of the explant (upper right) in 4 of the 5 cells. The main dendrite of the cell in the lower right comer points in the opposite direction. A small unstained hiatus (arrows) intervenes between the neuron soma and the stained part of the axon. X 980. M. K. WOLF Differentiation of Neuronal Types and Synapses 271 Published July 1, 1964 lqtIGURE ~1 Seven-day mouse, 1~ days in vitro. An array of Type I I I neurons. Some of the cells (arrows) have T-shaped bifurcations of their principal processes, while others have simpler bipolar shapes. X 460. 272 THE JOURNAL OF CELL BIOLOGY • VOLUME 2~, 1964 Downloaded from on June 16, 2017 FIGURE 20 Seven-day-old mouse, 1o. days in vitro. A Type III neuron which lies entirely within a single focal plane. Note the rounded, jet black soma, the same size as tim neuroglial nuclei among which it is embedded. Of tile three processes, the principal one undergoes a T-shaped bifurcation. Secondary branches are visible (arrows). X 4,60. Published July 1, 1964 r a n g e m e n t in loose, imperfectly ordered palisades resembles the a r r a n g e m e n t of i m m a t u r e Purkinje cells in n e w b o r n mouse. T h e i r large somas a n d nuclei, their tendency toward a flask shape with apically directed dendrites, a n d their axons, with argyrophobic initial segments a n d conspicuous recurrent collateral branches, are all typical of the m a t u r e Purkinje cell (17), a n d represent aspects of the m a t u r a t i o n of this cell proceeding in vitro. T h e tendency for the stainable neurofibrillar mate- rial in soma a n d dendrites to increase in a m o u n t from the 3rd week to the 7th week in vitro also suggests a gradual process of maturation. O n the other h a n d , cultured Purkinje cells never developed the two-dimensional b r a n c h i n g " a r b o r vitae" form of dendritic tree which they would have developed in situ. Possibly m a n y of the argyrophobic, radially oriented n e u r o n processes in the outgrowth of young cultures were Purkinje cell dendrites, which grew in the direction t h a t would have led to the developing molecular layer in situ, a n d then degenerated w h e n they failed to encounter oriented masses of developing granule cell Downloaded from on June 16, 2017 FmUR~ 22 Newborn cerebellum, 18 days in vitro. A group of weakly stained Type II neurons. An axon terminal enters the field from bottom right and terminates in a synaptic bouton at S I. A more intensely stained bouton is seen at S 2, though the axon leading to it is only partly in focus. X 1170. axons. Finally, Type I I I neurons are considered to represent granule cells (or, in some cases, basket cells) w h i c h have u n d e r g o n e their final mitosis a n d left the external g r a n u l a r layer, b u t are incompletely differentiated a n d r e m a i n so in vitro. T h e i r rarity in cultured n e w b o r n cerebellum a n d greater frequency in cultures from older mice support this FXGUP~ 23. Same culture as Fig. 22, same magnification. Additional synaptic boutons (S 3 and S 4). The axon terminal of S 3 is out of focus except for the last few micra leading up to the neurofibrillar ring. S 4 is seen edge-on and, therefore, appears darker and flatter than S 3. Its axon terminal is partially in focus and its course is indicated by the four unlabeled arrows. )< 1170. view. T h e forms they assume, especially in cultures from older mice (Figs. 20, 21), closely resemble R a m o n y Cajal's (17) drawings of Golgi preparations of the developing granule cell. Since T y p e I I I neurons were absent from the outgrowth zone, this interpretation tends to cast d o u b t on several previous accounts (13, 16, 20) of the migratory b e h a v i o r of cultured granule cells. F u r t h e r - M. K. WOLF Differentiation of Neuronal Types and Synapses 273 Published July 1, 1964 FICURE 25 Newborn cerebellum, another culture, 17 days in vitro. A synaptic bouton on the soma of a Type II neuron (arrow). >( 1170. more, the few granule cells seen in m y cultures could well have been present in the tissue at the time of explantation. Perhaps new granule cells do not form in the external g r a n u l a r layer u n d e r these culture conditions. It must be conceded t h a t neither the Holmes n o r a n y other silver stain c a n be taken as a self-sufficient diagnostic criterion of 274 T~E JOURNAL OF CELL BIOLOOY " VOLUME ~ , 1964 Downloaded from on June 16, 2017 FIGURE 24 Same culture, same magnification. Additional synaptic boutons. At S 5, two boutons originate from branches of a single axon terminal. S 6 is also a pair of boutons belonging to branches of the same axon. In both cases, additional boutons belonging to the same axons are out of focus and therefore invisible in the picture. X 1170. the neuron. If a n y small neurons stained as poorly as m a n y of the i m m a t u r e Purkinje cells, they would be unidentifiable, since their nuclei could not be distinguished from the thousands of glia] nuclei. T h e present study does not support, b u t does not rigorously disprove, previous accounts of the behavior of the granule cell in culture. Almost all the neurons w h i c h were favorably located for detailed study could be assigned to one of the three classes discussed above. A few cells resisted such classification, a n d it is possible t h a t some of the less n u m e r o u s types of cerebellar neuron, such as the Golgi Type II, were represented a m o n g these a b e r r a n t cells. It is also possible t h a t cells of the lateral vestibular nucleus m a y have been included in some explants. This nucleus, in the mouse, is located extremely close to the corpus cerebellum, a n d is difficult to separate from it with certainty since l a n d m a r k s are absent in the living tissues. These cells would h a v e been close to ependyma, like the roof nuclei, a n d m i g h t h a v e a p p e a r e d either as a b e r r a n t l y stained ceils or as regular Type I neurons. I t seemed advisable to limit the interpretive scheme to those neurons w h i c h a p p e a r e d regularly, in large numbers, a n d in evident good health. T h e Holmes stain has also provided w h a t appears to be the first light microscope demonstration of synapses in cultures of nervous tissues. Electrical recordings, suggesting the operation of synaptic networks in the living state, have been obtained by Crain a n d Peterson (4) from cultures of spinal cord, a n d by C r a i n (3) from cultures of cerebrum. Similar recordings (Crain, personal communication) were obtained only in restricted areas of a small proportion of cultures of cerebellum. This is consistent w i t h the rarity a n d irregular distribution of the synapses demonstrated in cultured cerebellum by the Holmes stain. M a n y interesting structures, such as synapses, are not only rare in cultures b u t are distributed in a nonr a n d o m m a n n e r which c a n n o t be anticipated by e x a m i n a t i o n of the living culture. Furthermore, the external landmarks w h i c h aid in the orientation of intact central nervous tissues are obliterated in cultures. Consequently, the study of these structures by electron microscopy is impeded by a formidable sampling problem. Ross, Bornstein, a n d Lehrer (18) reported neither synapses nor multiple types of neurons in their electron microscope study of cultured r a t a n d mouse cerebellum. Dr. O. Perier of Brussels has kindly c o m m u n i c a t e d u n - Published July 1, 1964 Downloaded from on June 16, 2017 FmtrRE 26 Newborn mouse, 49 days in vitro. A group of TypeII neurons, which aremore strongly stained in this culture than in most of the cultures fixed after shorter periods of time in vitro. Their dendrites are directed upward, toward the outgrowth zone. The palisade of Type II neurons is several ceils deep, but the cells in the deep part of the palisade also send dendrites toward the periphery (D). Two synaptic boutons (S i and S 2) are seen on the basal portions of the dendrites of two different neurons. An axosomatic bouton is also present in the field, but is out of focus (S 3). X 1660. M. K. WOLF Differentiation of Neuronal Types and Synapses 275 Published July 1, 1964 published electron micrographs, derived from three cultures of r a t a n d one of cat cerebellum, which show profiles consistent with axosomatic a n d axodendritic synapses. Similar profiles have been observed in cultures of spinal cord (Bunge a n d Bunge, cited in reference 4). I t is p e r t i n e n t to inquire w h i c h types of neurons m i g h t give rise to the myelinated axons a n d synaptic terminals found in vitro. T h e axons of some roof myelin content of a culture. Roof nuclear axons in situ form the superior cerebellar peduncle, w h i c h is one of the most heavily myelinated tracts in the rodent brain. A n o t h e r possibility is t h a t the roof nuclear neurons, in some m a n n e r w h i c h can only be speculated upon, favorably influenced the dev e l o p m e n t of the Purkinje cell axons a n d their sheaths in vitro. Miale a n d S i d m a n (11) have proposed t h a t the Purkinje cell from its earliest stages '~ 50/.z ® s'- .s2 FIGVnE 27 Newborn mouse, ~1 days in vitro. An unstained Type II neuron which sends its axon (Ef) towards top center of the field. This axon runs parallel to other, unrelated axons. An additional axon (Af) enters the field from top left, winds irregularly around the soma of the neuron, and gives rise to a synaptie bouton (S). X 460. FIOURE ~8 Drawing of the cell shown in Fig. 27. This and the two following figures were drawn with the aid of the Wild microscope drawing tube, and represent fairly accurate two-dimensional reconstructions of complicated synaptic structures. By this means, it is possible to show that the synaptic structure pictured in Fig. 27 actually bears three synaptic boutons (S 1, 2, 3), each in a different plane of focus and, therefore, impossible to record in one photograph. nuclear neurons were shown to be myelinated, b u t in the best cultures myelin was far too a b u n d a n t to be accounted for from this source alone, so t h a t a n u m b e r of Purkinje ceil axons must h a v e been myelinated as well. M y e l i n a t i o n was inferior, however, in those cultures containing only Purkinje cells a n d no roof nuclear cells. Possibly, the myelinated axons of roof nuclear cells were longer, thicker, more highly b r a n c h e d , a n d in general m a d e a more impressive c o n t r i b u t i o n to the total 275 of d e v e l o p m e n t m i g h t have a process w h i c h m a i n tains contact with a roof nuclear cell a n d w h i c h later becomes the definitive axon. If so, explants containing only Purkinje cells would h a v e all the i m m a t u r e axons cut, while explants containing b o t h roof nuclear a n d Purkinje cells m i g h t have m a n y of the Purkinje cell axons intact. T h e latter situation would presumably be more favorable for subsequent differentiation of Purkinje cells in vitro. However, in this case it m i g h t be expected t h a t ThE JOURNAL OF C]BLL BIOLOGY • VOLUME ~2, 1964 Downloaded from on June 16, 2017 iS5 Published July 1, 1964 synapses, if demonstrated at all in cultures, would be demonstrated regularly and in numbers on the surfaces of roof nuclear neurons, since the appropriate membrane contacts would already be present at the time of explantation and would require only to be maintained and differentiated. It is, therefore, somewhat disappointing to have found only one synapse on a roof nuclear neuron in all were traced to their cells of origin cannot be considered representative. It is conceivable that membrane contacts between Purkinje axons and roof nuclear neurons actually existed in cultures but failed to stain by Holmes' method. This matter might be worth re-investigating with the electron microscope if the sampling problem could be solved and the neuron types distinguished. Downloaded from on June 16, 2017 \ 50p. FmrYRE 29 Seven-day-old mouse, 12 days in vitro. Drawing of a Type III neuron with a T-shaped bifurcation of its principal process. A secondary process of this cell coils around a Type II neuron soma (dotted in) and bears structures interpreted as boutons en passant and boutons terminaux. the cultures examined, all other synapses being on Purkinje cells. Since synapses were found in cultures lacking granule cells, their cells of origin in those cultures must have been the roof nuclear or the Purkinje cells. In either case, the synapses demonstrated in culture do not correspond to the principal pathway of efferent impulses from those cells in situ. The two instances in which synapses The detailed anatomy of a culture of cerebellum is obviously simplified and randomized by comparison with the anatomy of intact cerebellum. Some investigators (7) have assumed that this simplification reduces central neurons in culture to a state of total isolation, in which their properties may be investigated and experimental results interpreted without any need to consider the M. K. WOLF Differentiation of Neuronal Types and Synapses 277 Published July 1, 1964 possibility of n e u r o n a l interactions. This assumption is u n w a r r a n t e d : the n e u r o n in vitro is embedded in a complex neuropil, the components of w h i c h must be analyzed in detail for each experim e n t to provide a basis for correct interpretation. This point requires special emphasis because of the faith w h i c h some workers a p p a r e n t l y have placed in phase microscopy as a self-sufficient m e t h o d of anatomical analysis. T h e present study has revealed serious limitations of phase microscopy, particularly the failure to detect axons w h i c h sub- sequently were revealed by silver i m p r e g n a t i o n a n d were presumably electrically excitable. Even in the thinnest a n d most t r a n s p a r e n t cultures, w h i c h contain the somas a n d dendrites of large neurons in phase-transparent areas, the phase images c a n n o t exclude the possible presence of axon terminals winding a r o u n d these neurons, perhaps synapsing u p o n them, a n d complicating the electrical recordings derived from them. Microscopic, cinematographic, a n d electrophysiological studies on living cultures of central nervous tissues • :2 -" ; ! 50/.L I FIGURE 30 Newborn mouse, 17 days in vitro. Drawing of a Type I I neuron. Soma and dendrites are dotted in, while the border of the nucleus and the axon with its branches are rendered as solid lines. The neuron has a flask-shaped soma and two apical dendrites. About 30 micra from its origin, the axon breaks up into a spray of six branches. A secondary collateral from one of these branches returns to the soma and gives rise to third- and fourth-order branches which surround the soma in a manner reminiscent of the basket ending around the Purkinje cell soma in vivo. This cell appears to be in synaptie eontaet with itself. 278 THE JOURNAL OF CELL BIOLOGY • VOLUME ~ , 1964 Downloaded from on June 16, 2017 y Published July 1, 1964 should, wherever technically possible, be complemented by study of the same cultures after the application of appropriate neurohistological methods. The author is grateful to Professor Margaret R. Murray, Dr. Murray B. Bornstein, and their col- leagues for generous advice concerning the maintenance of cultures. This work was supported by Special Fellowship No. 2 F l l NB 978-02 and Grant No. NB 04792-01 from the National Institute of Neurological Diseases and Blindness, United States Public Health Service. Received for publication, October 18, 1963. REFERENCES 13. 14. 15. 16. 17. 18. 19. 20. R. P., Some nutritional aspects of myelin sheath formation in cultures of central and peripheral nervous system, in Proceedings of the IVth International Congress of Neuropathology, Stuttgart, Georg Thieme Verlag, 1962, 2,267. NAKAI, J., and OrA~iOTO, M., Identification of neuroglial ceils in tissue culture, in Morphology of Neuroglia, (J. Nakai, editor), Japan, Igaku Shoin Ltd., 1963. PERIER, O., and DE HARVEN, E., Electron microscope observations on myelinated tissue cultures of mammalian cerebellum, in Proceedings of the Anatomical Society of Great Britain and Ireland, November 1961, Cytology of Nervous Tissue, London, Taylor & Francis Ltd., 1962. PIPER, R., Microtomy techniques and the problem of relation of cell outgrowth to events within explants in vitro, J. Med. Lab. Technol., 1962, 1. POMERAT, C. M., and Cosan~RO, I., Tissue cultures of cat cerebellum, Am. J. Anat., 1956, 99, 211. RAMON Y CAJAL, S., Histologie du systtme nerveux de l'homme et des vertebres, (Translated by Dr. L. Azoulay), Madrid, Instituto Ramon y Cajal, 1955. Ross, L. L., BORNSTEIN, M. B., and LEHRER, G. M., Electron microscopic observations of rat and mouse cerebellum in tissue culture, J. Cell Biol., 1962, 14, 19. SIDMAN, R. L., MOTTLA, P. A., and FEDER, N., Improved polyester wax embedding for histology, Stain Technol., 1961, 36, 279. YONEZAWA, T., BORNSTEIN, M. B., PETERSON, E. R., and MURRAY, M. R., A histochemical study of oxidative enzymes in myelinating cultures of central and peripheral nervous tissue, J. Neuropath. and Exp. Neurol., 1962, 21, 479. M. K. WoLF Differentiation of Neuronal Types and Synapses 279 Downloaded from on June 16, 2017 1. BORNSTEIN,M. B., and APPEL, S., The application of tissue culture to the study of experimental "allergic" encephalomyelitis. I. Patterns of demyelination, J. Neuropath. and Exp. Neurol., 1961, 20, 141. 2. BORNSTmN, M. B., and MURRAY, M. R., Serial observations on patterns of growth, myelin formation, maintenance and degeneration in cultures of newborn rat and kitten cerebellum, J. Biophysic. and Biochem. Cytol., 1958, 4, 499. 3. CRAIN, S. M., Bioelectric activity of mouse cerebral neocortex tissue during growth and differentiation in vitro, Excerpta Med., 1964, 18, Section 1, p. XI. 4. CRAm, S. M., and PETERSON, E., Bioelectric activity in long-term cultures of spinal cord tissues, Science, 1963, 141,427. 5. HILD, W., Myelinogenesis in cultures of mammalian central nervous tissue, Z. Zellforsch., 1957, 46, 71. 6. HILD, W., Ependymal cells in tissue culture, Z. Zellforsch., 1957, 46, 259. 7. HILl), W., and TASAKI, I., Morphological and physiological properties of neurons and glial cells in tissue culture, d. Neurophysiol., 1962, 25,277. 8. HOLMES, W., The peripheral nerve biopsy, in Recent Advances in Clinical Pathology, (S. C. Dyke, editor), Philadelphia, The Blakiston Co., 1st edition, 1947, 402. 9. KELTON, n. E,, and RAUCH, H., Myelination and myelin degeneration in the central nervous system of dilute-lethal mice, Exp. Neurol., 1962, 6,252. 10. MARGOLIS,G., and PICICETT,J. P., New applications of the Luxol Fast Blue myelin stain, Lab. Invest., 1956, 5, 459. 11. MIALE, I. L., and SIDMAN,R. L., An autoradiographic analysis of histogenesis in the mouse cerebellum, Exp. Neurol., 1961, 4, 277. 12. MURRAY, M. R., PETERSON, E. R., and BuNo~.,