- Email: [email protected]
Purkinje cell abnormalities and synaptogenesis in genetically jaundiced rats (Gunn rats)
Brain Research, 492 (1989) 116- 128
116
Elsevier BRE 14628
Purkinje cell abnormalities and synaptogenesis in genetically jaundiced rats (Gunn rats) Yoshiko Takagishi and Hideki Yamamura Department of Anatomy, Mie University School of Medicine, Mie (Japan) (Accepted 20 December 1988)
Key words: Gunn rat; Cerebellum; Synaptogenesis; Purkinje cell; Postnatal development
The course of cytological abnormalities and synaptogenesis of Purkinje cells were investigated in the culmen of cerebella from homozygous Gunn rats with hereditary hyperbilirubinemia from postnatal day 7 to adulthood (5-10 months old). The affected Purkinje cells were abundant at day 7. A large number of Purkinje cells reached the fully advanced stage of degeneration during the ensuing 16 days and disappeared between days 12 and 30. The Purkinje cells remaining at day 30 were less affected and recovered by the adult stage. Various abnormalities in Purkinje cell synaptogenesis with the parallel fibers, climbing fibers, and basket and stellate cell axons were observed. Primitive junctions between parallel fibers and Purkinje dendritic shafts were often found in adulthood. The parallel fiber boutons lacking postsynaptic partners and facing astrocytic processes were often noted from day 18 to adulthood. The persistence of such presynaptic elements suggests some mechanisms for stabilizing the synaptic elements once they have been formed. Many of the parallel fiber synaptic boutons with or without their postsynaptic partners were enlarged and were assumed to be transsynaptically affected by Purkinje cell damage. A number of climbing fiber synapses with perisomatic processes of Purkinje cells, which are transient in normal synaptogenesis, were present at day 30 and a few of them were still found even in adulthood. Basket and steltate cell synapses were often found in abundance on the remaining Purkinje cells in adulthood, though they were not frequently encountered during the developmental period.
INTRODUCTION D e v e l o p m e n t a l anomalies of the cerebellum have been extensively studied in m u t a n t mice 11A8"23'24'26 and animals with experimentally induced cerebellar m a l f o r m a t i o n s 2'6a5'26'3° and many kinds of synaptic d e r a n g e m e n t have been found. Thus, these studies have offered useful models for the analysis of synaptic organization in the developing brain. G u n n rats, a m u t a n t strain of Wistar rats 5, show a u t o s o m a l recessive h e r e d i t a r y hyperbilirubinemia 5" 8,28 T h e h o m o z y g o t e s Q)') develop jaundice soon after birth due to a deficiency of hepatic bilirubin: U D P - g l u c u r o n y l transferase 2°'z9 and exhibit the m a r k e d cerebellar hypoplasia caused by the hyperbilirubinemia 19'21. In the d e v e l o p m e n t of the cerebellar hypoplasia, Purkinje cells have been r e p o r t e d to be exclusively affected 19,21. The abnormalities of
Purkinje cells, which were first recognized at 30 a n d 72 h after birth in animals from jj × jj matings and those from h e t e r o z y g o t e 0 ' + ) x jj matings, respectively, were o b s e r v e d in detail by Schutta and Johnson 21 but the course of these abnormalities was not fully followed. F u r t h e r m o r e , to our knowledge, no observations have been m a d e on synaptogenesis in the jj rat cerebellum. In the present study, we first e x a m i n e d the course of the Purkinje cell abnormalities and then investigated the synaptogenesis of this n e u r o n in the culmen of jj rat cerebella. T h e culmen was chosen because Purkinje cells are severely affected in this lobule 19. O u r efforts focused on the synaptogenesis between parallel fibers and Purkinje cells because parallel fibers are the most n u m e r o u s afferents to the Purkinje cells. In addition, staggerer 12"25 as well as nervous 13'24"27 mutant mouse cerebella, in which
Correspondence: H. Yamamura, Department of Anatomy, Mie University School of Medicine, Tsu, Mie 514, Japan. 0006-8993/89/$03.50 (~ 1989 Elsevier Science Publishers B.V. (Biomedical Division~
117 Purkinje cells are affected, have presynaptic boutons of parallel fibers in the absence of their postsynaptic partners. Therefore, in the present study, special attention was paid to the differentiation and fate of parallel fiber presynaptic boutons in the jj cerebellum. MATERIALS AND METHODS Gunn rats bred in our laboratory were used. Homozygous (/7) and heterozygous (j+) Gunn rats were obtained from matings of a j + mother and a jj father. Heterozygous Gunn rats served as controls since they show neither hyperbilirubinemia nor cerebellar hypoplasia. Animals, aged 7, 12, 18, 23 and 30 days and 5-10 months, were examined. They were anesthetized with pentobarbital sodium and then perfused through the heart with 100-250 ml of a fixative solution containing 1% paraformaldehyde and 1.25% glutaraldehyde in 0.1 M phosphate buffer, pH 7.3, under pressure for 5-10 min immediately followed by 100-250 ml of another fixative solution containing 2% paraformaldehyde and 2.5% glutaraldehyde in the same buffer for 30-40 min. The fixative solutions were warmed to 38 °C just before use. After perfusion, the brain was excised. The cerebellum was isolated and weighed. Subsequently, it was immersed overnight in the first fixative solution at 4 °C. The culmen (lobules IV and V) were cut and postfixed with 2% OsO4 in 0.1 M phosphate buffer containing 4.5% glucose for 1 h, followed by dehydration in ethanol, and embedded in Epon. For light microscopic observation of the cerebellar cortex, five 1-#m-thick sagittal sections of the whole culmen per animal were prepared at 30-/zm intervals and stained with Toluidine blue. In these sections, the normal and affected Purkinje cells were counted. Thin sections were stained with uranyl acetate and lead citrate and then examined with a Hitachi H-700 electron microscope. The jj cerebellum was selected after weighing or histological observation. Criteria for the selection will be described in Results. In each age group, 3 cerebella in each genotype were subjected to light as well as electron microscope.
RESULTS There was considerable variation in the severity of cerebellar hypoplasia among jj rats at the same chronological age as reported by Katoh-Semba et al. 9. Therefore, at postnatal days 12 to 30 and the adult stage, ]] cerebella with about 1/5 to 2/5 mean weight of j + cerebella (99.3,164.3, 191.6, 198.3 and 295.5 mg at days 12, 18, 23, 30 and adult stage, respectively) were subjected to the examination. At day 7, however, the cerebellar weight of jj rats was not distinctly different from that of j+ rats, and hence the jj cerebellum in which over 80% of Purkinje cells in the five 1-#m-thick sagittal sections of the whole culmen (see Materials and Methods) contained osmiophilic granules, was subjected to further examination.
Course of Purkin]e cell abnormalities in j] cerebellum At day 7, 85% of Purkinje cell somata appeared abnormal, containing osmiophilic granules and small vacuoles. Some cells were lying along and/or buried in the internal granular layer, and consequently the Purkinje cell layer was disorganized (Fig. 1). However, there was no Purkinje cell loss in all the sections examined. By electron microscopy, membranous lamellar inclusions of various sizes were abundant in the perikaryal cytoplasm (Fig. 7). Mitochondria were often enlarged and included electron dense granules which were previously suggested to contain glycogen by a-amylase digestion procedures 22 (Fig. 8). At day 12, Purkinje cell loss was detected in every section from all three cerebella. Most of the remaining Purkinje cells were in a fully advanced stage of degeneration (Fig. 2). The soma of many Purkinje cells was swollen and filled with osmiophilic granules and the nucleus was often eccentric. More severely affected cells had vacuole-like soma. The molecular layer was rather thin, suggesting that the dendrites of Purkinje cells were poorly developed. There were scattered cells in the internal granular layer. At days 18 and 23, most of the remaining Purkinje cells were still severely affected (Fig. 3). The molecular layer and internal granular layer were strikingly thin. The external granular layer was uneven in thickness consisting of thin parts 1-2 cells deep and thick parts 3-5 cells deep.
118
Figs. 1-5. Light micrographs of a part of the cerebellar cortex from//" Gunn rats aged 7 days to 6 months. All 5 figures were prepared by photographing 1-/am-thick sagittal sections of the culmen, stained with Toluidine blue. EGL, external granular layer; ML, molecular layer; PL, Purkinje cell layer; IGL, internal granular layer. Bars = 20 pm. Fig. 1. Seven days of age. The external granular layer is thick, the molecular layer is thin, the Purkinje cell layer is disorganized, and in the internal granular layer, the cells are loosely packed. Almost all Purkinje cell somata contain osmiophilic granules and small vacuoles in the cytoplasm. Fig. 2. Twelve days of age. Most Purkinje cells are severely affected. Some Purkinje cell somata filled with osmiophilic granules are buried in the internal granular layer (arrows). Purkinje cells are absent in places from the Purkinje cell layer. Fig. 3. Eighteen days of age. Many Purkinje cells have been lost. One of the remaining Purkinje cells is vacuolated (arrow). The molecular and internal granular layers are quite thin and a paucity of cells is conspicuous. Fig. 4. Thirty days of age. Only one Purkinje cell with a small accumulation of osmiophilic granules in the periphery of the soma is present. Fig. 5. Six months of age. A remaining Purkinje cell in the atrophied cortex has obviously light cytoplasm in which no osmiophilic granules can be found. The cell body leans and obliquely gives off a proximal dendrite which appears rather stubby. The molecular and internal granular layers contain only a few cells.
119
1£
9
O.~
Postnatal days
(5-10 months old)
Fig. 6. Postnatal change in the ratio of Purkinje cell number in the culmen per section per cerebellum in jj rats to that in j+ rats.
At
day 30,
almost all the severely affected
Purkinje cells had disappeared and most of the remaining Purkinje cells appeared normal or mildly affected by light microscopy. The mildly damaged Purkinje cells contained a small a m o u n t of osmiophilic granules in the periphery of the cytoplasm
Fig. 7. An electron micrograph of a Purkinje cell from a jj Ounn rat aged 7 days. Membranous lamellar inclusions of various sizes and enlarged mitochondria (arrows) are present in the cytoplasm. Bar = 2/zm.
Fig. 8. A part of the Purkinje cell soma from a# Gunn rat aged 7 days. A mitochondrion (Mr) contains electron-dense granules. A climbing fiber (Cf) makes synaptic contacts with the perisomatic processes (Ps) and smooth surface of the soma. Bar = 1/zm.
Fig. 9. A Purkinje cell in the culmen from a jj Gunn rat aged 6 months. The cell organelles, especially free ribosomes and granular endoplasmic reticulum, are scarce in the periphery of the soma. Bar = 2/zm.
120 (Fig. 4). By electron microscopy, these less affected Purkinje cells were filled with organelles such as Golgi complex, granular endoplasmic reticulum, and mitochondria in which vesicles containing the abovementioned dense granules were often seen. These organelles were irregularly distributed. Ribosomes concentrated in the perinuclear zone. A cytoplasmic organization like this was observed in normal immature Purkinje cells at an early stage of development. Only a few Purkinje cells were found in thick sections of the culmen from ]j rats 5-10 months old (Fig. 5). Osmiophilic granules were seldom seen in these Purkinje cells. The cytoplasm was light especially in the periphery of the soma, where cell organelles, such as ribosomes and granular endoplasmic reticulum, were scarce (Fig. 9). The stubby dendritic trunk was often found in the strikingly thin molecular layer (Fig. 5).
Fig. 10. An apical part of a Purkinje cell from a jj Gunn rat aged 7 days. A slender Purkinje cell dendrite (Pcd) extends vertically from the apex of the soma. Parallel fibers are abundant around the dendrite. Some of them make synaptic contacts with the smooth surface or a process (P) of the dendrite (arrows). The synapses are immature with the inconspicuous synaptic membrane densities and a few synaptic vesicles. Bar = 1 pm.
The ratio of Purkinje cell number in the culmen in the jj cerebellum to that in the j + cerebellum was reduced by half between days 7 and 12, and continued to reduce gradually until day 30 (Fig. 6). There was little difference in the ratio between day 30 and the adult stage, suggesting the absence of Purkinje cell loss after day 30.
Development of parallel fiber-Purkinje cell synapses Development of Purkinje cell dendrites and synaptogenesis of Purkinje cells with parallel fibers in ] + cerebella were not basically different from those that have been described on the basis of observations of thin sections from normal rat a and mouse 1° cerebella. At day 7 postnatally, parallel fibers were closely packed in the molecular layer in both j + and jj cerebella. But in jj cerebella, the Purkinje cell dendrites, which came from the soma containing a large number of cytoplasmic m e m b r a n o u s b o d i e s , were rather slender (Fig. 10). The parallel f i b e r s were in synaptic contact with proximal dendritiC shafts or their processes (Fig. 10). The synapses
.......
A
Fig. l 1. Molecular layer of the cuimen from a jj Gtmn rat aged 12 days, A stubby Purkinje cell dendritic (Pcd) trunk faces parallel fibers without a glial sheath and projects spines (arrows). One of them is in synaptic contact with a parallel fiber (asterisk), whereas others lack their presynaptic partners in this section. Bar = 1 urn.
121 displayed a few synaptic vesicles and inconspicuous synaptic m e m b r a n e densities as those seen in j + cerebella but were found less frequently than in j + cerebella. At day 12, profiles of the Purkinje cell dendrite were not frequently encountered in the molecular layer of jj cerebella. This is due to a loss of many Purkinje cells (Fig. 6) and may be due to poor branching of the dendrites of the remaining cells. Moreover, dendritic trunks as well as branches appeared stubby (Fig. 11). They directly faced parallel fibers without a glial sheath from their proximal portions to the distal branches, whereas the proximal dendrites in j + cerebella were ensheathed by glial processes. Synapses with parallel fibers were found on the smooth surface of these dendrites and showed an immature appearance. Parallel fiberspine synapses, which were the only synapses between parallel fibers and Purkinje cells in adult j +
cerebella, were apparently seen but less frequent in jj cerebella than in j + cerebella. Spines emerged not only from the dendritic branches but also from the trunks (Fig. 11). The parallel fiber boutons contained aggregated synaptic vesicles and asymmetrical synaptic junctions were formed. In jj cerebella at days 18-30, the dendrites of the remaining Purkinje cells were still naked or covered partly by glial processes even in the rather thick portions (Fig. 12). The concentration of parallel fiber-spine synapses was unusually low. These synapses were found on the spines arising from rather thick dendrites as well as branchlet spines (Figs. 12 and 13). Parallel fiber boutons were often enlarged with a light matrix and contained a large number of synaptic vesicles (Fig. 14). The synaptic m e m b r a n e densities were asymmetrical and conspicuous (Fig. 14). The enlargement of parallel fiber synaptic boutons became prominent with age. Parallel fibers
Fig. 12. Molecular layer of the culmen from a jj Gunn rat aged 18 days. A thick Purkinje cell dendrite (Pcd) runs horizontally and gives off two branchlets up and down. The glial covering is incomplete (arrowheads), so most of the dendritic surface directly faces the parallel fibers. A few parallel fibers are in synaptic contact with spines arising from the thick dendrite (arrows 1 and 2) or a spine arising from one of the branchlets (arrow 3). A stellate cell synapse is seen on the thick dendrite (asterisk). Bar = 1/zm.
Fig. 13. Molecular layer of the culmen from a # Gunn rat aged 18 days. Three parallel fiber boutons are in synaptic contact with a thick Purkinje cell dendrite (Pcd). One of them (Pf 1) directly contacts the dendritic shaft. The synaptic membrane densities are prominent and a large number of synaptic vesicles are present. The second (Pf 2) is in contact with a spine which emerged from the dendrite. The third (Pf 3) is enlarged and is in contact with both the dendrite and a spine which presumably emerged from this dendrite. Bar = 0.5/zm.
122 were occasionally in synaptic contact with the thick dendritic shafts (Fig. 13). The parallel fiber boutons of these shaft junctions at this stage were larger than those at previous stages and contained aggregated synaptic vesicles. In addition, synaptic membrane densities became conspicuous. Thus, the parallel fiber presynaptic boutons seemed to mature on the dendritic shafts. Parallel fibers forming synaptic contact with the dendritic debris or atrophied spines of Purkinje cells were occasionally seen from days 18 to 23. Some parallel fiber synaptic boutons faced astrocytic processes in the absence of their postsynaptic partners (Fig. 15). The presynaptic membranes and synaptic vesicles appeared normal, although the latter varied in number. Electron-dense materials were usually present in the space between the presynaptic membrane of the parallel fiber and the
Fig. 14. Molecular layer of the culmen from ajj Gtmn rat aged 18 days. Several synapses between parallel fibers and Pu.rkinje cell dendritic spines (s) are seen and are surrounded by glial processes. They show a mature appearance with conspicuous synaptic membrane densities and increased number of synaptic vesicles. The synaptic boutons are unusually enlarged. A few profiles of enlarged parallel fiber boutons with synaptic vesicles but without synaptic membrane density indicate that these enlarged boutons were sectioned at a level other than that including the synaptic site. Bar = 0.5 gm.
glial process, but they were either not evident or less prominent in some cases. The parallel fiber boutons were usually as enlarged as those which were in synaptic contact with Purkinje cell spines. However, small parallel fiber synaptic boutons with a few synaptic vesicles in the absence of postsynaptic partners were present in the upper molecular layer from days 18 to 30. In cerebella from adult jj rats, parallel fiberPurkinje spine synapses were frequently seen only in the molecular layer above the remaining Purkinje cells (Fig. 16) but were seldom found in other parts of the molecular layer. The enlargement of parallel fiber boutons became prominent (Fig. 16). The parallel fiber synapses were also occasionally seen on the smooth surface of stubby dendrites, though some thick dendrites were often partially covered by glial processes except at the synaptic sites with climbing fibers or stetlate cell axons. These parallel fiber synaptic boutons appeared as mature as those seen in jj cerebella at day 18. Moreover, parallel fiber synaptic boutons without their postsynaptic partners
Fig. 15. Molecular layer of the culmen from ajj Gunn rat aged 30 days. Parallel fiber synaptic boutons (Pf) without their postsynaptic partners face glial processes (G1). They show the same features as those which are in contact with the postsynaptic partners. Bar = 0.5 pro.
123
Fig. 16. Molecular layer of the culmen from a # Gunn rat aged 8 months. Several parallel fiber-Purkinjc spine synapses (asterisks) are present. One of them is on the spine arising directly from the dendrite (Pcd) (arrow). Enlargement of parallel fiber boutons is prominent. Excessive glial processes (GI) segregate neuronal elements. Bar = 1/~m.
were encountered as frequently as at days 18-30. Their features were the same as those observed at days 18-30 (Fig. 17).
Development of climbing fiber-Purkinje cell synapses At day 7, climbing fiber synapses were found on the perisomatic processes of Purkinje cells in both jj (Fig. 8) and j + cerebella. The synapses displayed the same features in cerebella from both genotypes of rats. The perisomatic processes had a variety of shapes and were rather large in size. Climbing fiber synaptic boutons were also irregular in shape and contained a large number of synaptic vesicles, some of which had a dense core, and mitochondria. Some climbing fibers synapsed directly with the Purkinje cell soma. In j + cerebella, the climbing fiber synapses with
the perisomatic processes of Purkinje cells were most often observed at day 7, diminished in number thereafter and were never found after day 18. On the other hand, in jj cerebella they were most often found at day 12 and still frequently encountered from days 18 to 30 (Fig. 18). In ]+ cerebella, climbing fiber synapses with spines arising from Purkinje dendritic trunks began to be formed and increased in number beginning at day 12. In jj cerebella, climbing fiber/Purkinje dendritic spine synapses were rarely found at days 18-30 but were frequently found at the adult stage (Fig. 19). The synaptic boutons, in which round synaptic vesicles including several cored vesicles were densely packed, were unusually large. On the other hand, synapses with perisomatic processes, although small in number, were still found at the adult stage (Fig. 20). Therefore, the translocation of
124
Fig. 17. A parallel fiber bouton without its postsynaptic partner in the molecular layer of the culmen from a jj Gunn rat aged 6 months. Electron opaque material is present in the space between the synaptic membrane of a parallel fiber bouton and the plasma membrane of astrocytic process (arrow). Bar = 0.25/~m.
Fig. 19. Climbing fiber (Cf) and stellate celt (St) synapses on a Purkinje dendritic (Pcd) trunk in the culmen from a jj Gunn rat aged 6 months. The climbing fiber synaptic boutons contain densely packed round vesicles and several dense cored vesicles, and make conspicuous and asymmetrical junctions with Purkinje dendritic spines (s): The stellate cell axons make inconspicuous and symmetrical synaptic junctions with the smooth surface of the dendrite. However, their synaptic boutons are unusually large with a large number of round and ellipsoidal vesicles. Bar = 1 ~um.
the climbing fiber synaptic sites from the perisomatic processes to the spines of the dendritic trunks seemed to be delayed or arrested in jj cerebella.
Development of basket or stellate cell-Purkinje cell synapses In jj cerebella, cells in the molecular layer, most
Fig. 18. Climbing fiber (Cf) synapses on the soma of a Purkinje cell (Pc) in the culmen from a jj Gunn rat aged 18 days. Climbing fibers are in synaptic contact with both the perisomatic processes and the smooth surface of the soma. Bar = 1 /~m.
of which are basket or stellate cells, were scarce from day 12 onward. However, the basket cell synapses were found on the soma of Purkinje cells beginning at day 12, although they were not frequently encountered up to day 30 (Fig. 21). Most of the surface of the Purkinje cell soma, the main synaptic site of basket cell axons, was occupied by the perisomatic processes synapsing with the climbing fibers. At the adult stage, more basket cell axons terminated on the Purkinje cell soma which was mostly ensheathed by layers of glial processes except at the synaptic sites. Synaptic b o u t o n s of the basket cells appeared almost normal except that r o u n d and ellipsoidal vesicles were rather densely packed in occasionally enlarged synaptic boutons. Stellate c e l l - P u r k i n j e cell synapses were also rarely found from days 12 to 30 (Fig. 12). However,
125
Fig. 20. A climbing fiber (CO synapse with a perisomatic process of a Purkinje cell (Pc) in the culmen from a jj Gunn rat aged 6 months. The perisomatic process is small and has a spine-like shape. Bar = 0.5/zm.
Fig. 21. A basket cell (Bc) synapse on the soma of a Purkinje cell (Pc) in the culmen from a ]] Gunn rat aged 30 days. The synaptic bouton contains a few ellipsoidal and round synaptic vesicles and the synaptic membranes are inconspicuous. Bar = 1 ~m.
they were occasionally concentrated on the Purkinje cell dendritic trunk at the adult stage (Fig. 19). Stellate cell synaptic boutons were unusually large and were distinguished from the climbing fiber synaptic boutons by the presence of flattened vesicles and inconspicuous synaptic membranes. DISCUSSION
description by these authors. According to the present observation, the damage seems to increase rapidly in degree from postnatal days 7 to 23 in most Purkinje cells and severely damaged cells die and disappear by day 30. On the other hand, less affected cells may recover from the damage by day 30 and survive to the adult stage.
Course of Purkinje cell abnormalities
Development of parallel fiber-Purkinje cell synapses
The present study describes Purkinje cell abnormalities in the culmen of ]] Gunn rat cerebella. The Purkinje cell abnormalities found in the present observations were qualitatively the same as those reported by Schutta and Johnson 21. Despite their detailed observation on alterations in the Purkinje cells as compared with ours, modes of progression of the damage, especially in relation to phases in the maturation of this neuron and of the molecular layer of the cerebellum 1, are not necessarily clear from the
Homozygous (/7) Gunn rat cerebella showed various abnormalities in synaptic formation, such as persistent synaptic junctions between parallel fibers and the dendritic shaft of the Purkinje cells, the presence of presynaptic elements of parallel fibers in the absence of their postsynaptic partners, and an enlargement of parallel fiber synaptic boutons. Synaptic junctions between parallel fibers and Purkinje dendritic shafts were found in j + rats during the first four weeks in the present observa-
126 tion. These junctions were previously considered as transient L16 and to evolve to mature spine synapses 4 12 A recent investigation using a freeze-fracture technique suggests that these shaft junctions are not simply a transient intermediate in the formation of mature spine synaptic junctions, but rather represent a class of synaptic junction which has the capacity to dissociate in the remodeling process accompanying normal development 1°. In jj cerebella, the synaptic junctions between parallel fibers and Purkinje dendritic shafts were found at all stages examined. In these synaptic junctions, parallel fibers were immature with a few synaptic vesicles and inconspicuous synaptic membrane densities at a stage as early as day 7 but by day 18 some of them showed a mature appearance as seen in the mature spine synapses. The membrane density at synaptic sites on the Purkinje dendritic shaft was higher in later stages than in earlier stages. Thus, it seems to be likely that, in the jj cerebella, immature parallel fiber synaptic boutons mature on the outgrowing dendrite of Purkinje cells at early stages of development and that some of the parallel fiber-Purkinje dendritic shaft junctions do not or cannot dissociate in the later stage of development, but persist until the adult stage. The reason why the shaft junctions persist in the j] cerebellum is difficult to explain. They might have lost the capacity to dissociate due to the Purkinje cell abnormality. An alternative explanation is that the shaft junctions might supply a deficiency of mature spine synapses. Although there has been no direct evidence that the shaft junction can perform synaptic transmission in normal animals, an electrophysiological study in staggerer mutant mice aged 18-22 days demonstrated that parallel fibers were able to activate Purkinje cells 3. This suggests that the shaft junction fulfils its function as a synapse in this mutant animal, because in the staggerer mutant mouse cerebellum the only junctions between parallel fibers and Purkinje cells are on the dendritic shafts of these cells during the first 3 weeks after birth 12. In the pcd mutant mouse cerebellum in which virtually all Purkinje cells begin to degenerate at around postnatal days 15-18 (ref. 17), the appropriate afferent fibers have formed synapses on the Purkinje cells by day 15 (ref. 14). At day 18,
however, parallel fibers form synapses directly onto Purkinje dendritic shafts in an abnormal sequence of synaptogenesis, and these synapses have been thought to represent an aberration of parallel fiber synaptogenesis that is an early manifestation of the Purkinje cell damage in the pcd cerebellum 11. Thus, although the parallel fiber-Purkinje dendritic shaft junctions are found at unusually late stages in the pcd mutant mouse as well as jj Gunn rat cerebella, they are different in the mode and time of formation between both cerebella. Parallel fiber synaptic boutons devoid of their postsynaptic partners were often found in jj cerebella from day 18 through the adult stage. They were almost identical with presynaptic boutons of mature parallel fiber-Purkinje spine synapses and faced astrocytic processes. In the staggerer mutant mouse cerebellum in which dendritic branchlet spines are virtually not formed, similar parallel fiber synaptic boutons devoid of their postsynaptic partners have been observed 12"25. These synaptic boutons were found at postnatal day 26 (ref. 25) but were no longer seen at day 33 or 43 (ref. 12). However, in the nervous mutant mouse cerebellum, in which 90% of Purkinje cells selectively degenerate between days 23 and 50 (ref. 13), parallel fiber synaptic boutons devoid of their postsynaptic partners have been found as late as 6-14 months of age 27. Two hypotheses which are related to each other have been suggested concerning the development and fate of such presynaptic terminals in the absence of their postsynaptic partners 24'27. One is that parallel fibers may be able to produce autonomously, i.e., through an intrinsic process of differentiation, transient 'bouton en passant' with normal presynaptic organelles in the absence of their postsynaptic targets. The other is that presynaptic fibers in the cerebellum, which have been once stabilized by establishing function synapses, may be able to survive for long periods even if they lose their postsynaptic partners and those which have not made synaptic contacts may be unable to survive independently, going rapidly through the process of degeneration. The first hypothesis is based on the findings in the staggerer cerebellum 24"25, although a possibility that the presence of parallel fiber synaptic boutons devoid of their postsynaptic partners may be the
127 result of the detachment of parallel fibers from the Purkinje dendritic shaft after formation of the parallel fiber-Purkinje dendritic shaft junctions has been suggested 7'1°'12. The second hypothesis is based on the difference between the staggerer and nervous mutant mice in the fate of parallel fiber synaptic boutons in the absence of their postsynaptic partn e r s 24,27.
In the present observation of ]j cerebella, no marked difference in the concentration of parallel fiber synaptic boutons facing astrocytic processes was noted between postnatal day 30 and 5-10 months of age. This suggests that, in the j] cerebella, the parallel fiber synaptic boutons which were stabilized by making synaptic contacts with the Purkinje spines at the earlier stages lost their postsynaptic partners due to the progressive degeneration of Purkinje cells by day 30 and these boutons could survive to the adult stage without their postsynaptic partners, as in the nervous mutant mouse. This speculation is also supported by the fact that parallel fibers in contact with atrophied spines of Purkinje cells were found at days 18-23. However, the present study cannot exclude the possibility of the autonomous development of presynaptic specializations in the absence of their postsynaptic targets because the period of parallel fiber-Purkinje cell synaptogenesis overlapped almost completely with that of Purkinje cell degeneration in the jj cerebella. The unusual enlargement of parallel fiber synaptic boutons which were found in ]j cerebella after day 18 might represent a transsynaptic effect of the Purkinje cell degeneration. By electron microscopic observation of the molecular layer in j] cerebella, no abnormality was noted in non-synaptic segments of parallel fibers, suggesting that the possible transsynaptic effect on granule cells was limited to synaptic terminals of their axons.
Development of climbing fiber-Purkinje cell synapses Climbing fiber synapses were the most numerous synapses found in early stages of development of jj as well as j + cerebella. In the normal development of the rat cerebellum, climbing fiber synapses are first formed on the soma and/or perisomatic processes of Purkinje cells from postnatal day 5 and
then translocated to the spines of dendritic trunks from day 10 (ref. 1). In the present observation, the climbing fiber synaptogenesis in j + cerebella was similar to this. In ]] cerebella, however, the climbing fiber synapses with perisomatic processes still remained in appreciable numbers at day 30 and a few of them were even seen until the adult stage, whereas the synapses with Purkinje dendritic spines were rarely found until day 30. It has been suggested that the poor development of Purkinje cell dendrites deprives climbing fibers of the dendritic surface available for synapse formation and prevents the fibers from translocating themselves to the dendrites, so the climbing fiber synapses with perisomatic processes of Purkinje cells persist for an abnormally long time 11'12. The climbing fiber synaptogenesis in jj cerebella also seems to be delayed and/or arrested due to the Purkinje cell damage. However, the synaptogenesis might proceed when Purkinje cells have recovered from the damage, since climbing fiber synapses with dendritic spines were prominent yet only a few synapses with perisomatic processes were seen in less affected Purkinje cells present in adult j] cerebella.
Development of basket or stellate cell-Purkinje cell synapses The Purkinje cells in ]j cerebella received the synaptic inputs from both basket and stellate cells in accordance with the normal developmental course. However, the basket cell synapses were scarce until day 30. In the normally developing rat cerebellum, most of the temporary climbing fiber synapses on the soma and/or perisomatic processes of the Purkinje cell disappeared before basket cell axon terminals began to establish synapses on the Purkinje cell soma 1. In ]] cerebella, the delay or arrest of translocation of the climbing fiber synaptic sites from the Purkinje cell soma to the spines of the dendritic trunks and of disappearance of the perisomatic processes may prevent basket cell axons from making synaptic contact with the Purkinje cell soma. Stellate cell synapses were also poorly detected through the developmental period. This seems due to the fact that dendrites of the remaining Purkinje cells were underdeveloped and, in addition, that stellate cells were scarce. The synaptogenesis of
128
b a s k e t and stellate cells m i g h t p r o c e e d w h e n the
ACKNOWLEDGEMENT
P u r k i n j e cells h a v e r e c o v e r e d as m i g h t be the case of c l i m b i n g fiber s y n a p t o g e n e s i s .
T h e a u t h o r s t h a n k D r . C. S o t e l o for critically r e a d i n g this m a n u s c r i p t .
REFERENCES 1 Altman, J., Postnatal development of the cerebellar cortex in the rat. II. Phases in the maturation of Purkinje cells and of the molecular layer, J. Comp. Neurol., 145 (1972) 399-464. 2 Altman, J. and Anderson, W.J., Experimental reorganization of the cerebellar cortex. I. Morphological effects of elimination of all microneurons with prolonged X-irradiation started at birth, J. Comp. Neurol., 146 (1972) 355-406. 3 Crepel, E and Mariani, J., Anatomical, physiological and biochemical studies of the cerebellum from mutant mice. I. Electrophysiological analysis of cerebellar cortical neurons in the staggerer mouse, Brain Research, 98 (1975) 135-147. 4 Foelix, R.E and Oppenheim, R., The development of synapses in the cerebellar cortex of the chick embryo, J. Neurocytol., 3 (1974) 277-294. 5 Gunn, C.K., Hereditary acholuric jaundice in a new mutant strain of rats, J. Hered., 29 (1938) 137-139. 6 Herndon, R.M., Margolis, G. and Kilham, L., The synaptic organization of the malformed cerebellum induced by perinatal infection with the feline panleukopenia virus (PLV). II. The Purkinje cell and its afferents, J. Neuropathol. Exp. Neurol., 30 (1971) 557-570. 7 Hirokawa, N., A study of the synaptogenesis in the cerebellar cortex through chronic treatment and immunocytochemistry of fl-bungarotoxin, J. Comp. Neurol., 185 (1979) 107-120. 8 Johnson, L., Sarmiento, E, Blanc, W.A. and Day, R., Kernicterus in rats with an inherited deficiency of glucuronyl transferase, Am. J. Dis. Child., 97 (1959) 591-608. 9 Katoh-Semba, R., Aoki, E. and Kashiwamata, S., Cerebellar hypoplasia in Gunn rats: effects of bilirubin on postnatal changes in cerebellar constituents, Biomed. Res., 3 (1982) 319-329. 10 Landis, D.M.D., Initial junctions between developing parallel fibers and Purkinje cells are different from mature synaptic junctions, J. Comp. Neurol., 260 (1987) 513-525. 11 Landis, D.M.D. and Landis, S.C., Several mutations in mice that affect the cerebellum, Adv. Neurol., 21 (1978) 85-105. 12 Landis, D.M.D. and Sidman, R.L., Electron microscopic analysis of postnatal histogenesis in the cerebellar cortex of staggerer mutant mice, J. Comp. Neurol., 179 (1978) 831-864. 13 Landis, S.C., Ultrastructural changes in the mitochondria of cerebellar Purkinje cells of nervous mutant mice, J. Cell Biol., 57 (1973) 782-797. 14 Landis, S.C. and Mullen, R.J., The development and degeneration of Purkinje cells in pcd mutant mice, J. Comp. Neurol., 177 (1978) 125-144. 15 Llimls, R., Hillman, D.E. and Preeht, W., Neuronal circuit reorganization in mammalian agranular cerebellar cortex, J. Neurobiol., 4 (1973) 69-94. 16 Mugnaini, E., Ultrastructural studies on the cerebellar histogenesis. IL Maturation of nerve cell populations and
17
18
19
20
21
22
23
establishment of synaptic connections in the cerebellar cortex of the chick. In R. Llin~is (Ed.), Neurobiology of Cerebellar Evolution and Development, American Medical Association, Chicago, 1969, pp. 749-782. Mullen, R.J., Eicher, E.M. and Sidman, R.L., Purkinje cell degeneration, a new neurological mutation in the mouse, Proc. Natl. Acad. Sci. U.S.A., 73 (1976) 208-212. Rakic, P., Synaptic specificity in the cerebellar cortex: study of anomalous circuits induced by single gene mutations in mice. In The Synapse, Cold Spring Harbor Symposia on Quantitative Biology, Vol. 40, 1976, pp. 333-346. Sawasaki, Y., Yamada, N. and Nakajima, H., Developmental features of cerebeUar hypoplasia and brain bilirubin levels in a mutant (Gunn) rat with hereditary hyperbilirubinaemia, J. Neurochem., 27 (1976) 577-583. Schmid, R., Axelrod, J., Hammaker, L. and Swarm, R.L., Congenital jaundice in rats, due to a defect in glucuronide formation, J. Clin. Invest., 37 (1958) 1123-1130. Schutta, H.S. and Johnson, L., Bilirubin encephalopathy in the Gunn rat: a fine structure study of the cerebellar cortex, J. Neuropathol. Exp. Neurol., 26 (1967) 377-396. Schutta, H.S., Johnson, L. and Neville, H.E., Mitochondrial abnormalities in bilirubin encephalopathy, J. Neuropathol. Exp. Neurol., 29 (1970) 296-305. Sidman, R.L., Cell interactions in developing mammalian central nervous system. In L.G. Silvestri (Ed.), Cell
Interactions, Proceedings of the Third Lepetit Colloquium, North-Holland, Amsterdam, 1972, pp. 1-13. 24 Sotelo, C., Electrical and chemical communication in the central nervous system. In B.R. Brinkley and K.R. Porter (Eds.), International Cell Biology, Rockefeller Univ. Press, New York, 1977, pp. 83-92. 25 Sotelo, C. and Changeux, J.E, Transsynaptic degeneration 'en cascade' in the cerebellar cortex of staggerer mutant mice, Brain Research, 67 (1974) 519-526. 26 Sotelo, C. and Privat, A., Synaptic remodeling of the cerebellar circuitry in mutant mice and experimental cerebellar malformations: study 'in vivo' and 'in vitro', Acta Neuropathol., 43 (1978) t9-34. 27 Sotelo, C. and Triller, A., Fate of presynaptic afferents to Purkinje cells in the adult nervous mutant mouse: a model to study presynaptic stabilization, Brain Research, 175 (1979) 11-36. 28 Tanase, H., Sudo, S. and Suzuki, Y., Maintenance and propagation of jaundiced rat, Exp. Anita., 17 (1968) 23-25. 29 Weatherill, P.J., Kennedy, S.M.E. and Burchell, B., Immunochemical comparison of UDP-glucuronyltransferase from Gunn- and Wistar-rat livers, Biochem. J., 191 (1980) 155-163. 30 Yamano, T., Shimada, M., Nakao, K., Nakamura, T., Wakaizumi, S. and Kusunoki, T., Maturation of Purkinje cells in mouse cerebellum after neonatal administrations of cytosine arabinoside, Acta Neuropathol., 44 (1978) 41-45.
116
Elsevier BRE 14628
Purkinje cell abnormalities and synaptogenesis in genetically jaundiced rats (Gunn rats) Yoshiko Takagishi and Hideki Yamamura Department of Anatomy, Mie University School of Medicine, Mie (Japan) (Accepted 20 December 1988)
Key words: Gunn rat; Cerebellum; Synaptogenesis; Purkinje cell; Postnatal development
The course of cytological abnormalities and synaptogenesis of Purkinje cells were investigated in the culmen of cerebella from homozygous Gunn rats with hereditary hyperbilirubinemia from postnatal day 7 to adulthood (5-10 months old). The affected Purkinje cells were abundant at day 7. A large number of Purkinje cells reached the fully advanced stage of degeneration during the ensuing 16 days and disappeared between days 12 and 30. The Purkinje cells remaining at day 30 were less affected and recovered by the adult stage. Various abnormalities in Purkinje cell synaptogenesis with the parallel fibers, climbing fibers, and basket and stellate cell axons were observed. Primitive junctions between parallel fibers and Purkinje dendritic shafts were often found in adulthood. The parallel fiber boutons lacking postsynaptic partners and facing astrocytic processes were often noted from day 18 to adulthood. The persistence of such presynaptic elements suggests some mechanisms for stabilizing the synaptic elements once they have been formed. Many of the parallel fiber synaptic boutons with or without their postsynaptic partners were enlarged and were assumed to be transsynaptically affected by Purkinje cell damage. A number of climbing fiber synapses with perisomatic processes of Purkinje cells, which are transient in normal synaptogenesis, were present at day 30 and a few of them were still found even in adulthood. Basket and steltate cell synapses were often found in abundance on the remaining Purkinje cells in adulthood, though they were not frequently encountered during the developmental period.
INTRODUCTION D e v e l o p m e n t a l anomalies of the cerebellum have been extensively studied in m u t a n t mice 11A8"23'24'26 and animals with experimentally induced cerebellar m a l f o r m a t i o n s 2'6a5'26'3° and many kinds of synaptic d e r a n g e m e n t have been found. Thus, these studies have offered useful models for the analysis of synaptic organization in the developing brain. G u n n rats, a m u t a n t strain of Wistar rats 5, show a u t o s o m a l recessive h e r e d i t a r y hyperbilirubinemia 5" 8,28 T h e h o m o z y g o t e s Q)') develop jaundice soon after birth due to a deficiency of hepatic bilirubin: U D P - g l u c u r o n y l transferase 2°'z9 and exhibit the m a r k e d cerebellar hypoplasia caused by the hyperbilirubinemia 19'21. In the d e v e l o p m e n t of the cerebellar hypoplasia, Purkinje cells have been r e p o r t e d to be exclusively affected 19,21. The abnormalities of
Purkinje cells, which were first recognized at 30 a n d 72 h after birth in animals from jj × jj matings and those from h e t e r o z y g o t e 0 ' + ) x jj matings, respectively, were o b s e r v e d in detail by Schutta and Johnson 21 but the course of these abnormalities was not fully followed. F u r t h e r m o r e , to our knowledge, no observations have been m a d e on synaptogenesis in the jj rat cerebellum. In the present study, we first e x a m i n e d the course of the Purkinje cell abnormalities and then investigated the synaptogenesis of this n e u r o n in the culmen of jj rat cerebella. T h e culmen was chosen because Purkinje cells are severely affected in this lobule 19. O u r efforts focused on the synaptogenesis between parallel fibers and Purkinje cells because parallel fibers are the most n u m e r o u s afferents to the Purkinje cells. In addition, staggerer 12"25 as well as nervous 13'24"27 mutant mouse cerebella, in which
Correspondence: H. Yamamura, Department of Anatomy, Mie University School of Medicine, Tsu, Mie 514, Japan. 0006-8993/89/$03.50 (~ 1989 Elsevier Science Publishers B.V. (Biomedical Division~
117 Purkinje cells are affected, have presynaptic boutons of parallel fibers in the absence of their postsynaptic partners. Therefore, in the present study, special attention was paid to the differentiation and fate of parallel fiber presynaptic boutons in the jj cerebellum. MATERIALS AND METHODS Gunn rats bred in our laboratory were used. Homozygous (/7) and heterozygous (j+) Gunn rats were obtained from matings of a j + mother and a jj father. Heterozygous Gunn rats served as controls since they show neither hyperbilirubinemia nor cerebellar hypoplasia. Animals, aged 7, 12, 18, 23 and 30 days and 5-10 months, were examined. They were anesthetized with pentobarbital sodium and then perfused through the heart with 100-250 ml of a fixative solution containing 1% paraformaldehyde and 1.25% glutaraldehyde in 0.1 M phosphate buffer, pH 7.3, under pressure for 5-10 min immediately followed by 100-250 ml of another fixative solution containing 2% paraformaldehyde and 2.5% glutaraldehyde in the same buffer for 30-40 min. The fixative solutions were warmed to 38 °C just before use. After perfusion, the brain was excised. The cerebellum was isolated and weighed. Subsequently, it was immersed overnight in the first fixative solution at 4 °C. The culmen (lobules IV and V) were cut and postfixed with 2% OsO4 in 0.1 M phosphate buffer containing 4.5% glucose for 1 h, followed by dehydration in ethanol, and embedded in Epon. For light microscopic observation of the cerebellar cortex, five 1-#m-thick sagittal sections of the whole culmen per animal were prepared at 30-/zm intervals and stained with Toluidine blue. In these sections, the normal and affected Purkinje cells were counted. Thin sections were stained with uranyl acetate and lead citrate and then examined with a Hitachi H-700 electron microscope. The jj cerebellum was selected after weighing or histological observation. Criteria for the selection will be described in Results. In each age group, 3 cerebella in each genotype were subjected to light as well as electron microscope.
RESULTS There was considerable variation in the severity of cerebellar hypoplasia among jj rats at the same chronological age as reported by Katoh-Semba et al. 9. Therefore, at postnatal days 12 to 30 and the adult stage, ]] cerebella with about 1/5 to 2/5 mean weight of j + cerebella (99.3,164.3, 191.6, 198.3 and 295.5 mg at days 12, 18, 23, 30 and adult stage, respectively) were subjected to the examination. At day 7, however, the cerebellar weight of jj rats was not distinctly different from that of j+ rats, and hence the jj cerebellum in which over 80% of Purkinje cells in the five 1-#m-thick sagittal sections of the whole culmen (see Materials and Methods) contained osmiophilic granules, was subjected to further examination.
Course of Purkin]e cell abnormalities in j] cerebellum At day 7, 85% of Purkinje cell somata appeared abnormal, containing osmiophilic granules and small vacuoles. Some cells were lying along and/or buried in the internal granular layer, and consequently the Purkinje cell layer was disorganized (Fig. 1). However, there was no Purkinje cell loss in all the sections examined. By electron microscopy, membranous lamellar inclusions of various sizes were abundant in the perikaryal cytoplasm (Fig. 7). Mitochondria were often enlarged and included electron dense granules which were previously suggested to contain glycogen by a-amylase digestion procedures 22 (Fig. 8). At day 12, Purkinje cell loss was detected in every section from all three cerebella. Most of the remaining Purkinje cells were in a fully advanced stage of degeneration (Fig. 2). The soma of many Purkinje cells was swollen and filled with osmiophilic granules and the nucleus was often eccentric. More severely affected cells had vacuole-like soma. The molecular layer was rather thin, suggesting that the dendrites of Purkinje cells were poorly developed. There were scattered cells in the internal granular layer. At days 18 and 23, most of the remaining Purkinje cells were still severely affected (Fig. 3). The molecular layer and internal granular layer were strikingly thin. The external granular layer was uneven in thickness consisting of thin parts 1-2 cells deep and thick parts 3-5 cells deep.
118
Figs. 1-5. Light micrographs of a part of the cerebellar cortex from//" Gunn rats aged 7 days to 6 months. All 5 figures were prepared by photographing 1-/am-thick sagittal sections of the culmen, stained with Toluidine blue. EGL, external granular layer; ML, molecular layer; PL, Purkinje cell layer; IGL, internal granular layer. Bars = 20 pm. Fig. 1. Seven days of age. The external granular layer is thick, the molecular layer is thin, the Purkinje cell layer is disorganized, and in the internal granular layer, the cells are loosely packed. Almost all Purkinje cell somata contain osmiophilic granules and small vacuoles in the cytoplasm. Fig. 2. Twelve days of age. Most Purkinje cells are severely affected. Some Purkinje cell somata filled with osmiophilic granules are buried in the internal granular layer (arrows). Purkinje cells are absent in places from the Purkinje cell layer. Fig. 3. Eighteen days of age. Many Purkinje cells have been lost. One of the remaining Purkinje cells is vacuolated (arrow). The molecular and internal granular layers are quite thin and a paucity of cells is conspicuous. Fig. 4. Thirty days of age. Only one Purkinje cell with a small accumulation of osmiophilic granules in the periphery of the soma is present. Fig. 5. Six months of age. A remaining Purkinje cell in the atrophied cortex has obviously light cytoplasm in which no osmiophilic granules can be found. The cell body leans and obliquely gives off a proximal dendrite which appears rather stubby. The molecular and internal granular layers contain only a few cells.
119
1£
9
O.~
Postnatal days
(5-10 months old)
Fig. 6. Postnatal change in the ratio of Purkinje cell number in the culmen per section per cerebellum in jj rats to that in j+ rats.
At
day 30,
almost all the severely affected
Purkinje cells had disappeared and most of the remaining Purkinje cells appeared normal or mildly affected by light microscopy. The mildly damaged Purkinje cells contained a small a m o u n t of osmiophilic granules in the periphery of the cytoplasm
Fig. 7. An electron micrograph of a Purkinje cell from a jj Ounn rat aged 7 days. Membranous lamellar inclusions of various sizes and enlarged mitochondria (arrows) are present in the cytoplasm. Bar = 2/zm.
Fig. 8. A part of the Purkinje cell soma from a# Gunn rat aged 7 days. A mitochondrion (Mr) contains electron-dense granules. A climbing fiber (Cf) makes synaptic contacts with the perisomatic processes (Ps) and smooth surface of the soma. Bar = 1/zm.
Fig. 9. A Purkinje cell in the culmen from a jj Gunn rat aged 6 months. The cell organelles, especially free ribosomes and granular endoplasmic reticulum, are scarce in the periphery of the soma. Bar = 2/zm.
120 (Fig. 4). By electron microscopy, these less affected Purkinje cells were filled with organelles such as Golgi complex, granular endoplasmic reticulum, and mitochondria in which vesicles containing the abovementioned dense granules were often seen. These organelles were irregularly distributed. Ribosomes concentrated in the perinuclear zone. A cytoplasmic organization like this was observed in normal immature Purkinje cells at an early stage of development. Only a few Purkinje cells were found in thick sections of the culmen from ]j rats 5-10 months old (Fig. 5). Osmiophilic granules were seldom seen in these Purkinje cells. The cytoplasm was light especially in the periphery of the soma, where cell organelles, such as ribosomes and granular endoplasmic reticulum, were scarce (Fig. 9). The stubby dendritic trunk was often found in the strikingly thin molecular layer (Fig. 5).
Fig. 10. An apical part of a Purkinje cell from a jj Gunn rat aged 7 days. A slender Purkinje cell dendrite (Pcd) extends vertically from the apex of the soma. Parallel fibers are abundant around the dendrite. Some of them make synaptic contacts with the smooth surface or a process (P) of the dendrite (arrows). The synapses are immature with the inconspicuous synaptic membrane densities and a few synaptic vesicles. Bar = 1 pm.
The ratio of Purkinje cell number in the culmen in the jj cerebellum to that in the j + cerebellum was reduced by half between days 7 and 12, and continued to reduce gradually until day 30 (Fig. 6). There was little difference in the ratio between day 30 and the adult stage, suggesting the absence of Purkinje cell loss after day 30.
Development of parallel fiber-Purkinje cell synapses Development of Purkinje cell dendrites and synaptogenesis of Purkinje cells with parallel fibers in ] + cerebella were not basically different from those that have been described on the basis of observations of thin sections from normal rat a and mouse 1° cerebella. At day 7 postnatally, parallel fibers were closely packed in the molecular layer in both j + and jj cerebella. But in jj cerebella, the Purkinje cell dendrites, which came from the soma containing a large number of cytoplasmic m e m b r a n o u s b o d i e s , were rather slender (Fig. 10). The parallel f i b e r s were in synaptic contact with proximal dendritiC shafts or their processes (Fig. 10). The synapses
.......
A
Fig. l 1. Molecular layer of the cuimen from a jj Gtmn rat aged 12 days, A stubby Purkinje cell dendritic (Pcd) trunk faces parallel fibers without a glial sheath and projects spines (arrows). One of them is in synaptic contact with a parallel fiber (asterisk), whereas others lack their presynaptic partners in this section. Bar = 1 urn.
121 displayed a few synaptic vesicles and inconspicuous synaptic m e m b r a n e densities as those seen in j + cerebella but were found less frequently than in j + cerebella. At day 12, profiles of the Purkinje cell dendrite were not frequently encountered in the molecular layer of jj cerebella. This is due to a loss of many Purkinje cells (Fig. 6) and may be due to poor branching of the dendrites of the remaining cells. Moreover, dendritic trunks as well as branches appeared stubby (Fig. 11). They directly faced parallel fibers without a glial sheath from their proximal portions to the distal branches, whereas the proximal dendrites in j + cerebella were ensheathed by glial processes. Synapses with parallel fibers were found on the smooth surface of these dendrites and showed an immature appearance. Parallel fiberspine synapses, which were the only synapses between parallel fibers and Purkinje cells in adult j +
cerebella, were apparently seen but less frequent in jj cerebella than in j + cerebella. Spines emerged not only from the dendritic branches but also from the trunks (Fig. 11). The parallel fiber boutons contained aggregated synaptic vesicles and asymmetrical synaptic junctions were formed. In jj cerebella at days 18-30, the dendrites of the remaining Purkinje cells were still naked or covered partly by glial processes even in the rather thick portions (Fig. 12). The concentration of parallel fiber-spine synapses was unusually low. These synapses were found on the spines arising from rather thick dendrites as well as branchlet spines (Figs. 12 and 13). Parallel fiber boutons were often enlarged with a light matrix and contained a large number of synaptic vesicles (Fig. 14). The synaptic m e m b r a n e densities were asymmetrical and conspicuous (Fig. 14). The enlargement of parallel fiber synaptic boutons became prominent with age. Parallel fibers
Fig. 12. Molecular layer of the culmen from a jj Gunn rat aged 18 days. A thick Purkinje cell dendrite (Pcd) runs horizontally and gives off two branchlets up and down. The glial covering is incomplete (arrowheads), so most of the dendritic surface directly faces the parallel fibers. A few parallel fibers are in synaptic contact with spines arising from the thick dendrite (arrows 1 and 2) or a spine arising from one of the branchlets (arrow 3). A stellate cell synapse is seen on the thick dendrite (asterisk). Bar = 1/zm.
Fig. 13. Molecular layer of the culmen from a # Gunn rat aged 18 days. Three parallel fiber boutons are in synaptic contact with a thick Purkinje cell dendrite (Pcd). One of them (Pf 1) directly contacts the dendritic shaft. The synaptic membrane densities are prominent and a large number of synaptic vesicles are present. The second (Pf 2) is in contact with a spine which emerged from the dendrite. The third (Pf 3) is enlarged and is in contact with both the dendrite and a spine which presumably emerged from this dendrite. Bar = 0.5/zm.
122 were occasionally in synaptic contact with the thick dendritic shafts (Fig. 13). The parallel fiber boutons of these shaft junctions at this stage were larger than those at previous stages and contained aggregated synaptic vesicles. In addition, synaptic membrane densities became conspicuous. Thus, the parallel fiber presynaptic boutons seemed to mature on the dendritic shafts. Parallel fibers forming synaptic contact with the dendritic debris or atrophied spines of Purkinje cells were occasionally seen from days 18 to 23. Some parallel fiber synaptic boutons faced astrocytic processes in the absence of their postsynaptic partners (Fig. 15). The presynaptic membranes and synaptic vesicles appeared normal, although the latter varied in number. Electron-dense materials were usually present in the space between the presynaptic membrane of the parallel fiber and the
Fig. 14. Molecular layer of the culmen from ajj Gtmn rat aged 18 days. Several synapses between parallel fibers and Pu.rkinje cell dendritic spines (s) are seen and are surrounded by glial processes. They show a mature appearance with conspicuous synaptic membrane densities and increased number of synaptic vesicles. The synaptic boutons are unusually enlarged. A few profiles of enlarged parallel fiber boutons with synaptic vesicles but without synaptic membrane density indicate that these enlarged boutons were sectioned at a level other than that including the synaptic site. Bar = 0.5 gm.
glial process, but they were either not evident or less prominent in some cases. The parallel fiber boutons were usually as enlarged as those which were in synaptic contact with Purkinje cell spines. However, small parallel fiber synaptic boutons with a few synaptic vesicles in the absence of postsynaptic partners were present in the upper molecular layer from days 18 to 30. In cerebella from adult jj rats, parallel fiberPurkinje spine synapses were frequently seen only in the molecular layer above the remaining Purkinje cells (Fig. 16) but were seldom found in other parts of the molecular layer. The enlargement of parallel fiber boutons became prominent (Fig. 16). The parallel fiber synapses were also occasionally seen on the smooth surface of stubby dendrites, though some thick dendrites were often partially covered by glial processes except at the synaptic sites with climbing fibers or stetlate cell axons. These parallel fiber synaptic boutons appeared as mature as those seen in jj cerebella at day 18. Moreover, parallel fiber synaptic boutons without their postsynaptic partners
Fig. 15. Molecular layer of the culmen from ajj Gunn rat aged 30 days. Parallel fiber synaptic boutons (Pf) without their postsynaptic partners face glial processes (G1). They show the same features as those which are in contact with the postsynaptic partners. Bar = 0.5 pro.
123
Fig. 16. Molecular layer of the culmen from a # Gunn rat aged 8 months. Several parallel fiber-Purkinjc spine synapses (asterisks) are present. One of them is on the spine arising directly from the dendrite (Pcd) (arrow). Enlargement of parallel fiber boutons is prominent. Excessive glial processes (GI) segregate neuronal elements. Bar = 1/~m.
were encountered as frequently as at days 18-30. Their features were the same as those observed at days 18-30 (Fig. 17).
Development of climbing fiber-Purkinje cell synapses At day 7, climbing fiber synapses were found on the perisomatic processes of Purkinje cells in both jj (Fig. 8) and j + cerebella. The synapses displayed the same features in cerebella from both genotypes of rats. The perisomatic processes had a variety of shapes and were rather large in size. Climbing fiber synaptic boutons were also irregular in shape and contained a large number of synaptic vesicles, some of which had a dense core, and mitochondria. Some climbing fibers synapsed directly with the Purkinje cell soma. In j + cerebella, the climbing fiber synapses with
the perisomatic processes of Purkinje cells were most often observed at day 7, diminished in number thereafter and were never found after day 18. On the other hand, in jj cerebella they were most often found at day 12 and still frequently encountered from days 18 to 30 (Fig. 18). In ]+ cerebella, climbing fiber synapses with spines arising from Purkinje dendritic trunks began to be formed and increased in number beginning at day 12. In jj cerebella, climbing fiber/Purkinje dendritic spine synapses were rarely found at days 18-30 but were frequently found at the adult stage (Fig. 19). The synaptic boutons, in which round synaptic vesicles including several cored vesicles were densely packed, were unusually large. On the other hand, synapses with perisomatic processes, although small in number, were still found at the adult stage (Fig. 20). Therefore, the translocation of
124
Fig. 17. A parallel fiber bouton without its postsynaptic partner in the molecular layer of the culmen from a jj Gunn rat aged 6 months. Electron opaque material is present in the space between the synaptic membrane of a parallel fiber bouton and the plasma membrane of astrocytic process (arrow). Bar = 0.25/~m.
Fig. 19. Climbing fiber (Cf) and stellate celt (St) synapses on a Purkinje dendritic (Pcd) trunk in the culmen from a jj Gunn rat aged 6 months. The climbing fiber synaptic boutons contain densely packed round vesicles and several dense cored vesicles, and make conspicuous and asymmetrical junctions with Purkinje dendritic spines (s): The stellate cell axons make inconspicuous and symmetrical synaptic junctions with the smooth surface of the dendrite. However, their synaptic boutons are unusually large with a large number of round and ellipsoidal vesicles. Bar = 1 ~um.
the climbing fiber synaptic sites from the perisomatic processes to the spines of the dendritic trunks seemed to be delayed or arrested in jj cerebella.
Development of basket or stellate cell-Purkinje cell synapses In jj cerebella, cells in the molecular layer, most
Fig. 18. Climbing fiber (Cf) synapses on the soma of a Purkinje cell (Pc) in the culmen from a jj Gunn rat aged 18 days. Climbing fibers are in synaptic contact with both the perisomatic processes and the smooth surface of the soma. Bar = 1 /~m.
of which are basket or stellate cells, were scarce from day 12 onward. However, the basket cell synapses were found on the soma of Purkinje cells beginning at day 12, although they were not frequently encountered up to day 30 (Fig. 21). Most of the surface of the Purkinje cell soma, the main synaptic site of basket cell axons, was occupied by the perisomatic processes synapsing with the climbing fibers. At the adult stage, more basket cell axons terminated on the Purkinje cell soma which was mostly ensheathed by layers of glial processes except at the synaptic sites. Synaptic b o u t o n s of the basket cells appeared almost normal except that r o u n d and ellipsoidal vesicles were rather densely packed in occasionally enlarged synaptic boutons. Stellate c e l l - P u r k i n j e cell synapses were also rarely found from days 12 to 30 (Fig. 12). However,
125
Fig. 20. A climbing fiber (CO synapse with a perisomatic process of a Purkinje cell (Pc) in the culmen from a jj Gunn rat aged 6 months. The perisomatic process is small and has a spine-like shape. Bar = 0.5/zm.
Fig. 21. A basket cell (Bc) synapse on the soma of a Purkinje cell (Pc) in the culmen from a ]] Gunn rat aged 30 days. The synaptic bouton contains a few ellipsoidal and round synaptic vesicles and the synaptic membranes are inconspicuous. Bar = 1 ~m.
they were occasionally concentrated on the Purkinje cell dendritic trunk at the adult stage (Fig. 19). Stellate cell synaptic boutons were unusually large and were distinguished from the climbing fiber synaptic boutons by the presence of flattened vesicles and inconspicuous synaptic membranes. DISCUSSION
description by these authors. According to the present observation, the damage seems to increase rapidly in degree from postnatal days 7 to 23 in most Purkinje cells and severely damaged cells die and disappear by day 30. On the other hand, less affected cells may recover from the damage by day 30 and survive to the adult stage.
Course of Purkinje cell abnormalities
Development of parallel fiber-Purkinje cell synapses
The present study describes Purkinje cell abnormalities in the culmen of ]] Gunn rat cerebella. The Purkinje cell abnormalities found in the present observations were qualitatively the same as those reported by Schutta and Johnson 21. Despite their detailed observation on alterations in the Purkinje cells as compared with ours, modes of progression of the damage, especially in relation to phases in the maturation of this neuron and of the molecular layer of the cerebellum 1, are not necessarily clear from the
Homozygous (/7) Gunn rat cerebella showed various abnormalities in synaptic formation, such as persistent synaptic junctions between parallel fibers and the dendritic shaft of the Purkinje cells, the presence of presynaptic elements of parallel fibers in the absence of their postsynaptic partners, and an enlargement of parallel fiber synaptic boutons. Synaptic junctions between parallel fibers and Purkinje dendritic shafts were found in j + rats during the first four weeks in the present observa-
126 tion. These junctions were previously considered as transient L16 and to evolve to mature spine synapses 4 12 A recent investigation using a freeze-fracture technique suggests that these shaft junctions are not simply a transient intermediate in the formation of mature spine synaptic junctions, but rather represent a class of synaptic junction which has the capacity to dissociate in the remodeling process accompanying normal development 1°. In jj cerebella, the synaptic junctions between parallel fibers and Purkinje dendritic shafts were found at all stages examined. In these synaptic junctions, parallel fibers were immature with a few synaptic vesicles and inconspicuous synaptic membrane densities at a stage as early as day 7 but by day 18 some of them showed a mature appearance as seen in the mature spine synapses. The membrane density at synaptic sites on the Purkinje dendritic shaft was higher in later stages than in earlier stages. Thus, it seems to be likely that, in the jj cerebella, immature parallel fiber synaptic boutons mature on the outgrowing dendrite of Purkinje cells at early stages of development and that some of the parallel fiber-Purkinje dendritic shaft junctions do not or cannot dissociate in the later stage of development, but persist until the adult stage. The reason why the shaft junctions persist in the j] cerebellum is difficult to explain. They might have lost the capacity to dissociate due to the Purkinje cell abnormality. An alternative explanation is that the shaft junctions might supply a deficiency of mature spine synapses. Although there has been no direct evidence that the shaft junction can perform synaptic transmission in normal animals, an electrophysiological study in staggerer mutant mice aged 18-22 days demonstrated that parallel fibers were able to activate Purkinje cells 3. This suggests that the shaft junction fulfils its function as a synapse in this mutant animal, because in the staggerer mutant mouse cerebellum the only junctions between parallel fibers and Purkinje cells are on the dendritic shafts of these cells during the first 3 weeks after birth 12. In the pcd mutant mouse cerebellum in which virtually all Purkinje cells begin to degenerate at around postnatal days 15-18 (ref. 17), the appropriate afferent fibers have formed synapses on the Purkinje cells by day 15 (ref. 14). At day 18,
however, parallel fibers form synapses directly onto Purkinje dendritic shafts in an abnormal sequence of synaptogenesis, and these synapses have been thought to represent an aberration of parallel fiber synaptogenesis that is an early manifestation of the Purkinje cell damage in the pcd cerebellum 11. Thus, although the parallel fiber-Purkinje dendritic shaft junctions are found at unusually late stages in the pcd mutant mouse as well as jj Gunn rat cerebella, they are different in the mode and time of formation between both cerebella. Parallel fiber synaptic boutons devoid of their postsynaptic partners were often found in jj cerebella from day 18 through the adult stage. They were almost identical with presynaptic boutons of mature parallel fiber-Purkinje spine synapses and faced astrocytic processes. In the staggerer mutant mouse cerebellum in which dendritic branchlet spines are virtually not formed, similar parallel fiber synaptic boutons devoid of their postsynaptic partners have been observed 12"25. These synaptic boutons were found at postnatal day 26 (ref. 25) but were no longer seen at day 33 or 43 (ref. 12). However, in the nervous mutant mouse cerebellum, in which 90% of Purkinje cells selectively degenerate between days 23 and 50 (ref. 13), parallel fiber synaptic boutons devoid of their postsynaptic partners have been found as late as 6-14 months of age 27. Two hypotheses which are related to each other have been suggested concerning the development and fate of such presynaptic terminals in the absence of their postsynaptic partners 24'27. One is that parallel fibers may be able to produce autonomously, i.e., through an intrinsic process of differentiation, transient 'bouton en passant' with normal presynaptic organelles in the absence of their postsynaptic targets. The other is that presynaptic fibers in the cerebellum, which have been once stabilized by establishing function synapses, may be able to survive for long periods even if they lose their postsynaptic partners and those which have not made synaptic contacts may be unable to survive independently, going rapidly through the process of degeneration. The first hypothesis is based on the findings in the staggerer cerebellum 24"25, although a possibility that the presence of parallel fiber synaptic boutons devoid of their postsynaptic partners may be the
127 result of the detachment of parallel fibers from the Purkinje dendritic shaft after formation of the parallel fiber-Purkinje dendritic shaft junctions has been suggested 7'1°'12. The second hypothesis is based on the difference between the staggerer and nervous mutant mice in the fate of parallel fiber synaptic boutons in the absence of their postsynaptic partn e r s 24,27.
In the present observation of ]j cerebella, no marked difference in the concentration of parallel fiber synaptic boutons facing astrocytic processes was noted between postnatal day 30 and 5-10 months of age. This suggests that, in the j] cerebella, the parallel fiber synaptic boutons which were stabilized by making synaptic contacts with the Purkinje spines at the earlier stages lost their postsynaptic partners due to the progressive degeneration of Purkinje cells by day 30 and these boutons could survive to the adult stage without their postsynaptic partners, as in the nervous mutant mouse. This speculation is also supported by the fact that parallel fibers in contact with atrophied spines of Purkinje cells were found at days 18-23. However, the present study cannot exclude the possibility of the autonomous development of presynaptic specializations in the absence of their postsynaptic targets because the period of parallel fiber-Purkinje cell synaptogenesis overlapped almost completely with that of Purkinje cell degeneration in the jj cerebella. The unusual enlargement of parallel fiber synaptic boutons which were found in ]j cerebella after day 18 might represent a transsynaptic effect of the Purkinje cell degeneration. By electron microscopic observation of the molecular layer in j] cerebella, no abnormality was noted in non-synaptic segments of parallel fibers, suggesting that the possible transsynaptic effect on granule cells was limited to synaptic terminals of their axons.
Development of climbing fiber-Purkinje cell synapses Climbing fiber synapses were the most numerous synapses found in early stages of development of jj as well as j + cerebella. In the normal development of the rat cerebellum, climbing fiber synapses are first formed on the soma and/or perisomatic processes of Purkinje cells from postnatal day 5 and
then translocated to the spines of dendritic trunks from day 10 (ref. 1). In the present observation, the climbing fiber synaptogenesis in j + cerebella was similar to this. In ]] cerebella, however, the climbing fiber synapses with perisomatic processes still remained in appreciable numbers at day 30 and a few of them were even seen until the adult stage, whereas the synapses with Purkinje dendritic spines were rarely found until day 30. It has been suggested that the poor development of Purkinje cell dendrites deprives climbing fibers of the dendritic surface available for synapse formation and prevents the fibers from translocating themselves to the dendrites, so the climbing fiber synapses with perisomatic processes of Purkinje cells persist for an abnormally long time 11'12. The climbing fiber synaptogenesis in jj cerebella also seems to be delayed and/or arrested due to the Purkinje cell damage. However, the synaptogenesis might proceed when Purkinje cells have recovered from the damage, since climbing fiber synapses with dendritic spines were prominent yet only a few synapses with perisomatic processes were seen in less affected Purkinje cells present in adult j] cerebella.
Development of basket or stellate cell-Purkinje cell synapses The Purkinje cells in ]j cerebella received the synaptic inputs from both basket and stellate cells in accordance with the normal developmental course. However, the basket cell synapses were scarce until day 30. In the normally developing rat cerebellum, most of the temporary climbing fiber synapses on the soma and/or perisomatic processes of the Purkinje cell disappeared before basket cell axon terminals began to establish synapses on the Purkinje cell soma 1. In ]] cerebella, the delay or arrest of translocation of the climbing fiber synaptic sites from the Purkinje cell soma to the spines of the dendritic trunks and of disappearance of the perisomatic processes may prevent basket cell axons from making synaptic contact with the Purkinje cell soma. Stellate cell synapses were also poorly detected through the developmental period. This seems due to the fact that dendrites of the remaining Purkinje cells were underdeveloped and, in addition, that stellate cells were scarce. The synaptogenesis of
128
b a s k e t and stellate cells m i g h t p r o c e e d w h e n the
ACKNOWLEDGEMENT
P u r k i n j e cells h a v e r e c o v e r e d as m i g h t be the case of c l i m b i n g fiber s y n a p t o g e n e s i s .
T h e a u t h o r s t h a n k D r . C. S o t e l o for critically r e a d i n g this m a n u s c r i p t .
REFERENCES 1 Altman, J., Postnatal development of the cerebellar cortex in the rat. II. Phases in the maturation of Purkinje cells and of the molecular layer, J. Comp. Neurol., 145 (1972) 399-464. 2 Altman, J. and Anderson, W.J., Experimental reorganization of the cerebellar cortex. I. Morphological effects of elimination of all microneurons with prolonged X-irradiation started at birth, J. Comp. Neurol., 146 (1972) 355-406. 3 Crepel, E and Mariani, J., Anatomical, physiological and biochemical studies of the cerebellum from mutant mice. I. Electrophysiological analysis of cerebellar cortical neurons in the staggerer mouse, Brain Research, 98 (1975) 135-147. 4 Foelix, R.E and Oppenheim, R., The development of synapses in the cerebellar cortex of the chick embryo, J. Neurocytol., 3 (1974) 277-294. 5 Gunn, C.K., Hereditary acholuric jaundice in a new mutant strain of rats, J. Hered., 29 (1938) 137-139. 6 Herndon, R.M., Margolis, G. and Kilham, L., The synaptic organization of the malformed cerebellum induced by perinatal infection with the feline panleukopenia virus (PLV). II. The Purkinje cell and its afferents, J. Neuropathol. Exp. Neurol., 30 (1971) 557-570. 7 Hirokawa, N., A study of the synaptogenesis in the cerebellar cortex through chronic treatment and immunocytochemistry of fl-bungarotoxin, J. Comp. Neurol., 185 (1979) 107-120. 8 Johnson, L., Sarmiento, E, Blanc, W.A. and Day, R., Kernicterus in rats with an inherited deficiency of glucuronyl transferase, Am. J. Dis. Child., 97 (1959) 591-608. 9 Katoh-Semba, R., Aoki, E. and Kashiwamata, S., Cerebellar hypoplasia in Gunn rats: effects of bilirubin on postnatal changes in cerebellar constituents, Biomed. Res., 3 (1982) 319-329. 10 Landis, D.M.D., Initial junctions between developing parallel fibers and Purkinje cells are different from mature synaptic junctions, J. Comp. Neurol., 260 (1987) 513-525. 11 Landis, D.M.D. and Landis, S.C., Several mutations in mice that affect the cerebellum, Adv. Neurol., 21 (1978) 85-105. 12 Landis, D.M.D. and Sidman, R.L., Electron microscopic analysis of postnatal histogenesis in the cerebellar cortex of staggerer mutant mice, J. Comp. Neurol., 179 (1978) 831-864. 13 Landis, S.C., Ultrastructural changes in the mitochondria of cerebellar Purkinje cells of nervous mutant mice, J. Cell Biol., 57 (1973) 782-797. 14 Landis, S.C. and Mullen, R.J., The development and degeneration of Purkinje cells in pcd mutant mice, J. Comp. Neurol., 177 (1978) 125-144. 15 Llimls, R., Hillman, D.E. and Preeht, W., Neuronal circuit reorganization in mammalian agranular cerebellar cortex, J. Neurobiol., 4 (1973) 69-94. 16 Mugnaini, E., Ultrastructural studies on the cerebellar histogenesis. IL Maturation of nerve cell populations and
17
18
19
20
21
22
23
establishment of synaptic connections in the cerebellar cortex of the chick. In R. Llin~is (Ed.), Neurobiology of Cerebellar Evolution and Development, American Medical Association, Chicago, 1969, pp. 749-782. Mullen, R.J., Eicher, E.M. and Sidman, R.L., Purkinje cell degeneration, a new neurological mutation in the mouse, Proc. Natl. Acad. Sci. U.S.A., 73 (1976) 208-212. Rakic, P., Synaptic specificity in the cerebellar cortex: study of anomalous circuits induced by single gene mutations in mice. In The Synapse, Cold Spring Harbor Symposia on Quantitative Biology, Vol. 40, 1976, pp. 333-346. Sawasaki, Y., Yamada, N. and Nakajima, H., Developmental features of cerebeUar hypoplasia and brain bilirubin levels in a mutant (Gunn) rat with hereditary hyperbilirubinaemia, J. Neurochem., 27 (1976) 577-583. Schmid, R., Axelrod, J., Hammaker, L. and Swarm, R.L., Congenital jaundice in rats, due to a defect in glucuronide formation, J. Clin. Invest., 37 (1958) 1123-1130. Schutta, H.S. and Johnson, L., Bilirubin encephalopathy in the Gunn rat: a fine structure study of the cerebellar cortex, J. Neuropathol. Exp. Neurol., 26 (1967) 377-396. Schutta, H.S., Johnson, L. and Neville, H.E., Mitochondrial abnormalities in bilirubin encephalopathy, J. Neuropathol. Exp. Neurol., 29 (1970) 296-305. Sidman, R.L., Cell interactions in developing mammalian central nervous system. In L.G. Silvestri (Ed.), Cell
Interactions, Proceedings of the Third Lepetit Colloquium, North-Holland, Amsterdam, 1972, pp. 1-13. 24 Sotelo, C., Electrical and chemical communication in the central nervous system. In B.R. Brinkley and K.R. Porter (Eds.), International Cell Biology, Rockefeller Univ. Press, New York, 1977, pp. 83-92. 25 Sotelo, C. and Changeux, J.E, Transsynaptic degeneration 'en cascade' in the cerebellar cortex of staggerer mutant mice, Brain Research, 67 (1974) 519-526. 26 Sotelo, C. and Privat, A., Synaptic remodeling of the cerebellar circuitry in mutant mice and experimental cerebellar malformations: study 'in vivo' and 'in vitro', Acta Neuropathol., 43 (1978) t9-34. 27 Sotelo, C. and Triller, A., Fate of presynaptic afferents to Purkinje cells in the adult nervous mutant mouse: a model to study presynaptic stabilization, Brain Research, 175 (1979) 11-36. 28 Tanase, H., Sudo, S. and Suzuki, Y., Maintenance and propagation of jaundiced rat, Exp. Anita., 17 (1968) 23-25. 29 Weatherill, P.J., Kennedy, S.M.E. and Burchell, B., Immunochemical comparison of UDP-glucuronyltransferase from Gunn- and Wistar-rat livers, Biochem. J., 191 (1980) 155-163. 30 Yamano, T., Shimada, M., Nakao, K., Nakamura, T., Wakaizumi, S. and Kusunoki, T., Maturation of Purkinje cells in mouse cerebellum after neonatal administrations of cytosine arabinoside, Acta Neuropathol., 44 (1978) 41-45.