Synapses on the purkinje cell spines in the mouse an electronmicroscopic study

Synapses on the purkinje cell spines in the mouse an electronmicroscopic study

BRAIN RESEARCH 15 Research Reports SYNAPSES ON T H E P U R K I N J E CELL SPINES IN T H E MOUSE AN E L E C T R O N M I C R O S C O P I C STUDY L. M...

923KB Sizes 0 Downloads 6 Views

BRAIN RESEARCH

15

Research Reports

SYNAPSES ON T H E P U R K I N J E CELL SPINES IN T H E MOUSE AN E L E C T R O N M I C R O S C O P I C STUDY L. M. H. LARKAMENDI AND T. VICTOR Department of Anatomy, University of lllinois College of Medicine, Chicago, IlL (U.S.A.)

(Received October 21st, 1966)

INTRODUCTION The Purkinje cells of the cerebellum in their early developmental stages show, in Golgi impregnations, a very irregular surface from which many small cytoplasmic processes emerge which later disappear or resorb as noted by Caja111, Purpura et al. 1° and others. An electronmicroscopic study of the Purkinje cell of the mouse during the first and second postnatal weeks by Larramendi s has demonstrated that these cytoplasmic processes make synaptic contacts. Similar processes have been observed in the large dendrites and apical portion of the adult Purkinje cell by Larramendi and Victor 9. These processes make synaptic contact with morphologically distinct nerve terminals. We shall present evidence to support the interpretation that climbing fibers are the source of these terminals. It will also be shown that, in the glutaraldehyde-perfused mouse, parallel, stellate, basket and recurrent Purkinje collateral fiber endings can be identified by their morphological characteristics. MATERIALSAND METHODS Mice, 35 g in weight and over 90 days old, were used in this study. Following the Karlsson and Schultz 6 procedure of fixation, the mice were perfused in 2.5 ~ glutaraldehyde, the cerebellum removed, cut into small blocks and immersed in osmium for 2 h. The blocks were subsequently sectioned on a Porter-Blum ultramicrotome and stained, first with uranyl acetate for 15 min and then with lead citrate for 1.5 min. The sections were observed on a RCA EMU3 electron microscope. The sample analyzed is described in Tables I, II, and III. RESULTS (1) Soma and dendritic spines of the Purkinje cell The short cytoplasmic processes (spines) emerging from the soma and large Brain Research, 5 (1967) 15-30

16

L. M. H. LARRAMENDI AND T. VICTOR

Fig. 1. Purkinje cell, apical portion. Observe several terminals descending on the soma and some axodendritic synapses, presumably from the descending terminals. Notice near the apical portion of the cell a dark structure making synaptic contact with a spine profile. In the following figures the characteristics of the descending terminals and that of the dark structure will be shown in higher magnification.

Brain Research, 5 (1967) 15-30

SYNAPSESOF THE PURKINJECELLSPINES

17 r

dendrites have similar characteristics to those of the branchlets (dendrites with an average diameter of 1 #). They are filled with a homogeneous amorphous material and contain profiles of agranular cisternae. Most of the usual cytoplasmic organelles such as mitochondria, ribosomes, neurofilaments (75 A in diameter) or neurotubules (200 A in diameter) are absent. However, the spines are not exactly alike throughout the length of the Purkinje cell; those emerging from the soma, somatic spines (Ss), and large dendrites, dendritic spines (Ds), differ from those of the branchlets, branchlet spines (Bs), in that the Ss and Ds spines have a shorter stem, a more bulbous head, and a more dense amorphous material (Fig. 2). These differences appear to be related more to the distance of their point of emergence with respect to the nucleus than to the type of endings which make synaptic contact with them. Bs spines have been described in various species before by Gray 2, Fox 1, Kirsche et al. 7, Hamori and Szent~tgothai8, but Ss and Ds spines have not been reported previously, except in a preliminary form by Larramendi and Victor 9.

(2) Nerve ending profiles on Ss, Ds and Bs spines The nerve endings making synaptic contacts with spines are of two main types; we shall refer to them as p- and c-type termhzals. p-Type terminals. These are synapses 'en passage' of the parallel fiber 'crossingover synapse' of Hamori and Szent~gothai3. They contain large, round, synaptic vesicles, mostly aggregated close to the synaptic junction. The rest of the profile of the terminal contains a few mitochondria and three to four neurotubules within an abundant axoplasm. The synaptic thickenings are particularly pronounced on the postsynaptic surface (Fig. 3). These characteristics have been previously reported1-8, 7. p-Type terminals in the mouse are seen mostly on Bs spines and occasionally on Ds spines; they do not make contact with Ss or Ds spines from the main dendrites. c-Type terminals. The average profiles of these terminals are twice as large as the p-type terminals. They contain round, large, closely packed synaptic vesicles which fill the entire profile of the nerve terminal. The intervesicular axoplasm is very scarce, neither neurotubules nor neurofilaments are discernible, and only a few mitochondrial profiles are seen (Fig. 3). In the few profiles in which the axon connecting the enlargements has been observed, the axons appear packed with neurotubules (200 A in diameter); neurofilaments seem to be absent (Fig. 3). The synaptic thickenings of the pand c-type terminals are similar. Ss and Ds spines invaginate the c-type terminals more deeply than do Ds and Bs spines which associate with the p-type terminals, c-Type terminals make contact with the Purkinje cell exclusively through Ss and Ds spines. Simple appositions of c-type terminals on the smooth surface of dendrites have been observed, but always lacking synaptic thickenings. (3) Nerve endings on the non-spine surface of soma, dendrites, and branchlets Most of these nerve endings have similar morphological characteristics, whether found on the soma, the dendrites, or the branchlets of the Purkinje cell. They contain small, ellipsoidal vesicles, dispersed within a clear and abundant axoplasm, which sometimes aggregate near the synaptic junction but never are densely packed. A few Brain Research, 5 (1967) 15-30

18

k . M . H . LARRAMEND1 AND T. VI(IOR

neurotubules and more abundant neurofilaments follow the longitudinal orientation of the terminal. We shall refer to these as the b-s terminals. Although the appearance o1" these terminals is similar along the entire Purkioje cell, certain differences can bc

Fig. 2. Purkinje cell spines. Top, somatic spines (Ss). Middle, dendritic spines (Ds). Bottom, branchlet spines (Bs). Compare the morphological characteristics of spines and that of terminals making contact with Ss and Ds spines (c) and Bs spines (p).

Brain Research, 5 (1967) 15-30

SYNAPSESOF THE PURKINJE CELLSPINES

19

Fig. 3. p-, b-s-, and c-Type terminals. Top, parallel fibers (p-type) making synaptic contacts with Bs spines. Left bottom, terminal (b-s-type) making a large synaptic contact with large Purkinje dendrite. Right bottom, large c-type terminal near Purkinje dendrite making synaptic contact with Ds spine. Observe, in same inset, an axon profile (ax) containing synaptic vesicles of similar characteristics to these of c-type terminals packed with neurotubules. Compare the size, shape and aggregation of synaptic vesicles in p-, b-s-, and c-type terminals. Notice some very large vesicles in c-type terminal. We have evidence that these are depleted dense core vesicles. recognized between the ones on the soma and main emerging dendrites and those on the rest of the dendritic tree. These differences are mainly in the appearance of the synaptic contacts. Those on the soma and apical portion of the cell body have a slightly crenated surface and sometimes show small invaginations into the Purkinje cell body. The synaptic thickenings in these contacts are light and symmetrical and appear as small spotty areas (Fig. 5, top). The nerve endings on the rest of the dendritic tree make larger and smoother synaptic contacts. As in the soma contacts, the synaptic thickenings are light and more or less symmetrical (Fig. 5, middle and bottom). The basic similarities of all b-s terminals suggest that they may have their origin in cells of the same general type. This would be the case if basket and stellate cells were the cells of origin of these terminals. The Golgi study by Scheibel and Scheibe112 of the synaptic connections of the basket and stellate cells upon the Purkinje dendrites supports this interpretation. Further evidence was obtained by analyzing the morphological appearBrain Research, 5 (1967) 15-30

20

L. M. H. LARRAMENDI AND T. VICI'OR

Fig. 4. c-Type terminals (c) near the Purkinje cell at three levels. Top, near soma. Middle, near large dendrite. Bottom, near smaller dendrite. Notice c-type terminals making contact presumably with somatic (Ss) and dendritic spines (Ds).

Brain Research, 5 (1967) 15-30

SYNAPSES OF THE PURKINJE CELL SPINES

21

Fig. 5. b-s-Type terminals making synaptic contacts at three levels on the Purkinje cell. Top, soma. Middle, large dendrite. Bottom, branchlet. Observe the same type of synaptic vesicles in all three terminals and the different type of contacts made on the soma and on the dendrites. Notice the loose aggregation of synaptic vesicles, the abundance of neurofdaments and the scarcity of neurotubules in the terminal making contact on the soma. Compare this figure with Figs. 6 and 7.

Brain Research, 5 (1967) 15-30

22

Brain Research, 5 (1967) 15-30

L. M. H. LARRAMENDI AND T. VIC"It~)F,

SYNAPSES OF THE PURKINJE CELL SPINES

23

ance of axons crossing the molecular layer along the longitudinal axis of the folium. Whether in the upper molecular layer (axons from stellate cells) or in its lower portion (presumably basket axons), the contents of these axons had the characteristics of the b-s-type terminals described above (Fig. 6) indicating that these are, in fact, stellate and basket terminals. The observations by Kirsche et al. 7 and Hamori and Szenthgothai 4,5 on synapses on the non-spine surface of the Purkinje cells will be discussed later in this paper. In summary, we can conclude that in the mouse most of the synapses on the nonspine surface of the Purkinje cells are made by basket and stellate terminals.

(4) Identification of the terminals of the recurrent collaterals of the Purkinje cells Since one of the possible sources for terminals on the Purkinje dendrites is the recurrent collateral, a study was made of the myelinated nerve fibers in the molecular layer. Three myelinated fibers were traced to the point were the myelin ended. In all three instances the terminals appeared enlarged. Within this enlargement, mitochondria, small and numerous ellipsoidal vesicles similar in appearance to those seen in the b-s-type terminals, and very abundant profiles of agranular tubular structures, very infrequently seen in the other described nerve endings, were observed. Neurotubules and neurofilaments were also seen. Of the three myelinated fibers traced to their endings in the molecular layer, one made synaptic contact with the soma of a stellate cell; the remaining two did not make synaptic contacts (Fig. 7). Once the morphological characteristics of these terminals were recognized, others were identified although they lacked the preterminal myelin sheath. These endings were observed to terminate on the soma of stellate cells and on the non-spine dendritic surface of the Purkinje cells. We have the impression that these terminals are infrequent, as would be expected in the mouse where the myelinated fibers in the molecular layer are very scarce. It is reasonable to conclude, therefore, that recurrent collaterals are not the source of the abundant b-s terminals on the Purkinje dendrites.

(5) Quantitative distribution of spines and c-, p- and b-s-type terminals on the Purkinje cell surface From the evidence presented thus far, it is suggestive that the c-type terminals are climbing fibers. To substantiate this hypothesis, a quantitative study of the distribution of spines and nerve terminals on the soma and dendrites of the Purkinje cell was made. In Tables I, II and III the results of this analysis are presented. In order to compare the measurements of the soma and dendrites, all counts are expressed in the same unit, i.e. 100/z of surface of longitudinal dendritic profile or soma perimeter. Fig. 6. b-s-Type axons and c-type axons in molecular layer. Top, axon crossing the upper molecular layer in sagittal plane. Middle, similar axon and orientation in the lower molecularlayer. Notice that both axons contain the same type of synaptic vesicles axoplasm and neurofilaments. Few neurotubules are observed. These characteristics are the same as those seen in b-s-type terminals (Fig. 5). Compare them with bottom inset which shows an axon of the c-type terminals. In this axon the neurotubules are very predominant and the synaptic vesicles are different in size, shape and aggregation from those in b-s-type terminals. Compare also the spine profiles making contact with c-type terminals to those to the right which are p-terminals on Bs spines.

Brain Research, 5 (1967) 15-30

24

L. M. H. LARRAMENDI AND T. VICTOR

Fig. 7. Nerve terminal of myelinated fiber (r) near Purkinje cells in the molecular layer. Observe synaptic vesicles, mitochondria, a dark body, and agranular tubular structures within the r-type terminal. Notice similarities and differences between this terminal and nearby b-s-type terminal on Purkinje soma. r-Type terminals are very abundant on the middle and lateral cerebellar nuclei (Larramendi, Fickenscher and Lemkey, in preparation).

Brain Research, 5 (1967) 15-30

SYNAPSES OF THE PURKINJE CELL SPINES

25

TABLE I ANALYSIS

OF

B-S-

AND

P-TYPE

TERMINALS ON SMOOTH SURFACE OF PURKINJE

SOMA AND

DENDRITES

b-s-Type terminals interpreted as basket or stellate fibers; p-type terminals interpreted as parallel fibers. Dendrite average diameter (~J

Length of dendrite or soma surface analyzed (~)

1.1 1.4 2.1 3.4 Soma

450 116 214 672 624*

Number o f profiles per 100 # o f dendrite or soma surface b-s Terminals synapsing on dendrites or soma

p Terminals synapsing on dendrites or soma

1.30 3.45 3.75 4.15 11.85

0.4 0 0 0 0

* Total perimeter of 10 cells sectioned along axodendritic axis. TABLE II ANALYSISOF PURKINJEDENDRITICAND SOMATICSPINES p-Type terminals interpreted as parallel fibers; c-type terminals interpreted as climbing fibers. Dendrite average diameter (i~)

1.1 1.4 2.1 3.4 Soma

Length o f soma or dendrite surface analyzed (#)

Number o f profiles o f spines emerging per 109 kt o f soma or dendrite surface Total number

Synapsing with p terminals

Synapsing with c terminals

Without synaptic contacts**

450 116 214 672 3220"

32.95 8.25 7.95 1.55 0.19

5.2 3.2 1.4 0 0

0 0 1.85 0.95 0.16

27.75 5.05 4.70 0.60 0.03

* Total perimeter of 50 cells sectioned along axodendritic axis. ** No synapses present in the sections of the sample. T a b l e I shows t h a t b - s - t y p e t e r m i n a l s d i s t r i b u t e a c c o r d i n g to a gradient, with a m a x i m u m at the s o m a a n d a m i n i m u m at the b r a n c h l e t level. Between these two extremes, the n u m b e r o f b-s profiles on dendrites is a p p r o x i m a t e l y the same. This distrib u t i o n is consistent with d a t a o b t a i n e d in G o l g i studies o f b a s k e t a n d stellate cells by Scheibel a n d Scheibe112. A t the b r a n c h l e t level there are a few terminals c o n t a i n i n g large vesicles a n d p r o m i n e n t p o s t s y n a p t i c thickenings, which have the m o r p h o l o g i c a l a p p e a r a n c e o f p a r a l l e l fibers e n d i n g on the n o n - s p i n e surface o f the b r a n c h l e t a n d are referred to in the table as p terminals. In T a b l e II, the d i s t r i b u t i o n a l o n g the P u r k i n j e cell o f emerging spines is shown. T h e e m e r g i n g spines as a whole show a c o n t i n u o u s increase in n u m b e r f r o m the s o m a u p to the b r a n c h l e t level. T w o different g r a d i e n t s o f d i s t r i b u t i o n b e c o m e a p p a r e n t when spines are classified a c c o r d i n g to the type o f nerve ending with which they f o r m synapses. The c-type t e r m i n a l s y n a p s i n g with spines increases in frequency f r o m s o m a Brain Research, 5 (1967) 15-30

26

L.M.H. LARRAMENDIAND T. VICTOR

TABLE I11 ANALYSIS OF C-TYPE TERMINALS NEAR OR ON PURKINJE DENDRITES AND SOMA

c-Type terminals interpreted as climbing fibers; length of dendrite surface analyzed same as in Table ll; soma surface (perimeter) analyzed same as in Table 11. Dendrite average Number of c-type terminals per tO0 I~ of dendrite or soma diameter (!~) surface

1.1 1.4 2.1 3.4 Soma

Total number of spines synapsing with c terminals

Total number

Without synaptic contacts*

Synapsing Synapsing with emerging withisolated spine profiles spineprofiles* *

0

0

0

0

0

2.55 7.00 4.25 0.63

0.85 2.80 2.35 0.28

0 1.85 0.95 0.16

1.70 2.35 0.95 0.19

1.70 4.20 1.90 0.35

* No synapses present in the sections of the sample. ** Interpreted as the heads of spines from large dendrites and soma. to dendrites. The p-type terminal synapsing with spines increases in frequency from dendrite to branchlet. Both gradients overlap at dendrites of 2/~ in average diameter. In Table Ill, a more refined analysis of the c-type terminals is presented. The distribution of this type of terminal is of the somadendritic gradient mentioned above, whether the c-type terminals are isolated and close to the Purkinje cell or make contact with spines. This type of distribution is the one that could be expected if c-type terminals were climbing fibers, which according to Caja111, Scheibel and Scheibe112, and Szent~tgothai and Rajkovits 13, make contact with the soma and do not reach the branchlet level of the dendritic tree. (6) c-Type terminals in the molecular layer other than on the Purkinje cells

c-Type terminals have been seen occasionally in the molecular layer, making contact with the soma of stellate cells (Larramendi and Lemkey, in preparation). In most cases these terminals are larger than the parallel terminals on the same cells. Less frequently, c-type terminals are seen making synaptic contacts with dendrites from a source other than Purkinje cells. These observations are in agreement with those of Scheibel and Scheibe112 who demonstrate the termination of climbing fibers on the soma of stellate cells and the electron microscope observations of Hamori and Szentb, gothai 5. (7) Synaptic vesicle size and shape o f c-, p- and b-s-type terminals

Measurements of the major and minor diameters of vesicle profiles within several types of terminals in the mouse glutaraldehyde-perfused cerebellum have been made and they will be reported in a forthcoming publication (Larramendi, Fickenscher and Lemkey, in preparation). The analysis of these measurements indicates that the vesicles within the climbing and parallel fiber terminals are large and have a vesicle 'elongation index' (major over minor vesicle diameters) smaller than 1.2, and that the basket and stellate terminals show smaller vesicles, with an elongation index larger than 1.3. These Brain Research, 5 (1967) 15-30

SYNAPSES OF THE PURKINJE CELL SPINES

27

results confirm the suggestion made by Uchizono 14 that inhibitory terminals in the cerebellum have elongated vesicles whereas the excitatory terminals contain round vesicles. This confirmation results from the morphological identification of terminals based on the criteria presented in this paper, not upon vesicle size or shape, and supports the conclusions of this paper. In summary, from the analyzed sample, we conclude that in the glutaraldehydeperfused adult mouse, (1) various types of nerve terminals (c-, p-, b-s-, and r-types) can be recognized by their morphological characteristics, (2) Purkinje cells have somatic, dendritic and branchlet spines, increasing in number from soma to branchlet, (3) climbing fibers make synaptic contacts with the Purkinje cell exclusively on somatic

/ Parallel Climbing ~

I'~-" Stellate

Basket

Fig. 8. Synapseson mouse Purkinje cell. and dendritic spines, (4) parallel fibers make most of their synaptic contacts with branchlet spines, fewer with dendritic spines, and probably some with the smooth, non-spine surface of the branchlets, (5) most of the nerve terminals making synaptic contacts with the Purkinje cell smooth, non-spine surface, originate in basket and stellate cells, few probably in the Purkinje recurrent collaterals, and (6) excitatory terminals (parallel and climbing collaterals) contain large and round synaptic vesicles whereas inhibitory terminals (basket, stellates and the presumably inhibitory Purkinje recurrent collaterals) contain small, ellipsoidal vesicles. These conclusions are summarized in Fig. 8. Brain Research, 5 (1967) 15-30

28

L. M. H. L A R R A M E N D i A N D T. VICTOR

DISCUSSION

The evidence presented in this paper for the identification of climbing fibers differs from that obtained in other species. Neither Kirsche e t al. 7 nor Hamori and Szenfftgothai 5 have reported the presence of Ss or Ds spines in the rat or cat, respectively. In the mouse, however, Ss and Ds spines appear to be the exclusive sites of contact of the climbing fibers with the Purkinje cell. It is interesting to note that Kirsche e t al. 7 described synapses with small dendritic protuberances of the large Purkinje dendrites, which invaginated into the nerve terminals to form synaptic cups ('Kuppensynapsen'). Probably the small dendritic protuberances are resorbed Ds spines and, probably, they are making contacts with climbing terminals. Kirsche e t al. 7 had no suggestion as to the source of the 'Kuppensynapsen' terminals. They considered that the synaptic contacts on the smooth surface of the dendrites, with similar characteristics to the b-s-type terminals described in this paper, were the climbing terminals. Hamori and Szentfi.gothai 5 described two types of terminals on the smooth surface of the Purkinje cell dendrites in the cat: (I) a vesicular type, characterized by an abundance of synaptic vesicles, and (2) a second type, characterized by an abundance ofneurofilaments and scarce number of vesicles. These authors did not establish a difference between neurofilaments (75 A in diameter) and neurotubules (200 A in diameter), but presumably their 'course neurofilaments' were neurotubules. This distinction is an important one, since c-type and b-s-type terminals of the mouse differ in this respect. Hamori and Szentfigothai undercut the folium and observed that the terminals rich in neurofilaments had disappeared whereas the terminals rich in vesicles remained. Therefore, they concluded, the terminals rich in neurofilaments were from the climbing fibers, while the vesicular type originated in the basket, stellate, and recurrent Purkinje collaterals. Clearly, the above descriptions of the terminals to the Purkinje cells of the cat differ substantially from the conclusions reached in this paper for the mouse. The basis for this disagreement may be found in species differences. For instance, the Hungarian investigators do not report the presence of either protuberances or spines in the dendrites of the cat. If the dendritic protuberances described in the rat by Kirsche e t al. 7 are, in fact, the site of synaptic contact for the climbing fibers as we suggest, it can be speculated that these protuberances may represent a transitional form of receiving surface for the climbing fibers between the Ss and Ds spines present in the mouse and the smooth contacts on the dendrites reported by Hamori and Szentfigothai 5 in the cat. The clue to the understanding of these species differences, if they do exist, may be found in an ontogenetic study of the Ss and Ds spines in the rat and the cat. In the mouse, our studies8, 9 have shown that Ss and Ds spines are very abundant in the postnatal period (day 7) and that these spines make synaptic contacts with either climbing or basket terminals, or both. These spines, a few days later, decrease substantially in number. To account for these changes, three alternative explanations are possible: (1) Ss and Ds spines decrease in absolute numbers as a result of the stretch of the cell surface produced by the increase in soma and dendritic volumes with age, (2) the absolute numbers of Ss and Ds spines decrease by differential resorption; the spines making contact with the climbing fibers do not resorb whereas those with the basket terminals Brain Research,

5 (1967) 15-30

SYNAPSESOF THE PURKINJE CELL SPINES

29

do, and (3) Ss and Ds spines do not decrease in absolute numbers but only relatively per unit of surface due to the increase in cell surface and to the translocation upwards of the soma surface by the elongation of the cell. These main alternatives are under study in the mouse. In the cat, a similar situation may exist. However, since there are no electronmicroscopic reports of the presence of Ss or Ds spines in this species in the adult animals, it can be speculated that in the cat: (1) there is total resorption of somatic spine processes or, (2) the climbing fibers never come in synaptic contact with the somatic spine processes during development, and, therefore, make contacts with the Purkinje cell smooth surface. Whatever the case might be, it is obvious that a developmental study in the cat, as the one initiated a few years ago in the mouse by us, may help in the understanding of the species differences that apparently exist in the way climbing fibers end upon the Purkinje cell. The interesting possibility exists that the species transition from spine to smooth dendritic surface contact of the climbing fiber may result in a gain of synaptic surface per unit of axonal length. SUMMARY

In the mouse Purkinje cell, as seen in the electron microscope, spines emerge from the soma (Ss), large dendrites (Ds) and branchlets (Bs). Bs and some Ds spines make synaptic contacts with parallel fibers, but Ss and most Ds spines receive other types of terminals filled with large, round vesicles. A quantitative analysis of the distribution of these terminals has shown a clear correlation with the distribution of climbing fibers on the Purkinje soma and dendrites observed by other investigators in Golgi impregnations. It has been concluded that, in the mouse at least, climbing fibers make synaptic contacts with the Purkinje cell on spines; consideration is given to species differences in the way climbing fibers make synaptic contacts upon the Purkinje cell. The distribution of basket-stellate terminals on the Purkinje cell is reported. The suggestion made by Uchizono that synaptic vesicles of excitatory and inhibitory synapses have different morphological characteristics has been confirmed. ACKNOWLEDGEMENTS We thank Mrs. P. Larramendi for her technical assistance, Mr. L. Fickenscher for his help in tabulating data, and Dr. R. Greenberg for his advice preparing the manuscript. The electron microscope facilities of the Aero-Medical Laboratory of the University of Illinois College of Medicine were used. This work is supported by P. H. S. grant NB 05408-02. REFERENCES 1 Fox, C. A., The structure of the cerebellar cortex. In E. C. CROSBY,T. H. HUMPHREYANDE. W. LAUER(Eds.), Anatomy of the Nervous System, Macmillan, New York, 1962, p. 193. 2 GRAY,E. G,, The granule cells, mossy synapsesand Purkinje spine and synapsesof the cerebellum: light and electronmicroscopic observations, J. Anat. (Lond.), 95 0961) 345-356. Brain Research, 5 (1967) 15-30

30

L. M. H. LARRAMENDI AND T. VICTOR

3 HAMORI, J., AND SZENT,~GOTHAI, J., The 'crossing-over' synapse; an electronmicroscopic study of the molecular layer in the cerebellar cortex, Acta biol. Acad. Sci. hung., 15 (1964) 95-117. 4 HAMORI,J., AND SZENT,~GOTHAI,J., The Purkinje ceil baskets: ultrastructure of an inhibitory synapse, Acta biol. Acad. Sci. hung., 15 (1965) 465-479. 5 HAMORI,J., AND SZENT,~GOTHAI,J., Identification under the electron microscope of climbing fibers and their synaptic contacts, Exp. Brain Res., 1 (1966) 65-81. 6 KARLSSON, U., AND SCHULTZ, R. L., Fixation of the central nervous system for electronmicroscopy by aldehyde perfusion, J. Ultrastruct. Res., 12 (1965) 160-186. 7 KIRSCHE, W., DAVID, H., WINKELMANN, E., UND MARX, I., Electronmikroskopische Untersuchungen und synaptischen Formationen im Cortex cerebelli yon Rattus norvegicus, Berkenhoot, Z. mikr.-anat. Forsch., 72 (1964) 49-80. 8 LARRAMENDI, L. M. H., Purkinje axo-somatic synapses at seven and fourteen post-natal days in the mouse: an electronmicroscopic study, Anat. Rec., 151 (1965) 460. 9 LARRAMEND1,L. H. M., AND VICTOR, T., Soma-dendritic gradient of spine resorption in the Purkinje cell of the cerebellum of the mouse during postnatal development: an electronmicroscopic study, Anat. Rec., 154 (1966) 373. 10 PURPURA, D. P., SHOFER, R. J., HOUSEPIAN, E. M., AND NOVACK, C. R., Comparative ontogenesis of structure-function relations in cerebral and cerebellar cortex. In D. P. PURPURA AND J. P. SCHADE (Eds.), Progress in Brain Research, Vol. 4, Elsevier, Amsterdam, 1964, p. 187. I1 RAM6N v CAJAL, S., Histologie du Systdme Nerveux, Consejo Superior de Investigaciones Cientificas, lnstituto Ram6n y Cajal, Madrid, 1954, pp. 1-106. 12 SCHEIBEL,M., AND SCHEIBEL,A., Observations on the intracortical relations of the climbing fibers of the cerebellum: a Golgi study, J. comp. Neurol., 101 (1954) 733-764. 13 SZENT,~COTIaAI, J., END RAJKOVITS, K., ~)ber den Ursprung der Kletterfasern des Kleinhirns, Z. Anat. Entwickl.-Gesch., 121 (1959) 130-141. 14 UCHIZO~O, K., Characteristics of excitatory and inhibitory synapses in thecentral nervous system of the cat, Nature (Lond.), 207 (1965) 642-643.

Brain Research, 5 (1967) 15-30