- Email: [email protected]
Distribution of olivocerebellar fibers demonstrated by a radioautographic tracing method
Brain Research, 95 (1975) 253-263 © Elsevier Scientific Publishing Company, Amsterdam - Printed in The Netherlands
253
DISTRIBUTION OF O L I V O C E R E B E L L A R FIBERS D E M O N S T R A T E D BY A RADIOAUTOGRAPHIC TRACING METHOD
J. C O U R V I L L E
D@artement de Physiologie, Universitd de Montreal, Montreal (Canada)
SUMMARY
The method of injecting radioactive amino acids for the marking of neurons and the demonstration of their terminals by means of radioautography has been applied to the olivocerebellar projection. After injections in the olive, labeled fibers were observed in the molecular layer of the cerebellar cortex. The projections were distributed in specific regions of the cerebellum, depending on the position of the injections within the nucleus. A comparison of that topographical arrangement with Brodal's chart of olivocerebellar distribution indicates a close correspondence. In addition, the olivocerebellar climbing fibers terminate as a series of strips oriented perpendicularly to the long axes of the folia. In all regions observed so far, the strips of the projection alternate with empty spaces of the same width. It is postulated that another group of climbing fibers originating outside the olive interdigitate with the olivocerebellar fibers.
INTRODUCTION
The topographical organization of the projection from the inferior olive to the cerebellum was first studied in detail by Brodal 4. He showed, by means of retrograde chromatolysis, that the different regions of the inferior olive of the cat send their fibers to restricted portions of the cerebellar cortical surface. It was also concluded that all cells of the olive send fibers to the cerebellum and that all parts of the cerebellar cortex and nuclei receive olivary fibers. Apart from rare exceptions, the projections are crossed. The method of retrograde chromatolysis is eminently suited for studying the arrangement since controlled ablations of the cerebellar cortex are easier to obtain than localized lesions in the inferior olive. However, it would be interesting to confirm this pattern of distribution of the olivocerebellar projection by using another method which permits direct visualization of the fibers.
254 There are few recent studies in which the olivocerebellar connections were studied experimentally. Following lesions of the inferior olive, Szent/tgothai and Rajkovits is showed degenerated fibers in the granular layer of the cerebellar cortex and also identified a small number of degenerated fragments in the deep parts of the molecular layer. These authors concluded that the olivocerebellar fibers terminate as climbing fibers, but stressed the difficulty of staining their terminal portions in the molecular layer. Grant 7 also produced lesions of the olive and succeeded in demonstrating the terminal parts of the olivocerebellar fibers by the Fink-Heimer method, thus providing strong evidence for their identification as climbing fibers. Recently, Desclin 6 utilized 3-acetylpyridine to destroy the inferior olive and showed degenerated climbing fiber terminals in the molecular layer and climbing fiber collaterals in the granular layer. None of these authors, however, demonstrated a regional distribution over the cerebellar cortex, although Grant 7 indicated that small lesions of the olive do give rise to specific regions of degeneration. In the present study, small amounts of radioactive amino acids were injected into different regions of the olive allowing the projection to the cerebellar cortex to be mapped by radioautography. Data have so far only been obtained for a few regions of the olive and the distributions observed will be compared with the detailed chart of Brodal 4. In addition, the present study revealed an unsuspected feature of the distribution of climbing fibers. MATERIAL AND METHODS
In 17 cats, small injections of radioactive L-leucine or L-proline were made in the inferior olive. The injections were done with a 1 /~1 Hamilton syringe to which was sealed a glass pipette with a tip diameter of 100M 50/~m. The syringe and pipette were mounted in a modified stereotaxic holder. Volumes of 0.2-0.5 #l of the amino acids (25 #Ci/#l 0.9 ~ NaCI) were infused over periods varying between 1 and 4 h. The animals were sacrificed 2-48 h after the injection by intraventricular perfusion of 4 ~ formaldehyde. Serial paraffin sections were cut, dipped in Kodak NTB-2 emulsion and subsequently developed according to the technique of Cowan et al. 5. In each case, 2 or 3 parallel series were obtained and photographic development followed after 2, 3 and 5 weeks of exposure. Mapping of radioactivity at the site of injection and along the fiber paths was done by light microscopy using both bright-field and dark-field illuminations. RESULTS A N D D I S C U S S I O N
Pattern of distribution of olivocerebellar fibers The distributions of labeled material in the cerebella of two animals which survived for 48 h are presented. Similar results were obtained in two other animals with a 24 h survival period. The sites of injection are illustrated in two series of equally spaced sections through the inferior olive (Fig. 1). In each case, the first two sections correspond to the caudal third of the olive, sections 3 and 4 to the middle third, and
255
coudol
coudol
6
rostrQI
Fig. 1. Injections of L-leucine in the inferior olive. These two series of equally spaced sections through the lower brain stem represent levels through the caudal (1 and 2), middle (3 and 4) and rostral (5 and 6) portions of the inferior olive. In the first cat, the injection center is situated medially in the rostral third of the olive. In the second, the injection is centered laterally in the middle third of the olive. Both injections produced labeling of the central portions of the principal and accessory olives. In cat ACCM-5, radioactivity is present in the medial portions of the dorsal and medial accessory nuclei of the olive. In cat ACCM-6, cells in the lateral portions of the accessory nuclei are labeled. The black area represents total tissue destruction caused by the pipette. Hatchings surrounded by a continuous line indicate the adjacent zone where the tissue appears disrupted. Dots indicate heavily labeled cells in the olive, in adjacent structures and along the tract of the pipette.
sections 5 and 6 to the rostral third. In cat A C C M - 5 , the center o f the injection is within the rostral third o f the olive. The central portions o f the medial and dorsal accessory olives, as well as o f the dorsal and ventral lamellae o f the principal olive are labeled (Fig. 1, A C C M - 5 , levels 3-5). Subnucleus /~ and the ventrolateral outg r o w t h (not indicated on the drawings) are labeled in the middle third o f the nucleus. The other subnuclei (dorsomedial cell c o l u m n and dorsal cap) are n o t covered by the injection. In cat A C C M - 6 , the injection is centered in the middle third o f the inferior olive and somewhat laterally. Labeled cells in this case are located in the central part o f the principal and accessory subdivisions and in the lateral portions o f the accessory olives. There is also marking o f the ventrolateral region o f the principal olive (Fig. 1, A C C M - 6 , levels 2-5).
256
Fig. 2. Central zone of an injection in the inferior olive. This low power photomicrograph shows the region of maximal extent of label, following an injection of 0.5 ltl of e-leucine. This section is adjacent to level 5 in upper row of Fig. I. The hole left by the pipette is surrounded by a zone of disrupted tissue. A number of cells are heavily labeled within that zone. At the periphery, there is a ringshaped zone of very heavy deposit of radioactivity and intense labeling of the cell bodies of the olive. A rather sharp border can be seen between labeled cells and unlabeled portions of the ventrotateral part of the olive. Cells belonging to the reticular formation and to the nuclei of the rapbe are labeled at some distance from the injection center. A number of labeled fibers leaving the dorsal and medial aspects of the injection site stream towards the midline.
As shown in Fig. 2, the sites o f injection are seen as concentric zones a r o u n d the hole made by the tract of the pipette (black). There is first a roughly circular zone where the tissue is damaged (hatchings within a circle) a n d where the cells appear s h r u n k e n but still capable o f retaining the tracer, followed by a s u r r o u n d i n g region of dense deposits o f silver grains, in which the cells are intensely labeled (dots). A certain n u m b e r o f ceils o f the reticular f o r m a t i o n were marked along the passage o f the pipette a n d in the vicinity o f the olive (dots). It is assumed that only the cells o f the inferior olive labeled above b a c k g r o u n d levels give rise to t r a n s p o r t o f the tracer away from the site o f injection.
Course of the fibers F r o m the site o f injection, the labeled fibers traverse the midline as a series of
257 small bundles spread over the ventral half of the tegmentum (Fig. 3, sections on the left). A majority of fibers pass above and below the contralateral olive while some traverse this nucleus. The fibers then collect into a compact bundle underneath the spinal tract of the fifth nerve and course rostrally alongside that structure to reach the inferior cerebellar peduncle (not illustrated). The olivocerebellar fibers are spread rather uniformly within the latter structure (Fig. 3, middle section). In the white matter of the cerebellum, the fibers pass laterally to the rostral pole of the nucleus interpositus anterior. They then break into a number of wide independent fasciculi coursing towards the different sites of termination (Fig. 3, middle section of lower row).
Termination of olivocerebellarfibers In the cerebellar cortex of cat ACCM-5, accumulations of radioactive material are found in the molecular layer in lobules II-V of the anterior lobe and are restricted to the lateral portions of the rostral lobules (Fig. 3, upper row, section on the right). This distribution continues in the caudal folia of lobule V (upper row, middle section). A small amount of label is seen in lobule VI and the medial portion of crus I (upper row, middle section). In the posterior lobe, silver grains are distributed in the lateral part o f the vermian lobule VIII and over wide regions of the paramedian lobule (upper row, section on the left). There are also scattered accumulations in crus II and in the paraflocculus, but no label is present in the flocculus (middle section). In cat ACCM-6, the distribution of labeled material in the anterior lobe is also found over lobules II-V but in addition, it extends to the paravermian region (Fig. 3, i lower row, section on the right) and to lobule ! (middle section). An abundance of terminals are found in lobule VI and crus I (lower row I middle section). In the posterior lobe, a small amount o f radioactive material is distributed in the paramedian lobule. None is found in crus II and the vermis. There is a moderate amount of label in the caudal paraflocculus and in the flocculus.
Comparison with the results of Brodal The results described here, using radioautography, may now be compared i with Brodal's chart (Fig. 4) of the distribution of oli~¢ocerebellar fibers. The regions of the olive covered by injections of radioactive matbrial in part overlap in the two cases presented and in part extend to different areas. The affected areas common to both animals - - namely the central regions of the principal nucleus and of the accessory subdivisions of the inferior olive - - presun~bly project to identical regions of the cerebellum. On the other hand, the areas of distribution which are found to be different when the two cases are compared must correspond to the regions of the olive labeled exclusively in one or the other cat. The area, of label distribution which are identical in the two cats are the lateral part of th~ anterior lobe, lobule VI, the caudal regions of lobule V, crus I and the paramediai Llobule. These parts of the cerebellar cortex correspond to the distribution from he central regions of the olive indicated in Brodal's chart (Fig. 4). In cat ACCM- the portion of the ventral lamella labeled by the injection is more extensive than n cat ACCM-6. The distribution in the paramedian lobule is therefore spread over lore folia and labeled fibers are
258
ACCM" 5
%
• "-6 ACCM
Fig. 3. Distribution of label following injections in the olive. Three levels through the cerebellar cortex and brain stem of two experimental cases are shown. In the caudal sections on the left, the sites of injection in the olive appear as black patches. Wavy lines indicate the spread of olivocerebellar fibers coursing in the brain stem. Some of the fibers cross directly through the olive of the opposite side. The middle section shows the distribution of fibers in the inferior cerebellar peduncle and in the white matter of the cerebellum. A distribution of these fibers in a series of vertical bands is apparent in the middle section of the lower row. Abundant deposits of labeled material are seen in the molecular layer of the cortex, contralaterally to the lesion. Distributions in the lateral portion of the lobules of the anterior lobe, in lobule VI, crus 1, and the paraflocculus are seen in both cases, lncat ACCM-5, there are deposits of label in the posterior vermis, crus II, and an abundant distribution in the paramedian lobule. In cat ACCM-6, few silver grains are found in the paramedian lobule and there are distributions in lobule I and the paramedian region of the anterior lobe. Over the anterior lobe, the distribution takes the form of sagittal bands which are aligned from folium to folium. In the paramedian lobule, the bands are obliquely aligned and they are nearly horizontal in crus I1 (upper row, left section). The bands of labeled material alternate with empty spaces having approximately the same width in most regions. The lobules are identified according to the nomenclature of Larsell 8. present in crus I! a n d in the paraflocculus. In cat A C C M - 5 , there is labeling o f the medial portions of the medial accessory olive which is n o t present in cat A C C M - 6 . According to the data o f Brodal, this region sends fibers to the caudal vermis a n d in excellent confirmation, in cat A C C M - 5 , a d i s t r i b u t i o n of label in that region is seen. O n the other hand, only in cat A C C M - 6 does the injection involve the lateral p o r t i o n s of the medial a n d dorsal accessory olives. The presence of radioactivity over the paravermal zone o f the anterior lobe a n d in the flocculus thus also agrees with the dist r i b u t i o n expected from these regions according to the chart o f Brodal. The two cases presented here (and similar results in two other a n i m a l s with injections in the central region o f the olive) do not permit a detailed c o m p a r i s o n with the data o f Brodal. M a n y parts o f the olive have yet to be explored. The injections
259
I
couo'o/
II
IlI
eat
I)
.
~-. .'~
Y
Xl~ .
/')
•
x1v
V
X
xv
rostra/ Fig. 4. Topographical organization of the olivocerebellar projection. This chart summarizes the data obtained by Brodal 4 by the method of selective lesions of the cerebellar cortex in kittens and mapping of the cells of the inferior olive undergoing retrograde chromatolysis. Various symbols identify the areas of the olive which project to the cortical regions. The distributions encountered in the cerebellar cortex following injections in the olive correspond to this arrangement (see text). m a d e so far, being relatively large, cover m a n y regions o f the olive a n d consequently, there are fibers d i s t r i b u t e d to a n u m b e r o f cerebellar lobules. H o w e v e r , it is possible to c o n c l u d e at this stage t h a t the o r g a n i z a t i o n revealed with the m e t h o d o f r a d i o active a m i n o acids c o r r e s p o n d s very closely to the t o p o g r a p h i c a l d i s t r i b u t i o n d e m -
260 onstrated by retrograde chromatolysis. By injecting small volumes of tracer in various parts of the olive, it might be possible to establish a correspondence with the chart for all lobules and perhaps to obtain further data on this organization, especially with regard to the deeply situated folia. No projection to the cerebellar nuclei was reported in the present account and this point deserves a comment. Some of the regions projecting to the nuclei according to Brodal 4, e.g., dorsomedial region of the olive in the caudal third, were not involved by the injections. However, the dorsal lamella and the ventrolateral outgrowth have been labeled but no distribution has been found in the dentate nucleus. This negative finding may be due to the fact that the collaterals to the nuclei are presumably smaller than the main axons and therefore contain less radioactivity. Possibly, these small fibers will be demonstrated by lengthening the survival period and exposure time. A similar explanation must be given for the paradoxical finding that few labeled fibers are seen in the granular layer where the climbing fibers obviously have to pass to reach their sites of termination. It is known that the climbing fibers emit collaterals in the granular layer 17 and this perhaps causes a reduction of radioactivity in individual branches below the threshold level needed for the radioautographic reaction. With respect to the intense accumulations which were demonstrated in the molecular layer, this suggests that there is a selective accumulation of labeled material in terminal or preterminal structures.
D&tribution of the climbing fiber termina& in sagittal bands The present study has revealed for the first time, a very peculiar pattern of termination for the olivocerebellar fibers. In sections cut in the transverse plane (Figs. 3 and 5), the climbing fibers terminate in sagittal bands alternating with empty spaces. The width of the bands varies but, as a rule, the empty spaces appear to be of the same width as the labeled bands. Figs. 3 and 5 also demonstrate that the bands are often aligned correspondingly from folium to folium. In the anterior lobe, the strips have a longitudinal orientation but in the paramedian lobule they are distributed obliquely and in crus II they are nearly horizontal. The band pattern is less apparent in certain places (e.g. crus l, paraflocculus), although the distribution still appears to form clusters. In the anterior lobe of the cerebellum, the bands of projection and the void spaces occur with a certain regularity. The width of the average band is around 0.4-0.8 mm and therefore corresponds to 7 15 Purkinje cell layers in the mediolateral direction. In the sagittal plane, the bands extend over many sections. For instance, many of the bands in Fig. 3 extend throughout the anterior lobe. Many of the features of the olivoeerebellar projection demonstrated here, such as the termination in the molecular layer, the topographical distribution and the orientation of the bands across the axes of the folia, correspond to the characteristics expected for the climbing fibers. However, the presence of empty bands separating the strips of olivocerebellar fibers is a new information for which there is at present no obvious explanation. One may consider whether certain Purkinje cells are not innervated at all by climbing fibers. There is no information available on this point in the literature although it must be pointed out that negative findings such as an
261
Fig. 5. Distribution of labeled olivocerebellar fibers. This photographic montage illustrates an area of termination at the junction of two cerebellar folia of lobules IV and V, in a transverse section of the cerebellar cortex. The silver grains are concentrated in the molecular layer and few fibers are seen in the granular layer. The grains form a series of bands which are aligned in the two folia. The labeled bands alternate with unlabeled bands having the same width. Within a labeled band, linear accumulations oriented perpendicularly to the surface occur with a certain regularity and presumably correspond to the position of the dendritic trees of Purkinje cells. The amount of background radioactivity can be judged from the appearance of the space surrounding the folia. Dark-field illumination.
absence of climbing fibers ~in Golgi stained material or an absence of climbing fiber responses following stimulation of an afferent system to the cerebellum, would hitherto probably not have merited much attention. But if one assumes that there are climbing fibers terminating on every Purkinje cell, the present results strongly suggest that climbing fibers terminating on the Purkinje cells between the labeled bands do not have their origin in those parts of the olive into which the injection of label was made. Possibly, these fibers arise from other regions of the olive which were not labeled in the cases obtained so far although this seems unlikely in view of the known topographical arrangement of olivocerebellar fibers. A contribution of climbing fibers from the ipsilateral olive can also be dismissed since Brodal's results and the present data show a contralateral distribution. It is therefore proposed that some climbing fibers projecting to the cerebellar cortex originate outside the inferior olive. The pontine nuclei, particularly nucleus reticularis tegmenti pontis, seem one likely source one can suggest since Sasaki 16 reported that stimulation of this nucleus evokes complex spikes in the cerebellar cortex. Abundant previous evidence exists for a sagittal organization of the climbing fibers of the cerebellum. Voogd 19 has analyzed the distribution of degenerated fibers in the white matter of the cerebellum following lesions of the olivocerebellar fibers or of the olive. He observed that the degenerated climbing fibers, which could only be visualized before their entrance in the cortical layers, are grouped in a few longi-
262 tudinal zones. Such an arrangement of the climbing fibers in the white matter was also seen in the present study. Mapping of evoked potentials and of neural activity in the cerebellar cortex following stimulations of peripheral nerves also indicated a sagittal organization of the climbing fiber pathways 9-15. Oscarsson has demonstrated two sagittal zones in the vermis and 4 others in the paravermian region of the anterior lobe. These zones appear to coincide with the number of labeled bands seen in the anterior lobe of the present material. The recent papers of Armstrong et al. 1-3 on the organization of the climbing fiber projection present convincing evidence on the sagittal distribution of these fibers. The sagittal bands extending across a number of folia which have been observed in the cases presented (Figs. 3 and 5) seem to coincide remarkably well with the data of these authors. However, as pointed out above, the present results indicate that, in addition to the sagittal organization of the olivocerebellar fibers, there is a contribution to the cerebellar cortex of climbing fibers originating outside the olive. The two types of climbing fibers interdigitate in a very precise pattern. Further studies are needed to find the source of non-olivary climbing fibers and decipher the functional significance of this arrangement. ADDENDUM
Recent evidence obtained with the method of injection of horseradish peroxidase (Brodal, unpublished) and electrophysiological mapping (Armstrong, D. M., Harvey, R. J., and Schild, R. F., J. comp. Neurol., 154 (1974) 287-302) shows that the olivocerebellar localization is more complex than originally demonstrated by Brodal 4. Cells from a given portion of the olive send fibers to two or three different cerebellar destinations. The fact that fibers originating from various regions of the olive converge onto a part of the cerebellar cortex allows for other interpretations of the present data. Among other things, it appears still possible that all the climbing fibers originate from the inferior olive.
REFERENCES 1 ARMSTRONG, D. M., HARVEY, R. J., AND SCH1LD, R. F., Spino-olivo-cerebellar pathways to the posterior lobe of the cat cerebellum, Exp. Brain Res., 18 (1973) l-18. 2 ARMSTRONG, D. M., HARVEY, R. J., AND SCHILD, R. F., Cerebello-cerebellar responses mediated via climbing fibres, Exp. Brain Res., 18 (1973) 19-39. 3 ARMSTRONG, D. M., HARVEY, R. J., AND SCH1LD, R. F., The spatial organization of climbing fibre branching in cat cerebellum, Exp. Brain Res., 18 0973) 40-58. 4 BRODAL, A., Untersuchungen fiber die olivo-cerebell/~re Lokalisation, Z. ges. NeuroL PsTchiat., 169 (1940) 1-153. 5 COWAN, W. M., GOTTLIEB, D. 1., HENDRICKSON, A. E., PRICE, J. L., AND WOOLSEY, T. A., The autoradiographic demonstration of axonal connections in the central nervous system, Brain Research, 37 (1972) 21-51. 6 DESCLIN, J. C., Histological evidence supporting the inferior olive as the major source of cerebcllar climbing fibers in the rat, Brain Research, 77 (1974) 365-384. 7 GRANT, G., Demonstration of degenerating climbing fibers in the molecular layer of the cerebellum, Brain Research, 22 (1970) 236-242. 8 LARSELL, D., In J. JANSEN (Ed.), The Comparative Anatomy and Histology of the Cerebellum front Monotremes through Apes, University of Minnesota Press, Minneapolis, 1970, pp. 175-185.
263 9 LARSON,B., MILLER, S., AND OSCARS.SON,O., Spino-olivo-cerebeUar pathway ascending in the dorsolateral funiculus of the spinal cord, Acta physiol, scand., Suppl. 310 (1968) 37A-38A. 10 LARSON,B., MILLER, S., AND OSCARS,qON,O., Termination and functional organization of the dorso-lateral spino-olivocerebellar path, J. Physiol. (Lond.), 203 (1969) 611-640. 11 MILLER,S., NEZLINA,N., AND O$CARSSON,O., Projection and convergence patterns in climbing fibre paths to the cerebellar anterior lobe activated from the cerebral cortex and the spinal cord, Brain Research, 14 (1969) 230-233. 12 MILLER, S., NEZLINA,N., AND OSCARS,SON,O., Climbing fibre projection to cerebellar anterior lobe activated from structures in midbrain and from spinal cord, Brain Research, 14 (1969) 234-236. 13 OSCARSSON,O., Termination and functional organization of the ventral spino-olivo-cerebellar path, J. Physiol. (Lond.), 196 (1968) 453-478. 14 OSCARS,SON,O., Termination and functional organization of the dorsal spino-olivo-cerebellar path, J. Physiol. (Lond.), 200 (1969) 129-149. 15 OSCARSSON,O., The sagittal organization of the cerebellar anterior lobe as revealed by the projection patterns of the climbing fiber system. In R. LLINAS (Ed.), Neurobiology o f Cerebellar Evolution and Development, American Medical Association, Chicago, I11., 1969, pp. 525-537. 16 SASAKI,K., Mossy fiber and climbing fiber responses evoked in the cerebellar cortex by porltine stimulation. In R. LLIN,~S(Ed.), Neurobiology of Cerebellar Evolution and Development, American Medical Association, Chicago, Ill., 1969, pp. 629-638. 17 SCHEIBEL,M. E., ANDSCHEIBEL,A. B., Observations on the intracortical relations of the climbing fibers of the cerebellum, J. comp. Neurol., 101 (1954) 733-760. 18 SZENT.~,GOTHAI,J., UND RAJKOVITS,K., ~ber den Ursprung der Kletterfasern des Kleinhirns, Z. Anat. Entwickl.-Gesch., 121 (1959) 130-141. 19 VOOGD,J., The importance of fiber connections in the comparative anatomy of the mammalian cerebellum. In R. LLINAS(Ed.), Neurobiology of Cerebellar Evolution and Development, American Medical Association, Chicago, Ill., 1969, pp. 493-514.
253
DISTRIBUTION OF O L I V O C E R E B E L L A R FIBERS D E M O N S T R A T E D BY A RADIOAUTOGRAPHIC TRACING METHOD
J. C O U R V I L L E
D@artement de Physiologie, Universitd de Montreal, Montreal (Canada)
SUMMARY
The method of injecting radioactive amino acids for the marking of neurons and the demonstration of their terminals by means of radioautography has been applied to the olivocerebellar projection. After injections in the olive, labeled fibers were observed in the molecular layer of the cerebellar cortex. The projections were distributed in specific regions of the cerebellum, depending on the position of the injections within the nucleus. A comparison of that topographical arrangement with Brodal's chart of olivocerebellar distribution indicates a close correspondence. In addition, the olivocerebellar climbing fibers terminate as a series of strips oriented perpendicularly to the long axes of the folia. In all regions observed so far, the strips of the projection alternate with empty spaces of the same width. It is postulated that another group of climbing fibers originating outside the olive interdigitate with the olivocerebellar fibers.
INTRODUCTION
The topographical organization of the projection from the inferior olive to the cerebellum was first studied in detail by Brodal 4. He showed, by means of retrograde chromatolysis, that the different regions of the inferior olive of the cat send their fibers to restricted portions of the cerebellar cortical surface. It was also concluded that all cells of the olive send fibers to the cerebellum and that all parts of the cerebellar cortex and nuclei receive olivary fibers. Apart from rare exceptions, the projections are crossed. The method of retrograde chromatolysis is eminently suited for studying the arrangement since controlled ablations of the cerebellar cortex are easier to obtain than localized lesions in the inferior olive. However, it would be interesting to confirm this pattern of distribution of the olivocerebellar projection by using another method which permits direct visualization of the fibers.
254 There are few recent studies in which the olivocerebellar connections were studied experimentally. Following lesions of the inferior olive, Szent/tgothai and Rajkovits is showed degenerated fibers in the granular layer of the cerebellar cortex and also identified a small number of degenerated fragments in the deep parts of the molecular layer. These authors concluded that the olivocerebellar fibers terminate as climbing fibers, but stressed the difficulty of staining their terminal portions in the molecular layer. Grant 7 also produced lesions of the olive and succeeded in demonstrating the terminal parts of the olivocerebellar fibers by the Fink-Heimer method, thus providing strong evidence for their identification as climbing fibers. Recently, Desclin 6 utilized 3-acetylpyridine to destroy the inferior olive and showed degenerated climbing fiber terminals in the molecular layer and climbing fiber collaterals in the granular layer. None of these authors, however, demonstrated a regional distribution over the cerebellar cortex, although Grant 7 indicated that small lesions of the olive do give rise to specific regions of degeneration. In the present study, small amounts of radioactive amino acids were injected into different regions of the olive allowing the projection to the cerebellar cortex to be mapped by radioautography. Data have so far only been obtained for a few regions of the olive and the distributions observed will be compared with the detailed chart of Brodal 4. In addition, the present study revealed an unsuspected feature of the distribution of climbing fibers. MATERIAL AND METHODS
In 17 cats, small injections of radioactive L-leucine or L-proline were made in the inferior olive. The injections were done with a 1 /~1 Hamilton syringe to which was sealed a glass pipette with a tip diameter of 100M 50/~m. The syringe and pipette were mounted in a modified stereotaxic holder. Volumes of 0.2-0.5 #l of the amino acids (25 #Ci/#l 0.9 ~ NaCI) were infused over periods varying between 1 and 4 h. The animals were sacrificed 2-48 h after the injection by intraventricular perfusion of 4 ~ formaldehyde. Serial paraffin sections were cut, dipped in Kodak NTB-2 emulsion and subsequently developed according to the technique of Cowan et al. 5. In each case, 2 or 3 parallel series were obtained and photographic development followed after 2, 3 and 5 weeks of exposure. Mapping of radioactivity at the site of injection and along the fiber paths was done by light microscopy using both bright-field and dark-field illuminations. RESULTS A N D D I S C U S S I O N
Pattern of distribution of olivocerebellar fibers The distributions of labeled material in the cerebella of two animals which survived for 48 h are presented. Similar results were obtained in two other animals with a 24 h survival period. The sites of injection are illustrated in two series of equally spaced sections through the inferior olive (Fig. 1). In each case, the first two sections correspond to the caudal third of the olive, sections 3 and 4 to the middle third, and
255
coudol
coudol
6
rostrQI
Fig. 1. Injections of L-leucine in the inferior olive. These two series of equally spaced sections through the lower brain stem represent levels through the caudal (1 and 2), middle (3 and 4) and rostral (5 and 6) portions of the inferior olive. In the first cat, the injection center is situated medially in the rostral third of the olive. In the second, the injection is centered laterally in the middle third of the olive. Both injections produced labeling of the central portions of the principal and accessory olives. In cat ACCM-5, radioactivity is present in the medial portions of the dorsal and medial accessory nuclei of the olive. In cat ACCM-6, cells in the lateral portions of the accessory nuclei are labeled. The black area represents total tissue destruction caused by the pipette. Hatchings surrounded by a continuous line indicate the adjacent zone where the tissue appears disrupted. Dots indicate heavily labeled cells in the olive, in adjacent structures and along the tract of the pipette.
sections 5 and 6 to the rostral third. In cat A C C M - 5 , the center o f the injection is within the rostral third o f the olive. The central portions o f the medial and dorsal accessory olives, as well as o f the dorsal and ventral lamellae o f the principal olive are labeled (Fig. 1, A C C M - 5 , levels 3-5). Subnucleus /~ and the ventrolateral outg r o w t h (not indicated on the drawings) are labeled in the middle third o f the nucleus. The other subnuclei (dorsomedial cell c o l u m n and dorsal cap) are n o t covered by the injection. In cat A C C M - 6 , the injection is centered in the middle third o f the inferior olive and somewhat laterally. Labeled cells in this case are located in the central part o f the principal and accessory subdivisions and in the lateral portions o f the accessory olives. There is also marking o f the ventrolateral region o f the principal olive (Fig. 1, A C C M - 6 , levels 2-5).
256
Fig. 2. Central zone of an injection in the inferior olive. This low power photomicrograph shows the region of maximal extent of label, following an injection of 0.5 ltl of e-leucine. This section is adjacent to level 5 in upper row of Fig. I. The hole left by the pipette is surrounded by a zone of disrupted tissue. A number of cells are heavily labeled within that zone. At the periphery, there is a ringshaped zone of very heavy deposit of radioactivity and intense labeling of the cell bodies of the olive. A rather sharp border can be seen between labeled cells and unlabeled portions of the ventrotateral part of the olive. Cells belonging to the reticular formation and to the nuclei of the rapbe are labeled at some distance from the injection center. A number of labeled fibers leaving the dorsal and medial aspects of the injection site stream towards the midline.
As shown in Fig. 2, the sites o f injection are seen as concentric zones a r o u n d the hole made by the tract of the pipette (black). There is first a roughly circular zone where the tissue is damaged (hatchings within a circle) a n d where the cells appear s h r u n k e n but still capable o f retaining the tracer, followed by a s u r r o u n d i n g region of dense deposits o f silver grains, in which the cells are intensely labeled (dots). A certain n u m b e r o f ceils o f the reticular f o r m a t i o n were marked along the passage o f the pipette a n d in the vicinity o f the olive (dots). It is assumed that only the cells o f the inferior olive labeled above b a c k g r o u n d levels give rise to t r a n s p o r t o f the tracer away from the site o f injection.
Course of the fibers F r o m the site o f injection, the labeled fibers traverse the midline as a series of
257 small bundles spread over the ventral half of the tegmentum (Fig. 3, sections on the left). A majority of fibers pass above and below the contralateral olive while some traverse this nucleus. The fibers then collect into a compact bundle underneath the spinal tract of the fifth nerve and course rostrally alongside that structure to reach the inferior cerebellar peduncle (not illustrated). The olivocerebellar fibers are spread rather uniformly within the latter structure (Fig. 3, middle section). In the white matter of the cerebellum, the fibers pass laterally to the rostral pole of the nucleus interpositus anterior. They then break into a number of wide independent fasciculi coursing towards the different sites of termination (Fig. 3, middle section of lower row).
Termination of olivocerebellarfibers In the cerebellar cortex of cat ACCM-5, accumulations of radioactive material are found in the molecular layer in lobules II-V of the anterior lobe and are restricted to the lateral portions of the rostral lobules (Fig. 3, upper row, section on the right). This distribution continues in the caudal folia of lobule V (upper row, middle section). A small amount of label is seen in lobule VI and the medial portion of crus I (upper row, middle section). In the posterior lobe, silver grains are distributed in the lateral part o f the vermian lobule VIII and over wide regions of the paramedian lobule (upper row, section on the left). There are also scattered accumulations in crus II and in the paraflocculus, but no label is present in the flocculus (middle section). In cat ACCM-6, the distribution of labeled material in the anterior lobe is also found over lobules II-V but in addition, it extends to the paravermian region (Fig. 3, i lower row, section on the right) and to lobule ! (middle section). An abundance of terminals are found in lobule VI and crus I (lower row I middle section). In the posterior lobe, a small amount o f radioactive material is distributed in the paramedian lobule. None is found in crus II and the vermis. There is a moderate amount of label in the caudal paraflocculus and in the flocculus.
Comparison with the results of Brodal The results described here, using radioautography, may now be compared i with Brodal's chart (Fig. 4) of the distribution of oli~¢ocerebellar fibers. The regions of the olive covered by injections of radioactive matbrial in part overlap in the two cases presented and in part extend to different areas. The affected areas common to both animals - - namely the central regions of the principal nucleus and of the accessory subdivisions of the inferior olive - - presun~bly project to identical regions of the cerebellum. On the other hand, the areas of distribution which are found to be different when the two cases are compared must correspond to the regions of the olive labeled exclusively in one or the other cat. The area, of label distribution which are identical in the two cats are the lateral part of th~ anterior lobe, lobule VI, the caudal regions of lobule V, crus I and the paramediai Llobule. These parts of the cerebellar cortex correspond to the distribution from he central regions of the olive indicated in Brodal's chart (Fig. 4). In cat ACCM- the portion of the ventral lamella labeled by the injection is more extensive than n cat ACCM-6. The distribution in the paramedian lobule is therefore spread over lore folia and labeled fibers are
258
ACCM" 5
%
• "-6 ACCM
Fig. 3. Distribution of label following injections in the olive. Three levels through the cerebellar cortex and brain stem of two experimental cases are shown. In the caudal sections on the left, the sites of injection in the olive appear as black patches. Wavy lines indicate the spread of olivocerebellar fibers coursing in the brain stem. Some of the fibers cross directly through the olive of the opposite side. The middle section shows the distribution of fibers in the inferior cerebellar peduncle and in the white matter of the cerebellum. A distribution of these fibers in a series of vertical bands is apparent in the middle section of the lower row. Abundant deposits of labeled material are seen in the molecular layer of the cortex, contralaterally to the lesion. Distributions in the lateral portion of the lobules of the anterior lobe, in lobule VI, crus 1, and the paraflocculus are seen in both cases, lncat ACCM-5, there are deposits of label in the posterior vermis, crus II, and an abundant distribution in the paramedian lobule. In cat ACCM-6, few silver grains are found in the paramedian lobule and there are distributions in lobule I and the paramedian region of the anterior lobe. Over the anterior lobe, the distribution takes the form of sagittal bands which are aligned from folium to folium. In the paramedian lobule, the bands are obliquely aligned and they are nearly horizontal in crus I1 (upper row, left section). The bands of labeled material alternate with empty spaces having approximately the same width in most regions. The lobules are identified according to the nomenclature of Larsell 8. present in crus I! a n d in the paraflocculus. In cat A C C M - 5 , there is labeling o f the medial portions of the medial accessory olive which is n o t present in cat A C C M - 6 . According to the data o f Brodal, this region sends fibers to the caudal vermis a n d in excellent confirmation, in cat A C C M - 5 , a d i s t r i b u t i o n of label in that region is seen. O n the other hand, only in cat A C C M - 6 does the injection involve the lateral p o r t i o n s of the medial a n d dorsal accessory olives. The presence of radioactivity over the paravermal zone o f the anterior lobe a n d in the flocculus thus also agrees with the dist r i b u t i o n expected from these regions according to the chart o f Brodal. The two cases presented here (and similar results in two other a n i m a l s with injections in the central region o f the olive) do not permit a detailed c o m p a r i s o n with the data o f Brodal. M a n y parts o f the olive have yet to be explored. The injections
259
I
couo'o/
II
IlI
eat
I)
.
~-. .'~
Y
Xl~ .
/')
•
x1v
V
X
xv
rostra/ Fig. 4. Topographical organization of the olivocerebellar projection. This chart summarizes the data obtained by Brodal 4 by the method of selective lesions of the cerebellar cortex in kittens and mapping of the cells of the inferior olive undergoing retrograde chromatolysis. Various symbols identify the areas of the olive which project to the cortical regions. The distributions encountered in the cerebellar cortex following injections in the olive correspond to this arrangement (see text). m a d e so far, being relatively large, cover m a n y regions o f the olive a n d consequently, there are fibers d i s t r i b u t e d to a n u m b e r o f cerebellar lobules. H o w e v e r , it is possible to c o n c l u d e at this stage t h a t the o r g a n i z a t i o n revealed with the m e t h o d o f r a d i o active a m i n o acids c o r r e s p o n d s very closely to the t o p o g r a p h i c a l d i s t r i b u t i o n d e m -
260 onstrated by retrograde chromatolysis. By injecting small volumes of tracer in various parts of the olive, it might be possible to establish a correspondence with the chart for all lobules and perhaps to obtain further data on this organization, especially with regard to the deeply situated folia. No projection to the cerebellar nuclei was reported in the present account and this point deserves a comment. Some of the regions projecting to the nuclei according to Brodal 4, e.g., dorsomedial region of the olive in the caudal third, were not involved by the injections. However, the dorsal lamella and the ventrolateral outgrowth have been labeled but no distribution has been found in the dentate nucleus. This negative finding may be due to the fact that the collaterals to the nuclei are presumably smaller than the main axons and therefore contain less radioactivity. Possibly, these small fibers will be demonstrated by lengthening the survival period and exposure time. A similar explanation must be given for the paradoxical finding that few labeled fibers are seen in the granular layer where the climbing fibers obviously have to pass to reach their sites of termination. It is known that the climbing fibers emit collaterals in the granular layer 17 and this perhaps causes a reduction of radioactivity in individual branches below the threshold level needed for the radioautographic reaction. With respect to the intense accumulations which were demonstrated in the molecular layer, this suggests that there is a selective accumulation of labeled material in terminal or preterminal structures.
D&tribution of the climbing fiber termina& in sagittal bands The present study has revealed for the first time, a very peculiar pattern of termination for the olivocerebellar fibers. In sections cut in the transverse plane (Figs. 3 and 5), the climbing fibers terminate in sagittal bands alternating with empty spaces. The width of the bands varies but, as a rule, the empty spaces appear to be of the same width as the labeled bands. Figs. 3 and 5 also demonstrate that the bands are often aligned correspondingly from folium to folium. In the anterior lobe, the strips have a longitudinal orientation but in the paramedian lobule they are distributed obliquely and in crus II they are nearly horizontal. The band pattern is less apparent in certain places (e.g. crus l, paraflocculus), although the distribution still appears to form clusters. In the anterior lobe of the cerebellum, the bands of projection and the void spaces occur with a certain regularity. The width of the average band is around 0.4-0.8 mm and therefore corresponds to 7 15 Purkinje cell layers in the mediolateral direction. In the sagittal plane, the bands extend over many sections. For instance, many of the bands in Fig. 3 extend throughout the anterior lobe. Many of the features of the olivoeerebellar projection demonstrated here, such as the termination in the molecular layer, the topographical distribution and the orientation of the bands across the axes of the folia, correspond to the characteristics expected for the climbing fibers. However, the presence of empty bands separating the strips of olivocerebellar fibers is a new information for which there is at present no obvious explanation. One may consider whether certain Purkinje cells are not innervated at all by climbing fibers. There is no information available on this point in the literature although it must be pointed out that negative findings such as an
261
Fig. 5. Distribution of labeled olivocerebellar fibers. This photographic montage illustrates an area of termination at the junction of two cerebellar folia of lobules IV and V, in a transverse section of the cerebellar cortex. The silver grains are concentrated in the molecular layer and few fibers are seen in the granular layer. The grains form a series of bands which are aligned in the two folia. The labeled bands alternate with unlabeled bands having the same width. Within a labeled band, linear accumulations oriented perpendicularly to the surface occur with a certain regularity and presumably correspond to the position of the dendritic trees of Purkinje cells. The amount of background radioactivity can be judged from the appearance of the space surrounding the folia. Dark-field illumination.
absence of climbing fibers ~in Golgi stained material or an absence of climbing fiber responses following stimulation of an afferent system to the cerebellum, would hitherto probably not have merited much attention. But if one assumes that there are climbing fibers terminating on every Purkinje cell, the present results strongly suggest that climbing fibers terminating on the Purkinje cells between the labeled bands do not have their origin in those parts of the olive into which the injection of label was made. Possibly, these fibers arise from other regions of the olive which were not labeled in the cases obtained so far although this seems unlikely in view of the known topographical arrangement of olivocerebellar fibers. A contribution of climbing fibers from the ipsilateral olive can also be dismissed since Brodal's results and the present data show a contralateral distribution. It is therefore proposed that some climbing fibers projecting to the cerebellar cortex originate outside the inferior olive. The pontine nuclei, particularly nucleus reticularis tegmenti pontis, seem one likely source one can suggest since Sasaki 16 reported that stimulation of this nucleus evokes complex spikes in the cerebellar cortex. Abundant previous evidence exists for a sagittal organization of the climbing fibers of the cerebellum. Voogd 19 has analyzed the distribution of degenerated fibers in the white matter of the cerebellum following lesions of the olivocerebellar fibers or of the olive. He observed that the degenerated climbing fibers, which could only be visualized before their entrance in the cortical layers, are grouped in a few longi-
262 tudinal zones. Such an arrangement of the climbing fibers in the white matter was also seen in the present study. Mapping of evoked potentials and of neural activity in the cerebellar cortex following stimulations of peripheral nerves also indicated a sagittal organization of the climbing fiber pathways 9-15. Oscarsson has demonstrated two sagittal zones in the vermis and 4 others in the paravermian region of the anterior lobe. These zones appear to coincide with the number of labeled bands seen in the anterior lobe of the present material. The recent papers of Armstrong et al. 1-3 on the organization of the climbing fiber projection present convincing evidence on the sagittal distribution of these fibers. The sagittal bands extending across a number of folia which have been observed in the cases presented (Figs. 3 and 5) seem to coincide remarkably well with the data of these authors. However, as pointed out above, the present results indicate that, in addition to the sagittal organization of the olivocerebellar fibers, there is a contribution to the cerebellar cortex of climbing fibers originating outside the olive. The two types of climbing fibers interdigitate in a very precise pattern. Further studies are needed to find the source of non-olivary climbing fibers and decipher the functional significance of this arrangement. ADDENDUM
Recent evidence obtained with the method of injection of horseradish peroxidase (Brodal, unpublished) and electrophysiological mapping (Armstrong, D. M., Harvey, R. J., and Schild, R. F., J. comp. Neurol., 154 (1974) 287-302) shows that the olivocerebellar localization is more complex than originally demonstrated by Brodal 4. Cells from a given portion of the olive send fibers to two or three different cerebellar destinations. The fact that fibers originating from various regions of the olive converge onto a part of the cerebellar cortex allows for other interpretations of the present data. Among other things, it appears still possible that all the climbing fibers originate from the inferior olive.
REFERENCES 1 ARMSTRONG, D. M., HARVEY, R. J., AND SCH1LD, R. F., Spino-olivo-cerebellar pathways to the posterior lobe of the cat cerebellum, Exp. Brain Res., 18 (1973) l-18. 2 ARMSTRONG, D. M., HARVEY, R. J., AND SCHILD, R. F., Cerebello-cerebellar responses mediated via climbing fibres, Exp. Brain Res., 18 (1973) 19-39. 3 ARMSTRONG, D. M., HARVEY, R. J., AND SCH1LD, R. F., The spatial organization of climbing fibre branching in cat cerebellum, Exp. Brain Res., 18 0973) 40-58. 4 BRODAL, A., Untersuchungen fiber die olivo-cerebell/~re Lokalisation, Z. ges. NeuroL PsTchiat., 169 (1940) 1-153. 5 COWAN, W. M., GOTTLIEB, D. 1., HENDRICKSON, A. E., PRICE, J. L., AND WOOLSEY, T. A., The autoradiographic demonstration of axonal connections in the central nervous system, Brain Research, 37 (1972) 21-51. 6 DESCLIN, J. C., Histological evidence supporting the inferior olive as the major source of cerebcllar climbing fibers in the rat, Brain Research, 77 (1974) 365-384. 7 GRANT, G., Demonstration of degenerating climbing fibers in the molecular layer of the cerebellum, Brain Research, 22 (1970) 236-242. 8 LARSELL, D., In J. JANSEN (Ed.), The Comparative Anatomy and Histology of the Cerebellum front Monotremes through Apes, University of Minnesota Press, Minneapolis, 1970, pp. 175-185.
263 9 LARSON,B., MILLER, S., AND OSCARS.SON,O., Spino-olivo-cerebeUar pathway ascending in the dorsolateral funiculus of the spinal cord, Acta physiol, scand., Suppl. 310 (1968) 37A-38A. 10 LARSON,B., MILLER, S., AND OSCARS,qON,O., Termination and functional organization of the dorso-lateral spino-olivocerebellar path, J. Physiol. (Lond.), 203 (1969) 611-640. 11 MILLER,S., NEZLINA,N., AND O$CARSSON,O., Projection and convergence patterns in climbing fibre paths to the cerebellar anterior lobe activated from the cerebral cortex and the spinal cord, Brain Research, 14 (1969) 230-233. 12 MILLER, S., NEZLINA,N., AND OSCARS,SON,O., Climbing fibre projection to cerebellar anterior lobe activated from structures in midbrain and from spinal cord, Brain Research, 14 (1969) 234-236. 13 OSCARSSON,O., Termination and functional organization of the ventral spino-olivo-cerebellar path, J. Physiol. (Lond.), 196 (1968) 453-478. 14 OSCARS,SON,O., Termination and functional organization of the dorsal spino-olivo-cerebellar path, J. Physiol. (Lond.), 200 (1969) 129-149. 15 OSCARSSON,O., The sagittal organization of the cerebellar anterior lobe as revealed by the projection patterns of the climbing fiber system. In R. LLINAS (Ed.), Neurobiology o f Cerebellar Evolution and Development, American Medical Association, Chicago, I11., 1969, pp. 525-537. 16 SASAKI,K., Mossy fiber and climbing fiber responses evoked in the cerebellar cortex by porltine stimulation. In R. LLIN,~S(Ed.), Neurobiology of Cerebellar Evolution and Development, American Medical Association, Chicago, Ill., 1969, pp. 629-638. 17 SCHEIBEL,M. E., ANDSCHEIBEL,A. B., Observations on the intracortical relations of the climbing fibers of the cerebellum, J. comp. Neurol., 101 (1954) 733-760. 18 SZENT.~,GOTHAI,J., UND RAJKOVITS,K., ~ber den Ursprung der Kletterfasern des Kleinhirns, Z. Anat. Entwickl.-Gesch., 121 (1959) 130-141. 19 VOOGD,J., The importance of fiber connections in the comparative anatomy of the mammalian cerebellum. In R. LLINAS(Ed.), Neurobiology of Cerebellar Evolution and Development, American Medical Association, Chicago, Ill., 1969, pp. 493-514.