The Olivocerebellar Projection in the Cat Studied with the Method of Retrograde Axonal Transport of Horse radish Peroxid ase ALF BRODAL, FRED WALBERG AND GRETHE H . HODDEVIK Anatomical Institute, University of Oslo, Oslo I , N o w a y
ABSTRACT The distribution of labeled cells in the inferior olive of the cat has been mapped following injections of small amounts of horseradish peroxidase in the paramedian lobule of the cerebellum. The distribution of labeled cells was plotted in drawings of approximately serial transverse sections. The findings in each case were transferred to a standard diagram of the olive to facilitate comparison of cases. Previous studies of the distribution of retrograde cell loss in the inferior olive following cerebellar lesions (Brodal, '40b) showed that fibers ending in the paramedian lobule come from the caudal part of the ventral lamella of the principal olive. This was confirmed with the peroxidase method, but in addition three other separate and well circumscribed areas of the olive showed labeling: one in the dorsal accessory olive, another in the rostral part of the medial accessory olive, a third in the caudal part of the dorsal lamella of the principal olive (fig. 7). There is some degree of topical arrangement within the projection of each of these olivary areas to the paramedian lobule. It is particularly striking that the projection areas of the caudal one-third of the lobule are different from and overlap only little with those of the rostral two-thirds. On account of diffusion of the injected peroxidase solution in the folia it could not be decided whether the different olivary areas project to particular longitudinal zones in the paramedian lobule. The main findings can be correlated with the physiological observations of Armstrong et al. ('74). Some of the "paramedian" olivary areas are labeled also following peroxidase injections in other cerebellar parts, among them the nuclei interpositus anterior and posterior. The findings are compatible with the notion that olivocerebellar fibers branch to supply more than one cerebellar region. It is confirmed that the olivocerebellar projection, including that of the nuclei, is almost completely crossed. In the DISCUSSION it is emphasized that afferents from several sources converge on all four olivary regions projecting onto the paramedian lobule. The olivocerebellar projection obviously allows for divergence as well as convergence of impulses from the olive to the c e r e bellum. For further insight into the anatomical organization of the inferior olive, the entire olivocerebellar projection has to be mapped with the peroxidase method, and further studies of the afferents to the olive are needed. I n such studies, as well as in physiological ones, it is essential that findings are d e scribed with meticulous reference to the topography of the olivary subdivisions.
Studies of the distribution of the cell loss which occurs in the inferior olive following circumscribed lesions of the cerebellum in newly born cats and rabbits (modified Gudden method, Brodal, '40a) indicate that there exists a clearcut olivocerebellar localization (Brodal, '40b). Each lobule of the cerebellum appears to receive fibers J. COMP. NEUR., 164: 449470.
from a particular-region of the olivary complex (for further references see Jansen and Brodal, '58). Recent electrophysiological studies indicate, however, that the pattern in the olivocerebellar projection is far more complex than appears from the map constructed by Brodal ('40b) and reproduced in 449
450
ALF BRODAL. FRED WALBERG AND GRETHE H. HODDEVIK
figure 1. VanGilder and O'Leary ('70) following electrical stimulation of groups of olivary neurons studied the distribution within the cerebellum of the ensuing cortical potentials. Their results differ on certain points from those of Brodal ('40b). For example, the paramedian lobule was influenced from the medial accessory olive as well as from the principal olive. Further electrophysiological studies indicate that climbing fibers (assumed to be derived from the inferior olive) branch b e fore they reach the cerebellar cortex (Faber and Murphy, '69) and that a single climb ing fiber may give one branch to the paramedian lobule, another to the intermediate part of the anterior lobe (Armstrong et al., '71a,b; Cooke et al., '72). From antidromic recording from cells in the inferior olive following cerebellar stimulation Armstrong et al. ('73, '74) concluded that in many regions of the olive there are cells with branching axons which project to more than one area in the cerebellum, for example to lobules in the anterior as well as the posterior lobe. Further, a particular cerebellar lobule may receive fibers from several different parts of the olivary complex. When plotted in a diagram as shown in Armstrong et al.'s ('74) figure 5, a very complex pattern of the olivocerebellar projection emerges. Branching of cerebellar climbing fibers has so far not been observed anatomically, except for some fibers which branch within one folium (Scheibel and Scheibel, '54; Fox et al., '69), and it cannot yet be considered as decided that the inferior olive is the sole source of climbing fibers. It is clear, however, that the mapping of retrograde cell loss brings out only part of the pattern of the olivocerebellar projection. A possibility to obtain a more complete and detailed map of this projection has been opened in recent years. It has been amply documented that the retrograde axonal transport of horseradish peroxidase can be used to determine the origin of nerve fibers not only in peripheral nerves (Kristensson and Olsson, '71), but in central nerve fiber systems as well (La Vail and La Vail, '72; La Vail et al., '73; Warr, '73; Ralston and Sharp, '73; Jones and Leavitt, '74; Lynch et al., '73; and several others) Preliminary experiments in our laboratory on cats as well as ex-
periments published by LaVail et al. ('73, in the mouse) and Graybiel and Devor ('74, in the rat) make i t clear that the method can be used for studies of the olivocerebellar projection. The problems related to uptake of peroxidase in the cere bellar cortex and labeling of cells in the olive are considered elsewhere (Walberg et al., '76). The present study deals with the topographical organization of a r e stricted part of the olivocerebellar projection only, that onto the paramedian lobule. It will be shown i.a. that this lobule r e ceives fibers from four regions of the olive. Abbreviations
5,
Nucleus5 (Brodal, '40b) of inferior olive D, Dorsal accessory olive d.cap., Dorsal cap d.L, Dorsal lamella of principal olive dm.c.col., Dorsomedial cell column (Mareschal,
'34) Flocc., Flocculus F.pr., Primary fissure l., Lateral Lob.ant., Anterior lobe L.pm., Lobulus paramedianus M, Medial accessory olive m., Medial NIA, Nucleus interpositus anterior NIP, Nucleus interpositus posterior NL, Nucleus lateralis (dentatus) NM, Nucleus medialis (fastigii) P, Principal olive P.fl.d., ParafIocculus dorsalis P.fl.v., Paraflocculus ventralis v.L, Ventral lamella of principal olive v.l.o., Ventrolateral outgrowth
Fig. 1 A diagram of the olivocerebellar projection in the cat a s determined o n the basis of retrograde cell loss in the olive following cerebellar lesions in the kitten. Above a diagram of the cerebellar surface imagined unfolded. In the middle a series of drawings of equally spaced thionine stained transverse sections through the olive from caudal (I) to rostra1 (XV). On the left side in this series of drawings of the olive the different subdivisions of the complex are indicated. Black: dorsal accessory olive; white: principal olive, with the ventrolateral outgrowth and dorsal cap; hatchings: medial accessory olive and its subdivisions, the nucleus fl and the dorsomedial cell column. On the right side in this series of sections through the olive corresponding regions of the cerebellum and of the olive are marked with identical symbols. Below a graphic reconstruction of the olivary complex as it will appear when imagined unfolded i n one plane by pulling the lateral margin of the dorsal accessory olive laterally, the lateral margin of the medial accessory olive medially (see arrows i n lowermost drawing). (From Brodal, '40b). For Abbrevicitions see list above.
n
P
m
4 52
ALF BRODAL, FREII WALBERG AND GRETHE H. HODDEVIK
TABLE 1 Peroxidase solution Cat B.St.L.
618 619 620 622 633 634 636 647 649 650 652 653 656 662 663
Weight (kg)
Concentr.
2.1 2.1 2.7 2.1 3 3 3.2 1.8 2.7 0.750 2.7 2.3 1.6 1.8 2.7
Amount
survival time in days
0.5p1 0.5p1 0.5pl 0.25p1 0.5 pl
0.5 pl 0.5 p1 0.25p1
0.3pl 0.05p1 0.25pl 0.4p1 0.4p1 0.4p1 0.3p1 x2
2 3 1 2 3 2 2 1
cerebellum was separated from the brain stem. Both were divided into several blocks and further fixed, as described elsewhere (Walberg et al.. '76). Serial sections of 50-60 p were cut on the freezing microtome, in the transverse plane of the brain stem, in the sagittal plane of the cere bellum. From every group of five sections two were treated with 3,3-diaminobenzidine MATERIAL AND METHODS tetrahydrochloride as described by Graham The material used in the present study and Karnovsky ('66) and mounted on slides. consists of 13 adult cats and one kitten One of the sections was weakly stained with in which horseradish peroxidase (HRP) cresyl violet, the other was left unstained. solution was injected into various parts During the exposure of the cerebellum the of the paramedian lobule (table 1). Some site of injection was marked on a diagram cases had to be discarded because of tech- of the cerebellar surface imagined unfoldnical imperfections, especially inadvertent ed, taken from Larsell ('70). The stained spread of the solution into neighbouring parts of the cerebellar cortex were idenfolia. The technique employed is described tified in sagittal, approximately serial secelsewhere (Walberg et al., '76). Suffice it tions through the cerebellum and indicated to mention that after craniotomy under in the diagram of the cerebellar surface. Nembutal anaesthesia injections of small Usually the small tissue destruction caused amounts of a horseradish peroxidase solu- by the injection needle could also be astion (Sigma VI P 8375), most often 500 +Lgl certained. The mapping of the olivary labeling was pl, were made with a Hamilton syringe. To obtain access to the most caudal folia made by plotting the labeled cells as dots of the paramedian lobule the neck of the in drawings of the olive, made by means animal had to be strongly ventroflexed in of a projection apparatus. (Variations in a special manner in the Horsley-Clarke the density of dotting serve to give an apparatus. In two cats injections were impression of the relative density of cellular labeling.) All sections through the made in the nuclei interpositi. After one to four days the animals were olive were drawn. In order to facilitate killed under deep Nembutal anaesthesia a comparison between cases, the map obby intracardiac perfusion of a mixture of tained fmm an individual case was trans0.4% paraformaldehyde and 1.25% glu- ferred to a standard diagram of 15 equally taraldehyde with phosphate buffer. After spaced transverse sections through the olive removal of the brain from the skull the (figs. 1, 2) made on the basis of a study
The findings made considerably extend the information obtained with the retrograde cell degeneration method. On several points they are in agreement with the physiological results of Armstrong et al. ('74) and finally they bring forth supplementary evidence of the intricate organization of the olivocerebellar projection.
OLIVOCEREBELLAR LOCALIZATION
of the inferior olive in 2-3 weeks old kittens (Brodal, '40b). Apart from the fact that the olivary complex has a relatively greater rostrocaudal extent in adult than in very young cats (fig. 1, in Sousa-Pinto and Brodal, '69) the topography of the inferior olive, except for minor details, is the same. RESULTS
1. The staining of the injected
cerebellar cortex As described elsewhere (Walberg et al., '76) the density and spatial extent of staining in the cortex vary among cases, depending upon several factors such as amount and concentration of the peroxidase solution injected. In a stained area the Purkinje cells and their dendritic trees have a lightly granular brown color, and in sagittal sections pictures resembling a successful Golgi preparation are often seen. Usually the molecular as well as the granular layer are stained in these regions, and often also parts of the white matter. Occasionally the staining extends into the depth of the injected folia and reaches the cerebellar nuclei. When the injected fluid has not been deposited close to the top of the folium it is often seen that the cortex of the top of the folium is unstained, while a little deeper cortex on both sides of a sulcus may be clearly stained. 2. T h e labeling of cells in the inferior olive In sections stained lightly with cresylviolet and especially in unstained sections studied under dark field illumination, it is usually easy to distinguish labeled olivary cells from unlabeled ones. As in various other parts of the central nervous system, the labeled cells contain a varying amount of rather coarse brown granules, 0.5-0.1 p in diameter, in the cytoplasm of the soma and proximal parts of the dendrites. These cells can usually be easily distinguished from endothelial cells which have taken up peroxidase (see, for example, Nauta et al., '74) and which may occur in the olive as well as in other parts of the brain. The number and size of the granules, i.e. the intensity of staining, vary among the cells. The cells in the periphery of a labeled area of the olive have fewer, and often smaller granules
453
than those in the central part of the area. In some instances practically all cells are labeled within the central part of a labeled area, but often there are non-labeled cells intermingled with the positive area. 3. The distribution of labeled cells i n the olive This is the point of main interest in the present study. Since there are variations in this respect among cases with differently placed injections in the paramedian lobule, some of these cases will be presented below, as well as two cases with peroxidase injections into the cerebellar nuclei. Following a presentation of cases, mainly in the form of drawings, some conclusions which can be drawn from the experimental findings will be considered before the findings are discussed. Several cases with injections in the rostral two-thirds of the paramedian lobule give rather concordant results. (Concerning weight of the animals, the injections and the survival times, see table 1.) In cat B.St.L. 619 (fig. 2 ) the needle track in the left paramedian lobule can be identified in the third folium from above (figs. 2A,B). The cortex of the rostral 4 folia is stained (hatchings in figs. 2A,B), and in most of this cortex the molecular layer is stained (black in figs. 2A,B). Cell bodies and dendrites of Purkinje cells are likewise stained a finely granular brown. In the most intensely stained folia there is rather heavy staining of the white matter, and a fair number of brownish HRPimpregnated fibers (wavy lines in fig. 2A) can be traced from this region into the posterior pole of the nucleus interpositus posterior (NIP) where some nuclear cells are faintly brown. There is some diffuse staining of some dorsal regions of the NIP, and a little in the NIA (nucleus interpositus anterior). In the olive labeled cells are found only on the side contralateral to the injection. They are distributed to altogether four different, fairly well delimited regions (figs. 2C,D). In the principal olive there is conspicuous labelling of most of the cells in the caudal half of the ventral lamella, at levels IX-XI1 (covering most of the area found in retrograde studies to project onto the paramedian lobule, see fig. 1). In addition there is in the medial
454
ALF BRODAL. FRED WALBERG AND GRETHE H . HODDEVIK
part of the dorsal accessory olive (levels VIII-XIV) a large area with, on the whole more faintly labeled cells (about half of the total number), and a third in a small part of the dorsal lamella of the principal olive (levels VII-XII) and its transition to the ventrolateral outgrowth, and finally there is rather strong labelling of a number of cells in a restricted part of the rostral half of the medial accessory olive (levels IX-XII). An almost identical distribution of labeled cells in the contralateral olive is found in another experiment, cat B.St.L. 620 (fig. 3A) with a similar injection site. There is marked staining of most of the rostralmost four folia of the paramedian lobule, particularly the second and third. The white matter of the stained paramedim folia is heavily stained. Some stained fibers can be traced towards the NIP, but there is no staining of the nucleus. In the olive the distribution of labeled cells corresponds almost completely to that found in the case first described (fig. 2). In all cases with injections in the rostral two-thirds, approximately, of the paramedim lobule, labeling of cells in the olive is found within the areas indicated in figures 2 and 3A. However, in some instances the labeled areas are a little more extensive, in other cases they are more restricted. In cat B.St.L. 636 (fig. 3B), following injection into the fourth folium of the paramedian lobule, there is massive staining of most of the rostral half of the lobule, except for the medialmost parts of the caudalmost of the stained folia. There is some staining of the caudal part of the nucleus interpositus posterior. In the olive the areas containing labeled cells correspond closely to those found in the two preceding cases, but they are a little more extensive, especially in the ventral lamella, where in part of the area practically all cells are labeled. In cut B . S t . L . 647 (fig. 3C) the staining of the cortex is restricted to the two most caudal folia of the rostral half of the paramedian lobule. Labeling in the olive is found within the four areas indicated in figures 2 and 3A,B, but in all of them, except that in the medial accessory olive, it covers only part of them. In cat B.St.L. 634 (fig. 4A) rather similar cerebellar areas are stained as in the
previous case (B.St.L. 647). There is a trifling staining of the adjoing folia of lobulus VII A. Labeling in the olive resembles that found in cat B.St.L. 647 (fig. 3C) except for small differences. Thus there is only a very small patch of labeling in the dorsal accessory olive. In a third animal (cut I3.St.L. 633, not illustrated) with an injection similar to that in cat B.St.L. 647, there is labeling in the dorsal and ventral lamella as in cat B.St.L. 634 (fig. 4A) but not in the accessory olives. In a kitten (cut B.St.L. 650, not illustrated) a very restricted staining of the second and third folium of the paramedian lobule results in only two small patches of labeled cells in the contralateral olive, one in the dorsal, another in the medial accessory olive. Both patches are within the territories labeled in cat B.St.L. 636 (fig. 3B) and cover approximately the caudal half of each of them. In two cases injections limited to the rostral-most one or two folia of the paramedian lobule are followed by labeling in the areas in the medial and dorsal accessory olives stippled in figure 2 and figures 3A,B. Figure 4B shows one of them (cat B.St.L. 662). In both olivary subdivisions the labeled areas cover only parts of the more extensive areas labeled in cats B.St.L. 620 and 636 (figs. 3A,B), in the dorsal accessory olive mainly the rostrolateral parts of that area. A few cells are found in the dorsal lamella. Almost identical findings are made in another animal (cat B . S t . L . 663,not illustrated) where the staining of the cortex was restricted to the lateralmost part of the first paramedian folium. Fig. 2 Diagrammatic presentation of the findings in a cat with an injection of horseradish peroxidase into the paramedian lobule. A shows a parasagittal section through the cerebellum. B shows part of the cerebellar surface unfolded (after Larsell, '70). Arrows point to site of injection. In A and B the cerebellar areas stained are hatched, black indicates staining ofmolecular layer. C shows the distribution of peroxidase labeled cells (stippling) in the contralateral inferior olivary complex entered in a standard diagram of transverse sections (fig. 1). Variations in density of stippling in this and similar diagrams indicate varying degrees of cell labeling and density of labeled cells. D shows the labeled areas seen in the sections entered in a diagram of the olive unfolded in one plane as in figure 1 . For Abbreviations see list on p. 452.
455
OLIVOCEREBELLAR LOCALIZATION Lob. ant
A
m
rostra1
MEDIAL ACC OLIVE
D
PRINCIPAL OLIVE
rasfral
caudal
Figure 2
DORSAL ACC OLIVE
456
ALF BRODAL, FRED WALBERG AND GRETHE H. HODDEVIK
A
xv
XN
xm
w x X
M m
vn VI V N
m
I[
I
Fig. 3 Diagrams showing the findings in three cases with peroxidase solution injections of the rostral two-thirds of the paramedian lobule. The stained parts of the cerebellar cortex are indicated in a drawing of the paramedian lobule above to the left (black denotes staining of molecular layer). The distribution of labeled cells i n the olive is shown in a few drawings of transverse sections of the olive from the standard diagram shown in figure 2, and the total distribution is entered in the diagram of the olive imagined unfolded shown in figures 1 and 2. For Abbrevicctions see list on p. 452.
457
OLIVOCEREBELLAR LOCALIZATION
B.St.L.662
B
xm
Fig. 4 Diagrams showing the findings in two cases with peroxidase solution injections of middle and rostral parts of the paramedian lobule. Principles of presentation as in figure 3. For Abbrevirrtions see list on p. 452.
Injections in the folia of the rostral twothirds of the paramedian lobule thus give rise to labeling of olivary cells in one or more of four areas, situated in the medial and dorsal accessory olive, the ventral lamella and the dorsal lamella, respectively. Taken together the areas labeled in the various cases fill the territory indicated by horizontal hatching in figure 7. When injections are made in the caudal third, approximately, of t h e paramedian lobule, another pattern of distribution of labeled cells i s seen. Figure 5 shows two cases of this type. In cat B.St.L. 656 (fig. 5A) an injection in the second folium from below has resulted in a rather restricted staining of the molecular layer of two folia. In the oliue there are labeled cells in the dorsal
and medial accessory olives and the dorsal lamella, but not in the ventral lamella. In the dorsal and medial accessory olives the labeled areas are situated more laterally than following injections of the rostral two-thirds of the paramedian lobule with little overlapping, while the area in the dorsal lamella overlaps with the caudal part of the area sending fibers to the rostral paramedian lobule. In cat B.St.L. 653 (fig. 5B) a rather similarly placed injection, but with a more restricted staining of the molecular layer, has resulted in labeling of part of the areas in the dorsal and ventral accessory olive labeled in the previous case. A single labeled cell is observed caudally in the ventral lamella, none in the dorsal lamella. In another animal, cut B.St.L. 622 (not
458
ALF BRODAL, FRED WALBERG AND GRETHE H. HODDEVIK
B.St.L.656
A
X
xv
M
XIV
PI
D
xv
xu
@ w
vm w
m w
m
XI X M
VI nr V
XI
M
VI
V
nr
xv
XN
m xn XI
M
m
w
VI
V N
m
I
Fig. 5 Diagrams showing the findings in two cases with injections into the caudal part of the paramedian lobule, and one case with some staining of this part. Principles of presentation as in figure 3. For Abbrwitrtions see list on p. 452.
illustrated) a small quantity of peroxidase solution injected has resulted in staining of only small strands of cortex of the three caudalmost paramedian folia. There
are labeled cells only in part of the area of the medial accessory olive containing such cells in the preceding two animals. When the olivary areas with labeled
OLIVOCEREBELLAR LOCALIZATION
cells following injections of the caudal one-third of the paramedian lobule are added, they cover the territory shown by vertical hatching in figure 7. Even if i t is questionable whether these olivary areas can be taken to represent the total extent of the parts of the olive which send fibers to the caudal one-third of the paramedian lobule, they show that at least much of the olivary areas projecting to this part are differently situated than are those sending fibers to the rostral twothirds. A further case (cat B.St.L. 618) where the caudal part of the rostral twothirds as well as some folia of the caudal one-third of the paramedian lobule were stained is in agreement with this conclusion, as is seen from figure 5C. Since in some cases (cats B.St.L. 618, 619, 636) there was staining of the nuclei interpositi, especially the interpositus posterior (NIP), it was deemed desirable to determine whether the olivary areas labeled following injections of the paramedian lobule contain labeled cells also following injections of the nuclei. Two experiments in which the nuclei were injected stereotactically give information about this. In cut B.St.L. 649 (fig. 6A) the injection was made in the nucleus interpositus anterior (NIA). Transverse sections of the cerebellum show that the nucleus is heavily stained almost throughout except for a small dorsolateral area, which is faintly stained. The caudal border against the nucleus interpositus posterior (NIP) cannot be discerned, but from comparisons with sections from normal animals it appears that there is probably some staining of the rostro-medial part of the NIP. There is no staining of the fastigial or dentate (lateral) nuclei. In the olive labeled cells are found in the contralateral dorsal and medial accessory olive, a few also in the dorsal cap, as indicated in figure 6A. In cut B . S t . L . 652 (fig. 6B) the nucleus interpositus posterior (NIP) is heavily stained. There is some staining of the surrounding white matter. The border against NIA can not be identified. There is some weak staining of what apparently is the most caudal part of NIA. The fastigial and lateral nuclei are not stained. In the olive cell labeling is mainly faint and the labeled cells are distributed in a pattern which is largely different from that fol-
459
lowing injection in the NIA. In part, however, the areas labeled following injections of NIA and NIP overlap, mainly in the medial accessory olive. Since the injection in NIA in cat B.St.L. 649 has probably stained also the rostromedial part of NIP, it is possible that the weak labeling of the most lateral part of the accessory olives, particularly the dorsal, in this case is due to this fact. Further experiments are needed to decide this. For our present purpose it is of importance to note that the areas labeled in the two cases to some extent coincide with those labeled from the paramedian lobule. 4. Interpretation offindings A consideration of methodological problems (Walberg et al., '76), makes it clear
that the labeling of cells in the olive following injections of horseradish peroxidase in the cerebellar cortex (and nuclei) may be taken to show the sites of origin of olivocerebellar fibers to the part injected. Even if HRP is easily taken up by injured fibers (DeVito et al., '74) it is assumed in the following that the peroxidase is taken up, at least largely, by nerve terminals (LaVail et al., '73; Turner and Harris, '74; Bunt et al., '74) as more fully discussed elsewhere (Walberg et al., '76). In all cases reported above of injections in the paramedian lobule except three (cats B.St.L. 622, 647 and 650) 0.5 or 0.4 p1 of a solution of 500 ~gIp1was injected. The transferring of the topographical pattern of localization of labeled cells in the sections of the olive in a particular case to a standard diagram of the olive will of necessity often entail some slight displacement of the borders of a labeled zone. Even if errors of this kind are taken into account (they are not equally pertinent to all olivary regions) i t is striking to note the high degree of corresponding findings in the olive in cases with similarly placed injections. The findings permit some conclusions concerning the organization of the olivocerebellar projection of the paramedian lobule (and some other regions). In the first place they confirm the results obtained with other methods in showing that the olivocerebellar projection is precisely organized. This appears also from the fact
460
ALF BRODAL, FRED WALBERG AND GRETHE H. HODDEVIK
B
Fig. 6 Diagrams >.dowing the findings in two cases with injection of peroxidase solution of the nucleus interpositus anterior (A) and posterior (B), respectively. The stained p a t s Of the nuclei are shown in a transverse section of the cerebellum. Black denotes heavy staining hatching lighter staining. Findings in the olive presented according to the same principles as in figure 1. For Abbrevtcitions see list on p. 452.
that the borders between olivary regions containing labeled cells and those free from such cells are often astonishingly sharp. In other instances, however, the border zone of a labeled area contains more sparsely distributed labeled cells which are usually also more faintly labeled than those in the central part of the area. This presumably is due to the
fact that in the peripheral zones of the stained areas of the cerebellar cortex there will often be a lower concentration of the solution than more centrally, as appears also from the relative degrees of staining of the cortex seen in the sections. Or it may be due to a patchy staining of the folia (see Walberg et al., '76). Before proceeding it is appropriate to
OLIVOCEREBELLAR LOCALIZATION
consider the question whether some of the labeling found with injections of the paramedian lobule may be due to diffusion of peroxidase solution to the cerebellar nuclei and uptake by olivonuclear fibers, since in some cases there was some staining of the neuropil of the NIA or NIP. This, as well as the staining of the white matter between the injected cortical site and the nucleus in these cases indicates that there has been some extracellular spread of peroxidase solution in the white matter.' The olivary areas labeled following injections of the NIP and NIA overlap in part with those labeled from the paramedian lobule. Thus the medial part of the dorsal accessory olive is labeled following injections in the rostral part of the paramedian lobule (fig. 7) as well as of the NIA (fig. 6A). The common labeling, however, does not invalidate the conclusion that its cells project to the rostral twothirds of the paramedian lobule, since only in one of the cases with injections here, was there a slight staining of a small part of the NIA (cat B.St.L. 619, fig. 2). In the medial accessory olive the NIA projection area (which in part may supply also the NIP) largely spares the projection area of the rostral two-thirds of the paramedian lobule. The olivary areas labeled following injections in the NIP overlap only slightly with those labeled from the rostral twothirds of the paramedian lobule (compare figs. 6B and 7) and present no problem with regard to the projection of this. They overlap, however, to some extent with the projection area of the caudal one-third in the dorsal as well as the medial accessory olive. However, in none of the three cases where the injection was restricted to the caudal one-third of the paramedian lobule was there any staining of the NIP. The findings thus do not contradict the conclusions made about the olivary projection to the caudal one-third of the paramedian lobule. As seen from figure 7 the rostral twothirds of the paramedian lobule receive fibers (or collaterals) from four olivary areas, in the medial and dorsal accessory olive and the ventral and dorsal lamella. Except for some overlapping, especially in the dorsal lamella, the olivary areas pro-
461
jecting to the caudal one-third are largely different. They may be somewhat greater than shown in figure 7, since this cortex was never stained in its totality. The caudal one-third receives fibers from three olivary areas, but apparently not from the ventral lamella. From figure 7 it is further seen that in the medial and dorsal accessory olives the areas which project onto the caudal one-third are on the whole found more caudally and more laterally, and in the dorsal lamella more caudally, than the areas which project onto the rostral two thirds. The rostrocaudal folial sequence in the paramedian lobule thus appears to correspond to an arrangement of projection areas from rostromedial to caudolateral in the accessory olives, from rostral to caudal in the dorsal lamella (arrows in fig. 7). This raises the question: Is it possible to find evidence for a more detailed pattern in this projection, either in the longitudinal or transverse direction of the folia? In an analysis of this kind several factors must be taken into consideration. There may well be minor topical differences in the pattern which escape recognition. Our material is not suited to decide if there are topographical differences in the olivary projection of functionally discovered longitudinal zones of the paramedian lobule (DISCUSSION). In most cases the injected solution has spread more or less along the entire transverse extent of the small folia. There is slight evidence that the projection field in the dorsal accessory olive is related particularly to the medialmost part of the paramedian lobule as concluded by Armstrong et al. ('74). When the staining of the folia includes their middle and lateral parts labeled cells have been found in the regions of the olive indicated by Armstrong et al. ('74) as projecting to these two longitudinal strips, but nothing can be concluded from our material about possible differences between the olivary projections to these longitudinal zones. 1 A s described above, in some cases stained fibers could be traced from the injected parts of the paramedian lobule to the nuclei interpositi, indicating an anterograde transport of peroxidase in axons of Purkinje cells. The terminal regions for these fibers appear to be within those determined by Courville et al. ('73) to receive fibers from the paramedian lobule.
462
ALF BRODAL, FRED WALBERG AND GRETHE H. HODDEVIK
I
coudol QM
'
d.cap.
MEDIAL ACC.OLIVE
PRINCIPAL OLIVE
DORSAL ACC.OLIVE
rostrol
caudal
Fig. 7 Diagram showing the olivary areas projecting onto the rostral two-thirds of the contralateral paramedian lobule (horizontal hatchings) and onto the caudal one-third (vertical hatchings). The olive represented in a series of transverse sections from the standard dlagram (figs. 1, 2) and as imagined unfolded (below). Arrows indicate sequence of representation of folia of paramedian lobule from rostral to caudal.
Concerning differences in the projections of different folia of the paramedian lobule, better information can be obtained. It is striking that with similar extension of staining of the row of folia the labeling of the olive is usually almost identical, for example cats B.St.L. 619, 620 and 636, figures 2, 3A,B, survival times three, four and one day respectively; cats B.St.L. 647 and 634, figures 3C and 4A, survival times two and three days, respectively. It ap-
pears from our material that within the projection area in the ventral lamella, the paramedian folia are represented in a rostrocaudal sequence. When the rostralmost three or four folia of the lobule are stained, the labeling of the ventral lamella extends through approximately its caudal half (B.St.L. 619, 620, 636, figs. 2, 3A,B) while with more caudally placed injections (B.St. L. 647 and 634, figs. 3C, 4A, see also B.St.L, 618, fig. 5C) the labeling is found
OLIVOCEREBELLAR LOCALIZATION
only in the caudal part of this caudal area. As concerns the accessory olives, the evidence that the sequence from rostromedial to caudolateral, indicated by the arrows in figure 7, is valid also for minor divisions of the rostral two-thirds of the paramedian lobule is less clear. This is especially the case for the dorsal lamella where the projections of the rostral two-thirds and caudal one-third of the paramedian lobule overlap considerably. However, even if there is some overlapping between the projection areas of neighbouring folia there appears to be a main pattern. DISCUSSION
It has been amply demonstrated (La Vail and La Vail, ’72and others, see the beginning of this article) that the retrograde axonal transport of horseradish peroxidase can be used to determine the origin of nerve fibers in the central nervous system. It appears from these studies that conditions are not uniform in all situations (variations according to animal species, fiber systems, etc.). A study of how the cat’s olivocerebellar projection behaves in these respects (Walberg et al., ’76) showed that if an injection of HRP is made in a certain part of the cerebellar cortex, and the survival time is adequate, the distribution of labeled olivary cells appears to give a reliable picture of the site of origin of olivary fibers to that part. When the distribution of labeled cells in the olive is more restricted in one case than in another with staining of the same folia, the case with the most extensive labeling has been considered significant. There is reason to believe that the small particular subdivisions of the olive which can be differentiated on a morphological basis (such as the nucleus p, the dorsal cap, the dorsomedial cell column, see figs. 1, 2) are functionally dissimilar, likewise, for example, lateral and medial regions of the dorsal accessory olive. In mapping labeled cells in the olive care has, therefore, been taken to identify precisely the particular region(@ where they occur. The present study confirms that the olivocerebellar projection is topographically organized, as has been concluded from anatomical studies of human (Henschen, Jr., ’07; Holmes and Stewart, ’08) and experimental animal material (Brodal, 40b;
463
see also Jansen and Brodal, ’58). However, the pattern is clearly far more complex than was found in studies of the retrograde cellular changes in the olive. While the present study is restricted to a minor part only of the total olivocerebellar projection, that to the paramedian lobule, in the following discussion reference will occasionally be made to observations made with injections into other parts of the cerebellum (unpublished). Topography of the oliuocerebellar projection onto the paramedian lobule Following extirpation of the paramedian lobule in very young cats and rabbits (Brodal, ’40b) convincing cell loss was found only in the caudal part of the ventral lamella of the principal olive (fig. 1). The present study shows that the paramedian lobule in addition receives fibers from circumscribed parts of the dorsal and medial accessory olives and part of the dorsal lamella of the principal olive (fig. 7). The area in the ventral lamella corresponds almost completely to that determined as projecting onto the paramedian lobule in the study with the retrograde method (cp. fig. 1 with fig. 7). When the other olivary regions were not identified in 1940, the reason is that they did not show cellular changes or cell loss which were sufficiently clear to be considered as significant.* Why do lesions of the paramedian lobule result in clear-cut retrograde cellular changes only in its projection area in the ventral lamella? One possibility could be that the fibers of the other olivary areas projecting onto the paramedian lobule are collaterals of fibers whose main axon passes to other destinations. If so, destruction of their collaterals (to the paramedian lobule) might be insufficiently harmful to give rise to changes which can be recognized in Nissl sections in agreement with the 4 This does not exclude that these olivary areas may show less clearcut changes also with the retrograde method. A renewed examination of series from the old material shows that with guidance of the results of the present study one can discern some slight changes i n the “paramedian lobule area” in the medial accessory olive. It may be mentioned also, that in an experiment on the rabbit with isolated ablation of the paramedian lobule (rabbit 0.129) there was ,noted and marked i n the diagram (fig. 17 in Brodal, 40b) in addition to a heavy cell loss i n the ventral lamella slighter changes in the medial accessory olive i n an area corresponding to that s h o w n in figure 7. However, at that time these changes were assumed to be due to undercutting of some fibers to the vermis.
464
ALF BRODAI.. FRED WALBERG AND (;RETIIE 11. HODDEVIK
concept of “sustaining projections” of Rose and Woolsey (’43, see also Fry and Cowan, ’72). (Since the retrograde studies were made on newly born animals it may even be that the collaterals had not yet reached their site of destination when the animals were operated on. In adult animals the retrograde olivary changes are less clear cut than in young ones, Brodal, ’39). When taken together the olivary areas demonstrated here as projecting onto the paramedian lobule (fig. 7) cover territories of the entire olivary complex which are quite out of proportion to the modest share of the total cerebellar cortex which belongs to the paramedian lobule. One is almost forced to conclude, therefore, that these areas send fibers to other regions of the cerebellum as well. Our material shows that this is indeed the case. For example, the medial part of the dorsal accessory olive is labeled following injections in the rostral two-thirds of the paramedian lobule (figs. 2, 3), the nucleus interpositus anterior (fig. 6) and, in agreement with the findings from 1940 (fig. l ) , also the intermediate part of the anterior lobe (unpublished). This olivary area, therefore, influences at least three cerebellar regions. Correspondingly, the lateral part of the dorsal accessory olive is labeled following injections of the caudal third of the paramedian lobule (figs. 5, 7) as well as of the nucleus interpositus posterior (fig. 6B). From these and other (unpublished) findings it appears to be a general feature that at least many small parts of the olive have connections with several parts of the cerebellum, as has been demonstrated physiologicallv by Armstrong et al. (’74) for certain parts of the olivary complex. I t will require extensive studies to map these relations completely with the peroxidase method. In view of the rather precise borders of the various olivary fields projecting onto the paramedian lobule the question may be raised whether finer topographical patterns exist within these projections. As described above (4. Interpretation offindings) this appears to be the case. The folia, when followed from rostral to caudal, appear to receive their afferents from cells in regions situated successively more caudally in the ventral and dorsal lamella, and successively more caudolaterally in the
accessory olives (fig. 7), even if there is considerable overlapping. In the study of the retrograde changes following cerebellar lesions in the rabbit (Brodal, ’40b) the same pattern as here was found for the projection from the ventral lamella to the paramedian lobule. Judging from the well established pattern of somatotopical localization in the paramedian lobule this means that the rostral part of the projection area in the ventral lamella is related mainly to forelimb (and face), while caudal parts are related mainly to hindlimb (and tail). Our most caudally situated injections have been made in folia which scarcely are accessible to either recording or stimulation in physiological experiments. It is worthy of notice that the olivary projection areas of this part are largely separate from those of the rostral two-thirds of the lobule (fig. 7). The same appears to be the case for the regions of the pons projecting onto rostral two-thirds and caudal onethird, respectively, of the paramedian lobule (Hoddevik, ’75). Our study shows that the afferents to the caudal one-third come from at least three olivary regions, while Armstrong et al. (‘74) depict only the area found here in the lateral part of the dorsal accessory olive as sending fibers to this part. The caudal part of the paramedian lobule left white (i.e. not studied) in Armstrong et al.’s diagram (their fig. 5, ’74) probably includes more than the one or two folia shown in their diagram. It might well be that the caudal one-third of the paramedian lobule differs from the rostral two-thirds with regard to other connections than those from the pons and the olive and that the two parts are not hnctionally equivalent. As described above, differences in the projection of longitudinal zones of the paramedian lobule as found by Armstrong et al. (’74) could not be determined, except for suggestive evidence in favor of the projection to the medialmost zone from the dorsal accessory olive. Oliuary projections to the cerebellur
nuclei Our observations concerning these projections are incomplete. However, they provide clear evidence that at least the NIA and N I P receive fibers or collaterals from particular regions of the olive (fig. 6). The
OLIVOCEREBELLAR LOCALIZATION
main areas are found in the medial and dorsal accessory olives, while only a few labelled cells are found in the dorsal cap which was concluded from the retrograde studies (Brodal, '40b) to be the origin of fibers to these nuclei. It may be objected that following injections in the nuclei the labeling of olivary cells may be due to u p take by fibers passing to the cortex through the nuclei and injured during the injection (DeVito et al., '74). However, the vast majority of olivocerebellar fibers pass outside the nuclei, and the needle track causes only little destruction of tissue. It is extremely unlikely, therefore, that injury of fibers in the nuclei can explain the widespread and clearcut labeling of olivary cells in these cases (fig. 6). Only few labeled cells were found in the ipsilateral olive. The findings definitely fail to support the claim by Matsushita and Ikeda ('70) that fibers from the olive to the cerebellar nuclei are to a considerable extent bilateral. These authors base their opinion of the occurrence of widespread fiber degeneration in the nuclei following lesions (largely parasagittal incisions) of one olive. Such lesions, however, will inevitably transect fibers from the olive of the other side in addition to fibers from areas of the ipsilateral olive situated medial to the cut and therefore, do not permit conclusions like those made concerning an ipsilateral distribution. The question of branching of climbing fibers In the cat the inferior olives (both sides) contain some 121,000-145,000 cells (Escobar et al., '68; Mlonyeni, '72) while the number of Purkinje cells is given as 1.21.3 million (Palkovits et al., '71) or 1.5 million (Mlonyeni, '73). In anatomical studies branching of climbing fibers has been seen only in or just beneath the cortex and these branches supply two, three or four Purkinje cells not far removed from each other (Scheibel and Scheibel, '54; Fox et al., '69). Physiological investigations, however, show that a single climbing fiber may branch and supply folia at a considerable distance from each other (Faber and Murphy, '69; Armstrong et al., '71a, '73; Cooke et al., '72). Whether all climbing fibers are derived from the inferior olive is still not settled (see, for ex-
465
ample, OLeary et al., '70; and review by Brodal, '72). However, if so, the numerical relation between Purkinje cells and olive cells would require that in the cat one of the latter should supply altogether 10-11 Purkinje cells. If three or four of these are supplied by neighboring terminal collaterals in the cortex, it would still be necessary that there were about as many branchings of each parent fiber more proximally. In some parts of the olive Armstrong et al. ('74) indeed found cells which respond antidromically to stimulation of up to four different cortical zones. As mentioned above multiple projections from minor olivary regions have been found also with the peroxidase method. However, decisive anatomical evidence that multiple connections are established by branches or collaterals or individual parent axons would require that all, or practically all, cells in a particular olivary region were labeled following injection of each of the appropriate cerebellar sites. Following injections in the paramedian lobule such areas have been observed within all four paramedian projection areas in the olive, but it remains for future studies to decide if all cells can be labeled in the "paramedian" areas also following injections of other cerebellar lobules. The present findings are however, not incompatible with the idea of branching of the axons of olivary cells. Correlations of anatomical and physiological observations on the olivocerebellar topography In studies of the connections of the inferior olive it is essential to identify precisely the minor olivary parts dealt with. With regard to the olivocerebellar connections a precise identification of the cerebellar sites concerned is equally essential. In these respects there are some differences between what anatomical and physiological methods can yield. The peroxidase method permits the identification of even a very small region of the olive which contains labeled cells, while it is difficult to determine with equal precision the areas of the cortex fiom which terminals of olivocerebellar fibers have taken up the peroxidase solution. On the other hand, the use of antidromic stimulation of olivary cells has the advantage that the spot of the cortex stimulated can be determined ex-
466
ALF BRODAL, FRED WALBERG AND GRETHE H. HODDEVIK
actly, while the identification of recording points in the olive may not always meet the desirable degree of precision. The method of direct stimulation of the olive and recording from the cerebellum as used by VanGilder and O'Leary ('70) has a serious drawback in the approximate determination of the site of stimulation. However, the different approaches supplement each other, and in general there is good agreement between the results obtained. Thus VanGilder and O'Leary ('70) found projections to the paramedian lobule following stimulation of the principal olive and the medial accessory olive. The olivary sites indicated by Armstrong et al. ('74) as projecting to the rostral two-thirds of the paramedian lobule (their fig. 5) are found within the areas determined in the present study (fig. 7). Their area in the medial part of the dorsal accessory olive coincides almost completely with that outlined by us. However, the three other areas found here are more extensive than appears from the diagram of Armstrong et al., especially that in the dorsal lamella. With regard to determination of the extent of a projecting olivary area, the peroxidase method must be considered more reliable than the recordings of antidromic responses. The olivary projection to the caudal part of the paramedian lobule was found by Armstrong et al. ('74, their fig. 5) to be different from that from the anterior twothirds. Like us they found the lateral part of the dorsal accessory olive to project to the caudal one-third of the lobule. With the peroxidase method we found in addition an area in the medial accessory olive and another caudally in the dorsal lamella. That Armstrong et al. ('74) did not find the latter areas may be due to the fact that the caudalmost part of the lobule was not explored. It may well be that the three olivary projection areas show in fact slight topical differences in their sites of termination in the caudal onethird, a problem which could not be studied with our approach. For a study of the longitudinal pattern in the afferent cerebellar projections the method of antidromic stimulation has obvious advantages over the peroxidase method. On the other hand, rostrocaudal differences within the projection onto a single lobule are difficult to discern with the
former method, unless explicitly sought for. However, if two parts of a lobule receive their fibers from different olivary areas, they may be brought out, as exemplified by the finding of Armstrong et al. ('74) of the projections of the lateral and medial parts of the dorsal accessory olive to the caudal and rostral parts respectively, of the paramedian lobule, in agreement with our findings.
General considerations From the available data it is reasonable to assume that there is within the olivocerebellar projection a pattern which in part mimics the subdivision of the cerebellar cortex in lobules and to some extent even folia, in part it reflects the presence of longitudinal zones in the cerebellum. The evidence for the existence of the latter stems from physiological (Oscarsson, '69) as well as anatomical studies (Voogd, '64, '69; Korneliussen, '68, '69). Voogd ('69) in a study with silver impregnation methods found that olivocerebellar fibers originating in a certain small area of the olive terminate over the whole length of one or a few longitudinal zones. This is in agreement with physiological observations, but as rightly stated by Voogd ('69, p. 511) his observations "do not exclude the possibility that rostrocaudal differences in the projection of certain parts of the inferior olive are present as well, . . ." Some evidence supporting this has emerged from the present study (see above). Before the two principles of longitudinal and lobular organization can be synthesized, a first step will be to have a complete map of the entire olivocerebellar projection as studied with the horseradish peroxidase method. The recent anatomical findings explain why attempts to correlate physiological findings with the picture of the olivocerebellar projection as deduced from the study of retrograde changes (fig. 1) have been only partially successful. It is obvious from the present as well as from physiological studies (Armstrong et al., '74; and others) that each cerebellar lobule does not always receive its olivary afferents from one restricted part of the olive only. On the contrary, there are possibilities for a convergence of impulses from different parts of the olive to a particular part of the
OLIVOCEREBELLAR LOCALIZATION
cerebellum. On the other hand, impulses from many olivary parts may diverge to different cerebellar regions. The complexity of the anatomical organization of the olive is further increased by the fact that a particular olivary part receives afferents from many sources. As concerns the paramedian lobule, where there are at least three anatomically (Voogd, '69) and functionally (Szabo and Albe-Fessard, '54; Cooke et al., '72; Armstrong et al., '71a,b, '74) different longitudinal zones, its different olivary projection areas receive several contingents of afferents. Anatomically fibers have been traced to one or more of these olivary areas (fig. 7) from the spinal cord (Brodal et al., '50; Mehler et al., '60, and others), the dorsal column nuclei (Ebbesson, '68), the cerebral cortex (Walberg, '56; Sousa-Pinto and Brodal, '69; Sousa-Pinto, '69), the red nucleus (Walberg, '56; Hinman and Carpenter, '59; Edwards, '72), the caudate nucleus (Walberg, '56), certain regions of the mesencephalon as the reticular formation (Walberg, '56, '60, '74) and from the cerebellar nuclei (Graybiel et al., '73; Lacerda, '73). In addition there appear to be other afferents, thus probably from the pontine and medullary reticular formation. It is premature to attempt a correlation of these data with the physiological ones, but it may be mentioned that climbing fiber responses, assumed to be mediated via the inferior olive, have been observed in the paramedian lobule following stimulation of spinal nerves, some cranial nerves, the sensorimotor cerebral cortex, the caudate nucleus, the midbrain and some other sites (for a recent review, see Armstrong, '74). In general these findings are compatible with the anatomical data as are also physiological observations on the convergence on single olivary units of impulses from different sources (Armstrong, '74). The complexity in the anatomical and functional organization of the inferior olive appears to be ovenvhelming.3 To make possible a satisfactory integration, continued anatomical studies of afferents with detailed mapping of their olivary sites of termination will be necessary. In future physiological studies it will be particularly essential that sites stimulated or recorded from in the cere-
467
bellum or the olive are indicated with strict observation of their topography. LITERATURE CITED Armstrong, D. M. 1974 Functional significance of connections of the inferior olive. Physiol. Rev., 54: 358417. Armstrong, D. M., R. J. Harvey and R. F. Schild 1971a Climbing fibre pathways from the forelimb to the paramedian lobule of the cerebellum. Brain Res., 25: 199-202. 1971b Distribution in the anterior lobe of the cerebellum of branches from climbing fibres to the paramedian lobule. Brain Res., 25: 203-206. 1973 Branching of inferior olivary axons to terminate in different folia, lobules or lobes of the cerebellum. Brain Res., 54: 365-371. - 1974 Topographical localization in the olivocerebellar projection: An electrophysiological study i n the cat. J. Comp. Neur., 1 5 4 : 287-302. Brodal, A. 1939 Experimentelle Untersuchungen uber retrograde Zellveranderungen in der unteren Olive nach Lasionen des Kleinhirns. Z. ges. Neur. Psychiat., 166: 624-704. 1940a Modification of Gudden method for study of cerebral localization. Arch. Neurol. Psychiat. (Chic.), 43: 4&-58. - 1940b Experimentelle Untersuchungen uber die olivocerebellare Lokalisation. Z. ges. Neur. Psychiat., 169: 1-153. 1972 Cerebrocerebellar pathways. Anatomical data and some functional implications. Acta neurol. scand., suppl. 51: 153-196. Brodal, A,, F. Walberg and T. Blackstad 1950 Termination of spinal afferents to inferior olive in cat. J. Neurophysiol., 13: 431454. Bunt, A. H., R. D. Lund and J. S. Lund 1974 Retrograde axonal transport of horseradish peroxidase by ganglion cells of the albino rat retina. Brain Res., 73: 215-228. Cooke, J. D., 0. Oscarsson and B. Sjolund 1972 Termination areas of climbing fibre paths in paramedian lobule. Acta physiol., 8 4 : 37A-38A. Courville, J., N. Diakiw and A. Brodal 1973 Cerebellar corticonuclear projection in the cat. The paramedian lobule. An experimental study with silver methods. Brain Res., 50: 25-45. De Vito, J. L., K. W. Clausing and 0. A. Smith 1974 Uptake and transport of horseradish peroxidase by cut end of the vagus nerve. Brain Res., 8 2 : 269-271. Ebbesson, S. 0. E. 1968 A connection between the dorsal column nuclei and the dorsal accessory olive. Brain Res., 8 : 393-397. Edwards, S. B. 1972 The ascending and descending projections of the red nucleus in the cat: An experimental study using an autoradiographic tracing method. Brain Res., 48: 45-63. Escobar, A,, E. D. Sampedro and R. S. Dow 3 This applies as well to the finer histology of the olive as witnessed by light microscopical observations (Scheibel and Scheibel, '55; Scheibel et al., '56; and others) and to recent electron microscopical findings concerning special features in the synaptology of the
olive, such as gap junctions and glomeruli (Nemecek and Wolff, '69; Sotelo et al., '74; and others).
468
ALF BRODAL, FRED WALBERG AND GRETHE H. HODDEVIK
1968 Quantitative data on the inferior olivary nucleus i n man, cat and vampire bat. J. Comp. Neur., 132;397-404. Faber, D. S., and J. T. Murphy 1969 Axonal branching i n the climbing fiber pathway to the cerebellum. Brain Res., 15: 262-267. Fox, C. A,, A. Andrade and R. C. Schwyn 1969 Climbing fiber branching in the manular layer. In: Neurobiology of Cerebellar Evolution and Development. R. Llinas, ed. Chicago, Illinois, pp. 603-611. Fry, F. J., and W. M. Cowan 1972 A study of retrograde cell degeneration in the lateral mammillary nucleus of the cat, with special reference to the role of axonal branching in the preservation of the cell. J. Comp. Neur., 144: 1-24. Graham, R. C., Jr., and M. J. Karnovsky 1966 The early stages of absorption of injected horseradish peroxidase in the proximal tubules of mouse kidney: Ultrastructural cytochemistry by a new technique. J. Histochem. Cytochem., 14: 291-302. Graybiel, A. M., and M. Devor 1974 A microelectrophoretic delivery technique for use with horseradish peroxidase. Brain Res., 68: 167-173. Graybiel, A. M., H. J. W. Nauta, R. J. Lasek and W. J. H. Nauta 1973 A cerebello-olivary pathway in the cat: an experimental study using autoradiographic tracing techniques. Brain Res., 58: 205-211. Henschen, F. 1907 Serose Zyste und partieller Defekt des Kleinhirns. Zeitschr. f. klinische Medizin. 63: 115-1 52. Hinman, A., and M. B. Carpenter 1959 Efferent fiber projections of the red nucleus in the cat. J. Comp. Neur., I 1 3: 61-82. Hoddevik, G. H. 1975 The pontocerebellar projection onto the paramedian lobule i n the cat. An experimental study with the use of horseradish peroxidase as a tracer. Brain Res., 95: ~
291407. Holmes, G., and T. G. Stewart 1908 On the connection of the inferior olives with the cerebellum in man. Brain, 31: 125-137. Jansen, J., und A. Brodal 1958 Das Kleinhirn. In: v. Mollendorffs Handbuch der mikroskopischen Anatomie des Menschen 1V/8. Springer, Berlin-Gottingen-Heidelberg, 323 pp. Jones, E. G., and R. Y. Leavitt 1974 Retrograde axonal transport and the demonstration of nonspecific projections to the cerebral cortex and striatum from thalamic intralaminar nuclei i n the rat, cat and monkey. J. Comp. Neur., 154:
349-378. Korneliussen, H. K. 1968 Comments on the cerebellum and its division. Brain Res., 8: 229-236. - 1969 Cerebellar organization in the light of cerebellar nuclear morphology and cerebellar corticogenesis. In: Neurobiology of Cerebellar Evolution and Development. R. Llinas, ed., Chicago, Illinois, pp. 51%523. Kristensson, K., and Y. Olsson 1971 Retrograde axonal transport of protein. Brain Res., 29: 363-
365. Lacerda, A. M. de 1973 Alguns aspectos d a projecgBo dos nticleos centrais do cerebelo para a regiao da oliva bolbar. Um estudo experimental no gato. Fol. anat. Universitatis Conimbrigensis, 42: 1-19. Larsell, 0. 1970 The Comparativr: Anatomy
and Histology of the Cerebellum from Monotremes through Apes. J. Jansen, ed. The University of Minnesota Press, Minneapolis, 269 pp. LaVail, J . H., and M. M. LaVail 1972 Retrograde axonal transport in the central nervous system. Science, 176: 1416-1417. LaVail, J. H., K. R. Winston and A. Tish 1973 A method based on retrograde intraaxonal transport of protein for identification of cell bodies of origin of axons terminating within the CNS. Brain Res., 58: 4701177. Lynch, G., R. L. Smith, P. Mensah and C. Cotman 1973 Tracing the dentate gyrus mossy fiber system with horseradish peroxidase histochemistry. Exp. Neurol., 40: 516524. Mareschal, P. 1934 L’Olive Bulbaire, AnatomieOntog6nkse-Phylogeni!se-Physiologieet Physiopathologie. Pans. G. Doin & Cie, 216 pp. Matsushita, M., and M. Ikeda 1970 Olivary projections to the cerebellar nuclei in the cat. Exp. Brain Res., 10: 488500. Mehler, W. R., M. E. Feferman and W. J. H. Nauta 1960 Ascending axon degeneration following anterolateral cordotomy. An experimental study in the monkey. Brain, 83: 718-750. Mlonyeni, M. 1973 The number of Purkinje cells and inferior olivary neurones i n the cat. J. Comp. Neur., 147: 1-10. Nauta, H. J. W., M. B. Pritz and R. D. Lasek 1974 Afferents to the rat caudato-putamen studied with horseradish peroxidase. An evaluation of a retrograde neuroanatomical research method. Brain Res., 67: 219-238. Nemecek, S., and J. Wolff 1969 Light and electron microscopic evidence of complex synapses (glomeruli) in oliva inferior (cat). Experientia, 25: 6344335. OLeary, J. L., S. B. Dunsker, J. M. Smith, J. Inukai and M. O’Leary 1970 Termination of the olivocerebellar system in the cat. Arch. Neurol. (Chicago), 22: 193-206. Oscarsson, 0. 1969 The sagittal organization of the cerebellar anterior lobe as revealed by the projection patterns of the climbing fibre system. In: Neurobiology of Cerebellar Evolution and Development. R. LlinAs, ed. Chicago, Illinois, pp. 525-532. Palkovits, M., P. Magyar and J. Szentagothai 1971 Quantitative histological analysis of the cerebellar cortex in the cat. 1. Number and arrangement i n space of the Purkinje cells. Brain Res., 32: 1-13. Ralston, 111, H. J., and P. V. Sharp 1973 The identification of thalarnocortical relay cells i n the adult cat by means of retrograde axonal transport of horseradish peroxidase. Brain Res., 62: 273-278. Rose, J . E., and C. N. Woolsey 1943 A study of thalamo-cortical relations in the rabbit. Johns Hopkins Hosp. Bull., 73: 65-128. Scheibel, M. E., and A. B. Scheibel 1954 Observations on the intracortical relations of the climbing fibers of the cerebellum. A Golgi study. J. Comp. Neur., 1 0 1 : 73S764. - 1955 The inferior olive. A Golgi study. J. Comp. Neur., 102: 77-132. Scheibel, M. E., A. B. Scheibel, F. Walberg and A. Brodal 1956 Areal distribution of axonal and dendritic patterns in inferior olive. J. Comp. Neur., 106: 2149.
OLIVOCEREBELLAR LOCALIZATION Sotelo, C., R. Llinls and R. Baker 1974 Structural study of inferior olivary nucleus of the cat: Morphological correlates of electronic coupling. J. Neurophysiol., 37: 541-559. Sousa-Pinto, A. 1969 Experimental anatomical demonstration of a cortico-olivary projection from area 6 (supplementary motor area?) in the cat. Brain Res., 16: 73-83. Sousa-Pinto, A,, and A. Brodal 1969 Demonstration of a somatotopical pattern i n the cortico-olivary projection in the cat. An experimental-anatomical study. Exp. Brain Res., 8:
364-386. Szabo, T., et D. Albe-Fessard 1954 Repartition et caracteres des afferences somesthksiquks et dorigine corticale sur le lobe paramedian du cervelet du chat. J. de. Physiol., 46: 528-531. Turner, P. T., and A. B. Harris 1974 Ultrastructure of exogenous peroxidase in cerebral cortex. Brain Res., 74:305326. VanGilder, J. C., and J. L. OLeary 1970 Topical projection of the olivocerebellar system in the cat: A n electrophysiological study. J. Comp. Neur., 140:69-80. Voogd, J. 1964 The Cerebellum of the Cat. Structure and Fibre Connexions. Thesis. Assen, Van Gorcum, 215 pp.
469
1969 The importance of fiber connections in the comparative anatomy of the mammalian cerebellum. In: Neurobiology of Cerebellar Evolution and Development. R. Linls, ed., Chicago, Illinois, pp. 493-514. Walberg, F. 1956 Descending connections to the inferior olive. An experimental study i n the cat. J. Comp. Neur., 104: 77-174. 1960 Further studies on the descending connections to the inferior olive: Reticulo-olivary fibers: a n experimental study i n the cat. J. Comp. Neur., 114: 79-87. 1974 Descending connections from the mesencephalon to the inferior olive: An experimental study in the cat. Exp. Brain Res., 20: 145-156. Walberg, F., A. Brodal and G. H. Hoddevik 1976 Notes on the method of retrograde transport of horseradish peroxidase a s a tool in studies of afferent cerebellar connections, particularly those from the inferior olive, with comments on the orthograde transport of horseradish peroxidase. Exp. Brain Res., i n press. Warr, W. B. 1973 Localization of olivocochlear neurons by means of retrograde axonal transport of horseradish peroxidase. Anat. Rec., 175: 464.