The pontocerebellar projection onto the paramedian lobule in the cat: An experimental study with the use of horseradish peroxidase as a tracer

Brain Research, 95 (1975) 291-307 © Elsevier Scientific Publishing Company, Amsterdam - Printed in The Netherlands

291

THE PONTOCEREBELLAR PROJECTION ONTO THE PARAMEDIAN LOBULE IN THE CAT: AN EXPERIMENTAL STUDY WITH THE USE OF HORSERADISH PEROXIDASE AS A TRACER

GRETHE HELLSTROM HODDEVIK

Anatomical Institute, University of Oslo, Oslo (Norway)

SUMMARY

Horseradish peroxidase (HRP) was injected into cerebellar cortex of the paramedian lobule in 12 cats, and the ensuing distribution of labeled cells in the pontine nuclei was mapped in some detail. The cells in the pontine gray which give origin to fibers to the paramedian lobule lie together, in part in groups, and in part in columns. The columns are situated both medial and ventrolateral to the peduncle, as well as in the dorsolateral pontine nucleus. The projection is bilateral with a clearcut contralateral preponderance, except in the lateralmost region in the dorsolateral nucleus, which projects mainly ipsilaterally, The column medial to the peduncle projects in a topographical pattern to the paramedian lobule. The dorsal part of this column projects to the rostral folia of the paramedian lobule, while successively more ventral parts in the column project to more caudal paramedian lobules. Within the other columns only a faint sign of a topographical organization is found. The location of the pontine columns projecting onto the paramedian lobule largely corresponds to the pontine terminal areas of fibers from the sensory cerebral cortex (SmI and SmlI). The corresponding topography in these parts of the corticopontine and pontocerebellar pathways is suitable for a somatotopical impulse transmission from the sensory cortex to the paramedian lobule, in agreement with the results of physiological investigations. Furthermore, a correlation of the pontine areas projecting onto the paramedian lobule with the terminal areas of pontine afterents shows that the pons may be a relay station in mediating influences from other parts of the cortex (MsI, visual and acoustic), the cerebellar nuclei and the colliculi to the paramedian lobule.

INTRODUCTION

Physiological studies have elucidated various aspects of the influence of the

292 cerebral cortex on the cerebellum. Thus the impulse transmission from certain parts of the cerebral cortex to parts of the cerebellum, particularly the anterior lobe and the paramedian lobule, shows a somatotopical pattern (for a recent review see ref. 18). A number of pathways are available for impulse transmission from the cerebrum to the cerebellum (for a recent review, see ref. 4). However, until recently little has been known of whether any of these pathways are anatomically organized in a way that can explain the physiologically determined somatotopical cerebrocerebellar relations. The most massive cerebrocerebellar pathway is that via the pons. Earlier anatomical studies have brought forward evidence for a rather crude topographical relation only between the cerebral cortex and the pons 29,35 in the monkey and between the pons and the cerebellum in the rabbit and catn,3L Recently P. Brodal 8-13 has shown that by making small lesions in the cerebral cortex and tracing the ensuing degenerated fibers in silver impregnated sections, it is possible to map the corticopontine projection in considerable detail. His studies show that various cortical areas project to one or more well circumscribed cell columns in the pons. Within the projection from the cortical areas Ms, Sml 8, and SmlI 9 there is a somatotopical pattern. Since the first link of the cerebrocerebellar projection via the pons is thus precisely organized, it is indeed likely that a corresponding principle is valid for the second link, the pontocerebellar projection, more particularly the projection onto those cerebellar lobules which are known to be influenced in a somatotopical way from the cerebral cortex. Whether this is the case cannot be decided by means of retrograde cell degeneration studies (see Discussion). However, the recently introduced method of studying the retrograde transport of horseradish peroxidase (HRP) has made it possible to study the origin of fibers in a far more precise manner than has been possible so far. It was, therefore, decided to study the source of cerebellar afferents from the pons (and other brain stem nuclei) with this method, by injecting peroxidase solution in the cerebellum and mapping the labeled cells in the pons. The present study is restricted to a small part of this problem: the pontine projection onto the paramedian lobule. The reticular tegmental nucleus which most often is considered as a part of the reticular formation 3, is not included in this study. The aim of the present study has been the following. (1) To determine in as great detail as possible the pontine regions which send fibers to the paramedian lobule. (2) To determine if there is any topographical correlation of different parts of this lobule with certain parts of the pontine regions projecting onto it, i.e., to see if the somatotopical pattern in the paramedian lobule is reflected within its pontine projection area(s). (3) To determine if the pontine area(s) projecting onto the paramedian Iobule coincide with any of the areas receiving cortical afferents. (4) If questions 2 and 3 can be answered positively, to see if the somatotopical patterns in the corticopontine and pontocerebellar projection can be correlated.

293 MATERIAL AND METHODS

The material used in the present study is part of that used by other workers in our laboratory for studies of the olivocerebellar projection 7, of the cuneocerebellar projection ~1 and of the projection from the lateral reticular nucleus x4. It consists altogether of 12 adult cats, weighing from 1.8 to 3.2 kg and two kittens weighing 750 and 350 g, respectively. The latter and one of the adult cats were excluded because of inadvertent spread of the fluid injected. Solutions of H R P (Sigma VIP 8375) were injected in the paramedian lobule under Nembutal anesthesia with the cat fixed in a Horsley-Clarke frame. Injections were made with a Hamilton syringe coupled to a Braun Melsungen perfusor. The cannula was adjusted to extend for 1.5 mm intracortically. The amount of H R P solution injected varied from 0.2 to 0.5/A. The concentration of the solution was 50 % in most cases and in a few cases 75 or 100 %. All injections were carried out slowly in the course of 30 min. The survival time varied from 1 to 4 days, most often it was 2 days. The animals were killed in deep Nembutal anesthesia by means of intracardiac perfusion of a solution containing 0.4 % formaldehyde and 1.25 % glutaraldehyde in 0.1 M phosphate buffer at pH 7.4. The cerebellum and the brain stem were separated and immersed in the same solution for postfixation. The following day the preparation was moved to phosphate buffer with 30 % sucrose added. On the following day the blocks were sectioned on the freezing microtome, the cerebellum in the sagittal plane and the brain stem in the transverse plane. The 50 /~m thick sections were transferred to a solution containing 0.95% 3,3'diaminobenzidine tetrahydrochloride in Tris.HCl buffer (pH 7.6) for 5 min at room temperature. They were then incubated for a further 15 min at room temperature in a similar solution but with the addition of 66 #1 of 30 % hydrogen peroxide per 100 ml (see also ref. 20 and others). One series of sections was mounted unstained, another was counterstained with thionin or cresyl violet. The sections of the pons and the cerebellum were drawn under a projection apparatus. In the latter drawings the stained areas of the cerebellum were indicated as checked under the microscope and entered in a diagram of the paramedian lobule taken from Larsell zz. In the drawings of the pontine nuclei areas of labeled cells could often be seen under the projection apparatus. The distribution of labeled cells was subsequently checked under the microscope. To facilitate the comparison of cases, the findings in each case have been transferred to a standard diagram of the pontine nuclei in the cat taken from Brodal and Jansen 6. The various pontine nuclei are indicated in Figs. 2, 4, 5 and 7. The cytological appearance of the various nuclei was used as a guide in transferring the data to the diagram. The borders between the nuclei are to a large extent arbitrary. The most stable indicator of the pontine level is the appearance of the dorsolateral gray. Even if the areas of labeled cells are seldom confined to one pontine nucleus, but most often are lying on the borders between two or more nuclei, the subdivision into nuclei will facilitate the description of the results. RESULTS

Despite attempts to make all injections as identical as possible, there are some

294

Fig. 1. P h o t o m i c r o g r a p h s f r o m the pontine nuclei in cat B. St. L. 619. a: low power view of a transversely cut cell c o l u m n in the medial part of the pons at the site indicated in Fig. 2. Note the sharp limit to the s u r r o u n d i n g s . D a r k field illumination, x 50. b: part of the s a m e c o l u m n under higher power. M o s t of the cells appear spherical. D a r k field illumination. /, 260. c: high power view s h o w s labeling o f the dendrites o f the cells. Only the lower cell is in focus. D a r k field illumination, x 520. d: in the cell c o l u m n ventrolateral to the peduncle (see level VIII in Fig. 2) the cells are oriented parallel to the surface of the brain stem. D a r k field illumination. :, 260. e : the cells in the ventrolateral c o l u m n have a n elongated shape. Dark field illumination. ,,~ 520. f: cells f r o m the medial c o l u m n s h o w n in a. The H R P granules are clearly seen, m a n y of t h e m are aggregated close to the nucleus (n). Interference microscopy. × 1100.

295 variations among cases with regard to the extension of the staining of the cerebellar cortex. In the diagram of the paramedian lobule the entire area stained brown is shown hatched, and the region with a colored granular layer is indicated in black. In several cases the molecular and Purkinje cell layer are stained, while the granular layer is apparently not involved. Sometimes the granular layer is only partly brown (indicated by stippling in the figures). Sometimes the tops and/or the bottoms of the folia, apparently, are not stained, while the intermediate parts are brown. In contrast to what is the case in the cerebellum, the interpretation of the findings in the pons offers few problems. In most animals one is struck by the conspicuous labeling of the cells in certain areas of the pons (Fig. la). Cell somata as well as dendrites are labeled. Only rarely did a cell process which appears to be an axon contain brown granules. Occasionally labeled glial cells have been seen. Following injections of HRP in the paramedian lobule, labeled cells are found throughout the rostrocaudal extent of the pons. Their distribution is bilateral, but with a clear-cut contralateral preponderance. In all cases several distinct groups of labeled cells are found in the pons subsequent to a single HRP injection in the paramedian lobule. The groups can be followed in succeeding transverse sections, thus forming longitudinal columns which sometimes split into two or more parts. Often the columns have a tortuous course. Within the columns both small, medium sized and large cells are labeled, depending upon the normal cytology of the particular pontine nucleus where they are found. The different shapes of the cells in the various parts of the pons as seen in Nissl stained sections can clearly be recognized (compare Fig. lb and c with Fig. ld and e).

Presentation of cases A case in which labeled cells cover a considerable part of the pontine regions giving origin to fibers to the paramedian lobule will be described in some detail, since, in addition, it illustrates the principle of organization of the projection. In cat B.St.L. 619 (injection: 0.5 #1 50 ~ HRP, killed after 3 days, Fig. 2) the staining of the cortex is limited to the rostral 3folia of the paramedian lobule. The needle track is identified in the 3rd folium from above. As shown in the figure, the staining comprises both cortex and the white matter. Some homogeneous brown fibers can be followed to the nucleus interpositus anterior and posterior, where some cells contain granules. In most sections through the contralateral half of the pons two well circumscribed columns are found extending almost throughout its longitudinal extent. One (arrows with single head in Fig. 2) is found medial to the peduncle, and is mainly situated in the paramedian pontine nucleus, at some places extending into the peduncular nucleus. At rostral levels the column moves somewhat laterally and appears to split into two groups (Fig. 2, levels VIII and IX). Another column (arrows with double heads in Fig. 2) is found ventrolateral to the peduncle on the border between the lateral and peduncular nucleus, and likewise appears to split into smaller subdivisions at rostral levels (Fig. 2, levels VII, VIII and IX). In addition to these conspicuous columns, some smaller groups of labeled cells are present in the dorsolateral gray at some levels.

Lob rot.

L.pm.

~

//~ Lob poromed

B.St.L.619 X

ros'trol

V

N.ventr. ~

17i

Fig. 2. Diagrammatic representation of the findings in cat B. St. L. 619. Above to the left the injection site (white point) and the staining (hatchings) of the lobule is indicated in a diagram of the cerebellar surface. Black indicates staining of the granular layer. Above to the right a drawing of a section through the paramedian lobule shows the extension of the staining into the white matter. Wavy lines indicate homogeneously stained fibers passing to the nucleus interpositus anterior and posterior where a few cells contain granules. The areas of labeled cells found in the sections through the pons are transferred below to a diagram of the pontine nuclei, taken from Brodal and Jansen~. Density of dots indicates density of labeled cells. Abbreviations used in all figures: Flocc., flocculus; F. pr., primary fissure; L.pm., paramedian lobule; N. dorsolat., dorsolateral pontine nucleus; NIA, nucleus interpositus anterior; NIP, nucleus interpositus posterior; N. lat., lateral pontine nucleus; N. reed., medial pontine nucleus; N. ped., peduncular pontine nucleus; N. ventr., ventral pontine nucleus; P. ft. d., dorsal paraflocculus; P. ft. v., ventral paraflocculus; Ill-X, cerebellar lobules of Larsell.

297 One group is found dorsomedially at levels III-VI (Fig. 2), another is located more laterally at levels III and IV (Fig. 2). On the ipsilateral side the same columns and cell groups as described above are present, but they contain a smaller number of labeled cells, except for the lateralmost group in the dorsolateral nucleus which is more cell-rich than the corresponding group on the contralateral side. Three additional cases with injections in the rostral part of the paramedian lobule support the findings made in the case described. The stained parts of the paramedian lobule in these cases (cats B.St.L. 620, 650 and 636) are shown in Fig. 3. In all these cases labeled cells are found within the regions marked in cat B.St.L. 619 (Fig. 2). There is almost complete correspondence between the findings in that case and those in cats B.St.L. 620 and 636. Both of these had received injections similar in concentration and quantity to that of cat B. St. L 619, but the survival times were 4 and 2 days, respectively. In cat B.St.L. 650 injection of a smaller quantity of a concentrated H R P solution (0.05 ffl, 75 %, survival time 1 day) leads to a more restricted staining of the cortex than in the two other cases. In the pons the number of labeled cells is smaller than in the other two cases, but their distribution follows the same pattern. When different parts of the paramedian lobule are stained following an injection of HRP, the distribution of labeled cells in the pons varies somewhat. Thus in cat B.St.L. 662 (0.4 #1, 50% HRP, survival time 2 days, Fig. 4) the staining is almost completely confined to the lateral part of the first folium, with only very moderate spread to the crus II. In accordance with the very restricted staining of the cortex the areas of labeled cells in the pons are rather small, but they occur within the regions labeled in cat B.St.L. 619 (Fig. 2), and are distinct only contralaterally. Ipsilaterally, only scattered cells are seen in the corresponding regions, except for a cell group in the dorsolateral gray at level VII, which is most pronounced on this side. It should be noted that the medially situated group of labeled cells covers the dorsalmost region only of the medial column found in cat B.St.L. 619 (Figs. 2 and 4, levels II-VI) while a topical difference between the two cases is not clearly evident for the lateral column.

B.SIL.620

BStL.650

B.St.L.656

( tr~B

Fig. 3. Diagrammatic representation of injection site and stained part of the paramedian lobule in cats B.St.L. 620, 650 and 636. The findings in the pontine nuclei correspond almost completely to those made in cat B.St.L. 619, shown in Fig. 2. Symbols as in Fig. 2.

298

.-..__.

Z:7

B.St.L.662 .,~

rostral

V

I<. "" 2L_._.:2" '::

~-~

N.ventr.~

J

N.poromed.

N.j~t.~ "~ ~ ,, :~ "V'I

N.ped. v

--

,,~.~

N.dorsolat~

caudal Fig. 4. Diagrammatic representation of findings in cat B.St.L. 662 according to the same principle as used in Fig. 2. Note that the column medial to the peduncle at levels II-VI occupies the dorsal part of the corresponding column in cat B.St.L. 619 (Fig. 2). Symbols and abbreviations as in Fig. 2.

299

B.St.L. 647 X

rostral

k

7 .

.. ~.=!

V

.

,,,!:=c~3o'i'~J

N

caudal Fig. 5. Diagrammatic representation of findings in cat B.St.L. 647 according to the same principles as used in Fig. 2. Note the ventral position of the medialmost column in the medial part of the pons at levels I V . For further details concerning the columns see the text. Symbols and abbreviations as in Fig. 2. Cat B.St.L. 647 (Fig. 5) is an example of the findings made following an injection in the middle portion of the paramedian lobule. In the contralateral half of the pons labeled cells are found in areas which largely correspond to those in the preceding cases, but the position of the group medial to the peduncle is somewhat different. This group, which is present at levels I-VI, occupies a more ventral position than in the cases with injections in rostral parts of the paramedian lobule (cf levels I I - V I in Fig. 5 with Fig. 4). Furthermore, there is, at levels V-VIII, another group medial to the peduncle (arrows in Fig. 5) lying in part in the nucleus peduncularis and in part

300

B.St.L.655

B.St.L. 654

B.St.L.618

Fig. 6. Diagrammatic representation of injection sites and stained parts of the paramedian lobule in cats B.St.L. 633, 634 and 618. The findings in the pontine nuclei correspond almost completely to those made in cat B.St.L. 647 shown in Fig. 5. Symbols as in Fig. 2. in the nucleus paramedianus. Even if a cell collection is present at the same levels in cases with injections of the rostral third of the paramedian lobule, a columnar pattern cannot be discerned in these cases. In most sections between levels II and IX two groups, which at some levels appear to fuse, can be distinguished ventrolateral to the peduncle. These groups largely correspond to those found in cases with injections in the rostral third of the paramedian lobule (see Fig. 2). As in the preceding cases some labeled cells are found in the dorsolateral gray. On the ipsilateral side a few cells are found in places corresponding to the regions occupied by the columns on the contralateral side. As in the other cases there is an ipsilateral overweight only as concerns the group in the dorsolateral gray. Almost identical findings as in cat B.St.L. 647 are made in 3 other animals (Fig. 6) with similarly placed injections of the same amount and concentration of the H R P solution (cats B.St.L. 633, 634 and 618, survival times I, 3 and 2 days, respectively). When injections are made in the caudalmost part of the paramedian Iobule, the distribution of labeled cells occurs in a somewhat different pattern. As an example the findings in cat B.St.L. 656 (0.4/zl, 5 0 ~ HRP, survival time 2 days, Fig. 7) will be presented. The staining at the injection site extends considerably into the white matter, and involvement of the nucleus interpositus posterior cannot be entirely excluded. In the contralateral half of the pons the most conspicuous features are that the medialmost column, medial to the peduncle, is small and more ventrally situated than in any of the cases described above. Secondly, it is to be noted that there is a conspicuous cell column at the dorsomedial aspect of the peduncle (arrows in Fig. 7, levels I I - I X ) which has not been observed in any of the previous cases. In the dorsolateral nucleus there is labeling of a cell group with the same position as in all other cases, while there are no labeled ceils in this nucleus ipsilaterally. Corresponding findings are made in two more cases with injections in the caudal part of the paramedian lobule: cat B.St.L. 653 (0.4 ffl, 5 0 ~ HRP, survival time 3 days) and cat B.St.L. 622 (0.25 ffl, 50 ~ HRP, survival time 2 days). The staining of the cortex is illustrated in Fig. 8. In cat B.St.L. 622 most of the staining is confined to the

301

B . S t . L . 6 5 6

rostral

X

v _

_

IV"

-

Nventr.

N.ped. Ill

:7:

N.med.

N.paramed.

~,

N.iot.

-"~-'J

~

N.clorsolat. ~ - - -

v~

caudal

Fig. 7. Diagrammatic representation of findings in cat B.St.L. 656 according to the same principles as used in Fig. 2. Note the ventral position of the modest cell group close to the midline at levels ITVI. For further details concerning the columns see the text. Symbols and abbreviations as in Fig. 2.

molecular layer, but some absorption from the granular layer cannot be excluded. In the pons only scattered labeled cells are found in this case, but all of them lie within the regions containing labeled cells in cat B.St.L. 656. In some cases (among them B.St.L. 619 and 656) the fluid injected into the paramedian lobule has diffused so far into the white matter that possible absorption of H R P in the nucleus interpositus anterior or posterior cannot be excluded. However, following injections in these nuclei (cats B.St.L. 649 and 652, not illustrated) only a few scattered labeled cells are found in the pons. (See Discussion.)

302

B.St.L.622

B.St.L.655

/

,

Fig. 8. Diagrammatic representation of injection sites and stained parts of the paramedian lobule in cats B.St.L. 653 and 622. The findings in the pontine nuclei correspond almost completely to those made in cat B.St.L. 656, shown in Fig. 7. Symbols as in Fig. 2.

DISCUSSION

Since LaVail and LaVai126 first demonstrated that the retrograde axonal transport of H R P can be used to study the origin of fibers in the central nervous system, a number of students have used this method to map various fiber connections. Conditions appear to vary according to fiber systems, animal species and other factors. While the identification of labeled cells usually presents few problems, as regards the changes at the injection site, several factors at play are still incompletely known. A structure such as the cerebellar cortex presents particular problems with regard to the determination of the areas which may be assumed to have taken up the HRP. These questions have been discussed in some detail elsewhere 38. In the present study it is a consistent observation that the more extensive the cerebellar area which is stained brown, the greater is the number of labeled cells in the pontine nuclei. Further, survival times of 1--4 days appear to give approximately the same extension of staining in the cerebellum if other parameters are kept constant. It is still an open question whether all cortical areas stained really absorb the enzyme. Therefore, in order not to underestimate the area of the cerebellum which may have absorbed HRP, in the present study all parts of the cerebellar cortex stained brown to a greater or smaller degree, have been regarded as possible terminal fields of fibers emanating from the labeled cells in the pontine nuclei. The method of using HRP as a tracer has obvious advantages to the method of studying retrograde cell changes for the determination of the origin of nerve fibers. With the latter method it is often impossible to identify slightly changed cells, and the estimation of cell loss, if it is not complete, may be difficult. This method, therefore, cannot be expected to give as precise information about the sites of the cells giving origin to a projection as does the peroxidase method. This is clearly illustrated when the results of the present study are compared with those obtained with the methods of retrograde changes. Thus Brodal and Jansen 6, using the modified Gudden method 2, could only discern a rather crude pattern in the pontocerebellar projection. However,

303 concerning the projection onto the paramedian lobule it appears from their Fig. 17 (with a lesion restricted to the rostral three-quarters of this lobule) that the areas outlined as showing retrograde cell changes in the pons in this case correspond approximately to those determined here. However, from the present material it is obvious that the cells projecting to the paramedian lobule are not diffusely distributed but are arranged in particular groups, some of them forming longitudinal columns. In the present study the pontine projection to the paramedian lobule is found to be bilateral with a clear-cut contralateral preponderance. Only the dorsolateral nucleus gives off more uncrossed than crossed fibers. This agrees with the conclusions made previously concerning the pontocerebellar projection as a whole6,aL Previously available anatomical methods have not made it possible to decide whether there is a topographical pattern within the pontine projection onto the paramedian lobule. From the present study it is seen, at least concerning some of the cell groups (columns) of labeled cells, that there is a topographical relation between these and the paramedian lobule. This is particularly evident for the column situated medial to the peduncle (marked with arrows with a single head in Fig. 2). A comparison of the labeling of cells within this column following injections into the rostral (Fig. 4), middle (Fig. 5) and caudal (Fig. 7) parts of the paramedian lobule, respectively, shows that these columns occupy an increasingly more ventral position in the pons as the injection is placed at succeeding rostrocaudal levels of the paramedian lobule. When the staining of the cortex is confined to the caudal half of the lobule, there appears another group of labeled cells (arrows in Fig. 7) in the dorsolateral corner of the pontine gray, medial to the peduncle, not seen following injections in the rostral half. The size of this group increases as the injections are made increasingly further caudally. Some topographical differences in the projection onto various levels of the paramedian lobule may also be observed in the columns situated ventrolateral to the peduncle. However, the differences are small and the overlapping so extensive that conclusions as to a particular pattern can not be drawn. The above features are seen in the contralateral as well as the ipsilateral half of the pons. In the predominantly ipsilateral projection from the dorsolateral pontine nucleus, there is only faint evidence for a topographical relation with the paramedian lobule (cf. Figs. 2, 4 and 5). It may be mentioned that the dorsolateral nucleus has a heavy projection to lobule VII of the vermis (unpublished) in agreement with the observations of Brodal and Jansen 6. As mentioned above, in some cases there was some slight staining of small parts of the nuclei interpositi, obviously due to diffusion of the fluid injected. However, following injections into the NIP and NIA* only few and scattered labeled cells were present in the ports. Because of this, and since the distribution of labeled cells following injections of the paramedian lobule was similar, whether there was diffusion to the nuclei or not, it is concluded that the interpretations concerning the projection of the paramedian lobule made above are valid. In view of the topographical pattern in the pontine projection onto the parame* For abbreviations see Fig. 2 legend

304 dian lobule, demonstrated in the present study, it is of interest to compare the situation of the pontine cell columns projecting onto this lobule with the terminal fields of various contingents of pontine afferents. Some of these are known in considerable detail. It turns out that the best accordance is found with the terminal fields of fibers derived from the cerebral cortex, as might have been expected. Most of the pontine cell groups outlined in the present study coincide more or less with pontine regions receiving fibers from the cerebral cortex. The most clear-cut correspondence is found with regard to the medialmost pontine column (arrows with single head in Fig. 2) .This coincides almost completely with one of the pontine columns shown by P. Brodal 8,9 to receive fibers from SmI and SmII (first and second sensorimotor region, see Fig. 2 in ref. 9). Furthermore, a comparison of the findings in the present study with Fig. 1 l of P. Brodal s, shows that the topographical pattern of termination of cortical afferents in the column is the same as that deduced here on the basis of the well known somatotopical pattern in the paramedian lobule. Thus, there is a pathway from Sml and SmII through the pons to the paramedian lobule which is clearly somatotopically organized throughout, even if there is some overlap between the somatotopic subdivisions of this pontine cell column. However, the column is probably not to be considered purely as a link in a cerebrocerebellar pathway. Thus it also receives fibers from the cerebellar nuclei interpositus and dentatus via the descending brachium conjunctivum 5. It should be noted, as mentioned above, that injections in the very caudal part of the paramedian lobule in addition give rise to labeling of cells in another pontine area, a column at the dorsomedial aspect of the peduncle (arrows in Fig. 7). This region receives most of its cortical afferents from the motor cortex (Fig. 3 in ref. 10) and only to a slight extent fibers from SmII and SmI (see Fig. 4 in ref. 8 and Fig. 4 in ref. 9). Thus the caudalmost part of the paramedian lobule appears to have a particular route of cerebral cortical input via the pons which the rostral two-thirds lack. This is in agreement with the observation that the projection from the olive to the paramedian lobule 7 shows a similar difference between the caudalmost folia of the paramedian Iobule and the rest of it. Furthermore, the pontine cell group discussed here coincides with part of a column receiving fibers from the motor cortex Ms110 while the medialmost column, considered above, appears to receive its fibers mainly from SmI and SmII. The question may be raised whether this caudalmost part of the lobule, lying deeply covered, has ever been included in the physiological studies of the paramedian Iobule. Concerning the ventrolateral columns (arrows with double heads in Fig. 2) the situation is somewhat different. They appear to cover approximately an area which receives cerebral cortical fibers from SmI s as well as SmII 9 (see Fig. 2 in ref. 9). However, a somatotopical pattern in the projection onto the paramedian lobule is less clear, as are also the patterns in the cortical projection onto this cell column. However, these columns are also sites of convergence of afferents from other sources. Thus the orbital gyrus (its pontine projection coincides more or less with that of the cortical area SmII, see Fig. 4 in ref. 11) appears to project to a part of this column, as do the cerebellar nuclei 5 (whose pontine projection area on the whole coincides with that of Sml).

305 The pontine cell column labeled in the dorsolateral pontine nucleus is of particular interest since this nucleus has been shown to receive afferents from the superior colliculus and the inferior colliculus, as well as the auditory cortex 13,z3,24. Its projection to the vermis has been known for some time 6, and has been confirmed with the peroxidase method (ref. 23 and unpublished own experiments). A majority of its afferents appear to pass to the classical visual areas of the cerebellum (mainly lobulus VII, ref. 33). The present study shows that impulses of visual origin, in addition, may influence the paramedian lobule via the dorsolateral pontine nucleus. However, the projection onto the vermis is far more massive. It should be noted that both these projections are mainly ipsilateral, but there is in addition a topically closely related patch which has a mainly contralateral projection to the paramedian lobule. This last group corresponds to the dorsolateral column which receives afferents from the cortical area SmlIL It should be added that visual impulses from the striate cortex may be mediated to the paramedian lobule via restricted regions ventromedial to the peduncle in the rostral one-third of the pons (levels VII and IX in Fig. 2) since cortical fibers from the visual cortex end herelL It appears from the above that there are routes available for transmission of cortical impulses via the pons to the paramedian lobule mainly from SmI and SmlI as well as from the orbital gyrus, but also to some extent from motor, visual and acoustic cerebral cortical areas. Further, the paramedian lobule may be influenced via the pons from the cerebellar nuclei and the superior and inferior colliculus. Some physiological data are of relevance in this connection. The presence of a somatotopical arrangement in the cerebrocerebellar pathway to the paramedian lobule has been demonstrated repeatedly21,22,a2. This agrees with the present anatomical data. It is of particular interest that the somatotopical pattern was most clearly evident following stimulation of SmI and SmlI as shown by Hampson zl and Jansen Jr. 2z in the cat. On stimulation of MsI the pattern was less clear. This appears to correspond well with the fact that the pontine regions projecting onto the paramedian lobule receive a more massive and more clearly somatotopically organized projection from SmI and SmlI than from MsI. In his study of corticocerebellar relations in the cat Jansen Jr. ~2 found the responses in the paramedian lobule following stimulation of various cortical areas to differ in latency. Short latency responses were found when the sensory cortex was stimulated, while long latency responses occurred on stimulation o f the motor cortex. It is generally agreed upon to interpret long latency responses in the cerebellum as being due to climbing fiber activation and short latency responses to mossy fiber activation16,17, zT. Since it has been shown anatomically that the inferior olive receives its main cortical afferents from the motor cortex ~4, it is likely to suppose that in the case of the paramedian lobule the long latency responses recorded by Jansen Jr. were mediated via the inferior olive, the short latency ones via the pontine nuclei. Other physiological studies of less direct relevance to the present study have been published19, 2s, but they will not be discussed here. In recent years a longitudinal subdivision of the cerebellum has been demon-

306 strated by several a u t h o r s , for example in physiological studies o f afferents f r o m o t h e r sources16, 80. O n l y a few are concerned with the p a r a m e d i a n lobulel,l~, 36. However, a similar a r r a n g e m e n t within the p o n t o c e r e b e l l a r p r o j e c t i o n has n o t been described. Because o f the diffusion o f the p e r o x i d a s e solution in the cerebellar cortex, the present material is n o t suited for any conclusions as to the possible existence o f a longitudinal subdivision within the p o n t i n e p r o j e c t i o n o n t o the p a r a m e d i a n lobule. A s will be seen, the questions raised in the i n t r o d u c t i o n have been answered. The p o n t i n e regions sending their fibers to the p a r a m e d i a n lobule have been determined in great detail. They are f o u n d to have the shape o f cell columns. M o s t o f the c o l u m n s are located in areas which receive afferents f r o m the cortical areas S m I and S m l l , while a few coincide m o r e o r less with areas which receive fibers f r o m m o t o r , visual a n d a u d i t o r y cerebral cortical areas. In one o f the c o l u m n s the s o m a t o t o p i c a l p a t t e r n s in the c o r t i c o p o n t i n e and the p o n t o c e r e b e l l a r projections can be correlated. The p a t t e r n in the p o n t o c e r e b e l l a r p r o j e c t i o n o n t o the p a r a m e d i a n lobule is r a t h e r complex. There is reason to believe, however, t h a t further studies o f o t h e r p a r t s o f the p o n t o c e r e b e l l a r projection with the peroxidase m e t h o d will reveal it to be even m o r e complex than we m a y t h i n k t o d a y . Thus, it is likely t h a t a single c o l u m n m a y turn o u t to send fibers o r collaterals to m o r e t h a n one cerebellar subdivision, as indeed a p p e a r s f r o m some experiments (unpublished).

REFERENCES 1 ARMSTRONG,M. D., HARVEY,R. J., AND SCHILD, F. R., Topographical localization in the olivo cerebellar projection: an electrophysiological study in the cat, J. comp. Neuro[., 154 (1974) 287-302. 2 BRODAL, A., Modification of Gudden method for study of cerebral localization, Arch. Neurol. Psychiat. (Chic.), 43 (1940) 46-53. 3 BRODAL, A., The Reticular Formation of the Brain Stem. Anatomical Aspects and Functional Correlations, The Henderson Trust Lecture, Oliver and Boyd, Edinburgh, 1957. 4 BROOAL, A., Cerebrocerebellar pathways, anatomical data and some functional implications, Acta neurol, scand., Suppl. 51 (1972) 153-195. 5 BRODAL,A., DESTOMBES,J., LACERDA,A. M., AND ANGAUT,P., A cerebellar projection onto the pontine nuclei. An experimental anatomical study in the cat, Exp. Brain Res., 16 (1972) 115-139. 6 BROOAL,A., ANOJANSEN,J., The ponto-cerebellar projection in the rabbit and cat. Experimental investigations, J. comp. NeuroL, 84 (1946) 31-118. 7 BRODAL,A., WALBER6,F., ANDHODDEVIK,G. H., The olivocerebellarprojection in the cat studied with the method of retrograde axonal transport of horseradish peroxidase. 1. The projection to the paramedian lobule, J. comp. NeuroL, (1975) in press. 8 BROOAL P., The corticopontine projection in the cat. I. Demonstration of a somatotopically organized projection from the primary sensorimotor cortex, Exp. Brain Res., 5 (1968) 210-234. 9 BROOAL P., The corticopontine projection in the cat. Demonstration of a somatotopically organized projection from the second somatosensory cortex, Arch. itaL BioL, 106 (1968) 310-332. 10 BRODAL P., The corticopontine projection in the cat. I. The projection from the proreate gyrus, J. comp. Neurol., 142 (1971) 127-140. 11 BROOAL P., The corticopontine projection in the cat. 1I. The projection from the orbital gyrus, J. comp. Neurol., 142 (1971) 141-152. 12 BROOAL P., The corticopontine projection from the visual cortex in the cat. I. The total projection and the projection from area 17, Brain Research, 39 (1972) 297-317. 13 BRODAL P., The corticopontine projection in the cat. The projection from the auditory cortex, Arch. ital. BioL, 110 (1972) 119-144. 14 BRODAL P., Demonstration ofa somatotopically organized projection onto the paramedian lobule

307

15 16 17 18 19 20

21 22

23 24

25 26 27 28 29

30

31

32 33 34

35 36 37 38

and the anterior lobe from the lateral reticular nucleus: an experimental study with the horseradish peroxidase method, Brain Research, 95 (1975) 221-239. Coors, J. G., OSCARSSON,O., AND SJOLtJND, B., Termination areas of climbing fibre paths in paramedian lobule, Acta physiol, scand., 84 (1972) 37A-38A. DEtJRA, S., Long-latency cerebellar responses in cerebellar pedunculi and cortex, Neurology (Minn.), 11 (1961) 940-949. ECCL~S,J. C., ITO, M., AND SZENT.~GOTHAI,J., The Cerebellum as a Neuronal Machine, Springer, Berlin, 1967, 335 pp. EVARTS, E. V., AND TnAcrt, W. T., Motor mechanism of the CNS: cerebrocerebeUar interrelations, Ann. Rev. PhysioL, 31 (1969) 451-498. FAD~GA,E., AND PtJPILLt, G. C., Teleceptive components of the cerebeUar function, Physiol. Rev., 44 (1964) 432-486. GRAHAM,R. C., JR., AND KARNOVSKY,M. J., 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 (1966) 291-302. HAM~Or~, J. L., Relationships between cat cerebral and cerebeUar cortices, J. NeurophysioL, 12 (1949) 37-50. JANSEN, J., JR., Afferent impulses to the cerebellar hemispheres from the cerebral cortex and certain subcortical nuclei. An electro-anatomical study in the cat, Actaphysiol. scand., 41, Suppl. 143 (1957) 1-99. KAWAMUaA,K., The pontine projection from the inferior colliculus in the cat: an experimental anatomical study, Brain Research, 95 (1975) 309-322. KAWAMURA,K., AND BRODAL,A., The tectopontine projection in the cat: an experimental anatomical study with comments on pathways for teleceptive impulses to the cerebellum, J. comp. Neurol., 149 (1973) 371-390. LARSELL,O., In J. JANSEN(Ed.), The Comparative Anatomy and Histology of the Cerebellumfrom Monotremes through Apes, University of Minnesota Press, Minneapolis, Minn., 1970. LAVAIL, J. H., AND LAVAIL, M. M., Retrograde axonal transport in the central nervous system, Science, 176 (1972) 1416-1417. MAEKAWA, K., AND SIMPSON, J. L, Climbing fiber responses evoked in vestibulocerebellum of rabbit from visual system, J. Neurophysiol., 36 (1973) 649-666. MORTIMER,J. A., Cerebellar responses to teleceptive stimuli in alert monkeys, Brain Research, 83 (1975) 369-390. Nvav, O., AND JANSEN, J., An experimental investigation of the corticopontine projection in Macaca mulatta, Skr. norske Vidensk.-Akad., L Mat.-nat. Kl., 3 (1951) 47 pp. OSCARSSON,O., The sagittal organization of the cerebellar anterior lobes as revealed by theprojection patterns of the climbing fiber system. In R. LL1NAS(Ed.), Neurobiology of Cerebellar Evolution and Development, American Medical Association, Chicago, Ill., 1969, pp. 525-538. RINVIK, E., AND WALBERG, F., Studies on the cerebellar projections from the main and external cuneate nuclei in the cat by means of retrograde axonal transport of horseradish peroxidase, Brain Research, 95 (1975) 371-381. SNIDER, R. S., AND ELDRED, E., Cerebro-cerebellar relations in the monkey, J. NeurophysioL, 15 (1952) 27-40. SNIDER,R. S., ANDSTOWELL,A., Receiving areas of the tactile, auditory and visual systems in the cerebellum, J. Neurophysiol., 7 (1944) 331-358. SOUSA-PINTO,A., AND BRODAL, A., Demonstration of a somatotopical pattern in the corticoolivary projection in the cat. An experimental anatomical study, Exp. Brain Res., 8 (1969) 364-386. SUNDERLAND,S., The projection of the cerebral cortex on the pons and cerebellum in macaque monkey, J. Anat. (Lond.), 74 0940) 201-226. SZABO,I., ET ALEE-FESSARD,D., Repartition et caract6res des aff6rences somesth6siques et d'origine corticale sur le lobe param6dian du cervelet du chat, J. Physiol. Path. gdn., 46 (1954) 528-531. VOOGD,J., The Cerebellum of the Cat. Structure and Fibre Connexions, van Gorcum, Assen, 1964, 215 pp. WALBERG,F., BRODAL,A., AND HODDEVIK,G. H., The method of retrograde transport of horseradish peroxidase as a tool in studies of afferent cerebellar connections, particularly those from the inferior olives, with comments on the orthograde transport in Purkinje cell axons, Exp. Brain Res., (1975) in press.