The Olivocerebellar Projection Studied with the Method of Ret r og r ade Axon a I T r a n-spo r t of Ho r se r ad is h Per o x ida se V. THE PROJECTIONS TO THE FLOCCULONODULAR LOBE AND THE PARAFLOCCULUS IN THE RABBIT G . H. HODDEVIK A N D A. BRODAL Anatomical Institute, IJniuersLty of Oslo, Oslo 1. Niiru'ay

ABSTRACT HRP was injected in t h e flocculonodular lobe and the paraflocculus in t h e rabbit to determine t h e areas of the inferior olive which project onto these cerebellar regions. Following injections in the flocculus labeled cells occurred in the dorsal cap and t h e rostralmost tip of the medial accessory olive. Following injections in the nodulus labeled cells were likewise found in t h e dorsal cap, but in addition in the rostralmost part of the dorsomedial cell column and the adjoining part of the medial accessory olive. Injections in the dorsal paraflocculus gave rise to labeling in the rostrolateral part of the medial accessory olive, while injections in t h e ventral paraflocculus resulted in labeling in the principal olive, mainly in the lateral part of the ventral lamella. Injections in the lateral third of the dentate nucleus gave rise to labeling mainly in t h e dorsal lamella of the principal olive. The results are discussed with reference to those obtained by previous authors. There are both similarities and discrepancies. I t appears from what is known of afferents from areas mediating visual impulses to the inferior olive t h a t the olivary areas projecting onto the flocculonodular lobe, and possibly the dorsal paraflocculus, may mediate visual impulses to these lobules. The present paper is a link in a planned experimental mapping of t h e entire olivocerebellar projection with t h e method of retrograde axonal transport of horseradish peroxidase (HRP). So far the projections to the paramedian lobule (A. Brodal et al., '751, the uvula (A. Brodal, '76), the vermal visual area (Hoddevik et al., '76) and the anterior lobe (A. Brodal andwalberg, '77) have been described. These studies were all made in t h e cat. In the present study the rabbit was chosen as the experimental animal for two reasons. 1. The technical problems involved i n achieving HRP-injections localized t o t h e different parts of the paraflocculus-flocculus complex a r e less in the rabbit t h a n in the cat. 2. Recent physiological experiments on the olivary projections to t h e flocculus and nodulus (see below) have been made in t h e rabbit. The topography and cerebellar projection of the inferior olive are very similar in t h e rabbit and the cat (A. Brodal, '40b). A particular aim of the present study was to see if precise information could be obtained of t h e olivary relay for transmission of impulses J . COMP. N E U R . , 176: 269-280

of visual origin to the nodulus and flocculus, as a continuation of our previous study (Hoddevik et al., '76) on the olivary relay of the vermal visual area of Snider and Stowell ('42, '44). a visuo-cerebellar pathway to the vestibulo-cerebellum (flocculus and nodulus), relayed in t h e inferior olive, was first demonstrated by Maekawa and Simpson ('72, '73). They observed climbing fiber responses in t h e flocculus and nodulus following electrical stimulation of the optic disc and optic chiasm and following light flash stimulation. The impulses were found to be transmitted to the flocculus and nodulus via the tractus opticus accessorius, i t s terminal nucleus in t h e mesencephalon, and t h e inferior olive. The olivary region concerned was supposed to be the "rostromedial part of the contralateral inferior olive." The occurrence of visually evoked climbing fiber responses in the vestibulo-cerebellum has subsequently been studied physiologically and confirmed by several authors (Simpson e t al., '74; Simpson and Alley, '74; Maekawa a n d Takeda, '76; a n d o t h e r s ) . (Mossy fiber responses to visual stimulation

269

270

G. H. HODDEVIK AND A. BRODAL

have recently been described by Maekawa and Takeda " 7 5 , '761, but their precerebellar relay has not been identified.) While the primary aim of the present study was to clarify the olivary projection to t h e flocculonodular lobe it was deemed practical to include in the study also the parafloccular projections, especially since the ventral paraflocculus receives primary vestibular fibers (A. Brodal and Heivik, '64). Anatomical observations on t h e olivary projections to the flocculus, nodulus and paraflocculus are relatively scanty. Such fibers are described by A. Brodal ('40b) in studies of retrograde cellular changes in the olive following cerebellar lesions (modified Gudden method, A. Brodal, '40a). An olivary projection to t h e flocculus and nodulus is described in a more recent study of retrograde cell changes in t h e monkey (Lafleur e t al., ' 7 4 ) . In studies of fiber degeneration following lesions of t h e olive (Luthy, '32; Voogd, '64, '69)or following injections of tritiated leucine (Courville, ' 7 5 ) fibers passing to the flocculus, nodulus and parts of the paraflocculus have been found by one or more of these authors. Following t h e appearance of the papers of Maekawa and Simpson ('72, ' 7 3 ) renewed interest has been devoted to t h e determination of the site of t h e olivary relay for visual impulses. In studies appearing after the prese n t work was started Alley e t al. ('75) and Maekawa ( ' 7 5 ) describe labeling of neurons in t h e certain parts of the inferior olive of t h e rabbit following injections of HRP in the flocculus and nodulus. The findings made by t h e authors mentioned will be considered in t h e DISCUSSION. As will be seen our d a t a largely confirm those of Alley et al. ('75) and in addition bring some information of olivary projections to the paraflocculus and the dentate nucleus. MATERIALS A N D METHODS

I n 47 albino rabbits, weighing 1-2.7 kg, 50% HPR (500 pg/pl) of type Sigma VI or Serva was injected in amounts of 0.05-0.5 p1 under Mebumal (30 mg/kg) anesthesia. Twenty-six of the animals could be used in the present study. The injections in the flocculus were made stereotactically through the paramedian lobule. The nodulus was injected through lobule VIIIB or IX. The paraflocculus injections were made under visual control following surgical removal of the bony capsule covering the paraflocculus. All injections

were made with a I - p l Hamilton syringe coupled to a hydraulic perfusor. The needle was left in situ for 30 minutes after t h e stereotactic injections. One to three days postoperatively t h e animals were perfused intravitally under deep Mebumal anesthesia with a mixture of 0.4% formaldehyde and 1.25% glutaraldehyde in phosphate buffer. The brain was then dissected free, immersed for 24 hours in the fixative and transferred to a 30% sucrose phosphate buffer solution. After one to five days the brain stem together with the cerebellum were cut transversally on the freezing microtome. The 5 0 - p thick sections were activated with 3,3 diaminobenzidine and H,O, according to Graham and Karnovsky ('66) in groups of five. For further details of the technique employed, see Walberg et al. ('76). The sections were drawn in a projection apparatus, checked mciroscopically, and labeled cells were plotted in t h e drawings. In some cases a n X-Yrecorder was used. The map of A. Brodal ( ' 4 0 ~ )was used as a guide for identification of cerebellar lobules. The spreading of the injected fluid as plotted in t h e drawings of the sections was transferred to a map of t h e cerebellar surface taken from A. Brodal ('40~). To facilitate comparison among cases the findings in the particular cases were transferred to a diagram of the unfolded inferior olive taken from A. Brodal ('40b), since a comparison showed t h a t there are only minor differences between the olivary complex of the adult rabbit and the young rabbits used by A. Brodal ('40b). RESULTS

Following injections of HRP in the flocculonodular lobe and the paraflocculus labeled cells a r e always present in t h e inferior olive. However, their site in the olivary complex depends on the lobule injected. To simplify the description cases with injections in the various subdivisions will be described separately, and conclusions about the different projections will be made following the presentation of each group of cases. The weight, amount of HRP injected and t h e survival times are shown in table 1.

Flocculus In rabbit R 26 (fig. 1) the injection needle (inserted through the left paramedian lobule) has reached the center of t h e left flocculus.

271

OLIVARY PROJECTION TO VESTIBULO-CEREBELLUM

The cortex and the white matter of almost the entire flocculus a r e stained dark brown. Adjoining portions of the white matter of the paraflocculus and small patches of the parafloccular cortex are faintly stained, and there is a slight spreading of HRP to the lateral part of the dentate nucleus where the cells have a diffuse brown color. In the inferior olive labeled cells are present only contralaterally. There is a distinct difference in the distribution of heavily and faintly labeled neurons. In the dorsal cap, including the adjoining part of the ventrolateral outgrowth (levels 111-VIII in fig. l),most of the cells present are heavily labeled (fig. 2 ) . In addition some heavily labeled cells are found in the rostralmost portion of the medial accessory olive (levels XI-XI1 fig. 1). The lateral bend of the principal olive, mainly the ventral lamella, contains a number of faintly labeled cells. In two other cases with approximately identical extent of the cerebellar staining (rabbits R 17 and R 12, not illustrated) the findings in the contralateral inferior olive correspond closely to those described above. Since in these cases most of the cortex of the flocculus has taken up abundant HRP, while only small parts of the parafloccular cortex and part of the dentate nucleus are rather weakly brown, i t appears that most HRP which has been transported has passed along fibers ending in the flocculus. This suggests that the part of the olive where most of the neurons are heavily labeled (the dorsal cap and the adjoining ventrolateral outgrowth) projects to the flocculus. The faintly labeled cells in the ventral lamella of the principal olive are most likely due to the spreading to the ventral paraflocculus (see below) or to the dentate nucleus. Injections in the latter give origin to labeling mainly in the lateral half of the dorsal lamella, to some extent also in the lateral half of the ventral lamella (rabbits R 49 and 18, not illustrated, injections in the lateral third of the dentate nucleus.) The conclusion about the projection onto the flocculus is supported by the findings in a fourth case (rabbit R 28, fig. 3A) in which a circumscribed injection in the rostralmost folium of the flocculus was achieved. At the injection point of the needle of the paramedian lobule there is a tiny patch of stained cortex which is well stained. In this animal labeled cells are present in the dorsal cap only (levels 111-VII). (A single labeled cell is found

TABLE 1 Survival

Case

Weight ikgl

R2 R3 R4 R5 R 11 R 12 R 14 R 16 R 17 R 18 R 19 R 22 R 24 R 25 R 26 R 28 R 30 R 31 R 33 R 36 R 44 R 45 R 46 R 47 R 49 R 50

2.3 2.2 2.1 2.7 2.5 1.8

2.3 1.7 1.7 2.2 1.9 1.8 1.2 1.0 1.2 2.1 2.4 2.2 2.9 2.4 2.4 2.6 1.7 2.1 2.4 2.4

Amount of H R P injected igll

0.5 Sigma 0.5 Sigma 0.5 Sigma 0.5 Sigma 0.4 Sigma 0.5 Sigma 0.4 Sigma 0.3 Sigma 0.35 Sigma 0.3 Sigma 0.4 Sigma 0.3 Sigma 0.25 Serva 0.1 Serva 0.1 Serva 0.05 Serva 0.15 Serva 0.15 Serva 0.12 Serva 0.1 Serva 0.07 Serva 0.1 Serva 0.08 Sigma 0.08 Sigma 0.08 Sigma 0.08 Sigma

time

ldaysl

1 2 2 2 3 1 2 3 2 3 2 1 2 1 1 3 2 1 2 1 2 2 1

1 2 2

in the rostral half of the medial accessory olive, presumably due to the tiny staining in the paramedian lobule, see A. Brodal et al., '75.) Furthermore, in all cases with staining of the flocculus (in addition to other parts, see below) there is always labeling in the dorsal cap (rabbits R 5, 14,22 and 25). The number of labeled cells in these cases goes roughly parallel to the extent of staining of the floccular cortex. We feel safe, therefore, to conclude that the oliuary projection to the flocculus arises in the dorsal cap and the adjoining part of the uentrolateral outgrowth (fig. 41, and to a slight extent from the rostral pole of the medial accessory olive. The other areas containing labeled cell in the above cases project to other parts of the cerebellum as will be seen from the following.

Nodulus In rabbit R 47 (fig. 3B) the injection needle was inserted through the caudalmost folium of lobule VIIIB and has reached the nodulus. Most of the cerebellar cortex on the left side of the nodulus is stained. There is a slight spreading to the uvula but not to the cerebellar nuclei. In the right inferior olive labeled neurons are present in the dorsal cap (levels

272

C;

I 1 FfODDEVIK AND A BRODAL

R 26

F

B

Coudal Flocc

y

NI

LF

3

Flocc

-‘ d

P Pflv

~

Rostra1

C

XI

VI

Cnudal

U

m

m ri

?

j”

qFg2

m

Rostra1

D

Cadnl

I

Figure 1

OLIVARY PROJECTION TO VESTIRCTI,O-CEREBEI,L~JM A hhruciations

0.n u c l e u s 0 d. cap, dorsal cap d. I., dorsal lamella of t h e principal olive dm. cell col.. dorsomedial cell column Flocc., flocculus I.. lateral L. pm.. lobulus paramedianus m., medial N I.. nucleus interpositus NL., nucleus lateralis NM.. nucleus rnedialis Nod.. nodulus P. fl.d., paraflocculus dorsalis P fl v.. paraflocculus ventralis Uv.. uvula v.1.. ventral lamella of t h e principal olive v . l . 0 . . ventrolateral outgrowth VIIIB-X. cerebellar lobules of Larsell ('70)

273

nodulus received fibers from the whole extent of the dorsal cap and i n addition probably from a small area of the rostralmost part o f t h e dorsomedial cell column and the adjoining portion of the medial accessory olives. Ventral paroflocculus

In rabbit R 30 (fig. 3C) the staining of the cerebellum is entirely restricted to the ventral paraflocculus. The caudalmost portion is most strongly stained. In the inferior olive a fair number of labeled cells is present only in the lateral part of the ventral lamella contralaterally. Well labeled cells are concentrated about levels IX-X with faintly labeled cells in the periphery of the region. Corresponding findings are made in another 111-VII), in t h e rostralmost portion of the dor- case (rabbit R 31, inset in fig. 3 0 . The injecsomedial cell column (levels XI-XI11 and in tion is restricted to the rostra1 part of the left the rostralmost tip of t h e medial accessory ventral paraflocculus. A few, but strongly laolive (level XII). In another case, R 50 (not il- beled cells are found in the ventral lamella of lustrated), with only slight staining of the the right principal olive, particularly about cerebellar cortex in the nodulus, without level IX most laterally. These two cases demspreading to other cerebellar subdivisions, onstrate t h a t the lateralmost part of the vensome labeled neurons are present in the cau- tral lamella, mainly caudally, sends fibers to dal part of t h e dorsal cap only (levels 111-V). the ventral paraflocculus. I t may be noted that The findings made in these cases are sup- in all cases where there is concomitant stainported by those in five others (rabbits R 33, 36, ing of the ventral paraflocculus in addition to 44, 45, 461, all with injections in the nodulus, staining of the flocculus and/or the dorsal but with more or less spreading to the uvula. paraflocculus (R2, 5, 1 I , 12, 14, 17. 22, 24, 26) In addition to distinct labeling in the dorsal labeled cells are present in this part of the cap and in the rostralmost portion of the dor- olive. somedial cell column there are a few labeled Dorsa 1 pa raflocculus cells in the caudal part of the dorsomedial cell In rabbit R 4 (fig. 3D) practically the whole column and in the nucleus /3. Labeling in the two latter regions is most probably due to the spreading to the uvula (A. Brodal, '76). We conclude from our material that the Fig. 1 Diagrammatic representation of the findings in rabbit R 26. I n A the extent of HRP-positive area is indicated in drawings of t h e cerebellar surface and t h e cerebellar nuclei t a k e n from A. Brodal ('40b), in B in drawings of equally spaced transverse sections (1.6, from caudal t o rost r a l ) through t h e caudal part of t h e cerebellum. Black indicates heavy staining, hatchings faint staining. Dots in the nuclei indicate labeled neurons. The needle track is indicated (arrows). C: A series of drawings of equally spaced transverse sections through t h e inferior olive taken from A. Brodal ('40b). The olivary subdivisions a r e indicated on t h e left side of t h e series (medial accessory olive a n d nucleus p hatched; principal olive white; dorsal accessory olive black; dorsal cap and ventrolateral outgrowth crosshatched: dorsomedial cell column dotted). On t h e right side of t h e series labeled cells a r e indicated by dots. Different density of dotting indicates differences in number of labeled cells. Rectangle indicates t h e position of photomicrograph in figure 2 . D: Labeled cells a r e indicated by dots in a diagram of t h e unfolded olive t a k e n from A. Brodal ('40b, see explanatory drawings 1. 2 and 3 below)

Fig. 2 Photomicrograph showing t h e density of labeled cells in the dorsal cap in rabbit R 26 a t level I V (see r t v tangle in fig 1Cl X 105

274

G. H. HODDEVIK A N D A. BRODAL

28

MEDIAL ACC OLIVE

PRINCIPAL OLIVE

DORSAL ACC OLIVE

MEDIAL ACC OLIVE

PRINCIPAL OLIVE

DORSAL ACC OLIVE

MEDIAL ACC OLIVE

PRINCIPAL OLIVE

Rostrol XV

XN

xm xn XI

X

Tx

m w

VI

V N

m U

Caudal

I

R 47

C

DORSAL ACC OLIVE

N

m

Caudal

I

Fig. 3 Diagrammatic representation of cases with HRP-injections restricted to one lobule. Cortical staining is shown in t he drawings of the cerebellar surface to t h e left. Point of entrance of injection needle indicated by arrows. Labeled cells in t h e olive a re indicated by dots in drawings of selected transverse sections through the Inferior olive (Roman numbers refer to levels i n fig. 1) and in a diagram of t h e unfolded olive k p . fig. 1D). The olivary labeling in the cases shown in insets is practically identical with t h a t found in the corresponding case illustrated more fully. Symbols as in figure 1. Abbreviations: see list on p. 273.

275

OLIVARY PROJECTION T O VESTIBULO-CEREBELLUM Caudal

XI

I

A-2,-

n in la---/--,

xm

e

MEDIAL ACC OLIVE

PRINCIPAL OLIVE

DORSAL ACC OLIVE

Rostral X i

xiv Xm

xu x1 X

Ix

m

w vl V

n Caudal

I

Fig 4 Summarizing diagram showing our findings of the olivocerebellar projections onto the flocculus, nodulus and ventral and dorsal paraflocculus Abbreviations see list on p 273

left dorsal paraflocculus is stained without has proved to be valuable for studies of olivary spreading to neighboring structures. In the and pontine projections onto the cerebellum inferior olive weakly labeled cells are present in the cat (A. Brodal e t al., '75; A. Brodal, '76; only laterally in t h e rostral half of t h e medial Hoddevik et al., '76; A. Brodal and Walberg, accessory olive (levels VII-XI) on t h e right '77; Hoddevik, '75; Hoddevik et al., '77; P. side. Brodal and Walberg, '77, respectively). The Corresponding findings are made i n three method works well in rabbits (Alley et al., '75, other cases with small injections restricted to and others) a s is also our experience in t h e parts of the dorsal paraflocculus (R 3, 16 and present study. 19, inset in fig. 3D). Only in one of them (R 3) Following cerebellar injections the degree is there some staining of the lateralmost part of labeling of cells in the inferior olive varies of the dentate nucleus, which extends into the among cases, depending upon the amount of paraflocculus. In t h e inferior olive in these HRP injected and other factors. I n a n area cases there a r e labeled cells in t h e same area containing labeled cells, those in the center of the olive a s in rabbit R 4 (fig. 3D), and in R 3 are usually most heavily labeled. From our in addition a few labeled cells in the caudal present experience i t appears t h a t heavily lapart of t h e dorsal lamella a t the transition to beled cells reflect transport of HRP from the center of the HRP-positive cerebellar area, the ventrolateral outgrowth. These cases demonstrate t h a t t h e dorsal while transport from the periphery of the reparaflocculus receives its olivary afferents from gion results in faintly labeled cells (Walberg the lateral part of the rostral half of the medial e t al., '76). However, another explanation accessory olive. might be suggested, namely that well labeled Figure 4 shows a summarizing diagram of cells send their axons to t h e whole extent of our findings. t h e stained area of the cerebellar cortex, while the faintly labeled ones only give off a DISCUSSION collateral of the main axon to this area. The Methodological comments last possibility must especially be taken into The method of retrograde transport of HRP consideration in the evaluation of t h e findings

276

G. H HODDEVIK AND A. BRODAL

since some olivary regions supply more t h a n one cerebellar area (see below). Some sources of error of relevance for t h e interpretation of findings in studies with retrograde transport of HRP have recently been described. Following intravenous injections of HRP Broadwell and Brightman ('76) found widely dispersed labeled cells in t h e brain stem of mice. Although i t can scarcely be excluded that with t h e procedure employed in this study a n accidental intravascular injection may occur in addition to t h e intracortical one, this is probably a small source of error since there is always a good correlation between the extent of staining of a particular cerebellar cortical area and the number and location of labeled cells in the inferior olive. Further, endogenous HRP positive cells have been described (in t h e "extrapyramidal system" of the squirrel monkey, Wong-Riley, '76). As a control, therefore, we examined a n unoperated animal, treated in t h e same way as the others. No labeled cells were found in the inferior olive in this case.

Topography The present study was undertaken as part of a planned investigation of t h e entire olivocerebellar projection. For reasons mentioned in the introduction, the rabbit was chosen as the experimental animal, while the cat has been used in the other partial studies, to t h e paramedian lobule (A. Brodal et al., '751, uvula (A. Brodal, '76), vermal visual area (Hoddevik e t al., '76) and the anterior lobe (A. Brodal and Walberg, '77). It appears from A. Brodal's ('40b) findings t h a t the topography of the olive and t h e olivocerebellar projection in t h e cat and rabbit are essentially similar. This is supported by our HRP studies since results of injections of HRP in t h e paramedial lobule and vermal lobules VII, VIII and IX in t h e rabbit (unpublished) are in complete accordance with those previously made in the cat. Conversely, following injections in t h e dorsal and ventral paraflocculus in the cat the labeling in the inferior olive corresponds to t h a t found in the rabbit. We consider i t likely, therefore, t h a t the patterns found in the olivary projection onto the flocculus, nodulus and paraflocculus in the rabbit are valid for the cat as well. When t h e topography of the olivary projections to these lobules found in the present study (fig. 4) is compared with t h e results of

previous studies both similarities and discrepancies are present. From a n extensive study of the retrograde cellular changes in the olive following partial ablations of the cerebellum in rabbits and cats, A. Brodal ('40b) concluded t h a t the flocculus and nodulus receive fibers from two neighboring areas in the rostral half of the medial accessory olive, the paraflocculus from t h e lateral bend of the principal olive. The dorsal lamella and the ventrolatera1 outgrowth of t h e principal olive were concluded to project to the dentate nucleus. However, h e did not observe cell changes in the dorsal cap following lesions involving the flocculus or nodulus, nor in t h e rostral medial accessory olive following lesions involving the dorsal paraflocculus. The discrepancies between A. Brodal's ('40b) results and the prese n t a r e probably due to two factors. His negative findings (for example his failure to find the projection of t h e dorsal cap to t h e flocculus and nodulus) are probably due to the fact t h a t retrograde cellular changes are difficult to identify if they are moderate and result in only partial cell loss, or it may be that only few fibers to the flocculus or nodulus were injured. On the other hand, misleading conclusions concerning positive findings are probably explained by his lesions involving more t h a n one cerebellar lobule or part. For example, t h e involvement of the lateral part of the dentate nucleus in cases with lesions of the paraflocculus may have been more marked t h a n assumed, and may explain that the clearcut cell loss in t h e lateral part of the principal olive was attributed to lesions of the paraflocculus. In a more recent, less extensive study in monkey Lafleur et al. ('74) found almost total cell loss in t h e contralateral dorsal cap following lesions involving the flocculus. A small lesion in the nodulus, however, with extension into t h e fastigial and the interpositus nuclei, gave rise to cell loss in the medial and the dorsal accessory olive. In a case with a unilateral lesion which spared only the flocculus and the paraflocculus, intact cells occurred in the contralateral medial accessory olive, t h e dorsal cap and the principal olive. On the whole our results agree with those obtained by Alley et al. ('75) in a HRP study on the olivary projection onto the vestibulocerebellum in rabbits. Following injections in the flocculus (however, with spreading to the paraflocculus and the dentate nucleus) they

OLIVARY PROJECTION TO VESTIBULO-CEREBELLUM

found labeled neurons in the dorsal cap and the rostral part of t h e medial accessory olive. Injections in t h e nodulus (however, with spreading to the ventral part of the uvula) gave rise to labeled cells in t h e nucleus /3, and in a small area of t h e adjoining part of t h e dorsal cap and in the rostral third of the medial accessory olive. As assumed by the authors the labeling in the nucleus p is most probably due to the spreading to t h e uvula. To some extent data from studies with anterograde degeneration methods agree with our findings. Fibers to the flocculus were traced in Marchi studies by Luthy ('32) following midline incisions between the inferior olives. I n silver impregnated sections Voogd ('64) found degenerating fibers in t h e flocculus in three experiments with lesions of the inferior olive. Labeling was, however, not observed in the flocculus by Courville ('751, following a n injection of tritiated leucine in the rostral portion of the inferior olive, while a moderate amount of labeling was obtained in the flocculus (and in t h e caudal part of the paraflocculus) following a n injection in the lateral part of the middle third of t h e olive. None of Courville's ('75) injections involved the dorsal cap. Following lesions or injections in the inferior olive in the studies mentioned above neither degeneration nor labeling is described in the nodulus. However, Voogd ('69) briefly mentions such fibers. Neither Liithy nor Voogd traced degenerating fibers to t h e ventral paraflocculus. Courville, however, following injection in t h e lateral part of the middle third of t h e inferior olive, obtained a modera t e amount of labeling in this part of the paraflocculus. In none of the above studies was t h e dorsal paraflocculus found to receive fibers from t h e olive. Since the rostral part of t h e medial accessory olive and t h e dorsal paraflocculus both a r e extremely well developed in t h e whale, Jansen ('691, however, on the basis of a comparative anatomical study suggested t h a t this olivary region projects to the dorsal paraflocculus. The precise pattern of the olivocerebellar projection is a striking feature. From the present and from earlier studies with the H R P method in our institute i t is learned t h a t well circumscribed olivary areas project onto different lobules.' However, some olivary areas have been found to supply more t h a n one lobule, for example, t h e medial part of the

277

dorsal accessory olive projects to t h e rostral two-thirds of t h e paramedian lobule, to the intermediate part of the anterior lobe and to t h e nucleus interpositus anterior. Such multiple projection areas have been found in the present study a s well. Thus both the flocculus and the nodulus are supplied by fibers from the dorsal cap. In the olivary areas containing labeled cells following injections in the ventral paraflocculus (parts of the ventral lamella of t h e principal olive) labeled cells occur also following injections in the dentate nucleus (own results, Beitz, '76). Furthermore, areas projecting to different cerebellar parts may overlap more or less. Thus in the medial accessory olive there is considerable overlapping between the area labeled following injection in the dorsal paraflocculus (the rostrolatera1 part of the medial accessory olive) and t h a t labeled following injection in the nucleus interpositus anterior (Brodal et al., '75). There is, furthermore, a slight overlapping between this parafloccular region with t h a t projecting onto t h e uvula (Brodal, '76). When a n olivary region is found to project onto different cerebellar lobules this may indicate t h a t the axons of the cells in this region have branches which supply more than one lobule, or different cells in t h e region may supply different cerebellar parts. The evidence for a collateralization is rather strong when there is labeling of practically all cells in a region following injection in one lobule, as well as labeling of a fair number of cells within the same area following injection in another lobule. Only a s concerns t h e projections to t h e flocculus and the nodulus does the present study indicate a collateralization. Studies with anterograde, comparative anatomical and physiological methods have demonstrated a precise pattern of narrow sagitally oriented zones in the cerebellar cortex (Voogd, '64; Oscarsson, '68, '69; Korneliussen, '69; van Rossum, '69; Armstrong e t al., '74; Courville, '75; and others). In t h e paraflocculus Voogd ('64) described three parallel zones (his zones C2, D, and DP, Voogd, '691, which apparently extend into t h e flocculus. The nodulus appears to belong to his zone A. Arm-

' T h e e x a c t e x t e n t of a n olivary a r e a projecting onto a certain cerebellar lobule c a n ideally be determined only in cases where t h e e n zyme h a s been absorbed from t h e whole e x t e n t of t h e lobule without spreading to o t h e r cerebellar regiuns. Although t h i s ideal goal c a n only rarely be achieved with t h e procedure used here. a fairly good impression of t h e e x t e n t of a projecting olivary region c a n he obt a i n e d from cases where most of one lobule h a s absorbed t h e enzyme

278

G H. HODDEVIK A N D A. BRODAL

strong et al.'s ('74) study of the olivocerebellar projection, based on recording of antidromic potentials in the inferior olive following stimulation of the cerebellum covers large parts of t h e cerebellar cortex, but the flocculus, nodulus and most parts of t h e paraflocculus were not examined. However, one of their longitudinal zones is shown as comprising a part of t h e dorsal paraflocculus together with the lateral part of Crus I and 11. Following stimulation of this zone they obtained antidromic activation of neurons in the rostral medial accessory olive and the lateral part of the principal olive. The projection of t h e rostral medial accessory olive onto t h e dorsal paraflocculus is in complete agreement with our present results. However, although our injections have covered most parts of the dorsal paraflocculus, we have never obtained labeling in the principal olive in these cases. I t should be emphasized t h a t because of t h e spreading of the HRP our material is not suited for any conclusions about the existence of a longitudinal subdivision of t h e cerebellar cortex covering the paraflocculus and t h e flocculonodular lobe. The fact t h a t small injections in t h e flocculus and nodulus give rise only to labeling in t h e dorsal cap, while more extensive injections in addition give rise to labeling in the rostral medial accessory olive and the dorsomedial cell column, may indicate such a zonal subdivision. However, another possibility would be t h a t fibers from t h e medial accessory olive and t h e dorsomedial cell column do not terminate a s densely i n these lobules as do those from t h e dorsal cap.

Functional aspects The olivary regions which supply the flocculonodular lobe and t h e paraflocculus receive afferents from many and functionally different sources. The flocculus and t h e nodulus get most of their fibers from t h e dorsal cap which in anterograde degeneration studies has been shown to receive fibers from the pretectum Mizuno et al., '73 in t h e rabbit; Itoh and Kanaseki, '75 in the cat; Martin e t al., '75 in the opossum), according to A. Brodal e t al. ('50) and Mizuno ('66) some from the spinal cord, while Boesten and Voogd ('75) do not mention such fibers. A scanty projection has been described from t h e cerebral cortex, more particularly from t h e primary motor complex (Sousa-Pinto and A. Brodal, '69). Following injection of HRP involving the dorsal cap

Takeda and Maekawa ('76) obtained labeling of the dorsal and lateral terminal nuclei of the accessory optic tract. I n a n autoradiographic study Tolbert et al. ('751 found a projection from t h e nucleus interpositus anterior to the dorsal cap. According to Maekawa a n d Simpson ('73) visual impulses are transmitted along t h e posterior fascicle of the accessory optic tract to its terminal nucleus in the mesencephalon and from this to the dorsal cap and further to t h e flocculus. The function of this visual pathway has been discussed by many authors (Ron and Robinson, '73; Ito et al., '74; Maekawa and Kimura, '74; Simpson and Alley, '74; Wilson e t al., '75). Afferents to the olivary region which supply the dorsal paraflocculus, the rostralmost part of the medial accessory olive, arise in the pretectum (Mizuno e t al., '73, '74J, the mesencephalic reticular formation (Walberg, '56, '741, the caudate nucleus (Walberg, '561, and t h e cerebellar nuclei (Graybiel e t al., '73). In spite of the afferents from the pretectum, visually evoked responses have apparently not been recorded in the dorsal paraflocculus. However, recently Mortimer ('75), following auditory stimulation in monkeys, obtained climbing- and mossy fiber activation in a small area in the medialmost part of the dorsal paraflocculus (no responses occurred following visual stimulation). Previously Buser and Franchel ('60) following auditory stimulation recorded potentials in what they called t h e ansiform lobe, but what according to their figure appears to be the dorsal paraflocculus. Pathways mediating auditory impulses to the inferior olive have apparently not been described, and climbing fiber activation can probably not be taken a s proof t h a t the impulses are relayed in t h e inferior olive. The olivary region which supplies the ventral paraflocculus, the lateral part of the ventral lamella of the principal olive, has been found to receive fibers from t h e motor cortex (Sousa-Pinto and A. Brodal, '69),t h e red nucleus (Walberg, '56; Courville and O'Tabe, '741, and the cerebellar nuclei (Graybiel e t al., '731, more particularly the dentate nucleus (Tolbert et al., '75; Beitz, '76). Afferent fibers also emanate from the periaqueductal gray (Walberg, ' 5 6 ) . I t appears thus that teleceptive impulses a r e scarcely transmitted to the ventral paraflocculus via the inferior olive, while anatomical pathways which probably mediate mainly

OLIVARY PROJECTION TO VESTIBULO CEREBELLUM

visual impulses can be traced to the flocculus and the nodulus from approximately identical parts of the inferior olive. The auditory evoked climbing fiber potentials in the dorsal paraflocculus are probably not relayed in the inferior olive, but in another precerebellar nucleus. LITERATURE CITED Alley, K., R. Baker and J. 1. Simpson 1975 Afferents to the vestibulo-cerebellum and the origin of t h e visual climbing fibers in t he rabbit. Brain Res., 98: 582-589. Armstrong, D. M., R. J. Harvey a nd R. F. Schild 1974 Topographical localization in t h e olivo-cerebellar projection: An electrophysiological study in the cat. J . Comp. Neur., 152: 287-302. Beitz, A. J. 1976 The topographical organization of the olivo-dentate and dentato-olivary pathways i n the cat. Brain Res., 115: 311-317. Boesten, A. J . P., and J. Voogd 1975 Projections of the dorsal column nuclei and the spinal cord on the inferior olive in the cat. J . Comp. Neur., 161: 215-238. Broadwell, R. D., and M. W. Brightman 1976 Entry of peroxidase into neurons of t h e central and peripheral nervous systems from extracerebral and cerebral blood. J Comp. Neur., 166. 257-284. Brodal, A. 1940a Modification of Gudden method for study of cerebral localization. Arch. Neurol. Psychiat. (Chic.), 43: 46-58. 1940b Experimentelle Untersuchungen uber die olivocerebellare Lokalisation. Z. ges. Neurol. Psychiat., 169: 1-153. 1940c The cerebellum of t h e rabbit. A topographical atlas of t he folia a s revealed in transverse sections. J. Comp. Neur., 72: 63-81. 1976 The olivocerebellar projection in t h e c a t a s studied with t h e method of retrograde axonal transport of horseradish peroxidase. 11.The projection to the uvula. J Comp. Neur., 166: 417-426. Brodal, A., and B. Hsivik 1964 Site and mode and termination of primary vestibulocerebellar fibres in the cat. An experimental study with silver impregnation methods. Arch. ital. Biol., 102: 1-21. Brodal, A., and F. Walberg 1977 Theolivocerebellar projections in t h e cat studied with t h e method of retrograde axonal transport of horseradish peroxidase. IV. The projection to the anterior lobe. J . Comp. Neur , 172: 85-108. Brodal, A., F. Walberg and Th. Blackstad 1950 Termination of spinal afferents to inferior olive in cat. J . Neurophysiol., 13: 431-454. Brodal, A., F. Walberg and G €3. Hoddevik 1975 The olivocerebellar projection in the cat studied with the method of retrograde axonal transport of horseradish peroxidase. I The projection to the paramedian lobule. J. Comp. Neur.. 164: 449-470. Brodal, P., and F Walberg 1977 The pontine projection to the cerebellar anterior lobe An experimental study in the cat with retrograde transport of horseradish peroxidase. Exp. Brain Res., in press. Buser, P., and H. Franchel 1960 Existence d’un foyer de projection sensorielles acoustique a u niveau du lobe ansiforme du cervelet chez le Chat. C. R. Acad. Sci. (Paris), 251: 791-793. Courville, J. 1975 Distribution of olivocerebellar fibers demonstrated by a radioautographic tracing method. Brain Res., 95: 253-263. Courville, J., and S. Otabe 1974 The rubro-olivary projec-

279

tion in t h e macaque: An experimental study with silver impregnation methods. J. Comp. Neur., 158: 479-494. Graham, R. C. J r . , and M. J . Karnovsky 1966 The early stages of absorption of injected horseradish peroxidase in t h e proximal tubules of mouse kidney: Ultrastructural cytochemistry by a new technique. J. Histochem. Cytochem., 14: 291-302. Graybiel, A. M., H. J. W. Nauta, R . J. Lasek and W. J. H. Nauta 1973 A cerebello-olivary pathway in t h e cat: An experimental study using autoradiographic tracing techniques. Brain Res., 58: 205-211. Hoddevik. G. H. 1975 The pontocerebellar projection ont o t h e paramedian lobule i n t h e cat: An experimental study with t h e use of horseradish peroxidase a s a tracer. Brain Res., 95: 291-307. Hoddevik, G. H., A. Brodal, K. K a w a m u r a a n d T. Hashikawa 1977 The pontine projection t o t h e cerebellar vermal visual area studied by means of retrograde axonal transport of horseradish peroxidase. Brain Res., 123: 209-228. Hoddevik, G. H . , A. Brodal a n d F. Walberg 1976 The olivocerebellar projection i n the cat studied with the method of retrograde axonal transport of horseradish peroxidase. 111. The projection to t h e vermal visual area. J . Comp. Neur., 169: 155-170. Ito, M., T. Shiida, N. Yagi and M. Yamamoto 1974 Visual influence on rabbit horizontal vestibulo-ocular reflex presumably effected via t h e cerebellar flocculus. Brain Res., 65: 170-174. Itoh, K., and T. Kanseki 1975 The efferent projections of t h e pretectum in t h e cat. X ICA Kyoto Symp. Jansen, J . 1969 On cerebellar evolution a n d organization from t h e point of view of a morphologist. I n : Neurobiology of Cerebellar Evolution a n d Development. R. Llinas, ed. Amer. Med. Ass. Education & Research Foundation, Chicago, pp. 881-893. Korneliussen, H. K. 1969 Cerebellar organization in t h e light of cerebellar nuclear morphology and cerebellar corticogenesis. In: Neurobiology of Cerebellar Evolution and Development. R. Llinas, ed. Amer. Med. Ass. Education & Research Foundation, Chicago, pp. 515-523. Lafleur, J . , J. De Lean and L. J. Poirier 1974 Physiopathology of the cerebellum in t h e monkey. P a r t 1. Origin of cerebellar afferent nervous fibers from t h e spinal cord a n d brain stem. J . Neurol. Sci., 22: 471~490. Larsell, 0. 1970 The Comparative Anatomy a n d Histology of t h e Cerebellum from Monotremes through Apes. J . J a n s e n , ed. The University of Minnesota Press, Minneapolis, 269 pp. Luthy, F. 1932 Ueber anatomische Beziehungen der unteren Olive zum Kleinhirn. Proc. int. neurol. Congr., Bern 1931. Cit. from Zbl. ges. Neurol. Psychiat., 61: 498. Maekawa, K., and M. Kimura 1974 Inhibition of climbing fiber responses of rabbit’s flocculus Purkinje cells induced by light stimulation of t h e retina. Brain Res., 65: 347-350. Maekawa, K., and J. I. Simpson 1972 Climbing fiber activation of Purkinje cells in t h e flocculus by impulses transferred through t h e visual pathway. Brain Res.. 39: 245.251. 1973 Climbing fiber responses evoked in vestibulocerebellum of rabbit from visual system. J . Neurophysiol., 36: 649-666. Maekawa, K., and T. Takeda 1975 Mossy fiber responses evoked in t h e cerebellar flocculus of rabbits by stimulation of t h e optic pathway. Brain Res., 98: 590-595. 1976 Electrophysiological identification of t h e climbing and mossy fiber pathways from t h e rabbit’s reti-

280

G. H. HODDEVIK AND A. BRODAL

n a to t h e contralateral cerebellar flocculus. Brain Res., 109: 169-174. Martin, G. F., R. Dom.. J. S. King, M. RoBards and C. R. R. Watson 1975 The inferior olivary nucleus of t h e opossum (Didelphrs marsupials utrginianal, its organization and connections. J. Comp. Neur., 160: 507-534. Mizuno, N. 1966 An experimental study of t h e spino-olivary fibers in t he rabbit and t h e cat J. Comp. Neur., 127: 267-291. Mizuno. N., K. Mochizuki, C. Akimoto a nd R. Matsushima 1973 Pretectal projections to the inferior olive in t h e rabbit. Exp. Neurol.. 39: 498-506. Mizuno. N.. Y. Nakamura and N. Iwahori 1974 An electron microscopic study of the dorsal cap of t h e inferior olive in th e rabbit. with special reference to t h e pretecto-olivary fibers. Brain Res.. 77: 385-395. Mortimer. J. A. 1975 Cerebellar responses to teleceptive stimuli in alert monkeys. Brain Res., 83: 369-390. Oscarsson. 0. 1968 Termination and functional organization of the ventral spino-olivocerebellar path J . P h y s ~ iol. (London), 196: 453-478. 1969 The sagittal organization of the cerebellar anterior lobe a s revealed by the projection patterns of t h e climbing fiber system. In: Neurobiology of Cerebellar Evolution and Development. R. Llinas, ed. Amer. Med. Ass. Education & Research Foundation, Chicago, pp. 525537. Ron. S., and D. Robinson 1973 Eye movements evoked by cerebellar stimulation in the alert monkey. J. Neurophysiol.. 36: 1004-1022. Rossum. J. van 1969 Corticonuclear and corticovestibular projection of the cerebellum. An experimental investigation of t he anterior lobe, t h e simple lobule and t h e caudal vermis in t he rabbit. Proefschrift. Te Assen Bij. van Gorcum & Comp. N. V., Leiden, 169 pp. Simpson. J. 1.. and K. E Alley 1974 Visual climbing fiberinput to rabbit vestibulo-cerebellum: a source of direction-specific information. Brain Res., 82: 302-308. Simpson. J. I., W. Precht and R. Llinas 1974 Sensory separation in climbing a nd mossy fiber inputs to cat vestibulocerebellum. Pflugers Arch., 351: 183-193.

Snider, H.S.. and A. Stowell 1974 Evidence of a projection of t h e optic system to the cerebellum. Anat. Rec., 82: 448449. 1944 Receiving area of the tactile, auditory and visual systems in t h e cerebellum. J. Neurophysiol., 7. 331-358. Sousa-Pinto, A,, and A. Brodal 1969 Demonstration of a somatotopical pattern in t h e cortico-olivary projection in the cat. An experimental-anatomical study. Exp. Brain Res., 8: 364-386. Takeda, T., a n d K. Maekawa 1976 The origin of the pretecto-olivary tract. A study using t h e horseradish peroxidase method. Brain Res., 117: 319-325. Tolbert, D. L., M. G. Murphy, P. A. Young and L. C. Massopust 1975 The topographical organization of the cerehello-olivary pathway in t h e cat. Neurosci. Ahstr., 1: 216. Voogd, J. 1964 The Cerebellum of t h e Cat. Te Assen Bij, van Gorcum & Comp. N. V . . Leiden, 215 pp. 1969 The importance of fiber connections in t h e comparative anatomy of t h e mammalian cerebellum. I n : Neurobiology of Cerebellar Evolution and Development. R. Llinas, ed. Amer. Med. Ass. Education & Research Foundation. Chicago, pp. 493-514. Walberg. F. 1956 Descending connections to t h e inferior olive. An experimental study in the cat. J. Comp. Neur., 104: 77-174. 1974 Descending connections from t h e mesencephalon to t h e inferior olive: An experimental study in t h e cat. Exp. Brain Res.. 21: 145-156. Walberg. F., A. Brodal and G. H. Hoddevik 1976 A note on t h e method of retrograde transport of horseradish peroxidase as a tool in studies of afferent cerebellar connections, particularly those from the inferior olive; with comments on t h e orthograde transport in Purkinje cell axons. Exp. Brain Res., 24: 383-401. Wilson, V. J., M. Maeda and J. I. Franck 1975 Input from neck afferents to t h e cat flocculus. Brain Res., 89: 133138. Wong-Riley, M. T. 1976 Endogenous peroxidatic activity in brain stem neurons a s demonstrated by their staining with diaminobenzidine in normal squirrel monkeys Brain Res., 108: 257-277.

N ot e added in proof: Further studies in rabbits and cats have shown that there may be

minor species differences in the olivary projection onto the flocculo-nodular lobe and the paraflocculus.