- Email: [email protected]
The compartmentalization of the monkey and rat cerebellar cortex: zebrin I and cytochrome oxidase
Flscvh'r BRES 15087
The compartmentalization of the monkey and rat cerebellar cortex" zebrin I and cytochrome oxidase Nicole Leclerc t, Louise Dor6 ~, Andr6 Parent 2 and Richard Hawkes 1 Department of 1Biochemistry and 2Anatomy, and Laboratory of Neurobiology, Faculty of Medicine, Laval University, Que. (Canada) (Accepted 6 June 1989)
Key words: Cerebellar development; Primate; Monoclonal antibody; Cytochrome oxidase; Cerebellar compartmentalization
The cerebellar cortex of mammals is composed of parasagittal zones that encompass the afferent inputs, the efferent corticonuctear and corticovestibular projections, and a number of intrinsic molecular markers. One such marker is the polypeptide antigen zebrin I that is recognized by monoclonal antibody (mab) Ql13. In rodents, zebrin I immunocytochemistry reveals an array of parasagittat Purkinje cell compartments. In the present study, zebrin I has been used to reveal the molecular heterogeneity of the cerebellar cortex in the squirrel monkey (Saimiri sciureus). As in rodents, zebrin I is Purkinje cell specific in the primate cerebellum and not all Purkinje cells are immunoreactive. Immunocytochemistry on frontal or horizontal sections reveals a system of bands of zebrin I ÷ cells extending through the vermis of both anterior and posterior lobes. A midline (P1 ÷) band and two more lateral bands (P2 ÷ and P3 ÷) are found in all lobules. The situation in the paravermis and hemispheres is similar, with alternating zebrin I ÷ and zebrin I- compartments, but the complex lobulation obscures the precise band pattern: it seems probable that 4 additional bands are present in the hemispheres, as in rodents. Comparison of rat and monkey cerebellums suggests that the cortex has expanded in primates by the growth of the same individual bands found in rats rather than by the addition of supplementary compartments. The zebrin I compartmentalization revealed by using mab Q113 is reproducible from individual and thus provides a stable frame of reference that has been used to compare the different chemoarchitectonic patterns found in the cerebellar cortex. In the present study, the histochemical zonation of cytochrome oxidase (CO) is compared with the Purkinje cell compartmentalization revealed by zebrin I immunoreactivity in the rat and monkey cerebellar cortex. In the adult rat, the CO zonation is present in the molecular and granular layers and respects the same architectonic boundaries as zebrin I. The high CO-activity bands coincide with the zebrin ! and the CO-weak bands with the zebrin I ÷ compartments. In contrast, in the cerebellar cortex of the squirrel monkey, the CO band pattern occurs predominantly in the granular layer. Nevertheless, there is still a precise correspondence to the zebrin I zonation. However, in the monkey the CO-rich bands coincide with the zebrin I- compartments and CO-weak bands with zebrin I- zones. The differential CO expression between Purkinje cell compartments is present in rat from birth, suggesting that it is a fundamental characteristic of the zebrin I phenotype and not a secondary response to patterns of use. INTRODUCTION
expressed by a subset of Purkinje cells organized in parasagittal zones 11'13. In the vermis, zebrin I ÷ cells are
The m a m m a l i a n cerebellar cortex displays a zonal
aligned to form parasagittal bands (P+) that run through-
organization that involves afferents, efferents and intracortical compartments, as confirmed by studies of the myeloarchitecture 3°, the efferent Purkinje cell projections to the cerebellar nuclei (e.g. ref. 8), and the afferent mossy fibers 4'3°'36 and climbing fibers (see ref. 7 for
out the cortex interposed by similar bands of zebrin tcells (P-). In this report, the description of zebrin I distribution was extended to study the compartmentalization of the cerebellar cortex in primates. First, we wanted to make a preliminary survey of the primate cerebellar ground plan to learn how the cerebellar cortex may have enlarged during phylogenetic evolution. For example, the h u m a n cerebellum contains approximately 40 times more Purkinje cells than the rat 2°. How is this expansion achieved - - by the addition of new compartments or by the expansion of the same set seen in rodents? Second, we are engaged in a systematic comparison of the different cerebellar band markers in rodents by using zebrin I immunoreactivity as a stable frame of reference. We have extended the comparative analysis to the
refs.). Recently, immunocytochemistry has provided molecular markers that reveal the intrinsic compartmentalization of the cerebellar cortex. These fall into two broad classes: (1) those antigens expressed at different levels by m a n y cell types within a c o m p a r t m e n t (e.g. cysteine sulphinic acid decarboxylaseS; synaptophysin 11' 21), and (2) antigens specific to a particular cell type that is itself restricted to a subset of compartments. Most cell-specific markers are confined to the Purkinje cells. In primates, reported markers include motilin 3 and the antigen B1 ~8. Similarly in rodents, the zebrin I antigen is
Correspondence: R.B. Hawkes. Present address: Department of Anatomy and Neuroscience Research Group, University of Calgary, Calgary, Alberta. T2N 4N1, Canada. 0006-8993/90/$03.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)
71 primate cerebellar cortex. Given the marked differences
of CO staining and the zebrin I compartmentalization
b e t w e e n different m a m m a l s in the patterns of enzyme
revealed by zebrin 1 immunocytochemistry in the cere-
expression, broad assumptions about antigen distributions across species are hazardous. For example, in
bellar cortex of rats and monkeys.
mouse and rat 5'-nucleotidase (5'-N) is expressed in the molecular layer in a longitudinal zonal pattern of high and low activity23"28,29, whereas in cat, it is uniformly
MATERIALS AND METHODS
distributed. A similar parasagittal zonation was demonstrated by the acetylcholinesterase (ACHE) activity in the molecular layer of young and adult cats 23'24, and in the primate cerebellum, a n o n - u n i f o r m distribution of A C h E activity within the molecular layer, granular layer and white matter was described where AChE-rich bands in granular and molecular layer are congruent with ACHErich fiber c o m p a r t m e n t s in the white matter ~6"~. However, in rats the molecular layer is uniform but bands are seen in the granular layer ~. In the cerebellar cortex of the rat, it has already been d e m o n s t r a t e d that the m o d u l a r distributions of 5'-N and A C h E correspond precisely to the zebrin I compartmentalization, with high level of enzyme activity found in the zebrin I ÷ modules and low levels in the zebrin I modules 1"6. F u r t h e r m o r e , the olivocerebellar projection is organized in c o m p a r t m e n t s that are congruent with those revealed by using zebrin I ÷ immunocytochemistry7. In contrast to the c o m m o n plan revealed by the markers in rodents, the B1 bands do not seem to co-distribute with the motilin, or taurine bands, nor do they correspond to the A C h E staining pattern TM in primates. To extend these studies we have now compared the pattern
Monoclonal antibody (mab) QlI3 (anti-zebrin 1) is secreted by murine hybridoma cell line Ql13, produced from a mouse immunized with a crude subcellular membrane fraction from neonatal rat cerebellum. Technical details have been presented elsewhere H ~3 For immunocytochemistry 6 adult squirrel monkeys (Saimiri sciureus) were anesthetized with sodium pentobarbital and then fixed by transcardiac perfusion with 4% paraformaldehyde, 0.2% glutaraldehyde in 0.1 M phosphate buffer (pH 7.4) followed by overnight fixation by immersion in fixative without the glutaraldehyde. The cerebellum was stored in phosphate buffer plus 10 5 M sodium azide until sectioning up to 4 days later. Sections were cut at 50 ,urn with a freezing stage microtome and were incubated in anti-zebrin I overnight at room temperature on a rocker table. Both primary antibodies were used directly from spent culture medium and diluted as necessary into 10% normal horse serum in phosphate buffer. Antibody binding was detected with a second incubation for 2 h in rabbit anti-mouse immunoglobulin conjugated to horseradish peroxidase diluted 1/200 in blocking solution (Cedarlane Inc.). Peroxidase activity was revealed with a 15-min incubation in 0.5 mg/ml 4-chloro-l-naphthol, 0.01% v/v hydrogen peroxide in 0.1 M NaCI, 20 mM Tris-HCI, pH 7.5 w-~4. Sections were washed for 15 min in 3 changes of buffer between incubations. When the primary antibodies were replaced by either myeloma-conditioned medium or irrelevant mouse immunoglobulins, there was no specific staining. Cytochrome oxidase (CO) activity was visualized according to the method of Wong-Riley33. The animals were perfused through the heart with a cold saline solution followed by a solution of 4% paraformaldehyde, 0.2% glutaraldehyde in phosphate buffer, pH 7.4. The brains were post-fixed in 4% paraformaldehyde overnight. Horizontal sections (50 ~m thick) were cut on a freezing-stage microtome. For cytochrome oxidase histochemistry, the sections were preincubated in a 0.05-M Tris buffer (pH 7.6) containing 275
Fig. 1. A frontal section through the cerebellum of an adult rat immunoperoxidase stained for zebrin I. Only Purkinje cells are stained. Immunoreactive Purkinje cells are aligned to form a series of parasagittal bands (P1-P7) in both the vermis and the hemispheres. Bar = 1 mm.
72 mg/liter cobalt chloride followed by 2 washes of 10 min in Tris buffer alone. The sections were then incubated for 60 min at 40 °C in a reaction medium containing 50 mg diaminobenzidine tetrahydro-
chloride (DAB), 40 mg cytochrome oxidase per 90 ml of 0.1 M Tris buffer (pH 7.4). After 2 further washes in Tris buffer, the sections were postfixed in 10% formalin for 30 rain and dehydrated through a series of alcohols.
RESULTS The comparison
of c h e m o a r c h i t e c t o n i c
z o n a t i o n in
m o n k e y s and rats was r e s t r i c t e d to the v e r m i s f o r two reasons. First, the e l a b o r a t e h e m i s p h e r i c l o b u l a t i o n of the c e r e b e l l u m in p r i m a t e s h i n d e r s a s y s t e m a t i c c o m p a r ison of the parasagittal z o n a t i o n , and we h a v e not yet a t t e m p t e d a d e t a i l e d r e c o n s t r u c t i o n in t h e h e m i s p h e r e s . (In g e n e r a l , we can i d e n t i f y a l t e r n a t i n g b r o a d antigenic and n o n - a n t i g e n i c territories,
and p r e l i m i n a r y studies
suggest that 4 a d d i t i o n a l bands are p r e s e n t in t h e squirrel m o n k e y as in the rat.) S e c o n d , C O h i s t o c h e m i c a l z o n a t i o n in b o t h rat and squirrel m o n k e y is clear o n l y in the vermis.
Bi:
Z e b r i n I in the c e r e b e l l a r c o r t e x of the rat is c o n f i n e d i!!~
....
exclusively to a subset o f P u r k i n j e cells c l u s t e r e d to f o r m
i,!
Fig. 2. Zebrin I immunoreactivity in the cerebellum of the squirrel monkey. A: a sagittal view of immunoreactive Purkinje cells in the hemisphere. Only Purkinje cells are stained, and staining includes the dendritic arbor, the somata and the axons. Bar = 50 gm. B: a similar frontal view from the posterior lobe vermis. Immunoreactive Purkinje cell axons run through the granular layer (GL) to the white matter (WM): the Purkinje cell axons are purely ipsitateral and do not traverse the midline. Bar = I00 pm. C: Zebrin I-immunoreaclive Purkinje cell axon terminals in the dentate nucleus. Target cells are surrounded by immunoreactive punctae (arrows) but are themselves unstained. Bars = 50 ~m.
Fig. 3. Zebrin I is expressed differentially by Purkinje cell subsets in the squirrel monkey. Both in the vermis (A), and the hemispheres (B), clusters of immunoreactive Purkinje cells are interposed by similar weakly stained or unstained cells. Bar = 100 urn.
73 an array of parasagittal zebrin I+/zebrin I- compartments 13. T h e staining is restricted to a subset of Purkinje ceils a r r a n g e d in parasagittal bands extending rostrocaudally t h r o u g h o u t the cerebellar cortex and interposed by
Fig. 4. A series of low power photomicrographs to illustrate the zebrin I compartmentalization of the cerebellum in the squirrel monkey. Parasagittal bands (P1 ÷, P2 ÷, P3 +) are distinct in the vermis, less so in the hemispheres. Section A is more rostral (anterior lobe), section B lies roughly half way along the rostrocaudal extent of the cerebellum, and section C is caudal to show the posterior lobe vermis. The P1 +, P2 +, and P3 ~ compartments are tentatively indicated. The magnifications are different in panels A, B and C. Bar = 1.7 mm in A, and4 mmin B and C.
zebrin I- zones (Fig. 1). The antigenic zonation is symmetrical about the midline and is reproducible from individual to individual. T h e r e is a pair of contiguous midline bands (PI +) and 7 others (P2 +, P3 +, P4 +, P5 +, P6 ÷, P7 +, P8 ÷) distributed laterally in each hemicerebellum. In the vermis, the bands P1 +, P2 +, P3 + are continuous from lobule to lobule and b e c o m e n a r r o w e r and m o r e weakly immunoreactive as they extend rostrally through the anterior lobe. P4 + is found at the margin of the vermis and hemisphere. Z e b r i n I in the cerebellum of the squirrel m o n k e y is confined to the Purkinje cells as in rodents (Fig. 2 A , B , C ) . Within the zebrin I + Purkinje cells, peroxidase reaction p r o d u c t is found throughout the dendritic tree including the spines, the somata excluding the nucleus (Fig. 2A), the axons (Fig. 2B) and recurrent axon collaterals, and the presynaptic terminals in the deep cerebellar nuclei (Fig. 2C). Not all Purkinje cells are i m m u n o r e a c t i v e , and both in the vermis (Fig. 3A) and the hemispheres (Fig. 3B), stretches of zebrin I ÷ Purkinje cells alternate with zebrin I zones. The zebrin I- zones are less clear in m o n k e y than in rodents, and are
Fig. 5. A comparsion of zebrin I immunoreactivity (A) and CO activity (B) in adjacent frontal sections through the anterior lobe vermis of the cerebellar cortex of an adult rat. Lobules Iii, IV, and V are labelled in A. The midline P I ' , and first lateral P2 ~ compartments are labelled in A and the corresponding CO-weak regions indicatcd by arrowheads in B. Bar = 200/zm.
74 evident more in the molecular layer than in the Purkinje cell somata, which are often lightly immunoreactive. The distribution of zebrin I immunoreactivity in the vermis of the squirrel monkey is illustrated in a series of frontal sections in Fig. 4A,B,C. Clusters of antigenic Purkinje cells are arranged in parasagittal bands running throughout the vermis with a narrow band straddling the midlJne (P1 +) and 2 others running laterally to either side (P2 ÷ and P3+). The P+ bands are interposed by similar P bands. This numbering scheme follows that used for the rat cerebellum j3. Both the number and the position of the individual bands is highly reproducible from individual to individual, Consistent with the larger surface area of the cerebellar cortex in the squirrel monkey, the bands are all significantly wider. For example, in lobule VIII, P1 + and P2 ÷ average 280/~m in width in the rat and 345 ktm in the squirrel monkey. Similarly, in lobule V. P1 + in rat averages 62 /~m compared to 125 um in the squirrel monkey and the rat P2 ÷ averages 118/~m compared to 230/zm in the squirrel monkey.
Both the molecular and the granular layers of the rat cerebellar cortex contain CO activity. In the hemispheres, the activity is distributed uniformly, but in the vermis differences in staining intensity reveal alternating patches of low and high activity that include both granular and molecular layers. In the molecular layer, the staining is associated predominantly with the Purkinje cells, which are all reactive although the dendritic arbors display different levels of reactivity, in the granular layer, the patches of low and high activity align with those in the molecular layer. The CO patches are joined rostrocaudally to form two parasagittal bands in each hemivermis. Comparison with adjacent sections stained for zebrin 1 reveals that the CO-rich bands correspond precisely to the zebrin I- compartments and the CO-weak bands to zebrin 1+ compartments. Thus, in the anterior lobe of the vermis, we can distinguish 3 narrow immunoreactive bands, P1 + on the midline and two more lateral bands (P2 +, P3 +) on each side. The CO compartmentalization shares the same structural boundaries as the antigenic
Fig. 6. A comparison of zebrin I immunoreactivity (upper panels) and CO activity (lower panels) in two pairs of adjacent frontal sections through the posterior lobe vermis of the cerebellar cortex of an adult rat. Lobules VIII and IX are labelled. The midline P1 ~, and first lateral P2÷ compartments are labelled in the upper panels and the correspondingCO-weak regions indicated by arrowheads in the lower panels. Bars = 200 ,um.
75
Fig. 7. A comparison of zebrin I immunoreactivity (A) and CO activity (B) in adjacent frontal sections through the posterior lobe vermis of the cerebellar cortex of an adult squirrel monkey. The midline P1 ÷, and lateral P2 ÷ and P3 ÷ compartments are labelled (A) and the corresponding higher CO patches in the granular layer indicated by arrowheads in B. Bar = 100/~m.
zonation, with high C O in P1 and P2- and low C O in P1 ÷, P2 ÷ and P3 ÷. The same spatial correlation holds for the p o s t e r i o r vermis and again, high C O levels are associated with the zebrin I- c o m p a r t m e n t s (Figs. 5,6). In the cerebellar cortex of the squirrel m o n k e y , the C O activity is also non-uniformly distributed, but it is confined to the granular layer (Fig. 6). The same pattern occurred in the m a c a q u e 14. Despite this difference, the p r i m a t e C O c o m p a r t m e n t a l i z a t i o n shares the same structural b o u n d a r i e s as the zebrin I zonation. H o w e v e r , there is a striking difference between rat and p r i m a t e : in the squirrel m o n k e y the CO-rich bands c o r r e s p o n d to the zebrin I ÷ c o m p a r t m e n t s and, conversely, the C O - w e a k bands u n d e r l a y the zebrin I- c o m p a r t m e n t s (Fig. 7). The b a n d e d distribution of C O is only seen in the vermis. A t birth (P0) two populations of Purkinje cells can a l r e a d y be distinguished, one expressing high levels of C O reactivity and the other low levels. Fig. 8 shows two sections through the cerebellum of a newborn rat histochemically stained for CO. A t this stage of develo p m e n t , the Purkinje cells have not c o m p l e t e d their migration to the cerebellar cortex and form a multicellular layer. The staining is mainly localized in the Purkinje cell layer, but the resolution of the staining does not distinguish individual Purkinje cells clearly. Even at this early stage of d e v e l o p m e n t , the location of the C O c o m p a r t m e n t s is similar to that in the adult. As illustrated
Fig. 8. Two examples of CO activity in frontal sections through the cerebellar cortex of a newborn rat. In (A) the staining is in lobules V and VI, whereas in (B) it is in lobule IX. Most CO-activity is associated with the Purkinje cell layer (Pc). In all cases, clusters of Purkinje cells with high CO activity alternate with low activity patches: patch boundaries are indicated by arrowheads• In its general distribution, the patch arrangement is similar to that in the adult, with a low intensity patch at the midline, flanked by high activity patches to either side. Bar = 100 ltrn. in Fig. 7 A for lobule VI, a w e a k C O b a n d occupies the midline and wider clusters d e m o n s t r a t i n g a more intense staining are distributed to either side. As was previously seen by using an antiserum against cyclic G M P - d e p e n dent protein kinase 3~, there are thin acellular gaps usually separating clusters of different C O reactivity. The COweak bands in the vermis b e c o m e wider as they extend
76 into the posterior lobe, as in the adult. As far as we can determine, the congruence between zebrin I compartmentalization and CO zonation is preserved during cerebellar development: there is no zebrin I expression before P5 so no direct comparison is possible early on. DISCUSSION In the cerebellar cortex of the squirrel monkey, the immunocytochemical staining for zebrin I reveals an elaborate chemical architecture reminiscent of that in the rat. The only previous study of zebrin I in primates was an examination of postmortem human tissue that revealed Purkinje cell heterogeneity but no obvious parasagittal zonation 27. However, given the present results it seems plausible that an architecture common to rats and monkeys is also present in the human cerebellum. Direct comparison of ACHE, CO, and zebrin I compartmentalization has now revealed that all 3 share a common ground plan. Hess and Voogd 16 have described the non-uniform distribution of AChE and CO activities in the granular layer of the primate cerebellum, in particular they found a series of prominent AChE + bands in the vermis and paravermis, one astride the midline and two others parallel to either side. These bands align with bands in the white matter and, in some cases the molecular layer, and there is an exact correspondence between ACHE- and CO-rich zones. Our findings extend those of Hess and Voogd 16 by showing that CO + compartments and zebrin I + compartments are congruent. The boundaries of the CO patches in the vermis correspond to the zebrin I+/zebrin I- boundaries of the Purkinje cell compartments. The histochemical staining patterns of AChE and CO have also been shown directly to correspond to the myeloarchitectonic boundaries in the white matter 16. A similar congruence of compartments has been shown in rodents for zebrin 1 and AChE ~, the olivocerebellar projection 7, 5'-N 6, and zebrin II (unpublished observations). However, not all reported markers of chemical heterogeneity in the primate cerebellar cortex are consistent with a single underlying ground plan of the type described above for zebrin I, AChE and CO. For example, Ingram et al.lS concluded that the B1 antigenic compartments did not superimpose on those revealed by using AChE histochemistry. The results presented here have two novel features. First, the finding that high CO activity is associated with the rat P- compartments is the first report of a positive marker for this set - - all other staining protocols highlight the P+ set. Second, there is an obvious difference between rats and primates: in the monkey the CO + and zebrin I + zones overlap, whereas in the rat cerebellum they are interdigitated and CO + patches
overlap the zebrin 1 territories. This is the first time that such an interspecific difference is reported, and it suggests that care is needed in making interspecific comparisons of compartments based on their molecular phenotype. It should be born in mind that the compartmentalization revealed by CO histochemistry in rats and primates does not involve the same cell types: differential staining is molecular layer-associated in rats, and granular layer-associated in primates. In this respect, it is reminiscent of the species-dependent compartmentalization revealed by using AChE histochemistry (see ref. 1). The fact that differential CO levels are associated with cells in both the molecular layer and the granular layer may reflect differences in the pattern of use of the compartment as a whole, with high-CO compartments having either higher basal respiratory requirements or the need to respond more rapidly. This idea is consistent with the observations of Mjaatvedt and Wong-Riley 25, who found changes in the CO activity of Purkinje cells during development that reflected, the metabolic needs of the cell: when Purkinje cells receive excitatory input, their cell body CO level is high, and as the input changes to inhibitory, so the CO level declines. Such differences can arise in two ways - - either through genetic preprogramming or as a metabolic response to patterns of use. For example, CO has been reported to be a reliable marker of the neuronal metabolic activity in the visual cortex and in the SI somatosensory cortex 33-3s. Because differential CO expression by rat Purkinje cell clusters is already present at birth, a genetic explanation seems more probable. This is supported by our preliminary studies of deafferented cerebellums which have normal patterns of CO expression (Leclerc, Dor6 and Hawkes, unpublished). The same seems to be true for CO patches in the primate visual cortex 17. The cerebellum of the squirrel monkey contains many times more Purkinje cells than in the rat, and the number of Purkinje cells is even greater in humans. How is this expansion achieved? Two possibilities can be contrasted: either the cortical surface increases by the addition of new compartments or it is through the expansion in volume of a constant set of compartments. The former mechanism seems to be employed for example in the frontal cortex, where modular columns are of roughly similar sizes in many species and cortical expansion occurs by the addition of further modules (e.g. refs. 2,19). However, comparison of rat and monkey cerebellar cortices stained for zebrin I suggests that cerebellar expansion occurs not by the addition of new compartments but by the expansion in volume of a constant set. At least in the vermis, the zebrin 1 compartmentalization is rather similar in the two species, with clear P1 +, P2 + and P3 + bands identified in both, each set with similar
77 p r o p o r t i o n s . In the h e m i s p h e r e , the cerebellum of the m o n k e y has evidently u n d e r g o n e a massive expansion in surface area but we see no evidence of large numbers of new alternating zebrin I + and zebrin I- c o m p a r t m e n t s . R a t h e r , the individual zebrin I + c o m p a r t m e n t s seem to have increased in width such that very long continuous stretches of zebrin l + Purkinje cells are seen c o m p a r e d to the rat. This view is further s u p p o r t e d by comparison of the corticonuclear zonation of primates and rodents. In both the rat and the squirrel m o n k e y , the vermal A and B zones are both narrow, but in the paravermis and h e m i s p h e r e s there is a very significant relative b r o a d e n ing of the C1, C2, C3 and D zones of the squirrel m o n k e y 9. A l t h o u g h no detailed correlation has been m a d e b e t w e e n corticonuclear zones and antigenic comp a r t m e n t s , this is obviously consistent with what we see here with zebrin I, with relatively narrow vermal bands in both species and almost all the relative expansion in the m o n k e y cerebellar cortex c o n c e n t r a t e d in the zebrin I c o m p a r t m e n t s of the p a r a v e r m i s and hemispheres. Studies of the m a t u r a t i o n of antigen expression in the rat have shown that c o m p a r t m e n t a l i z a t i o n of the cerebellar cortex develops i n d e p e n d e n t l y from that of the afferent inputs 22"31'32. F u r t h e r m o r e , the crucial decisions are m a d e very early in d e v e l o p m e n t . O n e possibility is that bands of zebrin I + and zebrin 1 Purkinje cells arise by clonal expansion from a small n u m b e r of committed precursor cells, although chimera studies rather suggest a finer-grained mosaicism of n o n - r a n d o m clonal compartments in the m a t u r e cerebellar cortex 1s'26. How clonal c o m p a r t m e n t s might be related to expression of the zebrin I p h e n o t y p e is still unknown, but this general line of reasoning could explain the similarities and differences b e t w e e n rat and primate cerebellums. If all Purkinje cells arise from a small n u m b e r of early precursors, each a l r e a d y c o m m i t t e d to (but not yet expressing) its zebrin 1 p h e n o t y p e , then such a ground plan might well be very stable t h r o u g h o u t evolution: early embryonic patterns of d e v e l o p m e n t are difficult to change as their ramifications
for later d e v e l o p m e n t tend to be wide, and hence the consequences of change tend to be fatal. H o w e v e r , differences in the extent of clonal expansion can be a c c o m m o d a t e d more easily, and would result in different species having a c o m m o n set of c e r e b e l l a r c o m p a r t m e n t s , but of widely differing sizes. W h y do cerebellar and neocortical modes of expansion during evolution a p p e a r to be so different? The a p p a r e n t distinction may be because cerebellar expansion has involved only the a c c o m o d a t i o n of m o r e of the s a m e function, whereas additional cortical areas may subserve new and, perhaps, d i f f e r e n t functions. The obvious reason is that the underlying embryological mechanisms are quite different: in cerebellum, c o m p a r t m e n t s are o r d a i n e d very early in d e v e l o p m e n t and a p p e a r insensitive to postnatal influences; in cortex, columns are sculpted by competition b e t w e e n afferents and respond to postnatal stimuli. Thus, cerebellar c o m p a r t m e n t s reflect the mitotic histories of the cerebellar stem cells, whereas cortical columns are the result of the rules of e n g a g e m e n t between afferents and targets. Because of this distinction, perhaps the comparison is fundamentally misleading and the true correlation should be sought not between cerebellar c o m p a r t m e n t s and cortical modules but rather cerebellar c o m p a r t m e n t s and cortical architectonic zones. This perspective suggests that early structural subdivisions m a p out the territories in which subsequent fine afferent segregation is established. This opens the possibility of afferent segregation within cerebellar antigenic zones, This appears not to be the case for the climbing fiber p r o j e c t i o n 7, but is s u p p o r t e d by the recent finding that mossy fiber terminal fields may subdivide cerebellar antigenic c o m p a r t m e n t s both parasagittally and transversely ~. Acknowledgements. We wish to thank J. Rafrafi, R. Sasseville, L. Tremblay, and C. Harvey for technical assistance. This work was supported by grants from the Medical Research Council of Canada, the F.R.S.Q. and the F.C.A.R. (R.H., A.P.) and by a studentship from the F.R.S.Q.(N.L.).
REFERENCES 1 Boegman, R.J., Parent, A. and Hawkes, R., Zonation in the rat cerebellar cortex: patches of high acetylcholinesterase activity in the granular layer are congruent with Purkinje cell compartments, Brain Research, 448 (1988) 237-251. 2 Bugbee, N.M. and Goldman-Rakic, P.S., Columnar organization of corticocortical projections in squirrel and rhesus monkeys: similarity of column width in species differing in cortical volume, J. Comp. Neurol., 220 (1983) 355-364. 3 Chan-Palay, V., Nilaver, G., Palay, S.L., Beinfeld, M.C., Zimmerman, E.A., Wu, J.Y. and O'Donohue, T.L., Chemical heterogeneity in cerebellar Purkinje cells: existence and coexistence of glutamic acid decarbocylase-like and motilin-like immunoreactivities, Proc. Natl. Acad. Sci. U.S.A., 78 (1981) 7787-7791. 4 Chan-Palay, V., Palay, S.L.. Brown, J.T. and Van Itallie, C.,
5
6 7
8
Sagittal organization of olivocerebellar and reticulocerebellar projections: autoradiographic studies with [3SSlmethionine, Exp. Brain Res., 3(1 (1977) 561-576. Chan-Palay, V., Palay, S.L. and Wu, J.Y., Sagittal cerebellar microbands of taurine neurons: immunocytochemical demonstration by using antibodies against the taurinc-synthesizing enzyme cysteine sulfinic acid decarboxylase, Proc. Natl. Acad. Sci. U.S.A., 79 (1982) 4221-4225. Eisenman, L.M. and Hawkes, R., 5'-Nucleotidase and the mabQ113 antigen share a common distribution in the cerebellar cortex of the mouse, Neuroscience, 31 (1989) 231-235. Gravel, C., Eisenman, L., Sasseville, R. and Hawkes, R., Parasagittal organization of the rat ccrebellar cortex: a direct correlation between antigenic bands revealed by mabO113 and the topography of the olivoccrebellar projection, J. Comp. Neurol., 265 (1987) 294-310. Gravel, C. and Hawkcs, R., Parasagittal organization of the rat
7~
9
10 11
12
13
14
15
16 17
18
19
20 21
22
cerebellar cortex: direct comparison of Purkinje cell compartments and the organization of the spinocerebellar projection, J. Comp. Neurol., in press. Haines, D.E., Patrick, G.W. and Satrulee, P.. Organization of cerebellar corticonuclear fiber systems. In S.L. Palay and V. Chan-Palay (Eds.), The Cerebellum, New Vistas, Springer, Berlin, 1982, pp. 320-367, Hawkes, R., The dot immunobinding assay, Methods in F,nzvmology, 121 (1986) 484-491. Hawkes, R., Colonnier, M. and Leclerc, N., Monoclonal antibodies reveal sagittal banding in the rodent cerebellar cortex, Brain Research, 333 (1985) 359-365. Hawkes, R. and Leclerc, N., Immunocytochemical demonstration of topographic ordering of Purkinje cell axon terminals in the fastigial nuclei of the rat, J. Comp. NeuroL, 244 (1986) 481-491. Hawkes, R. and Leclerc, N., An antigenic map of the rat cerebellar cortex: the distribution of parasagittal bands as revealed by a monoclonal anti-Purkinje cell antibody mabQ113, J. Comp. Neurol., 256 (1987) 29-4l. Hawkes, R., Niday, E. and Gordon, J., A dot immunobinding assay for monoclonal and other antibodies, Anal. Biochem., 119 (1982) 142-147. Herrup, K., Cell lineage relationships in the development of the mammalian CNS: role of cell lineage in control of cerebellar Purkinje cell number, Dev. Biol., 115 (1986) 148-154. Hess, D.T. and Voogd, J., Chemoarchitectonic zonation of the monkey cerebellum, Brain Research, 369 (1986) 383-387. Horton, J.C., Cytochrome oxidase patches: a new cytoarchitectonic feature of monkey visual cortex, Phil. Trans. R. Soc. Lond., 304B (1984) 199-253. Ingrain, V.I., Ogren, M.P., Chatot, C.L., Gossels, J.M. and Owens, B.B., Diversity among Purkinje cells in the monkey cerebellum, Proc. Natl. Acad. Sci. US.A., 82 (1985) 7131-7135. Iscroff, A., Schwartz, M i . , Dekker, J,J. and Goldman-Rakic, P.S., Columnar organization of callosal and associational pro* jections from rat frontal cortex, Brain Research, 293 (1984) 213-223. Ito, M., The Cerebellum and Neural Control, Raven, New York. 1984, pp. 580. Leclerc, N., Beesley, P., Brown, I., Colonnier, M., Gurd, J.W., Paladino, T. and Hawkes, R., Synaptophysin expression during synaptogenesis in the rat cerebellar cortex, J. Comp. Neurol., 281) (1989) 197-212, Leclerc, N., Gravel, C. and Hawkes, R., Development of parasagittal zonation in the rat cerebellar cortex: mabQll3 bands are created postnatally by the suppression of antigen expression in a subset of Purkinje cells, J. Comp. Neurol., 273
(1988) 399-420. 23 Marani, E., Topographic Enzyme ttistochemistry of the Mammalian Cerebellum: 5'-Nucleotidase and Acetvlcholinestera~'e. Thesis, University of Leiden, 1982. 24 Marani, E. and Voogd, J,, An acetylcholinesterase band pattern in the molecular layer of the cat cerebellum, .L Anat,. 124 (1977) 335-345. 25 Mjaatvedt, A.E. and Wong-Riley, M.I.T., Relationship between synaptogenesis and cytochrome oxidase in Purkinje cells of the developing rat cerebellum, J. Comp. Neurol., 277 (1988) 155-182. 26 Oster-Granite, M.L. and Gearhart, J., Cell lineage analyses of Purkinje cells in murine chimeras. In 8.L. Palay and V, Chan-Palay (Eds.), The Cerebellum. New Vistas, Springer, New York, 1981, pp. 75-92. 27 Plioplys, A.V., Thibault, J. and Hawkes, R., Selective staining of a subset of Purkinje cells in the human cerebellum with monoclonal antibody mabQll3, J. Neurol. Sci., 70 (1985) 245-256. 28 Scott, E G . , A unique pattern of localization in the cerebellum, Nature (Lond.), 200 (1963) 793. 29 Scott, T.G., A unique pattern of localization within the cerebellum of the mouse. J. Comp. Neurol., 122 (1964) 1-8. 30 Voogd, J.~ The importance of fiber connections in the comparative anatomy of the mammalian cerebellum. In R. Llinas (Ed.), Neurobiology of Cerebellar Evolution and Development, Am. Med. Assoc., Chicago, 1969, pp. 493-514. 31 Wassef, M. and Sotelo, C., A synchrony in the expression of guanosine 3',5'-phostphate-dependent protein kinase by clusters of Purkinje cells during the perinatal development of rat cerebellum, Neuroscience, 13 (1984) 1217-1241. 32 Wassef, M , Zanetta, J.P., Brehier, A. and Sotelo, C., Transient biochemical compartmentalization of Purkinje celts during early cerebellar developmenL Dev. Biol., l l 1 (1985) 129-137. 33 Wong-Riley, M., Changes in the visual system of monocutarly sutured or enucleated cats demonstrable with cytoehrome oxidase histochemistry, Brain Research, 171 (1979) ll-28. 34 Wong-Riley, M.T.T. and Welt, C., Histochemical changes in cytochrome oxidase of cortical barrels following vibrissal removal in neonatal and adult mice, Proc. Natl. Acad. Sci. U.S.A., 77 (1980) 2333-2337. 35 Wong-Riley, M. and Riley, D.A., The effect of impulse blockage on cytochrome oxidase activity in the cat visual system, Brain Research, 261 (1983) 185-193. 36 Yaginuma, H. and Matsushita, M., Spinocerebellar projection fields in the horizontal plane of lobules of the cerebellar anterior lobe in the cat: an anterograde wheat germ agglutinin-horseradish peroxidase study, Brain Research. 365 (1986) 345-349.
The compartmentalization of the monkey and rat cerebellar cortex" zebrin I and cytochrome oxidase Nicole Leclerc t, Louise Dor6 ~, Andr6 Parent 2 and Richard Hawkes 1 Department of 1Biochemistry and 2Anatomy, and Laboratory of Neurobiology, Faculty of Medicine, Laval University, Que. (Canada) (Accepted 6 June 1989)
Key words: Cerebellar development; Primate; Monoclonal antibody; Cytochrome oxidase; Cerebellar compartmentalization
The cerebellar cortex of mammals is composed of parasagittal zones that encompass the afferent inputs, the efferent corticonuctear and corticovestibular projections, and a number of intrinsic molecular markers. One such marker is the polypeptide antigen zebrin I that is recognized by monoclonal antibody (mab) Ql13. In rodents, zebrin I immunocytochemistry reveals an array of parasagittat Purkinje cell compartments. In the present study, zebrin I has been used to reveal the molecular heterogeneity of the cerebellar cortex in the squirrel monkey (Saimiri sciureus). As in rodents, zebrin I is Purkinje cell specific in the primate cerebellum and not all Purkinje cells are immunoreactive. Immunocytochemistry on frontal or horizontal sections reveals a system of bands of zebrin I ÷ cells extending through the vermis of both anterior and posterior lobes. A midline (P1 ÷) band and two more lateral bands (P2 ÷ and P3 ÷) are found in all lobules. The situation in the paravermis and hemispheres is similar, with alternating zebrin I ÷ and zebrin I- compartments, but the complex lobulation obscures the precise band pattern: it seems probable that 4 additional bands are present in the hemispheres, as in rodents. Comparison of rat and monkey cerebellums suggests that the cortex has expanded in primates by the growth of the same individual bands found in rats rather than by the addition of supplementary compartments. The zebrin I compartmentalization revealed by using mab Q113 is reproducible from individual and thus provides a stable frame of reference that has been used to compare the different chemoarchitectonic patterns found in the cerebellar cortex. In the present study, the histochemical zonation of cytochrome oxidase (CO) is compared with the Purkinje cell compartmentalization revealed by zebrin I immunoreactivity in the rat and monkey cerebellar cortex. In the adult rat, the CO zonation is present in the molecular and granular layers and respects the same architectonic boundaries as zebrin I. The high CO-activity bands coincide with the zebrin ! and the CO-weak bands with the zebrin I ÷ compartments. In contrast, in the cerebellar cortex of the squirrel monkey, the CO band pattern occurs predominantly in the granular layer. Nevertheless, there is still a precise correspondence to the zebrin I zonation. However, in the monkey the CO-rich bands coincide with the zebrin I- compartments and CO-weak bands with zebrin I- zones. The differential CO expression between Purkinje cell compartments is present in rat from birth, suggesting that it is a fundamental characteristic of the zebrin I phenotype and not a secondary response to patterns of use. INTRODUCTION
expressed by a subset of Purkinje cells organized in parasagittal zones 11'13. In the vermis, zebrin I ÷ cells are
The m a m m a l i a n cerebellar cortex displays a zonal
aligned to form parasagittal bands (P+) that run through-
organization that involves afferents, efferents and intracortical compartments, as confirmed by studies of the myeloarchitecture 3°, the efferent Purkinje cell projections to the cerebellar nuclei (e.g. ref. 8), and the afferent mossy fibers 4'3°'36 and climbing fibers (see ref. 7 for
out the cortex interposed by similar bands of zebrin tcells (P-). In this report, the description of zebrin I distribution was extended to study the compartmentalization of the cerebellar cortex in primates. First, we wanted to make a preliminary survey of the primate cerebellar ground plan to learn how the cerebellar cortex may have enlarged during phylogenetic evolution. For example, the h u m a n cerebellum contains approximately 40 times more Purkinje cells than the rat 2°. How is this expansion achieved - - by the addition of new compartments or by the expansion of the same set seen in rodents? Second, we are engaged in a systematic comparison of the different cerebellar band markers in rodents by using zebrin I immunoreactivity as a stable frame of reference. We have extended the comparative analysis to the
refs.). Recently, immunocytochemistry has provided molecular markers that reveal the intrinsic compartmentalization of the cerebellar cortex. These fall into two broad classes: (1) those antigens expressed at different levels by m a n y cell types within a c o m p a r t m e n t (e.g. cysteine sulphinic acid decarboxylaseS; synaptophysin 11' 21), and (2) antigens specific to a particular cell type that is itself restricted to a subset of compartments. Most cell-specific markers are confined to the Purkinje cells. In primates, reported markers include motilin 3 and the antigen B1 ~8. Similarly in rodents, the zebrin I antigen is
Correspondence: R.B. Hawkes. Present address: Department of Anatomy and Neuroscience Research Group, University of Calgary, Calgary, Alberta. T2N 4N1, Canada. 0006-8993/90/$03.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)
71 primate cerebellar cortex. Given the marked differences
of CO staining and the zebrin I compartmentalization
b e t w e e n different m a m m a l s in the patterns of enzyme
revealed by zebrin 1 immunocytochemistry in the cere-
expression, broad assumptions about antigen distributions across species are hazardous. For example, in
bellar cortex of rats and monkeys.
mouse and rat 5'-nucleotidase (5'-N) is expressed in the molecular layer in a longitudinal zonal pattern of high and low activity23"28,29, whereas in cat, it is uniformly
MATERIALS AND METHODS
distributed. A similar parasagittal zonation was demonstrated by the acetylcholinesterase (ACHE) activity in the molecular layer of young and adult cats 23'24, and in the primate cerebellum, a n o n - u n i f o r m distribution of A C h E activity within the molecular layer, granular layer and white matter was described where AChE-rich bands in granular and molecular layer are congruent with ACHErich fiber c o m p a r t m e n t s in the white matter ~6"~. However, in rats the molecular layer is uniform but bands are seen in the granular layer ~. In the cerebellar cortex of the rat, it has already been d e m o n s t r a t e d that the m o d u l a r distributions of 5'-N and A C h E correspond precisely to the zebrin I compartmentalization, with high level of enzyme activity found in the zebrin I ÷ modules and low levels in the zebrin I modules 1"6. F u r t h e r m o r e , the olivocerebellar projection is organized in c o m p a r t m e n t s that are congruent with those revealed by using zebrin I ÷ immunocytochemistry7. In contrast to the c o m m o n plan revealed by the markers in rodents, the B1 bands do not seem to co-distribute with the motilin, or taurine bands, nor do they correspond to the A C h E staining pattern TM in primates. To extend these studies we have now compared the pattern
Monoclonal antibody (mab) QlI3 (anti-zebrin 1) is secreted by murine hybridoma cell line Ql13, produced from a mouse immunized with a crude subcellular membrane fraction from neonatal rat cerebellum. Technical details have been presented elsewhere H ~3 For immunocytochemistry 6 adult squirrel monkeys (Saimiri sciureus) were anesthetized with sodium pentobarbital and then fixed by transcardiac perfusion with 4% paraformaldehyde, 0.2% glutaraldehyde in 0.1 M phosphate buffer (pH 7.4) followed by overnight fixation by immersion in fixative without the glutaraldehyde. The cerebellum was stored in phosphate buffer plus 10 5 M sodium azide until sectioning up to 4 days later. Sections were cut at 50 ,urn with a freezing stage microtome and were incubated in anti-zebrin I overnight at room temperature on a rocker table. Both primary antibodies were used directly from spent culture medium and diluted as necessary into 10% normal horse serum in phosphate buffer. Antibody binding was detected with a second incubation for 2 h in rabbit anti-mouse immunoglobulin conjugated to horseradish peroxidase diluted 1/200 in blocking solution (Cedarlane Inc.). Peroxidase activity was revealed with a 15-min incubation in 0.5 mg/ml 4-chloro-l-naphthol, 0.01% v/v hydrogen peroxide in 0.1 M NaCI, 20 mM Tris-HCI, pH 7.5 w-~4. Sections were washed for 15 min in 3 changes of buffer between incubations. When the primary antibodies were replaced by either myeloma-conditioned medium or irrelevant mouse immunoglobulins, there was no specific staining. Cytochrome oxidase (CO) activity was visualized according to the method of Wong-Riley33. The animals were perfused through the heart with a cold saline solution followed by a solution of 4% paraformaldehyde, 0.2% glutaraldehyde in phosphate buffer, pH 7.4. The brains were post-fixed in 4% paraformaldehyde overnight. Horizontal sections (50 ~m thick) were cut on a freezing-stage microtome. For cytochrome oxidase histochemistry, the sections were preincubated in a 0.05-M Tris buffer (pH 7.6) containing 275
Fig. 1. A frontal section through the cerebellum of an adult rat immunoperoxidase stained for zebrin I. Only Purkinje cells are stained. Immunoreactive Purkinje cells are aligned to form a series of parasagittal bands (P1-P7) in both the vermis and the hemispheres. Bar = 1 mm.
72 mg/liter cobalt chloride followed by 2 washes of 10 min in Tris buffer alone. The sections were then incubated for 60 min at 40 °C in a reaction medium containing 50 mg diaminobenzidine tetrahydro-
chloride (DAB), 40 mg cytochrome oxidase per 90 ml of 0.1 M Tris buffer (pH 7.4). After 2 further washes in Tris buffer, the sections were postfixed in 10% formalin for 30 rain and dehydrated through a series of alcohols.
RESULTS The comparison
of c h e m o a r c h i t e c t o n i c
z o n a t i o n in
m o n k e y s and rats was r e s t r i c t e d to the v e r m i s f o r two reasons. First, the e l a b o r a t e h e m i s p h e r i c l o b u l a t i o n of the c e r e b e l l u m in p r i m a t e s h i n d e r s a s y s t e m a t i c c o m p a r ison of the parasagittal z o n a t i o n , and we h a v e not yet a t t e m p t e d a d e t a i l e d r e c o n s t r u c t i o n in t h e h e m i s p h e r e s . (In g e n e r a l , we can i d e n t i f y a l t e r n a t i n g b r o a d antigenic and n o n - a n t i g e n i c territories,
and p r e l i m i n a r y studies
suggest that 4 a d d i t i o n a l bands are p r e s e n t in t h e squirrel m o n k e y as in the rat.) S e c o n d , C O h i s t o c h e m i c a l z o n a t i o n in b o t h rat and squirrel m o n k e y is clear o n l y in the vermis.
Bi:
Z e b r i n I in the c e r e b e l l a r c o r t e x of the rat is c o n f i n e d i!!~
....
exclusively to a subset o f P u r k i n j e cells c l u s t e r e d to f o r m
i,!
Fig. 2. Zebrin I immunoreactivity in the cerebellum of the squirrel monkey. A: a sagittal view of immunoreactive Purkinje cells in the hemisphere. Only Purkinje cells are stained, and staining includes the dendritic arbor, the somata and the axons. Bar = 50 gm. B: a similar frontal view from the posterior lobe vermis. Immunoreactive Purkinje cell axons run through the granular layer (GL) to the white matter (WM): the Purkinje cell axons are purely ipsitateral and do not traverse the midline. Bar = I00 pm. C: Zebrin I-immunoreaclive Purkinje cell axon terminals in the dentate nucleus. Target cells are surrounded by immunoreactive punctae (arrows) but are themselves unstained. Bars = 50 ~m.
Fig. 3. Zebrin I is expressed differentially by Purkinje cell subsets in the squirrel monkey. Both in the vermis (A), and the hemispheres (B), clusters of immunoreactive Purkinje cells are interposed by similar weakly stained or unstained cells. Bar = 100 urn.
73 an array of parasagittal zebrin I+/zebrin I- compartments 13. T h e staining is restricted to a subset of Purkinje ceils a r r a n g e d in parasagittal bands extending rostrocaudally t h r o u g h o u t the cerebellar cortex and interposed by
Fig. 4. A series of low power photomicrographs to illustrate the zebrin I compartmentalization of the cerebellum in the squirrel monkey. Parasagittal bands (P1 ÷, P2 ÷, P3 +) are distinct in the vermis, less so in the hemispheres. Section A is more rostral (anterior lobe), section B lies roughly half way along the rostrocaudal extent of the cerebellum, and section C is caudal to show the posterior lobe vermis. The P1 +, P2 +, and P3 ~ compartments are tentatively indicated. The magnifications are different in panels A, B and C. Bar = 1.7 mm in A, and4 mmin B and C.
zebrin I- zones (Fig. 1). The antigenic zonation is symmetrical about the midline and is reproducible from individual to individual. T h e r e is a pair of contiguous midline bands (PI +) and 7 others (P2 +, P3 +, P4 +, P5 +, P6 ÷, P7 +, P8 ÷) distributed laterally in each hemicerebellum. In the vermis, the bands P1 +, P2 +, P3 + are continuous from lobule to lobule and b e c o m e n a r r o w e r and m o r e weakly immunoreactive as they extend rostrally through the anterior lobe. P4 + is found at the margin of the vermis and hemisphere. Z e b r i n I in the cerebellum of the squirrel m o n k e y is confined to the Purkinje cells as in rodents (Fig. 2 A , B , C ) . Within the zebrin I + Purkinje cells, peroxidase reaction p r o d u c t is found throughout the dendritic tree including the spines, the somata excluding the nucleus (Fig. 2A), the axons (Fig. 2B) and recurrent axon collaterals, and the presynaptic terminals in the deep cerebellar nuclei (Fig. 2C). Not all Purkinje cells are i m m u n o r e a c t i v e , and both in the vermis (Fig. 3A) and the hemispheres (Fig. 3B), stretches of zebrin I ÷ Purkinje cells alternate with zebrin I zones. The zebrin I- zones are less clear in m o n k e y than in rodents, and are
Fig. 5. A comparsion of zebrin I immunoreactivity (A) and CO activity (B) in adjacent frontal sections through the anterior lobe vermis of the cerebellar cortex of an adult rat. Lobules Iii, IV, and V are labelled in A. The midline P I ' , and first lateral P2 ~ compartments are labelled in A and the corresponding CO-weak regions indicatcd by arrowheads in B. Bar = 200/zm.
74 evident more in the molecular layer than in the Purkinje cell somata, which are often lightly immunoreactive. The distribution of zebrin I immunoreactivity in the vermis of the squirrel monkey is illustrated in a series of frontal sections in Fig. 4A,B,C. Clusters of antigenic Purkinje cells are arranged in parasagittal bands running throughout the vermis with a narrow band straddling the midlJne (P1 +) and 2 others running laterally to either side (P2 ÷ and P3+). The P+ bands are interposed by similar P bands. This numbering scheme follows that used for the rat cerebellum j3. Both the number and the position of the individual bands is highly reproducible from individual to individual, Consistent with the larger surface area of the cerebellar cortex in the squirrel monkey, the bands are all significantly wider. For example, in lobule VIII, P1 + and P2 ÷ average 280/~m in width in the rat and 345 ktm in the squirrel monkey. Similarly, in lobule V. P1 + in rat averages 62 /~m compared to 125 um in the squirrel monkey and the rat P2 ÷ averages 118/~m compared to 230/zm in the squirrel monkey.
Both the molecular and the granular layers of the rat cerebellar cortex contain CO activity. In the hemispheres, the activity is distributed uniformly, but in the vermis differences in staining intensity reveal alternating patches of low and high activity that include both granular and molecular layers. In the molecular layer, the staining is associated predominantly with the Purkinje cells, which are all reactive although the dendritic arbors display different levels of reactivity, in the granular layer, the patches of low and high activity align with those in the molecular layer. The CO patches are joined rostrocaudally to form two parasagittal bands in each hemivermis. Comparison with adjacent sections stained for zebrin 1 reveals that the CO-rich bands correspond precisely to the zebrin I- compartments and the CO-weak bands to zebrin 1+ compartments. Thus, in the anterior lobe of the vermis, we can distinguish 3 narrow immunoreactive bands, P1 + on the midline and two more lateral bands (P2 +, P3 +) on each side. The CO compartmentalization shares the same structural boundaries as the antigenic
Fig. 6. A comparison of zebrin I immunoreactivity (upper panels) and CO activity (lower panels) in two pairs of adjacent frontal sections through the posterior lobe vermis of the cerebellar cortex of an adult rat. Lobules VIII and IX are labelled. The midline P1 ~, and first lateral P2÷ compartments are labelled in the upper panels and the correspondingCO-weak regions indicated by arrowheads in the lower panels. Bars = 200 ,um.
75
Fig. 7. A comparison of zebrin I immunoreactivity (A) and CO activity (B) in adjacent frontal sections through the posterior lobe vermis of the cerebellar cortex of an adult squirrel monkey. The midline P1 ÷, and lateral P2 ÷ and P3 ÷ compartments are labelled (A) and the corresponding higher CO patches in the granular layer indicated by arrowheads in B. Bar = 100/~m.
zonation, with high C O in P1 and P2- and low C O in P1 ÷, P2 ÷ and P3 ÷. The same spatial correlation holds for the p o s t e r i o r vermis and again, high C O levels are associated with the zebrin I- c o m p a r t m e n t s (Figs. 5,6). In the cerebellar cortex of the squirrel m o n k e y , the C O activity is also non-uniformly distributed, but it is confined to the granular layer (Fig. 6). The same pattern occurred in the m a c a q u e 14. Despite this difference, the p r i m a t e C O c o m p a r t m e n t a l i z a t i o n shares the same structural b o u n d a r i e s as the zebrin I zonation. H o w e v e r , there is a striking difference between rat and p r i m a t e : in the squirrel m o n k e y the CO-rich bands c o r r e s p o n d to the zebrin I ÷ c o m p a r t m e n t s and, conversely, the C O - w e a k bands u n d e r l a y the zebrin I- c o m p a r t m e n t s (Fig. 7). The b a n d e d distribution of C O is only seen in the vermis. A t birth (P0) two populations of Purkinje cells can a l r e a d y be distinguished, one expressing high levels of C O reactivity and the other low levels. Fig. 8 shows two sections through the cerebellum of a newborn rat histochemically stained for CO. A t this stage of develo p m e n t , the Purkinje cells have not c o m p l e t e d their migration to the cerebellar cortex and form a multicellular layer. The staining is mainly localized in the Purkinje cell layer, but the resolution of the staining does not distinguish individual Purkinje cells clearly. Even at this early stage of d e v e l o p m e n t , the location of the C O c o m p a r t m e n t s is similar to that in the adult. As illustrated
Fig. 8. Two examples of CO activity in frontal sections through the cerebellar cortex of a newborn rat. In (A) the staining is in lobules V and VI, whereas in (B) it is in lobule IX. Most CO-activity is associated with the Purkinje cell layer (Pc). In all cases, clusters of Purkinje cells with high CO activity alternate with low activity patches: patch boundaries are indicated by arrowheads• In its general distribution, the patch arrangement is similar to that in the adult, with a low intensity patch at the midline, flanked by high activity patches to either side. Bar = 100 ltrn. in Fig. 7 A for lobule VI, a w e a k C O b a n d occupies the midline and wider clusters d e m o n s t r a t i n g a more intense staining are distributed to either side. As was previously seen by using an antiserum against cyclic G M P - d e p e n dent protein kinase 3~, there are thin acellular gaps usually separating clusters of different C O reactivity. The COweak bands in the vermis b e c o m e wider as they extend
76 into the posterior lobe, as in the adult. As far as we can determine, the congruence between zebrin I compartmentalization and CO zonation is preserved during cerebellar development: there is no zebrin I expression before P5 so no direct comparison is possible early on. DISCUSSION In the cerebellar cortex of the squirrel monkey, the immunocytochemical staining for zebrin I reveals an elaborate chemical architecture reminiscent of that in the rat. The only previous study of zebrin I in primates was an examination of postmortem human tissue that revealed Purkinje cell heterogeneity but no obvious parasagittal zonation 27. However, given the present results it seems plausible that an architecture common to rats and monkeys is also present in the human cerebellum. Direct comparison of ACHE, CO, and zebrin I compartmentalization has now revealed that all 3 share a common ground plan. Hess and Voogd 16 have described the non-uniform distribution of AChE and CO activities in the granular layer of the primate cerebellum, in particular they found a series of prominent AChE + bands in the vermis and paravermis, one astride the midline and two others parallel to either side. These bands align with bands in the white matter and, in some cases the molecular layer, and there is an exact correspondence between ACHE- and CO-rich zones. Our findings extend those of Hess and Voogd 16 by showing that CO + compartments and zebrin I + compartments are congruent. The boundaries of the CO patches in the vermis correspond to the zebrin I+/zebrin I- boundaries of the Purkinje cell compartments. The histochemical staining patterns of AChE and CO have also been shown directly to correspond to the myeloarchitectonic boundaries in the white matter 16. A similar congruence of compartments has been shown in rodents for zebrin 1 and AChE ~, the olivocerebellar projection 7, 5'-N 6, and zebrin II (unpublished observations). However, not all reported markers of chemical heterogeneity in the primate cerebellar cortex are consistent with a single underlying ground plan of the type described above for zebrin I, AChE and CO. For example, Ingram et al.lS concluded that the B1 antigenic compartments did not superimpose on those revealed by using AChE histochemistry. The results presented here have two novel features. First, the finding that high CO activity is associated with the rat P- compartments is the first report of a positive marker for this set - - all other staining protocols highlight the P+ set. Second, there is an obvious difference between rats and primates: in the monkey the CO + and zebrin I + zones overlap, whereas in the rat cerebellum they are interdigitated and CO + patches
overlap the zebrin 1 territories. This is the first time that such an interspecific difference is reported, and it suggests that care is needed in making interspecific comparisons of compartments based on their molecular phenotype. It should be born in mind that the compartmentalization revealed by CO histochemistry in rats and primates does not involve the same cell types: differential staining is molecular layer-associated in rats, and granular layer-associated in primates. In this respect, it is reminiscent of the species-dependent compartmentalization revealed by using AChE histochemistry (see ref. 1). The fact that differential CO levels are associated with cells in both the molecular layer and the granular layer may reflect differences in the pattern of use of the compartment as a whole, with high-CO compartments having either higher basal respiratory requirements or the need to respond more rapidly. This idea is consistent with the observations of Mjaatvedt and Wong-Riley 25, who found changes in the CO activity of Purkinje cells during development that reflected, the metabolic needs of the cell: when Purkinje cells receive excitatory input, their cell body CO level is high, and as the input changes to inhibitory, so the CO level declines. Such differences can arise in two ways - - either through genetic preprogramming or as a metabolic response to patterns of use. For example, CO has been reported to be a reliable marker of the neuronal metabolic activity in the visual cortex and in the SI somatosensory cortex 33-3s. Because differential CO expression by rat Purkinje cell clusters is already present at birth, a genetic explanation seems more probable. This is supported by our preliminary studies of deafferented cerebellums which have normal patterns of CO expression (Leclerc, Dor6 and Hawkes, unpublished). The same seems to be true for CO patches in the primate visual cortex 17. The cerebellum of the squirrel monkey contains many times more Purkinje cells than in the rat, and the number of Purkinje cells is even greater in humans. How is this expansion achieved? Two possibilities can be contrasted: either the cortical surface increases by the addition of new compartments or it is through the expansion in volume of a constant set of compartments. The former mechanism seems to be employed for example in the frontal cortex, where modular columns are of roughly similar sizes in many species and cortical expansion occurs by the addition of further modules (e.g. refs. 2,19). However, comparison of rat and monkey cerebellar cortices stained for zebrin I suggests that cerebellar expansion occurs not by the addition of new compartments but by the expansion in volume of a constant set. At least in the vermis, the zebrin 1 compartmentalization is rather similar in the two species, with clear P1 +, P2 + and P3 + bands identified in both, each set with similar
77 p r o p o r t i o n s . In the h e m i s p h e r e , the cerebellum of the m o n k e y has evidently u n d e r g o n e a massive expansion in surface area but we see no evidence of large numbers of new alternating zebrin I + and zebrin I- c o m p a r t m e n t s . R a t h e r , the individual zebrin I + c o m p a r t m e n t s seem to have increased in width such that very long continuous stretches of zebrin l + Purkinje cells are seen c o m p a r e d to the rat. This view is further s u p p o r t e d by comparison of the corticonuclear zonation of primates and rodents. In both the rat and the squirrel m o n k e y , the vermal A and B zones are both narrow, but in the paravermis and h e m i s p h e r e s there is a very significant relative b r o a d e n ing of the C1, C2, C3 and D zones of the squirrel m o n k e y 9. A l t h o u g h no detailed correlation has been m a d e b e t w e e n corticonuclear zones and antigenic comp a r t m e n t s , this is obviously consistent with what we see here with zebrin I, with relatively narrow vermal bands in both species and almost all the relative expansion in the m o n k e y cerebellar cortex c o n c e n t r a t e d in the zebrin I c o m p a r t m e n t s of the p a r a v e r m i s and hemispheres. Studies of the m a t u r a t i o n of antigen expression in the rat have shown that c o m p a r t m e n t a l i z a t i o n of the cerebellar cortex develops i n d e p e n d e n t l y from that of the afferent inputs 22"31'32. F u r t h e r m o r e , the crucial decisions are m a d e very early in d e v e l o p m e n t . O n e possibility is that bands of zebrin I + and zebrin 1 Purkinje cells arise by clonal expansion from a small n u m b e r of committed precursor cells, although chimera studies rather suggest a finer-grained mosaicism of n o n - r a n d o m clonal compartments in the m a t u r e cerebellar cortex 1s'26. How clonal c o m p a r t m e n t s might be related to expression of the zebrin I p h e n o t y p e is still unknown, but this general line of reasoning could explain the similarities and differences b e t w e e n rat and primate cerebellums. If all Purkinje cells arise from a small n u m b e r of early precursors, each a l r e a d y c o m m i t t e d to (but not yet expressing) its zebrin 1 p h e n o t y p e , then such a ground plan might well be very stable t h r o u g h o u t evolution: early embryonic patterns of d e v e l o p m e n t are difficult to change as their ramifications
for later d e v e l o p m e n t tend to be wide, and hence the consequences of change tend to be fatal. H o w e v e r , differences in the extent of clonal expansion can be a c c o m m o d a t e d more easily, and would result in different species having a c o m m o n set of c e r e b e l l a r c o m p a r t m e n t s , but of widely differing sizes. W h y do cerebellar and neocortical modes of expansion during evolution a p p e a r to be so different? The a p p a r e n t distinction may be because cerebellar expansion has involved only the a c c o m o d a t i o n of m o r e of the s a m e function, whereas additional cortical areas may subserve new and, perhaps, d i f f e r e n t functions. The obvious reason is that the underlying embryological mechanisms are quite different: in cerebellum, c o m p a r t m e n t s are o r d a i n e d very early in d e v e l o p m e n t and a p p e a r insensitive to postnatal influences; in cortex, columns are sculpted by competition b e t w e e n afferents and respond to postnatal stimuli. Thus, cerebellar c o m p a r t m e n t s reflect the mitotic histories of the cerebellar stem cells, whereas cortical columns are the result of the rules of e n g a g e m e n t between afferents and targets. Because of this distinction, perhaps the comparison is fundamentally misleading and the true correlation should be sought not between cerebellar c o m p a r t m e n t s and cortical modules but rather cerebellar c o m p a r t m e n t s and cortical architectonic zones. This perspective suggests that early structural subdivisions m a p out the territories in which subsequent fine afferent segregation is established. This opens the possibility of afferent segregation within cerebellar antigenic zones, This appears not to be the case for the climbing fiber p r o j e c t i o n 7, but is s u p p o r t e d by the recent finding that mossy fiber terminal fields may subdivide cerebellar antigenic c o m p a r t m e n t s both parasagittally and transversely ~. Acknowledgements. We wish to thank J. Rafrafi, R. Sasseville, L. Tremblay, and C. Harvey for technical assistance. This work was supported by grants from the Medical Research Council of Canada, the F.R.S.Q. and the F.C.A.R. (R.H., A.P.) and by a studentship from the F.R.S.Q.(N.L.).
REFERENCES 1 Boegman, R.J., Parent, A. and Hawkes, R., Zonation in the rat cerebellar cortex: patches of high acetylcholinesterase activity in the granular layer are congruent with Purkinje cell compartments, Brain Research, 448 (1988) 237-251. 2 Bugbee, N.M. and Goldman-Rakic, P.S., Columnar organization of corticocortical projections in squirrel and rhesus monkeys: similarity of column width in species differing in cortical volume, J. Comp. Neurol., 220 (1983) 355-364. 3 Chan-Palay, V., Nilaver, G., Palay, S.L., Beinfeld, M.C., Zimmerman, E.A., Wu, J.Y. and O'Donohue, T.L., Chemical heterogeneity in cerebellar Purkinje cells: existence and coexistence of glutamic acid decarbocylase-like and motilin-like immunoreactivities, Proc. Natl. Acad. Sci. U.S.A., 78 (1981) 7787-7791. 4 Chan-Palay, V., Palay, S.L.. Brown, J.T. and Van Itallie, C.,
5
6 7
8
Sagittal organization of olivocerebellar and reticulocerebellar projections: autoradiographic studies with [3SSlmethionine, Exp. Brain Res., 3(1 (1977) 561-576. Chan-Palay, V., Palay, S.L. and Wu, J.Y., Sagittal cerebellar microbands of taurine neurons: immunocytochemical demonstration by using antibodies against the taurinc-synthesizing enzyme cysteine sulfinic acid decarboxylase, Proc. Natl. Acad. Sci. U.S.A., 79 (1982) 4221-4225. Eisenman, L.M. and Hawkes, R., 5'-Nucleotidase and the mabQ113 antigen share a common distribution in the cerebellar cortex of the mouse, Neuroscience, 31 (1989) 231-235. Gravel, C., Eisenman, L., Sasseville, R. and Hawkes, R., Parasagittal organization of the rat ccrebellar cortex: a direct correlation between antigenic bands revealed by mabO113 and the topography of the olivoccrebellar projection, J. Comp. Neurol., 265 (1987) 294-310. Gravel, C. and Hawkcs, R., Parasagittal organization of the rat
7~
9
10 11
12
13
14
15
16 17
18
19
20 21
22
cerebellar cortex: direct comparison of Purkinje cell compartments and the organization of the spinocerebellar projection, J. Comp. Neurol., in press. Haines, D.E., Patrick, G.W. and Satrulee, P.. Organization of cerebellar corticonuclear fiber systems. In S.L. Palay and V. Chan-Palay (Eds.), The Cerebellum, New Vistas, Springer, Berlin, 1982, pp. 320-367, Hawkes, R., The dot immunobinding assay, Methods in F,nzvmology, 121 (1986) 484-491. Hawkes, R., Colonnier, M. and Leclerc, N., Monoclonal antibodies reveal sagittal banding in the rodent cerebellar cortex, Brain Research, 333 (1985) 359-365. Hawkes, R. and Leclerc, N., Immunocytochemical demonstration of topographic ordering of Purkinje cell axon terminals in the fastigial nuclei of the rat, J. Comp. NeuroL, 244 (1986) 481-491. Hawkes, R. and Leclerc, N., An antigenic map of the rat cerebellar cortex: the distribution of parasagittal bands as revealed by a monoclonal anti-Purkinje cell antibody mabQ113, J. Comp. Neurol., 256 (1987) 29-4l. Hawkes, R., Niday, E. and Gordon, J., A dot immunobinding assay for monoclonal and other antibodies, Anal. Biochem., 119 (1982) 142-147. Herrup, K., Cell lineage relationships in the development of the mammalian CNS: role of cell lineage in control of cerebellar Purkinje cell number, Dev. Biol., 115 (1986) 148-154. Hess, D.T. and Voogd, J., Chemoarchitectonic zonation of the monkey cerebellum, Brain Research, 369 (1986) 383-387. Horton, J.C., Cytochrome oxidase patches: a new cytoarchitectonic feature of monkey visual cortex, Phil. Trans. R. Soc. Lond., 304B (1984) 199-253. Ingrain, V.I., Ogren, M.P., Chatot, C.L., Gossels, J.M. and Owens, B.B., Diversity among Purkinje cells in the monkey cerebellum, Proc. Natl. Acad. Sci. US.A., 82 (1985) 7131-7135. Iscroff, A., Schwartz, M i . , Dekker, J,J. and Goldman-Rakic, P.S., Columnar organization of callosal and associational pro* jections from rat frontal cortex, Brain Research, 293 (1984) 213-223. Ito, M., The Cerebellum and Neural Control, Raven, New York. 1984, pp. 580. Leclerc, N., Beesley, P., Brown, I., Colonnier, M., Gurd, J.W., Paladino, T. and Hawkes, R., Synaptophysin expression during synaptogenesis in the rat cerebellar cortex, J. Comp. Neurol., 281) (1989) 197-212, Leclerc, N., Gravel, C. and Hawkes, R., Development of parasagittal zonation in the rat cerebellar cortex: mabQll3 bands are created postnatally by the suppression of antigen expression in a subset of Purkinje cells, J. Comp. Neurol., 273
(1988) 399-420. 23 Marani, E., Topographic Enzyme ttistochemistry of the Mammalian Cerebellum: 5'-Nucleotidase and Acetvlcholinestera~'e. Thesis, University of Leiden, 1982. 24 Marani, E. and Voogd, J,, An acetylcholinesterase band pattern in the molecular layer of the cat cerebellum, .L Anat,. 124 (1977) 335-345. 25 Mjaatvedt, A.E. and Wong-Riley, M.I.T., Relationship between synaptogenesis and cytochrome oxidase in Purkinje cells of the developing rat cerebellum, J. Comp. Neurol., 277 (1988) 155-182. 26 Oster-Granite, M.L. and Gearhart, J., Cell lineage analyses of Purkinje cells in murine chimeras. In 8.L. Palay and V, Chan-Palay (Eds.), The Cerebellum. New Vistas, Springer, New York, 1981, pp. 75-92. 27 Plioplys, A.V., Thibault, J. and Hawkes, R., Selective staining of a subset of Purkinje cells in the human cerebellum with monoclonal antibody mabQll3, J. Neurol. Sci., 70 (1985) 245-256. 28 Scott, E G . , A unique pattern of localization in the cerebellum, Nature (Lond.), 200 (1963) 793. 29 Scott, T.G., A unique pattern of localization within the cerebellum of the mouse. J. Comp. Neurol., 122 (1964) 1-8. 30 Voogd, J.~ The importance of fiber connections in the comparative anatomy of the mammalian cerebellum. In R. Llinas (Ed.), Neurobiology of Cerebellar Evolution and Development, Am. Med. Assoc., Chicago, 1969, pp. 493-514. 31 Wassef, M. and Sotelo, C., A synchrony in the expression of guanosine 3',5'-phostphate-dependent protein kinase by clusters of Purkinje cells during the perinatal development of rat cerebellum, Neuroscience, 13 (1984) 1217-1241. 32 Wassef, M , Zanetta, J.P., Brehier, A. and Sotelo, C., Transient biochemical compartmentalization of Purkinje celts during early cerebellar developmenL Dev. Biol., l l 1 (1985) 129-137. 33 Wong-Riley, M., Changes in the visual system of monocutarly sutured or enucleated cats demonstrable with cytoehrome oxidase histochemistry, Brain Research, 171 (1979) ll-28. 34 Wong-Riley, M.T.T. and Welt, C., Histochemical changes in cytochrome oxidase of cortical barrels following vibrissal removal in neonatal and adult mice, Proc. Natl. Acad. Sci. U.S.A., 77 (1980) 2333-2337. 35 Wong-Riley, M. and Riley, D.A., The effect of impulse blockage on cytochrome oxidase activity in the cat visual system, Brain Research, 261 (1983) 185-193. 36 Yaginuma, H. and Matsushita, M., Spinocerebellar projection fields in the horizontal plane of lobules of the cerebellar anterior lobe in the cat: an anterograde wheat germ agglutinin-horseradish peroxidase study, Brain Research. 365 (1986) 345-349.