Cholecystokinin corticostriatal pathway in the rat: Evidence for bilateral origin from medial prefrontal cortical areas

0306-4522(93)E0071-W

Neuroscience Vol. 59, No. 4, pp. 939-952, 1994 Elsevier Science Ltd Copyright 0 1994 IBRO Printed in Great Britain. All rights reserved 0306-4522/94 $6.00 + 0.00

CHOLECYSTOKININ CORTICOSTRIATAL PATHWAY IN THE RAT: EVIDENCE FOR BILATERAL ORIGIN FROM MEDIAL PREFRONTAL CORTICAL AREAS P.

MoRINo,*t$

F.

MASCAGNI,t

A.

MCDONALD?

and T.

H~KFELT*

of Neuroscience and Anatomy, Karolinska Institutet, Box 60400, S-10401 Stockholm, Sweden TDepartment of Cell Biology and Neuroscience, University of South Carolina, Columbia, SC 29208, U.S.A *Department

Abstract-The anterograde tracer Phuseolus vulgaris-leucoagglutinin was used to examine the organization of the projections to the striatum from medial prefrontal and frontal cortical areas in the rat with reference to their relation to cholecystokinin-like immunoreactivity in the striatum. Medial prefrontal cortical areas projected bilaterally, with an ipsilateral predominance, to the striatum. Most of the positive fibres were found in medial and ventral areas of the caudate-putamen and in the nucleus accumbens. Labelled fibres formed distinct patch-like arrangements throughout the dorsomedial striatum, whereas more ventrally the fibres were densely packed and spread to lateral areas. Almost no fibres were found in the dorsolateral aspects of the caudate-putamen. Cholecystokinin-like immunoreactivity in the striatum was diffusely distributed in the medial aspects, in fine punctate elements as well as in patches of fibres. Overlapping of corticostriatal clusters of fibres, from medial prefrontal cortex, with cholecystokinin-immunoreactive patches was found at all rostrocaudal levels studied, but predominantly in rostra1 areas. The overlap was present both in the ipsilateral and the contralateral side. Often the cluster of corticostriatal fibres was completely and precisely overlaid by a cholecystokinin-immunoreactive patch. At more caudal planes the overlap was only partial and in some instances cholecystokinin-positive patches “avoided” zones of dense corticostriatal fibre terminations. Frontal cortex injections of tracer gave rise to a network of fibres in the lateral aspects of the striatum, sparing the medial areas. No overlap with cholecystokinin-immunoreactive patches was found in these cases. These results suggest that a large number of cholecystokinin-containing striatal fibres originate in medial prefrontal cortical areas.

The cerebral cortex of the rat provides one of the major afferent inputs to the striatum, and all major regions of the cortex project bilaterally to this structure in a somatopically and topographically organized fashion.‘,9,‘8,28,46,4~5’ In general terms, each cortical region projects onto a defined longitudinally oriented region of the striatum with an anteroposterior and mediolateral topography.2* Furthermore, functionally different cortical areas project to largely separate striatal regions. 28,49In addition, it has been shown that different cortical areas project to chemically and functionally distinct striatal patch-matrix compartments.‘4 The peptide cholecystokinin (CCK), first discovered in the gastrointestinal system,34 is present in high concentrations in the rat brain,45 mainly in the form

$To whom correspondence should be addressed at: Institut D’Anatomie et D’Embryologie Speciale, 1 Rue Gockel, CH-1700 Fribourg, Switzerland. Abbreoiutions: CCK, cholecystokinin; KPBS, potassium phosphate-buffered saline; -LI,-like immunoreactivity; PBS, phosphate-buffered saline; PHA-L, Phase&s vulguris leucoagglutinin. 939

of the sulphated C-terminal octapeptide.8 The cerebral cortex is the region with the highest concentration of CCK in the central nervous system.2 In situ hybridization studies have demonstrated that high numbers of cortical neurons express CCK mRNA4*23*38 and retrograde tracing studies combined with in situ hybridization showed that several cortical cells expressing CCK mRNA project to subcortical areas, including the striatum.4 We have recently described, with immunohistochemical and lesioning techniques, the existence of a partially crossed corticostriatal projection which contains CCK-like immunoreactivity (CCK-LI).32.33 These studies suggest that CCK-positive corticostriatal fibres terminate in the medial aspects of the caudateputamen in a patchy fashion, since this CCK-LI was abolished on the ipsilateral side by an extensive cortical ablation followed by callosotomy. 32.33In agreement, after the same type of lesion there was a marked decrease in ipsilateral CCK release.20,52 Moreover we have shown that several cortical neurons retrogradely labelled from the medial striatum with wheatgerm agglutinin display immunoreactivity for an antibody to the CCK

P. MORINO el ul.

940

precursor.32 Most of these pro-CCK-immunoreactive cells were located bilaterally in rostra1 areas of the cingulate cortex, the medioventral orbital and frontal areas. In the present work, in an attempt to further characterize the CCK corticostriatal pathway, we have used the plant lectin Phaseolus vulgaris-leucoagglutinin (PHA-L). This substance is particularly useful for anterogradely tracing of axonal projections of neurons.” Among its properties PHA-L seems to have a negligible uptake by fibres of passage, and it labels neurons and fibres in a Golgi-like manner. We have thus unilaterally injected PHA-L in some areas of the rostra1 cerebral cortex and examined the projections to the striatum of both hemispheres. The tracing studies were then combined with immunohistochemical analysis of CCK-LI in the striatum, with the aim to examine the relationship of corticostriatal fibres with the patches of CCK-immunoreactive elements. EXPERIMENTAL

PROCEDURES

Ten male Sprague-Dawley rats (250g) received iontophoretic, unilateral, injections of PHA-L into the cerebral cortex. A 2.5% solution of PHA-L (Vector Labs) in 10 mM sodium phosphate buffer, pH 8, was loaded in micropipettes and was injected with a ~-PA positive current over a 45-min period. Each animal received four to six injections in different areas of the cerebral cortex (Fig. 1). Following a IO-20-day survival time the animals were deeply anaesthetized and perfused transcardially either with saline followed by a 4% paraformaldehyde solution in phosphate buffer 0.1 M2’ or with Tyrode’s solution followed by the same fixative solution containing in addition 0.3% picric acid. 54 Brains were postfixed in the same fixative for 2 h and transferred to phosphate-buffered saline (PBS) containing 10% sucrose and 0.01% sodium azide. Sections were cut in a cryostat (Microm) at 20 pm and incubated for 18 h at 4°C with an anti-PHA-L antiserum raised in goat (Vector Labs), diluted 1: 1,000 in potassium phosphatebuffered saline (KPBS) containing 2% normal donkey serum (Vector Labs) and 0.3% Triton X-100 (Sigma). After rinsing in KPBS, the sections were incubated with fluorescein isothiocyanate conjugated donkey anti-goat antibodies (1: 10; Nordic Immunology), rinsed in PBS and incubated for 18 h at 4’ C with a polyclonal rabbit antiserum (R-18) to the non-sulphated C-terminal octapeptide of CCK-33.‘) After rinsing in PBS, the sections were incubated with a lissamine rhodamine conjugated donkey anti-rabbit antibodies (I: 40; Jackson), rinsed, mounted and examined with a Nikon Microphot-FX microscope equipped with epifluorescence and proper filter combinations. For control purposes the CCK antiserum was preadsorbed with an excess ( 10m6 M) of the peptide (Peninsula). In addition, to exclude cross-reactivity between the two primary and secondary antisera, some sections were only incubated with a single primary antiserum and the two secondary antisera or the two primary and one of the secondarv antisera. Drawings of frontal planes at different levels of the rat striatum were taken from the atlas of Paxinos and Watson.j5 Cytoarchitectonic subdivisions of the prefrontal cortex were derived from Van Eden and UylingsN RESULTS

Anterogradely labelled fibres were analysed bilaterally in the caudate-putamen. We focused our atten-

tion on the medial aspects of this nucleus and on the relation of corticostriatal fibres with the pattern of CCK-LI. Three approximate levels with respect to bregma, according to the atlas of Paxinos and Watson,” were chosen as representative of the distribution of the PHA-L-labelled fibres in the striatum. Labelled fibres in the contralateral cortex or in subcortical areas other than the caudate-putamen were not analysed. Injection sites Animals were divided into three different groups, Rats in group A were injected at two sites in the anterior cingulate cortex and in two sites in the frontal cortex (Fig. 1A). Animals in group B were injected at four sites in the anterior cingulate cortex (Fig. 1B). Animals in group C received six injections in the frontal and forelimb area of the cortex. Groups A and B gave comparable results concerning the distribution of fibres in the striatum and nucleus accumbens, but with some differences in pattern mostly in the lateroventral parts of the striatum. Therefore the results of groups A and B will be presented together. The injection sites were centred both in deep (V and VI) and superficial (II and III) layers of the cortex (Fig. ID, E). The labelling was characterized by extensive filling of the neurons, both at the level of the cell body and the apical and basal dendrites, giving a Golgi-like appearance (Fig. lD, E). Fihre labelling in the striatum

in groups A und B

Labelling was found bilaterally in the caudate putamen. In general the contralateral component reached the same striatal regions and distributed according to the same pattern as the ipsilateral component. However, fibres were more numerous in the ipsilateral part. In an animal representative for these two groups. rostra1 levels (approximatively bregma + at 1.70 mm), labelled fibres were found in the corpus callosum at medial levels (Fig. 2B). On the ipsilateral side labelled fibres were entering the striatum by way of the bundles of the internal capsule (Fig. 2A) in the medial aspects of the striatum and were arranged in a patchy fashion. Most of the fibres were distributed in the medioventral areas of the caudate-putamen (Fig. 2C) and in the fundus striati. The anterograde label in the rostra1 part of the caudate nucleus did not involve the most dorsal part of this nuclear complex and, in addition, spared the extreme medial rim bordering the lateral ventricle. A few scattered fibres were seen in the lateral aspects (Fig. 2D). A dense network of labelled fibres was seen bilaterally in the nucleus accumbens (Fig. 2E), with an ipsilateral predominance. On the contralateral side the fibres were mostly found in medial aspects of the striatum, arranged in clusters (Fig. 2F). In the ventrolateral part labelled fibres were seen presumably entering the striatum via the external capsule (Fig. 2G).

Origin of the cholecystokinin corticostriatal pathway

Fig. 1. (AC) Drawings from the atlas of Paxinos and Watson 35 indicating the approximate sites of injection of PHA-L (stars) in three different groups of animals (A-C). n indicates the number of animals for each experimental group (D, E). Immunofluorescence micrographs of the injections sites in the anterior cingulate cortex (group B). The injection sites involve neurons in layers II and III as well as in layers V and VI. Cell bodies are filled with immunoreactivity (arrows). Apical dendrites are strongly labelled, and the dendritic arborization is clearly visible in layer I (curved arrows). Scale bars = 1OO~m.

At slightly more caudal levels (bregma + 1.20 mm) the distribution of the fibres was in general similar to the previous one. Both on the ipsilateral and con-

tralateral sides no fibres were seen in the lateral aspects of the striatum (Fig. 3A, B). Most of the fibres were found in the medial aspects of the striatum on

-l--l-.,-

-.-..-

---_

--.,.

-

----

-

-

.-

--

- __

- -,.

_

- -

Fig. 2. Inummofluorescence miCrographS showing different striatal areas ipsi- and contralateral to the PHA-L injection site. The frontal plane corresponds to bregma + 1.70,(modified from Paxinos and Watson”). The distribution of corticostriatal fibres in an animal representative for group B is shown. (A) On the ipsilateral side, in mediodorsal aspects, fibres enter the striatum via bundles of the internal capsule (arrows). They have a patchy distribution (arrowheads). (B) Numerous immunoreactive fibres are found in the corpus callosum (cc), especially at the midline level. (C) In the ventromedial aspects of the ipsilateral side, fibres are more densely packed than in dorsal areas. (0) Only a few scattered fibres are detected in the lateral aspects. (E) In nucleus accumbens (Acb), densely packed fibres can be detected on both sides, surrounding the anterior commissure (ac). (F) On the contralateral side, numerous immunoreactive fibres arranged in patches can be seen in medial aspects of the caudate putamen (CP) (arrowheads). (G) Thick. coarse fibres (arrows) enter the striatum via the external capsule (ec) on the contralateral side. Scale bars = 100 pm (A = C = D = E = F = G).

ipsi i

Fig. 3. Immuno~uores~n~ micrographs showing different striatal areas, in ipsi- and contralateral to the PHA-L injection site. The frontal plane corresponds to bregma + 1.20(modified from Paxinos and Watson3s). The distribution of corticostriatal fibres in an animal representative for group B is shown. Neither on the ipsilateral (A) nor on the contralateral (B) side any positive fibres are seen in the lateral aspects of the caudate-putamen (CP). On the ipsilateral side fibres enter the striatum via the internal capsule (arrows), both at mediodorsal (C) and ventromedial (D) levels. lmmunoreactive fibres are found arranged in a patchy fashion (arrowheads) both on the ipsilateral (D) and contralateral (E) side. (F) In the corpus callosum (cc), labelled fibres are running in two separate compartments, one just below the longitudinal sulcus (star) and the other ventrally, above the septum. Scale bars = 100 pm (A = B = C = D = E).

bregma +l.20

E

944

P.

ilkHUN

both sides (Fig 3C, D, E). Some fibres were thick and smooth and were arranged in small clusters with the same orientation, usually following the fibre bundles of the internal capsule (Fig. 3C, D). Other fibres were thinner and had varicosities giving them a beaded appearance (Fig. 3C-E). Terminal specializations could not be unequivocally assessed. On the ipsilatera1 side the fibres enter via the bundles of the internal capsule at dorsomedial levels (Fig. 3C) and terminate mostly in the me&o-ventral parts of the caudate-putamen (Fig. 3D). On the contralateral side patches of fibres with a dorsoventral orientation were seen (Fig. 3E). In the midline positive fibres were seen running in two separate compartments of the corpus callosum (Fig. 3F), one just below the longitudinal sulcus and the other in the ventral part, close to the septum. At levels around bregma + 0.70 mm (not illustrated) patches of positive fibres were seen ipsilaterally in medial aspects of the striatum. Clusters of fibres entering through the bundles of the internal capsule were detected in more ventral areas as compared to rostra1 levels. On the contralateral side a similar pattern was seen with fibres in the medioventral aspects of the caudate nucleus, but no fibres were detected in lateral areas. At caudal levels (bregma -0.26 mm) scattered fibres, arranged in small patches, were seen in mediodorsal aspects on the ipsilateral side (Fig. 4A). In ventromedial areas the fibres were more numerous and dense (Fig. 4B, C). Fibres were entering the striatum from the internal capsule also at these caudal levels (Fig. 4C). A similar distribution of fibres was seen in the contralateral side, but the fibres were less densely packed (Fig. 4D). No positive fibres were seen in the lateral part of the striatum (Fig. 4E). Correlation of Phaseolus vulgaris-leucoagglutinin labelled jibres with choleocystokinin -like immuno reactivity in groups A and B

CCK-LI in the medial striatum formed patches of densely packed flbres and terminals, often found in association with the bundles of the internal capsule. In addition a dense, diffuse, more weakly fluorescent network was seen all over the striatum. The size of the patches varied considerably; however, no obvious differences could be observed between rostrocaudal or dorsoventral areas. Overlapping clusters of PHA-L-labelled fibres with CCK-immunoreactive patches were found at all rostrocaudal levels studied, but were more evident in the rostra1 areas of the striatum. In general, the correspondence of CCK-

Pi ai

positive patches with corticostriatal fibre patches was found both on the ipsilateral and contralateral sides. However, the less dense network of PHA-L-positive fibres on the ~ontra1ateral side often gave a more exact overlap. Figure 5 shows examples of striatal sections double stained for PHA-L and CCK at rostra1 levels (bregma + 1.7 mm). Both on the ipsilatera1 (Fig. SA, D) and contralateral (Fig. 5B, E and C, F) side there was a considerable overlap between clusters of PHA-L-positive fibres, mostly longitudinally oriented, and the patches of fine, punctate CCK-immunoreactive elements. On the ipsilateral side (Fig. 5A) the clusters of PHA-L-positive fibres often encompassed the bundles of the internal capsule through which labelled fibres were entering the striatum. These fibres were not CCK-immunoreactive (Fig. 5D). CCK-LI was found around the fibre bundles corresponding to the area of dense PHA-Lpositive fibres. In some areas the patchy termination field of the corticostriatal fibres and the patch of CCK-LI had defined and completely matching borders (cf. Fig. 5B with 5E). In other cases the corticostriatal fibres were assembled in small clusters of few scattered fibres that coincided with the CCKimmunoreactive patch (cf. Fig. 6C with 6F). At more caudal levels a very dense network of PHA-L-labelled fibres was found (Fig. 6A-F) ipsilateralfy in ventromedial areas (Fig. 6A). In these cases the overlap with C~K-immunorea~tive patches was less evident, and sometimes (Fig. 6A, D) the area of dense CCK-LI covered a zone of less dense corticostriatal fibres. IIowever, in more dorsal ipsilateral areas (not shown) as well as on the contralateral side a positive correlation could be established (cf. Fig. 6B with 6E and 6C with 6F). Only partial overtapping was found at more caudal levels, both on ipsilateral (cf. Fig. 7A with 7E) and contralateral (cf. Fig. 7B with 7F; 7C with 7G; 7D with 7E) sides. Correlation of Phaseolus vuigaris-leucoaggiutinin1abeI~edfibres with choIe~ystokini~1 -like immunorea~tivity in group C

Corticostriatal fibres were mostly found in the lateral aspects of the caudate nucleus, throughout its rostrocaudal extent (Fig. 8A). They were densely packed, thick fibres, often with varicosities, arranged around the bundles of the internal capsule. In this area CCK-LI was found in small dot-like, weakly Auorescent structures (Fig. 8C). In the medial aspects of the striatum no PI-IA-L labelled fibres could be detected (Fig. 8B). On the other hand, strongly

Fig. 4. Immunofiuorescence micrographs showing different striatal areas ipsi- and contralateral to the PHA-L injection site. The frontal plane corresponds to bregma -0.26 (modified from Paxinos and Watson35). The distribution of corticostriatal fibres in an animal representative for group B is shown. Small clusters (arrowheads) of loosely arranged fibres can be seen in mediodorsal aspects (A) of the caudate-putamen (CP). In more ventral areas (C-E), the fibres are more densely packed. In D thick fibers (arrows) arise from the internal capsule (ic). No fibres can be seen in the lateral aspects (B). LV: lateral ventricle; cc: corpus callosum. Scale bar = lOO~$rnand applies to A~E.

Origin of the cholecystokinin corticostriatal pathway

.I

8 .I

945

frontal seetmns of m&d areas of to bregnxa + i .7 nm. ~lius~rdting the rclat~ons~p between PHA-L labelled fibres and the ~CK.irnrnun~~~c~i~e patches. Sections are from a rat belonging to group B. Both on the ipsilateral (A. D) and contralateral side (8, E: C. F) cluskr~ of

Ag. 5. Imrnuno~uo~~

the ~u~t~-put~rn~n

micrographs of pairs of ~~ub~e-sl~i~,

(A, D; B, E; C, F) at a level ~~r~~n~in~

PHA-L-positive Gbres(A, B, C) are overlapping with the patches of CCK-LI (D-F). Arrowheads delimil the borders of the cluster of corticostriatal fibres which overlays the corresponding area with the CCK-positive patch. Rhombi indicate the same blood vessels. Scale bar = IO0pm and applies to A F. immun~r~act~~e patches of CCK-LK were found in

this region (Fig. 80). No overlap of CCK-immunoreactive patches and PHA-L positive fibnls could be observed.

The present results confirm and extend our previous reports.” 33strong14 supporting the presence of a bilateral CCK-containing corticostrlatal pathway. In addition the present study indicates that a large part of these fibres originate in medial, prefrontal cartical areas.

The use of the PHA-L tracing technique offers many advantagw when studying anatomical connectionst5 Neurons at the injection site are labelled in a Go&i-like manner, allowing a rather precise determination of the limits of the emotive site of uptake of the tracer. The uptake of PHA-L was o~~nal~y shown to be restricted to cell bodies and dendrites at the injection Site.” However. more recent report@’ indicate that PHA-L can be taken up by fibres of passage. Such a possibility cannot be cotnpktety cncluded tn the prcscnt stud), Howevc~. our results

Origin

of the cholecystokinin

corticostnatal

pathway

Fig. 6. Immunotluorescence micrographs of pairs of double-stained frontal sections of medial areas of the cat&ate-putamen (A, D; 9, E; C, F) at a level corresponding to bregma + 0.70 mm, illustrating the relationship between PHA-L-labelled fibres and the CCK-immunoreactive patches. The sections are from a rat belonging to group 9. In ventromedial areas (A), in particular on the ipsilateral side, areas with less dense corticostriatal fibres (delineated by arrowheads) correspond to CCK-positive patches (arrowheads) (D). In other cases overlap of clusters of PHA-L-labelled fibres (arrowheads) (9, C) with CCK-immunoreactive patches (arrowheads) (E, F, respectively) is found as at more rostra1 levels. Rhombi indicate the same blood vessels. Scale bar = 1OOpm and applies to A F.

are in full agreement with previous detailed studies of the anatomical connections of the prefrontal and frontal cortex.‘.28.42 In addition, &sack er a1.42 have shown that injections of PHA-L in the white matter beneath the medial prefrontal cortex failed to produce any anterograde labelling. Organization

of prefrontal corticoswiutal projections

In line with previous reports,‘.3.‘n,42 the results of the present anterograde tracing study in the rat demonstrate that the medial prefrontal cortex projects bilat-

erally to the striatum, mostly to medial and ventral areas, sparing the lateral aspects. Sesack Ed aL4’ showed that moving from injection sites in the dorsal part of the medial prefrontal cortex, i. e. the dorsal anterior cingulate cortex, to injections in the ventral prelimbic area, terminal labelling in the striatum shifted gradually from a primarily dorsal position to a more ventromedial placement. This was particularly evident at rostra1 levels, while in the tail of the striatum there was a considerable overlap of medial prefrontal cortex afferents.42 Similar results were also

948

P. MOKINO er ai.

Fig. 7. lmmunofluorescence micrographs of pairs of double-stained, frontal sections at levels corresponding to bregma -0.26 mm (A, E; B, F) and -0.92 mm (C, G; D, H), illustrating the relationship between PHA-L-labelled fibres and CCK-immunoreactive patches. The sections are from a rat belonging to group B. Areas with a low density of corticostriatal fibres (A, B) correspond to CCK-immunoreactive patches (E, F, respectively). Some overlap of PHA-L-positive fibres (C) and a patch of CCK-LI (G) can be seen. In D, H a CCK-immunoreactive patch (H) overlays a zone almost devoid of corticostriatal fibres (D). Arrows indicate the same fibre bundle of the internal capsule. Arrowheads dekneate the patches. Rhombi indicate the same blood vessels. Scale bar = 1130pm and applies to A. H.

obtained by Berendse et al.’ In our material, injections sites in group B were centred at various levels of the anterior cingulate cortex, including a wide part of the prefrontal cortical area. Fibres were in fact found in the caudate-putamen on both sides, arranged in patches in dorsomedial areas, and becoming progressively more numerous and densely packed in more medioventral parts, spreading to ventrolatera1 aspects including the fundus striati. This pattern was rather constant in the rostrocaudal direction. The laterodorsal aspects were almost completely devoid

of labelled fibres: a few scattered ones were seen in rostra1 areas. This distribution of Iabelled fibres is in agreement with previous descriptions of corticostriatal projections from medial prefrontal cortical areas in the rat.3.9*‘4~42 In the present study PHA-L-positive fibres were seen in the corpus callosum only at rostra1 levets. Labelled fibres were also seen in ventral parts of the external capsule at rostra1 levels, entering the striatal neuropil, as shown in previous reports.3.42 In group A the injection at two sites of the frontal area may account for the larger contingent of %bres on the

Origin of the cholecystokinin corticostriatal pathway

949

Fig. 8. Two pairs of double-stained frontal sections (A, C; B, D), from a rat of group C. In the lateral aspect of the caudate putamen densely packed PHA-L-positive fibres from frontal cortical areas are seen (A). The same section shows weakly CCK-immunoreactive elements with no defined correspondence to the PHA-L-positive pattern of fibres (C). In medial aspects of the caudate-putamen no PHA-L-positive fibres can be detected (B). However, in this area strongly CCK-immunoreactive patches (arrowheads) can be seen (D). Stars indicate corresponding blood vessels. Scale bar = 100 pm and applies to A-D.

lateral aspects of the striatum.28 In group C the injection sites were centred in sensorimotor areas of the cortex. Corticostriatal fibres were found in the

lateral aspects of the striatum, while medial areas were completely devoid of labelhng. This pattern of distribution is in agreement with the one described in other reports for the rat.28*49,51 In groups A and B, the corticostriatal fibres entered the striatum on the ipsilateral side by way of the bundles of the internal capsule, often surrounded by terminal fibres arranged in patches. Here fibres entering the striatum in this way were seen exclusively on the ipsilateral side. This route is taken predominantly by fibres originating from deep layer V and layer VI3 These layers have been shown to project principally

to the striosome compartment, whereas superficial layer V as well as layers II and III project mainly to the matrix.16 Our results, however, include injections both in superficial and deep layers. Thus, the distribution of labelled fibres in the present material cannot be related to cortical lamination. Relationship of corticostriatal jbres with cholecystokinin-like immunoreactivity in the striatum

The pattern of CCK-LI has been described in detail in previous reports. 2’,22,33 In the present investigation our attention was focused on the patches of densely packed CCK-immunoreactive elements, which have been shown to be mostly of cortical origin, as they disappear after extensive cortical ablation followed

950

P. MORINOet al

by callosotomy.32,33 In the present investigation CCKimmunoreactive patches were found, in fact, to overlay the patchy termination fields of prefrontal cortical fibres, in medial areas of the caudate nucleus. The overlap was found bilaterally and was more evident at rostra1 levels. In caudal parts of the caudate putamen the overlap was less prominent, and sometimes areas with low density of corticostriatal fibres were corresponding to patches of CCK-LI. Taken together the present results suggest that a large part of CCK-LI concentrated in patches in the caudateputamen is originating bilaterally in medial prefrontal cortical areas. Indeed, retrograde tracing with wheatgerm agglutinin injected into the medial striaturn produced labelling of cells located mainly in the prefrontal cortex and predominantly on the ipsilatera1 side with a few cells on the contralateral one.32 The cortical lesions, which were performed in our previous studies, affected a large part of the neocortex, from prefrontal to sensorimotor areas.32.52 Most of the sensorimotor cortex, however, projects to lateral aspects of the striatum,” as is also shown in the present study in rats of group C. In these areas of the striatum CCK-LI is found in fine, weakly immunoreactive elements which in our studies were not significantly affected by decortication followed by callosotomy.32 Nevertheless, it cannot be excluded that immunohistochemical techniques are not sensitive enough to detect small changes in immunoreactivity after the related lesion. Indeed, it has been shown with radioimmunological techniques that after a lesion of the sensorimotor cortex followed by callosotomy3’ or after bilateral ablation of the cortex,‘” a decrease in CCK levels was seen in lateral aspects of the striatum, suggesting that in these areas CCK-LI might also be, at least partially, of cortical origin. Electron microscopical analysis of the CCKpositive patches32 revealed a number of immunoreactive terminals which morphologically resembled the corticostriatal terminals, as defined by Hassler et al.” In the present light microscopical investigation terminal specializations could not be unequivocally defined, neither for the traced corticostriatal fibres nor for the CCK-immunoreactive elements. In addition the fine and dense terminal labelling in the CCK-positive patches did not allow us, even at high magnifications, to assess undoubtly the co-localization of PHA-L labelled fibres with CCK-LI. However, the precise overlapping of prefrontal cortical fibres with CCK-immunoreactive elements in several instances, supports the presence of CCK-containing corticostriatal fibres. A considerable proportion of striatal CCK is supplied by the ventral tegmental area and substantia nigra CCK-containing dopaminergic cells.22,40.4’ The ascending CCK pathway terminates primarily in the posteromedial parts of the olfactory tubercle and the nucleus accumbens, and in the most ventral and medial aspects of the caudate-putamen. In these

areas part of CCK-LI is distributed in a patchy-like fashion.*’ However, in the caudate-putamen these patches were not affected by a 6-hydroxydopamine lesion in the medial forebrain bundle,21 while the dense, strongly fluorescent fibre networks in the periventricular zone disappeared. This area is indeed devoid of corticostriatal fibres from prefrontal areas3*42(also in the present study; however see Ref. 7). In the posterior, medial parts of the nucleus accumbens and the olfactory tubercle, a large part of the CCK-LI was depleted after 6-hydroxydopamine.** However, in both areas some CCK-immunoreactive fibres were still detectable. In particular in the anterior nucleus accumbens, patches of weakly fluorescent, dense fibre networks could also be seen after lesion of fibres of mesencephalic origin.‘? In the present report, the distribution of corticostriatal fibres in the nucleus accumbens was not carefully analysed. However, a dense network of PHA-Llabelled fibres was seen in this area, in correspondence with a strong CCK-immunoreactive fibre network (not shown). Thus it is likely that part of the CCK-positive fibres in ventral areas of the striatum are of cortical origin, as suggested also by the studies by Zaborsky et al. 53 Nevertheless, the projection to these areas might originate in large part from the lateral cortex, i.e. piriform cortex,26 from the entorhinal and perirhinal cortices,2s,43and from the insular cortex,37 in addition to a projection from the medial prefrontal cortex.‘,53 CCK-immunoreactive patches were found to overlap to a large extent with areas containing strong substance P and enkephalin immunoreactivities, though the latter have a wider distribution.” The striosome compartment of the striatum has been shown to be enriched in enkephalin and substance P cortical immunoreactivities,‘4s’7 where prefrontal areas mainly project.3~9,‘4~42 Patches of prefrontal cortical fibres, in fact, have been shown to overlap with zones of dense opiate receptor binding, surrounded by a receptor-poor matrix.9,14 In more medioventral aspects as well as in the fundus striati and in the nucleus accumbens, the pattern of PHA-Llabelled fibres was more diffuse, also probably involving the striatal matrix compartment.9,‘4 Thus, taken together, the present results suggest that CCK-containing corticostriatal fibres mainly project to the striosome compartment of the striatum, at least in medial areas. The compartmental organization of the ventral striatum seems to be, in fact, more complicated. The characteristics of the enkephalin and substance P patches in the caudate-putamen differ from those in the nucleus accumbens and olfactory tubercles, the latter having their own features in the distribution and correlation of these two substances as well as of dopamine. 47 In addition, at the border region between caudate nucleus and nucleus accumbens there is a complete reversal of distributional relations between enkephalin, substance P and calbindin immunoreactivities.47 Thus, in this ventral

Origin of the cholecystokinin corticostriatal region of the striatum the relation CCK-striosome might be less pronounced. Striosomes and matrix compartments in the striaturn are distinguished, in addition to chemical markers, on the basis of the input-output connectionsi This repartition of the striatum overlays the classical, topographically ordered dorsolateral (non limbic) and mediodorsal and ventral (limbic) segregation of striatal connectivity and function.18 Thus CCK corticostriatal fibres, which pr~ominantty terminate in the striosome compartment, would be found mostly in limbic regions of the striatum. CCK has been suggested to play some role in affective diseases, such as schizophrenia, through an action on antipsychotic drugs.48 The glutamatergic

pathway

951

corticostriatal pathway’t’2*24*27,36 might also be of clinical significance in this type of disease as it may interact with dopamine nerve terminals in the striaturn... CCK in corticostriatal fibres mi&t thus exert an additional control on dopamine transmission, in addition to the one acting at the mesolimbic level.

Acknowledgements-This study was supported by grants from the Swedish Medical Research Council (2887), Marianne och Marcus Wallenbergs Stiftelse, and the NIMH (MH43230). P.M. was a recipient of Wenner-Gren Center Foundation fellowship. The CCK-8 antiserum was a gift from Dr P. Frey, Sandoz Research Institute, Bern, Switzerland. We thank Ms. K. Aman and S. Nilsson for skilful technical assistance.

REFERENCES 1. Beckstead R. M. (1979) An autoradiographic examination of the corticocortical and subcortical projections of the mediodorsat-projection (prefrontal) cortex in the rat. J. camp. Neurol. 184, 4362. 2. Beinfeld M. C., Meyer D. K., R. L. E., Jensen R. T. and Brownstein M. J. (1981) The distribution of cholecystokinin immunor~cti~ty in the central nervous system of the rat as determined by rad~oimmunoassay. Bruin Res. 212, 51-57. 3. Berendse H. W., Gaiis-De Graaf Y. and Groenewegen H. J. (1992) Topogmphi~l organi~tion and relationship with ventral striatal compartments of prefrontal corticostriatal projections in the rat. J. camp. Neural. 316, 314-347. 4. Burgunder J.-M. and Young I. W. S. (1990) Cortical neurons expressing the cholecystokinin gene in the rat: distribution in the adult brain, ontogeny, and some of their projections. J. camp. Neural. 300, 26-16. 5. Carlsson M. and Carlsson A. (1990) Interactions between glutamatergic and monoaminergic systems within the basal ganglia-implications for schizophrenia and Parkinson’s disease. Trends Neurosci. 13, 272-276. 6. Cliffer K. D. and Giesler G. J. J. (1988) PHA-L can be transported anterogradely through fibers of passage. Bruin Res. 458, 185-191. 7. Dalsass M., Kiser S., Mendershau~n M. and German D. C. (1981) Medial prefrontal cortical projections to the region of the dorsal periventricular catecholamine system. Neuroscience 6, 657-665. 8. Dockray G. J. (1980) Cholecystokinins in rat cerebral cortex: identification, purification and characterization by immunochemical methods. Bruin Res. 188, 155-165. 9. Donoghue J. P. and Herkenham M. (1986) Neostriatal projections from individual cortical fields conform to histochemically distinct striatal compartments in the rat. Bruin Res. 365, 397-403. 10. Fagg G. E. and Foster A. C. (1983) Amino acid neurotransmitters and their pathways in the mammalian central nervous system. N~rosci~nce 9, 701-719. 11. Fonnum F., Fosse V. M. and Allen C. N. (1983) Identification of excitatory amino acid pathways in the mammalian nervous system. In Excitotoxins (eds K. Fuxe, P. J. Roberts and R. Schwartz), pp. 3-18. MacMillan Press, London. 12. Fonnum F., Storm-Mathisen J. and Divac I. (1981) Biochemical evidence for glutamate as neurotransmitter in corticostriatal and corticothalamic fibres in rat brain. Neuroscience 6, 863-873. 13. Frey P. (1983) Cholecystokinin octapeptide (CCK 26-33), nonsulfated octapeptide and tetrapeptide (CCK 30-33) in rat brain: analysis by high pressure liquid chromatography (HPLC) and radioimmunoassay (RIA). Neurochem. fnt. 5, 81 I-815. 14. Gerfen C. R. (1984) The neostriatal mosaic: compartmentalization of corticostriatal and striatonigral output systems, Nature 311, 46-464, 15. Gerfen C. R. and Sawchenko P. E. (1984) An anterograde neuroanatomical tracing method that shows the detailed morphology of neurons, their axon and terminals: immunohistochemical localization of an axonally transported plant lectin, Phuseolus vulguris leucoagglutinin (PHA-L). Bruin Res. 290, 219-238. 16. Gerfen C. R. (1989) The neostriatal mosaic: striatal patch-matrix organization is related to cortical lamination. Science 246, 385-388.

17. Graybiel A. M., Ragsdale C. W., Yoneoka E. S. and Elde R. P. (1981) An imm~ohist~hemical study of enkeph~ins and other neuropeptides in the striatum of the cat with evidence that the opiate peptides are arranged to form mosaic patterns in register with the striosomal compartments visible by acetylcholinesterase staining. Neuroscience 6, 377-397. 18. Graybiel A. M. and Ragsdale C. W. (1979) Fiber connections of the basal ganglia. Prog. Bruin Res. 51, 239-283. 19. Hassler R., Nitsch C. and Lee H. L. (1980) The role of the eight putative transmitters in the nine tvaes of svnaoses in rat caudate-putamen. In Purkinson’s Disease-Current Pr;?gre&, Problems and Management (e& Rinne*U.rK., Klinger M. and Stamms G.), pp. 61-91. Elsevier, Amsterdam. 20. Herrera-Marschitz M., Meana J. J., Hijkfelt T., You Z.-B., Morino P., Brodin E. and Ungerstedt U. (1992) ~hol~ystokinin is released from a crossed corticost~atal pathway. NeuroReporf 3, 905-908. 21. Hiikfelt T., Herrera-Marschitz M., Seroogy K., Gong J., Staines W. A., Holets V., Schalling M., Ungerstedt U., Post C., Rehfeld J. F., Frey P., Fischer J., Dockray G., Hamaoka T., Walsh H. H. and Goldstein M. (1988) Immunohistochemical studies on cholecystokinin (CCK)-immunoreactive neurons in the rat using sequence specific antisera and with special reference to the caudate nucleus and primary sensory neurons. J. them. Neuroanat. 1, 1l-52. 22. Hskfelt T., Skirboll L., Rehfeld J. F., Goldstein M., Markey K. and Dann 0. (1980) A subpopulation of mesencephalic dopamine neurons projecting to Iimbic areas contains a cholecystokinin-like peptide: evidence from immunohistochemistry combined with retrograde tracing. Neuroscience 5, 2093-2124.

952

P. MERINO et al.

23. Ingram S. M., Krause R. G., Baldino 1. F. Jr, Skeen L. C. and Lewis M. E. (1989) Neuronal localization of cholecystokinin mRNA in the rat brain by using in situ hybridization histochemistry. J. camp. Neural. 287, 260-272. 24. Kim J.-S., Hassler R., Haug P. and Paik K.-S. (1977) Effect of frontal cortex ablation on striatal giutamic acid level in rat. Brain Res. 132, 370&374. 25. Krayniak P. F., Meibach R. C. and Siegel A. (1981) A projection from the entorhinai cortex to the nucleus accumbens in the rat. Brain Res. 209, 42743 1. 26. Luskin M. B. and Price J. L. (1983) The topographic organization of associational fibers of the olfactory system in the rat, including centrifugal fibers to the olfactory bulb. J. camp. Neurol. 216, 264291. 27. McGeer P. L., McGeer E. G., Scherer V. and Singh K. (1977) A glutamatergic corticostriatal pathway, Bruin Res. 128, 369-373. 28. McGeorge A. J. and Faull R. L. M. (1989) The organization of the projection from the cerebral cortex to the striatum in the rat. Neuroscience 29, 503-538. 29. McLean I. W. and Nakane P. (1974) Periodate-lysine-paraformaldehyde fixative for immunoelectron microscopy. J. Histochem. Cytochem. 22, 1071P1083. 30. Meyer D. K., Schultheiss K. and Hardung M. (1988) Bilateral ablation of frontal cortex reduces concentration of cholecystokinin-like immunoreactivity in rat dorsolateral striatum. Brain Res. 452, 113~117. 31. Meyer D. K. and Protopapas Z. (1985) Studies on cholecystokinin-containing neuronal pathways in rat cerebral cortex and striatum. Ann. N.Y. Acad. Sci. 448, 133-143. 32. Morino P., Herrera-Marschitz M.. Caste1 M. N., Ungerstedt U., Varro A., Dockray G. and Hokfelt T. (1994) Cholecystokinin in cortico-striatal neurons in the rat: immunohistochemical studies at the light and electron microscopical level. Eur. J. Neurosci. (submitted). M., Meana J. J., Ungerstedt U. and Hiikfelt T. (1992) Immunohistochemical evidence 33. Morino P., Herrera-Marschitz for a crossed cholecystokinin corticostriatal pathway in the rat. Neurosci. Lefl. 148, 1333136. I. Cleavage with thrombin and 34. Mutt V. and Jorpes J. E. (1968) Structure of porcine cholecystokinin-pancreozymin. with trypsin. Eur. J. Biochem. 6, 156-162. 35. Paxinos G. and Watson C. (1986) The Rat Brain in Stereotaxic Coordinafes. Academic Press, San Diego. release in vitro from corticostriatal terminals. Bruin Res. 176, 185-188. 36. Reubi J. C. and C&nod M. (1979) Glutamate of autonomic and limbic connections in the insular cortex of the rat. J. camp. Neural. 37. Saper C. B. (1982) Convergence 210, 163-173. J.-J. (1991) Distribution of cells containing mRNA encoding cholecystokinin in 38. Schiffmann S. N. and Vanderhaeghen the rat central nervous system. J. romp. Neural. 304, 219-233. (PHA-L) by axons of passage. J. Neurosci. Mrth. 39. Schofield B. R. (1990) Uptake of Phaseolus vulgaris leucoagglutinin 35, 47. 56. K., Lim S., Haycock J. W. and Fallon J. H. (1989) Ventral mesencephahc neurons containing 40 Seroogy K. B., Dangaran both cholecystokininand tyrosine hydroxylase-like immunoreactivities project to forebrain regions. J. romp. Neural. 279, 397-414. 41 Seroogy K., Schalling M., Brent S., Dagerlind A., Chai S. Y., Hokfelt T., Persson H., Brownstein M., Huan R., Dixon J., Filer D., Schlessinger D. and Goldstein M. (1989b) Cholecystokinin and tyrosine hydroxylase messenger RNAs in neurons of rat mesencephalon: peptide/monoamine coexistence studies using in situ hybridization combined with immunocytochemistry. Expl Brain Rex 74, 149-162. organization of the efferent projections 42 Sesack S. R., Deutch A. Y., Roth R. H. and Bunney B. S. (1989) Topographical of the medial prefrontal cortex in the rat: an anterograde tract-tracing study with Phuseolus uulgaris leucoagghrtinin. J. camp. Neurol. 290, 213-242. Neurosci. Left. 35, 259-264. 43 Sorensen K. E. and Witter M. P. (1983) Entorhinal efferents reach the caudato-putamen. development of the prefrontal cortex in the rat. J. camp. 44 Van Eden C. G. and Uylings H. B. M. (1985) Cytoarchitectonic Neural. 241, 253-267. 45 Vanderhaeghen J. J., Signeau P. and Gepts W. (1975) New peptide in the vertebrate CNS reacting with antigastrin antibodies. Nature 257, 604-605. F. M. and Lieven P. A. J. M. (1980) The topical organization of the afferents to the 46 Veening J. G., Cornelissen caudato-putamen of the rat, A horseradish peroxidase study. Neuroscience 5, 1253-1268. H. J. (1989) Compartmental organization of the ventral striatum of the rat: 47 Voorn P., Gerfen C. R. and Groenewegen immunohistochemical distribution of enkephalin, substance P, dopamine. and calcium-binding protein. J. camp. Neural. zS9, 189.--20 1. and action on antipsychotic drugs. In The 48. Wang R. Y., Jiang L.-H., Ti X. and Kasser R. J. (1989) Cholecystokinin Neuropeptide Cholecystokinin (CCK) (eds Hughes J. J., Dockray G. and Woodruff G.), pp. 163-170. Ellis Horwood, Chichester. interrelations in the albino rat. J. Anal. 95, 532~544. 49. Webster K. E. (1961) Cortico-striate and synaptic connections of crossed corticostriatal neurons in the rat. J. c’omp. Neurok 50. Wilson C. J. (1987) Morphology 263, 567-580. of descending projections of the rat somatic 51. Wise S. p. and Jones E. G. (1977) Cells of origin and terminal distribution sensory cortex. J. camp. Neural. 175, 129-158. M., Brodin E., Meana J. J., Morino P., Hokfelt T., Silveira R., Goiny M. and Ungerstedt 52. You Z.-B., Herrera-Marschitz J. Neurochemistry U. (1994) On the origin of striatal cholecystokinin (CCK) release: studies with in oiuo microdialysis. (in press). L., Alheid G. F., Beinfeld M. C., Eiden L. E., Heimer L. and Palkovits M. (1985) Cholecystokinin 53 Zaborszky innervation of the ventral striatum: morphological and radioimmunological study. Neuroscience 14, 427453. a new rapid fixative for electron microscopy 54 Zamboni L. and De Martin0 C. (1967) Buffered picric acid formaldehyde: J. Cell Biol. 148A, 35. (Accepted 1 December 1993)