The organization of midbrain projections to the ventral striatum in the primate

Neuroscience Vol. 59, No. 3, pp. 609-623, 1994

Pergamon

03064522(93)E0001-7

Printed

in Great

Elsevier ScienceLtd Copyright 0 1994IBRO Britain. All rights reserved 0306-4522/94

THE ORGANIZATION OF MIDBRAIN PROJECTIONS THE VENTRAL STRIATUM IN THE PRIMATE

$6.00 + 0.00

TO

E. LYND-BALTA and S. N. HABER* Department of Neurobiology and Anatomy, University of Rochester School of Medicine, 601 Elmwood Avenue, Rochester, NY 14642, U.S.A. Abstract-Because the dopaminergic neurons of the midbrain form a continuum, boundaries between the ventral tegmental area, substantia nigra pars compacta, and retrorubral area are difficult to distinguish in the primate. Therefore, dopaminergic neurons have been subdivided into more readily discernible dorsal and ventral tiers. The projections from these dorsal and ventral tier neurons of the ventral mesencephalon to the ventral striatum were labeled by injections of horseradish peroxidase conjugated to wheatgerm agglutinin and Lucifer Yellow conjugated to dextran amines into different regions of the nucleus accumbens, the ventral caudate nucleus, and the rostral, ventral putamen in the primate. Neurons projecting to the ventral striatum are not topographically organized in the ventral mesencephalon. Retrogradely labeled neurons are found in the medial densocellular zone of the ventral tier following injections into all regions of the ventral striatum except the ventromedial shell region of the nucleus accumbens. These medial nigral neurons have diverging projections throughout the mediolateral extent of the ventral striatum. In addition, neurons of the dorsal tier project to all ventral striatal regions examined. Notably, neurons projecting to the shell region of the nucleus accumbens are limited to the dorsal tier, throughout the rostrocaudal extent of the substantia nigra. Both dorsal and ventral tier neurons innervate the ventral striatum. Not only do neurons of the ventral tegmental area project to the ventral striatum, but also many of the pars compacta. The projections to the shell region of the nucleus accumbens are more restricted, suggesting that the dopaminergic regulation of this accumbens subterritory is distinct from the rest of the ventral striatum.

The striatum is a major target of the midbrain dopaminergic neurons. These neurons have been divided into several subsets of cells based on morphology and connectivity. The pars compacta refers to the cell-dense region that occupies the dorsal area

of the substantia nigra. In addition, dopaminergic neurons are scattered more medially in the ventral tegmental area, and dorsolaterally in the retrorubral area. Together, these nuclei largely correspond to cell groups A9, A10 and A8 of Dahlstrijm and Fuxe.’ Early work done in the rodent distinguished two separate ascending dopaminergic projections from the midbrain to the striatum.’ The mesolimbic pathway was described as originating from neurons of the ventral tegmental area and projecting to the nucleus accumbens and the olfactory tubercle. The nigrostriatal pathway was described as originating from neurons of the substantia nigra, pars compacta and projecting to the dorsal caudatoputamen. Fallon and Moore modified this concept when they demonstrated that the ascending projections from the ventral tegmental area and the pars compacta are not two separate systems.’ Instead, the dopaminergic *To whom correspondence Abbreviations:

should be addressed.

BSA, bovine serum albumin: HRP-WGA.

horseradish peroxidase conjugated to wheatgerm agglu: tinin; LY, Lucifer Yellow coniueated to dextran amines: NGS, normal goat serum; dBS-T, phosphate-buffered saline_Triton X-100; TH, tyrosine hydroxylase.

projections form a dorsoventral, mediolateral, and rostrocaudal topographic continuum.2,9*44In particular, the ventral tegmental area and the very medial aspect of the substantia nigra pars compacta give rise to the dopaminergic innervation of the nucleus accumbens and the olfactory tubercle in the rodent. Dopaminergic neurons have also been clearly identified in the primate midbrain with histofluorescence techniques, and more recently using tyrosine hydroxylase (TH) immunocytochemistry.‘“J4~30~s2~s6 In the primate, the ventral tegmental area includes the midline cell groups, the paranigral nucleus, and the nucleus parabrachialis pigmentosus. These cells are continuous with the retrorubral cells and the cells of the pars mixta region of Francois et al., the pars dorsalis of Poirier et al., and the y sub-division of Olszewski and Baxter (Fig. 1).‘2,45,48 Together, these neurons constitute a morphologically similar group of cells situated dorsal to the majority of pars compacta neurons. These neurons stain positively for tyrosine hydroxylase and the calcium binding protein, calbindin-D2,,‘5,36*37 (Fig. 2). In contrast, the more ventrally located pars compacta consists of a densocellular zone that covers the mediolateral extent of the substantia nigra (Fig. 1, open arrowheads), and columns of dopaminergic neurons that invade areas of the pars reticulata and occupy very ventral regions of the substantia nigra (Fig. 1, closed arrows). The dendrites of most of the pars compacta neurons are

610

E. LYND-BALTA and S. N. HABER

oriented dorsoventrally, and fill much of the pars reticulata. These ventral compacta neurons stain positively for TH (Fig. 2A), but are calbindin-D,,, negative ‘5.36.37 (Fig. 2B). Thus, two tiers of dopaminergic neurons are recognizable in the ventral mesencephalon: the dorsal tier, characterized by its loose arrangement of dopaminergic neurons with processes oriented in a mediolateral direction; and the ventral tier, the pars compacta neurons of the densocellular zone including the columns of cells penetrating into the pars reticulata (Figs 1,2).9~‘5.‘6~22~36~37

The striatum is divided into functionally separate territories based on cortical inputs. The ventral striaturn is associated with the limbic system by virtue of its inputs from the limbic lobe and the amygdala.

18.19.25.27.29.32,33.40,47,49 51.57 in

the

primate,

these

) A

re_

gions innervate not only the nucleus accumbens, but also the medial and ventral caudate nucleus and the rostral, ventral putamen. 25,40,49,50*57 Thus, this entire region is referred to here as the ventral striatum. It is distinct from the dorsolateral striatum which receives input from sensorimotor cortices.“,3’~34.35

D

Fig. I. Many subdivisions are recognized in the substantia nigra and surrounding regions at rostra1 (A), central (B), and caudal (C) levels of the ventral mesencephalon: pc, pars compacta; a, 1, y, subdivisions of the pars compacta; pr, pars reticulata; pm, pars mixta; Pn, paranigral area; Pb, nucleus parabrachialis pigmentosus; pd, pars dorsalis. These subdivisions are compared to the organization and distribution of dorsal and ventral tier dopaminergic neurons in the ventral mesencephalon at similar rostra1 (D), central (E), and caudal (F) levels: filled circles, dorsal tier neurons; open circles + open arrowheads, ventral tier neurons, densocellular zone; open circles + closed arrows, ventral tier neurons, ventral columns.

Midbrain projections to primate ventral striatum

Fig. 2. The dopaminer~c neurons are organized as dorsal and ventral tiers in the midbrain. (A) TH immunore~ivity stains the dopamine neurons of the dorsal and ventral tiers in the ventral mesencephalon. Note the columns of TH-immunoreactive neurons penetrating ventral regions of the substantia nigra (SN; arrows). (B) A comparable section of the substantia nigra stained for the calcium binding protein calbindin-D,,. Dorsal tier neurons are distinct in that they contain this calcium binding protein. The area of virtually no calbindin-D2s,-positive staining ventral to the dorsal tier delineates the regions of the densocellular zone (small stars) and the ventral columns of the ventral tier (large stars). Calbindin-Disc-positive fibers, presumably of striatal origin, innervate the substantia nigra pars reticulata. (XW Recently, the nucleus accumhens has been further divided into core, shell, and rostra1 pole subterritories in the rodent.~s~20~28~58~60,62~~ The core is continuous with the rest of the striatum, while the shell surrounds the core at its medial, ventral, and ventrolateral borders. In rodents, several investigators have used different histochemical and immunocytochemical markers to delimit the core and shell regions. In primates, the shell region of the nucleus accumbens is best distinguished by calbindin-D,,, immunocytochemistry (Fig. 3D). 13,3WLike the rest of the rodent

striatum, including the core of the nucleus accumhens, the shell projects to both the globus pallidus and the substantia nigra. In addition, it also projects to a large region of the lateral hypothalamus?’ Of particular interest is whether the core and shell compartments can be distinguished with respect to separate populations of dopaminergic afferent projections in primates. Szabo described the general topography of ascending striatal afferents in the primate demonstrating that labeled cells projecting to the lateral putamen are

612

E.

LYND-BALTA and

found mainly in lateral regions of the substantia nigra. 54,55 In more recent studies, Parent et al. showed retrogradely labeled cells projecting to either the caudate nucleus or the putamen to be derived from different populations of pars compacta cells that are distributed as clusters throughout the substantia nigra.46,53These and other studies, however, did not examine the primate nigrostriatal pathway with respect to the organization of dorsal and ventral tiers of dopamine neurons. In addition, the distribution of nigrostriatal projections has not been related to the organization of cortical inputs to the striatum. In these respects, the present retrograde tracing study examined how the mesolimbic pathway is organized in its projection to the ventral striatum. The aim of these experiments was to compare the topography of midbrain neurons projecting to different parts of the ventral striatum including the caudate nucleus, the putamen, and the shell and core regions of the nucleus accumbens. The results are described with respect to the dorsal and ventral tier subdivisions of dopaminergic neurons in the midbrain. EXPERIMENTAL PROCEDURES

Two adult rhesus monkeys (Macacn mulutm) and four pig-tailed monkeys (Macacn nemesrrinu) were used for these retrograde tracing experiments. Prior to surgery, initial anesthesia was administered by an intramuscular injection of ketamine (10 mg/kg). A deep surgical level of anesthesia was maintained by intravenous injection of pentobarbital. Electrophysiological mapping was performed to locate appropriate injection sites. Serial electrode penetrations (glassinsulated etched platinum-iridium electrodes) were made throughout the rostrocaudal and mediolateral extent of the striatum to identify neuronal activity based on patterns of electrophysiological recordings.” The location of neurons encountered in a series of penetrations was used to prepare a map indicating the boundaries of different basal ganglia structures. The absence of cellular activity was noted in the area of fiber tracts, i.e. the corpus callosum, the internal capsule, and the anterior commissure. Accurate placement of the tracers was subsequently achieved by careful alignment of the injection cannulae with the electrode. Individual monkeys were injected with either Lucifer Yellow conjugated to dextran amines (LY) or horseradish neroxidase coniugated to wheatgerm agglutinin (HRPWGA). The retrograde tracers were pressure injected using a 0.5-u] Hamilton svrinee as follows: 20-40 nl of LY (10%. Molecular Probes): anvd 40--50 nl of HRP-WGA ‘(4%, Sigma). Following the injection, the syringe remained in place for 20-30 min to prevent leakage up the needle track. The protocol for LY as a tracer6 has been modified for use in primates. Ten to 12 days after surgery, the animals were again deeply anesthetized and perfused with saline followed bv a 4% naraformaldehyde solution in 0.1 M phosphate buffer, pH7.4. Brains were postfixed overnight and cryoprotected in increasing aradients of sucrose (10, 20, and 30% where they remainid-several days). Serial sections of 50 pm were cut on a freezing microtome into 0.1 M phosphate buffer or ~ryoprot~tant solution. Immun~yt~hemical techniques were used to visualize the tracers. Prior to incubation in primary antiserum, tissue was incubated in a solution of 10% methanol and 10% hydrogen peroxide in 0.1 M phosphate buffer to inhibit any endogenous peroxidase, followed by extensive rinsing for several hours. Sections to be immunoreacted with either anti-LY or anti-HRP-WGA were first rinsed in 0.1 M

S. N.

HABER

phosphate buffer @H 7.4) with 0.3% Triton X-100 (PBS-T). Sections were then preincubated in 10% normal goat serum diluted with PBS-T (NGS-PBS-T) for 30min. Tissue was placed in the primary antiserum, anti-LY 1:5000 (generously provided by Dr H. Chang), or anti-HRP-WGA I : 2000 (Dako Corp.) in NGS-PBS-;i’ for four nights at 4 C. The avidin-biotin reaction (rabbit Vectastain ABC kit. Vector Labs) was used to visualize the LY and HRP-WGA: Tissue was thoroughly rinsed in PBS-T, before incubating in biotinvlated aoat anti-rabbit 1:400 in PBS-T at room temperature I& 45 min. Following extensive rinsing, tissue was then incubated in the rabbit avidin-biotin complex 1: 200 at room temperature for I h. Staining was enhanced by incubating the tissue for IO-12 min in 3,3’-diaminobenzidine tetrahydrochloride and 0.01% H,O, intensified with 1% cobalt chloride and 1% nickel ammonium sulfate to yield a black reaction product. After thorough rinsing, sections were mounted on to slides and counterstained with Cresyl Violet. Immunocytochemist~ for the calcium binding protein ~alb~ndin-~*s~ was performed to determine the boundaries of the shell region of the nucleus accumbens and to illustrate the dorsal and ventral tier organization of ventral mesencephalic neurons (Fig. 2B). Immunocytochemistry for TH was done to visualize the dopaminergic neurons of the ventral mesencephalon (Fig. 2A). Antiserum to calbindinD,,, and TH was purchased from Sigma and Eugene Tech, respectively. The TH antiserum was generally used at 1: 10,000 in NGSPBS-T, although individual lots were always tested first for optimal staining. TH was visualized using the avidin-biotin reaction (rabbit elite Vectastain ABC kit, Vector Labs). The calbindin-D,,,_._antiserum generated in mouse was used at I : 10,000 in PBS-T with ~0.5% bovine serum albumin (BSA),~_ (Siama). Tissue was first incubated in PBS-T with 5% BSA for 1h, and then incubated in primary anti~rum for two nights at 4°C. The calbindin-D,sk staining was processed using the avidinbiotin reaction (mouse Vectastain ABC kit, Vector Labs) as described for LY and WGA above. The final steps for TH and calbindin-D,,, visualization were identical. Sections were rinsed in PBS 3 x 12 min and then Tris buffer (0.05 M, pH 7.6). Tissue was preincubated in a filtered 0.05% 3,3’-diamino~n~idine tetrahydrochloride solution in Tris for IOmin, before adding 0.01% H,Oz. Following an 8-l2-min incubation, tissue was rinsed thoroughly in PBS. RESULTS

Well-confined injections of HRP-WGA and LY were made in various regions of the ventral striatum providing a good basis for analyses of their midbrain afferent projections. Retrogradely labeled neurons were charted throughout the ventral mesencephalon and the pattern is described in relation to the dorsal and ventral tier organization of midbrain neurons (Figs 1, 2). The injection sites included the ventral caudate nucleus, the rostral-ventral putamen, and the core and shell regions of the nucleus accumbens (Fig. 3). Four injection sites that represent the different parts of the ventral striatum are described in detail. The distribution of retrogradely labeled neurons in the cortex was examined for each case to verify the organization of corticostriatal inputs. Experiments MS43 is a HRP-WGA injection site centered in the rostral, ventromedial portion of the caudate nucleus (Fig. 3A). The ventral region of the caudate nucleus receives inputs from the hippocampus, the amygdaia, the orbitofrontal cortex, the

Midbrain projections to primate ventral striatum

Fig. 3. Retrograde tracers were injected into different areas of the ventral striatum. (A, B) Schematics show the Iocation of injection sites in the ventral striatum; shaded area: shell region of the nucleus accumbens as identified by calbindin-D,,,-negative immunoreactivity. (C) An injection site in the shell region of the nucleus accumbens, and (D) an adjacent section stained for calbindin-D,,,. The calbindin-Dads-negative area delineates the shell region of the nucleus accumbens in the primate. (x 5.) C, caudate nucleus; P, putamen.

anterior cingulate cortex, the dorsomedial temporal tamen (Fig. 3A). Although much of the putamen is pole, and the superior temporal gyrus.‘7J5,40,49,50,57*s9 innervated by sensorimotor-related cortices, this rosCase MS43 is compared to experiment MS53, a tral, ventral area of the putamen receives inputs from lateral injection of LY into the rostra], ventral puthe amygdala, the entorhinal cortex, the orbitofrontal

614

E. LYND-BALTA and S. N. HABER

cortex and the anterior cingulate area.25@‘~so Thus, it is included as part of the ventral striatum. Experiment MS49 represents a LY injection centered in the core of the nucleus accumbens (Fig. 3B). Like the ventral caudate nucleus and the ventral putamen, this region of the nucleus accumbens is innervated by the orbitofrontal cortex, the anterior cingulate cortex, and parts of the temporal lobe.25*40~s”~57~59 Finally, experiment MS35 is a HRP-WGA injection site confined to the shell region of the nucleus accumbens (Fig. 3B,C), as determined by calbindin-D,, staining patterns (Fig. 3D). The organization of cortical inputs to the shell have not been documented in the primate. A large number of retrogradely labeled neurons, representing afferents to the shell, were found in the amygdala, the subiculum, and the medial orbitofrontal cortex. Projections to the ventromediai caudate nucleus

Case MS43 is a well-confined HRP-WGA injection into the ventromedial caudate nucleus (Fig. 3A). Labeled cells in the midbrain are found in both the dorsal and ventral tiers of neurons and are dist~buted throughout the rostrocaudal extent of the ventral mesencephalon. They extend over the medial half of the ventral midbrain (including the central portion). Rostrally, within the densocellular zone of the ventral tier, there are a number of labeled neurons in clusters, separated by areas containing no HRP-WGA-positive cells (Fig. 4A). Some retrogradely labeled neurons are found in the medial portion of the dorsal tier. Centrally, the retrogradely labeled neurons spread out dorsolaterally to fill the media1 half of the densocellular zone of the pars compacta (Fig. 4B). An equal number of cells are observed in the dorsal tier among the oculomotor rootlets (Fig. 5). Caudally, clusters of retrogradely labeled dorsal tier neurons are found along the midline, with many HRP-WGApositive cells spreading ventrolaterally between the interpeduncular nucleus and the pars compacta (Fig. 4C). A few labeled neurons are located in the medial pars compacta. Projections to the rostral, ventral putamen

Case MS53 is a LY injection placed laterally into the rostral, ventral putamen at the same rostrocaudal level as case MS43 described above (Fig. 3A). The pattern of retrogradely labeled neurons in the ventral mesencephalon is very similar to that seen after injections of retrograde tracers into the ventromedial caudate nucleus. At rostra1 levels, several retrogradely labeled cells are detected in the medial aspect of the densocellular zone of the pars compacta (Fig. 4D). Centrally, there are several LY-positive cells among the oculomotor rootlets, corresponding to the medial continuation of dorsal tier neurons (Fig. 4E). At this level, there are also many labeled neurons in the media1 densocellular zone of the ventral tier. A few labeled neurons are found in the ventral columns of cells that invade the medial half of the pars

reticulata. At caudal levels, some LY-positive neurons are seen along the midline (Fig. 4F). Many retrogradely labeled neurons are found lateral to the inte~~uncular nucleus within the paranigral region of the ventral mesencephalon. Some of these labeled neurons are within the medial portion of the pars compacta. Projections to the dorsolateral core of the nucleus a~eumbens

Case MS49 is a relatively large injection of LY centered in the core of the nucleus accumbens (Fig. 3B). Although centered in the nucleus accumbens, some tracer is found in its transition zone with the ventral caudate nucleus and the ventromedial putamen. The injection site however does not include the shell region of the nucleus accumbens. The distribution of midbrain neurons projecting here is similar to the distribution found after injections into the ventral caudate nucleus and the rostral, ventral putamen (compare Figs 6A-C and 4). At rostra1 levels, there are labeled neurons throughout the medial half of the pars compacta, filling the same mediolateral extent of the midbrain region as seen in experiment MS43 (Fig. 6A). Centrally, as with the previously described cases, many labeled neurons are noted in the dorsal tier, amid the emerging oculomotor rootlets (Fig. 6B). Clusters of LY-positive neurons are also scattered just lateral to the rootlets still within the dorsal tier of the pars compacta (Fig. 7). Several labeled neurons are also found in the medial half of the densocellular zone of the ventral tier. Thus, the medial densocellular region is a source of common inputs to the ventromedial caudate nucleus, the rostral, ventral putamen, and the core of the nucleus accumbens. In the caudal substantia nigra, neurons are labeled along the midline (Fig. 6C). The majority of LY-positive neurons are found lateral to the interpeduncular nucleus and scattered dorsolaterally in the ventral mesencephalon. Few, if any, labeled neurons are found in the pars compacta at this caudal level. Projections to the ventromedial shell region of the nucleus accumbens

Case MS35 is a well-confined HRP-WGA injection into the shell region of the nucleus accumbens (Figs 3B,C). The boundaries of the shell region were determined by patterns of calbindin-DZsk immunoreactivity (Fig, 3D). The distribution of calbindin-D,,, staining varies for the dorsal and ventral striatum. The dorsal striatum is noted for its patchy distribution of calbindin-Dzw immunoreactivity (Fig. 3D). In the ventral striatum, however, a large territory of the nucleus accumbens is conspicuously low in calbindin-D,,, immunoreactivity, forming a unique pattern, The shape of this pattern is consistent in all animals and occupies a ventral and medial region surrounding a central core. We designate this territorv, as the shell region of the nucleus accumbens. In

615

Midbrain projections to primate ventral striatum

MS53

MS43

n .

l

C Fig. 4. Chartings of the distribution of retrogradely labeled neurons in the ventral mesencephalon following injections into medial and lateral regions of the ventral striatum: (A-C) an injection of HRP-WGA into the ventromedial caudate nucleus, case MS43; (D-F) an injection of LY into the rostral, ventral putamen, case MS53).

sharp contrast to the labeling from the other ventral striatal experiments, retrogradely labeled neurons in case MS35 are confined almost entirely to the dorsal tier of mesencephalic neurons. The majority of the

retrogradely labeled neurons rostrally are located in a more dorsal position, within the dorsal tier of dopaminergic neurons (Fig. 6D). Only a few HRP-WGA-positive neurons are found ventrally in

E. LYNII-BALTA and S. N. HABER

616

the densocellular zone of the ventral tier. The labeled dorsal tier neurons cover the same mediolateral extent of the substantia nigra as the retrograde labeling seen in the other cases. At central levels. numerous labeled neurons are found very medially, distributed among the oculomotor rootlets (Fig. 6E). Several HRP-WGA-positive cells are also seen dorsolateral to these rootlets, in the central region of the dorsal tier (Fig. 8). Distinct from the distribution of cells described in the previous cases, few, if any labeled neurons occupy the densocellular zone of the ventral tier (cf. Fig. 9A and B). At caudal levels of the mesencephalon, many labeled neurons are concentrated along the midline (Fig. 6F). In addition, these cells spread out dorsolaterally, occupying a large region of the ventral mesencephalon. It is only in the caudal mesencephalon that the distribution of retrogradely labeled neurons is similar to that seen in experiments MS43, MS53, and MS49.

The nucleus accumbens is the region of the striaturn that is most often considered to be the limbic-related part of the basal ganglia. In primates, however. cortical and subcortical inputs that are considered to be related to the limbic system (the limbic lobe and the amygdala), project not only to the nucleus accumbens, but also to the ventral caudate nucleus and the rostral, ventral putamen (Fig. 10).25.4y.5”.“7 Furthermore, neither cytoarchitectonic nor histochemical markers can reveal a clear boundary between the nucleus accumbens and the caudate nucleus dorsomedially, nor the nucleus accumbens and the putamen laterally. Thus, the entire area that receives input from the limbic lobe and amygdala is referred to as the ventral striatum.23.2h The fact that the midbrain afferent projection does not provide a unique input to the nucleus accumbens further supports the idea that it is not a unique part of the striatum. Midbrain prqjections to the ventral striatum topographically organized in the primate

DISCUSSION

The results of the present tracing study demonstrate that the mesencephalic input to the ventral striatum in primates is not limited to the medial most portion of ventral midbrain neurons (the ventral tegmental area and the medial substantia nigra). Rather, it receives input from both medial and central parts of the ventral mesencephalon. There is little difference in the distribution pattern of cells projecting to the ventral caudate nucleus, to the ventral putamen, or to the core of the nucleus accumbens. In these three cases, labeled cells are found in the medial half of both the dorsal and ventral tiers. In contrast, a part of the shell of the nucleus accumbens receives inputs primarily from the dorsal tier of neurons.

Fig. 5. Retrogradely labeled

ure not

Rodent studies have suggested that there is a mediolateral topography of dopaminergic inputs to the ventral striatum.2.y.44,47 Thus, the majority of inputs to the nucleus accumbens come from the ventral tegmental area, with a much smaller projection originating in the medial substantia nigra. Furthermore, inputs to the medial and lateral nucleus accumbens arise from medial and lateral regions of the ventral tegmental area and medial substantia nigra. respectively.y.44,47 In contrast to the situation in the rodent, the midbrain projections to the ventral striatum in the primate do not exhibit a mediolateral topography. Retrogradely labeled neurons are found in similar mediolateral positions, following injections into the different regions of the ventral striatum. For example,

labeled neurons among the oculomotor rootlets. These dorsal tier neurons following an injection into the ventromedial caudate nucleus. ( x 220.)

were

Midbrain projections tu primate ventral striatum

c Fig. 6. C&flings of t&e &dribuGcsn af rztrcrgrzkly k&&d neurons ia the ventral meseucephalon following injdans into the sub-tertitwies of the nucleus accumbens: (A-C) an injwtion inta the nucleus accumbens core and its transition zone with the ventral csudate nucleus and the ventral putatnen, case MS49; [D-F) an injection into the shell region of the nucleus xcumbens, cilse MS35 The nqzons projecting to tfre shell region of the ~&ens accumbens are almost compiete~y confirmedto the dorsal tier,

618

E. LYND-BALTA and S. N. HABER

Fig. 7. Retrogradely labeled neurons located among and lateral to the oculomotor rootlets following an injection into the core region of the nucleus accumbens. All areas of the ventral striatum receive inputs from dorsal tier neurons. ( x 90.)

the injection into the ventral putamen (case MS53) is relatively lateral, and yet the mediolateral distribution of the labeled neurons is similar to that seen after an injection into the medial part of the caudate nucleus, case MS43 (cf. Fig. 4D,E and 4A,B). A comparison of cases MS53 and MS43 at caudal levels suggests that a mediolateral shift in retrograde labeling occurs (cf. Fig. 4F and C). However, when the labeling from all four injection sites is examined at caudal levels of the mesencephalon, no distinct mediolateral topography is apparent (cf. Figs 4C,F and

6C,F). Previously, large injections of retrograde tracers filling the entire caudate nucleus or putamen demonstrated an overlap of clusters of retrogradely labeled neurons in the substantia nigra.46*5JThe small, confined injection sites into different striatal regions in these experiments clearly illustrate the lack of a mediolateral topography. Similar to rodent studies, there is no apparent rostrocaudal topography.2,9,44 For example, in case MS43, an injection into the rostra1 ventromedial caudate nucleus results in dense labeling of neurons

Fig. 8. Dorsal tier neurons are labeled after an injection of HRP-WGA into the shell region of the nucleus accumbens. The dendrites of these neurons are oriented in a mediolateral direction. Projections to the shell originate from a restricted region of the ventral mesencephalon that largely corresponds to the dorsal tier neurons. ( x 220.)

Midbrain projections to primate ventral striatum from the rostra1 pole to the caudal aspect of the ventral mesencephalon (Fig. 4A-C). Similarly, retrogradely labeled neurons are found throughout the rostrocaudal extent of the midbrain following an injection into the caudal nucleus accumbens, experiment MS35 (Fig. 6D-F). This indicates that projections from widely separated rostrocaudal areas of the midbrain terminate in common areas of the striatum. The lack of a rostrocaudal topography is apparent for every ventral striatal locus examined. Mesencephalic projections to the entire striatum are considered to obey an inverse dorsoventral topography, with neurons in the dorsal portion of the ventral mesencephalon projecting to the ventral stria-

619

turn.* However, the organization of midbrain projections to different regions of the ventral striatum does not reflect an inverse dorsoventral topography. Case MS53, an injection into the ventral most portion of the putamen, labeled neurons of both the dorsal and ventral tiers (Fig. 4D-F). The pattern of labeling is very similar to that produced from injections into more dorsal regions of the ventromedial caudate nucleus, case MS43 (Fig. 4A-C). However, there is an important distinction for the mesencephalic projections to the shell region of the nucleus accumbens, as they are derived primarily from the dorsal tier. This may be related to specific features of the shell region.

Fig. 9. Retrogradely labeled neurons in the ventral mesencephalon following injections into the ventral striatum. (A) An injection into the shell region of the nucleus accumbens labels neurons in the dorsal tier, among and dorsolateral to the oculomotor rootlets. Few, if any, labeled neurons are found in the densocellular zone (arrows). (B) In contrast, an injection into the ventromedial caudate nucleus labels neurons in both the dorsal tier, among the oculomotor rootlets, and in the densocellular zone of the ventral tier, ventrolateral to the oculomotor rootlets (arrows). SN, substantia nigra. (x 90.)

E. LYNV-BALTAand S. N. HABEH

shell region in the primate, characterized by calbindin-D,,, negative immunoreactivity, receives a more limited midbrain projection than the core. The distinction of midbrain inputs is made with respect to the dorsal versus ventral tier of cells in the ventral mesencephalon and not by its mediolateral organization. The densocelfular zone of the ventral tier projects to all ventral striatal areas except the shell region of the nucleus accumbens. Although a subset of dorsal tier neurons projects throughout the ventral striatum, it is of interest that the shell region receives almost exclusive input from these cells. Electron microscopic studies in the rodent have shown that the dopaminergic fibers innervating the shell are characterized by smaller diameter fibers, fewer synapses associated with dendritic spines, and relatively fewer terminal boutons, than those terminating in the core of the nucleus accumbens and the dorsal striatum.SH.h’ Furthermore, pharmacological and environmental challenges produce different changes in dopamine

amygdala *** orbitofrontalcortex 0 0 parahippocampal car tex

***

l

Fig. 10. Afferents from the hmbic lobe and the amygdala innervate a large area of the striatum. The topographic organi~tion of these projections is shown schematic~Iy. C, caudate nucleus; P, putamen.

Projections to the shell region originate almost exlusivel_yfrom dorsal tier ~earo~s Within the nucleus accumbens of rats, a core and shell region have been identified that correspond to different connectivities.’ s,28,63 The core is continuous with the rest of the striatum and its afferent and efferent projections are similar to the dorsal striatum. The shell, on the other hand, has been distinguished in rats by inputs from the subiculum, the ventral prelimbic area, and the medial orbitofrontal cortex.“.*’ In addition to projecting to the substantia nigra and the globus pallidus, the shell also projects to ventral parts of the bed nucleus of stria terminalis, the lateral preoptic area, the sublenticular substantia innominata, and a large area of the lateral hypothalamus.ZR.63 Thus, the shell region of the nucleus accumbens has been associated most closely with the limbic system. This area in rats has been considered to belong to the so-called extended amygdaia: a basal forebrain region tightly connected to an amygdalobed nucleus of stria terminalis continuum.‘*28 It is of particular interest that the dopaminergic neurons of the dorsal and ventral tiers in the primate ventral mesencephalon are biochemically distinct. Expression of the calcium binding protein calbindinD,,, is limited to neurons of the dorsal tier, whereas dopamine D, receptor levels are higher in ventral tier neurons than in dorsal tier neurons of the primate.36”’ The results from this study indicate that a part of the

Fig. 11. The ventral striatum receives non-topographic inrmts from dorsal and ventral tier neurons of the medial and central ventral mesenceph~on. Notably, the she11region of the nucleus accumbens receives a more restricted projection, almost exclusively from dorsal tier neurons (large open diamonds). Neurons in the medial and central densocellular zone of the ventral tier (large filled diamonds) project throughout a large mediojateral extent of the ventral striatum. Small diamonds in the striatum represent corresponding terminal fields, C, caudate nucleus; P, putamen; SN, substantia nigra.

621

Midbrain projections to primate ventral striatum metabolism in the core and shell.’ This study indicates that the shell region of the nucleus accumbens in primates may also have a distinct dopaminergic input. The dorsal tier neurons are scattered in the dorsal substantia nigra, densely clustered among the oculomotor rootlets, and spread throughout much of the paranigral nucleus and the nucleus parabrachialis pigmentosus. This dorsal group of mesencephalic cells receives ventral striatal inputz3”** Furthe~ore in the rodent, the dorsal tier neurons receive afferents from the hypothalamus, and innervate the amygdaia and the olfactory tubercle. 9,42This particular group of dopamine neurons seems intimately connected with limbic-related structures, receiving inputs from the hypothaIamus. projecting to the amygdala, and being reciprocally connected with the ventral striatum. Recent studies have demonstrated that discrete areas of the ventral striatum send overlapping projections to a wide mediolateral extent of the substantia nigra pars compacta. 23 Thus, the midbrain area that is associated with the ventral striatum occupies the

entire medial region, and extends over the dorsal tier, throughout the length of the substantia nigra. Despite the suggestion that the shell region of the nucleus accumbens is more closely associated with the limbic system, a relativeiy large area of the primate striatum is connected with the limbic lobe, and therefore is likely to have functions related to motivation, affect, and reward.25,40*49*M,S7 Nevertheless, the restricted mesencephalic projections to the shell region of the nucleus accumbens may result in a more specific and segregated feedback mechanism to this accumbens subterritory (Fig. 11). In contrast, a divergent mesencephalic projection originating from the medial neurons of the densocellular zone may permit an integrated feedback mechanism linking together different areas of the ventral striatum (Fig. 11). Acknowledgements-The authors thank Marti Armour for her technical assistance and Jennifer Loughlin for her photographic assistance. The authors gratefully acknowledge the gift of Lucifer Yellow antisera from Dr H. Chang. This work was supported by NIH grant NS22511 and NIMH grant MH45573.

REFERENCES 1. Alheid G. F. and Heimer L. (1988) New perspectives in basal forebrain organization of special relevance for neuropsychiatric disorders: the striatopallidal, amvadaloid. and corticonetal comoonents of substantia innominata. __ Neuro&rce 27, l-39. 2 Beckstead R. M., Domesick V. B. and Nauta W. J. H. (1979) Efferent connections of the substantia nigra and ventral tegmental area in the rat. Bruin Res. 179, 191-217. 3 Berendse H. W., Galis de Graaf Y. and Groenewegen H. J. (1992) Topographical organization and relationship with ventral striatal compartments of prefrontal corticostriatal projections in the rat. J. camp. Neural. 316, 314347. 4 Berendse H. W. and Groenewegen H. J. (1990) The organization of the thalamostriatal projections in the rat, with special emphasis on the ventral striatum. J. camp. Neural. 299, 1877228. 5 Berendse H. W., Groenewegen H. J. and Lohman A. H. M. (1992) Compartmental distribution of ventral striatal neurons projecting to the mesencephalon in the rat. J. Neurosci. If, 20792103. 6 Chang H. T. (1991) Anterograde transport of Lucifer Yellow~extran conjugate. Brain Res. Bull. 26, 813-816. 7. Dahlstriim A. and Fuxe K. (1964) Evidence for the existence of monoamine neurons in the central nervous system. I. Demonstration of monoamines in the cell bodies of brain stem neurons. Acm. physiol. scand. 62, Suppi. 232, l-55. 8. Deutch A. Y. and Cameron D. S. (1992) Pharmacological characterization of dopamine systems in the nucleus accumbens core and shell. Neuroscience 46, 49-56. 9. Fallon J. H. and Moore R. Y. (1978) Catecholamine innervation of the basal forebrain. IV. Topography of the dopamine projections to the basal forebrain and neostriatum. J. camp. Neural. 180, 545-580. IO. Felten D. L. and Sladek J. R., Jr (1983) Monoamine distribution in primate brain. V. Monoaminer~c nuclei: anatomy. pathways and local organization. Brain Res. &IL 10, 171-284. 11. Flaherty A. W. and Graybeil A. M. (1991) Cortical transformations in the primate somatosensory system. Projections from physiologically mapped body-part representations. J. Neurophysiol. 66, 1249-1263. 12. Francois C., Percheron G., Yelnik J. and Heyner S. (1985) A histological atlas of the macaque (Macaca mulutta) substantia nigra in ventricular coordinates. Brain Res. Bull. 14, 349-367. 13. Friedman D. P., Porrino L. J. and Vinsant S. (1992) Anatomical analysis of the ventral striatum in the macaque monkey. Sot. Neurosei. Absrr. 18, 307. 14. Garver D. L. and Sladek J. R. (197.5)Monoamine dist~bution in primate brain. I. Catechoiamine~ontaining perikarya in the brain stem of Macaca speciosa. J. camp. Neural. 159, 289-304. 1.5. Gerfen C. R., Bairnbridge K. G. and Thibault J. (1987) The neostriatal mosaic: III. Biochemical and developmental dissociation of patch-matrix mesostriatal systems. J. Neurosci. 7, 3935-3944. 16. Gcrfen C. R., Herkenham M. and Thibault J. (1987) The neostriatal mosaic: II. Patch- and matrix-directed mesostriatal dopaminergic and non-dopaminergic systems. J. Neurosci. 7, 3915-3934. 17. Goldman P. S. and Nauta W. J. H. (1977) An intricately patterned prefrontocaudate projection in the rhesus monkey. J. camp. Neurol. X71, 369-386. 18. Gr~newegen H. J., Becker N. E. H. M. and Lohman A. H. M. (1980) Subcortical afferents of the nucleus accumbens septi in the cat, studied with retrograde axonal transport of horseradish peroxidase and bisbenzimid. Neuroscience 5, 1903-1916.

19. Groenewegen H. J., Room P., Witter M. P. and Lohman A. H. M. (1982) Cortical afferents in the cat, studied with anterograde and retrograde transport techniques. Neuroscience 7, 20. Groenewegen H. J. and Russchen F. T. (1984) Organization of the efferent projections of pallidal, hypothalamic, and mesencephahc structures: a tracing and immunohistochemical Neurot. 223, 347-367.

of the nucleus accumbens 977-996.

the nucleus accumbens to study in the cat. J. camp,

622

E. LYND-BALTA and S. N. HABER

21. Groenewegen H. J., Vermeulen-Van der Zee E., Te Kortschot A. and Witter M. P. (1987) Organization of the projections from the subiculum to the ventral striatum in the rat. A study using anterograde transport of Phaseolus vulgaris-leucoagglutinin. Neuroscience 23, 103-120. 22. Haber S. N. and Groenewegen H. J. (1989) Interrelationship of the distribution of neuropeptides and tyrosine hydroxylase immunoreactivity in the human substantia nigra. J. camp. Neurol. 290, 53-68. 23. Haber S. N., Lynd E., Klein C. and Groenewegen H. J. (1990) Topographic organization of the ventral striatal efferent projections in the Rhesus monkey: an anterograde tracing study. J. camp. Neural. 293, 282-298. 24. Haber S. N., Lynd-Balta E. and Mitchell S. J. (1993) The organization of the descending ventral pallidal projections in the monkey. J. camp. Neural. 329, 111-129. 25. Haber S. N. and Mizobuchi M. M. (1994) The orbitofrontostriatal projections in the monkey. Topography and relationship to chemical organization. (In preparation). 26. Heimer L., Switzer R. D. and Van Hoesen G. W. (1982) Ventral striatum and ventral pallidurn. Components of the motor system? Tren& Neurosci. March, 83-87. 27. Heimer L. and Wilson R. D. (1975) The subcortical projections of the allocortex: similarities in the neural associations of the hippocampus, the piriform cortex, and the neocortex. In Golgi Centennial Symposium: Perspectives in Neurobiology (ed. Santini M.), pp. 177-193. Raven Press, New York. C. (1991) Specificity in the projection patterns 28. Heimer L., Zahm D. S., Churchill L., Kalivas P. W. and Wohltmann of accumbal core and shell in the rat. Neuroscience 41, 89-125. 29. Hemphill M., Holm G., Crutcher M., Delong M. and Hedreen J. (1981) Afferent connections of the nucleus accumbens in the monkey. In The Neurobiology of the Nucleus Accumbens (eds Chronister R. B. and DeFrance J. F.), pp. 75-8 1. Hear Institute, Brunswick, Maine. histochemistry of monoamine-containing cell bodies in the brain 30. Hubbard J. E. and Di Carlo V. (1974) Fluorescence stem of the squirrel monkey (Suimiri sciureus) II. Catecholamine-containing groups. J. camp. Neural. 153, 369-384. of corticostriatal 31. Jones E. G., Coulter J. D., Burton H. and Porter R. (1977) Cells of origin and terminal distribution fibers arising in the sensory-motor cortex of monkeys. J. camp. Neural. 173, 53-80. of the projection from the hippocampal formation to the 32. Kelley A. E. and Domesick V. B. (1982) The distribution nuclebs accumbens in the rat: an anterograde and retrograde-horseradish peroxidase study. Neuroscience 7, 2321-2336. 33. Kellev A. E.. Domesick V. B. and Nauta W. J. H. (1982) The amygdalostriatal projection in the rat-an anatomical study* by anierograde and retrograde tracing methods. ieuroscierke 7, 615-630: projection in the monkey. Brain 93, 525-546. 34. Kemp J. M. and Powell T. P. S. (1970) The cortico-striate 35. Kiinzle H. (1975) Bilateral projections from precentral motor cortex to the putamen and other parts of the basal ganglia. An autoradiographic study in h4ucucu fusciculuris. Bruin Res. 88, 195-209. neurons expressing calbindin in normal and parkinsonian monkeys. 36. Lavoie B. and Parent A. (1991) Dopaminergic NeuroReport 2, 601604. E. and Haber S. N. (1992) Organization of dopaminergic striatal projections in relation to peptidergic 37. Lynd-Balta innervation of the substantia nigra in primate. Sot. Neurosci. Abstr. 18, 306. 38. Martin L. J., Hadfield M. G., Dellovade T. L. and Price D. L. (1991) The striatal mosaic in primates: patterns of neuropeptide immunoreactivity differentiate the ventral striatum from the dorsal striatum. Neuroscience 43, 397417. A., Groenewegen H. J. and Haber S. N. (1993) Shell and core in the primate ventral 39. Meredith G. E., Pattiselanno striatum identified with antibodies against calbindin. Sot. Neurosci. Abstr. 19, 1435 projection in the monkey: topography and relationship 40. Mizobuchi M. M. and Haber S. N. (7992) The cingulostriatal to chemical organization. Sot. Neurosci. Abstr. 18, 306. K., Lu W. X. and Haber S. N. (1992) Distribution of dopamine D2 receptors in subpopulations of primate 41 Nagamoto midbrain douaminergic neurons. Sot. Neurosci. Abstr. 18, 660. connections. 42. Nauta W. J. h. and Domesick V. B. (1978) Crossroads of limbic and striatal circuitry: hypothalamic-nigral In Limbic Mechanisms feds Livingston K. E. and Hornykiewicz O.), pp. 75-93. Plenum, New York. 43. Nauta W. J. H. and Domesick VT B. (1984) Afferent and efferent relationships of the basal ganglia. In Functions of the Basal Ganglia: Cibu Foundation Symposium, pp. 3-29. Pitman, London. 44. Nauta W. J. H., Smith G. P., Faull R. L. M. and Domesick V. B. (1978) Efferent connections and nigra afferents of the nucleus accumbens septi in the rat. Neuroscience 3, 385401. 45. Olszewski J. and Baxter D. (1954) Cytoarchitecture of the Human Brain Stem (ed. S. Karger). afferents to caudate nucleus and putamen in 46. Parent A., Mackey A. and De Bellefeuille L. (1983) The subcortical primate: a fluorescence retrograde double labeling study. Neuroscience 10, 113771150. 0. T. and Griffiths A. C. (1985) The topographic order of inputs to nucleus accumbens in the rat. 47. Phillipson Neuroscience 16, 275-296. R. (1983) Comparative morphology of the substantia nigra and ventral 48. Poirier L. J., Giguere M. and Marchand tegmental area in the monkey, cat and rat. Bruin Res. Bull. 11, 371-397. in the squirrel monkey: an experimental 49. Poletti C. E. and Cresswell G. (1977) Fornix system efferent projections degeneration study. J. camp. Neural. 175, 101-128. projections in the monkey. An 50. Russchen F. T., Bakst I., Amaral D. G. and Price J. L. (1985) The amygdalostriatal anterograde tracing study. Bruin Res. 329, 241-257. projections in the rat. Topographical organization and fiber 51. Russchen F. T. and Price J. L. (1985) Amygdalostriatal morphology shown using the lectin PHA-L as an anterograde tracer. Neurosci. Left. 47, 15-22. of catecholamine-containing neurons in the brain of the 52 Schofield S. P. M. and Everitt B. J. (1981) The organization rhesus monkey (Macucu mulatta). J. Anut. 132, 391-418. Differential connections of caudate nucleus and putamen in the squirrel monkey (Saimiri 53 Smith Y. and Parent A. (1986) ._ _ .- ^__ sciureus). Neuroscience 18, 347-j I I. 54. Szabo J. (1979) Cellular and synaptic organization of basal ganglia. Appl. Neurophysiol. 42, 9-12. 55. Szabo J. (1980) Organization of the ascending striatal afferents in monkeys. J. camp. Neural. 189, 3077321. 56. Tanaka C. (1982) Histochemical mapping of catecholaminergic neurons and their ascending fiber pathways in the rhesus monkey brain. Brain Res. Bull. 9, 255-270. 57. Van Hoesen G. W., Yeterian E. H. and Lavizzo-Mourney R. (1981) Widespread corticostriate projections from temporal cortex of the rhesus monkey. J. camp. Neural. 199, 205-219.

Midbrain projections to primate ventral striatum

623

58. Voom P., Gerfen C. R. and Groenewegen H. J. (1989) Compartmental organization of the ventral striatum of the rat: Immunohistochemical distribution of enkephalin, substance P, dopamine, and calcium-binding protein. J. camp. Neural. 2%). 189-201. 59. Yeterian E. H. and Van Hoesen G. W. (1978) Cortico-striate projections in the rhesus monkey: the organization of certain cortico-caudate connections. Brain Res. 139, 4343. 60. Zaborszky L., Alheid G. F., Beinfeld M. C., Eiden L. E., Heimer L. and Palkovits M. (1985) Cholecystokinin innervation of the ventral striatum: a morphological and radioimmunological study. Neuroscience 14, 427453. 61. Zahm D. S. (1992) An electron microscopic morphometric comparison of tyrosine hydroxylase immunoreactive innervation in the neostriatum and nucleus accumbens core and shell. Brain Res. 575, 341-346. 62. Zahm D. S. and Brog J. S. (1992) On the significance of subterritories in the “accumbens” part of the rat ventral striatum. Neuroscience 50, 751-767. 63. Zahm D. S. and Heimer L. (1990) Two transpallidal pathways originating in the rat nucleus accumbens. J. cotnp. Neural. 302, 437446. 64. Zahm D. S. and Heimer L. (1993) Specificity in the efferent projections of the nucleus accumbens in the rat: comparison of the rostra1 pole projection patterns with those of the core and shell. J. camp. Neural. 327, 220-232. (Accepted 18 October 1993)