THE JOURNAL OF COMPARATIW3NEUROLOGY 2991-16 (1990)

Immunohistochemical Study of the Serotoninergic Innervationofthe Basal Gangtiahthe SquirrelMonkey BRIGMTE LAVOIE AND AND& PARENT Centre de recherche en neurobiologie, Hapita1 de 1'Enfant-Jesus, Qubbec, Canada G1J 124

ABSTRACT A specific antibody raised against 5-hydroxytryptamine (5-HT)conjugated to bovine serum albumin was used to study the serotoninergic innervation of the basal ganglia in the squirrel monkey (Saimiri sciureus). At midbrain level, numerous fine 5-HT-immunoreactive axons were seen to arise from the immunopositive neurons of the dorsal raphe nucleus and less abundantly from those of the nucleus centralis superior. The bulk of these axom formed a rather loosely arranged bundle that arched ventrorostrally through the central portion of the midbrain tegmentum and ascended toward the ventral tegmental area. Several fascicles detached themselves from this bundle to reach the substantia nigra where they arborized into a multitude of heterogeneously distributed 5-HT terminals. The 5-HT innervation was particularly dense in the pars reticulata but much less so in the pars compacta of the substantia nigra. More rostrally other 5-HT fibers swept dorsolaterally and formed a remarkably dense network of varicose fibers within the subthalamic nucleus. A multitude of 5-HT axons continued their ascending course within the lateral hypothalamic area, and many of them swept laterally to invade the lenticular nucleus. At pallidal levels, the 5-HT axons arborized much less profusely in the external segment than in the internal segment, which contained numerous 5-HT varicose fibers and terminals arranged in a typical bandlike pattern. At striatal levels, the 5-HT terminals were particularly abundant in the ventral striatum, including the nucleus accumbens and deep layers of the olfactory tubercle. They also abounded in the ventrolateral region of the putamen and the ventromedial aspect of the caudate nucleus. Overall, the number of 5-HT fibers and terminals decreased progressively along the rostrocaudal axis of the striatum and several large and elongated zones rather devoid of 5-HT immunoreactivity were visualized, particularly in the caudate nucleus and the dorsal putamen. These zones of poor 5-HT immunoreactivity were in register with similar areas devoid of tyrosine hydroxylase immunoreactivity as seen on contiguous sections. These findings reveal that all the core structures of the basal ganglia in primates receive a significant serotoninergic input, but that the densities and patterns of innervation vary markedly from one structure to the other. Key words: 5-hydroxytryptamine,striatum, pallidum, subthalamic nucleus, substantia nigra, primates

Serotonin or 5-hydroxytryptamine (5-HT)is known to be widely distributed in the brain of mammals and other vertebrates (Parent et al., '84) and has been implicated in several physiological and behavioral functions (Azmitia and Gannon, '86; Fallon and Loughlin, '87). The involvement of this indolamine in the function of the basal ganglia is perhaps one of its most well documented roles (Dray, '81; Soubrie et al., '84). For instance, a multitude of experimental studies in animals revealed that 5-HT plays an important role as a neuromediator at basal ganglia levels. This idea is further supported by the variations in the concentration of serotonin and changes in the number of its specific receptors that are often encountered in postmortem examination of the basal forebrain of patients who suffered from (0

1990 WILEY-LISS, INC.

motor disorders such as Parkinson's and Huntington's diseases, in which the basal ganglia are known to be involved (Hornykiewicz, '76; Dray, '80, '81; Walsh et al., '82; Agid et al., '87; Reynolds and Pearson, '87). The first demonstration of the cellular localization of 5-HT in the central nervous system was achieved with the help of the Falck and Hillarp histofluorescence procedure ('62). This method has permitted detailed studies of the topographical distribution of 5-HT cell bodies, which were Accepted June 1,1990 Address reprint requests to Brigitte Lavoie, Centre de recherche en neurobiologie, HBpital de 1'Enfant-Jesus, 1401 18' rue, Quebec, Quebec, Canada, G1J 124.

B. LAVOIE AND A. PARENT

2

found to be mostly located in the brainstem raphe nuclei in various mammalian species (Dahlstrom and Fuxe, '64; Poitras and Parent, '78; Felten and Sladek, '82). However, this technique was not sufficiently sensitive for 5-HT so that several indirect approaches have had to be used to gain further knowledge of the anatomical organization of the serotoninergic neuronal system, particularly its poorly known axonal arborization (Fuxe, '65). Thus tritiated amino acids were utilized as anterograde tracers to investigate the origin and overall organization of the ascending serotoninergic projections in the rat (Conrad et al., '74; Fibiger and Miller, '77; Azmitia and Segal '78) and the cat (Bobillier et al., '75, '76). The tracers were injected into the dorsal raphe nucleus (DR) and/or the median raphe nucleus (MR) (nucleus centralis superior), which are the major sources of the ascending serotoninergic pathways. Retrograde labeling investigations with horseradish peroxidase (HRP) and fluorescence tracers served to complement the data obtained with anterograde tracers (Fibiger and Miller, '77; Van der Kooy and Hattori, '80a,b; Steinbusch et al., '80, '81). Another approach used the powerful reuptake and storage capability that neurons exhibit for their own neurotransmitter to study more specifically the serotoninergic system. Hence the radioautographic examination of brain sections of rats following L3H] 5-HT injections into the cerebral ventricles (Descarries and Beaudet, '78) helped to unravel some of the major organizational features of the ascending serotoninergic pathways (Parent et al., '81). A similar approach was also used to study the ultrastructural characteristics of the serotoninergic axon terminals in the basal ganglia of the rat (Soghomonian et al., '87, '89). However, the most crucial event in the search for tools to study the morphology of the serotoninergic neurons was certainly the development of an immunohistochemical procedure using specific antibodies raised against 5-HT conjugated to bovine serum albumin with paraformaldehyde (Steinbusch et al., '78). This approach has allowed a detailed mapping of the 5-HT cell bodies and axon terminals in the brain of the rat (Steinbusch, '81). Although this immunohistochemical technique has not yet been extensively applied to nonrodent mammals, the distribution of 5-HT-immunoreactive fibers in some components of the basal ganglia has been briefly examined in rats, cats, and macaque monkeys from a comparative point of view (Mori et al., '85a,b, '87). Furthermore, the synaptic organization

of the 5-HT-immunopositive terminals in various regions of the basal ganglia has been studied in macaques by Pasik and colleagues (Pasik and Pasik, '82; Pasik et al., '84a,b). Therefore, in order to further our knowledge of the monoaminergic innervation of the basal ganglia in primates, we thought it useful to investigate the organization of the ascending serotoninergic neuronal systems innervating the upper brainstem and the basal ganglia in the squirrel monkey by using 5-HT immunohistochemistry.

MATEXWWSANDMETHODS Preparationof tissue The observations were made on four squirrel monkeys (Saimiri sciureus), 550 to 800 g i n body weight, which were prepared for light immunohistochemical microscopy according to the protocol outlined in detail in a previous publication (Lavoie et al., '89). In brief, the animals were sacrificed by transcardial perfusion under deep barbiturate narcosis. The vascular system was first rinsed with 300 ml of a saline solution (0.9% NaCl) followed by 1 liter of a fixative containing 4% paraformaldehyde and 0.2% glutaraldehyde. Immediately after perfusion the head of the animal was placed in a stereotaxic apparatus and, after removal of the cranial vault, the brain was cut into 5- to 10-mm-thickslabs according to the stereotaxic coordinates of Emmers and Akert ('63). One brain was cut in the sagittal plane, whereas the others were sectioned in the transverse plane. The blocks were postfixed for 1hour in a 4% paraformaldehyde solution, kept overnight at 4°C in a phosphate buffer solution (PBS, 0.1 M, pH 7.41, and then cut at 50 p m with a Vibratome (Oxford Instruments). The sections were serially collected in a PBS solution prior to the immunohistochemical processing.

Itnmunohistochemistry Complete series of sections from the brainstem raphe nuclei to the rostra1 striatum were processed for the demonstration of 5-HT immunoreactivity according to the peroxidase anti-peroxidase (PAP) procedure of Sternberger ('86). The primary antibody was obtained from Immunonuclear Corp. (Stillwater, MN) and derived from rabbits

Abbreviations

AC AML AS AV CD CeM CM CP EML F GP GPe GPi GPi-1 GPi-m GPs IC ICA IML

IP L

anterior commissure accessory medullary lamina nucleus accumbens anterior ventral thalamic nucleus caudate nucleus nucleus centralis medialis centre median nucleus cerebral peduncle external medullary lamina fornix globus pallidus external segment of GP internal segment of GP lateral part of GPi medial part of GPi subcommissural part of GP internal capsule median island of Cajella internal medullary lamina interpeduncular nucleus lenticular fasciculus

LD LS LV NAC NST N 111 OT P PF PUT

R SI SNc SNr ST TM VA VL VP ZI

laterodorsal thalamic nucleus lateral septum lateral ventricle nucleus of the anterior commissure nucleus of the stria terminalis oculomotor nerve root fibers optic tract nucleus paraventricularis thalami parafascicular nucleus putamen nucleus reticularis thalami substantia innominata pars compacta of the substantia nigra pars reticulata of the substantia nigra subthalamic nucleus mammillothalamic tract ventralis anterior thalamic nucleus ventrolateral thalamic nucleus ventroposterior thalamic nucleus zona incerta

SEROTONIN IN PRIMATE BASAL GANGLIA immunized against serotonin conjugated to bovine serum albumin with paraformaldehyde (Ranadive and Sehon, '67). All the immunohistochemical reagents were diluted in a PBS solution containing 1%normal goat serum (NGS) and 0.1%Triton X-100. The free-floating sections were first preincubated for 30 minutes at room temperature in a solution containing 10%NGS, 0.4% Triton X-100, and PBS. Following this step, they were incubated overnight at 4°C with rabbit anti-5-HT diluted 1500. The sections were then rinsed, reincubated for 1 hour at room temperature in a goat antirabbit IgG (1:30, Sigma), rinsed again, and incubated with rabbit PAP complex (1:100, Sigma) for 1hour at room temperature. Finally, the bound peroxidase was revealed by a last incubation in a solution of 3-3' diaminobenzidine tetrahydrochloride (DAB, 0.05%, Sigma) and hydrogen peroxide (H,O,, 0.008%). The sections were mounted on gelatine-coated slides and observed with a Leitz microscope under both bright- and darkfield illuminations. In addition, some sections were taken at various levels throughout the rostrocaudal extent of the basal ganglia and immunostained with tyrosine hydroxylase (TH) antibodies (1:ZOO; Eugene Tech International, Allendale, NJ) for comparison. The protocol followed to reveal the TH immunoreactivity was basically the same as the one outline above (see Lavoie et al., '89).

Control experiments To test the specificity of the 5-HT immunostaining observed in the present study, two kinds of control experiments were undertaken. First, a series of contiguous sections were processed as above except that the antiserotonin serum was replaced by nonimmunized rabbit serum. Second, other sections were incubated in rabbit anti-5-HT serum preabsorbed for 16 hours at 4°C with serotonincreatine sulfate or dopamine hydrochloride at concentrations ranging from 10- to 10- M.

RESULTS visualizationof 5-IFI'immunoreactvity Only neuronal elements showed preferential immunostaining in the present material. The specificity of the immunostaining was ascertained most of all by the intense and selective labeling of nerve cells somata in the raphe nuclei compared to a negligible staining of known dopaminergic and noradrenergic cell groups lying in adjoining areas of the brainstem. Furthermore, sections incubated in normal rabbit serum remained virtually free of immunostaining, whereas a marked decrease of the immunostaining was observed in sections incubated with antiserotonin serum preabsorbed with serotonin. In these experiments the intensity of the immunostaining was inversely proportional to the concentration of the serotonin used for the immunoabsorption (see Steinbusch, '81).In contrast, no significant decrease in the 5-HT immunostaining was observed in sections incubated with the antiserotonin serum preabsorbed with dopamine. The overall pattern of distribution of 5-HT-immunoreactive fibers and axon terminals in the upper brainstem and basal ganglia of the squirrel monkey is illustrated schematically in Figures 1to 3, whereas photomicrographs of some of the most characteristic immunoreactive neuronal profiles encountered in these structures are shown in Figures 4 to 6. A general description of the organization of the 5-HT-immunoreactive fiber pathways is provided first,

3 followed by comments on the patterns of innervation of the substantia nigra, the subthalamic nucleus, the globus pallidus, and the striatum.

The ascending5-HTpathways In the present material the origin and initial trajectory of the ascending 5-HT fibers could be visualized in great detail in sections cut along the sagittal plane. These thin and nonvaricose immunopositive fibers were seen to originate from the darkly stained cell bodies of the DR and less abundantly from neurons of the nucleus centralis superior (or MR). They formed a multitude of small and compact fascicles that arched rostroventrally and traversed the central portion of the midbrain tegmentum where they appeared as a strikingly intricate network. These fiber fascicles collected themselves dorsomedially to the substantia nigra (SN) in the form of a rather diffused bundle (Fig. 1) that passed through the ventral tegmental area (VTA) and ascended within the lateral hypothalamic area (Fig. 2 ) . Several fascicles detached themselves from this bundle at different rostrocaudal levels and swept laterally to innervate structures such as the SN, the subthalamic nucleus (ST), the globus pallidus (GP) and the putamen (Figs. 1-3). Hence the bundle of 5-HT fibers tapered as it ascended within the lateral hypothalamic area and, because it was largely composed of very thin fibers, could not followed farther rostrally than midhypothalamic levels.

The substantianigra (SN) The SN was by far the most densely innervated components of the basal ganglia. This serotoninergic innervation derived from axons that reach the structure mainly from its caudal aspect and arborized profusely immediately upon entering the SN. In fact, very few distinct 5-HT fibers were found within the confines of the SN, although several immunoreactive fiber fascicles coursed along its dorsolatera1 surface (Fig. 1A). Within the SN the 5-HT axon terminals were heterogeneously distributed, being more abundant in the caudal third than in the rostral two-thirds of the structure. They were also significantly more numerous in the pars reticulata (SNr), especially its lateral portion (the pars lateralis), than in the pars compacta (SNc). This was particularly obvious in the caudomedial portion of the SN where there was a characteristic oval sector formed by densely packed reticulata cells surrounded by more scattered compacta neurons. This oval sector contained a multitude of 5-HTimmunoreactive axon terminals, whereas the surrounding areas were much less densely innervated (Fig. 1A). Conversely, a large oval sector rather devoid of 5-HT axon terminals was disclosed laterally in the caudal third of the SNr (Fig. 1A). In the rostral third of the SN, the 5-HT axonal varicosities were particularly numerous and densely packed ventromedially along the dorsal surface of the cerebral peduncle (Figs. IB, 2C).

The subthalamic nucleus (ST) In contrast to what has been noted in the SN, the serotoninergic innervation of the ST was composed almost

Figs. 1-3. Camera lucida drawings of transverse sections through the caudal (11, middle ( Z ) , and rostral (3) thirds of the basal ganglia in the squirrel monkey illustrating the distribution of 5-HT-immunoreactive fibers (sinuous lines) and axon terminals (dots). In each figure A is more caudal than B.

4

Figure I

5

Figure 2

SEROTONIN IN PRIMATE BASAL GANGLIA

Fig. 4. Darkfield photomicrographs depicting various features of the organization of the 5-HT-immunoreactive fibers and axon terminals encountered in the pallidum of the squirrel monkey. The pallidum is shown at a low power magnification in A. In this transverse section taken through the middle third of the structure, the dense and bandlike innervation of the internal segment compared to a much weaker innervation of the external segment can be appreciated. B shows part of

7

the dense terminal field observed in the peculiar caudoventral portion of the external pallidal segment. C to E depict the mediolateral gradient of the 5-HT innervation of the pallidum as seen in the middle portion of the structure (same level as in A). The medial and lateral parts of the internal segment of the pallidum are illustrated in C and D, respectively, whereas the dorsal region of the external segment of the pallidum is shown in E. Scale bars: 1mm (A), 100 p,m (B-E).

8

Fig. 5. Darkfield photomicrographs showing the dense network of 5-HT fibers pervading the subthalaniic nucleus (A), some of the immunoreactive fibers coursing through the internal capsule between the subthalamic nucleus on the right (visible in part) and the pallidum on the left (B),and the dense terminal field occurring in the rostral part of the substantia nigra pars reticulata (C). D and E compare striatal

B. LAVOIE AND A. PARENT

zones of poor 5-HT immunoreactivity (asterisks in D) in register with similar zones of poor tyrosine hydroxylase iTH) immunostaining (asterisks in E) as seen on adjacent sections. These zones occurred on each side of the internal capsule (IC) in the rostral part of the striatum; the caudate nucleus is in the upper portion of the figures. Scale bars: 200 Frn (A), 100 pm (B,C),500 km (D,E).

SEROTONIN IN PRIMATE BASAL GANGLIA

9

Fig. 6. Lightfield photomicrographsillustratingsome ofthe morphological characteristics of the 5-HT-immunopositivefibers present in the internal segment of the pallidum (A),the putamen (B),the subthalamic nucleus (Ci, and the caudate nucleus (D). Note that the 5-HT fibers

form typical plexuses in the caudate nucleus (D),whereas they remained rather linear in the putamen (B). Scale bars: 500 pm (A,B), 400 pm (C,D).

exclusively of fibers (Figs. lB, 5A,B). Only a few isolated axonal varicosities could be clearly delineated in this nucleus, principally in its ventrolateral part. The fibers innervating this component of the basal ganglia arose mainly from one of the fascicles that left the major 5-HT bundle in

the lateral hypothalamus and coursed along the dorsal surface of the ST. Within the nucleus itself, the immunoreactive fibers were either thin and varicose or thick and smooth (Fig. 6 0 . Both types of fibers were rather evenly distributed throughout the rostrocaudal extent of’ the nu-

B. LAVOIE AND A. PARENT

10 cleus, but appeared slightly more numerous laterally than medially.

The globus pallidus (GP) The GP received a strikingly massive serotoninergic innervation in the squirrel monkey. This innervation derived, as mentioned above, from several fascicles leaving the main 5-HT bundle at various levels along its course in the lateral hypothalamus. At caudal pallidal levels, some of the immunoreactive fibers that coursed along the dorsal surface of the ST continued their route laterally and pierced the internal capsule to reach the pallidum from its dorsal surface (Figs. lB, 2A). At more rostral levels, other fiber fascicles swept laterally and invaded the pallidum more directly from its ventromedial aspect (Fig. 2B). Other ventrally coursing fibers swept dorsally to invade the various medullary laminae separating the different components of the lenticular nucleus. These immunoreactive fibers were more numerous in the external than in the internal medullary lamina and appeared absent in the accessory medullary lamina (Figs. lB, 2,3A). Some of these fibers continued their route dorsally through the internal capsule to finally reach the caudate nucleus (Fig. 2B), whereas others turned perpendicularly to enter the putamen (Figs. lB, 2,3A). Overall, the immunoreactive fibers reaching the pallidum arborized more profusely in the internal (GPi) than in the external (GPe) pallidal segment, although the pattern significantly varied from rostral to caudal. In the caudal third of the pallidum, the 5-HT fibers were equally distributed in the two segments. Thin and varicose fibers were seen to arborize in all directions, although some displayed a dorsoventral orientation along the lateral border of each segment (Fig. 1B). In addition, a smaller number of thick and smooth fibers was disclosed in the pallidum, but these fibers appeared to follow a dorsolateral course en route to the putamen (Fig. 1B). In this caudal part of the external pallidum, a strikingly dense and well-circumscribed field of 5-HT axonal varicosities was observed ventrally, along the dorsolateral surface of the optic tract (Figs. lB, 4B). A much smaller number of axonal varicosities were also scattered in the dorsomedial part of the GPi. In the medial third of the pallidum, the pattern of 5-HT innervation was more complex than in the caudal or rostral third. The majority of the immunoreactive fibers reaching the pallidum at this level arborized within the GPi where they displayed a very typical bandlike pattern (Figs. 2A, 4A). The 5-HT-immunoreactive fibers were mostly fine and varicose and abounded particularly at the apex of the GPi and along the inner border of the internal medullary lamina (Figs. 2A, 4A, 6A). Very few immunoreactive profiles were present at the level of the accessory medullary lamina in the central portion of the GPi. A multitude of large and round axonal varicosities were scattered among the immunoreactive fibers (Fig. 2B), as well as in dorsomedial portion of the GPi. In contrast, very few 5-HT axonal varicosities were encountered in the GPe, which contained only a moderate number of immunopositive fibers mostly confined to its dorsal third (Figs. 2B, 4E). Thus in the middle third of the globus pallidus, the 5-HT-immunoreactive neuronal profiles were distributed according to a mediolateral gradient. The number of these profiles was very large in the medial part of the GPi, moderate in the lateral part of the GPi, and weak in most of the GPe (Figs. 4C-D). In the rostral third of the globus pallidus, numerous fine 5-HT axonal varicosities were present, particularly in the

medial half of the structure (Fig. 3A). A few thick and smooth immunopositive fibers oriented dorsolaterally toward the putamen were scattered among these 5-HT axonal varicosities in the dorsal part of the pallidum (Fig. 3A). The subcommissural pallidum (GPs) also contained numerous 5-HT axonal varicosities, but no immunoreactive fibers were encountered at this level (Fig. 3A). The density of the immunopositive axonal varicosities in the GPs was somewhat similar to that found in the dorsal pallidum at the same level. The adjoining substantia innominata also contained 5-HT axonal varicosities, but their number was smaller than in the overlying pallidal complex.

Thestriatum The serotoninergic innervation of the striatum consisted of a highly complex and heterogeneously distributed fiber network among which were scattered numerous isolated axonal varicosities. This network was composed of a multitude of very thin and varicose fibers among which were scattered a smaller number of thick and smooth fibers. The fibers of small caliber branched profusely within the striatum and bore varicosities that were more elongated than those encountered in the globus pallidus. In contrast, the thick and smooth fibers coursed in a much more linear fashion and branched only infrequently (Fig. 6B). In the head of the caudate nucleus some of the thin immunoreactive fibers formed several small plexuses (Fig. 6D), a feature that was not found elsewhere in the striatum. Topographically, there was a marked rostrocaudal gradient in the distribution of the 5-HT fibers, the rostral portions of the striatum being much more densely innervated than the caudal ones (Figs. 1-31. A dorsoventral gradient was also noted, particularly in the rostral (precommissural) striatum, where the number of immunoreactive fibers and axon terminals was found to be much larger in ventral than dorsal striatum (Fig. 3B). In the ventral striatum, which includes the nucleus accumbens, the ventralmost portion of the putamen and deep layers of the olfactory tubercle, dense terminal fields of 5-HT axonal varicosities occurred in the dorsal third of the nucleus accumbens and around the large median island of Cajella, the island itself remaining virtually devoid of 5-HT neuronal profiles (Fig. 3B). In the putamen, the number of 5-HT fibers was significantly larger in the lateral than in the medial half of the structure, and this pattern persisted along the entire rostrocaudal extent of the structure (Figs. lB, 2,3). In the head and the body of the caudate nucleus, a dense field of 5-HT axonal varicosities was present within the medial border of the structure along the inner surface of the lateral ventricle (Figs. 2B, 3). This field of dense terminal labeling in the caudate nucleus continued without a clear demarcation within the bed nucleus of the stria terminalis, which was also very densely innervated (Fig. 3A). In the remaining portion of the caudate nucleus, immunoreactive fibers predominated principally along the lateral border of the structure near the internal capsule. A closer examination of the serotoninergic innervation of the striatum in the squirrel monkey revealed several zones of poor 5-HT immunostaining that were particularly abundant on each side of the internal capsule at the level of the precommissural striatum (Fig. 3). These zones had a cross-sectional area ranging from 0.4 to 1.2 mm'; they contained only a few thin 5-HT fibers but were lying in a densely innervated matrix. Observations made on contiguous sections immunostained for tyrosine hydroxylase (TH) revealed that the zones of poor 5-HT immunoreactivity

SEROTONIN IN PRIMATE BASAL GANGLIA were largely in register with similar zones of poor TH immunostaining. The correspondence between some striatal zones of poor 5-HT and TH immunostaining is shown in Figure 5D,E. In addition, smaller striatal zones displaying a very dense 5-HT immunostaining were found particularly in the ventrolateral portion of the putamen (Fig. 3B). These hyperdense zones were often lying at the periphery of zones of poor 5-HT immunostaining. Some small zones of high TH immunostaining were also encountered within the ventrolateral region of the putamen, but a clear correspondence between them and the 5-HT hyperdense zones could not be established beyond doubt.

DISCUSSION The serotoninergc ascendingpathways In all major vertebrate groups, the 5-HT cell bodies form a somewhat continuous column that extends along the midline from caudal medulla to rostral midbrain levels (Parent et al., '84). This remarkable neuronal aggregate was first visualized in the rat by means of the histofluorescence method and subdivided into 9 more or less distinct entities termed groups B1 to B9 by Dalstrom and Fuxe ('64).A similar pattern of distribution has been described in cats (Parent and Poitras, '78; Wiklund et al., '81; Jacobs et al., '841, monkeys (Hubbard and DiCarlo, '74; Schofield and Everitt, '81; Schofield and Dixon, '82; Sladek et al., '82; Takeuchi et al., '82; Felten and Sladek, '83;Azmitia and Gannon, '861, and human fetuses (Nobin and Bjorklund, '73; Olson et al., '73). The most rostral components of the 5-HT cell column, the DR (groups B7 and B6) and the nucleus centralis superior (MR; groups B8 and part of B7), are know to project to the forebrain, whereas the most caudal elements innervate the spinal cord (see review by Azmitia, '78). Detailed anatomical studies of the ascending projections from the midbrain raphe nuclei revealed that the serotoninergic innervation of the basal ganglia derives largely from the DR with a lesser contribution from the MR (Conrad et al., '74; Bobillier et al., '75, '76; Fibiger and Miller, '77; Azmitia and Segal, '78; Moore et al., '78; Steinbusch et al., '80, '81). One of our previous studies in the rat showed that most of the 5-HT ascending fibers arising from the DR and MR formed a complex projection system that we termed the transtegmental pathway (Parent et al., '81). This system converges ventralward across the decussation of the superior cerebellar peduncles and sweeps rostralward to enter the VTA and ascend within the medial forebrain bundle in the lateral hypothalamus. The organization of the ascending 5-HT projections in the squirrel monkey, as visualized in the present study, appears strikingly similar to that in the rat. A multitude of immunoreactive fibers were seen to arise from the DR and less abundantly from the nucleus centralis superior at various rostrocaudal levels along the upper brainstem in Saimiri. These fibers formed several small and intertwined fascicles, which appeared in the sagittal plane as a remarkably complex reticulum covering much of the central core of the midbrain tegmentum. The bulk of these fibers arched ventrally to invade the ventral tegmental area and ascend within the lateral hypothalamic area. The ascending 5-HT fibers were more diffusely organized and slightly more ventrally located in the lateral hypothalamus than the dopaminergic fibers visualized on adjacent sections immunostained for TH. Because of their very small caliber, these

11

5-HT ascending fibers could not be followed farther rostrally than midhypothalamic levels. In macaque monkeys two ascending 5-HT bundles were visualized at caudal hypothalamic levels. As in the rat numerous fibers ascend within the medial forebrain bundle and another fiber contingent, the dorsal raphe cortical tract, courses more dorsally in the hypothalamus (Azmitia and Gannon, '86). In contrast to the situation in the rat, however, the dorsal raphe cortical tract, which ascends to the cerebral cortex via the internal capsule, was reported to be larger than the medial forebrain bundle in macaque monkeys (Azmitia and Gannon, '86). In the squirrel monkey the dorsal raphe cortical tract could not be identified with certainty. Furthermore, the number of myelinated 5-HT ascending fibers in the lateral hypothalamus appears much greater in macaques than in rats (Azmitia and Gannon, '83). This observation argues against 5-HT as having an exclusively diffuse, slow-acting function, with transmitter being release throughout its axons. Instead, it suggests that a specialized role for this neuromediator may have evolved in primates. The degree of myelinization of the ascending 5-HT fibers has not yet been investigated in the squirrel monkey.

Thesubstantianigra The SN was by far the most heavily innervated components of the basal ganglia in the squirrel monkey. This is in agreement with biochemical studies in rats (PalkoFlts et al., '74; Saavedra, '771, monkeys (Shannak and Hornykiewicz, 'SO), and humans (Fahn et al., '71; Mackay et al., '78), which designated the SN as one the brain structures most enriched with serotonin and its major metabolite 5Ghydroxyindolacetic acid (5-HIAA). The first axonal transport studies of the raphe-nigral projection suggested that it arose from both the DR and MR (Bobillier et al., '75, '76; Moore et al., '78) and terminated either within both the SNc and the SNr (Bobillier et al., '76; Fibiger and Miller, '77) or exclusively within the SNc (Conrad et al., '76). However, other investigations with more specific injection sites revealed that this projection arises mainly from the DR and terminate principally in the SNr (Azmitia and Segal, '78; Imai et al., '86; Wirtehafter et al., '87). In addition, double-labeling studies in the rat showed that the raphe-nigral projection is largely composed of collaterals of the raphe-striatal pathway (Van der Kooy and Hattori, 'SOa), appears to be strictly unilateral (Van der Kooy and Hattori, 'Sob), and contains a significant number of nonserotoninergic elements (Steinbusch et al., '80; Descarries et al., '86). In primates the exact cellular origin and degree of axonal collateralization of the raphe-nigral projection remain to be investigated. The demonstration by axonal transport studies that the SNr is the major target of the raphe-nigral projection confirms earlier histofluorescence findings (Fuxe, '65; Aghajanian et al., '73) and is in agreement with more recent data obtained by the immunohistochemical approach (Steinbusch et al., '81; Pasik et al., '84b; Mori et al., '87). In the squirrel monkey 5-HT-immunoreactive axonal varicosities were found to be highly heterogeneously distributed in the SNr. They abounded particularly in the lateral portion (pars lateralis) of the SNr and formed clusters of dense terminal fields distributed between the cell columns of the SNc more medially. Some 5-HT varicosities were also scattered within the SNc, and immunoreactive fibers ascended along the cell column of the pars compacta. A recent

B. LAVOIE AND A. PARENT

12

comparative study of the 5-HT immunoreactivity in the SN of rats, cats, and macaque monkeys revealed that the number and complexity of the distribution of serotoninergic fibers and axonal varicosities in this structure were much greater in primates than in nonprimates (Mori et al., '87). In the macaque monkey (Macaca fuscata), the pattern of distribution of the 5-HT immunoreactivity at SN levels was overall similar to that in the squirrel monkey. Ultrastructural studies in rats (Parizek et al., '71; Herve et al., '87; Nedergaard et al., '881, cats (Calas et al., '761, and monkeys (Pasik et al., '84b; Mori et al., '87) revealed that both synaptic and nonsynaptic axonal varicosities occurred in the SN neuropil. The axonal varicosities exhibiting junctional appositions were reported to be of the asymmetrical type (Pasik et al., '84b; Herve et al., '87; Nedergaard et al., '88; however, see Mori et al., '87) and some of these 5-HT synapses occurred directly on dopaminergic neurons in both the SN (Nedergaard et al., '88) and the VTA (Herve et al., '87) in the r a t . These findings suggest that serotonin can alter the activity of dopaminergic neurons projecting to the dorsal and ventral striatum by acting directly at the level of the SN-VTA complex. Receptors for serotonin have been classified into three main types referred to as 5-HT,-like, 5-HT, and 5-HT3 (Bradley et al., '86). The action of serotonin at the SN-VTA complex level is likely to be mediated by 5-HT, receptors, which are the most abundant 5-HT receptors in this brain area in both primates (Stuart et al., '86; Pazos et al., '87a,b) and rodents (Pazos and Palacios, '85; Pazos et al., '85). In primates the major serotonin receptor type in the SN is of the 5-HT,, subtype (Pazos et al., '87a), whereas the 5-HT1, receptor subtype predominates in the SN of rodents (Pazos and Palacios, '85).

The subthalamicnucleus In the squirrel monkey the ST contained both thin and varicose, as well as thick and smooth serotonin-immunoreactive fibers that formed a dense plexus rather uniformly distributed throughout the nucleus. The only significant heterogeneity was a slight preponderance of 5-HT fibers in the lateral portion of the nucleus. Biochemical investigations revealed significant concentrations of 5-HT and 5 - H I M in the ST of rodents (Palkovits et al., '74; Saavedra, '771, as well as in that of human and nonhuman primates (Cross and Joseph, '81; Walsh et al., '82). Furthermore, recent radioautographic binding studies showed low densities of 5-HT, receptors in this structure in macaque monkeys (Stuart et al., '861, but despite these findings very little is known of the role of serotonin in this component of the basal ganglia. Numerous 5-HT-immunoreactive fibers occurred in the ST of rodents (Steinbusch, '81; Mori et al., '85a), as well as in cats and macaque monkeys (Mori et al., '85a). In the ST of the macaque (Macaca fuscata) 5-HT fibers were more numerous and more heterogeneously distributed than in cats and rats (Mori et al., '85a). As in Saimiri thin varicose fibers forming a dense network together with straight, unbranched, thick fibers were visualized in the ST of M. fuscata (Mori et al., '85a). The 5-HT fibers were said to be more abundant in the ventromedial portion of the nucleus in the macaque, whereas they appeared much more uniformly distributed in the rat and the cat. Hence, a significant species difference between primates and non-primates seems to exist in respect to the density and patterns of distribution of 5-HT fibers in the ST.

The cellular origin of the 5-HT input to the ST is still a matter of controversy. Earlier anterograde tracing studies in cats reported that this innervation arises principally from the MR (Bobillier et al., '76), whereas retrograde labeling studies in cats and cynomolgus monkeys concluded that the DR was the major source of afferents (Rinvik et al., '79). Obviously further studies are needed to clarify the situation and to define the role that serotonin may play in the ST.Anatomical investigations designed to reveal the cellular origin, the degree of collateralization, and the ultrastructural features of the serotoninergic projection to the ST would be particularly useful. Electrophysiological and pharmacological studies are also a prerequisite to a proper understanding of the functional relationship between the 5-HT fibers and the ST neurons.

Theglobuspallidus The GP in the squirrel monkey receives a dense serotoninergic projection, which arborizes more profusely in the GPi than in the GPe, and the numerous thin and varicose fibers a t pallidal levels display a typical bandlike pattern. High concentrations of 5-HT and 5-HIAA were detected in the GP complex of rats (Palkovits et al., '74; Saavedra, '771, monkeys (Shannak and Hornykiewicz, '80), and humans where levels of the neurotransmitter and its major metabolite are significantly higher in the GPi than in GPe (Walsh et al., '82). The GP is also enriched with 5-HT, and poor in 5-HT2receptors in rats (Pazos and Palacios, '85; Pazos et al., '851, monkeys (Stuart et al., '861, and humans (Pazos et al., '87a,b). In humans the density of 5-HT, receptors, which were mainly of the 5-HTI,. subtype, was found to be higher in the GPi than GPe (Pazos et al., '87a), whereas the inverse was reported for macaques (Stuart et al., '86). In macaques the caudoventral portion of the GPe, which in Sainziri receives a particularly dense 5-HT innervation, displayed a very high density of 5-HT, receptors (Stuart et al., '86). Also worth noting is the fact that the GP in primates was reported to be the basal ganglia component that has the highest rate of 5-HT synthesis (Bacopoulos et al., '79). A comparative immunohistochemical study of the distribution of 5-HT fibers in the pallidal complex in rats, cats, and macaque monkeys revealed that both the entopednuclear nucleus (homologue of the GPi) and the globus pallidus (homologue of the GPe) in rats and cats receive a rather dense and homogeneous 5-HT innervation, whereas the GPe in the macaque was less densely innervated than the GPi (Mori et al., 85b). Other neuromediators are known to be differentially distributed in the pallidal complex of rats, cats, and monkeys. For instance, striatonigral fibers enriched with enkephalins terminate preferentially in the GPe, whereas striatofugal fibers expressing substance P arborize mainly within the GPi (Haber and Elde, '81; Gronewegen and Russchen, '84; Gerfen and Young, '88). Moreover, recent immunohistochemical studies showed that the GP in the squirrel monkey receives a significant dopaminergic innervation arborizing principally in the GPi (Lavoie et al., '89). This projection arises from dopaminergic neurons of the SN-VTA complex that are largely distinct from those projecting to the striatum (Smith et al., '89). The 5-HT afferent to the GP originates mostly from the DR (Bobillier et al., '76; Azmitia and Segal, '78; DeVito et al., 'go), but it is not yet known if this projection is a distinct subsystem or made up of collaterals from other major ascending 5-HT

SEROTONIN IN PRIMATE BASAL GANGLIA pathways. In any events, our data on the monoaminergic innervation of the GP in Saimiri indicate that, in addition to their well-established influence upon the striatum, the midbrain serotoninergic and dopaminergic neurons can act directly on the output neurons of the basal ganglia at pallidal levels. Most GP afferents, including the major striato- and subthalamopallidal projections, exhibit a typical bandlike pattern of termination (see Smith et al., '90 for references). This arrangement most likely reflects the fact that the large and discoidal dendritic arborization of pallidal neurons lie parallel to medullary laminae and perpendicular to the incoming striatal or subthalamic axons to which they present their greatest surface (Park et al., '82a; Yelnik et al., '84). The organization of the 5-HT innervation in bands within the GP agrees with the fact that 5-HT-immunoreactive axonal profiles at this level make contact mainly with the dendritic shafts and crests of pallidal neurons (Pasik et al., '84a,b). They form markedly asymmetric synapses similar to those observed in the substantia nigra and often display postsynaptic specializations (Pasik et al., 84a).

13

mammals and termed striosomes by Graybiel and her colleagues (Graybiel and Ragsdale, '78). In more recent studies, zones of poor TH immunoreactivity very similar to the one disclosed in the squirrel monkey (Lavoie et al., '89) were observed in the striatum of cats, macaque monkeys, and humans (Graybiel et al., '87). These zones of low TH immunostaining could be aligned with striosomes visible on contiguous sections stained for AChE or for met-enkephalin (Graybiel et al., '87). Since the zones of poor 5-HT immunostaining disclosed in the present study were in register with areas of poor TH immunoreactivity, it may be inferred that they correspond to the striosomal compartment of the striatum. Taken together, these findings suggest that both serotoninergic and dopaminergic inputs to the striatum in adult primates arborize more profusely in the extrastriosomal matrix than in striosomes. As in other components of the basal ganglia, the most abundant serotonin receptor in the striatum is of the 5-HT, type (Pazos and Palacios, '85; Pazos et al., '85, '87a,b; Stuart et al., '86). In rats and humans the newly characterized 5-HT,, subtype represent approximately 33% of all 5-HT, receptors in the striatum (Herrick-Davis and Titeler, The striaturn '88), and the newly discovered 5-HT3 receptor occur in Despite the fact that levels of 5-HT and 5-HIAA are significant number in the ventral portion of the striatum, significantly lower in the striatum than in the S N and GP in particularly in the nucleus accumbens (Kilpatrick et al., rats (Palkovits et al., '74; Saavedra, '771, monkeys (Shan- '87; Waeber et al., '88). Also worth noting is the highly nak and Hornykiewicz, '80), and humans (Walsh et al., '821, heterogeneous distribution of the 5-HT, receptors in the the serotoninergic innervation of the striatum has been the striatum of macaque monkeys, where numerous patches of subject of innumerable studies. Several methods were used high receptor density occurred (Stuart et al., '86). Although to visualized 5-HT fibers in the striatum, including histoflu- the correspondence between these patches of high receptor orescence (Fuxe, '651, radioautography after L3H1 5-HT binding and other chemical markers of the heterogeneity of uptake (Calas et al., '76; Chan-Palay, '77; Arluison and De the striatum has not yet been established, their size and la Manche, '80; Soghomonian et al., '87, '891, and immuno- configuration suggest that they may correspond to the histochemistry (Steinbusch, '81; Pasik et al., '82, '84b; Mori striosomes (Stuart et al., '86). If it is so, this would et al., '85b; O'Hearn et al., '88; Soghomonian et al., '87, represent another typical case of mismatch between the '89). These investigations revealed a multitude of thin localization of a neurotransmitter and its receptor (see varicose fibers forming a fine network that extends through- Herkenham, '87). High resolution radioautography after [3H]5-HT adminout the striatum. The density of the 5-HT innervation was found to be higher in ventral than dorsal portions of the istration or straight immunohistochemistry were used to striatum in all species. However, in contrast to the situation characterize the ultrastructural features of 5-HT axon in Saimiri, the number of 5-HT fibers and terminals was terminals in the striatum of rats (Arluison and De la reported to increase along the rostrocaudal axis of the Manche, '80; Soghomonian et al., '89), cats (Calas et al., '761, and monkeys (Pasik et al., '82,'84b). These studies striatum in macaque monkeys (Mori et al., '85b). Two types of 5-HT axons have been characterized in the revealed that striatal 5-HT axon terminals are generally striatum of the rat: numerous, fine, varicose axons and a small and contained small pleomorphic synaptic vesicles smaller number of larger and beaded fibers, which are with occasional large granular vesicles and mitochondria. believed to arise from the DR and MR, respectively (O'Hearn Junctional 5-HT terminals synapse exclusively on dendritic et al., '88).The fine but not the beaded axons were shown to spines or shafts and displayed asymmetrical membrane be highly vulnerable to the neurotoxic effects of certain differentiation. However, only 10 to 15%of 5-HT varicosipsychotropic drugs, such as methylenedioxyamphetamine ties in the rat striatum exhibited a synaptic junction, (MDA) and methylenedioxymethamphetamine (MDMA) compared to a junctional incidence of at least 70% for (O'Hearn et al., '88). A similar differential sensitivity to the nonlabeled varicosities in the surrounding neuropil (Arluisame drugs has been reported for 5-HT axons arborizing son and De la Manche, '80; Soghomonian et al., '89). The into the cerebral cortex in primates (Wilson et al., '88), but nonsynaptic appositions were mostly found on nonlabeled their effects on striatal fibers have not yet been investigated axon terminals. in monkeys. These findings indicate that spiny dendrites are the One of the most characteristic features of the 5-HT primary if not the exclusive target of 5-HT synaptic axon innervation of the striatum of the squirrel monkey is the terminals in the striatum. Hence, the medium spiny neupresence of zones of poor 5-HT immunoreactivity sur- rons are likely to be the major cellular target of 5-HT rounded by larger and more densely innervated areas. afferents. Likewise, the dopaminergic input to the rat These zones of poor 5-HT immunostaining were in register striatum was shown to terminate through symmetrical with similar zones of poor TH immunostaining as seen on junctional complexes mainly, if not solely, onto medium adjacent sections. The size and configuration of these zones spiny neurons (Freund et al., '84). These observations are very reminiscent of the zones of poor acetylcholines- together with the findings of the present study suggest that terase (AChE) staining disclosed in the striatum of several the bulk of the monoaminergic innervation of the striatum

B. LAVOIE AND A. PARENT

14 exerts its influence mostly upon medium spiny projection neurons, particularly those lying within the extrastriosoma1 matrix.

Functional considerations Despite the wealth of information on the anatomical organization of the 5-HT neuronal system innervating the basal ganglia, the function of serotonin within this set of structures is still not clear. Electrophysiological studies have yielded particularly confusing results. First, despite that most 5-HT synapses in the basal ganglia are of the asymmetrical type, which is usually indicative of excitatory processes, electrical stimulation of the DR was reported to inhibit striatal cells in extracellular recording experiments (Miller et al., '75; Olpe and Koella, '77; Davies and Tongroach, '78). Second, intracellular recordings in the striatum revealed that stimulation of the DR produces an excitatory postsynaptic potential followed by an inhibition and a late excitatory potential (Vandermaelen et al., '79; Park et al., '82b). It was suggested that only the initial EPSP is 5-HT-dependent, since it is the only component of the response that is altered by parachlorophenylalanine 5-HT depletion (Park et al., '82b). Third, both inhibitory and excitatory responses have been recorded after iontophoretic application of 5-HT in the rat striatum (Bevan et al., '75; Davies and Tongroach, '78).Fourth, 5-HT synapses onto dopaminergic neurons in the SN have been reported to have a site-dependent, receptor-mediated,facilitatory effect on a specific dendritic calcium-dependent potential in nigrostriatal neurons (Nedergaard et al., '88). This implies that 5-HT can cause a significant change in a local calciumdependent potential without provoking inhibition or excitation of the target cell as a whole. Several explanations may be provided for these discrepancies. First, the DR is not a chemically homogeneous entity; it contains a large proportion of nonserotoninergic neurons including dopaminergic cells (Steinbusch et al., '80; Descarries et al., '86). Furthermore, substance P, enkephalin, and glutamic acid decarboxylase were shown to coexist with 5-HT in several DR cells (see review by Soubrie et al., '84). In addition, the existence of both junctional and nonjunctional relationships between 5-HT terminals and their target cells may account for some of the differences reported above. For instance, direct depolarizing monosynaptic responses to DR stimulation could result from an axo-spinous synaptic action of 5-HT, whereas inhibitory or modulatory effects might be mediated through nonsynaptic, axo-axonic appositions. In fact, axo-axonic links could allow 5-HT to exert presynaptic influences on the other afferent systems, such as the important serotonin/dopamine (Nicolaou et al., '79; Stachowiak et al., '84; Snyder et al., '86; De Simoni et al., '87) and serotoniniacetylcholine (Jackson et al., '88) interactions that occur in the striatum. Finally, further studies on the localization and exact role of the multiple types of serotonin receptors are needed in order to reach a proper understanding of the function of serotonin in the basal ganglia.

ACKNOWLEDGMENTS The authors thank Carole Harvey and Lisette Bertrand for technical assistance and Suzanne Bilodeau for typing the manuscript. This research was supported by grant MT-5781 of the Medical Research Council of Canada to A. Parent. The financial support of the FRSQ and FCAR is

also acknowledged. Brigitte Lavoie was the recipient of a Studentship from the FRSQ.

LITERATURECITED Aghajanian, G.K., M.J. Kuhar, and R.H. Roth (1973) Serotonin-containing neuronal perikarya and terminals: Differential effects of p-chlorophenylalanine. Brain Res. 54:85-101. Agid, Y., F. Javoy-Agid, and M. Ruberg (1987) Biochemistry of neurotransmitters in Parkinson's disease. In C.D. Marsden and S. Fahn (eds): Movement Disorders-2. London: Butterworth, pp. 166-230. Arluison, M., and I S .De laManche (1980) High-resolution radioautographic study of the serotonin innervation of the rat corpus striatum after intraventricular administration of ?HI 5-hydroxytryptamine. Neuroscience 5:229-240. Azmitia, E.C. (1978) The serotonin-producing neurons of the midbrain median and dorsal raphe nuclei. In L.L. Iversen, S. Iversen and S.H. Snyder (eds): Handbook of Psychopharmacology, Vol. 9. New York Plenum Press, pp. 233-314. Azmitia, E.C., and P.J. Gannon (1983) The ultrastructural localization of serotonin immunoreactivity in myelinated and unmyelinated axons within the medial forebrain bundle of rat and monkey. J. Neurosci. 3.2083-2090. Azmitia, E.C., and P.J. Gannon (1986) The primate serotoninergic system: A review of human and animal studies and a report on Macaca fasczcularis. In S. Fahn (ed):Advances in Neurology, Vol. 43. New York: Raven Press, pp. 407-468. Azmitia, E.C., and M. Segal (1978) An autoradiographic analysis of the differential ascending projections of the dorsal and median raphe nuclei in the rat. J. Comp. Neural. 179t641-659. Bacopoulos, N.G., D.E. Redmond, and R.H. Roth (1979) Serotonin and dopamine metabolites in brain regions and cerebral spinal fluid of a primate species: Effects of ketamine and fluphenazine. J. Neurochem. 32:1215-1218. Bevan, P., C.W. Bradshaw, and E. Szabadi (1975) Effects of desimipramine on neuronal responses to dopamine, noradrenaline, 5-hydroxytryptamine and acetylcholine in the caudate nucleus of the rat. Br. J. Pharmacol. 54:285-293. Bobillier, P., F. Petitjean, D. Salvert, M. Ligier, and S. Seguin (1975) Differential projections of the nucleus raphe dorsalis and nucleus raphe centralis as revealed by autoradiography. Brain Res. 85205-210. Bobillier, P., S. Seguin, F. Petitjean, D. Salvert, M. Touret, and M. Jouvet (1976) The raphe nuclei of the cat brain stem: A topographical atlas of their efferent projections as revealed by autoradiography. Brain Res. 113;449-486. Bradley, P.B., G. Engel, W. Fenuik, J.R. Fozard, P.P.A. Humphrey, D.M. Middleniss, E.J. Mylecharane, B.P. Richardson, and P.R. Saxena (1986) Proposal for the classification of nomenclature of functional receptors for 5-hydroxytryptamine. Neuropharmacol. 25563-576. Calas, A., M.J.Besson, C. Gaudry, G. Alonso, J. Glowinski, and A. Cheramy (1976) Radioautographic study of in vivo incorporation of 'H-monoamines in the cat caudate nucleus: Identification of serotoninergic fibers. Brain Res. 118:l-13. Chan-Palay, V. (1977) Indoleamine neurons and their processes in the normal rat brain and in chronic diet-induced thiamine deficiency demonstrated by uptake of 3H-serotonin. J. Comp. Neurol. 176:467-494. Conrad, L.C.A., C.M. Leonard, and D.W. Pfaff (1974) Connections, of the median and dorsal raphe nuclei in the rat: An autoradiographic and degeneration study. J. Comp. Neurol. 156:179-206. Cross, A.J., and M.H. Joseph (1981) The concurrent estimation of the major monoamines metabolites in human and non-human primate brain by HPLC with fluorescence and electrochemical detection. Life Sci. 28;499505. Dahlstrom, A,, and K. Fuxe (1964) Evidence for the existence of monoaminecontaining neurons in the central nervous system. I. Demonstration of monoamines in the cell bodies of the brain stem neurons. Acta Physiol. Scand. 62 (Suppl. 232):l-55. Davies, J., and P. Tongroach (1978) Neuropharmacological studies on the nigrostriatal and raphe-striatal system in the rat. Europ. J. Pharmacol. 5I:91-100. De Simoni, M.G., G. Dal Toso, F. Fodritto, A. Sokola, and S. Algeri (1987) Modulation of striatal dopamine metabolism by the activity of dorsal raphe serotoninergic afferences. Brain Res. 411:81-88. Descarries, L., and A. Beaudet (1978) The serotonin innervation of adult rat hypothalamus. In J.D. Vincent and C. Kordon (eds): Biologie Cellulaire

SEROTONIN IN PRIMATE BASAL GANGLIA des Processus Neurosecretoires Hypothalamiques. Colloques internationauxduC.N.R.S.,Vol. 280. Paris:EditionduC.N.R.S.,pp. 135-153. Descarries, L., F. Berthelot, S. Garcia, and A. Beaudet (1986) Dopaminergic projection from nucleus raphe dorsalis to neostriatum in the rat. J. Comp. Neurol. 249:511-520. DeVito, J.H., M.E. Anderson, and K.E. Walsh (1980) A horseradish peroxidase study of afferent connections of the globus pallidus in Macaca mulatta. Exp. Brain Res. 3835-73. Dray, A. (1980) The physiology and pharmacology of mammalian basal ganglia. Prog. Neurohiol. 14:221-235. Dray, A. (1981) Serotonin in the basal ganglia: functions and interactions with other neurons pathways. J. Physiol. Paris 77:393403. Emmers, R., and K. Akert (1963) A stereotaxic atlas of the brain of the squirrel monkey (Sairniri sciureus). Madison: University of Wisconsin Press. Fahn, S., L.R. Libsch, and R.W. Cutler (1971) Monoamines in the human neostriatum: Topographic distribution in normals and in Parkinson’s disease and their role in akinesia, rigidity, chorea and tremor. J. Neurol. Sci. 14:427-455. Falck, B., N.A. Hillarp, G. Thieme, and A. Torp (1962) Fluorescence of catecholamines and related compounds condensed with formaldehyde. J. Histochem. Cytochem. 10:348-354. Fallon, J.H., and S.E. Loughlin (1987) Monoamine innervation of cerebral cortex and a theory of the role of monoamines in cerebral cortex and basal ganglia. In E.G. Jones and A. Peters (eds): Cerebral Cortex, Vol. 6. New York: Plenum Press, pp. 41-126. Felten, D.L. and J.R. Sladek J r (1983) Monoamine distribution in primate brain. V. Monoaminergic nuclei: Anatomy, pathways and local organization. Brain Res. Bull. 10:171-284. Fibiger, H.C., and J.J. Miller (1977) An anatomical and electrophysiological investigation of the serotoninergic projection from the dorsal raphe nucleus to the substantia nigra in the rat. Neuroscience 2975-987. Freund, T.F., J.F. Powell, and A.D. Smith (1984) Tyrosine hydroxylaseimmunoreactive houtons in synaptic contact with identified striatonigral neurons, with particular reference to dendritic spines. Neuroscience 13:1189-1215. Fuxe, K. (1965) The distribution of monoamines terminals in the central nervous system. Acta Physiol. Scand. 64 (Suppl. 247):37-85. Gerfen, C.R., and W.S. Young I11 (1988) Distribution of striatonigral and striatopallidal neurons in both patch and matrix compartments: An in situ hybridization histochemistry and fluorescent retrograde tracing study. Brain Res. 460:161-167. Graybiel, A.M., and C.W. Ragsdale Jr. (1978) Histochemically distinct compartments in the striatum of human, monkey and cat demonstrated by acetylcholinesterase staining. Proc. Natl. Acad. Sci. (USA) 75:57235726. Graybiel, A.M., E.C. Hirsch, and Y.A. Agid (1987) Differences in tyrosine hydroxylase-like immunoreactivity characterize the mesostriatal innervation of the striosomes and extrastriosomal matrix at maturity. Proc. Natl. Acad. Sci. (USA) 84r303-307. Gronewegen, E.K., and F.T. Russchen (1984) Organization of the efferent projections of the nucleus accumhens to pallidal, hypothalamic and mesencephalic structures: A tracing and immunohistochemical study in the cat. J. Comp. Neurol. 223t349-367. Haber, S., and R. Elde (1981) Correlation between met-enkephalin and substance P immunoreactivity in the primate globus pallidus. Neuroscience 6: 1291-1297. Herkenham, M.(1987) Mismatches between neurotransmitter and receptor localizations in brain: Observations and implications. Neuroscience 23:l-38. Herrick-Davis, K., and M. Titeler (1988) Detection and characterization of the serotonin 5-HT,, receptor in rat and human brain. J. Neurochem. 50:1624-1631. Herve, D., V.M. Pickel, T.H. Joh, and A. Beaudet (1987) Serotonin axon terminals in the ventral tegmental area of the rat: Fine structure and synaptic input to dopaminergic neurons. Brain Res. 43571-83. Hornykiewicz, 0. (1976) Neurohumoral interactions and basal ganglia function and dysfunction. In M.D. Yahr (ed): The Basal Ganglia. New York: Raven Press, pp. 269-280. Hubbard, J.E., and V. DiCarlo (1974) Fluorescence histochemistry of monoamine containing cell bodies in the brain stem of the squirrel monkey (Saimiri sciureus). 111. Serotonin-containing groups. J. Comp. Neurol. 153:385-398. Imai, H., D.A. Steindler, and S.T. Kitai (1986) The organization of divergent axonal projections from the midbrain raphe nuclei in the rat. J. Comp. Neurol. 243:363-380.

15 Jackson, D., J.P. Bruno, M.K. Stachowiak, and M.J. Zigrnond (1988) Inhibition of striatal acetylcholine release by serotonin and dopamine after the intracerebral administration of 6-hydroxydopamine to neonatal rats. Brain Res. 457:267-273. Jacobs, B.L., P.J. Gannon, and E. Azmitia (1984) Atlas of sercitonergic cell bodies in the cat brainstem: An immunohistochemical analysis. Brain Res. Bull. 13:l-31. Kilpatrick, G.J., B.J. Jones, and M.B. Tyers (1987) Identification and distribution of 5-HT, receptors in rat brain using radioligand binding. Nature 330:74&748. Lavoie, B., Y. Smith, and A. Parent (1989) Dopaminergic innervation of the basal ganglia in the squirrel inonkey as revealed by tyrosine hydroxylase immunohistochemistry. J. Comp. Neurol. 289:36-52. Mackay, A.V.P., C.M. Yates, A. Wright, P. Hamilton, and P. Davies (1978) Regional distribution of monoamines and their metabolites in the human brain. J. Neurochem. 30:841-848. Miller, J.J., T.L. Richardson, H.C. Fibiger, and H. McLennan (1975) Anatomical and electrophysiological identification of a projection from the mesencephalic raphe to the caudate-putamen in the rat. Brain Res. 97:133-138. Moore, R.Y., A.E. Halaris, and B.E. Jones (1978) Serotonin neurons of the midbrain raphe: Ascending projections. J. Comp. Neurol. 180:417-438. Mori, S., T.Matsuura, T . Takino, and Y.Sano (1987) Light and electron microscopic immunohistochemical studies of serotonin nerve fibers in the substantia nigra of the rat, cat and monkey. Anat. Embpol. 176:13-18. Mori, S., T. Takino, H. Yamada, and Y. Sano (1985a)Immunohistochemical demonstration of serotonin nerve fibers in the subthalamic nucleus of the rat, cat and monkey. Neurosci. Lett. 62305-309. Mori, S., S. Ueda, H. Yamada, T. Takino, and Y. Sano (1985b) Immunohistochemical demonstration of serotonin nerve fibers in the corpus striatum of rat, cat and monkey. Anat. Embryol. 173:l-5. Nedergaard, S., J.P. Bolam, and S.A. Greenfield (1988) Facilitation of a dendritic calcium conductance by 5-hydroxytryptamine in the substantia nigra. Nature 33:174-177. Nicolaou, N.M., M. Garcia-Munoz, G.W. Arbuthnott, and D. Eccleston (1979) Interactions between serotoninergic and dopaminergic systems in rat brain demonstrated by small unilateral lesions of the raphe nuclei. Europ. J. Pharmacol. 57:295-905. Nobin, A., and A. Bjorklund (1973) Topography of the monoamine neuron systems in the human brain as revealed in fetuses. Acta Phye,iol. Scand. [Suppl.] 388:1-40, O’Hearn, E., G. Battaglia, E.B. De Souza, M.J. Kuhar, and M.E. Molliver ( 1988)Methylenedioxyamphetamine (MDA)and methylenedioxymethamphetamine (MDMA) cause selective ablation of serotoninergic axon terminals in forebrain: Immunocytochemical evidence for neurotoxicity. J. Neurosci. 8.2788-2803. Olpe, H.R., and W.P. Koella (1977) The response of striatal cells upon stimulation of the dorsal and median raphe nuclei. Brain Res. 122357360. Olson, L., L.O. Boreus, and A. Seiger (1973) Histochemical demonstration and mapping of 5-hydroxytrygtamine- and catecholamine-containing neuron systems in the human fetal brain. 2. Anat. Entwick1.-Gesch. 139259-282. Palkovits, M., M. Brownstein, and J.M. Saavedra (1974) Serotonin content of the brain stem nuclei in the rat. Brain Res. 80:237-249. Parent, A,, L. Descarries, and A. Beaudet (1981) Organization of ascending serotonin systems in the adult rat brain. A radioautographic study after intraventricular administration of L3H1 5-hydroxytryptamine. Neuroscience 6:115-138. Parent, A,, D. Poitras, and L. Dube (1984) Comparative anatomy of central monoaminergic systems. I n A. Bjorklund and T. Hokfelt (eds): Handbook of Chemical Anatomy, Vol. 2: Classical Neurotransmitters in the CNS, Part 1.Amsterdam: Elsevier, pp. 409-439. Parizek, J., R. Hassler, and I.J. Bak (1971) Light and electron microscopic autoradiography of substantia nigra of rat after intraventricular administration of tritium labelled norepinephrine, dopamine, serotonin and the precursors. Z. Zellforsch. 115:137-148. Park, M.R., W.M. Falls, and S.T. Kitai (1982a) An intracellular HRP study of the rat globus pallidus. I. Responses and light microscopic analysis. J. Comp. Neurol. 211:284-294. Park, M.R., J.A. Gonzales-Vegas, and S.T. Kitai (1982b) Serotoninergic excitation from dorsal raphe stimulation recorded intracellularly from rat caudate-putamen. Brain Res. 243:49-58. Pasik, P., and T. Pasik (1982) Serotonergic afferents in the monkey neostriatum. Acta Biol. Acad. Sci. Hung. 33:277-288.

16 Pasik, P., T. Pasik, G.R. Holstein, and J.P. Saavedra (1984b) Serotoninergic innervation of the monkey basal ganglia: an immunocytochemical light and electron microscopy study. In: J.S. McKenzie, R.E. Kemn and L.N. Wilcock (eds): The Basal Ganglia: Structure and Function. New York: Plenum Press, pp. 115-129. Pasik, P., T. Pasik, J. Pecci-Saavedra, G.R. Holstein, and M.D. Yahr (1984a) Serotonin in pallidal neuronal circuits: An immunohistochemical study in monkeys. In R.G. Hassler and J.F. Christ (eds): Advances in Neurology, Vol. 40. New York Raven Press, pp. 63-76. Pazos, A,, R. Cortes, and J.M. Palacios (1985) Quantitative autoradiographic mapping of serotonin receptors in the rat brain. 11. Serotonin-2 receptors. Brain Res. 346231-249. Pazos, A,, and J.M. Palacios (1985) Quantitative autoradiographic mapping of serotonin receptors in the rat brain. I. Serotonin-1 receptors. Brain Res. 346:205-230. Pazos, A,, A. Probst, and J.M. Palacios (1987a) Serotonin receptors in the human brain, 111. Autoradiographic mapping of serotonin-1 receptors. Neuroscience 21;97-122. Pazos, A,, A. Probst, and J.M. Palacios (1987b) Serotonin receptors in the human brain. IV. Autoradiographic mapping of serotonin-2 receptors. Neuroscience 21:123-139. Poitras, D., and A. Parent (1978) Atlas of the distribution of monoaminecontaining nerve cell bodies in the brain stem of the cat. J. Comp. Neurol. I79:699-718. Ranadive, N.S., and A.H. Sehoii (1967) Antibodies to serotonin. Can. J. Biochem. 45:1701-1710. Reynolds, G.P., and S.J. Pearson (1987) Decreased glutamic acid and increased 5-hydroxytryptamine in Huntington’s disease brain. Neurosci. Lett. 78,233-238. Rinvik, E., I. Grofova, C. Hammond, J Feger, and J.M. Deniau (1979) A study of the afferent connections to the subthalamic nucleus in the monkey and the cat using the HRP technique. In L.J. Poirier, T.L. Sourkes and P.J. Bedard (eds): Advances in Neurology, Vol. 24. New York: Raven Press, pp. 53-70. Saavedra, J.M. (1977) Distribution of serotonin and synthesizingenzymes in discrete areas of the brain. Fed. Proc. 36.9134-2141. Schofield, S.P., and A.F. Dixon (1982) Distribution of catecholamine and indolamine neurons in the brain of the common marmoset (Callithrix jacchusj. J. Anat. 134:315-338. Schofield, S.P., and B.J. Everitt (1981) The organization of indolamine neurons in the brain of the rhesus monkey (Macaca nulatta). J. Comp. Neurol. 197:369-383. Shannak, K.S., and 0.Hornykiewicz (1980)Brain monoamines in the rhesus monkey during long-term neuroleptic administration. Adv. Biochem. Psychopharmacol. 24:315-323. Sladek, J.R. Jr, D.L. Garver, and J.P. Cummings (1982) Monoamine distribution in primate brain. IV. Indoleamine containing perikarya in the brain stem ofMacaca arctoides. Neuroscience 7:477493. Smith, Y., L.-N. Hazrati, and A. Parent (1990) Efferent projections of the suhthalamic nucleus in the squirrel monkey as studied by the PHA-L anterograde tracing method. J. Comp. Neurol. 294:306-323. Smith, Y., B. Lavoie, J. Dumas, and A. Parent (1989) Evidence for a separate nigropallidal dopaminergic projection in the squirrel monkey. Brain Res. 482381-386. Snyder, A.M., M.J. Zigmond, and R.D. Lund (1986) Sproutingof serotoninerpic afferents into striatum after dopamine-depleting lesions in infant rats. A retrograde transport and immunocytochemical study. J. Comp. Neurol. 245.274-281. Soghomonian, J.J., L. Descarries, and K.C. Watkins (1989) Serotonin innervation in adult rat neostriatum. 11. Ultrastructural features: A radioautographic and immunocytochemical study. Brain Res. 481:6786. Soghomonian, J.J.. G. Doucet. and L. Descarries (1987) Serotonin innervation in adult rat neostriatum. I. Quantified regional distribution. Brain Res. 42585-100.

B. LAVOIE AND A. PARENT Soubrie, P., T.D. Reisine, and J. Glowinski (1984) Functional aspects of serotonin transmission in the basal ganglia: A review and a ~n iizuo approach using the push-pull cannula technique. Neuroscience 13:605625. Stachowiak, M.K., J.P. Bruno, A.M. Snyder, E.M. Stricker, and M.J. Zigmond (1984) Apparent sprouting of striatal serotoninergic terminals after dopamine-depleting brain lesions in neonatal rats. Brain Res. 291: 164-167. Steinbusch, H.W.M. (1981j Distribution of serotonin-immunoreactivity in the central nervous system of the rat-cell bodies and terminals. Neuroscience 6:557-618. Steinbusch, H.W.M., R. Nieuwenhuys. A.A.J. Verhofstad, and D. Van der Kooy (1981) The nucleus raphe dorsalis of the rat and its projection upon the caudateputamen. A combined cytoarchitectonic, immunohistochemical and retrograde transport study. J. Physiol. Paris 77:157-174. Steinbusch, H.W.M., D. Van der Kooy, A.A.J. Verhofstad, and A. Pellepino (1980) Serotoninergic and non-serotoninergic projections from the nucleus raphe dorsalis to the caudate-putamen complex in the rat studied by a combined immunofluorescence and fluorescent retrograde axonal labeling technique. Neurosci. Lett. 19:137-142. Steinbusch, H.W.M.,A.A.J. Verhofstad, and H.W.M. Joosten (1978) Localization of serotonin in the central nervous system by immunohistochemistry: Description of a specific and sensitive technique and some applications. Neuroscience 3:811-819. Sternberger, L.A. (1986) Immunocytochemistry New York: John Wiley & Sons. Stuart, A.M., I.J. Mitchell, P. Slater, H.L.P. Unwin, and A.R. Crossman (1986) A semi-quantitative atlas of 5-hydroxytryptamine-1 receptors in the primate brain. Neuroscience 18:619-639. Takeuchi, Y., H. Kimura, T. Matsuura, and Y. Sano (1982) Immunohistochemical demonstration of the organization of serotonin neurnns in the brain of the monkey (Macaca fuscata). Acta Anat. 114:lOti-124. Van der Kooy, D., and T. Hattori (1980aj Dorsal raphe cells with collateral projections to the caudate-putamen and substantia nigra: A fluorescent retrograde double labeling study in the rat. Brain Res. 186:l-7. Van der Kooy, D., and T. Hattori (1980b) Bilaterally situated dorsal raphe cell bodies have only unilateral forebrain projections in rat. Brain Res. 192:550-554. Vandermaelen, C.P., A.C. Bonduki, and S.T. Kitai (1979) Excitation of caudate-putamen neurons following stimulation of the dorsal raphe nucleus in the rat. Brain Res. 175:356-361. Walsh, F.X., E.D. Bird, and J.T. Stevens (1982) Monoanline transmitters and their metabolites in the basal ganglia of Huntington’s disease and control postmortem brain. In A.J. Friedhoff and T.N. Chase (eds): Gilles de la Tourette Syndrome. Advances in Neurology, Vol. 35. New York: Raven Press, pp. 165-169. Waeber, C., K. Dixon, D. Hoyer, and J.M. Palacios (1988) Localisation by autoradiography of neuronal 5-HT3receptors in the mouse CNS. Europ. J. Pharmacol. 151:351-352. Wiklund, L., L. Leger, and M. Persson (1981)Monoamine cell distribution in the cat brain stem. A fluorescence histochemical study with quantification of indolaminergic and locus coeruleus cell groups. J. Comp. Neurol. 203:6 13-647. Wilson, M.A., G.A. Kicaurte, and M.E. Molliver (1989) Distinct morphologic classes of serotoninergic axons in primates exhibit differential vulnerability to the psychotropic drug 3-4-methylenedioxymethamphetamine. Neuroscience 28: 121-137. Wirtshafter, D., T.R. Stratford, andK.E. Anin (1987)Evidence that serotoninergic projections to the substantia nigra in the rat arise in the dorsal, but not the median raphe nucleus. Neurosci. Lett. 77%-266. Yelnik, J., G. Percheron, and C. Franqois (1984) A Golgi analysis of the primate globus pallidus. 11. Quantitative morphology and spatial orientation of dendritic arborization. J. Comp. Neurol. 227:200-213.