Pattern formation in the striatum: neurons with early projections to the substantia nigra survive the cell death period.

During the early postnatal period the striatum undergoes significant cell death. The specificity and regulation of this regressive event may be particularly interesting in the light of recent findings demonstrating that a developmentally organized compartmental architecture exists in the striatum. The striatum can be divided into two complementary and phenotypically distinct compartments, the patches and the matrix. In the adult, these two striatal compartments can be distinguished on the basis of their anatomy and a series of compartment-specific biochemical and hodological markers. We have previously demonstrated that the neurons within the patch and matrix compartments become postmitotic and make connections with the substantia nigra at distinct and sequential developmental times. The majority of patch neurons become postmitotic between embryonic days 12 and 15 and make a striatonigral connection prenatally. In contrast, a majority of matrix neurons become postmitotic between embryonic days 17 and 20 and do not form an efferent connection to the substantia nigra until the first postnatal week. Here we investigated whether either neuronal birthdate or time of making an efferent projection correlates with a neuron's probability of surviving the cell death period. We found that both the patch and matrix compartments undergo their entire cell death period by the end of the first postnatal week. During this period approximately 30% of striatal neurons are subject to cell death, regardless of striatal compartment. Neuronal counts within the striatal patch compartment suggest that both early born neurons (embryonic day 13) and early projecting neurons (to the substantia nigra) are preferentially spared. However, their considerable overlap (i.e., most early born neurons also have a nigral projection) prevents assessment of which feature is critical for survival. In contrast, there are small, but mostly separate, populations of early born and early projecting neurons within the matrix compartment. Quantitative analysis of these two distinct populations suggests that while early projection neurons within the matrix are spared, the early born matrix neurons lacking an early nigral projection undergo significant cell death. This proposal is further supported by the observation that the percentage of early born neurons in both the patch and matrix compartments that also have an early nigral projection increases from postnatal day 2 to 17. This finding suggests that among the early born striatal neurons in both compartments, those that do not project to the nigra selectively die during the cell death period. Together these results support the hypothesis that completion of an early projection to the substantia nigra gives neurons an advantage for surviving the cell death period.     

Differential maturation of cholinergic interneurons in the striatal patch versus matrix compartments

Striatal neurons are generated in two distinct phases. Neurons that become postmitotic early in embryonic development come to be located primarily in the patch compartment of the striatum, while the majority of the neurons situated in the striatal matrix compartment are generated later in embryogenesis. The cholinergic interneurons in the striatum, which have been reported to be more or less homogeneously distributed in the adult, are all generated early in development. Given that early generated neurons are expected to be situated primarily in the patch compartment, we investigated the apparently homogeneous distribution of cholinergic neurons by analysing their localizations in the patch and matrix compartments during striatal development. To selectively mark the striatal patch compartment we made injections of the retrograde fluorescent tracer True Blue in the substantia nigra on embryonic day 20 or postnatal day (P)1, and then stained for cholineacetyltransferase (ChAT) at different time-points in development. After P7, the distribution of the ChAT positive neurons changes from an earlier preference for the patch compartment to a preference for an area of the matrix just outside of the patches. Absolute counts show that this change in distribution is caused mainly by a late turn on of ChAT by the cholinergic neurons in the matrix compartment. These data suggest that there are different compartmental subpopulations of cholinergic neurons in the striatum. © 1996 Wiley-Liss, Inc.     

Pattern formation in the striatum: developmental changes in the distribution of striatonigral neurons

The striatum of the mammalian forebrain can be divided into 2 compartments, the patches and the matrix. We have investigated embryonic events involved in the formation of these compartments in rats. Early in development, dopamine fibers from the substantia nigra selectively innervate the patches. In the perinatal striatum, we observed a close match between the distributions of striatal cell bodies with axonal projections to the substantia nigra and patches of afferent dopamine fibers. Striatal cells projecting to the nigra are first seen in the ventrolateral striatum at embryonic day (E) 17. Striatonigral cell bodies are distributed homogeneously through the striatum from E18 to 19. At E20 and until postnatal day 4, these cell bodies are organized into discrete patches. After this time, striatonigral cell bodies assume the dense and homogeneous distribution characteristic of the adult striatum. A retrograde tracer injection in the nigra at E18 (during the early period of homogeneous striatonigral distribution) produces a patchy striatonigral distribution if the embryo is not sacrificed until E21. The number of retrogradely labeled striatonigral cell bodies in a midstriatal section, at times immediately before and after the early homogeneous to patchy changeover did not differ significantly. We suggest that the neurons of the patch compartment of the striatum are born first and project to the substantia nigra first. The patch neurons only become restricted to "patchy" areas as the later-born matrix neurons migrate out into the striatum.     

Neurogenesis and stereological morphometry of calretinin-immunoreactive GABAergic interneurons of the neostriatum

We determined the neurogenesis characteristics of a distinct subclass of rat striatum gamma-aminobutyric acidergic (GABAergic) interneurons expressing the calcium-binding protein calretinin (CR). Timed-pregnant rats were given an intraperitoneal injection of 5-bromo-2'-deoxyuridine (BrdU), a marker of cell proliferation, on designated days between embryonic day 12 (E12) and E21. CR-immunoreactive (-IR) neurons and BrdU-positive nuclei were labeled in the adult neostriatum by double immunohistochemistry, and the proportion of double-labeled cells was quantified. CR-IR interneurons of the neostriatum show maximum birth rates (>10% double labeling) between E14 and E17, with a peak at E15. CR-IR interneurons occupying the lateral half of the neostriatum become postmitotic prior to medial neurons. In the precomissural neostriatum, the earliest-born neurons occupy the lateral quadrants and the latest-born neurons occupy the dorsomedial sector. No significant rostrocaudal neurogenesis gradient is observed. CR-IR neurons make up 0.5% of the striatal population and are localized in both the patch and the matrix compartments. CR-IR neurons of the patch compartment are born early (E13-15), with later-born neurons (E16-18) populating mainly the matrix compartment. CR-IR cells of the neostriatum are a distinct subclass of interneurons that are born at an intermediate time during striatal development and share common neurogenesis characteristics with other interneurons and projection neurons produced in the ventral telencephalon.     

Compartmentalization of the embryonic striatum after intraocular transplantation

In order to assess the intrinsic potential of the isolated embryonic striatum to develop its adult patch and matrix compartments, embryonic day 16 striata were transplanted into the anterior eye chamber of adult host rats. After 2-12 weeks of survival the transplants showed heterogeneous and in the majority of cases complementary distributions of opiate receptor binding and acetylcholinesterase staining, which mark the patch and matrix compartments of the adult striatum, respectively. The complementarity of patch and matrix markers in the transplants shows that the transplants do compartmentalize. However, the density of the markers in the transplants did not reach the levels seen in the adult striatum. The results suggest that the commitment of cells to a striatal compartment is a very early event in the embryonic development of the forebrain.     

Pattern formation in the striatum: developmental changes in the distribution of striatonigral projections

The mammalian striatum (the major subcortical structure in the telencephalon) can be divided into two compartments, the patch and the matrix, on the basis of various neurochemical and hodological markers expressed in the adult. The primary efferent target of striatal neurons is the substantia nigra. We have previously shown that the patch compartment sends projections to the substantia nigra embryonically; whereas the matrix does not form a similar projection until the early postnatal period (Fishell and van der Kooy, J. Neurosci., 7 (1987) 1969-1978). The projection of patch neurons to the substantia nigra is the earliest developmental marker for the patch compartment. Here we ask about the early distribution of patch projections and their possible relation to striatal compartmentalization. Embryonic anterograde axonal tracing of the striatonigral pathway can take advantage of the temporal separation of patch versus matrix projections to reveal the terminal distribution of patch striatonigral neurons independent of the nigral terminal distribution from the striatal matrix. The anterograde tracer rhodamine isothiocyanate was shown in a model system to persist in labeled neurons for more than a week, but to be available for uptake into these neurons for a few days after injection at the most. These properties of rhodamine isothiocyanate were combined experimentally with short and long term survival periods. This allowed assessment of the changing developmental distribution of nigral fibers from specifically the striatal patch compartment. In all experimental cases the anterogradely labeled sections of the substantia nigra were also stained with antibodies to tyrosine hydroxylase, which permitted differentiation of the dopamine cell rich pars compacta from the dopamine cell poor pars reticulata. The results show that in the adult the majority of patch and matrix striatonigral projections are confined to the substantia nigra pars reticulata. Furthermore, their fiber distribution within the pars reticulata is overlapping rather than complementary. Most interestingly, in the late embryonic period (most noticeably at embryonic day 19) there is a marked overlap between patch striatonigral fibers and nigral dopamine perikarya. By early postnatal times, when the matrix compartment begins to form its striatonigral projection, the overlap of patch striatonigral fibers and dopamine cells has largely disappeared. The results suggest that a transient interaction between patch striatonigral fibers and dopamine neurons (which is concomitant with the formation of striatal compartments), may be an important developmental event in the phenotypic maturation of striatal pa     

Neuronal birthdate underlies the development of striatal compartments

The striatum of the mammalian forebrain is composed of two complementary functional compartments, the patches and the matrix. By injecting [3H]thymidine at different embryonic times and sacrificing the rats as young adults, we found that the earliest neurons to leave the mitotic cycle were restricted to the patch compartment. Neurons that became postmitotic at later times preferentially joined the matrix compartment. Distinctive periods of cell proliferation may underlie pattern formation throughout the developing forebrain.     

Pattern formation in the mammalian forebrain: patch neurons from the rat striatum selectively reassociate in vitro

Mechanisms involved in the developmental organization of the rat striatum were investigated in vitro. The neurons of the patch and matrix compartments were preferentially labeled in vivo with a [3H]thymidine injection on embryonic day (E) 13 or 18, respectively. Two or 7 days later the striatum was removed, dissociated into a single cell suspension and plated on a collagen-coated substrate. After 5 days in culture the neurons had migrated into aggregates. Within an individual aggregate, neurons labeled on E13 tended to clump together, whereas neurons labeled on E18 were randomly dispersed. Comparing between aggregates, [3H]thymidine-labeled E13 cells were located in aggregates containing numerous other labeled E13 cells, whereas [3H]thymidine-labeled E18 cells were dispersed randomly between aggregates. These results suggest that early born striatal neurons (primarily patch cells) selectively associate with each other, and that this process may be crucial to the developmental compartmentalization of the rat striatum.     

NGF facilitates the developmental maturation of the previously committed cholinergic interneurons in the striatal matrix.

Although all of the cholinergic interneurons of the striatum are generated early in development, the maturation of these neurons depends on striatal compartmental localization. The majority of the cholinergic neurons in the patches turn on choline acetyltransferase (CHAT) embryonically, whereas the majority of cholinergic neurons in the matrix turn on CHAT postnatally. To determine whether CHAT expression can be induced earlier in the cholinergic neurons and whether the facilitation is compartment specific, we infused nerve growth factor (NGF) into the lateral ventricle of either embryonic day 19 embryos or postnatal day 1 pups. We simultaneously marked the patch compartment by injecting the retrograde fluorescent tracer True Blue into the substantia nigra at the times of the NGF infusions. After a 2-day survival time, NGF induced a dramatic increase in the number of CHAT-immunoreactive neurons in the matrix compartment (up to adult levels), whereas the NGF infusions did not increase the number of CHAT neurons in the patch compartment. Analyses of the compartmental distributions of the p75 and trkA NGF receptors themselves do not provide an explanation for the differential cholinergic maturation in the compartments of the control striatum or for the upregulation of CHAT in the striatal matrix after the NGF infusion. We conclude that NGF infusion is capable of facilitating the normally slow cholinergic maturation of the cholinergic neurons in the matrix, whereas the cholinergic maturation of the CHAT cells in the patch compartment seems to be largely independent of NGF signalling.     

The development of a patchy organization of the rat striatum

The rat striatum can be divided into patch and matrix compartments. Patches, as marked by high opiate receptor binding, first emerge perinatally from a dense, diffuse field of striatal opiate binding. Our quantitative analysis revealed that the patch compartment formed its peak proportion of the total striatal area at postnatal day 7. After this time, patches occupied a smaller proportion of the striatum, reflecting the fact that the number of patches and mean area per patch reached near adult levels during the first postnatal week, yet the volume of the striatum as a whole continued to increase for several weeks postnatally. Results from transplant and early postnatal lesion experiments suggested that connections between the striatum and other brain areas are important for the formation and/or maintenance of the patch and matrix compartments. Transplants of embryonic striatum to cavities in the cortex of young adult hosts developed diffuse opiate receptor binding but not dopamine receptor binding. Significantly, the opiate receptor binding seen in the transplants was never organized into the dense patches normally seen in the adult striatum. In a few transplants areas of relatively higher opiate receptor binding occurred in areas of relatively low neuronal cell density, as is seen in early normal development, but the dense adult patches never developed. Coronal diencephalic hemisections, but not decortications, in the early postnatal period produced drastic shrinkage of the striatum and, more importantly, a large decrease in opiate receptor patches when expressed as a proportion of total striatal area. Neuronal connections with more caudal brain structures may play a role in the final differentiation and maintenance of the striatal compartments.     

The neostriatal mosaic. I. Compartmental organization of projections from the striatum to the substantia nigra in the rat

Combined neuroanatomical techniques were used to examine the organization of the striatal projection to the substantia nigra in the rat. Both double anterograde axonal tracing methods (Phaseolus vulgaris leuco-agglutinin (PHA-L) and 3H-amino acid tract tracing) and double fluorescent retrograde axonal transport tracing methods were used to examine the relationship among striatal neurons projecting to separate areas of the substantia nigra. Additionally, the distributions of retrogradely labeled striatonigral projection neurons were charted relative to the neurochemically distinct striatal "patch" compartment, identified by substance P- or leu-enkephalin-like immunoreactivity, and the complementary "matrix" compartment, identified by somatostatin-like immunoreactive fibers. These studies show two distinct types of organization in the striatonigral projections. One type is topographic in that the mediolateral relationships among these striatal efferent neurons are roughly maintained by their termination patterns in the substantia nigra, while the dorsoventral relationships are inverted. Projections from any part of the striatum, however, are distributed throughout the rostrocaudal axis of the substantia nigra. Despite their general topographic organization, the variable and dispersed nature of such projections from individual striatal loci results in partial overlap of afferent fields from separate striatal areas. The second type of organization is nontopographic and provides a different system for convergence of inputs from separated striatal areas that is superimposed on the rough topographic system. In this other projection system the mediolateral and dorsoventral relationships typical of the topographically ordered system are not maintained and are sometimes reversed. For example, PHA-L injected into the dorsal striatum labels a topographic (inverted relationship) projection to the ventral substantia nigra pars reticulata but also a smaller and separate projection to the dorsal pars reticulata and adjacent pars compacta. Retrograde tracer deposits in the pars compacta label neurons in the ventral striatum (the inverted relationship) but also clusters of neurons in the dorsal striatum. These clusters are in the neurochemically defined patch compartment whereas neurons in the matrix are labeled by injections into the pars reticulata. The dendrites of both retrogradely filled patch and matrix neurons are confined to the compartment containing their cell bodies, suggesting a restriction that would functionally segregate extrinsic striatal afferents shown in other studies to be confined to either patches or matrix.(ABSTRACT TRUNCATED AT 400 WORDS)     

Embryonic lesions of the substantia nigra prevent the patchy expression of opiate receptors, but not the segregation of patch and matrix compartment neurons, in the developing rat striatum.

Unilateral lesions of the substantia nigra on embryonic day 19 prevent the development of the normal patchy distribution of opiate receptors in the ipsilateral rat striatum. Independent, early and permanent labelling of patch compartment neurons in the same brains on embryonic day 14 with [3H]thymidine revealed that the substantia nigra lesions did not prevent the aggregation of early born neurons into patches, but rather blocked the normal expression of one phenotype (dense opiate receptor binding) of these patches. Thus, early nigrostriatal connections may not be critical for the fundamental patch/matrix compartmentation of the striatum, but may be important in the maturation of phenotypic markers of these compartments.     

The pallidostriatal projection in the rat: a recurrent inhibitory loop?

The pallidostriatal projection in the rat was investigated employing the PHA-L tracing technique. Following iontophoretic injections into the lateral aspect of the globus pallidus external segment, the ipsilateral striatum showed patches of dense anterograde labeling separated by areas containing sparse anterograde labeling and isolated retrogradely labeled neurons. The densely labeled patches did not correspond to any known compartments of the striatum. The retrogradely labeled neurons consistenly showed similar distribution of morphological features reminiscent of striatal type II projection neurons. As all projection neurons of the striatum and all pallidal neurons are GABAergic, the complementary pattern of anterogradely and retrogradely labeled profiles from the globus pallidus suggest a possible mechanism whereby a horizontal inhibition may be exerted on groups of striatal neurons via the striato-pallido-striatal pathway.     

Striatal neurons in striatal grafts are derived from both post-mitotic cells and dividing progenitors

Transplants of embryonic striatal tissue are characteristically heterogeneous, containing patches (P-zones) of striatal medium spiny projection neurons. It is not yet known how this morphology develops, and whether the striatal neurons in the grafts are derived from post-mitotic neuroblasts in the embryonic brain or from striatal progenitors that continue to divide after transplantation. To address this question we labelled dividing cells in the transplants with bromodeoxyuridine (BrdU), either prior to or after transplantation into the adult lesioned rat striatum. Cells for transplantation were either pre-labelled in utero by intraperitoneal (i.p.) injections of BrdU, or post-labelled after transplantation by i.p. injections to the hosts. Either two or six months after transplantation the brains were processed using double immunohistochemical techniques to detect BrdU and calbindin-positive neurons in the transplants. In the transplants pre-labelled with BrdU, approximately 30% of calbindin-positive cells were heavily labelled with BrdU, suggesting these had undergone a final division prior to transplantation. In transplants where cells had been labelled post-transplantation, approximately 17% of calbindin cells were heavily BrdU labelled. These results suggest that whereas a proportion of striatal medium spiny neurons in the striatal grafts were post-mitotic at the time of transplantation, other striatal progenitor cells can continue to divide after transplantation, and then complete an appropriate neuronal maturation programme in the adult host brain environment.     

Modular organization of projection neurons in the matrix compartment of the primate striatum.

It is well known that the striatum has a chemical architecture dividing it into striosomes and matrix, and that these compartments have different input-output connections. However, striatal afferent-fiber systems also form vividly patchy terminal fields in the matrix, and studies in the past year have uncovered instances of nonstriosomal clustering of striatal output neurons. In the experiments reported here, we systematically investigated this output-neuron clustering in the primate, using the striatopallidal system as a model. Our goals were to determine whether the modular organization is a general characteristic of projection neurons in the striatal matrix, whether the modularity occurs independent of striosomal boundaries, and whether the output modules are systematically organized. We studied the distribution of striatopallidal projection neurons in 9 adult squirrel monkeys by centering deposits of the retrograde tracer HRP-WGA in either the external segment or the internal segment of the globus pallidus. Following injections of each type, many retrogradely labeled neurons appeared in the striatal matrix in clusters and bands having cross-sectional diameters of 0.2-0.8 mm. Comparisons with adjoining sections stained to demonstrate striosomes established that the local groups of striatal output neurons sometimes abutted striosomes but often did not. The retrogradely labeled clusters and bands appeared both in the caudate nucleus and in the putamen. Their arrangements were regular and often periodic. These findings suggest that the large matrix compartment of the primate striatum, which is the primary site of origin of striatal outputs to the pallidum and the reticular part of the substantia nigra, contains systematic mosaics of projection neurons. We propose that this output-neuron modularity of the striatal matrix in the primate could serve as the template for redistribution of the massive afferent-fiber systems of the striatum into specialized striatopallidal output channels.     

Development of striatal compartmentalization following pre- or postnatal dopamine depletion.

Nigrostriatal dopamine (DA) projections terminate in distinct patches during the late prenatal and early postnatal period in the rat. During the first postnatal week, patches of DA fibers overlap with clusters of striatal neurons that share several identified characteristics. The early segregation of striatal cell types into either these patches or the surrounding matrix becomes a permanent organizational feature of the striatum. In order to determine whether the heterogeneous distribution of DA influences the formation of cellular patches, the developmental organization of chemically identifiable cell types was examined in normal rats and in rats DA depleted as infants (0 or 3 d) or in utero (embryonic days 17-18). During the first postnatal week, corresponding patches of DA afferents and substance P (SP)-immunoreactive neurons existed in the striatum of normal animals, and AChE-positive zones overlapped these patches in the lateral striatum. Injection of 6-hydroxydopamine into the lateral ventricles of fetal or infant rats produced a dramatic loss of striatal DA terminals. Neither the patchy distribution of SP-immunoreactive neurons nor the distinctive pattern of AChE staining present during the first 2 postnatal weeks was disrupted. During the third postnatal week, cells immunoreactive for leu-enkephalin or calbindin-D28k were confined to the matrix compartment, and this compartmentalization was also not noticeably changed by pre- or postnatal DA depletion. In adult animals, overlapping patches of leu-enkephalin- and SP-immunoreactive fibers were observed, regardless of whether any DA terminals remained. Thus, the basic organization of the striatal patch and matrix compartments develops normally in the absence of DA innervation through much of the formative period. Although these observations do not completely dismiss the possibility that the first DA afferents to appear in the striatal primordia influence contracted striatal cells to develop the patch phenotype, they suggest that the patchy distribution of DA afferents may be secondary to the early clustering of striatal neurons forming the patch compartment.     

Spatial distributions of chemically identified intrinsic neurons in relation to patch and matrix compartments of rat neostriatum

The spatial distributions and dendritic branching patterns of chemically identified subpopulations of striatal intrinsic neurons, defined by immunoreactivity for choline acetyltransferase (ChAT), neuropeptide Y or parvalbumin, were studied in relation to patch and matrix compartments of rat neostriatum. ChAT-immunoreactive cells and fibers showed an uneven pattern of distribution in the striatum. ChAT immunoreactivity was higher in the dorsolateral part and lower in the ventromedial part of the striatum. This regional gradient pattern is the inverse of the overall pattern of calbindin D_28k immunoreactivity. However, in small regions close to the lateral ventricle and globus pallidus, areas containing fewer ChAT-immunoreactive cells and fibers coincided with those containing low calbindin D_28k immunoreactivity. Neuropeptide Y immunoreactivity was uniform in the neostriatum. Certain neuropeptide Y cells (about 20%) were also immunoreactive for calbindin D_28k, indicating that at least a small population of calbindin D_28k-immunoreactive cells are medium aspiny cells. Parvalbumin immunoreactivity was not uniform in the striatum. A higher density of parvalbumin immunoreactivity was found in the neuropil in lateral and caudal parts than in the medial part. Small regions with weaker parvalbumin-immunoreactive neuropil partially corresponded to calbindin D_28k poor patches. Larger cells immunoreactive for parvalbumin were preferentially located in lateral and caudal parts of the striatum. Cells immunoreactive for ChAT, neuropeptide Y or parvalbumin showed basically similar distribution patterns in relation to the patch and matrix compartments. Most stained cells were located in the matrix, but some were located at the borders of patches and a few were inside patches. Most primary dendrites of stained cells in the matrix or patches remained confined to these compartments, but cells on the borders invariably extended dendrites into both compartments. The striatal intrinsic neurons form chemically differentiated neuronal circuits within the matrix, and the patches and those whose dendrites cross the borders may contribute to associational interconnections between the two compartments, unlike the spiny projection neurons whose dendrites are confined to one or the other compartment. © Wiley-Liss, Inc.     

Transient compartmental expression of a family of protein tyrosine phosphatases in the developing striatum

The expression of a family of intracellular protein tyrosine phosphatases (STEP) was studied in the striatum of rats during ontogeny. Links between the formation of dopamine islands and STEP immunoreactive patches in the striatum were examined since previous work had suggested that STEP isoforms were selectively expressed in dopaminoceptive brain regions. STEP protein and mRNAs were distributed in a patchy manner during the first postnatal week. By 2 weeks, STEP immunoreactivity was homogeneous, indicating that both patch and matrix neurons express STEP by maturity. Two-color immunofluorescent staining was also performed to compare STEP with specific markers for patch and matrix. Tyrosine hydroxylase immunoreactive fibers from the substantia nigra form distinctive dopamine islands in the striatum during late embryonic development, and occupy the sites of future patches [23,37,38,54]. These fiber islands align with STEP immunoreactive neuronal patches during the first two postnatal weeks, suggesting that STEP is a marker for patch neurons in early postnatal development. When STEP's distribution was compared with other markers for patch (substance P) or matrix (calbindin), STEP co-localized with substance P in most striatal neurons on postnatal days 1 through 7. However, STEP was also expressed within a subset of calbindin-positive neurons in the lateral striatum, but not with these neurons elsewhere in the striatum. By adulthood, STEP colocalized with both markers. These results suggest that STEP is expressed first within patch neurons but not matrix, and subsequently within both. The expression of STEP may be triggered by the arrival of striatal afferents or other regulatory factors.     

Convergent Synaptic Input From the Neostriatum and the Subthalamus Onto Identified Nigrothalamic Neurons in the Rat

The two major afferents of the substantia nigra pars reticulata are the subthalamic nucleus and the striatum. Stimulation of these afferents has opposing physiological effects on the output neurons of the substantia nigra pars reticulata. In order to better understand the role of these afferents in the flow of information through the basal ganglia and to better understand the ways in which they might interact, experiments have been performed to test the possibility that single-output neurons of the substantia nigra pars reticulata receive convergent synaptic input from the subthalamic nucleus and the neostriatum. To address this, rats received iontophoretic deposits of the anterograde tracer Phaseolus vulgaris leucoagglutinin in the subthalamic nucleus, injections of the anterograde tracer biocytin in the neostriatum and injections of the retrograde tracer horseradish peroxidase conjugated to wheat-germ agglutinin in the ventral medial nucleus of the thalamus. Following appropriate survival times the animals were perfusion-fixed and sections of the substantia nigra were processed to reveal the transported tracers and prepared for electron microscopy. Light microscopic examination revealed that the substantia nigra contained rich plexuses of anterogradely labelled subthalamic and striatal terminals, as well as many retrogradely labelled nigrothalamic neurons. The anterogradely labelled terminals were often seen apposed to the retrogradely labelled neurons. In the electron microscope the subthalamic terminals were seen to form asymmetrical synaptic contacts (subthalamic type 1) with the identified nigrothalamic neurons as well as unlabelled perikarya and both proximal and distal dendrites. In confirmation of previous findings, the striatal terminals made symmetrical synaptic contact with the nigrothalamic neurons as well as unlabelled neurons. In areas of overlap between the two classes of terminals, identified nigrothalamic neurons and unlabelled nigral neurons were found to receive convergent synaptic input from the subthalamic nucleus and the neostriatum. In addition to the anterogradely labelled subthalamic terminals that formed asymmetrical synaptic specializations, a second, much rarer class was also observed (subthalamic type 2). These terminals were much larger and formed symmetrical synapses; several lines of evidence suggest that they originated not in the subthalamic nucleus but in the globus pallidus. These terminals were found to make synaptic contacts with identified nigrothalamic neurons and non-labelled neurons and to form convergent synaptic contacts with subthalamic type 1 terminals and striatal terminals. It is concluded that the topographical and synaptic organization of the so-called direct (striatum to substantia nigra pars reticulata) and indirect pathways (i.e. pathways involving the subthalamic nucleus andlor the globus pallidus) of information flow through the basal ganglia underlies the inhibition and excitation of the output neurons of the substantia nigra pars reticulata that occur following stimulation of the striatum.