GABAergic input to cholinergic forebrain neurons: an ultrastructural study using retrograde tracing of HRP and double immunolabeling

Cat 'P300' and cholinergic septohippocampal neurons: depth recordings, lesions, and choline acetyltransferase immunohistochemistry

The role of septohippocampal circuits in the generation of the P300 response in cats (n = 12) was explored in a series of depth recording, tract-tracing and lesion experiments. Systematic mapping of the hippocampus in 1-mm increments from rostral to caudal extent revealed large positive potentials, greater in amplitude to rare than to frequent stimuli, within the 200-500 ms range. Each map revealed maximal amplitude responses at diverse, widely distributed hippocampus loci. Furthermore, these electrical responses displayed polarity inversion within the hippocampus that was generally localized to the pyramidal cell layer; polarity inversion was also observed in the adjacent entorhinal cortex and amygdala. Injections of propidium iodide, a tract-tracing agent, into these inversion sites resulted in retrograde labeling of small clusters of choline acetyltransferase (ChAT)-positive neurons in the medial septal nucleus and vertical limb of the diagonal band. Aspiration lesions that bilaterally destroyed large amounts of caudal hippocampus from stereotaxic levels A4 to A1 resulted in a decreased number of cells expressing ChAT in the rostral basal nuclear complex. In only 2 cats was the preoperative presence of a significant vertex P300 absent postoperatively. In the majority of cases (5 of 8 animals), hippocampal aspiration produced an enhancement of the preoperative P300 potential. We conclude that cholinergic mechanisms are importantly, albeit not exclusively, involved in the mediation of P300 potentials in cats. Neurons mediating P300 responses appear to be organized in diverse clusters of septal and diagonal band cells. These septal cells may facilitate, and in turn be facilitated or inhibited as a function of hippocampal, or other, allocortical feedback loops.

Projections from the nucleus accumbens to cholinergic neurons of the ventral pallidum: a correlated light and electron microscopic double-immunolabeling study in rat

A correlated light- and electron microscopic double-immunolabeling study combining choline acetyltransferase immunocytochemistry with anterograde tracing of Phaseolus vulgaris leucoagglutinin (PHA-L) revealed that axons of the nucleus accumbens terminate on cholinergic neurons of the ventral pallidum. These findings are discussed with respect to the possibility that these cholinergic neurons may be part of parallel circuits, providing feedback to the same cortical and amygdaloid areas which innervate the nucleus accumbens.

The development of neurons in the nuclei of the horizontal and vertical limb of the diagonal band of Broca of the rat: a qualitative and quantitative analysis of Golgi preparations

We have studied the morphological alterations of neurons in the nuclei of the horizontal (NHL) and vertical (NVL) limbs of the diagonal band of Broca of rats from late embryonic life to maturity using the Golgi-Stensaas and Golgi-Cox methods. During late embryonic life and in the first postnatal days, the two nuclei of the diagonal band of Broca were found to be located near the ventral surface of the brain. Shortly thereafter, neurons in the NHL and NVL gradually take up the positions which they normally occupy in adulthood. At this stage neurons were small with round or elongated somata and 1-3 primary dendrites that only occasionally bore spines and very seldom showed varicosities, features commonly shown by neurons at later postnatal ages. At birth, cells showing varying soma shapes and dendritic morphology were present, and by postnatal day 4 (P4) the three forms of neurons previously described in adult rats (Dinopoulos et al., J. Comp. Neurol., 272 (1988) 461-474) were readily distinguished. During the second postnatal week, the size of cell somata as well as the number, size and extent of dendritic branching underwent considerable increases in both nuclei and at P14 neurons showed features typical of their adult counterparts. In addition they showed a dramatic increase in the number of spines which was followed during the next 10 days by a substantial decrease. Overall, the dendritic geometry of neurons in the NHL and NVL did not change significantly after P14, although their cell bodies continued to increase in size until the middle of the fourth and fifth postnatal weeks respectively. These findings suggest that neurons in the nuclei of the diagonal band of Broca show continuous growth from embryonic life to the end of the second postnatal week when they acquire morphological features comparable to the adult. Thereafter they exhibit only minor morphological alterations with the exception of extensive spine elimination which is pronounced during the third postnatal week and continues until adulthood.

Disruptive effects of muscimol infused into the basal forebrain on conditional discrimination and visual attention: differential interactions with cholinergic mechanisms

The behavioral effects of GABAergic manipulation of the basal forebrain were investigated using two behavioral tasks, which previous studies have shown to yield dissociable effects following quisqualate-induced lesions of the basal forebrain: a five-choice serial reaction time task, involving approaching the location of a brief visual stimulus that is associated with reward; and a conditional visual discrimination task, requiring retrieval of information about a discriminative stimulus that stays constant over time. Following acquisition of the tasks, chronic guide cannulae were stereotaxically implanted into the basal forebrain. Those animals trained on the conditional visual discrimination task showed a dose-dependent reduction in choice accuracy and a lengthening of latency to respond correctly to the visual stimulus following administration of the GABA-A agonist, muscimol (1, 2, 3 ng/microliters/hem). While certain of these deficits, for example response latency, could be restored to control levels by co-administration of the GABA-A antagonist, bicuculline, none of the behavioural impairments could be significantly attenuated by systemic by systemic co-administration of the cholinesterase inhibitor, physostigmine (0.05, 0.1, 0.2 mg/kg, IP). Similarly, a dose dependent effect of muscimol (1, 1.5, 2 ng/microliters/hem) on choice accuracy and correct response latency was observed on performance of the five-choice attentional task. However, in contrast to the conditional task, significant attenuation of the impairment in choice accuracy was obtained following administration of physostigmine (0.05 and 0.1 mg/kg). Attenuation of muscimol-induced deficits by administration of bicuculline was also observed. It is therefore evident that although manipulation of GABAergic activity in the region of the basal forebrain produces profound deficits in different tasks of cognitive function, only some of these may be due to modulation of the magnocellular cholinergic projection to the neocortex.

Cholinergic neurons in the rat septal complex: ultrastructural characterization and synaptic relations with catecholaminergic terminals

Physiological and pharmacological studies have suggested that catecholamines modulate cholinergic neurons in the medial septal and diagonal band nuclei (i.e., the septal complex). Thus, the ultrastructural morphology of neurons containing choline acetyltransferase (ChAT), the biosynthetic enzyme for acetylcholine, and their relation to catecholaminergic terminals exhibiting immunoreactivity for the catecholamine synthesizing enzyme tyrosine hydroxylase (TH) were examined in the rat septal complex. Dual immunoautoradiographic and peroxidase anti-peroxidase labeling methods were used to simultaneously localize antibodies raised in rabbits against TH and from rat-mouse hybridomas against ChAT in single sections. At least two types of perikarya with ChAT-immunoreactivity (ChAT-I) were observed. The first type were large (20-30 microns), elongated or round, and contained a small indented nucleus with an abundant cytoplasm and an occasional lamellar body. The second type was also either ovoid or round but was medium-sized (15-20 microns) and contained a larger indented nucleus and a smaller amount of cytoplasm than the first type. Both types of perikarya as well as dendrites with ChAT-I were surrounded by astrocytic processes apposed to most of their plasmalemmal surfaces. The distribution and types of terminal associations (i.e., asymmetric synapses, symmetric synapses and appositions which lacked a membrane specialization in the plane of section analyzed) with ChAT-labeled perikarya and dendrites were quantitatively evaluated. The majority (68% of 197) of the presynaptic terminals were unlabeled; the remaining terminals were immunoreactive for TH (25%) or ChAT (7%). All three types of terminals contacted primarily the shafts of small dendrites and more rarely ChAT-labeled perikarya and large dendrites. ChAT-labeled terminals: (1) formed associations with unlabeled perikarya and dendrites (31% of 176); (2) formed associations with perikarya and dendrites with ChAT-I (7%); (3) contacted the same unlabeled perikarya and dendrite as a TH-containing terminal (21%); (4) were in apposition to TH-labeled terminals (25%); or (5) were either in apposition to unlabeled or ChAT-labeled terminals or lacked associations with any processes. The majority of associations formed by the terminals with ChAT-I were on the shafts of small dendrites. Moreover, most of the associations formed were either symmetric synapses or appositions not separated by astrocytes in the plane of section analyzed. These findings provide cellular substrates in the septal complex (1) for sparse synaptic input relative to astrocytic investment of cholinergic neurons and (2) for direct synaptic modulation of cholinergic and non-cholinergic neurons by catecholamines and/or acetylcholine. These findings have direct relevance to catecholaminergic-cholinergic interactions and to the neuropathological basis for Alzheimer's disease.

Ultrastructural localization of tyrosine hydroxylase immunoreactivity in the rat diagonal band of Broca

The present study sought to establish the cellular basis for the catecholaminergic (i.e., noradrenaline and dopamine) modulation of neurons in the horizontal limb of the diagonal band of Broca (HDB) in the rat brain. The light and electron microscopic localization of antigenic sites for a polyclonal antibody directed against the catecholamine synthesizing enzyme, tyrosine hydroxylase (TH), were examined in the HDB using a double-bridged, peroxidase-antiperoxidase method. By light microscopy, numerous punctate, varicose processes with intense TH-immunoreactivity (TH-I) were detected in the HDB. Additionally, a few small, bipolar, or multipolar TH-immunoreactive neurons were observed. Ultrastructural analysis of single sections revealed that the TH-labeled processes were axons and axon terminals. Axons (n = 134) with TH-I were primarily unmyelinated. Terminals with TH-I (n = 169) were 0.3-1.4 microns in diameter and contained many small, clear vesicles and 0-5 larger dense-core vesicles. The types of associations (i.e., asymmetric synapses, symmetric synapses, and appositions which lacked a membrane specialization in the plane of section analyzed) formed by the TH-labeled terminals were quantitatively evaluated. The TH-labeled terminals: (1) formed associations with unlabeled perikarya and dendrites (134 out of 169), (2) were closely apposed without glial intervention to unlabeled and TH-labeled terminals (11 out of 169), or (3) had no neuronal associations in the plane of section analyzed (24 out of 169). The relatively rare (n = 4) associations with unlabeled perikarya were mostly characterized by symmetric synaptic specializations. The majority of the TH-labeled terminals were associated with the shafts of small dendrites (66% of 134). Moreover, most of the associations on dendrites and dendritic spines were further characterized by asymmetric synaptic specializations; however, many were also appositions without any apparent glial intervention in the plane of section analyzed. Additionally, the TH-labeled terminals were often associated with only one dendrite, which, in the same plane of section, was sparsely innervated by other terminals. Astrocytic processes usually surrounded the portions of the terminals and dendrites not involved in the region of association. The TH-immunoreactive perikarya were small (7-12 microns), ovoid, and had an indented nucleus with some heterochromatin. Their scant cytoplasm contained mitochondria, Golgi complexes, and endoplasmic reticulum. A few immunoreactive dendrites, presumably derived from the local neurons, were also detected. Both TH-immunoreactive perikarya and dendrites were associated primarily with unlabeled terminals, although a few terminals with TH-I also contacted them.(ABSTRACT TRUNCATED AT 400 WORDS)

Modulation of cortical in vivo acetylcholine release by the basal nuclear complex: role of the pontomesencephalic tegmental area

Acetylcholine (ACh) release in vivo from rat cortices was determined by microdialysis either after injection of drugs into the basal nuclear complex (NBM) or after electrolytic lesion of the pontomesencephalic tegmental nucleus (PPT). Scopolamine (SCOP) (5-10 micrograms) increased and oxotremorine (10 micrograms) reduced cortical ACh release, indicating that an inhibitory mechanism operates within the area. The gamma-aminobutyric acid (GABA)ergic antagonist, picrotoxin (2.5 micrograms), by disinhibiting the cholinergic basocortical neurons, induced an increase that was not affected by SCOP. Acute lesion of the cholinergic PPT efferents to NBM raised cortical basal release. Thus, ACh released from the PPT terminals apparently modulates the function of basocortical neurons mainly through a polysynaptic link via GABAergic neurons.

Chronic administration of flumazenil (Ro 15-1788) enhances non-appetitive exploratory behavior of rats

The effects of chronic administration of the benzodiazepine receptor antagonist, flumazenil (Ro 15-1788; 4 mg/kg/day for 14 days in drinking water) on the performance of adult rats in the 12-arm radial maze were studied. Relative to controls, the animals treated with flumazenil showed an increase (P less than 0.002) in non-appetitively motived exploratory behavior, so called because it occurred in 88% of instances in non-baited alleys, facing the well-illuminated "enriched environment" of the center of the room, as opposed to the baited alleys, facing the "dull" corner of the room. This behavior emerged between day 5 and 7 of treatment with the drug, it continued to increase over the period of treatment with drug (P less than 0.002), and reached its peak at day 3, after withdrawal of the drug (P less than 0.008; a longer duration was not investigated). The occurrence of non-appetitively motivated exploratory behavior was inversely correlated with the scores for urination/defecation (P less than 0.003) and, therefore, most likely reflected the anxiolytic action of flumazenil. During treatment with drug or vehicle, the control and the drug groups made comparable numbers of "working memory" errors (P = 0.17). However, upon withdrawal of drug and introduction of alley gates (to confine the animal for 10 sec to the center platform, after an alley was explored), the working memory errors of the rats exposed to the drug, remained unchanged (P = 0.35), relative to the preceding three trials, while the performance of the control group was disrupted, as shown by an increase in the numbers of errors (P less than 0.004). At day seven of treatment with drug, the emergence of exploratory behavior was associated with an increased density and/or affinity of benzodiazepine receptors in cortex, hippocampus and brain stem, while three days after withdrawal of drug, when the exploratory behavior reached its peak, there was a reduction in GABA-enhanced binding of [3H]flunitrazepam in the cortex.

The basal forebrain cholinergic system: efferent and afferent connectivity and long-term effects of lesions

The first part of this article deals with several aspects of efferents and afferents of the rat basal forebrain cholinergic system (BFChS) studied with anterograde transport of Phaseolus vulgaris leucoagglutinin (PHA-L). PHA-L tracing of the BFChS efferents revealed topographically differentiated axonal trajectories and patterns of presynaptic endings to the neocortex, mesocortex, olfactory nuclei and hippocampus. Combining this method with second immunolabeling, we identified the muscarinic cholinoceptive neurons in the neocortex and the somatostatinergic neurons in the hippocampus as being directly innervated by the magnocellular basal nucleus and the medial septum, respectively. The prefrontal cortex was identified as a source of afferent input to the basal forebrain cholinergic neurons. This projection also exhibits a topographic organization, which shows a reciprocal relationship with the BFChS efferents to the cortex. The second part of this article describes the anatomical changes of cortical cholinergic and some other neurotransmitter systems after long-term cholinergic denervation in the aged rat cortex. The spared cholinergic projection in the largely denervated areas shows abundant malformations, which are similar in appearance to the anatomical alterations of the surviving cholinergic fibers in dementia of the Alzheimer type (AD). Hypertrophic changes also occur in the serotonergic system. The neuropeptide-Y- and somatostatin-containing cortical systems respond with an increment of their axonal densities, in contrast to the decline of these peptides in AD. Although transsynaptic effects are mediated by long-term cholinergic lesions, they do not support the hypothesis that the cholinergic deficiency is a primary event in the pathophysiology of AD.

Organization of ascending hypothalamic projections to the rostral forebrain with special reference to the innervation of cholinergic projection neurons

Axonal projections from hypothalamic nuclei to the basal forebrain, and their relation to cholinergic projection neurons in particular, were studied in the rat by using the anterograde tracer Phaseolus vulgaris-leucoagglutinin (PHA-L) in combination with choline acetyltransferase (ChAT) immunocytochemistry. Discrete iontophoretic PHA-L injections were delivered to different portions of the caudal lateral hypothalamus, as well as to various medial hypothalamic areas, including the ventromedial, dorsomedial, and paraventricular nuclei, and anterior hypothalamic and medial preoptic areas. The simultaneous detection of PHA-L-labeled fibers/terminals and ChAT-positive neurons was performed by using nickel-enhanced diaminobenzidine (DAB) and nonenhanced DAB as chromogens. Selected cases were investigated at the electron microscopic level. Ascending hypothalamic projections maintained an orderly lateromedial arrangement within the different components of the medial forebrain bundle, as well as with respect to their terminal projection fields (e.g., within the bed nucleus of the stria terminalis and lateral septal nucleus). The distribution pattern of hypothalamic inputs to cholinergic projection neurons corresponded to the topography of ascending hypothalamic axons. Axons originating from neurons in the far-lateral hypothalamus reached cholinergic neurons in a zone that extended from the dorsal part of the sublenticular substantia innominata (SI) caudolaterally, to the lateral portion of the bed nucleus of the stria terminalis rostromedially, encompassing a narrow band along the ventral part of the globus pallidus and medial portion of the internal capsule. Axons originating from cells in the medial portion of the lateral hypothalamus reached cholinergic cells primarily in more medial and ventral parts of the SI, and in the magnocellular preoptic nucleus and horizontal limb of the diagonal band nucleus (HDB). Axons from medial hypothalamic cells appeared to contact cholinergic neurons primarily in the medial part of the HDB, and in the medial septum/vertical limb of the diagonal band complex. Electron microscopic double-labeling experiments confirmed contacts between labeled terminals and cholinergic cells in the HDB and SI. Individual hypothalamic axons established synapses with both cholinergic and noncholinergic neuronal elements in the same regions. These findings have important implications for our understanding of the organization of afferents to the basal forebrain cholinergic projection system.

The role of interactions between the cholinergic system and other neuromodulatory systems in learning and memory

Extensive evidence indicates that disruption of cholinergic function is characteristic of aging and Alzheimer's disease (AD), and experimental manipulation of the cholinergic system in laboratory animals suggests age-related cholinergic dysfunction may play an important role in cognitive deterioration associated with aging and AD. Recent research, however, suggests that cholinergic dysfunction does not provide a complete account of age-related cognitive deficits and that age-related changes in cholinergic function typically occur within the context of changes in several other neuromodulatory systems. Evidence reviewed in this paper suggests that interactions between the cholinergic system and several of these neurotransmitters and neuromodulators--including norepinephrine, dopamine, serotonin, GABA, opioid peptides, galanin, substance P, and angiotensin II--may be important in learning and memory. Thus, it is important to consider not only the independent contributions of age-related changes in neuromodulatory systems to cognitive decline, but also the contribution of interactions between these systems to the learning and memory deficits associated with aging and AD.

GABAergic control of basal forebrain cholinergic neurons and memory

The involvement of the GABAergic innervation of basal forebrain neurons in the rats' conditional visual discrimination performance was examined. Performance in such a task is based on the subjects's ability to retrieve information about response rules, and previous experiments have demonstrated that basal forebrain lesions interfere with this ability. Following the acquisition of the task, chronic guide cannulae were stereotaxically implanted into the substantia innominata of both hemispheres, and the animals were retrained. Administration of the GABAA-agonist muscimol into the substantia innominata (0, 25, 50 ng/0.5 microliters/hemisphere) dose-dependently decreased the number of correct responses, increased the number of errors of omission, increased response latency, but did not affect side bias. Systemic co-administration of the cholinesterase inhibitor physostigmine (0, 0.1, 0.2 mg/kg; i.p.) exclusively interacted with the effects of muscimol on correct responding. Specifically, physostigmine dose-dependently intensified and attenuated the muscimol-induced reduction in correct responding. Although it cannot be excluded that alternative neuronal mechanisms were involved in the mediation of the effects of muscimol and their interaction with physostigmine, these findings support previous evidence indicating that the activity of basal forebrain cholinergic neurons is controlled by a GABAergic input, and that this neuronal link is involved in mnemonic processing.

The contribution of basal forebrain to limbic-motor integration and the mediation of motivation to action

The contribution of hippocampal glutamatergic and VTA dopaminergic inputs to the nucleus accumbens and the role of accumbens--ventral and subpallidal GABAergic pathway in integrating the limbic signals into motor responses via pedunculopontine nucleus were examined with electrophysiological and behavioural techniques. Stimulation of hippocampal input to the accumbens activates GABAergic output to the subpallidal area which leads to suppression of spontaneous firing of subpallidal neurons, while activation of dopamine receptors in the accumbens suppresses GABAergic output to subpallidal area and thus increases the firing of picrotoxin-sensitive ventral pallidal neurons. However, both treatments induced hypermotility suggesting the functional heterogeneity of the ventral and subpallidal areas in "limbic-motor integration". Furthermore, both hippocampal output signals and dopaminergic input to the accumbens descend via ventral and subpallidal areas serially to the pedunculopontine nucleus, the region of the mesencephalic locomotor region. In addition, a parallel ascending pathway from the subpallidal area to the mediodorsal nucleus, and subsequently to the medial prefrontal cortex, probably mediates behaviour, e.g. food hoarding, that requires higher cognitive processing.

Quantitative autoradiography of 11 different transmitter binding sites in the basal forebrain region of the rat--evidence of heterogeneity in distribution patterns

The distribution of 12 different binding sites for acetylcholine, L-glutamate, GABA, 5-hydroxytryptamine, dopamine and noradrenaline was measured with quantitative receptor autoradiography in four regions of the rat basal forebrain (medial septal nucleus including vertical and horizontal limbs of the diagonal band of Broca, magnocellular preoptic nucleus, substantia innominata and basal nucleus of Meynert, ventral pallidum). L-Glutamate binding sites represent the largest portion of the analysed receptors in all regions, followed by muscarinic2, 5-hydroxytryptamine1 and GABAA receptors. Muscarinic1, dopamine1, dopamine2 and 5-hydroxytryptamine2 receptors and alpha 1-, alpha 1A- and alpha 1B-adrenoceptors represent the minor receptor populations. The largest portion of the dopamine receptors is represented by the dopamine1 subtype, and the alpha 1B subtype dominates the alpha 1-adrenoceptor group. A heterogeneity of the distribution patterns of the different receptors throughout the basal forebrain regions is found. A comparison of the patterns shows that alpha 1-adrenoceptors have a similar regional distribution to that of the muscarinic2 receptors, but both receptor types have reciprocal distributions compared with the 5-hydroxytryptamine1 receptors. The results indicate that one transmitter may exert different effects in the basal forebrain regions depending on the densities of the respective receptor subtypes. Moreover, similar or reciprocal distribution patterns of some, but not all, analysed receptors point to a non-random association (co-distribution) of the different transmitter systems in the basal forebrain regions.

The effect of gamma-vinyl-GABA on the performance of nucleus basalis-lesioned rats in spatial navigation task

The present study investigates whether the stimulation of gamma-aminobutyric acid (GABA)ergic system affects spatial navigation deficits induced by the lesioning of the nucleus basalis (NB). Thus, the effect of gamma-vinyl-GABA treatment which elevates the GABA levels in brain was studied on water maze task both in unoperated and NB-lesioned (ibotenic acid) rats. The subchronic administration of gamma-vinyl-GABA aggravated dose-dependently NB lesion-induced deficits, although it did not impair the performance of unoperated rats in this task. The imbalance between GABAergic system and cholinergic or non-cholinergic systems of the NB may contribute to spatial navigation deficits in rats.

Catecholaminergic, GABAergic, and hippocamposeptal innervation of GABAergic "somatospiny" neurons in the rat lateral septal area

This study deals with the neurochemical characterization of the rat lateral septal area (LSA) somatospiny neurons and their innervation by hippocamposeptal, catecholaminergic, and GABAergic fibers. Electron microscopic single and double immunostaining methods were used to label catecholaminergic fibers and GABAergic cells and boutons. Axon terminals originating in the hippocampus were labeled by acute anterograde axon degeneration induced by fimbria-fornix transection 36 hours before sacrifice. Three types of experiments were performed. The convergent catecholaminergic and hippocamposeptal innervation of LSA somatospiny neurons was studied by combining immunostaining for tyrosine hydroxylase (TH) with fimbria-fornix transection. GABAergic neurons and their hippocamposeptal afferents were identified and characterized in colchicine pretreated animals immunostained for glutamic acid decarboxylase (GAD) combined with fimbria-fornix transection. The third experiment aimed at simultaneously visualizing the relationships between catecholaminergic boutons, hippocamposeptal excitatory amino acid containing axon terminals and GABAergic profiles by double immunostaining for TH (the PAP technique) and GAD (the immunogold method) combined with fimbria-fornix transection. The results are summarized as follows: 1) The same LSA somatospiny neurons receive synaptic inputs from the hippocampus and TH immunoreactive fibers which form pericellular baskets around these cells. 2) LSA somatospiny neurons are GABAergic and are postsynaptic targets of GABAergic boutons with unknown origin and hippocamposeptal axon terminals. 3) The double immunostaining experiment, finally, provided direct evidence that the same GABAergic somatospiny neurons are postsynaptic targets of both catecholaminergic and hippocamposeptal afferents. The synaptic interconnections described in this study provide anatomical basis for a better understanding of the action of catecholamines, excitatory amino acids, and GABA on the activity of LSA neurons.

Cholinergic synapses in the central nervous system: studies of the immunocytochemical localization of choline acetyltransferase

Cholinergic synapses can be identified in immunocytochemical preparations by the use of monoclonal antibodies and specific antisera to choline acetyltransferase (ChAT), the synthesizing enzyme for acetylcholine (ACh) and a specific marker for cholinergic neurons. Electron microscopic studies demonstrate that the fibers and varicosities observed in light microscopic preparations of many brain regions are small-diameter unmyelinated axons and vesicle-containing boutons. The labeled boutons generally contain clear vesicles and one or more mitochondrial profiles. Many of these boutons form synaptic contacts, and the synapses are frequently of the symmetric type, displaying thin postsynaptic densities and relatively short contact zones. However, ChAT-labeled synapses with asymmetric junctions are also observed, and their frequency varies among different brain regions. Unlabeled dendritic shafts are the most common postsynaptic elements in virtually all regions examined although other neuronal elements, including dendritic spines and neuronal somata, also receive some cholinergic innervation. ChAT-labeled boutons form synaptic contacts with several different types of unlabeled neurons within the same brain region. Such findings are consistent with a generally diffuse pattern of cholinergic innervation in many parts of the central nervous system. Despite many similarities in the characteristics of ChAT-labeled synapses, there appears to be some heterogeneity in the cholinergic innervation within as well as among brain regions. Differences are observed in the sizes of ChAT-immunoreactive boutons, the types of synaptic contacts, and the predominant postsynaptic elements. Thus, the cholinergic system presents interesting challenges for future studies of the morphological organization and related function of cholinergic synapses.

GABAergic synapses in the brain identified with antisera to GABA and its synthesizing enzyme, glutamate decarboxylase

GABA is a known inhibitory neurotransmitter in the mammalian brain. The site of GABAergic synapses can be determined with immunocytochemical methods that localize either GABA or its synthesizing enzyme, glutamate decarboxylase (GAD). In general, GABAergic axon terminals contain pleomorphic synaptic vesicles and form symmetric synapses. However, a small number of GABAergic axon terminals in selected brain regions (spinal cord, cerebellum, superior colliculus, striatum, globus pallidus, inferior olive, and substantia nigra) form asymmetric synapses. GAD- and GABA-immunoreactive processes that contain synaptic vesicles participate in every known morphological type of chemical synapse. These include axosomatic, axodendritic, axospinous, initial segment, axoaxonic, dendrodendritic, serial, reciprocal, and ribbon synapses. Although GABAergic synapses form a heterogeneous group, they most commonly form axosomatic, axodendritic, and initial segment synapses in the brain and spinal cord. These findings provide helpful guidelines for the identification of GABAergic synapses in future studies.

Medial forebrain bundle of the rat: III. Cytoarchitecture of the rostral (telencephalic) part of the medial forebrain bundle bed nucleus

The boundaries of the medial forebrain bundle (MFB) of the rat have been presented in previous work on the structure of this fiber system (Nieuwenhuys et al.: J. Comp. Neurol. 206:49-81, '82). Neuronal cell bodies within these outlines constitute the bed nucleus of the MFB. Many fiber components of the MFB appeared to be spatially arranged within the bundle and featured an orderly topography (Veening et al.: J. Comp. Neurol. 206:82-108, '82). As the fibers of the MFB are thought to be a major source of afferents to the bed nucleus (Millhouse: In P.J. Morgane and J. Panksepp (eds): Anatomy of the Hypothalamus, Vol. 1. New York: Marcel Dekker, pp. 221-265, '79), the latter has been subjected in this and the companion study (Geeraedts et al.: J. Comp. Neurol. 294:537-568, '90) to a detailed cytoarchitectonic analysis. This analysis is based on continuous series sectioned in the three conventional planes. On the basis of cytoarchitectonic characteristics, including size and shape, staining intensity, packing density, and spatial orientation of the cell bodies, it was found that the bed nucleus of the MFB as described in the literature is by no means a cytoarchitectonic unit per se. Rather, the neuronal cell population located within the telencephalic stream of the MFB can be parcellated into a number of cellular groups, which partly or entirely belong to more-or-less known basal telencephalic structures. These structures are designated here as the MFB-related areas. They correspond largely to the subcommissural substantia innominata (SIC), the sublenticular substantia innominata (SIL), the nucleus of the diagonal band of Broca, the olfactory tubercle, the magnocellular preoptic nucleus (POMA), the lateral preoptic area (LPOA), and the interstitial nucleus of the stria medullaris (ISM). The complex of the MFB-related areas is surrounded by the following cellular entities: the nucleus accumbens (ACB), the caudatus-putamen region (CPU), the globus pallidus (GP), the bed nucleus of the stria terminalis (BST), the anterior amygdaloid area (AAA), the amygdaloid nuclear complex (A), the medial preoptic area (MPOA) and the anterior hypothalamic area (AHA). Both MFB-related areas and their surroundings have been identified and delimited in this study. This resulted in a new cytoarchitectonic atlas of the rat's basal telencephalon. Our atlas does not only show the relative positions of the above mentioned cellular groups, but also those of their subdivisions.(ABSTRACT TRUNCATED AT 400 WORDS)

Activating the damaged basal forebrain cholinergic system: tonic stimulation versus signal amplification

The hypothesis that the cognitive decline in senile dementia is related to the loss of cortical cholinergic afferent projections predicts that pharmacological manipulations of the remaining cholinergic neurons will have therapeutic effects. However, treatment with cholinesterase inhibitors or muscarinic agonists has been, for the most part, largely unproductive. These drugs seem to disrupt the normal patterning of cholinergic transmission and thus may block proper signal processing. An alternative pharmacological strategy which focuses on the amplification of presynaptic activity without disrupting the normal patterning of cholinergic transmission appears to be more promising. Such a strategy may make use of the normal GABAergic innervation of basal forebrain cholinergic neurons in general, and in particular of the inhibitory hyperinnervation of remaining cholinergic neurons which may develop under pathological conditions. Disinhibition of the GABAergic control of cholinergic activity is assumed to intensify presynaptic cortical cholinergic activity and to enhance cognitive processing. Although the extent to which compounds such as the benzodiazepine receptor antagonist beta-carboline ZK 93,426 act via the basal forebrain GABA-cholinergic link is not yet clear, the available data suggest that the beneficial behavioral effects of this compound established in animals and humans are based on indirect cholinomimetic mechanisms. It is proposed that an activation of residual basal forebrain cholinergic neurons can be achieved most physiologically via inhibitory modulation of afferent GABAergic transmission. This modulation may have a therapeutic value in treating behavioral syndromes associated with cortical cholinergic denervation.

Ultrastructural localization of [125I]neurotensin binding sites to cholinergic neurons of the rat nucleus basalis magnocellularis

The distribution of specifically-labeled neurotensin binding sites was examined in relation to that of cholinergic neurons in the rat nucleus basalis magnocellularis at both light and electron microscopic levels. Lightly prefixed forebrain slices were either labeled with [125I](Tyr3) neurotensin alone or processed for combined [125I]neurotensin radioautography and acetylcholinesterase histochemistry. In light microscopic radioautographs from 1-microns-thick sections taken from the surface of single-labeled slices, silver grains were found to be preferentially localized over perikarya and proximal processes of nucleus basalis cells. The label was distributed both throughout the cytoplasm and along the plasma membrane of magnocellular neurons all of which were found to be cholinesterase-positive in a double-labeled material. Probability circle analysis of silver grain distribution in electron microscopic radioautographs confirmed that the major fraction (80-89%) of specifically-labeled binding sites associated with cholinesterase-reactive cell bodies and dendrites was intraneuronal. These intraneuronal sites were mainly dispersed throughout the cytoplasm and are thus likely to represent receptors undergoing synthesis, transport and/or recycling. A proportion of the specific label was also localized over the nucleus, suggesting that neurotensin could modulate the expression of acetylcholine-related enzymes in the nucleus basalis. The remainder of the grains (11-20%) were classified as shared, i.e. overlied the plasma membrane of acetylcholinesterase-positive neuronal perikarya and dendrites. Extrapolation from light microscopic data, combined with the observation that shared grains were detected at several contact points along the plasma membrane of cells which also exhibited exclusive grains, made it possible to ascribe these membrane-associated receptors to the cholinergic neurons themselves rather than to abutting cellular profiles. Comparison of grain distribution with the frequency of occurrence of elements directly abutting the plasma membrane of neurotensin-labeled/cholinesterase-positive perikarya indicated that labeled cell surface receptors were more or less evenly distributed along the membrane as opposed to being concentrated opposite abutting axon terminals endowed or not with a visible junctional specialization. The low incidence of labeled binding sites found in close association with abutting axons makes it unlikely that only this sub-population of sites corresponds to functional receptors. On the contrary, the dispersion of labeled receptors seen here along the plasma membrane of cholinergic neurons suggests that neurotensin acts primarily in a paracrine mode to influence the magnocellular cholinergic system in the nucleus basalis.

Immunocytochemical localization of cholinergic terminals in the region of the nucleus basalis magnocellularis of the rat: a correlated light and electron microscopic study

The cholinergic circuitry in the nucleus basalis magnocellularis of the rat was investigated in a correlated light and electron microscopic study by using monoclonal antibodies against the acetylcholine-synthesizing enzyme, choline acetyltransferase, following the unlabelled antibody peroxidase-antiperoxidase immunocytochemical procedure. After the immunocytochemical approach, large cholinergic cells and a few immunoreactive fibres exhibiting a varicose appearance, were detected by light microscopy in portions of the nucleus basalis magnocellularis located within the anatomical limits of the globus pallidus, mostly in its ventromedial part. Cholinergic neurons and fibre-like structures were also found within the substantia innominata on the edge of globus pallidus. The same material studied by light microscopy was analysed with the electron microscope. At the ultrastructural level, the immunopositive neurons showed the same cytological characteristics and pattern of synaptic input as cholinergic basal forebrain cells. Additionally, scarce immunoreactive preterminal axons and terminal boutons were detected in the region. The immunoreactive terminals were scattered or formed occasional clusters and appeared as heavily immunostained vesicle-filled boutons making exclusively axodendritic synaptic contacts principally with immunonegative distal dendrites. Both symmetric and asymmetric synaptic contacts established between these structures were detected, although the symmetric contacts were the more numerous. The surface of postsynaptic immunonegative dendrites in asymmetric synaptic contact with immunoreactive terminals was generally covered by terminals that lacked detectable immunoreactivity. In contrast, those in symmetric synaptic contact with labelled terminals showed much sparser input from immunonegative terminals, suggesting that they may belong to interneurons. Very rarely, cholinergic terminals were detected in asymmetric synaptic contact with dendrites which also contained positive immunoreaction product. Asymmetric contacts were frequently characterized by the presence of subjunctional dense bodies. The detection of cholinergic terminals in the region of the nucleus basalis magnocellularis of the rat indicates that this region not only contains cholinergic projecting neurons, but receives a cholinergic input itself. Results of this study provide evidence of the existence of a cholinergic transmission in the basal forebrain of the rat, and also that acetylcholine might play a role in the regulation of the extrinsic cortical cholinergic innervation. The possible sources of this innervation are discussed.

Organization of the septal region in the rat brain: cholinergic-GABAergic interconnections and the termination of hippocampo-septal fibers

This study deals with two characteristic cell types in the rat septal complex i.e., cholinergic and GABAergic neurons, and their synaptic connections. Cholinergic elements were labeled with a monoclonal antibody against choline acetyltransferase (ChAT), the acetylcholine synthesizing enzyme. Antiserum against glutamate decarboxylase (GAD), the GABA synthesizing enzyme, was employed to identify GABAergic perikarya and terminals, by using either the peroxidase-antiperoxidase (PAP) technique or a biotinylated second antiserum and avidinated gold or ferritin. With these contrasting immunolabels we have studied the cholinergic-GABAergic interconnections in double-labeled sections of intact septal regions and the GABAergic innervation of medial septal area cholinergic neurons in sections taken from animals 1 week following lateral septal area lesion. In other electron microscopic experiments we have studied cholinergic and GABAergic neurons in the septal complex for synaptic contacts with hippocamposeptal fibers, which were identified by anterograde degeneration following fimbria-fornix transection. Our results are summarized as follows: (1) GAD-positive terminals form synaptic contacts on ChAT-immunoreactive dendrites in the medial septum/diagonal band complex (MSDB), (2) surgical lesion of the lateral septal area resulted in a dramatic decrease of the number of GABAergic boutons on MSDB cholinergic neurons, (3) cholinergic terminals establish synaptic contacts with GAD immunoreactive cell bodies and proximal dendrites in the MSDB as well as in the lateral septum (LS), (4) degenerated terminals of hippocampo-septal fibers were mainly observed in the LS, where they formed asymmetric synaptic contacts on dendrites of GABAergic neurons and on nonimmunoreactive spines. We did not observe degenerated boutons in contact with ChAT-positive dendrites or cell bodies in the MSDB. From these results and from data in the literature we conclude that excitatory hippocampo-septal fibers activate GABAergic cells, and as yet unidentified spiny neurons in the LS, which may control the discharge of medial septal cholinergic neurons known to project back to the hippocampal formation.

GABAergic neurons in the primate basal forebrain magnocellular complex

Hybridization histochemistry was used to detect messenger ribonucleic acid (mRNA) coding for glutamic acid decarboxylase, the synthesizing enzyme for gamma-aminobutyric acid (GABA), in neurons of the nucleus basalis of Meynert and nucleus of the diagonal band of Broca of one rhesus monkey and 4 baboons. GABAergic neurons were distributed among the unlabeled large, hyperchromic Nissl-stained neurons characteristic of this basal forebrain magnocellular complex, although they were infrequent within the dense islands of large cells. Most GABAergic cells were small to medium in size, but some were large and hyperchromic. These findings demonstrate a heterogeneous population of presumably inhibitory neurons in the basal forebrain magnocellular complex of primates.

Release of acetylcholine from tissue slices of the rat nucleus basalis magnocellularis

We investigated the release of acetylcholine (ACh) from tissue slices obtained from the nucleus basalis magnocellularis (nbM) of the rat brain. Potassium (35 mM) depolarization produced a 10- to 12-fold increase in the release of endogenous ACh above spontaneous release. Potassium-evoked ACh release was Ca2+ dependent. Injection of the excitotoxin quinolinic acid into the nbM produced a 72.8 +/- 13.0% decrease in spontaneous ACh release and a 60.4 +/- 8.2% decrease in potassium-evoked release. A fourfold increase in ACh release was observed following perfusion of the tissue with 1 mM 3,4-diaminopyridine (3,4-DAP) whereas 10 mM 3,4-DAP caused a sevenfold increase. The increase in ACh release caused by 3,4-DAP was inhibited by tetrodotoxin. Tissue slices accumulated [3H]choline by high-affinity choline uptake and this could be inhibited by hemicholinium-3. These results indicate that ACh can be released from tissue slices of the nbM by a calcium-dependent process and that a part of this release appears to be from the cholinergic neurons of the nbM.

Peptidergic neurons in the basal forebrain magnocellular complex of the rhesus monkey

The basal forebrain magnocellular complex of primates is defined by the presence of large, hyperchromic, usually cholinergic neurons in the nucleus basalis of Meynert and nucleus of the diagonal band of Broca. Because there is growing evidence for noncholinergic neuronal elements in the basal forebrain complex, five neuropeptides and the enzyme choline acetyltransferase were studied immunocytochemically in this region of rhesus monkeys. Galaninlike immunoreactivity coexists with choline-acetyl-transferase-like immunoreactivity in most large neurons and in some smaller neurons of the primate nucleus basalis and nucleus of the diagnonal band. Four other peptides show immunoreactivity in more limited regions of the basal forebrain complex, usually in separate smaller, noncholinergic neurons. Numerous small, somatostatinlike-immunoreactive neurons occupy primarily anterior and intermediate segments of the nucleus basalis, especially laterally and ventrally. Somewhat fewer, small neuropeptide Y-like-immunoreactive somata are found in the same regions. Neurons that show neurotensinlike immunoreactivity are slightly larger than cells that contain immunoreactivity for somatostatin or neuropeptide Y, but these neurons also occur mainly in anterior and intermediate parts of the nucleus basalis. Overall, the usually small, leucine-enkephalin-like-immunoreactive neurons are infrequent in the basal forebrain complex and are most abundant in the rostral intermediate nucleus basalis. Thus, neurons that appear to contain somatostatin, neuropeptide Y, neurotensin, or enkephalin mingle with cholinergic/galaninergic neurons only in some subdivisions of the nucleus basalis/nucleus of the diagonal band, and their distributions suggest that some of these small neurons could be associated with structures that overlap with cholinergic neurons of the labyrinthine basal forebrain magnocellular complex. We also have found light microscopic evidence for innervation of basal forebrain cholinergic neurons by boutons that contain galanin-, somatostatin-, neuropeptide Y-, neurotensin-, or enkephalinlike immunoreactivity. The origins and functions of these putative synapses remain to be determined.

Hypothalamic axons terminate on forebrain cholinergic neurons: an ultrastructural double-labeling study using PHA-L tracing and ChAT immunocytochemistry

A correlated light and electron microscopic study combining choline acetyltransferase immunocytochemistry with anterograde tracing of Phaseolus vulgaris leucoagglutinin revealed that axons of lateral hypothalamic neurons terminate on forebrain cholinergic projection cells. This finding is discussed in relation to direct and indirect corticopetal systems.

Organization and synaptic interconnections of GABAergic and cholinergic elements in the rat amygdaloid nuclei: single- and double-immunolabeling studies

The aim of this study was to describe the localization of cholinergic and GABAergic neurons and terminals in the amygdaloid nuclei of the rat. Double immunolabeling was performed to study cholinergic-GABAergic synaptic interconnections. Cholinergic elements were labeled by using a monoclonal antibody to choline acetyltransferase (ChAT), the acetylcholine synthesizing enzyme. Antibodies against glutamate decarboxylase (GAD), the GABA- synthesizing enzyme, were employed to identify GABAergic perikarya and terminals. The tissue sites of the antibody bindings were detected by using either Sternberger's peroxidase-antiperoxidase (PAP) method or a biotinylated secondary antibody and avidinated ferritin. These two contrasting immunolabels allowed us to study GABAergic-cholinergic interconnections at the electron microscopic level. Our study revealed a characteristic distribution of GABAergic and cholinergic elements in the various amygdaloid nuclei: 1) Large, ChAT-immunopositive cells with heavily labeled dendrites were observed in the anterior amygdaloid area and in the lateral and medial zones of the central nucleus. These cells seem to constitute the intraamygdaloid extension of the magnocellular basal nucleus. Their dendrites invaded other amygdaloid nuclei, in particular the intercalated nuclei, the lateral olfactory tract nucleus, and the central zone of the central nucleus. These ChAT-immunoreactive dendrites formed synaptic contacts with GAD-positive terminals. GABAergic terminals probably thus exert an inhibitory amygdaloid influence onto cholinergic neurons of the magnocellular basal nucleus. 2) Two amygdaloid nuclei-the basal dorsal nucleus and the lateral olfactory tract nucleus-contained a dense network of ChAT-immunoreactive fibers and terminals, but they also contained numerous GAD-positive perikarya. Double-immunolabeling experiments revealed cholinergic terminals forming synaptic contacts on GAD-immunopositive cell bodies, dendritic shafts, and spines. 3) The central and medial nucleus seem to be the main target of GABAergic fibers to the amygdala. Both nuclei contained a dense plexus of GAD-immunoreactive terminals that may arise, at least in part, from the GABAergic neurons in the basal dorsal nucleus. Inhibition of the centromedial "excitatory" region through intraamygdaloid GABAergic connections may reduce excitatory amygdaloid influence onto hypothalamus and brainstem.

Scopolamine and benzodiazepine models of dementia: cross-reversals by Ro 15-1788 and physostigmine

The muscarinic antagonist scopolamine and the benzodiazepine lorazepam both produce transient impairments in memory and attention in normal volunteers. These impairments can be reversed by appropriate agents such as the cholinesterase inhibitor physostigmine in the case of scopolamine or the benzodiazepine antagonist Ro 15-1788 in the case of lorazepam. In this paper we investigated the pharmacological specificity of these reversals by examining the interactions of scopolamine and Ro 15-1788 and of lorazepam and physostigmine. There was no evidence that the effects of scopolamine and lorazepam on cognitive function could be attenuated by Ro 15-1788 and physostigmine, respectively. The results are discussed in terms of pharmacological models of Alzheimer's disease.

Afferent connections of the forebrain cholinergic projection neurons, with special reference to monoaminergic and peptidergic fibers

Earlier light microscopic data on afferent connections to the cholinergic forebrain neurons are reconsidered in the light of EM cross-identification of neurons and synapses by combinations of tracer and immunocytochemical techniques. Such studies suggest that brainstem monoaminergic afferents terminate on cholinergic forebrain neurons, and may modulate the activity of choline acetyltransferase levels in the postsynaptic neurons. A monosynaptic relationship between cholinergic forebrain neurons and neuropeptide Y and somatostatin containing axons is also supported by studies using double immunolabeling techniques at the EM level. These peptidergic afferents originate in part from locally arborizing neurons. Based upon the new data a circuit model for basal forebrain cholinergic neurons is proposed.

Neural associations of the substantia innominata in the rat: afferent connections

The afferent connections of the substantia innominata (SI) in the rat were determined employing the anterograde axonal transport of Phaseolus vulgaris leucoagglutinin (PHA-L) and the retrograde transport of wheat germ agglutinin conjugated to horseradish peroxidase (WGA-HRP), in combination with histochemical procedures to characterize the neuropil of the SI and identify cholinergic cells. Both neurochemical and connectional data establish that the SI is organized into a dorsal and a ventral division. Each of these divisions is strongly affiliated with a different region of the amygdala, and, together with its amygdalar affiliate, forms part of one of two largely distinct constellations of interconnected forebrain and brainstem cell groups. The dorsal SI receives selective innervation from the lateral part of the bed nucleus of the stria terminalis, the central and basolateral nuclei of the amygdala, the fundus of the striatum, distinctive perifornical and caudolateral zones of the lateral hypothalamus, and caudal brainstem structures including the dorsal raphe nucleus, parabrachial nucleus, and nucleus of the solitary tract. Projections preferentially directed to the ventral SI arise from the medial part of the bed nucleus of the stria terminalis, the rostral two-thirds of the medial nucleus of the amygdala, a large region of the rat amygdala that lies ventral to the central nucleus, the medial preoptic area, anterior hypothalamus, medialmost lateral hypothalamus, and the ventromedial hypothalamus. Both SI divisions appear to receive afferents from the dorsomedial and posterior hypothalamus, supramammillary region, ventral tegmental area, and the peripeduncular area of the midbrain. Projections to the SI whose selectivity was not determined originate from medial prefrontal, insular, perirhinal, and entorhinal cortex and from midline thalamic nuclei. Findings from both PHA-L and WGA-HRP experiments additionally indicate that cell groups preferentially innervating a single SI division maintain numerous projections to one another, thus forming a tightly linked assembly of structures. In the rat, cholinergic neurons that are scattered throughout the SI and in parts of the globus pallidus make up a cell population equivalent to the primate basal nucleus of Meynert (Mesulam et al.: Neuroscience 10:1185-1201, '83). PHA-L-filled axons, labelled from lectin deposits in the dorsal raphe nucleus, peripeduncular area, ventral tegmental area, or caudomedial hypothalamus were occasionally seen to approach individual cholinergic neurons int he SI, and to contact the surface of such cells with axonal varicosities (putative synaptic boutons.(ABSTRACT TRUNCATED AT 400 WORDS)

Efferent connections of the substantia innominata in the rat

The efferent connections of the substantia innominata (SI) were investigated employing the anterograde axonal transport of Phaseolus vulgaris leucoagglutinin (PHA-L) and the retrograde transport of wheat germ agglutinin conjugated to horseradish peroxidase (WGA-HRP). The projections of the SI largely reciprocate the afferent connections described by Grove (J. Comp. Neurol. 277:315-346, '88) and thus further distinguish a dorsal and a ventral division in the SI. Efferents from both the dorsal and ventral divisions of the SI descend as far caudal as the ventral tegmental area, substantia nigra, and peripeduncular area, but projections to pontine and medullary structures appear to originate mainly from the dorsal SI. Within the amygdala and hypothalamus, which receive widespread innervation from the SI, the dorsal SI projects preferentially to the lateral part of the bed nucleus of the stria terminalis; the lateral, basolateral, and central nuclei of the amygdala; the lateral preoptic area; paraventricular nucleus of the hypothalamus; and certain parts of the lateral hypothalamus, prominently including the perifornical and caudolateral zones described previously. The ventral SI projects more heavily to the medial part of the bed nucleus of the stria terminalis; the anterior amygdaloid area; a ventromedial amygdaloid region that includes but is not limited to the medial nucleus; the lateral and medial preoptic areas; and the anterior hypothalamus. Modest projections reach the lateral hypothalamus, with at least a slight preference for the medial part of the region, and the ventromedial and arcuate hypothalamic nuclei. Both SI divisions appear to innervate the dorsomedial and posterior hypothalamus and the supramammillary region. In the thalamus, the subparafascicular, gustatory, and midline nuclei receive a light innervation from the SI, which projects more densely to the medial part of the mediodorsal nucleus and the reticular nucleus. Cortical efferents from at least the midrostrocaudal part of the SI are distributed primarily in piriform, infralimbic, prelimbic, anterior cingulate, granular and agranular insular, perirhinal, and entorhinal cortices as well as in the main and accessory olfactory bulbs. The cells of origin for many projections arising from the SI were identified as cholinergic or noncholinergic by combining the retrograde transport of WGA-HRP with histochemical and immunohistochemical procedures to demonstrate acetylcholinesterase activity or choline acetyltransferase immunoreactivity. Most of the descending efferents of the SI appear to arise primarily or exclusively from noncholinergic cells.(ABSTRACT TRUNCATED AT 400 WORDS)