Cholinergic innervation of the mediodorsal thalamic nucleus in the monkey: ultrastructural evidence supportive of functional diversity

Organization of primate amygdalar-thalamic pathways for emotions

Studies on the thalamus have mostly focused on sensory relay nuclei, but the organization of pathways associated with emotions is not well understood. We addressed this issue by testing the hypothesis that the primate amygdala acts, in part, like a sensory structure for the affective import of stimuli and conveys this information to the mediodorsal thalamic nucleus, magnocellular part (MDmc). We found that primate sensory cortices innervate amygdalar sites that project to the MDmc, which projects to the orbitofrontal cortex. As in sensory thalamic systems, large amygdalar terminals innervated excitatory relay and inhibitory neurons in the MDmc that facilitate faithful transmission to the cortex. The amygdala, however, uniquely innervated a few MDmc neurons by surrounding and isolating large segments of their proximal dendrites, as revealed by three-dimensional high-resolution reconstruction. Physiologic studies have shown that large axon terminals are found in pathways issued from motor systems that innervate other brain centers to help distinguish self-initiated from other movements. By analogy, the amygdalar pathway to the MDmc may convey signals forwarded to the orbitofrontal cortex to monitor and update the status of the environment in processes deranged in schizophrenia, resulting in attribution of thoughts and actions to external sources.

Modeling Parkinson's disease falls associated with brainstem cholinergic systems decline

In addition to the primary disease-defining symptoms, approximately half of patients with Parkinson's disease (PD) suffer from postural instability, impairments in gait control and a propensity for falls. Consistent with evidence from patients, we previously demonstrated that combined striatal dopamine (DA) and basal forebrain (BF) cholinergic cell loss causes falls in rats traversing dynamic surfaces. Because evidence suggests that degeneration of brainstem cholinergic neurons arising from the pedunculopontine nucleus (PPN) also contributes to impaired gait and falls, here we assessed the effects of selective cholinergic PPN lesions in combination with striatal DA loss or BF cholinergic cells loss as well as losses in all 3 regions. Results indicate that all combination losses that included the BF cholinergic system slowed traversal and increased slips and falls. However, the performance of rats with losses in all 3 regions (PPN, BF, and DA) was not more severely impaired than following combined BF cholinergic and striatal DA lesions. These results confirm the hypothesis that BF cholinergic-striatal disruption of attentional-motor interactions is a primary source of falls. Additional losses of PPN cholinergic neurons may worsen posture and gait control in situations not captured by the current testing conditions.

Dopamine innervation in the thalamus: monkey versus rat

We recently identified the thalamic dopaminergic system in the human and macaque monkey brains, and, based on earlier reports on the paucity of dopamine in the rat thalamus, hypothesized that this dopaminergic system was particularly developed in primates. Here we test this hypothesis using immunohistochemistry against the dopamine transporter (DAT) in adult macaque and rat brains. The extent and density of DAT-immunoreactive (-ir) axons were remarkably greater in the macaque dorsal thalamus, where the mediodorsal association nucleus and the ventral motor nuclei held the densest immunolabeling. In contrast, sparse DAT immunolabeling was present in the rat dorsal thalamus; it was mainly located in the mediodorsal, paraventricular, ventral medial, and ventral lateral nuclei. The reticular nucleus, zona incerta, and lateral habenular nucleus held numerous DAT-ir axons in both species. Ultrastructural analysis in the macaque mediodorsal nucleus revealed that thalamic interneurons are a main postsynaptic target of DAT-ir axons; this suggests that the marked expansion of the dopamine innervation in the primate in comparison to the rodent thalamus may be related to the presence of a sizable interneuron population in primates. We remark that it is important to be aware of brain species differences when using animal models of human brain disease.

Ultrastructure of ascending cholinergic terminals in the anteroventral thalamic nucleus of the rat: a comparison with the mammillothalamic terminals

In this study, to identify the ultrastructure and distribution of ascending cholinergic afferent terminals in the anteroventral thalamic nucleus, we used an anti-vesicular acetylcholine transporter antibody as marker of cholinergic afferents, and characterized the immunoreactive terminals at the ultrastructural level. We then compared the distribution pattern of the cholinergic terminals and that of the mammillothalamic terminals identified by anterograde transport of a tracer injected into the mammillary body. The cholinergic terminals were small, and formed both symmetrical and asymmetrical synaptic contacts throughout the dendritic arborizations, particularly in the distal region. This distribution pattern differed from that of mammillothalamic terminals, that were of LR (large terminal containing round synaptic vesicles) type and were preferentially distributed in the proximal region of dendrites. We also found relatively numerous cholinergic terminals making contact directly with immunonegative excitatory terminals, both LR and SR (small terminal containing round vesicles) terminals, without clear postsynaptic specialization. A few cholinergic terminals even seemed to form a synaptic complex with the LR or SR terminals. These findings suggest that the ascending cholinergic afferents in the anteroventral thalamic nucleus can effectively modulate excitatory inputs from both the mammillothalamic and corticothalamic terminals, in close vicinity to a synaptic site.

Endogenous nicotinic cholinergic activity regulates dopamine release in the striatum

Dopamine is vital for coordinated motion and for association learning linked to behavioral reinforcement. Here we show that the precise overlap of striatal dopaminergic and cholinergic fibers underlies potent control of dopamine release by ongoing nicotinic receptor activity. In mouse striatal slices, nicotinic antagonists or depletion of endogenous acetylcholine decreased evoked dopamine release by 90%. Nicotine at the concentration experienced by smokers also regulated dopamine release. In mutant mice lacking the beta2 nicotinic subunit, evoked dopamine release was dramatically suppressed, and those mice did not show cholinergic regulation of dopamine release. The results offer new perspectives when considering nicotine addiction and the high prevalence of smoking in schizophrenics.

Multiple output pathways of the basal forebrain: organization, chemical heterogeneity, and roles in vigilance

Studies over the last decade have shown that the basal forebrain (BF) consists of more than its cholinergic neurons. The BF also contains non-cholinergic neurons, including gamma-aminobutyric acid-ergic neurons which co-distribute and co-project with the cholinergic neurons. Both types of neuron project, in variable proportions, to the cerebral cortex, hippocampus, thalamus, amygdala, and olfactory bulb, whereas descending projections to the posterior hypothalamus and brainstem nuclei are predominantly non-cholinergic. Some of the cholinergic and non-cholinergic projection neurons contain neuropeptides such as galanin, nitric oxide synthase, and possibly glutamate. To understand better the function of the BF, the organization of the multiple ascending and descending projections of BF neurons is reviewed along with their neurochemical heterogeneity, and possible functions of individual pathways are discussed. It is proposed that BF neurons belong to multiple systems with distinct cognitive, motivational, emotional, motor, and regulatory functions, and that through these pathways, the BF plays a role in controlling both cognitive and non-cognitive aspects of vigilance.

Ultrastructural organization of transmitters in the cat lateralis medialis-suprageniculate nucleus of the thalamus: an immunohistochemical study

The lateralis medialis-suprageniculate nuclear (LM-Sg) complex of the cat's posterior thalamus receives a rather wide variety of inputs from diverse cortical and subcortical areas. Previous ultrastructural studies of this nucleus demonstrated the presence of four types of vesicle-containing profiles and characterized some of these as gamma-aminobutyric acid (GABA)-containing terminals (Norita and Katoh [1987] J. Comp. Neurol. 263:54-67; Norita and Katoh [1988] Prog. Brain Res. 75:109-118). The present study has extended these observations by examining the immunoreactivity (ir) of LM-Sg, with antibodies raised against aspartate (Asp), glutamate (Glu), GABA, the acetylcholine (ACh) marker, choline acetyltransferase (ChAT), and substance P (SP), by using light and electron microscopy. Neuronal somata immunopositive for the excitatory amino acids (EAAs) Asp and Glu, were of medium size. EAA-ir terminals also were of medium size and contained round synaptic vesicles; they made asymmetrical synaptic contacts with dendritic profiles. Neuronal somata immunopositive for GABA were small. GABA-positive terminals also were small and contained pleomorphic synaptic vesicles; they formed symmetrical synaptic contacts with dendritic profiles. No neurons immunolabeled for ChAT were found. Terminals immunopositive for ChAT were small and contained round synaptic vesicles; these made symmetrical synaptic contacts, asymmetrical synaptic contacts, or both, of the en passant type with dendritic profiles. SP-immunolabeled neuronal somata were not found. Immunolabeled terminals were small, contained round synaptic vesicles, and made asymmetrical synaptic contacts with dendritic profiles. ChAT-ir and SP-ir axon terminals were not expressed evenly within LM-Sg. This difference in distribution suggests that within the LM-Sg, there may be a difference in specific sensory processing functions which correlate with transmitter type.

Cholinergic neurons of the nucleus basalis of Meynert receive cholinergic, catecholaminergic and GABAergic synapses: an electron microscopic investigation in the monkey

An electron microscopic analysis of the nucleus basalis in the macaque monkey was carried out following the immunohistochemical labeling of choline acetyltransferase, either by itself or in conjunction with glutamate decarboxylase or tyrosine hydroxylase. Cholinergic axon varicosities were frequently encountered, and formed large, usually asymmetric, synapses on both choline acetyltransferase-immunopositive and -immunonegative dendrites of nucleus basalis neurons. Catecholaminergic (tyrosine hydroxylase-immunoreactive) axon varicosities formed synapses which in most cases were classified as asymmetric, and glutamate decarboxylase-immunoreactive (GABAergic) axons formed clearly symmetric synapses, each on to choline acetyltransferase-immunopositive or -immunonegative dendrites. These findings indicate that cholinergic cells in the nucleus basalis of the monkey, also known as Ch4 neurons, receive numerous synaptic inputs from cholinergic, catecholaminergic and GABAergic axons.

An ultrastructural study of p75 neurotrophin receptor-immunoreactive fiber terminals in the reticular thalamic nucleus of young rats

The reticular thalamic nucleus (RT) receives cholinergic fibers from both the basal forebrain and the brainstem. Recent studies have shown that the p75 neurotrophin receptor (p75NTR) is synthesized in cholinergic neurons in the basal forebrain but not in those in the brainstem. In this study, to identify cholinergic fibers originating from the basal forebrain, we used a monoclonal antibody against p75NTR (192-IgG) and characterized the ultrastructure of the immunoreactive fiber terminals in the rostral part of the RT in 3-week-old rats. Light microscopy revealed that p75NTR-immunoreactive fine fibers and varicosities were distributed throughout the nucleus. From electron micrographs, three types of labeled terminals were identified. The first type of labeled fiber terminals (63 out of 106) was consistently small, contained densely packed vesicles, and established asymmetrical synaptic contacts with heavy and bushy postsynaptic thickening on distal dendritic profiles; the second type (18 out of 106) established asymmetrical synaptic contacts with very slight postsynaptic thickening; and the third type (25 out of 106) of labeled terminals contained pleomorphic vesicles and established symmetrical synaptic contacts with more proximal dendritic surfaces than the first two types. In addition to the above, labeled dendritic profiles receiving non-labeled asymmetrical and symmetrical synaptic contacts were identified. These findings suggest that the basal forebrain cholinergic system establishes a variety of synaptic connections in the RT and influences cortical activity indirectly via thalamocortical pathways, as well as via direct projections to the cortex.

A population of cholinergic neurons is present in the macaque monkey thalamus

Three new cholinergic markers were employed to study the cholinergic innervation in the thalamus of adult macaque monkeys. They were: two antibodies against choline acetyltransferase (ChAT), one polyclonal and one monoclonal; and a polyclonal antibody against the vesicular transporter of acetylcholine (VAChT), a powerful new marker that colocalizes with ChAT. This approach led to an unexpected finding: the three antibodies positively immunostained a population of neurons in the paracentral nucleus. The immunostained cells are confined to the dorsal region of this nucleus along its rostrocaudal extent. Measurement of the somatic areas of the immunostained neurons indicated that they correspond to a population of large neurons thought to be projection neurons. Because dorsal paracentral neurons are known to project to the dorsal striatum and specific cortical areas involved in visual and visuomotor mechanisms, these structures might be modulated by cholinergic thalamic neurons.

Diffuse transmission by acetylcholine in the CNS

Recent immunoelectron microscopic studies have revealed a low frequency of synaptic membrane differentiations on ACh (ChAT-immunostained) axon terminals (boutons or varicosities) in adult rat cerebral cortex, hippocampus and neostriatum, suggesting that, besides synaptic transmission, diffuse transmission by ACh prevails in many regions of the CNS. Cytological analysis of the immediate micro-environment of these ACh terminals, as well as currently available immunocytochemical data on the cellular and subcellular distribution of ACh receptors, is congruent with this view. At least in brain regions densely innervated by ACh neurons, a further aspect of the diffuse transmission paradigm is envisaged: the existence of an ambient level of ACh in the extracellular space, to which all tissue elements would be permanently exposed. Recent experimental data on the various molecular forms of AChE and their presumptive role at the neuromuscular junction support this hypothesis. As in the peripheral nervous system, degradation of ACh by the prevalent G4 form of AChE in the CNS would primarily serve to keep the extrasynaptic, ambient level of ACh within physiological limits, rather than totally eliminate ACh from synaptic clefts. Long-lasting and widespread electrophysiological effects imputable to ACh in the CNS might be explained in this manner. The notions of diffuse transmission and of an ambient level of ACh in the CNS could also be of clinical relevance, in accounting for the production and nature of certain cholinergic deficits and the efficacy of substitution therapies.

Cholinergic synapses in human cerebral cortex: an ultrastructural study in serial sections

Cholinergic axons in the human cerebral cortex were analyzed by electron microscopy. Choline acetyltransferase (ChAT) immunoreactivity was used to identify cholinergic axons in samples of anterior temporal lobe removed at surgery. A systematic survey of labeled axon varicosities, visualized in complete serial sections, showed that 67% of all varicosities formed identifiable synaptic specializations. These synapses were usually symmetric and quite small, often present in only one to two serial sections. However, an occasional synapse was asymmetric and larger, seen in five to seven serial sections. The postsynaptic processes at cholinergic synapses were often identified as spiny dendrites or spines. The existence of cholinergic axons in the human cerebral cortex has been demonstrated in numerous studies. Our findings provide the first ultrastructural evidence that these axons make synaptic contact with cortical neurons in the human brain.

Neuronal and synaptic organization of the centromedian nucleus of the monkey thalamus: a quantitative ultrastructural study, with tract tracing and immunohistochemical observations

The ultrastructure of the centromedian nucleus of the monkey thalamus was analysed qualitatively and quantitatively and projection neurons, local circuit neurons, and synaptic bouton populations identified. Projection neurons were mostly medium-sized, with oval-fusiform or polygonal perikarya, few primary dendrites, and frequent somatic spines; local circuit neurons were smaller. Four basic types of synaptic boutons were distinguished: (1) Small- to medium-sized boutons containing round vesicles (SR) and forming asymmetric contacts, identified as corticothalamic terminals. (2) Heterogeneous medium-sized boutons with asymmetric contacts and round vesicles, similar to the so-called large round (LR) boutons, which were in part of cortical origin. (3) Heterogeneous GAD-positive small- to medium-sized boutons, containing pleomorphic vesicles and forming symmetric contacts (F1 type), which included pallidothalamic terminals. (4) Presynaptic profiles represented by GAD-positive vesicle-containing dendrites of local circuit neurons. Complex synaptic arrangements, serial synapses and triads with LR and SR boutons engaging all parts of projection neuron dendrites and somata, were seen consistently, whereas classical glomeruli were infrequent. LR and SR boutons also established synapses on dendrites of local circuit neurons. F1 boutons established synapses on projection neuron somata, dendrites and initial axon segments. Compared to other previously studied motor-related thalamic nuclei, differences in synaptic coverage between proximal and distal projection neuron dendrites were less pronounced, and the density of synapses formed by local circuit dendrites on projection neuron dendrites was lower. Thus, compared to other thalamic nuclei, the overlap of different inputs was higher on monkey centromedian cells, and centromedian inhibitory circuits displayed a different organization.

Cholinergic synaptic circuitry in the macaque prefrontal cortex

Surprisingly little is known about the synaptic architecture of the cholinergic innervation in the primate cerebral cortex in spite of its acknowledged relevance to cognitive processing and Alzheimer's disease. To address this knowledge gap, we examined serially sectioned cholinergic axons in supra- and infragranular layers of the macaque prefrontal cortex by using an antibody against the acetylcholine synthesizing enzyme, choline acetyltransferase (ChAT). The tissue bound antibody was visualized with both immunoperoxidase and silver-enhanced diaminobenzidine sulfide (SEDS) techniques. Both methods revealed that cholinergic axons make synapses in all cortical layers and that these synapses are exclusively symmetric. Cholinergic axons formed synapses primarily on dendritic shafts (70.5%), dendritic spines (25%), and, to a lesser extent, cell bodies (4.5%). Both pyramidal neurons and cells exhibiting the morphological features of GABAergic cells were targets of the cholinergic innervation. Some spiny dendritic shafts received multiple, closely spaced synapses, suggesting that a subset of pyramidal neurons may be subject to a particularly strong cholinergic influence. Analysis of synaptic incidence of cholinergic profiles in the supragranular layers of the prefrontal cortex by the SEDS technique revealed that definitive synaptic junctions were formed by 44% of the cholinergic boutons. An unexpected finding was that cholinergic boutons were frequently apposed to spines and small dendrites without making any visible synaptic specializations. These same spines and dendrites often received asymmetric synapses, presumably of thalamocortical or corticocortical origin. Present ultrastructural findings suggest that acetylcholine may have a dual modulatory effect in the neocortex: one through classical synaptic junctions on dendritic shafts and spines, and the other through nonsynaptic appositions in close vicinity to asymmetric synapses. Further physiological studies are necessary to test the hypothesis of the nonsynaptic release of acetylcholine in the cortex.

Acetylcholinesterase histochemistry in the macaque thalamus reveals territories selectively connected to frontal, parietal and temporal association cortices

The patterns of histochemical staining for acetylcholinesterase (AChE) activity in the macaque thalamus were analyzed and compared with the distribution of cells and terminals labeled from injections of axonal tracers in the dorsolateral and orbital prefrontal cortex, in area 7a of the posterior parietal cortex and in the polysensory cortex of the superior temporal sulcus. AChE histochemistry is very useful in delineating the thalamic nuclei connected with the association cortex and in uncovering thalamic subdivisions that are barely evident on cytoarchitectonic grounds. Moreover, AChE activity reveals previously unrecognized heterogeneities within several thalamic nuclei, like the ventral anterior (VA), where a new ventromedial subdivision (VAvm) is described, the medial pulvinar (PulM) or the mediodorsal nucleus (MD). In this nucleus three distinct chemical domains are present: the medial, ventral and lateral sectors characterized by low, moderate and high AChE activities, respectively. The staining pattern of the lateral sector is markedly heterogeneous with patches of intense AChE activity surrounded by a moderately stained matrix. The MD medial sector is connected with the orbitofrontal cortex, whereas the AChE-rich patches in the lateral sector are selectively connected with the dorsolateral prefrontal, parietal and temporal association cortices. In the PulM, a dorsomedial AChE-rich patch is selectively connected with the orbitofrontal cortex, whereas the surrounding territory, which shows moderate AChE activity, is preferentially connected with the parietal and temporal cortices. Chemically specific domains in the anterior, ventral anterior, midline, and intralaminar thalamic nuclei are also connected with the examined association cortices. These findings indicate that the topographic patterns of the thalamo-cortical connections of primate association areas conform to the chemical architecture of the thalamus. This implies that because each cortical area is connected to a particular set of thalamic regions, the influence of the thalamus on cortical function is exclusive for each area, highly diverse among the various association areas, and subject to a wide range of modulation at the thalamic level.

Cerebellar-responsive neurons in the thalamic ventroanterior-ventrolateral complex of rats: light and electron microscopy

The morphology and synaptic organization of neurons in the ventroanterior-ventrolateral nucleus of rats was examined using in vivo intracellular staining techniques. Neurons were characterized electrophysiologically based on intrinsic membrane properties and synaptic responses to stimulation of motor cortex and cerebellar nuclei, as described in the companion paper. Cerebellar-responsive neurons were stained intracellularly with either horseradish peroxidase or biocytin. All stained ventroanterior-ventrolateral nucleus neurons were identified as thalamocortical neurons on anatomical (and often electrophysiological) grounds, consistent with previous findings that rat ventroanterior-ventrolateral nucleus is interneuron-sparse. Ventroanterior-ventrolateral nucleus neurons had three to eight thick primary dendrites. Proximal dendrites often exhibited a tufted branching pattern, from which many thinner, higher order dendrites arose. Dendrites branched to form a funnel-like infiltration of the neuropil that resulted in a spherical, roughly homogeneous dendritic field. The axon originated from the cell body or a proximal dendrite and coursed laterally and dorsally to innervate motor cortex. One to five axon collaterals were emitted in the rostral dorsolateral sector of the thalamic reticular nucleus; collaterals were not observed in the ventroanterior-ventrolateral nucleus or other nuclei in dorsal thalamus. The synaptic organization of the ventroanterior-ventrolateral nucleus was examined with electron microscopy, including two intracellularly labeled ventroanterior-ventrolateral nucleus neurons that were shown electrophysiologically to receive monosynaptic inputs from the cerebellum. The neuropil of rat ventroanterior-ventrolateral nucleus lacked the complexity and diversity found in corresponding thalamic nuclei of felines and primates, due to the paucity of interneurons. Vesicle-containing dendrites, dendrodendritic synapses and glomeruli were not observed. Three broad classes of presynaptic terminals were identified. (1) Small round boutons: small boutons containing densely-packed, small round vesicles that formed asymmetric synapses predominantly with the distal dendrites of thalamocortical neurons. These were the most prevalent type of bouton in the ventroanterior-ventrolateral nucleus (78% of presynaptic elements) and likely arose from the cerebral cortex. (2) Large round boutons: large terminals with loosely packed small round vesicles that made multiple asymmetric synapses with proximal and intermediate dendrites. Large round boutons comprised 8% of the neuropil, and likely arose from the cerebellar nuclei. (3) Medium size boutons with pleomorphic vesicles: medium-sized profiles containing pleomorphic vesicles that formed symmetric synapses with proximal, intermediate and distal dendrites and, less frequently, with cell bodies.(ABSTRACT TRUNCATED AT 400 WORDS)

The organization of projections from the mediodorsal nucleus of the thalamus to orbital and medial prefrontal cortex in macaque monkeys

The organization of interconnections between the mediodorsal nucleus of the thalamus (MD) and the orbital and medial prefrontal cortex and the agranular insular cortex in the monkey was studied by retrograde and anterograde tracing techniques. In addition to the magnocellular and parvicellular divisions of MD, three other subdivisions can be recognized on the basis of myeloarchitecture, cytoarchitecture, and connections. The first two of these represent a parcellation of the magnocellular division into a lateral, fiber-rich MD pars fibrosa and a medial, poorly myelinated MD pars paramediana adjacent to the midline. The third is a small, poorly myelinated area located at the caudomedial and dorsal edges of MD; it is referred to as MD pars caudodorsalis. MD pars fibrosa is reciprocally interconnected primarily with areas 11, 12 and 13 in the central and lateral part of the orbital cortex. There is a general organization within this projection, with the rostrocaudal axis of the cortex represented from dorsal to ventral in the pars fibrosa, and the mediolateral cortical axis represented from medial to lateral. Cells that project to area 12 also extend laterally into the adjacent pars parvicellularis. MD pars paramediana is more heavily interconnected with the caudal and medial portions of the orbital region, particularly the agranular insular areas and the caudal parts of areas 13 and 14. Cells that project to two caudal areas, 13a and Iad, do not fit with the general organization, in that they are located in the dorsomedial parts of the pars fibrosa and pars paramediana, where they overlap with cells that project to area 14. The pars fibrosa and pars paramediana receive inputs from areas of the ventral forebrain such as the amygdala, piriform (olfactory) cortex, and entorhinal cortex, which project directly to the orbital and agranular insular cortex, as well as from the ventral pallidum. MD pars caudodorsalis is reciprocally interconnected with areas 14, 24, and 32 on the medial surface of the prefrontal cortex. In this part of the nucleus the dorsoventral axis of the medial prefrontal cortex is represented from caudal to rostral in the thalamus. The amygdala and other ventral forebrain structures do not send fibers into the pars caudodorsalis, even though some of these structures project directly to the medial prefrontal cortex. Ventral to MD, and separated from it by the internal medullary lamina, a small region was recognized that appears to be comparable to the anteroventral part of the submedial nucleus previously defined in the rat and cat.(ABSTRACT TRUNCATED AT 400 WORDS)