Igor A. Ilinsky

Limbic Thalamus: Structure, Intrinsic Organization, and Connections

Limbic is one of the most elusive, equivocal, and, at the same time, the most useful and widely adopted terms in neuroscience. Its original usage derives from Paul Broca’s (1878) “grand lobe limbique,” referring to the ringlike structures that form the hilus of the cerebral hemisphere, including the hip-pocampal formation and pericallosal cingu-late cortex. The notion of a limbic system, “which may elaborate the functions of central emotion” (Papez, 1937), has since then evolved to include subcortical and neocor-tical structures that are believed to play a role in visceral functions, in the emotional aspects of behavior, and in learning and memory. As discussed in the introduction to this volume by Paul MacLean, the term limbic has been exploited in the anatomical and functional definition of a number of structures and no better term has hitherto emerged as its substitute.

Reevaluation of the primary motor cortex connections with the thalamus in primates.

Six injections (∼1 mm in diameter) of biotinylated dextran amine (BDA) were placed in different locations of the primary motor cortex of the rhesus monkey. Anterograde and retrograde labeling patterns in the thalamus were charted and individual labeled axons traced in continuous serial sections. Both anterograde and retrograde labeling in the thalamus was extensive, spanning several millimeters mediolaterally and including ventral lateral, ventral anterior, centromedian, and centrolateral nuclei. Paracentral, mediodorsal, lateral posterior, and medial pulvinar nuclei were also labeled. Two basic types of corticothalamic axons were identified: small to medium‐width, type 1 axons that formed large terminal fields with small boutons, and thick, type 2 axons that formed small terminal fields with large boutons. Within each group, subtypes were identified based on specific features of the axons and terminals: two subtypes of type 1 axons and four subtypes of type 2 axons. The results revealed multiple modes of corticothalamic connectivity: sparsely distributed type 1 axons, dense plexuses of type 1 axons, type 2 axon terminal fields either singly or in clusters, and mixed plexuses of type 1 and type 2 axons. Only some cells in the plexuses were retrogradely labeled; some plexuses did not contain any labeled neurons, and many retrogradely labeled neurons were in the regions devoid of anterograde labeling. These connectivity patterns differed between thalamic nuclei. The results revealed much more complex relationships between M1 and thalamus than were previously thought to exist. It is suggested that this connectivity is neither of exclusively a feedback nature nor perfectly reciprocal but is subserved by a multitude of channels, most likely originating from different populations of cortical neurons, and feeding into a variety of functionally different neuronal networks, with each processing specific information. J. Comp. Neurol. 457:133–158, 2003.

Motor thalamic circuits in primates with emphasis on the area targeted in treatment of movement disorders

The ventral region of the motor thalamus that receives cerebellar afferents has been and still is the target of stereotactic interventions for movement disorders. According to Hassler, this area includes ventro‐oralis posterior (Vop) and ventral intermedius (Vim) nuclei, although some investigators believe that Vop is associated with the pallidothalamic pathway. We sought to correlate our experimental data on distribution of nigral, pallidal, and cerebellar afferents to the monkey thalamus with Hasslers motor thalamic parcelations. We concluded that Hasslers parcelations retained their value, although some adjustments were needed to relate them to the current neuroanatomic data; particularly, the cerebellothalamic zone that represents the monkey ventral lateral nucleus (VL) corresponds topographically to Hasslers Vop, Vim, and most of Voi. Electron microscopic tracing studies have shown very complex circuitry in this region of the monkey thalamus, as the cerebellar and cortical afferents innervating it are engaged in complex synapses with thalamocortical projection neurons, and this interaction is strongly modulated by local circuit neurons and the input from the reticular thalamic nucleus, which are both inhibitory and γ‐aminobutyric acid (GABA)ergic. Spinothalamic afferents also reach the VL, but this input is less studied in the monkey. The circuitry subserving the activity of thalamocortical projection neurons in the VL should be considered while interpreting the functional data obtained in stereotactic surgery.

Mode of termination of pallidal afferents to the thalamus: a light and electron microscopic study with anterograde tracers and immunocytochemistry in Macaca mulatta.

The mode of termination of individual pallidothalamic fibers in the densicellular subdivision of the ventral anterior thalamic nucleus (VAdc) of Macaca mulatta was analyzed with light and electron microscopy after injections of anterograde tracers in the medial globus pallidus. Three tracers were utilized: tritiated leucine, biotinylated dextran amine, and wheat germ agglutinin conjugated to horseradish peroxidase in combination with postembedding immunocytochemsitry for gamma‐aminobutyric acid (GABA).

High-frequency stimulation of the subthalamic nucleus suppresses experimental resting tremor in the monkey

The effect of high-frequency stimulation of the subthalamic nucleus on parkinsonian-like resting tremor was investigated in two monkeys (Macaca fascicularis). Unilateral tremor of the arm and leg was induced by electrical coagulation of the brainstem area including the substantia nigra and the red nucleus. The tremor was only seen at rest condition with a very stable frequency of 4.46+/-0.59 Hz (mean+/-S.D.). Apomorphine (0.10-0.4 mg/kg, s.c.) completely blocked the tremor, suggesting that it was a dopaminergic-dependent symptom just like the parkinsonian tremor. When the stimulating frequency varied from 20 to 1000 Hz, both mono- and bipolar stimulation (square pulses, 0-5 mA, 0.06 ms) of the subthalamic nucleus suppressed resting tremor in a frequency-dependent manner but monopolar stimulation was more effective. These effects remained stable for more than two years. The present results suggest that the subthalamic nucleus is involved in the control and mechanism of resting tremor and that the high-frequency stimulation of the subthalamic nucleus can be used as an alternative therapy in parkinsonian patients with akinesia, rigidity and resting tremor.

Anatomical distinctions between the two basal ganglia afferent territories in the primate motor thalamus.

In primates, the efferents of the two basal ganglia output structures, medial globus pallidus and substantia nigra pars reticularis, are completely segregated and target different thalamic regions. Despite similarities demonstrated earlier in non-primate species in the functional properties and ultrastructural features of the terminals of the two pathways, the present findings suggest significant differences between thalamic circuits associated with the two systems in primates. The data presented in this report further support the concept on functional and anatomical diversity of the subdivisions of the primate motor thalamus proposed in 1990 by Hinsky.

Organization of projections from the anterior pole of the nucleus reticularis thalami (NRT) to subdivisions of the motor thalamus: Light and electron microscopic studies in the Rhesus monkey

Projections to the motor‐related thalamic nuclei from the anterior pole of the reticular thalamic nucleus (NRT) were studied after injections of biotinylated dextran amine and wheat germ agglutinin conjugated horseradish peroxidase at light and electron microscopic levels, respectively. Each injection resulted in anterograde labeling in the three subdivisions of the ventral anterior nucleus (pars parvicellularis, VApc; pars densicellularis, VAdc; and pars magnocellularis, VAmc) and in the ventral lateral nucleus (VL). NRT fibers had beaded shapes and coursed in a posterior direction giving rise to relatively diffuse terminal plexuses. The average size of the beads (0.7μm2) and their density per 100μm of fiber length (23.7–25.7) were similar between the nuclei studied. At the electron microscopic level, anterogradely labeled boutons displayed positive immunoreactivity for γ‐aminobutyric acid (GABA), contained pleomorphic synaptic vesicles, and formed relatively long (∼0.4μm) symmetric synaptic contacts. Usually, a single terminal formed synapses on more than one postsynaptic structure. Synaptic contacts were on projection and local circuit neurons and targeted mainly their distal dendrites. In the VAmc, synapses on local circuit neurons composed 48% of the total sample, in the VAdc/VApc and in the VL the proportion was higher, 65% and 62%, respectively. The results suggest that the input from the anterior pole of the monkey reticular nucleus to the motor‐related thalamic nuclei is organized differently from what is known on the organization of connections of NRT with sensory thalamic nuclei in other species in that the terminal fields of individual fibers are diffuse rather than focal and that at least 50% of synapses are established on GABAergic local circuit neurons. J. Comp. Neurol. 409:369–384, 1999.

Structural and Connectional Diversity of the Primate Motor Thalamus: Experimental Light and Electron Microscopic Studies in the Rhesus Monkey

In this study we analyzed neuronal and synaptic organization of nigral and cerebellar afferent territories in the rhesus monkey motor thalamus using a variety of electron microscopic techniques. Significant qualitative and quantitative differences were revealed between the two regions with respect to features of nerve cells, their relationships to one another, and organization of afferent inputs on them. The data suggest that information arriving to the thalamus via nigral and cerebellar afferents is processed by different mechanisms.

Nucleus reticularis thalami connections with the mediodorsal thalamic nucleus : a light and electron microscopic study in the monkey

Wheat germ agglutinin conjugated horseradish peroxidase (WGA-HRP) and biotinylated dextran amine (BDA) were used as tracers to study nucleus reticularis (NRT) connections with the mediodorsal nucleus (MD). Injections of WGA-HRP in the MD resulted in retrograde labeling of cells in the anteromedial segment of the NRT, the so-called rostral NRT pole. Injections of WGA-HRP and BDA in this NRT region resulted in dense anterograde labeling in the MD. Labeled NRT fibers gave off several collaterals to different MD regions ending with terminal plexuses of thin varicose fibers. In the neuropil, the varicosities were distributed at random, and no tendency to form pericellular baskets was noted. Postembedding immunocytochemistry for GABA was performed on the tissue containing anterograde WGA-HRP label for identification of NRT boutons under electron microscope. The double-labeled boutons were of small to medium size, contained a large number of pleomorphic vesicles, few mitochondria, and formed multiple symmetric synaptic contacts. The number of contacts established by one bouton ranged from 1 to 4 with an average of 1.8 per bouton. About 60% of these boutons made synapses on distal dendrites of GABAergic local circuit neurons; 33% of synaptic contacts were on distal dendrites of thalamocortical neurons, and the rest on their proximal dendrites and soma. NRT boutons were also found in serial synapses and triads. The results demonstrate that the NRT input to the MD is organized so that a single fiber innervates; different MD regions and its terminals form numerous synaptic contacts mostly on the distal dendrites of a large number of local circuit neurons and projection neurons.

Thalamic terminal fields of individual axons from the ventral part of the dentate nucleus of the cerebellum in Macaca mulatta

This study examined organization of the projection from the dentate nucleus of the cerebellum to the ventral lateral nucleus (VL) of the thalamus in Macaca mulatta. Small injections of biotinylated dextran amine were placed in the ventral parts of dentate nuclei. The distribution of all contralateral terminal fields in the thalamus was charted, and representative individual axons that terminated in the VL were traced in serial sections under the light microscope. These axons were reconstructed with all their branches and terminal fields in the thalamus. The geometry and size of the terminal fields as well as the number and distribution of boutons and neurons in them were analyzed. The terminal fields of all labeled axons were distributed widely over the VL either singly or in clusters. Two types of axons were found: simple axons formed only one terminal field and complex axons formed multiple terminal fields at a distance. Individual terminal fields were focal, had the form of flattened discs, and generated up to 200 boutons distributed between 10 and 29 nerve cell bodies. These findings suggest that a simple axon activates a small group of neurons at one site. The complex axons, in turn, influence similar size cell groups at different VL locations. The total number of boutons generated by a single complex axon was up to 300. Future studies should determine whether simple axons could be branches of complex axons that took off below the thalamus. The results reveal a complex organization of the input from the ventral dentate to the VL that only partially fits into the traditional concept of somatotopic organization of the nucleus. J. Comp. Neurol. 421:412–428, 2000.