Hypothalamic and ventral thalamic projections to the superior colliculus in the cat

Brainstem sources of input to the central mesencephalic reticular formation in the macaque

Physiological studies indicate that the central mesencephalic reticular formation (cMRF) plays a role in gaze changes, including control of disjunctive saccades. Neuroanatomical studies have demonstrated strong interconnections with the superior colliculus, along with projections to extraocular motor nuclei, the preganglionic nucleus of Edinger-Westphal, the paramedian pontine reticular formation, nucleus raphe interpositus, medullary reticular formation and cervical spinal cord, as might be expected for a structure that is intimately involved in gaze control. However, the sources of input to this midbrain structure have not been described in detail. In the present study, the brainstem cells of origin supplying the cMRF were labeled by retrograde transport of tracer (wheat germ agglutinin conjugated horseradish peroxidase) in macaque monkeys. Within the diencephalon, labeled neurons were noted in the ventromedial nucleus of the hypothalamus, pregeniculate nucleus and habenula. In the midbrain, labeled cells were found in the substantia nigra pars reticulata, medial pretectal nucleus, superior colliculus, tectal longitudinal column, periaqueductal gray, supraoculomotor area, and contralateral cMRF. In the pons they were located in the paralemniscal zone, parabrachial nucleus, locus coeruleus, nucleus prepositus hypoglossi and the paramedian pontine reticular formation. Finally, in the medulla they were observed in the medullary reticular formation. The fact that this list of input sources is very similar to those of the superior colliculus supports the view that the cMRF represents an important gaze control center.

Altered effective connectivity within an oculomotor control network in individuals with schizophrenia

Rapid inhibition or modification of actions is a crucial cognitive ability, which is impaired in persons with schizophrenia (SZP). Primate neurophysiology studies have identified a network of brain regions that subserves control over gaze. Here, we examine effective connectivity within this oculomotor control network in SZP and healthy controls (HC). During fMRI, participants performed a stop-signal task variant in which they were instructed to saccade to a visual target (no-step trials) unless a second target appeared (redirect trials); on redirect trials, participants were instructed to inhibit the planned saccade and redirect to the new target. We compared functional responses on redirect trials to no-step trials and used dynamic causal modelling (DCM) to examine group differences in network effective connectivity. Behaviorally, SZP were less efficient at inhibiting, which was related to their employment status. Compared to HC, they showed a smaller difference in activity between redirect trials and no-step trials in frontal eye fields (FEF), supplementary eye fields (SEF), inferior frontal cortex (IFC), thalamus, and caudate. DCM analyses revealed widespread group differences in effective connectivity across the task, including different patterns of self-inhibition in many nodes in SZP. Group differences in how effective connectivity was modulated on redirect trials revealed differences between the FEF and SEF, between the SEF and IFC, between the superior colliculus and the thalamus, and self-inhibition within the FEF and caudate. These results provide insight into the neural mechanisms of inefficient inhibitory control in individuals with schizophrenia.

Hypothalamic Projections to the Optic Tectum in Larval Zebrafish

The optic tectum of larval zebrafish is an important model for understanding visual processing in vertebrates. The tectum has been traditionally viewed as dominantly visual, with a majority of studies focusing on the processes by which tectal circuits receive and process retinally-derived visual information. Recently, a handful of studies have shown a much more complex role for the optic tectum in larval zebrafish, and anatomical and functional data from these studies suggest that this role extends beyond the visual system, and beyond the processing of exclusively retinal inputs. Consistent with this evolving view of the tectum, we have used a Gal4 enhancer trap line to identify direct projections from rostral hypothalamus (RH) to the tectal neuropil of larval zebrafish. These projections ramify within the deepest laminae of the tectal neuropil, the stratum album centrale (SAC)/stratum griseum periventriculare (SPV), and also innervate strata distinct from those innervated by retinal projections. Using optogenetic stimulation of the hypothalamic projection neurons paired with calcium imaging in the tectum, we find rebound firing in tectal neurons consistent with hypothalamic inhibitory input. Our results suggest that tectal processing in larval zebrafish is modulated by hypothalamic inhibitory inputs to the deep tectal neuropil.

The efference cascade, consciousness, and its self: naturalizing the first person pivot of action control

The 20 billion neurons of the neocortex have a mere hundred thousand motor neurons by which to express cortical contents in overt behavior. Implemented through a staggered cortical "efference cascade" originating in the descending axons of layer five pyramidal cells throughout the neocortical expanse, this steep convergence accomplishes final integration for action of cortical information through a system of interconnected subcortical way stations. Coherent and effective action control requires the inclusion of a continually updated joint "global best estimate" of current sensory, motivational, and motor circumstances in this process. I have previously proposed that this running best estimate is extracted from cortical probabilistic preliminaries by a subcortical neural "reality model" implementing our conscious sensory phenomenology. As such it must exhibit first person perspectival organization, suggested to derive from formating requirements of the brain's subsystem for gaze control, with the superior colliculus at its base. Gaze movements provide the leading edge of behavior by capturing targets of engagement prior to contact. The rotation-based geometry of directional gaze movements places their implicit origin inside the head, a location recoverable by cortical probabilistic source reconstruction from the rampant primary sensory variance generated by the incessant play of collicularly triggered gaze movements. At the interface between cortex and colliculus lies the dorsal pulvinar. Its unique long-range inhibitory circuitry may precipitate the brain's global best estimate of its momentary circumstances through multiple constraint satisfaction across its afferents from numerous cortical areas and colliculus. As phenomenal content of our sensory awareness, such a global best estimate would exhibit perspectival organization centered on a purely implicit first person origin, inherently incapable of appearing as a phenomenal content of the sensory space it serves.

The dorsal tectal longitudinal column (TLCd): a second longitudinal column in the paramedian region of the midbrain tectum

The tectal longitudinal column (TLC) is a longitudinally oriented, long and narrow nucleus that spans the paramedian region of the midbrain tectum of a large variety of mammals (Saldaña et al. in J Neurosci 27:13108–13116, 2007). Recent analysis of the organization of this region revealed another novel nucleus located immediately dorsal, and parallel, to the TLC. Because the name “tectal longitudinal column” also seems appropriate for this novel nucleus, we suggest the TLC described in 2007 be renamed the “ventral tectal longitudinal column (TLCv)”, and the newly discovered nucleus termed the “dorsal tectal longitudinal column (TLCd)”. This work represents the first characterization of the rat TLCd. A constellation of anatomical techniques was used to demonstrate that the TLCd differs from its surrounding structures (TLCv and superior colliculus) cytoarchitecturally, myeloarchitecturally, neurochemically and hodologically. The distinct expression of vesicular amino acid transporters suggests that TLCd neurons are GABAergic. The TLCd receives major projections from various areas of the cerebral cortex (secondary visual mediomedial area, and granular and dysgranular retrosplenial cortices) and from the medial pretectal nucleus. It densely innervates the ipsilateral lateral posterior and laterodorsal nuclei of the thalamus. Thus, the TLCd is connected with vision-related neural centers. The TLCd may be unique as it constitutes the only known nucleus made of GABAergic neurons dedicated to providing massive inhibition to higher order thalamic nuclei of a specific sensory modality.

Honeycomb-like structure of the intermediate layers of the rat superior colliculus: afferent and efferent connections

There is increasing evidence that acetylcholinesterase is organised in a lattice-like fashion in the intermediate layers of the mammalian superior colliculus. In a recent study, we described this organisation in rat by showing that it comprises a well formed honeycomb-like lattice with about 100 cylindrical compartments or modules occupying both the intermediate collicular layers. Considering this enzyme domain as a reference marker for comparing the organisation of collicular input-output systems, the present study investigates whether the principal sensori-motor systems in intermediate layers also have honeycomb-like arrangements. In 33 animals, the distributions of afferents (visual from extrastriate cortex; somatic from the primary somatosensory cortex, the trigeminal nucleus and the cervical spinal cord) and efferents (cells of origin of the crossed descending bulbospinal tract and uncrossed pathway to the pontine gray, the ascending system to the medial dorsal thalamus) were examined in a tangential plane following applications of horseradish peroxidase-wheatgerm agglutinin conjugate (used as an anterograde and retrograde tracer). In 22 of the 33 rats, axonal tracing was made within single tangential sections also stained for cholinesterasic activity in order to compare the neuron profiles with the cholinesterasic lattice.The results show that these afferent and efferent systems are also organised in honeycomb-like networks. Moreover, those related to the cortical, trigeminal and some of the spinal afferents are aligned with the cholinesterasic lattice. Likewise most of colliculo-pontine, colliculo-bulbospinal and half of colliculo-diencephalic projecting cells also tend to be in spatial register with the enzyme lattice. This indicates that the honeycomb-like arrangement is a basic architectural plan in the superior colliculus for the organisation of both acetylcholinesterase and major sensori-motor systems for orientation.

Honeycomb-like structure of the intermediate layers of the rat superior colliculus, with additional observations in several other mammals: AChE patterning

The aim of the present study was to reinvestigate the stereometric pattern of acetylcholinesterase (AChE) activity staining in the intermediate layers of the superior colliculus in several mammalian species. A pioneering study in the cat and the monkey by Graybiel (1978) stressed the regular arrangement of AChE staining in the deep collicular layers. According to her description, made in the frontal plane, the enzyme was arranged in a mediolateral series of patches, the cores of which tended to line up in the longitudinal axis of the structure, so they formed roughly parallel bands. As exhaustive a description as possible of the AChE distribution was undertaken in the rat by compiling observations in the frontal, sagittal, and tangential planes. It emerged that AChE-positive elements are organized in the form of a conspicuous honeycomb-like network that is divided into about 100 rounded compartments, over virtually the full extent of the intermediate layers. The generality of the rat model was then tested in other rodents such as mouse and hamster and also in cat and monkey. For these species we resorted to a single tangential cutting plane, which proved to be more appropriate for disclosing such a modular arrangement. The data revealed that in all species AChE staining followed the same architectural plan and identified the striking similarity in the number of compartments that compose the various honeycomb-like lattices. In conclusion, the present findings support a unified model of the AChE arrangement within the intermediate layers of the mammalian colliculus; the model comprehensively incorporates the classical description of the patchy and stripy features of the enzyme distribution. We hypothesize here that the modular AChE arrangement might be the anatomical basis for collicular vectorial encoding of orienting movements.

Role of the superior colliculus in the motor effects of cannabinoids and dopamine

We studied the cellular distribution of CB1 cannabinoid receptors in the superior colliculus of the rat using an antibody raised against the N-terminal of the receptor. The effect of unilateral cannabinoid receptor stimulation in the intermediate layers of the superior colliculus on rotational behavior in rats was also explored. The antibody against CB1 receptors outlined the crossed descending system of the superior colliculus (predorsal bundle output system) as well as the collicular commisure. The potent cannabinoid agonist CP55,940 (5 microgram/0.25 microliter) induced strong contralateral turning when microinjected unilaterally into the lateral intermediate layers of the superior colliculus. The levels of turning obtained with the intracollicular administration of the cannabinoid were comparable to the highest levels obtained with dopamine agonists in the basal ganglia. The D(2) dopamine agonist quinpirole or the D(1) dopamine agonist SKF82958 reversed this contralateral rotation but failed to affect motor behavior on their own. A new motor pathway for cannabinoids is discussed.

Hippocampal theta-related cellular activity in the superior colliculus of the urethane-anesthetized rat

In the present study, 93 cells in the superior colliculus (SC) were recorded extracellularly during the simultaneous occurrence of spontaneous theta field activity, sensory-induced (tail pinch) theta field activity, and large amplitude irregular (LIA) field activity, recorded from an electrode located in the stratum moleculare of the hippocampal formation (HPC). The effect of the intravenous administration of atropine sulfate (ATSO4) was also tested on SC cellular activity. The field activities of theta and LIA were recorded from all layers of the SC and were found to be temporally coherent with the same activities recorded simultaneously from the HPC, during all conditions tested. By using the criteria of Colom and Bland (1987) for the classification of theta-related cells, 75 of 93 cells (81%) were found to be related to the generation of theta field activity in the HPC and 18 of 93 (19%) were nonrelated. All cells recorded discharged in a tonic, nonrhythmic pattern during the theta HPC field states. Of the 75 theta-related cells, 61 (81%) were classified as tonic theta-ON cells and 14 (19%) as tonic theta-OFF cells. Although these cell types were found in all three layers of the SC, the majority of tonic theta-ON cells were recorded in the intermediate layer, and the tonic theta-OFF cells were dispersed evenly between the intermediate layer and the deep layer of the SC. The intravenous administration of ATSO4 abolished theta field activity in the HPC and SC, and the theta-related increase in the discharge rate of all tonic theta-ON cells tested. However, the same treatment did not have any effect on the discharge properties of tonic theta-OFF cells. The same stimuli that resulted in the inhibition of the discharge rates of these cells (tail pinch and electrical stimulation of the PH) in the predrug condition did so after the administration of ATSO4.

Patterns of connections between zona incerta and brainstem in rats

To understand better the organisation of zona incerta of the thalamus, this study has examined the patterns of connections that this nucleus has with various nuclei of the brainstem. Injections of biotinylated dextran or cholera toxin subunit B were made into the dorsal raphe, midbrain reticular nucleus, pedunculopontine tegmental nucleus, periaqueductal grey matter, pontine reticular nucleus, substantia nigra, superior colliculus, and ventral tegmental area of Sprague-Dawley rats, and their brains were processed by using standard tracer-detection methods. In general, our results show that zona incerta forms the major zone in the thalamus where these ascending brainstem axons terminate and from which descending axons that travel back to these same brainstem centres originate. These incertal inputs and outputs are limited largely to a distinct sector of zona incerta, the dorsal sector. An exception to this pattern is evident in the incertal projection to the deep layers of the superior colliculus; this projection, unlike all of the others, arises from cells in the ventral sector of zona incerta. Our results also show little evidence for a well-defined topography of projection between the brainstem and the zona incerta. For instance, small injections into each brainstem nucleus result in labelled terminals and in cells spread throughout much of the dorsal sector of zona incerta, with no local zone of concentration within the sector. Again, an exception to this pattern is seen in the incertal projection to the superior colliculus. This projection, unlike the others, shows a clear topographical organisation: A medial-lateral shift in the injection site in the colliculus results in a lateral-medial shift in the position of labelled cells in zona incerta. Curiously, even though the incertal projection to the colliculus appears to be mapped, the collicular projection back to zona incerta is not mapped. In conclusion, then, a number of brainstem nuclei (except for the deep collicular layers) have strong and overlapping connections within the same sector of zona incerta. This convergence of many functionally diverse brainstem afferents within zona incerta places this nucleus in a strategic position to sample the general activity of the brainstem and, perhaps, acts as a relay of this information to higher centres, such as the dorsal thalamic relay nuclei and the cerebral hemispheres.

Topographic and laminar organizations of the incertocortical pathway in rats

The topographic and laminar organizations of the projection system from the zona incerta to the neocortex were studied by using both retrograde and anterograde methods in the rat. Injections of retrograde fluorescent tracers into different cortical areas revealed that the incertocortical projection neurons have a rough topographic organization with respect to their cortical targets. Furthermore, the incertocortical projecting neurons were found mainly in the dorsal and rostral subdivisions of the zona incerta, and none were found in the ventral subdivision. In cases which included three different fluorescent tracers injected into the frontal, the parietal and the occipital cortices, retrogradely single-labelled cells were found intermingled within the dorsal zona incerta. Very few double-labelled cells were noted, and triple-labelled cells were absent. Injections of anterograde tracers into the dorsal zona incerta demonstrate that labelled fibres traverse the striatum and terminate most densely in the outer half of layer I of the neocortex. The density of incertocortical terminals was greatest in the somatosensory cortex, while the innervation of visual cortical areas was sparse. Very fine and sparse bouton-like swellings of labelled incertocortical fibres were found running parallel along the pial surface. Since it has recently been shown that the incertocortical projections derive from GABAergic neurons, the present results suggest that the diffuse and roughly topographic projection from the zona incerta to the cerebral cortex may play an inhibitory role in widespread areas of cerebral cortex. This inhibitory action may preferentially target the distal dendrites of cortical neurons, since the majority of incertocortical terminals were found in the outer part of layer I of the neocortex.

Organization of the zona incerta in the macaque: an electron microscopic study

BACKGROUND

The zona incerta (ZI) receives projections from many telencephalic and brainstem structures. On the basis of its connectivity and physiology, this nucleus has been implicated in the control of saccadic eye movements. Because of the complexity of its afferent signals and its simple efferent signal, there must be much local interaction within the ZI to integrate these various afferents. The purpose of this study was to investigate, at the ultrastructural level, whether the ZI contains the anatomical substrata which could subserve the control of eye movements.

METHODS

Blocks of tissue from the ZI of macaque monkeys were prepared for electron microscopy using standard techniques. Some of these animals were taken specifically for electron microscopy. Others had received injections of tracer substances and were prepared for electron microscopy subsequent to tracer visualization.

RESULTS

Cell bodies of medium-large neurons were found in our preparations. They have large nucleoli and relatively small volumes of karyoplasm. Cell bodies and dendrites of all sizes have many synaptic contacts. Three types of synaptic profiles were found, designated Types 1, 2, and 3. Type 1 profiles are symmetrical and contact cell bodies and small dendrites. Type 2 profiles are thought to be presynaptic dendrites and may have symmetrical or asymmetrical contacts. Type 3 profiles are asymmetrical and primarily contact small dendrites. Many synapses contacted vesicle-containing profiles. In some cases, it was clear that these profiles participated in serial synapses on presumptive presynaptic dendrites. Other profiles appeared to be axoaxonic contacts.

CONCLUSIONS

Afferent and efferent signals are likely to be modulated extensively within the ZI. Therefore, there needs to be complex interactions between neuronal elements of the ZI and its afferents. This study demonstrates that this nucleus possesses the structural substrata to subserve diverse roles, such as the gating of saccadic eye movements.

Reciprocal connections between the zona incerta and the pretectum and superior colliculus of the cat

The goal of the present experiments was to examine the relationships of the zona incerta with two structures associated with visuomotor behavior, the superior colliculus and pretectum. The experiments were carried out in the cat, a species commonly used in studies of visuomotor integration, and utilized wheat germ agglutinin horseradish peroxidase and biocytin as retrograde and anterograde neuronal tracers. Retrograde axonal transport demonstrated that most cells in the ventral subdivision of the zona incerta project to the superior colliculus. Anterograde tracers demonstrated that the incertotectal terminal field is most dense in the intermediate gray layer, which is the primary source of the descending pathway from the superior colliculus to brainstem gaze centers. Further experiments showed that scattered cells within the intermediate gray layer give rise to a reciprocal pathway that terminates in both the dorsal and ventral subdivisions of the zona incerta. The distribution of both labeled incertotectal cells and tectoincertal terminals extends dorsolateral to the zona incerta proper, between the reticular thalamic nucleus and the external medullary lamina. Electron microscopic examination of labeled tectoincertal terminals demonstrated that they contain mainly spherical vesicles and have slightly asymmetric to symmetric synaptic densities. Labeled terminals were observed contacting labeled cells in the zona incerta, suggesting that the reciprocal pathway may be monosynaptic. The zona incerta is also reciprocally interconnected with the pretectum. The anterior pretectal nucleus provides a dense projection to the ventral part of the zona incerta and receives a sparse reciprocal projection. The posterior pretectal nucleus and nucleus of the optic tract may also project to the zona incerta. The pretectoincertal fibers form terminals that contain primarily spherical vesicles and make distinctly asymmetric synaptic contacts. In summary, these results indicate that the deep layers of the superior colliculus, which are important for controlling saccades, are the target of a projection from the ventral subdivision of the zona incerta. Like the substantia nigra, the zona incerta may play a permissive role in the tectal initiation of saccadic eye movements. The incertotectal terminal field in the cat is less dense than that observed previously in the rat, suggesting species differences in the development of this pathway. An additional finding of this study is that one of the main sources of input to these incertotectal cells is the anterior pretectal nucleus. This pretectal incertal tectal pathway is likely to play a role in the guidance of tectally initiated saccades by somatosensory stimuli.

Diencephalic connections of the superior colliculus in the hedgehog tenrec

Using different tracer substances the pathways connecting the superior colliculus with the diencephalon were studied in the Madagascan hedgehog tenrec (Echinops telfairi), a nocturnal insectivore with tiny eyes, a small and little differentiated superior colliculus and a visual cortex with no obvious fourth granular layer. The most prominent tecto-thalamic projection terminated in the ipsilateral dorsal lateral geniculate nucleus. The entire region receiving contralateral retinal afferents was labeled with variable density. In addition, there was a widespread, homogeneously distributed collicular input to the lateralis posterior-pulvinar complex and a distinct tectal projection to the suprageniculate nucleus. The latter projections were bilateral with a clear ipsilateral predominance. Among the intra- and paralaminar nuclei the centralis lateralis complex was most heavily labeled on both sides, followed by the nucleus centralis medialis. The paralamellar portion of the nucleus medialis dorsalis and the nucleus parafascicularis received sparse projections. A clear projection to the nucleus ventralis medialis could not be demonstrated but its presence was not entirely excluded either. There were also projections to medial thalamic nuclei, particularly the reuniens complex and the nucleus paraventricularis thalami. The main tecto-subthalamic target regions were the zona incerta, the dorsal hypothalamus and distinct subdivisons of the ventral lateral geniculate nucleus. These regions also gave rise to projections to the superior colliculus, as did the intergeniculate leaflet. The pathways oriented toward the visual or frontal cortex and the projections possibly involved in limbic and circadian mechanisms were compared with the connectivity patterns reported in mammals with more differentiated brains. Particular attention was given to the tenrec's prominent tecto-geniculate projection, the presumed W- or K-pathway directed toward the supragranular layers.

Reticular thalamic region in the rabbit: organisation of efferents to the superior colliculus

Although it is well-established that the reticular thalamic nucleus provides a strong GABAergic input to the dorsal thalamus, the existence of reticular efferents to other subcortical centres is less certain. In this study, we investigate whether the reticular nucleus projects to a major brainstem centre, the superior colliculus. The neuronal tracer, biotinylated dextran, was injected into superficial and deep layers of the superior colliculus of rabbits and the resultant labelling in the reticular region was examined. After large injections, which encompassed both superficial and deep collicular layers, two discrete populations of retrogradely labelled cells are seen in the region of the reticular nucleus. One population of retrogradely labelled cells lies in the dorsocaudal regions of the reticular nucleus, the classically defined visual sector. This group of retrogradely labelled reticular cells is also seen after injections into the superficial layers of the superior colliculus, but not after injections limited to the deeper collicular layers. The other population lies close to the ventromedial edge of the main body of the reticular nucleus, within a region referred to as the inner small-celled region. This group of small cells has been commonly thought to be part of the reticular nucleus, but our immunohistochemical studies suggest that this is a clearly separate region, a region continuous ventrally with zona incerta. The retrogradely labelled cells in the inner small-celled region are seen also after injections limited to the deeper collicular layers, but not after injections limited to the superficial collicular layers. Our results suggest functional heterogeneity within the reticular nucleus: Specifically, it suggests that the nucleus is in a position to influence the processing of visual information at both the dorsal thalamic and midbrain levels.

Development of direct GABAergic projections from the zona incerta to the somatosensory cortex of the rat

The postnatal development of direct thalamocortical projections from the zona incerta of the ventral thalamus to the whisker representation area of the rat primary somatosensory cortex was investigated. Cytoarchitectonic analysis based on Nissl staining, cytochrome oxidase histochemistry and immunohistochemistry for glutamic acid decarboxylase, GABA, parvalbumin and calbindin D28K revealed that the zona incerta can be clearly distinguished from surrounding diencephalic structures from the day of birth. Moreover, four distinct anatomical subdivisions of this nucleus were identified: the rostral, dorsal, ventral and caudal. Of these, the ventral subdivision is by far the most conspicuous, containing the highest density of neurons, and the highest levels of cytochrome oxidase, glutamate decarboxylase, GABA, parvalbumin and calbindin D28K. In contrast, the dorsal, rostral and caudal subdivisions contain fewer cells, lower levels of glutamic acid decarboxylase and GABA and very few parvalbumin-positive and calbindin-positive neurons. Small injections of rhodamine coated microspheres or Fluoro-gold in the primary somatosensory cortex of animals at different stages of development revealed the existence of retrogradely labeled neurons in the rostral and dorsal subdivisions of the zona incerta from postnatal day 1. At this age, retrogradely labeled cells were also found in the ventral lateral, ventral posterior medial, posterior medial, centrolateral, ventral medial and magnocellular subdivision of the medial geniculate nuclei of the dorsal thalamus. The density of the incertocortical projection reaches its maximum between the first and second postnatal weeks, decreasing subsequently, until an adult pattern of labeling is achieved. Tracer injections combined with immunohistochemistry revealed that the majority of the incertocortical projection derives from GABAergic neurons, implying a potentially inhibitory role for the incertocortical projection. These results demonstrate that the rat trigeminal system contains parallel thalamocortical pathways of opposite polarity, emerging from both the dorsal (glutamatergic, excitatory) and ventral (GABAergic, inhibitory) thalamus since the day of birth. As such, these findings suggest that, contrary to the classical notion, not only the dorsal but also the ventral thalamus may play a special role in both cortical maturation and function.

Organization of projections from the anterior hypothalamic nucleus: a Phaseolus vulgaris-leucoagglutinin study in the rat

Anterior hypothalamic nucleus (AHN) projections were examined with the Phaseolus vulgaris-leucoagglutinin (PHA-L) method in adult male rats. Labeled axons from the AHN follow three major routes. 1) A large ascending pathway ends densely in the telencephalon, particularly in the lateral septal nucleus. Axons along this route provide moderate to dense input to the medial and lateral preoptic areas, and a few are also observed in the septofimbrial nucleus and fimbria; the latter end in the temporal hippocampus. A few axons reach the amygdala through the bed nuclei of the stria terminalis, which receive a moderate input, and then the stria terminalis, and others reach it by way of the ansa peduncularis. 2) The second pathway travels dorsal to the AHN, ending densely in rostral perifornical regions of the lateral hypothalamic area, and the rostral ventrolateral tip of the nucleus reuniens. The parataenial and rostral paraventricular thalamic nuclei also receive a significant input. Some fibers and boutons were also observed in the rhomboid, interanterodorsal, and mediodorsal nuclei, and others course through the stria medullaris to the lateral habenula. 3) the largest pathway descends through dorsal and ventral routes in the medial hypothalamic zone before ending massively in the periaqueductal gray. Dorsal route fibers provide inputs to the zona incerta and posterior hypothalamic nucleus, whereas more ventral axons generate dense terminal fields in the ventromedial nucleus capsule and core, and dorsal premammillary nucleus. The retrochiasmatic area, dorsomedial nucleus, and medial supramammillary nucleus also receive significant inputs, and a few axons end in the subparafascicular nucleus, superior colliculus, and mammillary body. The caudalmost axons were seen in the pontine central gray and reticular formation. These pathways are bilateral, usually with a distinct ipsilateral predominance. The overall pattern of efferents from anterior, central, and posterior parts of the AHN is similar, whereas the relative densities of particular terminal fields may vary considerably. Projections from adjacent parts of the retrochiasmatic and perifornical areas are also described. The results are discussed in terms of neural circuitry that may be involved in mediating interactions between animals.

The distribution of histaminergic axons in the superior colliculus of the cat

The histaminergic projection from the hypothalamus to the superior colliculus was examined immunohistochemically in the cat brain using an antibody to histamine. The source of histaminergic fibers in the brain is a group of neurons in the posterior hypothalamus, located primarily in ventrolateral and periventricular regions and collectively referred to as the tuberomammillary nucleus. All laminae of the superior colliculus--including the superficial, intermediate, and deep layers, as well as the central gray--were blanketed with histamine-immunoreactive axonal fibers. Overall, labeling in the superior colliculus was moderately dense compared to other locations in the cat brain, with some variation in fiber density. Individual labeled fibers resembled histaminergic fibers described previously in the brain. Labeled axonal fibers showed infrequent branching and were beaded with numerous en passant varicosities that were typically 1 micron or smaller, but as large as 2.5 micron in diameter. Varicosity size differed significantly at different depths in the colliculus. The histaminergic projection appears to be separate from a previously reported, apparently non-histaminergic projection from neurons in the dorsal hypothalamic area to discrete regions of intermediate and deep colliculus. These results indicate that the histaminergic projection from the tuberomammillary nucleus of the hypothalamus projects extensively throughout the superior colliculus. Histamine, which is believed to act as a neuromodulator in the brain, is in a position to influence sensory and motor-related processes in every layer of the cat superior colliculus.

Corticotectal projections in the cat: anterograde transport studies of twenty-five cortical areas

Retrograde transport studies have shown that widespread areas of the cerebral cortex project upon the superior colliculus. In order to explore the organization of these extensive projections, the anterograde autoradiographic method has been used to reveal the distribution and pattern of corticotectal projections arising from 25 cortical areas. In the majority of experiments, electrophysiological recording methods were used to characterize the visual representation and cortical area prior to injection of the tracer. Our findings reveal that seventeen of the 25 cortical areas project upon some portion of the superficial layers (stratum zonale, stratum griseum superficiale, and stratum opticum, SO). These cortical regions include areas 17, 18, 19, 20a, 20b, 21a, 21b, posterior suprasylvian area (PS), ventral lateral suprasylvian area (VLS), posteromedial lateral suprasylvian area (PMLS), anteromedial lateral suprasylvian area (AMLS), anterolateral lateral suprasylvian area (ALLS), posterolateral lateral suprasylvian area (PLLS), dorsolateral lateral suprasyvian area (DLS), periauditory cortex, cingulate cortex, and the visual portion of the anterior ectosylvian sulcus. While some of these corticotectal projections target all superficial laminae and sublaminae, others are more discretely organized in their laminar-sublaminar distribution. Only the corticotectal projections arising from areas 17 and 18 are exclusively related to the superficial layers. The remaining 15 pathways innervate both the superficial and intermediate and/or deep layers. The large intermediate gray layer (stratum griseum intermedium; SGI) receives projections from almost every cortical area; only areas 17 and 18 do not project ventral to SO. All corticotectal projections to SGI vary in their sublaminar distribution and in their specific pattern of termination. The majority of these projections are periodic, or patchy, and there are elaborate (double tier, bridges, or streamers) modes of distribution. We have attempted to place these findings into a conceptual framework that emphasizes that the SGI consists of sensory and motor domains, both of which contain a mosaic of connectionally distinct afferent compartments (Illing and Graybiel, '85, Neuroscience 14:455-482; Harting and Van Lieshout, '91, J. Comp. Neurol. 305:543-558). Corticotectal projections to the layers ventral to SGI, (stratum album intermediale, stratum griseum profundum, and stratum album profundum) arise from thirteen cortical areas. While an organizational plan of these deeper projections is not readily apparent, the distribution of several corticotectal inputs reveals some connectional parcellation.

Cytoarchitectonic heterogeneities in the thalamic reticular nucleus of cats and ferrets

The thalamic reticular nucleus has been classically defined as a group of cells surrounding most of the rostral and lateral surfaces of the dorsal thalamus, lateral to the fibres of the external medullary lamina and medial to those of the internal capsule. With the use of Nissl staining and antibodies to gamma-aminobutyric acid (GABA), somatostatin, and parvalbumin, this study describes the cytoarchitecture of the thalamic reticular nucleus of cats and ferrets. In cats, three subdivisions of the nucleus are distinguished, two of which are distinct in ferrets also. First, the main body of the reticular nucleus lies lateral to the fibres of the external medullary lamina (except ventrally) and medial to those of the internal capsule. In both cats and ferrets, this structure is heterogeneous, consisting of distinct layers, the details of which vary along the dorsoventral axis. A prominent rostroventral portion of comparatively small rounded cells is also apparent within the main body. Most reticular cells in all areas of the main body are labelled with all of the above mentioned antibodies. Second, the inner small-celled region is a group of small cells located between the external medullary lamina (ventrally) and the medial margin of the ventral regions of the main body of the reticular nucleus: the inner small-celled region is clearly differentiated in cats only. Previous studies have referred to this area as being part of the main body of the reticular nucleus, but we suggest that it may form a separate subnucleus. For example, the inner small-celled region stands in striking contrast to the main body of the reticular nucleus in that none of its cells are GABA immunoreactive and only a small caudal subpopulation are parvalbumin immunoreactive. A very similar pattern of immunostaining is apparent for the cells in the zona incerta, although the latter contains a small rostral subpopulation of GABA immunoreactive cells. Furthermore, although morphologically distinct from the zona incerta, the inner small-celled region fuses with it ventrocaudally. We suggest that the inner small-celled region may constitute a previously undescribed dorsal extension of the zona incerta, rather than a subdivision of the reticular nucleus. Third, the perireticular nucleus, hitherto unidentified, is a discrete group of small cells lateral to the main body of the reticular nucleus and medial to the corpus striatum (globus pallidus and caudate-putamen). It is apparent throughout most of the dorsoventral extent of the main body of the reticular nucleus of cats and ferrets.(ABSTRACT TRUNCATED AT 400 WORDS)

Pathway from the zona incerta to the superior colliculus in the rat

In order to test the proposal that the zona incerta contributes to the generation of orienting movements, we examined the synaptic relationships between the incertotectal pathway and the cells of origin of the predorsal bundle. The predorsal bundle cells give rise to the major premotor pathway from the superior colliculus to the brainstem gaze centers. First, cytochrome oxidase histochemistry, gamma-aminobutyric acid (GABA), and glutamic acid decarboxylase (GAD) immunocytochemistry, and the axonal transport of markers were used to define the borders of a ventral subdivision of the zona incerta. This subdivision projects topographically to the same sublamina of the intermediate grey layer of the superior colliculus that contains the vast majority of the predorsal bundle cells. Experiments in which incertotectal cells were labeled by both retrograde transport and immunocytochemistry showed that this pathway is GABAergic. Retrograde and anterograde experiments also showed that this pathway is reciprocated by a pathway from the intermediate grey layer of the superior colliculus to the same ventral subdivision of the zona incerta. Finally, experiments combining axonal transport and electron microscopic methods showed that the incertotectal pathway is the source of a monosynaptic GABAergic input to the cells of origin of the predorsal bundle. The ventral subdivision of the zona incerta is contrasted with a second source of GABAergic input to the predorsal bundle cells, the substantia nigra pars reticulata.

AMPA glutamate receptor activation in the posterior zona incerta inhibits amphetamine- and apomorphine-induced stereotypy

Previous work demonstrated that alpha-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid (AMPA) glutamate receptor antagonism in the zona incerta (ZI) dorsal to the subthalamic nucleus inhibits stereotypy in rats. The current investigation was undertaken to determine if AMPA receptors in a more caudal portion of the ZI have a role in the expression of stereotyped behavior. Rats were injected bilaterally with AMPA into the posterior ZI dorsal to the substantia nigra, and immediately given a systemic injection of d-amphetamine (10 mg/kg, s.c.) or apomorphine (1 mg/kg s.c.). AMPA produced a dose-dependent inhibition of stereotypy induced by both drugs which was prevented by the coadministration of the AMPA/kainic acid antagonist, 6,7-dinitroquinoxaline-2,3-dione (DNQX) (0.5 microgram/0.5 microliter). A dose of AMPA as low as 62.5 ng completely abolished the oral component of stereotypy induced by both apomorphine and amphetamine. This dose of AMPA alone had no significant effect on spontaneous locomotor activity but enhanced the locomotor response stimulated by amphetamine (10 mg/kg, s.c.) due to an inhibition of stereotypy. The finding that activation of AMPA receptors in the posterior ZI inhibits stereotypy shows a contrast to results in the neighboring medial ZI dorsal to the subthalamic nucleus, where blockade of AMPA/kainic acid glutamate receptors with DNQX inhibits stereotypy.

Patterns of antigenic expression in the thalamic reticular nucleus of developing rats

The present study describes the development of the thalamic reticular nucleus in rats with the use of Nissl staining and antibodies to parvalbumin and pro-alpha-thyrotropin-releasing hormone (alpha TRH). Two major subdivisions of the reticular nucleus are apparent: 1) the main body, which is itself heterogeneous and lies for the most part between the fibres of the internal capsule and external medullary lamina, and 2) the perireticular nucleus, which lies lateral to the main body and medial to the globus pallidus. In the main body of the reticular nucleus of adults, most cells in all regions are immunoreactive to parvalbumin and alpha TRH. During development there are two waves of parvalbumin and alpha TRH expression. The first wave occurs between postnatal day (P) 0 and P10, and labelled cells are apparent in rostrolateral areas of the main body of the nucleus only. At P10, such cells are not apparent. From P7 to adult, there is a second wave of parvalbumin and alpha TRH expression: labelled cells emerge first in central, then in caudal, and finally in rostral areas of the nucleus. In adults, the perireticular nucleus is made up of a few small cells which are immunostained for parvalbumin and alpha TRH. These cells are more frequent in areas of the internal capsule adjacent to the ventral regions of the main body of the reticular nucleus, rostrodorsal to the entopeduncular nucleus. From E (embryonic day) 17 to about P10, the perireticular nucleus consists of a surprisingly large population of neurones, many of which are parvalbumin and alpha TRH immunoreactive. By about P10, as in adults, there are few perireticular cells.

AMPA/kainic acid glutamate receptor antagonism in the zona incerta dorsal to the subthalamic nucleus inhibits amphetamine-induced stereotypy bur not locomotor activity

The effect of 6,7-dinitroquinoxaline-2,3-dione (DNQX), an alpha-amino-3- hydroxy-5-methyl-4-isoxazole-propionate (AMPA)/kainic acid glutamate receptor antagonist, injected into the zona incerta (ZI) was investigated to determine whether the behavioral responses to systemic amphetamine involve AMPA/kainic acid receptors in this brain region. Rats were injected bilaterally in the ZI with either vehicle or DNQX (1 microgram/0.5 microliter) and immediately given a systemic injection of D-amphetamine (0.5, 1.0 or 10.0 mg/kg, s.c.). Locomotor activity was recorded for 1 h. DNQX did not significantly affect hypermotility stimulated with 0.5 and 1.0 mg/kg amphetamine, but markedly increased the level of locomotor activity elicited by the higher dose, 10 mg/kg. To test the hypothesis that the enhanced locomotor response to high dose amphetamine was due to an inhibition of stereotyped behavior, the effect of DNQX in the ZI on amphetamine and apomorphine-induced stereotypy was investigated. DNQX significantly inhibited stereotypy induced by amphetamine (10 mg/kg) and apomorphine (1 mg/kg), with the onset of inhibition of amphetamine-induced stereotypy corresponding to the onset of enhanced locomotor activity. Ibotenic acid lesions of the ZI produced similar results, having an insignificant effect on locomotor activity stimulated by low dose amphetamine (1 mg/kg) and an attenuation of apomorphine-induced stereotypy which was of a magnitude comparable to that produced by DNQX. Thus, the AMPA/kainic acid subtypes of glutamate receptors in the ZI may be involved in the regulation of motor function mediated via striatal output but not mesolimbically generated locomotor activity.

The organization of GABAergic neurons in the mammalian superior colliculus

GABA is an important inhibitory neurotransmitter in the mammalian superior colliculus. As in the lateral geniculate nucleus, GABA immunoreactive neurons in SC are almost all small and are distributed throughout the structure in all mammalian species studied to date. Unlike the LGN, GABA-labeled neurons in SC have a variety of morphologies. These cells have been best characterized in cat, where horizontal and two granule cell morphologies have been identified. Horizontal cells give rise to one class of presynaptic dendrite while granule C cells give rise to another class of spine-like presynaptic dendrite. Granule A cells may be the origin of some GABAergic axon terminals. GABA containing synaptic profiles form serial synapses, providing a possible substrate for disinhibition. The distribution of GABAA and GABAB receptor subtypes appears similar to that of GABA neurons, with the densest distribution found within the superficial gray layer. However, antibody immunocytochemistry of the beta 2 and beta 3 subunits of the GABAA receptor reveals that it is located at both synaptic and non-synaptic sites, and may be associated with membrane adjacent to terminals with either flattened or round vesicles. A few GABA containing neurons in SC colocalize the pentapeptide leucine enkephalin or the calcium binding protein calbindin. However, none appear to co-localize parvalbumin, a situation different from GABA containing interneurons in the LGN and visual cortex. The diversity of GABA neurons in SC rivals that found in visual cortex, although unlike visual cortex, the pattern of co-occurrence does not distinguish GABA cell types in SC. The superior colliculus also differs from both LGN and visual cortex in that GABA and calbindin immunoreactivity is not altered by either long-term occlusion and/or short-term enucleation in adult Rhesus monkeys. No consistent differences have been found in the optical density of GABA labeling in either cells or neuropil. To conclude, GABA neurons in the superior colliculus share some properties like those in LGN and others like those in visual cortex. In other properties, they differ from GABA neurons in both the LGN and visual cortex. The GABA systems in the superior colliculus are similar in all mammalian species studied, suggesting that they are phylogenetically conserved systems which are not amenable to plastic alterations, a situation different to that in the geniculostriate system.

Preoccipital cortex receives a differential input from the frontal eye field and projects to the pretectal olivary nucleus and other visuomotor-related structures in the rhesus monkey

The bidirectional axonal transport capabilities of the horseradish peroxidase (HRP) technique facilitated the study of the frontal-eye-field (FEF) input and pretectal output of two regions of extrastriate preoccipital cortex (POC). Following horseradish peroxidase (HRP) gel implants into the middle and dorsal POC in two rhesus monkeys, the middle POC implant demonstrated retrograde frontal cortical labeling largely restricted to the inferior frontal eye field (iFEF) and adjacent inferior prefrontal convexity, whereas the dorsal POC implant showed labeling in the caudal ventral bank of the superior ramus of the arcuate sulcus (sas) and middle-to-dorsal region of the rostral bank of the concavity of the arcuate sulcus (dorsal FEF). Prominent anterogradely labeled efferent preoccipital projections were observed to the ipsilateral pretectal olivary nucleus (PON) and to a lesser extent the anterior pretectal nucleus. Although the middle POC case had heavier projections to the lateral PON, the dorsal case projected more heavily to the medial PON. In addition, both implants demonstrated subcortical connections with the lateral and dorsal inferior pulvinar nuclei, central superior lateral thalamic intralaminar nucleus, caudate nucleus, and middle-to-ventral claustrum. However, while the middle POC implant had efferent projections to the superficial superior colliculus (SC), pregeniculate nucleus (PGN), lateral terminal accessory optic nucleus (LTN), and dorsolateral pontine nucleus (DLPN), resembling those previously reported for the middle temporal (MT) visual area (Maunsell & Van Essen, 1982; Ungerleider et al., 1984), the dorsal implant had projections to the lateral intermediate SC, zona incerta (ZI), PGN, a notably lesser projection to the LTN, and basilar pontine projections to the lateral and lateral dorsal pontine subnuclei (not including the extreme dorsolateral DLPN). These preliminary results suggest that the preoccipital cortex, which reportedly functions in pupillary constriction, accommodation, and convergence, entertains connections with the PON and other visuomotor-related structures, and thus could act as an intermediary in the pathway between the iFEF and PON, and provide a possible explanation for pupillary effects that occur with stimulation of the FEF (Jampel, 1960) and within the contex of other oculomotor activities. The findings shed light on certain differences in connections of middle vs. dorsal POC with visuomotor-related nuclei, and appear to suggest that the middle region, which receives input from the iFEF, has greater access to the optokinetic (OKN) system by virtue of its projection to the LTN, and to the smooth-pursuit system b

Organization of the projections from the trigeminal brainstem complex to the superior colliculus in the rat and hamster: anterograde tracing with Phaseolus vulgaris leucoagglutinin and intra-axonal injection

Anterograde tracing with Phaseolus vulgaris leucoagglutinin (PHA-L) and intra-axonal recording and injection techniques were employed to describe the projection from the trigeminal (V) brainstem complex to the deep laminae of the superior colliculus (SC) in the hamster and the rat. The organization of these projections was the same in the two species. Deposits of PHA-L into V nucleus principalis (PrV) produced labelled axons and boutonlike swellings in the lower stratum griseum intermediale (SGI) and upper stratum album intermedium (SAI) in the SC bilaterally. Plots of boutonlike swellings indicated that the terminals of this projection were arrayed in clusters. Nucleus principalis also projected to the stratum griseum profundum (SGP) and stratum album profundum (SAP). This deeper projection did not terminate in clusters and it was most prominent in the lateral SC. The ipsilateral PrV-SC projection appeared to arise mainly from axons that recrossed the midline at the level of the SC commissure. Reconstruction of individual PHA-L labelled fibers demonstrated that single axons gave rise to terminals on both sides of the midline. Deposits of PHA-L into V subnucleus interpolaris (SpI) yielded results that were identical to those obtained with PrV injections with one exception: none of these deposits produced any labelled terminals in the ipsilateral SC. Deposits of PHA-L into V subnucleus caudalis (SpC) produced only sparse labelling in SC. Most labelled swellings were located in the SGP and SAP and they were visible only in the SC contralateral to the PHA-L injection site. Single axons arising from cells in SpI were recorded and injected with horseradish peroxidase (HRP) in the hamster's SC. These fibers all responded to stimulation of multiple mystacial vibrissae and gave rise to 2-5 clusters of bouton-like swellings in the lower SGI and upper SAI.

Possible origin of glutamatergic projections to the midbrain periaqueductal gray and deep layer of the superior colliculus of the rat

The possible origin of glutamatergic input to the rodent periaqueductal gray (PAG) was analyzed utilizing a combined retrograde transport-immunocytochemical technique. Injections of wheat germ agglutinin-horseradish peroxidase were made into the PAG of 12 adult rats and into the deep layer of the superior colliculus in 2 rats. The brain tissue was first reacted histochemically to demonstrate the retrograde tracer and subsequently processed with immunohistochemical techniques using a recently developed monoclonal glutamate antibody. Following PAG injections, several brain areas were found to contain double-labeled neurons. The greatest number of double-labeled glutamate-like immunoreactive neurons were observed in the zona incerta, spinal trigeminal nucleus, cuneiform nucleus, cingulate cortex, cerebellar interpositus nucleus, deep mesencephalic nucleus and the PAG itself. Double-labeled neurons were also observed in several other nuclei including the pretectal nuclei, the frontal and occipital cortex, several reticular nuclei, the dorsomedial hypothalamic nucleus, and the substantia nigra. Many of the same nuclei contained double-labeled neurons following collicular injections, but in addition, double-stained cells were found in the primary visual cortex, lateral dorsal and lateral posterior thalamic nuclei, nucleus of the posterior commissure, ventral lateral geniculate nucleus, dorsal column nuclei and several additional pretectal nuclei. The results of this double-labeling study raise the possibility that these nuclei may provide glutamatergic inputs to the midbrain PAG and/or superior colliculus. These putative glutamatergic afferent projections may ultimately influence the PAG's role in several important functions including antinociception, defensive mechanisms or vocalization and may also play a role in the superior collicular involvement in defensive mechanisms, in visuo-motor integration in the orienting reflex and in facilitating shifts in gaze.

The neurons of the substantia nigra and zona incerta which project to the cat superior colliculus are GABA immunoreactive: a double-label study using GABA immunocytochemistry and lectin retrograde transport

The neurotransmitter cytochemistry of neurons in the substantia nigra and zona incerta which project to the cat superior colliculus was examined. Neurons in both structures were double-labeled with an antibody to the transmitter GABA and a retrograde tracer, [3H]n-acetylated wheat germ agglutinin, injected into the superior colliculus. All cells in the zona incerta and substantia nigra which projected to the superior colliculus were labeled by the GABA antiserum. Most other neurons within the zona incerta and virtually all within the substantia nigra pars reticulata and pars lateralis were also labeled by the GABA antibody. By contrast, neurons in the substantia nigra pars compacta were not labeled by either the GABA antibody or wheat germ agglutinin. Nigrotectal cells in the substantia nigra were of medium to large size and most had stellate-shaped cell bodies. Zona incerta cells projecting to the superior colliculus were also of medium to large size, but most had horizontal fusiform cell bodies. This study demonstrates two new findings: (1) that all nigrotectal neurons in cat are immunoreactive to a GABA antibody and probably contain the neurotransmitter GABA; and (2) that these GABA immunoreactive neurons in cat are found not only in the substantia nigra pars reticulata but also within the pars lateralis. Zona incerta cells projecting to the superior colliculus have a different morphology but are also apparently GABAergic. These data provide an anatomical substrate for the known inhibitory action of the nigrotectal pathway on superior colliculus neurons.

The distribution pattern of alpha-MSH-like immunoreactivity in the cat central nervous system

The distribution pattern of alpha-melanocyte stimulating hormone-like immunoreactivity (alpha-MSH-Li) was studied in cats using avidin-biotin modification of immunocytochemical method. Cell bodies containing alpha-MSH-Li were observed in the medial basal hypothalamus, especially in the infundibular nucleus, the lateral hypothalamus and near zona incerta. Fibers with alpha-MSH-Li extended beyond the hypothalamus, into the paraventricular nucleus of the thalamus, rostral amygdala, periaqueductal gray, locus ceruleus, parabrachial nucleus and medial nucleus of the nucleus tractus solitarius. Axons with alpha-MSH-Li were also seen diffusely in various cortical areas, but more extensively in the limbic cortical regions. The distribution pattern of the cell bodies and fibers containing alpha-MSH-Li bears several similarities to that seen in rats, but differs in that the alpha-MSH-Li was not observed in cell bodies in locations other than the medial basal and lateral hypothalamus.

Frontal eye field as defined by intracortical microstimulation in squirrel monkeys, owl monkeys, and macaque monkeys: I. Subcortical connections

Intracortical microstimulation was used to define the borders of the frontal eye fields in squirrel, owl, and macaque monkeys. The borders were marked with electrolytic lesions, and horseradish peroxidase conjugated to wheat germ agglutinin was injected within the field. Following tetramethyl benzidine histochemistry, afferent and efferent connections of the frontal eye field with subcortical structures were studied. Most connections were ipsilateral and were similar in all primates studied. These include reciprocal connections with the following nuclei: medial dorsal (lateral parts), ventral anterior (especially with pars magnocellularis), central lateral, paracentral, ventral lateral, parafascicular, medial pulvinar, limitans, and suprageniculate. The frontal eye field also projects to the ipsilateral pretectal nuclei, subthalamic nucleus, nucleus of the posterior commissure, superior colliculus (especially layer four), zona incerta, rostral interstitial nucleus of the medial longitudinal fasciculus, nucleus Darkschewitsch, dorsomedial parvocellular red nucleus, interstitial nucleus of Cajal, basilar pontine nuclei, and bilaterally to the paramedian pontine reticular formation and the nucleus reticularis tegmenti pontis. Many of these structures also receive input from deeper layers of the superior colliculus and are known to participate in visuomotor function. These results reveal connections that account for the parallel influence of the superior colliculus and the frontal eye field on visuomotor function; suggest that there has been little evolutionary change in subcortical connections, and therefore function, of the frontal eye fields since the time that these lines of primates diverged; and support the conclusion that the frontal eye fields are homologous in New and Old World monkeys.