The cells of origin of the incertofugal projections to the tectum,thalamus, tegmentum and spinal cord in the rat: A study using the autoradiographic and horseradish peroxidase methods

Efferent projections of the zona incerta were examined in the rat using the autoradiographic and horseradish peroxidase methods, with special reference to the cytoarchitectonic structure of the zona incerta.Autoradiographic experiments showed that the incertofugal fiber systems reach ipsilaterally to the thalamus (lateral dorsal, central lateral, ventral lateral geniculate, parafascicular, subparafascicular and reuniens nuclei, and posterior nuclear complex), to the hypothalamus (dorsal, lateral and posterior hypothalamic areas), to the tectum (medial pretectal area, deep pretectal and pretectal nuclei, superior colliculus and periaqueductal gray) and to the midbrain tegmentum, pons and medulla oblongata (subcuneiform, cuneiform and red nuclei, nuclei of the posterior commissure and Darkschewitsch, interstitial nucleus of Cajal, pedunculopontine tegmental nucleus, oral and caudal pontine reticular nuclei, nucleus raphe magnus, gigantocellular reticular nucleus, pontine gray and inferior olivary complex). Contralaterally, incertal efferent fibers reach to the zona incerta.Cells of origin of the incertofugal fiber systems to the tectum, thalamus, tegmentum and spinal cord were examined using the retrograde horseradish peroxidase method. Cells of origin of the incertotectal pathway are located mainly in the ventral and caudal parts of the zona incerta and partly in the antero-polar, dorsal and postero-polar parts. Cells projecting to the thalamus (at least to the lateral dorsal and central lateral nuclei) are situated in the ventral and caudal parts of the zona incerta, but they are rare in the other incertal structures. Cells of origin of the incertotegmental system are located mainly in the dorsal, magnocellular and caudal parts and partly in the antero- and postero-polar parts, but they are not situated in the ventral part. Cells of the magnocellular part project more caudally to the medulla oblongata and spinal cord than those of the other parts of the zona incerta. Forel's field contains many cells projecting to the tegmentum.The results provide good evidence that the cells of origin of efferent projections are topographically organized and are related to cytoarchitectonic areas within the zona incerta.     

The projection of the fovea to the superior colliculus in rhesus monkeys

Horseradish peroxidase was injected into the superior colliculus of seven Rhesus monkeys to determine whether retinal ganglion cells concerned with the fovea project to the superior colliculus or whether, as results of several experiments using other methods suggest, only the parafoveal retina projects directly to this region. The retinal distribution of ganglion cells labelled by retrograde transport showed unequivocally that there is a direct projection from the fovea to the superior colliculus. However, there was no convincing evidence that the nasotemporal overlap that exists in the retinogeniculate projections also occurs in the projection to the superior colliculus. The results cast serious doubt on schemes that attribute little or no role to the superior colliculi in foveal vision and which emphasize the projection of the fovea to the visual cortex and the peripheral retina to the superior colliculi.     

The pontocerebellar projection in the rhesus monkey: an experimental study with retrograde axonal transport of horseradish peroxidase.

The pontocerebellar projection has been studied in the Rhesus monkey by use of the retrograde axonal transport of horseradish peroxidase. The parts of the cerebellum investigated receive afferents from pontine cell groups arranged as rostro-caudally oriented lamella or slab-like regions. As a rule labelled cells are found at all rostro-caudal levels in discrete groups, but their number at various levels, and particularly their distribution in the transverse plane, differ according to which part of the cerebellum has been injected. Although the same cell group may apparently be labelled after injections of different parts of the cerebellum, each cerebellar region has its own characteristic territory in the pontine nuclei. The anterior lobe receives fibres from lamella-like regions mainly in the caudal half of the pons, with cells projecting to the vermis and intermediate parts of the anterior lobe somewhat differently situated. Crus I is connected mainly with a region rostromedially in the pons, while crus II receives fibres from all levels of the pons, the cells being located medially and ventrally at caudal levels and more laterally at rostral levels. The paramedian lobule is supplied from pontine cells restricted to the rostral two-thirds of the pons and located more laterally than those projecting to crus II. Lobules VII and VIIIA (the main part of the vermal visual area) receive fibres mainly from two long column-like regions located dorsomedially and dorsolaterally, while lobulus VIIIB is supplied mainly from other areas in the pons.Counting of labelled cells indicates that about 30% of the fibres to the vermis, and about 10% to the hemispheres, are uncrossed. The density of the projection to the cerebellar hemispheres is found to be about three times as high as to the vermis. In two cases injections of horseradish peroxidase were placed in the cerebral cortex as well as in the cerebellum, enabling a direct comparison of sites of termination of corticopontine fibres (orthograde transport of horseradish peroxidase) with location of pontocerebellar neurons.It is suggested that the cortico-ponto-cerebellar pathway is organized so as to bring about convergence in the cerebellar cortex from several parts of the cerebral cortex, and that this takes place in a highly organized manner so that each small part of the cerebellar cortex has its own characteristic set of inputs.     

Dorsal root projections in the clawed toad (Xenopus laevis) as demonstrated by anterograde labeling with horseradish peroxidase

Horseradish peroxidase was applied to the proximal stumps of severed cervical, thoracic and lumbar dorsal roots in the clawed toad, Xenopus laevis, in order to study the course, distribution and site of termination of dorsal root fibers in the spinal cord and brain stem. The anterograde transport of horseradish peroxidase as applied in the present study proved to be a useful and reliable technique. Results show that on entering the spinal cord, dorsal root fibers segregate into a medially placed component entering the dorsal funiculus and a more laterally situated bundle in the dorsal part of the lateral funiculus. As regards its position the latter bundle presumably represents the anuran homologue of the mammalian tract of Lissauer. Moreover, a small intermediate bundle of fibers directly enters the spinal gray matter. The labeled fibers entering the dorsal funiculus and the tract of Lissauer ascend and descend in the spinal cord, displaying a longitudinal arrangement resembling that of higher vertebrates.In the spinal gray, dorsal root fibers terminate in the dorsal, central and lateral fields of Ebbesson, 25 with the last field being a major terminus for dorsal root fibers originating in the limb-innervating segments. No dorsal root fibers were found to project to the motoneuron fields. A dorsal column nucleus, which is divisible into medial and lateral compartments, is present in the obex region and extends from the level of the second spinal nerve to that of the entrance of the vagus and glossopharyngeal nerves. Dorsal root fibers from the lumbar and all thoracic segments project to the medial compartment of the dorsal column nucleus, whereas those of the cervical enlargement project to the lateral compartment.Although the anuran dorsal column nucleus appears to be less differentiated than that of higher vertebrates, its medial and lateral compartments can be considered to be the forerunners of the mammalian nucleus gracilis and nucleus cuneatus, respectively.     

Cortical projections to the periaqueductal grey in the cat: a retrograde horseradish peroxidase study

Stereotaxic fluid microinjections of horseradish peroxidase into different parts of the rostral and caudal periaqueductal grey (PAG) in cats have provided substantial retrograde evidence that the somatosensory cortex (I and II), frontal cortex, insular and cingular cortex are the principal sources of cortical-PAG projections. The somatosensory cortex II projects to all the regions of the rostral and caudal PAG. The frontal cortex projects to dorso-lateral quadrant of the PAG. The same projections were determined from insular and cingular cortex to PAG. The findings revealed a morphological substratum of corticofugal effects on PAG.     

Widespread cortical projections of the hippocampal formation in the cat

Efferent projections from the hippocampal formation to the cat's cortex were traced with the retrograde horseradish peroxidase technique. Different areas of the cortex of 31 cats were injected with small amounts of horseradish peroxidase. All subregions of the hippocampal formation were screened for labeled cells. It was found that, with the exception of the entorhinal injections, only subicular areas of the hippocampal formation contain labeled neurons. When HRP was injected into the entorhinal cortex, labeled cells are also found in the hippocampus proper. The most dense projection from the subicular cortex is directed to the medial part of the cortical hemisphere. Here, cingulate, retrosplenial and medial prefrontal fields receive a substantial number of subicular efferents. Furthermore, the entorhinal cortex is reached by a number of axons originating in the subicular area. Scarce projections from the subicular cortex terminate in the dorsal prefrontal, temporal, parietal and prepiriform cortex. It is suggested that the projection from the subicular cortex to the neocortical areas of the frontal pole (medial prefrontal cortex) is of special importance as it may constitute a link between the association areas of the neocortex and those regions of the limbic system thought to play a role in memory (subicular cortex, mamillary bodies, anterior thalamus, cingulate gyrus).     

The rapid anterograde transport of horseradish peroxidase

Intravitreal injections of horseradish peroxidase in the rat consistently resulted in the labelling of contralateral and ipsilateral efferents of the retina. Simultaneous intravitreal injections of horseradish peroxidase and tritiated amino acids yielded identical projection patterns. Intravitreal injections of colchicine and pentobarbital reversibly blocked this transport of horseradish peroxidase from the eye to the brain. Furthermore, silver impregnation procedures indicated that this transport occurs in the absence of significant neurol damage. The anterograde transport of horseradish peroxidase along the optic pathways occurs at a rate between 288 and 432 mm per day.These observations show that the anterograde migration of horseradish peroxidase is subserved by fast axonal transport and that its occurrence does not depend on neurol injury or passive diffusion. The anterograde transport of horseradish peroxidase thus offers a valid and reliable anatomical method for tracing efferent connections in the nervous system.     

Selective uptake and anterograde transport of horseradish peroxidase by hippocampal granule cells

Introduction of horseradish peroxidase into the ventriculocisternal system results in selective labeling of the granule cells of the dentate gyrus and their axons, the mossy fibers. This labeling pattern is not seen after direct injections of horseradish peroxidase into the dorsal hippocampus. The density of the granule cell labeling appears to be related to their proximity to the site of highest horseradish peroxidase concentration. The combined distribution of horseradish peroxidase in the granule cells and mossy fibers strongly suggests that the latter element is labeled as the result of anterograde transport of horseradish peroxidase taken up by the granule cell perikarya or dendrites. This labeling was found in the absence of injury to the hippocampus, suggesting that neuronal damage is not necessary for anterograde transport horseradish peroxidase to occur.     

Periaqueductal grey projection to the ventrobasal complex in the cat: An horseradish peroxidase study

Horseradish peroxidase (HRP) was injected within the thalamic ventrobasal complex of 14 cats. The aim was to ascertain whether the periaqueductal grey matter (PAG) sends fibres to this complex. Retrogradely labelled cells were found within the PAG following HRP delivery either in the nucleus ventralis posterolateralis (VPL) or ventralis posteromedialis (VPM). PAG-VPL projection is only ipsilateral and arises mainly from lateral PAG. PAG-VPM projection is bilateral and originates from latero-ventral regions of the central grey. The hypothesis that PAG might control the activity of ventrobasal nociceptive neurones is proposed.     

Identification of cerebellar corticovestibular neurons retrogradely labeled with horseradish peroxidase.

The projection from the cerebellar cortex of the anterior lobe to the vestibular nuclei has been studied in adult cats by using the retrograde axonal transport of horseradish peroxidase. The corticocerebellar projection to the vestibular nuclei, namely to the lateral vestibular nucleus of Deiters, originates from Purkinje cells located within a narrow longitudinal zone of the ipsilateral vermis. This cortical zone has an average medio-lateral dimension of 0.4–0.6 mm, and extends laterally to the midline. with its medial border being located on the average 0.5–0.8 mm lateral to the midsagittal plane within lobuli I–V (ventral part); this distance, however, increases to 1.5–1.6 mm at the level of the dorsalmost part of lobule V. The presence of a narrow medial zone and a larger lateral zone in the cerebellar anterior lobe, which did not contain labeled Purkinje neurons, suggests that these cortical areas project to cerebellar nuclei rather than to the vestibular nuclei.These findings confirm the principle of longitudinal organization of the corticofugal projection from the cerebellum, and indicate that this projection is arranged in a discrete and subtle way. The results obtained are discussed with particular regard to the problem of the selective afferent inputs impinging on the longitudinal corticocerebellar zone projecting to the lateral vestibular nucleus as well as to the possible influences exerted on this zone by the different inputs impinging on the neighboring cortical compartments.     

Retinal ganglion cells that project to the superior colliculus and pretectum in the macaque monkey

Horseradish peroxidase was injected into the superior colliculus or pretectum or both in order to label, by retrograde axoplasmic transport, the retinal ganglion whose cells axons innervate the dorsal midbrain. The dendrites of ganglion cells were sufficiently well-labelled to reveal their overall morphological characteristics. It was therefore possible to compare the number and form of ganglion cells projecting to the midbrain with the total population of ganglion cells as revealed by optic nerve injections, and with ganglion cells labelled by injections in the lateral geniculate nucleus. We found that not more than 10% of all retinal ganglion cells project to the superior colliculus in the macaque monkey. This percentage varies little over the retina, being about 6% of all ganglion cells near the fovea and increasing slightly with eccentricity. The superior colliculus does not receive a projection from P beta cells and only a few P alpha cells terminate there. The majority of cells which project to the superior colliculus have a small- to medium-sized cell body and sparsely branched dendritic tree. We have called them P gamma and P epsilon cells by analogy with the gamma cells and epsilon cells in the cat's retina. Anatomically the P gamma and P epsilon cells are heterogeneous, which would be consistent with the physiological heterogeneity found for ganglion cells which project to the midbrain in monkeys.     

Cells of origin of brainstem afferents to lobules I and II of the cerebellar anterior lobe in the cat

Cells of origin of the brainstem afferents to lobules I and II of the cerebellar anterior lobe were identified by means of the retrograde horseradish peroxidase technique. In order to avoid diffusion into other lobules, injections were made under direct visual control through the fourth ventricle, after having removed ventral parts of the posterior lobe.With clearcut localization, the major projections originated from neurons of the following nuclei; the pontine nuclei (dorsal to the lateral nucleus, and the lateral and dorsal part of the peduncular nucleus) and nucleus corporis pontobulbaris; vestibular nuclear complex (the superior, medial and descending vestibular nuclei and group x), nucleus of Martin and interstitial nucleus of the vestibular nerve; the ventrolateral part of the external cuneate nucleus; lateral reticular nucleus (mainly the parvocellular portion); the inferior olivary complex (the caudal and central parts of the medial accessory nucleus and the lateral part of the dorsal accessory nucleus at middle levels). Small projections originated from the paramedian reticular nucleus, prepositus hypoglossi nucleus, nuclei raphe obscurus and pallidus, and the gracile and main cuneate nuclei.It was suggested that lobules I and II function not only as representations of the hindlimb-tail regions but also of the neck region by receiving afferents from the central cervical nucleus and the ventrolateral part of the external cuneate nucleus that receives dorsal root afferents C1 to C4.     

Afferent connections of the zona incerta: a horseradish peroxidase study in the rat

Restricted microelectrophoretic injections either of free horseradish peroxidase or of horseradish peroxidase conjugated with wheat germ agglutinin were given to albino rats in order to study the afferent connections of structures of the subthalamic region. The results suggest that the zona incerta receives its main input from several territories of the cerebral cortex, the mesencephalic reticular formation, deep cerebellar nuclei, regions of the sensory trigeminal nuclear complex and the dorsal column nuclei. Substantial input to the zona incerta appears to come from the superior colliculus, the anterior pretectal nucleus and the periaqueductal gray substance, whereas many other structures, among which hypothalamic nuclei, the locus coeruleus, the raphe complex, the parabrachial area and medial districts of the pontomedullary reticular formation, seem to represent relatively modest but consistent additional input sources. The afferentation of neurons in Forel's fields H1 and H2 appears to conform to the general pattern outlined above. As pointed out in the Discussion, the present results provide hodological support for the classic concept according to which the zona incerta can be regarded as a rostral extent of the midbrain reticular core. Some of the possible physiological correlates of the fiber connections of the zona incerta in the context of the sleep-waking cycle, ingestive behaviors, somatic motor mechanisms, visual functions and nociceptive behavior are briefly discussed.     

The septohippocampal projection in the rat: an electron microscopic horseradish peroxidase study

In order to identify the specific targets of the septohippocampal projection in the rat, horseradish peroxidase localization at the electron microscopic level was used. Following injections of free horseradish peroxidase into the medial septum, sections of the dorsal hippocampal formation were reacted with diaminobenzidine and processed for electron microscopy by routine methods. Sections were viewed unstained. Horseradish peroxidase labeling in the dentate gyrus was predominantly in the supra- and infragranular layers. All postsynaptic elements were neuronal. They included granule cell somata and somata and dendrites of hilar cells; these may include pyramidal basket cells. No synaptic contacts with vascular or glial elements were found. These results provide a basis for comparing the specific targets of the septohippocampal projection with those of the sympathohippocampal pathway, which innervates the dentate following lesions of the septohippocampal projection.     

Afferent connections of the mesencephalic reticular formation: A horseradish peroxidase study in the rat

The afferent connections of the mesencephalic reticular formation were studied experimentally in the rat by the aid of the retrograde horseradish peroxidase tracer technique. The results suggest that the rostral portion of the mesencephalic reticular formation receives its main input from the cerebral cortex, the zona incerta and the fields of Forel, the central gray substance, the nuclei reticularis pontis oralis and caudalis, and the deep cerebellar nuclei. Substantial input to the same territory of the mesencephalic reticular formation appears to come from the superior colliculus, the substantia nigra, the parabrachial area, the spinal trigeminal nucleus, and the nucleus reticularis gigantocellularis, whereas several other brain structures, among which the locus coeruleus and the raphe complex, seem to represent modest but consistent additional input sources. The afferentation of more caudal portions of the mesencephalic reticular formation appears to conform to the general pattern outlined above with only three exceptions: the cerebral cortex, the deep cerebellar nuclei and the spinal trigeminal nucleus seem to be relatively modest sources of projections to these levels.Considering that the mesencephalic reticular formation is a critical structure in the “ascending activating systems”, the present results, confirming and extending those of many other investigators, characterize a set of pathways that seem to be an important part of the anatomical substrate of the sleep-waking cycle.     

Endocytosis in glial cells (pituicytes) of the rat neurohypophysis demonstrated by incorporation of horseradish peroxidase.

The uptake and intracellular disposition of an extracellular marker, horseradish peroxidase, was traced at various time intervals in glial cells of the neurohypophysis, the pituicytes. Glands from normal rats and rats in which neurohypophysial secretion was activated by various stimuli were studied.Following an intravenous injection of the enzyme, reaction product rapidly appeared in the pituicyte cytoplasm, within numerous membrane-bound vacuoles of various size and morphology. The vacuoles presumably arose from large bristle coated invaginations of the plasmalemma that were marked by the tracer as soon as 5 min after injection. By 24 h, the reaction product was sequestered mainly in lysosomal multivesicular bodies and dense bodies.The observations suggest that endocytosis normally occurs in pituicytes. Furthermore, this endocytotic activity appears to be related to neurohypophysial secretion: morphometric analysis indicated that at each time period studied, the relative volume occupied by the peroxidase-labelled structures in cells from stimulated preparations was significantly greater than their corresponding controls. The possible role of uptake of extracellular fluid and accompanying plasma membrane that result from endocytosis is discussed.     

Morphology of migrating trigeminal motor neuroblasts as revealed by horseradish peroxidase retrograde labeling techniques

The trigeminal motor sensory roots were severed in chick embryos on days 2.5 4.5 of incubation and horseradish peroxidase applied to the wound. This procedure retrogradely labels developing trigeminal motor neuroblasts whose axons are at the level of the incision. In day 2.5 embryos, migrating and lateral trigeminal motor neuroblasts were labeled only when the incision was 8 μm from the metencephalon. Migrating cells did not have somatic processes whereas cells of the lateral nucleus had one dendritc-like process extending dorsally. No cells of the medial column, a cluster of premigratory trigeminal motor neuroblasts, were labeled at this age.In the 3-, 3.5-, 4- and 4.5-day embryos, medial column cells, migratory cells and lateral nucleus cells were retrogradely labeled by this procedure. At all these ages, medial column cells tend to have few somatic filopodia migratory cells tend to have increased filopodia as they proceed laterally, and lateral nucleus cells arc characterized either by a single, long dendrite-like process dorsally directed, or a radiation of short, stubby processes. Axons of medial column and migratory cells take a sinuous course across the metencephalon and frequently exhibit small branches and localized swellings. Axons of lateral nucleus cells rarely have branches within the brain stem and usually enter the motor root at an acute angle from their origin at the soma.The central processes of the trigeminal ganglion cells are also labeled with this procedure. In all embryos these fibers were confined to the trigeminal spinal tract at the level of the trigeminal motor nucleus. Caudally, small fibers were observed to exit this tract in the presumed region of the developing trigeminal spinal nucleus.This study demonstrates that axonal outgrowth into the periphery precedes somatic migration and translocation in the trigeminal motor nucleus During their migration, trigeminal motor neuroblasts appear to be in proximity to axons of adjacent migrating cells. Considerable differentiation occurs during this process.