Cortical projects of two types of principal cells of the rat lateral geniculate body

Morphology, classification, and distribution of the projection neurons in the dorsal lateral geniculate nucleus of the rat

The morphology of confirmed projection neurons in the dorsal lateral geniculate nucleus (dLGN) of the rat was examined by filling these cells retrogradely with biotinylated dextran amine (BDA) injected into the visual cortex. BDA-labeled projection neurons varied widely in the shape and size of their cell somas, with mean cross-sectional areas ranging from 60–340 µm². Labeled projection neurons supported 7–55 dendrites that spanned up to 300 µm in length and formed dendritic arbors with cross-sectional areas of up to 7.0×10⁴ µm². Primary dendrites emerged from cell somas in three broad patterns. In some dLGN projection neurons, primary dendrites arise from the cell soma at two poles spaced approximately 180° apart. In other projection neurons, dendrites emerge principally from one side of the cell soma, while in a third group of projection neurons primary dendrites emerge from the entire perimeter of the cell soma. Based on these three distinct patterns in the distribution of primary dendrites from cell somas, we have grouped dLGN projection neurons into three classes: bipolar cells, basket cells and radial cells, respectively. The appendages seen on dendrites also can be grouped into three classes according to differences in their structure. Short “tufted” appendages arise mainly from the distal branches of dendrites; “spine-like” appendages, fine stalks with ovoid heads, typically are seen along the middle segments of dendrites; and “grape-like” appendages, short stalks that terminate in a cluster of ovoid bulbs, appear most often along the proximal segments of secondary dendrites of neurons with medium or large cell somas. While morphologically diverse dLGN projection neurons are intermingled uniformly throughout the nucleus, the caudal pole of the dLGN contains more small projection neurons of all classes than the rostral pole.

Consistent phosphenes generated by electrical microstimulation of the visual thalamus. An experimental approach for thalamic visual neuroprostheses

Most work on visual prostheses has centered on developing retinal or cortical devices. However, when retinal implants are not feasible, neuroprostheses could be implanted in the lateral geniculate nucleus (LGN) of the thalamus, the intermediate relay station of visual information from the retina to the visual cortex (V1). The objective of the present study was to determine the types of artificial stimuli that when delivered to the visual thalamus can generate reliable responses of the cortical neurons similar to those obtained when the eye perceives a visual image. Visual stimuli {S(i)} were presented to one eye of an experimental animal and both, the thalamic {RTh(i)} and cortical responses {RV1(i)} to such stimuli were recorded. Electrical patterns {RTh(i)*} resembling {RTh(i)} were then injected into the visual thalamus to obtain cortical responses {RV1(i)*} similar to {RV1(i)}. Visually- and electrically generated V1 responses were compared.

RESULTS

During the course of this work we: (i) characterized the response of V1 neurons to visual stimuli according to response magnitude, duration, spiking rate, and the distribution of interspike intervals; (ii) experimentally tested the dependence of V1 responses on stimulation parameters such as intensity, frequency, duration, etc., and determined the ranges of these parameters generating the desired cortical activity; (iii) identified similarities between responses of V1 useful to compare the naturally and artificially generated neuronal activity of V1; and (iv) by modifying the stimulation parameters, we generated artificial V1 responses similar to those elicited by visual stimuli. Generation of predictable and consistent phosphenes by means of artificial stimulation of the LGN is important for the feasibility of visual prostheses. Here we proved that electrical stimuli to the LGN can generate V1 neural responses that resemble those elicited by natural visual stimuli.

Electrophysiological and neurochemical study of the rat geniculo-cortical pathway. Evidence for glutamatergic neurotransmission

The projection from the dorsal lateral geniculate nucleus to the primary visual cortex of the rat was studied electrophysiologically. Electrical stimulation of the dorsal lateral geniculate nucleus and the optic tract produced three types of responses on neurons of area 17: excitation followed by inhibition, excitation and inhibition. These results extend and confirm, in adult rats, previous studies done in rat geniculate-visual cortex cocultures preparations in vitro. The role of glutamate in the neurotransmission of the rat geniculo-cortical pathway was also investigated. In a first set of experiments, the effects of kynurenate, an antagonist of glutamate receptors, on visual cortex neurons with a monosynaptic excitatory response to dorsal lateral geniculate nucleus stimulation were studied. Microiontophoresis of kynurenate in area 17 neurons selectively suppressed the excitatory response to dorsal lateral geniculate nucleus and optic tract stimulation. In a second set of experiments, the effects of electrical stimulation of the dorsal lateral geniculate nucleus and the optic tract on the release of amino acids in the rat visual cortex in vivo were studied. Using the push-pull method, we perfused a discrete region of the visual cortex with artificial cerebrospinal fluid (CSF), and the amino acid content of the perfusates was analysed by high performance liquid chromatography (HPLC). Stimulation of either the dorsal lateral geniculate nucleus or the optic tract significantly increased glutamate release in area 17. The rest of the amino acids studied did not show significant changes. The results provide evidence for the participation of glutamate in the neurotransmission of the geniculo-cortical pathway in the rat.

Rat and human visual-evoked potentials recorded under comparable conditions: a preliminary analysis to address the issue of predicting human neurotoxic effects from rat data

Pattern-onset visual-evoked potentials (VEPs) were recorded from rats and humans in order to perform cross-species comparison of neuronal functional properties reflected by the early VEP components. The spatial frequency of a sinusoidal test grating was varied in Experiment 1. For both species, amplitude of the first positive VEP component was larger at low spatial frequency and decreased as spatial frequency increased. The immediately succeeding negative component was small at low spatial frequency and was of maximal amplitude at moderate spatial frequency. The effects of stationary pattern adaptation on these components were investigated in Experiment 2. Subjects viewed either a blank field or the test grating prior to recording VEPs. For both species, adaptation had no effect on the positive component but strongly attenuated the negative component. Experiment 3, in which only humans were tested, indicated that the negative component was of cortical origin. Only cortical neurons are known to be orientation selective, and the effect of adaptation diminished as the orientation difference between the adaptation and test gratings increased. These results suggest that the early positive and negative components arise from parallel visual pathways, and that the rat components may reflect visual processes qualitatively similar to those of humans.

Metabolic activity in striate and extrastriate cortex in the hooded rat: contralateral and ipsilateral eye input

The extent of changes in glucose metabolism resulting from ipsilateral and contralateral eye activity in the posterior cortex of the hooded rat was demonstrated by means of the C-14 2-deoxyglucose autoradiographic technique. By stimulating one eye with square wave gratings and eliminating efferent activation from the other by means of enucleation or intraocular TTX injection, differences between ipsilaterally and contralaterally based visual activity in the two hemispheres were maximized. Carbon-14 levels in layer IV of autoradiographs of coronal sections were measured and combined across sections to form right and left matrices of posterior cortex metabolic activity. A difference matrix, formed by subtracting the metabolic activity matrix of cortex contralateral to the stimulated eye from the ipsilateral "depressed" matrix, emphasized those parts of the visual cortex that received monocular visual input. The demarcation of striate cortex by means of cholinesterase stain and the examination of autoradiographs from sections cut tangential to the cortical surface aided in the interpretation of the difference matrices. In striate cortex, differences were maximal in the medial monocular portion, and the lateral or binocular portion was shown to be divided metabolically into a far lateral contralaterally dominant strip along the cortical representation of the vertical meridian, and a more medial region of patches of more or less contralaterally dominant binocular input. Lateral peristriate differences were less than those of striate cortex, and regions of greater and lesser monocular input could be distinguished. We did not detect differences between the two hemispheres in either anterior or medial peristriate areas, thus indicating either completely binocular input (which seems unlikely given the retinotopic organization of these regions), or a greater dependence than in the lateral peristriate on inputs that were not affected by the visual manipulations.

Two types of thalamic reticular cells in relation to the two visual thalamocortical systems in the rat

We found in urethane-anesthetized rats that thalamic reticular (TR) cells responding to an electrical stimulus of the optic tract (OT) can be further subdivided into two types, viz. S- and L-type cells. S-type cells, which were selectively excited from area 17 of the visual cortex, were characterized by short latency responses (2.3-6.1 ms) to OT stimulation. TR cells activated antidromically from the dorsal lateral geniculate nucleus were all classified as S-type. Long OT latencies (5.2-15.3 ms) and selective excitation from area 18a were peculiar to L-type cells, which showed antidromic responses to the lateral posterior nucleus stimulation. Mapping studies documented that cells belonging to each type were segregated in the thalamic reticular nucleus; L-type cells were located in the most posterior part. It is suggested that S- and L-type cells are inhibitory interneurons modulating activity of geniculocortical and extrageniculocortical projection cells, respectively.

'Hidden lamination' in the dorsal lateral geniculate nucleus: the functional organization of this thalamic region in the rat

The cyto-and myeloarchitecture of the rat's dorsal lateral geniculate nucleus (dLGN) display none of the laminar features characteristic of this thalamic region in carnivores and primates. Despite this, the rodent's nucleus contains a segregation of functionally and ocularly distinct afferents--organizational properties manifested in the prominent lamination of these other mammalian forms. The rat's dLGN can be divided into two main regions: an inner core and an outer shell. The inner core contains two ocular laminae receiving direct retinotopic projections from the contralateral nasal and ipsilateral temporal retinae, mapping the contralateral visual hemifield. The outer shell receives a retinotopic projection from the complete contralateral retina only, the representation of the ipsilateral hemifield being extremely compressed at the medial edge of this lamina. The retinotopic maps in these three ocular laminae (contra, ipsi, contra) are in conjugate register, so that lines of projection course rostro-ventro-medially from the optic tract at the thalamic surface through these laminae. Three morphologically distinct retinal ganglion cell types project to the dLGN, and the axons of these ganglion cells are partially segregated within the optic tract in anticipation of their segregation within the nucleus, where they terminate at distinct locations along the lines of projection. Type I and III cells terminate in the inner core of the nucleus, while type II and III cells terminate in the outer shell. The outer shell also receives a direct projection from the superior colliculus. These characteristics of the afferent termination within the rat's dLGN support the view of a general mammalian plan for the organization of this thalamic region, and provide a basis for further experimentation to test speculations about potentially homologous subdivisions of this nucleus. Conclusions regarding functionally analogous pathways are proposed with less confidence, due to the paucity of definitive evidence for physiologically distinct cell classes. The type I cells in the rat's retina are the likely homologues of the cat's alpha-cell. Geniculocortical relay cells driven by them have properties similar to the cat's Y-cell. The inner core of the nucleus then may transmit information of a Y-like nature onto striate cortex. The outer shell of the rat's nucleus, a portion of which receives collicular as well as retinal innervation, may convey W-like information onto striate cortex. The rat's retinogeniculate projection appears to be lacking a beta-cell-like pathway that may subserve X-cell function altogether.

Complex convolutions in neurons of the dorsal lateral geniculate nucleus of the normal albino rat

The postnatal development of complex convolutions (CCs) of the dorsal lateral geniculate nucleus (LGNd) in normal rats has been studied quantitatively with light microscopy. We report that immature neurons do not contain these scarcely understood organelles, since they can be seen for the first time in very few, mature neurons of the 30 day rat; their number constantly increases during the following 4 months. These cytoplasmic inclusions can be equally seen in the aged rat. CCs are present in neurons of all sizes, except the smallest, which correspond to the interneuron population. Although, morphologically, CCs of the LGNd of the rat are similar, but not identical, to the cytoplasmic multilaminated bodies of the cat, intermediate forms are described.

Neurons of the dorsal lateral geniculate nucleus of the hereditary microphthalmic rat: a Golgi study

Neurons of the dorsal lateral geniculate nucleus (LGNd) in the microphthalmic rat were examined by the Golgi-Cox method. LGNd neurons in the microphthalmic rat were classified into the multipolar (I) and bipolar (II) types as in the normal rat. The multipolar type was further divided into two subclasses (Ia and Ib) on the basis of their dendritic patterns. The proximal portion of their primary dendrites was thinner than in normal LGNd neurons. The Ia cells had 6-7 primary dendrites extending radially, while the Ib cells had 3-4 primary dendrites spreading primarily parallel to the optic tract. Type II cells had two or three primary dendrites emerging from the cell bodies. In both types, primary dendrites were shorter in length or less branched than usual. These results suggested that LGNd neurons in the microphthalmic rat had smaller dendritic fields than those in the normal rat.

Bilateral projections of neurons in the lateral geniculate nucleus and nucleus lateralis posterior to the visual cortex in the neonatal rat

The projection from the lateral geniculate nucleus (LGN) and nucleus lateralis posterior (LP) to the visual cortex was examined in rat pups 3-7 days of age using the fluorescent tracers True Blue, Fast Blue and Nuclear Yellow. Our data provide the first evidence that (1) these projections are bilateral, (2) the ipsilateral projection from these nuclei to the visual cortex in the neonatal rat is well localized and is similar in distribution and organization to that reported by others in the adult and (3) bilaterally projecting geniculocortical cells are morphologically heterogeneous; bilaterally projecting cells in LP are morphologically homogeneous.

Morphology of identified relay cells and interneurons in the dorsal lateral geniculate nucleus of the rat

The morphology of neurons in the lateral geniculate nucleus of the rat has been examined in both Golgi impregnated and in horseradish peroxidase (HRP) filled material. Two major classes of neurons are seen in Golgi material which encompass the variety of cells described in previous reports. Cells of one group (class A) are routinely labelled by injections of HRP into the visual cortex or optic radiations. This group also displays some morphological variation which may be related to the presence of parallel information channels in the retino-fugal pathway, but clear subgroups cannot be identified on the basis of morphological criteria alone. Cells of the other group (class B) are not labelled by HRP injections into visual cortex or the optic radiations, and are probably local circuit interneurons.

Quantitative analysis of the striate cortex in the mutant microphthalmic rat

A quantitative analysis of the striate cortex of the mutant microphthalmic rat was conducted to determine whether or not transneuronal changes of the visual cortex were induced following the loss of eyes. The area of the striate cortex in the microphthalmic rat was approximately 60% of that in the normal rat. As for the thickness of each layer of the striate cortex, many layers of microphthalmia tended to be thin in comparison with the normal animal, except for layers I and III: the thickness of layers II, IV, V, and VI was about 74, 62, 82, and 82% of normal values, respectively. There was fractically no difference between the number of neurons of each layer of the microphthalmic and the normal striate cortex per unit (10(4) microns2), except for layer IV, in which the density had increased to 117% of the normal value. In many layers, the neurons of the microphthalmic striate cortex were smaller than normal and they had narrow neuroplasmic space. Our study demonstrated that the striate cortex of the microphthalmic rat underwent quantitative and morphometric transneuronal changes. Especially striking changes of the striate cortex were found in the inner granular layer with a reduction in thickness and a diminution of cell size.

Bilateral changes in soma size of geniculate relay cells and corticogeniculate cells after neonatal monocular enucleation in rats

Soma areas of relay cells of the lateral geniculate nucleus and of corticogeniculate cells of normal rats (n = 4) were compared with those of neonatally unilaterally eye-enucleated adult rats (n = 13). These cells were labeled by retrogradely transported HRP. Monocular enucleation was performed on postnatal days 1 (PND 1) (n = 4), 3 (PND 3) (n = 5) and 6 (PND 6) (n = 4). The results are summarized as follows. In PND 1 rats soma areas of relay cells were 12-16% smaller than those of normal rats, but only for the geniculate nucleus ipsilateral to the remaining eye. In PND 3 and 6 rats the areal shrinkage of relay cells was 27-39% of the normal control for both hemispheres, though it was less marked in the hemisphere contralateral to the remaining eye. The corticogeniculate cells were distributed in layers V and VI in eye-enucleated rats as well as in normal rats. Soma areas of both layer V and VI cells increased in PND 1 rats for both hemispheres by about 15-47% of the normal control. In PND 3 rats increase in soma size tended to occur for layer VI cells, although the data varied from animal to animal. In summary, it was established that unilateral eye-enucleation in rats at birth induced soma size changes of the geniculate relay cells and of the corticogeniculate cells in the non-deafferented as well as in the deafferented hemisphere. Possible mechanisms for the bilateral changes in soma area of central visual cells after neonatal monocular enucleation are discussed.

Projection lines and the ipsilateral retino-geniculate pathway in the hooded rat

The organization of the hooded rat's dorsal lateral geniculate nucleus was studied with anatomical techniques, with particular regard to the representation of temporal retina and the binocular field. The ipsilateral and contralateral retinal terminal fields were examined in three stereotaxic planes following injections of horseradish peroxidase into the eye. Projections arising from the temporal crescent of the retina were studied with silver staining techniques for anterograde degeneration products. Following discrete retinal lesions there was clear evidence that the temporal retina projects in a topographic fashion both ipsilaterally and contralaterally. The orientation of the lines of projection in the dorsal lateral geniculate nucleus was assessed by retrograde labelling of cells after cortical implants of horseradish peroxidase. Although both the lines of projection and the ipsilateral terminal field extend rostro-caudally in the dorsal lateral geniculate nucleus, their paths are oblique rather than parallel. Their intersection appears to correspond to the representation in this nucleus of conjugate retinal points. This was confirmed by administering horseradish peroxidase by iontophoresis in either the binocular or monocular representation of the primary visual cortex, while one eye received an injection of [3H]proline. Only those cortical injections in the binocular region gave rise to labelled projection lines passing through the autoradiographically-labelled ipsilateral terminal field. The rat's dorsal lateral geniculate nucleus displays none of the cytoarchitectural lamination which is so prominent in the primate and cat. Even after labelling the input to the nucleus from one eye, there is still no obvious laminar relationship between the terminal fields from the two eyes. Despite the absence of lamination, the current results suggest that the principle of apposing the representation of conjugate retinal points in the dorsal lateral geniculate nucleus is the same in the rat as in cat and monkey.

Visual pathways and acuity hooded rats

Three experiments on the effects of lesions of the visual system on contrast-detection in hooded rats are described, in which the ability of rats to detect stationary high-contrast square-wave gratings of various fundamental frequencies presented in the central visual field was measured before and after operation. The results suggested the following conclusions: (i) The pathways from retina to striate cortex via dorsal lateral geniculate nucleus (dLGN) conveys information about high spatial frequencies sufficient for normal detection of these gratings, that is up to about 1 cycles/deg. It may be the only pathway to carry this information, and may thus play a unique role in the analysis of fine detail. The high-frequency information is probably relayed from striate cortex to extrastriate cortex, rather than to subcortical sites. (ii) The superior colliculus receives information from the retina up to at least 0.7 cycles/deg, which it then relays to extrastriate visual cortex, probably via the lateral posterior nucleus of the thalamus. (iii) Neither the projections from superior colliculus to other, non-thalamic sites nor the remaining pathways from the retina (e.g. to ventral LGN) appear to carry contrast information higher than 0.3 cycles/deg. These sets of projections therefore do not appear to be used for precise analysis of stationary scenes. These findings suggest that there are considerable similarities between the visual systems of rats and other mammals with respect to the routing of information about stationary spatial contrast, and may help to explain the results of some experiments that have used tasks besides contrast-detection to assess the visual capacities of rats after lesions.

Plasticity and interaction after ablations of visual or somatosensory motor cortex or retina in neonatal rats

The results obtained following eye enucleation and dorsal occipital cortical (DOC) lesion in normal adult animals are in general agreement with those reported in previous work. The details, however, vary somewhat. A DOC projection to the claustrum and reticularis thalami is reported here for the first time in the rat. Six experiments were performed: (1) DOC lesions followed by eye enucleations. (2) Eye enucleations followed by DOC lesions. (3) Eye enucleations followed by sensorimotor cortical (SMC) lesions. (4) SMC lesions followed by eye enucleations. (5) DOC lesions followed by SMC lesions. (6) SMC lesions followed by DOC lesions. The results obtained from the experimental animals demonstrate a certain degree of plasticity and interaction between the DOC and retinal efferents to the superior colliculus, lateral geniculate nucleus, the pretectal region and the superior colliculus. Thus, following a neonatal left dorsal occipital cortical lesion, the retinal efferents from the right eye produce a denser projection to the contralateral lateralis posterior (LP), nucleus lateralis of the optic tract (NLOT), and the caudal part of the lateral division of the lateral geniculate nucleus pars ventralis (VGN). Following neonatal left eye enucleation, increased ipsilateral projection from the right DOC was noted in the superficial part of the stratum griseum superficiale (SGS), NLOT, pretectal nucleus (PT) and the adjoining stratum griseum intermediate (SGI) of the superior colliculus, lateral geniculate nucleus, pars dorsalis (DGN), and both the medial and lateral halves of the caudal portion of VGN. Following a neonatal DOC lesion a denser projection from the remaining ipsilateral SMC was seen in the medial pretectal nucleus (PTM), PT and deep pretectal nucleus (PTP). Following a neonatal SMC lesion a denser projection from the ipsilateral DOC was seen in the SGI, PT, nucleus medialis of the optic tract (NMOT) and PTM. There was no change in the retinal efferents following neonatal SMC lesions. Nor was there any change in the SMC efferents following neonatal eye enucleations.

Visual acuity in hooded rats: effects of superior collicular or posterior neocortical lesions

The visual resolution acuity of hooded rats was measured with an avoidance technique, using large, high contrast square-wave gratings of high mean luminance. Measurements were taken before and after ablation of either posterior cortex or the superior colliculus. The cortical lesions included both striate and temporal cortex, and caused retrograde degeneration throughout the dorsal lateral geniculate nucleus. Neither group showed signs of detecting even coarse square-wave gratings when first tested after operation. The animals with collicular lesions quickly relearnt, and their acuity was unaltered. After extensive training 3 out of 4 cortical animals relearnt to detect gratings, and their acuity was reduced to about one-third of its preoperative value. It seems likely that in rats the geniculocortical pathway carries sufficient information for the normal detection of high spatial frequencies. Whether a pathway from superior colliculus to neocortex via a thalamic relay also carries this information is uncertain.

Architectonic map of neocortex of the normal mouse

The neocortex of the normal mouse has been subdivided into architectonic fields on the basis of its cellular and fiber patterns. The map of medial, retrohippocampal, frontal and insular regions is little different from that of Brodmann as modified in minor ways by Krieg. The map of parietal, occipital and temporal regions follows closely the major rearrangements introduced to Brodmann's map by Krieg. Krieg's map has been modified to give individual status to the barrel fields and to disignate occipital fields around the full circumference of field 17, and temporal fields circumferentially around field 41.