Ultrastructural study of large efferent neurons in the superior colliculus of the cat after retrograde labeling with horseradish peroxidase

Excitotoxic lesions of the superior colliculus preferentially impact multisensory neurons and multisensory integration

The superior colliculus (SC) plays an important role in integrating visual, auditory and somatosensory information, and in guiding the orientation of the eyes, ears and head. Previously we have shown that cats with unilateral SC lesions showed a preferential loss of multisensory orientation behaviors for stimuli contralateral to the lesion. Surprisingly, this behavioral loss was seen even under circumstances where the SC lesion was far from complete. To assess the physiological changes induced by these lesions, we employed single unit electrophysiological methods to record from individual neurons in both the intact and damaged SC following behavioral testing in two animals. In the damaged SC of these animals, multisensory neurons were preferentially reduced in incidence, comprising less than 25% of the sensory-responsive population (as compared with 49% on the control side). In those multisensory neurons that remained following the lesion, receptive fields were nearly twofold larger, and less than 25% showed normal patterns of multisensory integration, with those that did being found in areas outside of the lesion. These results strongly suggest that the multisensory behavioral deficits seen following SC lesions are the combined result of a loss of multisensory neurons and a loss of multisensory integration in those neurons that remain.

Velocity response profiles of collicular neurons: parallel and convergent visual information channels

We have recorded from single neurons in the retinorecipient layers of the superior colliculus of the cat. We distinguished several functionally distinct groups of collicular neurons on the basis of their velocity response profiles to photic stimuli. The first group was constituted by cells responding only to photic stimuli moving at slow-to-moderate velocities across their receptive fields (presumably receiving strong excitatory W-type input but not, or only subthreshold, Y-type input). These cells were recorded throughout the stratum griseum superficiale and stratum opticum and constituted 50% of our sample. The second group of cells exhibited excitatory responses only at moderate and fast velocities (presumably receiving excitatory Y-type but not W-type input). These cells constituted only about 7% of the sample and were located principally in the lower stratum griseum superficiale. The third group of cells was constituted by cells excited over the entire range of velocities tested (1-2000 /s) and presumably received substantial excitatory input from both W- and Y-channels. These cells constituted almost 26% of our sample and were located in the lower stratum griseum superficiale, stratum opticum and the upper part of the stratum griseum intermediale. Overall, cells receiving excitatory Y-type input, i.e. the sum of group two and group three cells, constituted about a third of the sample and their excitatory discharge fields were significantly larger than those of cells receiving only W-type input. A fourth distinct group of collicular neurons was also constituted by cells responding over a wide range of stimulus velocities. These cells were excited by slowly moving stimuli, while fast-moving photic stimuli evoked purely suppressive responses. The excitatory discharge fields of these cells (presumably, indicating the spatial extent of the W-input) were located within much larger inhibitory fields, the extent of which presumably indicates the spatial extent of the Y-input. These low-velocity-excitatory/high-velocity-suppressive cells were recorded from the stratum griseum superficiale, stratum opticum and stratum griseum intermediale and constituted about 17% of the sample. The existence of low-velocity-excitatory/high-velocity-suppressive cells in the mammalian colliculus has not been previously reported. Low-velocity-excitatory/high-velocity-suppressive cells might play an important role in activating "fixation/orientation" and "saccade" premotor neurons recorded by others in the intermediate and deep collicular layers. Overall, in the majority (57%) of collicular neurons in our sample there was no indication of a convergence of W- and Y-information channels. However, in a substantial minority of collicular cells (about 43% of the sample) there was clear evidence of such convergence and about 40% of these (low-velocity-excitatory/high-velocity-suppressive cells) appear to receive excitatory input from the W-channel and inhibitory input from the Y-channel.

Cortical somatosensory and trigeminal inputs to the cat superior colliculus: light and electron microscopic analyses

Two different axonal transport tracers were used in single animals to test the hypothesis that the expansive intermediate gray layer of the cat superior colliculus (stratum griseum intermediale, SGI) is composed of sensorimotor domains. The results show that two sensory pathways, the trigeminotectal and the corticotectal arising from the fourth somatosensory area, commingle in patches across the middle tier of the SGI. Furthermore, the data reveal that tectospinal cells are distributed within these patches. Taken together, these results show a commingling of functionally related afferents and a consistent spatial relationship between these afferents and tectospinal neurons. These relationships indicate that the SGI consists of domains that can be distinguished by their unique combinations of afferent and efferent connections. The ultrastructural characteristics and synaptic relationships of these somatosensory afferent pathways suggest that they have distinct roles within the sensorimotor domain of the SGI. The trigeminotectal terminals are relatively small, contain round vesicles and make asymmetrical synapses on small, presumably distal, dendrites. We submit that these trigeminal terminals bestow the basic receptive field properties upon SGI neurons. In contrast, the somatosensory corticotectal terminals are relatively large, contain round vesicles, make asymmetrical synapses, participate in triads, and are presynaptic to proximal dendrites. We suggest that these cortical terminals bestow integrative abilities on SGI neurons.

Choline acetyltransferase-immunoreactive patches overlap specific efferent cell groups in the cat superior colliculus

Fibers containing acetylcholine (ACh) form distinct patches in the dorsal intermediate gray layer (IGL) of the cat superior colliculus (SC). Although these patches are known to overlap several afferent projections to SC, it is not known whether they are associated with specific postsynaptic cell groups. We have examined the relationship of these ACh fiber patches to specific efferent cell groups by combining retrograde transport of horseradish peroxidase (HRP) with choline acetyltransferase (ChAT) immunocytochemistry. Successful HRP injections were made into the predorsal bundle (PB), the tecto-pontine-bulbar pathway (TPB) and the cuneiform region (CFR), the inferior olive (IO), the dorsolateral pontine gray nucleus (PGD), and the pedunculopontine tegmental nucleus (PPTN). The distribution of HRP-labeled neurons which project to these targets was mapped by a computer-based microscope plotter. Distinct clusters of HRP-labeled neurons in the IGL were seen after three injections into the mesencephalic reticular formation that involved the caudal TPB and cuneiform region (CFR), and after one injection into the medial accessory nucleus of IO. As many as seven clusters of labeled neurons were found in some sections through the caudal one-half of SC after the TPB/CFR injections. Each cluster consisted of 3-20 cells, all of which were small to medium in size. In sections also tested for ChAT, the cell clusters in the TPB/CFR cases were found to overlap precisely the ACh patches in the IGL. In addition, SC neurons projecting to the IO formed clusters above the ChAT patches and in the intermediate white layer (IWL) of SC. None of the other HRP injections produced any obvious cell clusters in the deep layers of SC. These results are the first to show that specific cell groups, distinguished by size and projection site, form clusters that match the patch-like innervation of cholinergic afferents to SC. This modular organization may correspond to saccade-related cells that have also been reported to be organized into clusters in the cat SC.

Correlations between the receptive field properties and morphology of neurons in the deep layers of the hamster's superior colliculus

Extracellular and intracellular recording, receptive field mapping, and intracellular HRP injection techniques were used to define the morphological classes of cells in the deep laminae of the hamster's superior colliculus and to determine whether there are any correlations between the structural and functional characteristics of these neurons. A total of 110 neurons were characterized and reconstructed. Of these, 23.6% (N = 26) were visual, 60% (N = 66) were somatosensory, 0.9% (N = 1) were bimodal (visual-somatosensory), and 15.4% (N = 17) were unresponsive. Of the somatosensory neurons, 72.7% (N = 48) were low threshold, 4.5% (N = 3) had a wide dynamic range, 9.1% (N = 6) responded only to noxious stimulation, and 13.6% (N = 9) had complex somatosensory receptive fields. Deep layer cells were divided into eight morphological classes. These classes were multipolar cells (26.4%, N = 29), bipolar cells (9.1%, N = 10), widefield vertical cells (7.3%, N = 8), horizontal cells (13.6%, N = 15), stellate cells (10.9%, N = 12), ventrally directed cells (5.5%, N = 6), sparse radial cells (17.3%, N = 19), and small sparse radial cells (6.4%, N = 7). Four cells (3.6%) did not fit into this classification scheme. Univariate and multivariate analyses of variance of properties such as soma area, number of branch points, total dendritic length, and volume and orientation of dendritic arbor indicated that these classes were significantly different. However, chi 2 analysis and multivariate analysis of variance indicated no significant relationships between morphological class and either laminar location or receptive field type. There was a significant positive relationship between the possession of dendrites that extended into the superficial laminae and visual responsivity.

Association of efferent neurons to the compartmental architecture of the superior colliculus

The superior colliculus is a layered structure in the mammalian midbrain serving multimodal sensorimotor integration. Its intermediate layers are characterized by a compartmental architecture. These compartments are apparent through the clustering of terminals of major collicular afferents, which in many instances match the heterogeneous distribution of tissue components such as acetylcholinesterase, choline acetyltransferase, substance P, and parvalbumin. The present study was undertaken to determine whether efferent cells observe this compartmental architecture. It was found that subpopulations of both descending and ascending collicular efferents originate from perikarya situated in characteristic positions relative to the collicular compartments defined by elevated acetylcholinesterase activity and that their dendrites appear to be specifically coordinated with the heterogeneous environment. With the specific interlocking of afferent and efferent neurons through spatially distinguished neural networks, the compartmental architecture apparently constitutes an essential element for the determination of information flow in the superior colliculus.

The nigral projection to predorsal bundle cells in the superior colliculus of the rat

Predorsal bundle cells give rise to the major efferent pathway from the superior colliculus to the premotor centers of the brainstem and spinal cord responsible for initiating orienting movements. The activity of predorsal bundle cells is profoundly influenced by an inhibitory pathway from substantia nigra pars reticulata that uses gamma aminobutyric acid (GABA) as a neurotransmitter. The present study examines the morphological basis for this influence of substantia nigra on predorsal bundle cells in the rat. In the first experiments, the laminar distributions of the nigrotectal tract terminals and the predorsal bundle cells were compared. The predorsal bundle cells were labeled by the retrograde axonal transport of horseradish peroxidase from either the decussation of the predorsal bundle or the cervical spinal cord, while the terminations of the pathway from substantia nigra pars reticulata were labeled by anterograde axonal transport from the substantia nigra. Either horseradish peroxidase, wheat germ agglutinin conjugated to horseradish peroxidase, or Phaseolus vulgaris leucoagglutinin were used as anterograde tracers. The results showed that the distributions of both the predorsal bundle cells and the nigrotectal terminals are restricted almost entirely to the intermediate grey layer and that they overlap extensively. Predorsal bundle cells varied in size. Within the areas of maximum overlap, the majority, regardless of size, was closely apposed by nigrotectal terminals. In a second series of experiments, the synaptic contacts between nigrotectal terminals and the tectospinal component of the predorsal bundle were examined in tissue in which both the terminals and the tectospinal cells were labeled for electron microscopy. In the final experiments, the distribution and fine structure of the nigrotectal terminals were compared to those of terminals that had been labeled immunocytochemically with an antibody to glutamic acid decarboxylase, the synthesizing enzyme for GABA. The results showed that nigrotectal terminals contain large numbers of mitochondria and pleomorphic vesicles, and form synaptic contacts with the somas and proximal dendrites of tectospinal cells. These synapses have modest postsynaptic densities. In both their distribution and fine structure, these terminations resemble the glutamic acid decarboxylase immunoreactive terminals that contact tectospinal cells. Taken together, these results support the view that the nigrotectal tract is an important source of GABAergic input to most, if not all, predorsal bundle cells.

Visual, auditory and somatosensory convergence in output neurons of the cat superior colliculus: multisensory properties of the tecto-reticulo-spinal projection

A select population of superior colliculus (SC) neurons receives and integrates information from the visual, auditory and somatosensory systems. Determining which SC neurons comprise this population and where they send their multisensory messages is important in understanding the functional impact of the SC on attentive and orientation behavior. One of the major routes by which the SC influences these behaviors is the tecto-reticulo-spinal tract, a descending pathway that plays an integral role in the orientation of the eyes, ears and head. Of the 182 tecto-reticulo-spinal neurons (TRSNs) encountered in the present study, almost all (94%) responded to sensory stimuli and the overwhelming majority (84%) were multisensory. The present results demonstrate that the TRSN serves as an important link among the different sensory systems and provides a substrate through which they may gain access to the circuitry mediating orientation behavior.

Dendritic and synaptic properties of collicular neurons: a quantitative light and electron microscopical study of Golgi-impregnated cells

The present study deals with a light- and electron microscopic morphometric analysis of Golgi-impregnated neurons in the superior colliculus of rats with the purpose to unravel inter- and intralaminar differences in their dendritic and synaptic organization. In particular, layer IV was studied and compared with its boundary layers III and V. The results show that collicular cells in layer IV basically form a homogeneous population with respect to the number of primary dendrites, the total length of impregnated dendrites, and the diameter, ellipticity, and orientation of dendritic fields and somata of Golgi-impregnated neurons. Somata of reconstructed small cells in layer III and IV as well as V have all a similar density of about 40 synaptic contacts per 100 microns2 surface. However, the cell bodies of large multipolar cells in layer V have a slightly but significantly larger synaptic density (about 50 per 100 microns2). Dendrites of large and small collicular cells had no significantly different synaptic densities (43 and 48 per 100 microns2, respectively). In conclusion, the present results show only minor dendritic and synaptic differences between individual cells in the same layer, as well as in neighboring layers, which implies a low degree of cellular and synaptic intra- and interlaminar differentiation. It is discussed that this organization differs markedly from that in other visual centers, including the collicular homologue, the tectum of lower vertebrates, and the mammalian visual cortex, where pronounced inter- and intralaminar differentiations exist. Such an organization may provide a framework of laminar specificity by which distinct cell types may select a restricted set of input out of all information available. The present quantitative investigation suggests that a similar framework is not present in the superior colliculus.

Spatial relationships of axons arising from the substantia nigra, spinal trigeminal nucleus, and pedunculopontine tegmental nucleus within the intermediate gray of the cat superior colliculus

We have utilized two different anterograde transport methods (Phaseolus vulgaris leucoagglutinin [PHA-L] immunocytochemistry and autoradiography) in the same experiment to compare the sublaminar location and arrangement of tectopetal axons arising from the substantia nigra pars reticulata, the spinal trigeminal nucleus, and the pedunculopontine tegmental nucleus. Our findings reveal that the nigrotectal projection terminates in a patchy fashion within three horizontally oriented sublaminae of the stratum griseum superficiale (SGI), the dorsal, middle and ventral. The middle tier of nigrotectal axons exhibits an exquisite, puzzle-like, complementary spatial relationship with trigeminotectal axons. In contrast, axons arising from the pedunculopontine tegmental nucleus overlap with patches of nigrotectal axons within the middle tier. Thus the middle tier of the SGI consists of domains of overlapping nigral and pedunculopontine tegmental inputs which interdigitate with domains rich in somatosensory inputs.

Sources of subcortical GABAergic projections to the superior colliculus in the cat

The goal of this study was to identify GABAergic input to the cat superior colliculus from neurons located in the caudal diencephalon, mesencephalon, pons and medulla. Cells efferent to the superior colliculus were labeled retrogradely with the tracer horseradish peroxidase, and an antibody to gamma-aminobutyric acid was used to label GABAergic neurons in the same sections. The results indicate that neurons in several distinct areas of the caudal diencephalon and brainstem are both immunocytochemically labeled for GABA and retrogradely labeled with horseradish peroxidase. These areas include zona incerta, nucleus of the posterior commissure, anterior and posterior pretectal nuclei, nucleus of the optic tract, superior colliculus, cuneiform nucleus, subcuneiform area, substantia nigra pars reticulata and pars lateralis, periparabigeminal area, external nucleus of the inferior colliculus, the area ventral to the external nucleus of the inferior colliculus, mesencephalic reticular formation, dorsal and ventral nuclei of the lateral lemniscus, and the perihypoglossal nucleus. The role that such diverse inhibitory input to the superior colliculus might play, particularly in influencing eye movements, is discussed.

A light- and electron-microscopic investigation of the optic tectum of the frog, Rana pipiens, II: The neurons that give rise to the crossed tecto-bulbar pathway

The superficial layers of the frog's optic tectum, Potter's (1969) layers A-G, comprise a complex neuropil made up of many afferent axons, the somata of a few neurons, and many dendrites from the neurons located in the deeper layers. Different types of retinal axons are believed to terminate in different layers (Maturana et al., 1960; Kuljis & Karten, 1988; Sargent et al., 1989), but little is known about the relationships between each type of input and the dendrites of the deep tectal neurons that extend into these superficial layers. The present study used the method of retrograde transport of horseradish peroxidase to study the synaptic contacts on the dendrites of the neurons that give rise to the crossed tecto-bulbar pathway. These cells have apical dendrites that ascend through the superficial retino-recipient layers. The somata of the cells that give rise to the crossed tecto-bulbar pathway are located in the superficial half of layer 6, preferentially clustered along the caudal, lateral, and rostral margins of the tectum. The somata of these cells range from 8-30 microns in diameter. Their axons are large (2-4 microns in diameter) myelinated fibers that arise from either their somata or proximal dendrites. Their axons travel within the deep medullary layer to leave the tectum at the lateral margin. Their dendritic arbors extend obliquely through the superficial layers to reach layer B where they turn and extend within the layer for up to 0.5 mm. The somata of these cells receive only a scant synaptic input. In contrast, their dendrites receive input in every layer, but the nature of this input varies from layer to layer. Synaptic terminals that resemble retinal ganglion cell boutons contact the labeled dendrites in layers B, F, and G. This indicates that the dendrites may receive monosynaptic input from several types of retinal ganglion cells. Terminals with small, flattened vesicles also contact the dendrites of these cells in each layer. In layer F and below, the terminals with flattened vesicles constitute 15% of the contacts; above layer F they constitute only 5-8% of the contacts. Terminals with medium-sized, flattened vesicles also contact the dendrites of these cells in every layer and constitute a large proportion of their input (33-95%). The latter terminals resemble those that are often postsynaptic to retinal terminals.