A projection by the superior colliculus to the supraoculomotor area (SOA) located dorsal to the oculomotor complex was first described in 1978. This projection’s targets have yet to be identified, although the initial study suggested that vertical gaze motoneuron dendrites might receive this input. Defining the tectal targets is complicated by the fact the SOA contains a number of different cell populations. In the present study, we used anterograde tracers to characterize collicular axonal arbors and retrograde tracers to label prospective SOA target populations in macaque monkeys. Close associations were not found with either superior or medial rectus motoneurons whose axons supply singly innervated muscle fibers. S-group motoneurons, which supply superior rectus multiply innervated muscle fibers, appeared to receive a very minor input, but C-group motoneurons, which supply medial rectus multiply innervated muscle fibers, received no input. A number of labeled boutons were observed in close association with SOA neurons projecting to the spinal cord, or the reticular formation in the pons and medulla. These descending output neurons are presumed to be peptidergic cells within the centrally projecting Edinger–Westphal population. It is possible the collicular input provides a signaling function for neurons in this population that serve roles in either stress responses, or in eating and drinking behavior. Finally, a number of close associations were observed between tectal terminals and levator palpebrae superioris motoneurons, suggesting the possibility that the superior colliculus provides a modest direct input for raising the eyelids during upward saccades.

Comparative studies have greatly contributed to our understanding of the organization and function of visual pathways of the brain, including that of humans. This comparative approach is a particularly useful tactic for studying the pulvinar nucleus, an enigmatic structure which comprises the largest territory of the human thalamus. This review focuses on the regions of the mouse pulvinar that receive input from the superior colliculus, and highlights similarities of the tectorecipient pulvinar identified across species. Open questions are discussed, as well as the potential contributions of the mouse model for endeavors to elucidate the function of the pulvinar nucleus.

Metabotropic glutamate receptors (mGluRs) have been shown to be involved in the modulation of retinocollicular neurotransmission. In glaucoma, retinal ganglion cells (RGCs) degenerate, which may have an implication on this transmission as the superior colliculus is their major central target in the much-used rodent models of the disease. We have investigated this using an in vitro slice preparation of the superior colliculus by eliciting field excitatory postsynaptic potentials (fEPSPs) through optic tract stimulation in a rat ocular hypertension model of glaucoma. Application of the group III mGluR agonist L-AP4 reduced the peak amplitude of the fEPSP in superior colliculus slices through presynaptic mechanisms as previously shown in our lab. At 3 and 16 weeks after surgery, there were no significant differences in the effect of L-AP4 on fEPSP peak amplitude in the superior colliculus slices receiving input from the glaucomatous eyes [elevated intraocular pressure (IOP)] compared to those with input from the unoperated eyes (normal IOP). However, at 32 weeks, the fEPSP peak amplitude was reduced to a significantly greater degree during L-AP4 application in the elevated IOP slices compared to normal IOP slices. At all time points, there were no significant changes in the baseline amplitudes of fEPSPs or the stimulus intensities required to evoke fEPSPs. These results suggest that the modulation of synaptic transmission through group III mGluRs on RGC terminals to the superior colliculus is changed at later stages due to RGC degeneration through IOP elevation. These changes may be compensatory changes possibly through plasticity in the RGC terminals of surviving cells, which may be due to increases in the numbers of group III mGluRs. This result may have implications on further treatment studies carried out using these models of glaucoma as changes in the central visual system may need to be considered along with the retinal changes that occur.

Precise targeting of retinal projections is required for the normal development of topographic maps in the mammalian primary visual system. During development, retinal axons project to and occupy topographically appropriate positions in the dorsal lateral geniculate nucleus (dLGN) and superior colliculus (SC). Phr1 retinal mutant mice, which display mislocalization of the ipsilateral retinogeniculate projection independent of activity and ephrin-A signaling, were found to have a more global disruption of topographic specificity of retinofugal inputs. The retinocollicular projection lacks local refinement of terminal zones and multiple ectopic termination zones originate from the dorsal–nasal (DN) retinal quadrant. Similarly, in the dLGN, the inputs originating from the contralateral DN retina are poorly refined in the Phr1 mutant. These results show that Phr1 is an essential regulator of retinal ganglion cell projection during both dLGN and SC topographic map development.

Recent studies on neural pathways in a broad spectrum of vertebrates suggest that, in addition to migration and an increase in the number of certain select neurons, a significant aspect of neural evolution is a “parcellation” (segregation-isolation) process that involves the loss of selected connections by the new aggregates. A similar process occurs during ontogenetic development. These findings suggest that in many neuronal systems axons do not invade unknown territories during evolutionary or ontogenetic development but follow in their ancestors' paths to their ancestral targets; if the connection is later lost, it reflects the specialization of the circuitry.

The pattern of interspecific variability suggests (1) that overlap of circuits is a more common feature in primitive (generalized) than in specialized brain organizations and (2) that most projections, such as the retinal, thalamotelencephalic, corticotectal, and tectal efferent ones, were bilateral in the primitive condition. Specialization of these systems in some vertebrate groups has involved the selective loss of connections, resulting in greater isolation of functions. The parcellation process may also play an important role in cell diversification.

The parcellation process as described here is thought to be one of several underlying mechanisms of evolutionary and ontogenetic differentiation.

Theories of superior collicular and hippocampal function have remarkable similarities. Both structures have been repeatedly implicated in spatial and attentional behaviour and in inhibitory control of locomotion. Moreover, they share certain electrophysiological properties in their single unit responses and in the synchronous appearance and disappearance of slow wave activity. Both are phylogenetically old and the colliculus projects strongly to brainstem nuclei instrumental in the generation of theta rhythm in the hippocampal EEC

On the other hand, close inspection of behavioural and electrophysiological data reveals disparities. In particular, hippocampal processing mainly concerns stimulus ambiguity, contextual significance, and spatial relations or other subtle, higher order characteristics. This requires the use of largely preprocessed sensory information and mediation of poststimulus investigation. Although collicular activity must also be integrated with that of “higher” centres (probably to a varying degree, depending on the nature of stimuli being processed and the task requirements), its primary role in attention is more “peripheral” and specific in controlling orienting/localisation via eye and body movements toward egocentrically labelled spatial positions. In addition, the colliculus may exert a nonspecific influence in alerting higher centres to the imminence of information potentially worthy of focal attention. Nevertheless, it is noteworthy that collicular and hippocampal lesions produce deficits on similar tasks, although the type of deficit is usually different (often opposite) in each case. Functional overlap between hippocampus and colliculus (i.e., strategically synchronised or mutually interdependent activity) is virtually certain vis-à-vis stimulus sampling, for example in the acquisition of information via vibrissal movements and visual scanning. In addition, insofar as stimulus significance is a factor in collicular orienting mechanisms, the hippocampus — cingulate – cortex — colliculus pathway may play a significant role, modulating collicular responsiveness and thus ensuring an attentional strategy appropriate to current requirements (stimulus familiarity, stage of learning). A tentative “reciprocal loop” model is proposed which bridges physiological and behavioural levels of analysis and which would account for the observed degree and nature of functional overlap between the superior colliculus and hippocampus.

The present study describes the patterns of NADPH-diaphorase reactivity in the ventral and dorsal lateral geniculate nuclei of rats. In the ventral lateral geniculate nucleus, two distinct populations of NADPH-diaphorase reactive cells are apparent. One population is deeply stained, generally larger in somal size and located in the more superficial or dorsolateral regions of the nucleus. The second population of reactive cells in the nucleus is lightly labeled, small in somal size, and found in deeper or more ventromedial regions of the nucleus. Double labeling with an antibody to GAB A revealed that neither cell class is GABAergic.

In the dorsal lateral geniculate nucleus, reactivity is apparent in lightly labeled small cells only, most of which are GABA immunoreactive also. The NADPH-diaphorase reactive cells, however, form only a small proportion of the total population of GABAergic cells in the nucleus. The striking feature of the NADPH-diaphorase reactive cells in the dorsal lateral geniculate nucleus is their spatial distribution. Most cells are located in the more superficial or dorsolateral areas: very few are apparent in deeper or more ventromedial areas of the nucleus. This distribution closely parallels the location of the outer “shell” region of the nucleus (see Reese, 1988), which receives most of its afferents from the smaller class II and III ganglion cells of the retina and from the superior colliculus.

Ganglion cells of the cat retina that are neither alpha nor beta cells are often lumped for convenience into a single anatomical group—the gamma cells (Boycott & Wässle, 1974; Stone, 1983; Wässle & Boycott, 1991). Defined in this way, gamma cells are the morphological counterpart to the physiological W-cell class, which includes all ganglion cells that are neither Y (alpha) nor X (beta) cells. We have estimated the retinal distribution of gamma cells by using retrograde transport to label ganglion cells innervating the superior colliculus and by assuming that these included virtually all gamma cells and no beta cells. We excluded labeled alpha cells on the basis of soma size. Our data suggest that gamma cells represent just under half of the ganglion cells in most of the nasal retina, but only about a third of those in the area centralis and temporal retina. Gamma cells do not appear to be more highly concentrated in the nasal visual streak than are other ganglion cells. In the temporal retina, gamma cells with crossed projections to the brain are apparently at least twice as common as those with uncrossed projections.

The distribution of neurons and fibers immunoreactive (ir) to the three calcium-binding proteins parvalbumin (PV), calbindin D-28k (CB), and calretinin (CR) was studied in the human lateral geniculate nucleus (LGN), lateral inferior pulvinar, and optic radiation, and related to that in the visual cortex. In the LGN, PV, CR, and CB immunoreactivity was present in all laminae, slightly stronger in the magnocellular than in the parvocellular laminae for CB and CR. PV-ir puncta, representing transversally cut axons, and CR-ir fibers were revealed within the laminae and interlaminar zones, and just beyond the outer border of lamina 6 in the geniculate capsule. In the optic radiation both PV- and CR-immunoreactive neurons, puncta, and fibers were present. CB immunoreactivity was revealed in neurons of all laminae of the lateral geniculate nucleus, including S lamina and interlaminar zones. There were hardly any CB-ir puncta or fibers in the laminae, interlaminar zones, geniculate capsule, or optic radiation. In the lateral inferior pulvinar, immunoreactive neurons for the three calcium-binding proteins were present in smaller number than in the LGN, as well as PV-ir puncta and CR-ir fibers within the nucleus and in the pulvinar capsule. In the white matter underlying area 17, fibers intermingled with a few scattered neurons were stained for both PV and CR, but very rarely for CB. These fibers stopped at the limit between areas 17 and 18. Area 17 showed a dense plexus of PV-ir puncta and neurons in the thalamo-receptive layer IV and CR-ir puncta and neurons both in the superficial layers I-II, IIIC, and in layer VA. Cajal-Retzius CR-ir neurons were present in layer I. CB-ir puncta were almost confined to layer I-III and CB-ir neurons to layer II. Finally the superior colliculus exhibited mostly populations of PV and CR pyramidal-like immunoreactive neurons, mainly in the intermediate tier. These data suggest that in the visual thalamus most calcium-binding protein immunoreactive neurons project to the visual cortex, while in the superior colliculus a smaller immunoreactive population represent projection neurons.

Eye movements were recorded with the scleral search coil method while striate cortex (area 17) was stimulated in alert cats with their heads fixed. Regardless of where stimulation was applied in the retinotopic map, eye position at the onset of stimulation strongly affected the amplitudes of evoked saccades, but had much less influence on their directions. Application of long stimulus trains evoked repeated saccades at all sites tested. Highly convergent or goal-directed saccades were not observed. Cortically evoked saccades appeared to habituate with repeated stimulation and had higher thresholds and longer latencies that those reported for saccades evoked from the superior colliculus. The directions of cortically evoked saccades generally agreed with those predicted from the retinotopic coordinates of the stimulus sites, but saccade amplitudes were usually lower than expected. It is suggested that these findings are consistent with certain characteristics of eye-head coordination in the cat's normal visual orienting behavior. The results are difficult to reconcile with the hypothesis that goal-directed saccades are a normal response to targets outside the cat's oculomotor range.

The neurotrophins, including nerve growth factor (NGF), brain-derived neurotrophic factor (BDNF), neurotrophin-3 (NT-3) and NT-4/5 are important in a variety of developmental processes in the peripheral and central nervous systems. These molecules bind to a low-affinity receptor and to distinct high-affinity receptors. The high-affinity receptor for NGF is the proto-oncogene product, p140trkA(trkA). Isoforms of p140trkA, p145trkB(trkB), and p140trkC(trkC), are the primary high-affinity receptors for BDNF and NT-3, respectively. We evaluated the developmental regulation of the high-affinity neurotrophin receptors in the rat retina using polyclonal antibodies directed to a highly conserved region of the C-terminus of the p140trkA isoforms (pantrk) and antibodies directed to unique amino-acid sequences of p140trkA, p145trkB, and p140trkC. Immunoreactivities for trkA and trkB, as well as pantrk, were detected in the developing retina and showed similar distributions. At similar antibody concentrations, trkC immunoreactivity was not detected. In the embryo, immunoreactivities were present in cells located throughout the neuroblastic retina, especially in the inner retinal layers, and in fibers in the nerve fiber layer and optic nerve. In the newborn retina, immunoreactivities for these two receptor isoforms were localized to numerous somata in the inner nuclear layer (INL), as well as to cells in the ganglion cell layer (GCL) and axons in the nerve fiber layer and optic nerve. A similar pattern of immunostaining persisted throughout the first postnatal week. By postnatal day-10, immunostaining was confined to large-diameter cells in the GCL, both heavily stained and lightly stained cells in the INL and a plexus of processes in the inner plexiform layer (IPL). In the adult retina, specific immunoreactivity was present in sparsely distributed, lightly and moderately stained, large cells in the GCL, numerous lightly and moderately stained cells throughout the INL and in plexuses of processes in the IPL and outer plexiform layer. Specific immunostaining of photoreceptor cells was not observed. These observations indicate that high-affinity receptors for the neurotrophins are expressed in cells of the inner retina, including ganglion cells, during the period of retinal development. This is congruent with roles for neurotrophins in such processes as survival, differentiation and synapse formation of cells in the developing visual system.

Immunocytochemistry and retrograde labeling were used to define the thalamic projections of calbindin- and parvalbumin-containing cells in superficial layers of the rat's superior colliculus (SC). Quantitative analysis revealed that 90.8 ± 2.2% (mean ± standard deviation) of the calbindin-immunoreactive neurons in the stratum griseum superficiale (SGS) projected to the dorsal lateral geniculate nucleus (LGNd) and that 91.3 ± 4.3% of calbindin-immunoreactive neurons in the stratum opticum (SO) projected to the lateral posterior nucleus (LP). In contrast, only 17.3 ± 2.5% of parvalbumin-immunoreactive neurons in the SGS were found to project to the LGNd and 16.5 ± 3.1% of the parvalbumin-immunoreactive SO cells were retrogradely labeled after LP injections. Few of the parvalbumin-immunoreactive neurons in either the SGS (7.2 ± 2.5%) or the SO (9.2 ± 2.5%) were GABA positive. The retrograde-labeling results suggest that parvalbumin-immunoreactive neurons in the rat's SO and SGS may either be primarily interneurons or have descending projections, while calbindin-containing cells are primarily thalamic projection neurons. These results are consistent with data from other rodents, but almost exactly the opposite of data that have been reported for the cat for these same populations of SC projection neurons. Such interspecies differences raise questions regarding the functional importance of expressing one calcium-binding protein versus another in a specific neuronal population.

Electrical stimulation of the omnipause neuron region (OPN) at saccade onset results in interrupted saccades (IS) —eye movements which pause in midflight, resume after a brief period, and end near the target location. Details on the endpoint accuracy of IS do not exist, except for a brief report by Becker et al. (1981). Their analysis emphasized the accuracy of IS relative to the visual target which remained on during the interrupted period. We instead quantified the metric properties of IS relative to nonstimulated saccades during a target flash paradigm. Our results show that IS tend to be slightly hypermetric relative to the nonstimulated saccades to the same target location. The amount of overshoot is not correlated with target eccentricity. Detailed analyses also indicate that the standard deviations of the endpoint in IS are not significantly larger than those for nonstimulated saccades, although there was a much larger variability produced in eye position during the interruption. Both these latter observations support the notion that saccades are controlled by an internal negative feedback system. Also, the size of the remaining motor error during the interrupted period is one factor influencing when an IS resumes, but the variability in this measure is large particularly for smaller motor errors. Recent results have suggested that the resettable neural integrator involved in the feedback loop may be reset after each saccade through an exponential decay process. To probe the properties of the neural integrator, we varied the duration of interruption between the initial and resumed saccades and sought a systematic overshoot in the final eye position with increasing interruption period and variable initial saccade size. Our results showed the neural integrator does not decay during the pause period of interrupted saccades.

Inferior frontal eye field (FEF) projections to the dorsolateral pontine nucleus (DLPN), and corticocortical connections with the superior temporal sulcal (STS) cortex, were studied in five macaque monkeys which had received horseradish peroxidase (HRP) gel implants into the inferior prearcuate cortex (including area 45 of Walker, 1940). These connections were contrasted with those from the dorsal FEF (area 8a) in another macaque monkey. Findings of heavy inferior FEF projections to the ipsilateral DLPN (light to the contralateral DLPN) and reciprocal connections with the deep caudal bank and fundus of the superior temporal sulcus (STS), presumed to be the middle temporal (MT) visual area (Maunsell & Van Essen, 1983a), appeared to go hand in hand with more pronounced projections to the stratum superficialis of the superior colliculus (SC). In contrast, the HRP gel implant in the dorsal prearcuate cortex (area 8a of Walker, 1940) resulted in only very light projections to the ipsilateral DLPN, more pronounced projections to the dorsomedial pontine nucleus (DMPN), almost no projection to the stratum superficialis (SS), and more pronounced reciprocal connections with the upper bank of the STS, presumed to be the medial superior temporal (MST) area (Maunsell & Van Essen, 1983a). Both the inferior and dorsal FEF also had extensive reciprocal connections with the ventral intraparietal area (VIP; Maunsell & Van Essen, 1983a) in the caudal bank of the intraparietal sulcus. The correlated projections of the inferior FEF to the DLPN, MT area, and SS may explain its reported role in smooth pursuit (Lynch, 1987), in addition to its well-established role in the production of voluntary purposeful saccadic eye movements (Bruce et al., 1985).

The projection of the perihypoglossal (PH) complex to the superior colliculus (SC) in the rhesus monkey was investigated using the retrograde transport of wheat germ agglutinin conjugated to horseradish peroxidase (WGA-HRP). Following physiological identification by electrical stimulation and multiunit recording, small injections of the tracer were placed within the SC of three monkeys. The largest numbers of retrogradely labeled neurons within the PH complex were found in the contralateral nucleus prepositus hypoglossi (NPH), in the laterally adjacent medial vestibular nucleus, and in the ventrally adjacent reticular formation (the nucleus reticularis supragigantocellularis). These labeled neurons are strikingly heterogeneous in size and morphology. The nuclei supragenualis and intercalatus also contain numerous labeled neurons in the 2 cases in which the injections involve the caudal SC. Large numbers of retrogradely labeled neurons as well as anterogradely transported WGA-HRP are observed alo throughout the pontine and medullary reticular formation, including the midline raphe. The PH complex, particularly the NPH, is known to be involved in the coding of eye position and has been hypothesized to be a critical component of the “neural integrator.” Our data demonstrate the existence of a robust projection from the PH complex to the contralateral SC in the rhesus monkey. This projection may serve as the anatomical substrate by which a corollary of eye position could reach the SC. Such a signal is a prerequisite for the computation, at the collicular level, of saccadic motor error signals observed in the SC of rhesus monkeys.

This paper reviews evidence that the superior colliculus (SC) of the midbrain represents visual direction and certain aspects of saccadic eye movements in the distribution of activity across a population of cells. Accurate and precise eye movements appear to be mediated, in part at least, by cells of the SC that have large sensory receptive fields and/or discharge in association with a range of saccades. This implies that visual points or saccade targets are represented by patches rather than points of activity in the SC. Perturbation of the pattern of collicular discharge by focal inactivation modifies saccade amplitude and direction in a way consistent with distributed coding. Several models have been advanced to explain how such a code might be implemented in the colliculus. Evidence related to these hypotheses is examined and continuing uncertainties are identified.

In visual neurones of the superficial layers of the superior colliculus (SSC), repetitive stimulation causes a progressive decline in the size of the response to the stimulus, usually known as response habituation or response adaptation. A mechanism has been proposed in which habituation results from coactivation of excitatory and inhibitory neurones, and the responses of the inhibitory neurones block the response to subsequent stimulus presentations. Excitatory amino acid (EAA) neurotransmitters mediate visual responses via NMDA and non-NMDA receptors in cat SSC. We have investigated the role of these receptors in the generation of response habituation. Following the iontophoretic application of the EAA antagonists CNQX, APS or CPP, repetitive visual stimulation paradigms which normally produce response habituation no longer do so. Indeed the response to each presentation of the stimulus is similar. Intravenous administration of the dissociative anesthetic ketamine (2–10 mg/kg) had similar actions to iontophoretically applied NMDA antagonists. The data imply that intracollicular mechanisms activated by NMDA and non-NMDA receptors contribute to the generation of the inhibitory responses in SCC which lead to response habituation. Furthermore, the effects seen with ketamine anesthesia suggest that the use of ketamine in studies of sensory systems may result in the lack of habituation.

The early development of the optic tract in hamsters was studied by labeling retinal axons with Dil applied to the eye, and then examining the labeled axons in flatmount preparations of the rostral brain stem. This technique permits a panoramic view of the entire retinal projection, from the chiasm to the caudal end of the superior colliculus. In the Ell embryo, retinal axons have reached the chiasm. They defasciculate as they emerge from the nerve, prior to reaching the ventral midline of the diencephalon, then converge again as they pass over to the opposite side. At the midline, many axonal trajectories crisscross, implying some shuffling of relative positions. Retinal axons are tightly bundled within the optic tract. Upon reaching the ventral border of the lateral geniculate body (LGB), they splay out over the nucleus, revealing a wavefront of pioneer axons individually distributed across the rostro-caudal extent of the LGB. Later-emerging retinal axons course over the surface of the thalamus in waves; subsequent waves of axons interdigitate between the lead fibers without fasciculating along them. Past the LGB, the axons undergo a second change in relative positions as the ribbon of fibers swerves caudally, prior to entering the superior colliculus. Retinal axons are tipped with growth cones of varying morphologies. No strong correlation is evident between the Structural complexity of the growth cone and its position within the tract. In the majority of cases, ipsilaterally and contralaterally directed axons follow a similar developmental course along the optic tract, without any indication of a temporal lag in the ipsilateral projection as claimed in earlier reports. Understanding the changes in spatial distribution of embryonic retinal axons as they navigate along the optic tract provides a further step towards elucidating how point-to-point projections form in developing sensory systems.

While much attention has been given to the correlation between cytochrome-oxidase (CO) compartments and patterns of cortico-cortical projections originating from supragranular layers in the striate cortex, little is known in this regard about patterns of cortico-subcortical projections originating from infragranular cortex. We studied the tangential distribution of the striate cortex neurons projecting to the superior colliculus and used two approaches to analyze the relationship of this distribution to the arrangement of CO “blobs.” First, chi-square analysis indicated that significantly fewer labeled neurons were found within the CO blob compartment than the number expected for a random distribution. Second, spatial cross-correlation analysis – which circumvents the inherent subjectivity of delineating blob boundaries – revealed an area around blob centers in which there was a decreased probability of encountering labeled cells. The size of this area compared well with that of our outlines of CO blobs. We conclude that corticotectal projection neurons in the striate cortex are distributed preferentially within the interblob compartment of the infragranular striate cortex. These results demonstrate that the spatial distribution of cortico-subcortical projection neurons within infragranular cortex can correlate with the CO architecture of the primary visual cortex.