Printed in Great Britain. All rights reserved 0306-4522/99 $20.00+0.00
VELOCITY RESPONSE PROFILES OF COLLICULAR NEURONS: PARALLEL AND CONVERGENT VISUAL INFORMATION CHANNELS
W. J. WALESZCZYK, C. WANG, W. BURKE and B. DREHER
Department of Anatomy and Histology, Department of Physiology, Institute for Biomedical Research, The University of Sydney, Sydney, NSW 2006, Australia
Abstract�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 tnput). 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 �2000û/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.
Key words: cat�s superior colliculus, excitatory responses. suppressive responses, superficial and intermediate layers, parallel channels, convergence of W- and Y-channels.
The superior colliculus (SC), the principal retinorecipient nucleus of the mammalian mesencephalon, plays an impor-tant role in the selection of targets for visually guided beha-viour, integration of sensory information and initiation of goal-directed orientation responses towards novel sensory stimuli (for reviews see Refs 81, 82 and 93). The superficial layers of the SC (stratum zonale; stratum griseum superfi-ciale, SGS; and stratum opticum, SO) receive massive visual inputs which originate either directly in the retinas or are relayed mainly, but not exclusively, via the dorsal lateral geniculate nuclei (LGNd) and a number of visual cortical areas, again mainly but not exclusively in the ipsilateral hemisphere (for reviews see Refs 34, 42 and 82). In the cat, the species in which the connectional and functional organization of the superficial collicular layers has been studied most thoroughly, these layers receive their direct retinal input from only two (the so-called W- and Y-channels) of the Abbreviations: HVE, high-velocity-excitatory cells; HVS. high-velocity-suppressive; LGNd, dorsal lateral geniculate nucleus; LGNv, ventral lateral geniculate nucleus; LP, lateral posterior nucleus; LyE, low-velocity-excitatory cells; RGCs, retinal ganglion cells; RRZ, retinorecipient zone of the pulvinar; SC, superior colliculus; 5g. suprageniculate nucleus; SGI, stratum griseum intermediale; SGS, stratum griseum superficiale; SO, stratum opticum.
three principal parallel information channels present in the visual system of this species. 11,14a,20,38,52.71,83.87.90 The W-type retinal ganglion cells (RGCs), especially the W-2 subtype, which provide the principal retinal input to the SC,�0-�2�~ may be involved in the detection of local motion in the environment77 but respond poorly or not at all to fast-moving photic stimuli. 18.19.50.55,87.88 Although Y-type RGCs, like the W-2 subgroup of the RGCs, are presumed to be mainly involved in processing information about motion, unlike W-cells they exhibit good responsiveness to photic stimuli moving at high velocities and poor responsiveness to slowly moving photic stimuli.25,29,32,50 There is very little overlap in the laminar distribution of W-type and Y-type retinotectal terminals. 11,28,38,55,57.67 Furthermore, although there is a strong excitatory convergence on single collicular neurons of retinal W-type input and corticotectal input from the primary visual cortices, only a small proportion of collicular neurons which receive W-type retinal input also receives Y-type input via the corticotectal projection.12 It is not clear therefore to what extent, if any, the convergence of retinal and corticotectal inputs onto single collicular neurons represents the conver-gence of different visual information channels. Indeed, despite clear-cut differences in the receptive field properties of W-type and Y-type RGCs (especially in the velocity sensitivity profiles) it has been claimed in numerous studies that
1064 W. J. Waleszczyk et al. the receptive field properties of collicular neurons throughout the retinorecipient layers are rather homogeneous (for review see Ref. 82).
It has been shown recently that contrary to previous asser-tions there is a substantial degree of excitatory convergence of different information channels in different visual cortical areas of mammals with well-developed visual systems such as cats and macaque monkeys (for review see Ref. 16). In the present study, in order to assess the proportion of collicular neurons in which there is convergence (or lack of convergence) of different visual information channels we have examined quantitatively the velocity response profiles of single neurons in the retinorecipient layers of the cat�s SC to photic stimuli. Our results indicate that in almost half of the collicular neurons there is a detectable convergence of different visual information channels.
A preliminary report of this work has been published in abstract form.26
EXPERIMENTAL PROCEDURES Animals, anaesthesia and surgical procedures
All efforts were made to minimize animal suffering and to reduce the number of animals used. Experiments adhered strictly to the guidelines of the National Health and Medical Research Council of Australia and all procedures were approved by the Sydney University Animal Care and Ethics Commitee. Five adult, normally pigmented, female cats weighing from 2.2 to 3.5 kg were provided by the Laboratory Animal Services of the University of Sydney. The initial surgery which included intravenous and tracheal cannulation, bilateral cervical sympathectomy and craniotomy was carried out under 1—1.5% halothane in N20/02 (67%/33%) gaseous mixture. During the recording sessions, the animals were paralysed with gallamine triethiodide at the rate of 7.5 mg/kg/h in a mixture of equal parts of 5% dextrose and sodium lactate (Hartmanns) solutions and artificially ventilated. Anaesthesia was maintained with a gaseous mixture of N20102 (67%/33%) and halothane (0.5%). Expired-CO2 was maintained at 3.7-4.0% by adjusting the rate of the pulmonary pump. Body temperature was maintained at about 37.50C. Heart rate and electroencephalogram were monitored continuously. Each day antibiotic (amoxycillin trihydrate, 75 mg), dexamethasone phosphate (4 mg) and atropine sulphate (0.3 mg) were injected intramuscularly. Furthermore, each day 10 ml of 25% solution of D-mannitol was injected intravenously.
The corneas were protected with zero-power, air-permeable plastic contact lenses. Pupils were dilated and accommodation paralysed with 1% atropine sulphate solution. The nictitating membranes were retracted with 0.128% phenylephrine hydrochloride. Artificial pupils, 3 mm in diameter, were placed in front of the contact lenses. Correct-ing lenses were used, if required, to focus the eyes on a tangent screen located 57 cm in front of eyes. The locations of the optic discs and the areae centrales were plotted daily using a fibre-optic projector. Recording
Craniotomy for recording the activity of single neurons in the superior colliculus was made at Horsley-Clarke coordinates P2-A4 and L0-5. A plastic cylinder was mounted around the opening and glued to the skull with dental acrylic to form a well around the hole. A smaller opening in the dura was made above the recording site. A platinum! iridium glass-coated microelectrode was lowered until its tip was positioned 14mm below the cortical surface. The cylinder was then filled with 4% agar gel and sealed with warm wax to reduce brain pulsations. The microelectrode was advanced into the colliculus with an hydraulic micromanipulator. The action potentials of single neurons were recorded extracellularly and amplified conventionally; triggered standard pulses were then fed to a microcomputer. Photic stimulation and data analysis
The excitatory receptive fields (discharge fields) of recorded units were first plotted with a hand-held light slit projector and remapped with hand-held black bars and spots. The size of the excitatory
discharge field was defined as the area of visual space within which moving bars, lighter or darker than the background elicited an increase in the cell�s firing rate which was detectable by ear.5�23 The positions and dimensions of receptive fields were corrected for tangential distortion. The relative magnitude of the responses to photic stimuli presented separately through either eye (occular dominance class) was evaluated quantitatively by the analysis of the peristimulus time histograms)� Computer-controlled light slits or spots from a slide projector, used for studying, quantitatively, the properties of the recep-tive fields. had a luminance of 15 cd/m2/m2 against a background luminance of 0.9 cd!m2/m2. The peristimulus time histograms were usually constructed by summing the responses to 30 successive stimu-lus sweeps at each test condition. The temporal base of each histogram was divided into 150 bins. The bin width varied depending on stimulus velocity, the amplitude of the sweep and the time the stimulus remained stationary outside the receptive field (see Figs 2, 3, 4, 6,7). The responses were then smoothed using a Gaussian weighted average over five neighbouring bins. The smoothed values were used for peristimulus time histograms and evaluation of the peak discharge rates. Localization of recording sites
At the end of long penetrations microlesions were made with 20 µA current passing through the microelectrode for 20 s (electrode negative). At the end of the recording session (lasting four to six days) the animal was deeply anaesthetized with an intravenous injection of 120 mg of sodium pentobarbitone and perfused transcardially with warm (370C) Hartmann�s solution followed by a 4% solution of para-formaldehyde in 0.1 M phosphate buffer (pH 7.4). The midbrains were stereotaxically blocked and sectioned coronally at 50 ~.m on a freezing microtome. mounted on gelatinized slides and counterstained with Cresyl Violet. The reconstruction of the locations of the cells studied by us was based on the depth recorded from the microdrive and the positions of the microlesions. Statistics
Statistical significance of differences between different groups of cells was assessed using non-parametric tests: the Mann�Whitney U-test or χ2-test.80 Statistical differences were considered significant when P at two-tailed criterion was 0.05 or less.
RESULTS Binocularity and spatial characteristics of collicular neurons
All our electrode penetrations were through the central part of the SC where, in the superficial layers at least, the binocular part of the visual field is represented (Fig. lA; see Refs 9,27 and 49). Consistent with this, as indicated in Fig. 1B, the great majority of cells (50/58; about 86%) in our sample were binocular cells, that is, they could be activated by stimuli presented via either eye. All monocular cells (8/58; about 14%) in our sample could be activated by photic stimuli presented via the contralateral, but not the ipsilateral, eye (class I cells). Furthermore, the majority of binocular cells was either dominated by the contralateral eye (class 2 cells) or responded equally well to photic stimuli presented through either eye (class 3 cells). Thus, the eye dominance distribution in our sample is very similar to those reported in previous studies of the cat�s collicular visual neurons, with receptive fields located within the binocular part of the visual field (see Refs 4, 9, 20, 36, 38, 56, 79 and 85). For all binocular cells the velocity response profiles were virtually the same irrespective of the eye through which the stimuli were presented.
Consistent with numerous previous reports9,20,22,36,56,75,85 all neurons but one in our sample were characterized (at least at some velocities, see further) by the presence of excitatory discharge regions in their receptive fields. Thus, as the narrow slit of light (or a light spot of small diameter) crossed the
Convergence of channels in superior colliculus 1065
Fig. 1. (A) The positions of electrode penetrations superimposed onto the dorsal view of cat�s SC with projections of horizontal and vertical meridians and some of their parallels (modified after Ref. 27). Note that all penetrations were in that part of the colliculus in which the central part of the binocular visual field is represented. Note also that only the most rostral penetrations encroach onto the representation of the ipsilateral visual field (indicated by gray area in the rostral colliculus). HM, zero horizontal meridian; VM, zero vertical meridian. (B) Percentage histogram of eye dominance classes of the present sample of collicular neurons. Class I cells are monocular neurons which respond only to photic stimuli presented via the contralateral eye; class 2 cells are binocular neurons which give stronger excitatory response when stimulated via the contralateral eye; class 3 cells are binocular neurons which are excited equally strongly when stimulated through either eye; class 4 cells are binocular neurons excited more strongly when stimulated via the ipsilateral eye. Note that there were few class 4 neurons and that monocular neurons which could be excited only via the ipsilateral eye (class 5 cells) were not present in our sample. excitatory discharge region there was a clear increase in the cell�s firing rate (Figs 2A, B, C, 3B, D, F. 4B, D, F). This increase in firing rate often consisted of a single peak in the peristimulus time response histogram (Fig. 2A and B, top and bottom histograms, respectively; Figs 3B, D. F, 4B, D, F). For most cells, however, at least at certain stimulus velocities, the peristimulus time response histograms comprised two or more distinct peaks (Figs 2B, C, 4F; see Ref. 22) suggesting that there were several discharge subregions that were partly or completely separated from one another.
Most cells in our sample exhibited a degree of direction selectivity, that is, they responded more vigorously to stimuli moving across their receptive field in one (preferred) direction than to stimuli moving in the opposite direction. However, for a given cell the degree of direction selectivity was usually velocity-dependent (Figs 2A, B, C, 3D, 4D, F; see Refs 22, 56, 85 and 89). Furthermore, at a given velocity some cells exhibited a substantial degree of direction selectivity only when short stimuli, not extending beyond the the excitatory discharge regions, were applied (Fig. 3B, D) but not when the photic stimuli extended well into the parts of the suppressive fields which did not overlap the excitatory discharge region (Fig. 3F). In other cells however the degree of direction selec-tivity was not dependent on the length of the photic stimuli applied (Fig. 4D, F).
Based on their responses to photic stimuli moving at various velocities across their receptive fields we have distin-guished four classes of cells in our sample of collicular neurons. Cells responding only at low to moderate velocities
Low-velocity-excitatory cells. These cells gave their strongest excitatory responses at stimulus velocities not exceeding 400/s. They invariably appeared to have strong suppressive regions in their receptive fields, at least partly overlapping with the excitatory discharge regions and
responded poorly, if at all, to long (12-28û) bars moving across their receptive fields (Fig. 3B bottom two histograms, Fig. 3F; see Refs 4, 22, 56, 82 and 85). Indeed, in all neurons constituting this group (28/58 cells, about 48% of our sample) light bars much shorter than the width of the excitatory discharge region evoked stronger responses than stimuli approximating the width of the excitatory discharge region (Fig. 3B, G; see Refs 22, 36 and 85).
Low-velocity-excitatory (LVE) cells responded well (as indicated by a substantial increase in firing rate) to short-to-medium size (1-6û) bars moving at slow-to-moderate (l-l00û/s) velocities, but poorly, if at all, to fast-moving short or long bars or spots (over l000/s; Figs 2A, D, 3B, C, D, F, G). Furthermore, most of the LVE cells (23/28 about 82%) gave their strongest responses to short bars, or small diameter spots, moving at velocities not exceeding 100/s (Figs 2A, D, 3D, G, 5A). Indeed, the mean preferred velocity for LVE cells was 9.60/s (± 8.80/s). Consistent with the velocity� response profiles based classification of neurons in the mammalian visual cortex developed by Orban68 we have found that LVE collicular cells could be classified as either velocity-low-pass (14/27; 52%) or low-velocity-tuned (13/27; 48% of the sample). In several earlier studies of single neurons in the superior colliculus of the cat (see Refs 22, 78 and 79) cells with very similar velocity�response profiles constituted the majority of cells recorded from the superficial layers.
As indicated in Fig. SB, LVE cells were characterised by: (i) low �spontaneous� activities (mean 1.3 spikes/s; range 0�6 spikes/s; see Ref. 22); and (ii) low-to-moderate peak discharge rates (mean 19.7 spikes/s; range 5�47.2 spikes/s). Furthermore, as indicated in Fig. SC, the excitatory discharge regions of cells responding poorly to fast-moving stimuli tended to be relatively small (mean area 42.9±46.2 deg2 mean diameter 6.0±3.0 deg).
Finally, LVE cells constituted the majority of cells (17/25; 68%) recorded from the SGS and a large minority of cells (11/28; about 39%) recorded from the SO (Fig. 9).
1066 W. J. Waleszczyk et al.
Fig. 2. Patterns of discharges of collicular neurons to photic stimuli moving at different velocities. In A, B and C are shown peristimulus time histograms for three neurons. In each case only the responses to stimuli presented via the dominant eye are presented. In each histogram the stimulus (a light bar) moves forwards and backwards across the receptive field. The angle of the bar and the direction of movement are indicated beneath the bottom histogram. The velocity of movement of the bar is indicated above each histogram. The bar moves only during the time indicated by the filled rectangles beneath the histograms and then remains stationary for 400 ma (at 40û/S and lOOû/s) or 800 ins (at velocities >1000/s), outside the receptive field before moving back in the opposite direction (right half of each histogram). The period of time necessary to complete single forward and backward sweeps plus delays is indicated on the right of each histogram. Each peristimulus histogram was compiled from responses to 30 successive stimulus sweeps. In D are shown the graphs of peak response vs velocity for three LVE cells. These cells (including the cell whose responses are illustrated in A) are low-pass or low-velocity-tuned cells (see Ref. 68). In E are shown the graphs of peak response vs velocity for three cells which respond only at moderate and high velocities but not at low velocities. These cells (including the cell whose responses are illustrated in B) are velocity high-pass (HVE) cells (see Ref. 68). In F are shown the graphs of peak response vs velocity for three cells which respond over a wide range of velocities. These cells (including the cell whose responses are illustrated in C) are velocity broad-band (LVE/HVE) cells (see Ref. 68). In D, E and F the responses have been corrected for �spontaneous� activity, the levels of which are indicated by the horizontal lines through the symbols at the bottom right of each graph. All the graphs in D, E and F are based on the responses to stimuli presented via the dominant eyes.
“Suppressed-by-contrast” cell. One of the cells in our sample had a relatively high spontaneous firing rate (22 spikes/s) which was always reduced, rather than increased, by the introduction into the receptive field of any sort of photic stimulus (stationary flashing light spots or slits, slowly or moderately but not fast-moving light or dark spots or bars). This cell was recorded from the border of the SGS and SO (Fig. 9) and responded only to photic stimuli presented via the contralateral eye. High-velocity-excitatory cells
A small number of the cells which constituted this group (4/58; about 7% of the sample) did not respond to slowly (1-10û/s) moving photic stimuli but exhibited clear-cut excitatory responses at moderate (2O-4Oû/s) and high velocities (l00-
2000û/s; Fig. 2B, E). These cells did not exhibit strong suppressive regions within and/or around the excitatory discharge regions and could be classified as velocity high-pass cells (see Ref. 68). High-velocity-excitatory (HVE) cells were recorded from the lowest part of the lower SGS or from the SO (Fig. 9). Low-velocity-excitatory/high-velocity-excitatory cells
Cells constituting this group (15/58; about 26% of the sample) gave clear-cut excitatory responses over a very wide range of stimulus velocities employed by us 1-2OOOû/s; Fig. 2C, F). Although some of these cells were tuned for
Convergence of channels i n superior colliculus 1067
Fig. 3. Characteristics of LVE cells. A and E show outlines of the excitatory discharge fields for two LVE neurons, units 104.17 and 104.16. respectively. Note that the nearest border of the excitatory discharge field of cell 104.17 (A) is located about 50 from the area centralis (AC, center indicated by 0) while the excitatory discharge field of cell 104.16 (E) encompasses the area centralis (AC, center indicated by®). B shows peristimulus time histograms of responses of unit 104.17 to light bars of different length moving at 40/s across the cell�s receptive field. The peak discharge rates of unit 104.17 to bars of different length moving in the preferred direction at 4û/s across the cell�s receptive field are graphed in C. Note that responses illustrated in B and C are strongest for the shortest bar (1.2û length x 1û width) and progressively weaken as the bar is lengthened until at a length of 120 (10 width) the response disappears. Note also that responses to the shorter bars (1.20 and 30) are clearly direction selective while those to longer bars are not. The responses of cell 104.16 to a short bar (30x1û) moving at progressively increased velocities (from 4û/s to 400û /s) are illustrated in D. The responses of the cell in D to a long bar (220x1û) moving at progressively increased velocities are illustrated in F. The amplitudes of responses of unit 104.16 to narrow bars of different length (indicated) moving at different velocities are graphed in G. Note that at all velocities the responses of unit 104.16 are substantially weaker for the longer bars, and for both the short and long bars the response diminishes with increase in stimulus velocity. In each case only the responses to stimuli presented via the dominant eye are illustrated. For other details see legend of Fig. 2. particular velocities (Figs 2C, F, 4D, F, G) in general these cells could be classified as velocity broad-band cells (see Ref. 68). In a significant proportion of cells constituting the third group, there were also strong suppressive regions within and/or around the excitatory discharge regions (Fig. 4). Nevertheless, all LVE/HVE cells gave clear-cut excitatory responses not only to fast-moving (l00-2000û/s) short-to-medium (1-6û) bars (Fig. 4D, G) but also to fast-moving long (l2-28û) bars (Figs 2C, F, 4F, C).
As indicated in Fig. 9, LVE/HVE cells were recorded mainly (12/15) from the SO and the deepest part of the SGS with only an occasional cell of this kind recorded from the upper part of the lower SGS or the uppermost part of the stratum griseum intermediale (SGI). The laminar distribution of LVE/HVE cells was significantly different from that of LVE cells (P<0.05; χ2-test).
Overall, the majority (12/19) of cells which were excited at high stimulus velocities (combined HVE and LVE/HVE cells) responded optimally at velocities exceeding 100û/s (Fig. 5A). Not surprisingly, the preferred velocities of cells excited at high velocities (mean 450±693û/s) were significantly (P<0.0001; Mann�Whitney U-test) higher than those of LVE cells.
As indicated in Fig. SB, HVE and LVE/HVE cells taken
together were characterized by: (i) moderate-to-high sponta-neous activities (mean 12.8 spikes/s; range 0.4�35 spikes/s; see Ref. 22); and (ii) moderate-to-high peak discharge rates (mean 74.8 spikes/s; range 10.3�165.2 spikes/s). Thus, cells responding to higher stimulus velocities differed significantly in these parameters from LVE cells (in either case P<0.000l; Mann�Whitney U-test). Furthermore, despite the fact that the receptive fields of cells responding at high stimulus velocities were located at the eccentricities not significantly different from those of the receptive fields of the LVE cells (P>0.12; Mann�Whitney U-test) the excitatory discharge regions of HVE and LVE/HVE cells (mean area 185.1±248.7 deg2 mean diameter 11.7±7.5 deg) were significantly (P<0.002; Mann�Whitney U-test) larger than those of LVE cells (Fig. 5C). Low-velocity-excitatory/high-velocity-suppressive cells
Cells constituting this group had rather unusual receptive field properties and as far as we know collicular neurons with such properties have not been described before. In particular: (i) presentation of slowly moving (2-10û/s) short-to-medium (2-6û) or long (12-28û) bars or small (2-4û in diameter) spots
1068 W. J. Waleszczyk et al.
Fig. 4. Characteristics of LVE/HVE cells. A and E show outlines of the excitatory discharge fields for two LVE/HVE neurons, units 104.12 and 104.15 respectively. Note that the nearest border of the excitatory discharge field of cell 104. 12 (A) is located about 40 from the area centralis (AC, center indicated by ®) while the excitatory discharge field of cell 104.15 (E) encompasses the area cent rails (AC, center indicated by ®). B shows peristimulus time histograms of responses of unit 104.12 to light bars of different length moving at 4û/s across the cell�s receptive field. The peak discharge rates of unit 104.12 in the preferred direction to bars of different length moving at 4û/s across the cell�s receptive field are graphed in C. Note that responses illustrated in B and C are strongest for the shortest bar (3û length x10 width) and weaken as the bar is lengthened; however a substantial response remains for the bars of l20x10 and 220x 10. The responses of cell 104.15 to a short bar (30x 1û) moving at progressively increased velocities (from l00/s to l0000/s) are illustrated in D. The responses of the same cell as in D to a long bar (220 x1û) moving at progressively increased velocities are illustrated in F. The peak responses of unit 104.15 to narrow bars of different length (indicated) moving at different velocities are graphed in G. Note that although at most velocities the peak responses of unit 104.15 for long and short bars are quite different, the overall shape of the velocity-response curves as well as the range of best velocities for the long and short bars are very similar. .As in Figs 2 and 3 in each case only the responses to stimuli presented via the dominant eye are illustrated. For other details see legend of Fig. 2. revealed excitatory discharge regions in their receptive fields (Figs 6A, B, 7A, B, C); (ii) most of the cells responded poorly or not at all to photic stimuli moving across their receptive fields at moderate velocities (20-40˚/s; Figs 6A, 7A, B); (iii) both long and short bars moving at high velocities (over 100˚/s) evoked purely suppressive responses which tended to be non-direction selective (Figs 6A, B, C, D, 7B, C): (iv) photic stimuli moving slowly outside the excitatory discharge region did not evoke any responses (Fig. 6C and D upper histograms); (v) suppressive, but not excitatory, responses could be evoked by fast-moving stimuli which were positioned well outside the excitatory discharge regions (Fig. 6C and D, lower histograms); and (vi) while there could be some axis-of-movement preference in the excitatory responses to slowly moving stimuli (Fig. 7C, upper histograms) such preferences were not apparent in the suppressive responses to fast-moving stimuli (Fig. 7C, lower histograms).
As indicated in Figs SB, 6 and 7, LVE/high-velocity-suppressive (HVS) cells were characterized by moderate background activities (mean 11.0 spikes/s; range 7-17 spikes/s) and moderate peak discharge rates (mean 26.8 spikes/s; range 15-43.3 spikes/s).
Consistent with the fact that in all five cats we have recorded from the central part of the superior colliculi where in the superficial layers the central 20-25˚ of the visual field
are represented (Fig. 1A, see Refs 9, 27 and 49) the excitatory discharge regions of a substantial proportion (3/10) of our LVE/HVS cells were centered or encroached on the area centralis (Fig. 8). Furthermore, the excitatory discharge regions of all our LVE/HVS cells were either centered or at least encroached on the vertical and/or horizontal meridians (Fig. 8).
The excitatory discharge regions of LVE/HVS cells were located at eccentricities not significantly different from those of the excitatory discharge regions of the LVE cells (P>0.7; Mann�Whitney U-test). Similarly, the sizes of the excitatory discharge regions of LVE/HVS cells (mean area 48.6±28.6 deg2 mean diameter 6.8±2.3 deg) were not significantly different (P>0.3; Mann�Whitney U-test) from those of LVE cells (Figs 5C, 8). At the same time the suppressive regions had diameters six to eight times those of the excitatory discharge regions and invariably included a portion of the visual hemifield contralateral to the excitatory discharge regions (Figs 6, 7).
Cells constituting the LVE/HVS group (10/58; about 17% of the sample) were recorded in the deep SGS, SO or the upper part of the SGI (Fig. 9) of all five animals studied by us. However, one LVE/HVS cell was recorded above the deepest part of the SGS (Fig. 9).
Convergence of channels in superior colliculus 1069
Fig. 5. Properties of collicular neurons of different classes. (A) Percentage histogram of preferred velocities of LVE, HVE and LVE/HVE neurons. Note that while over 85% of LVE cells responded maximally at velocities not exceeding l00/s, only 20% of LVE/HVE cells responded maximally at such velocities. Note also that all HVE cells and over 53% of LVE/HVE cells responded optimally at velocities exceeding 1000/s. (B) Peak discharge rate to presumably optimal stimulus vs level of �spontaneous� activity. Note that cells with higher spontaneous activity tend to respond more strongly (higher peak discharge rates) to photic stimuli. Note also that LVE neurons exhibit very little or no spontaneous activity and tend to respond weakly (low peak discharge rates) to photic stimuli. (C) Excitatory receptive field areas of collicular neurons vs distance of the centre of the excitatory discharge region from the area centralis (AC). Note lack of clear correlation between the receptive field size and the distance from the AC. In each case the data are based on the responses to stimuli presented via the dominant eyes.
DISCUSSION
Parallel information channels in superior colliculus. Low-velocity-excitatory cells. In view of the clear preference of LVE cells (constituting almost 50% of cells in our sample) for slowly moving stimuli and their lack of responsiveness for fast-moving photic stimuli, it appears that these cells receive their principal visual input from the W-channel and do not receive a significant excitatory input from the Y-channel. There are additional lines of evidence which support this conclusion. First, about 75% of collicular neurons which receive W-type input as indicated by the conduction velocity of their retinal afferents respond poorly or not at all to photic stimuli moving at velocities exceeding l000/s.20,38,79 Second, all but three cells recorded from that part of the SGS where W- but not Y-type RGCs terminate (between 200 and 600 p.m from the collicular surface; Fig. 9), like W-type RGCs, not only responded poorly to fast-moving photic stimuli but had low spontaneous activities, low peak discharge rates and contained strong suppressive regions in their receptive fields. 18.19,77.87,88 Third, the velocity low-pass cells recorded from area 19 (which receive their principal thalamic input from the W-type cells23) are also characterized by low spontaneous activities, low peak discharge rates and in most cases contained strong suppressive regions in their receptive fields (hypercomplex or end-stopped cells, see Refs 23, 41 and 68). Fourth, as mentioned in the introduction, although most, if not
all, neurons in the SGS which receive W-type excitatory input from the retina receive also excitatory input from the visual cortex, this input is not of the Y-type. 12 Fifth, although cells responding poorly to fast-moving photic stimuli were also encountered in the deeper part of the SGS and the SO which receive substantial Y-type retinotectal and corticotectal inputs, the receptive field properties of those cells were indis-tinguishable from those of their counterparts in the upper part of the SGS. Furthermore, the majority of morphologically identified neurons encountered in the deeper part of the SGS receive W-type but not Y-type inputs.67 Indeed, this finding is consistent with earlier findings of Berson12 that not all W-driven cells in the lower SGS and SO receive Y-type input.
In view of the fact that the excitatory discharge regions of collicular neurons with W-type excitatory input are substan-tially larger than those of W-type retinal ganglion cells18,19,76,77,87,88 it appears that there is a substantial excitatory convergence of W-type RGCs on single collicular neurons. Since ablation of the ipsilateral visual cortices does not affect the size of the collicular excitatory discharge fields75,92 it is unlikely that the large size of the excitatory discharge fields of collicular cells reflects the convergence of corticotectal fibers.
However, a word of caution is needed here. Thus, lack of responsiveness to fast-moving photic stimuli does not neces-sarily imply that LVE cells receive only W-type excitatory input. First of all, in view of the fact that area 17 which is dominated by X-input (see Refs 23 and 86) provides strong
1070 W. J. Waleszczyk et al.
Fig. 6. Responses of a LVE/HVS cell. The hatched oval in the centre of the diagram shows the excitatory discharge field (ellipsoid hatched area) which is embedded in the much larger suppressive field (grey), the outer boundary of which has not been accurately determined. Note however that the suppressive field involves not only a large part of the visual field ipsilateral to the excitatory discharge field but also extends into the visual field contralateral to the excitatory discharge region. AC, area centralis; ® indicates the center of the area centralis. (A) The peristimulus time histograms on the left show the responses to a long bar (200 x1.20 moving forward and backward across the entire receptive field. Note the excitation at low velocities (40/s and l00/s) and a clear suppression at high velocities (1000/s or more) with virtually no response at the intermediate (400/s) velocities. (B) A short light bar (about 40, similar to the width of the excitatory discharge field) produces strong excitation at a low velocity (100/s) and clear suppression at a high velocity (2000/s). (C and D) When the long bar is moved outside the excitatory discharge region there is no response at low velocities (l00/s) but again a clear suppression at high velocities (2000/s). As in Figs 2, 3 and 4 all illustrated responses are responses to stimuli presented via the dominant eye. As in Figs 2 and 3 in each case only the responses to stimuli presented via the dominant eye are illustrated. Other conventions as in Figs 2. 3 and 4. non-Y corticotectal input to the superficial collicular layers12,34
it is very likely that LVE cells also receive some excitatory input from the X-channel. Second, lack of responsiveness to fast-moving photic stimuli does not necessarily imply that LVE cells do not receive any excitatory input from the Y-channel. Indeed, 20�25% of cells in the primary visual cortices of the cat (areas 17 and 18) which receive excitatory Y-type input as indicated by the conduction velocity of their retinogeniculate afferents respond poorly to photic stimuli moving at velocities exceeding 1000/s.23 Similarly, about 10% of collicular neurons which receive excitatory Y-type Input via the ipsilateral visual cortex respond poorly to photic stimuli moving at velocities exceeding 1000/s.20,38 Further-more, in 40% of layer V cells recorded from area 17 of the cat, blockade (with lignocaine) of visuotopically corresponding parts of layer V of area 18 revealed responses to high-velocity photic stimuli which were ineffective before the blockade.2
The suppressed-by-contrast cell recorded by us is strongly reminiscent of the W-type �suppressed-by-contrast� or �uniformity detectors� ganglion cells described in the cat�s retina. 19,74.76,88 A few collicular cells with such properties have been reported previously (see Refs 22, 35, 36 and 38).
Taking into account the laminar distribution of collicular cells projecting to the cat�s dorsal and/or ventral lateral geni-culate nuclei (LGNd and LGNv) and/or the retinorecipient
zone of the pulvinar1,33 and the fact that geniculate cells in those regions which receive collicular input as well as cells in the retinorecipient zone of the pulvinar (RRZ) exhibit W-like properties (see for reviews Refs 30 and 86) it is likely that collicular neurons receiving W- but not Y-input project to the LGNd, LGNv and the RRZ.
High-velocity-excitatory cells. HVE cells which constituted about 7% of the sample appeared to receive Y-type but not W-type inputs. Unlike cells receiving only W-input these cells were characterized by: strong excitatory responses to fast-moving photic stimuli; poor, if any, responses to slowly moving photic stimuli; moderate-to-high spontaneous activity; moderate-to-high peak discharge rates; and absence of strong suppressive regions in their receptive fields. All these cells were located in the deep SGS or SO where there is a strong Y-type input coming directly from the retina and/or indirectly via the visual corticesi.13,20,38 Indeed, it has been reported by Hoffmann38 that virtually all collicular neurons which receive Y-type input directly from the retina respond vigorously at stimulus velocities exceeding 2000/s. Again, a word of caution is needed here. Thus, despite the lack of responses to the slowly moving photic stimuli a subthreshold W-input to HVE cells is a distinct possibility.
Convergence of channels in superior colliculus 1071
Fig. 7. A further example of receptive field organization of LVE/HVS cells. (A) Shows graphs of the change in discharge rate vs velocity of bar movement for three cells which show excitation at low velocities and suppression at high velocities (0, no change in cells� ongoing discharge rates; positive values, increase in discharge rates; negative values, decrease in discharge rates). B and C show responses of a cell when stimulated via the contralateral eye (B) or via the ipsilateral eye (C; excitatory discharge field and surround from stimulation of this eye). C. illustrates the fact that the excitatory discharge region is axis-of-movement-selective (low velocities), while the suppressive region is not (high velocities). Further explanations in text. Other conventions as in Figs 2, 3, 4 and 6. Convergence of different information channels
Low-velocity-excitatory/high-velocity-excitatorv cells. Over a third (12/34) of the cells recorded from the lower part of the lower SGS and SO as well as one cell recorded from the SGI and two cells recorded in the upper half of the lower SGS were LVE/HVE cells characterized by good excitatory responses over a wide range of stimulus velocities (Fig. 9). Those cells appear to have received their excitatory visual inputs from both W- and Y-channels. Further support for this conclusion comes from the fact that up to 25% of collicular neurons which receive direct excitatory input from W-type retinal ganglion cells responded not only to photic stimuli moving at slow-to-moderate velocities but also quite clearly to stimuli moving at velocities exceeding 100û/s.38,79 Like other cells receiving excitatory Y-input (see above) cells responding well over a wide range of stimulus velocities were characterized by moderate-to-high spontaneous activities; moderate-to-high peak discharge rates and usually weak suppressive regions in their receptive fields. Furthermore, their excitatory receptive fields tend to be larger than those of the LVE neurons (see above). The laminar distribution of these cells suggests that their Y-type input originates directly from the retina and/or is relayed via the visual cortex. Note that a substantial degree of excitatory convergence of different information channels is apparent in all visual cortical areas of the cat so far studied (see for review Ref. 16).
Taking into account: (i) the laminar distribution of collicu-lar cells projecting to the suprageniculate nucleus (Sg37,47) and/or lateral posterior nucleus of the cat (LP17,48,54); (ii) a clear preference for high-velocity motion of Sg cells8 and many LP neurons,1 it is likely that collicular neurons receiving excitatory Y-input project to the Sg and/or the tectorecipient
part of the LP complex. The receptive field sizes versus velocity response profiles of low-velocity-excitatory cells, high-velocity-excitatory cells and low-velocity-excitatory/high-velocity-excitatory cells
At any given eccentricity the excitatory discharge regions of collicular cells presumably receiving Y-input (HVE and
Fig. 8. Plots of the excitatory receptive fields of LVE/HVS cells. Although only the receptive fields plotted through the dominant eye are illustrated, all 10 cells could be activated through either eye. Note that the excitatory receptive fields of three LVE/HVS cells encroach onto the area central is and that the receptive fields of all cells encroach on the horizontal and/or vertical meridian. Note also that the excitatory receptive fields of four cells encroach onto the ipsilateral visual field. X indicate the positions of the centres of two excitatory receptive fields which were not completely outlined.
1072 W. J. Waleszczyk et al.
Fig. 9. Depth and laminar distribution of the present sample of collicular neurons. The anatomical subdivisions shown at right are: SAl, stratum album intermediale; SGI, stratum griseum intermediale; SGP, stratum griseum profundum; SGS1, stratum griseum superficiale lower; SGSu, stratum griseum superficiale upper; SO, stratum opticum; SZ, stratum zonale. The laminar distributions of the W- and Y-terminals from the retina are shown under �retinotectal input�, while the laminar distribution of the corticotectal terminals (Y and non-Y) originating from a number of the ipsilateral visual cortices is shown under �corticotectal input� to the right (after Refs 10, 11, 14a, 28, 38, 45, 55, 57, 67 and 84). LVE/HVE cells) are substantially larger than these of cells presumably receiving only W-input (LVE cells). Similar positive eorelations between the size of the excitatory recep-tive fields of collicular neurons and their responsiveness to the fast moving photic stimuli were observed in earlier reports.22,70
Thus, one might argue that in order to excite the cell the stimulus has to spend some minimal time in the cell�s receptive field. To a certain extent it might be so. However, the differences in the sizes of the excitatory receptive fields of the HVE and LVE/HVE cells vs those of LVE cells are probably not sufficient to account for the good responses of HVE and LVE/HVE cells to stimuli moving at velocities exceeding 100û/s. There are several other lines of argument, indicating that the positive correlation between the size of the excitatory discharge field and the responsiveness to fast-moving stimuli is not generally applicable to the visual system at least in the cat. Thus, although at a given eccentricity receptive fields of W-type retinal ganglion cells are at least as large as those of the Y-type RGCs (see for review Ref. 86), the W-type ganglion cells, unlike Y-type cells, respond poorly or not at all to fast-moving photic stimuli.18,19,50,55,87,88 Second, although the excitatory receptive fields of neurons in cat�s visual cortical area 2la are at any excentricity substantially larger than those of area 18 neurons, hardly any area 2la neurons respond to fast-moving (over l000/s) photic stimuli24
while a very high proportion of area 18 neurons does so.23,24,68
Low-velocity-excitatory/high-velocity-suppressive cells. In a distinct group of collicular cells (about 17% of the sample)
there appears to be a special type of convergence of the W- and Y-channels. Activation of the W-channel (by slowly moving photic stimuli) resulted in excitation (increase in firing rate) while activation of the Y-channel (by fast-moving photic stimuli) resulted in inhibition (reduction in firing rate). Poor responsiveness of these cells to stimuli moving at moderate velocities suggests that the two inputs act in an antagonistic manner. Furthermore, the suppressive fields revealed by the activation of the Y-input were much larger than those revealed by the activation of the W-input and in addition to the large visual field ipsilateral to the excitatory discharge field the suppressive field usually included a region of the visual field contralateral to the excitatory discharge field of the cell.
In view of the fact that the region of LVE/HVS receptive field which appears to receive the Y-input is very large and invariably extends into the ipsilateral visual hemifield, it is likely that morphological counterparts of our LVE/HVS cells located in the lower SO and upper SGI (6/10 cells in our sample of LVE/HVS cells) are �medium-sized trapezoid radiating�, or �T type� efferent cells present in these layers of cat�s superior colliculus.7,60 Indeed, the dendritic trees of the T cells located in the lower SO or upper SGI are: (i) very large in the dorsoventral axis (600�1 600 p.m) with dendrites extending into the SGS7,60; and (ii) extend to over a 1000 p.m mediolaterally and rostrocaudally.7,60 The W-type input to these cells is likely to be relayed via the dendrites in the SGS, while the Y-input might be relayed not only via their dendrites in the SOS and SO but also more directly
Convergence of channels in superior colliculus 1073 via the Y-type retinal and cortical terminals in the 561 (see Fig. 9). Furthermore, Y-input to LVEIHVS cells is presum-ably relayed via inhibitory interneurons and indeed, GAB Aer-gic neurons are plentiful in both the superficial and deep layers of cat�s SC (see for review Ref. 59).
Another possibility, especially in relation to the LVE/HVS cells located in the upper SO and the 565 is that the suppres-sive Y-input to these cells is relayed via a pretectocollicular projection. Indeed, the pretectal complex, especially the nucleus of the optic tract, is characterized by a large propor-tion of non-direction-selective cells with large receptive fields and responding well (or exclusively) to fast-moving photic stimuli (see Refs 39 and 79). Furthermore, the pretectal complex projects to both the superficial and, to a lesser extent, the intermediate collicular layers (see Ref. 42). Many of the pretectal neurons which project to the superior colliculi are GABAergic.3 Finally, cells with receptive field properties very similar to these LVE/HVS collicular cells are present in the pretectal nucleus of optic tract of cats,40 macaque monkeys39
and diprotodont marsupials-tammar wallabies.43,44 Cells with excitatory input activated by the W-channel and
inhibitory input activated by the Y-channel might be more common than our data would suggest. In particular, the inhi-bitory input would be revealed by our procedures only when the cell had a substantial spontaneous activity in the absence of specific photic stimuli. However, most of the cells respond-ing preferentially at low stimulus velocities but not increasing their firing rates when fast-moving stimuli were presented had very low, if any, background activity. Further support for the idea that collicular cells with very extensive suppressive fields comes from the study of Rizzolatti and his colleagues72,73 in which they reported that in the great majority of neurons recorded from the retinorecipient layers of the superior colli-culus of the cat the magnitude of responses to a photic stimu-lus moving across the cell�s excitatory discharge field was significantly (but only transiently for 2�3 s) decreased by a second stimulus moving at 30—1000/s in a region 300 to 1200 from the border of the excitatory discharge region.72,73 When the second stimulus was introduced into the field ipsilateral to the cell�s excitatory receptive field the decrease of the response was stronger and apparent in 95% of cells, while introduction of the second stimulus into the field contra-lateral to the excitatory receptive field produced a statistically significant decrease of the response in only 77% of the cells tested.
In many neurons in the intermediate and deep layers of superior colliculi of cats and macaque monkeys there is a substantial cross-modality excitation (see for recent review Ref. 91) and, to a lesser degree, cross-modality suppression (see for recent review Ref. 46). It is possible that LVE/HVS neurons recorded by us in the intermediate layers also receive a convergent auditory and/or somatosensory inputs. If so, the nature of interactions between the visual and other sensory modalities (excitatory vs suppressive) might depend on the velocity of the photic stimuli applied. Functional role of low-velocity-excitatory/high-velocity-suppressive cells
Cells with excitatory input activated by the W-channel and inhibitory input activated by the Y-channel might play a unique functional role. One possibility is that these cells innervate, with presumable involvement of some
interneurons, two types of premotor (usually tectoreticular and tecto-reticulo-spinal) cells described in the intermediate and deep collicular layers of alert cats and macaque monkeys: (i) cells which discharge tonically when the visual axis is approximately stationary reduce their discharge before the onset of the saccade which breaks the fixation and pause for most of the saccade; and (ii) �saccade� neurons which discharge before and during saccades to a specific region in the visual field and pause during visual fixation.31,53,61-66,69 The premotor cells which discharge tonically when the visual axis is stationary have been subdivided into two subclasses: (i) �fixation� neurons which have visual receptive fields that include the area centralis and discharge maximally when the animal fixates attentively the target of interest; and (ii) �orien-tation� neurons whose receptive fields do not include the area centralis.63 We would argue that slowly moving photic stimuli which evoke the excitatory responses of the LVE/HVS cells would in the case of the LVE/HVS cells with excitatory receptive fields involving the area centralis evoke excitatory responses in �fixation� neurons. In the case of the LVE/HVS cells with excitatory receptive fields not encroaching onto the area centralis slowly moving stimuli would evoke excitatory responses in �orientation� neurons. At the same time the LVE/HVS cells excited by slowly moving stimuli would inhi-bit, presumably via interneurons, the �saccade� neurons. By contrast, fast-moving photic stimuli by suppressing the ongoing activity of the LVE/HVS cells would in turn reduce the ongoing activity of inhibitory interneurons innervating saccade neurons. There is emerging evidence that at least in macaque and ferret the reciprocal inhibition between fixation and saccade-related neurons is mediated by interneurons located in the intermediate and deep collicular layers58,64 and indeed some of the LVE/HVS neurons in our sample were located in the SHI. However, most of the LVE/HVS cells in our sample were located in the superficial layers. contrary to some previous assertions, small groups of cells in the superficial collicular layers project to the intermediate and deep layers (see Ref. 6). Furthermore, a recent study of Lee and co-workers51 indicates that, at least in the tree shrew, cells in the superficial collicular layers directly excite cells in the underlying intermediate layers. It has been postulated that some of the so-called retinal-slip neurons in the nucleus of optic tract, which have receptive field properties very similar to those of LVE/HVS collicular neurons (excitatory responses to slowly moving photic stimuli and suppressive nondirectional responses to fast-moving stimuli) and are believed to drive the ocular following components of optokinetic nystagmus, function as blockers of �ocular following in response to the visual disturbances caused by saccades�.43
CONCLUSIONS
In the majority (57%) of collicular neurons in our sample of cells recorded from the retinorecipient layers there was no indication of convergence of W- and Y-channels. However, in about 43% of the cells there was clear evidence of conver-gence of W- and Y-channels. Furthermore, over 15% of colli-cular neurons (about 40% of these which receive convergent input from both W- and Y-channels) appear to receive exci-tatory input from the W-channel and inhibitory input from the Y-channel (LVE/HVS cells). We postulate that the LVE/HVS cells (which constitute a new class of collicular neurons) play
1074 W. J. Waleszczyk et al. an important role in activating �fixation/orientation� and�saccade� premotor neurons recorded by others in the intermediate and deep collicular layers. Acknowledgements—
The work was supported by a grant from the National Health and Medical Research Council of Australia. We would like to thank anonymous reviewers for their helpful comments.
REFERENCES 1. Abramson B. P. and Chalupa L. M. (1988) Multiple pathways from the superior colliculus to the extrageniculate visual thalamus of the cat. J. Comp.
Neurol. 271, 397�4 18. 2. Alonso J. M., Cudeiro J., Perez R., Gonzalez F. and Acuńa C. (1993) Influence of layer V of area 18 of the cat visual cortex on responses of cells in layer
V of area 17 to stimuli of high velocity. Exp.Brain Res. 93, 363�366. 3. Appell P. P. and Behan M. (1990) Sources of subcortical GABAergic projections to the superior colliculus in the cat.]. comp. Neurol. 302, 143�158. 4. Bacon B. A., Villemagne J., Bergeron A., Lepore F. and Guillemot J.-P. (1998) Spatial disparity coding in the superior colliculus of the cat. Exp. Brain
Res. 119, 333�344. 5. Barlow H., Blakemore C. and Pettigrew J. D. (1967) The neural mechanism of binocular depth discrimination.J. Physiol., Lond. 193, 327�342. 6. Behan M. and Appell P. P. (1992) Intrinsic circuitry in the cat superior colliculus: projections from the superficial layers.J. Comp. Neurol. 315, 230�243.
7. Behan M., Appell P. P. and Graper M. J. (1988) Ultrastruetural study of large efferent neurons in the superior colliculus of the cat after retrograde labeling with horseradish peroxidase J. Comp. Neurol. 270, 17 1�184.
8. Benedek G., Perény J., Kovacs G., Fischer-Szátmari L. and Katoh Y. Y. (1997) Visual, somatosensory, auditory and nociceptive modality properties in the feline suprageniculate nucleus. Neuroscience 78, 179�189.
9. Berman N. and Cynader M. (1972) Comparison of receptive field organization of the superior colliculus in Siamese and normal cats.]. Physiol., Lond.224, 363�389.
10. Berson D. M. (1987) Retinal W-cell input to the upper superficial gray layer of the cat�s superior colliculus: a conduction velocity analysis. J. Neurophysiol. 58, 1035�1051.
11. Berson D. M. (1988) Retinal and cortical inputs to cat superior colliculus, convergence and laminar specificity. Prog. Brain Res. 75, 17�26. 12. Berson D. M. (1988) Convergence of retinal W-cell and corticotectal input to cells of the cat superior colliculus. J. Neurophysiol. 60, 1861�1873. 13. Berson D. M. and Mcllwain J. (1982) Retinal Y-cell activation of deep-layer cells in superior colliculus of the cat. J. Neurophysiol. 47, 700�7 14. 14. Berson D. M. and Mcllwain J. (1983) Visual cortical inputs to deep layers of cafs superior colliculus. J. Neurophysiol. 50, 1143�1155. 14a. Berson D. M., Isayama T. and Pu M. (1999) The eta ganglion cell type of cat retina.]. Comp Neurol. 408, 204�219. 15. Berson D. M., Pu M. and Famiglietti E. V. (1998) The zeta cell: a new ganglion cell type in cat retina. J. Comp. Neurol. 399, 269�288. 16. Burke W., Dreher B. and Wang C. (1998) Selective block of conduction in Y optic nerve fibres: significance for the concept of parallel processing.
Eur.J.Neurosci. 10, 8�19. 17. Caldwell R. B. and Mize R. R. (1981) Superioreolliculus neurons which project to the cat lateral posterior nucleus have varying morphologies. J. Comp.
Neurol. 203, 53�66. 18. Cleland B. G. and Levick W. R. (1974) Brisk and sluggish concentrically organized ganglion cells in the cat�s retina. J. Physiol., Lond. 240, 421�456. 19. Cleland B. G. and Levick W. R. (1974) Properties of rarely encountered types of ganglion cells in the cat�s retina and an overall classification.]. Physiol.,
Lond. 240, 457�492. 20. Crahtree J. W., Spear P. D., McCall M. A., Jones K. R. and Komguth S. E. (1986) Contributions of Y - and W-cell pathways to response properties of cat
superior colliculus neurons: Comparison of antibody- and deprivation-induced altemation.. J. Neurophvsiol. 56, 1157�1173. 21. Distler C. and Hoffmann K.-P. (1991) Depth perception and cortical physiology in normal and innate microstrabismic cats. Vis. Neurosci. 6,25�41. 22. Dreher B. and Hoffmann K-P. (1973) Properties of excitatory and inhibitory regions in the receptive fields of single units in the cat�s superior colliculus.
Exp. Brain Res. 16, 333�353. 23. Dreher B., Leventhal A. G. and Hale P. T. (1980) Geniculate input to cat visual cortex: a comparison of area 19 with areas 17 and 18.]. Neurophy.ciol. 44,
804�826. 24. Dreher B., Michalski A., Ho R. H. T., Lee C. W. F. and Burke B. (1993) Processing of form and motion in area 21 a of cat visual cortex. Vis. Neurosci.
10, 93�115. 25. Dreher B. and Sanderson K. J. (1973) Receptive field analysis: responses to moving visual contours by single lateral geniculate neurones in the cat.
J.Physiol., Lond. 234, 95�118. 26. Dreher B.. Waleszczyk W. J., Wang C., Burke W. and Davey D. F. (1998) Velocity response profiles of collicular neurons: Parallel and convergent visual
channels. Soc. Neurosci. Abstr. 24, 1633. 27. Feldon S., Feldon P. and Kruger L. (1970) Topography of the retinal projection upon the superior colliculus of the cat. Vision Res. 10, 135�143. 28. Freeman B. and Singer W. (1983) Direct and indirect visual inputs to superficial layers of cat superior colliculus: a current source � density analysis of
electrically evoked potentials.J. Neurophysiol. 49, 1075�109 1. 29. Frishman L. J., Schweitzer-Tong D. E. and Goldstein E. B. (1983) Velocity tuning of celL in the dorsal lateral geniculate nucleus and retina of the cat. J.
Neurophysiol. 50, 1393�1414. 30. Garey L. I., Dreher B. and Robinson S. R. (1991) The organization of the visual thalamus. In The Neuroanatomy of the Visual Pathways and Their
Development, (eds Dreher B. and Robinson S. R.), pp. 176�234. In Vision and Visual Dysfunction, Vol. 3 (series ed. Cronly-Dillon J.) Macmillan, London.
31. Guitton D. and Munoz D. P. (1991) Control of orienting gaze shifts by the tectoreticulospinal system in the head-free cat. I. Identification, localization, and effects of behavior on sensory responses.J. Neurophvsiol. 66, 1605�1623.
32. Hamasaki D. I. and Cohen H. I. (1977) Differential response of X and Y retinal ganglion cells to moving stimuli results from a difference in the surround mechanism. Brain Res. 122, 157�161.
33. Harrell J. V., Caldwell R. B. and Mize R. R. (1982) The superior colliculus neurons which project to the dorsal and ventral lateral geniculate nuclei in the cat. Exp. Brain Res. 46, 234�242.
34. Harting J. K., Updyke B. V. and Van Lieshout D. P. (1992) Corticotectal projections in the cat: anterograde transport studies of twenty-five cortical areas. J. Comp. Neurol. 324, 379�4 14.
35. Harutiunian-Kozak B., Kozak W. and Dec K. (1970) Visually evoked potentials and single unit activity in the superior colliculus of the cat. Acta Neurobiol. Exp. 30, 211�232.
36. Hayashi Y., NagataT., Tamaki Y. and Iwama K. (1973) Binocular interaction in the superior colliculus of chronic cats. Expl Brain Res. 18, 531�547. 37. Hicks T. P., Stark C. A. and Fletcher W. A. (1986) Origins of afferents to visual supra-geniculate nucleus of the cat. J. Comp. Neural. 246, 544�554. 38. Hoffmann K-P. (1973) Conduction velocity in pathways from retina to superior colliculus in the cat: a correlation with receptive-field properties. J.
Neurophysiol. 36, 409�424. 39. Hoffmann K-P. and Distler C. (1989) Quantitative analysis of visual receptive fields of neurons in nucleus of the optic tract and dorsal terminal nucleus of
the accessory optic tract in macaque monkey. J. Neurophvsiol. 62, 4 16�428. 40. Hoffmann K-P. and Schoppmann A. (1981) A quantitative analysis of the direction specific response of neurons in the cat�s nucleus of the optic tract. Exp.
1075 Convergence of channels in superior colliculus Brain Res. 42, 146�157.
41. Hubel D. H. and Wiesel T. N. (1965) Receptive fields and functional architecture in two nonstriate visual areas (18 and 19) of the cat. .1. Neurophysiol. 28, 229�289.
42. Huerta M. F. and Harting J. K. (1984) The mammalian superior colliculus: studies of its morphology and connections. In Comparative Neurology of the Optic Tectum, (ed. Vanegas H.), pp. 687�773. Plenum, London.
43. Ibbotson M. R. and Mark R. F. (1994) Wide-field nondirectional visual units in the pretectum: do they suppress ocular following of saccade-induced visual stimulation. J. Neurophysiol. 72, 1448�1450.
44. Ibbotson M. R., Mark R. F. and Maddess T. I. (1994) Spatiotemporal response properties of direction-selective neurons in the nucleus of the optic tract and dorsal terminal nucleus of the wallaby Macro~ms eugenii. .1. Neurophysiol. 72, 2927�2943.
45. Itoh K., Conley M. and Diamond 1. T. (1981) Different distributions of large and small ganglion cells in the cat after HRP injections of single layers of the lateral geniculate body and the superior colliculus. Brain Res. 207, 147�152.
46. Kadunce D. C., Vaughan J. W., Wallace M. T., Benedek G. and Stein B. E. (1997) Mechanisms of within- and cross-modality suppression in the superior colliculus. J. Neurophysiol. 78, 2834�2847.
47. Katoh Y. Y. and Benedek G. (1995) Organization of the colliculo-suprageniculate pathway in the cat: a wheat germ agglutinin�horseradish peroxidase study. J. Comp. Neurol. 352, 38 1�397.
48. Kawamura S., Fukushima N., Hattori S. and Kudo M. (1980) Laminar segregation of cells of origin of ascending projections from the superficial layers of the superior colliculus in the cat. Brain Res. 184, 486�490.
49. Lane R. H.. Kaas J. H. and Allman J. M. (1974) Visuotopic organization of the superior colliculus in normal and Siamese cats. Brain Res. 70, 413�430. 50. Lee B. B. and Willshaw D. J. (1978) Responses of the various types of cat retinal ganglion cells to moving contours. Vision Res. 18, 757�765. 51. Lee P. H., Helms M. C., Augustine G. J. and Hall \V. C. (1997) Role of intrinsic synaptic circuitry in collicular sensorimotor integration. Proc. natn.
Acad. Sci. U.S.A. 94, 13,299�13,304. 52. Leventhal A. 0., Rodieck R. W. and Dreher B. (1985) Central projections of cat retinal ganglion cells.J. Comp. Neural. 237, 216�226. 53. Mays L. E. and Sparks D. L. (1988) Dissociation of visual and saccade-related responses in superior colliculus neurons. 1. Neurophysiol. 43,
207�232. 54. Mcllwain J. T. (1978) Properties of cells projecting rostrally from the superficial layers of the cat�s superior colliculus. Brain Res. 143, 445�457. 55. Mcllwain J. T. (1978) Cat superior colliculus: extraeellular potentials related to W-cell synaptic actions.]. Neurophysiol. 41. 1343�1358. 56. Mcllwain J. T. and Buser P. (1968) Receptive fields of single cells in the cat�s superior colliculus. Expl Brain Res. 5, 314�325. 57. Mcllwain J. T. and Lufkin R. B. (1976) Distribution of direct Y-cell inputs to the cat�s superior colliculus: are there spatial gradients? Brain Res. 103,
133�138. 58. Meredith M. A. and Ramoa A. 5. (1998) Intrinsic circuitry of the superior colliculus: pharmacological identification of horizontally oriented inhibitory
intemeurons.]. Neurophysiol. 79, 1597�1602. 59. Mize R. R. (1992) The organization of GABAergic neurons in the mammalian superior colliculus. Prog. Brain Res. 90, 2 19�248. 60. Moschovakis A. K. and Karabelas A. B. (1985) Observations on the somatodendritic morphology and axonal trajectory of the intracellularly HRPlabelled
efferent neurons located in the deeper layers of the superior colliculus of the cat. ]. comp. Neurol. 239, 276�308. 61. Munoz D. P. and Guitton D. (1985) Tectospinal neurons in the cat have discharges coding gaze position error. Brain Res. 341, 184�188. 62. Munoz D. P. and Guitton D. (1985) Presaccadic burst discharges of tecto-reticulo-spinal neurons in the alert head-free cat. Brain Res. 398, 185�190. 63. Munoz D. P. and Guitton D. (1991) Control of orienting gaze shifts by the tectoreticulospinal system in the head-free cat. II. Sustained discharges during
motor preparation and fixation. J. Neurophysiol. 66, 1624� 1641. 64. Munoz D. P. and Istvan P. J. (1998) Lateral inhibitory interactions in the intermediate layers of the monkey superior colliculus. J. Neurophysiol. 79,
1193�1209. 65. Munoz D. P. and Wurtz R. H. (1993) Fixation cells in monkey superior colliculus. Characteristics of cell discharge. J. Neurophysiol. 70, 559�575. 66. Munoz D. P. and Wurtz R. H. (1995) Saccade related activity in monkey superior colliculus. Characteristics of burst and buildup cells. J. Neurophvsiol.
73, 2313�2333. 67. Ogawa T. and Takahashi Y. (1981) Retinotectal eonneetivities within the superficial layers of the cat�s superior colliculus. Brain Res. 217, I�Il. 68. Orban G. A. (1984) Neuronal Operations in the Visual Cortex. Springer�Verlag. Berlin. 69. Peck C. K. (1989) Visual responses of neurones in cat superior colliculus in relation to fixation of targets.J. Physiol., Lond. 414, 301�315. 70. Pinter R. B. and Harris L. R. (1981) Temporal and spatial response characteristics of the cat superior colliculus. Brain Res. 207, 73�94. 71. Rhoades R. W., Mooney R. D. and Fish S. E. (1991) Retinotopic and visuotopic representations in the mammalian superior colliculus. In The
Neuroanatomv of the Visual Pathways and their Development. (eds. Dreher B. and Robinson S. R.), pp. 150�175. In Vision and Visual Dysfunction, Vol. 3 (series ed. Cronly-Dillon J.) Macmillan, London.
72. Rizzolatti G., Camarda R., Grupp L. A. and Pisa M. (1973) Inhibition of visual responses of single units in the cat superior colliculus by the introduction of a second visual stimulus. Brain Res. 61, 390�394.
73. Rizzolatti G., Camarda R., Grupp L.A. and Pisa M. (1974) Inhibitiory effect of remote visual stimuli on visual responses of cat superior colliculus: spatial and temporal factors. J. Neurophysiol. 37, 1262�1275.
74. Rodieck R. W. (1967) Receptive fields in the cat retina: a new type. Science 157, 90�92. 75. RosenquistA. C. and PalmerL. A. (1971) Visual receptive field properties of cells of the superiorcolliculus aftercortical lesions in the cat. Exp. Neural.
33, 629�652. 76. Rowe M. H. and Cox J. F. (1993) Spatial receptive-field structure of cat retinal W cells. Vis. Neurosci. 10, 765�779. 77. Rowe M. H. and Palmer L. A. (1995) Spatio-temporal receptive-field structure of phasic W cells in the cat retina. Vis. Neurosci. 12, 117�139. 78. Sarna M. and Dec K. (1984) The velocity�response curves of the cat�s superior colliculus neurons. Acta Neurobiol. Exp. 44, 89�103. 79. Schoppmann A. and Hoffmann K.-P. (1979) A comparison of visual responses in two pretectal nuclei and in the superior colliculus of the cat. Exp. Brain
Res. 35, 495�510. 80. Siegel S. (1956) Nonparametric Statistics for the Behavioral Sciences. McGraw�Hill, New York. 81. Sparks D. L. (1988) Neural cartography: sensory and motor maps in the superior colliculus. Brain Behav. Evol. 31, 49�56. 82. Stein B. B. and Meredith M. A. (1991) Functional organization of the superior colliculus. In The Neural Basis of Visual Function, (ed. Leventhal A. 6.),
pp. 85�110. In Vision and Visual Dysfunction, Vol. 4 (series ed. Cronly-Dillon J.), Macmillan, London. 83. Stein J. J. and Berson D. M. (1995) On the distribution of gamma cells in the cat. Vis. Neurosci. 12, 687�700. 84. Sterling P. (1971) Receptive fields and synaptic organization of the superficial gray layer of the cat superior colliculus. Vision Res. 3, (Suppl.) 309�328. 85. Sterling P. and Wickelgren B. G. (1969) Visual receptive fields in the superior colliculus of the cat.J. Neurophysiol. 32, 1�15. 86. Stone J., Dreher B. and Leventhal A. 0. (1979) Hierarchical and parallel mechanisms in the organization of visual cortex. Brain Res. Rev. 1,
345�394. 87. Stone i. and Fukuda Y. (1974) Properties of cat retinal ganglion cells: a comparison of W-cells with X- and Y- cells. J. Neurophysiol. 37, 722�748. 88. Stone J. and Hoffmann K.-P. (1972) Very slow-conducting ganglion cells in the cat�s retina: a major, new functional type? Brain Res. 43, 610�616. 89. Straschill M. and Taghavy A. (1967) Neuronale reactionen im tectum opticum der katze auf bewegte und stationäre lichtreize. Exp. Brain Res. 3,
353�367. 90. Tamamaki N., Uhlrich D. J. and Sherman S. M. (I 995j Morphology of physiologically identified retinal X and Y axons in the cat�s thalamus and midbrain
as revealed by intraaxonal injection of biocytin. J. Comp. Neural. 354, 583�607. 91. Wallace M. T. and Stein B. E. (1996) Sensory organization of the superior colliculus in cat and monkey. Prog. Brain Res. 112, 301�323.
1076 W. J. Waleszczyk et al.
92. Wickelgren B. G. and Sterling P. (1969) Influence of visual cortex on receptive fields in the superior colliculus of the cat. J. Neurophysiol. 32, 16-23. 93. Wurtz P. H. and Albano J. E. (1980) Visual-motor function of the primate superior colliculus. A. Rev. Neurosci. 3, 189�226.
