Distribution across cortical areas of neurons projecting to the superior colliculus in new world monkeys

Authors


Abstract

Surprisingly little is known about the proportions of projections of different areas and regions of neocortex to the superior colliculus in primates. To obtain an overview of such projection patterns, we placed a total of 10 injections of retrograde tracers in the superior colliculus of three New World monkeys (Callithrix, Callicebus, and Aotus). Because cortex was flattened and cut parallel to the surface, labeled corticotectal neurons could be accurately located relative to architectonic boundaries and surface features. While there was variability across cases and injection sites, the summed results clearly support several conclusions. One, three well-defined visual areas, V1 (18%), V2 (14%), and MT (11%), contributed nearly half of the total of labeled cells. Two, several other visual areas (V3, DL, DM, and FST) that are early in the processing hierarchy provided another fifth of the total. Three, inferior temporal visual areas of the ventral stream provided only minor projections. Four, visuomotor fields (FEF, FV, cortex in the region of SEF, and posterior parietal cortex) contained less than 10% of the labeled neurons. Five, few labeled neurons were in auditory or somatosensory areas. The results indicate that cortical inputs to the superior colliculus originate predominantly from early visual areas rather than from multimodal or visuomotor areas. © 2005 Wiley-Liss, Inc.

The mammalian superior colliculus is a midbrain structure importantly involved in eye and head movements that direct gaze toward objects of interest (McPeek and Keller, 2004). This function depends on multiple sources of information. Inputs include direct retinal projections, ascending auditory and somatosensory projections, and contributions from a number of cortical areas (Huerta and Harting, 1984; Kaas and Huerta, 1988; Harting, 2004). As the number and types of subdivisions of cortex vary across mammalian taxa (Merzenich and Kaas, 1980), the sources of cortical projections to the superior colliculus also vary. In addition, as cortical areas have different functional roles, their relevance to superior colliculus functions should vary. Thus, the proportional contributions of different cortical areas must vary within a species.

While there is considerable evidence that a number of cortical areas across a range of studied mammals project to the superior colliculus, there is yet little understanding of the relative strengths of these projections. One reason for this is that most studies of cortical projections to the superior colliculus have been based on placing injections of tracers into subdivisions of cortex. The projections of cortical areas are generally studied one by one. In this way, projections can be demonstrated, but it is very difficult to judge the relative contributions of different cortical areas. While injections of retrograde tracers into the superior colliculus have been used to provide a more global overview of the projection pattern to the superior colliculus, this approach has shortcomings as well. As cortical projections terminate at different depths in the superior colliculus, with caudal visual areas terminating more superficially and rostral visuomotor areas more deeply (Graham et al., 1979; Harting et al., 1992; Lock et al., 2003), injections at different depths in the superior colliculus favor some cortical areas over others in labeling corticotectal neurons. Another complication is that the superior colliculus contains a visuomotor map, and injections of various sizes and locations in the map impact differently in cortical areas of precise or crude visuotopic organization. In both types of studies, cortical areas can be difficult to delimit so that labeled corticotectal cells often can be assigned to specific cortical areas with only limited levels of confidence.

While it is important to recognize these difficulties, quantitative comparisons of projection strengths across cortical areas and across species would seem to depend on studies using retrograde labeling of corticotectal neurons. Using this approach, labeled neurons can be assigned to cortical areas or regions, counted, and the numbers compared across areas to provide a quantitative measure of relative connection strengths. Surprisingly, this has not been done, although there is good evidence that the densities of labeled neurons do vary considerably across the cortical sheet after such injections. Most notably, in an early study in macaque monkeys, Fries (1984) injected horseradish peroxidase (HRP) into the superior colliculus and plotted the locations of labeled neurons throughout the labeled hemisphere. The results across cases clearly indicated that large regions of neocortex contributed little to the superior colliculus, and that much of the input came from caudal visual areas. In addition, deeper injections labeled more cells in posterior parietal cortex and frontal cortex. Relative cell densities were determined for some regions of cortex and found to be similar, but relative projection strengths for cortical areas were not determined. As the study was completed before current understandings of visual cortex organization emerged, visual cortex was only divided into areas 17, 18, and 19. More recently, this type of study was repeated in macaques, and distributions of labeled cells were displayed in unfolded surface views of cerebral cortex, with estimates of the locations and boundaries of currently proposed visual areas (Lock et al., 2003). In this recent study, single injections in each colliculus of three monkeys labeled large numbers of neurons in visual areas V1, V2, V3, and MT. Several other visual areas contained few or no labeled neurons (MST, VIP, 7a, MIP, TE). However, proportions of neurons in various visual areas were not determined. In addition, the uncertainties that exist about the boundaries and even the validity of a number of the proposed extrastriate visual areas (Kaas, 1997) remain a problem.

The present study was directed toward providing a quantitative overview of the areal distribution of corticotectal projections in New World monkeys. While injections in several visual or visuomotor areas have demonstrated cortical projections to the superior colliculus in New World monkeys (Graham et al., 1979; Tigges and Tigges, 1981; Huerta et al., 1986), nothing is known about the total cortical projection pattern. As three species of small New World monkeys have been used in recent studies of visual cortex organization in our laboratory (Lyon and Kaas, 2002), and cortical organization has been studied extensively in two of these species (owl monkeys and marmosets), we included a marmoset, owl monkey, and titi monkey in the present study. Rather than including 2–4 individuals of each species in the study, we injected 2–4 distinguishable tracers into the superior colliculus of each individual. While results varied across injections, total results from 10 injections provided a comprehensive view of the projection pattern. Although this approach has limitations, it addressed our goal of providing an overview for these monkeys.

As an important component of the present study, cortex was separated from the rest of the brain, manually flattened, and cut parallel to the surface. Some of the sections were processed for cytochrome oxidase or myelin, as these markers allowed several areas of cortex to be accurately outlined. The boundaries of other areas could be closely estimated by reference to the histologically defined areas and other landmarks. In this way, we were able to locate labeled neurons accurately in flattened views of cortex, assign many of them to specific areas of cortex with certainty, and attribute others to proposed areas and functional regions with some confidence.

MATERIALS AND METHODS

In this study, injections of two to four different tracers were made into the superior colliculus of a single adult of three different species of New World monkeys (a marmoset, Callithrix jacchus; owl monkey, Aotus trivirgatus; and titi monkey, Callicebus moloch). Injections were placed under aseptic conditions into the medial aspect of the right superior colliculus (upper visual field representation) under visual guidance after the left cerebral hemisphere was retracted and partly aspirated in the marmoset and owl monkey. Injections were placed into the lateral aspect of the right superior colliculus (lower visual field representation) under visual guidance after retracting the caudal pole of the hemisphere in the titi monkey. Injections of fluorescent tracers (Table 1) were large enough (0.4–0.8 μl) to involve both superficial and deep layers. After a survival time of 7–10 days, monkeys were given a lethal dose of sodium pentobarbital (50 mg/kg or more) and perfused transcardially with 0.9% phosphate-buffered saline (PBS; pH 7.4) followed by 2% paraformaldehyde in PBS and then 2% paraformaldehyde in PBS with 10% sucrose. Cortex was separated from the brainstem, flattened, submerged for 12–36 hr in 30% sucrose in PBS under a weighted glass slide. The cortex was frozen and cut into 40 μm sections parallel to the surface so that distributions of labeled corticotectal neurons could be located throughout the total neocortical surface (Lyon and Kaas, 2002). One set of sections was mounted without further processing to determine the distribution of neurons labeled with fluorescent tracers. Adjacent sets of sections were processed for cytochrome oxidase (Wong-Riley, 1979) or myelinated fibers (Gallyas, 1979) so cortical areas could be identified histologically. The midbrain was cut in the coronal plane and processed as above.

Table 1. Numbers of labeled cells per area per injection in three species of New World monkeys*
 V1V2V3DLDMMTFSTFEFFVOtherTotal
  • *

    Of the sections examined from 10 tracer injections, the total number of cells labeled in the cortex of all three species was 14,341. Of the total cells labeled, 61% of cells were in one of the nine visual and visuomotor areas listed.

Callithrix           
 Fast blue466372568598184141591531,0192,633
 Fluororuby329146142149711899221122331,465
 Diamidino yellow771690312821201451357
 Fluoroemerald821120427161937271415
 Total880555300237185482261803161,5744,870
 % per area1811.46.24.93.89.95.41.66.532.3100
Callicebus           
 Fast blue7564001056199203803624442,186
 Fluororuby799490154671372821120985502,689
 Diamidino yellow26241810400163137
 BFITC (green)0761113673047100
 Total1,581921283140249495199391011,1045,112
 % per area31185.52.74.99.73.90.8221.5100
Aotus           
 Diamidino yellow97253108225572967596361,4992,742
 Fluororuby88257194237812556495363101,617
 Totals185510302462138551139191721,8094,359
 % per area4.211.76.910.63.212.63.24.41.741.5100
All species           
 Grand total2,6461,9868858395721,5285993104894,48714,341
 Grand % per area18.413.86.25.9410.74.22.23.431.2100

The locations and numbers of neurons labeled by retrograde transport of fluorescent tracers were determined relative to the section outline using an XY encoder attached to the stage of a fluorescent microscope. A Macintosh computer running Igor Pro software (Wavemetrics) was used to plot the locations of labeled cells. Architectonic borders derived from sections processed for CO or myelin were carefully aligned with fluorescence sections using blood vessel patterns and section contours as reference points. Photographic images of superior colliculus sections were captured using a Spot 2 digital camera (Diagnostic Instruments) and Adobe Photoshop software (Adobe Systems). In some cases, images were adjusted for brightness and contrast, but no other alterations were made.

Labeled neurons in the cortex were related to previously defined areas and regions. Three areas of visual cortex V1, V2, and the middle temporal area, MT, were myelin- and CO-dense and were easily delimited (Lyon and Kaas, 2001, 2002). The primary auditory area, A1, combined with the rostral auditory area, R (Morel and Kaas, 1992), was also easily identified by dense myelin and CO, as was the primary somatosensory field, area 3b (Jain et al., 1998). The dorsal and ventral divisions of the fundal area of the superior temporal sulcus, FSTd and FSTv, were estimated from their known locations relative to MT and the superior temporal sulcus (Kaas and Morel, 1993). Estimates of the boundaries of V3, the dorsomedial area, DM, and the dorsolateral visual area, DL (equivalent to V4), follow those of Lyon and Kaas (2001). The estimated locations of the frontal eye field, FEF, and the associated frontal visual area, FV, are based on Krubitzer and Kaas (1990). Inferior temporal cortex (IT) is a region of several visual areas (Weller and Kaas, 1987), but IT was not subdivided here. The region of the medial visual area, M (Allman and Kaas, 1976), was noted but not outlined.

RESULTS

The results are based on 10 injections of fluorescent tracers into the superior colliculus of three monkeys (Table 1). The injections varied in size and effectiveness in labeling cortical neurons. They labeled a total of 14,341 neurons that were identified and located in the examined brain sections. As sets of sections were processed for architectonic features, or specific tracers, series of one of three or one of four sections were used for counts. The number of counted labeled cells was 4,359 in the owl monkey (Aotus), 4,870 in the marmoset monkey (Callithrix), and 5,112 in the titi monkey (Callicebus). In Table 1, cell numbers are presented for each tracer injection by cortical area, so that the magnitude of areal projections could be compared. Differences in the amount of cortical surface area occupied by each cortical area can further inform quantitative comparisons of the superior colliculus projection across areas. By considering areal size (see Krubitzer and Kaas, 1990), the density of the superior colliculus projection can be compared across areas. For example, about 10% of labeled cells in the cortex were found in MT, which occupies about 2% of the total cortical surface area. Thus, labeled cells were densely distributed in MT, a relatively small cortical area. Approximately 18% of labeled cells were located in V1, which occupies about 14–18% of the total cortical surface area. Thus, while the largest number of superior colliculus projecting cells were located in V1, the density of the projection is relatively low due to the large size of V1. Together, these results provide a comprehensive overview of the cortical projection pattern to the superior colliculus. While cortical sections cut parallel to the surface are the most suitable for revealing areal patterns of labeled neurons, they are not the most suitable for revealing the laminar locations of labeled neurons. However, the labeled neurons were in the deeper sections and were likely in layer 5 and layer 6 (for V1 Meynert cells) as expected from previous results (Catman-Berrevoets et al., 1979; Fries and Distel, 1983; Fries, 1984). Results from each case are considered separately below.

Callicebus

In the titi monkey, four injections of tracers were placed in the superior colliculus using a lateral approach. As a result, the lateral superior colliculus, representing the lower visual quadrant, was more involved in the injection cores than the medial superior colliculus representing the upper visual quadrant (Fig. 1B). The injections appeared to include both superficial and intermediate layers of the superior colliculus. Although the fluorochromes caused some damage to the superior colliculus, the fast blue and fluororuby injections labeled large numbers of cortical neurons (2,186 and 2,689, respectively). In contrast, the injections of diamidino yellow (DY) and CTB conjugated to BFITC green labeled relatively few cells (137 and 100, respectively). Overall, half of the labeled neurons were in V1 (31%), V2 (18%), while nearly 10% were in MT (Fig. 1C). Large numbers of labeled cells were also located in other caudal visual areas (V3, DL, DM, and FST), cortex between MT and A1 and R, including MST, secondary auditory cortex, and multisensory cortex of the superior temporal sulcus. A few labeled cells were in inferior temporal cortex. Other labeled neurons, but relatively few, were in posterior parietal cortex, a region that includes the lateral intraparietal area (LIP) in macaque monkeys, an area described as having dense connections with the superior colliculus (Lynch et al., 1985). Regions of the frontal lobe that appear to correspond to the frontal eye field (FEF) and the associated frontal visual area (FV) collectively had nearly 3% of the labeled neurons. Large parts of neocortex, including much of the frontal lobe, limbic cortex, the ventral temporal lobe, and somatosensory cortex, were nearly devoid of labeled neurons.

Figure 1.

A: Dorsolateral view of the titi monkey brain with several of the proposed visual areas indicated. Scale bar = 1 cm. B: Photographs of the superior colliculus of the titi monkey with extents of injection sites approximated by color outlines. Colors of outlines are matched to colors of labeled cells in cortex. SC sections were stained for either cytochrome oxidase (light brown) or acetylcholinesterase (dark brown). The diagonal dotted lines on the photos indicate the approximate position of the representation of the horizontal meridian, with the upper visual field (+) represented medially and the lower visual field (−) laterally. Fluororuby and Fast blue injections included upper and lower layers of the superior colliculus and were considerably overlapped. C: Reconstruction of flattened cortex from a titi monkey with labeled corticotectal cells resulting from four fluorescent tracer injections in the superior colliculus. Injections were placed using a lateral approach, so most are centered in the representation of the lower visual field.

Callithrix

Four injections were also placed in the superior colliculus of the marmoset (Fig. 2A). As the injections were placed via a medial approach, injections were largely confined to the medial half of the superior colliculus representing the upper visual quadrant (Fig. 2B). While injection sites overlapped, they were placed in a rostrocaudal sequence that progressively involved the representations of less central parts of the upper visual quadrant. The injections involved both superficial and intermediate layers. Again, the fast blue and fluororuby injections were the most effective, labeling 2,633 and 1,465 counted cells, respectively (Table 1). The injections of diamidino yellow and fluoroemerald labeled far fewer neurons (357 and 415, respectively). As in the titi monkey, V1, V2, and MT were the three areas with the most labeled neurons (Fig. 2C). Together, they contained nearly 40% of the labeled cells. Visual areas V3, DL, DM, and FST accounted for another 20% of labeled neurons. A moderately dense focus of labeled neurons was in posterior parietal cortex. The visual and visuomotor areas of the frontal lobe, FEF and FV, contributed 8% of the labeled neurons. Other labeled neurons were scattered in prefrontal cortex and limbic cortex of the medial wall. A few labeled cells were in somatosensory and auditory cortex. Large regions of the temporal lobe and orbital frontal cortex were free of labeled neurons. As the four injection sites were placed in rostrocaudal sequence in the superior colliculus, they provided some information on the existence of topographic patterns of projections of central areas. As expected from injections placed in the medial half of the superior colliculus, representing the upper visual quadrant, labeled neurons for all four injections were concentrated in the lateral halves of V1, V2, V3, DL, and MT, which also represent the upper visual quadrant (Lyon and Kaas, 2001). The four injections overlapped each other in the superior colliculus, so labeled regions of cortex for each injection also overlapped. In addition, some of the labeled neurons were in the medial representations of the lower visual quadrant in V1, V2, V3, and MT, indicating that some spread of the effective injection sites into the lateral superior colliculus was likely. Nevertheless, the most rostral injection of fluororuby in the superior colliculus, corresponding to the most central part of the upper quadrant representation, labeled the most medial parts of lateral, V1, V2, V3, and DL, and the most caudal part of lateral MT, all corresponding to the more central parts of the representations of the upper visual quadrant. Thus, there is some evidence for topographic patterns of projections from visuotopically organized visual areas to the superior colliculus.

Figure 2.

A: Dorsolateral view of the marmoset monkey brain with proposed visual areas indicated. Scale bar = 1 cm. B: The approximate retinotopic positions of injection sites in the superior colliculus are shown on a schematic drawing of the dorsal surface of the SC (based on Lane et al., 1973). C: Surface-view distribution of labeled corticotectal cells in a marmoset monkey shown on a reconstruction of flattened cortex after injection of four distinguishable fluorescent tracers into the SC. Areal boundaries were determined from adjacent cortex sections stained for myelin or cytochrome oxidase.

Aotus

The owl monkey had only two injections into the superior colliculus, but both the diamidino yellow and fluororuby injections labeled large numbers of cortical neurons (DY, 2,742; FR, 1,617; Table 1). As the injections were focused in the deeper layers of the superior colliculus (Fig. 3C), V1 was not densely labeled. However, large numbers of neurons were labeled in V2, V3, DL, MT, and DM; Fig. 3D), visual areas that project deeper into the superficial gray (see Graham et al., 1979; Cusick, 1988; Kaas and Huerta, 1988). V2, DL, and MT contained 35% of the labeled neurons. Labeled cells were also present in caudal IT, FST, and DM, but few were in posterior parietal cortex. In the frontal lobe, both FEF and FV contained concentrations of labeled neurons, and other labeled neurons were in caudal cortex of the medial wall, just rostral to V2. Except for a few labeled neurons just medial to the FEF, and those in FEF and FV, the frontal lobe was devoid of labeled neurons, as was much of the ventral temporal lobe. Only an occasional neuron was labeled in somatosensory or auditory cortex.

Figure 3.

A: Dorsolateral view of an owl monkey brain with proposed visual areas. Scale bar = 1 cm. B: Schematic representation of the visuotopic map in the superior colliculus of owl monkey (based on Lane et al., 1973), with approximate locations of tracer injections indicated. C: Four coronal sections from the right superior colliculus of owl monkey 00-77 demonstrating tracer injection locations. The cytochrome oxidase section (top left) shows the depth of the diamidino yellow injection, shown in the top right panel. The tracer did not spread to include the superficial-most layers of the superior colliculus. The Nissl-stained section (bottom left) illustrates the location of the fluororuby injection site, shown in the bottom right panel. Again, the tracer appears to be found mainly in the deeper layers of the superior colliculus. Scale bar = 1 mm. D: Distributions of labeled corticotectal cells on a reconstruction of flattened cortex from this owl monkey following injections of diamidino yellow (DY) and fluororuby (FR). Injection sites were placed in the superior colliculus using a midline approach. Both injections were centered in the lower layers of the superior colliculus in the representation of the upper visual field and were heavily overlapped. Scale bar = 2 mm.

As expected from injections placed in the medial part of the superior colliculus (Fig. 3B) representing the upper visual quadrant, the most dense clusters of labeled neurons in areas V2, V3, DL, and MT were in the lateral portions of these areas representing the upper visual quadrant. As the two injections were very close to each other, the dense clusters of labeled neurons tended to overlap. In addition, labeled neurons were scattered within medial parts of V1, V2, V3, DL, and MT, suggesting a spread of injected tracers into the lateral half of the superior colliculus.

DISCUSSION

The primary contribution of the present study is to provide an overview of the corticotectal projection pattern in New World monkeys. Results are based on a total of 10 injections and 14,341 corticotectal neurons that were localized relative to brain surface landmarks and architectonic boundaries. Most of the injections were large enough to involve the superficial to deeper layers I–IV and thus label neurons in visual as well as visuomotor areas. As cortex was manually flattened after the brain was perfused, and cut parallel to the surface, surface-view distributions of labeled neurons were accurately reconstructed. This study is the first in New World monkeys that examines the total cortical projection pattern to the superior colliculus, and the first study where the proportions of labeled neurons in a number of well-defined cortical areas are compared to each other and to the total distribution of labeled neurons. The most comparable previous results were obtained after superior colliculus injections in macaque monkeys (Fries, 1984; Lock et al., 2003). In these studies, labeled neurons were localized relative to the surface of the brain (Fries, 1984), or proposed visual areas on a reconstructed, flattened surface view of visual cortex (Lock et al., 2003). While cell counts by visual area were not provided in these studies, illustrations reflecting cell numbers allow useful comparisons with present results. Distributions of labeled cortical cells after superior colliculus injections have also been shown for tree shrews (Casseday et al., 1979), cats (Hollander, 1974; Kawamura and Konno, 1979; Tortelly et al., 1980; Meredith and Clemo, 1989), and rats (Olavarria and Van Sluyters, 1982; Thong and Dreher, 1986).

Distributions of Corticotectal Cells in New World and Old World Monkeys, Cats, and Rats

The present results suggest that primary visual cortex, V1, is the major contributor of corticotectal projections in New World monkeys. In both the titi monkey and the marmoset, three of the four injections labeled more neurons in V1 than any other area. Only the least effective injection in each of these cases failed to label the most neurons in V1, probably because the injection site was deeper than most of the terminations from V1 (Graham et al., 1979). The two injections in the owl monkey were deep and also failed to label many cells in V1 for the same reason. Injections in V1 of owl monkeys densely label the superior colliculus (Graham et al., 1979). The present results suggest that 20–30% of the corticotectal cells are located in V1 of titi and marmoset monkeys, and this may be the case for owl monkeys as well. Judging from the corticotectal cell distributions illustrated by Fries (1984) and especially Lock et al. (2003), more projections originate in V1 than any other visual area in macaques. Large numbers of corticotectal neurons, perhaps more than any other cortical area, are also found in V1 of tree shrews (Casseday et al., 1979), cats (Hollander, 1974; Kawamura and Konno, 1979), and rats (Olavarria and Van Sluyters, 1982; Thong and Dreher, 1986). Part of the reason for this may simply be that V1 is the largest of visual cortical areas. However, the functional implication of this conclusion is that relatively unprocessed visual information from V1 is an important component of the corticotectal system. The large V1 projection to the superior colliculus in monkeys may compensate in part for a proportionally limited retinal projection. As few as 10% of retinal ganglion cells appear to project to the superior colliculus of monkeys (Perry and Cowey, 1984; Weller and Kaas, 1989).

The second visual area, V2, provides the next largest projection to the superior colliculus. Present results from the titi and marmoset monkeys suggest that the V2 projection is in the range of one-half to two-thirds the size of the projection from V1. As V2 in these New World monkeys is about half the size of V1 (Krubitzer and Kaas, 1990), V2 projections to the superior colliculus are at least as dense as those from V1. Macaques also appear to have large numbers of corticotectal neurons in V2 (Fries, 1984; Lock et al., 2003). As V2 is described as nearly the size of V1 in macaques (Sincich et al., 2003), the number of corticotectal cells in V2 may approach those in V1. In macaques, corticotectal neurons appear to be located preferentially in the thick cytochrome oxidase stripes of V2 that are associated with the M cell, dorsal stream pathway (Abel et al., 1997). In cats, V2 also projects densely to the superior colliculus (Hollander, 1974; Kawamura and Konno, 1979). In rats, there are uncertainties about the location and extent of V2, but a large number of projection cells have been described (Olavarria and Van Sluyters, 1982) in the cortical zone identified as V2 of rats by Rosa and Krubitzer (1999).

While the middle temporal area, MT, is about 1/10 the size of V1 of New World monkeys (Allman and Kaas, 1971; Tootell et al., 1985; Krubitzer and Kaas, 1990), MT appears to have one-half to one-third as many corticotectal neurons as V1. Thus, the projections from MT are denser than expected from the premise that early visual areas project with nearly equal density to the superior colliculus. In macaques, only V1, V2, V3, and MT were labeled by an injection restricted to the superficial layers of the superior colliculus (Lock et al., 2003), and MT appears to be a major source of corticotectal projections (Fries, 1984; Lock et al., 2003). MT is a key visual area of the dorsal stream of processing concerned with visuomotor behavior (action) (Ungerleider and Mishkin, 1982; Goodale and Milner, 1992), and it is almost exclusively activated by relays of the M-cell retinal input (Maunsell et al., 1990) that also directly activates the superior colliculus (Perry and Cowey, 1984). Microstimulation of MT affects visual saccades (Groh et al., 1997), possibly through projections to the superior colliculus. Thus, it is not surprising that MT contributes a very dense projection to the superior colliculus. Current evidence suggests that MT is a subdivision of visual cortex unique to primates (Kaas, 2002), but the posteromedial area of the lateral suprasylvian sulcus (PMLS or the Clare-Bishop area) of cats is functionally similar to MT (Payne, 1993), and PMLS has dense connections with the superior colliculus (Kawamura and Konno, 1979).

Other visual areas also project to the superior colliculus. In the present cases, limited proportions (3–7%) of the corticotectal neurons originated from areas V3, DL, DM, and FST. Some labeled cells were also in the regions of M, MST, and caudal IT. Surprisingly few such neurons were in posterior parietal cortex, including the regions corresponding to the lateral intraparietal area, LIP, and the ventral intraparietal area, VIP, of macaques. LIP and VIP are areas with direct connections with MT (Maunsell and van Essen, 1983; Blatt et al., 1990), and LIP reportedly has dense projections to the superior colliculus (Lynch et al., 1985). Area DL of New World monkeys is largely equivalent to area V4 as defined in macaques (Stepniewska et al., 2005), which is part of the ventral stream for object vision (Ungerleider and Mishkin, 1982). Being part of the ventral stream may account for the limited projection of DL to the superior colliculus (3–5% in titi and marmoset monkeys; 10% in the owl monkey). However, a smaller rostral division of DL appears to be part of the dorsal stream (Weller and Kaas, 1987; Cusick and Kaas, 1988). Yet rostral DL was not more densely populated with corticotectal neurons than caudal DL. Approximately 6% of the labeled neurons were in V3. As V3 is about half the width of V2, and not as long (Lyon and Kaas, 2002), the proportion of labeled neurons of roughly half of that of V2 indicates that projections of V2 and V3 to superior colliculus are equally dense. Lock et al. (2003) illustrated many labeled neurons in V3 of macaques after superficial superior colliculus injections. About 4% of the labeled neurons in the present cases were in each of areas DM and FST. As these areas are densely interconnected with area MT (e.g., Krubitzer and Kaas, 1990; Kaas and Morel, 1993), they likely contribute to the dorsal stream of processing and have functional roles that complement MT. Thus, it is not surprising that these areas together have about the same proportion of superior colliculus projecting cells as MT. In macaques, superior colliculus injections labeled neurons in both the FST and DM (V3A) regions (Lock et al., 2003). While dorsal FST and DM did have a considerable number of corticotectal cells, another MT target, MST, contained only moderate to sparse distributions of labeled neurons. In macaques, superior colliculus injections labeled a substantial number of neurons in LIP, but not in MST (Lock et al., 2003). Finally, ventral stream visual areas of the ventral temporal lobe appear to be nearly devoid of corticotectal neurons in both New World and Old World monkeys.

Other important sources of projections to the superior colliculus in New World monkeys are the regions of the frontal eye field (FEF) and the frontal visual area (FV). The FEF has been identified in New World monkeys by intracortical microstimulation (Huerta et al., 1986; Krubitzer and Kaas, 1990), and the FV area was defined as the region just ventral to the FEF with inputs from MT (Krubitzer and Kaas, 1990). As FV may correspond to a part of the FEF devoted to visual fixation, the areas may be parts of a single complex. Together, they accounted for nearly 6% of the corticotectal projections. Although labeled neurons were also located in prefrontal cortex, no other portion of the frontal cortex, including the medial region of the supplementary eye field, was as densely or as consistently labeled as FEF plus FV. In macaques, Lock et al. (2003) did not illustrate distributions of labeled corticotectal neurons in the frontal lobe, but mentioned that labeled neurons were found in the frontal eye field. Fries (1984) portrayed a moderately dense distribution of labeled neurons in and around the FEF of macaques after superior colliculus injections. Other labeled neurons were medial to the FEF in the region of the supplementary motor area (SMA) and the supplementary eye field (SEF). Injections placed in the FEF of either New World monkeys or macaques labeled terminations in the deeper layers of the superior colliculus (Leichnetz et al., 1981; Huerta et al., 1986). In macaques, projections to the superior colliculus from the supplementary eye field were described as sparser and deeper than those from the FEF (Huerta and Kaas, 1990). Possibly sparse SEF projections in New World monkeys were not labeled by the superior colliculus injections in the present cases. Projections from the FEF and SEF to the superior colliculus are consistent with the major involvement of both of these areas and the superior colliculus in mediating eye movements (Schiller and Tehovnik, 2001).

Neurons in the deeper layers of the superior colliculus are known to be responsive to auditory and somatosensory stimuli in monkeys (Updyke, 1974; Wallace et al., 1996) and other mammals (Stein et al., 2004). Some of the input to the deeper layers comes from subcortical relays, but cortical sources have also been demonstrated, especially in cats (Stein et al., 1983; Meredith and Clemo, 1989; Harting et al., 1992). In the present study of New World monkeys, there was only evidence of very sparse projections from auditory and somatosensory cortical areas to the superior colliculus. In macaques, Fries (1984) did not find labeled neurons in somatosensory areas 3a, 3b, 1, and 2, but some labeled neurons were in the S2 region. Other labeled neurons were in the auditory belt bordering primary auditory cortex. Lock et al. (2003) did not describe the connections of somatosensory and auditory fields. In cats, the main auditory fields failed to project to the superior colliculus (Meredith and Clemo, 1989) and of the somatosensory areas; only the fourth somatosensory area (S4) appears to project to the superior colliculus (Stein et al., 1983). In rats, primary somatosensory cortex may project sparsely to the superior colliculus (Thong and Dreher, 1986). Thus, much of the auditory and somatosensory activation in the superior colliculus appears to be from subcortical sources, and very little from primary and secondary somatosensory and auditory areas across these mammals.

Acknowledgements

Supported by the National Institutes of Health grant EY002686 (to J.H.K.).

Ancillary