There are many levels of disorder in amblyopic vision, from basic acuity and contrast sensitivity loss to abnormal binocular vision and global perception of motion and form. Amblyopia treatment via patching to restore acuity often leaves other aspects of vision deficient. The source for these additional deficits is unclear. Neural correlates of poor binocular function and acuity loss are found in V1 and V2. However, they are generally not sufficient to account for behaviorally measured vision loss. This review summarizes the known cortical correlates of visual deficits found in association with amblyopia, particularly those relevant to binocular vision and higher-order visual processing, in striate and extrastriate cortex. Recommendations for future research address open questions on the role of suppression and oculomotor abnormalities in amblyopic vision, and underexplored mechanisms such as top-down influences on information transmission in the amblyopic brain.

The short-wavelength-sensitive (S) cones play an important role in color vision of primates, and may also contribute to the coding of other visual features, such as luminance and motion. The color signals carried by the S cones and other cone types are largely separated in the subcortical visual pathway. Studies on nonhuman primates or humans have suggested that these signals are combined in the striate cortex (V1) following a substantial amplification of the S-cone signals in the same area. In addition to reviewing these studies, this review describes the circuitry in V1 that may underlie the processing of the S-cone signals and the dynamics of this processing. It also relates the interaction between various cone signals in V1 to the results of some psychophysical and physiological studies on color perception, which leads to a discussion of a previous model, in which color perception is produced by a multistage processing of the cone signals. Finally, I discuss the processing of the S-cone signals in the extrastriate area V2.

White noise stimuli were used to estimate second-order kernels for complex cells in cortical area VI of the macaque monkey, and drifting grating stimuli were presented to the same sample of neurons to obtain orientation and spatial-frequency tuning curves. Using these data, we quantified how well second-order kernels predict the normalized tuning of the average response of complex cells to drifting gratings.

The estimated second-order kernel of each complex cell was transformed into an interaction function defined over all spatial and temporal lags without regard to absolute position or delay. The Fourier transform of each interaction function was then computed to obtain an interaction spectrum. For a cell that is well modeled by a second-order system, the cell’s interaction spectrum is proportional to the tuning of its average spike rate to drifting gratings. This result was used to obtain spatial-frequency and orientation tuning predictions for each cell based on its second-order kernel. From the spatial-frequency and orientation tuning curves, we computed peaks and bandwidths, and an index for directional selectivity.

We found that the predictions derived from second-order kernels provide an accurate description of the change in the average spike rate of complex cells to single drifting sine–wave gratings. These findings are consistent with a model for complex cells that has a quadratic spectral energy operator at its core but are inconsistent with a spectral amplitude model.

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

The effects of monocular deprivation on cytochrome-oxidase (CO) expression were used to reveal ocular dominance columns in flatmounts of the striate cortex in macaque (Macaca fascicularis) and talapoin (Miopithecus talapoin) monkeys. This procedure allowed the first direct visualization of the complete pattern of ocular dominance bands in macaque monkeys, and less complete reconstructions in talapoin monkeys. In a second macaque monkey, the ocular dominance organization was revealed by injecting wheat germ agglutinin conjugated to horseradish peroxidase (WGA-HRP) into one eye.

The organization of ocular dominance columns in the macaque monkeys conforms to previous descriptions, but the flat-mounted hemispheres provide accurate details on (1) the arrangement of columns, (2) the orientation of the representation of the optic disc, and (3) the breakdown of the bands in the cortex between the optic disc and monocular representations into a pattern of dots activated by the ipsilateral eye and larger surrounds related to the contralateral eye.

Talapoin monkeys, the smallest of Old World monkeys, have sharply segregated ocular dominance bands. The columns in talapoins are narrower than those in macaques, so that even with less striate cortex than macaques, talapoins have more ocular dominance hypercolumns.

Layers II/III of the primary visual cortex contain a regular pattern of histochemically detectable cytochrome oxidase (CO)-rich “puffs,” which differ from the interpuff regions in their thalamo-cortical and cortico-cortical connectivity, receptive-field properties, and the density of inhibitory GABA-containing synaptic terminals. We used an immunocytochemical method, in combination with cytochrome oxidase histochemistry, to analyze the spatial relationship between neurons that contain neuropeptide Y (NPY) and the CO puffs. Of a total of 606 neurons, only 2.6% of the NPY-containing cells are located in the puffs, whereas the rest are situated in the interpuffs, or at the interface between puffs and interpuffs. The number of NPY-containing neurons in the puffs is substantially less than that expected in an equal volume of the interpuffs (X2 = 13.86; df = 1; P < 0.001).

These observations indicate that columns containing the puffs may differ also from those in the interpuff regions in that they contain a unique array of chemically and morphologically distinct local circuit neurons.

Responses to full-field colored flashes (red, blue, and green) were compared with those to illuminancematched white flashes in area V1, optic radiations, and the lateral geniculate nucleus of two alert macaques. Laminar profiles of visual evoked potentials (VEPs), current source density, and multiunit activity were obtained using multicontact electrodes capable of sampling from all layers of cortex or lateral geniculate nucleus, simultaneously. In striate cortex, stimulation with colored flash enhanced transmembrane current flow dramatically in both layer 4c and the supragranular laminae. Stimulation with red evoked the largest enhancement in every electrode penetration. The mean peak amplitudes of current sinks evoked by red were 203% and 537% of those evoked by white light in layer 4c and the supragranular laminae, respectively. Color effects in VI were preceded by an initial epoch of wavelength-insensitive activity. In layer 4c, the red effect reached significance, on average, at 47 ms, or ≈24 ms after the onset of transmembrane current flow. In the supragranular layers, the red effect reached significance, on average, at 55 ms, or ≈14 ms after the onset of current flow. Recordings from optic radiations in the white matter below V1 and from lateral geniculate nucleus showed no significant difference in the responses to color and illuminance-matched white light. Enhancement of supragranular current flow with color stimulation increased the contribution of these laminae to the generation of the surface VHP. Comparison of the surface VHP wave forms evoked by white and color stimuli may, therefore, help to differentiate the responses of the granular and supragranular laminae.

Measurements were made of the optimal velocity for drifting bar-shaped stimuli to excite striate cortex neurons of the cat. These data were compared to the optimal spatial and temporal frequencies of the same neurons, as determined with drifting sine-wave grating stimuli. A systematic relationship was revealed, whereby those neurons preferring higher velocities of bar motion also preferred lower spatial and higher temporal frequencies of gratings. The optimal bar velocity for a given neuron could be quantitatively predicted from the ratio of that neuron's optimal temporal frequency to its optimal spatial frequency.

It is believed that spatial summation in most simple cells is a linear process. If this were so, then the Fourier transform of a simple cell's line weighting function should predict the cell's spatial frequency tuning curve. We have compared such predictions with experimental measurements and have found a consistent discrepancy: the predicted tuning curve is much too broad. We show qualitatively that this kind of discrepancy is consistent with the well-known threshold nonlinearity shown by most cortical cells. We have tested quantitatively whether a response threshold could explain the observed disagreements between predictions and measurements: a least-squares minimization routine was used to fit the inverse Fourier Transform of the measured frequency tuning curve to the measured line weighting function. The fitting procedure permitted us to introduce a threshold to the reconstructed line weighting function. The results of the analysis show that, for all of the cells tested, the Fourier transforms produced better predictions when a response threshold was included in the model. For some cells, the actual magnitude of the response threshold was measured independently and found to be compatible with that suggested by the model. The effects of nonlinearities of spatial summation are considered.

To study the relationship between ocular dominance columns (ODCs) and axonal projections of individual layer 6 pyramidal neurons in the primary visual cortex, neurons were intracellularly labeled with biocytin in live slices prepared from macaque monkeys that had received an intravitreal injection of tetrodotoxin (TTX). The TTX injection indirectly causes a decrease in cytochrome oxidase (CO) expression in the cortical ODCs corresponding to the treated eye (Wong-Riley & Carroll, 1984). Sections from slices with labeled layer 6 neurons were double stained for biocytin and CO, to allow visualization of neuronal processes as well as ODCs. Twenty-seven layer 6 pyramidal neurons in ODC-labeled slices were analyzed. These neurons were classified according to the criteria of Wiser and Callaway (1996). Eight of these are class I neurons, which have dense axonal projections to the monocular layer 4C. The remaining 19 are class II neurons which project primarily to the binocular layers outside 4C. Among class I neurons, two have dense axonal arbors in layer 4Cα (type Iα), one in layer 4Cβ (type Iβ), and two throughout the depth of layer 4C (type IC). None of these neurons have ODC-specific axonal arbors. The remaining three class I neurons have focused axonal projections in layers 4Cβ and 4A (type IβA). All three appear to have axonal arbors predominantly within their home ODC in layer 4C. The axonal arbors of class II neurons do not appear to relate to ODCs in any specific fashion.

The compartmental organization of visual cortical neurons was examined across species of primates by directly comparing the pattern of immunoreactivity for the 28-kD vitamin D-dependent calcium-binding protein (calbindin) in area 17 of squirrel monkeys, macaques, and neurologically normal adult humans. Area 17 of macaques and squirrel monkeys was similar in that somata and processes intensely immunoreactive for calbindin were present in the same layers (II-III, IVB, and V) and in both species formed a well-stained matrix that surrounded the CO-rich puffs in layer III. These intensely calbindin-immunoreactive neurons were identified as subpopulations of GABA-immunoreactive neurons. Among the most obvious differences in the two monkey species was the distribution of calbindin-positive elements outside of layer III: a dense immunostained matrix surrounded the puffs in layers II, IVB, V, and VI of squirrel monkeys but the immunostained neurons adopted no regular pattern outside layer III in macaques. In addition, although somata lightly immunoreactive for calbindin were present in both species, they were much more abundant in squirrel monkeys than macaques. The pattern of calbindin immunostaining in human area 17 resembled that of macaques in forming an intense matrix that surrounded puffs only in layer III, yet also resembled that of squirrel monkeys by including large numbers of lightly immunoreactive somata. These lightly immunostained somata included a very dense population forming a prominent band in layer IVA of human visual cortex. We conclude that for layer III of primary visual cortex, a similar pattern of neuronal chemistry exists across species of primates which is related to this layer's compartmental organization. Yet for other layers, the expression of calbindin immunoreactivity varies from one species to the next, perhaps reflecting variations in other neuronal properties.

The monoclonal antibody Cat-301 was used to stain neurons and neuropil in the visual thalamus and cortex of tree shrews —small, highly visual mammals that are closely related to primates. Previously, this antibody has been found to label neurons associated with the Y-cell stream of processing in cats and the magnocellular or M-cell stream in macaque monkeys. In tree shrews, the antibody selectively labeled layers 1, 2, 4, and 5 of the dorsal lateral geniculate nucleus, layers that are likely to contain neurons previously classified as Y-cells. Of the two layers that contain W-cells, layer 3 was unlabeled and layer 6 was lightly labeled. In area 17, layer 3c was densely stained, as in cats and macaque monkeys. The external half of layer 5 was also densely stained, in contrast to cats where the internal half of layer 5 is stained and macaques where layer 5 is sparsely stained. Area 18 was characterized by dense, uniform staining of inner layer 3 and outer layer 5, but no pattern of alternating light and dense bands crossed the width of area 18 as in macaques. Dense labeling of these same sublayers occurred in cortical areas TA and TD just lateral to area 18. Area TD may be the homologue of area MT of primates, which also stains densely with Cat-301 in macaques. These results indicate that Cat-301 differentially labels layers and areas in the visual system of tree shrews, and raise intriguing issues of comparison among tree shrews, primates, and cats.

The lateral geniculate nucleus (LGN) of primates contains three classes of relay cells, the magnocellular (M), parvocellular (P), and koniocellular (K) cells. At present, very little is known about either the structure or function of the K relay cells in New or Old World monkeys (simian primates). In monkeys, K cells are located between the main LGN layers and adjacent to the optic tract. For convenience, these intercalated cell layers are numbered K1-K4 starting closest to the optic tract with K1. The objective of this study was to examine the details of K axon morphology in the primary visual cortex (V1) of owl monkeys and to determine if different K layers give rise to distinct axon types. For this purpose, injections of WGA-HRP or PHA-L were made into specific K LGN layers and the distribution and morphology of the resulting labeled axons were analyzed. Injections of fluorescent tracers also were made within the superficial layers of V1 to further document connections via analysis of the patterns of retrogradely labeled cells in the LGN. Our main finding is that K axons in owl monkeys terminate as delicate focused arbors within single cytochrome oxidase (CO) blob columns in cortical layer III and within cortical layer I. Overall, the morphology of the K axons in these monkeys is quite similar to what we described previously for K geniculocortical axons in the distantly related bush baby (prosimian primate), suggesting that the basic features of this pathway are common to all primates. Our results also provide evidence that the axon arbors from different K layers are morphologically distinct; axons from LGN layer K1 project mainly to cortical layer I, while axons from LGN layer K3 chiefly terminate in cortical layer III. Taken together, these results imply that the basic features of axons within the K pathway are conserved across primates, and that the K axons from different K layers are likely to differ in function based upon their different morphologies.

To characterize the postnatal development of geniculocortical axon arbor morphology in owl monkeys at a series of ages from birth to adulthood, individual arbors were bulk-filled with HRP in brain slice preparations and were reconstructed from serial sections. At all ages, cortical layers and sublayers were obvious. Presumed M or magnocellular arbors were largely confined to layer IVα, but they also extended into layer IIIc (IVB of Brodmann, 1909); presumed P or parvocellular arbors were almost exclusively confined to layer IVβ. Other axons that may reflect feedback projections from MT terminated in layer IIIc. Overall, M axon arbors increased in size and complexity from birth to adulthood with mean surface-view arbor areas ranging from 0.08 ± 0.01 mm2 in newborns to 0.24 ± 0.02 mm2 in adults. The developing P arbor areas were, on average, as large or larger than adult (newborn = 0.07 ± 0.01 mm2, adult = 0.047 ± 0.01 mm2; n.s.) but the arbors were somewhat less complex. Since the brain and area 17 increase in size postnatally, the proportion of area 17 subserved by each P arbor would decrease in postnatal development. Terminal boutons with immature features were evident in both M and P populations at all developmental ages. The results indicate that, while both LGN axon types in monkeys undergo morphological changes postnatally, M arbors appear to mature by increasing arbor size and terminal branching complexity, whereas P arbors increase in complexity but not in size. These distinct programs of axon arbor development suggest that the periods of susceptibility of geniculocortical axon arbors to postnatal influences of the environment, and the types of plastic responses they potentially exhibit, are class-specific.

We investigated the dynamics of neurons in the striate cortex (V1) and the lateral geniculate nucleus (LGN) to study the transformation in temporal-frequency tuning between the LGN and V1. Furthermore, we compared the temporal-frequency tuning of simple with that of complex cells and direction-selective cells with nondirection-selective cells, in order to determine whether there are significant differences in temporal-frequency tuning among distinct functional classes of cells within V1. In addition, we compared the cells in the primary input layers of V1 (4a, 4cα, and 4cβ) with cells in the layers that are predominantly second and higher order (2, 3, 4b, 5, and 6). We measured temporal-frequency responses to drifting sinusoidal gratings. For LGN neurons and simple cells, we used the amplitude and phase of the fundamental response. For complex cells, the elevation of impulse rate (F0) to a drifting grating was the response measure. There is significant low-pass filtering between the LGN and the input layers of V1 accompanied by a small, 3-ms increase in visual delay. There is further low-pass filtering between V1 input layers and the second- and higher-order neurons in V1. This results in an average decrease in high cutoff temporal-frequency between the LGN and V1 output layers of about 20 Hz and an increase in average visual latency of about 12–14 ms. One of the most salient results is the increased diversity of the dynamic properties seen in V1 when compared to the cells of the lateral geniculate, possibly reflecting specialization of function among cells in V1. Simple and complex cells had distributions of temporal-frequency tuning properties that were similar to each other. Direction-selective and nondirection-selective cells had similar preferred and high cutoff temporal frequencies, but direction-selective cells were almost exclusively band-pass while nondirection-selective cells distributed equally between band-pass and low-pass categories. Integration time, a measure of visual delay, was about 10 ms longer for V1 than LGN. In V1 there was a relatively broad distribution of integration times from 40–80 ms for simple cells and 60–100 ms for complex cells while in the LGN the distribution was narrower.

We studied excitatory local circuits in the macaque primary visual cortex (V1) to investigate their relationships to the magnocellular (M) and parvocellular (P) streams. Sixty-two intracellularly labeled spiny neurons in layers 2–5 were analyzed. We made detailed observations of the laminar and columnar specificity of axonal arbors and noted correlations with dendritic arbors. We find evidence for considerable mixing of M and P streams by the local circuitry in V1. Such mixing is provided by neurons in the primary geniculate recipient layer 4C, as well as by neurons in both the supragranular and infragranular layers. We were also interested in possible differences in the axonal projections of neurons with different dendritic morphologies. We found that layer 4B spiny stellate and pyramidal neurons have similar axonal arbors. However, we identified two types of layer 5 pyramidal neuron. The majority have a conventional pyramidal dendritic morphology, a dense axonal arbor in layers 2–4B, and do not project to the white matter. Layer 5 projection neurons have an unusual “backbranching” dendritic morphology (apical dendritic branches arc downward rather than upward) and weak or no axonal arborization in layers 2–4B, but have long horizontal axonal projections in layer 5B. We find no strong projection from layer 5 pyramidal neurons to layer 6. In macaque V1 there appears to be no single source of strong local input to layer 6; only a minority of cells in layers 2–5 have axonal branches in layer 6 and these are sparse. Our results suggest that local circuits in V1 mediate interactions between M and P input that are complex and not easily incorporated into a simple framework.

Striate cortical cells were classified according to whether or not their preferred orientation was close to one of the “primary” orientations (horizontal, vertical or radial, i.e. directed to the area centralis) and according to their ordinal position on the afferent pathway from the dorsal lateral geniculate nucleus (dLGN). Among the neurones that could be driven monosynaptically from the dLGN, there was a high representation of those with a preference for the primary orientations. This was particularly evident in the case of C (complex) cells. There was no such preponderance of primary orientations among the polysynaptically activated cells. It is proposed that the asymmetry of distribution seen among the first-order cells reflects the asymmetry seen subcortically in neurones that show orientation biases. It may be that the cortex elaborates a more uniform representation of orientations only at the higher ordinal levels.

The thalamic reticular nucleus (TRN) supplies an important inhibitory input to the dorsal thalamus. Previous studies in non-primate mammals have suggested that the visual sector of the TRN has a lateral division, which has connections with first-order (primary) sensory thalamic and cortical areas, and a medial division, which has connections with higher-order (association) thalamic and cortical areas. However, the question whether the primate TRN is segregated in the same manner is controversial. Here, we investigated the connections of the TRN in a New World primate, the marmoset (Callithrix jacchus). The topography of labeled cells and terminals was analyzed following iontophoretic injections of tracers into the primary visual cortex (V1) or the dorsal lateral geniculate nucleus (LGNd). The results show that rostroventral TRN, adjacent to the LGNd, is primarily connected with primary visual areas, while the most caudal parts of the TRN are associated with higher order visual thalamic areas. A small region of the TRN near the caudal pole of the LGNd (foveal representation) contains connections where first (lateral TRN) and higher order visual areas (medial TRN) overlap. Reciprocal connections between LGNd and TRN are topographically organized, so that a series of rostrocaudal injections within the LGNd labeled cells and terminals in the TRN in a pattern shaped like rostrocaudal overlapping “fish scales.” We propose that the dorsal areas of the TRN, adjacent to the top of the LGNd, represent the lower visual field (connected with medial LGNd), and the more ventral parts of the TRN contain a map representing the upper visual field (connected with lateral LGNd).

The preferred stimulus size of a V1 neuron decreases with increases in stimulus contrast. It has been supposed that stimulus contrast is the primary determinant of such spatial summation in V1 cells, though the extent to which it depends on other stimulus attributes such as orientation and spatial frequency remains untested. We investigated this by recording from single cells in V1 of anaesthetized cats and monkeys, measuring size-tuning curves for high-contrast drifting gratings of optimal spatial configuration, and comparing these curves with those obtained at lower contrast or at sub-optimal orientations or spatial frequencies. For drifting gratings of optimal spatial configuration, lower contrasts produced less surround suppression resulting in increases in preferred size. High contrast gratings of sub-optimal spatial configuration produced more surround suppression than optimal low-contrast gratings, and as much or more surround suppression than optimal high-contrast gratings. For sub-optimal spatial frequencies, preferred size was similar to that for the optimal high-contrast stimulus, whereas for sub-optimal orientations, preferred size was smaller than that for the optimal high-contrast stimulus. These results indicate that, while contrast is an important determinant of spatial summation in V1, it is not the only determinant. Simulation of these experiments on a cortical receptive field modeled as a Gabor revealed that the small preferred sizes observed for non-preferred stimuli could result simply from linear filtering by the classical receptive field. Further simulations show that surround suppression in retinal ganglion cells and LGN cells can be propagated to neurons in V1, though certain properties of the surround seen in cortex indicate that it is not solely inherited from earlier stages of processing.