The corticogeniculate circuit is an evolutionarily conserved pathway linking the primary visual cortex with the visual thalamus in the feedback direction. While the corticogeniculate circuit is anatomically robust, the impact of corticogeniculate feedback on the visual response properties of visual thalamic neurons is subtle. Accordingly, discovering the function of corticogeniculate feedback in vision has been a particularly challenging task. In this review, the morphology, organization, physiology, and function of corticogeniculate feedback is compared across mammals commonly studied in visual neuroscience: primates, carnivores, rabbits, and rodents. Common structural and organizational motifs are present across species, including the organization of corticogeniculate feedback into parallel processing streams in highly visual mammals.

Although the core functions and structure of the lateral geniculate nucleus (LGN) are well understood, this core is surrounded by questions about the integration of feedforward and feedback connections, interactions between different channels of information, and how activity dependent development restructures synaptic networks. Our understanding of the organization of the mouse LGN is particularly limited given how important it has become as a model system. Advances in circuit scale electron microscopy (cellular connectomics) have made it possible to reconstruct the synaptic connectivity of hundreds of neurons within in a circuit the size of the mouse LGN. These circuit reconstructions can reveal cell type-to-cell type canonical wiring diagrams as well as the higher order wiring motifs that are only visible in reconstructions of intact networks. Connectomic analysis of the LGN therefore not only can answer longstanding questions about the organization of the visual thalamus but also presents unique opportunities for investigating fundamental properties of mammalian circuit formation.

Synchronized oscillatory discharge in the visual cortex has been proposed to underlie the linking of retinotopically disparate features into perceptually coherent objects. These proposals have largely relied on the premise that the oscillations arise from intracortical circuitry. However, strong oscillations within both the retina and the lateral geniculate nucleus (LGN) have been reported recently. To evaluate the possibility that cortical oscillations arise from peripheral pathways, we have developed two plausible models of single cell oscillatory discharge that specifically exclude intracortical networks. In the first model, cortical oscillatory discharge near 50 Hz in frequency arises from the integration of signals from strongly oscillatory cells within the LGN. The model also predicts the incidence of 50-Hz oscillatory cells within the cortex. Oscillatory discharge around 30 Hz is explained in a second model by the presence of intrinsically oscillatory cells within cortical layer 5. Both models generate spike trains whose power spectra and mean firing rates are in close agreement with experimental observations of simple and complex cells. Considered together, the two models can largely account for the nature and incidence of oscillatory discharge in the cat's visual cortex. The validity of these models is consistent with the possibility that oscillations are generated independently of intracortical interactions. Because these models rely on intrinsic stimulus-independent oscillators within the retina and cortex, the results further suggest that oscillatory activity within the cortex is not necessarily associated with the processing of high-order visual information.

Using a combination of retrograde horseradish peroxidase (HRP) labelling, silver staining, and electron microscopy, we have assessed the relationship between retinal ganglion cell soma size and axon diameter in the adult ferret (Mustela putorius furo). Retinal ganglion cells were labelled following injections of HRP into the lateral geniculate nucleus (LGN), superior colliculus (SC), or LGN+SC. The soma size distributions following LGN, SC, or LGN+SC injections were all unimodal showing considerable overlap between different cell classes. This was confirmed for alpha cells, identified on the basis of dendritic filling or from neurofibrillar-stained retinae. Analysis of the soma size and axon diameters of a population of heavily labelled retinal ganglion cells showed a significant correlation between the two. However, the overall distribution of intraretinal axon diameter was bimodal with an extended tail. Analysis of the ganglion cell distributions in the adult ferret indicates that beta cells comprise about 50.5–55%, gamma 42.5–47%, and alpha 2.5% of the ganglion cell population. This implies that the proportion of gamma, beta, alpha cells in both cat and ferret retina is highly conserved despite differences in visual specialization in the two species.

arises from the integration of signals from strongly oscillatory cells within the LGN. The model also predicts the incidence of 50-Hz oscillatory cells within the cortex. Oscillatory discharge around 30 Hz is explained in a second model by the presence of intrinsically oscillatory cells within cortical layer 5. Both models generate spike trains whose power spectra and mean firing rates are in close agreement with experimental observations of simple and complex cells. Considered together, the two models can largely account for the nature and incidence of oscillatory discharge in the cat's visual cortex. The validity of these models is consistent with the possibility that oscillations are generated independently of intracortical interactions. Because these models rely on intrinsic stimulus-independent oscillators within the retina and cortex, the results further suggest that oscillatory activity within the cortex is not necessarily associated with the processing of high-order visual information.

Recent work has revealed multiple pathways for cross-orientation suppression in cat and human vision. In particular, ipsiocular and interocular pathways appear to assert their influence before binocular summation in human but have different (1) spatial tuning, (2) temporal dependencies, and (3) adaptation after-effects. Here we use mask components that fall outside the excitatory passband of the detecting mechanism to investigate the rules for pooling multiple mask components within these pathways. We measured psychophysical contrast masking functions for vertical 1 cycle/deg sine-wave gratings in the presence of left or right oblique (±45 deg) 3 cycles/deg mask gratings with contrast C%, or a plaid made from their sum, where each component (i) had contrast 0.5Ci%. Masks and targets were presented to two eyes (binocular), one eye (monoptic), or different eyes (dichoptic). Binocular-masking functions superimposed when plotted against C, but in the monoptic and dichoptic conditions, the grating produced slightly more suppression than the plaid when Ci ≥ 16%. We tested contrast gain control models involving two types of contrast combination on the denominator: (1) spatial pooling of the mask after a local nonlinearity (to calculate either root mean square contrast or energy) and (2) “linear suppression” (Holmes & Meese, 2004, Journal of Vision4, 1080–1089), involving the linear sum of the mask component contrasts. Monoptic and dichoptic masking were typically better fit by the spatial pooling models, but binocular masking was not: it demanded strict linear summation of the Michelson contrast across mask orientation. Another scheme, in which suppressive pooling followed compressive contrast responses to the mask components (e.g., oriented cortical cells), was ruled out by all of our data. We conclude that the different processes that underlie monoptic and dichoptic masking use the same type of contrast pooling within their respective suppressive fields, but the effects do not sum to predict the binocular case.

The work of Mircea Steriade demonstrated that the neocortex could synchronize large regions of the thalamus within 10−100 msec. Unlike the synchrony generated by the cortex, the retinal afferents synchronize a restricted group of neighboring thalamic neurons with <1-msec precision. Here, we use a large sample (n = 372) of simultaneous recordings from neighboring neurons in the lateral geniculate nucleus (LGN) to illustrate the high specificity of the synchrony generated by retinal afferents and its dependency on sensory stimulation. First, we demonstrate that cells sharing a retinal afferent show a balanced receptive field diversity: although slight receptive field mismatches are common, the largest mismatches in a specific property (e.g. receptive field size) are restricted to cells that are precisely matched in other properties (e.g. receptive field overlap). Second, we show that these receptive field mismatches are functionally important and can lead to a 5-fold variation in the percentage of synchronous spikes driven by the shared retinal afferent under different stimulus conditions. Based on these and other findings, we speculate that the precise synchronous firing of cells sharing a retinal afferent might serve to amplify local stimuli that might be too brief and small to generate a large number of thalamic spikes.

This study examines the temporal properties of geniculocortical and corticogeniculate (CG) pathways that link the lateral geniculate nucleus (LGN) and primary visual cortex in the ferret. Using electrical stimulation in the LGN to evoke action potentials in geniculocortical and CG axons, results show that conduction latencies are significantly faster in geniculocortical neurons than in CG neurons. Within each pathway, axonal latency and visual physiology support the view of sub-classes of neurons. By examining the timing of visual responses and the latency of CG feedback, estimates indicate that visual information can reach the cortex and return to the LGN as early as 60 msec following the onset of a visual stimulus. These findings place constraints on the functional role of corticogeniculate feedback for visual processing.

Recurrent projections comprise a universal feature of cerebral organization. Here, we show that the corticofugal projections from the striate cortex (V1) to the lateral geniculate nucleus (LGN) robustly and multiplicatively enhance the responses of parvocellular neurons, stimulated by gratings restricted to the classical receptive field and modulated in luminance, by over two-fold in a contrast-independent manner at all but the lowest contrasts. In the equiluminant plane, wherein stimuli are modulated in chromaticity with luminance held constant, such enhancement is strongly contrast dependent. These projections also robustly enhance the responses of magnocellular neurons but contrast independently only at high contrasts. Thus, these results have broad functional significance at both network and neuronal levels by providing the experimental basis and quantitative constraints for a wide range of models on recurrent projections and the control of contrast gain.

Thalamic relay cells fire in two distinct modes, burst or tonic, and the operative mode is dictated by the inactivation state of low-threshold, voltage-gated, transient (or T-type) Ca2+ channels. Tonic firing is seen when the T channels are inactivated via membrane depolarization, and burst firing is seen when the T channels are activated from a hyperpolarized state. These response modes have very different effects on the relay of information to the cortex. It had been thought that only tonic firing is seen in the awake, alert animal, but recent evidence from several species suggests that bursting may also occur. We have begun to explore this issue in macaque monkeys by recording from thalamic relay cells of unanesthetized, behaving animals. In the lateral geniculate nucleus, the thalamic relay for retinal information, we found that tonic mode dominated responses both during alert behavior as well as during sleep. We nonetheless found burst firing present during the vigilant, waking state. There was, however, considerably more burst mode firing during sleep than wakefulness. Surprisingly, we did not find the bursting during sleep to be rhythmic. We also recorded from relay cells of the somatosensory thalamus. Interestingly, not only did these somatosensory neurons exhibit much more burst mode activity than did geniculate cells, but bursting during sleep was highly rhythmic. It thus appears that the level and nature of relay cell bursting may not be constant across all thalamic nuclei.

Signals relayed through the magnocellular layers of the LGN travel on axons with faster conduction speeds than those relayed through the parvocellular layers. As a result, magnocellular signals might reach cerebral cortex appreciably before parvocellular signals. The relative speed of these two channels cannot be accurately predicted based solely on axon conduction speeds, however. Other factors, such as different degrees of convergence in the magnocellular and parvocellular channels and the retinal circuits that feed them, can affect the time it takes for magnocellular and parvocellular signals to activate cortical neurons. We have investigated the relative timing of visual responses mediated by the magnocellular and parvocellular channels. We recorded individually from 78 magnocellular and 80 parvocellular neurons in the LGN of two anesthetized monkeys. Visual response latencies were measured for small spots of light of various intensities. Over a wide range of stimulus intensities the fastest magnocellular response latencies preceded the fastest parvocellular response latencies by about 10 ms. Because parvocellular neurons are far more numerous than magnocellular neurons, convergence in cortex could reduce the magnocellular advantage by allowing parvocellular signals to generate detectable responses sooner than expected based on the responses of individual parvocellular neurons. An analysis based on a simple model using neurophysiological data collected from the LGN shows that convergence in cortex could eliminate or reverse the magnocellular advantage. This observation calls into question inferences that have been made about ordinal relationships of neurons based on timing of responses.

Reverse correlation was used in conjunction with ternary white noise to estimate the first-order spatiotemporal receptive-field structure of LGN cells in the anesthetized, paralyzed cat. Based on a singular-value decomposition of these data, we conclude that most LGN cells are approximately space–time separable. An analysis of the timecourses of the first singular values revealed a strongly bimodal but continuous distribution of rise times and waveforms. The two modes represented cells generally associated with the lagged and nonlagged classes of Mastronarde (1987a,b), and this was confirmed by their responses to step and sine-modulated spots in their field centers. The intermediate cells, rather than appearing to constitute a separate group, smoothly filled the region between the obviously lagged and nonlagged cells in every respect. These conclusions are limited to X-cells although the data from a much smaller population of Y-cells conform to the same scheme. We conclude that lagged and nonlagged cells represent the modes of a continuous and very broad distribution of temporal responses in the cat LGN.