Origin of orientation-selective EPSPs in simple cells of cat visual cortex

According to current theory, orientation selectivity in cortical simple cells is critically dependent on intracortical synaptic inhibition. In particular, it is thought that IPSPs evoked by stimuli of the nonpreferred orientation are required to prevent a neuron from responding to a broadly tuned excitatory input at any but the preferred orientation. Yet EPSPs recorded in simple cells are in themselves highly orientation-selective. How is this possible, when excitation arises primarily from relay cells of the lateral geniculate nucleus (LGN), which are largely insensitive to orientation? In this paper, the properties of EPSPs are compared with the predictions of a model of geniculate excitation of simple cells. The model, which is derived from the suggestions of Hubel and Wiesel (1962), relies on the now familiar arrangement of the receptive fields of presynaptic geniculate cells into lines parallel to the axis of orientation of the postsynaptic cell. Several properties of the EPSPs observed in simple cells, and the orientation tuning of simple cells observed in extracellular experiments, can be accounted for without resorting to intracortical inhibition.     

A detailed model of the primary visual pathway in the cat: comparison of afferent excitatory and intracortical inhibitory connection schemes for orientation selectivity

In order to arrive at a quantitative understanding of the dynamics of cortical neuronal networks, we simulated a detailed model of the primary visual pathway of the adult cat. This computer model comprises a 5 degrees x 5 degrees patch of the visual field at a retinal eccentricity of 4.5 degrees and includes 2048 ON- and OFF-center retinal beta-ganglion cells, 8192 geniculate X-cells, and 4096 simple cells in layer IV in area 17. The neurons are implemented as improved integrate-and-fire units. Cortical receptive fields are determined by the pattern of afferent convergence and by inhibitory intracortical connections. Orientation columns are implemented continuously with a realistic receptive field scatter and jitter in the preferred orientations. We first show that realistic ON-OFF-responses, orientation selectivity, velocity low-pass behaviour, null response, and responses to spot stimuli can be obtained with an appropriate alignment of geniculate neurons converging onto the cortical simple cell (Hubel and Wiesel, 1962) and in the absence of intracortical connections. However, the average receptive field elongation (length to width) required to obtain realistic orientation tuning is 4.0, much higher than the average observed elongation. This strongly argues for additional intracortical mechanisms sharpening orientation selectivity. In the second stage, we simulated five different inhibitory intracortical connection patterns (random, local, sparse-local, circular, and cross-orientation) in order to investigate the connection specificity necessary to achieve orientation tuning. Inhibitory connection schemes were superimposed onto Hubel and Wiesel-type receptive fields with an elongation of 1.78. Cross-orientation inhibition gave rise to different horizontal and vertical orientation tuning curves, something not observed experimentally. A combination of two inhibitory schemes, local and circular inhibition (a weak form of cross-orientation inhibition), is in good agreement with observed receptive field properties. The specificity required to establish these connections during development is low. We propose that orientation selectivity is caused by at least three different mechanisms ("eclectic" model): a weak afferent geniculate bias, broadly tuned cross-orientation inhibition, and some iso-orientation inhibition. The most surprising finding is that an isotropic connection scheme, circular inhibition, in which a cell inhibits all of its postsynaptic target cells at a distance of approximately 500 microns, enhances orientation tuning and leads to a significant directional bias. This is caused by the embedding of cortical cells within a columnar structure and does not depend on our specific assumptions.(ABSTRACT TRUNCATED AT 400 WORDS)     

Relationship between preferred orientation and receptive field position of neurons in cat striate cortex

It has been known for two decades that neurons in mammalian visual cortex respond selectively to stimuli falling on the retina at a particular angular orientation (Hubel and Wiesel, '62). Recent evidence suggests that most cat retinal ganglion cells (Levick and Thibos, '82) and relay cells (Vidyasagar and Urbas, '82) in the cat's dorsal lateral geniculate nucleus are also orientation selective. In the retina there is a systematic relationship between receptive field position (polar angle) and preferred orientation. Outside of the area centralis, most retinal ganglion cells have oriented dendritic fields (Leventhal and Schall, '83) and respond best to stimuli oriented radially, i.e., oriented parallel to the line connecting their receptive fields to the area centralis (Levick an Thibos, '82). This relationship is strongest close to the horizontal meridian (the visual streak) of the retina (Leventhal and Schall, '83). To determine if a relationship between preferred orientation and polar angle exists in visual cortex, the preferred orientations and receptive field positions of 768 striate cortical neurons were studied. As in the retina, a systematic relationship exists between preferred orientation and visual field position in area 17. In parts of striate cortex 15--80 degrees from the area centralis projection there is a strong tendency for cells to respond best to lines oriented radially. In regions 4--15 degrees from the area centralis projection this relationship appears weaker. In regions subserving the central 4 degrees of visual angle no such relationship exists. Throughout area 17 the relationship between preferred orientation and polar angle is strongest in regions subserving the horizontal meridian.(ABSTRACT TRUNCATED AT 250 WORDS)     

TrkB-like immunoreactivity is present on geniculocortical afferents in layer IV of kitten primary visual cortex.

Exogenous administration of the neurotrophins brain-derived neurotrophic factor (BDNF) or neurotrophin-4/5 (NT-4/5), or blockade of their endogenous actions, have been reported to affect the anatomic organization and physiological responses of neurons in developing mammalian primary visual cortex. Experimental alteration of levels of these neurotrophic factors can also influence the morphology of the geniculocortical afferents that project from the lateral geniculate nucleus (LGN) to primary visual cortex. BDNF and NT-4/5 are ligands of the TrkB tyrosine kinase receptor. Although multiple populations of cortical neurons express TrkB, it is not known whether geniculocortical afferents express this receptor on their axon branches in visual cortex. We have anatomically labeled geniculocortical afferents of postnatal day 40 kittens with the anterograde neuronal tracer Phaseolus vulgaris leucoagglutinin (PHA-L) and performed double-label immunofluorescence with a panel of anti-TrkB antibodies. Confocal microscopy and object-based colocalization analysis were used to measure levels of TrkB-like immunoreactivity (IR) on geniculocortical afferents in layer IV of primary visual cortex. By using a conservative analysis involving a comparison of measured colocalization with the amount of colocalization expected based on random overlap of TrkB puncta and PHA-L--labeled afferents, 3 of 5 anti-TrkB antibodies tested showed significant colocalization with the geniculocortical axons. Results for the other two antibodies were indeterminate. The indices obtained for colocalization of TrkB and geniculocortical afferents were also compared with the equivalent index obtained for GAD65, a protein that has a similar overall expression pattern to that of TrkB but is not expressed on geniculocortical axons. This analysis indicated that TrkB was present on geniculocortical axons for all five TrkB antibodies tested. TrkB-like IR was also observed on neuronal somata in the LGN. These results indicate that TrkB receptors on geniculocortical afferents are potential mediators of the actions of BDNF and NT-4/5 in developing visual cortex.     

Modeling LGN responses during free-viewing: a possible role of microscopic eye movements in the refinement of cortical orientation selectivity.

Neural activity appears to be essential for the normal development of the orientation-selective responses of cortical cells. It has been proposed that the correlated activity of LGN cells is a crucial component for shaping the receptive fields of cortical simple cells into adjacent, oriented subregions alternately receiving ON- and OFF-center excitatory geniculate inputs. After eye opening, the spatiotemporal structure of neural activity in the early stages of the visual pathway depends not only on the characteristics of the environment, but also on the way the environment is scanned. In this study, we use computational modeling to investigate how eye movements might affect the refinement of orientation tuning in the presence of a Hebbian scheme of synaptic plasticity. Visual input consisting of natural scenes scanned by varying types of eye movements was used to activate a spatiotemporal model of LGN cells. In the presence of different types of movement, significantly different patterns of activity were found in the LGN. Specific patterns of correlation required for the development of segregated cortical receptive field subregions were observed in the case of micromovements, but were not seen in the case of saccades or static presentation of natural visual input. These results suggest an important role for the eye movements occurring during fixation in the refinement of orientation selectivity.     

Contribution of feedforward thalamic afferents and corticogeniculate feedback to the spatial summation area of macaque V1 and LGN

Neurons in the primary visual cortex (V1) respond best to oriented gratings of optimal size within their receptive field (RF) and are suppressed by larger gratings involving the nonclassical RF surround. A V1 neuron's optimal stimulus size is larger at lower stimulus contrast. A central question in visual neuroscience is what circuits generate the size tuning of V1 cells. We recently demonstrated that V1 horizontal connections integrate signals within a region of the RF center corresponding to the V1 neuron's optimal stimulus size at low contrast; extrastriate feedback connections to V1, instead, are longer range and can integrate signals from the most distant regions of the V1 cell's RF surround. Here, we have determined the contribution of geniculocortical feedforward and corticogeniculate feedback connections to the size-tuning of macaque V1 and lateral geniculate (LGN) neurons, respectively. Specifically, we have quantitatively compared the visuotopic extent of geniculate feedforward afferents to V1 with the size of the RF center and surround of neurons in the V1 input layers and the visuotopic extent of V1 feedback connections to the LGN with the RF size of cells in V1 layer 6, where these connections originate. We find geniculate feedforward connections to provide visuotopic information to V1 that is spatially coextensive with the V1 neuron's optimal stimulus size measured with high-contrast gratings. V1 feedback connections restrict their influence to an LGN region visuotopically coextensive with the size of the minimum response field (or classical RF) of V1 layer 6 cells and commensurate with the LGN region from which they receive feedforward connections.     

Expansion of the ipsilateral visual corticotectal projection in hamsters subjected to partial lesions of the visual cortex during infancy: Electrophysiological experiments

Single unit recording from cells in the superior colliculus ipsilateral to the damaged hemisphere in hamsters subjected to unilateral removal of a part of the posterior neocortex during infancy was combined with electrical stimulation of the cortical remnant and the visual cortex in the undamaged hemisphere. Cells activated by stimulation of the cortical remnant were recorded in all portions of the colliculus. No differences in percentages of driven cells or threshold current intensities were noted between electrode penetrations in which collicular neurons having receptive fields within the remaining visual cortical representation were recorded and tracks where units with receptive fields outside this region were isolated. In the medial part of the tectum ipsilateral to the damaged hemisphere cells driven by stimulation of either cortex were encountered. It was also demonstrated that stimulation of the ipsilateral cortical remnant and/or the contralateral cortex was capable of suppressing discharges normally elicited by optic chiasm or visual stimulation in a manner qualitatively similar to that observed for collicular cells in normal hamsters. The response properties of cells functionally influenced by the ipsilateral and/or contralateral corticles were not different from those of neurons which received no demonstrable cortical input. The receptive field characteristics of the sample of neurons recorded were, on the whole, quite similar to those of collicular neurons in hamsters subjected to lesions of the visual cortex as adults.     

Learning the invariance properties of complex cells from their responses to natural stimuli

Neurons in primary visual cortex are typically classified as either simple or complex. Whereas simple cells respond strongly to grating and bar stimuli displayed at a certain phase and visual field location, complex cell responses are insensitive to small translations of the stimulus within the receptive field [Hubel & Wiesel (1962) J. Physiol. (Lond.), 160, 106-154; Kjaer et al. (1997) J. Neurophysiol., 78, 3187-3197]. This constancy in the response to variations of the stimuli is commonly called invariance. Hubel and Wiesel's classical model of the primary visual cortex proposes a connectivity scheme which successfully describes simple and complex cell response properties. However, the question as to how this connectivity arises during normal development is left open. Based on their work and inspired by recent physiological findings we suggest a network model capable of learning from natural stimuli and developing receptive field properties which match those of cortical simple and complex cells. Stimuli are drawn from videos obtained by a camera mounted to a cat's head, so they should approximate the natural input to the cat's visual system. The network uses a competitive scheme to learn simple and complex cell response properties. Employing delayed signals to learn connections between simple and complex cells enables the model to utilize temporal properties of the input. We show that the temporal structure of the input gives rise to the emergence and refinement of complex cell receptive fields, whereas removing temporal continuity prevents this processes. This model lends a physiologically based explanation of the development of complex cell invariance response properties.     

Effects of monocular deprivation on geniculocortical synapses in the cat

In monocularly deprived (MD) cats, many cells in the lateral geniculate nucleus (LGN) but few cells in the visual cortex respond to input from the deprived eye, suggesting that the connections to visual cortex from the deprived geniculate laminae may have been disrupted. I have examined these connections in MD cats by using electron microscopic autoradiography of visual cortex after injections of tritiated lysine into single laminae of LGN. After injections into either deprived or experienced laminae, there was label over terminals that contained mitochondria and round synaptic vesicles and that made asymmetric contacts with dendritic profiles. However, the terminals of deprived afferents differed from those of experienced afferents: They were 25% smaller, contained 33% fewer mitochondria, were more likely to make synapses that were presynaptically convex (and thus, perhaps, immature), and synapsed onto smaller spines. These morphological changes were greater for afferents to upper layer IV than for afferents to lower layer IV. The geniculocortical synapses from deprived laminae were also reduced in number. To correct for variations in injection size and for a probable reduction in protein synthesis by cells in the deprived laminae, I computed the ratio of labeled synaptic terminals to labeled myelinated axons. Injections into the deprived laminae labeled 43% fewer synaptic terminals per labeled myelinated axon than did injections into the experienced lamina. The finding that the synaptic terminals of deprived afferents are both abnormal morphologically and fewer in number can help to explain the reduced effectiveness of the deprived eye in driving cortical cells.     

Contrast invariance of orientation tuning in the lateral geniculate nucleus of the feline visual system

Responses of most neurons in the primary visual cortex of mammals are markedly selective for stimulus orientation and their orientation tuning does not vary with changes in stimulus contrast. The basis of such contrast invariance of orientation tuning has been shown to be the higher variability in the response for low‐contrast stimuli. Neurons in the lateral geniculate nucleus (LGN), which provides the major visual input to the cortex, have also been shown to have higher variability in their response to low‐contrast stimuli. Parallel studies have also long established mild degrees of orientation selectivity in LGN and retinal cells. In our study, we show that contrast invariance of orientation tuning is already present in the LGN. In addition, we show that the variability of spike responses of LGN neurons increases at lower stimulus contrasts, especially for non‐preferred orientations. We suggest that such contrast‐ and orientation‐sensitive variability not only explains the contrast invariance observed in the LGN but can also underlie the contrast‐invariant orientation tuning seen at the level of the primary visual cortex.     

Ordinal position and afferent input of neurons in monkey striate cortex

From the extracellular recording of single units in the monkey striate cortex and electrical stimulation at two selected sites in the optic radiations it was possible to estimate 1) the ordinal position of striate neurons (i.e., whether they received a monosynaptic, disynaptic or polysynaptic input from the thalamus) and 2) the nature of the afferent input to these neurons (i.e., whether it came from the magnocellular or parvocellular subdivision of the lateral geniculate nucleus (LGN)). Based on receptive field properties six major classes of striate neuron were identified--three which lacked orientation specificity (the ON-center, the OFF-center, and the ON/OFF or nonoriented (N-0) receptive fields) and three with orientation specific responses (the S, the C, and the B categories of receptive field). Units lacking orientation specificity were concentrated in laminae 4A, 4C beta and 6 while, for the cells with orientation specificity, C cells were found in laminae 4B and 6, B cells in 2/3 and 5, and S cells predominantly in laminae 2/3, 4C alpha, and 5. The results of electrical stimulation indicated that cell-to-cell transmission time in the monkey striate cortex is 1.5 msec, and latency measures showed that cells with a monosynaptic drive from the thalamus were confined to laminae 4 and 6 while disynaptically driven cells were found principally in upper lamina 4 (4A and 4B). No cell class was identified exclusively with a given ordinal position and there were many types of potential first-order neurons. The conduction time from one stimulating electrode to the next in the optic radiation was used to identify the afferent input to each striate neuron. The input to color-coded neurones was found to come exclusively from parvocellular layers while the C cells and two subclasses of the S cell (S2 and S3) were driven predominantly by the magnocellular subdivision. For other cell types (those with ON-center, N-0, and S1 receptive fields) the input came from either type of LGN neuron. The laminar distribution of neurons receiving a direct input from the magnocellular and parvocellular streams is in accord with the results of anatomical studies into the site of termination of the LGN input. The cell types receiving these direct inputs vary in the two streams so that the parvocellular input terminates on cells with ON-center and N-0 receptive fields in lamina 4C beta while the magnocellular input goes to cells with S, ON-center, N-0, and C receptive fields in lamina 4C alpha and the lower part of 4B. Consideration is given to the influence of these results on models for neural processing in monkey striate cortex and a comparison is drawn with the results of similar studies in the cat.     

Organized arrangement of orientation-sensitive relay cells in the cat's dorsal lateral geniculate nucleus

We studied the physiological orientation biases of over 700 relay cells in the cat's dorsal lateral geniculate nucleus (LGNd). Relay cells were sampled at regular intervals along horizontally as well as vertically oriented electrode penetrations in a fashion analogous to that used previously in studies of visual cortex (Hubel and Wiesel, 1962). The strengths of the orientation biases and the distributions of the preferred orientations were determined for different classes of relay cells, relay cells in different layers of the LGNd, and relay cells subserving different parts of the visual field. We find that, at the population level, LGNd cells exhibit about the same degree of orientation bias as do the retinal ganglion cells providing their inputs (see also Soodak et al., 1987). Also, as in the retina (Levick and Thibos, 1982; Leventhal and Schall, 1983), most LGNd cells tend to prefer stimuli oriented radially, i.e., parallel to the line connecting their receptive fields to the area centralis projection. However, the radial bias in the LGNd is weaker than in the retina. Moreover, there is a relative overrepresentation of cells preferring tangentially oriented stimuli in the LGNd but not in the retina. As a result of the overrepresentation of cells preferring radial and tangential stimuli, the overall distribution of preferred orientations varies in regions of the LGNd subserving different parts of the visual field. Reconstructions of our electrode penetrations provide evidence that, unlike in the retina, cells having similar preferred orientations are clustered in the LGNd. This clustering is apparent for all cell types and in all parts of laminae A and A1. The tendency to cluster according to preferred orientation is evident for cells preferring radially, intermediately, and tangentially oriented stimuli and thus is not simply a reflection of the radial bias evident among retinal ganglion cells at the population level. It is already known that cells having inputs from different eyes, on-center, off-center, X-, Y-, W-type, and color-sensitive ganglion cells are distributed nonrandomly in the LGNd of cats and monkeys (for review, see Rodieck, 1979; Stone et al., 1979; Lennie, 1981; Stone, 1983). The finding that relay cells having similar preferred orientations are also distributed nonrandomly suggests that the initial sorting of virtually all properties segregated in visual cortex may begin in the LGNd.     

Visual cortical cell classification criteria: Reliability and equivalence of the quantitative dynamic — and static — field plotting procedures

There is in vision research a general unwillingness to classify or define visual cortical cell types, particularly new cell types, outside the classical simple/ complex dichotomy. Cells lacking clear-cut characteristics are, therefore, considered simple by some authors and complex by others. The present unsatisfactory state has largely arisen because of the absence of any rigorous, generally accepted, classification scheme of visual cortical neurons. Actually the present two classification schemes of visual cortical neurons, that is the Hubel and Wiesel and the Bishop schemes, are generally considered to be not comparable since the former is based on the cell qualitative static-field properties as revealed by hand-held stationary flashing stimuli, whereas the latter is based on the cell quantitative dynamic-field plotting properties as revealed by moving light stimuli. Since receptive fields lacking clear-cut characteristics of simple and complex cells have been observed in area 18 of the cat as well, all the receptive field types of this area have been classified either qualitatively or quantitatively using both the Hubel and Wiesel and the Bishop classifying procedures. It has been observed, at least as far as simple cells are concerned, that the two schemes are not antithetic but, on the contrary, equivalent if averaging procedures are taken into consideration.     

Morphology of single, physiologically identified retinogeniculate Y-cell axons in the cat following damage to visual cortex at birth

It has been reported previously that neurons in the dorsal lateral geniculate nucleus (LGN) of cats with neonatal damage to visual cortex (KVC cats) have receptive fields that are abnormally large and that the receptive fields of these neurons sometimes do not appear to conform to the normal retinotopic order in the LGN. A primary aim of this study was to determine if these physiological abnormalities are related to inappropriate patterns of retinogeniculate connections. We therefore have analyzed the terminal arbors of retinogeniculate axons in adult cats that had received a lesion of visual cortex (areas 17, 18, and 19) on the day of birth. Single retinogeniculate axons were characterized physiologically and injected intracellularly with horseradish peroxidase. Consistent with earlier reports that neonatal removal of visual cortex results in a retrograde loss of retinal X-cells, all of the retinogeniculate axons that we recorded were from Y-cells. While the visual responses of these Y-cell axons were normal, the morphology of their terminal arbors in the LGN was abnormal. Retinal Y-cell axons in KVC cats have terminal fields in the A laminae of the LGN that are as large or larger than those of normal Y-cells. However, since the LGN in KVC cats is severely degenerated, single Y-cell arbors occupy a proportional volume of the LGN that is 12 times greater than normal. Thus an early lesion of visual cortex produces a severe mismatch between retinogeniculate axon arbor size and target size. Also, despite the normal size of retinogeniculate axon arbors in KVC cats, the number and density of terminal boutons are greatly decreased. Thus our morphological results suggest that the unusually large receptive fields of LGN cells in KVC cats and the relative lack of retinotopic precision in the LGN are due, at least in part, to anomalies in the relative size and distribution of retinogeniculate axon arbors that develop after neonatal removal of visual cortex.     

Development of NMDA and non-NMDA receptor-mediated excitatory synaptic transmission in geniculocortical and corticocortical connections in organotypic coculture preparations

Development of N-methyl-D-aspartate (NMDA) and non-NMDA receptor-mediated excitatory synaptic transmission was studied in the visual cortex using organotypic slice cocultures. A slice of visual cortex (VC) dissected from newborn rats was cocultured with either a chunk of embryonic lateral geniculate nucleus (LGN) or another VC. During 7-38 days in vitro (DIV), geniculocortical monosynaptic excitatory postsynaptic potentials (EPSPs) were recorded from layer IV neurons in response to stimulation of the LGN in cocultures of the VC with the LGN. Similarly, corticocortical monosynaptic EPSPs were recorded from layers II/III and V/VI neurons in cocultures of two VCs when stimulating the partner VC. The initial slopes of the non-NMDA and NMDA receptor-mediated components of the EPSPs, which were dissociated pharmacologically, were assessed and compared among three different culture stages, early (7-11 DIV), middle (12-15 DIV) and late (17-38 DIV). With progression of the culture stage, the non-NMDA component tended to increase in both the geniculocortical and corticocortical connections. In contrast, the NMDA component exhibited distinct developmental changes. The NMDA component in layer IV neurons, which receive geniculate inputs, showed a transient increase in the middle stage. In the corticocortical connection, the magnitude of the NMDA component was large in the early stage and maintained through all culture stages in layer V/VI cells, whereas in layer II/III cells it decreased sharply by the late stage. Our results suggest that glutamatergic transmission in the visual cortex develops differently in the geniculocortical and corticocortical connections.     

Temporal interactions in the cat visual system. II. Suppressive and facilitatory effects in the lateral geniculate nucleus

Extracellular responses were recorded from single neurons in the lateral geniculate nucleus (LGN) of the cat during presentation of pairs of brief visual stimuli identical to those that produce orientation-selective paired-pulsed suppression in the visual cortex. LGN neurons also show paired-pulse suppression, but the suppression is not orientation selective, and it occurs only for short interstimulus intervals (ISIs; usually less than 200 msec). At longer ISIs, most LGN neurons show a period of facilitation. Thus, the paired-pulse suppression in the LGN cannot account for that seen in the visual cortex. Paired-pulse suppression in the LGN was found to be enhanced by stimulation of the receptive field surround. LGN neurons also showed a second type of suppression, termed "offset suppression," which consisted of a more long-lasting suppression of spontaneous activity following the offset of an excitatory visual stimulus. The suppression of spontaneous activity was accompanied by a reduction of the antidromic excitability, assessed by stimulating LGN axons within the cortex or optic radiation. Unlike paired-pulsed suppression, offset suppression was not enhanced by increased stimulation of the receptive field surround. Paired-pulse suppression and offset suppression are most likely due to different mechanisms because they have different time courses and depend differently on the spatial properties of the stimuli. Functionally, paired-pulse suppression may be related to the reduced visual sensitivity that accompanies eye movements, while offset suppression may serve to enhance temporal contrast.     

Orientation selectivity of synaptic potentials in neurons of cat primary visual cortex

Neurons of the visual cortex of the cat were penetrated with intracellular electrodes and postsynaptic potentials evoked by visual stimuli recorded. By alternately polarizing the cell with steady current injected through the recording electrode, IPSPs and EPSPs could be recorded and analyzed independently. Hyperpolarizing current suppressed IPSPs and enhanced EPSPs by moving the membrane potential toward the IPSP equilibrium potential. Depolarizing the cell toward the EPSP equilibrium potential enhanced IPSP. The responses to electrical stimulation of the LGN, where EPSPs and IPSPs could be distinguished easily by virtue of their characteristic latencies and shapes, were used to set the current injection to the appropriate level to view the two types of synaptic potential. EPSPs were found to be well oriented in that maximal depolarizing responses could be evoked at only one stimulus orientation; rotating the stimulus orientation in either direction produced a fall in the EPSP response. IPSPs were also well tuned to orientation, and invariably the preferred orientations of EPSPs and IPSPs in any one cell were identical. In addition, no systematic difference in the width of tuning of the two types of potential was seen. This result has been obtained from penetrations of over 30 cortical cells, including those with simple and complex receptive fields. It is concluded that orientation of cortical receptive fields is neither created nor sharpened by inhibition between neurons with different orientation preference. The function of inhibition evoked simultaneously with excitation by optimally oriented stimuli has yet to be determined, though it is likely to be the mechanism underlying other cortical receptive field properties, such as direction selectivity and end-stopping.     

The afferent connections and laminar distribution of cells in the cat striate cortex.

A laminar distribution of different functional cell types in the striate cortex of the cat is drawn up from the visual responses of single cells recorded in 64 electrode penetrations in 38 cats. In summary, S cells were found to be concentrated in laminae 4 and 6; SH cells in laminae 2, 3 and 4; C cells in laminae 5 and lower 3; B cells in laminae 3 and upper 5 and cells with non-oriented receptive fields in lamina 4. In addition, the nature of afferent innervation to striate neurons was derived from the latency of the orthodromic response to electrical stimulation in the optic chiasm and optic radiations in 19 cats. An analysis of latency values allowed the afferent innervation to a cell to be classed as belonging either to fast or slow conducting streams in the population of dLGN axons and also permitted a decision to be made on whether or not the afferent path passed directly to the cell. Direct afferent innervation from the dLGN was not found to be confined to a single class of striate neuron. Instead, examples of cells with S, SH, C, B and non-oriented receptive fields all had orthodromic latencies that met the requirement for direct innervation. Instances of cells with orthodromic latencies suggestive of indirect innervation were also found for most receptive field classes but these cells were encountered less frequently than those with a direct afferent input. It is argued that a variety of different cell types may act as first order neurons in the striate cortex and that cells occurring at later stages in the sequence of cortical processing may have been incompletely studied because they are more difficult to stimulate either visually or electrically.     

Functional organization of primary visual cortex in the mink (Mustela vison), and a comparison with the cat

The functional organization of geniculocortical afferents and the visual responses of neurons in primary visual cortex (area 17) were studied in barbiturate-anesthetized, paralyzed minks and cats. Responses of the afferents were studied after silencing intrinsic cortical activity with injections of kainic acid. In both species, afferents were segregated into patches on the basis of eye of origin. In the mink, but not in the cat, there was a further segregation on the basis of center type, with on- and off-center afferents terminating in alternating, partially overlapping patches. The visual responses of cortical neurons in the mink showed many similarities to those in the cat. Nearly all units were orientation-selective, and there was a columnar organization for preferred orientation. Many units were selective for one direction of movement. Within the binocular segment of cortex, although many units could be driven from either eye, there was a marked bias toward the contralateral eye compared to the cat. There was a columnar system for ocular dominance, but contralateral eye columns were wider than ipsilateral. In both species, a quantitative study was made of the responses of cortical neurons to stationary, flashing slits as a function of position in the receptive field. In the mink, and less clearly in the cat, units could be identified as simple or complex on the basis of the spatial separation or overlap of "on" and "off" discharge zones. In both species, simple cells were found most commonly in layers IV and VI, while layer V contained the greatest proportion of complex cells. The relative strengths of the on and off discharges of single cells were also measured. In the mink, many units gave better overall responses to the on or off phase of the stimulus, and 15% showed a strong (greater than 9:1) preference for one or the other, compared to 4% in the cat. In the mink, units with a common preference for the on or off phase of stationary stimuli were arranged in columnar aggregates, a feature of cortical organization that was not found in the cat. These columns probably result from the partial segregation of on-center and off-center geniculate afferents within layers IV and VI of the mink's cortex. On-dominated columns were, however, wider or more numerous than off-dominated columns.     

Topographic organization of the orientation column system in the striate cortex of the tree shrew (Tupaia glis). I. Microelectrode recording

Microelectrode recordings were made in the binocular portion of the tree shrew striate cortex to determine how orientation selective cells are distributed topographically in area 17 of this species. Seventy-five percent of the cells sampled were activated well by elongated visual stimuli and were quite selective for stimulus orientation. Ninety-five percent of the orientation-selective cells had orientation tuning ranges (Wilson and Sherman, '76) between ± 5° and ± 40° from their optimal orientation. Orientation-selective cells with the same or similar optimal orientations were distributed in cortex in a columnar manner (Hubel and Wiesel, '62), as determined from electrode penetrations nearly normal to the cortical surface. Penetrations parallel to the cortical surface revealed a highly ordered representation of optimal stimulus orientation, generally characterized by sequential changes in optimal orientation with electrode movement across the striate cortex. In addition, relatively consistent differences were observed in the rates and patterns of orientation shift on these penetrations depending on the direction of electrode movement across the cortex. Penetrations parallel to the 17–18 border yielded moderate-to-high rates of orientation change (mean slope = 434°/mm), with the changes genearlly progressing through a complete clockwise or counterclockwise cycle of 180° or more before a major reversal in the direction of orientation shift was encountered. In contrast, penetrations perpendicular to the border yielded low-to-moderate slopes (mean slope = 239°/mm). On these penetrations a more limited range of optimal orientations (<180°) was usually encountered, due to frequent reversals in the direction of orientation shift. Also, extended regions (100–200 μm long) of constant optimal orientation were observed in these penetrations. The different patterns of orientation change found on these orthogonal penetrations across the striate cortex indicate that the orientation column system in this species is anisotropically organized with respect to the 17–18 border. Further, the regions of constant optimal orientation frequently encountered on penetrations perpendicular to the 17–18 border suggest that the anisotropy is subserved by a system of elongated zones of iso-orientation arranged approximately perpendicular to the 17–18 border.