Y. Hata

Mechanisms underlying orientation selectivity of neurons in the primary visual cortex of the macaque.

1. Effects of blocking intracortical inhibition by microiontophoretic administration of bicuculline methiodide (BMI), a selective antagonist for GABAA receptors, on orientation selectivity of 109 neurones were studied in the primary visual cortex (V1) of anaesthetized and paralysed monkeys. 2. The averaged orientation tuning of visual responses of cells was poor in cytochrome oxidaserich blobs of layer II/III and in layer IVc beta, moderate in layers IVb, IVc alpha and V, and sharp in the interblob region of layer II/III and in layers IVa and VI. 3. Iontophoretic administration of BMI reduced the sharpness of orientation tuning of cells to a varying extent in each layer. In most cells, furthermore, the originally ineffective stimuli induced visual responses during the BMI administration, suggesting that excitatory inputs evoked by the non‐optimally oriented stimuli were masked by GABAergic inhibition. Nevertheless, the maximal facilitation was observed in the response to the optimally or near‐optimally oriented stimuli. 4. There was a difference in such an effect of BMI among layers. Orientation selectivity of cells in interblobs in layer II/III and in layer IVb was sensitive to BMI whereas that of cells in layer VI was relatively insensitive to BMI, suggesting a larger contribution of excitatory mechanisms to the orientation selectivity in this layer. 5. In the orientation‐selective cells, an analysis of the magnitude of excitation and inhibition evoked by stimuli at various orientations suggests that both inputs tune around the optimal orientation and their magnitudes are almost proportional to each other except at the optimal orientation. This analysis also indicates that the orientation tuning of inhibition had a less prominent peak around the optimal orientation than that of excitation. This dominance of excitation over inhibition around the optimal orientation may function to accentuate the response to the optimally oriented stimulus. 6. These results suggest that, in the monkey V1, the orientation selectivity of cells is largely dependent on the orientation‐biased excitatory and inhibitory inputs which have a broader tuning profile, covering from the optimal to null‐orientation, than that observed in extracellularly recorded responses at the control level.

Horizontal interactions between visual cortical neurones studied by cross-correlation analysis in the cat.

1. To explore the functional significance of horizontal neural connections in the extent of a ‘hypercolumn’ of the cat visual cortex, we carried out cross‐correlation analysis of spike trains recorded simultaneously from a pair of neurones separated horizontally by less than 1 mm. 2. Significantly correlated firings, which were found in sixty‐eight pairs of cells among 327 pairs analysed, were classified into three types on the basis of their functional implications: (1) excitatory interactions, (2) inhibitory interactions and (3) common inputs to both neurones of a pair from other sources. 3. Of these three types, common inputs were encountered most frequently. Excitatory interactions were always accompanied by common inputs. Inhibitory interactions were observed least frequently. 4. The proportion of cell pairs with correlated firings was high in pairs with a horizontal separation of less than 200 microns and decreased markedly with a horizontal separation of more than 400 microns. 5. Regarding laminar locations of cells, common inputs and excitatory interactions were often observed in layers II + III and V, whereas laminar bias was not seen in inhibitory interactions. 6. With respect to difference in orientation preference between two cells, all the three types of correlations were observed, mostly in cell pairs with a difference of less than 45 deg. In particular, common inputs and excitatory interactions were often seen in cell pairs with matched orientation preferences, but inhibitory interactions were found mostly in those with slightly different orientation preferences. In addition, common inputs and excitatory interactions tended to be found between cells with the same eye preference. 7. These results suggest that horizontal functional interactions exist mainly in a range of up to 400 microns as far as the extent of a hypercolumn of the visual cortex is concerned, and these interactions operate effectively between cortical cells with similar receptive field properties except for inhibitory interactions.

Selective Pruning of More Active Afferents When Cat Visual Cortex Is Pharmacologically Inhibited

Activity-dependent competition is thought to guide the normal development of specific patterns of neural connections. Such competition generally favors more active inputs, making them larger and stronger, while less active inputs become smaller and weaker. We pharmacologically inhibited the activity of visual cortical cells and measured the three-dimensional structure of inputs serving the two eyes when one eye was occluded. The more active inputs serving the open eye actually became smaller than the deprived inputs from the occluded eye, which were similar to those in normal animals. These findings demonstrate in vivo that it is not the amount of afferent activity but the correlation between cortical and afferent activity that regulates the growth or retraction of these inputs.

Actions of excitatory amino acid antagonists on geniculo-cortical transmission in the cat's visual cortex

SummaryTo test the possibility that glutamate and aspartate are transmitters at geniculo-cortical synapses and to elucidate which type of receptors for the excitatory amino acids (EAA) operate at these synapses, we studied effects of microionophoretic administration of EAA antagonists on the responses of visual cortical neurons to afferent electrical and visual stimulation in the cat. The antagonists used were kynurenate, a non-selective antagonist for all classes of EAA receptors and 2-amino-5-phos-phonovalerate (APV), a selective antagonist for N-methyl-D-aspartate (NMDA)-preferring receptors. The administration of kynurenate suppressed responses elicited by electrical stimulation of the dorsal lateral geniculate nucleus (LGN) and optic chiasm (OX) of 65% of the cells tested. This suppression was more marked for the short-latency responses which were evoked monosynaptically from the LGN, than for the longer-latency responses. In contrast with the effectiveness of kynurenate, APV failed to suppress electrically and visually elicited responses in 66% of the cortical cells. Such differences between kynurenate and APV were particularly prominent in layers IV and VI, which receive direct inputs from the LGN, but were less marked or were not recognizable in layers II + III and V. These results support previous suggestions that EAAs may be excitatory transmitters in the cerebral cortex, at least at geniculo-cortical synapses, and indicate further that EAA receptors of the “non-NMDA type” may be involved in afferent synaptic transmission in the cats visual cortex.

Release of glutamate and aspartate from the visual cortex of the cat following activation of afferent pathways.

SummaryTo test the possibility that glutamate (Glu) and aspartate (Asp) are transmitters at geniculo-cortical synapses in the visual cortex of the cat, we studied the release of amino acids from the striate cortex consequent upon visual and electrical stimulation of the dorsal lateral geniculate nucleus (LGN) and of the optic tract, using push-pull cannulae. We perfused a discrete region that included layer IV of the cortex with an artificial cerebrospinal fluid (aCSF) and analysed the amino acid content of these perfusates by high-performance liquid chromatography (HPLC). Significant increases only of Glu and Asp were obtained among all 17 amino acids measured, except for gamma-aminobutyric acid (GABA), during electrical stimulation of the afferent pathways. Visual stimulation by stroboscopic diffuse flashes of light increased the level of Glu released, but did not change that of Asp significantly. The level of GABA released did not change during diffuse flash stimulation, suggesting that the increase in Glu was not derived from cortical neurons. The increases in release of Glu/Asp were not seen when the perfusion medium was replaced with a Ca2+-free, high-Mg2+-containing solution. The basal (resting) release of Glu/Asp in the absence of stimulation also was decreased during perfusion with Ca2+-free/high-Mg2+ solutions. Intraocular injections of a sodium channel blocker, tetrodotoxin (TTX), resulted in a remarkable decrease in the basal release of Glu. These results suggest that Glu is released as an excitatory synaptic transmitter at least from terminals of geniculo-cortical afferents and Asp from axons of a certain type of visual cortical neuron.

Less Segregated Processing of Visual Information in V2 than in V1 of the Monkey Visual Cortex

To test the possibility of cross‐talk between parallel pathways dealing with different aspects of visual information, such as orientation, direction of motion and colour in cortical area V2, we quantitatively analysed visual responses of 121 V2 cells recorded from anaesthetized and paralysed macaques and compared them with those of 147 V1 cells. A selectivity index of visual responses was calculated for each neuron, which was then classified as selective or not to a particular attribute of visual stimuli. Twenty‐one percent of the V2 neurons had dual selectivity to both colour and direction of stimulus motion (C&D cells). In V1, only 5% of the cells were C&D cells. Thus, the proportion of C&D cells significantly increased from V1 to V2. We also carried out cross‐correlation analysis of spike trains recorded simultaneously from pairs of V2 neurons or pairs of V1 neurons. In V2, correlated firings could be observed between cells with completely different optimal orientation, such as orthogonal, while it was never observed in V1. The cross‐correlation analysis further indicated that functional interactions in V2 were more widespread than those in V1. These results suggest that neurons which have different functional properties become less segregated, and that functional interactions become more widespread in V2 than in V1.

Brain‐derived neurotrophic factor induces long‐lasting potentiation of synaptic transmission in visual cortex in vivo in young rats, but not in the adult

Brain‐derived neurotrophic factor (BDNF) rapidly enhances excitatory synaptic transmission in cortical slices. To date, however, a question of how long such an action persists remains unanswered as it is hard to record synaptic responses longer than several hours in slice preparations. To address this question and to investigate possible age‐dependency of the action, we analysed effects of a brief application of BDNF and nerve growth factor (NGF) on field potentials of visual cortex in rats of postnatal days 13–17 and 19–24 and in the adulthood for 10–24u2003h. Evoked potentials to stimulation of the lateral geniculate nucleus were recorded simultaneously from two cortical sites into which the neurotrophin and control solution were injected. An application of BDNF induced a slowly developing increase in the field potential amplitude in young rats. The amplitude attained a plateau level 3–4u2003h after the infusion; 139u2003±u200326% (meanu2003±u2003SD) and 132u2003±u200321% of the baseline in the rats at P13‐17 and P19‐24, respectively. This potentiation remained stable from 4 to 8u2003h, then gradually decreased to the baseline 15–16u2003h after the infusion. NGF applied in the same way did not induce potentiation. An inhibitor of BDNF receptors blocked the potentiation when it was applied immediately after the BDNF application, but was not effective about 2u2003h later. In the adults, BDNF did not potentiate field potentials. These results indicate that BDNF induces synaptic potentiation lasting for several hours only in the developing cortex through processes downstream of receptor activation.

Imaging of calcineurin activated by long-term depression-inducing synaptic inputs in living neurons of rat visual cortex

Long‐term depression (LTD) of synaptic transmission is induced by low‐frequency stimulation (LFS) of afferents lasting for a long time, typically for 10–15u2003min, in neocortical and hippocampal slices. It is suggested that calcineurin, Ca2+/calmodulin‐dependent protein phosphatase, plays a role in the induction of LTD, based on the results that pharmacological or genetic manipulation of calcineurin activity interfered in its induction. However, questions as to why it takes so long to induce LTD and in which compartment of neurons calcineurin is activated remain unanswered. With a fluorescent indicator for calcineurin activity, we visualized the spatiotemporal pattern of its activation in living neurons in layer II/III of visual cortical slices of rats during the LFS of layer IV that induced LTD of synaptic responses. During LFS, the fluorescence intensity gradually increased with a latency of a few minutes in dendrites and soma of neurons, and remained increased during the whole observation period (10–25u2003min) after LFS. The onset latency of the increase in the soma was slower than that in the distal dendritic region. The LFS‐induced rise in fluorescence was not observed in neurons which were loaded with inhibitors of calcineurin, indicating that the intensity of fluorescence reflects calcineurin activity. Control stimulation at 0.05u2003Hz and θ‐burst stimulation did not significantly change the intensity of fluorescence. Only LFS‐type inputs effectively activate calcineurin in postsynaptic neurons in an augmenting manner, and such a time‐consuming activation of calcineurin may be a reason why long‐lasting LFS is necessary for the induction of LTD.

Activity-dependent change in the protein level of brain-derived neurotrophic factor but no change in other neurotrophins in the visual cortex of young and adult ferrets

Neurotrophins are suggested to play a role in activity-dependent plasticity of visual cortex during the critical period of postnatal development. Thus, the concentration of neurotrophins in the cortex is expected to change with development and/or with alteration in neuronal activities. To test this, we measured protein levels of nerve growth factor, brain-derived neurotrophic factor, neurotrophin-3 and neurotrophin-4/5 in visual cortex of young (postnatal day 38-46, at the peak of the critical period) and adult ferrets with two-site enzyme-immunoassay systems. Measurements were carried out also in somatosensory cortex, hippocampus and cerebellum as control. With development the level of brain-derived neurotrophic factor did not significantly change, while those of the other neurotrophins changed in the visual cortex. A blockade of visual inputs for 24 h by an injection of tetrodotoxin into both eyes significantly decreased brain-derived neurotrophic factor protein level in the visual cortex, but not in the other regions in both young and adult ferrets. On the other hand, no significant decrease was seen in the protein level of the other neurotrophins in the visual cortex of young and adult ferrets. A monocular injection of tetrodotoxin in young ferrets resulted in the reduction of brain-derived neurotrophic factor by approximately half that by binocular injection. The degree of the decrease in the contralateral cortex to the injected eye was significantly larger than that in the ipsilateral cortex, reflecting that the contralateral eye is dominantly represented in the cortex in ferrets. Blockade of cortical neuronal activities by a GABA(A) receptor agonist led to a remarkable reduction of brain-derived neurotrophic factor protein in the visual cortex. These results suggest that the level of brain-derived neurotrophic factor protein in visual cortex is regulated by activities of cortical neurons.

Role of NMDA receptors in the propagation of excitation in rat visual cortex as studied by optical imaging

To examine the role of the N-methyl-D-aspartate (NMDA) type of glutamate receptors in the propagation of information in visual cortex, optical imaging with high spatial and temporal resolution of neuronal activity was used in cortical slices of rats. Single-shock stimulation of the white matter elicited a vertical propagation of excitation toward the cortical surface simultaneously with a horizontal spread of excitation in lower layers. The horizontal spread in upper layers occurred subsequent to the vertical spread reaching these layers. The results from perfusion of Ca2+-free medium and application of an antagonist of non-NMDA receptors indicated that this intracortical propagation of signals is due mostly, if not exclusively, to the postsynaptic excitation of cortical neurons. Blockade of NMDA receptors attenuated the rising and peak phases of the upper horizontal spread, but did not affect those of the lower horizontal or vertical propagation of excitation. Perfusion with Mg2+-free solution enhanced the upper horizontal spread, but in most cases did not significantly change the spread of excitation in the other pathways. These results indicate that NMDA receptors are involved in the flow of information in the upper layers of visual cortex, and further suggest that this propagation of activity is mediated mainly by horizontal connections intrinsic to the upper layers.