Loading...
Please wait, while we are loading the content...
Subcolumnar structures in local inputs of pyramidal neurons onto corticothalamic neurons in rat barrel cortex
| Content Provider | Semantic Scholar |
|---|---|
| Author | Tanaka, Yasuhiro Tanaka, Yasuyo Konno, Michiteru Fujiyama, Fumino Kaneko, Takeshi |
| Copyright Year | 2010 |
| Abstract | s / Neuroscience Research 68S (2010) e55–e108 e73 ing in the neurons’ preferred direction was much higher, the activities were almost dominated by the component. These results suggest that the nonlinear interaction between the components that moved in the preferred and anti-preferred directions underlies the activities of the MT/MST neurons. doi:10.1016/j.neures.2010.07.086 O1-8-2-1 Temporal dynamics of orientation and spatial frequency tuning of neurons in the cat lateral geniculate nucleus Tomoyuki Naito 1 , Naofumi Suematsu 2, Hiromichi Sato 1 1 Grad Sch Med, Osaka Univ, Toyonaka 2 Sch Eng Sci, Osaka Univ, Toyonaka Orientation and spatial frequency selectivities are fundamental properties of neurons in the early visual system. In this study, we addressed temporal development of neural selectivity to orientation and spatial frequency of the stimulus in the cat lateral geniculate nucleus (LGN). In order to measure the temporal dynamics of both tunings, we used a recently developed subspace reverse correlation method. Additionally, we measured the spatiotemporal structure of the classical receptive field of the same neuron using reverse correlation technique with dynamic dense white noise stimuli. Most of LGN neurons exhibited clear orientation and spatial frequency tunings and both tunings evolved over the time course of response. For many cells, the optimal spatial frequency shifted from low to high, while the optimal orientation generally remained stable throughout the duration of response. These results are basically consistent with the results reported in the primary visual cortex (V1) of the cat. We also computed the prediction of temporal dynamics of both tunings from the measured classical receptive field. However, the temporal changes of predicted tuning were sometimes not conformable to those measured. Our results suggest that previously reported coarse-to-fine tunings in V1 already exist in LGN and that some nonlinear mechanisms may contribute to generating temporal dynamics of orientation and spatial frequency tunings in the LGN. doi:10.1016/j.neures.2010.07.087 O1-8-2-2 Subcolumnar structures in local inputs of pyramidal neurons onto corticothalamic neurons in rat barrel cortex Yasuhiro Tanaka 1 , Yasuyo Tanaka 1, Michiteru Konno 1, Fumino Fujiyama 1,2, Keiko Okamoto-Furuta 1, Takahiro Sonomura 3, Hiroshi Kameda 1, Hiroyuki Hioki 1, Takahiro Furuta 1, Kouichi C Nakamura 1, Takeshi Kaneko 1 1 Department Morphological Brain Science, Kyoto University, Kyoto 2 JST, CREST, Tokyo 3 Department Anatomy for Oral Sciences, Kagoshima University, Kagoshima The thalamocortical reciprocal circuit is one of principal input/output organizations in the mammalian neocortex and the local inputs of pyramidal neurons to corticothalamic projection neurons are considered to play a key role in the reciprocal circuit. In the present study, we first retrogradely visualized almost all the dendrites and cell bodies of layer (L) 6 corticothalamic neurons in rat barrel cortex by injection of a newly developed adenovirus vector into the ventrobasal thalamic complex. With the cortical slices containing retrogradely visualized corticothalamic neurons, we morphologically examined the local inputs of intracellularly labeled L2/3, L4, L5, L6 pyramidal neurons to the corticothalamic neurons. Of the examined pyramidal neurons, the L6 corticothalamic neurons, which were labeled intracellularly and retrogradely, supplied the most abundant inputs to the retrogradely labeled dendrites within a horizontal spread of a columnar size (within 200 m from the neurons). In contrast, the L6 putative corticocortical neurons, which were only intracellularly labeled, provided the richest inputs to the retrogradely labeled dendrites more than 200 m away from the neurons. Interestingly, the L4 pyramidal neurons were the most intense source of the local inputs to the retrogradely labeled dendrites within a restricted domain from the neurons (<50 m). This restricted L4-to-L6 connection was not formed on the apical dendrites but on the basal dendrites of the corticothalamic neurons. These results suggest that the contrasted organization between the vertical inputs of L4 pyramidal neurons and horizontal inputs of L6 corticocortical pyramidal neurons of corticothalamic neurons, although corticothalamic neurons make the most intense connection with each other within a cortical column. doi:10.1016/j.neures.2010.07.088 O1-8-2-3 Theory of correlations in neural network model of V1 with synaptic depression Yasuhiko Igarashi 1 , Masafumi Oizumi 1,2, Masato Okada 1,3 1 Graduate School of Frontier Sciences, The University of Tokyo 2 Research Fellow of the Japan Society for the Promotion of Science 3 RIKEN Brain Science Institute, Wako, Japan Neurophysiological experiments have shown that brief adaptation to a stimulus of fixed orientation not only changes tuning curves but also reduces mean and variability of correlations between neurons in the macaque primary visual cortex. The post-adaptation changes in neural correlations lead to stimulus dependent changes in the efficiency of a population code. We investigated a mechanism of the brief adaptation in primary visual cortex (V1) using a theoretical approach. A possible mechanism of these adaptation effects is synaptic depression, because the brief adaptation to a stimulus of fixed orientation leads to depression of feed forward synapses and intracortical synapses. Some researchers have shown that a V1 network model with asymmetric synaptic depression, that is, inhibitory synapses depress less than excitatory ones, causes experimentally observed postadaptation changes of tuning curves. However, whether the asymmetric synaptic depression causes the decrease of neural correlation, which is associated with the efficiency of a population code, remains unknown. We studied the effects of synaptic depression on the neural correlations by using a stochastic V1 network model with asymmetric synaptic depression. We constructed a theory of neural correlation in stochastic neural network model with synaptic depression. A result obtained from the present theory and computer simulation coincided with each other. Based on this theory, we will calculate the change of the neural correlation and the efficiency of population code in the V1 network model after brief adaptation. This approach allows us to evaluate the effects of synaptic depression on the efficiency of a population code and examine the functional consequences for brief adaptation. doi:10.1016/j.neures.2010.07.089 O1-8-2-4 Investigation of Transform Domain Reverse Correlation by simulation of model neurons with the position invariance in the visual cortex Toshiya Arai 1 , Izumi Ohzawa 1,2 1 Graduate School of Fronteir Biosciences, Osaka University, Osaka, Japan 2 CREST, Japan Science and Technology Agency, Tokyo, Japan Responses of neurons in the early visual cortex depend generally on the exact location of visual stimuli. On the other hand, neurons in higher order visual areas increasingly gain invariance to stimulus position. Such characteristics may prevent standard reverse correlation methods from revealing internal spatial organization of receptive fields in these neurons. Therefore, we have developed an extension of a reverse correlation technique, named Transform Domain Reverse Correlation (TDRC). It computes the spike-triggered average of stimuli after transforming them to spectral and curvature domains.To examine the ability of TDRC for obtaining the response profiles, simulations using model neurons were performed. Models of simple and complex cells in V1 were used. And inputs from complex cell models with different preferred orientations at different but neighboring locations were combined to form model V2 neurons. We also examined the applicability of the method to neurons with a spatial invariance property.V1 models were selective to straight components, and response profiles of the model for V2 neurons showed a preference to curved contours than straight lines, in both spectral and curvature domains. However, for the spatially invariant models, the spectral domain analysis failed to reveal any internal structure of the model, because a part of the model neuron was always responsive to local orientations in the stimuli. On the other hand, the curvature domain analysis correctly recovered the structure of the model. These results suggest that TDRC using curvature domain can reveal detailed curvature selectivity of actual neurons with position invariance. Furthermore, since the stimuli are not optimized for any specific neurons, the TDRC method can simultaneously obtain detailed structures of response profiles from multiple visual neurons, even for neurons with the position invariance. doi:10.1016/j.neures.2010.07.090 |
| File Format | PDF HTM / HTML |
| DOI | 10.1016/j.neures.2010.07.088 |
| Alternate Webpage(s) | https://api.elsevier.com/content/article/pii/S0168010210002592 |
| Alternate Webpage(s) | https://www.sciencedirect.com/science/article/pii/S0168010210002592?dgcid=api_sd_search-api-endpoint |
| Alternate Webpage(s) | https://doi.org/10.1016/j.neures.2010.07.088 |
| Volume Number | 68 |
| Journal | Neuroscience Research |
| Language | English |
| Access Restriction | Open |
| Content Type | Text |
| Resource Type | Article |
