However, there was a considerable variability between different neuron types in the strength and frequency of lateral inhibition. Strong lateral inhibition was mostly found in neurons locked to the first half of the respiration cycle. In contrast, weak inhibition arriving from many surrounding glomeruli was relatively more common in neurons locked to the late phase of the respiration cycle.
Proximal neurons could receive different levels of inhibition. In many sensory systems, sensory information is processed locally and then transmitted to higher cortical regions in parallel by different cell types. This notion has been supported by recent findings showing that mitral and tufted cells differ in some of their odor response properties and project to distinct cortical regions 1 , 2 , 3 , 4 , 5. However, it is still unclear how neurons that share similar inputs vary in their output information, and which mechanisms govern this process.
Recent work has shown that back projections can decorrelate mitral cell odor responses, but less in the case of tufted cells 6. Lateral inhibition is a prominent motif in many sensory systems, in which neurons receive inhibition from adjacent cells at the same processing level. To better understand how this takes place, the frequency, strength and spatial distribution of lateral inhibition and the way it differentially affects different cell types need to be determined.
Despite extensive research in this field, these parameters are not fully known. It also remains unclear whether all projection neurons receive lateral inhibition and whether there are differences in the level of lateral inhibition between cell types.
Several studies have attempted to characterize lateral inhibition in-vivo in anesthetized animals using odor stimulation combined with imaging and electrophysiology techniques. These have led at times to contradictory conclusions regarding strength, frequency and spatial distribution. However, odor typically excites several glomeruli and may also directly inhibit some glomeruli due to receptor-odorant interactions or local circuits; hence, using odors to elucidate the inhibitory contribution of each glomerulus and their spatial organization is an indirect form of assessment.
In-vitro studies typically cannot examine all spatial connections 5 , Thus, to date there is no direct estimation of the amount, frequency and spatial extent of lateral inhibition in the olfactory bulb in-vivo.
Here we used precise optogenetic stimulations in a transgenic mouse strain to systematically characterize the frequency, strength and spatial distribution of lateral inhibition in different cell types in the mouse OB. These results support a model in which there are several classes of output neurons, each of which receives different levels of surround inhibition, which provides a mechanism for sending different odor features to the cortex in several parallel pathways.
Twenty-three transgenic male and female mice aged 3 to 6 months were used. The animals were housed in a group cage and received no experimental treatment except genotyping. The bone overlying the dorsal OB was removed. Ketamine is a known antagonist of NMDA receptors, which might hinder inhibition by granule cells. It is therefore possible that the result here reflects inhibitory circuits that are less affected by ketamine e. It is unlikely that we recorded granule cells as previous studies have noted that granule cells are not visible to extracellular electrodes 33 , 34 , Furthermore, there was no response to strong light stimulation when the electrode was in the granule cell layer.
Optical stimulation of the OB has been described in detail in To avoid possible bias due to respiration phase 37 , 38 , optical stimulation was randomly applied with respect to respiration cycles, with typically 20 repetitions of each spot stimulation. On average The expression of Chr2 in the OMP-Chr2 mice varies which result in different light power required to elicit a clear response near the recording electrode.
We scanned the full extent of the exposed dorsal OB divided into an N by M non-overlapping spots. N and M were determined by the size of the craniotomy. The result did not change when we only used experiments with the same spot sizes e.
Finally, due to the spatial organization of the axons of passage i. To avoid this, in a few neurons in which there was a clear response from the anterior OB parts we tended to focus our light stimulation on the more posterior parts of the OB and ignored the anterior parts. Data from all animals tested were included in the analysis. We termed spots that caused a significant reduction in firing rate as strong inhibitory spots. Although statistical significant does not necessarily imply large normalized effect size because statistical significant depends also on the sample size and variance, since in our experiments all spots were stimulated for a similar number of times the P values obtained are comparable to each other and can be used to assess the normalized effect size.
Therefore a significant inhibitory spot is likely to have relatively stronger inhibitory effect than a non-significant inhibitory spot see also text and Fig. A Strong lateral inhibition is scarce.
Schematic of the experiment for optogenetic stimulation of Omp-ChR2 based mice. ChR2 is expressed in olfactory sensory neurons OSNs. Right: PETHs and raster plots of one excitatory and one inhibitory spot marked with numbers. The P values of these excitatory and inhibitory spots are marked Mann-Whitney U-test. Light stimulations were randomized on the respiration cycle.
C Distribution of the percentage of significant inhibitory spots. E The reduction in baseline activity firing rate by the significant inhibitory spots blue and the non-significant inhibitory spots red.
F Same as E but for the percent of change from baseline activity. Only non-excitatory spots were considered. Since we typically light- excited many spots On average To estimate the number of false positive detected spots we repeated the same analysis but shuffled the spike count before and after light stimulation of each spot for all trials and recomputed the heat map and the rate of spots that evoked a significant change in firing rate. This shuffling procedure assumes that light stimulation has no effect on the evoked firing rate.
The rate of significant lateral inhibition spots reported in this study is slightly lower than what we reported in This is because we used a smaller spot size in the range of the size of a glomerulus and scanned the whole dorsal OB in the current study. We recorded respiration using a thermocouple. Because the transition between inhalation and exhalation was the most salient feature in the respiration recording trace, we defined a respiration cycle as being from the beginning of exhalation to the beginning of the next exhalation.
The beginning of respiration was thus the same as in 2. The piezoelectric and pressure sensor signals were very similar, with 5. We therefore shifted our respiration estimates by These thresholds detected all phase-locked neurons. Changing the threshold parameters e. Not all recorded neurons were locked to the respiration phase. However, the result of this study did not change if we restricted our analysis to only neurons that were significantly locked to respiration and we therefore did not exclude neurons from our analysis when the neuron preferred phase was not relevant to the analysis.
Each recorded neuron was typically excited strongly by one-two spots adjacent to the recording electrode and a few spots weakly around it. The magnitude of the light-evoked response was comparable to that observed during odor stimulation About 3. This however was expected from estimates of the false discovery rate 3. We found that contrary to what might be expected, inhibitory responses were quite rare.
Second, on average, only 5. This estimate is unlikely to be an underestimate since we used spot sizes and a light intensity that elicited clear excitatory responses that should have driven the inhibitory glomeruli. To estimate the strength of these significant inhibitory spots we plotted the absolute and relative reduction in baseline firing rate of all the significant inhibitory spots and the non-significant inhibitory spots Fig.
The mean reduction in baseline activity caused by the significant inhibitory spots was 6. The mean relative reduction was This analysis further suggests that the number of strong inhibitory spots is low.
Hence, while this analysis cannot exclude the possibility that there is a group of neurons with low baseline activity that receives strong lateral inhibition, the majority of the recorded neurons received strong lateral inhibition from only a small set of surrounding glomeruli and there was a substantial number of neurons that did not receive strong inhibition from any of the surrounding glomeruli in anesthetized mice.
The above analysis shows that strong lateral inhibition from surrounding glomeruli is infrequent. Many neurons do not receive strong lateral inhibition at all. Those that do, experience inhibition from a handful of surrounding glomeruli. However, our estimation of the percentage of surrounding inhibitory glomeruli was based on a threshold test. Thus, this procedure is biased towards detecting spots that exert a relatively strong inhibition and is more likely to fail when there is weak inhibition.
Examining more carefully the mean reduction in baseline activity caused by the none-significant inhibitory spots Figs. This strongly suggests that our classification method of inhibition does not capture all inhibitory spots and it is biased towards strong ones.
This method provides a stronger statistical power as it pools all average spot responses and therefore can capture small reductions in baseline activity as long as it is caused by a substantial number of spots.
To see this, consider the response map example shown in Fig. In this example, none of the light- activated spots crossed the threshold to be considered inhibitory per se.
Cells using LI have existed mainly in the cerebral cortex and thalamus. LI has been observed in the retina and lateral geniculate nuclei of the animals in experimental studies. Although LI has been identified primarily in the processings in visual sensation, it also occurs during sensory procedures such as touch, hearing, and smell.
RLI creates a stimulation contrast allowing increased sensory perception and enhances the contrast between the center and the periphery in a stimulated region. If activated at the same time, neighboring photoreceptors react less, although they are activated alone. Thus, when fewer neighboring neurons are stimulated, a neuron reacts more strongly. RLI is that the rod and cone photoreceptors in the perception zone interfere with each other to be active, inhibiting the response to central illumination by an increase in environmental illumination.
RLI is the main mechanism for achieving high visual acuity, sharpening the sensory location and color discrimination, which is involved in the transmission of contrasting edges in the visual image and increasing the contrasting sharpness.
After the light beam comes to the retina via crossing the cornea, the pupil and the lens, it then bypasses ganglion cells GCs , amacrine cells, bipolar cells, and HCs to reach the rod photoreceptors. The rods are stimulated by light and give a neural signal to stimulate the HCs. However, this stimulating signal will only be transmitted to the GCs by rod cells in the middle of the GC receiving area, because the HCs respond by sending an inhibitory signal to the neighboring rod photoreceptors.
The central rod cells send light signals directly to bipolar cells transmitting the signal to the GCs. Amacrine cells also provide LI to bipolar cells and GCs for various visual calculations such as image sharpening. Lastly, visual inputs are sent to the thalamus and cerebral cortex.
LI is directed by HCs in the vertebrate retina. The mutual synapse between cone cells and horizontal cells mediates negative feedback. RLI allows visual images to be transmitted to the central nervous system with appropriate visual contrast.
Before reaching the photoreceptors, light must travel through all the other layers of the retina. The reason for this somewhat paradoxical arrangement is that the light-sensitive pigments in the photoreceptors must be in contact with the layer of epithelial cells at the back of the eye, which provide them with a continuous supply of retinene , a light-sensitive derivative of Vitamin A.
Also, after the structural arrangement of the retinene molecules has been changed by the light energy, they are recycled in this epithelium. The dark pigmentation of this epithelium also prevents unabsorbed photons from being reflected back onto the photoreceptors and thus creating light interference that would degrade the image. Horizontal cells share a special characteristic with amacrine cells: the lack of any extension resembling an axon.
These cells in fact have only dendrites, some of which are presynaptic, that is, play the role of axons. The extensions of these cells thus apparently play both roles. Within the retina, information travels from the photoreceptors to the bipolar cells and then on to the ganglion cells. At each stage along this most direct visual pathway , the responses are modified by the activation of lateral connections involving horizontal and amacrine cells.
Thus the analysis of visual stimuli begins even in the retina. Every type of cell in these layers has its own distinctive distribution pattern and physiological characteristics. Regardless of whether nerve impulses from the photoreceptor cells are following the retina's direct pathway or its indirect pathway, in which the horizontal cells are involved, these signals must pass through the bipolar cells to reach the ganglion cells.
These signals are transmitted to the bipolar cells in the form of graduated potentials, which may be either depolarizations or hyperpolarizations, depending on whether the bipolar cell is of the ON or OFF type. Because the horizontal cells are connected laterally to many rods, cones, and bipolar cells, their role is to inhibit the activity of neighbouring cells. This selective suppression of certain nerve signals is called lateral inhibition, and its overall purpose is to increase the acuity of sensory signals.
In the case of vision, when light reaches the retina, it may illuminate some photoreceptors brightly and others much less so. By suppressing the signal from these less illuminated photoreceptors, the horizontal cells ensure that only the signal from the well lit photoreceptors reaches the ganglion cells, thus improving the contrast and definition of the visual stimulus. The retina's amacrine cells have highly diverse morphologies and employ an impressive number of neurotransmitters.
Their cell bodies are all located in the inner nuclear layer , while their synaptic endings are located in the inner plexiform layer. By connecting bipolar cells with ganglion cells, they provide an alternative, indirect path between them.
Amacrine cells appear to have many functions, most of them as yet unknown. As regards ganglion cells , various types with distinct functions have been characterized.
Lateral inhibition occurs in cells of the retina resulting in enhancement of edges and increased contrast in visual images. This type of lateral inhibition was discovered by Ernst Mach, who explained the visual illusion now known as Mach bands in In this illusion, differently shaded panels placed next to each other appear lighter or darker at the transitions despite uniform color within a panel.
Panels appears lighter at the border with a darker panel left side and darker at the border with a lighter panel right side. The darker and lighter bands at the transitions are not really there but are the result of lateral inhibition. Retinal cells of the eye receiving greater stimulation inhibit surrounding cells to a greater degree than cells receiving less intense stimulation.
Light receptors receiving input from the lighter side of the edges produce a stronger visual response than receptors receiving input from the darker side. This action serves to enhance contrast at the borders making the edges more pronounced.
Simultaneous contrast is also the result of lateral inhibition. In simultaneous contrast, the brightness of a background affects the perception of brightness of a stimulus. The same stimulus appears lighter against a dark background and darker against a lighter background. In the image above, two rectangles of different widths and uniform in color gray are set against a background with a gradient of dark to light from the top to the bottom.
Both rectangles appear lighter at the top and darker at the bottom. Due to lateral inhibition, light from the top portion of each rectangle against a darker background produces a stronger neuronal response in the brain than the same light from the lower portions of the rectangles against a lighter background.
Lateral inhibition also occurs in tactile, or somatosensory perception. Touch sensations are perceived by activation of neural receptors in the skin. The skin has multiple receptors that sense applied pressure.
Lateral inhibition enhances the contrast between stronger and weaker touch signals. Stronger signals at the point of contact inhibit neighboring cells to a greater degree than weaker signals peripheral to the point of contact. This activity allows the brain to determine the exact point of contact.
Areas of the body with greater touch acuity, such as the fingertips and tongue, have a smaller receptive field and a greater concentration of sensory receptors. Lateral inhibition is thought to play a role in hearing and the auditory pathway of the brain.
0コメント