Neuronal coding in the insect lamina

The large monopolar cells are specialised to respond to changes, or contrasts, in the visual signal. Like the photoreceptor cells, they convey signals as small, graded changes in membrane potential and do not generate trains of spikes. They are, therefore, particularly sensitive to small fluctuations in light intensity about an average value. Their responses also depend on the contrast between light received by their own photoreceptors and those of neighbouring cartridges.

Characteristic intracellular responses from a photoreceptor and a large monopolar cell to a pulse of light are shown in Fig. 5.2. Photoreceptors

Figure 5.2 The transfer of signals across the first synapse in an insect visual pathway. The drawing on the left shows how the retinula cells of an omma-tidium send short axons down to the layer of second-order neurons in the lamina. Intracellular recordings of responses to a 0.5s flash of light were recorded separately from the receptor cell body and axon, and from a large monopolar cell (LMC). (Recordings from Laughlin, 1981; copyright © 1981 Springer-Verlag.)

Figure 5.2 The transfer of signals across the first synapse in an insect visual pathway. The drawing on the left shows how the retinula cells of an omma-tidium send short axons down to the layer of second-order neurons in the lamina. Intracellular recordings of responses to a 0.5s flash of light were recorded separately from the receptor cell body and axon, and from a large monopolar cell (LMC). (Recordings from Laughlin, 1981; copyright © 1981 Springer-Verlag.)

make inhibitory synapses with large monopolar cells which, therefore, respond to an increase in light with a hyperpolarising signal, a response of the opposite polarity to that of photoreceptors. The response by a large monopolar cell is not, however, a mirror image of the response in a photoreceptor. These second-order neurons respond much more phasically than photoreceptors to a change in light, which is well illustrated in the waveforms of the responses to a pulse of light shown in Fig. 5.2. The response of the photoreceptor shows an initial peak depolarising potential, followed by a sustained, smaller depolarisation that lasts until the end of the stimulus. The large monopolar cell marks the start of the stimulus by a large, transient hyperpolarising signal, and the end of the stimulus by a transient depolarising signal. During the light stimulus, its membrane potential repolarises almost to its level in darkness. Thus, information about mean

Relative light intensity

0 10 100

Relative light intensity

0 10 100

Mean background intensity

Figure 5.3 Coding of light intensity by large monopolar cells (LMCs) of dragonfly (Hemicordulia) or blowfly (Calliphora). (a) Peak response amplitudes plotted as a function of the intensity of illumination delivered from a small point source of light, centred on the receptive field of a LMC. For the curve on the left, the LMC is dark adapted (•), whereas for the others it is light adapted to various different background intensities. For each curve, background intensity can be read from the intersect with the axis that shows zero response. The responses to both light-on and light-off were measured relative to the largest hyperpolarising response to light-on. (b) A schematic diagram to show coding of light-intensity in retinula cells and LMCs. The light signal varies sinusoidally and gradually increases in mean intensity, representing the envelope of light-intensity variations that might be experienced by scanning a visual scene at different intensities of ambient light. Because of adaptation, when the mean light level increases, the retinula cell response does not grow as dramatically as the light-intensity signal. The LMC response is proportional to the contrast in the light signal: most of the signal about ambient light intensity is removed, and the remaining signal is amplified. (Modified after Laughlin & Hardie, 1978.)

levels of illumination is lost in passing from photoreceptors to large monopolar cells, but information about changes in illumination is enhanced. The loss of information about the mean level of illumination is a good example of adaptation in a sensory system.

The way in which the large monopolar cells code visual stimuli has been studied by Simon Laughlin and colleagues in the same way as has been done for photoreceptors (Fig. 5.3). First, in the dark-adapted state, responses are measured to brief light stimuli of different intensities delivered from a dark background. Then a series of intensity-response curves is plotted, each one recording responses to increases and decreases in light from a particular mean, or background, level. Because adaptation quickly returns the membrane potential of a large monopolar cell to near its dark resting potential following a change in light, there is no steady-state response to background illumination. Consequently, the large monopolar cell response to an increase or decrease in light always starts from the same value of membrane potential, and the intensity-response curve moves horizontally by an amount equal to the change in background light intensity. Small increases and decreases in intensity produce, respectively, hyperpo-larising and depolarising departures from resting potential.

The size of a response to a particular change in light intensity is always greater in a large monopolar cell than in a photoreceptor. In other words, as the signal is passed across the first synapse in the pathway it is amplified as well as being inverted. In blowflies, the signal is amplified about six times. The relative change in light during a particular stimulus is expressed as its contrast, and so large monopolar cells generate larger responses than photoreceptors to light signals of particular contrasts.

Quantitatively, the relationship between contrast and response can be measured as the gradient of the intensity-response curve; the steeper the gradient, the greater the sensitivity to small changes in light intensity. Large monopolar cells have steeper intensity-response curves than photorecep-tors, reflecting their greater sensitivity to small fluctuations in light from a particular mean intensity. As a result, for a particular background level of light, the response by a large monopolar cell to increases in light intensity saturates at a lower intensity than the response by a photoreceptor. Saturation in the response is shown by a levelling-off in the intensity-response curve; it is the intensity at which a further increase in light will no longer lead to an increase in response amplitude.

The range of decreases in intensity that a light-adapted, large monopolar cell can cover is also reduced compared with a photoreceptor. Consequently, there is a trade-off in response characteristics: the price of an increase in sensitivity to changes in light is a reduction in the range of changes in intensity that can be covered. The response by a large monopolar cell to a signal with a particular contrast is the same, irrespective of the background, adapting light level.

When an insect experiences a large change in background light intensity, for example by flying out of dense shade into bright sunlight, its large monopolar cells can be driven beyond the range of their intensity-response curves. However, light adaptation is sufficiently rapid that large monopolar cell membrane potential usually settles to near its original value within less than a second, and the neuron's ability to respond to small changes in light intensity is restored. Following a change in light intensity of 100 times, the initial, large response by a large monopolar cell has almost completely decayed within 200 ms. Very large changes in light intensity such as this are extremely rare in the day-to-day life of an insect, and the intensity-response curve of a large monopolar cell spans the range of changes in intensity that the insect will encounter most often as it moves around.

Thus, the synapses that connect a photoreceptor with a large monopolar cell process the visual signal in two important ways. They filter it, so that information about background intensity is subtracted; and they amplify it, so that the signal about small changes in intensity is transmitted. Amplification, to make the signal as large as possible at an early stage in the visual pathway, is important because every time the signal passes across a synapse from one neuron to another it can become contaminated with noise (Laughlin, Howard & Blakeslee, 1987).

Another kind of transformation that occurs involves mutual inhibition between nearby cartridges. This kind of inhibition between neighbouring units in a sensory system is widespread, and is known as lateral inhibition. It is a mechanism for sharpening the receptive field of a neuron so that it responds well to small stimuli centred on its own receptive field, but not to large stimuli that fall on the receptive fields of many neurons. The signal about light levels detected by surrounding ommatidia is subtracted from the signal in the cartridge of a central ommatidium.

Lateral inhibition was first investigated in a compound eye of the horseshoe crab (Limulus) and was subsequently demonstrated to occur in vertebrate retinas (Box 5.1). Its action can be shown in an insect large monopolar cell by stimulating photoreceptors with a spot of light that is small enough to illuminate only a single photoreceptor. The spot can be directed at different angles to the ommatidium, and the responses by a large monopolar cell to the light change with the angle of the stimulus (Fig. 5.4). When the light is centred on the central axis of the large monopolar cell's receptive field, the cell's biggest responses are recorded, and the response amplitude

Box 5.1. Processing and filtering in the vertebrate retina

Although the insect eye is quite different in structure from the vertebrate retina, both have the function of converting a physical image into a neuronal representation that the brain can act upon, and there are striking parallels in the way the two operate, particularly in the early stages. The figure summarises some of the principal types of cells and their signals in response to a short pulse of light directed to the photoreceptor on the left, based mainly on work on the amphibian Necturus (see Dowling, 1987). Rod and cone photoreceptors transduce light into receptor potentials, and signals pass first to bipolar cells and then to retinal ganglion cells, which have axons that travel into the brain (about 1.2 million in humans; Sterling, 1998). Signals are modified by two layers of horizontally oriented cells, horizontal cells and amacrine cells. Adaptation to changes in background light levels occurs at several stages, including in the machinery responsible for transduction in the photoreceptors, in the strengths of electrical synapses that link photoreceptors, and in the chemical output

synapses from photoreceptors. Bipolar cells have steeper intensity-response curves than photoreceptors, and these curves remain centred on the mean background light level as it alters so, like the large monopolar cells in the fly eye, bipolar cells signal contrasts in the visual stimulus (Laughlin, 1994). Bipolar cells also have a centre-surround organisation to their receptive fields, in which the action of the photoreceptors directly above a bipolar cell is opposed by the surrounding photoreceptors, an action mediated via the horizontal cells. (Figure redrawn after Dowling, 1970)

decreases as the light is moved off axis. When the spot is shone from about 1.5° off axis, it begins to illuminate neighbouring ommatidia, and the large monopolar cell starts to generate a depolarising rather than a hyperpolarising response. The receptive field of the large monopolar cell, therefore, has two different regions: a central region where stimuli elicit the largest responses (+ sign in Fig 5.4), and a surrounding region that elicits responses with the opposite polarity (— sign in Fig. 5.4). These two regions of the receptive field work in opposition to each other, and the organisation of the receptive field is described as centre-surround. This type of organisation is extremely common for neurons in sensory systems. It acts as a type of sensory filter because it enables neurons to react strongly to small stimuli that are centred on their own receptive field, but less strongly or not at all to stimuli that are large enough to enter the receptive fields of several neurons.

A centre-surround receptive field organisation is an efficient way of processing visual information because it enables the size of neuronal signals to be related to the likelihood of a particular stimulus occurring. Two different points in space are more likely to be equally illuminated if they are close together than if they are far apart. This means that if we know how bright one point is, we could predict how bright the second is with a certainty that is related to their separation. In terms of recognising objects, we do not need to know the exact brightness of the two points, but whether they are different from each other - information about their exact brightness is redundant. Careful study of the responses by large monopolar cells in the fly has shown that the strengths of feedback and lateral inhibition are tuned to reduce redundant signals and enhance the detection of contrasts between neighbouring cartridges (Srinivasan, Laughlin & Dubs, 1982).



Angle of illumination (degrees off axis)

Figure 5.4 The sensitivity of a large monopolar cell to light incident upon the retina from different directions, expressed as a percentage of the maximum response. Zero response level is the membrane potential in the dark. Through lateral inhibition, off-axis light elicits a small response of opposite polarity to on-axis light, so that the receptive field has an excitatory centre and an inhibitory surround, as shown in the inset. (Redrawn after Srinivasan et al., 1982.)

Essentials of Human Physiology

Essentials of Human Physiology

This ebook provides an introductory explanation of the workings of the human body, with an effort to draw connections between the body systems and explain their interdependencies. A framework for the book is homeostasis and how the body maintains balance within each system. This is intended as a first introduction to physiology for a college-level course.

Get My Free Ebook

Post a comment