|John Dowling (2007), Scholarpedia, 2(12):3487.||doi:10.4249/scholarpedia.3487||revision #91715 [link to/cite this article]|
The retina is a thin (~0.25 mm) layer of neural tissue that lines the back of the eye. It is part of the brain (central nervous system) displaced into the eye during development. In addition to the light-sensitive photoreceptor cells – the rods and cones - the retina contains four basic classes of neurons – horizontal, bipolar, amacrine and ganglion cells – along with one major type of glial cell – the Müller cell. The neurons are organized into three cellular (nuclear) layers which are separated by two synaptic (plexiform layers). Virtually all retinal synapses are found in the plexiform layers, and all visual information passes across at least two synapses, one in the outer plexiform layer and another in the inner plexiform layer, before leaving the eye.
Analysis of the visual image occurs in both plexiform layers. The outer plexiform layer (OPL) separates visual information into ON- and OFF-channels and carries out a spatial type analysis on the visual input. The output neurons of the OPL, the ON- and OFF-bipolar cells demonstrate a center-surround antagonistic receptive field organization. The inner plexiform layer (IPL) is concerned more with the temporal and complex aspects of visual stimuli. Many cells receiving input in the IPL respond with transient responses and are highly sensitive to moving stimuli. The output neurons of the INL – the ganglion cells – carry the visual message to the rest of the brain; some ganglion cell responses reflect more OPL processing, others IPL processing.
Most vertebrate retinas contain two basic classes of photoreceptor cells, rods and cones. Rods mediate dim-light vision, whereas cones function in bright light and are responsible for color vision. The light sensitivity of the photoreceptors results from the presence of visual pigment molecules contained within a specialized region of the cell, called the outer segment. One rod pigment called rhodopsin, and three cone pigments which absorb light maximally in the red, green and blue regions of the spectrum, are in the human retina; non-mammalian species may have 4 or more cone visual pigments.
All visual pigments have a similar chemistry. They consist of two components, retinal (vitamin A aldehyde), termed a chromophore, bound to a protein called opsin. Different visual pigments have different opsins, and this accounts for the variations in their color sensitivity. The light sensitivity of the visual pigments is due to the retinal chromophore. When a visual pigment molecule absorbs a quantum of light, several conformational (shape) changes occur, first in the chromophore and then in the protein (opsin) part of the molecule. The opsin conformational changes lead to a series of intermediates called metarhodopsins ( Figure 2), and one, metarhodopsin II, results in excitation of the photoreceptor cell via a cascade of molecular interactions.
All vertebrate photoreceptors hyperpolarize in response to light (i.e., the inside of the cell becomes more negative). Why photoreceptors hyperpolarize in the light is as follows: In darkness, the membrane of the outer segment is permeable (leaky) to Na+. Because Na+ levels are higher outside the cell than inside, positively charged Na+ ions enter the cell in darkness, causing the cell to be partially depolarized (i.e., the cell’s inside is more positive than is typically the case for neurons at rest). Light decreases the permeability of the outer segment membrane to Na+, thereby decreasing the flow of positive ions into the cell and causing the inside of the cell to become more negative (i.e., to hyperpolarize).
In many primates including humans, some birds and reptiles, a small region of the retina is specialized for high-acuity vision. It is called the fovea and is centrally located (i.e. on the visual axis of the eye). The layers of the retina below the photoreceptors are displaced laterally around the fovea so that light can impinge more directly on the photoreceptors. This results in a retinal thinning centrally, forming the so-called foveal pit. Only cones are present in the central region of the fovea, and the foveal cones are the thinnest and longest photoreceptors in the retina. The rod-free area in humans is about 0.3 mm in diameter and contains approximately 35,000 cones. No blood vessels are found in the human fovea and, furthermore, there are few blue-absorbing cones in the center of the fovea. These specializations serve to improve the visual resolution of the fovea.
Neurons: Cellular and synaptic organization
Although there are just four basic classes of retinal neurons in addition to the photoreceptors, many neuronal types exist in each of these classes. Retinas typically have one type of rod cell and three or four cone types which are maximally responsive to light of different wavelengths (e.g. red, green, blue light). Three to four horizontal cell types exist in most retinas and as many as nine to eleven bipolar cell types. Amacrine cell types tend to be most numerous with retinas having as many as twenty or more. Finally, twelve to fifteen ganglion cell types have been identified in various species. Two kinds of chemical synapses have been identified in all retinas – ribbon synapses made by photoreceptor and bipolar terminal endings, and conventional synapses made by amacrine and horizontal cells ( Figure 3). Ribbon synapses are characterized by a dense ribbon-like structure found presynaptically that serves as a conveyor belt to bring synaptic vesicles to the presynaptic membrane where they bind and release their contents. Conventional synapses are characterized by a cluster of synaptic vesicles apposed to the membrane to which they bind. In addition, cone photoreceptors make an unusual chemical synapse, called a flat or basal junction mainly onto OFF-type bipolar cells, and finally, electrical synapses are seen between many types of retinal neurons.
Figure 4 depicts in a highly simplified way the cellular and synaptic organization of the primate retina. The photoreceptors, rods (R) and cones (C) provide the input to the OPL, contacting horizontal cells (H) via ribbon synapses and bipolar cells via ribbon or flat/basal synapses. One type of horizontal cell is depicted along with four bipolar cell types; flat (OFF-) and invaginating (ON-) midget bipolar cells that contact a single cone (FMB and IMB), diffuse bipolars that contact several cones (IDB), and rod bipolars (RB) that contact only rods. In the IPL the bipolar cells via ribbon synapses contact midget ganglion cells (MG), larger, diffuse ganglion cells (G) and amacrine cells (A). The rod bipolar cells contact amacrine cells that pass rod information onto cone bipolar cell terminals and the large ganglion cells. Other amacrine cells contact midget bipolar terminals, and midget and diffuse ganglion cells. Finally, a specialized type of amacrine cell, called an interplexiform cell (I), receives input in the INL and feeds information back to cells of the OPL ( Figure 4).
Light responses and retinal processing of information
The distal retinal neurons – photoreceptor, horizontal and bipolar cells, respond to light with sustained graded membrane potential (voltage) changes. Most of these cells also respond by hyperpolarizing in response to light stimuli – the membrane potential becomes more negative ( Figure 5). The more proximal retinal neurons respond to light mainly by depolarizing – the membrane potential becomes more positive – and they generate action potentials like most brain neurons. Distal retinal neurons use sustained graded potentials to carry visual signals, probably for two reasons: 1) they have relatively short processes and do not need to carry information long distances; hence, the electrotonic spread of potential along the cell membrane is sufficient to transmit information from one end of the cell to the other; and 2) graded potentials are capable of discriminating a wider range of signals than can all-or-none events (i.e. action potentials).
All photoreceptors hyperpolarize in response to light; via their synapses they cause horizontal and OFF-bipolar cells to hyperpolarize and ON-bipolar cells to depolarize. The horizontal cells act as lateral inhibitory neurons, establishing the classic center-surround receptive field organization first seen prominently in the bipolar cells. The center response reflects the direct photoreceptor-bipolar cell input; the antagonistic surround response reflects inhibitory horizontal cell input from more distal photoreceptors. Two types of amacrine cell responses – transient and sustained – have been observed in most retinas. Transient amacrine cells may respond by depolarizing at the ON-, OFF- or at both the ON- and OFF- of illumination. Sustained amacrine cells may either depolarize or hyperpolarize to light. Action potentials are often observed on the depolarizing responses of amacrine cells.
All ganglion cells generate action potentials and to full-field illumination, three basic types of responses are observed; ON, OFF, and ON-OFF cells. However, by probing the retina with small spots of light, the receptive field organizations of ganglion cells are found to be varied and complex, reflecting excitatory, inhibitory and modulatory interactions occurring in the retina, especially in the inner plexiform layer. Most prominent are the contrast-sensitive ON-center OFF-surround ganglion cells, and OFF-center ON-surround cells ( Figure 6a). A recent study identified ten subtypes of ganglion cell responses in the rabbit retina and correlated these with various morphological ganglion cell types. Each of these ganglion cell types provides higher visual centers with a somewhat different representation of the image falling on the retina. Thus, much visual processing occurs in the retina, the extent of which is almost certainly not fully recognized.
An intriguing type of ganglion cell found in many retinas is a movement-sensitive cell that is directionally selective – it responds well when a spot of light moves across the retina in one direction, but not at all when the spot moves in the opposite direction ( Figure 6b). A special type of amacrine cell – the starburst amacrine cell – is involved in the generation of the directionally-selective responses, and understanding of the synaptic interactions that underlie direction-selectivity in these ganglion cells is well along.
Transmitters, modulators, and receptors
The retina, like other regions of the brain, employs a large number of neuroactive substances (Figure 7). They include both neurotransmitters – substances that act rapidly and directly on the retinal neurons by altering membrane permeability to one or several ions; and neuromodulators – substances that usually act more slowly by altering the neurons via intracellular biochemical cascades. Relatively few substances appear to serve as retinal neurotransmitters; the majority of neuroactive substances released from retinal neurons (mainly from amacrine cells) appear to be neuromodulatory in nature although little is known about the action or role of most of these substances.
Amino acids and acetylcholine are the principal neurotransmitters used in the retina. L-glutamate is used by both photoreceptors and bipolar cells as their excitatory transmitter, whereas acetylcholine is used as an excitatory neurotransmitter in the IPL, principally by starburst amacrine cells. GABA and glycine are the major inhibitory transmitters used in the retina, and it is believed that 80% of the amacrine cells release one or another of these substances. A major area of study is to identify the various receptors/channels present on the neurons that mediate the actions of the neurotransmitters/modulators in the retina. For example, it is clear that ON- and OFF-bipolar cells, and perhaps even the horizontal cells, express different glutamate receptors, and how these receptors/channels give rise to the responses of the various neurons is beginning to be understood. A particularly interesting challenge is to understand how the glutamate released from photoreceptor cells gives rise to the ON-bipolar cell response. Two types of molecule have been implicated – a metabotropic glutamate receptor expressed by rod bipolar cells in all species and a transporter-like receptor expressed by ON-cone bipolar cells in the teleost retina.
A number of substances are believed to function as neuromodulatory agents in the retina, although how they function and their role is poorly understood. The proposed neuromodulatory substances include a number of amines, neuropeptides and other substances such as nitric oxide, retinoic acid, and zinc. About a dozen neuropeptides have been identified in the retina, mainly in amacrine cells and often they are colocalized with GABA. Their roles and the significance of the colocalization with GABA is not known.
Of all the proposed neuromodulatory substances, we know most about the role of dopamine which has been shown to modulate electrical (gap) junctions throughout the retina and the glutamate receptors on horizontal and bipolar cells. Dopamine, through cyclic AMP and protein kinase A (PKA), alters the conductance of many retinal gap junctions, and the sensitivity of certain glutamate receptors to glutamate. Dopamine in the retina is also believed to play a role in establishing retinal circadian rhythmicity, and regulating retinal motor movements in fish and other non-mammalian species.
Baylor, Denis A. (1987) Photoreceptor signals and vision, Investigative Ophthalmology and Visual Science. 28, 34-49.
Dowling, John E. (1987) The Retina: An Approachable Part of the Brain. Harvard University Press.
Fried, S. I., Münch, T. A. and Werblin, Frank S. (2005) Directional selectivity is formed at multiple levels by laterally offset inhibition in the rabbit retina. Neuron. 46, 117-127.
Kolb, Helga, Ripps, H. and Wu S. (eds.) (2001) Concepts and Challenges in Retinal Biology. Amsterdam:Elsevier.
Polyak, Steven L. (1941) The Retina. Chicago: Chicago University Press.
Rodieck, Robert W. (1998) The First Steps in Seeing. Sunderland, MA: Sinauer Associates.
Roska, Botond, Molnar, A., Werblin, Frank S. (2006) Parallel processing in retinal ganglion cells: How integration of space-time patterns of excitation and inhibition form the spiking output, Journal of Neurophysiology, 95, 3810-3822.
Sterling, P. and Demb, J. B. (2004) Retina. In Shepherd, G. M. (ed) The Synaptic Organization of the Brain. pp. 217-269. Oxford: Oxford University Press.
- Valentino Braitenberg (2007) Brain. Scholarpedia, 2(11):2918.
- Keith Rayner and Monica Castelhano (2007) Eye movements. Scholarpedia, 2(10):3649.
- Eugene Roberts (2007) Gamma-aminobutyric acid. Scholarpedia, 2(10):3356.