SIGNALING BETWEEN RODS, CONES AND BIPOLAR CELLSRESPONSES TO LIGHT AND DARKNESSPHASIC AND TONIC RESPONSESOVERALL CODINGS OF THE RETINAL OUTPUTCOVERAGELecture 2 Richard MaslandNovember 2, 2005PRIMARY VISUAL CODINGThe last lecture was concerned primarily with phototransduction--how a rod or cone catches a photon and transmits a signal about it from the outer segment to the synapse. Today wewill consider the subsequent processing of information by the retina. The retina is not simply a bank of photodiodes. It contains a variety of internal circuits, which compress and shape information before transmitting it to the optic nerve. At the end of the last lecture we saw one example. In the central retina, one cone drives one bipolar cell which drives one ganglion cell: this creates a region of maximal acuity. In fact, the acuity in the central retina reaches the theoretical maximum, which is limited by the packing density of individual cones. In the peripheral retina there is major convergence of rods and cones upon ganglion cells, such that acuity is traded for sensitivity and for economy in the number of optic nerve fibers. SIGNALING BETWEEN RODS, CONES AND BIPOLAR CELLSHow do rods and cones signal to the subsequent neurons? In general, the answer is: just like any other neurons. However, there are some unusual features. One was encountered in the previous lecture. Rods and cones hyperpolarize in response to light, and there is no sign of action potentials in them. It turns out that the synapses of rods and cones release neurotransmitter, in just the same way as any other cell. When the cell hyperpolarizes, less neurotransmitter is released by their synapses. Why do they not make action potentials? The reason is that rods and cones are very tiny cells. A garden variety neuron of the brain – say a cortical pyramidal cell – might extend for around 2 millimeters. A rod or cone spans only 20-30 microns from tip to base. Since the rod does not have to transmit information very far, there is no need for action potentials. The electrotonic decrement between the end of the outer segment and the base of the synapse is so small, that no regenerative activity is required.The same is true for retinal bipolar cells, horizontal cells, and many amacrine cells, whichare equally tiny. The cells of the outer retina do not need to make action potentials, because electrotonic decrements are very small. In contrast, the retinal ganglion cell must send an axon several centimeters to the brain. There, action potentials are required.RESPONSES TO LIGHT AND DARKNESSWhen a rod or cone hyperpolarizes, it releases less neurotransmitter. What effect does this have on the subsequent cells? At this point--the synapse between rods or cones and bipolar cells--a major divergence in the handling of information is created. Different bipolar cells contain different kinds of neurotransmitter receptors. The neurotransmitter released by rods and cones is glutamate. Some bipolar cells contain kainate-type glutamate receptor, which gates a cation channel. When glutamate falls on this receptor, the cation channel is opened. Ions, predominantly sodium, flow into the bipolar cell. The bipolar cell will thus follow fairly faithfully the activity of the rod or cone. When the cone is very hyperpolarized it releases little neurotransmitter. The bipolar cell cation channel is thus mostly closed, and the bipolar cell also remains relatively hyperpolarized. When the photoreceptor cell depolarizes, the bipolar cell also 1depolarizes. Accordingly, this synapse is sign-conserving: when light falls on the rod or cone, the bipolar cell hyperpolarizes (because the photoreceptor hyperpolarizes).However, there is a second class of bipolar cell, which contains a different type of glutamate receptor. This receptor is a metabotropic glutamate receptor, MGluR6. Binding of glutamate to the receptor activates a g protein. Through its downstream effectors (still incompletely known) it holds a cation channel closed. The net effect is that the synapse is sign-reversing. When the concentration of glutamate is high, the channel tends to stay closed. When the concentration of glutamate is low, the channel opens. When light hyperpolarizes the rod or cone, the concentration of glutamate in the synapse falls, the channel opens and the cell depolarizes. Thus are created two major classes of bipolar cell. One depolarizes when light fallson the rods or cones and the other hyperpolarizes. These are conventionally called ON and OFF bipolar cells. The neurotransmitter released by bipolar cells onto ganglion cells is also glutamate. The synapse of bipolar cells upon ganglion cells are not understood in as much detail as those of rods and cones on bipolar cells, but it is certain that they are sign-conserving synapses. As a final consequence, two classes of retinal ganglion cells are created. One responds to light with depolarization and one with hyperpolarization. Ganglion cells generate action potentials; when the cell is depolarized above a certain threshold, regenerative sodium spikes are caused. The net result is that one class of ganglion cells fires action potentials at an increased rate down the optic nerve to the brain, while the other decreases its rate.The dichotomy between ON and OFF responses is a central one in the early stages of vision. About half of the cells in the early visual system respond to light by increasing their rate of firing and half by decreasing it. One may imagine the situation as being a push-pull one. Retinal ganglion cells have fairly restricted rates of firing. Their operating range is from around 0 to around 1,000 Hz. The cells that are inhibited by light (OFF cells) tend to have a higher levelof spontaneous activity in the dark. They fire steadily even in the absence of a stimulus. This means that they have a working range at “negative” rates of firing--rates below their resting rate. One interpretation is that the overall range of signaling is thus expanded by having cells that work in two directions.Another way to think about it is to consider the situation at an edge between a light and a dark zone. What the visual system really cares about is transitions between light and dark. Uniform areas of illumination carry little information; it is the points of change where information if contained. If one has a light-dark edge, is the information contained in the