The processing of light by the retina and brain provides the basis for visual perception. Photons are captured and converted to electrical signals by rod and cone photoreceptor cells in the retina. These electrical signals are converted to chemical signals for transmission to downstream neurons. This article provides an overview of the mechanisms involved in transmitting light responses from rods and cones. Chemical signalling occurs at synapses between neurons. In keeping with many other neurons, the chemical messenger released by photoreceptors is the amino acid glutamate, which is packaged into small spherical vesicles. Each photoreceptor synaptic terminal has thousands of synaptic vesicles. Some of these vesicles are attached to the face of planar structures known as ribbons. Ribbons capture and deliver vesicles to release sites at the bottom of the ribbon. Upon stimulation, vesicles fuse with the cell membrane and release their contents. Glutamate molecules diffuse through the extracellular space to reach specialized receptors that regulate the activity of downstream neurons. We discuss how rates of vesicle attachment to ribbons, delivery of vesicles down the ribbon and the release of glutamate shape the information provided to downstream neurons in the visual system.
The duty of primary sensory cells in the body is to transform an external stimulus (i.e. chemical odorant, tastant, light, sound wave) into a transmembrane voltage change and then convert that voltage change into a chemical signal understood by neighbouring neurons. Vision begins with phototransduction: the capture of light and its transformation into membrane voltage changes by rod and cone photoreceptor cells of the retina. Rods operate at low light levels and can respond to single photons (particles of light), whereas cones operate at higher light levels. In both cell types, phototransduction occurs in an extensively modified cilium known as the outer segment. Photons absorbed by light-sensitive opsin molecules in outer segments activate an enzyme cascade that regulates membrane-spanning pore proteins known as ion channels. Movement of charged molecules (ions) through the channel changes the electrical potential (voltage) across the membrane. In darkness, positively charged ions enter the photoreceptor, neutralizing the negative charge on the inside of the membrane to produce a less-polarized or ‘depolarized’ transmembrane potential. With increasing light levels, channels close, reducing the entry of positively charged ions into the photoreceptor, leading to a more strongly polarized or ‘hyperpolarized’ membrane potential. The membrane potential of rods and cones varies continuously, hyperpolarizing with increasing light and depolarizing with decreasing light.
The process of phototransduction is understood in detail, down to the molecular structure of key proteins. But phototransduction is only the first step in vision. Light-evoked voltage responses of rods and cones must also be transmitted to downstream retinal neurons. Signals transmitted to second-order retinal bipolar cells are in turn transmitted to third-order retinal ganglion cells. Visual signals are modified by inhibitory horizontal and amacrine interneurons as they are transmitted through the retina. The axons of retinal ganglion cells bundle together to form the optic nerve (Figure 1), carrying light information to higher visual centres in the brain as a train of action potentials.
Structure of photoreceptor synapses
For transmission to second-order neurons, voltage changes in rods and cones generated by alteration in light levels are transformed into chemical signals. Neurons communicate with other neurons at specialized synapses (Figure 1b). Like many other neurons, rods and cones signal to downstream neighbours by releasing the amino acid, glutamate. Glutamate molecules are packaged into small, spherical, membrane-delimited synaptic vesicles. Vesicle release increases with membrane depolarization and diminishes with membrane hyperpolarization. Thus, in photoreceptor cells, release is maximal in darkness when rods and cones are depolarized and minimal in bright light when they are hyperpolarized. While the rate of release varies continuously, release from photoreceptors almost never ceases. This contrasts with most synapses in the brain where release occurs intermittently, triggered by brief depolarizing action potentials. The ability of rods and cones to sustain release almost endlessly is facilitated by their possession of many thousands of vesicles at each synapse. Most other neurons typically have fewer than a hundred vesicles per synapse (Figure 2).
A prominent feature of rod and cone synapses in electron micrographs is a structure known as the ribbon. In cross-section, ribbons look like dark, narrow bars jutting up from the plasma membrane, but three-dimensional reconstructions show them to be planar structures (Figure 2). Similar but smaller ribbons are present in the synaptic terminals of retinal bipolar cells, and spherical ribbons are found in auditory and vestibular sensory hair cells. Like photoreceptors, these other ribbon-bearing cells are sensory neurons that encode sensory information by varying the rates of neurotransmitter release.
The core of each ribbon is constructed from tens of thousands of copies of a single protein, Ribeye. Ribeye was originally isolated from cow eyes and is found only at ribbon synapses, hence its name. Ribeye proteins self-assemble into spherical structures, but the presence of an accessory protein, piccolino, helps ribbons assume a planar structure in photoreceptors. Ribbons are anchored to the plasma membrane through another protein with a musically inspired name, bassoon (Figure 3). Beneath each ribbon sits an arciform density, a trough-shaped structure that consists of multiple proteins including Munc13-2, RIM2 and CAST (Figure 3).
Along with thousands of cytoplasmic vesicles, hundreds of vesicles are attached to each ribbon, forming hexagonal arrays along both faces. Vesicles at the bottom of the ribbon can also contact the cell membrane. These membrane-associated vesicles form a ‘readily releasable pool’ that can be released almost immediately upon membrane depolarization. Other vesicles further up the ribbon form a reserve pool used to replenish release sites at the base.
Rods typically have a single large ribbon, whereas cones have 10–50 small ribbons. Cones release vesicles exclusively at ribbons, but rods can also release vesicles at more distant, non-ribbon sites. How use of these different release sites impacts vision is not known.
After release at a ribbon, glutamate molecules diffuse through the space between neurons (the synaptic cleft) to influence as many as a dozen different types of bipolar cells, each with distinct response properties. Each bipolar cell type appears specialized to extract particular features of the visual image, but the details of this processing are not fully understood.
Ca2+-dependent neurotransmitter release
Like other neurons, release of vesicles at photoreceptor ribbons is regulated by the entry of Ca2+ ions through Ca2+-selective, voltage-gated ion channels that open upon depolarization. The CaV1.4 Ca2+ channels that regulate neurotransmitter release in photoreceptors are clustered in the plasma membrane just beneath the arciform density. There are roughly three channels for each vesicle at the base of the ribbon. CaV1.4 Ca2+ channels are related to Ca2+ channels that regulate muscle contraction. However, unlike muscle subtypes, channels in rods and cones show little inactivation with continued membrane depolarization and Ca2+ influx. This property is essential for maintaining release continuously in darkness.
Changes in membrane potential accompanying changes in illumination are translated into changes in Ca2+ channel activity. Rates of synaptic vesicle release in rods and cones, in turn, vary linearly with the number of Ca2+ channel openings. Light-evoked changes in the membrane voltage of rods and cones are, therefore, encoded at the synapse by varying the number of Ca2+ channel openings to regulate the rate of synaptic vesicle fusion.
Similar to other membrane trafficking events, fusion between the lipid bilayers of a synaptic vesicle and cell membrane is driven by the coiling together of three proteins (synaptobrevin, SNAP25 and syntaxin) into a SNARE complex (Figure 4). Synchronizing fusion with membrane voltage changes requires a Ca2+ sensor. The principal sensor that controls fusion in rods and cones is the protein synaptotagmin 1. Upon Ca2+ binding, domains of synaptotagmin insert themselves into the cell membrane, unclamping the SNARE complex to allow fusion to proceed.
The typical membrane potential of rods and cones is near the midpoint of the voltage-sensitive activation range for CaV1.4 Ca2+ channels. This means that Ca2+ channel activity and, thus, glutamate release from photoreceptors is very sensitive to small light-dependent changes in membrane potential. The opening of only a few Ca2+ channels is sufficient to trigger vesicle fusion, further enhancing sensitivity to small changes in membrane voltage. The exquisite sensitivity of photoreceptor synaptic output to small contrast changes is well matched to the real world which is dominated by weak contrasts (i.e. lots of greys in a black and white image).
Synaptic ribbon function
It has long been thought that ribbons were required to maintain continuous vesicle release, but recent studies in which Ribeye was genetically eliminated caused surprisingly small deficits in the total amount of vesicle release, suggesting that ribbons are not simply vesicle repositories but play other roles.
Ribbons tether vesicles just above clusters of Ca2+ channels. After vesicles at the bottom of a ribbon fuse in response to Ca2+ channel opening, vesicles further up the ribbon move down to take their place. Vesicle movement along the ribbon does not require molecular motors fuelled by adenine 5′-triphosphate (ATP), but instead occurs passively. Each vesicle is attached to a ribbon by three to five fine tethers. The molecular composition of these tethers is unknown. The combined force of these tethers keeps the vesicles firmly attached to the ribbon, but a tether must detach in order for the vesicles to move along the ribbon, suggesting that each individual tether provides only a weak attachment force. Vesicles depart ribbons only by fusion with the cell membrane at the bottom of the ribbon. Vesicle fusion thus creates vacancies at the ribbon base that can be reoccupied by vesicles descending the ribbon. The population of vesicles, therefore, tends to migrate towards the base during sustained release.
In order to be released, vesicles need to be made fusion-competent in a process known as priming. At most synapses, vesicle priming occurs at the plasma membrane and involves the proteins Rim, Rab3A and Munc13. Surprisingly, these proteins are not required for priming vesicles at photoreceptor ribbon synapses, and priming appears to occur as vesicles descend the ribbon, prior to membrane association.
Because vesicles are primed during descent, vesicles that reach the base of a ribbon can be released almost immediately. Thus, the rates of maintained release in darkness are not governed directly by the rates of vesicle fusion, but by the rate at which vesicles can replenish empty release sites. Replenishment is, in turn, limited by the rate at which cytoplasmic vesicles attach to the ribbon.
Unlike most synapses in which vesicles are largely immobilized by attachment to the cytoskeleton, vesicles at photoreceptor ribbon synapses are free to move throughout the terminal. Cytoplasmic vesicles occasionally collide with ribbons and may attach if there is sufficient space available. The likelihood that a vesicle will attach can be enhanced by Ca2+ ions acting through a protein, calmodulin, to make ribbon-tethering molecules ‘stickier’. Ca2+-dependent enhancement allows rods and cones to adjust replenishment rates in response to changes in intracellular Ca2+.
Maintained release of vesicles in darkness depletes vesicles from the ribbon, and the slowing of release when photoreceptors hyperpolarize to light allows ribbons to be replenished. A ribbon, therefore, operates somewhat like an electrical capacitor, recharging with vesicles in bright light and then discharging them in a burst upon return to darkness. Rapid changes in release rate at light onset and offset preferentially encode information about contrast (changes in light intensity), whereas the maintained rate of release transmits information about steady light levels.
Another proposed function of ribbons is that they facilitate fusion between neighbouring vesicles by situating them close to one another. Vesicles can fuse with one another prior to exocytosis or sequentially during exocytosis. The SNARE protein, syntaxin 3, appears necessary for fusion between neighbouring vesicles on ribbons. The capability for vesicle–vesicle fusion allows ribbon synapses to alter the amount of glutamate released during each release event. This may provide another means of encoding information about light and dark to supplement the information coded by changes in release rate.
Researchers have made considerable progress in understanding how visual information is transmitted at rod and cone ribbon synapses. Studies have characterized the calcium channels that regulate release from photoreceptors, the behaviour of vesicles on ribbons and how these properties shape perceptions of luminance and contrast. However, many questions remain. For example, how are vesicles primed at ribbons? What is the molecular identity of vesicle tethers? Why are certain protein isoforms preferred at rod and cone synapses? In addition to revealing the role of photoreceptor synapses in shaping vision, these basic studies are essential for understanding the mechanisms underlying various retinal diseases. For example, mutations in CaV1.4 Ca2+ channels and their accessory proteins have been shown to cause night blindness by limiting neurotransmission from rods. Further study will allow researchers to identify and understand the synaptic pathologies that underlie other visual disorders.
Cooper, B., Hemmerlein, M., Ammermüller, J., et al. (2012) Munc13-independent vesicle priming at mouse photoreceptor ribbon synapses. J Neurosci. 32, 8040–8052. DOI: 10.1523/JNEUROSCI.4240-11.2012
Grabner, C.P., Ratliff, C.P., Light, A.C. and DeVries, S.H. (2016) Mechanism of high-frequency signaling at a depressing ribbon synapse. Neuron. 91, 133–415. DOI: 10.1016/j.neuron.2016.05.019
Jackman, S., Choi, S.-Y., Thoreson, W. B., et al. (2009) Role of the synaptic ribbon in transmitting the cone light response. Nature Neurosci. 12, 303–310. DOI: 10.1038/nn.2267
Moser, T., Grabner, C.P. and Schmitz, F. (2020) Sensory processing at ribbon synapses in the retina and the cochlea. Physiol Rev. 100, 103–144. DOI: 10.1152/physrev.00026.2018
Müller, T.M., Gierke, K., Joachimsthaler, A., et al. (2019) A multiple Piccolino-RIBEYE interaction supports plate-shaped synaptic ribbons in retinal neurons. J Neurosci. 39, 2606–2619. DOI: 10.1523/JNEUROSCI.2038-18.2019
Pangrsic, T., Singer, J.H. and Koschak, A. (2018) Voltage-gated calcium channels: key players in sensory coding in the retina and the inner ear. Physiol Rev. 98, 2063–2096. DOI: 10.1152/physrev.00030.2017
Schmitz, F., Königstorfer, A. and Südhof, T.C. (2000) RIBEYE, a component of synaptic ribbons: a protein’s journey through evolution provides insight into synaptic ribbon function. Neuron. 28, 857–872. DOI: 10.1016/s0896-6273(00)00159-8
Snellman, J., Mehta, B., Babai, N., et al. (2011) Acute destruction of the synaptic ribbon reveals a role for the ribbon in vesicle priming. Nature Neurosci. 14, 1135–1141. DOI: 10.1038/nn.2870
Thoreson, W.B. The physiology of photoreceptor synapses and other ribbon synapses. Reference Module in Neuroscience and Biobehavioral Psychology, Elsevier, Inc., N.Y., 2017. DOI: 10.1016/B978-0-12-809324-5.01484-X
Sterling, P. and Matthews, G. (2005) Structure and function of ribbon synapses. Trends Neurosci. 28, 20–29. DOI: 10.1016/j.tins.2004.11.009
Wallace B. Thoreson is the Gilmore Professor of Ophthalmology and Visual Sciences and the Director of Research in the Truhlsen Eye Institute at the University of Nebraska Medical Center (UNMC), Omaha, NE. After graduate and post-doctoral studies in neuroscience at the University of Minnesota, he joined the faculty at UNMC in 1993. His research focuses on mechanisms of early visual processing in the vertebrate retina. Email: email@example.com
Cassandra L. Hays received her PhD from UNMC, studying how rods transmit information about the detection of single photons. She also received an MS degree from UNMC for studies on eye pressure and glaucoma. She has taught classes in physiology and biology at the University of Nebraska-Omaha and Linn-Benton Community College, Albany, OR. Email: firstname.lastname@example.org