Concerted signaling by retinal ganglion cells.

To analyze the rules that govern communication between eye and brain, visual responses were recorded from an intact salamander retina. Parallel observation of many retinal ganglion cells with a microelectrode array showed that nearby neurons often fired synchronously, with spike delays of less than 10 milliseconds. The frequency of such synchronous spikes exceeded the correlation expected from a shared visual stimulus up to 20-fold. Synchronous firing persisted under a variety of visual stimuli and accounted for the majority of action potentials recorded. Analysis of receptive fields showed that concerted spikes encoded information not carried by individual cells; they may represent symbols in a multineuronal code for vision.

Synchronous bursts of action potentials in ganglion cells of the developing mammalian retina.

The development of orderly connections in the mammalian visual system depends on action potentials in the optic nerve fibers, even before the retina receives visual input. In particular, it has been suggested that correlated firing of retinal ganglion cells in the same eye directs the segregation of their synaptic terminals into eye-specific layers within the lateral geniculate nucleus. Such correlations in electrical activity were found by simultaneous recording of the extracellular action potentials of up to 100 ganglion cells in the isolated retina of the newborn ferret and the fetal cat. These neurons fired spikes in nearly synchronous bursts lasting a few seconds and separated by 1 to 2 minutes of silence. Individual bursts consisted of a wave of excitation, several hundred micrometers wide, sweeping across the retina at about 100 micrometers per second. These concerted firing patterns have the appropriate spatial and temporal properties to guide the refinement of connections between the retina and the lateral geniculate nucleus.

Mechanisms of rhodopsin inactivation in vivo as revealed by a COOH-terminal truncation mutant.

Although biochemical experiments suggest that rhodopsin and other receptors coupled to heterotrimeric guanosine triphosphate-binding proteins (G proteins) are inactivated by phosphorylation near the carboxyl (COOH)-terminus and the subsequent binding of a capping protein, little is known about the quenching process in vivo. Flash responses were recorded from rods of transgenic mice in which a fraction of the rhodopsin molecules lacked the COOH-terminal phosphorylation sites. In the single photon regime, abnormally prolonged responses, attributed to activation of individual truncated rhodopsins, occurred interspersed with normal responses. The occurrence of the prolonged responses suggests that phosphorylation is required for normal shutoff. Comparison of normal and prolonged single photon responses indicated that rhodopsin begins to be quenched before the peak of the electrical response and that quenching limits the response amplitude.

Role for the target enzyme in deactivation of photoreceptor G protein in vivo.

Heterotrimeric guanosine 5'-triphosphate (GTP)-binding proteins (G proteins) are deactivated by hydrolysis of the GTP that they bind when activated by transmembrane receptors. Transducin, the G protein that relays visual excitation from rhodopsin to the cyclic guanosine 3',5'-monophosphate phosphodiesterase (PDE) in retinal photoreceptors, must be deactivated for the light response to recover. A point mutation in the gamma subunit of PDE impaired transducin-PDE interactions and slowed the recovery rate of the flash response in transgenic mouse rods. These results indicate that the normal deactivation of transducin in vivo requires the G protein to interact with its target enzyme.

Origin and functional impact of dark noise in retinal cones.

Spontaneous fluctuations in the electrical signals of the retina's photoreceptors impose a fundamental limit on visual sensitivity. While noise in the rods has been studied extensively, relatively little is known about the noise of cones. We show that the origin of the dark noise in salamander cones varies with cone type. Most of the noise in long wavelength-sensitive (L) cones arose from spontaneous activation of the photopigment, which is a million-fold less stable than the rod photopigment rhodopsin. Most of the noise in short wavelength-sensitive (S) cones arose in a later stage of the transduction cascade, as the photopigment was relatively stable. Spontaneous pigment activation effectively light adapted L cones in darkness, causing them to have a smaller and briefer dim flash response than S cones.

Rapid and reproducible deactivation of rhodopsin requires multiple phosphorylation sites.

Efficient single-photon detection by retinal rod photoreceptors requires timely and reproducible deactivation of rhodopsin. Like other G protein-coupled receptors, rhodopsin contains multiple sites for phosphorylation at its COOH-terminal domain. Transgenic and electrophysiological methods were used to functionally dissect the role of the multiple phosphorylation sites during deactivation of rhodopsin in intact mouse rods. Mutant rhodopsins bearing zero, one (S338), or two (S334/S338) phosphorylation sites generated single-photon responses with greatly prolonged, exponentially distributed durations. Responses from rods expressing mutant rhodopsins bearing more than two phosphorylation sites declined along smooth, reproducible time courses; the rate of recovery increased with increasing numbers of phosphorylation sites. We conclude that multiple phosphorylation of rhodopsin is necessary for rapid and reproducible deactivation.

Receptive-field microstructure of blue-yellow ganglion cells in primate retina.

We examined the functional microcircuitry of cone inputs to blue-ON/yellow-OFF (BY) ganglion cells in the macaque retina using multielectrode recording. BY cells were identified by their ON responses to blue light and OFF responses to red or green light. Cone-isolating stimulation indicated that ON responses originated in short (S) wavelength-sensitive cones, whereas OFF responses originated in both long (L) and middle (M) wavelength-sensitive cones. Stimulation with fine spatial patterns revealed locations of individual S cones in BY cell receptive fields. Neighboring BY cells received common but unequal inputs from one or more S cones. Inputs from individual S cones differed in strength, indicating different synaptic weights, and summed approximately linearly to control BY cell firing.

Interactions between divalent cations and the gating machinery of cyclic GMP-activated channels in salamander retinal rods.

The effects of divalent cations on the gating of the cGMP-activated channel, and the effects of gating on the movement of divalent cations in and out of the channel's pore were studied by recording macroscopic currents in excised membrane patches from salamander retinal rods. The fractional block of cGMP-activated Na+ currents by internal and external Mg2+ as well as internal Ca2+ was nearly independent of cGMP concentration. This indicates that Mg2+ and Ca2+ bind with similar affinity to open and closed states of the channel. In contrast, the efficiency of block by internal Cd2+ or Zn2+ increased in proportion to the fraction of open channels, indicating that these ions preferentially occupy open channels. The kinetics of block by internal Ni2+, which competes with Mg2+ but blocks more slowly, were found to be unaffected by the fraction of channels open. External Ni2+, however, blocked and unblocked much more rapidly when channels were mostly open. This suggests that within the pore a gate is located between the binding site(s) for ions and the extracellular mouth of the channel. Micromolar concentrations of the transition metal divalent cations Ni2+, Cd2+, Zn2+, and Mn2+ applied to the cytoplasmic surface of a patch potentiated the response to subsaturating concentrations of cGMP without affecting the maximum current induced by saturating cGMP. The concentration of cGMP that opened half the channels was often lowered by a factor of three or more. Potentiation persisted after the experimental chamber was washed with divalent-free solution and fresh cGMP was applied, indicating that it does not result from an interaction between divalent cations and cGMP in solution; 1 mM EDTA or isotonic MgCl2 reversed potentiation. Voltage-jump experiments suggest that potentiation results from an increase in the rate of cGMP binding. Lowering the ionic strength of the bathing solution enhanced potentiation, suggesting that it involves electrostatic interactions. The strong electrostatic effect on cGMP binding and absence of effect on ion permeation through open channels implies that the cGMP binding sites on the channel are well separated from the permeation pathway.

Modification of cyclic nucleotide-gated ion channels by ultraviolet light.

We irradiated cyclic nucleotide-gated ion channels in situ with ultraviolet light to probe the role of aromatic residues in ion channel function. UV light reduced the current through excised membrane patches from Xenopus oocytes expressing the alpha subunit of bovine retinal cyclic nucleotide-gated channels irreversibly, a result consistent with permanent covalent modification of channel amino acids by UV light. The magnitude of the current reduction depended only on the total photon dose delivered to the patches, and not on the intensity of the exciting light, indicating that the functionally important photochemical modification(s) occurred from an excited state reached by a one-photon absorption process. The wavelength dependence of the channels' UV light sensitivity (the action spectrum) was quantitatively consistent with the absorption spectrum of tryptophan, with a small component at long wavelengths, possibly due to cystine absorption. This spectral analysis suggests that UV light reduced the currents at most wavelengths studied by modifying one or more "target" tryptophans in the channels. Comparison of the channels' action spectrum to the absorption spectrum of tryptophan in various solvents suggests that the UV light targets are in a water-like chemical environment. Experiments on mutant channels indicated that the UV light sensitivity of wild-type channels was not conferred exclusively by any one of the 10 tryptophan residues in a subunit. The similarity in the dose dependences of channel current reduction and tryptophan photolysis in solution suggests that photochemical modification of a small number of tryptophan targets in the channels is sufficient to decrease the currents.

Recoverin regulates light-dependent phosphodiesterase activity in retinal rods.

The Ca2+-binding protein recoverin may regulate visual transduction in retinal rods and cones, but its functional role and mechanism of action remain controversial. We compared the photoresponses of rods from control mice and from mice in which the recoverin gene was knocked out. Our analysis indicates that Ca2+-recoverin prolongs the dark-adapted flash response and increases the rod's sensitivity to dim steady light. Knockout rods had faster Ca2+ dynamics, indicating that recoverin is a significant Ca2+ buffer in the outer segment, but incorporation of exogenous buffer did not restore wild-type behavior. We infer that Ca2+-recoverin potentiates light-triggered phosphodiesterase activity, probably by effectively prolonging the catalytic activity of photoexcited rhodopsin.

Photoreceptor signals and vision. Proctor lecture.

In recent years, there has been rapid progress in understanding the properties and mechanism of generation of the light-evoked electrical signals of vertebrate rods and cones. The graded hyperpolarization that carries information over the length of the cell is generated by closure of cation-selective aqueous pores in the surface membrane of the outer segment. These pores are controlled cooperatively by cyclic GMP, which acts continuously in darkness to keep the pores open. Photoisomerization of rhodopsin or cone pigment produces the rapid amplified activation of phosphodiesterase, which lowers the concentration of cGMP, thereby lowering the conductance of the surface membrane. Calcium ions, once thought to relay excitation to the light-sensitive channels, do not play this role. Instead, they appear to participate in a feedback control mechanism that regulates the nucleotide cascade. Although some general features of the transduction mechanism are now understood, a number of important questions remain. How is the nucleotide cascade shut off? Where does Ca act? What is the structure of the light-sensitive channel? How are stereotyped single photon responses produced? Primate photoreceptors are no longer off limits to single cell electrophysiology. Analysis of the response properties and dark noise of primate rods gives a physiological basis for several fundamental features of human rod vision: single photon detection, poor temporal resolution, the "dark light," rod saturation, scotopic spectral sensitivity, and, perhaps, after-image signals. Primate cones show less sensitive but faster responses shaped by a resonance which may figure in the flicker sensitivity of human cone vision. The spectral sensitivity of the three types of primate cones has been determined over the entire visible region. These sensitivities satisfactorily predict human color matching. The spectral sensitivity curves indicate that the pigment in a given cone is very pure, and that individual cones of a given type normally contain pigments with very similar or identical spectral properties.

Multi-neuronal signals from the retina: acquisition and analysis.

Throughout the central nervous system, information about the outside world is represented collectively by large groups of cells, often arranged in a series of 2-dimensional maps connected by tracts with many fibers. To understand how such a circuit encodes and processes information, one must simultaneously observe the signals carried by many of its cells. This article describes a new method for monitoring the simultaneous electrical activity of many neurons in a functioning piece of retina. Extracellular action potentials are recorded with a planar array of 61 microelectrodes, which provides a natural match to the flat mosaic of retinal ganglion cells. The voltage signals are processed in real time to extract the spike trains from up to 100 neurons. We also present a method of visual stimulation and data analysis that allows a rapid characterization of each neuron's visual response properties. A randomly flickering display is used to elicit spike trains from the ganglion cell population. Analysis of the correlations between each spike train and the flicker stimulus results in a simple description of each ganglion cell's functional properties. The combination of these tools will allow detailed study of how the population of optic nerve fibers encodes a visual scene.

Light path and photon capture in turtle photoreceptors.

1. The directional selectivity of individual cones was examined by intracellular recording in the eye of the turtle. Sensitivites were determined from linear responses to dim flashes of monochromatic light incident on a cell over a range of angles to its long axis. 2. With light near the optimum wave-length, some red- and green-sensitive cones showed a high sensitivity for light entering axially and lower sensitivities for light entering obliquely. In contrast, other cells had lower peak sensitivities and less pronounced directional selectivities. The highest axial sensitivities observed in red receptors were about 320 muV photon(-1) mu2; in these cells, the sensitivity declined to half for rays 6-9 degrees off the axis as measured in the retina. Green receptors had lower axial sensitivities and broader angular profiles. 3. On the assumption that rays at all angles contribute independently to the over-all sensitivity, the sensitivity of a cell to large cones of rays was successfully predicted from the angular selectivity determined with a narrow pencil of rays. The shape of small responses to dim stimuli delivered on and off the axis of the cell was invariant, implying that a cone signals the number of photons absorbed but not their angle of incidence. 4. Short wave-lengths have previously been shown to be filtered out by the oil droplets present in turtle cones. At short wave-lengths, the angular profiles showed a depression in axial sensitivity consistent with this filtering action. 5. Diameters of inner segments, oil droplets, and outer segments were measured in red-, green-, and blue-sensitive cones, since these dimensions are expected to influence the cones' angular acceptances and ability to collect light. The diameters of the structure were in approximately the same proportions for each type of receptor, but the absolute values of the diameters were found to be scaled in relation to the wave-length of maximum sensitivity. 6. Optical determinations of the efficiency with which axial rays are concentrated by red receptors gave a mean value of 55%. 7. Receptors in histological sections of the whole eye were found to be oriented with their long axes directed approximately toward the pupil. 8. The observed directional selectivities and collecting efficiencies agree well with the behaviour of a model retinal cone developed by Winston & Enoch (1971) on a geometrical optical treatment. 9. Effective collecting areas are derived for red-, green- and blue-sensitive cones; these permit conversion of observed flash sensitivities into the mean peak hyperpolarization produced by isomerization of a visual pigment molecule. The figure obtained is about 25 muV for red-sensitive cones and 21muV for green-sensitive cones.