Color vision is spectrally opponent, suggesting that spectrally opponent neurons, such as the horizontal cells in fish and turtle retinae, play a prominent role in color discrimination. In the accompanying paper (Kraaij et al., 1998), it was shown that the output signal of the horizontal cell system to the cones is not at all spectrally opponent. Therefore, a role for the spectrally opponent horizontal cells in color discrimination seems unlikely. In this paper, we propose that the horizontal cells play a prominent role in color constancy and simultaneous color contrast instead of in color discrimination. We have formulated a model of the cone/horizontal cell network based on measurements of the action spectra of the cones and of the feedback signal of the horizontal cell system to the various cone types. The key feature of the model is (1) that feedback is spectrally and spatially very broad and (2) that the gain of the cone synapse strongly depends on the feedback strength. This makes the synaptic gain of the cones strongly dependent on the spectral composition of the surround. Our model, which incorporates many physiological details of the outer retina, displays a behavior that can be interpreted as color constancy and simultaneous color contrast. We propose that the horizontal cell network modulates the cone synaptic gains such that the ratios of the cone outputs become almost invariant with the spectral composition of the global illumination. Therefore, color constancy appears to be coded in the retina.

The spectral sensitivity of cones in isolated goldfish retina was determined with whole-cell recording techniques. Three spectral classes of cones were found with maximal sensitivities around 620 nm, 540 nm, and 460 nm. UV-cones were not found because our stimulator did not allow effective stimulation in the UV range. The spectral sensitivity of the cones closely matched the cone photopigment absorption spectra at the long wavelength side of the spectrum, but deviated significantly at shorter wavelengths. Surround stimulation induced an inward current in cones due to feedback from horizontal cells. The spectral sensitivity of this feedback signal was determined in all three cone classes and found to be broader than the spectral sensitivity of the cones recorded from, and to be spectrally nonopponent. These data are consistent with a connectivity scheme between cones and horizontal cells in which the three horizontal cell systems feed back to all cone systems and in which all horizontal cell systems receive input from more than one cone system.

Mammalian rods respond to single photons with a hyperpolarization of about 1 mV which is accompanied by continuous noise. Since the mammalian rod bipolar cell collects signals from 20–100 rods, the noise from the converging rods would overwhelm the single-photon signal from one rod at scotopic intensities (starlight) if the bipolar cell summed signals linearly (Baylor et al., 1984). However, it is known that at scotopic intensities the retina preserves single-photon responses (Barlow et al., 1971; Mastronarde, 1983). To explore noise summation in the rod bipolar pathway, we simulated an array of rods synaptically connected to a rod bipolar cell using a compartmental model. The performance of the circuit was evaluated with a discriminator measuring errors in photon detection as false positives and false negatives, which were compared to physiologically and psychophysically measured error rates. When only one rod was connected to the rod bipolar, a Poisson rate of 80 vesicles/s was necessary for reliable transmission of the single-photon signal. When 25 rods converged through a linear synapse the noise caused an unacceptably high false positive rate, even when either dark continuous noise or synaptic noise where completely removed. We propose that a threshold nonlinearity is provided by the mGluR6 receptor in the rod bipolar dendrite (Shiells & Falk, 1994) to yield a synapse with a noise removing mechanism. With the threshold nonlinearity the synapse removed most of the noise. These results suggest that a threshold provided by the mGluR6 receptor in the rod bipolar cell is necessary for proper functioning of the retina at scotopic intensities and that the metabotropic domains in the rod bipolar are distinct. Such a nonlinear threshold could also reduce synaptic noise for cortical circuits in which sparse signals converge.

The appearance of cGMP-gated cation channel protein in the postnatal rat retina has been studied by fluorescence immunocytochemistry of radial retinal sections and immunoblots of retinal membrane proteins. Channel immunoreactivity was first detectable with RCNGC1-7H2 monoclonal antibody at postnatal day 7 (PN7) by both methods. Immunocytochemical label in retinal sections was localized to the outer segments, and immunoreactivity increased with increasing age. We also compared the developmental appearance of the cGMP-gated cation channel to that of other phototransduction proteins and developmental markers. RET-P2, a monoclonal antibody recognizing the 39-kDa rds/peripherin disc protein, first labeled outer segments at PN7, coincident with cGMP-gated cation channel expression. Double labeling of the same section of PN7 rat retina with RET-P2 and R309 (a polyclonal antiserum against the rod cGMP-gated cation channel) revealed identical patterns of labelling. Similarly, double labeling with RCNGC1-7H2 and an antibody against the rod cGMP-phosphodiesterase gave coincident labeling, suggesting coordinate expression mechanisms of phototransduction proteins with each other and with outer segment structural proteins.

Development and growth of V1 begins during embryogenesis and continues postnatally. The growth of V1 has direct implications on the organization of features such as the retinotopic map and the pattern of visual cortical columns. We have examined the postnatal growth and two-dimensional shape of V1 in macaque monkeys, cats, and rats. The perimeter, area, and anterior–posterior length of V1 were measured from unfolded and flattened sections from neonatal and adult animals from each of these species. Although there were substantial differences in the overall amount of postnatal growth, from 18% in macaque monkeys to more than 100% in cats, in all three species the shape of V1 did not change during development. Thus, growth of the mammalian visual cortex is well described as an isotropic expansion, so the layout of the global features, such as the arrangement of ocular dominance columns and the retinotopic map, does not need to change during development. Furthermore, quantification of the shape confirms the observations that there is a similar, egg-like oval shape to the visual cortex of these mammalian species.

Controversy over the relationship between ocular pigmentation and absolute dark-adapted light sensitivity has persisted for over two decades. Previous electrophysiological experiments in hypopigmented mammals (mice, rats, rabbits) show increased thresholds in the dark-adapted state proportional to the deficit in ocular melanin. Animals with the least amount of ocular melanin have the most elevated thresholds. Dark-adapted thresholds in hypopigmented mice show similar threshold elevations in behavioral tests. The present study extends these findings to show that a specific increase in ocular pigmentation results in the converse effect, lowered absolute dark-adapted thresholds. The increase in ocular melanin was accomplished by keeping Himalayan mice in the cold (4°C) for 6 weeks. Himalayan mice (C57BL/6J cH/cH) were compared to black mice (C57BL/6J +/+) and albino mice (C57BL/6J c2J/c2J) after 6 weeks at either 4°C or 20°C in 12-h cycling light (<1 cd/m2). The Himalayan mice that were kept in the cold exhibited a 44% increase in ocular melanin compared to Himalayan mice kept at room temperature. Cold rearing did not effect ocular melanin or visual thresholds in control animals (black mice = 10−5.9 cd/m2 and albino mice = 10−4.4 cd/m2). In contrast, the Himalayan mice maintained at 4°C had thresholds of 10−5.7 cd/m2 compared to 10−5.1 cd/m2 for Himalayan mice kept at 20°C. This represents compelling evidence of a direct relationship between ocular melanin concentration and absolute dark-adapted light sensitivity.

We have recently developed a behavioral assay, based on the escape response of fish to a threatening object, to analyze quantitatively the visual sensitivity of zebrafish. During the course of dark adaptation, we measure the threshold light intensity required to evoke an escape response. Under a normal light–dark (LD) cycle, thresholds for both the cone and rod systems are considerably lower in late afternoon hours than in early morning hours. Over a period of 24 h, zebrafish are most sensitive to visual stimuli prior to light off and least sensitive prior to light on. Under conditions of constant illumination, this rhythm of visual sensitivity persists for several days but is gradually lost. In constant light (LL), the rhythm persists 1–2 days; thereafter, visual thresholds at all times of the day converge at a level similar to thresholds measured in late afternoon hours in control animals. In constant darkness (DD), the rhythm persists at least 5 days; thereafter, it dampens to a level about a half-log unit less sensitive to that measured in the late afternoon hours in control animals. These data suggest that visual sensitivity in zebrafish is regulated by an endogenous circadian clock which functions to decrease the visual sensitivity.

We measured the photopic spectral sensitivity of multiunit activity in the torus longitudinalis and optic tectum of goldfish. Since negative contrast stimuli are most effective for exciting torus longitudinalis, spiking activity was evoked by the shadow of a disc moving through a monochromatic light beam projected upon a screen. The amount of activity evoked in torus longitudinalis generally increased with the monochromatic stimulus radiance at the same rate for all wavelengths, indicating a univariant response. Spiking activity in tectum, however, increased at different rates across the spectrum, indicating color-dependent responses. The action spectra for torus longitudinalis were all similar and relatively flat as expected of a homogeneous, broad-band luminance processing system, and about 1 log unit more sensitive than the tectal action spectra. The latter generally displayed sharp peaks and dips in sensitivity indicative of opponent processing.

Cyclic GMP has been shown in recent years to directly activate ion channels in bipolar and ganglion cells, and to indirectly regulate coupling between horizontal cells, and between bipolar and amacrine cells. In all of these cases, the effects of cyclic GMP are mimicked by nitric oxide. An increase in calcium concentration stimulates the production of nitric oxide by neuronal and endothelial forms of nitric oxide synthase, which in turn activates soluble guanylate cyclases, enhancing the synthesis of cyclic GMP. Though some effects of nitric oxide do not involve cyclic GMP, the nitric oxide-cyclic GMP cascade is well recognized as a signaling mechanism in brain and other tissues. The widespread occurrence of nitric oxide/cyclic GMP-regulated ion channel activity in retinal neurons raises the possibility that nitric-oxide-sensitive soluble guanylate cyclases play an important role in cell–cell communication, and possibly, synaptic transmission. Immunohistochemical studies have indicated the presence of soluble guanylate cyclase in retinal synaptic layers, but such studies are not suitable for determination of the density or quantitative subcellular distribution of the enzyme. Microanalytical methods involving microdissection of frozen retina also showed the presence of cyclase activity in retinal plexiform layers but these methods did not permit distinction between nitric oxide-sensitive and insensitive cyclases. In this study, we fractionated retinal homogenate into the cytosolic and synaptosomal fractions and investigated the specific activity and distribution of soluble guanylate cyclase and nitric oxide synthase. The results show that both enzymes are present in the synaptosomal fractions derived from inner and outer plexiform layers. The synaptosomal fraction derived from inner retina was highly enriched in cyclase activity. Nitric oxide synthase activity was also higher in the inner than outer retinal synaptosomal fraction. The results suggest that the nitric oxide-cyclic GMP system is operational in both synaptic layers of retina and that it may play a more significant role in the inner retina.

The sudden displacement of the retinal image during a saccade raises the visual threshold of human observers to foveal stimuli. The fall in visual sensitivity observed during this phenomenon, known as saccadic suppression, seems to occur very early in the visual processing chain. The lateral geniculate nucleus (LGN) is a likely locus for the multiple retinal and extraretinal interactions occurring during saccadic eye movements, therefore we used the responses of relay cells of adult cats to simulate a pychophysical experiment. We first measured the responses of X and Y relay cells (27 X and 13 Y) to central spots of optimal size and different contrasts. The spots were presented either alone or time locked with the rapid movement of a large, high-contrast peripheral pattern, referred to as shift. We measured the percentage of trials on which the relay cell fired more spikes when the spot (contrast: 0.03–1.0) was present than when it was absent. In experiments with human observers the task was to indicate, by a keypress, which of two otherwise identical temporal intervals contained the spot. The shift reduces the sensitivity (raises the contrast threshold) of neurones in the cat relay cells to brief, stationary targets presented to the receptive-field center. The suppression of visual sensitivity is significantly greater in Y cells than in X cells (average sensitivity ratios 5.6 ± 5.4 in Y cells, 1.59 ± 0.9 in X cells: P < 0.001, U test). The shift also reduces the sensitivity of human observers to the same target. This suggests that the LGN is a potential locus for the modulation of visual responses that leads to saccadic suppression.

Nitric oxide (NO) acts as a neuronal messenger which activates soluble guanylyl cyclase (SGC) in neighboring cells and produces a wide range of physiological effects in the central nervous system (CNS). Using immunocytochemical and histochemical stains, we have characterized the NO/SGC system in the rabbit retina and to a lesser extent, in monkey retina. Based on staining patterns observed with an antibody to nitric oxide synthase (NOS) type I and a histochemical marker for NADPH diaphorase, a metabolic intermediate required for NOS activity, three major classes of neurons appear to generate NO in the rabbit retina. These include two subclasses of sparsely distributed wide field amacrine cells, rod and cone photoreceptors, and a subpopulation of ganglion cells. Equivalent cell populations were labled in monkey retina. An antibody to SGC (tested only in rabbit retina), labeled large arrays of cone photoreceptors in the outer nuclear layer, both amacrine and bipolar cells in the inner nuclear layer (INL), as well as populations of neurons in the ganglion cell layer. These data suggest that the ability to generate NO is restricted to relatively few neurons in the inner retina and to photoreceptor cells in the outer retina; while presumptive target cells, containing pools of SGC, are widespread and form contiguous fields across the inner and outer nuclear layers (ONL) as well as the ganglion cell layer.

Previous physiological studies have shown that serotonin (5-HT) reciprocally modulates ON and OFF channels in the mammalian retina. This study was undertaken to determine if the serotoninergic system is exclusively associated with the rod pathway. We tested drugs specific to 5-HT3 receptor, a serotonin-gated ion channel, in both dark- and light-adapted retina. Consistent with previous studies, we demonstrated that 5-HT3 receptors modulate the light-evoked responses of ganglion cells in the dark-adapted state. Moreover, we have extended these prior studies and shown that activation of the 5-HT3 receptor is capable of completely blocking the light-evoked response of OFF-center cells whereas inactivation of the 5-HT3 receptor is capable of completely blocking the light-evoked responses of ON-center cells. In contrast, in light-adapted retinae, serotonin agents failed to have any effect on retinal processing. These data suggest that the serotoninergic system in retina is (1) specifically associated with rod-related pathways; and (2) exerts a powerful modulatory force over information transfer in the retina. Together these observations suggests that serotonin plays an important physiological role in modulating retinal processing.

The presence of a commissure connecting the two superior colliculi suggests they do not act independently, but the function of the tectotectal connection has never been firmly identified. To develop a better understanding of this commissural system, the present study determined the distribution and morphology of tectotectal neurons in the cat and macaque monkey, two animals with well-studied, but different orienting strategies. First, we compared the distribution of tectotectal cells retrogradely labeled following WGA-HRP injections into the contralateral superior colliculus. In monkeys, labeled tectotectal cells were found in all layers, but were concentrated in the intermediate gray layer (75%), particularly dorsally, and the adjacent optic layer (12%). Tectotectal cells were distributed throughout nearly the entire rostrocaudal extent of the colliculus. In cats, tectotectal cells were found in all the layers beneath the superficial gray, but the intermediate gray layer contained the greatest concentration (56%). Labeled cells were almost exclusively located in the rostral half of the cat superior colliculus, in contrast to the monkey distribution. In the context of the representation of visuomotor space in the colliculus, the distribution of monkey and cat tectotectal cells suggests a correspondence with oculomotor range. So these neurons may be involved in directing orienting movements performed within the oculomotor range. The somatodendritic morphology of tectotectal cells in these two species was revealed by homogeneous retrograde labeling from injections of biocytin or biotinylated dextran amine into the contralateral colliculus. The cell classes contributing to this pathway are fairly consistent across the two species. A variety of neuronal morphologies were observed, so there is no single tectotectal cell type. Instead, cell types similar to those found in each layer, excepting the largest neurons, were present among tectotectal cells. This suggests that a sample of each layer's output is sent to the contralateral colliculus.

Although differences in visual pigments between developmental stages of the European eel are well known, the expected differences in spectral sensitivity have not been demonstrated at the electrophysiological level. In fact, one past electroretinographic study led to the conclusion that in eels there is no change in scotopic sensitivity, with increasing sexual maturity. In the present experiments, electroretinograms (ERGs) were recorded from in situ eyecups of immobilized eels Anguilla anguilla (L.) caught in coastal running waters. It was shown that the ERG b-wave is as good an indicator of spectral sensitivity as the unmasked late receptor potential (LRP) which directly reflects the responsiveness of photoreceptors. Complete spectral-sensitivity curves, based on b-wave thresholds and on thresholds of LRP subsequently isolated by means of sodium iodate, have been obtained in the same eel. Using fitted amplitude-log intensity functions for threshold calculation, and two models for computer-assisted fitting of spectral-sensitivity curves, significant differences in λmax were found between yellow and silver developmental stages of the eel, identified by ocular index measurements.

Parvocellular (P-) and magnocellular (M-) cells in the marmoset LGN can receive prominent rod input up to relatively high illuminance levels (Kremers et al., 1997b). In the present paper, we quantify rod and cone input strengths under different retinal illuminance levels. The stimulus was based on the so-called “silent substitution” method. The activities of P- and M-cells of dichromatic animals were recorded extracellularly. We were able to adequately describe the response amplitudes and phases by a vector summation of rod and cone signals. At low retinal illuminance levels, the cells' responses were determined by rod and cone inputs. With increasing illuminances the strength of the cone input increased relative to the rod strength. But, we often found significant rod inputs up to illuminances equivalent to 700 td in the human eye or more. Rod input strength was more pronounced in cells with receptive fields at large retinal eccentricities. The phase differences between rod and cone inputs suggest that the rod signals lag about 45 ms behind the cone signals.

In the mammalian retina, neuronal nitric oxide synthase (NOS) is mainly localized in subpopulations of amacrine cells. One function of nitric oxide (NO) is to stimulate soluble guanylate cyclases which in turn synthesize cGMP. We used an antibody specific for cGMP to demonstrate cGMP-like immunoreactivity (cG-IR) in bovine, rat, and rabbit retinae and investigated the effects on cGMP levels of both exogenously applied NO and of endogenously released NO. We found that cGMP levels in inner and outer retina were controlled in opposite ways. In the presence of the NO-donors SNP, SIN-1 or SNAP, cG-IR was prominent in neurons of the inner retina, mainly in cone bipolar cells, some amacrine and ganglion cells. Retinae incubated in IBMX showed weak cG-IR in bipolar cells. Glutamate increased cG-IR in the inner retina, presumably by stimulating endogenous NO release, whereas NOS inhibitors or GABA and glycine decreased cG-IR in bipolar cells by reducing NO release. In somata, inner segments and spherules of rod photoreceptors the situation was reversed. cG-IR was undetectable in the presence of NO-donors or glutamate, was moderate in IBMX-treated retinae, but increased strongly in the presence of NOS inhibitors or GABA/glycine. We conclude that NO is released endogenously in the retina. In the presence of NO, cGMP levels are increased in neurons of the inner retina, but are decreased in rods.

When a temporally fluctuating background is rapidly modulated (e.g. 30 Hz), the threshold variation of a superimposed flash (the probe) is approximately sinusoidal and in phase with the stimulus. But, with low rates of sinusoidal modulation (e.g. 1 Hz), the threshold variation is distinctly nonsinusoidal in shape. The bases of these aspects of the data, as well as an unmodulated, dc, threshold elevation, are poorly understood. Here 30-Hz and 1-Hz conditions are simulated using a new model of light adaptation (Wilson, 1997). By assuming that the OFF pathway is twice as sensitive as the ON pathway, the model correctly captured the key aspects of both conditions. The results suggest that the 1-Hz data are mediated by a mixture of ON and OFF pathways while the 30-Hz data are largely mediated by the OFF pathway. The probe thresholds on the 30-Hz background appear approximately sinusoidal and approximately in phase with the background stimulus. A number of factors contribute to this deceptively simple observation.

Both local cell–cell interactions and lineage bias have roles in determining the different retina cell phenotypes. In this study, subpopulations of amacrine cells that dually express GABA or serotonin (5-HT) and dopamine (DA) are identified in the early Xenopus tadpole (stages 42–48) retina. GABA is first detected by immunocytochemistry in amacrine cells at stage 35/36, 5-HT at stage 39, and DA at stage 41. As the number of these subtypes of amacrine cells increases by differentiation, a subset of them transiently express two neurotransmitters. GABA/DA double-labeled amacrine cells are detected first at stage 42, at which time they constitute 52% of the DA-containing population; this percentage decreases to only 3% by stage 48. 5-HT/DA amacrine cells are detected only at stage 44, constituting about 20% of the DA-containing cells and 4% of the small-dim 5-HT-containing cells. Regional location does not differentially affect the differentiation of these three types of amacrine cells (DA only, GABA/DA, and 5-HT/DA cells); each type is found more in the anterior and dorsal than the posterior and ventral quadrants, and their overall distribution patterns are statistically indistinguishable. However, these subtypes of amacrine cells reside in different sublamina of the inner nuclear layer. DA-only amacrine cells are located predominantly in the inner sublayer of the 2–3 cell thick amacrine cell layer, closest to the inner plexiform and the ganglion cell layers. Both types of double-labeled cells are located mostly in the outer sublayer of the amacrine cell layer, closest to other interneurons in the inner nuclear layer. This distinct sublaminar location of different neurotransmitter phenotypes suggests that local laminar cues influence the coexpression of neurotransmitters in amacrine cells.

Sharp electrodes were used to record light-evoked postsynaptic potentials (PSPs) from neurons in turtle visual cortex in an in vitro preparation of the geniculocortical pathway. Neurons were placed into four groups based on the firing patterns produced by intracellular current injections: regular spiking (RS), fast spiking (FS), intrinsic bursting (IB), and chattering (CH) cells. RS cells have been shown to be pyramidal cells while FS cells are typically interneurons. Light stimuli were diffuse, 1-s flashes of 640-nm light with intensities (I) varying from 0 to 104 photons μm−2 s−1. The response (R) in each case was the maximal amplitude of the light-evoked depolarizing PSP. Cells of all four types showed sigmoidal intensity–response (IR) functions with a linear rising phase for stimuli above the intensity threshold followed by saturation at high light intensities. Responses at high intensities were variable and some cells showed indications of supersaturation. Light-evoked PSPs had longer latencies and times-to-peak response in RS cells than they did in FS cells. RS cells fired action potentials as much as 200 ms later than did FS cells. Since responses recorded in RS cells at light intensities just above threshold are unlikely to involve contributions from other pyramidal cells, these data indicate that the geniculocortical or feedforward pathway to pyramidal cells has a high gain. The fact that FS cells fire well before RS cells suggests that feedforward inhibition plays a role in controlling the gain of the geniculocortical pathway.