A labeled ON parasol ganglion cell from a macaque retina was analyzed in serial, ultrathin sections. It received 13% of its input from diffuse bipolar cells. These directed a large proportion of their output to amacrine cells but received a relatively small proportion of their amacrine cell input via feedback synapses. In these respects, they were similar to the DB3 bipolar cells that make synapses onto OFF parasol cells. Bipolar cell axons that contacted the ON parasol cell in stratum 4 of the inner plexiform layer always made synapses onto the dendrite, and therefore, the number of bipolar cell synapses onto these ganglion cells could be estimated reliably by light microscopy in the future. Amacrine cells provided the majority of inputs to the ON parasol cell. Only a few of the presynaptic amacrine cell processes received inputs from the same bipolar cells as the parasol cells, and most of the presynaptic amacrine cell processes did not receive any inputs at all within the series. These findings suggest that most of the inhibitory input to the ON parasol cell originates from other areas of the retina. Amacrine cells presynaptic to the parasol ganglion cell interacted very infrequently with other neurons in the circuit, and therefore, they would be expected to act independently, for the most part.

Horizontal cells are second-order retinal neurons that play a key role in spatial information processing. In some cold-blooded vertebrates such as turtles, a subtype of these cells, the chromaticity horizontal cells exhibit color-opponent responses and therefore are considered to be important also for color information processing. To reveal spatial and color interactions, the receptive-field properties of Red/Green and Yellow/Blue chromaticity horizontal cells in the retina of the turtle Mauremys caspica were studied by intracellular recordings from the everted eyecup preparation. We found that the polarity of the photoresponses depended not only upon the wavelength and intensity of the stimulus, but also upon its spatial configuration. Thus, a hyperpolarizing photoresponse that was elicited by full-field stimulation with bright light of wavelength close to the “neutral” one was reversed in polarity to a pure depolarizing one when a small spot or a thin annular pattern were used for stimulation. This finding could not be explained either by different balances between depolarizing and hyperpolarizing inputs to different cells or by stray light that effectively reduced the light intensity in the center of the small spot. Rather, it was found that the depolarizing and hyperpolarizing components were characterized by different receptive-field size and that these differences could account for the dependency of response polarity upon the spatial pattern of the stimulus. These findings indicate that color information processing in turtle C-type horizontal cells is a complex process that depends upon the wavelength and intensity of the light stimulus as well as upon its spatial properties.

Temperature effects on spectral properties of the two types of rod photoreceptors in toad retina, “red” and “green” rods, were studied in the range 0–38°C. Absorbance spectra of the visual pigments were recorded by single-cell microspectrophotometry (MSP) and spectral sensitivities of red rods were measured by electroretinogram (ERG) recording across the isolated retina. The red-rod visual pigment is a usual rhodopsin (λmax = 503.6 nm and 502.3 nm at room temperature (21°C) in, respectively, Bufo marinus and Bufo bufo), that of green rods (λmax = 432.6 nm in Bufo marinus) belongs to the “blue” cone pigment family. In red rods, λmax depended inversely and monotonically on temperature, shifting by −2.3 nm when temperature was raised from 0°C to 38°C. Green-rod λmax showed no measurable dependence on temperature. In red rods, warming caused a relative increase of sensitivity in the long-wavelength range. This effect can be used for estimating the energy needed for photoexcitation, giving Ea = 44.3 ± 0.6 kcal/mol for Bufo marinus rhodopsin and 48.8 ± 0.5 kcal/mol for Bufo bufo rhodopsin. The values are significantly different (P < 0.001), although the two rhodopsins have very similar absorption spectra and thermal isomerization rates. Our recording techniques did not allow measurement of the corresponding effect at long wavelengths in green rods. Although spectral effects of temperature changes in the physiological range are small and of little significance for visual function, they reveal information about the energy states and different spectral tuning mechanisms of the visual pigments.

Neither discrete peripheral retinal lesions nor the normal optic disk produces obvious holes in one's percept of the world because the visual brain appears to perceptually “fill in” these blind spots. Where in the visual brain or how this filling in occurs is not well understood. A prevailing hypothesis states that topographic map of visual cortex reorganizes after retinal lesions, which “sews up” the hole in the topographic map representing the deprived area of cortex (cortical scotoma) and may lead to perceptual filling in. Since the map reorganization does not typically occur unless retinotopically matched lesions are made in both eyes, we investigated the conditions in which monocular retinal lesions can induce comparable map reorganization. We found that following monocular retinal lesions, deprived neurons in cat area 17 can acquire new receptive fields if the lesion occurred relatively early in life (8 weeks of age) and the lesioned cats experienced a substantial period of recovery (>3 years). Quantitative determination of the monocular and binocular response properties of reactivated units indicated that responses to the lesioned eye for such neurons were remarkably robust, and that the receptive-field properties for the two eyes were generally similar. Moreover, excitatory or inhibitory binocular interactions were found in the majority of experimental units when the two eyes were activated together. These results are consistent with the hypothesis that map reorganization after monocular retinal lesions require experience-dependent plasticity and may be involved in the perceptual filling in of blind spots due to retinal lesions early in life.

We performed a nonradioactive in situ hybridization histochemistry (ISH) study of the lateral geniculate nucleus (LGN) and the primary visual area (area 17) of the macaque monkey to investigate mRNA expression of the myristoylated alanine-rich C-kinase substrate (MARCKS), a major protein kinase C (PKC) substrate. In the LGN, intense hybridization signals were observed in both magnocellular neurons (layers 1 and 2) and parvocellular neurons (layers 3 to 6). Double labeling using ISH and immunofluorescence revealed that MARCKS mRNA was coexpressed with the α-subunit of type II calcium/calmodulin-dependent protein kinase, indicating that MARCKS mRNA is also expressed in koniocellular neurons in the LGN. GABA-immunoreactive neurons in the LGN did not contain MARCKS mRNA, indicating that MARCKS mRNA is not expressed in inhibitory interneurons. The signals were generally weak in area 17, and intense signals were restricted to large neurons in layers IVB, V, and VI. GABA-immunoreactive neurons in layers II–VI of area 17 did not contain MARCKS mRNA. Double-label ISH revealed that MARCKS mRNA was coexpressed with mRNA of GAP-43, another PKC substrate, in neurons of both the LGN and area 17. To determine whether the expression of MARCKS mRNA is regulated by retinal activity, we performed ISH in the LGN and area 17 of monkeys deprived of monocular visual input by tetrodotoxin. After monocular deprivation for 5 to 30 days, MARCKS mRNA was down-regulated in the LGN, but not in area 17. These results suggest that MARCKS mediates the activity-dependent changes in the excitatory relay neurons in the LGN.

The retinas of anchovies have two unique photoreceptor types: “bifid” and “long” cones (Fineran & Nicol, 1976). The outer segments of these cells contain multiple layers of membranes (lamellae) oriented longitudinally (axially). This orientation is distinct from that in all other vertebrate rods and cones, where the lamellae are stacked transversely with their planes perpendicular to the incident light path. Although the common arrangement provides optimal absorption for normally incident light rays, it is also insensitive to the rays' direction of vibration (i.e. their polarization). In contrast, the two mutually perpendicular sets of axially oriented lamellae segregated into bifid and long cones could function as the principal analyzers for linearly polarized light, as previously hypothesized (Fineran & Nicol, 1976, 1978). Here, we report on a microspectrophotometric study that shows (1) the presence of two spectrally distinct visual pigments in the three photoreceptor types of the bay anchovy retina; these are typical vertebrate pigments in that they bleach, when exposed to light, and have absorption spectra like all other vitamin A1-based visual pigments; (2) that the rods and cones exhibit dichroic absorption of light in accordance with their lamellar orientation, and (3) that the two cone types of the retina contain a spectrally indistinguishable pigment with peak absorbance (λmax) around 540 nm, while the rods contain a rhodopsin-like pigment with λmax near 500 nm. Compared to other vertebrates, anchovies are remarkable for using a monochromatic cone system with unusual specializations supportive of a polarization detection system.

Using autoradiographic and biochemical techniques, we studied the sodium-dependent forward and reverse transport of the neurotransmitter histamine in an arthropod photoreceptor in order to test whether the transporter plays a central role in visual signal transfer at this synapse. In particular, we asked whether the histamine transporter might be the important factor in synaptic adaptation, the process by which the operating range of the synapse adapts to increasing depolarizations of the photoreceptor in increasing background light. Drugs known from electrophysiological observations to interfere with synaptic adaptation blocked the uptake of [3H]histamine into photoreceptors. These drugs also blocked the sodium (Na)-triggered efflux of [3H]histamine, previously loaded into photoreceptors, via the histamine transporter. Several lines of evidence showed that efflux of [3H]histamine did not occur via calcium-dependent exocytosis. First, efflux occurred when the preparation was bathed in calcium (Ca)-free/EGTA salines or in cobalt (Co)-containing salines. Even more importantly, efflux could be elicited from axons, whose membranes must contain the transporter protein since they take up [3H]histamine independently from the presynaptic terminals. Since both adaptation and the histamine transporter are blocked by the same agents, the transporter may underlie adaptation by maintaining the cleft histamine concentration in a particular range independent of light intensity. We also characterized the transporter further and found that it is partially dependent on chloride ions, and that neither [3H]norepinephrine nor [3H]dopamine are transported (at 20 μM), adding to evidence that the transporter is highly selective for histamine.

There is considerable controversy over the existence of orientation and direction sensitivity in lateral geniculate nucleus (LGN) neurons. Claims for the existence of these properties often were based upon data from cells tested well beyond their peak spatial frequencies. The goals of the present study were to examine the degree of orientation and direction sensitivity of LGN cells when tested at their peak spatial and temporal frequencies and to compare the tuning properties of these subcortical neurons with those of visual cortex. For this investigation, we used conventional extracellular recording to study orientation and direction sensitivities of owl monkey LGN cells by stimulating cells with drifting sinusoidal gratings at peak temporal frequencies, peak or higher spatial frequencies, and moderate contrast. A total of 110 LGN cells (32 koniocellular cells, 34 magnocellular cells, and 44 parvocellular cells) with eccentricities ranging from 2.6 deg to 27.5 deg were examined. Using the peak spatial and temporal frequencies for each cell, 41.8% of the LGN cells were found to be sensitive to orientation and 19.1% were direction sensitive. The degree of bias for orientation and direction did not vary with eccentricity or with cell class. Orientation sensitivity did, however, increase, and in some cases orientation preferences changed, at higher spatial frequencies. Increasing spatial frequency had no consistent effect on direction sensitivity. Compared to cortical cell orientation tuning, the prevalence and strength of LGN cell orientation and direction sensitivity are weak. Nevertheless, the high percentage of LGN cells sensitive to orientation even at peak spatial and temporal frequencies reinforces the view that subcortical biases could, in combination with activity-dependent cortical mechanisms and/or cortical inhibitory mechanisms, account for the much narrower orientation and direction tuning seen in visual cortex.

Contrast adaptation occurs in both the retina and the cortex. Defining its spatial dependence is crucial for understanding its potential roles. We thus asked to what degree contrast adaptation depends on spatial frequency, including cross-adaptation. Measuring the pattern electroretinogram (PERG) and the visual evoked potential (VEP) allowed separating retinal and cortical contributions. In ten subjects we recorded simultaneous PERGs and VEPs. Test stimuli were sinusoidal gratings of 98% contrast with spatial frequencies of 0.5 or 5.0 cpd, phase reversing at 17 reversals/s. Adaptation was controlled by prolonged presentation of these test stimuli or homogenous gray fields of the same luminance. When adaptation and test frequency were identical, we observed significant contrast adaptation only at 5 cpd: an amplitude reduction in the PERG (−22%) and VEP (−58%), and an effective reduction of latency in the PERG (−0.95 ms). When adapting at 5 cpd and testing at 0.5 cpd, the opposite effect was observed: enhancement of VEP amplitude by +26% and increase in effective PERG latency by +1.35 ms. When adapting at 0.5 cpd and testing at 5 cpd, there was no significant amplitude change in PERG and VEP, but a small effective PERG latency increase of +0.65 ms. The 0.5-cpd channel was not adapted by spatial frequencies of 0.5 cpd. The adaptability of the 5-cpd channel may mediate improved detail recognition after prolonged blur. The existence of both adaptable and nonadaptable mechanisms in the retina allows for the possibility that by comparing the adaptational state of spatial-frequency channels the retina can discern between overall low contrast and defocus in emmetropization control.

The mesencephalic tectum plays a prominent role in integrating both visual and multimodal sensory information essential for normal behavior in amphibians. Activity in the mesencephalic tectum is thought to be modulated by the influence of distinct neurochemical inputs, including the catecholaminergic and the cholinergic systems. In the present study, we have investigated the distribution and the origin of the catecholaminergic innervation of the mesencephalic tectum in two amphibian species, the anuran Rana perezi and the urodele Pleurodeles waltl. Immunohistochemistry for dopamine and two enzymes required for the synthesis of catecholamines, tyrosine hydroxylase (TH) and dopamine β-hydroxylase (DBH), revealed a complex pattern of catecholaminergic (CA) innervation in the anuran and urodele mesencephalic tectum. Dopaminergic fibers were primarily present in deep tectal layers, whereas noradrenergic (DBH immunoreactive) fibers predominated in superficial layers. Catecholaminergic cell bodies were never observed within the tectum. To determine the origin of this innervation, applications of retrograde tracers into the optic tectum were combined with immunohistochemistry for TH. Results from these experiments demonstrate that dopaminergic neurons in the suprachiasmatic and juxtacommissural nuclei (in Rana) or in the nucleus pretectalis (in Pleurodeles), together with noradrenergic cells of the locus coeruleus, are the sources of CA input to the amphibian mesencephalic tectum. The present results suggest that similar CA modulatory inputs are present in the mesecencephalic tectum of both anurans and urodeles.

The aim of this study was to assess the plasticity of human voluntary fixational eye movement control in relation to visual experience/chronic visual deprivation. Twelve blind adults participated (self-reported vision ≤ light perception in each eye; age range = 23–56 years; visual experience range = 0–28 years; blindness duration range = 6–55 years). Infrared-based recordings of horizontal eye movements were made before, during, and immediately after three 30-s periods of auditory ocular motor feedback, while participants were instructed to look straight ahead and keep their eyes as steady as possible. Percent change in horizontal displacement of the eye during and after feedback was compared with the no-feedback baseline. Eleven of the 12 individuals demonstrated feedback-mediated increase in eye stability, which improved as a function of visual experience. Improved eye stability was inversely related to duration of blindness. Clearly, blind adults can use nonvisual external feedback to stabilize gaze. Thus, the fixational subsystem can exhibit improved voluntary control despite chronic visual deprivation. Possible cortical and subcortical mechanisms are discussed.

We examined intraretinal [H+] in the intact retina of anesthetized cats using H+-sensitive microelectrodes to obtain spatial profiles of extracellular [H+]. One H+ is produced when an anaerobically generated ATP is utilized. We theorized that H+ production directly reflects anaerobic glucose consumption. From the choroid (pH ∼7.40), [H+]o steadily increased to a maximum concentration in the proximal portion of the outer nuclear layer (pH∼7.20). The shape of the profile was always concave down, indicating that a net production of H+ occurred across the avascular outer retina. A three-layer diffusion model of the outer retina was developed and fitted to the data to quantify photoreceptor H+ extrusion into the extracellular space (QOR-H+). It was determined that the outer segment (OS) layer had negligible H+ extrusion. The data were then refitted to a special three-layer model in which the OS layer QH+ was set equal to zero, but in which the inner segments and outer nuclear layer produced H+. The resulting QOR-H+ was several orders of magnitude lower than previous measurements of QOR-lactate, which were based on choroidal mass balances of lactate.

The color vision of Old World primates and humans uses two cone-opponent systems; one differences the outputs of L and M cones forming a red–green (RG) system, and the other differences S cones with a combination of L and M cones forming a blue–yellow (BY) system. In this paper, we show that in human vision these two systems have a differential distribution across the visual field. Cone contrast sensitivities for sine-wave grating stimuli (smoothly enveloped in space and time) were measured for the two color systems (RG & BY) and the achromatic (Ach) system at a range of eccentricities in the nasal field (0–25 deg). We spatially scaled our stimuli independently for each system (RG, BY, & Ach) in order to activate that system optimally at each eccentricity. This controlled for any differential variations in spatial scale with eccentricity and provided a comparison between the three systems under equivalent conditions. We find that while red–green cone opponency has a steep decline away from the fovea, the loss in blue–yellow cone opponency is more gradual, showing a similar loss to that found for achromatic vision. Thus only red–green opponency, and not blue–yellow opponency, can be considered a foveal specialization of primate vision with an overrepresentation at the fovea. In addition, statistical calculations of the level of chance cone opponency in the two systems indicate that selective S cone connections to postreceptoral neurons are essential to maintain peripheral blue–yellow sensitivity in human vision. In the red–green system, an assumption of cone selectivity is not required to account for losses in peripheral sensitivity. Overall, these results provide behavioral evidence for functionally distinct neuro-architectural origins of the two color systems in human vision, supporting recent physiological results in primates.

Modulation of responses elicited by moving bars within the classical receptive fields (CRF) of cat area 17 neurons were studied as a function of the direction and velocity of drifting gratings in the surrounds. Several different types of modulation were observed; collectively, the responses of most cells, both simple and complex, were strongly modulated by surround motion. None of these cells appear to signal relative velocity between the CRF and its surround. The gain and spatiotemporal structure of the CRF mechanism were estimated using contrast-response functions and reverse correlation with spatiotemporal ternary white noise, respectively. These measurements were made in the presence of surround gratings shown to significantly modify responses elicited from the CRF. In all cases, the gain of the CRF mechanism was driven up or down relative to controls but the receptive-field structure did not change in any way. We conclude that neurons in cat area 17 act like scalable filters, meaning that their gains can be adjusted by stimuli in the surrounds without altering the properties of the CRF. This was verified by showing that velocity tuning curves were also unmodified by stimuli in the surround that did change the gain. Based in part on these data, we discuss the notion that primary visual cortex makes use of a double-opponent mechanism for the representation of local discontinuities in motion and orientation.

Unlike simple cells, complex cells of area 18 give a directionally selective response to motion of random textures, indicating that they may play a special role in motion detection. We therefore investigated how texture motion, and especially its velocity, is represented by area 18 complex cells. Do these cells have separable spatial and temporal tunings or are these nonseparable? To answer this question, we measured responses to moving random pixel arrays as a function of both pixel size and velocity, for a set of 63 directionally selective complex cells. Complex cells generally responded to a fairly wide range of pixel sizes and velocities. Variations in pixel size of the random pixel array only caused minor changes in the cells' preferred velocity. For nearly all cells the data much better fitted a model in which pixel size and velocity act separately, than a model in which pixel size and velocity interact so as to keep temporal-frequency sensitivity constant. Our conclusion is that the studied population of special complex cells in area 18 are true motion detectors, rather than temporal-frequency tuned neurons.

A possible effect of dopamine on the temporal resolution of goldfish was investigated in a behavioral, two-alternative, forced-choice procedure. Flicker fusion frequency (FFF) was measured before and after bilateral intravitreal injections of D1- or D2-dopamine receptor (D1-/D2-R) antagonists, or after depletion of retinal dopamine by bilateral intravitreal injections of the dopaminergic neurotoxin 6-hydroxydopamine (6-OHDA). Prior to drug injections, fish achieved FFFs of 33–39 Hz. A D1-R antagonist, SCH 23390, reduced FFF by about 12% (P > 0.1), whereas a D2 antagonist, sulpiride, reduced the relative FFF by 25% (P < 0.03). Depletion of retinal dopamine with 6-OHDA induced a gradual reduction in the FFF to a maximal reduction of 50% (P < 0.001) at 2 weeks postinjection. There was recovery to control levels after 3–4 weeks postinjection. The recovery of FFF, at least in one animal, was due to the return of retinal dopamine because FFF could be reduced by intravitreal injections of sulpiride during the recovery phase. These experiments demonstrate that retinal dopamine, particularly acting on D2-R, is important for photopic temporal resolution.

Ganglion cells in an isolated wholemount preparation of the rat retina were labeled using the “DiOlistic” labeling method (Gan et al., 2000) and were classified according to their morphological properties. Tungsten particles coated with a lipophilic dye (DiI) were propelled into the wholemount retina using a gene gun. When a dye-coated particle contacted the cell membrane, the entire cell was labeled. The ganglion cells were classified into four types based on their soma size, dendritic-field size, branching pattern, and level of stratification. Broadly monostratified cells were classified into three types: RGA cells (large soma, large dendritic field); RGB cells (small- to medium-sized soma, small- to medium-sized dendritic field); and RGC cells (small- to medium-sized soma, medium-to-large dendritic field). Bistratified cells were classified as RGD. Several subtypes were identified within each ganglion cell group. A number of new subtypes were discovered and added into the existing catalog, among them were two types of bistratified cells. This study therefore represents the most complete morphological classification of rat retinal ganglion cells available to date.

Axonal regeneration of retinal ganglion cells (RGCs) into a normal or pre-degenerated peripheral nerve graft after an optic nerve pre-lesion was investigated. A pre-lesion performed 1–2 weeks before a second lesion has been shown to enhance axonal regeneration in peripheral nerves (PN) but not in optic nerves (ON) in mammals. The lack of such a beneficial pre-lesion effect may be due to the long delay (1–6 weeks) between the two lesions since RGCs and their axons degenerate rapidly 1–2 weeks following axotomy in adult rodents. The present study examined the effects of the proximal and distal ON pre-lesions with a shortened delay (0–8 days) on axonal regeneration of RGCs through a normal or pre-degenerated PN graft. The ON of adult hamsters was transected intraorbitally at 2 mm (proximal lesion) or intracranially at 7 mm (distal lesion) from the optic disc. The pre-lesioned ON was re-transected at 0.5 mm from the disc after 0, 1, 2, 4, or 8 days and a normal or a pre-degenerated PN graft was attached onto the ocular stump. The number of RGCs regenerating their injured axons into the PN graft was estimated by retrograde labeling with FluoroGold 4 weeks after grafting. The number of regenerating RGCs decreased significantly when the delay-time increased in animals with both the ON pre-lesions (proximal or distal) compared to control animals without an ON pre-lesion. The proximal ON pre-lesion significantly reduced the number of regenerating RGCs after a delay of 8 days in comparison with the distal lesion. However, this adverse effect can be overcome, to some degree, by a pre-degenerated PN graft applied 2, 4, or 8 days after the distal ON pre-lesion enhanced more RGCs to regenerate than the normal PN graft. Thus, in order to obtain the highest number of regenerating RGCs, a pre-degenerated PN should be grafted immediately after an ON lesion.

Activation of Group III metabotropic glutamate receptors (mGluRs) by L(+)-2-amino-4-phosphonobutyric acid (L-AP4) has different effects on in vitro slice preparations of visual cortex (Jin & Daw, 1998) as compared with in vivo recordings from somatosensory cortex (Wan & Cahusac, 1995). To investigate the role of Group III mGluRs in the cat visual cortex, in vivo recordings were made of neurons in area 17 of the visual cortex of kittens and adult cats at different ages and the effect of iontophoretic application of L-AP4 (100 mM) was examined. Application of L-AP4 resulted in an increase of the spontaneous activity and visual response of neurons to visual stimulation, the former more than the latter. The effect of L-AP4 was greatest at 3–5 weeks of age with the effect on the visual response declining more rapidly than the effect on spontaneous activity. Consistent with work in rat cortex (Jin & Daw, 1998), the effect of L-AP4 was significantly greater in upper and lower layers than in middle layers. Whole-cell in vitro recordings from slices of rat visual cortex indicated that L-AP4 (50 mM) did not increase the number of spikes elicited by increasing levels of current injections. These results confirm that L-AP4 increases activity in vivo and reasons for the discrepancy with the in vitro results are discussed.

The effects of intraocular injections of ethylcholine mustard aziridinium ion (AF64A), an irreversible inhibitor of choline uptake, on the rabbit retina were assessed electrophysiologically, pharmacologically, anatomically, and behaviorally. Survival times from 1 day to 30 days were investigated. After 24 h, the shortest time tested, the directional selectivity of On–Off responding ganglion cells having the characteristic morphology of On–Off directionally selective directionally selective (DS) ganglion cells, as revealed by intracellular dye injection, was significantly reduced, both by an apparent decrease of preferred direction responses and an increase in responses to null-direction movement. No toxin-mediated changes in the dendritic trees of these cells were noted. Cells in AF64A-affected retinas having the DS morphology did not respond significantly to GABAergic or cholinergic agents such as picrotoxin and eserine, but did respond to nicotine. Recordings from a small random sample of other ganglion cell classes in the same retinas yielded no obvious changes in response properties. The direct effects on starburst (cholinergic) amacrine cells, which were identified by intraocular injection of the fluorescent dye DAPI with the AF64A, were investigated by intracellular injections of Lucifer yellow, and by immunohistochemical staining with antibodies to choline acetyltransferase (ChAT). Although starburst amacrine cell somas survived the AF64A treatment for at least several days, the dendrites could not be visualized by fluorescent dye injection in affected retinas due to dye leakage of the injected fluorescent dye from either the soma or proximal dendritic region. ChAT staining revealed a sequence in which ChAT-positive cells were undetectable first in the inner nuclear layer, and then in the ganglion cell layer. Cholinergic amacrine cells in the central retina were also affected before those in the periphery. The electrophysiological changes observed typically preceded the loss of ChAT activity. Behavioral tests for optokinetic nystagmus responses also revealed a lack of such responses in the affected eyes.